The Skeletal Muscle Calcium Release Channel

Cell, Vol. 102, 499–509, August 18, 2000, Copyright 2000 by Cell Press

The Skeletal Muscle Calcium Release Channel: Coupled O2 Sensor and NO Signaling Functions Jerry P. Eu,*§ Junhui Sun,†§ Le Xu,† Jonathan S. Stamler,*‡ and Gerhard Meissner†‡ * Howard Hughes Medical Institute Department of Medicine Divisions of Pulmonary and Cardiovascular Medicine and Department of Biochemistry Duke University Medical Center Durham, North Carolina 27710 † Departments of Biochemistry and Biophysics and Cell and Molecular Physiology University of North Carolina Chapel Hill, North Carolina 27599

Summary Ion channels have been studied extensively in ambient O2 tension (pO2), whereas tissue pO2 is much lower. The skeletal muscle calcium release channel/ryanodine receptor (RyR1) is one prominent example. Here we report that pO2 dynamically controls the redox state of 6–8 out of 50 thiols in each RyR1 subunit and thereby tunes the response to NO. At physiological pO2, nanomolar NO activates the channel by S-nitrosylating a single cysteine residue. Among sarcoplasmic reticulum proteins, S-nitrosylation is specific to RyR1 and its effect on the channel is calmodulin dependent. Neither activation nor S-nitrosylation of the channel occurs at ambient pO2. The demonstration that channel cysteine residues subserve coupled O2 sensor and NO regulatory functions and that these operate through the prototypic allosteric effector calmodulin may have general implications for the regulation of redox-related systems. Introduction Redox-related modifications of protein cysteine residues have emerged as molecular mechanisms behind many cellular processes, including DNA transcription (Hausladen et al., 1996; Aslund et al., 1999), protein folding (Bader et al., 1999), chaperone activity (Jakob et al., 1999), and enzyme function (Mannick et al., 1999). A remaining question is whether such redox-related activity results from a change in the redox state of the system (i.e., from an oxidative stress) or whether “redox” is also a regulated mechanism for control-of-protein function (i.e., may serve as a redox signal) analogous to phosphorylation. S-nitrosylation has emerged as a prototype redox-related signal. It is still the only redoxrelated posttranslational modification that has been both detected in vivo and identified with reversible changes in protein function on a physiologically relevant time scale (Jia et al., 1996; Stamler et al., 1997a; Mannick et al., 1999). ‡ To whom correspondence should be addressed (e-mail: meissner@

med.unc.edu [G. M.], [email protected] [J. S. S.]). § These authors contributed equally to this work.

A growing body of work indicates that redox-related modifications of ion channels may be involved in a broad spectrum of complex physiological responses. Examples of these responses include vasomotion (Bolotina et al., 1994), muscle contraction (Meszaros et al., 1996; Stoyanovsky et al., 1997; Xu et al., 1998), baroreceptor reflexes (Li et al., 1998), carotid body/glomus cell responses (Summers et al., 1999), and synaptic plasticity (Lipton et al., 1993; Kim et al., 1999). However, the nature of the redox-related protein modification is in question in every case, and the critical issue of “redox signal” versus “oxidative stress response” has not been clarified in these systems. It has also been shown that O2 tension can modulate ion channel activities (Lopez-Barneo et al., 1988; Youngson et al., 1993) and that NO may play a role in O2 chemoreception (Prabhakar, 1999), but again, the molecular details are not known. The skeletal muscle isoform of the intracellular calcium release channel/ryanodine receptor (RyR1) is an archetype redox-sensitive channel. RyR1 releases intracellular calcium from the sarcoplasmic reticulum (SR) in response to an action potential of the surface membrane, resulting in muscle contraction (reviewed by Franzini-Armstrong and Protasi, 1997). Many endogenous effectors and associated proteins, including Ca2⫹, Mg2⫹, ATP, calmodulin (CaM), and triadin modulate RyR1 activity (reviewed by Meissner, 1994; FranziniArmstrong and Protasi, 1997). RyR1 also contains many thiol residues whose modification by reactive oxygen species (ROS) and reactive nitrogen species (RNS) alters channel function (Abramson and Salama, 1989; Kobzik et al., 1994; Xu et al., 1998). It has been reported that the cardiac isoform of RyR (RyR2) colocalizes with nitric oxide synthase (NOS) in cardiac SR (Xu et al., 1999), that RyR2 is endogenously S-nitrosylated (Xu et al., 1998), and that NOS inhibition affects both RyR function (Meszaros et al., 1996) and muscle contractility (Kobzik et al., 1994) in isolated muscle preparations. Moreover, RyR2 responds differently to nitrosative modifications (covalent linkages), which reversibly activate the channel, than to oxidative modifications (such as intramolecular disulfide), which do not, in keeping with the notion that S-nitrosylation may be a signaling event (Xu et al., 1998). Strikingly, ion channels (including RyR1) have been studied almost exclusively in ambient oxygen tension (pO2 ⵑ150 mm Hg), whereas tissue pO2 is ⵑ10–20 mm Hg and even lower in exercising muscle (Gorczynski and Duling, 1978; Honig and Gayeski, 1993). We therefore considered the possibility that standard experimental protocols in vitro have not accurately simulated the physiological conditions in vivo. In particular, NO and O2 preferentially partition in lipid bilayers, where their local concentrations and chemistry may be different from solution phase (Liu et al., 1998). Thus, models for soluble proteins may not be relevant for ion channels and other membranous proteins. In this study, the possible role of pO2 in NO/redox control of RyR1 channel activity was explored under ambient O2 tension (21% O2 or ⵑ150 mm Hg) and under the much lower pO2 (ⵑ10 mm Hg) normally found in muscle. Given the stated importance of lipid environment and the existence of endogenous modulators and

Cell 500

Figure 1. Purification of RyR1 and Thiol Quantification SR vesicles were labeled with [3H]ryanodine or with the thiol-specific agent monobromobimane (mBB) in excess prior to treatment with CHAPS and fractionation of the solubilized proteins by sucrose density gradient centrifugation. [3H]ryanodine binding and thiol/bimane fluorescence intensity are enhanced in fraction 7 (A, inset), which contained purified RyR1 (⬎95%) as determined by densitometry analysis of 3%–12% Coomassie blue–stained polyacrylamide gradient gels (B). Arrows indicate RyR1 in lane 1 (unfractionated skeletal SR) and lane 2 (fraction 7). Fractions 12–19 in (A) contained unbound [3H]ryanodine.

associated channel proteins (Meissner, 1994; FranziniArmstrong and Protasi, 1997), SR vesicles were used exclusively in place of purified RyR1. We report that exposing RyR1 within native SR to ambient oxygen leads to oxidative changes in structure and function that are likely to be entirely artificial. Moreover, we find that the NO concentrations that are required to activate the channel in room air are exceedingly high and are not physiologically attainable. In contrast, NO activates RyR1 very efficiently at physiological muscle pO2, with maximal effects seen at submicromolar NO concentrations. Oxygen sensitivity is conferred by 6–8 cysteines (comprising the O2 sensor) that maintain the RyR in a NO-responsive state. In this state, RyR1 is selectively S-nitrosylated at 1 of the 50 free cysteines/subunit. Modulation of RyR1 activity by submicromolar NO is dependent on calmodulin. Remarkably, RyR1 purified from skeletal muscle is endogenously S-nitrosylated. Our studies of RyR1 thus (1) demonstrate the fundamental importance of tissue O2 tension in channel function; (2) implicate 6–8 cysteines in sensing O2 (the first molecular description of an O2 sensor in a mammalian protein) and perhaps also in determination of channel conformation; (3) show that O2 sensitivity of the channel is transduced by S-nitrosylation of a single cysteine; (4) demonstrate that S-nitrosylation is a dynamic regulatory modification of RyR1 in the truest sense; and (5) provide a model for redox-related signaling that may broadly apply to other ion channels and membrane proteins, in which redoxrelated posttranslational modifications operate through allosteric effectors (such as calmodulin). Results Oxygen Tension and Redox State of RyR1 The redox state of RyR1 was determined using the lipophilic, thiol-specific probe monobromobimane (mBB) in a chamber equilibrated at an O2 tension of ⵑ10 mm Hg or in room air as described in Experimental Procedures.

Briefly, SR vesicles were probed with an excess of mBB, solubilized with detergent (CHAPS), and then fractionated by sucrose density gradient centrifugation to isolate RyR1; the fluorescence intensity (thus thiol content) of each fraction was determined. A representative experiment is shown in Figure 1A. During the experiment, matched SR preparations were probed with [3H]ryanodine to identify the fraction most enriched with RyR1. Fraction 7 consistently contained the channel with purity ⬎95%, as determined by densitometry of Coomassie blue–stained polyacrylamide gradient gels (Figure 1B). The number of thiols per RyR1 subunit was determined from the thiol content and the protein concentration of fraction 7 (n ⫽ 3–4 preparations). Under pO2 ⵑ10 mm Hg, RyR1 has 35.4 ⫾ 0.4 (SEM) free thiols per RyR1 subunit, whereas the thiol content of RyR1 was 29.1 ⫾ 0.5, or ⵑ6 fewer per subunit, when assays were carried out in ambient O2 tension (Table 1). Since SR vesicles in both groups were exposed to room air during the initial isolation steps from skeletal muscle, the data suggest that some SR component(s) can actively reduce RyR1 and thereby maintain it in the reduced state at pO2 of ⵑ10 mm Hg. To confirm that this is indeed the case, SR vesicles were exposed to room air for 1 hr and then divided into two aliquots: half of the aliquots were kept under room air for another 1 hr while the other half were placed under pO2 of ⵑ10 mm Hg. Thiol determinations for the room air group and for the physiological muscle O2 group were 29.8 ⫾ 0.3 and 36.1 ⫾ 0.5 (mean ⫾ SEM, n ⫽ 4), respectively. In a complementary experiment, SR aliquots were first exposed to pO2 of ⵑ10 mm Hg for 1 hr and then either switched to room air or maintained at the same low O2 tension for a second hour. Almost identical results were obtained: the number of thiols per RyR1 subunit in the room air-exposed group was 29.9 ⫾ 0.6 and in the group maintained at low O2, 35.9 ⫾ 0.5 (mean ⫾ SEM, n ⫽ 3). The results of O2 switching experiments show that in the presence of other SR constituents, O2 tension dynamically controls

Activation of Ryanodine Receptor by NO 501

Table 1. Effect of O2 Tension on Redox State of RyR1

Exogenous Agents Added to SR

Thiols per RyR1 Subunit (SH/RyR1) Muscle O2 tension (pO2 ⵑ10 mm Hg)

None DTT GSH GSSG

(Control) 0.1 mM 5.0 mM 5.0 mM 5.0 mM

Approximate Numbers of Free Thiols per RyR1 Subunit Lost by Ambient Air Exposure

35.4 37.1 52.8 49.1 27.7

⫾ ⫾ ⫾ ⫾ ⫾

0.4 1.0 2.7 2.0 0.3

(4) (3) (3) (3) (3)

Ambient O2 tension (pO2 ⵑ150 mm Hg) 29.1 31.5 44.9 40.8 25.2

⫾ ⫾ ⫾ ⫾ ⫾

0.5 0.4 1.3 1.5 1.1

(4) (4) (3) (3) (3)

6 6 8 8 2

The experimental procedures are the same as described in the legend of Figure 1. Values are the mean ⫾ SEM of three to four experiments.

the redox state of RyR1. The presence of other SR components is necessary because the purified RyR1 did not revert to a more reduced state at low O2 tension (data not shown). The isolated SR vesicles contained 0.58 ⫾ 0.09 (mean ⫾ SD, n ⫽ 3) ␮mol of glutathione/g SR protein, ⵑ90% of which was in the reduced form (ⵑ10 GSH per RyR1 subunit). To see whether the effect of pO2 on RyR1’s redox state is dependent on glutathione/thiol, the SR vesicles were treated with various amounts of GSH or dithiothreitol (DTT) and the redox state of RyR1 was reassessed. As shown in Table 1, ⵑ50 of 101 cysteine residues per RyR1 subunit (100 cysteines per 565 kDa RyR1 peptide [Takeshima et al., 1989] and 1 per FK506 binding protein [Jayaraman et al., 1992]) could be maintained in the reduced state if a physiological concentration of thiol (5 mM GSH) was added to the SR at an O2 tension of ⵑ10 mm Hg. In contrast, there were 6–8 fewer (mBB-reactive) thiols per RyR1 subunit when SR vesicles were treated with exogenous thiol in ambient O2 tension. That is, physiological concentration of GSH could not fully protect RyR1 within native SR from oxidation in room air; these O2-sensitive RyR1 thiols were oxidized by glutathione disulfide (GSSG). Specifically, 5 mM GSSG decreased the number of pO2-responsive RyR1 thiols from 6–8 to 2 when added to the SR vesicles. The number of free thiols per RyR1 subunit

under this condition was 27.7 ⫾ 0.3 at pO2 ⵑ10 mm Hg versus 25.2 ⫾ 1.1 in ambient air (Table 1). Oxidation of RyR1 by Ambient O2 Artificially Increases RyR1 Channel Activity It is well known that exposure of RyR1 to oxidizing agents leads to alterations of channel activity. Enhanced [3H]ryanodine binding is the functional equivalent of increased open probability (Po) in lipid bilayer studies and is increased by thiol reactive disulfides (Zaidi et al., 1989) and hydrogen peroxide (Favero et al., 1995; Aghdasi et al., 1997). In this light, the effects of O2 tension were investigated. Air exposure of native SR vesicles resulted in oxidation of 6 RyR1 thiols per subunit (Table 1) and enhanced [3H]ryanodine binding to RyR1 across the spectrum of physiological calcium concentrations (Figure 2A). Thus, 21% O2, which far exceeds the O2 tension in muscle, caused an artificial activation of the channel. Modulation of RyR1 by NO at Physiological and Ambient O2 Tension NO-related molecules are physiological modulators of E-C coupling in striated muscle (Kobzik et al., 1994; Meszaros et al., 1996). Reactions of NO with thiol-containing proteins are influenced by O2 and transition metals (Stamler, 1994). To determine whether NO is able

Figure 2. Effects of O2 Tensions on NO Modulation of RyR1 Channel Activity (A) Room air enhances RyR1 channel activity as indicated by increases in [3H]ryanodine binding to RyR1 over a wide range of [Ca2⫹] as compared to pO2 ⵑ10 mm Hg. Solid lines were obtained by fitting data to a two site (one activation, one inactivation) logistic function (Tripathy et al., 1995). Values are the mean ⫾ SD from at least three experiments. (B) and (C) show the dose-dependent effect of NO on [3H]ryanodine binding to RyR1 at pO2 of ⵑ10 mm Hg (B) and ⵑ150 mm Hg (C). Values are the mean ⫾ SD of three experiments. Asterisk, p ⬍ 0.05; double asterisk, p ⬍ 0.01, as compared to the controls (NO ⫽ 0).

Cell 502

Figure 3. NO Consuming Activity of the SR, Endogenous S-Nitrosylation of RyR1, and the Specificity of S-Nitrosylation at Physiological pO2 (A and B) Acceleration of NO metabolism by SR vesicles. Representative NO electrode tracings of 0.125 ␮M NO (A) and 0.75 ␮M NO (B) added to 0.15 M KCl and 20 mM K-PIPES (pH 7.4) buffer in the absence (solid line) and presence (dotted line) of SR vesicles (0.25 mg/ml protein) at pO2 of ⵑ10 mm Hg. [In (A) d[NO]/dt was ⫺0.13 A.U./min in the buffer versus ⫺0.31 in the presence of SR; in (B) d[NO]/dt was ⫺1.19 A.U./min in the buffer versus ⫺1.30 in the presence of SR.] (C and D) Selective S-nitrosylation of RyR1 in SR vesicles at pO2 of ⵑ10 mm Hg. Every other fraction from the sucrose gradients was assayed for endogenous SNO (C) or amounts of SNO following treatment of SR vesicles with 0.75 ␮M NO (D). The results are the average of two independent experiments.

to modulate RyR1 in physiological muscle O2 tension, SR vesicles were treated with a range of NO concentrations, and the amount of [3H]ryanodine binding at equilibrium was determined. Parallel [3H]ryanodine binding studies were carried out in ambient O2 tension (ⵑ150 mm Hg) for comparison. We observed that physiological muscle O2 tension sensitizes RyR1 to NO activation. At a pO2 of ⵑ10 mm Hg, 500 nM NO was sufficient to maximally activate the RyR1 (Figure 2B), whereas 100fold higher NO concentrations (50 ␮M) were required to achieve maximum activation in ambient O2 tension (Figure 2C). Notably, these elevated NO concentrations (10–100 ␮M) had the opposite effect of inhibiting RyR1 at pO2 of 10 mm Hg (Figure 2B). pO2 therefore dictates the outcome of the interaction between RyR1 and NO. Biphasic activation/inactivation of RyR by NO is reminiscent of the action of other sulfhydryl-modifying reagents, such as H2O2, dithiopyridines, or thimerosal (Favero et al., 1995; Aghdasi et al., 1997). In this regard, submicromolar NO is physiological in muscle (Balon and Nadler, 1994; Kobzik et al., 1994; Malinski et al., 1996; Poderoso et al., 1998), whereas higher concentrations may be viewed as a nitrosative stress (Eu et al., 2000). The NO-enhanced RyR1 channel activity at physiological NO concentrations was not only seen in ryanodine binding studies (Figure 2B) but also in 45Ca2⫹ uptake studies and single-channel recordings (see below), and then only at physiological pO2. The NO concentration which activates RyR1 at pO2 of ⵑ10 mm Hg is comparable to that which activates guanylate cyclase (Stone and Marletta, 1996).

S-Nitrosylation of One Cysteine by Submicromolar NO Reversibly Activates RyR1 at Tissue pO2 An NO electrode was used to measure the lifetime of NO at pO2 of 10 mm Hg in solutions lacking and containing SR vesicles. Figure 3A shows that SR vesicles accelerate the rate of NO decay. At 0.125 ␮M NO, the peak height of the NO signal was reduced by SR and the rate of decay was significantly increased (d[NO]/dt was ⫺0.14 ⫾ 0.03 [A.U./min] in the buffer group versus ⫺0.28 ⫾ 0.04 in the SR group; n ⫽ 3, p ⫽ 0.02). At 0.75 ␮M NO, the peak height of the NO signal was still reduced but rates of NO decay were only slightly faster in the presence of SR vesicles (Figure 3B), indicating saturation of the NO consuming activity (at this concentration of SR vesicles). Figure 3B further shows that substantial levels of NO were present during the duration of the Ca2⫹ uptake and single-channel measurements (see below). Taken together, these data show that SR vesicles exhibit NO consuming activity under physiological conditions (NO ⬍ 1 ␮M; pO2 of ⵑ10 mm Hg). The ramifications of NO consumption by SR vesicles and the molecular mechanism of RyR1 interaction with NO were further investigated. Native skeletal SR vesicles and vesicles treated with 0.75 ␮M NO (as in Figure 3B) at both low (i.e., physiological) and ambient O2 tension (as in the [3H]ryanodine binding studies of Figure 2) were subjected to sucrose gradient centrifugation following their solubilization with CHAPS. Fraction 7 (by our convention), which contained RyR1 with ⬎95% purity (Figure 1), was injected into a NO/chemiluminescence detector to determine the amount of RyR1 S-nitrosothiol

Activation of Ryanodine Receptor by NO 503

Table 2. Number of SNO/RyR1 Subunit and [3H]ryanodine Binding with or without 0.75 ␮M of NO Groups

Muscle O2 Tension (pO2 ⵑ10 mm Hg) SNO/RyR1 subunit (SNO/RyR1)

Control ⫹5 mM GSH NO (0.75 ␮M) ⫹5 mM GSH

0.25 0.09 1.32 0.14

⫾ ⫾ ⫾ ⫾

0.09 0.06 0.34 0.08

(4) (7) (6) (3)

Ambient O2Tension (pO2 ⵑ150 mm Hg)

3

Bound [ H]ryanodine (pmol/mg Protein) 0.67 0.18 1.35 0.27

⫾ ⫾ ⫾ ⫾

0.05 0.02 0.06 0.04

(4) (3) (4) (3)

SNO/RyR1 subunit (SNO/RyR1) 0.35 0.06 0.09 0.10

⫾ ⫾ ⫾ ⫾

0.14 0.05 0.05 0.08

(4) (4) (3) (3)

Bound [3H]ryanodine (pmol/mg Protein) 0.95 0.25 0.86 0.31

⫾ ⫾ ⫾ ⫾

0.08 0.04 0.17 0.06

(9) (3) (7) (3)

Shown are endogenous amounts of SNO-RyR1, SNO formed after exposure to 0.75 ␮M NO, and the functional correlations, as measured by [3H]ryanodine binding (with 10 ␮M free Ca2⫹). S-Nitrosylation and increase of RyR1 channel activity by NO were reversed by 5 mM GSH. Values are the mean ⫾ SEM and number of experiments is given in parenthesis.

(SNO) formed endogenously and following NO treatment. As shown in Table 2, significant amounts (ⵑ0.3 SNO/RyR1 subunit) of endogenous SNO-RyR1 were found in control samples. The levels of endogenously S-nitrosylated RyR1 are higher than those modifying the cardiac isoform (ⵑ0.04 SNO/RyR2 subunit [Xu et al., 1998]), but whether this is due to higher NOS activity, SNO stability, or protocol differences is not clear. Moreover, ⵑ1 additional cysteine per RyR1 subunit became S-nitrosylated upon exposure to 0.75 ␮M NO in low (tissue) O2 tension but not in ambient O2 tension (Table 2). This S-nitrosylation of ⵑ1 RyR1 cysteine correlated with an ⵑ2-fold increase in [3H]ryanodine binding (1.35 ⫾ 0.06 versus 0.67 ⫾ 0.05 pmol/mg protein in control, p ⬍ 0.01). When 5 mM GSH was added to the SR vesicles 30 min after the addition of 0.75 ␮M NO, both the activation of RyR1 and the increased level of S-nitrosylation were completely reversed. Remarkably (and in keeping with all above studies), submicromolar NO did not increase [3H]ryanodine binding or S-nitrosylation of RyR1 in ambient O2. We also considered the possibility that alternative redox-related modifications of RyR1 thiol(s) such as formation of disulfide bond(s) (Aghdasi et al., 1997) might explain the alteration in channel activities. However, this was not the case. Specifically, we measured 34.2 ⫾ 0.3 (n ⫽ 4) and 28.9 ⫾ 0.8 (n ⫽ 4) free thiols per RyR1 subunit after the exposure to 0.75 ␮M NO at pO2s of ⵑ10 mm Hg and ⵑ150 mm Hg, respectively, as compared with 35.4 ⫾ 0.4 and 29.1 ⫾ 0.5 without NO (Table 1). That is, 0.75 ␮M NO led to loss of 1 reactive thiol at low pO2 but did not affect RyR1 thiol content in ambient air. S-nitrosylation of one cysteine residue per RyR1 subunit is therefore the only redox-related modification of RyR1 thiol occurring under physiological NO/O2 concentrations (0.75 ␮M NO at a pO2 of ⵑ10 mm Hg). S-Nitrosylation Is Specific to RyR1 Among SR Proteins Since RyR1 constitutes only a small fraction of SR protein thiol (Figure 1), and it has been suggested that proteins such as the Ca2⫹-ATPase may be S-nitrosylated (Viner et al., 1999), we screened constituents of the SR for S-nitrosylation. As shown in Figure 3C, only the RyR1-enriched fraction (Figure 1, no. 7) contained significant endogenous SNO. Further, the RyR1-enriched fraction was preferentially S-nitrosylated by the addition of submicromolar NO (at pO2 ⵑ10 mm Hg) (Figure 3D). S-nitrosothiol stoichiometry corresponded to ⵑ1 cysteine per RyR1 subunit, consistent with our previous data (Table 2). (S-nitrosylation of other minor constituents is not precluded by these observations.) Taken together

with previous data, these results strongly suggest that at physiological NO and O2 concentrations, S-nitrosylation of one RyR1 cysteine per RyR1 subunit is a specific signaling event and not a result of random nitrosative modification. Modulation of RyR1 Activity by NO Is CaM Dependent It has been suggested that superoxide anion activates the cardiac RyR channel and that such modulation is CaM dependent (Kawakami and Okabe, 1998). Oxidation of the RyR1 alters its interaction with CaM; conversely, CaM may protect RyR1 from oxidative modification (Zhang et al., 1999). The physiological relevance of these observations is unknown. We examined whether CaM has a role in the modulation of RyR1 by NO. The amount of endogenous CaM in isolated SR vesicles was 0.14 ⫾ 0.03 ␮g/mg SR protein or 0.10–0.15 CaM per RyR1 subunit (n ⫽ 7). At pO2 of ⵑ10 mm Hg, sequestration of endogenous CaM in the SR with a specific CaM binding peptide (CaMBP) increased [3H]ryanodine binding to the same level as 0.75 ␮M NO throughout the range of Ca2⫹ tested (Figure 4A). NO- and CaMBP-mediated increases in channel activity exhibited the same pattern of calcium dependence. Moreover, [3H]ryanodine binding induced by CaMBP was not increased further by subsequent addition of 0.75 ␮M (at pO2 ⵑ10 mm Hg). Therefore, modulation of RyR1 by submicromolar NO appears to be CaM dependent. The interactions of RyR1, NO and CaM were further explored in 45Ca2⫹ uptake and single-channel studies. The channel activities of RyR1 were inferred from 45Ca2⫹ uptake studies of SR vesicles (Figure 4B, pO2 ⵑ10 mm Hg). The baseline maximum 45Ca2⫹ uptakes of SR vesicles were established in each group by pretreatment of SR with 0.3 mM ryanodine, which completely abolishes RyR1 channel activity (dotted lines in Figure 4B, maximum Ca2⫹ uptakes of SR vesicles among different groups were seen to be nearly identical). Consistent with the [3H]ryanodine binding results (Figure 4A), 1 ␮M CaMBP or 0.75 ␮M NO decreased 45Ca2⫹ uptake of SR to comparable degrees, indicating an equivalent activation of RyR1. NO (0.75 ␮M) also reversed the inhibitory effect of 2 ␮M CaM—which when added with 1 ␮M Ca2⫹, increased 45Ca2⫹ uptake of SR above that of the control—and suppressed 45Ca2⫹ uptake close to the control level (Figure 4B). The effect of 0.75 ␮M NO on 45 Ca2⫹ uptake was reversed by GSH at low O2 tension (data not shown). In contrast, 0.75 ␮M NO had no effect on 45Ca2⫹ uptake of the SR at ambient O2 tension (data not shown). The 45Ca2⫹ uptake results in Figure 4B could also be

Cell 504

Figure 4. Effects of CaM and NO on RyR1 Channel Activity at pO2 of ⵑ10 mm Hg Specific [3H]ryanodine binding (A) and 45Ca2⫹ uptake of SR vesicles (B) were determined at the indicated free [Ca2⫹] in the absence and presence of NO (0.75 ␮M), CaM (2 ␮M), and CaMBP (1 ␮M). (A) At pO2 of ⵑ10 mm Hg, either 0.75 ␮M NO or CaMBP (which sequesters CaM) enhanced RyR1 channel activity. Additive effects of NO and CaMBP were not seen. When RyR1 channel activity was suppressed by 2 ␮M exogenous CaM, 0.75 ␮M NO reversed the inhibition at low but not elevated Ca2⫹ concentrations. Data are the mean ⫾ SD of four experiments. (B) No differences in 45Ca2⫹ uptake were observed in SR vesicles that were pretreated with 0.3 mM ryanodine to completely close the SR Ca2⫹ release channel (dotted lines). The effects of NO, CaM, and CaMBP on 45Ca2⫹ uptake by SR vesicles at pO2 of ⵑ10 mm are consistent with [3H]ryanodine binding results (see Results and Discussion). Data are the mean ⫾ SD of three experiments.

explained by NO modulation of the SR Ca2⫹-ATPase (Viner et al., 1999; Xu et al., 1999). To test this possibility, Ca2⫹-dependent ATPase activities in SR vesicles were determined before and after the addition of 0.75 ␮M NO at both physiological and ambient pO2 using a malachite green ATPase method adapted from Kirchgesser and Dahlmann (1990) in the presence of 1 ␮M ionomycin. NO (0.75 ␮M) did not change the activity of Ca2⫹-ATPase irrespective of the pO2 (n ⫽ 8–9, data not shown). The ability of submicromolar NO to antagonize CaM’s inhibitory effect on RyR1 under physiological muscle O2 tension was also confirmed in single-channel recordings. SR vesicles were incorporated into planar lipid bilayers and single RyR1 channels were recorded with Cs⫹ as the current carrier. As shown in Figures 5A and 5B, the channel open probability (Po) of RyR1 in the presence of 100 nM CaM and 4 ␮M free Ca2⫹ was nearly doubled after the addition of 0.75 ␮M NO at pO2 of ⵑ10 mm Hg. In contrast, 0.75 ␮M NO did not change Po in ambient O2 tension (Figure 5B). A second control experiment in Figure 5B shows that at pO2 of ⵑ10 mm Hg, 0.75 ␮M NO was without effect on channels obtained from SR vesicles preincubated with CaMBP. Thus, [3H]ryanodine binding, 45Ca2⫹ uptake, and single-channel studies all show that modulation of RyR1 by submicromolar NO at pO2 of ⵑ10 mm Hg is dependent on endogenous CaM. Scatchard analysis of [35S]CaM binding to RyR1 was performed to determine whether S-nitrosylation of RyR1 activates channels by displacing CaM from RyR1. At pO2 ⵑ10 mm Hg, 0.75 ␮M NO only modestly decreased Bmax (68.4 ⫾ 2.2 pmol/mg protein with NO versus 74.0 ⫾

2.6 pmol/mg protein for control) and KD (12.8 ⫾ 0.1 nM with NO versus 15.6 ⫾ 1.1 nM for control, p ⬍ 0.05) of [35S]CaM binding. Thus, S-nitrosylation of one cysteine per RyR1 subunit did not activate the channel by displacing CaM. It has been shown that adding a saturating amount of CaM (2 ␮M) to SR can inhibit [3H]ryanodine binding to RyR1 (Tripathy et al., 1995). The effect of submicromolar NO on [3H]ryanodine binding in the presence of 2 ␮M CaM was therefore examined under physiological muscle O2 tension. NO (0.75 ␮M) (at O2 tension ⵑ10 mm Hg) antagonized the inhibitory effect of exogenously added CaM by partially restoring [3H]ryanodine binding at [Ca2⫹] ⬍ 10 ␮M but NO had a slightly inhibitory effect at [Ca2⫹] ⬎ 100 ␮M (Figure 4A). In particular, in the presence of 2 ␮M CaM, 0.75 ␮M NO shifts the peak of [3H]ryanodine binding to a 10-fold lower Ca2⫹ concentration (from pCa 5 to 6) at physiological O2 tension. This effect of 0.75 ␮M of NO was not observed when matching experiments were carried out in room air (data not shown).

Discussion The nature of the “redox/O2 sensor” that couples membrane excitability to intracellular redox chemistry is a mystery (Goldberg et al., 1988; Semenza, 1999). Only recently has one prevalent theory, involving the NADPH oxidase, been questioned (Archer et al., 1999) and another, based on structural homology between an NAD(P)H-oxidoreductase and a subunit of a K⫹ channel, been proposed

Activation of Ryanodine Receptor by NO 505

Figure 5. Effect of NO on RyR1 Single-Channel Activities (A) Channel currents of two RyR1s, shown as downward deflections from closed (c) levels, were recorded in a lipid bilayer chamber preequilibrated to pO2 ⵑ10 mm Hg with voltage held at ⫺35 mV. Top tracing: control with 100 nM of CaM and 4 ␮M free Ca2⫹, Po ⫽ 0.06; second trace: 2 min after the addition of 0.75 ␮M NO to cytosolic side, Po ⫽ 0.13. (B) As compared to controls (open bars), 0.75 ␮M NO (shaded bars) significantly increased Po of RyR1 (asterisk, p ⬍ 0.05) at pO2 ⵑ10 mm Hg but not pO2 ⵑ150 mm Hg. In the presence of 1 ␮M CaMBP, 0.75 ␮M NO did not significantly alter Po of RyR1 at pO2 of ⵑ10 mm Hg. Data are mean ⫾ SEM of the number of experiments shown in parentheses.

(Gulbis et al., 1999). It has been alternatively surmised that a heme protein may be used to sense O2 (Huang et al., 1999; Semenza, 1999), but none has been discovered in multicellular organisms. We uncovered a striking plasticity of RyR1’s redox state that was dictated by O2 tension. The channel does not contain heme or transition metals or a flavin domain but is unusually rich in free cysteines. Exposure of native SR vesicles to ambient air resulted in oxidation of 6–8 free thiols per RyR1 subunit (or 24–32 per channel) that reverted to the reduced state upon lowering of the pO2 into the physiological range (ⵑ10 mm Hg). Thus, pO2 appears to be the main determinant of the redox status of 6–8 RyR1 thiols, which are thereby identified with an O2 sensing function. Evidently, this thiol-based “O2 sensor” is coupled to a redox system within the SR (that provides the reducing equivalents for switching between “oxidized” and “reduced” states of the channel); it is also intimately linked to NO function; that is, only at physiological pO2 is the channel responsive to physiological concentrations (submicromolar) of NO. Channel activation is the result of S-nitrosylation of one cysteine. Thus, the redox state of the channel (i.e., its responsiveness) is set by O2 tension, whereas redox regulation of the channel is mediated by NO. Our demonstration of dynamic changes in the thiol redox status of an ion channel in response to pO2 is novel, but unlikely to be unique to RyR1. In addition to the special cases of O2-responsive chemoreceptors in the aortic arch (Lopez-Barneo et al., 1988) and airway (Youngson et al., 1993), it has been reported that a voltage- and Ca2⫹-dependent Cl⫺ channel in SR is inhibited by pO2 of ⬍1 mm Hg and reactivated by air oxidation. These effects of O2 concentration were mimicked by reducing and oxidizing agents (Kourie, 1997) that have

analogous effects on various ion channels, among which are included the NMDA receptor/Ca2⫹ channel (Lipton et al., 1993; Gbadegesin et al., 1999), baroreceptor Na⫹ channel (Drummond et al., 1998), and smooth muscle K⫹ channel (Ruppersberg et al., 1991). But whereas activity at low pO2 has been viewed as a compensatory or potentially adverse response to “hypoxia,” it may be more appropriately viewed as physiological. We note that all studies previously performed with the RyR1 have been done at ambient pO2 and thus with an oxidized form of the channel. Inasmuch as skeletal muscle is never exposed to such high pO2, this oxidized state of the channel is nonphysiological. Air exposure of RyR1 manifested as an increase in channel activity (Figure 2A) and as a loss of responsiveness to physiological effectors (i.e., NO)—it has been common practice to further subject air-exposed RyR1 to high concentrations of oxidants in order to simulate oxidative damage. Our data suggest that the standard RyR1 preparation may, in fact, be representative of channel behavior in oxidative stresses such as muscle fatigue or reperfusion injury and more generally highlight the importance of studying redox-based cellular functions under redox conditions (e.g., pO2) that closely mimic the physiological situation. RyR1 is most efficiently activated at low NO and O2 concentrations. Since the reactions of NO with cysteine redox centers of proteins occur through intermediates formed by NO and O2 and/or are facilitated by redox reactions that require the presence of oxidants (Stamler, 1994; Stamler et al., 1997b), one might have expected to see a more vigorous effect of NO at high pO2 and a greater dependence on the NO concentration. Indeed, the notion that NO reactions with cysteine do not occur

Cell 506

at submicromolar NO concentrations shapes the current perspective (Wink et al., 1994; Zhao et al., 1999). It is important to appreciate, however, that this perspective, which derives from reactions of NO in model aqueous systems, does not imply that NO reactions in membranous compartments or complex systems are similarly dependent on high NO/O2 and thus cannot be assumed to apply to the physiological situation. In fact, NO reactions with thiols are probably greatly accelerated in membranes where NO and O2 concentrate (Liu et al., 1998), and there is no reason to believe that products of NO/O2 reactions in solution should interact with the same class of RyR thiols as those modified by NO following its entry into the membrane (or other hydrophobic pockets). In addition, the kinetics of NO/O2 metabolites such as NO2 and N2O3 or the relative distribution among them may be significantly altered at different pO2, perhaps even by virtue of the existence of an O2-sensitive enzymatic mechanism in the SR. Whatever the mechanism, NO metabolism is accelerated by the presence of SR (Figure 3) and the S-nitrosylation reactions in SR vesicles (Table 2) are facilitated at the low O2 and NO concentrations present in vivo. NO responsiveness of the channel is thus tuned by the pO2. The enhanced sensitivity of RyR1 to NO modulation at pO2 of ⵑ10 mm Hg is identified with 6–8 thiols. NO may react with one of them. Alternatively, the reduced thiols could maintain the RyR1 in an NO-sensitive conformation (i.e., in a “reduced” state) by exposing an NO-reactive cysteine—a model we favor because stable SNOs do not form at thiols that oxidize to disulfide (Stamler and Hausladen, 1998). S-nitrosylation of RyR1 would then be allosterically linked to pO2/redox, as in the precedent setting cases of hemoglobin, which binds NO groups preferentially in the R (or oxy) structure (Stamler et al., 1997a) and methionine adenosyl transferase (Perez-Mato et al., 1999). Interestingly, these latter proteins contain cysteines that conform to a motif (Stamler et al., 1997b). The sensitivity of RyR1 to NO is relatively specific, as submicromolar NO has no noticeable effect on another reputed redox-sensitive SR transporter, the Ca2⫹-ATPase (Viner et al., 1999), irrespective of pO2. The physiological relevance of this type of redoxrelated modification is proven by our finding that RyR1, like the cardiac isoform of RyR (Xu et al., 1998), is endogenously S-nitrosylated; a major fraction of purified RyR1, about 30%, was S-nitrosylated (“control groups” in Table 2). The significance of this amount is underscored by the instability of SNOs, which are easily lost during the lengthy purification of this protein, and the fact that nNOS is often expressed only in type II (fast twitch) muscle fibers. Thus, S-nitrosylation may be restricted to fast-twitch fibers in order to facilitate a more rapid release of SR Ca2⫹. Notably, in the presence of saturating levels of CaM (2 ␮M), NO shifted the peak of [3H]ryanodine binding (as a function of Ca2⫹) leftward (Figure 4A). The data suggest that RyR1 may be activated by S-nitrosylation at less than 10 ␮M Ca2⫹ and inhibited at ⬎100 ␮M Ca2⫹. In other words, NO may facilitate both contraction and relaxation by modulating RyR1 channel activity. Our studies suggest that S-nitrosylation of RyR1 activates the channel by antagonizing the inhibition by CaM. At pO2 of ⵑ10 mm Hg, 0.75 ␮M NO had virtually identical enhancing effects on channel activity in both [3H]ryanodine binding (Figure 4A) and 45Ca2⫹ uptake (Figure 4B)

studies. NO did not, however, activate RyR1 that had been stripped of calmodulin by CaMBP (Figures 4A, 4B, and 5B). Scatchard analysis showed that 0.75 ␮M NO at pO2 of ⵑ10 mm Hg did not significantly change Bmax value and only very modestly decreased KD of [35S]CaM binding to RyR1. Thus, we conclude that although S-nitrosylation affects the interaction of CaM with RyR1, it does not activate the channel by displacing it. We observed that RyR1 activity can be increased by high doses of NO in ambient pO2 (i.e., at high pO2). It is possible that the channel has adapted this means to respond to redox-related signals in conditions of oxidative and nitrosative stress. However, this “fall back” mechanism comes with a price: only at physiological NO/O2 is the biphasic dose response to NO (Figure 2) preserved. That is, increased channel activity mediated by the higher doses of NO or O2 comes at the expense of loss of control. In summary, our studies with RyR1 demonstrate that the redox state of an ion channel can be dynamically controlled by pO2, which thereby tunes the response to other modulators, particularly NO. We further show that the NO signal is itself transduced by the allosteric effector calmodulin. Thus, redox control may be thought of in terms of NO/O2-based posttranslational modifications of proteins. Since NO/O2 concentrate in hydrophobic environments, the prototype modification, S-nitrosylation, may serve as a ubiquitous membrane signal. Our work provides a model for future studies of ion channels and other membrane proteins in which critical cysteines subserve coupled O2 sensor and NO regulatory functions. Experimental Procedures Isolation of SR Vesicles and Purification of RyR1 SR fractions enriched with RyR1 were prepared from skeletal muscle in the presence of protease inhibitors as previously described (Lai et al., 1988). The maximum numbers of high-affinity [3H]ryanodine binding sites obtained ranged from 12–20 pmol/mg protein. The purification of RyR1 from SR vesicles using a sucrose gradient centrifugation method (Lai et al., 1988) was modified for each assay (see below). Quantification of RyR1 Thiols The free thiol content of the rabbit skeletal RyR1 was determined by the monobromobimane (mBB, Calbiochem) fluorescence method of Kosower and Kosower (Xu et al., 1998). For low O2 tension, the experiments were set up in a chamber preequilibrated to pO2 of ⵑ10 mm Hg. Skeletal SR vesicle preparations, treated with or without redox agents for 1 hr at 24⬚C, were centrifuged (100,000 ⫻ g) at 4⬚C for 1 hr. The pellets were brought back to the hypoxic chamber and washed with buffer preequilibrated to pO2 of ⵑ10 mm Hg and then probed with an excess (500 ␮M) of lipophilic, thiol-specific agent mBB for 1 hr in the dark at 24⬚C. Matched experiments were carried out in room air (pO2 of ⵑ150 mm Hg). Following the mBB treatment, SR vesicles were solubilized with 1.5% CHAPS, and bimane-labeled RyR1 was isolated by sucrose density gradient centrifugation. The fluorescence intensities of bimane (i.e., the thiol content) in the sucrose gradient fractions most enriched with RyR1 (⬎95%) were determined and normalized for protein concentrations as previously described (Xu et al., 1998). NO Solutions and Lifetime in the Absence and Presence of SR NO gas (purity ⬎99%, National Wielders) was scrubbed to remove O2 and nitrite by passing it through an argon-purged column filled with KOH pellets followed by bubbling the gas through a solution of NaOH. The concentration of NO dissolved in argon-purged buffer was determined by using a hemoglobin titration assay. To determine the rates of NO metabolism with or without SR, the NO was added to a rapidly stirred RyR1 buffer preequilibrated to pO2 ⵑ10 mm Hg

Activation of Ryanodine Receptor by NO 507

in the hypoxic chamber with or without 0.2 mg/ml to 2.0 mg/ml of SR vesicle proteins. Decomposition of NO was recorded with a NO electrode (WPI Instruments). S-Nitrosylation of RyR1 Both native and NO-treated SR aliquots (0.75 ␮M NO for 1 hour) were subjected to sucrose density gradient centrifugation and aliquots of RyR1-enriched (⬎95% purity) and -poor fractions were assayed for SNO by a photolysis/chemiluminescence-based NO detection assay (Xu et al., 1998). [3H]Ryanodine Binding SR vesicles, prepared as described in “thiol assays” (at controlled pO2) were exposed to various concentrations of NO and incubated at 24⬚C for 5 hr with 5 nM [3H]ryanodine and the indicated concentrations of CaM, CaMBP, and free Ca2⫹. The amounts of [3H]ryanodine binding were determined as described (Tripathy et al. 1995). Ca2ⴙ Uptake Ca2⫹ uptake by SR vesicles was determined by a filtration method (at both high and low oxygen tension). Samples were incubated in 0.15 M KCl, 20 mM K-HEPES, pH 7.4, 1 ␮M free Ca2⫹, 0.3 mM Pefabloc, and 30 ␮M leupeptin for 1 hr at 24⬚C in the absence and in the presence of 0.3 mM ryanodine to completely close the SR Ca2⫹ release channel. To investigate the effects of CaM, samples were preincubated with 2 ␮M CaM or 1 ␮M CaMBP. The final SR protein concentration was 0.4 mg/ml. 45Ca2⫹ uptake was started at 24⬚C by addition of the incubated samples to 10 volumes of 0.15 M KCl, 20 mM K-HEPES, pH 7.4, 8 mM MgCl2, 5 mM ATP, 5 mM NaN3, and 1 ␮M free 45Ca2⫹ (NEN Life Science Products) in the presence or absence of 0.75 ␮M NO. Aliquots of the assay media were placed on 0.45 ␮m Millipore filters under vacuum at various times and rinsed with 3 ml of washing buffer containing 0.15 M KCl, 5 mM K-PIPES, pH 6.8, 0.1 mM EGTA, and the two SR Ca2⫹ release channel inhibitors Mg2⫹ (10 mM) and ruthenium red (10 ␮M). The radioactivity remaining with the vesicles on the filters was counted by liquid scintillation.

at 100⬚C for 3 min, placed on ice and centrifuged at 7,000 ⫻ g for 20 min. The supernatant was used to determine amounts of CaM associated with SR membranes. Ca2⫹-ATPase activity in SR vesicles was assayed in 10 mm Hg or ambient O2 tension by malachite green ATPase method adapted from Kirchgesser and Dahlmann (1990) in the presence of a Ca2⫹ ionophore (1 ␮M ionomycin) and the presence and absence of 0.75 ␮M NO. Mg2⫹-ATPase remaining in the SR preparation was subtracted from the total ATPase by adding 1 mM EGTA to assay media. [35S]Calmodulin Binding SR vesicles were incubated with 1–100 nM [35S]CaM at 24⬚C in 0.15 M KCl, 20 mM K-PIPES, pH 7.0, 0.3 mM Pefabloc, 30 ␮M leupeptin, and 1 ␮M free Ca2⫹. After 2 hr of incubation at 24⬚C in 10 mm Hg or ambient O2 tension, aliquots of vesicle suspensions were assayed for total radioactivity and centrifuged for 30 min at 30 psi in a Beckman Airfuge. Bound CaM was derived from free and total [35S]CaM measured by scintillation counting.

45 45

Single-Channel Recordings Single-channel measurements were performed on skeletal SR vesicles (fused with Mueller-Rudin-type bilayers) as described (Tripathy et al., 1995). The side of the bilayer to which the SR vesicles were added was defined as the cis (cytoplasmic) side. The trans (SR lumenal) side of the bilayer was defined as ground. Single-channel recordings were done in ambient O2 or in a hypoxic chamber at ⵑ10 mm Hg of O2 tension. All buffers were bubbled with nitrogen gas for 30 min before being placed in the hypoxic chamber. SR vesicles were preincubated with CaM or CaMBP for 30 min before being added to the cis chamber of the bilayer set-up. Single channels were recorded in a symmetric CsCH3SO3 buffer solution (0.25 M CsCH3SO3, 10 mM Cs-HEPES, pH 7.4) containing the additions indicated in the text. 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). Glutathione, CaM Content, and Ca2ⴙ-ATPase Activity in SR Vesicles Total glutathione (GSSG ⫹ GSH) and GSSG fractions were determined using a glutathione reductase-recycling assay modified from Griffith (Eu et al., 2000). The SR vesicles were treated with 0.5% Triton-X 100 and centrifuged at 14,000 ⫻ g for 10 min. To determine the oxidized GSSG fraction, reduced GSH was first derivatized with 2-vinylpyridine for 40 min prior to the recycling assay. SR-associated CaM was assayed by its ability to stimulate phosphodiesterase (PDE) activity, as measured by hydrolysis of cAMP to 5⬘-AMP, which was further decomposed by 5⬘-nucleotidase to orthophosphate (Pi) and adenosine (Garg et al., 1988). The Pi generated was measured using the method of Lanzetta et al. (1979). Briefly, aliquots of SR vesicles were denatured with 3 volumes 80 mM Tris-HCl, 80 mM imidazole, pH 7.8, 6 mM MgCl2, 0.2 mM CaCl2

Ca2ⴙ Assays Free Ca2⫹ concentrations were obtained by including in the solutions Ca2⫹ and EGTA as published by Schoenmakers et al. (1992). Free Ca2⫹ concentrations of ⬎1 ␮M were verified with the use of a Ca2⫹ selective electrode (World Precision Instruments, Sarasota, FL). Acknowledgments Supported by grants from the National Institutes of Health to J. P. E. (HL04053), J. J. S. (HL52529, ES-09206, and HL59130), and G. M. (AR18687 and HL27430). Received January 10, 2000; revised June 20, 2000. References Abramson, J.J., and Salama, G. (1989). Critical sulfhydryls regulate calcium release from sarcoplasmic reticulum. J. Bioenerg. Biomembr. 21, 283–294. Aghdasi, B., Reid, M.B., and Hamilton, S.L. (1997). Nitric oxide protects the skeletal muscle Ca2⫹ release channel from oxidation induced activation. J. Biol. Chem. 272, 25462–25467. Archer, S.L., Reeve, H.L., Michelakis, E., Puttagunta, L., Waite, R., Nelson, D.P., Dinauer, M.C., and Weir, E.K. (1999). O2 sensing is preserved in mice lacking the gp91 phox subunit of NADPH oxidase. Proc. Natl. Acad. Sci. USA 96, 7944–7949. Aslund, F., Zheng, M., Beckwith, J., and Storz, G. (1999). Regulation of the OxyR transcription factor by hydrogen peroxide and the cellular thiol-disulfide status. Proc. Natl. Acad. Sci. USA 96, 6161–6165. Bader, M., Muse, W., Ballou, D.P., Gassner, C., and Bardwell, J.C. (1999). Oxidative protein folding is driven by the electron transport system. Cell 98, 217–227. Balon, T.W., and Nadler, J.L. (1994). Nitric oxide release is present from incubated skeletal muscle preparations. J. Appl. Physiol. 77, 2519–2521. Bolotina, V.M., Najibi, S., Palacino, J.J., Pagano, P.J., and Cohen, R.A. (1994). Nitric oxide directly activates calcium-dependent potassium channels in vascular smooth muscle. Nature 368, 850–853. Drummond, H.A., Price, M.P., Welsh, M.J., and Abboud, F.M. (1998). A molecular component of the arterial baroreceptor mechanotransducer. Neuron 21, 1435–1441. Eu, J.P., Liu, L., Zeng, M., and Stamler, J.S. (2000). An apoptotic model for nitrosative stress. Biochemistry 39, 1040–1047. Favero, T.G., Zable, A.C., and Abramson, J.J. (1995). Hydrogen peroxide stimulates the Ca2⫹ release channel from skeletal muscle sarcoplasmic reticulum. J. Biol. Chem. 270, 25557–25563. Franzini-Armstrong, C., and Protasi, F. (1997). Ryanodine receptors of striated muscles: a complex channel capable of multiple interactions. Physiol. Rev. 77, 699–729. Garg, U.C., Rai, N., Singh, Y., Dhaunsi, G.S., Sidhu, G.S., Ganguly, N.K., and Bhatnagar, R. (1988). A spectrophotometric method for calmodulin assay. Biotechniques 6, 294–296.

Cell 508

Gbadegesin, M., Vicini, S., Hewett, S.J., Wink, D.A., Espey, M., Pluta, R.M., and Colton, C.A. (1999). Hypoxia modulates nitric oxideinduced regulation of NMDA receptor currents and neuronal cell death. Am. J. Physiol. 277, C673–C683. Goldberg, M.A., Dunning, S.P., and Bunn, H.F. (1988). Regulation of the erythropoietin gene: evidence that the oxygen sensor is a heme protein. Science 242, 1412–1415. Gorczynski, R.J., and Duling, B.R. (1978). Role of oxygen in arteriolar functional vasodilation in hamster striated muscle. Am. J. Physiol. 235, H505–H515. Gulbis, J.M., Mann, S., and MacKinnon, R. (1999). Structure of a voltage-dependent K⫹ channel beta subunit. Cell 97, 943–952. Hausladen, A., Privalle, C.T., Keng, T., DeAngelo, J., and Stamler, J.S. (1996). Nitrosative stress: activation of the transcription factor OxyR. Cell 86, 719–729. Honig, C.R., and Gayeski, T.E. (1993). Resistance to O2 diffusion in anemic red muscle: roles of flux density and cell pO2. Am. J. Physiol. 265, H868–H875. Huang, I.E., Willmore, W.G., Gu, J., Goldberg, M.A., and Bunn, H.F. (1999). Inhibition of hypoxia-inducible factor 1 activation by carbon monoxide and nitric oxide. J. Biol. Chem. 274, 9038–9044. Jakob, U., Muse, W., Eser, M., and Bardwell, J.C. (1999). Chaperone activity with a redox switch. Cell 96, 341–352. Jayaraman, T., Brillantes, A.M., Timerman, A.P., Fleischer, S., Erdjument, B.H., Tempst, P., and Marks, A.R. (1992). FK506 binding protein associated with the calcium release channel (ryanodine receptor). J. Biol. Chem. 267, 9474–9477. Jia, L., Bonaventura, C., Bonaventura, J., and Stamler, J.S. (1996). S-nitrosohaemoglobin: a dynamic activity of blood involved in vascular control. Nature 380, 221–226. Kawakami, M., and Okabe, E. (1998). Superoxide anion radical-triggered Ca2⫹ release from cardiac sarcplasmic reticulum through ryanodine receptor Ca2⫹ channel. Mol. Pharmacol. 53, 497–503. Kim, W., Choi, Y., Rayudu, P., Das, P., Assad, W., Arnelle, D.R., Stamler, J.S., and Lipton, S.A. (1999). Attenuation of NMDA receptor activity and neurotoxicity by nitroxyl anion (NO). Neuron 24, 461–469. Kirchgesser, M., and Dahlmann, N. (1990). A colorimetric assay for the determination of acid nucleoside triphosphatase activity. J Clin. Chem. Clin. Biochem. 28, 407–411. Kobzik, L., Reid, M.B., Bredt, D.S., and Stamler, J.S. (1994). Nitric oxide in skeletal muscle. Nature 372, 546–548. Kourie, J.I. (1997). A redox O2 sensor modulates the SR Ca2⫹ countercurrent through voltage- and Ca2⫹-dependent Cl channels. Am. J. Physiol. 272, C324–C332. Lai, F.A., Erickson, H.P., Rousseau, E., Liu, Q.Y., and Meissner, G. (1988). Purification and reconstitution of the calcium release channel from skeletal muscle. Nature 331, 315–319. Lanzetta, P.A., Alvarez, L.J., Reinach, P.S., and Candia, O.A. (1979). An improved assay for nanomole amounts of inorganic phosphate. Anal. Biochem. 100, 95–97. Li, Z., Chapleau, M.W., Bates, J.N., Bielefeldt, K., Lee, H.C., and Abboud, F.M. (1998). Nitric oxide as an autocrine regulator of sodium currents in baroreceptor neurons. Neuron 20, 1039–1049. Lipton, S.A., Choi, Y.B., Pan, Z.H., Lei, S.Z., Chen, H.S., Sucher, N.J., Loscalzo, J., Singel, D.J., and Stamler, J.S. (1993). A redoxbased mechanism for the neuroprotective and neurodestructive effects of nitric oxide and related nitroso-compounds. Nature 364, 626–632. Liu, X., Miller, M.J.S., Joshi, M.S., Thomas, D.D., and Lancaster, J.R., Jr. (1998). Accelerated reaction of nitric oxide with O2 within the hydrophobic interior of biological membranes. Proc. Natl. Acad. Sci. USA 95, 2175–2179. Lopez-Barneo, J., Lopez-Lopez, J.R., Urena, J., and Gonzalez, C. (1988). Chemotransduction in the carotid body: K⫹ current modulated by pO2 in type I chemoreceptor cells. Science 241, 580–582. Malinski, T., Mesaros, S., and Tomboulian, P. (1996). Nitric oxide measurement using electrochemical methods. Methods Enzymol. 268, 58–69.

Mannick, J.B., Hausladen, A., Liu, L., Hess, D.T., Zeng, M., Miao, Q.X., Kane, L.S., Gow, A.J., and Stamler, J.S. (1999). Fas-induced caspase denitrosylation. Science 284, 651–654. Meissner, G. (1994). Ryanodine receptor/Ca2⫹ release channels and their regulation by endogenous effectors. Annu. Rev. Physiol. 56, 485–508. Meszaros, L.G., Minarovic, I., and Zahradnikova, A. (1996). Inhibition of the skeletal muscle ryanodine receptor calcium release channel by nitric oxide. FEBS Lett. 380, 49–52. Perez-Mato, I., Castro, C., Ruiz, F.A., Corrales, F.J., and Mato, J.M. (1999). Methionine adenosyltransferase S-nitrosylation is regulated by the basic and acidic amino acids surrounding the target thiol. J. Biol. Chem. 274, 17075–17079. Poderoso, J.J., Peralta, J.G., Lisdero, C.L., Carreras, M.C., Radisic, M., Schopfer, F., Cadenas, E., and Boveris, A. (1998). Nitric oxide regulates oxygen uptake and hydrogen peroxide release by the isolated beating rat heart. Am. J. Physiol. 274, C112–C119. Prabhakar, N.R. (1999). NO and CO as second messengers in oxygen sensing in the carotid body. Respir. Physiol. 115, 161–168. Ruppersberg, J.P., Stocker, M., Pongs, O., Heinemann, S.H., Frank, R., and Koenen, M. (1991). Regulation of fast inactivation of cloned mammalian IK(A) channels by cysteine oxidation. Nature 352, 711–714. Schoenmakers, J.M., Visser, G.J., Flik, G., and Theuvene, A.P.R. (1992). CHELATOR: an improved method for computing metal ion concentrations in physiological solutions. Biotechniques 12, 870–879. Semenza, G.L. (1999). Perspectives on oxygen sensing. Cell 98, 281–284. Stamler, J.S. (1994). Redox signaling: nitrosylation and related target interactions of nitric oxide. Cell 78, 931–936. Stamler, J.S., and Hausladen, A. (1998). Oxidative modifications in nitrosative stress. Nature Struct. Biol. 5, 247–249. Stamler, J.S., Jia, L., Eu, J.P., McMahon, T.J., Demchenko, I.T., Bonaventura, J., Gernert, K., and Piantadosi, C.A. (1997a). Blood flow regulation by S-nitrosohemoglobin in the physiological oxygen gradient. Science 276, 2034–2037. Stamler, J.S., Toone, E.J., Lipton, S.A., and Sucher, N.J. (1997b). (S)NO signals: translocation, regulation, and a consensus motif. Neuron 18, 691–696. Stone, J.R., and Marletta, M.A. (1996). Spectral and kinetic studies on the activation of soluble guanylate cyclase by nitric oxide. Biochemistry 35, 1093–1099. Stoyanovsky, D., Murphy, T., Anno, P.R., Kim, Y.-M., and Salama, G. (1997). Nitric oxide activates skeletal and cardiac ryanodine receptors. Cell Calcium 21, 19–29. Summers, B.A., Overholt, J.L., and Prabhakar, N.R. (1999). Nitric oxide inhibits L-type Ca2⫹ current in glomus cells of the rabbit carotid body via a cGMP-independent mechanism. J. Neurophysiol. 81, 1449–1457. Takeshima, H., Nishimura, S., Matsumoto, T., Ishida, H., Kangawa, K., Minamino, N., Matsuo, H., Ueda, M., Hanaoka, M., and Hirose, T. (1989). Primary structure and expression from complementary DNA of skeletal muscle ryanodine receptor. Nature 339, 439–445. Tripathy, A., Xu, L., Mann, G., and Meissner, G. (1995). Calmodulin activation and inhibition of skeletal muscle Ca2⫹ release channel (ryanodine receptor). Biophys. J. 69, 106–119. Viner, R.I., Ferrington, D.A., Williams, T.D., Bigelow, D.J., and Schoneich, C. (1999). Protein modification during biological aging: selective tyrosine nitration of the SERCA2a isoform of the sarcoplasmic reticulum Ca2⫹-ATPase in skeletal muscle. Biochem. J. 340, 657–669. Wink, D.A., Nims, R.W., Darbyshire, J.F., Christodoulou, D., Hanbauer, I., Cox, G.W., Laval, F., Laval, J., Cook, J.A., Krishna, M.C., et al. (1994). Reaction kinetics for nitrosation of cysteine and glutathione in aerobic nitric oxide solutions at neutral pH. Insights into the fate and physiological effects of intermediates generated in the NO/O2 reaction. Chem. Res. Toxicol. 7, 519–525. Xu, L., Eu, J.P., Meissner, G., and Stamler, J.S. (1998). Activation of

Activation of Ryanodine Receptor by NO 509

the cardiac calcium release channel (ryanodine receptor) by polyS-nitrosylation. Science 279, 234–237. Xu, K.Y., Huso, D.L., Dawson, T.M., Bredt, D.S., and Becker, L.C. (1999). Nitric oxide synthase in cardiac sarcoplasmic reticulum. Proc. Natl. Acad. Sci. USA 96, 657–662. Youngson, C., Nurse, C., Yeger, H., and Cutz, E. (1993). Oxygen sensing in airway chemoreceptors. Nature 365, 153–155. Zaidi, N.F., Lagenaur, C.F., Abramson, J.J., Pessah, I., and Salama, G. (1989). Reactive disulfides trigger Ca2⫹ release from sarcoplasmic reticulum via an oxidation reaction. J. Biol. Chem. 264, 21725–21736. Zhang, J.Z., Wu, Y., Williams, B.Y., Rodney, G., Mandel, F., Strasburg, G.M., and Hamilton, S.L. (1999). Oxidation of the skeletal muscle Ca2⫹ release channel alters calmodulin binding. Am J. Physiol. 276, C46–C53. Zhao, Y., Brandish, P.E., Ballou, D.P., and Marletta, M.A. (1999). A molecular basis for nitric oxide sensing by soluble guanylate cyclase. Proc. Natl. Acad. Sci. USA 96, 14753–14758.