Nitric Oxide, NOC-12, and S -Nitrosoglutathione Modulate the Skeletal Muscle Calcium Release Channel/Ryanodine Receptor by Different Mechanisms AN ALLOSTERIC FUNCTION FOR O 2 IN S -NITROSYLATION OF THE CHANNEL*

The skeletal muscle Ca 2 (cid:1) release channel/ryanodine receptor (RyR1) contains (cid:1) 50 thiols per subunit. These thiols have been grouped according to their reactivity/ responsiveness toward NO, O 2 , and glutathione, but the molecular mechanism enabling redox active molecules to modulate channel activity is poorly understood. In the case of NO, very low concentrations (submicromo-lar) activate RyR1 by S -nitrosylation of a single cysteine residue (Cys-3635), which resides within a calmodulin binding domain. S -Nitrosylation of Cys-3635 only takes place at physiological tissue O 2 tension (pO 2 ; i.e. (cid:1) 10 mm Hg) but not at pO 2 (cid:1) 150 mm Hg. Two explanations have been offered for the loss of RyR1 responsiveness to NO at ambient pO 2 , i.e. Cys-3635 is oxidized by O 2 versus O 2 subserves an allosteric function (Eu, J. P., Sun, J. H., Xu, L., Stamler, J. S., and Meissner, G. (2000) Cell 102, 499– 509). Here we report that the NO donors NOC-12 and

The large, homotetrameric skeletal muscle Ca 2ϩ release channel/ryanodine receptor (RyR1) 1 contains several classes of regulatory thiols. These classes are distinguished by reactivity or responsiveness to O 2 tension (pO 2 ) (1, 2), redox active molecules such as glutathione (3) and nitric oxide (NO) (1), transmembrane glutathione redox potential (4), and allosteric effector molecules (Ca 2ϩ , Mg 2ϩ ) (5). It has recently been shown that cysteine 3635, which is localized to the calmodulin (CaM) bind-ing domain of RyR1 (6 -8), confers responsiveness to NO. In contrast, the identities of the remaining regulatory thiols are not known. NO forms a covalent bond with the thiol group of Cys-3635 (i.e. S-nitrosylation) in vivo and thereby reverses the inhibitory effect of CaM on the channel (6). Full-length RyR1 channels with an alanine residue substituted for Cys-3635 are not S-nitrosylated by physiological concentrations of NO, and channel activity is unaffected by NO. S-nitrosylation of Cys-3635 only occurs at low O 2 tension (pO 2 ϳ10 mm Hg, comparable with that found in skeletal muscle in vivo) (1,6). At this pO 2 , 6 -8 (of ϳ50) thiols per RyR1 subunit are actively maintained in the reduced state (1). Thus, one explanation for the failure of NO to S-nitrosylate RyR1 at ambient pO 2 is that Cys-3635 is oxidized. An alternative possibility is that the oxidation of pO 2 -sensitive thiols leads to a change in channel conformation; in this state S-nitrosylation of Cys-3635 is unfavorable. Alternatively stated, O 2 is either serving as an oxidant (of Cys-3635) or as an allosteric effector (of Cys-3635 reactivity).
NO donors, compounds capable of donating NO and redox active forms thereof, are widely used to mimic the effects of NO synthase (9). A number of these compounds are capable of modulating RyR1 activity (1, 10 -15). RyR1 contains a large number of reactive thiols (1,2), and the action of NO donors may differ widely depending on the mechanisms and rates of NO release, the chemistry of NO group transfer, the base structure of the NO donor compound, and the reactivity of substrate thiol. In particular, members of the S-nitrosothiol (SNO) class of NO donors can modulate protein function by transnitrosylation as well as NO release (16,17). In contrast, the NONOate class of NO donors is thought to be less susceptible to transnitrosylation chemistry (18). It is important to note, however, that NONOate compounds may directly interact with proteins through polyamine recognition sites and/or through ionic interactions.
In the present study, we examined the activation of the skeletal muscle Ca 2ϩ release channel by NOC-12 and GSNO, an endogenous S-nitrosothiol, and compared their effects to solutions of NO. We found that both NOC-12 and GSNO activated RyR1 independently of O 2 tension and that the NO scavenger, C-PTIO, blocked the effects of both. But whereas NOC-12 mediated its effects by S-nitrosylation of a single cysteine (Cys-3635), GSNO activation involved the S-nitrosylation and oxidation of multiple thiols. Moreover, Cys-3635 was not required for activation by GSNO. Thus, NO, NOC-12, and GSNO activate the prototypic redox-sensitive RyR1 channel by different mechanisms, and the effect of O 2 tension on S-nitrosylation by NO is best rationalized by an allosteric mechanism.

Materials-[ 3 H]
Ryanodine was a product of PerkinElmer Life Sciences. CaM was obtained from Sigma. NO donors, monobromobimane, myosin light chain kinase-derived CaM binding peptide and anti-Snitrosocysteine polyclonal antibody were from Calbiochem, and leupeptin and Pefabloc (protease inhibitors) were from Roche Molecular Biochemicals. An ECL detection reagent kit was from Amersham Biosciences. NO gas (purity Ͼ99%, National Welders) was scrubbed to remove O 2 and nitrite by passing through an argon-purged column filled with KOH pellets and then a solution of NaOH. The concentration of NO was determined by a hemoglobin titration assay and an NO electrode (WPI Instruments) as described (1). All other chemicals were of analytical grade.
Sample Preparations-Skeletal muscle sarcoplasmic reticulum (SR) vesicles enriched in RyR1 were prepared from rabbit skeletal muscle in the presence of protease inhibitors (19). The construction and expression of wild type (WT) and C3635A mutant RyR1s have been described (6). WT and C3635A RyR1s were expressed in HEK293 cells, and crude membrane fractions were prepared as described (6).
Quantification of RyR1 Free Thiols and S-Nitrosothio1s-RyR1 free thiol (SH) and SNO contents were determined by the monobromobimane fluorescence method and a photolysis/chemiluminescence-based NO detection assay, respectively (1).
Electrophoresis and Detection of S-Nitrosocysteine on Western Blots-All procedures were performed under non-reducing conditions (6). Membranes were incubated in 0.125 M KCl, 20 mM imidazole, pH7.0, and 8 M free Ca 2ϩ for 1 h at 24°C in room air in the absence and presence of NOC-12 or GSNO. Protein samples were separated by 3-20% SDS-PAGE under non-reducing conditions and transferred to polyvinylidene difluoride membranes. The membranes were blotted with 5% nonfat milk in 0.05% Tween 20 phosphate-buffered 0.1 M saline solution at 24°C for 2 h and probed with anti-S-nitrosocysteine polyclonal antibody (Calbiochem; 1:500) and secondary peroxidase-conjugated anti-rabbit IgG antibody (Calbiochem; 1:2000). Anti-S-nitrosocysteine signals were detected with an ECL kit (Amersham Biosciences). After that, the membranes were re-probed with anti-RyR1 monoclonal antibody D110 (1:10) and peroxidase-conjugated anti-mouse IgG (Calbiochem, 1:2000) using the ECL detection method. Single Channel Recordings-Single channel measurements were performed at room air by fusing RyR1-containing membrane fractions with Mueller-Rudin-type bilayers containing phosphatidylethanolamine, phosphatidylserine, and phosphatidylcholine in the ratio 5:3:2 (25 mg of total phospholipid per milliliter of n-decane) (1,20). The side of the bilayer to which the RyR1-containing membrane fractions were added was defined as the cis (cytoplasmic) side. The trans (lumenal) side of the bilayer was defined as ground. Single channels were recorded in the buffer solutions given in the legends to Figs. 3 and 5. Measurement of the sensitivity of the channels to cytosolic Ca 2ϩ indicated that in a majority of recordings (Ͼ98%) the cytosolic side of RyR1 faced the cis side and the lumenal side faced the trans side of the bilayer. Electrical signals were filtered at 2 kHz, digitized at 10 kHz, and analyzed with a commercially available software package (pClamp 8.2, Axon Instruments, Foster City, CA). P o values in multichannel recordings were calculated using the equation P o ϭ ⌺ iP o,i /N, where N is the total number of channels, and P o,i is channel open probability of the i th channel.
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 Schoenmakers et al. (21). Free Ca 2ϩ concentrations were verified with the use of a Ca 2ϩ selective electrode. The protein concentrations were determined by the Amido Black method (22).
Data Analysis-Results are given as means Ϯ S.D. unless otherwise indicated. 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.  half-life times of ϳ6.5 and 2.8 h, respectively (Table I). In a majority of the experiments, we matched the NO peak concentrations by comparing the groups treated with 0.1 mM NOC-12 with those treated with 0.2 mM GSNO. There was no difference in peak concentrations of NO released by either donor as a function of pO 2 (pO 2 ϳ10 mm Hg versus ϳ150 mm Hg) (data not shown). The half-life time of NO was ϳ10 min.

Release of NO by NOC-12 and GSNO-NOC
O 2 Tension-independent Modulation of RyR1 by NOC-12 and GSNO-Modulation of RyR1 by NO is O 2 tension dependent; only at a pO 2 comparable with that found in skeletal muscle in vivo (pO 2 ϳ10 mm Hg) can physiological amounts of NO (submicromolar) S-nitrosylate and activate RyR1 (1). In Fig. 1, SR vesicles were treated with increasing concentrations of NO, NOC-12, or GSNO, and RyR1 activities were determined by [ 3 H]ryanodine binding at pO 2 ϳ10 mm Hg (Fig. 1A) or at pO 2 ϳ150 mm Hg (Fig. 1B). Ryanodine is a highly specific plant alkaloid that is widely used as a probe of channel activity because of its preferential binding to the open channel states (25,26). As shown previously (1), only at pO 2 ϳ10 mm Hg did NO (1-10 M) cause a significant increase in [ 3 H]ryanodine binding (Fig. 1A) (Fig. 2A) and a decrease at 0.3 M free Ca 2ϩ (Fig. 2B). NOC-12 caused an increase in [ 3 H]ryanodine binding in the presence of CaM but not after CaM had been sequestered (Fig. 2, A and B). In contrast, GSNO caused an additional enhancement of RyR1 channel activity even after endogenous CaM sequestration. These results support the idea that NOC-12 controls RyR1 via the S-nitrosylation of Cys-3635, which is found in the CaM binding region of RyR1. On the other hand, redox modulation by GSNO does not appear to be dependent on S-nitrosylation or oxidation of Cys-3635. More definitive evidence for the role of Cys-3635 in the redox modulation of RyR1 is given below using a RyR1 construct with a Cys-3635 to Ala substitution.
Modulation of RyR1 Single Channel Activities by NOC-12 and GSNO-The ability of the two NO donors to activate RyR1 under ambient oxygen tension was confirmed in single channel recordings. Skeletal SR vesicles were incorporated into planar lipid bilayers, and single RyR1 channels were recorded with Cs ϩ as the current carrier. As shown in Fig. 3A activated the channel (Fig. 3B). Fig. 3C shows that the averaged channel open probability (P o ) of RyR1 tripled after the addition of 0.1 mM NOC-12 or 0.2 mM GSNO. Thus, both [ 3 H]ryanodine binding and single channel measurements show that under comparable conditions these two NO donors activate the RyR1 to the same extent.
Redox-related Basis of RyR1 Modulation by NOC-12 and GSNO-We next determined whether modulation of RyR1 by NO donors involved the formation of a single SNO per RyR1 subunit, as was shown previously for NO at pO 2 ϳ10 mm Hg (1). We thus determined both the free thiol and SNO content of RyR1s treated with NOC-12 or GSNO at pO 2 ϳ150 mm Hg. Exposure of SR vesicles to 0.1 or 1.0 mM NOC-12 increased [ 3 H]ryanodine binding to a similar extent and reduced the RyR1 thiol content by ϳ1 per RyR1 subunit, which was accounted for by the formation of ϳ1 SNO per RyR1 subunit (Table II). The stoichiometry of 1 SNO/RyR1 subunit agreed with that obtained by exposure to 0.75 M NO at pO 2 ϳ10 mm Hg (1). 0.1 mM NOC-12 optimally activated RyR1 in single channel recordings in less than 1 min (Fig. 3A).
In contrast to NOC-12, 0.2 mM GSNO activated RyR1 at ambient O 2 tension via the S-nitrosylation or oxidation of multiple thiols or a combination of both redox-based modifications. As shown in Table II We considered the possibility that GSNO S-nitrosylates RyR1 via transnitrosylation using C-PTIO, a NO scavenger, and NOC-12 as a control. NOC-12 (0.1 mM) no longer had any effect on RyR1 in the presence of 0.1 mM C-PTIO, neither S-nitrosylating nor activating RyR1 (not shown). Similarly, 0.1 mM C-PTIO eliminated RyR1 S-nitrosylation and activation by 0.2 mM GSNO (not shown). These results suggest that S-nitrosylation of RyR1 by GSNO is dependent on release of NO, as is the release of NO from NOC-12. We caution, nevertheless, that C-PTIO may have other effects, including scavenging and generating additional reactive radicals.
Cysteine 3635 Is Critical for RyR1 Modulation by NOC-12 but Not by GSNO-The aforementioned data using SR vesicles suggest that at ambient pO 2 NOC-12 S-nitrosylates Cys-3635 and activates RyR1 by antagonizing the inhibitory effect of CaM. In contrast, GSNO works by a different mechanism. We tested this hypothesis using a strategy that was previously employed to demonstrate selective modification of Cys-3635 by NO (6). Full-length WT or single-site C3635A RyR1 mutant channels were expressed in HEK293 cells. Membranous fractions containing WT and C3635A mutant RyR1s were isolated from the HEK293 cells, and the effects of the two NO donors were assessed at pO 2 ϳ150 mm Hg in [ 3 H]ryanodine binding (Fig. 4) and in single channel measurements (Fig. 5). NOC-12 had no effect on the mutant RyR1 (Figs. 4A and 5, A and C), whereas the GSNO effect was preserved (Figs. 4B and 5, B and  C). The failure of NOC-12 to activate RyR1 C3635A was not due to a lack of CaM binding, because the C3635A mutation does not eliminate modulation of RyR1 activity by CaM (6,8).
We used an anti-nitrosocysteine polyclonal antibody to determine whether NOC-12 and GSNO S-nitrosylated the RyR1 C3635A mutant channel. We first confirmed that NO increased the immunoreactivity of the native and WT RyR1s in pO 2 ϳ10 mm Hg but not pO 2 ϳ150 mm Hg (6) (not shown). NO did not, however, increase immunoreactivity of the C3635A mutant RyR1 at either oxygen tension. A weak signal was detected by the antibody in the control samples (without NO donor) in a region of the immunoblots containing the RyR1 (Fig. 6, left panel), as determined by an anti-RyR1 antibody (Fig. 6, right  panel). NOC-12 (0.1 and 1.0 mM) produced virtually the same signal as NO (6) but in ambient pO 2 , thus increasing the level of S-nitrosylation of native and WT RyR1s but not of C3535A RyR1. Specificity of S-nitrosylation was proven by showing that prior treatment with HgCl 2 nearly eliminated the signal (not shown). In contrast, 0.2 and 1.0 mM GSNO did not noticeably increase the low levels of endogenous immunoreactivity. Taken together, the data of Figs. 4 -6 suggest that NOC-12 and GSNO affect the RyR1 by two different mechanisms, i.e. NOC-12 by S-nitrosylation of Cys-3635 and GSNO by S-nitrosylation and/or oxidation of an additional/alternative class of RyR1 thiols.

DISCUSSION
The massive (ϳ2,200 kDa) ryanodine receptors contain numerous allosteric sites subserving multiple levels of control (25). It has been firmly established that all three mammalian ryanodine receptor isoforms are redox sensitive, i.e. the channels contain regulatory thiols whose oxidation or covalent modification alters their activities (1, 2, 29 -32). These thiols (ϳ50/ subunit) have been grouped according to their differential reactivities toward NO, O 2 , and glutathione, which in turn may be linked to binding of allosteric effectors (1,2). We have recently shown that NO, at low pO 2 , selectively modifies Cys-3635 (6). PO 2 is dynamically linked to the redox state of a class of 6 -8 thiols. However, the identities of these regulatory thiols and the mechanistic basis of the pO 2 regulation of NO binding (homotropic versus heterotropic) remain to be determined. Here we have probed this question by taking advantage of the different reactivities and properties of alternative classes of NO donors.
Cysteine 3635 is part of a predicted hydrophobic motif for S-nitrosylation (33) located within the RyR1 CaM binding domain (7,8); NO regulation of RyR1 activity is thus CaM dependent (1,6). We posited that the inability of NO to S-nitrosylate Cys-3635 at ambient O 2 tension is either due to Cys-3635 being oxidized (i.e. Cys-3635 is one of the 6 -8 thiols) or to a change in channel conformation brought about by the oxidative posttranslational modification. As a first step to address this question, we determined the dependence of NOC-12 and GSNO on Cys-3635 and pO 2 . NOC-12 and GSNO had very similar effects on RyR1 channel activity (at concentrations matched for NO release), and neither compound showed O 2 dependence (pO 2 ϳ10 mm Hg versus pO 2 ϳ150 mm Hg). However, the underlying mechanism of activation was quite different in each case. GSNO activated RyR1 via poly-S-nitrosylation and/or oxidation of RyR1 thiols. Cys-3635 and CaM were not essential for activation. These data are highly reminiscent of the effects of GSNO on cardiac muscle isoform of RyR (RyR2), except that O 2 and CaM dependence were not explored at that time (20). NO and NOC-12 have little effects on RyR2. 2