A Role for Cysteine 3635 of RYR1 in Redox Modulation and Calmodulin Binding*

Oxidation of the skeletal muscle Ca2+ release channel (RYR1) increases its activity, produces intersubunit disulfide bonds, and blocks its interaction with calmodulin. Conversely, bound calmodulin protects RYR1 from the effects of oxidants (Zhang, J.-Z., Wu, Y., Williams, B. Y., Rodney, G., Mandel, F., Strasburg, G. M., and Hamilton, S. L. (1999)Am. J. Physiol. 276, Cell Physiol. C46–C53). In addition, calmodulin protects RYR1 from trypsin cleavage at amino acids 3630 and 3637 (Moore, C. P., Rodney, G., Zhang, J.-Z., Santacruz-Toloza, L., Strasburg, G. M., and Hamilton, S. L. (1999) Biochemistry 38, 8532–8537). The sequence between these two tryptic sites is AVVACFR. Alkylation of RYR1 with N-ethylmaleimide (NEM) blocks both35S-apocalmodulin binding and oxidation-induced intersubunit cross-linking. In the current work, we demonstrate that both cysteines needed for the oxidation-induced intersubunit cross-link are protected from alkylation with N-ethylmaleimide by bound calmodulin. We also show, using N-terminal amino acid sequencing together with analysis of the distribution of [3H]NEM labeling with each sequencing cycle, that cysteine 3635 of RYR1 is rapidly labeled by NEM and that this labeling is blocked by bound calmodulin. We propose that cysteine 3635 is located at an intersubunit contact site that is close to or within a calmodulin binding site. These findings suggest that calmodulin and oxidation modulate RYR1 activity by regulating intersubunit interactions in a mutually exclusive manner and that these interactions involve cysteine 3635.

Reactive oxygen species and nitric oxide (NO) 1 are produced by skeletal muscle even at rest, but their levels are dramatically increased by muscle activity (1)(2)(3)(4)(5). Both reactive oxygen species and NO alter muscle function (6), possibly by altering excitation-contraction coupling (7). One of the proteins involved in excitation-contraction coupling, the skeletal muscle Ca 2ϩrelease channel, is modulated by both oxidants (8 -10) and NO (6,11). This regulation appears to be controlled, at least in part, by the binding of calmodulin (CaM) to RYR1 (12,13). Ca 2ϩ -free CaM is a partial agonist of RYR1, whereas Ca 2ϩ -CaM is an inhibitor (14). In addition to these direct functional effects, CaM bound to RYR1 protects the channel from oxidation-induced intersubunit cross-linking (12), and conversely, oxidation can block CaM binding to RYR1. In contrast to the effects of oxidation, alkylation of RYR1 with NEM rapidly destroys the ability of RYR1 to bind 35 S-Ca 2ϩ -free CaM (apoCaM) but does not alter its ability to bind 35 S-Ca 2ϩ -CaM. Alkylation also prevents oxidation-induced intersubunit crosslinking. Our studies suggest that there is only one CaM site per subunit of RYR1 at either high or low Ca 2ϩ (13). Some indication of the location of the CaM binding site has been obtained from examination of sites on RYR1 protected from tryptic cleavage by CaM (13). Treatment of RYR1 with trypsin rapidly destroys its ability to bind CaM, but CaM bound to RYR1 can protect its binding site from tryptic digestion. The sites protected by Ca 2ϩ -CaM or apoCaM are at amino acids 3630 and 3637, suggesting that either the binding sites for CaM in both the Ca 2ϩ -free and Ca 2ϩ -bound forms are physically close to this region of RYR1 or that the binding of both forms of CaM produces a conformational change that buries this region of RYR1. The latter possibility seems unlikely because the functional effects of apoCaM are opposite those of Ca 2ϩ -CaM. The sequence between these two sites is AVVACFR, suggesting that cysteine 3635 may be one of the cysteines that, in the absence of CaM, can form the intersubunit disulfide bond. This intersubunit contact site is likely to represent an important site for regulating RYR1 activity. In the current study, we demonstrate that cysteine 3635 is one of the amino acids involved in redox modulation of RYR1.
SR Membrane Preparation-SR membranes were prepared from rabbit leg and backstrap white skeletal muscle and were purified using sucrose gradient centrifugation as described previously (15).
NEM Alkylation and Diamide Oxidation of SR Membranes-SR membranes were reduced with 0.1 mM DTT for 30 min at room temperature. Excess DTT was removed by pelleting in an Airfuge for 5 min at 30 p.s.i. The membranes (1 mg/ml) were then labeled with 10 M [ 3 H]NEM in the presence or absence of CaM for 10 min on ice. Reactions and CaM incubations were carried out in either high Ca 2ϩ buffer (300 mM NaCl, 50 mM MOPS (pH 7.4), 1 mM EGTA, 1.2 mM CaCl 2 ) or low Ca 2ϩ buffer (300 mM NaCl, 50 mM MOPS (pH 7.4), 1 mM EGTA). The reactions were stopped by the addition of DTT to 0.1 mM. Protein was centrifuged in a Beckman TL100.3 rotor for 45 min at 30,000 rpm and resuspended to remove CaM. Diamide was added to 100 M and incubated for 30 min on ice to induce intersubunit cross-links of RYR1. NEM was added to 5 mM to stop the reaction with diamide. Equal amounts of protein were electrophoresed on a 5% Laemmli gel (16).
Generation of Trypsin-digested Complexes of RYR1-SR membranes were trypsinized for 10 min at 37°C at a protease to protein ratio of 1:1000. Digestion was halted by the addition of 20ϫ soybean trypsin inhibitor. After solubilization in 2% digitonin, proteolyzed complexes were purified over a 5-20% linear sucrose gradient containing 0.4% CHAPS for 18 h. Fractions were collected, and peak protein was assessed by a Bradford assay (18).
Protein Sequencing-Proteins were prepared for sequencing as described (13). N-terminal sequencing was performed in the laboratory of Dr. Richard Cook at Baylor College of Medicine.

CaM Protects Both Cysteines Required for Oxidation-induced
Intersubunit Cross-linking from NEM Alkylation-The activity of RYR1 is modulated by modification of its cysteine residues by oxidation (8 -10), alkylation (11,12,19) or nitrosylation (6,7,11,20). The cysteine residues involved in these reactions have not previously been identified. The identification of the cysteines involved in these modulations is difficult because RYR1 has 100 cysteines per subunit. However, the functionally important cysteines appear to be hyper-reactive (21)(22)(23). We have previously shown that the N terminus of RYR1 contains sulfhydryls that react rapidly with NEM (19). Other sulfhydryls also react rapidly with this reagent, in particular, the cysteines that alter the ability of RYR1 to bind CaM. Alkylation of less than 5% of the total RYR1 sulfhydryls completely blocks 35 S-apoCaM binding to RYR1, without altering the ability of RYR1 to bind either 35 S-Ca 2ϩ -CaM or [ 3 H]ryanodine (12). However, both Ca 2ϩ -CaM and apoCaM can protect this site from NEM inactivation (data not shown), suggesting a steric inhibition of 35 S-apoCaM, but not 35 S-Ca 2ϩ -CaM binding.
Are the sulfhydryls that are alkylated under these conditions also those that form the intersubunit disulfide bond? To assess this, we made use of our previous observation that NEM blocks diamide-induced intersubunit cross-linking and asked whether bound CaM could protect both cysteines needed for the crosslink from NEM alkylation (Fig. 1). For this experiment, apoCaM was first bound to RYR1, and the membranes were then treated with NEM. To avoid the effect of CaM on the diamide cross-linking, the bound CaM was removed by washing. The membranes were then treated with diamide under conditions previously shown to generate disulfide bonds be-tween neighboring subunits (19). Alkylation in the absence of apoCaM prevented diamide-induced intersubunit cross-linking (Fig. 1, lane 2). If, however, apoCaM was bound at the time of NEM alkylation and the apoCaM was then removed by washing, diamide-induced intersubunit cross-link was detected (Fig.  1, lane 3). The same result was obtained using Ca 2ϩ -CaM (data not shown), suggesting that CaM (either form) either protects both of the cysteines involved in the intersubunit cross-link or one of the cysteines does not react with NEM even in the absence of CaM. These findings again support a model in which CaM binds at a site of intersubunit contact.
CaM Protects Sulfhydryls within the 160-kDa Tryptic Fragment (N-terminal Amino Acid is 3119) from Alkylation-The CaM binding sites on RYR1 are extremely sensitive to tryptic cleavage (13). We have previously identified the fragments of RYR1 generated by trypsin cleavage (24). One fragment is a 160-kDa polypeptide that begins at amino acid 3119 and extends to the C terminus. A second cleavage generates a smaller fragment that probably ends at amino acid 4475. Bound CaM protects both of these fragments from additional cleavages at amino acids 3630 and 3637. Bound CaM also protects these fragments from labeling with [ 14 C]NEM (Fig. 2), suggesting that these fragments contain the cysteine(s) that regulates CaM binding. Also seen in Fig. 2 are two bands (bands c and f) that are very heavily labeled by [ 14 C]NEM under these conditions, but their labeling is not altered by the presence of calmodulin. We have previously identified these bands as fragments of the N terminus (24). The bands obtained after labeling in low Ca 2ϩ have also been excised and digested, and the radiolabel was quantified. We find a 65 Ϯ 13% (n ϭ 3) decrease in radioactivity in the 160-kDa band and a 41 Ϯ 19% (n ϭ 3) decrease in the 150-kDa when labeling is carried out in the presence of CaM.
Localization of the RYR1 Cysteine Protected by Bound CaM-An examination of the sequence between the two CaM protected cleavage sites shows a cysteine at position 3635. To test whether this cysteine modulates CaM binding, we incubated membranes in the presence and absence of CaM at M Ca 2ϩ and then reacted with [ 3 H]NEM. In these experiments, CaM was then removed by successive washes, and the membranes were digested with trypsin. This treatment cleaves RYR1 after amino acids 3630 and 3637, producing two fragments in approximately equal amounts that differ by 7 amino acids. The N-terminal sequence of fragment 1 that begins at amino acid 3630 is AVVACFRMTPLYNLPT, and fragment 2 that begins at amino acid 3637 is MTPLYNLPTHRAC. In fragment 1 there are cysteines that will be detected at sequencing cycles 5 and 20, whereas fragment 2 has a cysteine that would be detected at cycle 13. Half of the sample from each round of sequencing was collected for scintillation counting. The 3 H radiolabel was detected only in the fifth cycle of sequencing, corresponding to cysteine 3635 in fragment 1 (Fig. 3). Fragment 2 has a tyrosine at cycle 5. The incorporation was greatly decreased by the presence of CaM during the labeling (Fig. 3). No radiolabel was detected in cycle 13 or 20, the position of cysteine 3650 in fragments 2 and 1, respectively.
Summary and Conclusions-Our data suggest that one potential site of redox modulation of RYR1 is cysteine 3635 and that the reactivity of this residue is regulated by the binding of CaM to RYR1. CaM protects RYR1 from oxidation-induced intersubunit cross-linking and alkylation-induced inactivation of the apoCaM-binding site. Our findings support the assignment of cysteine 3635 as the site of NEM alkylation that blocks apoCaM binding. The question is whether this cysteine is also one of the two cysteines involved in the intersubunit cross-link. NEM can prevent the formation of the diamide-induced disul- fide, but not if CaM is bound to RYR1, demonstrating that either both of the intersubunit disulfide bond-forming cysteines are protected by CaM or that one of these cysteines does not react with NEM even in the absence of CaM. Although we cannot completely eliminate the possibility that the cysteines involved in the intersubunit cross-link are allosterically modulated by CaM binding, this seems unlikely because both apoCaM and Ca 2ϩ -CaM, which have opposite effects on chan-nel activity, can protect against oxidation-induced intersubunit cross-linking, alkylation of cysteine 3635, and trypsin cleavage at amino acids 3630 and 3637. We propose that cysteine 3635 can be disulfide bonded to a cysteine on a neighboring subunit of the RYR1 tetramer.
Another question that arises is why alkylation of cysteine 3635 leads to selective loss of 35 S-apoCaM binding when either apoCaM or Ca 2ϩ -CaM can protect the site from alkylation. One possibility is that the introduction of the bulky maleimide group sterically hinders apoCaM but not Ca 2ϩ -CaM binding. The conformation of CaM bound to RyR1 is unknown, but the amino acids on RYR1 participating in the binding of these two forms of CaM are likely to be quite different.
Another unknown is the identity of the cysteine on the neighboring subunit that forms the oxidation-induced disulfide bond with cysteine 3635. Bound CaM can protect this site to some extent from NEM alkylation, but we have not yet determined where it is located. Earlier cross-linking studies have suggested a location between amino acids 2100 and 2840 (19). There are, however, 18 cysteines in this region of RYR1 and none that are close enough to a tryptic cleavage site for Nterminal sequencing. We are currently exploring the use of other proteolytic enzymes to identify this cysteine.
In summary, our data suggest that CaM, bound at a site of intersubunit contact, protects RYR1 from oxidation and possibly nitrosylation. High concentrations of oxidants could, however, lead to a loss of the interaction between RYR1 and CaM. Such an effect might contribute to altered muscle function or damage during periods of high oxidative stress, such as in fatigue.  (25). Identification of these bands was on the basis of N-terminal sequence and Western blotting with sequence-specific antibodies. The relative mobility of the fragments on this gel system does not accurately reflect their mass.
FIG. 3. Alkylation of cysteine 3635 is blocked by CaM. As in Figs. 1 and 2, SR membranes were reduced, washed by centrifugation, and resuspended at 1 mg/ml in high Ca 2ϩ buffer. Samples were incubated for 30 min at room temperature in the presence and absence of 10 M CaM. [ 3 H]NEM was added to both samples to 1 M. After incubation for 10 min on ice, the alkylation was terminated with 5 mM DTT. Trypsin treatment and isolation of tryptic complexes was performed as described previously (24). Samples were electrophoresed on Schagger gels (17) and transferred to Immobilon-S for sequencing. Half of each sample at each cycle of sequencing was collected for scintillation counting. The figure shows the [ 3 H]NEM cpm per cycle of sequencing for both samples. Open symbols (E) represent samples obtained from the alkylation in the presence of CaM, closed symbols (q) indicate those alkylated in the absence of CaM. This experiment was performed twice with similar results.