The α-Bungarotoxin-binding Nicotinic Acetylcholine Receptor from Rat Brain Contains Only the α7 Subunit

When expressed in Xenopus oocytes, the rat α7 subunit forms homo-oligomeric nicotinic acetylcholine receptors, which are blocked by α-bungarotoxin. Since the pharmacological and physiological properties of the α7 receptor expressed in oocytes are similar to those of the α-bungarotoxin-sensitive nicotinic currents recorded from neuronal preparations and the distribution patterns of α7 mRNA and α-bungarotoxin-binding sites in the rat brain are very similar, α7 is thought to be the main component of the α-bungarotoxin-binding nicotinic receptor in the mammalian brain. However, while α7 is found in purified α-bungarotoxin-binding complexes from rat brain or PC12 cells, other proteins copurify with it. Therefore, the question whether α7 forms a homo-oligomeric α-bungarotoxin-binding nicotinic receptor in the mammalian brain remains. We have developed and characterized affinity-purified polyclonal antibodies and used these antibodies in Western blot analyses of α-bungarotoxin-binding proteins purified from rat brains. We report here that our experimental data support the current working hypothesis that the α-bungarotoxin-binding nicotinic receptor is a homo-oligomer of α7 subunits in the rat brain.

neuronal nAChR subunits have been identified to date. Seven encode ␣ subunits (␣2-␣7 and ␣9), and three encode ␤ subunits (␤2-␤4). When expressed in Xenopus oocytes, only ␣7 and ␣9 form homo-oligomeric nAChRs, and activation of these receptors can be blocked by ␣-BTX (9,10). Since in situ hybridization has shown the ␣9 subunit to be present in hair cells but not in the brain and the ␣9 receptor expressed in oocytes has unique pharmacological properties that are similar to those of the cholinergic receptor observed in vertebrate cochlear hair cells, the ␣9 receptor is thought to be involved primarily in the cholinergic efferent innervation of the cochlear hair cells (10). On the other hand, ␣7 mRNA is widely expressed in the rat brain and is thought to be the main component of the neuronal ␣-BTX-binding nAChR (9).
It is unclear, however, whether the rat neuronal ␣-BTXbinding nAChR consists of only the ␣7 subunit. Data supporting this notion are as follows. 1) ␣7 forms homo-oligomeric ligand-gated ion channels when expressed in oocytes, and the function of these receptors can be blocked by nanomolar concentrations of ␣-BTX (9,11,12); 2) the distribution of ␣7 mRNA in the adult rat brain is similar to that of the ␣-125 I-BTXbinding sites previously documented (2); 3) both ␣7 mRNA and the ␣-BTX-binding sites are transiently expressed in primary sensory cortical and thalamic regions of developing rat and mouse brains (13,14); 4) currents elicited from the ␣7 receptor expressed in oocytes have fast kinetic profiles (12), similar to those of ␣-BTX-blockable nicotinic currents recorded from cultured rat hippocampal neurons that have ␣7 expressed on the surface (15); and 5) the ␣7 receptor expressed in oocytes has high Ca 2ϩ permeability (9), which is also observed with the ␣-BTX-sensitive nicotinic receptor found in cultured hippocampal neurons (16).
However, ␣-BTX-binding proteins purified from neuronal preparations show more than one band on SDS gels stained with Coomassie Blue, possibly because the neuronal ␣-BTXbinding nAChR is a hetero-oligomer (17)(18)(19)(20). In the chick brain, some ␣-BTX-binding proteins are hetero-oligomers composed of both ␣7 and ␣8 subunits (21). Although a rat homologue of chick ␣8 has not been identified, it is possible that ␣7 associates with other neuronal nicotinic receptor subunits to form heterooligomeric ␣-BTX-binding receptors in the rat brain. There are areas in the rat brain where the expression of other receptor subunit genes overlaps with the expression of the ␣7 gene (22). Furthermore, different subunits can express in a single cell. PC12 cells, for example, have ␣-BTX-binding sites on their surface and express mRNAs encoding ␣3, ␣5, ␤2, and ␤4 in addition to ␣7 (23).
To determine whether the ␣7 subunit in the ␣-BTX-binding nAChR from rat brain is associated with other known neuronal nicotinic receptor subunits, we developed affinity-purified polyclonal antibodies and used them in Western blot analyses of affinity-purified ␣-BTX-binding sites. We found no evidence that ␣7 is associated with other subunits in the ␣-BTX-binding nAChR. The data from our experiments support the current working hypothesis that, in the mammalian brain, the ␣-BTXbinding nAChR is a homo-oligomer of ␣7 subunits.

EXPERIMENTAL PROCEDURES
Production of Sheep Antisera-To produce antibodies that recognize rat neuronal nAChR subunits, recombinant proteins were made as follows. Recombinant proteins representing the N-terminal domains of all subunits except ␣7 were made as reported previously (24). To produce a recombinant protein representing the N-terminal domain of ␣7, a DNA fragment encoding the N terminus of ␣7 was generated through polymerase chain reactions and cloned into the bacterial expression vector pTrxFus (Invitrogen). The recombinant ␣7 N terminus-thioredoxin fusion protein was produced according to the manufacturer's instructions. Antisera were custom-produced in sheep by Bethyl Laboratories, Inc.
Purification of Sheep Antisera Using Affinity Chromatography-Hi-Trap columns coupled with either recombinant proteins or synthetic peptides were washed with phosphate-buffered saline. Antiserum was recirculated through the column at 4°C overnight. The column with bound antibodies was then washed with ice-cold 10 mM sodium phosphate, pH 6.8, and antibodies were eluted from the column with 100 mM glycine, pH 2.5. The eluate was collected into 0.1 volume of 1 M sodium phosphate, pH 8.0. Purified antibodies were quickly frozen in liquid nitrogen and stored at Ϫ80°C.
Isolation of Rat Brain Membrane ␣-BTX-binding Sites-Rats (male Sprague-Dawley, 70 -100 g) were decapitated, and the cortex, cerebellum, hippocampus, and the remainder of the brains were dissected and immediately placed in tubes containing ice-cold Buffer A (50 mM sodium phosphate, 50 mM NaCl, 2 mM EDTA, 2 mM EGTA, and 1 mM phenylmethylsulfonyl fluoride, pH 7.4). They were kept on ice until homogenized in a Teflon-glass Dounce homogenizer (8 -10 strokes) using an overhead electronic stirrer (Wheaton) at setting 4. Homogenates were sedimented in an SW 28 rotor at 28,000 rpm (ϳ100,000 ϫ g) for 1 h at 4°C. Supernatants were discarded. The resulting pellets were then resuspended in Buffer A, homogenized, and resedimented under the same conditions. Pellets were resuspended in ice-cold Buffer A containing 2% Triton X-100 and protease inhibitors (5 g/ml each aprotinin, pepstatin, bestatin, and leupeptin). Suspension mixtures were mixed for 2 h at 4°C and sedimented in an SW 28 rotor at 28,000 rpm for 1 h at 4°C, and supernatants (solubilized brain membrane protein fractions) were collected. Since neuronal ␣-BTX-binding proteins can be purified using ␣-cobratoxin-coupled affinity columns (17), supernatants were incubated overnight at 4°C with aliquots of agarose beads coupled with ␣-cobratoxin (Sigma). The beads were washed with ice-cold Buffer A, first with and then without 2% Triton X-100. Proteins bound to the beads were eluted with 6 ϫ SDS-polyacrylamide gel electrophoresis (PAGE) sample buffer and analyzed by Western blotting.
Western Blot Analysis-Proteins in SDS-PAGE sample buffer were separated on 9% SDS-polyacrylamide gels and transferred to nitrocellulose membranes using a transfer unit manufactured by Hoefer Scientific Instruments. The nonspecific protein-binding sites on membranes were blocked with Blotto (Tris-buffered saline and 0.05% Tween 20 (TBST) plus 5% nonfat dry milk) at room temperature for 2 h. Membranes were then incubated at 4°C overnight in affinity-purified antibodies diluted 1:100 in Blotto. The membranes were washed in TBST at least three times before being incubated for 2 h at room temperature in horseradish peroxidase-coupled rabbit anti-sheep IgG (H ϩ L; Organon Teknika Corp.) diluted 1:20,000 in Blotto. Membranes were washed again in TBST at least three times and then incubated in ECL solution (Amersham Corp.) or SuperSignal chemiluminescent substrate for Western Blotting (Pierce) at room temperature for 1 min.
Proteins recognized by the antibody were visualized by exposing the membrane to x-ray films.
The concentration of antibodies was determined for each purified antibody to give the best result with respect to the signal/noise ratio. The same concentration was used in all Western blot analyses performed in this study.
Stripping and Reprobing of Blots-Blots were incubated in the stripping buffer (100 mM 2-mercaptoethanol, 2% SDS, and 62.5 mM Tris-HCl, pH 6.7) at 50°C for 30 min with slow agitation, washed in large volumes of TBST at room temperature with several changes of the buffer, and reprobed with the desired antibodies.
125 I-␣-BTX Binding Assay-Aliquots of 100 l of solubilized brain membrane proteins (ϳ3.5 mg/ml) from the cortex, cerebellum, hippocampus, and the remainder (before and after incubation with ␣-cobratoxin beads) were incubated with 10 nM 125 I-␣-BTX (0.8 -1 ϫ 10 9 cpm/nmol) at room temperature for 2 h in the presence (nonspecific binding) or absence (total binding) of 1 M unlabeled ␣-cobratoxin. 80 l of each sample was spotted on a DE81 disc (Whatman) as described previously (30). Discs were washed four times for 5 min each in ice-cold 0.1% Triton X-100, 3% bovine serum albumin (Sigma), 50 mM NaCl, and 10 mM sodium phosphate, pH 7.4, at 4°C. Discs were air-dried on paper towels and counted in a Beckman ␥-counter (GAMMA310). Specific binding was calculated by subtracting nonspecific binding from total binding.
Determining the Concentration of Recombinant Proteins-The concentration of recombinant proteins was determined by the biuret reaction-based BCA kit and bovine serum albumin standards purchased from Pierce.
The neuronal nAChR subunits are homologues; thus, it is important to determine the specificity of the antibody used in these experiments. We performed Western blot analysis on a panel of recombinant proteins representing nAChR subunits. The recombinant proteins were size-fractionated by SDS-PAGE and transferred to nitrocellulose membranes. Each panel was probed with one of the four antibodies. The results shown in Fig. 1  To unambiguously identify the ␣7 subunit, ␣7 antiserum was purified on an affinity column covalently coupled with peptide ␣7EC1 as described under "Experimental Procedures." This 13-amino acid peptide corresponds to a sequence in the ␣7 N-terminal domain that aligns with the main immunogenic region of the muscle nAChR ␣1 subunit (for review, see Ref. 31). The specificity of the purified antibody, anti-␣7EC1, was determined by Western blotting as described above. The blot shown in Fig. 1E demonstrates that anti-␣7EC1 recognizes only the ␣7 recombinant protein.
Purified Antibodies Detect Subunit Recombinant Proteins at Low Picomole Range on Western Blots-The limit of detecting receptor subunits using purified antibodies by Western blot analysis was determined as follows. Known quantities of ␣5, ␣7, ␤2, or ␤4 recombinant protein were size-fractionated by SDS-PAGE and transferred to nitrocellulose membranes. The membranes were then probed with the corresponding purified antibodies. The results shown in Fig. 2  ␣-BTX-binding Sites Are Expressed at Different Levels in Various Brain Regions-␣-BTX-binding sites were isolated from rat brain membrane proteins by affinity chromatography. Solubilized membrane proteins were prepared from four rat brain fractions: cortex; cerebellum; hippocampus; and the remainder, which corresponds to the rat brain lacking the cortex, hippocampus, and cerebellum. ␣-BTX-binding proteins in each solubilized membrane fraction were purified with ␣-cobratoxincoupled agarose beads as described under "Experimental Procedures." The amount of ␣-BTX-binding sites in solubilized membrane proteins before the incubation with ␣-cobratoxin beads (starting material) was determined by the standard filter binding assay and normalized to the amount of protein. The results are illustrated in Fig. 3 (hatched bars), indicating that ␣-BTXbinding sites were present in membrane proteins from all four brain fractions. The amount of ␣-BTX-binding sites in different fractions varies: 208 Ϯ 24.2 fmol/mg in the hippocampus, 84.8 Ϯ 9.5 fmol/mg in the cortex, 38.3 Ϯ 11.5 fmol/mg in the cerebellum, and 130.8 Ϯ 22.6 fmol/mg in the remainder (data are presented as mean Ϯ S.E.).
To determine how efficiently ␣-BTX-binding sites bound to ␣-cobratoxin beads, the amount of ␣-BTX-binding sites in the membrane proteins after the incubation with ␣-cobratoxin beads (flow-through material) was determined as before and normalized to the amount of protein. The results are presented in Fig. 3 (shaded bars). Comparing the amount of ␣-BTXbinding sites in the starting (hatched bars) and flow-through (shaded bars) materials, it is clear that ␣-BTX-binding sites in all four brain membrane fractions were bound to ␣-cobratoxin beads.
␣7 Is the Only Subunit Detected in the Brain Membrane ␣-BTX-binding Sites-To identify the subunits present in the ␣-BTX-binding nAChR, proteins bound to ␣-cobratoxin beads were eluted with SDS-PAGE sample buffer. Samples from each fraction were divided into four aliquots, and each was analyzed Since ␣7 expressed either in COS cells or in oocytes is a 57-kDa protein and the size of a major protein in ␣-BTX-binding protein affinity-purified from brains is also 57 kDa (32)(33)(34), the simplest interpretation of the data is that the immunoreactive protein band of 57 kDa is rat ␣7. Therefore, ␣7, of the four subunits tested, is the only subunit detected in proteins bound to ␣-cobratoxin beads.
Recombinant proteins served as positive controls in these experiments. On the SDS gels, 0.5-0.8 pmol of the respective recombinant protein was added to each gel next to ␣-BTXbinding proteins. As seen in Fig. 4A, although no immunoreactive bands are present in lanes Cx, Cb, Hp, and R of panels 1, 3, and 4, recombinant protein in each of these panels shows a clear positive signal. In addition, known amounts of ␣5, ␣7, ␤2, and ␤4 recombinant proteins were resolved on separate SDS gels and analyzed by Western blotting under the same conditions as described for Fig. 4A to provide a quantitative estimate of the sensitivity of the assay. The data are presented in Fig.  4B. The results in Fig. 4 (A and B) demonstrate that the antibodies used in these experiments consistently detect their respective subunit recombinant proteins at low picomole range.
To determine if the negative results seen in Fig. 4A were due to a failure of transferring proteins from the resolving gel to the membrane, the ␤2 blot (Fig. 4A, panel 3) was stripped and reprobed with anti-␣7EC1 as described under "Experimental Procedures." As shown in Fig. 5, all four lanes (Cx, Cb, Hp, and R) on the blot have a major immunoreactive protein band of 57 kDa, thus eliminating that possibility. Since the 57-kDa protein is recognized by both anti-␣7-rp (Fig. 4A, panel 2) and anti-␣7EC1 (Fig. 5) and the binding of anti-␣7EC1 to the protein was competed away by the excess amount of peptide ␣7EC1 (Fig. 5, inset), the 57-kDa protein is most likely the ␣7 subunit. Minor immunoreactive bands at 52 kDa are observed in the blots probed with both anti-␣7-rp and anti-␣7EC1. Since these proteins are recognized by both antibodies against ␣7 and, in some experiments, the amount of the 52-kDa protein increased while the amount of the 57-kDa protein decreased, it is likely that these 52-kDa proteins are partially degraded ␣7 subunits.
Another explanation for the negative results is that purified polyclonal antibodies may not react to the native nAChR subunits on Western blots even though they recognize recombinant proteins. However, several lines of evidence suggest otherwise. We showed that both anti-␣7-rp and anti-␣7EC1 reacted with the native ␣7 subunit (Figs. 4A and 5). We also showed that anti-␤2-rp recognized proteins expressed on the surface of cultured hippocampal neurons that express ␤2 mRNA, suggesting that the antibody does recognize the native ␤2 subunit. 2 Recently, anti-␤4-rp was shown to recognize the ␤4 subunit purified from SN-56 cells. 3 Based on these data, it is likely that anti-␣5-rp also recognizes the native proteins. Therefore, the results from the Western blot analyses indicate that ␣5, ␤2, and ␤4 are unlikely to be present in the neuronal ␣-BTX-binding nAChR.  Fig. 4A, panel 3) was stripped and reprobed with anti-␣7EC1 as described under "Experimental Procedures." The ␤2 recombinant protein (Rp), the positive control on the original blot, was not recognized by anti-␣7EC1 as expected. However, a major immunoreactive protein band of 57 kDa in each lane is clearly present on the blot. The ␤4 and ␤2 blots were stripped again and reprobed with anti-␣7EC1 in the presence (ϩ) and absence (Ϫ) of 10 g/ml peptide ␣7EC1, respectively. The binding of anti-␣7EC1 to the 57-kDa band was competed away by the peptide as shown in the inset. Cx, cortex; Cb, cerebellum; Hp, hippocampus; R, the remainder.  A and B) were used in the assay. The protein size markers in A were rainbow markers from Amersham Corp.
The above experiments not only show that ␣5, ␤2, and ␤4 are not detected in the neuronal ␣-BTX-binding nAChR, but also suggest that ␣3, ␣6, and ␤3 are not present in the receptor. As shown in Fig. 1, anti-␣5-rp also recognizes ␣3 and ␤3 with similar sensitivities as ␣5. Therefore, if ␣3 and ␤3 were present, they would be recognized by anti-␣5-rp. The fact that no immunoreactive proteins are observed on the ␣5 blot (Fig. 4A,  panel 1) suggests that ␣3 and ␤3 are not present in the ␣-BTXbinding proteins. Anti-␣7-rp cross-reacted with the ␣6 recombinant protein, but anti-␣7EC1 did not (Fig. 1, B and E). Therefore, if the neuronal ␣-BTX-binding nAChR contains ␣6 and the size of native ␣6 is different from that of native ␣7, the blot probed with anti-␣7-rp should show an immunoreactive band that is not present on the blot probed with anti-␣7EC1. If the size of native ␣6 is also 57 kDa, the intensity of the 57-kDa immunoreactive bands on the blot probed with anti-␣7-rp (Fig.  4A, panel 2) should be stronger than that on the blot probed with anti-␣7EC1 (Fig. 5). Since the same number of immunoreactive protein bands is found on the blots probed with anti-␣7-rp (Fig. 4A) and anti-␣7EC1 (Fig. 5) and the intensity of the 57-kDa immunoreactive protein band on these two blots is similar, it is unlikely that ␣6 is present in the ␣-BTX-binding nAChR.
Determining the Ratio of ␣-BTX-binding Sites to ␣7 in the ␣-BTX-binding nAChR of the Rat Brain-The above results are consistent with the current working hypothesis that ␣7 forms homo-oligomeric ␣-BTX-binding nAChRs in the rat brain. One of the predictions derived from this hypothesis is that the ratio of ␣-BTX-binding sites to the ␣7 subunit in the ␣-BTX-binding nAChR should be a constant for all brain fractions. To test this idea, the amount of ␣7 protein in the ␣-BTX-binding nAChR from each brain fraction was determined, and the ratio was calculated.
First, to determine the amount of ␣7 protein in the ␣-BTXbinding nAChR of each brain fraction, a standard curve was established as follows. Known amounts of ␣7 recombinant protein in each lane showed immunoreactive bands with characteristic intensities (Fig. 4B). The integrated density of each band was measured as described under "Experimental Procedures," and a standard curve of integrated density versus protein amount was generated. Over the range of protein amount used, a linear relationship of integrated density to protein amount was observed (Fig. 6A). Integrated densities of ␣7 immunoreactive bands in the ␣-BTX-binding nAChR (Fig. 4A) were measured, and the amount of ␣7 protein was calculated using the standard curve and normalized to the amount of solubilized membrane protein. Normalized data in Fig. 6B show that the expression of ␣7 in different brain fractions varies: 1100 fmol/mg of membrane protein in the hippocampus, 470 fmol/mg of membrane protein in the cortex, 700 fmol/mg of membrane protein in the remainder, and 200 fmol/mg of membrane protein in the cerebellum. The amount of ␣-BTX-binding sites (presented also in femtomoles/mg of protein) from each brain fraction was then calculated by subtracting the amount of ␣-BTX-binding sites detected in the flow-through material (Fig. 3, shaded bars) from that detected in the starting material (Fig. 3, hatched bars). When the calculated amounts of ␣-BTXbinding sites (in femtomoles/mg of protein) are compared with the amounts of ␣7 (in femtomoles/mg of protein; Fig. 6B), it is clear that the amounts of ␣-BTX-binding sites correspond well to the amounts of ␣7 detected in different regions of the rat brain.
To determine the ratio of ␣-BTX-binding sites to the ␣7 subunit in the neuronal ␣-BTX-binding nAChR, the amount of ␣-BTX-binding sites (Fig. 7, y axis) was plotted against the amount of ␣7 protein (x axis). The data from the two separate experiments shown in Fig. 7 indicate that the relationship between the number of ␣-BTX-binding sites and that of the ␣7 subunit in ␣-BTX-binding nAChRs from all brain fractions can be fitted by a straight line with a slope of 0.20 (r ϭ 0.85). The data were also analyzed for the significance of regression with Student's t test. The analysis showed that the regression (0.2) was statistically significant (p Ͻ 0.01), suggesting that the ratio of ␣-BTX-binding sites to the ␣7 subunit in the ␣-BTXbinding receptors of all brain fractions is a constant, consistent with our prediction. If the ␣-BTX-binding nAChR is a pentamer, then the ratio suggests that each receptor consists of five ␣7 subunits with one detectable ␣-BTX-binding site. FIG. 6. Quantifying ␣7 in the ␣-BTX-binding nAChR of different brain fractions. A, integrated densities of ␣7 recombinant protein bands (see Fig. 4B, panel 2) were measured using an Agfa studio scanner and NIH Image software. The standard curve was generated by plotting the integrated density of each band against the amount of protein loaded in that lane. B, integrated densities of ␣7 bands on the blot (see Fig. 4A, panel 2) were measured in the same way as described for A. The amount of ␣7 in the ␣-BTX-binding proteins of each brain fraction was calculated from the standard curve in A, normalized to total membrane proteins, and presented as a bar graph in femtomoles/mg of membrane protein. Cx, cortex; Cb, cerebellum; Hp, hippocampus; R, the remainder.

FIG. 7.
Determining the molar ratio of ␣-BTX-binding sites to ␣7 in the neuronal ␣-BTX-binding nAChR. The number of ␣-BTXbinding sites in each brain fraction was plotted against that of the ␣7 subunit in the corresponding fraction. Data from two experiments (indicated by shaded and white squares) are plotted in the graph. The relationship between the number of ␣-BTX-binding sites and that of ␣7 in different brain fractions is linear with a slope of 0.20 (r ϭ 0.85; p Ͻ 0.01, Student's t test for significance of regression).

DISCUSSION
We conducted experiments to determine the subunit composition of the ␣-BTX-binding nAChR in the rat brain. Using Western blot analysis, we showed that only the ␣7 subunit was detected in neuronal ␣-BTX-binding receptors purified from all four brain fractions. In addition, we calculated the ratio of ␣-BTX-binding sites to the ␣7 subunit in the receptors from these fractions and showed that it was constant (0.20). Therefore, the most straightforward interpretation for these observations is that the rat neuronal ␣-BTX-binding nAChR is a homo-oligomer of ␣7 subunits with one detectable ␣-BTX-binding site.
Several groups have reported that other proteins copurify with ␣7 from brains using ␣-BTX affinity columns (17,35,36). Therefore, it is possible that in vivo, the neuronal ␣-BTXbinding nAChR is a hetero-oligomer of ␣7 plus other subunits. We performed Western blot analyses and showed that ␣3, ␣5, ␣6, ␤2, ␤3, and ␤4 were not detected in the ␣-BTX-binding nAChR purified from rat brains. Although the antibodies used in our experiments did not recognize ␣2 and ␣4, both subunits are unlikely to be present in the neuronal ␣-BTX-binding nAChR due to the observations that ␣2 mRNA is expressed in very restricted brain areas and ␣4 mRNA is hardly detected in the hippocampus, where ␣-BTX-binding sites are abundant (2,37), and that neither ␣2 nor ␣4 mRNAs are found in PC12 cells, which also express ␣-BTX-binding sites (23).
A rat homologue of the chicken ␣8 subunit, which also binds ␣-BTX, has not been found. Nevertheless, it is unlikely that an ␣8-like subunit is present in the rat neuronal ␣-BTX-binding nAChR. As shown in Fig. 1, anti-␣7-rp, a polyclonal antibody against the N terminus of ␣7, also recognizes the N termini of ␣3, ␣6, ␤2, and ␤4, subunits that share only 40% amino acid identity with ␣7 in the N-terminal region. On the other hand, the N termini of the chicken ␣7 and ␣8 subunits share 78% amino acid identity. Therefore, if an ␣8-like subunit were in the rat neuronal ␣-BTX-binding nAChR, it is likely that the antibody anti-␣7-rp would recognize it on Western blots. Chicken and rat ␣7 subunits have the same molecular mass (38). If an ␣8-like protein existed in the rat brain and were the same size as chicken ␣8, it might also be expected to migrate as a 60-kDa protein on SDS gels (38). The blots probed with anti-␣7-rp were carefully examined at different exposure times, and no immunoreactive protein band of 60 kDa was observed. The ␣9 subunit, expressed only in epithelial cells, is unlikely to be present in the ␣-BTX-binding nAChR in the brain (10). Therefore, the results from our Western blot analyses indicate that subunits other than ␣7 are unlikely to be present in the ␣-BTX-binding nAChR, consistent with the current working hypothesis that the ␣-BTX-binding nAChR is a homo-oligomer of ␣7 subunits in the rat brain. A plausible explanation for the proteins copurifying with ␣7 is that they may be associated proteins such as cytoskeletal proteins and/or degradation products of ␣7.
The number of ␣-BTX-binding sites per rat neuronal ␣-BTXbinding nAChR is not known. However, given the assumption that the receptor is a homo-oligomer of five ␣7 subunits, several predictions can be made. 1) If each ␣7 subunit binds ␣-BTX in the mature receptor, the predicted number would be five; 2) if only a correctly folded ␣7 dimer forms an ␣-BTX-binding site and two such dimers plus a monomer assemble to form a mature ␣7 pentamer, then the predicted number would be two; 3) if only one ␣-BTX-binding site is formed on a mature ␣7 receptor or if a receptor has five possible ␣-BTX-binding sites, but binding of the first ␣-BTX sterically prevents further ␣-BTX binding, the predicted number would be one.
We calculated the molar ratio of ␣-BTX-binding sites to the ␣7 subunit in ␣-BTX-binding nAChRs from all brain fractions (Fig. 7) and found that it is 0.2. This suggests that the number of detectable ␣-BTX-binding sites per pentameric ␣7 receptor is one, consistent with the third prediction above. Since, under certain experimental conditions, only one of the two ␣-BTXbinding sites of the muscle-type nAChR can be detected (39,40), it is possible that the ␣7 receptor consists of two or more ␣-BTX-binding sites, but only one is detected under our experimental conditions. In either case, the estimated molar ratio supports the current working hypothesis that the ␣-BTX-binding nAChR is a homo-oligomer of ␣7 subunits in the rat brain.