T-tubule disorganization and defective excitation-contraction coupling in muscle fibers lacking myotubularin lipid phosphatase

Abnormal Distribution of Triads in Myotubularin-Deficient Myofibers.

In this study, we used a constitutive Mtm1 KO line in the 129PAS background, because mutant mice develop a clinically more homogenous and severe CNM than previously characterized in C57/Bl6 KO mice (20). These mutant animals develop a progressive growth impairment, with a 10–15% reduction of weight during the first 4 postnatal weeks as compared to wild-type (WT) littermates, reaching about 30% from 5 weeks of age (see Fig. S1). Mtm1-deficient mice manifest a progressive muscle weakness, starting clinically at around 3–4 weeks, and most of them are dead by the age of 7 weeks (mean = 38.6 ± 7.1 days, range = 28–54, n = 11). In 5-week-old mutant mice, the weight of the tibialis anterior (TA) muscle is lower by 50% compared to controls, the corresponding mean cross-sectional area of myofibers is significantly smaller (554 μm² in KO versus 1,260 μm² in WT), and the percentage of internalized nuclei is 10 ± 3.9 compared to 1.6 ± 0.3 in WT mice (see Fig. S1). Fig. S1 also shows that the distribution of oxidative staining, which reflects the position of mitochondria and ER, is abnormal in mutant TA myofibers and adopts a “necklace” aspect with accumulation of staining in the subsarcolemmal region, as observed in XLMTM patients (21) and a “core-like” aspect, with reduction of staining in the center of fibers, as seen in myopathies due to RYR1 mutations (22). Central aggregation of NADH-TR staining (mitochondria), a hallmark of XLMTM pathology, is also observed.

Since myotubularin is partly associated with triads in skeletal muscle (18) and transcriptome studies from XLMTM murine muscle revealed disturbances in expression of some genes implicated in Ca²⁺ homeostasis including Cacng1 (2.2-fold up), Homer1 (2.4-fold down), Camk2D (3.2-fold up), Chrna1 (6.7-fold up), Rrad (61-fold up), and Sln (72-fold up), we investigated the E-C coupling machinery in this animal model. We analyzed at 5 weeks the distribution of triads by immunolabeling markers of T-tubules (DHPR) and terminal cisternae of SR (RyR1). The longitudinal SR membrane was monitored by SERCA1 ATPase labeling. We observed abnormal localization of some of these markers compared to WT muscles where they aligned as a typical striated pattern reflecting the organization of myofibrils (Fig. 1A). In Mtm1 KO muscles, DHPRα and RyR1 striations were disorganized with absence of labeling in some regions and occasional divergence from the transversal orientation, while SERCA1 distribution appeared mostly normal. This suggests a disorganization of both T-tubules and terminal SR cisternae.

We further analyzed the morphology of triads at the ultrastructural level by electron microscopy using potassium ferrocyanide as a selective staining procedure for T-tubules. In 5-week-old mutant TA myofibers, we found a significant proportion of tubules with longitudinal orientation that are aligned to the direction of myofibrils (Fig. 1B). This distribution is very rare in normal muscle at this age (less than 6% of total tubules) and was found in 11 out of 16 mutant fibers. Longitudinal (L)-tubules represents 25% of total tubules in analyzed KO myofibers (n = 16 fibers from both four WT and four Mtm1 KO mice) (Fig. 1C). We also measured the number of tubules (T- and L-tubules) per Z-line, which reflects the distribution of triads along myofibrils, and corresponds to about two in normal mature muscle cells. Fig. 1D shows that there is a 40% decrease in the number of triads in mutant fibers at this age (mean 1.09 ± 0.1 in KO versus 1.86 ± 0.03 in WT, P < 0.001).

To know whether these T-tubule abnormalities are already present at early stages of disease progression, we investigated 2-week-old KO mice. At this age, weakness could not be observed in hind limbs, but muscle weight was already decreased, mitochondria appeared partly mislocalized, and the percentage of internal nuclei was slightly increased in KO muscle (see Fig. S2). Indeed, L-tubules were also more frequent in mutants at this early stage of the disease from about 4% in WT to a mean value of 15% in KO fibers (Fig. 1E). In addition, we found that the number of tubules per Z-line was already lower by about 25% in Mtm1-deficient muscle fibers at this age (1.37 ± 0.07 and 1.81 ± 0.06 in KO and WT fibers, respectively, P < 0.001) (Fig. 1F).

Altogether, these results indicate that myotubularin is important for the proper organization of the T-tubules and triads in skeletal muscle and that triad disorganization is an early event in the pathogenesis of the disease in these mice that progresses over time.

Absence of Myotubularin Leads To Dysregulation of Genes Involved in Calcium Homeostasis.

We also examined the expression level of key regulators of the E-C coupling process in TA muscle. By using quantitative RT-PCR, we measured the mRNA level of the α₁, β₁, and γ₁ subunits of the voltage-sensing DHPR (Cacna1s, Cacnb1, and Cacng1), the SR Ca²⁺ release channel type 1 ryanodine receptor (RyR1), type 1 and 2 SERCA pumps (Atp2a1, Atp2a2), and calsequestrin 1 and 2 (Casq1, Casq2), which encode the major Ca²⁺-buffering proteins in the lumen of striated muscle SR (23). In 5-week-old-mice, we found that Cacnb1 and Cacng1 mRNAs were significantly increased by 1.3- and 3-fold, respectively, whereas transcript level of Casq2 was decreased by about 30% (Fig. 2A). The expression of Cacng1 was already augmented in mutant muscle at 2 weeks of age (SI Methods and Fig. S3).

Given these data, we quantified DHPRα₁, DHPRβ₁, RyR1, SERCA1, and calsequestrin protein level in TA muscle of Mtm1 null mice during disease evolution. We found that the ryanodine receptor was strikingly reduced by 3-fold in microsomal preparations of KO muscles at 5 weeks of age and DHPRα₁ was decreased by about 30%, whereas DHPRβ₁ level was about 6-fold higher (Fig. 2B and Fig. S3). Discrepancies between the mRNA and protein levels of Ryr1 and Cacna1s may result from decreased protein stability, possibly due to an alteration in RYR1-DHPR interaction. All these proteins were also analyzed in muscle from 2-week-old mice, when weakness was not visible, and we found no significant differences between genotypes (see Fig. S3). These results demonstrate that dysregulation of components of the E-C machinery is a secondary and progressive process in murine myotubular myopathy.

Impaired SR Calcium Release in Muscle Fibers from Mtm1 KO Mice.

These observations led us to investigate the functional properties of E-C coupling in muscle fibers from 5-week-old mice. For these measurements, short muscle fibers that are easily amenable to voltage-clamp, were isolated from the flexor digitorum brevis (FDB) and interosseus muscles. In these muscles, we also obtained evidence for partial loss of T-tubules in mutant fibers using confocal imaging of di-8-anneps fluorescence (SI Methods and Fig. S4).

Figure 3A shows indo-1 calcium signals recorded from a representative control and Mtm1 mutant fiber, respectively. Ca²⁺ transients were triggered by voltage-clamp depolarizations of 5, 10, and 20 ms duration from −80 to +10 mV. The peak amplitude of the indo-1 transients was much smaller in the Mtm1-deficient fiber than in the control one, while the resting indo-1 level and the overall time-course of the transients were similar in the two fibers. Mean results from the indo-1 measurements on 14 fibers each revealed that the peak change in cytosolic calcium concentration (Δ[Ca²⁺]) was strongly depressed for all pulse durations in Mtm1-deficient fibers as compared to the values in control fibers (P ≤ 0.001) (Fig. 3B). Conversely resting [Ca²⁺] levels were unaffected (Fig. 3C). The overall calcium removal capability of the cytosol including SR Ca²⁺ uptake was estimated from the rate of [Ca²⁺] decay after the end of the pulses. Exponential fits to the [Ca²⁺] decay time course showed that the rate was significantly depressed (by ≈30%, P < 0.001) following 5-ms-long pulses in mutant muscle cells, but this was not the case after longer pulses (Fig. 3D). This lower rate at 5-ms-long pulses may result from the very low peak [Ca²⁺] levels (< 0.5 μM) reached in response to this pulse in KO fibers, considering the strong [Ca²⁺] dependence of the SR Ca²⁺ ATPase activity. To know whether the depressed peak Δ[Ca²⁺] in the Mtm1-deficient fibers resulted from a decreased SR calcium content, we determined the total amount of releasable calcium by measuring indo-1 signals in fibers equilibrated with a large concentration of intracellular EGTA and challenged by a series of voltage-clamp depolarizations applied in the presence of a SR Ca²⁺ pump inhibitor. Results showed that the mean SR Ca²⁺ content did not significantly differ between control and Mtm1-deficient fibers (SI Methods and Fig. S4).

We next tested whether the depressed Ca²⁺ release in Mtm1-deficient fibers was related to an alteration of the properties of the voltage-sensor of E-C coupling, the DHPR. The voltage-activated Ca²⁺ current through the DHPR was measured in WT and Mtm1-deficient fibers. Figure 4A shows representative calcium current traces from a WT and from an Mtm1-deficient fiber in response to 0.5-s-long depolarizing pulses to various levels. The corresponding mean peak Ca²⁺ current density versus voltage relationships from identical measurements performed in 16 WT and 20 KO fibers are shown in Fig. 4B. Surprisingly, the peak amplitude of the Ca²⁺ current was strongly reduced in the Mtm1-deficient fibers. Fitting the individual current-voltage curves by a Boltzmann function scaled by the driving force (24) showed that this reduction corresponded to a 37% depression in the mean value for the maximal conductance of the calcium channels (P < 0.03). Overall results demonstrate that myotubularin deficiency is responsible for a severe functional alteration of both DHPR-mediated Ca²⁺ entry and RyR1-mediated SR Ca²⁺ release.

Skeletal Muscle T-Tubule Labeling.

WT and Mtm1 KO mice in the 129PAS background were used in this study. Care and manipulation of mice were performed in accordance with national and European legislations on animal experimentation.

Immunohistochemistry.

Anesthetized mice [by i.p. injection of 5 μL per body gram of ketamine (20 mg/mL; Virbac) and xylazine (0.4%, Rompun; Bayer)] were perfused with 4% paraformaldehyde before muscle dissection. Semithin sections (500-nm) of muscle were prepared as previously described (18). We used monoclonal antibodies directed against DHPRα₁ (Ca_v1.1) subunit (MA3–920; Affinity Bioreagents), RyR (clone 34C; Sigma), and SERCA1 ATPase (MA3–911; ABR).

Electron Microscopy.

Selective staining of T-tubules in TA muscle was modified from (26). Briefly, muscles were dissected from anesthetized mice and fixed in 2.5% paraformaldehyde, 2.5% glutaraldehyde, and 50 mM CaCl₂ in 0.1 M cacodylate buffer (pH 7.4). Samples were postfixed with 2% OsO₄, 0.8% K₃Fe(CN)₆ in 0.1 M cacodylate buffer (pH 7.4) for 2 h at 4 °C and incubated with 5% uranyl acetate for 2 h at 4 °C. Muscles were dehydrated in a graded series of ethanol and embedded in epon resin. Thin (70-nm) sections were stained with uranyl acetate and lead citrate and examined by transmission electron microscope. The length of longitudinally oriented T-tubules and number of triads per Z-line were quantified with Metamorph 3 software from electron micrographs (magnification 9,800× and 5,600× for muscles of 5- and 2-week-old mice, respectively).

Quantitative RT-PCR Analysis.

Total RNA was purified from muscles of 2- and 5-week-old male mice using TRIzol reagent (Invitrogen) according to manufacturer's instructions. cDNA was synthetized from 2–5 μg total RNA using SuperScript II reverse transcriptase (Invitrogen) and random hexamers. Quantitative PCR amplifications of cDNA were performed on Light-Cycler 480 and Light-Cycler 24 instruments (Roche) using 58 °C as melting temperature. Primer sequences for amplification are provided as SI Methods.

Immunoblot Analysis

Electrophysiology and Intracellular Calcium Measurements.

Experiments were performed on single skeletal fibers from the FDB and interosseus muscles using previously described methods and analytical procedures (42–44), see SI Methods for further details.

Statistical Analysis.

Data were statically analyzed using paired Student t-test. Values were considered significant when P ≤ 0.05. Growth curves of mice were analyzed by ANOVA.