Ryanodine Receptor ( -RyR) - The Journal of Biological Chemistry

THE JOURNAL OF BIOLOGICAL CHEMISTRY © 2001 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 276, No. 4, Issue of January 26, pp. 2953–2960, 2001 Printed in U.S.A.

Selectively Suppressed Ca2ⴙ-induced Ca2ⴙ Release Activity of ␣-Ryanodine Receptor (␣-RyR) in Frog Skeletal Muscle Sarcoplasmic Reticulum POTENTIAL DISTINCT MODES IN Ca2⫹ RELEASE BETWEEN ␣- AND ␤-RyR* Received for publication, July 3, 2000, and in revised form, October 5, 2000 Published, JBC Papers in Press, October 27, 2000, DOI 10.1074/jbc.M005809200

Takashi Murayama and Yasuo Ogawa‡ From the Department of Pharmacology, Juntendo University School of Medicine, Tokyo 113-8421, Japan

We reported earlier that the two ryanodine receptor (RyR) isoforms (␣- and ␤-RyR) purified from frog skeletal muscle were equipotent in the Ca2ⴙ-induced Ca2ⴙ release (CICR) activity (Murayama, T., Kurebayashi, N., and Ogawa, Y. (2000) Biophys. J. 78, 1810 –1824). Whether this is also the case with the native Ca2ⴙ release channel in the sarcoplasmic reticulum (SR), however, remains to be determined. Taking advantage of the facts that [3H]ryanodine binds only to the open form of the channels and that it is practically irreversible at 4 °C, we devised a method to separate the total binding to contributions of ␣- and ␤-RyR, using immunoprecipitation with an ␣-RyR-specific monoclonal antibody. Surprisingly, the binding of ␣-RyR was strongly suppressed to as low as ⬃4% that of ␤-RyR in the SR vesicles. The two isoforms, however, showed no difference in sensitivity to Ca2ⴙ, adenine nucleotides, or caffeine. This reduced binding of ␣-RyR was ascribed to the low affinity for [3H]ryanodine, with no change in the maximal binding sites. Solubilization of SR with 3-[(3-cholamidopropyl) dimethylammonio]-1-propanesulfonic acid partly remedied this nonequivalence, whereas 1 M NaCl was ineffective. 12-kDa FK506-binding protein (FKBP12), however, could not be responsible for it, because FK506 treatment did not eliminate the suppression, in contrast to marked removal of 12-kDa FK506-binding protein from ␣-RyR. These results suggest that ␣-RyR in the SR may serve Ca2ⴙ release in a mode other than CICR, being selectively suppressed in CICR. Ryanodine receptor (RyR)1 is a large (⬃2.3-MDa) homotetrameric Ca2⫹ release channel in the sarcoplasmic reticulum (SR) membrane in vertebrate skeletal muscles and plays a critical role in excitation-contraction coupling (1– 4). The RyR channel is mainly activated by two distinct modes: depolariza* 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. 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. ‡ To whom correspondence should be addressed: Dept. of Pharmacology, Juntendo University School of Medicine, 2-1-1 Hongo, Bunkyo-ku, Tokyo 113-8421, Japan. Tel.: 81-3-5802-1034; Fax: 81-3-5802-0419; Email: [email protected]. 1 The abbreviations used are: RyR, ryanodine receptor; AMPPCP, ␤,␥-methylene adenosine 5⬘-triphosphate; CHAPS, 3-[(3-cholamidopropyl) dimethylammonio]-1-propanesulfonic acid; CICR, Ca2⫹-induced Ca2⫹ release; DHPR, dihydropyridine receptor; DICR, depolarizationinduced Ca2⫹ release; FKBP12, 12-kDa FK506-binding protein; PAGE, polyacrylamide gel electrophoresis; SR, sarcoplasmic reticulum; pCa, ⫺log [Ca2⫹]. This paper is available on line at http://www.jbc.org

tion-induced Ca2⫹ release (DICR) and Ca2⫹-induced Ca2⫹ release (CICR). DICR, which is the primary mechanism in skeletal muscle contraction, is triggered directly or indirectly by the conformational change of the voltage sensor, the dihydropyridine receptor (DHPR), on depolarization of the transverse tubule membrane. On this occasion, extracellular Ca2⫹ entry is not necessarily required. In contrast, CICR is caused by activation of the RyR channel by micromolar or greater concentrations of Ca2⫹, which is attained by the Ca2⫹ influx through DHPR, although tight association with DHPR is not required for this mode. Adult mammalian skeletal muscles predominantly express the type 1 isoform of RyR (RyR1). Some specific muscles, e.g. diaphragm and soleus, however, also have a minuscule amount (⬍1– 4%) of the type 3 isoform (RyR3; Refs. 5–7). Recent studies using gene-targeted mice revealed that RyR1 could mediate both DICR and CICR, whereas RyR3 showed CICR but not DICR (8, 9). Frog and other nonmammalian vertebrate skeletal muscles, in contrast, express nearly equal amounts of the two isoforms of RyR, referred to as ␣- and ␤-RyR (10, 11), which are homologues of RyR1 and RyR3, respectively (12, 13). Because of an RyR1 homologue in the primary structure, ␣-RyR is believed to mediate DICR in nonmammalian skeletal muscles. This is supported by the fact that a Crooked Neck Dwarf mutant of chicken that lacks normal ␣-RyR fails to exhibit DICR (14). The role and significance of ␤-RyR in these muscles, however, are still unclear. ␣- and ␤-RyR purified from frog skeletal muscle demonstrated the CICR channel activity, and the Ca2⫹ release appeared to be a simple summation of each contribution (11, 15). Further investigations on [3H]ryanodine binding activity revealed that these isoforms were very similar in activity and that their responses to CICR modulators were indistinguishable under conditions simulating the myoplasm, suggesting an equal assignation of ␣- and ␤-RyR to CICR in situ in frog skeletal muscle (16, 17). These experiments were conducted using purified proteins that were solubilized with a detergent such as CHAPS to separate them from each other. It is well known that the RyR channel activity is modulated by several accessory proteins (e.g. 12-kDa FK506-binding protein (FKBP12) and calmodulin; Refs. 4, 18, 19). However, no or only a minor amount of such accessory proteins was detected in the purified RyR preparations (20, 21), probably because of dissociation of these proteins from RyR in the presence of CHAPS during the solubilization and purification procedure; this, in turn, might cause changes in properties of these isoforms. Recent investigation revealed that CHAPS greatly enhanced [3H]ryanodine binding to RyRs in SR vesicles of rabbit and frog skeletal muscles (22). In addition, the affinities for divalent

2953

2954

Suppressed CICR Activity of Frog ␣-RyR

cations of the activating and inactivating Ca2⫹ sites of the purified ␣- and ␤-RyR were found to be lower than those of the Ca2⫹ release channel of RyR in situ (17). Determination of the activity of each of ␣- and ␤-RyR in situ are therefore required for better understanding of their roles. In the present study, we established a method for separating the total [3H]ryanodine binding activity into those of ␣- and ␤-RyR in frog skeletal muscle SR vesicles where the native organization of the Ca2⫹ release channel including RyR and accessory proteins was still maintained. This success is attributable entirely to two findings: first, an extremely slow dissociation of ryanodine from RyR, practically irreversible at a low temperature (23, 24); and second, separation of the two isoforms without loss of their binding activity by immunoprecipitation with a specific monoclonal antibody after solubilization of the SR with CHAPS (6, 25). The results, contrary to our expectations, demonstrate that [3H]ryanodine binding activity of ␣-RyR is much lower than that of ␤-RyR in the SR vesicles, suggesting a selective suppression of the CICR activity of ␣-RyR in frog skeletal muscle. EXPERIMENTAL PROCEDURES

Materials—[3H]Ryanodine (60 –90 Ci/mmol) was purchased from NEN Life Science Products, Inc. Goat anti-rat IgG-agarose, nonimmune rat IgG, and monoclonal anti-calmodulin antibody (C-7055) were obtained from Sigma. Anti-FKBP12 antibody (PA1-026) was from Affinity Bioreagents Inc. Egg lecithin (egg total phosphatide extract) was from Avanti Polar Lipids. All other reagents were of analytical grade. Isolation of SR Vesicles—SR vesicles were prepared from bullfrog leg muscle (11). The isolated vesicles were quickly frozen in liquid N2 and stored at ⫺80 °C until used. Membrane protein was measured by the biuret method using bovine serum albumin as a standard. SDS-Polyacrylamide Gel Electrophoresis (PAGE) and Western Blotting—SDS-PAGE was performed with 2–12% or 5–15% linear gradient gels (21). Gels were stained with Coomassie brilliant blue. For Western blotting, gels were electrophoretically transferred onto polyvinylidene difluoride membranes. Western blotting was performed with rabbit anti-FKBP12 antibody (1:1000 dilution), rat anti-␣-RyR antibody (1H7, 1:100 dilution), and mouse anti-calmodulin antibody (1:100 dilution), and positive bands were detected by an ECL system using peroxidaseconjugated secondary antibodies (21). Antibody Production and Immunoprecipitation—A monoclonal antibody against frog ␣-RyR, 1H7, was produced in rats according to the method of Kishiro et al. (26) using the purified ␣-RyR as an antigen. This antibody selectively recognized ␣-RyR among proteins of SR vesicles prepared from frog skeletal muscle and did not react with ␤-RyR (see “Results”). It also did not react with any RyR isoforms of mammals (RyR1–3) or ␣-RyR of chicken or fish (data not shown). Immunoprecipitation was performed using 1H7-bound agarose beads, which were prepared from anti-rat IgG-agarose beads (6, 25). Briefly, an aliquot of 100 ␮l of the antibody solution was incubated for 2 h at 4 °C with 30 ␮l of goat anti-rat IgG-agarose beads in a buffer containing 0.5 M NaCl, 20 mM Tris-HCl, pH 7.5, and 0.05% Tween 20. Nonimmune rat IgG was similarly adsorbed to the beads for control experiments. These beads were washed three times with the buffer and stored at 4 °C until use. SR vesicles were solubilized with 1% CHAPS and 0.5% egg lecithin in the above buffer and incubated with the antibody beads for 2 h at 4 °C. After washing three times with the buffer, proteins bound to the beads were subjected to SDS-PAGE. [3H]Ryanodine Binding Assay—SR vesicles were usually incubated with 8.5 nM [3H]ryanodine for 5 h at 25 °C in 200 ␮l of a buffer containing 0.17 M NaCl, 20 mM 3-(N-morpholino)-2-hydroxypropanesulfonic acid/NaOH, pH 6.8, and 2 mM dithiothreitol. 1 mM ␤,␥-methylene adenosine 5⬘-triphosphate (AMPPCP) and various concentrations of Ca2⫹ buffered with 10 mM EGTA were supplemented unless otherwise indicated. Free Ca2⫹ concentrations were calculated using the value of 8.79 ⫻ 105 M⫺1 as the apparent binding constant for Ca2⫹ of EGTA (27). After 5 h of incubation when the [3H]ryanodine binding reached nearly steady state, the vesicles were supplemented with 20 ␮M nonradioactive ryanodine to terminate further incorporation of [3H]ryanodine, followed by immediate cooling to 4 °C, and solubilized with 1% CHAPS and 0.5% egg lecithin. For separation of the total binding into that for ␣- and ␤-RyR, ␣-RyR was immunoprecipitated by incubating for 2 h at 4 °C with the 1H7-bound agarose beads, which had been washed by the

FIG. 1. Time course of [3H]ryanodine binding to frog skeletal muscle SR vesicles and resources for maintaining [3H]ryanodine binding at the steady state value after solubilization. 100 ␮g of SR vesicles was incubated with 8.5 nM [3H]ryanodine at 25 °C in medium containing 0.17 M NaCl, 20 mM 3-(N-morpholino)-2-hydroxypropanesulfonic acid/NaOH, pH 6.8, 2 mM dithiothreitol, 1 mM AMPPCP, and 0.1 mM free Ca2⫹. Inset, time course of the binding reaction. After 5 h of incubation (arrow), 20 ␮M nonradioactive ryanodine, 1% CHAPS, and 0.5% egg lecithin were added to the medium to solubilize the SR with cessation of further incorporation of [3H]ryanodine. Further incubation at 25 °C (open circles) or 4 °C (filled circles) was continued for a period as indicated on the abscissa. The radioactivity derived from bound [3H]ryanodine was determined as described under “Experimental Procedures” and plotted against incubation time. The data are mean ⫾ S.E. (n ⫽ 3). Symbols without bars mean that the magnitude of S.E. was within the size of the symbols. Note that the bound [3H]ryanodine was gradually decreased at 25 °C, whereas it was unchanged at 4 °C. reaction medium containing 1% CHAPS and 0.5% egg lecithin. The precipitated agarose beads were washed three times with the same washing buffer, and the radioactivity responsible for ␣-RyR was recovered by incubating the beads with 0.1 M glycine-HCl, pH 1.5, which was used to disrupt the antigen-antibody complex. The resultant supernatant, on the other hand, gave [3H]ryanodine binding to ␤-RyR which was determined by the gel filtration method using a small-scale column in a centrifuge (21). Total binding to SR vesicles was separately determined in a similar way from the supernatant after immunoprecipitation using nonimmune rat IgG-agarose beads instead of 1H7-agarose beads. Nonspecific radioactivity was determined in the presence of 20 ␮M unlabeled ryanodine at the onset of the binding reaction. RESULTS

Rationale for Separation of Total [3H]Ryanodine Binding to Frog SR Vesicles into Contributions of ␣- and ␤-RyR—In previous experiments (6), we successfully determined [3H]ryanodine binding to RyR3 in the mixture of solubilized RyR1 and RyR3. In the present experiments, however, total binding to SR vesicles must be separated into ␣- and ␤-RyR by immunoprecipitation after solubilization of SR vesicles by CHAPS and phospholipids. These reagents, however, enhanced [3H]ryanodine binding at steady state as shown in Fig. 6 (also see Ref. 22). Because the immunoprecipitation procedure requires 3– 4 h of additional incubation, [3H]ryanodine binding might increase during this period. Resources to avoid this are required. Fig. 1, inset, shows the time course of [3H]ryanodine binding to frog skeletal muscle SR vesicles at 25 °C in an isotonic medium containing 0.17 M NaCl, 8.5 nM [3H]ryanodine, and the optimal Ca2⫹ (pCa 4.0; see “Experimental Procedures”). Because frog SR vesicles show very low [3H]ryanodine binding activity without added ligands other than Ca2⫹ (24), 1 mM AMPPCP, a nonhydrolyzable ATP analog, was supplemented to the medium to stimulate the binding. The binding followed an apparently exponential time course (24) and was close to near steady state at 4 h (Fig. 1., inset). After 5 h of incubation (Fig. 1, upward arrow), incorporation of [3H]ryanodine was terminated by addition of 20 ␮M unlabeled ryanodine, and the vesicles were solubilized with 1% CHAPS and 0.5% egg lecithin

Suppressed CICR Activity of Frog ␣-RyR to separate ␣- and ␤-RyR as described below. The following change of bound [3H]ryanodine after solubilization is shown in Fig. 1. The radioactivity gradually decreased with time at 25 °C (open circles) because of replacement of the bound [3H]ryanodine by the unlabeled one. At 4 °C, in contrast, it was unchanged up to 3.5 h, because the bound [3H]ryanodine hardly dissociated from RyR at the low temperature (filled circles; Refs. 23, 24). Thus, by incubating at 4 °C, we were able to hold the [3H]ryanodine binding at the steady state under a specified condition for several hours after solubilization. This allowed us to separate ␣- and ␤-RyR retaining their [3H]ryanodine binding. The immunoprecipitation method with a monoclonal antibody against ␣-RyR, 1H7, was used to separate ␣- and ␤-RyR but to retain their [3H]ryanodine binding activity as shown previously (21, 28). Fig. 2A shows an SDS-PAGE pattern for the immunoprecipitation experiments. In frog skeletal muscle SR vesicles, ␣- and ␤-RyR were detected as two bands of nearly equal intensity (Fig. 2A, SR). The ratio of intensity of these bands (␣:␤) was estimated by densitometry to be 45:55. When SR proteins after solubilization with CHAPS were immunoprecipitated with control nonimmune IgG, both isoforms remained in the supernatant, and neither was detected in the precipitated beads (Fig. 2A, IgG). In contrast, when immunoprecipitated with 1H7, ␣-RyR was recovered all in a single band in the beads and was not found in the supernatant, whereas ␤-RyR remained in the supernatant (Fig. 2A, 1H7). These indicate that 1H7 specifically and completely immunoprecipitated ␣-RyR from the solubilized SR proteins. This complete separation of the two RyRs was consistently observed, irrespective of the presence or absence of the CICR modulators (Ca2⫹, AMPPCP, caffeine, and ryanodine) that were used in the [3H]ryanodine binding experiments (data not shown). Model experiments using the purified ␣- and ␤-RyR clearly proved that the method of immunoprecipitation with 1H7 as described under “Experimental Procedures” properly determined [3H]ryanodine binding of individual isoforms (Fig. 2B). The purified ␣- and ␤-RyR were prelabeled with [3H]ryanodine up to steady state and then immunoprecipitated with 1H7 (see “Experimental Procedures”). The [3H]ryanodine binding activity of the supernatant after immunoprecipitation with nonimmune IgG, which was separately determined, is referred to as total binding, because the supernatant should contain both isoforms (Fig. 2B, hatched columns). Total binding of ␣- and ␤-RyR was 134 ⫾ 6 and 131 ⫾ 3 pmol/mg protein, respectively, and they were similar as shown previously (16, 17). By immunoprecipitating with 1H7, almost all of the radioactivity for ␣-RyR was recovered in the beads (Fig. 2B, filled columns), whereas all the binding for ␤-RyR remained in the supernatant (Fig. 2B, open columns). With each isoform, the sum of radioactivity in the supernatant and the precipitate (beads) was consistent with the total binding, indicating no significant loss of bound [3H]ryanodine during immunoprecipitation (Fig. 2B, columns in ␣-RyR and ␤-RyR). With a mixed preparation of equal amounts of ␣- and ␤-RyR, furthermore, approximately half of the total binding was detected in each of the supernatant and the precipitate (Fig. 2B, ␣ ⫹ ␤). Thus, solubilization with supplement of nonradioactive ryanodine followed by immunoprecipitation with the 1H7 antibody works well to determine the individual [3H]ryanodine binding of ␣- and ␤-RyR in the SR vesicles where both isoforms occur. [3H]Ryanodine Binding Activity of Two RyR Isoforms in Frog Skeletal Muscle SR—A three-step procedure was used to determine the binding of ␣- and ␤-RyR in SR vesicles of frog skeletal muscles. First, the SR vesicles were labeled with [3H]ryanodine to determine total [3H]ryanodine binding at the steady state

2955

FIG. 2. Separation of ␣- and ␤-RyR by immunoprecipitation with an ␣-RyR-specific monoclonal antibody, 1H7. A, SDS-polyacrylamide gel electrophoresis of ␣- and ␤-RyR separated. 15 ␮g of frog SR vesicles was solubilized (SR), followed by immunoprecipitation with nonimmune IgG (IgG) or 1H7, a monoclonal antibody against ␣-RyR (1H7), as described under “Experimental Procedures.” The resultant supernatant (lane S) and precipitate (beads; lane B) were subjected to SDS-PAGE on a 2–12% linear gradient gel together with 15 ␮g of unprocessed SR vesicles (SR) and stained with Coomassie brilliant blue. H and L, bands for heavy and light chains of IgG, respectively. Molecular masses (in kDa) of standards are indicated on the left. Note that the supernatant (S) showed the band for ␤-RyR (␤) without the contaminating band of ␣-RyR (␣), and that the beads (B) showed the reverse when immunoprecipitated with 1H7. With IgG, in contrast, both ␣- and ␤-RyR were retained in the supernatant. B, model experiments for determination of [3H]ryanodine binding to ␣- and ␤-RyR by immunoprecipitation. 1 ␮g of each purified ␣- and ␤-RyR and a mixture of them (0.5 ␮g each) were incubated with [3H]ryanodine as in Fig. 1 in the presence of 1% CHAPS and 0.5% egg lecithin and then processed to immunoprecipitation with 1H7. The radioactivity of protein-bound ryanodine in the resultant supernatant (open columns) and beads (filled columns) were determined. Total binding (hatched columns) was separately determined from the supernatant of immunoprecipitation with nonimmune IgG instead of 1H7. The data are mean ⫾ S.E. (n ⫽ 3).

under a specified condition. Second, they were supplemented with nonradioactive ryanodine, cooled down to 4 °C, and solubilized with CHAPS and phospholipids. Finally, the solubilized specimen was incubated at 4 °C with 1H7 to separate ␣- and ␤-RyR into the precipitate of agarose beads and the supernatant, respectively, and the [3H]ryanodine binding of each isoform was determined. Total [3H]ryanodine binding to SR vesicles was separately determined using nonimmune IgGagarose beads as mentioned above, confirming the validity of our results. Fig. 3 shows the stimulatory effect of various amounts of AMPPCP on ␣- and ␤-RyR in SR vesicles. The ryanodine binding reaction was carried out at the optimal

2956


FIG. 3. [3H]Ryanodine binding activity of ␣- and ␤-RyR in the SR vesicles. [3H]Ryanodine binding to the SR vesicles was carried out at 25 °C as in Fig. 1 in the presence of 0.2 and 1 mM AMPPCP. Then SR vesicles were rapidly cooled to 4 °C and solubilized with 1% CHAPS and 0.5% egg lecithin in the presence of 20 ␮M supplemented nonradioactive ryanodine. The two isoforms (␣- and ␤-RyR) were separated by immunoprecipitation with 1H7 as described under “Experimental Procedures.” The radioactivity in the precipitate (beads; filled columns) and the supernatant (open columns) represents the ryanodine binding activity of ␣- and ␤-RyR, respectively. Total binding (hatched columns), which was obtained separately from the supernatant of immunoprecipitation with nonimmune IgG, represents the binding of the sum of the two isoforms. Inset, results for ␣-RyR redrawn on an extended scale. The data are mean ⫾ S.E. (n ⫽ 4). Note that ␣-RyR shows much lower [3H]ryanodine binding activity than ␤-RyR, although ␣-RyR is as sensitive to AMPPCP as ␤-RyR. The results in the presence of 5 mM AMPPCP were similar.

concentration of pCa 4.0 in an isotonic medium containing 0.17 M NaCl in the presence of 0.2 and 1 mM AMPPCP. Surprisingly, the [3H]ryanodine bound to ␣-RyR was much less than that bound to ␤-RyR; the radioactivities in the beads were ⬍1 and 4% of those in the supernatant at 0.2 mM and 1 mM AMPPCP, respectively. This low binding is not attributable to loss of the bound [3H]ryanodine from the beads during the separation procedure, because the sum of the binding in the supernatant and the beads was almost equal to the total binding. This reduced value of ␣-RyR (⬃4% of ␤-RyR) was consistently observed at a longer incubation period (over 15 h) of the binding reaction, excluding the possibility of a shorter incubation time for this reason. The increase in AMPPCP from 0.2 to 1 mM enhanced the total binding ⬃3-fold, which was primarily ascribed to an increase in the binding to ␤-RyR. It should be noted, however, that ␣-RyR is also sensitive to AMPPCP in this stimulation of [3H]ryanodine binding (Fig. 3, inset). The enhancement factor for ␣-RyR appeared to be somewhat greater than that for ␤-RyR with the increase from 0.2 to 1 mM AMPPCP. This is also the case with 5 mM AMPPCP (data not shown). Similar results were also obtained by immunoprecipitation using a polyclonal antibody against ␤-RyR (25) instead of 1H7 (data not shown). Consequently, these results indicate that ␣-RyR intrinsically has much lower ryanodine binding than ␤-RyR in the SR vesicles. Ca2⫹ dependence of the [3H]ryanodine binding to each isoform is shown in Fig. 4. For clear observation, 1 mM AMPPCP was supplemented to the reaction medium to enhance the binding. ␤-RyR showed a biphasic Ca2⫹ dependence, with apparent EC50 values of ⬃16 ␮M and ⬃0.8 mM for Ca2⫹ activation and Ca2⫹ inactivation, respectively (Fig. 4, triangles). ␣-RyR also showed a similar biphasic Ca2⫹ dependence, although the relationship seemed to be slightly shifted to a lower Ca2⫹ concentration range (Fig. 4, squares, also see inset). There was only a slight difference, however, between ␣- and ␤-RyR in the Ca2⫹ sensitivity in either Ca2⫹ activation or Ca2⫹ inactivation. ␣-RyR, in contrast, displayed much less binding at all of the Ca2⫹ concentrations examined (Fig. 4, squares). The peak

FIG. 4. Ca2ⴙ dependence of the [3H]ryanodine binding for ␣and ␤-RyR in SR vesicles. [3H]Ryanodine binding experiments were carried out in the presence of 1 mM AMPPCP with varied free Ca2⫹ concentrations and processed as in Fig. 3. Inset, results for ␣-RyR redrawn on an extended scale. The data are mean ⫾ S.E. (n ⫽ 3). Note that ␣-RyR showed biphasic Ca2⫹ dependence similar to that of ␤-RyR, although the amount of binding was much smaller.

value for ␣-RyR (0.036 pmol/mg protein) at pCa 4.0 was ⬍4% of that for ␤-RyR (0.95 pmol/mg). Fig. 5 demonstrates the effect of caffeine on ␣-RyR (Fig. 5A) and ␤-RyR (Fig. 5B) in the SR vesicles. At pCa 5.6, which is near the threshold for Ca2⫹ activation without caffeine, 10 mM caffeine enhanced the binding several 10-fold. At the optimal concentration of pCa 4.0, however, the enhancement by caffeine was 3-fold at most. In the presence of caffeine, the binding at pCa 5.6 was significantly higher than that at pCa 4.0. These findings were common to the two isoforms and consistent with the well-known modification of the RyR channel activity by caffeine: increased sensitivity to Ca2⫹ in the Ca2⫹ activation and enhancement of the peak activity of CICR and [3H]ryanodine binding (17, 22). It should be noted that ␤-RyR showed [3H]ryanodine binding 10 –30 times as great as that of ␣-RyR under the same conditions. These results indicate that ␣-RyR is sensitive to caffeine, as is the case with ␤-RyR, in SR vesicles, although the former is much lower than the latter in ryanodine binding activity even in the presence of caffeine. Effects of High Salt and CHAPS on [3H]Ryanodine Binding to ␣- and ␤-RyR in SR Vesicles—In isotonic media, ␣-RyR had [3H]ryanodine binding activity much lower than ␤-RyR, although the sensitivity to Ca2⫹, AMPPCP, and caffeine was still retained. The effects of high salt and CHAPS, both of which potently stimulate [3H]ryanodine binding to SR vesicles (22, 29), were examined (Fig. 6). In medium containing 1 M NaCl, the total [3H]ryanodine binding was nearly 5-fold higher than that in medium containing 0.17 M NaCl. A marked enhancement in binding was observed not only with ␤-RyR but also with ␣-RyR. The nonequivalence of the two RyRs, however, was still remarkable: ␤-RyR showed 6-fold higher [3H]ryanodine binding than ␣-RyR. The addition of 1% CHAPS with 0.5% egg lecithin into the isotonic medium increased the total binding 4-fold. Under this condition, the [3H]ryanodine binding to ␣-RyR was more enhanced, reaching nearly half of the binding to ␤-RyR. Thus, [3H]ryanodine binding activity of ␣-RyR is effectively increased by solubilization of the SR vesicles with CHAPS. These findings indicate that some molecular interaction that is weakened by CHAPS may selectively suppress the CICR activity of ␣-RyR. FKBP12 Homologue Cannot Be a Candidate for Selective Suppression—It has been reported that FKBP12 selectively inhibits the Ca2⫹ release channel activity of RyR1 among three RyR isoforms in mammals (30, 31). Possible involvement of FKBP12 in the suppressed activity of ␣-RyR was therefore tested using FK506, which specifically removed the FKBP12


2957

FIG. 5. Effect of caffeine on the [3H]ryanodine binding to ␣and ␤-RyR in SR vesicles. Determinations were carried out with or without 10 mM caffeine as in Fig. 4. A, ␣-RyR; B, ␤-RyR. The data are mean ⫾ half the range of deviation of duplicate determinations. Note that the enhancement by caffeine was more marked at pCa 5.6 than at pCa 4.0. In the presence of 10 mM caffeine, the binding at pCa 5.6 was much higher than that at pCa 4.0, although these Ca2⫹ concentrations corresponded to those near the threshold and the optimum for Ca2⫹ activation, respectively, in the absence of caffeine. These findings are consistent with the following conclusions on the effect of caffeine: increase in the sensitivity to Ca2⫹ in Ca2⫹ activation and enhanced peak value.

FIG. 6. Effects of high salt and CHAPS on the [3H]ryanodine binding to ␣- and ␤-RyR in SR vesicles. [3H]Ryanodine binding to SR vesicles was carried out as in Fig. 1 under the conditions indicated. The following procedures were performed as in Fig. 3. The data are mean ⫾ S.E. (n ⫽ 3– 6). Note that the presence of 1 M NaCl (center) instead of 0.17 M (left) greatly enhanced the binding of ␣-RyR as well as ␤-RyR, but nonequivalence of the two isoforms was still prominent. Addition of 1% CHAPS with 0.5% egg lecithin (right) more markedly increased the binding of ␣-RyR, reaching approximately half that of ␤-RyR.

(32). Fig. 7A demonstrates Western blots of the fraction immunoprecipitated with 1H7 of solubilized frog SR vesicles. A small protein of ⬃15 kDa was positively reacted with anti-FKBP12 antibody (Control lane). This band disappeared from the SR vesicles treated with 100 ␮M FK506 before immunoprecipitation, without change in the amount of precipitated ␣-RyR (⫹FK506 lane). The reacted band was closer in its partial N-terminal amino acid sequence to human FKBP12.6 than to FKBP12.0 (Fig. 7B). The [3H]ryanodine binding of SR vesicles treated with 100 ␮M FK506 was compared with that of nontreated ones (Fig. 7C). The selective suppression of ␣-RyR was consistently observed even after the SR vesicles were pretreated with FK506. These results suggest that although a homologue of mammalian FKBP12 is tightly associated with ␣-RyR in frog skeletal muscle, it cannot be responsible for the selective suppression of ␣-RyR.

FIG. 7. Effect of FK506 on the dissociation of the FKBP12 homologue from ␣-RyR and the suppressed binding of ␣-RyR. A, Western blots of the fraction of frog SR vesicles that was immunoprecipitated with 1H7. Immunoprecipitation was carried out as shown in Fig. 2 in the presence (⫹FK506) and absence (Control) of 100 ␮M FK506. Molecular masses (in kDa) of standards are indicated on the left. Note that a band of ⬃15 kDa, which reacted with anti-human FKBP12, specifically disappeared after incubating the SR with FK506 (lower panel). FK506 did not affect the amount of ␣-RyR immunoprecipitated (upper panel). B, N-terminal partial amino acid sequence of frog FKBP12 homologue and its alignment with those of human FKBP12.0 and FKBP12.6. C, [3H]ryanodine binding of ␣- and ␤-RyR in the SR vesicles with (⫹FK506) or without (Control) 100 ␮M FK506. Assays were carried out as in Fig. 3. The data are mean ⫾ S.E. (n ⫽ 3). The binding of each isoform was normalized with the total binding under respective conditions. 100% denotes 0.81 and 2.3 pmol/mg protein for Control and ⫹FK506, respectively. Note that FK506 did not eliminate the suppression of ␣-RyR in contrast to its marked removal of the ⬃15-kDa protein.

It is known that calmodulin is associated with RyR and affects its Ca2⫹ release channel activity (1–3). We therefore tested the presence of calmodulin in our SR vesicles by Western blotting with commercial anti-calmodulin antibody. No positive bands were detected, indicating the undetectable amount of calmodulin in the SR (data not shown). This suggests a minor possibility of involvement of calmodulin in the suppression of ␣-RyR. Scatchard Plot Analysis of [3H]Ryanodine Binding to ␣- and ␤-RyR in SR Vesicles—Fig. 8 demonstrates dose-dependent [3H]ryanodine binding to ␣- and ␤-RyR in the SR vesicles together with total binding in medium containing 1 M NaCl. In isotonic medium, the binding to ␣-RyR was too little to enable us to carry out reliable analysis. The reaction was therefore made under a more favorable condition for ryanodine binding (16, 29); the medium contained 1 M NaCl, 1 mM AMPPCP, and 10 mM caffeine, and not only ␤-RyR but also ␣-RyR showed more enhanced binding (see Fig. 6). The amounts of [3H]ryano-

2958

Suppressed CICR Activity of Frog ␣-RyR nM). The Kd values of the purified ␣- and ␤-RyR under similar conditions were 2– 4 nM (11). Thus, the affinity of ␣-RyR for [3H]ryanodine was significantly lower in the SR vesicles than in the purified preparations, whereas ␤-RyR showed similar affinities between native and purified preparations. The maximal binding sites of ␣- and ␤-RyR (Bmax ⫽ 3.9 ⫾ 0.3 and 4.9 ⫾ 0.2 pmol/mg SR protein, respectively) were consistent with the content of each isoform in the SR vesicles (45:55 for ␣- and ␤-RyR; see Fig. 2). These results indicate that the suppressed [3H]ryanodine binding of ␣-RyR is primarily attributable to a reduced affinity but not to a decreased Bmax value. The total [3H]ryanodine binding to the SR vesicles (Fig. 8, circles) corresponded well to the calculated sum of binding amounts of ␣- and ␤-RyR as shown by the dashed line (Fig. 8A). In Scatchard plot analysis, the results could also be fitted well by an assumption of a single class of homogeneous binding sites for ryanodine (Fig. 8B). Kd and Bmax values were similarly calculated by the results in Fig. 8A to be 3.8 ⫾ 0.5 nM and 7.6 ⫾ 0.3 pmol/mg SR protein, respectively. This Bmax value was slightly lower than the sum (8.8 ⫾ 0.4 pmol/mg protein) of the two Bmax values for ␣- and ␤-RyR. The upper limit of available concentrations of [3H]ryanodine used here was 21 nM, which may be insufficient for saturable binding of ␣-RyR. If higher concentrations of the ligand were feasible, the Scatchard plot would be downward convex with an intercept at ⬃8.8 pmol/mg protein, as shown by the dashed line (Fig. 8B). DISCUSSION

FIG. 8. Dose-dependent [3H]ryanodine binding to ␣- and ␤-RyR in SR. Frog skeletal muscle SR vesicles (100 ␮g) were incubated with 0.5–21 nM [3H]ryanodine for 5 h at 25 °C in 1 M NaCl, 20 mM 3-(Nmorpholino)-2-hydroxypropanesulfonic acid/NaOH, pH 6.8, 2 mM dithiothreitol, 1 mM AMPPCP, 10 mM caffeine, and 0.1 mM free Ca2⫹. The results for total SR (circles) and individual isoforms (squares, ␣-RyR; triangles, ␤-RyR) were determined as in Fig. 3. A, presentation in common coordinates to show the range of deviation. The data are mean ⫾ half the range of deviation of duplicate determinations. Symbols without bars mean that the deviation was within the size of the symbol. B, Scatchard plot for the average at each ligand concentration. Linear Scatchard plots indicate that each isoform had a single class of [3H]ryanodine binding sites. Binding parameters (Kd and Bmax) were obtained by fitting the data in Fig. 8A to the equation B ⫽ Bmax ⫻ [free ryanodine]/(Kd ⫹ [free ryanodine]); 12.8 ⫾ 1.3 nM and 3.9 ⫾ 0.3 pmol/mg protein for ␣-RyR and 2.3 ⫾ 0.1 nM and 4.9 ⫾ 0.2 pmol/mg protein for ␤-RyR, respectively. The results of total binding to SR vesicles (␣ ⫹ ␤) could also be fitted by a single class of binding sites with Kd of 3.8 ⫾ 0.5 nM and Bmax of 7.6 ⫾ 0.3 pmol/mg protein. The continuous lines in A and B were drawn using these values. The dashed lines represent the calculated sum using the parameters determined for ␣- and ␤-RyR. Note that the directly determined and calculated results at the lowest concentration of the ligand were actually indistinguishable (A), although they appeared very diverse in the Scatchard plot (B). See “Results” for details.

dine bound to ␣-RyR (Fig. 8, squares) and ␤-RyR (Fig. 8, triangles) increased with increase in free [3H]ryanodine concentration (Fig. 8A). Scatchard plot analysis revealed that the results could be expressed by a linear line in either case, indicating a single class of homogeneous binding sites (Fig. 8B). The best fit binding parameters (Kd and Bmax) were obtained by fitting the data to an equation B ⫽ Bmax ⫻ [free ryanodine]/(Kd ⫹ [free ryanodine]) according to the Marquardt-Levenberg algorithm using SigmaPlot for Macintosh, version 5. The obtained sets of binding parameters were as follows: Kd ⫽ 12.8 ⫾ 1.3 nM, and Bmax ⫽ 3.9 ⫾ 0.3 pmol/mg protein for ␣-RyR; and Kd ⫽ 2.3 ⫾ 0.1 nM, and Bmax ⫽ 4.9 ⫾ 0.2 pmol/mg protein for ␤-RyR (expressed as mean ⫾ S.E. for duplicate determinations at each of six different ligand concentrations). ␣-RyR (Kd ⫽ 12.8 ⫾ 1.3 nM) showed 6-fold lower affinity than ␤-RyR (Kd ⫽ 2.3 ⫾ 0.1

In the present study, we investigated the [3H]ryanodine binding to the isolated SR vesicles as an index of the CICR activity of the two RyR isoforms (␣- and ␤-RyR) in frog skeletal muscle, because ryanodine can bind only the open form of RyR. After the binding reached the steady state, solubilization of SR vesicles by CHAPS and phospholipids at a low temperature in the presence of an excess amount of nonradioactive ryanodine and immunoprecipitation with the monoclonal antibody against ␣-RyR (1H7) enabled us to identify each contribution of ␣- and ␤-RyR to the total [3H]ryanodine binding. This is the first report to demonstrate the properties of the individual RyR isoforms of nonmammalian skeletal muscles under conditions in which these proteins exist in the intact SR membranes. Lipid bilayer experiments with SR vesicles from skeletal muscles of nonmammalian vertebrates including frog reported two distinct types of channel activity (3, 33–35): low peak open probability with marked Ca2⫹ inactivation and high peak open probability with no or slight Ca2⫹ inactivation. No evidence, however, can identify which isoform shows which type of channel activity. Marengo et al. (35) showed that the two types of channel activity were convertible to each other by sulfhydryl oxidation and reduction of channels. Note that the experiments presented here were carried out in the presence of 2 mM dithiothreitol. We demonstrated that the [3H]ryanodine binding to ␣-RyR is much lower (only ⬃4% at most) than that of ␤-RyR in SR vesicles in isotonic medium. Neither isoform, however, showed a difference in its Ca2⫹ dependence (Fig. 4) or sensitivity to AMPPCP (Fig. 3) and caffeine (Fig. 5), which are well known stimulants of CICR. These findings lead us to conclude that ␣-RyR may have CICR activity selectively suppressed to a great extent in SR vesicles from frog skeletal muscle. This is in marked contrast to the results obtained with the purified ␣and ␤-RyR, which showed nearly equal [3H]ryanodine binding in the physiological milieu (Refs. 16, 17; see Fig. 2). Consistently, the affinity for [3H]ryanodine of ␣-RyR was considerably lower (Kd ⫽ 12.7 nM) than that of ␤-RyR (Kd ⫽ 2.3 nM) in the SR in the medium containing 1 M NaCl, 1 mM AMPPCP, and 10 mM caffeine (Fig. 8), whereas the purified ␣- and ␤-RyR showed

Suppressed CICR Activity of Frog ␣-RyR similar Kd values of 2– 4 nM (11), which are close to the value of ␤-RyR in the SR. The selective suppression of ␣-RyR may also be true with the native Ca2⫹ release channel in SR of intact or skinned fibers. Scatchard plot analysis revealed that the suppression of ␣-RyR is primarily caused by a marked decrease in affinity for ryanodine, but not a reduction in Bmax (Fig. 8). This excludes the possibility that some populations of ␣-RyR would be completely inactive or silent, whereas others would be active. Alterations of ␣-RyR in sensitivity to CICR modulators, such as Ca2⫹, AMPPCP, and caffeine, also cannot be the cause for the reduced activity of ␣-RyR in view of the results shown in Figs. 3–5. All of the ␣-RyRs are potentially viable, but they may be in a suppressed state. We have recently demonstrated that the CICR activity can be defined by at least three independent parameters: activation by Ca2⫹, inactivation by Ca2⫹ or Mg2⫹, and the potential peak activity (17). An adenine nucleotide dose-dependently increased the potential peak activity that determined the upper limit of CICR activity at the optimum Ca2⫹ concentration, but the stimulating reagent did not change affinities for Ca2⫹ or Mg2⫹ of RyR in Ca2⫹ activation or inactivation (17). Caffeine showed dual effects of Ca2⫹ sensitization in Ca2⫹ activation and an increase in the peak activity in CICR. Separate experiments showed that these stimulators also dosedependently increased affinity for [3H]ryanodine of purified RyR and SR vesicles. These effects, although in the reverse direction, may be reflected in the low affinity for [3H]ryanodine of ␣-RyR in the present case. This type of inhibition seems more suitable for stable suppression of the activity, because the extent of inhibition will be consistent regardless of free Ca2⫹ concentrations. It is interesting to consider the mechanism of how ␣-RyR is suppressed in SR vesicles. Because the suppression is not observed in the purified proteins or their mixtures (Fig. 2), it is likely that some factors in tight association with ␣-RyR may suppress its CICR activity in these vesicles. This is strongly supported by the fact that addition of CHAPS, which solubilizes the SR to weaken protein-protein or protein-lipid interactions, more selectively enhanced the [3H]ryanodine binding activity of ␣-RyR than that of ␤-RyR (Fig. 6). The putative factors should be associated with ␣-RyR alone or should suppress the activity of this isoform even if they bind to both isoforms, because the suppression was effective only on ␣-RyR. To date, many proteins have been proposed to interact with the RyRs to modulate their channel activity, including calmodulin, FKBP12, triadin, and calsequestrin (4, 18, 19). Among them, FKBP12 appeared to be a probable candidate as a suppressive factor, because it was reported to selectively inactivate RyR1, although it can bind to both RyR1 and RyR3 (21, 30). Consistently, our results indicate that a homologue of mammalian FKBP12 may be tightly associated with ␣-RyR (Fig. 7). However, this does not seem to be the case, because addition of FK506, which effectively purged the protein from the immunoprecipitated ␣-RyR, did not eliminate the suppression of the [3H]ryanodine binding activity of ␣-RyR. Because calmodulin was not detected in our SR vesicles by Western blotting with anti-calmodulin antibody, involvement of calmodulin in the suppression also may be unlikely. Alternatively, it is also possible that the lowered activity of ␣-RyR in the SR membrane may be an inherent nature of ␣-RyR itself. It has been demonstrated that the purified RyR1 channels displayed heterogeneous populations of low open probability (⬍0.1) and high open probability (⬃ 0.3), whereas the purified RyR3 channels showed a homogeneous population of open probability (⬃1) in an almost all-or-none Ca2⫹-dependent manner in planar lipid bilayers (21). Both isoforms, in

2959

contrast, showed similar activity in [3H]ryanodine binding in the presence of CHAPS (21). This different behavior of RyR isoforms has generally been explained by the difference in vulnerability to some oxidative conditions. However, this might reflect an association state of the RyR monomer in a planar lipid bilayer, which is different from that in an aqueous solution containing a detergent, because tetramer formation is indispensable for activity. Then, ␣-RyR might be inclined to be at the lowered state in CICR activity in the SR membrane, although the relationship between the extent of association and activity has yet to be determined. Further experiments are necessary to evaluate this possibility. It has been reported that the CICR activity may be under suppressive control by DHPR in mammalian skeletal muscles (36, 37). Suda (36) observed that caffeine-induced Ca2⫹ release in rat skeletal muscle cells was terminated on repolarization of the membrane, probably through DHPR. Shirokova et al. (37) recently demonstrated with cultured mouse myotubes that discrete Ca2⫹ release events, which were probably activated by Ca2⫹, occurred primarily at locations where depolarization of the membrane did not elicit the continuous Ca2⫹ release, irrespective of the presence of RyR3. These findings suggest that the CICR through RyR1 could be under the negative control of DHPR. On the basis of the homology between RyR1 and ␣-RyR in the structure and function, it can reasonably be presumed that ␣-RyR might also be selectively suppressed by interaction with DHPR in frog skeletal muscle. Sucrose density gradient ultracentrifugation revealed that SR vesicles used in this study were hardly associated with the transverse tubule (data not shown). This may exclude the possibility of direct suppressive control of the CICR of ␣-RyR by DHPR. Furthermore, it should be noted that there has been no evidence to date that indicates the occurrence of suppressed RyR1 in SR vesicles from rabbit skeletal muscles, unlike ␣-RyR. Recent studies demonstrated that Ca2⫹ sparks, discrete Ca2⫹ release events probably activated by Ca2⫹, are clearly observed in frog skeletal muscle at rest and on depolarization (38, 39). Shirokova et al. (40) reported that no or only sparse sparks were detected in adult mammalian skeletal muscles, which primarily express RyR1 with very little RyR3 at most (40, 41), whereas they were easily detectable in frog skeletal muscle where ␤-RyR occurred in an amount almost equal to that of ␣-RyR. They proposed that ␤-RyR might be essential for the production of the Ca2⫹ sparks in frog skeletal muscle. This appears to be compatible with our results; however, there is one essential difference. They assumed an intrinsic difference between RyR1 homologues (including ␣-RyR) and RyR3 homologues (including ␤-RyR), whereas we showed that there was no difference in CICR activity between the two isoforms. ␣-RyR, however, is selectively suppressed in SR vesicles but not in the purified state. They also indicated that RyR3 might inactivate RyR1, because the time course of the discrete Ca2⫹ release was prolonged by knockout of RyR3 (37). Bertocchini et al. (42), on the contrary, proposed possible amplification by RyR3 of CICR through RyR1 in mammalian neonate skeletal muscle because of reduction greater than that expected from the RyR3 content on its knockout. Cooperative interaction between the two isoforms is an attractive hypothesis, and further investigations are required. The results of the present study provide important information regarding the roles of the two RyR isoforms in skeletal muscles. The [3H]ryanodine binding activity of ␣-RyR is much lower than that of ␤-RyR in SR vesicles in an isotonic milieu, regardless of the presence or absence of CICR modulators. This suggests that the CICR may be mainly mediated by ␤-RyR in situ in frog skeletal muscle. ␣-RyR is thought to function pri-


2960

marily as a DICR channel, at least in nonmammalian skeletal muscles (2, 3). Thus, the two RyR isoforms may play distinct roles in Ca2⫹ release from SR: ␣-RyR as a DICR channel and ␤-RyR as a CICR channel. Further investigations will clarify the significance of the two RyR isoforms in nonmammalian vertebrate skeletal muscles. Acknowledgments—We thank Drs. H. Abe and T. Obinata (Chiba University) for kind help with producing the monoclonal antibody. REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9.

10. 11. 12. 13. 14.

15. 16. 17.

Meissner, G. (1994) Annu. Rev. Physiol. 56, 485–508 Ogawa, Y. (1994) Crit. Rev. Biochem. Mol. Biol. 29, 229 –274 Sutko, J. L., and Airey, J. A. (1996) Physiol. Rev. 76, 1027–1071 Franzini-Armstrong, C., and Protasi, F. (1997) Physiol. Rev. 77, 699 –729 Conti, A., Gorza, L., and Sorrentino, V. (1996) Biochem. J. 316, 19 –23 Murayama, T., and Ogawa, Y. (1997) J. Biol. Chem. 272, 24030 –24037 Jeyakumar, L. H., Copello, J. A., O’Malley, A. M., Wu, G. M., Grassucci, R., Wagenknecht, T., and Fleischer, S. (1998) J. Biol. Chem. 273, 16011–16020 Takeshima, H., Iino, M., Takekura, H., Nishi, M., Kuno, J., Minowa, O., Takano, H., and Noda, T. (1994) Nature 369, 556 –559 Takeshima, H., Ikemoto, T., Nishi, M., Nishiyama, N., Shimuta, M., Sugitani, Y., Kuno, J., Saito, I., Saito, H., Endo, M., Iino, M., and Noda, T. (1996) J. Biol. Chem. 271, 19649 –19652 Olivares, E. B., Tanksley, S. J., Airey, J. A., Beck, C. F., Ouyang, Y., Deerinck, T. J., Ellisman, M. H., and Sutko, J. L. (1991) Biophys. J. 59, 1153–1163 Murayama, T., and Ogawa, Y. (1992) J. Biochem. 112, 514 –522 Oyamada, H., Murayama, T., Takagi, T., Iino, M., Iwabe, N., Miyata, T., Ogawa, Y., and Endo, M. (1994) J. Biol. Chem. 269, 17206 –17214 Ottini, L., Marziali, G., Conti, A., Charlesworth, A., and Sorrentino, V. (1996) Biochem. J. 315, 207–216 Airey, J. A., Baring, M. D., Beck, C. F., Chelliah, Y., Deerinck, T. J., Ellisman, M. H., Houenou, L. J., McKemy, D. D., Sutko, J. L., and Talvenheimo, J. (1993) Dev. Dyn. 197, 169 –188 Murayama, T., and Ogawa, Y. (1993) Biomed. Res. 14, 71–72 Murayama, T., and Ogawa, Y. (1996) FEBS Lett. 380, 267–271 Murayama, T., Kurebayashi, N., and Ogawa, Y. (2000) Biophys. J. 78, 1810 –1824

18. Wagenknecht, T., and Radermacher, M. (1997) Curr. Opin. Struct. Biol. 7, 258 –265 19. Ogawa, Y., Kurebayashi, N., and Murayama, T. (1999) Adv. Biophys. 36, 27– 64 20. Wagenknecht, T., Grassucci, R., Berkowitz, J., Wiederrecht, G. J., Xin, H. B., and Fleischer, S. (1996) Biophys. J. 70, 1709 –1715 21. Murayama, T., Oba, T., Katayama, E., Oyamada, H., Oguchi, K., Kobayashi, M., Otsuka, K., and Ogawa, Y. (1999) J. Biol. Chem. 274, 17297–17308 22. Ogawa, Y., Murayama, T., and Kurebayashi, N. (1999) Mol. Cell. Biochem. 190, 191–201 23. Lai, F. A., Erickson, H. P., Rousseau, E., Liu, Q.-Y., and Meissner, G. (1988) Nature 331, 315–319 24. Ogawa, Y., and Harafuji, H. (1990) J. Biochem. 107, 887– 893 25. Murayama, T., and Ogawa, Y. (1996) J. Biol. Chem. 271, 5079 –5084 26. Kishiro, Y., Kagawa, M., Naito, I., and Sado, Y. (1995) Cell Struct. Funct. 20, 151–156 27. Harafuji, H., and Ogawa, Y. (1980) J. Biochem. 87, 1305–1312 28. Percival, A. L., Williams, A. J., Kenyon, J. L., Grinsell, M. M., Airey, J. A., and Sutko, J. L. (1994) Biophys. J. 67, 1834 –1850 29. Ogawa, Y., and Harafuji, H. (1990) J. Biochem. 107, 894 – 898 30. Copello, J. A., Jeyakumar, L., and Fleischer, S. (1999) Biophys. J. 76, A469 (abstr.) 31. Barg, S., Copello, J. A., and Fleischer, S. (1997) Am. J. Physiol. 272, C1726 –C1733 32. Timerman, A. P., Ogunbumni, E., Freund, E., Wiederrecht, G., Marks, A. R., and Fleischer, S. (1993) J. Biol. Chem. 268, 22992–22999 33. Bull, R., and Marengo, J. J. (1993) FEBS Lett. 331, 223–227 34. Oba, T. (1997) Am. J. Physiol. 273, C1588 –C1595 35. Marengo, J. J., Hidalgo, C., and Bull, R. (1998) Biophys. J. 74, 1263–1277 36. Suda, N. (1995) J. Physiol. (Lond.) 486, 105–112 37. Shirokova, N., Shirokov, R., Rossi, D., Gonzalez, A., Kirsch, W. G., Garcia, J., Sorrentino, V., and Rios, E. (1999) J. Physiol. (Lond.) 521, 483– 495 38. Tsugorka, A., Rios, E., and Blatter, L. A. (1995) Science 269, 1723–1726 39. Klein, M. G., Cheng, H., Santana, L. F., Jiang, Y. H., Lederer, W. J., and Schneider, M. F. (1996) Nature 379, 455– 458 40. Shirokova, N., Garcia, J., and Rios, E. (1998) J. Physiol. (Lond.) 512, 377–384 41. Conklin, M. W., Barone, V., Sorrentino, V., and Coronado, R. (1999) Biophys. J. 77, 1394 –1403 42. Bertocchini, F., Ovitt, C. E., Conti, A., Barone, V., Scholer, H. R., Bottinelli, R., Reggiani, C., and Sorrentino, V. (1997) EMBO J. 16, 6956 – 6963