Reduced voltage-activated Ca²⁺ release flux in muscle fibers from a rat model of Duchenne dystrophy

A number of mechanisms involved in the control of Ca²⁺ homeostasis and excitation–contraction (EC) coupling have been reported altered in dystrophin-deficient muscle cells. This is now to the point that emphasizing abnormal or defective Ca²⁺ homeostasis in the associated Duchenne muscular dystrophy (DMD) disease situation has become a common place, with cytosolic Ca²⁺ overload being one frequently assumed pathogenic driver of multiple detrimental effects. Specifically, altered Ca²⁺ fluxes at the plasma/t-tubule membrane level have been reported to be carried by (1) membrane tearing (Mokri and Engel, 1975; Pestronk et al., 1982; Petrof et al., 1993; Pasternak et al., 1995), (2) TRPC or other types of sarcolemmal Ca²⁺ leak channels (Fong et al., 1990; Franco and Lansman, 1990; Mallouk and Allard, 2002; Vandebrouck et al., 2002; Ducret et al., 2006; Gervásio et al., 2008; Krüger et al., 2008; Teichmann et al., 2008; Millay et al., 2009; Matsumura et al., 2011; Lin et al., 2022), (3) the store-operated Ca²⁺ entry complex (Boittin et al., 2006; Edwards et al., 2010; Cully et al., 2012; Zhao et al., 2012; García-Castañeda et al., 2022), (4) connexin hemichannels (Cea et al., 2016; Nouet et al., 2020), and (5) P2X purinoreceptors (Jiang et al., 2005; Yeung et al., 2006); in intracellular compartments by sarcoplasmic reticulum (SR) ryanodine receptors (Woods et al., 2004; Hollingworth et al., 2008; Bellinger et al., 2009; Cully and Launikonis, 2016; Cully and Rodney, 2020), inositol 1,4,5-trisphosphate receptors (Balghi et al., 2006a, 2006b; Mondin et al., 2009; Cárdenas et al., 2010), and the SERCA pump (Divet and Huchet-Cadiou, 2002; Morine et al., 2010; Schneider et al., 2013; Kargacin and Kargacin, 2016); and there are also reported changes in the expression level of Ca²⁺-binding proteins (Niebroj-Dobosz et al., 1989a, 1989b; Doran et al., 2004; Lohan and Ohlendieck, 2004; Balghi et al., 2006b; Pertille et al., 2010). Still, in most cases, there remains considerable uncertainty as to up to what extent these changes affect physiological Ca²⁺ cycling in intact muscle fibers and whether or not they play a significant role in either muscle fiber weakness or muscle fiber necrosis. Detection of membrane current and cytosolic Ca²⁺ under voltage control in intact isolated muscle fibers offers exclusive integrative insights into the plasma membrane integrity, ion channel activity, and intracellular Ca²⁺ regulation with specific relevance to EC coupling function. We have used this approach, complemented by confocal imaging of associated relevant parameters, in muscle fibers isolated from DMD^mdx rats (Larcher et al., 2014). The rationale is that most previous studies relied on the C57BL/10ScSn mdx (from here on referred to as mdx) mouse model of DMD, which is commonly admitted to exhibit a mild, nonprogressive disease phenotype (see for review Willmann et al., 2009; McGreevy et al., 2015; Wilson et al., 2017). In comparison, the DMD^mdx rat exhibits a progressively aggravating status characterized, between 3 and 7 mo of age, by a loss of body weight and of muscle mass and myo-necrosis/-degeneration (Larcher et al., 2014). DMD^mdx rat muscles also present with fibrosis that expands with time, resulting in marked endomysial and perimysial infiltration of connective tissue in 7-mo-old animals (Larcher et al., 2014). In contrast, fibrosis remains minimal in muscles other than the diaphragm in the mdx mouse (Tanabe et al., 1986; Stedman et al., 1991). DMD^mdx rats suffer from defects in cardiac structure and function in 3-mo-old animals, which become more severe with time (Larcher et al., 2014), with the extent of cardiac fibrous area being doubled between 3 and 7 mo of age. In mdx mice, although MRI is able to identify early cardiac abnormalities, cardiac defects are usually reported less pronounced and/or to develop over a more delayed time frame (Quinlan et al., 2004; Van Erp et al., 2010; Stuckey et al., 2012). At last, the mdx mouse exhibits a slightly reduced lifespan relative to WT mice (Pastoret and Sebille, 1995; Chamberlain et al., 2007), whereas median survival of the DMD^mdx rat is ∼1 year (Huchet, C., personal communication). Of importance, the severity of the DMD disease in the rat has been confirmed in the other DMD rat models developed by Nakamura et al. (2014) and Taglietti et al. (2022), which also showed progressive and severe skeletal muscle loss associated with fibrotic deposition and fat infiltration leading to severe muscle, respiratory, and cardiac dysfunction, resulting in premature death (at ∼1 year of age, Taglietti et al., 2022). Altogether, there are thus clear indications that the DMD^mdx rat reproduces classical features of Duchenne dystrophy patients better than the mdx mouse. We speculated that it could help enlighten the DMD Ca²⁺ issue.

This work was achieved following the ethics principles of the French Department of Veterinary Services and the French Ministry for Higher Education, Research and Innovation, in accordance with the guidelines of the local committee in charge of animal well-being (SBEA) at the University of Nantes, the French Ministry of Agriculture (decree 87/848), and the revised European Directive 2010/63/EU. Animals were housed at the accredited Therassay (University of Nantes) animal facility. Investigations were approved by the Ethics Committee on Animal Experimentation of the Pays de la Loire Region, in accordance with the guidelines from the French National Research Council for the Care and Use of Laboratory Animals (APAFIS#10792–2017061316021120).

Experiments were performed on muscle fibers isolated from the flexor digitorum brevis (FDB) muscles of 3-mo-old WT and DMD^mdx littermate male rats (Larcher et al., 2014). Rats were sacrificed by intravenous administration of sodium pentobarbital (300 mg, Dolethal; Vetoquinol UK Ltd.). All procedures for fiber isolation and preparation for silicone-clamp electrophysiology were identical to those described in our previous works (e.g., Sanchez et al., 2021; Jaque-Fernandez et al., 2023). In brief, FDB muscles were dissected and stored at 4°C in Tyrode solution during 6 h, the time needed for dedicated express transport from Nantes to Lyon. Muscles were then incubated for 60 min at 37°C in Tyrode solution containing 2 mg/ml collagenase (Sigma-Aldrich type 1). Single isolated fibers were obtained by gentle mechanical trituration of the muscles within the experimental chamber (50-mm-wide culture μ-dish; Ibidi). For this, the μ-dish had its floor covered with a thin layer of silicone grease and was then filled with culture medium containing 10% bovine fetal serum (MI199; Eurobio). Following mechanical trituration of an FDB muscle, single isolated fibers were partially covered with silicone so that an end portion of fiber ∼100–200 µm long remained silicone free, in contact with the extracellular medium. For voltage clamp experiments, the culture medium was then replaced by our standard TEA-containing extracellular solution. The tip of a borosilicate glass micropipette filled with a solution mimicking the composition of the intracellular medium (intracellular-like solution, see Solutions section) and also containing the Ca²⁺-sensitive dye rhod-2 and 15 mM EGTA was inserted in the silicone-embedded portion of fiber. The chloride silver wire in the micropipette was connected to an RK400 amplifier (Bio-Logic). A 30-min period of equilibration was allowed before delivering voltage clamp pulses from a holding voltage of −80 mV. For action potential (AP) measurements, the culture medium was replaced by Tyrode solution and the micropipette was filled with the intracellular-like solution. Following impalement of the silicone-embedded portion of fiber, the current-clamp mode of the amplifier was established and a 5-min period of equilibration was allowed before delivering short (typically 0.5-ms-long) suprathreshold depolarizing current pulses to trigger APs at frequencies ranging between 1 and 80 Hz over a 3-s duration. The resting membrane potential was maintained at −80 mV by passing a constant hyperpolarizing current. All experiments were performed at room temperature (20–22°C).

Procedures were similar to those described previously for mouse muscle fibers (e.g., Kutchukian et al., 2016; Jaque-Fernandez et al., 2020). A Zeiss LSM 800 microscope equipped with a 63× oil immersion objective (numerical aperture 1.4) was used. Measurements of voltage clamp–activated rhod-2 Ca²⁺ transients were achieved with the line-scan (x,t) mode of the microscope: a 50.7-µm-long line was scanned every 1.02 ms using excitation from the 561-nm diode laser. Rhod-2 fluorescence transients were expressed as F/F₀, F₀ being the pre-pulse baseline fluorescence. Quantification of the Ca²⁺ release flux (dCa_Tot/dt) from the rhod-2 fluorescence transients was achieved according to previously described procedures (e.g., Kutchukian et al., 2016). In brief, the time derivative of the total released Ca²⁺ ([Ca_Tot]) was calculated from the occupancy of intracellular Ca²⁺-binding sites and Ca²⁺ transported back across the SR membrane by the SERCA pump. For this, changes in (Ca²⁺) were calculated from the rhod-2 transients using the pseudo-ratio equation described by Cheng et al. (1993), assuming a resting [Ca²⁺] level of 0.1 μM and an equilibrium dissociation constant of 1.2 μM. From this, the time course of Ca²⁺ bound to troponin C, to parvalbumin, and of Ca²⁺ back-transported across the SR membrane by the SERCA pump were calculated using values described by Sanchez et al. (2021) for binding sites concentrations, equilibrium dissociation constants, and on and off rate constants, and with the value for maximum pump rate assumed to be 5 µM.ms⁻¹. This value was adapted from the 3.14–4.49 µM.ms⁻¹ range estimated by Garcia and Schneider (1993) in rat fast-twitch fibers, taking into account the difference in temperature (16°C versus 20–22°C in our conditions). For Ca²⁺ binding to EGTA, the total concentration of effective binding sites was assumed to be 7.2 mM, and values for on and off rate constants and corresponding equilibrium dissociation constant were taken from the Ca²⁺-removal analysis performed by Schuhmeier and Melzer (2004): 0.02 µM⁻¹.ms⁻¹, 0.00271 ms⁻¹, and 0.1355 µM, respectively. With the large concentration of EGTA used in our experiments, Ca-EGTA makes a predominant contribution to the calculated Ca²⁺ release flux.

This was achieved with the 63× oil immersion objective of the same above-mentioned Zeiss LSM 800 microscope. For t-tubule staining, single isolated FDB muscle fibers dispersed within an Ibidi μ-dish filled with Tyrode solution were incubated for 60 min in the presence of 10 µM of Di-8-anepps. For mitochondria staining, separate batches of fibers were incubated under the same conditions in the presence of 10 nM tetramethyl rhodamine methyl ester (TMRM) for 10 min. Di-8-anepps and TMRM fluorescence were imaged using excitation from the 488 and 561-nm diode lasers, respectively. Quantification was achieved with ImageJ (National Institutes of Health) from frames of 512 × 512 pixels at 0.2 and 0.1 μm per pixel for Di-8-anepps and TMRM and Di-8-anepps, respectively.

The t-tubule density was estimated following the approach described by Kutchukian et al. (2017). The largest possible region of interest excluding the plasma membrane was selected. Automatic Otsu’s method thresholding was used to create a binary image of the surface area occupied by t-tubules. A t-tubule density index was calculated as the percentage of positive pixels within the region. Analysis was performed within two images taken at distinct locations of each fiber, and the two corresponding index values were averaged. The coefficient of variation for the t-tubule density index was 10.2% and 7.9% in the WT and DMD^mdx groups, respectively. Of note, different from the initial procedure described by Kutchukian et al. (2017), the “skeletonize” function was not used to delineate the t-tubule network (as also done in Ghasemizadeh et al. [2021]). For measurements of the distance between adjacent t-tubules, average fluorescence profiles were generated from a large region of interest (e.g., Kutchukian et al., 2017) in two images taken at distinct locations of each fiber. The average peak-to-peak distance was measured from each profile and the mean from the two images was calculated. Under the conditions used, lateral resolution is ∼200 nm. The coefficient of variation for t-tubule spacing was 2.9% and 3.1% in WT and DMD^mdx groups, respectively.

From the TMRM images, ImageJ isodata method thresholding followed by particle analysis was used to compare the area of what is referred to as mitochondrial objects, between the two groups.

Single isolated FDB muscle fibers dispersed within an Ibidi μ-dish filled with Tyrode solution were incubated for 30 min in the presence of 10 µM Fluo-4 acetoxymethyl ester. Recording and quantification of Ca²⁺ release events were achieved similarly to previously described procedures (see Jaque‐Fernandez et al., 2020). In brief, two series of 40 consecutive confocal frames of Fluo-4 fluorescence (512 × 512 pixels at 0.2 μm per pixel, 636 ms per frame) were acquired in each fiber. Images in each stack were smoothed and the SD of fluorescence intensity at each pixel position, along the stack, was calculated. The 20% largest values in the SD image were removed to calculate the mean SD of silent areas. Active areas were then defined as pixel positions exhibiting at least 1.5× larger values of SD than the mean SD value from silent areas.

Tyrode solution contained (in mM) 140 NaCl, 5 KCl, 2.5 CaCl₂, 2 MgCl₂, and 10 HEPES. The standard extracellular solution for voltage clamp experiments contained (in mM): 140 TEA-methane-sulfonate, 2.5 CaCl₂, 2 MgCl₂, 1 4-aminopyridine, 10 HEPES, and 0.002 tetrodotoxin. The standard intracellular-like solution contained (in mM): 120 K-glutamate, 5 Na₂-ATP, 5 Na₂-phosphocreatine, 5.5 MgCl₂, 5 glucose, and 5 HEPES. For measurements of rhod-2 Ca²⁺ transients, it also contained 15 EGTA, 6 CaCl₂, and 0.1 rhod-2. All solutions were adjusted to pH 7.20.

Randomization was not used, and experimenters were not blinded to the source of the muscle fibers being studied. Individual data points and corresponding mean ± SD values are presented for groups of fibers from a same animal. Statistical comparison was achieved using a hierarchical/nested analysis implemented in GraphPad Prism 10 software (Eisner, 2021) preventing pseudoreplication: measurements from fibers from the same animal were considered as technical replicates; clustering of replicates from the same animal was corrected for by the hierarchical approach (see Sikkel et al., 2017). Unless otherwise stated, this procedure was used for comparison of all parameters between the WT and DMD^mdx groups. For negative results (P > 0.05), the 95% confidence interval for the difference between the mean value in the DMD^mdx group minus the mean value in the WT group is given. In all presented bar graphs, each individual bar represents one animal and each symbol represents a fiber from that animal.

Enzymatic dissociation of the FDB muscles generated generous batches of healthy-looking muscle fibers, the WT and DMD^mdx groups being qualitatively indistinguishable, except from the routine presence of 10–15% branched DMD^mdx fibers. Those were not included in any of the present analysis. No difference in apparent fiber width could be detected between the WT and DMD^mdx groups (Fig. 1). Staining of the plasma membrane/t-tubules with Di-8-anepps revealed a globally unaltered network (Fig. 2 A) in DMD^mdx fibers, with no sign of change in t-tubule density nor in t-tubule transverse spacing (Fig. 2 B). Dystrophin-deficient muscles were reported to suffer from a decline in mitochondria function and content (see Reid and Alexander, 2021), although not all studies concur on this issue, possibly because of the heterogeneous nature of mitochondrial stress during progression of the disease (Bellissimo et al., 2022). We took advantage of our isolated fiber preparation to examine the mitochondrial network pattern using TMRM staining (Desai et al., 2023).

The red images in Fig. 3 A correspond to the TMRM fluorescence from the above shown transmitted light images of a WT and a DMD^mdx muscle fiber. Overall, there was no striking difference that could be visually pinpointed between the two groups of images. The density of staining, normalized to fiber surface area, did not differ between the two groups (P = 0.16, nested t test, 95% confidence interval for the difference between mean values: −2.8 to 13.9), with mean (±SD) values from the four WT animals being 20.2 ± 2.5% (n = 5 fibers), 20.7 ± 3.9% (n = 6 fibers), 27.1 ± 4.6 (n = 5 fibers), and 20.3 ± 5.2% (n = 10 fibers), and mean values from the four DMD^mdx animals being 28.3 ± 4.5% (n = 5 fibers), 27.6 ± 8.1% (n = 6 fibers), 34.5 ± 9.3% (n = 5 fibers), and 20.7 ± 5.8% (n = 10 fibers).

Particle analysis showed that the large majority of spatially individualized objects occupied an area <5 µm² in both groups (Fig. 3 B), likely corresponding to single pairs of mitochondria on both sides of the Z line. Mean values for the area of this class of objects did not differ between WT and DMD^mdx fibers (Fig. 3 C). When examining larger objects, the frequency was increased in DMD^mdx fibers with 117 objects larger than 20 µm² detected in a total of 21 out of 28 fibers, as compared with 38 detected in a total of 15 out of 26 WT fibers (Fig. 3, D–E). A nested t test comparing the number of large events (>20 µm², Fig. 3 E) detected in each tested fiber from the different animals in the two groups showed that the mean number was significantly elevated in the DMD group (P = 0.043). It should be acknowledged that since TMRM accumulation is driven by mitochondrial membrane potential, we cannot exclude a difference in the density of undetected, depolarized mitochondria between WT and DMD^mdx fibers. Nevertheless, we can at least ascertain that the density of polarized mitochondria is similar in the two groups.

Dystrophin, together with its associated protein complex, is believed to play an important role in securing plasma membrane stability, to the point of preventing membrane rupturing/tearing, which are commonly admitted consequences of dystrophin deficiency. Measurements of ion current across the plasma membrane should help define whether leakiness is enhanced in the DMD muscle fibers. Using voltage clamp, we compared the properties of the membrane current at rest and in response to small hyperpolarizing pulses between WT and DMD^mdx fibers. The capacitance of each voltage-clamped fiber portion was estimated from the running integral of the transient current change elicited by a −20-mV pulse from −80 mV. The inset in Fig. 4 A shows the capacitive current surges at the onset and termination of the pulse, averaged from all tested WT (black) and DMD^mdx fibers (blue). The two bar-and-dot graphs in Fig. 4 A show individual and mean values for capacitance in fibers from each animal in the two groups. Unexpectedly, values were significantly larger in DMD^mdx than in WT fibers (Fig. 4 A, P = 0.006) with the overall mean value being 1.65 times that in WT fibers. Since fiber apparent width did not differ between the two groups (Fig. 1) and a similar length of fiber portion was always left outside the silicone (typically ∼150–200 µm), this suggests an elevated density of the plasma membrane in fibers from the DMD^mdx model. The bottom graph in Fig. 4 A shows another representation of this dataset with capacitance values plotted versus the corresponding amount of surface membrane approximated from the length and apparent width (assumed to be representative of the diameter) of each fiber, assuming it is a cylinder. Fitting a linear relationship in each group gave values consistent with commonly admitted levels of t-tubule/surface membrane ratio, but with a value 1.5 times larger in the DMD^mdx group. Fig. 4, B and D, shows the individual and mean absolute values for holding current (at −80 mV) and for the steady change in current elicited by the −20-mV pulse, respectively, from the same fibers. Those did not differ between the two groups. However, when normalized to the corresponding capacitance, values for both holding current (Fig. 4 C) and steady change upon −20-mV hyperpolarization (Fig. 4 E) are significantly lower in DMD^mdx fibers as compared with WT fibers (P = 0.005 and 0.02, respectively).

Ca_V1.1 is the most well accredited (Robin and Allard, 2015) and characterized Ca²⁺ entry pathway in differentiated muscle fibers. Considering the amount of literature pointing to exacerbated Ca²⁺ entry in dystrophin-deficient muscle cells, assessing Ca_V1.1 channel function in DMD^mdx fibers was of prime interest. Our measurements of the membrane current taken in response to 0.5-s-long depolarizing pulses of increasing amplitude showed that the current exhibited very similar properties in the two groups (Fig. 5 A), and none of the parameters fitted from the voltage dependency of the peak current differed between WT and DMD^mdx fibers (Fig. 5, B and C).

Fig. 6 shows representative Ca²⁺ release datasets from a WT fiber (left) and from a DMD^mdx fiber (right). A line-scan F/F₀ image taken during application of a pulse to −10 mV is shown in Fig. 6 A. Fig. 6 B shows the superimposed line-averaged F/F₀ traces from the two fibers in response to all tested pulses. The corresponding calculated traces for the rate of Ca²⁺ release (Fig. 6 C) displayed the classical early peak, the amplitude of which increased with the pulse amplitude, that declined rapidly toward a very small level in both groups. For intermediate voltages, a hump-like component was routinely detected most likely reflecting the Ca²⁺ storage/release function of calsequestrin in the SR lumen (Royer et al., 2010). Overall, the time course of the fluorescence transients and of the corresponding release fluxes presented no specific feature in DMD^mdx fibers that would visually distinguish them from those in WT fibers. But, the maximum amplitude of the peak Ca²⁺ release was lower in the diseased condition. The running integral of each Ca²⁺ release trace during the voltage pulses is shown in Fig. 6 D. Despite a maximum peak amplitude of Ca²⁺ release smaller by ∼25% in the DMD^mdx fiber, the maximum amount of released Ca²⁺ (expressed with respect to the myoplasmic volume) was similar in the two fibers. All values for peak Ca²⁺ release and total amount of released Ca²⁺ from the tested 29 WT fibers and 27 DMD^mdx fibers are presented, versus voltage, in Fig. 7, A and B, respectively. The voltage dependency of peak Ca²⁺ release in each fiber was fitted with a Boltzmann function, the result of which is shown as a superimposed curve on each corresponding dataset in Fig. 7 A. Individual values (circles) and corresponding mean (±SD) values (from groups of fibers from the same animal, bars) for maximum peak Ca²⁺ release, mid-activation voltage, and slope factor are presented in Fig. 8, A–C, respectively, whereas values for the total amount of released Ca²⁺ for the largest pulse (to +60 mV) are in Fig. 8 D. Statistical analysis showed that the maximum peak amplitude of Ca²⁺ release was significantly depressed in the DMD^mdx fibers by 25% as compared with WT fibers, whereas all other parameters did not differ. The preserved total amount of released Ca²⁺ in the DMD^mdx group indicates that the time course of the release flux must differ from that in the WT group. To test for this, in each fiber, the decay of Ca²⁺ release triggered by the pulse to +60 mV was fitted with a two-exponential plus constant function. Values for the parameters (amplitude and time constant for the fast and slow components, and final level/constant) were compared using the nested analysis. Respective values in WT (n = 7) and DMD^mdx animals (n = 7) for the amplitude (mean of 36.6 ± 10.3 versus 32.2 ± 3.8 µM.ms⁻¹) and time constant (mean of 4.3 ± 1.2 versus 6.0 ± 1.9 ms) of the fast component and for the final level (mean of 0.34 ± 0.18 versus 0.34 ± 0.17 µM.ms⁻¹) did not differ. In contrast, the slow exponential component in the DMD^mdx group exhibited a reduced amplitude (mean of 26.7 ± 6.7 versus 49.1 ± 14.8 µM.ms⁻¹, P = 0.002) and a larger time constant (mean of 53.7 ± 14.9 versus 30.2 ± 9.6 ms, P = 0.007). Thus, the slower late decay of Ca²⁺ release in the DMD^mdx fibers may well explain for the lack of significant difference in total Ca²⁺ released between the two groups, despite the lower peak amplitude in the DMD^mdx group.

Enhanced SR Ca²⁺ leak through RYR1 channels was proposed to contribute to dysfunction and damage of dystrophic muscle fibers (Bellinger et al., 2009). Because we have previously revealed RYR1 overactivity at rest, in the form of Ca²⁺ sparks, in intact muscle fibers from other models of muscle diseases (Kutchukian et al., 2017, 2019), we tested whether this would be also the case in the DMD^mdx condition. The bottom images in Fig. 9 A correspond to the SD of Fluo-4 fluorescence intensity at each pixel position, along a sequence of 40 consecutive x,y confocal frames. Local spots of elevated SD correspond to locations of Ca²⁺ sparks occurrence. In these conditions, a few sparks were routinely detected, in both the WT and the DMD^mdx fibers. The relative fiber area showing sparks activity was quantified. Individual values from each fiber and corresponding mean values from fibers from a same animal are shown in Fig. 9 B. Values were very low: <0.5% of the fiber’s area presented with Ca²⁺ release activity, consistent with the notorious very low incidence or absence, of Ca²⁺ sparks in healthy intact mammalian muscle fibers. There was no significant difference between the WT and the DMD^mdx groups.

Lack of dystrophin in muscle cells is presumed to impair plasma membrane stability so that contraction would induce membrane rupture and/or impairment of excitability; this would be involved in muscle strength loss (Petrof et al., 1993; Call et al., 2013; Roy et al., 2016). With the aim of identifying functional signs of such processes, we challenged WT and DMD^mdx muscle fibers with trains of APs, in the absence of intracellular Ca²⁺ buffering, thus allowing full contraction of the fiber portion under study. Representative examples of trains of APs are shown in Fig. 10, from a WT and a DMD^mdx muscle fiber at the different tested frequencies. Overall, we found no indication that electrical activity of the diseased fibers would be specifically sensitive to either repeated or tetanic contractions. Fig. 11 presents values for the voltage level preceding each AP and for the peak amplitude of each AP, during the course of the trains, at the different frequencies, in each fiber tested. At 80 Hz, two out of six WT fibers and four out of eight DMD^mdx fibers presented with a substantial drop in AP amplitude. Otherwise, in no case did we detect depolarization of the resting (pre-AP) levels that would witness substantial membrane damage and or loss of excitability. We also quantified the rate of voltage return to the resting level, at the end of the 20, 50, and 80 Hz trains, by fitting an exponential function to this time course (highlighted in red in Fig. 10): there was no significant difference between the groups of WT and DMD^mdx fibers. With respect to single AP properties, peak values did not differ between the two groups but the half-width was significantly enhanced (P = 0.02) in the DMD^mdx condition with mean (±SD) values of 1.45 ± 0.27 (n = 4 fibers), 1.46 ± 0.1 (n = 3), and 1.24 ± 0.13 ms (n = 3) in the fibers from three WT animals, and 1.61 ± 0.19 (n = 4), 1.76 ± 0.34 (n = 4), and 1.60 ± 0.09 ms (n = 3), in fibers from three DMD^mdx animals.

Multiple molecular and cellular consequences of dystrophin deficiency have been documented (see for instance Allen et al. [2016]), but the exact mechanisms that determine muscle fibers weakness and muscle fibers necrosis in DMD remain controversial. Altered Ca²⁺ fluxes affecting intracellular Ca²⁺ handling has been proposed, for >40 years (e.g., Bodensteiner and Engel, 1978), to underlie or at least contribute. However, despite recurrent authoritative statements in the literature regarding the critical role of Ca²⁺ entry across the plasma membrane or of Ca²⁺ leak across the SR membrane, direct evidence of how exactly (and quantitatively to what extent and along which time frame) these mechanisms affect muscle fiber function and integrity remains nebulous. Further understanding of these issues requires detailed knowledge of the functional properties of the diseased muscle fibers, keeping in mind that conditions allowing either control or reading of the plasma membrane voltage are precious when it comes to investigate the properties of transmembrane Ca²⁺ fluxes and of SR Ca²⁺ release. We used here a combination of electrophysiology and imaging that has proved efficient in the past to identify and quantify defects in Ca²⁺ signaling and EC coupling in other muscle diseases (e.g., Kutchukian et al., 2016, 2017, 2019; Huntoon et al., 2018; Pelletier et al., 2020; Silva-Rojas et al., 2022). We applied this combination to muscle fibers isolated from a rat model (DMD^mdx) that reproduces the DMD patient’s condition better than the standard mdx mouse. One limitation of the preparation is that collagenase treatment was used for isolation, so that the extracellular matrix is largely removed. Intact isolated DMD^mdx muscle fibers looked fine, with no sign of widespread or local sarcomere shortening (Fig. 2 B) that would be expected in the presence of membrane damage. Measurements of the apparent width of the FDB muscle fibers showed no difference between the WT and DMD^mdx groups, which contrasts with the smaller diameter values reported by Larcher et al. (2014) in DMD^mdx fibers, from analysis of frozen sections from the biceps femoris muscles. It cannot be excluded that the FDB muscles exhibit a milder phenotype than the biceps femoris muscles. However, diameter values determined from a muscle section cannot distinguish between non-branched and branched fibers, and it is very possible that this contributed to the difference between the two studies, as we did not consider branched fibers in our analysis. Membrane current measurements also provided no indication for an increase in the resting membrane conductance that would also be anticipated from a leaky or damaged plasma membrane. WT and DMD^mdx fibers contracting in response to trains of APs behaved very similarly, with no sign that the contractile activity would specifically affect plasma membrane polarization in the diseased fibers. Of course, one cannot exclude that more stringent conditions, for instance eccentric contractions (Call et al., 2013), are required to induce membrane damage and/or electrophysiological alterations in the diseased fibers. However, single muscle fiber measurements of either the membrane current or membrane voltage are very sensitive to the integrity of the plasma membrane and, with respect to the widely acknowledged concepts of membrane frailty and rupturing, that are traditionally associated with dystrophin deficiency, we find it puzzling that this does not show in the electrical measurements.

In fully differentiated fibers, t-tubules and mitochondria form dense, well-organized networks. The t-tubule network integrity is mandatory for proper EC coupling function. Our di-8-anepps images did not show any obvious alteration in the DMD^mdx fibers, at least nothing that would come any close to the t-tubule loss observed, for instance, in muscle fibers from the mouse model of myotubular myopathy (Al Qusairi et al., 2009; Kutchukian et al., 2016), which make fibers suffer from EC coupling failure (Szentesi et al., 2023). On the contrary and quite intriguing, we found that DMD^mdx fibers presented with a larger capacitance than WT fibers. This may result from enhanced t-tubule tortuosity at a resolution level inaccessible to di-8-anepps imaging. Interesting in this respect are results from early electron microscopy studies of biopsies from DMD patients, which described alterations of the t-tubule system taking the form of abnormal course and tangle formation (Oguchi and Tsukagoshi, 1980; Oguchi et al., 1982). We are currently exploring the possibility that this is happening in the DMD^mdx muscle fibers.

With respect to mitochondria, there have been reports of changes in mitochondria morphology and of concurrent metabolic defects in dystrophic muscle, possibly related to altered mitochondrial Ca²⁺ content (Pant et al., 2015; Moore et al., 2020; see Mareedu et al., 2021), although absence of differences in aspects of mitochondrial content and dynamics has also been reported (Barker et al., 2018). Our TMRM measurements showed that the global mitochondrial network pattern is similar in WT and DMD^mdx fibers, with a majority of small single objects in the two populations. Interesting though was the propensity for more numerous large objects (area >20 µm²) in DMD^mdx fibers that, we speculate, may be related to the previously acknowledged changes in mitochondria dynamics and mitophagy in dystrophin-deficient muscle cells (Pant et al., 2015; Reid and Alexander, 2021).

Properties of the Ca_V1.1 Ca²⁺ current are unaffected in DMD^mdx fibers. Although this Ca²⁺ entry is not mandatory for muscle function (Dayal et al., 2017), our data at least prove that there is no enhanced Ca²⁺ entry through Ca_V1.1 in the DMD^mdx fibers, which does not exclude that it may occur through alternate pathways. This result may also be viewed with reference to muscle disease conditions affecting EC coupling, with depressed SR Ca²⁺ release being paralleled by depressed Ca_V1.1 Ca²⁺ current (Al Qusairi et al., 2009; Kutchukian et al., 2017; Huntoon et al., 2018; Pelletier et al., 2020; Silva-Rojas et al., 2022). The situation is obviously different here, even though the maximum peak amplitude of Ca²⁺ release was depressed by a similar extent as, for instance, in the model of myopathy due to reduction in RYR1 content (Pelletier et al., 2020). This indicates that the reason for reduced Ca²⁺ release in DMD^mdx fibers is likely to be mechanistically distinct. In comparison, previous studies using the mdx mouse provided contrasting results, some groups reporting reduced SR Ca²⁺ release in the mdx fibers (Woods et al., 2004; Hollingworth et al., 2008), whereas other reported either minor or no differences with the WT fibers (Head, 1993; Collet et al., 1999; Tutdibi et al., 1999). Here, the reduced peak SR Ca²⁺ release occurred with no significant change in the total amount of released Ca²⁺ at the end of the 0.5-s-long pulses. A lower SR content, which could occur as a consequence of an increased SR Ca²⁺ leak, was thus likely not involved. Also arguing against an enhanced SR Ca²⁺ leak in the DMD^mdx fibers was the absence of any difference in the spontaneous Ca²⁺ release activity at rest.

Altogether, the present characterization of intact DMD^mdx muscle fibers failed to identify functional signs of plasma membrane compromised integrity, fragility, unspecific leakiness, or enhanced Ca²⁺ entry across Ca_V1.1 channels. It though revealed a significant reduction of peak SR Ca²⁺ release during EC coupling, the extent of which (25%) likely contributes to muscle weakness, without being expected to dramatically compromise muscle function (if remaining at that level). We found no indication that this decrease is related to an enhanced SR Ca²⁺ leakage. We cannot exclude that the previously proposed enhanced plasma and SR membrane leakiness to Ca²⁺ occur within a range and/or time scale that are not accessible to our measurements. We also cannot exclude that the pathophysiological impact of such alterations is of limited relevance with respect to the dramatic outcome of the disease in terms of muscle function and integrity.

In summary, muscle fibers from 3-mo-old DMD^mdx rats present no obvious deleterious alteration of plasma/t-tubule membrane properties that would be consistent with fragility, compromised excitability, or excess Ca²⁺ entry. Yet, these fibers exhibit a reduced maximum rate of voltage-activated SR Ca²⁺ release, which likely contributes to depressed force production, without being expected to dramatically impair muscle function. Future investigations at later stages of the disease should help define whether these features remain stable or deteriorate as the disease progresses.

All generated datasets are presented in the article. Raw data will be made available upon reasonable request to the corresponding author.

Eduardo Ríos served as editor.

This work was supported by grants from the Centre National de la Recherche Scientifique, the Institut national de la santé et de la recherche médicale, and Université Claude Bernard - Lyon 1 to Institut NeuroMyoGène—Pathophysiology and Genetics of Neuron and Muscle (CNRS UMR5261- INSERM U1315). This work was also supported by the Association Française contre les Myopathies (AFM-Téléthon: Alliance MyoNeurALP2 program, project # 4.1.1 to V. Jacquemond).

Author contributions: J. Schreiber: formal analysis, investigation, and writing—review and editing. L. Rotard: formal analysis, investigation, visualization, and writing—review and editing. Y. Tourneur: formal analysis and software. A. Lafoux: resources. C. Berthier: project administration and supervision. B. Allard: conceptualization and writing—review and editing. C. Huchet: methodology, resources, validation, and writing—review and editing. V. Jacquemond: conceptualization, data curation, formal analysis, funding acquisition, investigation, methodology, project administration, resources, supervision, validation, visualization, and writing—original draft, review, and editing.