Properties of Ryr3 ryanodine receptor isoform in mammalian brain.

Although the RNA for the third isoform (Ryr3) of ryanodine receptor (RyR), a Ca2+ release channel, is detected in specific regions of mammalian brain, little is known about the protein. We investigated Ryr3 in rabbit brain, using an antibody raised against the synthetic peptide corresponding to amino acid sequence 4375-4387 of rabbit Ryr3, the homologue of bullfrog beta-RyR. The antibody which reacted with bullfrog beta-RyR, but not with the other isoforms, Ryr1 or Ryr2, specifically precipitated a single polypeptide from rabbit brain microsomes having a size similar to beta-RyR. Sucrose gradient ultracentrifugation revealed that Ryr3 forms a homotetramer, as true of the other isoforms. Being consistent with the distribution of its RNA, Ryr3 was abundantly expressed in hippocampus, corpus striatum, and diencephalon. Ryr3 demonstrated Ca2+-dependent [3H]ryanodine binding, and caffeine increased its Ca2+ sensitivity. The Ca2+ sensitivity of Ryr3 was also enhanced in a medium containing 1 m NaCl, as observed with beta-RyR. [3H]Ryanodine binding gave an estimate of Ryr3 which would be only 2% or less of total RyR in rabbit brain. These results confirm the expression of functional Ryr3 in mammalian brain which is similar to nonmammalian beta-RyR.

Increase in cytoplasmic Ca 2ϩ in neurons plays an important role in the regulation of neuronal functions (1). Ca 2ϩ can be released from endoplasmic reticulum via two different types of release channel: inositol 1,4,5-trisphosphate receptors (IP 3 Rs) 1 (2,3) and ryanodine receptors (RyRs) (4 -8). In brain, IP 3 R is more abundant than RyR. However, the contribution of IP 3 R and RyR may be nearly equipotent in the regulation of cytoplasmic Ca 2ϩ in neurons because the unit conductance of RyR is about 4 times as large as that of IP 3 R (2). The two receptors also differed in their distribution in the brain.
The major isoform of RyR isolated from brain is very similar to Ryr2 (16 -19), although all three ryr genes have been found in the central nervous system by Northern blot (13) or in situ hybridization (20,21). It has been shown that Ryr1 is exclusively localized in cerebellar Purkinje cells (17, 20 -26; see also Ref. 8). mRNA of the ryr3 gene has been detected in specific regions of brain, e.g. hippocampus, thalamus, and corpus striatum (13,20,21), suggesting that this isoform, i.e. Ryr3, may have an important role in specific brain functions. The properties of Ryr3, however, are still unclear.
We purified two isoforms of RyR (␣and ␤-RyR) from bullfrog skeletal muscle which coexist in non-mammalian vertebrate skeletal muscle (27). The sequencing of their cDNAs revealed that these two isoforms are most homologous to Ryr1 and Ryr3, respectively (28). Taking full advantage of this similarity between mammalian Ryr3 and bullfrog ␤-RyR, we here studied the properties of the former in rabbit brain using the polyclonal antibody specific for Ryr3.

Materials-
The peptides corresponding to the amino acid sequences 4375-4387 of the rabbit Ryr3 (13) and 4460 -4472 of rabbit Ryr2 (11) were synthesized at the Central Laboratory of Medical Sciences, Division of Biochemical Analysis, Juntendo University School of Medicine. Monoclonal antibody RY-3, which is specific for rabbit Ryr2, was a generous gift from Dr. Munekazu Shigekawa of the National Cardiovascular Center Research Institute. [ 3 H]Ryanodine (74.8 Ci/mmol) was purchased from DuPont NEN. Goat anti-rabbit IgG-agarose and antimouse IgG-agarose were from Sigma. Egg lecithin (egg total phosphatide extract) was from Avanti Polar Lipids. All other reagents were of analytical grade.
Isolation of Rabbit Brain Microsomes-Microsomal membranes were prepared from rabbit brain according to McPherson and Campbell (16) with minor modifications. Briefly, adult rabbit brain was homogenized in a buffer containing 50 mM Tris-HCl, pH 7.4, 2 mM dithiothreitol, and a mixture of protease inhibitors (2 g/ml aprotinin, 2 g/ml leupeptin, 1 g/ml antipain, 2 g/ml pepstatin A, and 2 g/ml chymostatin). The homogenate was centrifuged at 10,000 ϫ g for 10 min, and the pellet was discarded. The supernatant was centrifuged at 28,000 ϫ g for 70 min, and the pellet was collected, resuspended in the above buffer, and centrifuged again. The microsomal pellet was resuspended in the above buffer containing 0.3 M sucrose, quickly frozen in liquid nitrogen, and stored at Ϫ80°C until use.
Solubilization of Microsomal Proteins and Sucrose Gradient Ultracentrifugation-Brain microsomal membranes were suspended in 2% CHAPS and 1% egg lecithin, 0.5 M NaCl, 20 mM Tris-HCl, pH 7.4, 2 mM dithiothreitol, and the above mixture of protease inhibitors. The suspension was placed on ice for 15 min and then centrifuged at 100,000 ϫ g for 30 min. The supernatant was used for detection and characterization of Ryr3. Sucrose gradient ultracentrifugation with 5-20% linear gradients was carried out as described previously (27).

SDS-Polyacrylamide Gel Electrophoresis (SDS-PAGE) and
Western Blot Analysis-SDS-PAGE was performed by the buffer system of * This work was supported in part by a grant-in-aid for scientific research from the Ministry of Education, Science, Sports, and Culture of Japan, by the Uehara Memorial Foundation, and by the Suzuken Memorial Foundation. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Production and Purification of Anti-Ryr3 Antibody-Anti-Ryr3 antibody was produced in rabbits according to standard protocols against a synthetic peptide corresponding to rabbit Ryr3 sequence (13 amino acids, 4375-4387) (13), KKRRRGQKVEKPE-C. A cysteine was added to the C terminus for linking to bovine serum albumin via lysines using the bifunctional agent m-maleimidobenzoyl-N-hydroxysuccinimide ester. To purify the antibody, the collected antiserum was applied to an EAH-Sepharose column (Pharmacia Biotech Inc.) which had been conjugated with the antigen peptide using m-maleimidobenzoyl-N-hydroxysuccinimide ester. The column was washed extensively with 0.5 M NaCl, 50 mM Tris-HCl, pH 7.5, 0.05% Tween 20 (TTBS), and the specific antibody was eluted with 0.1 M glycine-HCl, pH 2.8, immediately neutralized, and stored at 4°C. For immunoprecipitation, the antibody was further purified with protein-bound PVDF membranes of ␤-RyR from bullfrog skeletal muscle (27).
Detection of Ryr3-Ryr3 in solubilized microsomal membranes was detected by immunoprecipitation using the purified anti-Ryr3 antibody and goat anti-rabbit IgG-agarose beads. The anti-rabbit IgG-agarose beads were washed three times with TTBS and incubated with the purified anti-Ryr3 antibody overnight at 4°C. After the three washings, 50 l of the beads were incubated overnight at 4°C with solubilized microsomal membranes. The beads were washed five times with TTBS containing 0.1% CHAPS and 2 mM dithiothreitol, and all excess fluid was removed. They were resuspended in 40 l of 2 ϫ Laemmli's sample buffer (29) containing 0.1 M dithiothreitol. After being left for at least 1 h at room temperature, aliquots of 15-20 l were subjected to SDS-PAGE.
[ 3 H]Ryanodine Binding Assay-Solubilized brain microsomal proteins (2-5 mg) were incubated with 10 nM [ 3 H]ryanodine for 4 h at 25°C in 1 ml of a binding buffer containing 1 M NaCl, 10 mM MOPSO/NaOH, pH 6.8, 2 mM dithiothreitol, 2% CHAPS, 1% egg lecithin, 1 mM AMP-PCP, the mixture of protease inhibitors, and specified Ca 2ϩ concentration. The total binding was measured by filtering aliquots of the samples (100 -500 g of protein) through polyethyleneimine-treated Whatman GF/B glass filters (27). The binding specific for Ryr3 was determined by immunoprecipitation. After incubation with [ 3 H]ryanodine, the samples were incubated at 4°C overnight with 10 -100 l of the anti-rabbit IgG-agarose beads that had been pre-equilibrated with the anti-Ryr3 as described above. The beads were washed five times with the binding buffer without [ 3 H]ryanodine or AMP-PCP, placed on a filter, rinsed twice with ice-cold water, and dried. The radioactivity remaining on the filter was determined in a liquid scintillation counter.

Specificity of Anti-Ryr3
Antibody-To detect the Ryr3 in mammalian brain, we produced a polyclonal antibody against a synthetic peptide which was found to be specific to this isoform. We chose the amino acid sequence, KKRRRGQKVEKPE, which corresponds to 4375-4387 of rabbit Ryr3 (13), and is different from rabbit Ryr1 or Ryr2 (Fig. 1). This portion is highly hydrophilic and is reasonably assumed to be exposed on the surface of the protein molecule. This sequence was also highly homologous to that in ␤-RyR from bullfrog skeletal muscle: only one amino acid Val 4383 is replaced by Lys in ␤-RyR. This was known by the finding that ␤-RyR is the most homologous to Ryr3 among the three mammalian RyR isoforms (28). Now that mammalian Ryr3 is not available, ␤-RyR could serve as its substitute in immunologic reactions. As shown in Fig. 2, the antibody reacted with bullfrog ␤-RyR, but not with bullfrog ␣-RyR which is the most homologous to mammalian Ryr1. Neither rabbit Ryr1 nor Ryr2 reacted to the antibody. This strongly indicates that the antibody can specifically recognize Ryr3 among the three isoforms in rabbit.
Detection of Ryr3 in Rabbit and Other Mammalian Brains-First, we attempted to detect the Ryr3 directly by Western blot analysis, but without success. As shown in Fig. 3A, monoclonal antibody RY-3 (anti-Ryr2) reacted with a single polypeptide of approximately 560 kDa in rabbit whole brain microsomes, indicating the expression of Ryr2. In contrast, no polypeptide bands were recognized by anti-Ryr3 antibody. The negative reaction was either due to the lack of Ryr3 or the epitopes for the antibody, or to an amount of expression of Ryr3 too small to be detectable by Western blot. To clarify this point, we carried out an immunoprecipitation assay using antibody-conjugated agarose beads and highly purified antibody (see "Experimental Procedures"). Fig. 3B demonstrates the Coomassie Brilliant Blue-stained SDS-PAGE pattern (left panel) of the immunoprecipitated products of rabbit whole brain microsomes and corresponding Western blot (right panel). The products contained many protein bands. A single band marked by an asterisk specifically disappeared in the presence of the antigen peptide. In contrast, it did not disappear in the presence of the counterpart of Ryr2 (Fig. 1), QKLRQLHTHRYG-C (data not shown). The immunoprecipitated band was clearly and specifically recognized by anti-Ryr3 antibody (Fig. 3B). The others may be due to polypeptides of the antibodies or nonspecific precipitated products. The mobility of the band was similar to that of bullfrog ␤-RyR and slightly larger than that of rabbit Ryr1 or Ryr2 (see Fig. 2A). These results confirm that Ryr3 with a molecular mass similar to that of bullfrog ␤-RyR is definitely expressed in rabbit brain and that the content is small in contrast to the major isoform of Ryr2 in the brain.
FIG. 1. Amino acid sequence for antigen peptide in rabbit Ryr3 and its alignment with counterparts for the other RyR isoforms of rabbit and bullfrog. The amino acid sequence of rabbit Ryr3 used to immunize rabbits is shown at the top (Rabbit Ryr3), and corresponding sequences of rabbit Ryr1 and Ryr2, and bullfrog ␣and ␤-RyRs are aligned below. Number of amino acid residues of each sequence is given at either end. Dots represent residues identical with those of rabbit Ryr3. It is well known that RyR forms a tetramer showing a large sedimentation coefficient (about 30 S) (31,32). Fig. 4 shows the pattern of sucrose density gradient ultracentrifugation of microsomal proteins from rabbit brain. Using monoclonal antibody RY-3, we observed Ryr2 in brain which sedimented to high density fractions as reported previously (16,18) (Fig. 4A). Immunoprecipitation of the identical gradient fractions showed that Ryr3 also sedimented to the same fractions as Ryr2 (Fig.  4B). Monoclonal antibody RY-3 did not react with the product immunoprecipitated by anti-Ryr3 antibody (data not shown), nor did anti-Ryr3 antibody react with the product immunoprecipitated by monoclonal antibody RY-3 (data not shown). Thus it appears that Ryr3 as well as Ryr2 forms a homotetramer in brain as is true of Ryr1 in skeletal muscles and Ryr2 in cardiac muscle.
The distribution of Ryr3 in rabbit brain was determined by immunoprecipitation assay of solubilized microsomal proteins prepared from each region of the brain (Fig. 5). The isoform was abundantly expressed in hippocampus, diencephalon, and corpus striatum and slightly in cerebral cortex and mesencephalon. In cerebellum, no bands could be detected. This distribution well corresponded to the previous results of Northern blots (13) and in situ hybridization (20) in rabbit brain. We determined the cross-reactivity of anti-Ryr3 antibody with Ryr3 in mammalian brains other than rabbit. A single band showing mobility similar to rabbit Ryr3 was detected in the immunoprecipitated products of mouse, rat, and guinea pig whole brain microsomes. The intensities of the reacted bands appeared to be similar among these animals (data not shown). These results indicate that Ryr3 exists in similar quantities in the brains of these animals.
Estimation of the Amount of the Expressed Ryr3 in Rabbit Brain-In the course of the investigation, it was of interest to learn whether Ryr3 is functional. We approached this problem through determination of the [ 3 H]ryanodine binding. Solubilized rabbit brain microsomes were incubated with [ 3 H]ryanodine to achieve ryanodine binding, and Ryr3 was specifically immunoprecipitated by anti-Ryr3 antibody. As shown in Table  I, small but significant binding was observed in the presence of Ca 2ϩ , but not in its absence. Addition of excess concentrations of unlabeled ryanodine caused almost total loss of the radioactivity. Immunoprecipitation in the presence of 30 M of the antigen peptide failed to detect the binding. These results indicate that Ryr3 has Ca 2ϩ -dependent [ 3 H]ryanodine binding activity, suggesting that the isoform may constitute the functional Ca 2ϩ -induced Ca 2ϩ release (CICR) channel.
As shown in Fig. 6, the immunoprecipitated radioactivity due to [ 3 H]ryanodine bound in 2.5 mg of solubilized brain microsomes increased with the added amounts of antibodyagarose beads and reached a plateau around 350 cpm at 50 l of beads or more. Radioactivity was no longer immunoprecipitated from the supernatant that had been incubated with 100 l of the beads (data not shown). Thus, 100 l of the antibodyagarose beads immunoprecipitated almost all Ryr3 in rabbit brain microsomes. This allowed us to estimate the relative amount of Ryr3 to total RyRs in rabbit brain. As shown in  FIG. 4. Sedimentation of RyR isoforms expressed in rabbit brain through sucrose gradient ultracentrifugation. 20 mg of solubilized rabbit whole brain microsomes was sedimented through 5-20% sucrose gradient by ultracentrifugation and divided into 8 fractions (5 ml each, numbered from the top of the gradient). A, 30 l of each fraction was resolved on 2-12% SDS-polyacrylamide gel, transferred onto a PVDF membrane, and probed with monoclonal antibody RY-3. B, the immunoprecipitated products from 4 ml of each gradient fraction with anti-Ryr3 antibody were separated, transferred, and probed with anti-Ryr3 antibody as described under "Experimental Procedures." increased linearly with the amount of microsomes, which gave the specific binding activity of 8200 cpm/mg of microsomal protein. The immunoprecipitated activities were also linearly proportional to the amount of solubilized microsomes, the specific activity of which was 160 cpm/mg of protein (Fig. 7B). Ratio of the immunoprecipitated ryanodine binding activity to total activity was calculated to be about 0.02. The fraction of 1.6% was obtained with another lot of microsomal preparation (e in Fig. 7). Assuming that all RyR isoforms have almost identical properties of ryanodine binding, the fraction of Ryr3 would be only 1.6 -2% of total RyRs in rabbit whole brain.

Characterization of [ 3 H]Ryanodine
Binding to Ryr3-Ryr3 is homologous to ␤-RyR in bullfrog skeletal muscle with more than 85% identity of the deduced amino acid sequences (28). One of the properties characteristic of ␤-RyR is enhancement of Ca 2ϩ sensitivity, which is distinct from that by caffeine, by high concentrations of certain salts such as 1 M NaCl in the presence of CHAPS, and phospholipids (33). This nature is not observed in either ␣-RyR of bullfrog skeletal muscle or in rabbit Ryr1 or Ryr2. We therefore examined the Ca 2ϩ sensitivity in [ 3 H]ryanodine binding of Ryr3 in low and high salt media (Fig. 8). In low salt medium containing 0.17 M NaCl (Fig. 8A), there was little [ 3 H]ryanodine binding at pCa 5.4. About 0.1 mM Ca 2ϩ (pCa 3.9) may be optimal for ryanodine binding, and 3 mM Ca 2ϩ (pCa 2.5) reduced the binding. In 1 M NaCl medium, in contrast, about 70% of the maximum binding was observed at pCa 5.4 (Fig. 8B). The Ca 2ϩ sensitivity of Ryr3 was thus enhanced by high concentrations of NaCl as is the case with bullfrog ␤-RyR. 1 M NaCl also caused remarkable enhancement of ryanodine binding activity at pCa 3.9 by a factor of about 8 as shown in Fig. 8, A and B. Although this enhancement was marked in the absence of AMP-PCP, it should not be as enormous as shown in Fig. 8 where AMP-PCP was present (33). The enhancement may be partly due to the larger extent of solubi-  Table I in the presence of 10 mM Ca-EGTA (pCa 4). Immunoprecipitation was performed using antibody-agarose beads in the amount indicated on the abscissa. The data are the means of duplicate determinations.

FIG. 7. Estimation of the fractional amount of Ryr3 in rabbit brain based on [ 3 H]ryanodine binding activity. [ 3 H]Ryanodine
binding to solubilized rabbit whole brain microsomes was carried out as in Table I in the presence of 10 mM Ca-EGTA (pCa 4). A, total [ 3 H]ryanodine binding determined by filtering samples through polyethyleneimine-treated Whatman GF/B glass filters (27). B, the [ 3 H]ryanodine binding immunoprecipitated by 100 l of anti-Ryr3 agarose beads. The specific activities for total and immunoprecipitated [ 3 H]ryanodine binding (circles) were calculated from each slope to be 8200 and 160 cpm/mg of microsomes, respectively. This indicates that the immunoprecipitable fraction, i.e. that of Ryr3, is 2.0% of total. Similar determinations were carried out using a different microsomal preparation at a fixed amount of microsomes (squares), giving an estimate of 1.6%. Error bars indicate half-range of deviations of duplicate determinations (circles) or S.E. of four determinations (squares). No error bars were shown in cases where the deviation was smaller than the diameter of a circle. lization of RyR in 1 M NaCl than in 0.17 M NaCl (34). Disinhibition of ryanodine binding by high Ca 2ϩ was also noted (Fig.  8B) (33). These characteristic modifications by 1 M NaCl were also observed with ␤-RyR from bullfrog skeletal muscle (33).
We examined the effect of caffeine on [ 3 H]ryanodine binding to Ryr3, since several groups have claimed that the isoform may be insensitive to this substance (14,15). As shown in Fig.  8B, 10 mM caffeine markedly increased the [ 3 H]ryanodine binding to near maximum at pCa 6 in 1 M NaCl media. Thus, caffeine did increase Ca 2ϩ sensitivity for Ryr3; this effect was also observed in the 0.17 M NaCl medium (data not shown). The results in Fig. 8B are consistent with the finding that the effects of caffeine and 1 M NaCl were additive with respect to the increase in Ca 2ϩ sensitivity (35). DISCUSSION The results of this study have revealed several new findings about mammalian Ryr3 in the brain. First, mammalian Ryr3 is expressed in specific regions (hippocampus, diencephalon, and corpus striatum) of rabbit brain; this distribution is consistent with that of mRNA. Second, Ryr3 was estimated to be only 2% or less of the total RyR in the rabbit brain, while the major isoform was Ryr2. Third, Ryr3 forms a homotetramer in spite of being a minor fraction as do the other RyR isoforms and may constitute a functional Ca 2ϩ release channel. Finally, but no less important, mammalian Ryr3 shares the functional characteristics of Ca 2ϩ -induced Ca 2ϩ release channel with bullfrog ␤-RyR: enhancement of Ca 2ϩ sensitivity in the presence of 1 M NaCl and response to caffeine action. This functional coincidence is consistent with the homologous primary structures of both.
The sequence of the antigen peptide was very similar to the corresponding portion of bullfrog ␤-RyR, but was not homologous to those of rabbit Ryr1 or Ryr2 nor to bullfrog ␣-RyR (Fig.  1). Furthermore, this portion is highly conserved in the deduced amino acid sequence of Ryr3 of mink lung epithelial cell and human Jurkat T cell; in mink, the sequence is identical (14) and in human only one amino acid is replaced (15). The antibody recognized Ryr3s not only of rabbit but also of mouse, rat, and guinea pig. Chicken skeletal muscle ␤-RyR, which is homologous to bullfrog ␤-RyR (36,37), also reacted positively with the antibody (data not shown). Thus, this sequence may be highly conserved among Ryr3 of various species.
The anti-Ryr3 antibody was able to precipitate the cognate RyR while retaining its Ca 2ϩ -dependent [ 3 H]ryanodine binding activity from solubilized microsomes that had been incubated with [ 3 H]ryanodine. Similar results of ryanodine binding were obtained when immunoprecipitation preceded the [ 3 H]ryanodine binding assay (data not shown). Thus, the binding of the antibody may not affect the ryanodine binding activity of RyR. We could not exclude the possibility that the antibody might alter the properties of RyR channels as previously reported (38,39), because we could not obtain Ryr3 without complex formation with the antibody. However, well-consistent properties of ryanodine binding with those of bullfrog ␤-RyR (Fig. 8) suggest that the antibody may not significantly affect the RyR channel properties.
The response to caffeine of mammalian Ryr3 has, to date, been controversial. Giannini et al. (14) first reported that caffeine did not release Ca 2ϩ from intracellular stores in transforming growth factor ␤-treated mink lung epithelial cells which express Ryr3. Recently, Hakamata et al. (15) demonstrated that human Jurkat T-cell expresses only Ryr3 among the three isoforms, and that an increase of intracellular Ca 2ϩ was evoked by ryanodine but not by caffeine. In contrast, skinned skeletal muscle fibers of "dispedic mouse" which lack Ryr1 have been shown by Takeshima et al. (40) to release Ca 2ϩ from sarcoplasmic reticulum by caffeine. They concluded on the basis of mRNA analysis that the Ca 2ϩ release was probably mediated by Ryr3. The [ 3 H]ryanodine binding of this study (Fig. 8) clearly indicated that caffeine markedly enhanced the Ca 2ϩ sensitivity of Ryr3 as well as other isoforms reported to date. The lack of caffeine response, therefore, may be due to the low levels of expression of the RyR in those cells, as suggested by Giannini et al. (21). Alternatively, it is possible that Ryr3 expressed in peripheral tissues may not be identical because the transcription product may be subject to alternative splicing (13).
It has been widely accepted that the major isoform of RyR in mammalian brain is Ryr2 (11,12,18,19,41) and that the other isoforms are expressed in specific regions (13,20,21). Very recently, Giannini et al. (21) reported low levels of Ryr3 in the mouse brain, and we have estimated that this fraction is about 2% or less of total RyRs expressed in rabbit brain (Fig. 7). This is also the case with other mammalian brains. This limited amount can clearly explain the failure to detect the isoform by Western blot (Fig. 2) (see also Ref. 37), as well as earlier observations that RyR purified from rabbit brain was very similar to Ryr2 (19). This is also consistent with the results of Furuichi et al. (20), who reported exposure time of 5 days for ryr2 and 18 and 14 days for ryr1 and ryr3, respectively, for radioautography of in situ hybridization.
Study of the distribution of mRNA indicated that Ryr3  Table I  should be expressed in specific regions (hippocampus, thalamus, and corpus striatum) of rabbit brain (13,20). The enriched expression of Ryr3 in hippocampus, corpus striatum, and diencephalon was confirmed as demonstrated by immunoprecipitation (Fig. 5). More recently, expression of Ryr3 in specific regions of mouse brain has been demonstrated by Western blot analysis using isoform-specific antibody (21). The results between rabbit and mouse are similar in some points: abundant expression in hippocampus and corpus striatum, suggesting their important functions in these regions. Several differences are also noted, however. In cerebral cortex, marked expression was observed in mouse, while only a small amount was expressed in rabbit. Expression in diencephalon (or thalamus) seems more abundant in rabbit than in mouse. These might be due to the difference of the animal species examined.
In the central nervous system, the dominant Ca 2ϩ release channel protein is IP 3 R which also has at least three isoforms and forms a tetrameric complex of these subunits. The possibility of heterotetramer as well as homotetramer has recently been suggested, resulting in potential diversity of the function as the Ca 2ϩ release channels, although this theory is yet to be proved (42). In contrast, our evidence shows that RyR isoforms, other Ca 2ϩ release channel proteins, make only homotetramers even in a fraction as small as 2% or less of total RyR. The ␣and ␤-isoforms (Ryr1 and Ryr3, respectively) of bullfrog skeletal muscle were found to have very similar properties of CICR channel in the myoplasm, although they differed in the 1 M NaCl medium (33) (concerning RyR isoforms of chicken and fish, see Refs. 43-45, respectively). It might be generalized that all three isoforms of RyR have similar properties of CICR channels. Then, what is the biological reason that the three isoforms occur and coexist in some places? Why is the Ryr3 isoform localized in three specific and very interesting regions of the brain, hippocampus, thalamus, and corpus striatum?
The key to the answers to these questions may be found in the mechanism(s) of signal transduction like that between the voltage sensor in the T-tubule (dihydropyridine receptor) and RyR in the sarcoplasmic reticulum in the skeletal muscle. Further investigations will clarify the functional role of the Ryr3 in specific neuronal cells.