Sarcoplasmic reticulum Ca²⁺ permeation explored from the lumen side in mdx muscle fibers under voltage control

Duchenne muscular dystrophy is a very severe muscle disease that is characterized by progressive skeletal muscle wasting. Duchenne muscular dystrophy is provoked by mutations in the gene encoding the protein dystrophin, which lead to the total absence of this protein in skeletal muscles. In normal skeletal muscle, dystrophin is located underneath the sarcolemma, and interacts with the F-actin component of the intracellular cytoskeleton at its N-terminal extremity and with a sarcolemmal-embedded glycoprotein complex at its C-terminal extremity, which itself is associated with the extracellular matrix (Blake et al., 2002). Lack of dystrophin is assumed to destabilize this architecture and to promote disruption of the linkage between the subsarcolemmal cytoskeleton and the extracellular matrix, but the functional consequences of the absence of dystrophin that contribute to muscle degeneration still remain elusive. Mainly with the help of the mdx mouse model, which also lacks dystrophin, several studies have nevertheless put forward the idea that degeneration of dystrophin-deficient skeletal muscle may result from a chronic intracellular Ca²⁺ overload that initiates massive protein degradation (Mallouk et al., 2000; Gailly, 2002; Ruegg et al., 2002; Allen et al., 2010). Several lines of evidence support the notion that this Ca²⁺ overload is the consequence of a chronic and exacerbated sarcolemmal Ca²⁺ influx. Initially, this Ca²⁺ influx was described to occur through spontaneously active leaky channels or through mechano-gated channels that become overactive in the absence of dystrophin (Fong et al., 1990; Franco and Lansman, 1990; Allard, 2006). More recently, up-regulated store-operated Ca²⁺ entry (SOCE) has been proposed to correspond to the Ca²⁺ influx pathway that contributes to detrimental Ca²⁺ excess in dystrophic muscle fibers (Vandebrouck et al., 2002; Boittin et al., 2006; Edwards et al., 2010). SOCE is thought to be triggered by Ca²⁺ depletion within the SR so that up-regulation of SOCE in dystrophic muscle implies that either SOCE is hyperactive or hypersensitive to SR depletion or that SR depletion is more pronounced in dystrophin-deficient muscle. In support of the first possibility, Orai1 associated to stromal interacting molecule 1 (STIM1) and the transient receptor potential canonical 1 (TRPC1), two candidate molecules that have been proposed to support SOCE, were found to be overexpressed in mdx muscle fibers (Gervásio et al., 2008; Edwards et al., 2010). Possible reduced SR Ca²⁺ content provoked either by an enhanced SR Ca²⁺ leak or by a decreased SR Ca²⁺ filling process has also been investigated in dystrophic muscle, but the results obtained were contradictory. Using chemically skinned muscle fibers, Takagi et al. (1992) first reported an increased SR Ca²⁺ leak in mdx muscle with no change in SR Ca²⁺ uptake, whereas Divet and Huchet-Cadiou (2002) described a reduced SR Ca²⁺ uptake and an increased SR Ca²⁺ leak in mdx muscle. Using mechanically skinned fibers, Plant and Lynch (2003) reported no difference in SR Ca²⁺ reloading and in the SR Ca²⁺ leak between control and mdx muscle fibers. However, because of the loss of intracellular components, skinned fibers do not reproduce the native intracellular environment of the SR so that leakage could be distorted. Intracellular Ca²⁺ sparks at rest, which are thought to reflect a resting SR Ca²⁺ leak, have also been measured in intact and in permeabilized muscle fibers from control and mdx mice. Wang et al. (2005) showed that osmotic shock induced irreversible intracellular Ca²⁺ spark activity in intact muscle fibers from mdx mice, likely to cause an enhanced SR Ca²⁺ leak, and more recently the frequency of spontaneous Ca²⁺ sparks in permeabilized fibers was found to be significantly higher in mdx as compared with control muscle (Bellinger et al., 2009). Nevertheless these experiments were performed in the absence of voltage control. Yet it has been demonstrated that the dihydropyridine receptor (DHPR), whose conformation changes elicited by depolarization promote the opening of the SR Ca²⁺ release channel, regulates the SR passive leak (Eltit et al., 2011). These data suggest that muscle membrane potential may control not only SR active Ca²⁺ release but also SR passive Ca²⁺ leak, further emphasizing the need to control membrane potential to carefully investigate the SR Ca²⁺ permeation process.

In this paper, we investigate SR Ca²⁺ permeation at resting potential in intact voltage-controlled skeletal muscle fibers from control and mdx mice. Using cytosolic and intraluminal SR Ca²⁺ dyes, we demonstrate that the passive SR Ca²⁺ leak is significantly higher in mdx fibers and is associated with an elevated SR Ca²⁺ reuptake activity and a likely up-regulation of SOCE. Our study also shows that SR Ca²⁺ depletion induced by depolarizing pulses of controlled amplitude is not altered in dystrophin-deficient fibers.

All experiments were performed in accordance with the guidelines of the French Ministry of Agriculture (87/848) and of the European Community (86/609/EEC). Interosseal and fdb (flexor digitorum brevis) muscles were removed from 1–2-mo-old male C57BL6 or OF1 (as indicated in the figure legends) and mdx-5cv mice. Single skeletal muscle fibers were isolated by a 50-min enzymatic treatment at 37°C using a Tyrode solution containing 2 mg/ml collagenase type I (Sigma-Aldrich).

The major part of a single fiber was electrically insulated with silicone grease, and a micropipette was inserted into the fiber through the silicone layer to current clamp or voltage clamp the portion of the fiber free of grease (50–100 µm in length) using a patch-clamp amplifier (RK-400; Bio-Logic) in a whole-cell configuration (Pouvreau et al., 2007). Command voltage or current pulse generation and data acquisition were done using the pClamp10 software (Axon Instruments) driving an A/D converter (Digidata 1400A; Axon Instruments). Analogue compensation was systematically used to decrease the effective series resistance. Currents or membrane potentials were filtered at 3 kHz and acquired at a sampling frequency of 10 kHz. The tip of the micropipette was then crushed into the dish bottom to allow intracellular dialysis of the fiber with the intra-pipette solution (see Solutions; Berbey and Allard, 2009).

Cells were dialyzed through the micropipette during 20 min using an intracellular solution containing 100 µM Fura-2. Fura-2 was excited at 340 and 380 nm, and fluorescence was measured between 520 and 560 nm on images from a region of 20 µm diameter on the silicone-free extremity of the cell. Images were captured with a charge-coupled device camera (Roper Scientific) driven by the Metavue software and using a 40× oil immersion objective. After background subtraction, the ratio F340/F380 was calculated to estimate [Ca²⁺]. Images were captured at a frequency of 0.5 Hz to limit photobleaching.

Subcellular distribution of Fluo-5N was observed with a confocal microscope (LSM 5; Carl Zeiss) using a 63× oil immersion objective lens.

Fibers were incubated in a Tyrode solution containing 10 µM Fluo-5N in the acetoxymethyl ester form at room temperature for 2 h to allow entry of the dye into the SR, and subsequent de-esterification. Cells were then dialyzed for 20 min through the micropipette used for voltage clamping with an internal solution containing 50 mM EGTA (see Solutions) to prevent contraction during imaging and avoid possible contribution of cytosolic Ca²⁺ changes to SR Ca²⁺ signals. Fluo-5N was excited at 488 nm, and the emitted fluorescence was measured at 520 nm after background subtraction. Changes in SR Ca²⁺ content were expressed as F/F₀, where F₀ is the measured fluorescence at the steady state at the beginning of the experiment. The frequency of images capture ranged between 0.2 and 25 Hz depending on the recording duration. For each experiment, this frequency is indicated in the figure legends.

For experiments using Fura-2 and experiments using Fluo-5N under current clamp conditions, the Tyrode external saline contained 140 mM NaCl, 5 mM KCl, 2.5 mM CaCl₂ (or 0, compensated by 2.5 mM MgCl₂), 2 mM MgCl₂, and 10 mM Hepes adjusted to pH 7.2 with NaOH. For experiments using Fluo-5N under voltage clamp conditions, the external solution contained 140 mM TEA-methanesulfonate, 2.5 mM CaCl₂, 2 mM MgCl₂, and 10 mM Hepes adjusted to pH 7.2 with TEA-OH. For experiments using Fura-2, the intrapipette solution contained 120 mM potassium glutamate, 5 mM Na₂-ATP, 5 mM Na₂-phosphocreatine, 5.5 mM MgCl₂, 5 mM glucose, and 5 mM Hepes adjusted to pH 7.2 with KOH. For experiments using Fluo-5N, the intrapipette solution contained 50 mM EGTA-KOH, 50 mM potassium glutamate, 5 mM Na₂-ATP, 5 mM Na₂-phosphocreatine, 5.5 mM MgCl₂, 5 mM glucose, and 5 mM Hepes adjusted to pH 7.2 with KOH. Stock solution of Fura-2 was dissolved in distilled water at 10 mM. Stock solutions of Fluo-5N AM, 4-chloro-m-cresol (CmC), and cyclopiazonic acid (CPA) were dissolved in dimethylsulfoxide at 1, 1,000, and 50 mM, respectively. Cells were exposed to different solutions by placing them in the mouth of a perfusion tube from which flowed by gravity the rapidly exchanged solutions. Experiments were performed at room temperature.

Fits were performed with Microcal Origin software (GE Healthcare). Data are given as means ± SEM and compared using different tests indicated in the figure legends or in the text. Differences were considered significant when P < 0.05. In the figures, *, P < 0.05; **, P < 0.005; ***, P < 0.0005.

Fig. S1 depicts the intracellular distribution of Fluo-5N in a control fiber using confocal imaging and the corresponding average fluorescence intensity profile from a small region of this fiber.

In a previous paper, we showed that pharmacological inhibition of the SR Ca²⁺-ATPases by CPA induced a progressive increase in cytosolic Ca²⁺ at resting potential that mainly results from the passive Ca²⁺ leak from the SR (Berbey et al., 2009). We used this pharmacological tool here to determine if the magnitude of the SR Ca²⁺ leak was different in control and in mdx muscle fibers. Fig. 1 A shows that CPA poisoning led to an increase in intracellular Ca²⁺ monitored by Fura-2 at a higher rate in an mdx fiber as compared with a control one at a holding potential of −80 mV. Fitting a linear regression to the fluorescence records 1 min after CPA addition in every cell tested indicated that the mean rate of fluorescence increase was significantly higher in mdx (0.51 ± 0.07 ratio [F340/F380]/min) as compared with control fibers (0.11 ± 0.01 ratio [F340/F380]/min; Fig. 1 B), which suggests that SR Ca²⁺ leak is higher in dystrophin-deficient fibers. However, it cannot be excluded that an influx of external Ca²⁺ contributed to the measured Ca²⁺ changes, as the Ca²⁺ leak from the SR, no longer compensated for by Ca²⁺ uptake, may be responsible for a substantial depletion of Ca²⁺ in the SR that in turn may activate SOCE. Thus, in order to exclude the contribution of SOCE to CPA-induced cytosolic Ca²⁺ increase and focus on the SR Ca²⁺ leak, fibers were bathed in a Ca²⁺-free external solution before the addition of CPA. Unexpectedly, under these conditions of continuous exposure of fibers to the Ca²⁺-free medium, the rate of CPA-induced Ca²⁺ increase was significantly augmented in control fibers, whereas it was significantly decreased in mdx fibers (Fig. 1 B). There was also no significant difference in the rate of CPA-induced Ca²⁺ increase between control and mdx fibers in the Ca²⁺-free medium. Collectively, these data suggest that removal of external Ca²⁺ has important effects on intracellular Ca²⁺ handling that are not simply limited to suppression of SOCE, and provokes an increase in the SR Ca²⁺ leak (see results in Fig. 6 A). To further investigate the SR Ca²⁺ leak in control and in mdx fibers, and to circumvent the use of Ca²⁺-free solution for preventing the possible involvement of SOCE, we next investigated SR Ca²⁺ permeation by measuring the changes in intraluminal Ca²⁺ using a low-affinity Ca²⁺ dye loaded in the SR.

After loading of the dye, confocal images revealed a striated pattern of Fluo-5N fluorescence (Fig. S1). A fluorescence profile of a selected region revealed the presence of doublets consistent with accumulation of the de-esterified form of the dye mainly at the level of the junctional SR (Fig. S1). The presence of the dye in the SR lumen was confirmed by measuring the changes in Fluo-5N fluorescence induced by various stimuli of SR Ca²⁺ release in the presence of a high intracellular concentration of EGTA. Fig. 2 A shows that a train of action potentials delivered at 50 Hz during 1 s under the current clamp condition in the presence of Tyrode evoked a marked decrease in Fluo5N fluorescence. Fluorescence had recovered 2 min later, and exposure of the same fiber to 1 mM of the ryanodine receptor agonist CmC at −80 mV gave rise to a decrease in Fluo5N fluorescence of a similar amplitude, which was partially reversible upon washout of the drug (Fig. 2 B).

The Fluo5N signals were explored in more detail under voltage clamp conditions in response to depolarizing pulses of increasing amplitudes in control fibers from OF1 mice. Fig. 3 A shows that depolarizing pulses >−40 mV evoked a drop in Fluo-5N fluorescence, whose magnitude and kinetics increased with the amplitude of the depolarization. For pulses >−20 mV, the decrease in fluorescence was associated with the activation of inward membrane currents that shared all the characteristics of voltage-activated L-type Ca²⁺ currents, routinely recorded in this preparation under similar experimental conditions. The best fit to the mean decrease in Fluo-5N fluorescence as a function of voltage using a Boltzmann equation indicated a V_1/2 of −23mV, a steepness factor of 3 mV, and a maximal decrease of fluorescence of 40% (Fig. 3 B). The magnitude of the decrease in Fluo5N fluorescence induced by trains of action potentials (52 ± 8%, n = 5), CmC (45 ± 6%, n = 5), and a single voltage pulse to +20 mV (42 ± 4%, n = 12) were found to not be significantly different (Student’s unpaired t test).

As already illustrated in Fig. 2, Fig. 3 A shows that, despite the presence of a high intracellular concentration of EGTA, Fluo-5N fluorescence recovery occurred during the 2-min intervals that separated voltage pulses. This recovery phase that followed voltage-activated Fluo-5N fluorescence decrease could be totally inhibited by the presence of the SR Ca²⁺-ATPase inhibitor CPA. At a holding potential of −80 mV, exposing the fiber to CPA induced a progressive decrease in Fluo-5N fluorescence that likely revealed an SR passive Ca²⁺ leak (see also Fig. 6), no longer balanced by the activity of Ca²⁺-ATPases (Fig. 4 A). A depolarizing pulse to 0 mV evoked an abrupt decrease in Fluo-5N fluorescence that was maintained during the 50-s duration pulse but also upon repolarization of the fiber. Washout of CPA led to a progressive return of Fluo-5N fluorescence toward its initial level. Collectively, these results indicate that the changes in Fluo-5N fluorescence elicited by depolarizing pulses mainly report decreases in luminal [Ca²⁺], which results from voltage-activated SR Ca²⁺ release and subsequent depletion. To further examine the recovery phase of Fluo-5N fluorescence after voltage-activated Ca²⁺ release, depolarizing pulses of increasing duration were applied on the same fiber. As illustrated in Fig. 4 B, a first pulse of 1-s duration to +30 mV induced a rapid decrease of Fluo-5N fluorescence, which remained stable during depolarization and which was then followed upon repolarization at −80 mV by a fast recovery phase (τ = 1.4 s in this cell). 2 min later, Fluo-5N fluorescence had returned to the initial level, and a second, longer pulse of 5-s duration gave rise to a comparable initial drop in fluorescence, but this time followed, after a 1-s stabilization, by a recovery phase approximately five times slower during depolarization (τ = 4.8 s) than during repolarization. Upon repolarization to −80 mV, Fluo-5N fluorescence rapidly recovered with a kinetic measurement similar to the one measured after the first 1 s-pulse (τ = 1 s). Finally a third pulse maintained during 20 s gave rise to a change in Fluo-5N fluorescence of similar amplitude and kinetics compared with the one measured during the 5-s pulse, except that repolarization after a 20-s depolarization did not induce any change in the recovery phase kinetic. Similar results were obtained in two other cells. These data likely indicate that depolarization leads to SR depletion that is maintained as long as SR Ca²⁺ release channels remain open, and slowly recovers because of the progressive closure of Ca²⁺ release channels induced by inactivation of excitation–contraction coupling together with SR Ca²⁺ refilling. Upon repolarization or after total inactivation of the Ca²⁺ release channels, the fast recovery phase mainly resulted from the Ca²⁺ reloading process.

A set of experiments next aimed at comparing the magnitude, the voltage dependence, and the kinetics of the recovery phase of Ca²⁺ depletion in the SR lumen induced by depolarizing pulses in control (C57BL6 mice) and in mdx fibers. Fig. 5 (A and B) shows that the mean maximal amplitude of the voltage-induced Fluo-5N fluorescence decrease (0.29 ± 0.03 ΔF/F₀ in control versus 0.31 ± 0.03 ΔF/F₀ in mdx fibers), the V_1/2 (−22 ± 2.5 mV in control vs. −24 ± 2 mV in mdx fibers), and the steepness factor (2.1 ± 0.1 mV in control vs. 2.3 ± 0.4 mV in mdx fibers) of the voltage dependence of the depolarization-induced Fluo-5N fluorescence decrease were not significantly different in control and in mdx fibers. However, as clearly illustrated in the mean Ca²⁺ signals elicited by voltage pulses in Fig. 5 A, the recovery phase of fluorescence that followed voltage pulses was faster in mdx fibers than in control fibers. Because some cells displayed a very slow recovery phase that precluded any satisfactory fitting of the recovery time course with an exponential function, we chose to measure the fraction of fluorescence that had recovered 1 s after the end of the pulse in every cell depolarized between +10 and +40 mV. On average, the fraction of fluorescence recovery was quite significantly higher in mdx (26 ± 3%, n = 48) as compared with control fibers (10 ± 2%, n = 44), which suggests that SR Ca²⁺ reuptake subsequent to voltage-induced SR depletion is more efficient in dystrophin-deficient fibers.

Measurements of cytosolic Ca²⁺ changes in response to CPA poisoning in Fig. 1 indicated that the rate of CPA-induced Ca²⁺ increase was augmented in control fibers in the absence of external Ca²⁺, which suggests an increased SR Ca²⁺ leak provoked by the absence of external Ca²⁺. To test if a potentiating effect of a Ca²⁺-free external solution could be revealed on the SR Ca²⁺ leak, control fibers were loaded with Fluo-5N, voltage-clamped at −80 mV, and exposed to CPA in the presence and in the absence of external Ca²⁺. Fig. 6 A shows that in the presence of external Ca²⁺, CPA induced a progressive decrease in Fluo-5N fluorescence as already presented in Fig. 4 A, the rate of which increased upon removal of external Ca²⁺. Fitting a linear regression to the fluorescence decline in every cell tested indicated that the mean rate of CPA-induced fluorescence decrease was significantly higher in the absence of external Ca²⁺ (0.056 ± 0.0117 [F/F₀]min) than in its presence (0.034 ± 0.0057 [F/F₀]/min), which indicates a potentiation of the SR Ca²⁺ leak upon removal of external Ca²⁺ (Fig. 6 B). The rate of CPA-induced Fluo-5N fluorescence decrease was then compared in control and in mdx fibers in the presence of external Ca²⁺ (Fig. 6 C). As illustrated in Fig. 6 D, the rate of CPA-induced Fluo-5N fluorescence decrease was found to be significantly doubled in mdx as compared with control fibers, which suggests that the passive SR Ca²⁺ leak is higher in dystrophin-deficient fibers.

In this paper, we investigated the SR Ca²⁺ leak at resting membrane potential in dystrophin-deficient fibers. Up to now, the SR leak had been explored in isolated SR vesicles, skinned fibers, or intact fibers, though in the absence of voltage control. The use of intact fibers under acute voltage control in the present study allowed us to avoid possible distortion provoked by the loss of intracellular components in isolated SR vesicles or skinned fibers and by uncontrolled voltage-induced SR Ca²⁺ fluxes. Voltage control also offers the advantage of being able to monitor the background membrane current, which, when stabilized to a low level, represents a reliable index of fiber integrity. At a resting state, the Ca²⁺ leak is constantly balanced by the influx of Ca²⁺ into the SR created by the pumps so that inhibition of the SR Ca²⁺-ATPases reveals, at least in part, the leaky efflux of Ca²⁺ from the SR. Our first set of measurements of Ca²⁺ changes in the cytosol indicated that inhibition of the SR pumps by CPA led to an increase in cytosolic [Ca²⁺] at a very significant higher rate in mdx fibers in the presence of physiological external [Ca²⁺]. However, under these experimental conditions, the SR Ca²⁺ leak induced by CPA may progressively lead to SR Ca²⁺ depletion, which in turn may provoke SOCE activation. The cytosolic Ca²⁺ increase induced by CPA may then result both from Ca²⁺ efflux from the SR and from Ca²⁺ influx through SOCE. However, when cells were bathed in a Ca²⁺-free external solution to prevent any SOCE contribution to the cytosolic Ca²⁺ changes induced by CPA, the rate of CPA-induced Ca²⁺ increase was augmented in control fibers and decreased in mdx fibers. This result suggested that SR Ca²⁺ leak is potentiated in the absence of external Ca²⁺ in control fibers, giving rise to a rate of increase in cytosolic Ca²⁺ that is higher than the one measured in the presence of external Ca²⁺; i.e., under experimental conditions that allow SOCE to occur. In mdx fibers, the observed decrease of the rate of CPA-induced Ca²⁺ increase in the absence of extracellular Ca²⁺, which did not became significantly different from the rate measured in control fibers in Ca²⁺-free solution, gives evidence that the higher rate of CPA-induced Ca²⁺ increase in mdx fibers in the presence of physiological Ca²⁺ is principally caused by an elevated sarcolemmal influx, likely via SOCE. This interpretation is consistent with the recent data obtained by Edwards et al. (2010) in skinned muscle fibers, which provides evidence for an up-regulation of SOCE in mdx muscle. Nevertheless, this series of cytosolic Ca²⁺ measurements on its own did not allow for a straightforward conclusion about the respective contribution of the SR leak and sarcolemmal Ca²⁺ influx in the elevated rate of Ca²⁺ increase in response to CPA poisoning in mdx muscle fibers.

To circumvent limitations associated with the use of Ca²⁺-free external solution, we used a low-affinity Ca²⁺ indicator, Fluo5N, which is able to report Ca²⁺ changes in the SR lumen. Kabbara and Allen (2001) were the first to use Fluo5N to measure SR calcium in muscle fibers of the cane toad, and more recently, Ziman et al. (2010) successfully used this Ca²⁺ dye in mouse skeletal muscle fibers. We obtained here similar confocal images of subcellular distribution of Fluo5N fluorescence as Ziman et al. (2010), which indicates the presence of the indicator mainly in the junctional SR. As mentioned by these authors, after loading, the dye may also have accumulated in the cytosol and in other organelles such as mitochondria. However, the very high internal concentration of EGTA that we used should considerably minimize the possible contamination of the recorded Fluo5N fluorescence by signals originating from non-SR compartments. Overall, we showed that stimuli eliciting SR Ca²⁺ release, such as trains of action potentials, addition of a potent agonist of ryanodine receptors, and depolarization pulses all evoked a transient decrease in Fluo5N fluorescence, demonstrating that Fluo5N mainly reports Ca²⁺ changes in the SR lumen. Although the sampling rate of our fluorescence imaging (25 Hz for the highest) did not allow us to acutely analyze the kinetics of the Fluo5N signal, the profile of our Fluo5N responses induced by 1-s depolarizations were very comparable to the profile of responses of the Ca²⁺ sensor cameleon D4cpv targeted to the SR, recently described by Sztretye et al. (2011), exhibiting a rapid phase of decrease upon depolarization followed by a plateau phase and then a slow recovery phase upon repolarization. The fact that our response profiles matched the ones obtained with cameleon proteins that specifically express in the SR further demonstrates that our Fluo5N signals originate from the SR lumen.

Despite the presence of a high concentration of Ca²⁺ buffer in the cytosol, we also observed that, after depolarization-induced depletion, the SR Ca²⁺ level returned to close to the initial value during the 2-min intervals that separated pulses, and this recovery phase was inhibited by the SR Ca²⁺ pump blocker CPA. Although this recovery phase takes place on a nonphysiological time scale because of the high concentration of EGTA, these data indicate that SR Ca²⁺-ATPases have the capacity to actively reuptake Ca²⁺ into the SR by overcoming the Ca²⁺-chelating action of the buffer.

Acute control of the fiber membrane potential allowed us to investigate the voltage dependence of the depolarization-induced SR Ca²⁺ changes. We obtained voltage values of half activation ranging −23 to −22 mV and a steepness factor ranging 2 to 3 mV in the series of experiments performed in fibers from three different mouse lineages. These values are very close to the corresponding values obtained for the voltage dependence of Ca²⁺ release measured with a cytosolic Ca²⁺ indicator and the same silicone clamp technique (Pouvreau et al., 2006). Using the cameleon D1ER to monitor SR Ca²⁺ changes, Jiménez-Moreno et al. (2010) obtained more depolarized values of half activation (∼−15 mV) and less steep voltage dependence (close to 6 mV). However, they also obtained comparable fit parameters for the voltage dependence of Ca²⁺ release measured with a cytosolic Ca²⁺ dye. The use of a different voltage clamp technique in this study, the whole-cell configuration of the patch clamp technique on whole fdb adult muscle fibers, could explain the different values reported for SR Ca²⁺ changes as compared with ours.

Our measurements of voltage-induced Fluo5N decrease performed on mdx and C57BL6 control fibers indicated that neither the magnitude of the Ca²⁺ depletion nor its voltage dependence were significantly changed in dystrophin-deficient fibers. This result is in concordance with several studies that reported no significant difference in the amplitude (Turner et al., 1988; Tutdibi et al., 1999) and in the voltage dependence (Collet et al., 1999) of cytosolic Ca²⁺ transients recorded from normal and mdx muscles but contrasts with the observations of DiFranco et al. (2008) and Hollingworth et al. (2008), who found a significant reduction of the amplitude of the action potential–evoked Ca²⁺ transients in mdx mice. However, the low sampling rate of our SR Ca²⁺ measurements did not allow us to explore the Ca²⁺ signal during the first 50 ms, so we cannot exclude the possibility that a transient phase of depletion that may precede the plateau phase could be altered in mdx muscle.

Upon repolarization to −80 mV, we observed a quite significantly faster rate of Fluo5N recovery in mdx fibers, which suggests a higher reuptake efficiency of SR Ca²⁺-ATPases. In the literature, either no change or a reduced velocity of Ca²⁺ reuptake has been described in mdx muscle (Turner et al., 1988; Takagi et al., 1992; Kargacin and Kargacin, 1996; Collet et al., 1999; Divet and Huchet-Cadiou, 2002; Plant and Lynch, 2003; Woods et al., 2004). Different experimental conditions might explain the different results obtained. To our knowledge, our study is indeed the first to explore the SR Ca²⁺ reuptake process from the SR lumen side under voltage control in intact fibers. These conditions offer the advantage of preserving the native cellular environment of the SR and avoiding the possible involvement of cytosolic Ca²⁺ binding sites that may compete with cytosolic Ca²⁺ dyes in a different manner in control and mdx muscles. The higher reloading efficiency of SR Ca²⁺-ATPases in dystrophin-deficient fibers may be seen as a compensatory mechanism in response to chronic Ca²⁺ overload and an elevated SR Ca²⁺ leak, as further suggested by our results.

Measurement of Fluo5N fluorescence in the presence of CPA led us to directly monitor the Ca²⁺ leak from the SR lumen at resting membrane potential. We first confirmed that the SR Ca²⁺ leak revealed by CPA poisoning of SR Ca²⁺-ATPases is significantly increased in the absence of external Ca²⁺, as suggested by experiments performed using Fura-2. The mechanisms involved in the up-regulation of the SR Ca²⁺ leak in control fibers after external Ca²⁺ removal are not clear. Eltit et al. (2011) proposed that the DHPR regulates the SR passive leak at a resting state. It thus can be postulated that removal of external Ca²⁺ may induce changes in the configuration of the DHPR that could weaken the repressive action exerted by its presence on the leak state of the ryanodine receptor, resulting in an elevated leak. Overall, we demonstrated that CPA-induced SR Ca²⁺ leak occurred at a significant faster rate in mdx fibers together with an elevated sarcolemmal influx. This result is in agreement with results obtained in skinned fibers (Takagi et al., 1992; Divet and Huchet-Cadiou, 2002) but also in intact and permeabilized fibers from mdx muscles in which a higher frequency of osmotic shock-induced and spontaneous Ca²⁺ sparks was, respectively, reported (Wang et al., 2005; Bellinger et al., 2009). The SR Ca²⁺ leak has been postulated to occur through ryanodine receptors in the absence of releasing signals, but other pathways could be potentially involved (Camello et al., 2002; Guerrero-Hernandez et al., 2010). Among these pathways, the TRPC1 channel is a potential candidate because this channel is overexpressed in mdx muscle (Gervásio et al., 2008), and we demonstrated that it operates as a SR Ca²⁺ leak channel in muscle fibers overexpressing TRPC1 (Berbey et al., 2009).

In conclusion, this paper demonstrates that the magnitude and the voltage dependence of depolarization-activated SR Ca²⁺ depletion are not modified in mdx muscle but that the reuptake activity of SR Ca²⁺-ATPases is elevated. More importantly, this study also shows straightforwardly that the passive SR Ca²⁺ leak is higher in mdx muscle at resting membrane potential. A higher SR Ca²⁺ leak has important physiopathological consequences in dystrophin-deficient muscle that could be modeled as follows on the basis of our data (Fig. 7). The SR Ca²⁺ content is conditioned by the rate of the leak and by the rate of the reuptake that operates in the SR membrane, whereas, in the long term, cytosolic [Ca²⁺] is strictly dependent on sarcolemmal influx and efflux (Ríos, 2010). In control fibers, it can be assumed that SR leak and reuptake and sarcolemmal influx and efflux are balanced so that cytosolic [Ca²⁺] remains in the physiological resting range. In mdx fibers, the elevated SR leak that we measured may be no more balanced by reuptake, although this uptake was increased, probably by compensatory mechanisms. This unbalanced elevated leak may result in a chronic SR depletion that in turn overactivates sarcolemmal Ca²⁺ influx via SOCE, the activation threshold of which is, moreover, reduced (Edwards et al., 2010). It also cannot be excluded that an elevated SR leak could directly influence sarcolemmal Ca²⁺ influx independently of the SR Ca²⁺ content and SOCE. Eventually, the elevated sarcolemmal Ca²⁺ influx that transiently dominates over the efflux brings cytosolic [Ca²⁺] to a new sustained elevated set point in mdx fibers. Elevated sarcolemmal influx and efflux then again balance each other, maintaining chronic Ca²⁺ overload in mdx fibers.

We thank Vincent Jacquemond and Maelle Jospin for critical comments on the manuscript.

This work was supported by grants from the Université Lyon 1, the Centre National de la Recherche Scientifique, and the Association Française contre les Myopathies.

Richard L. Moss served as editor.