ANT-dependent MPTP underlies necrotic myofiber death in muscular dystrophy

Abstract

Mitochondrial permeability transition pore (MPTP) formation contributes to ischemia-reperfusion injury in the heart and several degenerative diseases, including muscular dystrophy (MD). MD is a family of genetic disorders characterized by progressive muscle necrosis and premature death. It has been proposed that the MPTP has two molecular components, the adenine nucleotide translocase (ANT) family of proteins and an unknown component that requires the chaperone cyclophilin D (CypD) to activate. This model was examined in vivo by deleting the gene encoding ANT1 (Slc25a4) or CypD (Ppif) in a δ-sarcoglycan (Sgcd) gene–deleted mouse model of MD, revealing that dystrophic mice lacking Slc25a4 were partially protected from cell death and MD pathology. Dystrophic mice lacking both Slc25a4 and Ppif together were almost completely protected from necrotic cell death and MD disease. This study provides direct evidence that ANT1 and CypD are required MPTP components governing in vivo cell death, suggesting a previously unrecognized therapeutic approach in MD and other necrotic diseases.

Deletion of genes encoding ANT1 and CypD synergistically desensitizes the MPTP and blocks myofiber necrosis in muscular dystrophy.

INTRODUCTION

The mitochondrial permeability transition pore (MPTP) is a megachannel in the inner mitochondrial membrane that opens in response to high matrix Ca²⁺ and oxidative stress (1). MPTP activation permeabilizes the inner mitochondrial membrane to solutes of up to 1.5 kDa, induces mitochondrial swelling and membrane rupture, dissolves the adenosine triphosphate (ATP) producing H⁺ gradient, and initiates necrotic cell death (2). MPTP-dependent cell death contributes to several important human diseases including cardiac ischemia-reperfusion injury, muscular dystrophy (MD), and neurodegenerative diseases such as Huntington’s disease, amyotrophic lateral sclerosis, and Alzheimer’s disease (3–10).

Despite a wealth of data describing the MPTP phenomenon, the molecular identity of the MPTP itself remains controversial (1). One of the early proposed molecular candidates for the MPTP was the adenine nucleotide translocase (ANT) family of proteins (11, 12); however, genetic experiments were less definitive (13). We revisited this model and observed that the ANT family of proteins do comprise one part of the MPTP (14). Specifically, deletion of the three mouse ANT isoform encoding genes resulted in desensitization of the MPTP, although activation was not blocked (14). However, when liver mitochondria lacking all three ANT isoforms were also treated with cyclosporin A (CsA), an inhibitor of the MPTP regulatory protein cyclophilin D (CypD) (15–18), MPTP activation was abolished (14). Simultaneous deletion of Ppif (CypD gene) with the three ANT isoform encoding genes also abolished MPTP opening (14). These results suggested that the MPTP is composed of ANT and another unknown protein species that requires CypD (14, 19). The mitochondrial ATP synthase complex may also generate MPTP-like activity, either at the level of the c-ring subunit within the adenosine triphosphatase (ATPase) core or at the interface between the ATPase core and the oligomycin sensitivity conferral protein (OSCP) subunit (20, 21). A recent study of ATP synthase c-subunit–deleted mitochondria displayed a substantial change in MPTP conductance but retained a smaller MPTP-like pore that could be inhibited by the ANT-specific antagonist bongkrekic acid (BKA) (22). Another study testing the role of ATPase subunits in MPTP found that genetic ablation of the OSCP, b, and g ATPase subunits in HeLa cells caused altered CsA-sensitive MPTP activity but retained ANT-dependent MPTP activity that could be induced by atractyloside, an ANT-specific MPTP-sensitizing agent, or inhibited by BKA (23). A recent proteomics study of the mitochondrial interactome in the failing heart identified an ANT1 species that is consistent with an open pore-forming conformation providing evidence that ANT proteins form MPTP pores in a pathophysiological context (24). Collectively, these studies provide evidence for ANT-dependent, two-component model of the MPTP. However, the role of ANT-dependent MPTP has only been investigated in reductionist systems, so it remains unclear whether it serves as a bona fide regulator of necrotic cell death in vivo.

MD is a family of genetic disorders characterized by progressive muscle wasting, decreased muscle function, and premature death (25). At the myofiber level, many sMD result in sarcolemma instability and rupture resulting in uncontrolled Ca²⁺ influx leading to myofiber necrotic cell death and progressive pathology (26). Previous studies have demonstrated that genetic or pharmacological inhibition of CypD reduces myofiber cell death and decreases MD disease burden in animal models (5, 27–29). Here, we generated Slc25a4^−/− (ANT1 protein) mice as well as mice deficient in both Slc25a4^−/− and Ppif^−/− (CypD protein), which when crossed into the δ-sarcoglycan gene deletion (Sgcd^−/−) mouse model of MD (30) leads to a stepwise reduction in myofiber necrosis and MD disease manifestation. These results indicate that the ANT-dependent component of the MPTP underlies necrotic cell death in vivo. Furthermore, because simultaneous deletion of Slc25a4 with Ppif caused an almost complete inhibition of muscle histopathology in these Sgcd^−/− mice, it supports the two-component model of the MPTP as contributing to necrosis in a disease relevant model.

RESULTS

Mitochondria-dependent myofiber necrosis is a major underlying mechanism leading to pathology in MD. To assess MPTP activity in dystrophic mitochondria, we simultaneously measured the Ca²⁺ retention capacity (CRC) and absorbance-based swelling in mitochondria isolated from either healthy wild-type (WT) muscle or Sgcd^−/− dystrophic muscle (30). We observed that Sgcd^−/− myofiber mitochondria exhibited substantially reduced CRC compared with WT mitochondria (Fig. 1A), as has been observed in isolated skeletal muscle mitochondria from mdx mice, a model of Duchenne MD (31). We also observed that mitochondria from quadricep of Sgcd^−/− mice released more Ca²⁺ than WT mitochondria, which suggests that dystrophic mitochondria are in a state of Ca²⁺ overload (Fig. 1B). This increase in matrix Ca²⁺ is known to cause the increased MPTP activation that underlies myofiber necrosis in dystrophic muscle, as mice lacking the Ppif gene are partially protected (5, 28).

Fig. 1. Slc25a4 gene deletion desensitizes MPTP activity in dystrophic muscle.

(A) Representative mitochondrial CRC assay of isolated C57Bl/6J WT and Sgcd^−/− quadricep mitochondria. The arrow and each inflection point represent 20 μM additions of CaCl₂. The y axis is arbitrary units (AU). (B) Mitochondrial Ca²⁺ overload assay of isolated WT and Sgcd^−/− quadricep mitochondria. Two milligrams of mitochondria was treated with 40 μM alamethicin (ALM, red arrow). WT n = 4; Sgcd^−/− n = 3 (*P ≤ 0.05 versus WT). The y axis is arbitrary units. (C) Western blots of quadricep mitochondria for the proteins: ANT1, ANT2, ANT4, CypD, and the protein-loading controls electron transport chain complex II (CII) and voltage-dependent anion channel 1 (VDAC1). Testis is a positive control for ANT4 and liver for ANT2. (D) Representative mitochondrial CRC assay (top) and simultaneous swelling (bottom) as absorption (Abs) in relative units (RU) with (E) average CRC measurements of isolated quadricep mitochondria from the indicated genotypes of mice. The y axis is arbitrary units. The legend shows genotypes of mice used by color coding in the figure panels. WT n = 10; Sgcd^−/− n = 4; Slc25a4^−/−;Sgcd^−/− n = 3; Slc25a5^MKO;Sgcd^−/−, n = 3; Ppif^−/−;Sgcd^−/−, n = 4 (*P ≤ 0.05 versus WT). (F) Mitochondrial Ca²⁺ load assay from quadricep of the indicated genotypes of mice. Forty micromolar alamethicin was used. The y axis is arbitrary units. WT n = 6; Sgcd^−/− n = 3; Slc25a4^−/−;Sgcd^−/− n = 15; Ppif^−/−;Sgcd^−/− n = 8 (*P ≤ 0.05 versus WT). (G) Representative transmission electron micrographs of quadricep from mice with the indicated genotypes. Scale bar, 800 nm. White arrows indicate swollen mitochondria with few cristae.

To interrogate the role and identity of the MPTP in MD pathology, we investigated mice with genetic deletion of the genes encoding the ANT isoforms. Healthy skeletal muscle almost exclusively expresses Slc25a4 (ANT1), and previous studies of Slc25a4 gene–deleted mice did not show up-regulation of Slc25a5 (ANT2) expression in skeletal muscle (32, 33). However, Slc25a5 expression was shown to be up-regulated in the muscles of mdx mice (31), suggesting that it may contribute to MD pathology. Accordingly, here, we generated Slc25a4 gene–deleted mice crossed into the δ-sarcoglycan model of limb-girdle MD (Slc25a4^−/−;Sgcd^−/−) (30). We also analyzed mice with muscle-specific deletion of the Slc25a5 gene [LoxP (fl) targeted] using the Myl1^Cre-allele herein called Slc25a5^MKO [germline loss of Slc25a5 in mice is embryonic lethal (34)]. We investigated the previously published Ppif^−/− (CypD) gene deleted mice (Ppif^−/−;Sgcd^−/−) in the same dystrophic background for direct experimental comparison (5). All mouse models were viable, and there was no difference in total body weight between groups (Fig. 1C and fig. S1A), although muscle weights changed on the basis of the degree of ongoing dystrophic damage (fig. S1, B to E), see below.

Analysis of protein expression in isolated quadricep mitochondria showed a complete loss of ANT1 protein in Slc25a4^−/−;Sgcd^−/− mice, without an increase in ANT2 or ANT4 expression (Fig. 1C). However, we did observe an increase in CypD protein levels in Slc25a4^−/−;Sgcd^−/− mitochondria isolated from quadriceps (Fig. 1C). We also observed an increase in CypD protein in nondsytrophic Slc25a4^−/− quadriceps mitochondria, as well as an increase in ANT1 protein in nondystrophic Ppif^−/− quadriceps mitochondria, suggesting potential compensation between these two effectors in regulating MPTP activity (fig. S2A). ANT2 protein was not observed in isolated mitochondria from adult WT or Sgcd^−/− quadricep (Fig. 1C).

Muscle mitochondria from these same groups of mice were assessed for CRC and absorbance-based mitochondrial swelling. We observed that quadricep-derived mitochondria from Slc25a4^−/−;Sgcd^−/− mice had increased CRC and delayed swelling compared to Sgcd^−/− mitochondria (Fig. 1, D and E), suggesting that MPTP activation is desensitized compared to mitochondria from Sgcd^−/− muscle. Slc25a4^−/−;Sgcd^−/− mitochondria CRC levels and swelling were indistinguishable from Ppif^−/−;Sgcd^−/− mitochondria, demonstrating that Slc25a4 gene deletion yielded similar MPTP desensitization to Ppif gene deletion as previously reported (5). Consistent with the lack of ANT2 protein expression in mouse skeletal muscle, CRC and swelling in isolated mitochondria from Slc25a5^MKO;Sgcd^−/− muscle were identical to Sgcd^−/− mitochondria (Fig. 1, D and E). Direct analysis of mitochondrial matrix Ca²⁺ in quadricep from these groups showed that both Slc25a4^−/−;Sgcd^−/− and Ppif^−/−;Sgcd^−/− mice released significantly more Ca²⁺ than either WT or Sgcd^−/− muscle mitochondria (Fig. 1F), suggesting some degree of mitochondrial Ca²⁺ overload despite MPTP desensitization. This result suggests the CRC experiments (Fig. 1, D and E) likely underestimate the degree of muscle MPTP desensitization in Slc25a4^−/−;Sgcd^−/− and Ppif^−/−;Sgcd^−/− mice.

To assess the in vivo status of Slc25a4^−/−;Sgcd^−/− mitochondria, we performed transmission electron microscopy of quadricep muscle. As expected, intramuscle mitochondria from Sgcd^−/− mice showed swelling, disorganized cristae, and loss of membrane integrity (Fig. 1G). However, electron micrographs from quadriceps of both Slc25a4^−/−;Sgcd^−/− and Ppif^−/−;Sgcd^−/− mice showed improved mitochondria morphology, albeit with some baseline swelling. We also assessed mitochondrial metabolic function using a Seahorse bioanalyzer and found that Sgcd^−/− quadricep mitochondria had significantly reduced maximal oxygen consumption rates (OCRs) compared to WT mitochondria (fig. S2B). Unexpectedly, Slc25a4^−/−;Sgcd^−/− but not Ppif^−/−;Sgcd^−/− mitochondria had maximal OCR rates restored to WT levels under optimal metabolic substrate conditions (fig. S2B). Collectively, these results demonstrate that Slc25a4 (ANT1) gene deletion desensitizes MPTP activation and improves mitochondrial function in vivo in skeletal muscle with MD.

Quadricep-derived mitochondria from Ppif^−/−;Sgcd^−/− mice released significantly more Ca²⁺ than Slc25a4^−/−;Sgcd^−/− mitochondria (Fig. 1F). We therefore measured the CRC of quadricep-derived mitochondria from nondystrophic Slc25a4^−/− and Ppif^−/− mice and observed that while Slc25a4^−/− mitochondria had greater CRC than WT control mitochondria, they had significantly less CRC than Ppif^−/− mitochondria (fig. S2, C and D). These data suggest that the mitochondrial Ca²⁺ content observed in dystrophic mitochondria is due to differences in overall CRC capacity between these groups. To further investigate the greater levels of quadricep mitochondrial Ca²⁺ released from Ppif^−/−;Sgcd^−/− mice, we performed Western blotting for the Ca²⁺ handling effectors of the mitochondrial Ca²⁺ uniporter (MCU) complex. We observed that mitochondria from Slc25a4^−/−;Sgcd^−/− mice had increased levels of the dominant-negative MCU-complex subunit MCUb, as well as the MCU-complex gatekeeping protein MICU1 compared to Sgcd^−/− mitochondria (fig. S2E). Conversely, quadricep mitochondria from Ppif^−/−;Sgcd^−/− mice do not show increased levels of MCUb and low levels of MICU1, despite maintaining a robust level of MCU expression (fig. S2E). Both a reduction of MCUb (35, 36), and a decrease in the ratio of MICU1:MCU (37) are consistent with a more active MCU complex. This suggests that Ppif^−/−;Sgcd^−/− mitochondria, but not Slc25a4^−/−;Sgcd^−/− mitochondria, have a larger CRC and greater mitochondrial Ca²⁺ uptake, which likely underlies the markedly increased Ca²⁺ levels observed in the Ppif^−/−;Sgcd^−/− mitochondria (Fig. 1F).

To determine whether ANT-dependent MPTP formation contributed to in vivo necrosis in MD, we assessed several aspects of MD pathophysiology in these models. As alluded to above, we observed that pseudo-hypertrophy for the tibialis anterior (TA) and quadricep was reduced in Slc25a4^−/−;Sgcd^−/− and Ppif^−/−;Sgcd^−/− mice compared with Sgcd^−/− controls at 16 weeks of age, although there was no difference at 8 weeks (fig. S1, B to E). Histological analysis revealed that diaphragms and quadriceps from Slc25a4^−/−;Sgcd^−/− mice displayed substantially less pathology than Sgcd^−/− control or Slc25a5^MKO;Sgcd^−/− mice (Fig. 2A and fig. S3A). Specifically, diaphragms and quadriceps from Slc25a4^−/−;Sgcd^−/− mice did not have large expanding necrotic zones and instead had small localized regions of muscle fiber death. The same type of small localized fiber death was observed in Ppif^−/−;Sgcd^−/− muscle, while muscles from Slc25a5^MKO;Sgcd^−/− mice had large necrotic regions similar to Sgcd^−/− mice. Combinatorial deleted Slc25a4^−/−;Slc25a5^MKO;Sgcd^−/− (ANT1 and ANT2) mice and Slc25a4^−/−;Slc25a31;Sgcd^−/− (ANT1 and ANT4) mice showed the same amount of histopathology and myofiber central nucleation as Slc25a4^−/−;Sgcd^−/− mice (fig. S3, A and B), indicating that additionally targeting the ANT2 or ANT4 encoding genes did not alter necrosis in MD beyond the effect of ANT1 loss. Consistent with this, there was no additional increase in CRC in Slc25a4^−/−;Slc25a5^MKO;Sgcd^−/− or Slc25a4^−/−;Slc25a31;Sgcd^−/− isolated muscle mitochondria versus Slc25a4^−/−;Sgcd^−/− (fig. S3, C and D).

Fig. 2. Slc25a4 (ANT1) gene deletion reduces histopathology in MD.

(A) Representative hematoxylin and eosin (H&E)–stained diaphragm sections from 8-week-old mice of the indicated genotypes. Scale bar, 200 μm. (B) Quantification of myofiber central nucleation in quadricep sections from 8-week-old mice of the indicated genotypes of mice in the legend by color coding. WT n = 4; Sgcd^−/− n = 18; Slc25a4^−/−;Sgcd^−/− n = 22; Slc25a5^MKO;Sgcd^−/−, n = 7; Ppif^−/−;Sgcd^−/− n = 13 (*P ≤ 0.05 versus WT; #P ≤ 0.05 versus WT and Sgcd^−/−). (C) Quantification of myofiber central nucleation in quadricep sections from 16-week-old mice. WT n = 3; Sgcd^−/− n = 9; Slc25a4^−/−;Sgcd^−/− n = 21; Slc25a5^MKO;Sgcd^−/−, n = 6; Ppif^−/−;Sgcd^−/− n = 6 (*P ≤ 0.05 versus WT; #P ≤ 0.05 versus WT and Sgcd^−/−; §P ≤ 0.05 versus WT, Sgcd^−/−, and Slc25a4^−/−;Sgcd^−/−). (D) Average Feret’s diameter of quadricep myofibers in sections from 8-week-old mice of the indicated genotypes. WT n = 13; Sgcd^−/− n = 18; Slc25a4^−/−;Sgcd^−/− n = 21; Slc25a5^MKO;Sgcd^−/−, n = 7; Ppif^−/−;Sgcd^−/− n = 8 (*P ≤ 0.05 versus WT). (E) Average Feret’s diameter of quadricep myofibers in sections from 16-week-old mice of the indicated genotypes. WT n = 7; Sgcd^−/− n = 9; Slc25a4^−/−;Sgcd^−/− n = 15; Slc25a5^MKO;Sgcd^−/−, n = 6; Ppif^−/−;Sgcd^−/− n = 6 (*P ≤ 0.05 versus WT; #P ≤ 0.05 versus WT and Sgcd^−/−).

To better characterize the degree of MD protection associated with Slc25a4 deletion, we quantified the rate of myofiber central nucleation and observed that quadricep from Slc25a4^−/−;Sgcd^−/− mice had increased levels compared with WT but significantly less than Sgcd^−/− or Slc25a5^MKO;Sgcd^−/− mice at both 8 and 16 weeks of age (Fig. 2, B and C). Central nucleation in Slc25a4^−/−;Sgcd^−/− quadricep was also significantly lower than in Ppif^−/−;Sgcd^−/− muscle, suggesting that Slc25a4 gene deletion may be more protective than Ppif gene deletion in MD. Consistent with this, we observed that Slc25a4^−/−;Sgcd^−/− quadricep had partially preserved myofiber size that was not observed in Ppif^−/−;Sgcd^−/− muscle (Fig. 2, D and E). Both Slc25a4^−/−;Sgcd^−/− and Ppif^−/−;Sgcd^−/− quadriceps displayed significantly reduced embryonic myosin heavy chain (Myh3) expression (Fig. 3, A and B), which is another indicator of reduced myofiber degeneration-regeneration. Similarly, both Slc25a4^−/−;Sgcd^−/− and Ppif^−/−;Sgcd^−/− quadriceps showed lower rates of acute Evan’s blue dye (EBD) uptake following a downhill treadmill challenge compared with Sgcd^−/− or Slc25a5^MKO;Sgcd^−/− mice, demonstrating increased myofiber membrane stability (Fig. 3C). Changes in muscle histology, central nucleation, or myofiber size were not observed in nondystrophic Slc25a4^−/− or Ppif^−/− control animals (fig. S4, A to C). Collectively, these data demonstrate that Slc25a4 (ANT1) gene deletion reduces MD pathology and in vivo necrosis.

Fig. 3. Slc25a4 (ANT1) gene deletion reduces muscle degeneration in MD.

(A) Quadricep sections immunostained for embryonic myosin heavy chain (Myh3; red) and costained with wheat germ agglutinin (WGA; green) from 8-week-old mice of the indicated genotypes. Scale bar, 1 mm. (B) Quantification of Myh3⁺ area in quadricep sections in 8-week-old mice of the indicated genotypes. WT n = 6; Sgcd^−/− n = 5; Slc25a4^−/−;Sgcd^−/− n = 8; Slc25a5^MKO;Sgcd^−/−, n = 6; Ppif^−/−;Sgcd^−/− n = 4 (*P ≤ 0.05 versus WT; #P ≤ 0.05 versus Sgcd^−/−). (C) Quantification of EBD-positive area in quadricep sections in 8-week-old mice of the indicated genotypes. WT n = 4; Sgcd^−/− n = 6; Slc25a4^−/−;Sgcd^−/− n = 17; Slc25a5^MKO;Sgcd^−/−, n = 5; Ppif^−/−;Sgcd^−/− n = 11 (*P ≤ 0.05 versus WT). (D) Average treadmill running distance in meters for 8-week-old mice of the indicated genotypes. Maximum possible running distance is 270 m. WT n = 4; Sgcd^−/− n = 7; Slc25a4^−/−;Sgcd^−/− n = 20; Slc25a5^MKO;Sgcd^−/−, n = 5; Ppif^−/−;Sgcd^−/− n = 9 (*P ≤ 0.05 versus WT; #P ≤ 0.05 versus Sgcd^−/−). (E) Average treadmill running distance in meters for 16-week-old mice of the indicated genotypes. WT n = 4; Sgcd^−/− n = 6; Slc25a4^−/−;Sgcd^−/− n = 18; Slc25a5^MKO;Sgcd^−/−, n = 8; Ppif^−/−;Sgcd^−/− n = 13 (*P ≤ 0.05 versus WT; #P ≤ 0.05 versus Sgcd^−/−).

The impact of Slc25a4 gene deletion on muscle performance in MD was examined. Sgcd^−/− dystrophic mice ran a significantly shorter distance before exhaustion on a treadmill compared with WT mice at 8 and 16 weeks of age (Fig. 3, D and E). Slc25a4^−/−;Sgcd^−/− and Ppif^−/−;Sgcd^−/− mice ran significantly further than Sgcd^−/− control animals (Fig. 3, D and E), however neither group were able to complete the running protocol unlike healthy WT mice (Fig. 3, D and E). This is likely due to metabolic dysfunction as Slc25a4 gene deletion causes exercise intolerance and mitochondrial myopathy in mice and humans (32, 38, 39), and Ppif gene deletion in mice causes cardiac metabolic dysfunction and vulnerability to cardiac stress (40). Both nondystrophic Slc25a4^−/− and Ppif^−/− control animals struggle to complete the downhill treadmill running protocol in contrast to WT mice (fig. S4D). These experiments reveal that while Slc25a4 (ANT1) gene deletion reduces MD pathology and provides some functional muscle improvement, metabolic exercise intolerance likely prevents greater rescue of muscle performance.

In our previous cell-reductionist work, we observed that the full extent of MPTP activity required ANT and another component separately regulated by CypD (14). Here, we observed that CypD protein levels were substantially increased in Slc25a4^−/− mitochondria from dystrophic muscle, consistent with this dual regulation model (Fig. 1C and fig S2A). Skeletal muscle mitochondria from Slc25a4^−/−;Sgcd^−/− mice showed a synergistically greater CRC when treated with CsA to inhibit CypD, and mitochondrial swelling was similarly desensitized (Fig. 4, A and B). We therefore examined mice in which the Slc25a4 gene was deleted in combination with the Ppif gene in the Sgcd^−/− background. Consistent with these results, mitochondria isolated from skeletal muscle of Slc25a4^−/−;Ppif^−/−;Sgcd^−/− mice demonstrated the same synergistic increase in CRC and swelling desensitization as the CsA-treated Slc25a4^−/−;Sgcd^−/− muscle mitochondria (Fig. 4, A and B). Analysis of matrix Ca²⁺ release from Slc25a4^−/−;Ppif^−/−;Sgcd^−/− muscle mitochondria was markedly higher compared with WT or Sgcd^−/− muscle-derived mitochondria (Fig. 4C). This indicates that Slc25a4^−/−;Ppif^−/−;Sgcd^−/− mitochondria are in a state of Ca²⁺ overload and that the CRC data likely under-represent the true desensitization of the MPTP, which is almost completely inhibited.

Fig. 4. Combined Slc25a4 and Ppif gene deletion synergistically desensitizes MPTP activity in dystrophic muscle mitochondria.

(A) Representative CRC assay (top) and parallel absorbance-based mitochondrial swelling (bottom) of quadricep mitochondria from WT, Sgcd^−/−, CsA-treated Slc25a4^−/−;Sgcd^−/− mice, and Slc25a4^−/−;Ppif^−/−;Sgcd^−/− mice. Two milligrams of mitochondria was given repetitive 20 μM additions of CaCl₂ at each inflection point. The genotypes of mice are color coded in the legend. The y axis is arbitrary units or relative units (RU). (B) Average CRC measurements of total Ca²⁺ uptake before cessation from quadricep mitochondria of the indicated genotypes of mice. WT n = 10; Sgcd^−/− n = 4; Slc25a4^−/−;Sgcd^−/− + CsA n = 3; Slc25a4^−/−;Ppif^−/−;Sgcd^−/− n = 3 (*P ≤ 0.05 versus Sgcd^−/−). (C) Mitochondrial Ca²⁺ assay of quadricep mitochondria from mice of the indicated genotypes. One milligram of mitochondria was treated with 40 μM alamethicin (red arrow). The y axis is arbitrary units. WT n = 6; Sgcd^−/− n = 3; Slc25a4^−/−;Ppif^−/−;Sgcd^−/− n = 5 (*P ≤ 0.05 versus WT). (D) Representative transmission electron micrographs of quadricep ultrastructure (top) and close-up of mitochondria (bottom) from 8-week-old mice. Scale bars, 2 μm (top) and 800 nm (bottom). (E) Western blots from isolated quadricep mitochondria for ANT1, ANT2, ANT4, CypD, VDAC1, and the protein-loading controls (CII in the figure). Red box indicates induction of ANT2. (F) Oxygen consumption rate (OCR) of quadricep mitochondria from 8-week-old mice of the indicated genotypes. Respirating mitochondria were treated sequentially with 10 mM adenosine diphosphate (ADP), 2 μM oligomycin (Oligo), 5 μM carbonyl cyanide-p-trifluoromethoxyphenylhydrazone (FCCP), and 0.5 μM of a mixture of rotenone and antimycin A (Rote/AA) in a Seahorse bioanalyzer. WT n = 4; Sgcd^−/− n = 4; Slc25a4^−/−;Ppif^−/−;Sgcd^−/− n = 4 (*P ≤ 0.05 versus WT).

In vivo examination of skeletal muscle from Slc25a4^−/−;Ppif^−/−;Sgcd^−/− mice by transmission electron microscopy revealed the same state of mitochondrial swelling observed in Slc25a4^−/−;Sgcd^−/− and Ppif^−/−;Sgcd^−/− mice but with substantially improved morphology, cristae organization, and membrane integrity compared to Sgcd^−/− dystrophic mitochondria (Fig. 4D). Unexpectedly, we observed that quadricep mitochondria from Slc25a4^−/−;Ppif^−/−;Sgcd^−/− mice uniquely showed up-regulation of ANT2 protein (Fig. 4E, red box), suggesting potential metabolic compensation. Measurement of mitochondrial respiration using a Seahorse analyzer showed reduced capacity in Sgcd^−/− mitochondria compared with WT but that mitochondria from quadricep of Slc25a4^−/−;Ppif^−/−;Sgcd^−/− mice had rescued maximal respiration under ideal substrate conditions (Fig. 4F). Collectively, these experiments show that combined Slc25a4 (ANT1) and Ppif (CypD) gene deletion causes a “synergistic” desensitization of MPTP opening not observed in muscle from either Slc25a4^−/−;Sgcd^−/− or Ppif^−/−;Sgcd^−/− mice. The unique induction of the ANT2 expression by double deletion of ANT1 and CypD encoding genes likely explains the enhanced metabolic function and other improvements (see below).

The electron micrographs from Slc25a4^−/−;Ppif^−/−;Sgcd^−/− mice appeared to show an expansion both in the number and size of quadricep mitochondria compared to both Sgcd^−/− and WT controls (Fig. 4D). We therefore quantified mitochondrial DNA (mtDNA) content and observed reductions in mtDNA from Sgcd^−/− quadricep compared to WT, but that muscle from both Slc25a4^−/− and Slc25a4^−/−;Ppif^−/−;Sgcd^−/− mice had significantly more mtDNA than Sgcd^−/− quadricep tissue (fig. S5A). Furthermore, we observed that Slc25a4^−/−;Ppif^−/−;Sgcd^−/− had more mtDNA than WT quadricep tissue (fig. S5A), suggesting that mitochondrial content is preserved but also expanded in the Slc25a4^−/−;Ppif^−/−;Sgcd^−/− mice. To assess whether the increase in mtDNA was due to increased mitochondrial biogenesis, we measured peroxisome proliferator-activated receptor-gamma coactivator (PGC)-1α (PGC1α), nuclear respiratory factor 1 (NRF1), and transcription factor A, mitochondrial (TFAM) levels in quadricep tissue lysates, which were not substantially altered between the genotypes examined (fig. S5B). ANT proteins have also been shown to promote PTEN-induced kinase 1 (PINK1)-dependent mitophagy (41), hence we also measured markers of mitophagy in isolated muscle mitochondria. We observed that mitochondria from Slc25a4^−/−;Sgcd^−/− and Slc25a4^−/−;Ppif^−/−;Sgcd^−/− mice had sharply increased levels of PINK1, PARKIN, and p62 compared to both Sgcd^−/− and WT mitochondria (fig. S5C). These results are most consistent with impaired mitophagy wherein mitochondria are being marked for degradation but are unable to complete the process, resulting in an accumulation and enlargement of the mitochondrial compartment.

We next assessed whether combined Slc25a4 and Ppif gene deletion improved MD pathology and muscle function. We observed that histopathology in both the diaphragm and quadricep was markedly improved in Slc25a4^−/−;Ppif^−/−;Sgcd^−/− mice compared with Sgcd^−/− controls (Fig. 5A). Specifically, regions of active muscle necrosis in Slc25a4^−/−;Ppif^−/−;Sgcd^−/− mice were not observed, except in rare small, localized instances. Quantification showed that quadriceps from Slc25a4^−/−;Ppif^−/−;Sgcd^−/− mice had low levels of myofiber central nucleation that were significantly lower than those of both Sgcd^−/− and Slc25a4^−/−;Sgcd^−/− mice (Fig. 5B and fig. S5D). Furthermore, muscle from Slc25a4^−/−;Ppif^−/−;Sgcd^−/− mice had significantly improved myofiber size distribution characterized by a loss of very small regenerating fibers (0 to 20 μm) and a significant preservation of larger fibers (40 to 60 μm) (Fig. 5C), and a near complete loss of Myh3 expression in muscle histological sections, suggesting substantially less necrotic cell death compared with Sgcd^−/− controls (Fig. 5, D and E). Immunoglobulin M (IgM)–positive myofibers in Slc25a4^−/−;Ppif^−/−;Sgcd^−/− quadriceps were also not readily observed, indicating a near-complete absence in necrotic myofibers compared in Sgcd^−/− mice when the genes for ANT1 and CypD were deleted (Fig. 5F). To investigate whole-body muscle damage, we also measured serum levels of muscle creatine kinase (CK) that is released by necrotic muscle fibers in MD. We observed that Sgcd^−/− mice had significantly elevated levels of serum CK compared to WT mice and that Slc25a4^−/−;Ppif^−/−;Sgcd^−/− mice had CK levels significantly lower than Sgcd^−/− mice, comparable to WT controls (Fig. 6A). We also measured muscle performance by downhill treadmill running and observed that Slc25a4^−/−;Ppif^−/−;Sgcd^−/− mice ran significantly further than Sgcd^−/− mice and were not statistically different from WT (Fig. 6B and fig. S5E). Isometric force production showed that the TA muscle from Slc25a4^−/−;Ppif^−/−;Sgcd^−/− rescued the reduction in active force observed in Sgcd^−/− mice, which was indistinguishable from WT controls (Fig. 6C). These data suggest that the combined Slc25a4 (ANT1) and Ppif (CypD) gene deletion virtually eliminates ongoing necrosis and MD pathology in the Sgcd^−/− background.

Fig. 5. Combined Slc25a4 (ANT1) and Ppif (CypD) gene deletion prevents tissue histopathology and myofiber necrosis and normalizes muscle function in MD.

(A) H&E-stained quadricep (left) and diaphragm (right) histological sections from 8-week-old mice of the indicated genotypes. Scale bar, 200 μm. (B) Quantification of myofiber central nucleation in quadricep sections from 8-week-old mice with the indicated genotypes by color coding to the legend. WT n = 4; Sgcd^−/− n = 18; Slc25a4^−/−;Sgcd^−/− n = 22; Ppif^−/−;Sgcd^−/− n = 13; Slc25a4^−/−;Ppif^−/−;Sgcd^−/− n = 15 (*P ≤ 0.05 versus WT; #P ≤ 0.05 versus WT and Sgcd^−/−; &P ≤ 0.05 versus Sgcd^−/−, Ppif^−/−;Sgcd^−/−, and Slc25a4^−/−;Sgcd^−/−). (C) Average Feret’s diameter of myofibers measured in quadricep sections from 8-week-old mice of the indicated genotypes. The y axis is percentage of total fiber diameters for each genotype up to 100% total. WT n = 13; Sgcd^−/− n = 18; Slc25a4^−/−;Ppif^−/−;Sgcd^−/− n = 14 (*P ≤ 0.05 versus WT). (D) Quadricep histological sections immunostained for Myh3 (red) and costained with WGA (green) from 8-week-old mice of the indicated genotypes. Scale bar, 1 mm. (E) Quantification of the percent Myh3-positive area in quadricep sections taken from 8-week-old mice of the indicated genotypes. WT n = 6; Sgcd^−/− n = 5; Slc25a4^−/−;Ppif^−/−;Sgcd^−/− n = 5 (*P ≤ 0.05 versus WT). (F) Quantification of IgM-positive area in quadricep histological sections taken from 8-week-old mice of the indicated genotypes. WT n = 13; Sgcd^−/− n = 21; Slc25a4^−/−;Sgcd^−/− n = 19; Ppif^−/−;Sgcd^−/− n = 8; Slc25a4^−/−;Ppif^−/−;Sgcd^−/− n = 10 (*P ≤ 0.05 versus WT; #P ≤ 0.05 versus Sgcd^−/−).

Fig. 6. Slc25a4 / Ppif double-null mice show reduced necrosis and normalized muscle function in MD.

(A) Serum CK levels for 8-week-old mice of the indicated genotypes. The y axis is units of CK per liter of serum. WT n = 4; Sgcd^−/− n = 14; Slc25a4^−/−;Sgcd^−/− n = 8; Ppif^−/−;Sgcd^−/− n = 14; Slc25a4^−/−;Ppif^−/−;Sgcd^−/− n = 15 (*P ≤ 0.05 versus WT; #P ≤ 0.05 versus Sgcd^−/−). (B) Average treadmill running distance in meters for 8-week-old mice of the indicated genotypes. Maximum possible distance is 270 m. WT n = 4; Sgcd^−/− n = 7; Slc25a4^−/−;Ppif^−/−;Sgcd^−/− n = 14 (*P ≤ 0.05 versus WT). (C) Measurement of isometric force in millinewtons generated from the TA muscle of 8-week-old mice with the indicated genotypes. WT n = 8; Sgcd^−/− n = 5; Slc25a4^−/−;Ppif^−/−;Sgcd^−/− n = 5 (*P ≤ 0.05 versus WT).

DISCUSSION

This study contains direct evidence that ANT is a core MPTP component in mediating necrotic cell death in vivo. Slc25a4 gene deletion desensitized MPTP activation in isolated diseased muscle mitochondria and markedly reduced MD pathology in vivo. These findings are consistent with a pore-forming role for ANT isoforms in the MPTP that most likely responds to Ca²⁺ overload, although reactive oxygen species generation could also be involved (42–44), both of which are reported to underlie myofiber loss in MD (45, 46). Our results further demonstrate that ANT proteins function in concert with an unknown CypD-dependent protein species in generating MPTP activity, which we and others have previously proposed on the basis of studies in cells or isolated mitochondria (14, 22). Here, genetically targeting both Slc25a4 and Ppif provides synergistic desensitization of MPTP activation that almost completely prevents necrosis in MD.

It is interesting that deletion of the genes encoding ANT1 and CypD provide some level of sarcolemmal protection compared with Sgcd^−/− controls even though these animals still lacked the membrane stabilizing δ-sarcoglycan protein. We believe this is due to less mitochondrial dysfunction caused by reduced MPTP activation in these groups, because MPTP activation can generate reactive oxygen species leading to secondary sarcolemmal instability (47, 48). While CsA and its derivative compounds effectively inhibit CypD-dependent MPTP in vivo and are protective in animal models of MD (5, 27–29), these agents are needed at high concentrations and can have serious side effects with chronic use (49). Our results suggest that ANT inhibitors could also be used as a therapeutic in MD if a nontoxic agent could be identified (50). However, another therapeutic strategy could be to combine appropriate CypD and ANT inhibitors, potentially at lower dosages and with synergistic benefits, either independently or in combination with other gene therapy strategies (51).

Our results show that Slc25a4 gene–deleted mice have metabolic dysfunction as observed in previous studies (32). Mitochondrial myopathies in humans caused by autosomal dominant mutations in Slc25a4 are characterized by exercise intolerance and progressive external ophthalmoplegia that manifests in middle age (39, 52, 53). There are documented cases of humans with homozygous Slc25a4 loss-of-function mutations that cause childhood exercise intolerance, mild muscle myopathy, and mild hypertrophic cardiomyopathy, although by 21 and 25 years of age the patients did not yet present with cardiac disease manifestations (38, 54), leaving open the possibility that a properly designed nontoxic ANT1-specific antagonist might offer some therapeutic benefit to skeletal muscle necrotic disease in MD.

ANT proteins can be involved in mitophagy (41), and our results provide further evidence for this concept. While more complex experiments that directly assess mitophagic flux are needed, we believe that the increase in muscle mitochondrial content coupled with the increase in mitophagy markers in Slc25a4^−/−;Sgcd^−/− and Slc25a4^−/−;Ppif^−/−;Sgcd^−/− mice reflects a deficit in mitophagy wherein mitochondria are marked for degradation by PINK1 but are unable to complete the process and therefore accumulate (Fig. 4D and fig. S5, A to C). The increase in mtDNA is greater in muscle from Slc25a4^−/−;Ppif^−/−;Sgcd^−/− mice than Slc25a4^−/−;Sgcd^−/− mice, suggesting the possibility that MPTP activity itself may be important to mitophagy. It has been previously observed that CypD inhibition with CsA can reduce mitophagy in some contexts (55–57).

Last, results shown here add evidence to the ongoing controversy as to the composition of the MPTP. ANT1 and CypD each separately have some role in facilitating MPTP activity leading to myofiber death in MD. Previous research showed that MPTP has both an ANT-dependent component and an unknown component that requires CypD (14). Here, genetically targeting both Slc25a4 (ANT1) and Ppif (CypD) provides synergistic desensitization of MPTP activation that almost completely prevents necrosis in MD. This finding is consistent with previous work and confirms that MPTP either operates by the proposed two-component MPTP model (19) or functions by a dual-activation model wherein ANT or CypD are required to activate a single, still unidentified MPTP pore. Another model has been proposed whereby the F₁F₀ATPase can generate MPTP activity in cells and isolated mitochondria (20, 21). However, the concept of the mitochondrial ATPase serving as a direct component of the MPTP in vivo has not been evaluated in a relevant disease setting like MD; thus, the physiologic relevance of the mitochondrial ATPase as a bona fide MPTP component in vivo remains untested.

MATERIALS AND METHODS

Animal models

Germline Slc25a4^−/− (ANT1 protein), Ppif^−/− (CypD protein), and Slc25a31^−/− (ANT4 protein) mice have been described previously (17, 32, 58). To target the ANT2 encoding gene in a muscle-specific manner, Slc25a5-LoxP (fl) allele–targeted mice (Slc25a5^fl/fl) (13) were crossed with a line expressing Cre recombinase under the control of the muscle-specific myosin, light polypeptide 1 promoter (Myl1^{tm1(cre)Sjb/J}) (59). The combination of the Slc25a5^fl/fl alleles with the Myl1-Cre line was abbreviated Slc25a5^MKO. Because both Slc25a5 and the dystrophin gene (Dmd) reside on the X chromosome, we used a δ-sarcoglycan gene–deleted mouse model (Sgcd^−/−) of MD throughout this study (30). To assess the effect of deleting the genes encoding ANT1 and CypD in combination, we generated dystrophic mice lacking both genes (Slc25a4^−/−;Ppif^−/−;Sgcd^−/−). To assess the role of different ANT isoforms, we generated dystrophic mice lacking Slc25a4 and Slc25a5 (Slc25a4^−/−;Slc25a5^MKO;Sgcd^−/−) or Slc25a4 and Slc25a31 (Slc25a4^−/−;Slc25a31^−/−;Sgcd^−/−) by crossing the relevant mouse lines.

All experimental procedures with animals were approved by the Institutional Animal Care and Use Committee of Cincinnati Children’s Medical Center, protocols IACUC 2019-0047 and 2021-0047. We have complied with the relevant ethical considerations for animal usage overseen by this committee. The number of mice used in this study reflects the minimum number needed to achieve statistical significance based on experience and previous statistical power analysis. Blinding was performed for some experimental procedures with mice, although blinding was not possible in every instance. Only male mice were used in this study to model the human disease in X-linked Duchenne MD that afflicts primarily males.

Mitochondrial isolation

Quadricep muscle mitochondria were isolated in MS-EGTA buffer [225 mM mannitol, 75 mM sucrose, 5 mM Hepes, and 1 mM EGTA (pH 7.4); Sigma-Aldrich]. Quadricep tissue was minced into 2-mm by 2-mm pieces in 4 ml of MS-EGTAS supplemented with trypsin (0.2 mg/ml; Worthington; 3703; 187 U/mgP) and digested for 5 min. Digestion was stopped by the addition of 10 ml of 0.2% bovine serum albumin (BSA) MS-EGTA. Seven milliliters of the resulting buffer containing the majority of tissue was then homogenized using a glass and Teflon Potter-Elvehjem tissue homogenizer on ice (8 to 15 strokes, depending on the tissue). Tissue homogenates were then subjected to a 2000g centrifugation for 5 min (1×), and the supernatants were then centrifuged at 11,500g for 10 min. Following centrifugation, mitochondria were suspended in KCl buffer [125 mM KCl, 20 mM Hepes, 2 mM MgCl₂, 2 mM KH₂PO₄, and 40 μM EGTA (pH 7.2); Sigma-Aldrich] and either used in a functional assay or spun down (5000g for 5 min) and snap frozen in liquid nitrogen for further analysis.

Western blotting

All Western blots were performed using isolated mitochondrial lysates unless otherwise specified in the figure caption. Following mitochondrial isolation, the snap-frozen mitochondrial pellets were suspended in radioimmunoprecipitation assay buffer (RIPA buffer) [10 mM tris-HCl, 150 mM NaCl, 4% glycerol, 0.5 mM NaMetabisulfite, 1% Triton X-100, 0.1% NaDeoxycholate, and 0.05% SDS (pH 7.5); Sigma-Aldrich] containing protease inhibitor cocktails (Roche; 4693124001). The samples were then sonicated and centrifugated to remove the insoluble fraction. The insoluble fractions were discarded following centrifugation. SDS sample buffer [50 mM tris-HCl, 2% SDS, 10% glycerol, 1% β-mercaptoethanol, 12.5 mM EDTA, and 0.02% bromophenol blue (pH 6.8); Sigma-Aldrich] was added to the lysates, and samples were boiled for 5 min. The samples were then loaded onto 10% acrylamide gels and transferred onto polyvinylidene fluoride transfer membranes (MilliporeSigma). For Western blots of whole quadricep, the tissue was snap frozen in liquid nitrogen and then dry homogenized using a chilled mortar and pestle. The resulting muscle homogenate powder was resuspended in RIPA buffer and then treated as above to generate Western samples. The following commercially available primary antibodies were used: Slc25a4 (ANT1) (Signalway Antibody; 32484; 1:500), Slc25a5 (ANT2) (Cell Signaling Technology; 14671; 1:500), Slc25a31 (ANT4) polyclonal antibody (Signalway Antibody; 40596-1; 1:500), Ppif (CypD) monoclonal antibody (Affinity BioReagents; PA1-028; 1:1000), PGC1α/β monoclonal antibody (Abcam; ab72230; 1:1000), NRF1 polyclonal antibody (Abcam; ab34682; 1:1000), TFAM polyclonal antibody (Proteintech; 22586-1-AP; 1:1000), PINK1 polyclonal antibody (Cayman Chemical; 10006283; 1:1000), PARKIN monoclonal antibody (Cell Signaling Technology; 4211; 1:1000), p62 polyclonal antibody (Sigma-Aldrich; P0067; 1:1000), MCU monoclonal antibody (Cell Signaling Technology; 14997; 1:1000), MICU1 monoclonal antibody (Cell Signaling Technology; 12524; 1:500), VDAC1 monoclonal antibody (Abcam; ab14734; 1:1000), glyceraldehyde-3-phosphate dehydrogenase monoclonal antibody (Fitzgerald; 0RG109A; 1:5000), and Total OXPHOS rodent WB antibody cocktail (that contained antibodies to electron transport chain complexes 1 to 5; specifically succinate dehydrogenase complex iron-sulfur subunit B from complex II was measured) (Abcam; ab110413; 1:1000). In addition, we used a custom-made MCUb rabbit primary polyclonal antibody (produced by YenZym Antibodies LLC; 1:250) (36). Fluorescence-based secondary antibodies were used against different species of antibodies and incubated at room temperature for 1 hour (LI-COR Biosciences; 926-32211 or 925-68070; 1:10,000). Membranes were imaged using an Odyssey CLx Imaging System (LI-COR Biosciences).

Mitochondrial CRC, swelling, and calcium overload assays

The CRC assay and the mitochondrial swelling assay were performed simultaneously using a dual-detector (one to measure fluorescence and the other to measure absorbance), single-cuvette–based fluorimetric system (Horiba Scientific, PTI QM-800). Two milligrams of isolated quadricep mitochondria was loaded into the cuvette along with 250 nM Calcium Green-5 N (Invitrogen; C3737), 7 mM pyruvate (Sigma-Aldrich), and 1 mM malate (Sigma-Aldrich) and brought up to 1 ml using KCl buffer (as above). Mitochondria were then pulsed with sequential additions of CaCl₂ (20 μM) until MPTP opening occurred as assessed by mitochondrial Ca²⁺ release and/or swelling.

Calcium overload was measured using the above system and experimental conditions; however mitochondria were suspended in modified KCl buffer lacking EGTA [125 mM KCl, 20 mM Hepes, 2 mM MgCl₂, and 2 mM KH₂PO₄ (pH 7.2); Sigma-Aldrich]. To assess free matrix Ca²⁺, 1 or 2 mg of isolated quadricep mitochondria was treated with a membrane-permeabilizing agent, 40 μM alamethicin (Santa Cruz Biotechnology). The relative amount of matrix Ca²⁺ was assessed by the increase in Calcium Green fluorescence following permeabilization.

Metabolic analysis

OCR was determined using an XF24 Extracellular Flux Analyzer (Agilent Technologies, Seahorse). Isolated quadricep mitochondria (100 μg) were loaded onto XF24 cell culture microplates in KCl buffer containing 7 mM pyruvate (Sigma-Aldrich) and 1 mM malate (Sigma-Aldrich) and subjected to the Mito Stress Test Kit using the standard protocol provided (Agilent Technologies; 103015-100). Briefly, after basal respiration was measured, the mitochondria were treated sequentially with 10 mM adenosine diphosphate (Sigma-Aldrich), 2 μM oligomycin (Agilent Technologies), 5 μM carbonyl cyanide-p-trifluoromethoxyphenylhydrazone (Agilent Technologies), and 0.5 μM of a mixture of rotenone and antimycin A (Agilent Technologies).

mtDNA quantification

mtDNA was measured using quantitative polymerase chain reaction (qPCR) methodology (60). DNA was isolated from ~10 mg of quadricep tissue using a KingFisher Flex benchtop automated extraction instrument (Thermo Scientific). Briefly, tissue was digested overnight at 55°C in Nuclei Lysis Solution (Promega; PR-A7943) supplemented with 0.1 M EDTA (VWR), Proteinase K (1 mg/ml; Thomas Scientific; C755F45), and RNAse A (2 mg/ml; Qiagen; 19101). To capture nuclear DNA and mtDNA, digested tissue lysates were incubated with MagneSil Paramagnetic Particles (Promega; MD1370) suspended in Lysis Buffer, Blood solution (Promega; PRM1392). DNA-complexed MagneSil PMPs were then loaded into a KingFisher Flex system, washed three times with Alcohol Wash, Blood solution (Promega; MD1412), and eluted into tris-EDTA buffer at pH 8. Isolated quadricep DNA was then diluted to 20 ng/μl, and qPCR was conducted on both Cox1 to measure mtDNA content (primers: F: 5′-TCTGTTCTGATTCTTTGGGCAC-3′; R: 5′-AGCCAATAGACATTATTGCTCATAC-3′) and Actb to measure genomic content (primers: F: 5′-CTGTCGAGTCGCGTCCACC-3′; R: 5′-CAGTGAGGTACTAGCCACGAGA-3′) for each sample using the SsoAdvanced Universal SYBR qPCR Green Supermix reagent (Bio-Rad; 1725274). The qPCR used the following protocol: a 98°C 3-min incubation followed by 40× cycles of 98°C for 15 s and 60°C for 25 s in a Bio-Rad CFX96 Real-Time System Thermocycler (Bio-Rad). mtDNA (Cox1) was quantified using the Pffafl method (61) with Atcb as the reference control gene and then normalized relative to WT levels.

Immunohistochemistry and histological analysis

Histological sections (10 μm) were collected from frozen skeletal muscles using a cryostat. For laminin/IgM staining frozen sections were fixed in 4% paraformaldehyde for 10 min, washed in 1× phosphate buffer saline (1× PBS), and maintained in blocking buffer for 30 min [10% goat serum and 0.4% Triton X diluted in 1× PBS]. Slides were stained overnight at 4°C with antilaminin antibody (Sigma-Aldrich; L9393; 1:200) diluted in staining solution [1% BSA, 0.04% Triton X diluted in 1× PBS]. Slides were then stained with IgM primary antibody conjugated to fluorescein isothiocyanate (Sigma-Aldrich; F9259; 1:100) and Alexa Fluor 568 goat anti-rabbit IgG secondary antibody (Invitrogen, A-11036; 1:300) for 1 hour at room temperature (RT). Nuclei were visualized by a 15-min 4′,6-diamidino-2-phenylindole (DAPI) stain (Invitrogen; D3571; 1:5000). For Myh3 staining, frozen muscle sections were fixed with cold acetone for 10 min, air dried for 30 min, washed with 1× PBS, and blocked with 2% BSA in 1× PBS for 1 hour. Slides were stained overnight at 4°C with anti-Myh3 antibody (Developmental Studies Hybridoma Bank; F1.652, 1:20) diluted in staining solution. Slides were then stained with Alexa Fluor 488 goat anti-mouse IgG1 secondary antibody (Invitrogen; A-21121; 1:400) for 1 hour at RT. Slides were subsequently stained with wheat germ agglutinin (WGA) conjugated to Alexa Fluor 488 (Invitrogen, W11261; 1:200) and DAPI for 20 min to visualize myofiber outlines and nuclei. Immunofluorescence images were captured using a Nikon Eclipse Ti microscope using NIS Elements AR 4.13 software. Regenerating myofibers were assessed by measuring Myh3-positive muscle area from an entire quadricep section at mid-belly. IgM-positive muscle area was assessed from entire quadricep sections taken at the mid-belly of the muscle. All analyses were performed using NIS Elements AR 4.13 software and conducted in a blinded fashion whereby the experimenter was only aware of the genotypes following quantification.

Histological cross sections (10 μm) were collected from quadricep or diaphragm using a cryostat and stained for hematoxylin and eosin to assess mitochondrial muscle myopathy. The rate of myofiber central nucleation was quantified from one 400× magnification micrograph taken from histological sections of the quadricep at the mid-belly. One thousand fibers were measured per quadricep section. The minimal feret’s diameter was determined from laminin-stained whole quadricep muscle using NIS Elements AR 4.13 software. One thousand fibers were measured per quadricep section. All analyses were performed in a blinded fashion whereby the experimenter was only aware of the genotypes following quantification.

Forced treadmill running and EBD uptake

To assess the exercise capacity of mice and sarcolemma stability, mice were subjected to forced treadmill running in the presence of EBD as previously described (62). Briefly, adult mice were intraperitoneally injected with EBD (10 mg/ml; 0.1 ml per 10-g body weight) and 24 hours later subjected to forced downhill treadmill running to measure membrane rupture events. Mice were run in an electrically driven six-lane treadmill (Columbus Instruments International; Exer-3/6 Treadmill) at a 5° decline. Mice were given a 10-min training regimen at 6 m/min to familiarize them with the environment and shock grids adjustable from 0 to 2.0 mA. The speed was then increased in increments of 2 m/min every 3 min to a maximum speed of 18 m/min. Exhaustion was assessed as greater than 10 consecutive seconds on the shock grid without attempting to reengage the treadmill which was considered an experimental endpoint. Mice were euthanized 24 hours following the running experiment and isolated muscles were frozen for histological analysis. Histological cross sections (10 μm) were collected from quadricep using a cryostat and subsequently stained with Alexa Fluor 488 WGA and DAPI (as above) for 20 min to visualize myofiber outlines and nuclei, respectively, as described above. Immunofluorescence images were captured using a Nikon Eclipse Ti microscope using NIS Elements AR 4.13 software to determine the percentage of EBD-positive quadricep muscle area. EBD area analysis was performed in a blinded fashion.

Isolated muscle force measurements

Mice were anesthetized by intraperitoneal injection of pentobarbital (6 mg/100 g body weight). The distal tendon of the TA muscle was exposed, and 4-0 nylon suture was tied at the muscle-tendon interface. The knee and foot of the mouse were secured to a platform, and the tendon was mounted to a servomotor (Aurora Scientific; Model 305C). The TA muscle was stimulated using two intramuscular electrodes placed on either side of the peroneal nerve at 200 Hz with a biphase high-power stimulator (Aurora Scientific; Model 701C). Stimulation voltages and optimal muscle length were adjusted to produce maximum isometric twitch force. A series of six consecutive tetanic isometric contractions were performed with a 2-min rest period between each contraction. Specific force was measured by dividing the active force by the muscle’s physiologic cross-sectional area, and the percentage drop in force was determined by taking the difference in force between contractions 1 and 6. The experimenter was blinded to the genotypes during the collection of these data.

Transmission electron microscopy

Fresh quadricep was harvested and immediately immersed in relaxing buffer (0.15% sucrose, 5% dextrose, and 100 mM KCl in 1× PBS), subsequently fixed overnight [3.5% glutaraldehyde and 0.15% sucrose in 0.1 M sodium cacodylate (pH 7.4)] and postfixed in 1% OsO₄ (in water) for 2 hours at RT. Samples were then washed, dehydrated and embedded using epoxy resin. Ultrathin sections of tissues were counterstained with 1.5% uranyl acetate, 70% ethanol, and lead nitrate/Na citrate. Images were obtained using a Hitachi 7600 transmission electron microscope connected to an AMT digital camera.

Serum CK measurement

Blood samples from mice were collected from the abdominal aorta of anesthetized mice, and plasma was isolated by centrifugation at 5000g for 15 min at 4°C. Serum CK measurement was performed on a Roche Cobas c311 clinical chemistry analyzer (Roche) by a trained observer blinded to genotype.

Statistics

All results are presented as means ± SEM. Data from each experiment was subjected to a Brown-Forsyth test for unequal variance before additional statistical analysis. For datasets with equal variance, statistical analysis was performed with unpaired two-tailed Student’s t test for two independent groups or one-way analysis of variance (ANOVA) with post hoc Dunnett’s test for multiple comparisons of three or more independent groups. Two-way ANOVA with post hoc Dunnett’s or Tukey’s tests for multiple comparisons were used where appropriate. For datasets with unequal variance, a nonparametric Kruskal-Wallis test was used to assess statistical significance. All statistics were performed using GraphPad Prism 9, and values were considered statistically significant when P ≤ 0.05. No statistical analysis was used to predetermine sample size although previous power analysis of similar experimental variables in mice did direct selection of group sizes. The experiments were not randomized, and no animals were excluded from analysis. The investigators were not blinded to allocation during all experiments but were blinded in specific experimental outputs as noted. The exact number of animals or biological replicates for each experiment is indicated in the figure legends.

Supplementary Materials

This PDF file includes:

Figs. S1 to S5