Azumolene Inhibits a Component of Store-operated Calcium Entry Coupled to the Skeletal Muscle Ryanodine Receptor*

Dantrolene reduces the elevated myoplasmic Ca2+ generated during malignant hyperthermia, a pharmacogenetic crisis triggered by volatile anesthetics. Although specific binding of dantrolene to the type 1 ryanodine receptor (RyR1), the Ca2+ release channel of skeletal muscle sarcoplasmic reticulum, has been demonstrated, there is little evidence for direct dantrolene inhibition of RyR1 channel function. Recent studies suggest store-operated Ca2+ entry (SOCE) contributes to skeletal muscle function, but the effect of dantrolene on this pathway has not been examined. Here we show that azumolene, an equipotent dantrolene analog, inhibits a component of SOCE coupled to activation of RyR1 by caffeine and ryanodine, whereas the SOCE component induced by thapsigargin is not affected. Our data suggest that azumolene distinguishes between two mechanisms of cellular signaling to SOCE in skeletal muscle, one that is coupled to and one independent from RyR1.

Malignant hyperthermia (MH) 2 is a potentially fatal pharmacogenetic syndrome in which exposure to volatile anesthetics triggers uncontrolled elevation of myoplasmic Ca 2ϩ concentrations ([Ca 2ϩ ] i ), skeletal muscle hypercontracture, and hypermetabolism, resulting in a dramatic rise in body temperature (1,2). Mutations in the type 1 ryanodine receptor (RyR1), the major Ca 2ϩ release channel in skeletal muscle, are linked to MH susceptibility in pigs in an autosomal recessive manner (3)(4)(5). In humans, MH is transmitted as an autosomal dominant trait with incomplete penetrance, and the more than 80 mutations in RyR1 that have been identified appear in only about 50% of affected families (6,7). Muscle bundles from MH-susceptible patients are hypersensitive to RyR1 agonists, including caffeine, Ca 2ϩ , and halothane (8,9), the latter a member of the class of volatile anesthetics that triggers MH. Thus, the loss of control of RyR1-mediated Ca 2ϩ release likely contributes to the elevation of [Ca 2ϩ ] i observed in MH patients. To date, the only effective treatment for MH is dantrolene sodium, a skeletal muscle relaxant, which suppresses the uncontrolled rise in myoplasmic Ca 2ϩ , presumably by targeting RyR1 and suppressing its Ca 2ϩ channel activity (10 -12). Azumolene sodium is a structurally similar, equipotent analog of dantrolene, with an ϳ30-fold greater water solubility (13,14).
Whereas hyperactivity or leakiness of the RyR1 channel has been described as the primary physiological defect in MH-susceptible muscle, the molecular mechanism underlying the suppression of [Ca 2ϩ ] i by dantrolene remains controversial. Functional studies demonstrate partial inhibition of Ca 2ϩ release from isolated sarcoplasmic reticulum (SR) (11,15), and partial suppression of the elemental Ca 2ϩ spark signals in adult muscle fibers (16). However, functional studies have been unable to unequivocally demonstrate direct inhibition of the RyR1 Ca 2ϩ channel activity by dantrolene (12,17).
Sustained opening of RyR1 Ca 2ϩ channels leads to reduction of the Ca 2ϩ store within the SR lumen, a signal that activates store-operated Ca 2ϩ entry (SOCE) in skeletal muscle (18,19). Recent evidence from heterologous expression systems supports a role for the amino-terminal cytoplasmic foot structure of RyR1 in coupling to channels possibly involved in SOCE (20). Additionally, elevated Ca 2ϩ entry through cell surface Ca 2ϩ channels with pharmacology similar to SOCE has been linked to the elevation of [Ca 2ϩ ] i in muscular dystrophy (21). The contribution of SOCE to MH and the potential effect of dantrolene on SOCE have not been examined. In this study, we test the hypothesis that azumolene can influence SOCE function. Using Ca 2ϩ -sensitive fluorescence measurements of RyR1-dependent intracellular Ca 2ϩ transients and extracellular Ca 2ϩ entry via SOCE, we show that azumolene inhibits SOCE both in skeletal muscle fibers and in cultured cells expressing RyR1. Our results reveal two modes of SOCE activation. One mode is RyR1dependent and can be inhibited by the action of azumolene; the other is RyR1-independent and is insensitive to azumolene.

EXPERIMENTAL PROCEDURES
Cell Culture-C1148 cells, a Chinese hamster ovary (CHO) cell line stably expressing RyR1, were maintained in Ham's F-12 medium supplemented with 10% fetal bovine serum, 1% penicillin and streptomycin, and 0.5 mg/ml G-418 (22). Culture of C2C12 myogenic cells was described previously (23). Myotubes derived from C2C12 were used in experiments at day 5 of differentiation.
Intracellular Ca 2ϩ Measurement-C1148 cells were loaded with 10 M Fura-2 AM (Invitrogen) for 45 min at 37°C and allowed to de-esterify for 15 min at 25°C. Cells were then harvested and resuspended in balanced salt solution (BSS) containing (in mM) the following: 140 NaCl, 2.8 KCl, 2 CaCl 2 , 2 MgCl 2 , 10 HEPES, pH 7.2. 2.5 ϫ 10 6 cells were transferred into the cuvette system of a PTI spectrofluorometer (Photon Technology International, Princeton, NJ), and the changes in [Ca 2ϩ ] i were measured as changes in the ratio of Fura-2 fluorescence at excitation wavelength of 350 nm (F 350 ) and 380 nm (F 380 ), following exposure to various concentrations of caffeine and ryanodine (C/R). For measurement in 0.5 mM EGTA, cells were centrifuged and resuspended in BSS without CaCl 2 , and 0.5 mM EGTA was added immediately before recordings. Measurement of Ca 2ϩ in individual C2C12 myotubes was performed as described before (18). All experiments were conducted at 25 Ϯ 2°C. SOCE Assay; Mn 2ϩ Quenching of Fura-2-Mn 2ϩ is known to be able to permeate into cells via store-operated Ca 2ϩ channels (SOC), but it is impervious to surface membrane extrusion processes or SR uptake by Ca 2ϩ pumps. Hence, Mn 2ϩ fluorescence quenching represents a measurement of unidirectional Ca 2ϩ flux into cells via SOC (18,19). Briefly, to measure the Mn 2ϩ influx rate through the SOC machinery, thapsigargin (TG), or C/R, was applied to C1148 cells or C2C12 myotubes to induce SR Ca 2ϩ depletion in 0 mM extracellular Ca 2ϩ ([Ca 2ϩ ] o ), and 0.5 mM Mn 2ϩ was then added to the extracellular solution. The quenching of Fura-2 fluorescence by Mn 2ϩ was measured at the Ca 2ϩ -independent excitation wavelength of Fura-2 (360 nm). The decay of Fura-2 fluorescence upon Mn 2ϩ addition was expressed as percent decrease in Fura-2 fluorescence per unit time (the initial fluorescence is set to be equal to 100%). For all measurements of SOCE by Mn 2ϩ quenching, the maximally quenched fluorescence signal was established at the end of the experiment by lysing the cells with 1% Triton and was set equal to 0% fluorescence.
Dissociation of Individual Flexor Digitorum Brevis (FDB) Fibers and Measurement of SOCE-FDB fibers were enzymatically dissociated from 2-to 4-month-old C57Bl6/J male mice, following the procedure described in our previous study (24). For experiments performed at room temperature (25 Ϯ 2°C), individual muscle fibers were plated onto either uncoated (for TG treatment) or a silicon-coated ⌬TC3 dish (for C/R treatment) and loaded with 10 M Fura-2 AM at room temperature for 1 h. FDB fibers were then fastened by silicon drops at both ends to avoid contraction induced by C/R (25). For experiments performed at 35°C, FDB fibers were plated on uncoated ⌬T4 dishes, and the temperature was controlled by a thermal controller (Bioptechs Inc., Butler, PA). To prevent motion artifact in fibers associated with intracellular Ca 2ϩ release, 20 M N-benzyl-p-toluene sulfonamide (Sigma), a specific myosin II inhibitor (26), was applied.

Spatial and Temporal Resolution of SOCE in Skinned Muscle
Preparation-The detailed procedure for the application of this methodology in SOCE measurement has been described before (27). Briefly, single muscle fibers are dissected from extensor digitorum longus muscle of C57Bl6/J mice and cultured for 72 h. For SOCE measurements, fibers were mechanically skinned in the presence of Rhod-5N potassium salt (Invitrogen) to trap the dye conjugated with Ca 2ϩ into transverse tubules (T-tubule). For SR Ca 2ϩ content assessment, the fiber was mechanically skinned in the absence of Rhod-5 and Ca 2ϩ and then incubated with an intracellular-like solution (in mM, 140 potassium glutamate, 6.5 MgCl 2 , 6 creatine phosphatase, 0.5 CaCl 2 , 20 2-bromoethanesulfonate/BES-KOH) containing 0.2 mM EGTA and 20 M Rhod-5N AM (Invitrogen) for 1 h at room temperature to load the SR, followed by extensive washes and incubation for an additional 30 min to allow the complete deesterification of the dye. Treatment with 5 M p-trifluoromethoxy carbonyl cyanide phenylhydrazone (FCCP) (Sigma) effectively eliminated the fluorescence signal from mitochondria. A Bio-Rad 2100 confocal microscope (Zeiss, Thornwood, NY) was used to resolve the spatial and temporal distribution of Rhod-5N inside the T-tubules or SR compartment as described in Zhao et al. (27), with the following exception. The fiber was perfused with SR loading solution plus 20 M azumolene or 0.1% Me 2 SO for 2 min. After that, the fiber was exposed to an SR-depleting solution (in mM, 100 potassium glutamate, 16 sodium glutamate, 20 EGTA-KOH, 5 1,2-bis(o-aminophenoxy)ethane-N,N,NЈ,NЈ-tetraacetic acid, 0.07 MgCl 2 , 0.25 ATP, 1 creatine phosphatase, 10 BES-KOH) with C/R (30 mM/5 M), in the presence of azumolene or Me 2 SO carrier for 1000 s. These experiments were repeated six times, and mean values of 10 regions of interest per fiber were analyzed. Rhod-5N intensity was normalized to the maximal loading intensity, prior to the onset of SR Ca 2ϩ depletion. The above experiments were conducted at 25 Ϯ 2°C.
Statistics-Values are mean Ϯ S.E. Significance was determined by Student's t test or one-way analysis of variance. A value of p Ͻ 0.05 was used as criterion for statistical significance.

Azumolene Inhibits Extracellular Ca 2ϩ Entry in CHO Cells
Stably Transfected with RyR1-C1148 is a cell line derived from CHO cells that are stably transfected with RyR1 (22). These cells contain functional RyR1 channel on the endoplasmic reticulum (ER) membrane, in addition to the presence of an endogenous SOCE pathway. Thus, cell population assays (e.g. 2.5 ϫ 10 6 cells) using a cuvette system can be applied to evaluate RyR1-mediated changes in intracellular Ca 2ϩ signaling. As shown in Fig. 1, exposure of these cells to C/R leads to release of Ca 2ϩ from the ER. C/R treatment results in complete depletion of the ER Ca 2ϩ store, since it has been demonstrated previously that no further Ca 2ϩ release is observed with subsequent addition of TG or ionomycin (28). When the bath solution contains 2 mM Ca 2ϩ , the C/R-induced Ca 2ϩ transient displays an initial peak followed by a sustained tail component. The peak is somewhat attenuated, and the sustained Ca 2ϩ elevation is absent when the bath solution contains BSS without Ca 2ϩ and 0.5 mM EGTA, which results in a nominally 0 mM [Ca 2ϩ ] o . This indicates that extracellular Ca 2ϩ entry contributes to the development of both the peak and the sustained component of the C/R-induced Ca 2ϩ transient, although quite significantly to the latter. The sustained Ca 2ϩ elevation observed in 2 mM [Ca 2ϩ ] o could be substantially inhibited by 20 M 2-aminodiphenyl borate, an inhibitor of SOCE (29) (Fig. 1A). Taken together, these results suggest that an endogenous SOCE pathway likely mediates the entry of extracellular Ca 2ϩ following depletion of the C/R-dependent ER Ca 2ϩ store in C1148 cells.
To test the effects of azumolene on the RyR1-dependent Ca 2ϩ signaling, azumolene was added to C1148 cells from a Me 2 SO stock solution. In the presence of 20 M azumolene, the peak amplitude of C/R-induced Ca 2ϩ transients (in 2 mM [Ca 2ϩ ] o ) was significantly reduced from F 350 /F 380 ϭ 0.37 Ϯ 0.06 (in Me 2 SO carrier) to 0.20 Ϯ 0.03 (plus azumolene). This is consistent with the previously reported partial inhibitory effect of dantrolene on RyR1-dependent Ca 2ϩ transients in cells (17). Significantly, the sustained post-peak elevation of Ca 2ϩ , which our results suggest is because of SOCE, was significantly reduced by the presence of azumolene.
Note that the addition of azumolene leads to apparent instantaneous elevation of the Fura-2 signal, as shown in Fig. 1A, inset. This is because of the intrinsic autofluorescence of azumolene. The autofluorescence of azumolene displays a sharp peak at an excitation wavelength of 348 nm, and thus likely contributes to the apparent elevation of F 350 /F 380 Fura-2 signal shown in Fig. 1A. To correct for the autofluorescence of azumolene, we subtracted the component of F 350 Fura-2 fluorescence associated with azumolene addition in the calculation of F 350 /F 380 . As shown in Fig.  1B, the sustained component of Ca 2ϩ transients was completely inhibited by azumolene. This observation led us to hypothesize that azumolene may suppress extracellular Ca 2ϩ entry through SOCE.
We next used Mn 2ϩ quenching of Fura-2 fluorescence to test whether azumolene directly affects SOCE in C1148 cells. In this assay, Mn 2ϩ is supplied to the extracellular solution, and its entry through SOC quenches the intracellular Ca 2ϩ -dependent Fura-2 fluorescence (30 -32). Measurement of the decrease in Fura-2 fluorescence at an excitation wavelength of 360 nm, the Ca 2ϩindependent isosbestic point of Fura-2, provides an assessment of SOCE function. We found that preincubation of C1148 cells with 20 M azumolene significantly reduced the rate of Mn 2ϩ entry following depletion of the ER Ca 2ϩ store with C/R (Fig. 1,  C and D). These results directly demonstrate that azumolene is capable of inhibiting SOCE induced by C/R in CHO cells expressing RyR1.
Azumolene Inhibits SOCE in C2C12 Myotubes-The effect of azumolene on SOCE was further evaluated in myotubes derived from the C2C12 mouse myogenic cell line. Differentiated C2C12 myotubes were treated with C/R for 5 min in the absence of extracellular Ca 2ϩ to allow for complete depletion of the SR Ca 2ϩ stores (18). As shown in Fig. 2A, individual myotubes treated with Me 2 SO (vehicle control) exhibited steep quenching of Fura-2 fluorescence because of Mn 2ϩ influx. This defines the maximal C/R-triggered activation of SOCE measured in our system. C2C12 myotubes exposed to 20 M azumolene prior to C/R stimulation exhibited an ϳ70% reduction of SOCE compared with that of vehicle-treated control cells (Fig. 2, B and D). The peak amplitude of C/R-induced Ca 2ϩ  NOVEMBER 3, 2006 • VOLUME 281 • NUMBER 44 JOURNAL OF BIOLOGICAL CHEMISTRY 33479 release in the absence of extracellular Ca 2ϩ in C2C12 myotubes was not altered by azumolene (Fig. 2, A and B), with F 360 /F 390 ϭ 0.32 Ϯ 0.03 in Me 2 SO control versus 0.28 Ϯ 0.03 in azumolenetreated myotubes, a change that was not statistically significant (n ϭ 8 -11, p ϭ 0.38). Thus, this decrease in SOCE in azumolenetreated C2C12 myotubes does not appear to result from an inhibition of SR Ca 2ϩ release by azumolene.

Azumolene Inhibits Store-operated Calcium Channel
Interestingly, changing the order of drug treatment in the experimental protocol produces a significantly different result (Fig. 2C). Here, incubation of C2C12 myotubes with azumolene after C/R-initiated SR Ca 2ϩ depletion resulted in loss of the inhibitory effect of the drug on SOCE (Fig. 2D, also compare Fig. 2,  C and A). Me 2 SO carrier alone had no effect on the slope of Mn 2ϩ quenching, regardless of the order of application. Because a significant decrease in the slope of Mn 2ϩ quenching of Fura-2 fluorescence occurred only when azumolene was added prior to C/R stimulation, and not after, the effect of azumolene on SOCE in C2C12 myotubes may depend on the conformation of the RyR1 channel at the time of incubation with azumolene, if azumolene acts by binding to RyR1.
Dose-dependent Effects of Azumolene on SOCE in C2C12 Myotubes-Our previous study showed that SOCE in fetal skeletal muscle can be activated in a graded manner by the reduction of SR Ca 2ϩ store (25). To determine the effects of azumolene on the graded activation of SOCE in C2C12 myotubes, the quenching of Fura-2 by Mn 2ϩ was monitored from the beginning of C/R-initiated SR Ca 2ϩ release. As shown in Fig. 3A, myotubes pretreated with Me 2 SO exhibited a sigmoidal Mn 2ϩ quench curve, reflecting the graded activation of SOCE that follows reduction of the SR Ca 2ϩ store. Prior incubation of myotubes with 10 M azumolene for 2 min led to significant delay in activation of SOCE and an altered slope of the Mn 2ϩ quench curve.
To quantify the graded activation of SOCE, the first-order derivative of changes in F 360 , dF 360 /dt, was determined (Fig. 3B). This analysis led to the definition of the following two kinetic parameters of SOCE in skeletal muscles: m max , the peak slope of Mn 2ϩ quenching, reflecting the maximum degree of SOCE activation; and ⌬, the delay time to reach m max from the onset of Ca 2ϩ release from SR after addition of C/R. Using this analysis, we conducted systematic studies to resolve the dose-dependent effect of azumolene on m max and ⌬ of SOCE in C2C12 myotubes. As shown in Fig. 3, C and D, the steepest range of azumolene effect on ⌬ and m max was observed between concentrations of 0.1 to 20 M, a clinically relevant concentration range. Note that there appears to be a biphasic effect of azumolene on SOCE, e.g. a high affinity effect with an apparent K d close to 2 M, and a low affinity one that does not saturate under our experimental conditions.
Azumolene Does Not Affect TG-induced Activation of SOCE-SR Ca 2ϩ store in skeletal muscle can be depleted using other methods besides activation of RyR1 by C/R. TG, an inhibitor of the Ca 2ϩ -ATPase on SR, has been classically used to passively deplete the SR Ca 2ϩ store to activate SOCE (19). As shown in Fig. 4A, treatment of C2C12 myotubes with 10 M TG for 5 min in a bath solution containing 0 [Ca 2ϩ ] o led to depletion of SR Ca 2ϩ stores that is not affected by azumolene treatment ( p Ͼ 0.05). The F 360 /F 390 equaled 0.27 Ϯ 0.02 in Me 2 SO control versus 0.24 Ϯ 0.02 in the azumolene-treated group. This SR Ca 2ϩ depletion leads to maximum activation of SOCE, which is reflected in the steep Mn 2ϩ quenching of Fura-2 fluorescence. In contrast to the results shown in Fig. 2A, we found that the TG-induced activation of SOCE was not affected by azumolene, e.g. the slope of Mn 2ϩ quenching of Fura-2 fluorescence did not change significantly with the addition of azumolene (Fig. 4A).
The ability of azumolene to discriminate between C/R-and TG-induced SR Ca 2ϩ depletion suggests the existence of at least two pathways of SOCE activation in skeletal muscle, one RyR1-dependent and the other RyR1-independent. To further test this hypothesis, we performed the following studies. First, we tested whether there are additive effects of TG and C/R on SOCE activation in C2C12 myotubes. As shown in Fig. 4B, the slope of Mn 2ϩ quenching induced by prior exposure to C/R did not increase following subsequent addition of TG, suggesting that TG does not induce additional activation of SOCE. The lack of additive effects of TG-and C/R-induced activation of SOCE in C2C12 myotubes was further demonstrated in Fig. 4C, where addition of C/R to a C2C12 myotube that was previously treated with TG also did not elicit additional activation of SOCE. Second, we tested whether C/R and TG mediation of SOCE involve interacting or parallel pathways in C2C12 myotubes by determining the effect of order of addition of the two sets of drugs on the parameters of SOCE and the effect of azumolene on these. As shown in Fig. 4B, the inhibitory effect of azumolene on C/R-induced SOCE in C2C12 myotubes could be completely overcome by subsequent treatment with TG.  Horizontal dashed lines represent the basal Mn 2ϩ entry rate, whereas oblique dashed lines represent the Mn 2ϩ entry activated by SR store depletion. B, Me 2 SO or azumolene was added to C2C12 myotubes before TG-induced SR Ca 2ϩ depletion. Cells were treated with caffeine/ryanodine for 5 min, followed by Mn 2ϩ addition to the perfusate for 4 min, and then TG plus Mn 2ϩ was added to the perfusate. C, identical to B, with the order of caffeine/ryanodine and TG addition reversed. D, average data for the slope of F 360 derived from A to C. The two sequential slopes after each of the two treatment regimens in B and C were designated as S1 (boxed) and S2 (short dashed line), respectively (n ϭ 10 for each group tested). *, p Ͻ 0.05. Experiments were performed at room temperature (25 Ϯ 2°C).
This result suggests that TG can maximally activate SOCE despite azumolene inhibition of RyR1-mediated SOCE, indicating the two pathways to SOCE are parallel and access the same store of SR Ca 2ϩ . Indeed, in the converse experiment, azumolene does not inhibit TGinduced SOCE, and subsequent addition of C/R does not appear to affect the degree of SOCE activation (Fig. 4C), supporting the argument for parallel pathways to SOCE activation.
The data from multiple measurements summarized in Fig. 4D substantiate our conclusions. Although TG and C/R may share a final common target that leads to activation of SOCE, only the C/R-induced SOCE pathway is significantly affected by azumolene.

Differential Effects of Azumolene on SOCE Activation in FDB Muscle
Fibers-To complement our cell culture-based measurements of azumolene effects on SOCE, we tested azumolene in enzymatically dissociated FDB fibers from mice. For these measurements, we used 20 mM/5 M C/R or 20 M TG to deplete the SR Ca 2ϩ store for maximum activation of SOCE. To prevent motion artifacts, we employed two techniques. First, the silicongrease method was used to immobilize the individual FDB fibers onto the culture dish (25). Second, N-benzyl-p-toluene sulfonamide, a specific myosin II inhibitor with minimum alteration of Ca 2ϩ signaling (26), was used to prevent muscle contraction associated with intracellular Ca 2ϩ release. Similar to our results with C2C12 myotubes, we found that 20 M azumolene could significantly inhibit C/R-activated SOCE (Fig. 5A) but not the TG-induced SOCE (Fig. 5B).
The above analyses of the effect of azumolene on SOCE in FDB muscle fibers were all performed at room temperature (25 Ϯ 2°C) (Fig. 5C). Because previous studies have suggested that the effect of dantrolene on RyR-mediated Ca 2ϩ release may display temperature dependence (33), we performed a series of experiments at 35°C. As shown in Fig. 5D, azumolene affected SOCE in these experiments in a manner virtually identical to those observed at 25°C; the drug inhibited C/R-induced SOCE without significantly affecting TGinduced SOCE. Similar to our results in the C2C2 myotube, the SR Ca 2ϩ release in response to both C/R (0.47 Ϯ 0.06 in control group versus 0.46 Ϯ 0.06 in azumolene group, p Ͼ 0.05) and TG (0.48 Ϯ 0.05 in control group versus 0.46 Ϯ 0.09 in azumolene group, p Ͼ 0.05) was unaltered by azumolene.

Azumolene Uncouples Graded Activation of SOCE from SR Ca 2ϩ Release in Adult Skeletal Muscle
Fibers-To investigate the relationship between SR store depletion and SOCE activation, we monitored SOCE and SR Ca 2ϩ release by C/R in mechanically skinned skeletal muscle fibers using a recently adapted confocal microscopy methodology (27). The membrane-impermeant salt of Rhod-5N can be trapped inside the sealed T-tubule compartments of skinned skeletal muscle fibers (Fig. 6A). Upon initiation of Ca 2ϩ release from SR with C/R, activation of SOCE allows the flow of Ca 2ϩ ions from the sealed T-tubule compartment to the myoplasm, resulting in decreased Rhod-5N fluorescence. In the presence of azumolene, the rate and extent of decrease in Rhod-5N fluorescence are significantly reduced compared with the Me 2 SO control group. At the end of the exposure to the SR Ca 2ϩ depletion solution (1000 s), Rhod-5N fluorescence intensity equaled 0.40 Ϯ 0.03 in the Me 2 SO control group, whereas in the azumolene-treated fibers this value was 0.68 Ϯ 0.02 ( p Ͻ 0.05). This demonstrated that SOCE was reduced by azumolene.
In parallel experiments, we determined the effect of azumolene on C/R-induced Ca 2ϩ release by loading Rhod-5N AM, the membranepermeable form of the dye, into the SR of skinned muscle fibers, rather than the T-tubule system (Fig. 6B). To eliminate potentially confounding changes resulting from a mitochondrial Rhod-5N Ca 2ϩ signal, the mitochondrial electron transport inhibitor, 5 M FCCP, was added to the skinned muscle fiber prior to induction of SR Ca 2ϩ release by C/R and was present throughout the Ca 2ϩ depletion process. In the presence of FCCP, C/R induces SR Ca 2ϩ depletion that is demonstrated by a time-dependent decrease in SR Rhod-5N fluorescence. Fibers pretreated with azumolene display a slightly slower rate of fluorescence decrease (0.63 Ϯ 0.01), but no significant difference in final fluorescence was found when compared with the Me 2 SO control (0.56 Ϯ 0.01, p ϭ 0.26). This result suggests that the inhibitory effect of azumolene on SOCE does not result from its direct inhibition of the SR Ca 2ϩ release process.  6). B, left panel shows confocal images of skinned extensor digitorum longus fibers with Rhod-5N AM loaded into the SR compartment at three experimental time points as follows: maximal SR Ca 2ϩ loading, after treatment with FCCP, and after depleting the Ca 2ϩ store with C/R. Right panel illustrates the average cumulative change in Rhod-5N fluorescence following C/R treatment in the presence of either 0.1% Me 2 SO (green) or 20 M azumolene (red) (n ϭ 6). C, correlated changes of Rhod-5N fluorescence in SR (x axis) and T-tubule (y axis) illustrates graded activation of SOCE in response to SR Ca 2ϩ depletion under control conditions (green). SOCE, represented as loss of T-tubule Ca 2ϩ fluorescence, is significantly uncoupled from C/R-induced Ca 2ϩ release, as represented by loss of SR Ca 2ϩ fluorescence, after pretreatment with azumolene (red).
The correlation between changes in SR Ca 2ϩ content and SOCE activation is illustrated in Fig. 6C, where the normalized Rhod-5N fluorescence intensity of the T-tubule compartment is plotted against the intensity in the SR compartment over a time interval of 500 s following addition of C/R. During this period, close coupling between changes in SR Ca 2ϩ release and SOCE activation are observed under control conditions. In the presence of azumolene, however, this close coupling is disrupted, as reflected by the shallower correlation between SR Ca 2ϩ release and SOCE activation.

DISCUSSION
Elucidating the cellular mechanism(s) of dantrolene action on skeletal muscle Ca 2ϩ signaling is of great interest from both physiological and pathophysiological points of view. Because MH syndromes are linked to mutations in the RyR1 channel in various vertebrates, as well as in humans (2,4,34,35), and because dantrolene binds to a specific site on RyR1 (36), previous studies have focused on the role of dantrolene in modulating RyR1 channel activity. Functional studies demonstrate only partial inhibition of Ca 2ϩ release from isolated SR membrane vesicles (11,12,15) and partial suppression of the elemental Ca 2ϩ spark signals in adult muscle fibers (16). As a muscle relaxant, dantrolene can suppress the elevation of [Ca 2ϩ ] i in intact muscle fibers, yet conclusive evidence for direct inhibition of RyR1 channel activity by dantrolene is lacking. Our data suggest that a significant portion of the action of dantrolene and related compounds in the therapy of MH may stem from their inhibition of RyR1-coupled SOCE.
First, we demonstrate that azumolene disrupts the tight Ca 2ϩ release-coupled graded activation of SOCE that is normally seen in adult mouse skeletal muscle fibers without substantially inhibiting SR Ca 2ϩ release. Second, we show that azumolene can substantially inhibit the C/R-triggered SOCE in heterologous cells expressing RyR1, in cultured C2C12 myotubes, and in adult mouse skeletal muscle fibers, suggesting that this effect is dependent on RyR1. Third, we found that although TG-induced depletion of SR Ca 2ϩ stores leads to maximal rates of SOCE, this process was not susceptible to azumolene inhibition, indicating that azumolene does not inhibit all signals that can lead to the stimulation of SOCE. Fourth, we show that substantial inhibition of RyR1-coupled SOCE by azumolene occurs only when cells are treated with this drug prior to C/R-induced RyR1 activation and SR Ca 2ϩ depletion. Because C/R treatment produces prolonged RyR1 channel opening, the inability of azumolene to inhibit SOCE when added after C/R treatment is consistent with previously published in vitro studies demonstrating that dantrolene interacts preferentially with the closed state of RyR1 (36,37). Therefore, we hypothesize that the inhibitory effect of azumolene and, by extension, of dantrolene on SOCE results from drug binding to the closed state of RyR1 The discordance between the ability of azumolene to inhibit SOCE versus SR Ca 2ϩ release in FDB muscle fibers suggests that Ca 2ϩ itself is not the direct signal from RyR1 that stimulates SOCE. Furthermore, because azumolene does not inhibit TG-induced SOCE, azumolene cannot be acting at the level of the SOCE machinery itself. It therefore follows that azumolene is likely uncoupling the efficiency of a Ca 2ϩ -dependent RyR1 signal coupled directly or indirectly to the SOCE machinery and represents a novel hypothesis for the mechanism of action of this drug.
Because under control conditions there is no additivity between saturating effects of C/R-and TG-induced Ca 2ϩ release in their effects on SOCE, it is likely that the two systems for activating SOCE result from competition for the same intracellular Ca 2ϩ store in the cells examined here. Even when RyR1-coupled SOCE is inhibited by azumolene, TG is still able to activate SOCE at nearly the same rate as if azumolene is absent. Taken together, our data suggest that at least two different mechanisms, either through RyR1 or ER/SR Ca 2ϩ -ATPase, are capable of activating the SOCE machinery in mammalian skeletal muscle, which is consistent with previous studies (20,38).
Recent studies from Pessah and co-workers (39,40) have demonstrated a process of excitation-coupled Ca 2ϩ entry (ECCE) experimentally distinct from SOCE in cultured myotubes. ECCE is not sensitive to Ca 2ϩ store depletion but is activated by membrane depolarization and is sensitive to RyR1 conformation and mutations (39,40). Furthermore, they have presented preliminary evidence that dantrolene also affects ECCE, but not TG-induced SOCE, thereby discriminating between the two processes (41). It is possible that the integral membrane machinery and/or the attendant signaling mechanisms underlying these pathways of extracellular Ca 2ϩ entry (i.e. SOCE and ECCE) may be similar, if not identical, because both involve coupling to RyR1 and azumolene/dantrolene sensitivity.
Because both dantrolene and azumolene are therapeutic in the treatment of MH, the novel mechanism of drug action described here leads us to suggest that an elevated RyR1coupled signal to SOCE may contribute appreciably to the pathophysiology of MH, i.e. MH is as much a syndrome of exaggerated Ca 2ϩ entry as it is of exaggerated Ca 2ϩ release. By extension then, the therapeutic activity of dantrolene in MH may result from its ability to inhibit exaggerated Ca 2ϩ influx, rather than from its ability to inhibit SR Ca 2ϩ efflux. Further defining the role of dantrolene and azumolene in modulating various Ca 2ϩ -dependent aspects of muscle physiology should improve our knowledge of the machinery responsible for cellular Ca 2ϩ homeostasis. This may provide novel therapeutic targets for various human disorders linked to dysfunctional Ca 2ϩ signaling involving susceptible RyR isoforms and SOCE mechanisms.