Type 1 Inositol (1,4,5)-Trisphosphate Receptor Activates Ryanodine Receptor 1 to Mediate Calcium Spark Signaling in Adult Mammalian Skeletal Muscle*♦

Background: Osmotic stress triggers RyR-mediated Ca2+ sparks that are spatially confined at the periphery of muscle fibers. Results: Pharmacological intervention of IP3 production and genetic ablation of IP3R suppressed RyR-mediated Ca2+ sparks. Conclusion: Activation of the type 1 IP3R is necessary to induce Ca2+ sparks. Significance: This work highlights a potential interaction between IP3R and RyR in mediating Ca2+ signaling in adult skeletal muscle fibers. Functional coupling between inositol (1,4,5)-trisphosphate receptor (IP3R) and ryanodine receptor (RyR) represents a critical component of intracellular Ca2+ signaling in many excitable cells; however, the role of this mechanism in skeletal muscle remains elusive. In skeletal muscle, RyR-mediated Ca2+ sparks are suppressed in resting conditions, whereas application of transient osmotic stress can trigger activation of Ca2+ sparks that are restricted to the periphery of the fiber. Here we show that onset of these spatially confined Ca2+ sparks involves interaction between activation of IP3R and RyR near the sarcolemmal membrane. Pharmacological prevention of IP3 production or inhibition of IP3R channel activity abolishes stress-induced Ca2+ sparks in skeletal muscle. Although genetic ablation of the type 2 IP3R does not appear to affect Ca2+ sparks in skeletal muscle, specific silencing of the type 1 IP3R leads to ablation of stress-induced Ca2+ sparks. Our data indicate that membrane-delimited signaling involving cross-talk between IP3R1 and RyR1 contributes to Ca2+ spark activation in skeletal muscle.

Ca 2ϩ sparks are localized Ca 2ϩ release events originating from the opening of clustered ryanodine receptors (RyR) 4 on the sarcoplasmic reticulum in muscle cells (1)(2)(3). In cardiac muscle, these spontaneous Ca 2ϩ release events underlie the rhythmic contractile activity of the heart. In skeletal muscle, opening of the type 1 ryanodine receptor (RyR1) is strictly controlled by the voltage sensor located on the sarcolemmal membrane, and thus spontaneous Ca 2ϩ sparks are rarely observed during resting conditions. We previously showed that application of transient osmotic stress led to robust Ca 2ϩ sparks that are predominantly distributed at the periphery of the fibers from young, normal skeletal muscles (4). These Ca 2ϩ spark events involve Ca 2ϩ release from RyR1 as knock-out of RyR3 does not prevent the activation of Ca 2ϩ sparks (5). Immunostaining and electron microscopy studies established that RyR1 is distributed throughout the junctional sarcoplasmic reticulum membrane in association with the transverse tubular invagination of the sarcolemmal membrane (6); thus, activation of these spatially confined Ca 2ϩ sparks must reflect localized activation of RyR1 channels near the periphery of the muscle fiber. At present, it is not known whether physical perturbation of the junctional membrane structure and/or production of a local second messenger are essential for the initiation of stressinduced Ca 2ϩ sparks in skeletal muscle.
It is known that application of osmotic stress can affect stretch-activated phospholipases on the plasma membrane and lead to production of phosphatidylinositol 4,5-bisphosphate (PI(4,5)P 2 ) and inositol (1,4,5)-trisphosphate (IP 3 ), which could impact intracellular Ca 2ϩ signaling (7,8). Several studies in smooth muscle and cardiomyocytes demonstrated that activation of IP 3 receptors (IP 3 R) can contribute to RyR-mediated Ca 2ϩ signaling in these cells and shape the spatial distribution and kinetic properties of Ca 2ϩ sparks in smooth and cardiac muscle cells (9 -14). Although some studies have shown that the IP 3 R can affect intracellular Ca 2ϩ signaling in skeletal muscle (15)(16)(17) and potentially contribute to excitationtranscription coupling (18), the influence of IP 3 R on RyRmediated Ca 2ϩ spark signaling in skeletal muscle has not been clearly resolved.
In this study, we tested the hypothesis that cross-talk between IP 3 R and RyR underlies activation of stress-induced Ca 2ϩ sparks in mammalian skeletal muscle. We found that inhibition of IP 3 R by pharmacological agents suppresses osmotic stress-induced Ca 2ϩ sparks and that UV photo-release of caged IP 3 could increase Ca 2ϩ spark activity in intact skeletal muscle fibers. Selective activation of RyR1 near the periphery of the sarcolemma through coupling with IP 3 R appears to contribute to the onset of localized Ca 2ϩ sparks as RNAi silencing of IP 3 R1 abolishes stress-induced skeletal muscle Ca 2ϩ sparks. Our data provide direct evidence that IP 3 R1 modulates RyR1mediated Ca 2ϩ spark signaling in skeletal muscle.

Isolation of Flexor Digitorum Brevis (FDB) Muscle Fibers-
The method was described in previous work (4,19). Briefly, FDB muscles were surgically removed from mice in zero Ca 2ϩ Tyrode's buffer (in mM) 140 NaCl, 5 KCl, 10 HEPES, 2 MgCl 2 (pH 7.2) and incubated in Tyrode's buffer containing 2 mg/ml type I collagenase (Sigma C-5138) for 90 min at 37°C. After two washes in isotonic Tyrode's buffer containing 2.5 mM Ca 2ϩ (osmolality ϭ 290 mosM), muscle fibers were gently dissociated by several passages through a series of micropipette tips of gradually decreasing diameter. Individual FDB muscle fibers were plated onto ⌬TC3 glass-bottomed Petri dishes (Fisher Scientific) in Tyrode's buffer and used for experimentation within 6 h.
Confocal Ca 2ϩ Imaging and Spark Analysis-Imaging of intracellular Ca 2ϩ levels used previously described methods (4,19). Individually isolated muscle fibers were loaded with Fluo4-AM (10 M) for 60 min at room temperature. Measurements of Ca 2ϩ sparks were performed using a Bio-Rad Radiance-2100 confocal microscope equipped with an argon laser (488 nm) and a ϫ40, 1.3 NA oil immersion objective. For Ca 2ϩ spark measurements, fibers were perfused with hypotonic solution (osmolality ϭ 170 mosM) containing (in mM) 70 NaCl, 5 KCl, 10 Hepes, 2.5 CaCl 2 , 2 MgCl 2 (pH 7.2), for 100 s to induce swelling before perfusion was switched back to the initial isotonic Tyrode's buffer (osmolality ϭ 290 mosM). The osmolality of all solutions was measured using an Advanced model 3300 micro-osmometer (Advanced Instruments). Line scan images of 512 pixels in length were acquired at a sampling rate of 2 ms/line, and serial x-y images of muscle fibers were acquired at 3.08 s/frame.
Patch Clamp Electrophysiology-This approach was adapted from the methods developed by Jacquemond and colleagues (20,21) and adapted in our recent publication (19).
Photo-release of IP 3 -ciIP 3 /PM (caged membrane-permeant derivative of IP 3 ) was dissolved in DMSO to yield a final concentration of 10 M. FDB fibers were loaded with 10 M ciIP 3 /PM for 45 min and an additional 30 min with 10 M Fluo-4 AM. Photo-uncaging was performed according to Zhang et al. (12). The caged IP 3 was released into the fibers by photolysis of the compound using a UV lamp following the protocol previously described (12). Only one photolysis was induced in each fiber.
shRNA Design-Multiple short hairpin RNA (shRNA) probes targeting common sequences on the mRNA for IP 3 R1 and IP 3 R2 were screened for their efficacy in knocking down IP 3 R1 and IP 3 R2 protein expression using transfection in C2C12 cells. We found one shRNA IP 3 R probe, 5Ј-GATCGACTACAGGA-AGAACCAGGAGTACTTCAAGAGAGTACTCCTGGTTC-TTCCTGTAGTCTTTTTT-3Ј, that was effective in knocking down the expression of IP 3 R1 and IP 3 R2. shRNA IP 3 R1 for specific knockdown of IP 3 R1 expression was according to the published sequence (22). Bold underlined nucleotides represent the conserved sequence on the target mRNAs. shRNA control targets sequence in the luciferase cDNA as described previously (19). These shRNA probes were annealed to the pU6-mRFP expression vector (19).
Electroporation of Plasmid DNA into Adult Muscle-Plasmid delivery followed previously published methods (19). For all experimental results reported here, mice were sacrificed at 14 days after electroporation, and FDB muscles were surgically removed. IP 3 R2 Ϫ/Ϫ mice and age-matched wild-type control mice (11) were kindly provided by Dr. Ju Chen.
Western Blot Preparation-RFP expression was used as an indicator to measure the transfection efficiency of shRNA into the FDB muscle. Untransfected muscle bundles were trimmed out from the RFP-positive muscle for preparation of tissue lysates. The lysates were separated on a 6 -10% SDS-PAGE gel. Antibodies used for immunoblots were as follows: rabbit polyclonal antibodies against IP 3 R1 and IP 3 R2 (gifts from Drs. J. Maxwell and G. Mignery (Loyola University) and J. Chen (University of California, San Diego (UCSD)); polyclonal anti-CSQ1 (from Dr. E. K. Kim (Gwangju Institute of Science and Technology)); mouse monoclonal antibodies against RyR1 (34c-ABR from Affinity BioReagents), actinin (from Sigma), and sarcoplasmic reticulum Ca 2ϩ -ATPase (SERCA) (from Affinity BioReagents); and goat polyclonal antibody of phospholipase C␦ (PLC␦) (from Dr. L. Runnels (University of Medicine and Dentistry of New Jersey (UMDNJ)). HRP-conjugated RFP antibody was from Abcam. Densitometry analysis was performed using ImageJ software.
Resting and Store Ca 2ϩ Measurement-Individual RFP-positive FDB fibers (as indicator for transfection of shRNA) were loaded with 10 M fura-2 AM. The ratio of fura-2 fluorescence at excitation wavelength of 340 and 380 nm was measured using a PTI spectrofluorometer (Photon Technology International) to assess the resting [Ca 2ϩ ] i level. Zero Ca 2ϩ Tyrode's solution was perfused onto the fiber before adding either caffeine/ryanodine or ionomycin (Sigma) to induce Ca 2ϩ store release.
Statistical Analysis-Results are presented as mean Ϯ S.E. as tested for statistical significance by Student's t test, *, p Ͻ 0.01.

Interference with IP 3 R Signaling Affects Stress-induced Ca 2ϩ
Sparks in Skeletal Muscle-We showed in our previous publication (4) that a characteristic feature of osmotic stress-induced Ca 2ϩ sparks in intact skeletal muscle is their spatial distribution at the periphery of the muscle fiber directly underneath the sarcolemma (Fig. 1a). These osmotic stress-induced Ca 2ϩ sparks are mediated by activation of RyR channel because preincubation of the FDB muscle fibers with 10 M ryanodine could completely abolish the appearance of Ca 2ϩ sparks (4). To test whether changes in sarcolemmal membrane potential could influence the onset of Ca 2ϩ sparks in skeletal muscle, we used electrophysiological means to clamp the membrane potential at Ϫ70 mV following the method developed by Jacquemond and colleagues (20,21). As shown in Fig. 1b, an apparently normal Ca 2ϩ spark response was observed when the muscle fibers were clamped at Ϫ70 mV during application of osmotic stress. This suggests that changes in sarcolemmal membrane potential are not required in activation of osmotic stressinduced Ca 2ϩ sparks in muscle fibers. Additional studies show that changes in Cl Ϫ flux associated osmotic stress did not appear to contribute to the onset of Ca 2ϩ sparks (supplemental Fig. S1a).
Exposure of cells to osmotic stress has been shown to increase PI(4,5)P 2 production (8) and to activate PLC that can hydrolyze membrane-delimited PI(4,5)P 2 to IP 3 and diacylglycerol (7). Given the well known contribution of this signaling cascade to Ca 2ϩ spark activation in various tissues (9 -14), we used pharmacological inhibitors of PLC to test whether IP 3 R was involved in the generation of osmotic stress-induced Ca 2ϩ sparks. Preincubation of FDB fibers with U73122 (5 M), a pharmacological inhibitor of PLC, led to complete suppression of the osmotic stress-induced Ca 2ϩ spark activity (Fig. 1d). As a control, FDB fibers treated with U73343 (an inactive chemical analog for U73122) displayed normal onset of osmotic stress Ca 2ϩ sparks (Fig. 1c). Moreover, pretreatment of FDB fibers with xestospongin C (10 M) (Fig. 1e) that specifically inhibited IP 3 R channel activation (18) greatly suppressed the Ca 2ϩ spark activity following osmotic stress. Quantification analysis showed an average of 0.5 Ϯ 0.1 Ca 2ϩ spark events/min when compared with 4 Ϯ 0.2 Ca 2ϩ spark events/min in DMSOtreated fibers (Fig. 1a). Kinetic analysis of the amplitude (⌬F/F 0 ) and full duration at half-maximum (FDHM) of the remaining Ca 2ϩ sparks showed similar Ca 2ϩ spark amplitude (0.85 Ϯ 0.02; 0.92 Ϯ 0.03) but reduced FDHM (67 Ϯ 21.0ms; 102.2 Ϯ 7.6 ms) in xestospongin C-treated fibers (Fig. 1e) when compared with DMSO-treated fibers (Fig. 1a). These findings provide initial evidence that activation of IP 3 R might be required for RyRmediated Ca 2ϩ sparks in skeletal muscle.
Facilitation of Ca 2ϩ Spark Activity through Uncaging of IP 3 -To test whether local activation of IP 3 R could produce a Ca 2ϩ spark response in intact skeletal muscle fibers, we used UV flash photo-uncaging to produce local elevation of IP 3 by photo-un-caging of a caged IP 3 analog, ciIP 3 /PM (12). A pulse of UV flash photo-uncaging of IP 3 was applied at 20 s after the start of a 60-s confocal line scan of the fibers. Application of UV photo-uncaging in the DMSO-treated fibers did not affect the resting cytosolic Ca 2ϩ level (Fig. 2a, top panel). Uncaging of IP 3 in a resting muscle fiber bathed in an isotonic solution also did not produce any immediate Ca 2ϩ release (Fig. 2a, middle panel).
Other muscle fibers were challenged with osmotic stress, and IP 3 uncaging was performed at 7 min after osmotic stress, a point when Ca 2ϩ spark activity has greatly subsided. In greater than 50% of the fibers tested, we observed a significant increase in Ca 2ϩ spark events within 20 s of the uncaging of IP 3 in fibers that had previously been exposed to osmotic stress (Fig. 2a,  lower panel, and 2b). The average amplitude of Ca 2ϩ sparks was  5, and 19). a, spatially confined Ca 2ϩ sparks near the sarcolemmal region were observed in FDB muscle fiber following transient exposure to a hypo-osmotic stress (170 mosM) with a normal Tyrode's solution containing 2.5 mM Ca 2ϩ (290 mosM) and 0.1% DMSO (n ϭ 20 fibers). b, clamping of FDB fiber at Ϫ70 mV did not affect osmotic stress-induced Ca 2ϩ sparks in a normal Tyrode's solution supplemented with 0.1 mM anthracene 9-carboxylic acid (n ϭ 10 fibers). c, FDB fibers treated with U73343 (as a negative control) showed abundant Ca 2ϩ sparks similar to untreated controls (n ϭ 15 fibers). d, application of PLC inhibitor (5 M U73122) prevented onset of osmotic stress-induced Ca 2ϩ sparks (n ϭ 12 fibers). e, application of 10 M xestospongin C, an IP 3 R antagonist, prevented appearance of Ca 2ϩ sparks (n ϭ 15 fibers). similar before and after uncaging of IP 3 (Fig. 2c). In both resting (Fig. 2a, middle panel) and osmotic stress- (Fig. 2a, bottom  panel) treated fibers, we observed diffuse Ca 2ϩ release along with gradual elevation of cytosolic Ca 2ϩ at later stages of the experiment (beyond 20 s after UV flash). This Ca 2ϩ release could result from diffusion of IP 3 to other IP 3 R targets (18,23) rather than direct photo-damage to the muscle fiber as application of UV flash to the DMSO only-treated fibers did not result in any distinct Ca 2ϩ release (Fig. 2a). Considering the immediate effect of IP 3 R in activating RyR-associated Ca 2ϩ sparks in smooth muscle cells (12), we therefore focused our analysis on the production of Ca 2ϩ sparks immediately following uncaging of IP 3 by analyzing the characteristics of Ca 2ϩ sparks appearing within 20 s of photo-uncaging of IP 3 . In this experiment, line scans were conducted at the periphery of the fiber as we normally observed the Ca 2ϩ sparks in this region. Additional line scan recording at the middle of fiber upon photo-uncaging of IP 3 did not reveal any increase in Ca 2ϩ release events over base line (not shown). These findings provide direct evidence that IP 3 can act as a diffusible second messenger to facilitate activation of Ca 2ϩ sparks in skeletal muscle.
Our results are consistent with early studies from Pozzan and colleagues (16), who demonstrated that injection of IP 3 in skinned muscle preparation could elicit intracellular Ca 2ϩ release from skeletal muscle. Many subsequent studies from other investigators have revealed a role for IP 3 R in facilitating the RyR-mediated Ca 2ϩ release in smooth muscle and cardiac muscle. Recent studies from Jaimovich and colleagues (18) demonstrated a role for IP 3 R in modulation of excitation-transcription coupling in skeletal muscle; however, controversy exists in this topic as Blaauw et al. (24) did not find a significant role for IP 3 in skeletal muscle Ca 2ϩ signaling. The observation that uncaging of IP 3 alone could not produce Ca 2ϩ spark events in muscle fibers without pre-exposure to osmotic stress suggests the possibility that other factors such as physical uncou-pling of voltage sensor from RyR during osmotic stress could be required for induction of the robust Ca 2ϩ spark events (4,25).

Knockdown of IP 3 R Ablates Stress-induced Ca 2ϩ Sparks in Skeletal
Muscle-Toward understanding the role of IP 3 Rs in regulation of Ca 2ϩ spark signaling in skeletal muscle, we examined the distribution of IP 3 R isoforms in skeletal muscle. Western blots showed that both IP 3 R1 and IP 3 R2 were present in the skeletal muscle and that IP 3 R3 was absent (Fig. 3a). Immunolocalization revealed staining of IP 3 R1 and IP 3 R2 in the perinuclear and subsarcolemmal region where Ca 2ϩ sparks appear (supplemental Fig. S2). To directly test whether IP 3 R is required for induction of Ca 2ϩ sparks in skeletal muscle, we used shRNA to simultaneously knock down both IP 3 R1 and IP 3 R2 in the FDB muscle using a probe targeting conserved sequences in the two mRNAs (Fig. 3b). As a control, we used an shRNA sequence targeting the luciferase cDNA (19). Plasmid DNA was delivered to FDB muscles in mice using electroporation-mediated transfection (19), and after 14 days, FDB muscles were removed. Western blotting showed that both IP 3 R1 and IP 3 R2 expression was significantly reduced in these muscles, whereas the expression of other Ca 2ϩ regulatory proteins was not affected (Fig. 3c). The shRNA sequence was cloned into an RFP expression vector, allowing for identification of FDB muscle fibers with targeted knockdown of IP 3 R1 and IP 3 R2 (Fig. 3d). Muscle fibers with knockdown of both IP 3 R1 and IP 3 R2 (based on the presence of RFP fluorescence) were challenged with osmotic stress. Although robust Ca 2ϩ sparks were observed in the shRNA control fibers (Fig. 4a, top panel), shRNA IP 3 R fibers displayed very few, if any, Ca 2ϩ sparks following osmotic stress (Fig. 4a, bottom panel). Quantification of Ca 2ϩ spark frequency showed that these events were extremely rare in the shRNA IP 3 R when compared with the shRNA control fibers (Fig. 4c). This result was in accordance with our data shown in Fig. 1 where pharmacological interference of IP 3 R signaling had a significant impact on the activation of Ca 2ϩ sparks in skeletal muscle. Kinetic . shRNA-mediated knockdown of IP 3 R1 and IP 3 R2 in skeletal muscle. a, Western blots were performed using lysates from gastrocnemius muscle, lung, and brain tissues derived from WT mice (n ϭ 6 mice). 10-fold of muscle lysates when compared with brain and lung lysates were loaded. IP 3 R1 and IP 3 R2, but not IP 3 R3, were present in the skeletal muscle. b, shRNA IP 3 R was designed to target a common sequence on IP 3 R1 and IP 3 R2. Knocking down both IP 3 Rs did not alter the expression of other Ca 2ϩ regulatory proteins. c, densitometry analysis showed that shRNA IP 3 R resulted in significant knockdown of IP 3 R1 and IP 3 R2 in FDB muscle (n ϭ 8 mice). A. U., arbitrary units. d, FDB fibers with knockdown of IP 3 Rs could be identified by the appearance of red fluorescence.
analysis of the remaining Ca 2ϩ sparks showed that knockdown of IP 3 R1 and IP 3 R2 did not affect the Ca 2ϩ spark amplitude (⌬F/F 0 ) between shRNA control (0.96 Ϯ 0.04, n ϭ 20) and shRNA IP 3 R (0.91 Ϯ 0.04, n ϭ 30), whereas FDHM changed from 88.3 Ϯ 5.5 ms in the shRNA control to 44.9 Ϯ 21.0 ms in the shRNA IP 3 R fibers. The reduction in frequency of Ca 2ϩ sparks following knockdown of IP 3 R was not the result of changes in the resting intracellular Ca 2ϩ levels or loading of the internal Ca 2ϩ stores. fura-2 ratiometric measurements showed that the resting [Ca 2ϩ ] i levels were similar between shRNA IP 3 R (0.47 Ϯ 0.01, n ϭ 15) and shRNA control (0.43 Ϯ 0.02, n ϭ 15) (Fig. 4c). Ionomycin and caffeine plus ryanodine-induced release of intracellular Ca 2ϩ stores showed no significant difference between shRNA control and shRNA IP 3 R fibers (Fig. 4c).
Activation of IP 3 R1, but Not IP 3 R2, Is Involved in RyR1-mediated Ca 2ϩ Spark in Skeletal Muscle-Previous studies have shown that different IP 3 R isoforms exhibit different functional properties in Ca 2ϩ signaling (22,26,27). We next asked whether a specific IP 3 R isoform is required for RyR-mediated Ca 2ϩ sparks. Genetic ablation of IP 3 R2 produced a viable mouse model (11), allowing for examination of stress-induced Ca 2ϩ sparks in adult skeletal muscle. When examined under our experimental conditions, we found that the IP 3 R2 Ϫ/Ϫ skeletal muscle fibers displayed robust peripheral Ca 2ϩ sparks upon osmotic stress stimulation (Fig. 5b), which was similar to the results in wild-type (WT) control or fibers transfected with the shRNA control (Fig. 5a).
We then tested whether IP 3 R1 was required for the onset of Ca 2ϩ sparks in skeletal muscle. We used electroporation to deliver an shRNA probe that specifically targets knockdown of IP 3 R1 (shRNA IP 3 R1) (22), without affecting the expression of IP 3 R2 (Fig. 5e). We found that knocking down IP 3 R1 alone in FIGURE 5. IP 3 R1 but not IP 3 R2 contributes to regulation of RyR-mediated Ca 2؉ spark activity. FDB muscle derived from the IP 3 R2 Ϫ/Ϫ mice displayed robust peripheral Ca 2ϩ sparks following osmotic stress (b, n ϭ 18 fibers), similar to those observed in age-matched WT muscle (a, n ϭ 18 fibers). c, knockdown of IP 3 R1 alone suppressed Ca 2ϩ sparks in FDB muscle following osmotic stress (n ϭ 22 WT/shRNA IR 3 P1 fibers). d, compromised Ca 2ϩ spark response was also observed in the IP 3 R2 Ϫ/Ϫ muscle fiber following electroporation of shRNA IP 3 R1 (n ϭ 16 fibers). e, Western blots showed the specificity of shRNA IP 3 R1 in knocking down IP 3 R1, but not IP 3 R2 (n ϭ 4 mice). f, timedependent changes in Ca 2ϩ spark activity in FDB fibers following hypo-osmotic stress were quantified. Knocking down of IP 3 R1 from WT muscle led to near complete suppression of Ca 2ϩ spark activity (red) when compared with robust Ca 2ϩ spark events in WT muscle that decline with time (black). Muscle fibers derived from IP 3 R2 Ϫ/Ϫ mice showed normal Ca 2ϩ spark response (blue), and knockdown of IP 3 R1 led to complete ablation of Ca 2ϩ events (aqua). FDB muscle derived from WT mice could completely suppress Ca 2ϩ sparks following osmotic stress stimulation (Fig. 5c). Similar results were observed with electroporation of shRNA IP 3 R1 into the IP 3 R2 Ϫ/Ϫ FDB muscle, where Ca 2ϩ spark response was nearly completely ablated (Fig. 5, d and f). No significant differences were observed in the average Ca 2ϩ spark amplitude (⌬F/ F 0 ) of WT (0.97 Ϯ 0.05, n ϭ 18), IP 3 R2 Ϫ/Ϫ (0.93 Ϯ 0.04, n ϭ 18), shRNA IP 3 R1 (0.89 Ϯ 0.06, n ϭ 22), and shRNA IP 3 R1 in IP 3 R2 Ϫ/Ϫ background (0.85 Ϯ 0.08, n ϭ 16). Although the average FDHM was similar between WT (105 Ϯ 10.7 ms) and IP 3 R2 Ϫ/Ϫ (100 Ϯ 11.3 ms), it was significantly reduced in the shRNA IP 3 R1 in WT (45.7 Ϯ 10.9 ms) or in IP 3 R2 Ϫ/Ϫ fibers (42.5 Ϯ 15.4 ms). Overall, our findings indicate that activation of IP 3 R1, and not IP 3 R2, was required to trigger RyR-mediated Ca 2ϩ sparks in skeletal muscle.

DISCUSSION
We present evidence for a role for IP 3 R in regulation of RyRmediated Ca 2ϩ sparks in skeletal muscle. The Ca 2ϩ spark response in skeletal muscle was absent with pharmacological intervention that inhibited activation of the IP 3 R channels. It was also absent in muscle fibers following knockdown of IP 3 R1; thus, activation of IP 3 R1 is required for the onset of osmotic stress-induced Ca 2ϩ sparks in mammalian skeletal muscle. Based on our findings, we propose that initial Ca 2ϩ release through IP 3 R1 channel coupled with transient membrane deformation activates neighboring RyR1, leading to the production of peripherally localized Ca 2ϩ sparks in intact skeletal muscle fibers.
Although transient elevation of IP 3 through photo-uncaging could amplify RyR-mediated Ca 2ϩ release, the frequency of Ca 2ϩ sparks triggered by IP 3 was much less when compared with the osmotic stress-induced Ca 2ϩ sparks. Moreover, significant elevation of Ca 2ϩ sparks associated with photo-uncaging of IP 3 was only observed in muscle fibers following exposure to osmotic stress. Thus, stress-induced Ca 2ϩ sparks in mammalian skeletal muscle would likely require two cellular events, one involving production of a local IP 3 second messenger and the other involving structural changes at the triad junction that would lead to uncoupling of RyR1 channel inhibition by the voltage sensor on the sarcolemmal membrane (4,25). Our study added additional evidence that osmotic stress-induced Ca 2ϩ sparks in adult skeletal muscle does not involve changes in ion flux across the sarcolemmal membrane or changes in membrane potential.
A distinct feature of the stress-induced Ca 2ϩ sparks in skeletal muscle is their confinement to the periphery of the muscle fiber. We show that the IP 3 R channels are preferentially localized near the subsarcolemmal region of the muscle fiber, which could potentially facilitate local activation of the RyR1 channel through Ca 2ϩ as an intermediate messenger. The asymmetrical distribution of phospholipids on the sarcolemmal membrane (28,29) could result in localized activation of PLC and IP 3 production, which could represent a peripherally confined signal for activation of the RyR1 channel in skeletal muscle. Although we showed that IP 3 R2 is not required for the osmotic stressinduced Ca 2ϩ sparks in skeletal muscle, a specific role for IP 3 R1 in the activation of Ca 2ϩ sparks is established through our shRNA silencing approaches. Although the different Ca 2ϩ sig-naling properties of the different IP 3 R isoforms (26,27) could contribute to the specificity of IP 3 R1 in regulating RyR1-mediated Ca 2ϩ sparks, it is not clear why IP 3 R2 is not required for Ca 2ϩ spark activation. Future studies are necessary to test whether a direct interaction between the different isoforms of IP 3 R and RyR plays a role in their functional cross-talk in modulating the spatial and temporal aspects of Ca 2ϩ signaling in skeletal muscle or whether other intermediate messengers that favor the IP 3 R1 activation pathway could participate in modulation of the stress-induced changes in Ca 2ϩ signaling in skeletal muscle.
The role for IP 3 in Ca 2ϩ signaling in skeletal muscle was discovered over 25 years ago (15,16); however, cross-talk between IP 3 R and RyR in modulation of Ca 2ϩ signaling has mostly been studied in cardiac and smooth muscles (9 -13). Altered Ca 2ϩ release mediated by cross-talk between IP 3 R and RyR has been linked to hypertrophic cardiomyopathy and arrhythmia (9, 11, 13, 14, 30 -33), indicating the role of such cross-talk in physiology and pathophysiology. Peripherally confined Ca 2ϩ sparks in smooth muscle have been linked to activation of Ca 2ϩ -dependent K ϩ channel, which is important for vasodilation (2,34). Numerous studies in other cell types have shown the physiological relevance of localized Ca 2ϩ signals in vesicle fusion (28) and cytoskeletal reorganization (35). Our findings of IP 3 R cross-talk with RyR1 in skeletal muscle suggest that this may be a mechanism governing normal Ca 2ϩ regulation in skeletal muscle, and these findings also open a new line of investigation into potential pathophysiologic mechanism during muscle diseases. Several studies showed that the pathological consequence of altered Ca 2ϩ signals observed in the dystrophic skeletal muscle fibers was associated with abnormal IP 3 R distribution (36,37). Knockdown of IP 3 R1 could minimize cell death in dystrophin-deficient muscle fibers (38). Our previous studies showed that Ca 2ϩ spark in the dystrophic (mdx) muscle fibers was irreversible and no longer peripherally confined (4). Hence, aberrant Ca 2ϩ sparks observed in the dystrophic skeletal muscle could be associated with altered IP 3 R1 expression, localization, or activation.