Identification of apocalmodulin and Ca2+-calmodulin regulatory domain in skeletal muscle Ca2+ release channel, ryanodine receptor.

Fusion proteins and full-length mutants were generated to identify the Ca(2+)-free (apoCaM) and Ca(2+)-bound (CaCaM) calmodulin binding sites of the skeletal muscle Ca(2+) release channel/ryanodine receptor (RyR1). [(35)S]Calmodulin (CaM) overlays of fusion proteins revealed one potential Ca(2+)-dependent (aa 3553-3662) and one Ca(2+)-independent (aa 4302-4430) CaM binding domain. W3620A or L3624D substitutions almost abolished completely, whereas V3619A or L3624A substitutions reduced [(35)S]CaM binding to fusion protein (aa 3553-3662). Three full-length RyR1 single-site mutants (V3619A,W3620A,L3624D) and one deletion mutant (Delta4274-4535) were generated and expressed in human embryonic kidney 293 cells. L3624D exhibited greatly reduced [(35)S]CaM binding affinity as indicated by a lack of noticeable binding of apoCaM and CaCaM (nanomolar) and the requirement of CaCaM (micromolar) for the inhibition of RyR1 activity. W3620A bound CaM (nanomolar) only in the absence of Ca(2+) and did not show inhibition of RyR1 activity by 3 microm CaCaM. V3619A and the deletion mutant bound apoCaM and CaCaM at levels compared with wild type. V3619A activity was inhibited by CaM with IC(50) approximately 200 nm, as compared with IC(50) approximately 50 nm for wild type and the deletion mutant. [(35)S]CaM binding experiments with sarcoplasmic reticulum vesicles suggested that apoCaM and CaCaM bind to the same region of the native RyR1 channel complex. These results indicate that the intact RyR1 has a single CaM binding domain that is shared by apoCaM and CaCaM.

CaM is a ubiquitous cytosolic protein that has a critical role in regulating cellular functions by altering the activity of a large number of proteins. CaM regulates all three RyR isoforms. RyR1 and RyR3 are activated by Ca 2ϩ -free CaM (apoCaM) and are inhibited by Ca 2ϩ -bound CaM (CaCaM) (5)(6)(7)(8), whereas RyR2 is not activated by apoCaM but is inhibited by CaCaM (8 -10). Determination of the number of CaM binding sites and their location has been the focus of several studies. Early studies using [ 125 I]CaM (6,11) or fluorescence-labeled CaM (12) showed a stoichiometry of 1 CaCaM and 2-6 apoCaM binding sites/RyR1 subunit. More recent studies using metabolically 35 S-labeled CaM showed one binding site/RyR1 monomer for both of apoCaM and CaCaM (8,10,13,14). Binding site localization studies with fusion proteins and synthetic peptides revealed up to seven candidate CaM binding sites in RyR1 (15)(16)(17)(18), clearly exceeding the number of 1 [ 35 S]apoCaM and 1 [ 35 S]CaCaM binding site/RyR polypeptide. To resolve this discrepancy, full-length RyR1 mutants were generated focusing on two CaM binding domains identified in [ 35 S]CaM overlays of fusion proteins spanning the full-length RyR1 (10). The RyR1 mutants were expressed in HEK293 cells, and their [ 35 S]CaM binding properties and regulation by CaM were determined. We found that two amino acid substitutions (W3620A,L3624D) resulted in a loss of high affinity CaCaM binding and inhibition of RyR1 by CaCaM (nanomolars). The L3624D substitution also resulted in a loss of apoCaM binding and activation of RyR1 by apoCaM. Portions of this study have been published previously in abstract form (19).

Materials-[ 3 H]ryanodine was obtained from PerkinElmer
Life Sciences, Tran 35 S-label was from ICN Radiochemicals (Costa Mesa, CA), unlabeled ryanodine was from Calbiochem (La Jolla, CA), unlabeled CaM was from Sigma, and complete protease inhibitors were from Roche Molecular Biochemicals.
Construction of Wild Type and Mutant cDNA Plasmids-cDNAs for RyR1 fusion proteins tagged with trpE and GST were constructed using pATH and pGEX-5X vectors, respectively. The plasmids were transformed into BL21 Escherichia coli cells, and protein expression was induced by manufacturer's protocol (for GST) and as described previously (for trpE) (20). FPI (3225-3662), FPI-2 (3352-3392), FPI-3 (3391-3554), and FPI-4 (3553-3662) were expressed as trpE fusion proteins, and FPI-1 (3225-3353) and FPM (4302-4430) were expressed as GST fusion proteins (amino acid sequences are shown in parentheses). The full-length rabbit RyR1 cDNA (ClaI/XbaI) was constructed and cloned into expression vector pCMV5 as described previously (21). Single and multiple base changes were introduced by pfu polymerase-based chain * The work was supported by National Institutes of Health Grant AR18687. 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.
§ Postdoctoral Fellow of Japan Society for the Promotion of the Science.
Expression of Full-length RyR1 in HEK293 Cells-RyR1 cDNAs were transiently expressed in HEK293 cells with the LipofectAMINE Plus (Life Technologies, Inc.) or Fugene6 (Roche Molecular Biochemicals) methods according to the manufacturers' instructions. Cells were maintained at 37°C and 5% CO 2 in high glucose Dulbecco's modified Eagle's medium containing 10% fetal bovine serum and plated the day before transfection. For each 10-cm tissue culture dish, 3-6 g cDNA was used. Cells were harvested 42-48 h after transfection. Cells were washed twice with 3-ml ice-cold phosphate-buffered saline containing 1 mM EDTA and Complete protease inhibitors and harvested in the same solution by removal from the plates by scraping. Cells were collected by centrifugation, washed in the same buffer without EDTA, and stored at Ϫ80°C. Sarcoplasmic reticulum (SR) vesicles were prepared from rabbit skeletal muscle as described previously (6).
[ 35 S]Calmodulin Overlay-CaM binding to RyR1 fusion proteins was assayed by [ 35 S]CaM overlays using whole cell preparations or inclusion bodies. [ 35 S]CaM was prepared using Tran 35 S label as described previously (10). Proteins were separated by SDS-polyacrylamide gel electrophoresis and transferred to nitrocellulose membranes. Nonspecific binding sites were blocked by treating membranes with a solution, 150 mM KCl, 20 mM KPipes, pH 7.0, containing 1 mg/ml bovine serum albumin (BSA) and 100 M Ca 2ϩ (blocking buffer) for 1 h. CaCaM binding was analyzed by incubating membranes with 100 nM [ 35 S]CaM in 150 mM KCl, 20 mM KPipes, pH 7.0, 0.04% Tween 20, and 100 M Ca 2ϩ for 1 h and washing with blocking buffer 4 times. Dried membranes were exposed to x-ray film, and radioactivity was determined by autoradiography. ApoCaM binding was analyzed in buffer solutions containing 5 mM EGTA instead of 100 M Ca 2ϩ .
[ 35 S]Calmodulin Binding-Crude membrane fractions prepared as described below were incubated for 2 h at room temperature with solutions, 5 or 15 nM

[ 35 S]Calmodulin Overlays of Wild Type and Mutant RyR1
Fusion Proteins-In a previous study, we used 15 fusion proteins spanning the full coding sequence of the RyR1 polypeptide to identify candidate CaM binding domains (10). We found that two fusion proteins including amino acids 3225-3662 of RyR1 (FPI) and amino acids 4302-4430 (FPM) specifically bound [ 35 S]CaM in a Ca 2ϩ -dependent and independent manner, respectively ( Fig. 1) (10). In this study, we further subdivided the larger of the two fusion proteins (FPI) into four fragments (FPI-(1-4)) using specific restriction enzyme sites. The fragments were expressed as trpE fusion proteins. FPI-1 (3225-3353) was also expressed as a GST fusion protein because the expression level of the trpE fusion protein was very low. Since all fusion proteins were insoluble, [ 35 S]CaM overlays were done with whole cell fractions in Fig. 1. The amounts of proteins on the gels were adjusted to show similar Coomassie Blue staining for the fusion proteins (Fig. 1A). Fig. 1B  Primary sequence predictions suggest the presence of several CaM binding sites in RyR1 (23,24). One of these sites was predicted to be present in FPI-4 (3614 -3637). Using nnPredict (University of California, San Francisco, CA), we identified a stretch of amino acids (aa 3617-3628) predicted to form an amphipathic ␣-helical structure but not in perfect agreement with reported CaM binding motifs. Therefore, we somewhat arbitrarily mutated three hydrophobic amino acid residues (Val 3619 to Ala, Trp 3620 to Ala, and Leu 3624 to Ala and Asp) lying on one face of the helix. We also substituted cysteine 3635 with an alanine because CaM blockage of N-ethylmaleimide alkylation of Cys 3635 suggested that this residue may be important for CaM binding (25). All of the mutant fusion proteins including wt were isolated as inclusion bodies and tested for [ 35 S]CaCaM binding using the overlay assay. Equivalent amounts of wt and mutated FPI-4s were used based on Coomassie Blue staining of SDS gels. The results of the overlay assay are shown in Fig. 2. The strongest binding was observed for wt and C3635A mutant proteins. Mutant proteins with V3619A or L3624A substitutions showed reduced binding, whereas mutant proteins with W3620A or L3624D substitutions barely showed detectable binding. The results identify two amino acid residues (Trp 3620, Leu 3624) that are critical for CaCaM binding to FPI-4. However, it was unclear whether the results with the fusion protein were directly applicable to the full-length RyR1. The information gained was limited because FPI-4 did not bind apoCaM and, therefore, could not be used to locate the apoCaM binding sites in RyR1. Also, [ 35 S]CaM overlays revealed two candidate CaCaM binding sites as opposed to one site/subunit in the native RyR1. Therefore, we extended our mutant studies to the intact RyR1.
[ 35 S]Calmodulin Binding to Wild Type and Mutant RyR1s-We introduced three site-specific mutations in the full-length RyR1 that led to nearly a complete loss (W3620A,L3624D) or a reduction (V3619A) of [ 35 S]CaM binding to FPI-4 (Fig. 2). We also generated a deletion mutant (RyR1⌬4274 -4535) to address the significance of a Ca 2ϩ -independent CaM binding site detected in the overlays in FPM (aa 4302-4430) (Fig. 1, B and  C). The mutant RyR1s were expressed in HEK293 cells, and crude membrane fractions were prepared to determine their CaM binding properties. In parallel experiments, the RyR1 expression levels were quantified by a ligand binding assay using saturating [ 3 H]ryanodine concentrations as described under "Experimental Procedures." Expression of full-length wt and mutant RyR1s was confirmed by Western blot analysis using anti-RyR1 monoclonal antibody D110 (26) (data not shown). In Fig. 3 These studies provided information beyond that obtained with the fusion proteins. The results of Fig. 3 indicate that in the intact RyR1 amino acid residues 4274 -4535 are not important for high affinity apoCaM and CaCaM binding. Rather, they suggest that Leu 3624 constitutes a part of both the apoCaM and the CaCaM binding site in the intact RyR1, whereas Trp 3620 appeared to be only a part of the CaCaM binding site, results that could not be obtained with the mutant fusion proteins because FPI-4 did not show apoCaM binding. inactivation profile comparable with wt-RyR1 with the exception of RyR1⌬4274 -4535, which showed an ϳ10-fold increased sensitivity to activating Ca 2ϩ in agreement with a previous report (22) (data not shown). Fig. 4A shows that [ 3 H]ryanodine binding to wt-RyR1 was inhibited by CaCaM in a concentrationdependent manner with an IC 50 ϳ50 nM. The maximal extent of inhibition (60% by ϳ1 M CaM) was comparable with that observed for native RyR1s (6). The deletion mutant (RyR1⌬4274 -4535) showed a response to CaCaM essentially identical to wt-RyR1. V3619A required a higher CaCaM concentration for the inhibition of [ 3 H]ryanodine binding (IC 50 ϳ200 nM as compared with IC 50 ϳ50 nM for wt-RyR1). L3624D exhibited a greatly reduced apparent affinity for CaCaM as indicated by the requirement of 3 M CaCaM for partial inhibition of RyR1 activity, whereas W3620A did not show any inhibition at 3 M CaCaM. Fig. 4B shows that, in agreement with the apoCaM binding data of Fig. 3B, 1 M apoCaM significantly increased [ 3 H]ryanodine binding to wt and V3619A, W3620A, and ⌬4274 -4535 RyR1s but not L3624D. Taken together, the results of the [ 35 S]CaM (Fig. 3) and [ 3 H]ryanodine binding (Fig. 4) experiments suggest that Leu 3624 constitutes a part of the CaCaM inhibiting and apoCaM activating sites of RyR1, whereas Trp 3620 appears to be only essential for Ca-CaM inhibition.
[ 35 S]Calmodulin Binding to Native RyR1-Dissociation and chase experiments were performed to determine whether Ca-CaM and apoCaM share a common binding domain in native RyR1s using a filtration assay. As shown in Fig. 5A, the dissociation of [ 35 S]CaM from skeletal muscle SR vesicles enriched in RyR1 is not largely dependent of whether CaM is bound in the presence or absence of Ca 2ϩ but rather on whether Ca 2ϩ is present in the dissociation buffer with apoCaM dissociating at a significantly greater rate than CaCaM. In a similar set of experiments, SR vesicles were preincubated with or without non-radioactive CaM in either the presence or absence of Ca 2ϩ followed by the binding of radioactive CaCaM. The resulting rates of [ 35 S]CaM binding were dramatically slower to vesicles pretreated with non-radioactive CaM (Fig. 5B, open symbols) than the rates of binding to vesicles not pretreated with CaM (Fig. 5B, closed symbols). Furthermore, the rates were relatively independent of whether the preincubation had been performed in the absence or presence of Ca 2ϩ . These experiments strongly support the mutant results that CaCaM and apoCaM bind to a common region of RyR1. DISCUSSION Calmodulin has a dual effect on skeletal muscle Ca 2ϩ release channel activity. CaM activates the channel at Ca 2ϩ concentrations below 1 M, whereas at Ca 2ϩ concentrations above 1 M, the channel activity is inhibited by CaM. The data we have presented here indicate that these effects are mediated through Several studies have reported the stoichiometry of CaM binding to RyR1 using SR vesicles (6, 8, 10 -14) and purified RyR1 preparations (6,10). The initial studies using either 125 I (6, 11) or fluorescently (12) labeled CaM revealed that the native RyR1 binds with nanomolar affinity 1 CaM/subunit in the presence of Ca 2ϩ , and that there are as many as six high affinity binding sites for apoCaM on each of the four RyR1 subunits that comprise the functional channel. More recent studies using 35 S metabolically labeled CaM indicate that the tetrameric skeletal muscle channel complex binds 4 CaM molecules both in the absence and presence of Ca 2ϩ or 1 CaM/subunit (8,10,13). These results imply that chemical modification of CaM increases the number of CaM binding sites of RyR1.
Previous studies performed to localize the CaM binding sites relied on the use of fusion proteins and synthetic peptides (Fig.  6). CaM overlays of RyR1 fusion proteins using 125 I (15) (15,16).
Studies with fusion proteins show that the fragmentation of the 565 kDa of RyR peptide into smaller pieces unmasks CaM binding sites not detected in the large channel complex. It is therefore necessary that full-length RyR1 mutants lacking putative CaM binding sites are constructed and that the functional consequences of these mutations are examined. Deletion of one of the potential CaM binding sites identified in the [ 35 S]CaM overlays, RyR1⌬4274 -4535, was without effect on high affinity CaM binding and the inhibition and activation of [ 3 H]ryanodine binding by CaCaM and apoCaM, respectively (Figs. 3 and 4). In this study, we therefore focused on amino acid residues covered by FPI-4 (aa 3553-3662), which contained a CaM binding site implicated in all previous studies (Fig. 6). site that interacts with both apoCaM and CaCaM. Using cryoelectron microscopy and three-dimensional reconstruction, Samso et al. (27) showed that apoCaM and CaCaM bind to two near but distinct cytoplasmic locations on each of the four subunits of the RyR1. This observation suggests that apoCaM and/or CaCaM binding induce major RyR1 protein conformational changes given that it is unlikely that a shift of CaM by several amino acids can be detected at the resolution achievable by electron microscopy.
Our data are in good agreement with a recent report by Moore et al. (13) who suggested that the region of the RyR1 identified in this study binds both apoCaM and CaCaM as both CaM forms were capable of protecting RyR1 from trypsin cleavage at arginines 3630 and 3637. Furthermore, these investiga-tors showed that a synthetic peptide (aa 3614 -3643), which included the two trypsin cleavage sites, bound both apoCaM and CaCaM (18). A shorter peptide (aa 3614 -3635) bound CaCaM but showed a loss of apoCaM binding, whereas another peptide including neither Trp 3620 nor Leu 3624 (aa 3625-3644) bound apoCaM and with a reduced affinity CaCaM (18). Therefore, the results obtained with synthetic peptides (18) and the intact RyR1 in this study do not agree entirely.
The functional consequences of our mutations were assessed by determining their Ca 2ϩ dependence and [ 3 H]ryanodine binding properties. The RyR1 mutants bound [ 3 H]ryanodine with an affinity and showed a Ca 2ϩ dependence comparable with wt-RyR1 with the exception of RyR1⌬4274 -4535, which showed an ϳ10-fold increased sensitivity to activating Ca 2ϩ , as previously reported (22). Therefore, the mutations did not introduce major global conformational changes, but rather they appeared to be mostly limited to the CaM binding sites. The functional studies also allowed tests of the effects of micromolar concentrations of CaM as opposed to the binding studies that are limited to nanomolar CaM concentrations due to experimental restraints. [ 3 H]Ryanodine binding to W3620A was not inhibited by 3 M CaCaM, which suggests a complete loss or at least a very large reduction of CaCaM binding affinity. L3624D and V3619A were inhibited by CaM with IC 50 ϳ3 M and ϳ200 nM, respectively, as compared with IC 50 ϳ50 nM for wt in agreement with the binding studies, which showed nearly a complete loss of CaCaM binding for L3624D but not for V3619A.
In addition to regulating the Ca 2ϩ release channel, CaM probably also influences Ca 2ϩ release through other proteins that interact with the release channel. Potential targets of CaM regulation are the transverse tubule Ca 2ϩ channel, which via a direct interaction controls the SR Ca 2ϩ release channel, calmodulin-dependent protein kinase, and calmodulin-stimulated protein phosphatase (calcineurin) (1-3, 28, 29). Our work provides information for future studies of distinguishing CaM regulation of the RyR1 from that of other proteins. We show that two single amino acid substitutions distinctly change the regulation of the skeletal muscle Ca 2ϩ release channel by CaM; one of which (L3624D) results in a loss of activation by apoCaM and an inhibition by CaCaM, whereas the other (W3620A) specifically abolishes CaCaM inhibition.