Impaired Orai1-mediated Resting Ca2+ Entry Reduces the Cytosolic [Ca2+] and Sarcoplasmic Reticulum Ca2+ Loading in Quiescent Junctophilin 1 Knock-out Myotubes*

In the absence of store depletion, plasmalemmal Ca2+ permeability in resting muscle is very low, and its contribution in the maintenance of Ca2+ homeostasis at rest has not been studied in detail. Junctophilin 1 knock-out myotubes (JP1 KO) have a severe reduction in store-operated Ca2+ entry, presumably caused by physical alteration of the sarcoplasmic reticulum (SR) and T-tubule junction, leading to disruption of the SR signal sent by Stim1 to activate Orai1. Using JP1 KO myotubes as a model, we assessed the contribution of the Orai1-mediated Ca2+ entry pathway on overall Ca2+ homeostasis at rest with no store depletion. JP1 KO myotubes have decreased Ca2+ entry, [Ca2+]rest, and intracellular Ca2+ content compared with WT myotubes and unlike WT myotubes, are refractory to BTP2, a Ca2+ entry blocker. JP1 KO myotubes show down-regulation of Orai1 and Stim1 proteins, suggesting that this pathway may be important in the control of resting Ca2+ homeostasis. WT myotubes stably transduced with Orai1(E190Q) had similar alterations in their resting Ca2+ homeostasis as JP1 KO myotubes and were also unresponsive to BTP2. JP1 KO cells show decreased expression of TRPC1 and -3 but overexpress TRPC4 and -6; on the other hand, the TRPC expression profile in Orai1(E190Q) myotubes was comparable with WT. These data suggest that an important fraction of resting plasmalemmal Ca2+ permeability is mediated by the Orai1 pathway, which contributes to the control of [Ca2+]rest and resting Ca2+ stores and that this pathway is defective in JP1 KO myotubes.

tein of the plasma membrane (9 -11). Support for the interaction of Stim1 and Orai1 being one mechanism for SOCE is found from experiments that show that Orai1 knock-out/ knockdown or the expression of dominant negative forms of Orai1 almost completely prevents SOCE despite normal clustering of Stim1 (10 -12). Although it has been proposed that Orai1 by itself could be the Ca 2ϩ channel responsible for the SOCE current, the participation of canonical transient receptor potential (TRPC) channels alone or in concert with Orai1 in this process has also been suggested (13)(14)(15)(16). In immortal cell lines, it has been shown that an Orai1-dependent mechanism orchestrated with a Ca 2ϩ store sensor works as a feedback loop, which might contribute to keep [Ca 2ϩ ] rest and the SR Ca 2ϩ stores at normal levels, suggesting a link between SOCE and the control of the resting Ca 2ϩ homeostasis (17).
The membrane system of skeletal muscle is highly organized in triads where the junctional SR and the T-tubule membrane are only ϳ10 nm apart. The proximity of these two membranes is critical for the physical coupling between the dihydropyridine receptor and the ryanodine receptor 1 (RyR1) for EC coupling and likewise for the coupling of Stim1 and Orai1 for SOCE (18). Orai1 and several TRPC channels (TRPC1, -2, -3, -4 and -6) are expressed in skeletal muscle as well as in cultured myotubes (18 -20), and most of these channels have been localized in or near the triad junction. It is believed that junctophilins (JP1 and JP2) are the specialized proteins in the triad that maintain the proper spatial disposition of the triad membranes (21)(22)(23). Decreased expression of JP proteins by knockdown/knock-out strategies has been shown to cause altered triad formation, and this structural alteration is thought to be responsible for the decreased SOCE found in JP1 KO muscle cells (23). However, the mechanism by which the absence of JP1 affects SOCE other than the presumed association with this structural change has not been investigated.
In the current study, we explored the possible failure of Orai1-mediated sarcolemmal Ca ϩ2 entry at rest and the concomitant alterations in the resting Ca ϩ2 homeostasis in JP1 KO skeletal myotubes. Using the Ca 2ϩ entry blocker BTP2 and Orai1 pore mutant (E190Q), we demonstrated that an Orai1-mediated entry pathway at rest is blocked by BTP2 and that this pathway is absent in JP1 KO myotubes. In addition, we show that the Orai1-dependent R CaE is fundamental to maintain levels of physiological [Ca 2ϩ ] rest in the cytosol and to maintain the Ca 2ϩ content in intracellular stores.

EXPERIMENTAL PROCEDURES
Cell Culture-WT or JP1 KO myoblasts were plated either on 96-well imaging plates or 35-mm plates in Ham's F-10 medium (Invitrogen) supplemented with 20% bovine growth serum and 5 ng/ml basic fibroblast growth factor (Thermo Fisher Scientific). The following day, differentiation was started by changing the media to DMEM (Invitrogen) supplemented with 2% heat inactivated horse serum (Sigma).
Permanent Expression of Orai1 and Orai1(E190Q) in Myoblasts-Orai1 and Orai1(E190Q) cDNAs cloned into a retroviral expression vector with a bicistronic eGFP expression cassette and puromycin resistance were obtained from Addgene (Cambridge, MA, plasmids 12199 and 21662, respectively). These plasmids were kindly deposited by Dr. Anjana Rao (Harvard Medical School and Immune Disease Institute, Boston, MA) (11). We verified the correct sequence of both inserts and retroviral particles were packaged in HEK 293 helper cells, and myoblasts were infected with either construct at a multiplicity of infection of 5. Cells were allowed to recover for 12 h and then selected with puromycin (0.5 g/ ml) for 1 week. After the selection period, all remaining cells on the plate show the expression of the eGFP marker. An uninfected cell culture did not survive to the selection protocol, and wild type cells were used as controls.
Double-barreled Ca 2ϩ Microelectrodes and [Ca 2ϩ ] rest Measurements-Double-barreled Ca 2ϩ -selective microelectrodes were prepared using thin-walled borosilicate glass capillaries with filament (PB150F-4, World Precision Instruments, Sarasota, FL) as described previously (2). They were back-filled first with the neutral carrier ETH 129 (21193 Fluka-Sigma-Aldrich) and then with pCa 7 solution. Each Ca 2ϩ -selective microelectrode was individually calibrated as described previously (4), and only those with a linear relationship between pCa 3 and pCa 7 (Nernstian response, 28.5 mV per pCa unit) were used experimentally. After making measurements of resting [Ca 2ϩ ], all electrodes were then recalibrated, and if the two calibration curves did not agree within 3 mV, data from that microelectrode were discarded. Before starting the studies, we determined by direct calibration that the Ca 2ϩ sensitivity of Ca 2ϩ microelectrodes was not affected by any of the drugs used in the present study.
Microelectrode recordings were performed as described previously (2). Single myotubes were impaled with the double-barreled Ca 2ϩ -selective microelectrode, and the potentials were recorded via high impedance amplifier (Duo 773 electrometer, World Precision Instruments, Sarasota, FL). The potential from the 3 M KCl microelectrode (V m ) was subtracted electronically from the potential of the Ca 2ϩ electrode (V CaE ), to produce a differential Ca 2ϩ -specific potential (V Ca ) that represents [Ca 2ϩ ] rest . V m and V Ca were filtered  to improve the signal-to-noise ratio and stored in a computer for further analysis.
Store-operated Ca 2ϩ Entry Measurements-5 to 6 days after differentiation, the myotubes were loaded with 5 M Fura2-AM (Invitrogen) for 30 min at 37°C. After three washes with imaging solution (140 mM NaCl, 5 mM KCl, 1 mM MgCl 2 , 1 mM CaCl 2 , 5,5 mM glucose, 10 mM Hepes, pH 7.4), the cells were subjected to a Ca 2ϩ store depletion protocol, incubating the cells in zero Ca 2ϩ solution (140 mM NaCl, 5 mM KCl, 2 mM MgCl 2 , 5,5 mM glucose, 1 mM EGTA, 10 mM Hepes, pH 7.4) supplemented with 5 M thapsigargin. After 20 min, the cells were placed on the stage of a TE2000 epifluorescence microscope (Nikon) coupled to a digital acquisition system (Stanford Photonics, Stanford CA). The cells were illuminated at the isosbestic Fura2 excitation wavelength (360 nm), and the fluorescence was measured at 510 nm. Cells were then perfused with zero Ca 2ϩ solution and after 30 s switched to Mn 2ϩ -containing solution (140 mM NaCl, 5 mM KCl, 0.5 mM MnCl 2 , 5.5 mM glucose, 10 mM Hepes, pH 7.4), and the changes in Fura2 fluorescence were monitored.
Resting Ca 2ϩ Entry Measurements-Fura2-loaded myotubes were perfused with imaging solution for 1 min, and then the perfusion system switched to Mn 2ϩ -containing solution for 3 min. Some recordings showed motion artifact due to perfusion switching and to minimize errors in calculating the rate of decrease in Fura2 fluorescence after Mn 2ϩ exposure, the rate was measured when the signal was linear and stable (30 s after solution twitching). To calculate the fluorescence quench rate, the stable part of the signal was fitted to a linear regression (y ϭ a ϩ bx). The slope derived is expressed as fluorescence arbitrary units (f.a.u.) per second.
Sarcoplasmic Reticulum Loading Capacity-Myotube preparations were loaded with 5 M Fluo-4 AM for 30 min at 37°C. The myotubes were placed on the stage of a Nikon TE2000 epifluorescence microscope coupled to a digital acquisition system (Stanford Photonics). A filter set with an excitation of 480/30 nm and emission of 535/40 nm was used. The emission signal was acquired at a frequency of 30 frames per second. The amount of SR Ca 2ϩ was estimated by taking the combined area under the curve of three sequential exposures to 20 mM caffeine in Ca 2ϩ -free medium to prevent Ca 2ϩ entry to get the total area under the curve to assess the total amount of Ca 2ϩ released. To estimate the total intracellular Ca 2ϩ store, Fluo-4 AM-loaded myotubes were exposed to the Ca 2ϩ ionophore 4Br-A23187 in Ca 2ϩ -free medium. The area of the released Ca 2ϩ was computed for each myotube type studied.
Statistics-All values are expressed as mean Ϯ S.E., with the numbers in parentheses indicating the number of myotubes tested. Statistical analysis was performed using one-way analysis of variance and Tukey's test for multiple measurements to determine significance (p Ͻ 0.05).
After a depletion protocol with thapsigargin, WT myotubes showed robust Mn 2ϩ entry, which was strongly affected by 5 M BTP2 (Fig. 2, left). In contrast, JP1 KO myotubes showed markedly decreased Mn 2ϩ entry compared with WT myotubes; however, the small amount of Mn 2ϩ entry that remained could also be blocked by BTP2 (Fig. 2, right).    Fig. 3).
Orai1 and Stim1 Expression Are Decreased in JP1 KO Myotubes-We evaluated the expression of two key proteins involved in Ca 2ϩ entry, Stim1 and Orai1, in WT and JP1 KO myotubes. Western blot analysis shows that expression of both Orai1 and Stim1 are significantly decreased in JP1 KO cells (ϳ65 and ϳ60% decrease, respectively, n ϭ 5, p Ͻ 0.01), whereas the expression of JP2 remained unchanged (Fig. 5).

DISCUSSION
The mechanisms that regulate the intracellular Ca 2ϩ homeostasis in quiescent skeletal muscle are not fully understood. Under resting conditions, any overall passive inward flux from the extracellular space and leak from the SR is compensated by active efflux by the plasma membrane Ca 2ϩ ATPase and Na ϩ /Ca 2ϩ exchanger and Ca 2ϩ pumping to the SR by SERCA (Fig. 10, left). Maintaining this equilibrium is essential because chronic Ca 2ϩ overload would activate some of the intracellular proteases that can trigger apoptosis (25). On the other hand, the opposite scenario,

Orai1 Mediates Resting [Ca 2؉ ] and SR Loading
chronic Ca 2ϩ depletion, could compromise muscle performance.
Hypothetically, to keep the perfect balance between influx and efflux, this system needs mechanisms to sense and modulate fluxes as is reflected in the stability of the actual values of [Ca 2ϩ ] rest . However, the fact that a brief external Ca 2ϩ withdraw is sufficient to affect [Ca 2ϩ ] rest suggests that mechanisms (SERCA, Na ϩ /Ca 2ϩ exchanger, and plasma membrane Ca 2ϩ ATPase) that have been indicated to be the major contributors are insufficient to explain how this can happen. Thus, it is clear that in addition to what been previously believed, other rapidly responding "on-line" mechanisms must be operating continuously at rest to set the proper [Ca ϩ2 ] rest and the correct loading in the SR.
In this study, we addressed the participation of extracellular Ca 2ϩ influx at rest to maintain the proper [Ca 2ϩ ] rest in the  DECEMBER 10, 2010 • VOLUME 285 • NUMBER 50 cytosol and in the SR. We used myotubes that lack the expression of JP1 protein, which has been proposed to be responsible for the interaction between the membranes of the triad junction. If this is the case, the lack of JP1 expression could affect the correct communication between proteins present in the SR membrane and proteins in the T-tubule, and this failure is the presumed explanation for the defect in SOCE demonstrated here and elsewhere (23) in JP1 KO myotubes. Thus, it was not surprising that our results showed that a reduction in SOCE in JP1 myotubes by Ͼ60% compared with WT myotubes was accompanied by a reduction in [Ca 2ϩ ] rest .

Orai1 Mediates Resting [Ca 2؉ ] and SR Loading
To further characterize the contributors to the resting Ca 2ϩ entry pathway, WT and JP1 KO myotubes were pretreated with BTP2, a cation entry blocker previously shown to block SOCE in T cells, DT40, and HEK293 cells (26,27). The exact target of BTP2 is unclear, but it has been shown that it does not interfere with Ca 2ϩ ATPase-coupled pumps, mitochondrial Ca 2ϩ signaling, endoplasmic reticulum Ca 2ϩ release, and K ϩ channels in T cells (26). Incubation of myotubes with BTP2 blocks SOCE in WT cells and also eliminates the remaining SOCE in JP1-null myotubes.
In addition, it has been demonstrated that BTP2 blocks Sr 2ϩ entry mediated by TRPC3 expressed in HEK cells when induced by activation of phospholipase C-coupled receptors using carbachol (27). Phospholipase C activation generates diacylglycerol, which is a known agonist of TRPC3 and TRPC6 channels (28). In addition, it has been shown that BTP2 blocks TRPC3 channels activated by the diacylglycerol analog OAG (27). BTP2 has also been shown to block Sr 2ϩ entry through TRPC5 after carbachol stimulation (27), but as TRPC5 is not activated by diacylglycerol, this suggests that BTP2 is a general TRPC blocker regardless of the mechanism of activation.
The channels responsible of SOCE current in skeletal muscle are still controversial (18), but there is compelling data to suggest that Orai1/STIM1 are the major molecular components of SOCE in skeletal muscle (12,19). BTP2 is very effective (IC 50 ϭ 0.1-0.3 M) in blocking thapsigargin-induced Ca 2ϩ entry (27). In this case, the BTP2 target is unknown. The finding that BTP2 "selectively" blocks iCRAC (Ca 2ϩ releaseactivated Ca 2ϩ current) in lymphocytes (26) and that both BTP2 and expression of the E190Q dominant negative form of Orai1 have a similar effect on SOCE and R CaE in muscle cells strongly suggest that BTP2 acts on Orai1-mediated Ca 2ϩ entry either directly or indirectly through an Orai1-TRPC channel complex rather than through a TRPC channel mechanism alone. Further investigation will be required to directly address the precise mechanism of action of BTP2 on Orai1 channel function.
A different outcome was obtained when we measured Mn 2ϩ entry in cells that were not subjected to Ca 2ϩ store depletion where the rate of entry was at least 20 times smaller than cells with empty Ca 2ϩ stores. In WT myotubes, the estimated R CaE was decreased by 50% by BTP2, suggesting an overlapping mechanism between SOCE and R CaE . However, in JP1 KO myotubes, R CaE is similar to that seen in WT myotubes exposed to BTP2, and further exposure of JP1 KO myotubes to BTP2 has a negligible effect. These data show for the first time that in WT, at least two independent mechanisms work in tandem to control the fine tuning of R CaE : one that is sensitive to BTP2 and another that is refractory BTP2. On the other hand, in JP1 KO myotubes, the primary mechanism for Ca 2ϩ entry is via the BTP2-insensitive pathway, whereas the role of the BTP2-sensitive pathway is almost insignificant. In addition, the BTP2-sensitive Ca 2ϩ entry pathway is also responsible for fine tuning the Ca 2ϩ content of SR, as demonstrated by the fact that JP1 KO myotubes and WT myotubes

Orai1 Mediates Resting [Ca 2؉ ] and SR Loading
pretreated with BTP2 have less Ca 2ϩ in their stores than untreated WT myotubes.
Western blot analysis clearly shows a significant decrease of Orai1 and Stim1 protein levels in JP1 KO myotubes. This alteration itself could impair the regulatory feedback between Ca 2ϩ stores in the SR and the permeability of the plasma membrane to Ca 2ϩ both at rest and after store depletion. To test this hypothesis, we compared these results with those in cells that overexpressed the pore-defective E190Q Orai1 pro-tein as a dominant negative in WT myotubes. Orai1(E190Q) completely blocks SOCE in cultured myotubes as was previously shown by Lyfenko and Dirksen (12). In addition, Orai1(E190Q) decreases R CaE , the amount of Ca 2ϩ in internal stores and myoplasmic [Ca 2ϩ ] rest , mimicking the JP1 KO phenotype and that of WT myotubes after exposure to BTP2. More importantly, similar to what is seen in JP1 KO myotubes, BTP2 has no effect on reducing [Ca 2ϩ ] rest , maintenance of internal stores, and R CaE in Orai1(E190Q)-expressing myotubes. Thus, these data demonstrate that the BTP2-sensitive   TRPC4 showed two bands that may correspond to splicing forms described previously (31). In this case, we measured both bands together as an estimation of total TRPC4 expression. Densitometric quantification of Western blots against the indicated proteins are shown; each value was corrected with GAPDH band intensity as a loading correction, and the mean Ϯ S.E. of at least three independent experiments are shown as % of the WT intensity (right panel). The dashed line shows the level of expression of WT myotubes. DECEMBER 10, 2010 • VOLUME 285 • NUMBER 50 JOURNAL OF BIOLOGICAL CHEMISTRY 39177 aspect of R CaE at rest is dependant on Orai1 and suggests that the decreased expression of Orai1 in JP1 KO myotubes or blocking the Orai1-dependent fraction of the R CaE by BTP2 in WT myotubes leads to altered resting Ca 2ϩ homeostasis.

Orai1 Mediates Resting [Ca 2؉ ] and SR Loading
Since TRPCs are known targets of BTP2, the alternative hypothesis was that the unresponsiveness of JP1 KO as well as Orai1(E190Q) myotubes to BTP2 treatment was due to altered expression of TRPCs. Although TRPC1 and TRPC3 expression is reduced in JP1 KO myotubes compared with WT cells, the expression of TRPC4 and TRPC6 (which are similar in structure an function to TRPC1 and TRPC3, respectively (29,30)) are compensatorily increased, suggesting that low expression of TRPC1 and TRPC3 is not the primary cause of BTP2 unresponsiveness. Furthermore, the fact that Orai1(E190Q)-expressing myotubes are refractory to BTP2 but have a TRPC expression profile that is almost identical to WT cells suggests that changes in TRPC expression cannot account for the alterations in resting Ca 2ϩ homeostasis that we observed.
In a previous report (23), it was shown that in muscle fibers in which adenoviral-shRNA was used to knock down both JP1 and JP2 decreases both SOCE and SR Ca 2ϩ content in agreement with this work. However, these results differ with our study in that a slight increase in the basal Fura2 F340/F360 ratio in muscle cells treated with shRNA compared with controls was found. At first, this last result could appear to be contradictory with the present work, but these results are not comparable because JP2 expression was not altered in the current study. The simultaneous knockdown of JP1 and JP2 proteins may cause additional alterations in Ca 2ϩ -handling proteins that could shift the [Ca 2ϩ ] equilibrium at rest to higher levels in the cytosol, which could explain the previously observed small increase in resting Ca 2ϩ . In addition, in the present study, calibrated microelectrodes were used, which is a much more precise and exact method to determine resting [Ca 2ϩ ]. These electrodes allow discrimination of differences in [Ca 2ϩ ] of Ͻ5 nM in the linear portion of the curve response (pCa3-7), which gives further confidence that the differences that we observed are correct.
In summary, our data show that under resting conditions at least two Ca 2ϩ entry pathways are present in myotubes. One of these pathways involves the participation of Orai1, which is sensitive to BTP2. The ablation of Orai1-mediated R CaE results in an impaired control of the Ca ϩ2 homeostasis in myotubes, resulting in a decrease in [Ca 2ϩ ] rest and SR Ca 2ϩ content (Fig. 10, right panel).