Classes of Thiols That Influence the Activity of the Skeletal Muscle Calcium Release Channel*

The skeletal muscle Ca 2 1 release channel/ryanodine receptor (RyR1) is a prototypic redox-responsive ion channel. Nearly half of the 101 cysteines per RyR1 subunit are kept in a reduced (free thiol) state under conditions comparable with resting muscle. Here we as-sessed the effects of physiological determinants of cellular redox state (oxygen tension, reduced (GSH) or oxidized (GSSG) glutathione, and NO/O 2 . (released by 3-morpholinosydnonimine)) on RyR1 redox state and activity. Oxidation of ; 10 RyR1 thiols (from ; 48 to ; 38 thiols/RyR1 subunit) had little effect on channel activity. Channel activity increased reversibly as the number of thiols was further reduced to ; 23/subunit, whereas more extensive oxidation (to ; 13 thiols/subunit) inactivated the channel irreversibly. Neither S -nitrosylation nor tyrosine nitration contributed to these effects. The results identify at least three functional classes of RyR1 thiols and suggest that 1) the channel may be protected from oxidation by a large reservoir of functionally inert thiols, 2) the channel may be designed to respond to moderate oxidative stress by a change in activation setpoint, and 3) the channel is susceptible to oxidative in-jury SR vesicles, or without SIN-1 in the or of GSH or GSSG at p O 2 of ; or mm Hg, centrifuged (100,000 3 g ) at 4 for 1 h. The pellets were and resuspended and probed with an excess (500 m M ) of the lipophilic, thiol-specific agent mBB fo r 1 h in thedark at 24 °C. Following mBB treatment, SR vesicles were solubilized with 1.5% CHAPS, and the bimane-labeled RyR1 was isolated by sucrose density gradient centrif-* statis-tically significant

Ca 2ϩ release channels/ryanodine receptors (RyRs) 1 are the largest known ion channels, consisting of four ϳ565-kDa RyR subunits and four associated 12-kDa FK506-binding protein subunits (1,2). Following an action potential, the cardiac and skeletal muscle RyR isoforms release Ca 2ϩ from an intracellular Ca 2ϩ storing membrane compartment, the sarcoplasmic reticulum (SR), in a process known as excitation-contraction coupling. Numerous endogenous effectors are known to regulate RyR function and therefore muscle contractility. These effectors range from ions (Ca 2ϩ and Mg 2ϩ ) to other small molecules (adenine nucleotides) to polypeptides such as calmodulin (1,2). Recent work has also established RyRs as prototypic redox-sensitive ion channels. The skeletal muscle isoform of RyR (RyR1) contains a large number of free thiols: as many as 50 out of a total of 101 cysteine residues/subunit (100 cysteines/ RyR1 subunit (3) and 1 cysteine/FK506-binding protein subunit (4)) (5). RyR1 channel activity is dramatically altered by redox modifications of critical thiols (oxidation, S-nitrosylation, or alkylation) (5)(6)(7)(8)(9)(10)(11)(12)(13)(14). Conversely, RyR1 has thiols whose redox potential is dependent on effectors that regulate RyR1 activity such as Ca 2ϩ and Mg 2ϩ (15). In a physiological context, nitric oxide (NO) and reactive oxygen species are produced in contracting muscle and have been shown to modulate in vitro RyR redox state and channel activity (5, 12, 16 -21). Remarkably, RyR1 redox state and function are dependent on O 2 tension (5). Altering O 2 tension alone dynamically reduced/oxidized as many as 6 -8 thiols/RyR1 subunit and thereby modified channel responsiveness to physiological concentrations of NO (5).
In this study, we varied reducing and oxidizing conditions to explore in detail the effect of RyR1 redox state on channel function. At one end of the redox spectrum, RyR1 was maintained in a highly reduced state by GSH at low pO 2 (ϳ10 mm Hg). At the other end of the spectrum, we tested the effects of the strongly oxidizing conditions produced by high concentrations of NO/O 2 . (released by 3-morpholinosydnonimine SIN-1).
Our studies demonstrate that 1) RyR1 thiols can be grouped into at least three distinct functional classes, 2) RyR1 channel function peaks in moderately oxidizing conditions (ϳ23 free thiols/RyR1 subunit), 3) channel closure is favored by strongly reducing (Ն38 free thiols/RyR1 subunit) or oxidizing (Յ15 free thiols/RyR1 subunit) conditions, and 4) NO and O 2 . , which are produced in vivo (17), have concentration-dependent effects on channel function that are mediated by thiol oxidation and may have both physiological and pathological relevance.

Materials-[ 3 H]Ryanodine was obtained from
PerkinElmer Life Sciences; unlabeled ryanodine and monobromobimane (mBB) from Calbiochem; SIN-1 from Molecular Probes, Inc. (Eugene, OR); and CHAPS, leupeptin, and Pefabloc (a protease inhibitor) from Roche Molecular Biochemicals. All other chemicals were of analytical grade.
Quantification of RyR1 Thiol and S-Nitrosothiol Contents-The number of free thiols in RyR1 was determined as described previously (5). Briefly, skeletal SR vesicles, treated with or without SIN-1 in the absence or presence of GSH or GSSG at pO 2 of ϳ10 or 150 mm Hg, were centrifuged (100,000 ϫ g) at 4°C for 1 h. The pellets were washed and resuspended and then probed with an excess (500 M) of the lipophilic, thiol-specific agent mBB for 1 h in the dark at 24°C. Following mBB treatment, SR vesicles were solubilized with 1.5% CHAPS, and the bimane-labeled RyR1 was isolated by sucrose density gradient centrif-* 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.This work was supported by National Institutes of Health Grants HL04053 (to J. P. E.); HL52529, ES-09206, and HL59130 (to J. J. S.); and AR18687 and HL27430 (to G. M. . , superoxide anion; SIN-1, ugation. The fluorescence intensity of bimane (i.e. the thiol content) in the sucrose gradient fraction most enriched in RyR1 (Ͼ95% purity) was determined and normalized for protein concentration as described (5). The S-nitrosothiol content of RyR1 was determined by isolating the receptor without prior mBB treatment and using a photolysis/chemiluminescence method (5,12).
[ 3 H]Ryanodine Binding-Unless otherwise indicated, skeletal SR vesicles, treated with various concentrations of SIN-1 in the absence or presence of GSH or GSSG at pO 2 of ϳ10 or 150 mm Hg, were incubated with 5 nM [ 3 H]ryanodine at 24°C in media containing 0.125 M KCl, 20 mM imidazole, pH 7.0, 0.3 mM Pefabloc, 30 M leupeptin, and the indicated concentrations of free Ca 2ϩ . Nonspecific binding was determined using a 1000-fold excess of unlabeled ryanodine. After 5 h, aliquots of the samples were diluted with 20 volumes of ice-cold water and placed on Whatman GF/B filters soaked with 2% (w/w) polyethyleneimine. Filters were washed with three 5-ml volumes of ice-cold buffer, and the radioactivity remaining on the filters was determined by liquid scintillation counting to obtain bound [ 3 H]ryanodine.
Single Channel Recordings-Single channel measurements were performed by fusing skeletal SR vesicles with Mueller-Rudin type bilayers containing phosphatidylethanolamine, phosphatidylserine, and phosphatidylcholine in the ratio 5:3:2 (25 mg of total phospholipid/ml n-decane) (5). The side of the bilayer to which the SR vesicles were added was defined as the cis (cytoplasmic) side. The trans (SR luminal) side of the bilayer was defined as ground. Single channels were recorded in a symmetric CsCH 3 SO 3 buffer solution (0.25 M CsCH 3 SO 3 , 10 mM Cs-HEPES, pH 7.3) containing the additions indicated (see Fig. 4). Measurement of the sensitivity of the channels to cytosolic Ca 2ϩ indicated that in a majority of recordings (Ͼ98%) the cytosolic side of the RyRs faced the cis side, and the luminal side faced the trans side of the bilayer. Electrical signals were filtered at 2 kHz, digitized at 10 kHz, and analyzed. Data acquisition and analysis were performed with a commercially available software package (pClamp 6.0.4; Axon Instruments, Burlingame, CA) with an IBM-compatible Pentium II computer and 12-bit A/D to D/A converter (Digidata 1200, Axon Instruments).
Other Biochemical Assays-Free Ca 2ϩ concentrations were obtained by including in the solutions the appropriate amounts of Ca 2ϩ and EGTA as determined using the stability constants and computer program published by Shoenmakers et al. (24). Free Ca 2ϩ concentrations Ͼ1 M were verified with the use of a Ca 2ϩ -selective electrode (World Precision Instruments, Inc.). The protein concentrations were determined by the Amido Black method (25).
Ca 2ϩ -ATPase activity in SR vesicles was assayed by malachite green ATPase method (26) in the presence of a Ca 2ϩ ionophore (1 M ionomycin) and the absence or presence of SIN-1. Mg 2ϩ -ATPase remaining in the SR preparations was subtracted from the total ATPase by adding 1 mM EGTA to assay media.
Data Analysis-Results are given as means Ϯ S.D. with the number of experiments in parentheses. Significance of differences of data was analyzed with Student's t test. Differences were regarded to be statistically significant at p Ͻ 0.05 (*) and p Ͻ 0.01 (**).

Correlation of RyR1
Redox State and Activity-The highly specific plant alkaloid ryanodine is widely used as a probe of RyR channel activity because of its preferential binding to open channel states (1, 2). In Fig. 1 oxygen tension (Fig. 1B) but had only moderate effects at pO 2 ϳ10 mm Hg (Fig. 1A). Finally, in the presence of 5 mM GSH, SIN-1 effects on [ 3 H]ryanodine binding were almost entirely suppressed. This latter observation is consistent with GSH being a potent scavenger for reactive nitrogen and oxygen species (31).
Recent studies have indicated that oxidation and S-nitrosylation affect channel activity by altering RyR1's interaction with calmodulin (5,13). SR vesicles contain 0.10 -0.15 calmodulin/RyR1 subunit (5). To eliminate the effects of endogenous calmodulin, SR vesicles were pretreated with 2 M myosin light chain kinase-derived calmodulin binding peptide, followed by centrifugation through a layer of 0.3 M sucrose to remove complexed calmodulin and calmodulin binding peptide. After centrifugation, the endogenous SR-associated calmodulin content was reduced to ϳ5% of the control value, as determined by a phosphodiesterase activation assay (5). The effects of SIN-1 on RyR1 activity, as measured by [ 3 H]ryanodine binding under the conditions in Fig. 1, were then studied by exposing control and pretreated vesicles to 0, 0.2, and 1.0 mM SIN-1 at pO 2 ϳ150 mm Hg. An essentially identical activation by 0.2 mM SIN-1 and inactivation by 1.0 mM SIN-1 for vesicles pretreated and not pretreated with the calmodulin binding peptide (data not shown) indicated that SIN-1 did not transduce its effects in Fig.  1 via the small amounts of calmodulin associated with the SR vesicles.
The oxidation or reduction of a large number of thiols is the principle mechanism by which O 2 tension and reducing agents such as glutathione modulate RyR1 channel activity in SR vesicles. A specific free thiol-labeling agent, mBB, was used to correlate RyR1 free thiol content and activities in the presence of redox active molecules. As reported previously (5), RyR1 in SR vesicles exposed to 5 mM GSH had ϳ48 and ϳ40 free thiols/subunit at pO 2 ϳ10 or ϳ150 mm Hg, respectively. Expo The effects of SIN-1 on RyR1 activity were also explored in single channel measurements using maximally activating concentrations of Ca 2ϩ (10 M; Fig. 3) and SIN-1 (0.2 mM; Fig. 1). Skeletal SR vesicles were incorporated into planar lipid bilayers, and RyR1 single channel activity was recorded at ambient oxygen tension with Cs ϩ as the current carrier to eliminate other ion currents also present in SR vesicles (32). In the  (27,28). Therefore, peroxynitrite is probably a dominant oxidative species in experiments involving SIN-1. Rapid formation of peroxynitrite was supported by the finding that a NO electrode with a high sensitivity (5) failed to detect any NO release from SIN-1 in our assay conditions at either oxygen tension (data not shown). Peroxynitrite oxidizes thiols reversibly to disulfide bonds or sulfenic (SOH) acids or irreversibly to sulfinic (SO 2 H) or sulfonic acids (SO 3 H) (33)(34)(35). The reversibility of RyR1 oxidation was determined at ambient oxygen tension from the thiol content of SR vesicles that were first treated with 0, 0.2, or 1.0 mM SIN-1 at 24°C for 5 h and then exposed to 5 mM GSH (experimental group) or no reducing equivalent (control) for another 5 h. [ 3 H]Ryanodine binding and free RyR1 thiol content were determined after the final 5 h of incubation. The results are summarized in Fig. 5, with the number of free thiols per RyR1 subunit (mean Ϯ S.D., n Ն 3) labeled at the top of each column. Channel activation and thiol oxidation by 0.2 mM SIN-1 were nearly completely reversed by 5 mM GSH. In contrast, 5 mM GSH could not reverse the effects of 1.0 mM SIN-1 on RyR1 channel activity or redox state (Fig. 5, third pair of columns). These results suggest, but do not prove, that at high concentrations of peroxynitrite, numerous RyR1 thiols (ϳ10/RyR1 subunit) proceeded to high degrees of oxidation.
Peroxynitrite may also S-nitrosylate free thiols in proteins (33,34). We have shown previously that S-nitrosylation of a single thiol per RyR1 can activate the channel (5). To see if RyR1 is S-nitrosylated by SIN-1, control and SIN-1 (0.2 mM)treated samples were assayed for S-nitrosothiol content using a photolysis/chemiluminescence method (5,12). In the control group, there were ϳ0.4 S-nitrosothiol/RyR1 subunits, a level comparable with the endogenous amount of S-nitrosylation found in our previous study (5). SIN-1/peroxynitrite did not S-nitrosylate any additional RyR1 thiols (Table I). Therefore, unlike NO (5), SIN-1/peroxynitrite did not activate RyR1 by S-nitrosylation.
Peroxynitrite can also modify proteins by the addition of a nitro group to the ortho position of tyrosine to form nitrotyrosine (33,34). In the case of the SR Ca 2ϩ -ATPase, in vitro exposure of skeletal SR vesicles to peroxynitrite resulted in both S-nitrosylation (36) and nitrotyrosine formation (37). Both modifications contributed to inhibition of SR Ca 2ϩ -ATPase activity. Nitrotyrosine formation could therefore represent an additional mechanism by which SIN-1 modulates RyR1, and this possibility was examined in immunoblots. A polyclonal antibody recognizing nitrotyrosine (Fig. 6A, lanes 4 -6) did not detect any nitrotyrosine formation in the region of the blots corresponding to RyR1 (Fig. 6A, lanes 1-3) in controls and SR vesicles exposed to 0. kDa. The results suggested that a 100-kDa protein such as the skeletal muscle SR Ca 2ϩ -ATPase has endogenous nitrotyrosine(s) and that its level is increased by exogenous SIN-1/ peroxynitrite (37). In favor of this interpretation, SIN-1 (0.2 or 1.0 mM) inhibited SR Ca 2ϩ -ATPase activity by about ϳ30 and 50%, respectively (Fig. 6B).
Taken together, the results suggest that SIN-1 acts in our assay conditions as an oxidant rather than a NO donor. In the absence of glutathione, NO/O 2 . reversibly activates RyR1 at low concentrations and irreversibly inactivates the channel at high concentrations. Furthermore, modulation of RyR1 by NO/O 2 .
appears to be due to oxidation of thiols, since neither S-nitrosylation nor tyrosine nitration of RyR1 was detected.

DISCUSSION
Contracting muscle produces reactive oxygen and nitrogen species (16,17,38,39). A functional role of these molecules is indicated by the finding that force development in muscle is affected using nitric-oxide synthase inhibitors and scavengers of superoxide. The data we have presented here imply that the effects of NO/O 2 . -related species are mediated, at least in part, through the SR Ca 2ϩ release channel in skeletal muscle. The RyR1 is exquisitely sensitive to redox modulation. Ample evidence indicates that RyRs are activated or inactivated or both by sulfhydryl-modifying molecules (5-15, 18 -21, 40, 41). We previously uncovered a striking plasticity of redox state in RyR1 channel activity (5). RyR1 free thiol content and channel activity were dynamically controlled by GSH/GSSG, oxygen tension, and NO. Other recent studies have shown that RyR1 responds to changes in transmembrane glutathione redox potential (14) and contains thiols whose redox potential is dependent on ligands (Ca 2ϩ , Mg 2ϩ ) that control RyR1 activity (15).
In this study, we used SIN-1 to probe the effect of redox state on RyR1 activity and have done so under a spectrum of redox conditions that are encountered in muscle, by further varying GSH/GSSG and pO 2 . SIN-1 spontaneously generates NO and O 2 . (27), both of which rapidly combine in a 1:1 stoichiometry to form the highly oxidative species peroxynitrite, although NO/O 2 . chemistry may produce other species as well (28,33).
While NO and O 2 . are modulators of muscle contractility (16, 17, 38 -40), we consider it unlikely that NO reacted with the RyR1, since we were unable to detect it with a NO electrode. Furthermore, SIN-1 and free NO modulate the skeletal muscle Ca 2ϩ release channel by two very different redox-related mechanisms. NO activates the channel at pO 2 ϳ10 mm Hg by S-nitrosylation of a single cysteine per RyR1 subunit (5), whereas SIN-1 affected channel activity by the oxidation of a large number of thiols. NO/O 2 . generation is likely to be encountered in normally functioning muscle and in pathological states. Indeed, immunoblotting with an anti-nitrotyrosine antibody revealed endogenous and SIN-1-mediated nitration in a protein that co-migrated with the SR Ca 2ϩ -ATPase, confirming a previous report (37). Notably, the anti-nitrotyrosine antibody failed to detect endogenous or SIN-1-mediated nitration of tyrosines in RyR1. Exposure to the three redox variables O 2 , glutathione, and SIN-1 revealed three distinct redox states of RyR1. The free thiol content ranged from as low as 13 to as high as 48 thiols/ RyR1 subunit (SH/RyR1 s ) (see scheme below).   1 and 4) or presence of 0.2 mM (lanes 2 and 5) or 1.0 mM (lanes 3 and 6) SIN-1. Proteins separated by 3-15% gradient SDS-PAGE were transferred to nitrocellulose membranes and probed with an anti-RyR1 (lanes 1-3) or anti-nitrotyrosine (lanes 4 -6) antibody. The anti-nitrotyrosine antibody did not detect nitrotyrosines in RyR1. However, a protein with a molecular mass of ϳ100 kDa was recognized by the antibody, and SIN-1 significantly increased the level of nitrotyrosines. B, SIN-1 (0.2 or 1.0 mM) significantly decreased SR Ca 2ϩ -ATPase activity. Data are the means Ϯ S.D. of three experiments. Compared with control, double asterisks represent p Ͻ 0.01. the whole Ca 2ϩ concentration range). Extensive oxidative inactivation (1.0 mM SIN-1 in the absence and presence of 5 mM GSSG) was attributed to the loss of up to 10 additional thiols (ϳ13 free thiols/RyR1 subunit remain) and was irreversible, making it likely to be of pathological rather than physiological significance. Therefore, RyR1 redox state and function appear to be intimately linked.

Moderate oxidation
This analysis of RyR1 is reminiscent of our previous studies of RyR2 (12). In the case of RyR1, however, the oxidation of up to 25 thiols was reversible (suggesting the formation of disulfides and/or sulfenic acids), whereas only 5-6 thiols could be oxidized in the cardiac channel before irreversible changes were encountered. Irreversible thiol oxidation (suggesting oxidation to sulfinic or sulfonic acids) was dependent on the concentration of the oxidant (1.0 mM SIN-1) and required the absence of GSH. Although the (patho)physiological correlative of thiol oxidation remains to be elucidated, the finding of endogenous nitration in the SR (Ref. 37 and this study) indicates that such oxidative modifications are likely (i.e. thiols are generally more reactive toward NO/O 2 . and peroxynitrite than tyrosines (42)). Intracellular Ca 2ϩ concentration, the main determinant of skeletal muscle contractile function, is controlled by the RyR1 and an ATP-driven Ca 2ϩ pump, with the former releasing the stored Ca 2ϩ from SR to initiate contraction and the later sequestering Ca 2ϩ back in SR to initiate relaxation. We previously showed that O 2 tension dynamically reduced/oxidized 6 -8 thiols/RyR1 subunit. The alteration of channel redox state determined its responsiveness to S-nitrosylation by NO of one cysteine per RyR1 subunit, and the effect of S-nitrosylation on channel activity was transduced via calmodulin (5). We believe that such regulation may impact on excitation-contraction coupling. On the other hand, we now show that NO/O 2 . can transduce its effects on RyR1 independently of calmodulin, thus providing an additional means by which the redox state of the cell may influence RyR1 function. But these findings probably bear on situations characterized by oxidative stress, where we envision that the RyR1 would counter impairments of force production imposed by oxidizing conditions with oxidation-induced activation. Ultimately, however, irreversible oxidation of RyR1 thiols may contribute to dysfunction of muscle. The relevance of these findings to conditions such as in fatigue, spasm, and rhabdomyolysis remains to be shown.