Research ArticleMYOPATHIES

Amphiphysin 2 modulation rescues myotubular myopathy and prevents focal adhesion defects in mice

See allHide authors and affiliations

Science Translational Medicine  20 Mar 2019:
Vol. 11, Issue 484, eaav1866
DOI: 10.1126/scitranslmed.aav1866

A muscle-building interaction

Centronuclear myopathies (CNMs) are rare genetic disorders characterized by severe muscle weakness. Mutations in myotubularin 1 (MTM1) and in amphiphysin 2 (BIN1) are responsible for two different forms of the disease. BIN1 and MTM1 have been shown to interact in skeletal muscles. Now, Lionello et al. investigated the role of this interaction in a model of Mtm1-mediated CNM. Postnatal BIN1 overexpression improved survival and muscle strength in Mtm1 knockout mice. The treatment also restored myofiber integrity and rescued extracellular matrix and focal adhesion defects in myofibers. The results suggest that BIN1-MTM1 interaction plays a role in CNM and could be targeted for treating CNM due to MTM1 mutations.


Centronuclear myopathies (CNMs) are severe diseases characterized by muscle weakness and myofiber atrophy. Currently, there are no approved treatments for these disorders. Mutations in the phosphoinositide 3-phosphatase myotubularin (MTM1) are responsible for X-linked CNM (XLCNM), also called myotubular myopathy, whereas mutations in the membrane remodeling Bin/amphiphysin/Rvs protein amphiphysin 2 [bridging integrator 1 (BIN1)] are responsible for an autosomal form of the disease. Here, we investigated the functional relationship between MTM1 and BIN1 in healthy skeletal muscle and in the physiopathology of CNM. Genetic overexpression of human BIN1 efficiently rescued the muscle weakness and life span in a mouse model of XLCNM. Exogenous human BIN1 expression with adeno-associated virus after birth also prevented the progression of the disease, suggesting that human BIN1 overexpression can compensate for the lack of MTM1 expression in this mouse model. Our results showed that MTM1 controls cell adhesion and integrin localization in mammalian muscle. Alterations in this pathway in Mtm1−/y mice were associated with defects in myofiber shape and size. BIN1 expression rescued integrin and laminin alterations and restored myofiber integrity, supporting the idea that MTM1 and BIN1 are functionally linked and necessary for focal adhesions in skeletal muscle. The results suggest that BIN1 modulation might be an effective strategy for treating XLCNM.


Centronuclear myopathies (CNMs) are severe congenital disorders characterized by muscle weakness, hypotonia, respiratory insufficiency, myofiber atrophy, and abnormal nuclei position (1). Currently, no specific therapy is available for patients, and the pathophysiology of these disorders is not well understood.

The most severe form of CNM is the X-linked form, also called myotubular myopathy (XLCNM; OMIM 310400). XLCNM is caused by mutations in the phosphoinositide 3-phosphatase myotubularin (MTM1) (25). Other forms of CNM are mainly due to mutations in bridging integrator 1 (BIN1), ryanodine receptor 1 (RYR1), and dynamin 2 (DNM2). Mutations in BIN1 cause both autosomal recessive and dominant forms (OMIM 255200) (6, 7). BIN1 encodes amphiphysin 2, a Bin/amphiphysin/Rvs domain protein that senses and induces membrane curvature and remodeling (811). Among the various tissue-specific isoforms of BIN1, the skeletal muscle isoform 8 contains a phosphoinositide binding domain important for the formation of transverse tubules (T-tubules), a highly specialized muscle structure crucial for excitation-contraction coupling (8, 12). BIN1 is also important for clathrin-mediated endocytosis (13, 14) and is linked to nuclei positioning (15).

No similarity has been identified between MTM1 and BIN1 protein structures (16). MTM1 and BIN1 bind directly, and this interaction is important for membrane tubulation (16). Furthermore, BIN1 mutations leading to CNM disrupt this interaction (16), suggesting that the MTM1-BIN1 binding might be important for normal muscle function.

The muscle weakness of XLCNM correlates with a strong reduction in muscle fiber size, rounder fibers with increased interfiber space, abnormal centralization of nuclei or organelles such as mitochondria, and altered T-tubule structure (17). These hallmarks have been reproduced faithfully in the Mtm1−/y knockout mouse, which develops a progressive muscle weakness starting at 2 to 3 weeks of age, leading to death by 1 to 3 months of age (18, 19).

We hypothesize that the main defects of fiber shape and size are due to an alteration in focal adhesions. One of the key focal adhesion components and regulators are integrins (20, 21). Recently, MTM1 has been implicated in the exit of β1 integrin from endosomes (22), and depletion of mtm, the ortholog of MTM1, in drosophila muscle caused accumulation of integrin on endosomes (23). These results raise the possibility that MTM1, through integrins, is required for myofiber attachments. Integrins are part of the focal adhesion complex responsible for maintaining the connection between the cytoskeleton and the extracellular matrix (24). Integrins bind extracellular matrix proteins (such as fibronectin, laminin, and collagen), change their conformation, and recruit cytoskeletal regulators (such as vinculin) and kinases [such as focal adhesion kinase (FAK)] to activate downstream pathways (25). Integrins are actively recycled and can also signal from endosomes (26). Hence, they are key regulators of mechanotransduction, tissue integrity, cell shape, and migration (20). In muscle, β1 integrin is important for myoblast fusion, nuclei peripheral positioning, and sarcomere assembly (2729) and is part of the costamere that couples sarcomeric forces to the extracellular matrix (30).

The aim of this study was to investigate the relationship between MTM1 and BIN1 in skeletal muscle, under normal and pathological conditions. Because loss-of-function mutations in MTM1 or BIN1 cause CNM, we hypothesized that increasing BIN1 expression may compensate for loss of MTM1. This hypothesis was tested by increasing BIN1 either through a genetic cross with a humanized BIN1 transgenic mouse or by exogenous expression of human BIN1 in Mtm1−/y mice using adeno-associated virus (AAV) and by characterizing the CNM motor and histological phenotypes. We also investigated the focal adhesion pathway in the pathological model and upon BIN1-mediated rescue.


Increased expression of human BIN1 rescues Mtm1−/y survival

To study the epistasis between MTM1 and BIN1, we modulated the expression of these genes in mice. We first investigated the effect of down-regulation of Bin1 in Mtm1−/y mice. Mtm1−/y mice start developing a progressive muscle disease closely resembling XLCNM from 3 to 4 weeks and die usually by 2 months of age (18). Mtm1+/− females were crossed with Bin1+/− males to obtain wild-type (WT), Mtm1−/y, Bin1+/−, and Mtm1−/y Bin1+/− mice. At embryonic day 18.5 (E18.5), 8.3% of embryos were Mtm1−/y Bin1+/−, whereas no Mtm1−/y Bin1+/− pups were obtained at 10 days after birth, showing that concomitant down-regulation of Mtm1 and Bin1 was not compatible with postnatal life (table S1).

Conversely, to up-regulate BIN1, we created transgenic mice expressing the human BIN1 gene (TgBIN1) by insertion of a human bacterial artificial chromosome (BAC) containing the human BIN1 gene with its flanking sequence into the mouse genome (fig. S1A). Reverse transcription polymerase chain reaction, cloning, and sequencing from tibialis anterior (TA) muscles showed the presence of the human BIN1 isoform 8 in five of six clones containing human BIN1 (fig. S1B). BIN1 isoform 8 is the main muscle isoform (31). TgBIN1 mice are viable with no overt motor phenotypes (movie S1). Crossing TgBIN1 with Bin1−/− mice that die at birth from muscle defects (31) efficiently rescued the lethality at birth (table S2). Furthermore, no difference was observed in body weight, TA weight, and specific muscle force between WT and Bin1−/− TgBIN1 mice at 4 months (fig. S1, C to I), suggesting that human BIN1 is functional in a mouse context.

To investigate whether increased expression of BIN1 rescues the survival of Mtm1−/y mice, we generated Mtm1−/y TgBIN1 mice. Most Mtm1−/y TgBIN1 mice survived more than 12 months, and survival was indistinguishable from WT and TgBIN1 mice (Fig. 1A). There was no difference in body weight between WT, TgBIN1, and Mtm1−/y TgBIN1 mice throughout their 2-year life span, whereas Mtm1−/y mice weighed significantly less before dying (P < 0.0001) (Fig. 1B, fig. S2A, and movie S2). BIN1 expression was increased about fourfold in TgBIN1 and Mtm1−/y TgBIN1 mice compared to WT (Fig. 1, C and D). These results show that increased expression of BIN1 alone is sufficient to rescue the postnatal lethality and growth defects observed in Mtm1−/y mice.

Fig. 1 BIN1 overexpression rescues the life span of Mtm1−/ymice.

(A) Life span represented as the percentage of survival. (B) Body weight measured at different time points (n > 4). (C) Western blot from TA muscles probed with anti-BIN1 and MTM1 antibodies. (D) BIN1 quantification normalized on TCE (2,2,2-trichloroethanol) fluorescence labeling all tryptophan-containing proteins. Statistical analysis: One-way analysis of variance (ANOVA) and Bonferroni post hoc test were performed for data represented in the graph; *P < 0.05.

Increased expression of BIN1 rescues muscle strength and coordination in Mtm1−/y mice

We next investigated whether muscle strength was rescued in Mtm1−/y mice by BIN1 expression. At 5 weeks, Mtm1−/y mice had strong defects in rotarod (neuromuscular analysis) and footprint (gait analysis) tests, whereas the Mtm1−/y TgBIN1 mice performed similarly to WT and TgBIN1 mice in rotarod, footprint, bar, and grip tests (Fig. 2A and fig. S2, A to E). Mtm1−/y mice could not perform the bar test. At 5 months of age, Mtm1−/y TgBIN1 were still able to perform all tests (rotarod and footprint), indicating a long-term improvement in their motor function (fig. S2, F and G). Mtm1−/y mice had difficulties to perform the hanging test (a test indicating whole-body strength), whereas the Mtm1−/y TgBIN1 mice performed similarly to WT.

Fig. 2 Overexpression of BIN1 rescues muscle force of Mtm1−/ymice.

(A) Hanging test. Mice were suspended from a cage lid for a maximum of 60 s, and each mouse repeated the test three times (n ≥ 4). (B) Weight of TA muscle normalized on total body weight (n > 5). (C) Specific muscle force (sPo) of the TA at 2 months (n > 5), 7 months (n > 5), and 24 months (n = 4). NA, not applicable as Mtm1−/y mice died before. Statistical analysis: One-way ANOVA and Bonferroni post hoc test were performed for data represented in (A). Nonparametric Kruskal-Wallis and Dunn’s post hoc tests were performed for data represented in (B) and (C). ns, not significant; *P < 0.05, **P < 0.01, ***P < 0.001.

Overexpression of BIN1 in Mtm1−/y mice, which normally present with strong muscle atrophy, rescued the TA muscle atrophy back to WT (Fig. 2B). Specific muscle force, measured in situ in the TA muscle, was extremely low in Mtm1−/y mice at 2 months and rescued to WT levels in 2-, 7-, and 24-month-old Mtm1−/y TgBIN1 mice (Fig. 2C and fig. S2H). No difference was detected in the muscle force between TgBIN1 and WT mice. The time to muscle exhaustion during continuous stimulation was similar between Mtm1−/y TgBIN1, TgBIN1, and WT mice at 2 and 7 months of age (fig. S2I). Overall, the severe muscle weakness phenotype of Mtm1−/y mice was fully rescued by increased expression of BIN1 in Mtm1−/y mice.

Increased BIN1 rescues histological and ultrastructural defects in Mtm1−/y mice

At 8 weeks, Mtm1−/y TA muscles present with small-rounded fibers with abnormal subsarcolemmal and central accumulation of oxidative staining as it was previously described (Fig. 3A and fig. S3A) (18, 19). Fiber size distribution (minimum Feret diameter) was shifted toward smaller fibers (peak diameter, 20 to 25 μm), whereas it increased from 25 to 30 μm in Mtm1−/y TgBIN1, similar to WT and TgBIN1 muscles (Fig. 3B and fig. S3, A and B). Mtm1−/y TA muscles exhibit about 22% of fibers with abnormal nuclei position, whereas Mtm1−/y TgBIN1 were indistinguishable from WT (Fig. 3C). Similar defects in Mtm1−/y mice were found in other muscles [gastrocnemius (GAS) and diaphragm] and were efficiently rescued in Mtm1−/y TgBIN1 mice (fig. S3, C to F). Later at 7 months of age, no differences were found in TA and GAS muscles between the Mtm1−/y TgBIN1 and WT mice (fig. S3, G to J).

Fig. 3 BIN1 overexpression rescues muscle histology and ultrastructure in 8-week-old Mtm1−/ymice.

(A) Transversal TA muscle sections stained with hematoxylin and eosin (H&E) and succinate dehydrogenase (SDH). Scale bars, 50 μm. (B) Diameter of TA fibers grouped into 5-μm intervals (n = 5). (C) Frequency of fibers with abnormal (internalized and centralized) nuclei position in TA (n = 5). (D) TA muscle ultrastructure observed by electron microscopy (EM). Scale bars, 1 μm. High-magnification insert for triads. (E) Longitudinal TA muscle sections stained for DHPR (red) and BIN1 (green). The asterisk indicates disorganized DHPR staining. Images were taken with different laser intensity. Scale bars, 10 μm. Statistical analysis: Nonparametric Kruskal-Wallis and Dunn’s post hoc tests were performed; *P < 0.05. The black, blue, and green asterisks correspond to the significant difference observed between the Mtm1−/y group and the WT, TgBIN1, or Mtm1−/y TgBIN1 groups, respectively.

We next investigated myofiber organization by transmission electron microscopy in TA at 8 weeks (Fig. 3D). In contrast to Mtm1−/y mice that had misaligned Z lines and general sarcomere disorganization, Mtm1−/y TgBIN1 mice displayed normal myofiber ultrastructure (fig. S3K). Because several studies highlighted an important role for BIN1 in T-tubule biogenesis (8, 12) and because T-tubule defects were observed in several forms of CNMs (32), we next analyzed the triads that are composed of one T-tubule and two sarcoplasmic reticulum cisternae. Mtm1−/y triads were barely distinguishable (Fig. 3D and fig. S3L). However, normal triads and T-tubule shape and localization were observed in TgBIN1 and Mtm1−/y TgBIN1 mice (fig. S3M). BIN1 and the T-tubule L-type Ca2+ channel or dihydropyridine receptor (DHPR) colocalized at T-tubules in WT, TgBIN1, and Mtm1−/y TgBIN1 mice, whereas Mtm1−/y mice had some fibers with disorganized staining (Fig. 3E and fig. S3N). Overall, increased expression of human BIN1 rescued the muscle atrophy, histopathology, and ultrastructure alterations observed in Mtm1−/y mice.

Postnatal muscle overexpression of human BIN1 rescues muscle force and myofiber organization in Mtm1−/y mice

The expression of BIN1 during development in Mtm1−/y mice by genetic cross rescued the muscle strength and all CNM hallmarks. We next investigated whether postnatal BIN1 expression was sufficient to rescue the Mtm1−/y defects. The human BIN1 isoform 8 was the main isoform expressed in the rescued Mtm1−/y TgBIN1 mice and was cloned into AAV serotype 9. AAV-BIN1 was injected intramuscularly in TA muscles of 3-week-old Mtm1−/y mice, and analysis was performed 2 or 4 weeks after injection, compared to empty AAV. BIN1 was expressed fourfold higher than in control muscles (Fig. 4A). Two weeks after injection, there was no significant difference in TA muscle weight in Mtm1−/y injected with AAV-BIN1 and the Mtm1−/y injected with the AAV Ctrl (P = 0.1216; Fig. 4, C and D).Whereas the specific muscle force of Mtm1−/y mice injected with the AAV Ctrl was significantly different compared to the WT control (P = 0.0056), no difference was detected between the WT and the Mtm1−/y injected with AAV-BIN1 (P = 0.9288; Fig. 4E). AAV-BIN1 greatly improved the general aspect of the muscle (H&E) including fiber size and to a lesser extent nuclei position and the oxidative staining (SDH; Fig. 4, F to H, and fig. S4A). Myofiber organization was improved in Mtm1−/y TA muscles injected with AAV-BIN1 compared with AAV-empty control and WT, and Mtm1−/y injected with AAV-BIN1 had the same number of triads per sarcomere (Fig. 4I and fig. S4B). Similar effects on improvement of skeletal muscle force and muscle histology were also observed 4 weeks after injection (fig. S4, C to H). Overall, these results show that the intramuscular overexpression of BIN1 after birth is sufficient to improve muscle force and myofiber organization in the Mtm1−/y mice. These results suggest that increasing BIN1 postnatally is sufficient to reduce the myopathic phenotype in Mtm1−/y mice.

Fig. 4 Postnatal intramuscular BIN1 overexpression rescues muscle force and myofiber organization in Mtm1−/ymice.

(A) Western blot with anti-BIN1 antibody (n = 5). (B) Quantification of BIN1 on glyceraldehyde-3-phosphate dehydrogenase (GAPDH). (C) TA muscle weight normalized on total body weight (n ≥ 5). (D) Representative photograph of Mtm1−/y TA muscles injected with AAV-BIN1 (left) or AAV empty control (right) 5 weeks after injection. (E) Specific TA muscle force (n ≥ 5). (F) Transversal TA sections stained with H&E and SDH. Scale bars, 50 μm. (G) Frequency of fibers with abnormal (internalized and centralized) nuclei position in TA of WT mice and Mtm1−/y mice (n ≥ 2). (H) Diameter of TA fibers grouped into 5-μm intervals (n ≥ 4). (I) TA muscle ultrastructure observed by electron microscopy (n = 1). Scale bars, 1 μm. Statistical analysis: Nonparametric Kruskal-Wallis and Dunn’s post hoc tests were performed; *P < 0.05, **P < 0.01. The black and green asterisks correspond to the significant difference observed between the WT AAV Ctrl and Mtm1−/y AAV Ctrl and between Mtm1−/y AAV-BIN1 and Mtm1−/y AAV Ctrl, respectively.

Postnatal systemic overexpression of BIN1 prolongs Mtm1−/y life span and rescues CNM muscle defects

Because increasing BIN1 by intramuscular injection of AAV-BIN1 improved CNM features in Mtm1−/y muscles, we tested whether systemic AAV-BIN1 transduction after birth could rescue the muscle defects and extend life span of Mtm1−/y mice. AAV-BIN1 or AAV-empty control was injected into pups at postnatal day 1 by intraperitoneal injection, and the effects of the injection were analyzed until 10 weeks of age, an age never reached by untreated Mtm1−/y mice. The systemic overexpression of BIN1 after birth rescued the premature death of Mtm1−/y mice (Fig. 5A). The oldest treated Mtm1−/y mouse died at 1.5 years old (movie S3). A slight amelioration of body weight was noted for Mtm1−/y mice injected with AAV-BIN1 compared to mice injected with AAV Ctrl (Fig. 5B). In addition, no difference in organ weight (brain, heart, and liver) was observed between Mtm1−/y mice injected with AAV-BIN1 and WT mice at 10 weeks (fig. S5A).

Fig. 5 Postnatal systemic BIN1 overexpression rescues the survival and muscle defects of Mtm1−/ymice.

(A) Percentage of surviving animals until 10 weeks of age. (B) Body weight measurements. (C) Western blot probed with anti-BIN1 and GAPDH antibodies. (D) BIN1 quantification normalized to GAPDH. (E) Ratio of TA muscles weight on total body weight (n ≥ 5). (F) Absolute (Abs) TA muscle force. (G) Specific TA muscle force (n ≥ 4). (H) Transversal TA muscle sections stained with H&E and SDH. Scale bars, 50 μm. (I) Percentage of fibers with abnormal (internalized and centralized) nuclei position (n = 5). (J) Minimum Feret of TA fibers grouped into 5-μm intervals (n = 5). (K) TA muscle ultrastructure observed by electron microscopy. Scale bars, 1 μm. High-magnification insert for triads. (L) Frequency of triads per sarcomere. (M) Longitudinal TA muscle sections stained with DHPR and BIN1 antibodies. Scale bar, 10 μm. Statistical analysis: One-way ANOVA and Tukey’s post hoc test were performed in (B) (until 8-week time point), and Student’s t test was performed in (B) (for 9- and 10-week time points), (C) to (F), (H), and (I); *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. The black and green asterisks correspond to the significant difference observed between the WT AAV Ctrl and Mtm1−/y AAV Ctrl and between Mtm1−/y AAV-BIN1 and Mtm1−/y AAV Ctrl, respectively.

To evaluate whether the positive effect on growth and survival correlated with an increase in muscle function and structure, we dissected TA and GAS muscles. BIN1 overexpression was confirmed in the AAV-BIN1–injected animals from TA muscle lysate (Fig. 5, C and D). TA muscle weight relative to body weight of surviving treated Mtm1−/y mice was smaller than WT control, whereas no difference was observed in the GAS weight between the WT and the treated Mtm1−/y (Fig. 5E and fig. S5B). Mtm1−/y mice presented a severe TA muscle weakness at 7 weeks of age, whereas no difference was observed in situ in TA absolute or specific muscle force or time to fatigue between Mtm1−/y mice injected with AAV-BIN1 and WT injected with AAV Ctrl at 10 weeks (Fig. 5, F and G, and fig. S5C), indicating a complete rescue in muscle force. We analyzed the histology at 10 weeks when no untreated Mtm1−/y mice were alive for direct comparison. In AAV-BIN1–treated Mtm1−/y mice, the general organization was normal. There was no difference in nuclei position, and most of the fibers had a muscle diameter between 30 and 50 μm as the WT control; however, the number of fibers with diameter between 60 and 95 μm was increased in AAV-BIN1–treated Mtm1−/y mice (Fig. 5, H to J). Furthermore, only an average of 5% of abnormal oxidative staining fibers were counted in the treated Mtm1−/y (fig. S5D). Ultrastructural analysis revealed that the sarcomere organization was rescued by AAV-BIN1 injection and the number of triads per sarcomere was normalized, with most triads presenting a normal shape and localization (Fig. 5, K and L). The correct T-tubule organization was confirmed through DHPR and BIN1 immunofluorescence (Fig. 5M and fig. S5E). In conclusion, systemic injection of AAV-BIN1 extended Mtm1−/y life span, normalized TA muscle force, and rescued the main histological and ultrastructural defects of the TA muscle.

Focal adhesion is impaired after loss of Mtm1 and rescued in mice overexpressing BIN1

Hypotrophic (smaller) and rounder fibers with increased interfiber space are the main histological defects in patients with XLCNM (33), suggesting a defect in cell adhesion (Fig. 6A). The Mtm1−/y mouse faithfully recapitulates these hallmarks. To better understand whether these defects correlate with an increase of extracellular matrix, we stained transverse TA sections from WT and Mtm1−/y mice with collagen and laminin, two main components of the extracellular matrix in skeletal muscle (fig. S6A) (34). Mtm1−/y muscle presented an increased interfiber space occupied by collagen labeled with Masson staining or a specific antibody (Fig. 6, B to D, and fig. S6, A and B). We next quantified the distance between muscle fibers using collagen fluorescence staining. Mtm1−/y TA muscle presented an increase in the thickness of collagen interfiber staining compared to WT control (fig. S6C). Proteins from the extracellular matrix connect to proteins of focal adhesion including α71 integrins and adaptor proteins such as vinculin (fig. S6C) (35). In WT, vinculin and integrins localize at the sarcolemma, specifically at the costamere. Mtm1−/y muscles exhibited internalized vinculin (Fig. 6E and fig. S7, A to C) and β1 integrin (Fig. 6F).

Fig. 6 MTM1 and BIN1 are essential for β1 integrin trafficking in mammalian muscle.

(A) Transversal muscle sections from a control human skeletal muscle and a patient with XLCNM (mutation c.141-144delAGAA p.Glu48LeufsX24 in MTM1). Scale bar, 20 μm. (B) Eight-week WT and Mtm1−/y TA muscle stained with Masson trichrome. Scale bar, 10 μm. (C to F) Transversal WT and Mtm1−/y TA muscles stained for the extracellular matrix proteins (C) collagen, (D) laminin, (E) vinculin, and (F) β1 integrin. Scale bars, 10 μm. (G) Eight-week transversal Mtm1−/y TgBIN1 TA muscle sections stained with Masson trichrome. Scale bar, 10 μm. (H to K) Transversal WT and Mtm1−/y TA muscles stained for the extracellular matrix proteins (H) collagen, (I) laminin, (J) vinculin, and (K) β1 integrin. Scale bars, 10 μm. (L) Percentage of muscle fibers with β1 integrin internalized. (M) Transversal TA muscle sections stained for β1 integrin (green) and EEA1 (red). Scale bars, 10 and 1 μm [for the Mtm1−/y (zoom) panel]. Arrows point to abnormal intracellular accumulation of β1 integrin on EEA1-positive endosomes. Scale bars, 10 and 1 μm (zoom). (N) Western blot probed with anti-FAK and anti–pY397-FAK antibodies. (O) Quantification of FAK on TCE and (P) quantification of pY397-FAK on total FAK. Statistical analysis: One-way ANOVA and Tukey’s post hoc test were performed; *P < 0.05, ***P < 0.001.

We next investigated whether the modulation of BIN1 expression rescued the localization defects in the extracellular matrix and focal adhesion proteins. The muscle of Mtm1−/y TgBIN1 mice showed a normalization of interfiber space and consequently reduced collagen accumulation between fibers (Fig. 6, G to I), in addition to a rescue in fiber size. We next verified the localization of vinculin and β1 integrin in Mtm1−/y TgBIN1 skeletal muscle. Vinculin and β1 integrin localized on the plasma membrane in Mtm1−/y TgBIN1 as observed in WT (Fig. 6, J and K). Twenty percent of muscle fibers had abnormal β1 integrin localization in Mtm1−/y, whereas no fibers with abnormal staining were identified in Mtm1−/y TgBIN1 (Fig. 6L).

To investigate in which intracellular compartment β1 integrin accumulates, we labeled endosomal markers on muscle sections and found the early endosome marker EEA1 (early endosomal antigen 1) to colocalize with β1 integrin (Fig. 6M and fig. S7, D to E). EEA1-positive endosomes also aggregated in several Mtm1−/y myofibers compared to WT (fig. S7F). Overall, these results highlight that β1 integrin abnormally accumulates at early endosomes in Mtm1−/y muscles, indicating a defect in β1 integrin turnover that may induce the abnormality in fiber shape and the increase in interfiber space.

To further decipher the mechanisms responsible for the defects in the focal adhesion pathway, we measured the activation of FAK, a downstream effector of β1 integrin. No difference in FAK protein quantity was detected between the Mtm1−/y and WT (Fig. 6, N and O). The activation of the focal adhesion complex leads to the autophosphorylation of FAK on tyrosine 397 (Y397) (25). Mtm1−/y muscles exhibited decreased autophosphorylation of FAK, confirming that activation of the focal adhesion pathway is altered (Fig. 6P). We also verified whether phospho-Y397 (pY397)–FAK was present on endosomes accumulating β1 integrin in Mtm1−/y skeletal muscle fibers. A colocalization of internalized β1 integrin and pY397-FAK was detected in some Mtm1−/y fibers, suggesting that β1 integrin signaling was still maintained in endosomes (fig. S7G).

Consequences in focal adhesion defects in Mtm1−/y skeletal muscle

To better understand the functional impact of focal adhesion defects observed in skeletal muscle, we conducted experiments on primary myoblasts. Mtm1−/y myoblasts exhibited larger β1 integrin vesicles than WT myoblasts (Fig. 7A and fig. S7H), confirming the accumulation of β1 integrin observed in adult skeletal muscle. No difference was detected in β1 integrin plasma membrane signal between WT and Mtm1−/y skeletal muscles (fig. S7I). Focal adhesions are important for cell adhesion, migration, and fusion (27, 28, 36). Cell adhesion was checked by allowing WT and Mtm1−/y myoblasts to adhere for 10, 20, and 40 min on laminin-coated dishes. The surface area of Mtm1−/y myoblasts was lower than that of WT cells at 20 min, whereas no difference was detected at 1, 40, and 60 min (Fig. 7B). Because studies showed that defects in β1 integrin localization affected cell migration (36), a migration assay was performed with WT and Mtm1−/y myoblasts plated on laminin-coated dishes. A significant (P = 0.0126) reduction of migrating distance was observed in Mtm1−/y myoblasts compared to WT (Fig. 7C). Last, the ability of myoblasts to fuse was analyzed over time on Matrigel-coated dishes. A significant (P = 0.0001) defect in myoblast fusion was identified at 24 and 48 hours of differentiation in Mtm1−/y cells but not at 72 hours (Fig. 7D). Overall, these results suggest that Mtm1−/y myoblasts exhibit defects in cell adhesion, migration, and fusion that are related to a defect in β1 integrin localization and turnover in muscle.

Fig. 7 BIN1 overexpression rescues focal adhesion deficit in Mtm1−/ymuscle.

(A) WT and Mtm1−/y primary myoblasts probed for β1 integrin. Scale bars, 10 μm2. (B) Adhesion assay: adherent surface of primary myoblasts at different time points after plating (n ≥ 25 from n ≥ 2 mice). (C) Migration assay, distance (micrometers) migrated by WT and Mtm1−/y myoblasts during 24 hours (n ≥ 20 from n = 3 mice). (D) Fusion index, number of nuclei in WT and Mtm1−/y myotubes at three time points after differentiation was started (24, 48, and 72 hours; n > 36 from n ≥ 2 mice). (E) Western blot probed with anti-laminin antibodies. (F) Quantification of laminin on TCE. (G) Western blots probed for β1 integrin. (H) Quantification of β1 integrin normalized to the TCE of the same gel. (I) Transversal TA muscle of WT mice injected systemically with AAV empty as control and Mtm1−/y mice injected systemically with AAV-BIN1 and probed with anti–β1 integrin antibody. Scale bars, 10 μm. Statistical analysis: Student’s t test, one-way ANOVA, and Bonferroni post hoc test were performed; *P < 0.05, **P < 0.01, ***P < 0.001.

We next analyzed the protein levels of laminin, vinculin, and β1 integrin and found an increase in their amount in the Mtm1−/y skeletal muscle, which was normalized to WT levels upon BIN1 expression (Fig. 7, E to H, and fig. S7J). Mtm1−/y skeletal muscle exhibited a higher β1 integrin protein level compared to the WT control. No difference in β1 integrin transcription was identified between WT and Mtm1−/y skeletal muscles (fig. S7K). We next investigated whether the postnatal overexpression of BIN1 rescued the β1 integrin intracellular accumulation observed in Mtm1−/y mice. A normalization of β1 integrin localization in Mtm1−/y muscle was observed upon AAV-BIN1 systemic injection after birth (Fig. 7I and fig. S7L), suggesting that this rescue was not dependent on the methodology used for BIN1 expression. Overall, BIN1 expression efficiently rescued the defects of extracellular matrix and focal adhesion in addition to the muscle weakness and fiber histopathology in the Mtm1−/y mouse model of XLCNM.


This study reports a genetic and functional link between MTM1 and BIN1 in skeletal muscle. Increased expression of BIN1 by genetic cross or viral delivery after birth prolongs the life span of Mtm1−/y mice and rescued the muscle force and the main histological hallmarks of CNM in this mouse model. Mtm1−/y mice showed defects in integrin turnover and focal adhesion functions, myofiber hypotrophy and abnormal shape, and these phenotypes were rescued upon BIN1 overexpression.

Loss-of-function mutations in BIN1 and MTM1 cause different forms of CNM (7, 37); however, whether a functional connection existed between these genes was not clear. Although Mtm1−/y mice present with progressive muscle weakness from 2 to 3 weeks of age and die by 2 months (18), increased expression of human BIN1, either by transgenesis or AAV-mediated transduction, rescued the life span, the motor defects, most of the histological and ultrastructural defects, and the molecular alterations. These results show that increasing BIN1 compensates for the lack of MTM1, suggesting that MTM1 and BIN1 might be in a common pathway where MTM1 is a positive regulator of BIN1. Previously, we showed that decreased expression of DNM2, a third protein mutated in CNM, rescued both the CNM phenotypes due to MTM1 or BIN1 loss (31, 38), supporting that MTM1 and BIN1 are negative regulators of DNM2. Together with the present data, we propose a CNM pathway where MTM1 would activate BIN1 that, in turn, inhibits DNM2. This hypothesis is supported by the fact that MTM1 binds BIN1 (16) and that BIN1 is a known interactor of DNM2 (7, 8). In addition, BIN1 mutations reduce the interaction with MTM1 or DNM2 (7, 16). However, we cannot rule out that MTM1 and BIN1 act on parallel pathways converging to regulate DNM2. We observed a 1.9-fold increase in BIN1 protein amount in Mtm1−/y muscle at 8 weeks (advanced disease stage) but normal quantity at 5 weeks, suggesting a potential compensatory mechanism that is insufficient for reaching a rescue that can be obtained by increasing exogenous BIN1 by about 3.5-fold through transgenesis or AAV injection.

Here, we identified BIN1 as a modifier for MTM1-related CNM and thus BIN1 as a potential therapeutic target. We showed that BIN1 expression could rescue the postnatal muscle maintenance defects linked to MTM1 loss. After the positive proof of concept based on a transgenesis approach, we used AAV delivery of human BIN1 after birth to validate a potentially translational approach. BIN1 was first overexpressed intramuscularly and then through systemic delivery, and both strategies rescued muscle force and myofiber structural defects; in addition, the systemic injection greatly prolonged the life span. AAV injection at 3 weeks, after the onset of the disease, was sufficient to reduce disease severity, suggesting that early treatment in symptomatic patients may provide a benefit. The AAV injection does not allow treatment interruption, and it is still not clear how long the AAV-mediated protein expression remains present in the body. In addition, we cannot exclude the possibility of a toxic response to BIN1 overexpression. In addition, our proof of concept was obtained in mice that differ in size and physiology to human. However, the positive results obtained here using a human BIN1 transgene in an AAV9 vector, a serotype already used in clinical trial, suggest that this approach could be tested in larger animals that more closely mimic the human condition.

DNM2 modulation also rescued BIN1 or MTM1 loss in animal models (31, 39, 40). In this study, we identified an additional “cross-therapy” concept where modulation of a CNM gene (BIN1) rescues the loss of another CNM gene (MTM1). Moreover, AAV-mediated MTM1 gene therapy was shown to be effective in animal models of XLCNM and is currently in clinical trials (41). Patients with XLCNM barely express MTM1 protein, and its delivery might trigger an immune response against an unknown protein. Using AAV-BIN1 strategy could avoid a potential immune response as in the case of AAV-MTM1 or the potential secondary effects of DNM2 reduction (42).

Small rounded fibers and increased interfiber space are main histological hallmarks in patients for the diagnosis of XLCNM and suggest a defect of adhesion to the extracellular matrix. β1 Integrin is the major integrin molecule of skeletal muscle and links the extracellular matrix with the intracellular cytoskeleton and the sarcomeres at focal adhesions termed costameres (34). Focal adhesion integrity is important for muscle as they mediate mechanotransduction and are a platform for intracellular signaling (43). Alteration of β1 integrin quantity and localization in muscle was seen in 5- and 8-week-old Mtm1−/y mice, together with increased collagen and interfiber space. Correct sarcomere alignment and integrity depends on costameres. Myofibril formation can be inhibited by antagonizing integrin dimers alone, suggesting that integrin–extracellular matrix interaction is important for correct sarcomere formation during muscle development (28). Sarcomeres are greatly altered in XLCNM and Mtm1−/y muscles, which probably contribute to the severe muscle weakness (33, 44). In addition, a recent report proposed that integrins regulate peripheral nuclear positioning in myofibers differentiated in vitro, suggesting that integrin defects may also mediate the centralization of nuclei in CNM (29). All these defects were rescued by increased BIN1 expression, supporting that defects in the focal adhesion pathway are an important cause of the disease. MTM1 and BIN1 thus appear as important regulators of focal adhesions, in addition to their recognized roles in maintenance of the T-tubule structure (8, 12, 45, 46). Mice that lack β1 integrin specifically in skeletal muscle had reduced muscle mass and alteration of sarcomere ultrastructure and died at birth with noninflated lungs (28); such phenotypes are typical from patients with XLCNM (1). In addition, compound heterozygous mutations in α7 integrin cause muscular dystrophy (47). Together, the literature supports the rationale that defects in the functions of focal adhesions are an important component of the pathomechanism leading to the MTM1-related myopathy.

Both MTM1 and BIN1 are involved in membrane remodeling and recycling in cells (22, 23, 48), and we observed that β1 integrin is blocked in EEA1-positive endosomes in Mtm1−/y muscles. This defect is potentially due to the fact that MTM1 is implicated in the conversion of early to late or recycling endosomes (22). This function appears conserved in evolution as a previous study found that the drosophila ortholog of MTM1 was necessary for integrin turnover (23). In this study, we showed that integrin downstream effectors as vinculin and FAK were altered in Mtm1−/y muscles, supporting that integrin trafficking defects lead to alteration of focal adhesion signaling. Subsequently, alteration of focal adhesions caused a defect in adhesion, migration, and fusion of myoblasts lacking MTM1, leading to a decrease in myoblast-to-myotube fusion index that is reminiscent of the myofibers hypotrophy typically seen in patient’s muscles (33).

Overall, this study underlines a key role for MTM1 and BIN1 in the regulation of integrin trafficking and focal adhesion in skeletal muscle and points to the defect in these mechanisms as an important cause of XLCNM that can be efficiently rescued by increasing BIN1 expression through viral delivery. Because MTM1 homologs and BIN1 are implicated in other diseases, especially in peripheral neuropathies or late-onset Alzheimer’s disease and arrhythmia, these findings suggest that a similar mechanism as proposed here might be relevant in other disorders.


Study design

The sample size for each experiment is included in the figure legends. In this study, we used mice (WT and Mtm1−/y TgBIN1 mice) and primary myoblasts obtained from WT and Mtm1−/y pups. The number of mice used was selected on the basis of previous phenotyping analyses conducted in the same model and calculating the statistical power of the experiment. Mtm1−/y mice died before 2 to 3 months of age and were analyzed in this study until 2 months. The other genotypes studied (WT, TgBIN1, Mtm1−/y, Mtm1−/y TgBIN1, Mtm1−/y injected with AAV Ctrl, and Mtm1−/y injected with AAV-BIN1 systemically) were phenotyped and euthanized at the ages noted. Blind phenotyping tests were conducted on mice (WT, TgBIN1, Mtm1−/y TgBIN1, and Mtm1−/y) and on primary cells. Each phenotyping experiment was repeated at least three times. Sample collection, treatment, and processing information are included in the result section or in other sections of Material and Methods. No outliers were excluded in the study.


Primary antibodies used were mouse anti-DHPRα1 (Cav1.1) subunit (abcam2862, Abcam), BIN1 (C99D, Abcam), GAPDH (MAB374, Chemicon), β1 integrin (MAB1997, Chemicon), vinculin (V9131, Sigma-Aldrich), laminin (ab11575, Abcam), collagen VI (NB120-6588, Novus Biologicals), FAK (3285S, Cell Signaling Technology), pY397-FAK (44-624G, Invitrogen), EEA1 (sc-137130, Santa Cruz Biotechnology Inc.), dystrophin (ab15277, Abcam), and rhodamine phalloidin (PHDR1, Cytoskeleton), and anti-BIN1 (R2405) and rabbit anti-DNM2 antibodies (R2680 and R2865) were made onsite at the polyclonal antibody facility of the Institut de Génétique et de Biologie Moléculaire et Cellulaire (IGBMC). Alexa Fluor–conjugated secondary antibodies were purchased from Invitrogen (Alexa Fluor 488, Alexa Fluor 594, and Alexa Fluor 647). Secondary antibodies against mouse and rabbit immunoglobulin G conjugated with horseradish peroxidase were purchased from Jackson ImmunoResearch Laboratories (catalog numbers 115-035-146 and 111-036-045). An enhanced chemiluminesence kit was purchased from Pierce.

Primary myoblasts

Primary myoblasts from WT and Mtm1−/y newborn mice were prepared as previously described in Cowling et al. (31). After extraction, primary cells were plated in Iscove’s Modified Dulbecco’s Medium (IMDM) with 20% fetal calf serum (FCS) and 1% chicken embryo extract (MP Biomedicals) onto 1:200 Matrigel reduced factor (BD Biosciences) and laminin (354232, Corning).

Primary myoblasts adhesion experiments

The experiments were conducted after the protocol adapted from Ratcliffe et al. (36). WT and Mtm1−/y primary myoblasts were trypsinized and resuspended in IMDM with 20% FCS and 1% chicken embryo extract. Primary myoblasts (2.5 × 104) were diluted in 500-μl media and plated in laminin-coated dishes. Cells were allowed to adhere for 10, 20, 30, and 60 min. Primary myoblasts were then washed with warmed medium and fixed with 4% paraformaldehyde (PFA). Immunofluorescence was conducted, and cells were stained with rhodamine phalloidin (Cystoskeleton). After confocal acquisition, cell surface was measured using ImageJ program.

Primary myoblasts fusion index

Primary myoblasts were plated at 4 × 104 on Matrigel. Primary myoblasts differentiation was triggered when cells reach 70% by switching the medium to IMDM with 2% horse serum, and 24 hours later, a thick layer of Matrigel (1:3 in IMDM) was added. Bright-field image was acquired in living myotubes at 24, 48, and 72 hours after differentiation.

Primary myoblasts migration

Primary myoblasts (2 × 104) were plated in IMDM with 20% FCS and 1% chicken embryo extract on laminin-coated dishes. Migration of cells was observed by time lapse with a Leica microscope for 24 hours. Images were taken every 15 min. The migration velocity was measured using Fiji program.

Mouse lines

Mtm1−/y mouse line (129PAS) was previously generated and characterized (18, 39). TgBIN1 (B6J) mice were obtained by the insertion of human BAC (no. RP11-437K23; Grch37 chromosome 2: 127761089-127941604) encompassing the full BIN1 gene with 180.52 kb of genomic sequence. The Bin1+/− mice were previously published (31). Mtm1 heterozygous female mice were crossed with TgBIN1 males to generate four genotypes: WT, TgBIN1, Mtm1−/y TgBIN1, and Mtm1−/y. Animals were maintained at room temperature with 12-hour light/12-hour dark cycle. Animals were euthanized by cervical dislocation following European legislation on animal experimentation and experiments approved by ethical committees (APAFIS #5640-2016061019332648 and 2016031110589922; Com’Eth 01594).

Animal phenotyping

The phenotyping experiments were conducted blinded, and all the experiments were repeated three times for each mouse to ensure reproducibility. The tests were performed by the same examiners to avoid stress and to ensure reproducibility. The daily phenotyping experiments were performed at the same time of the day for all the mice in the cohort, whereas the weekly experiments were always performed on the same day of the week. The following phenotyping tests were performed hanging, grip, rotarod, bar, and footprint tests. The hanging test was performed each week from 3 to 16 weeks of age and monthly from 4 to 12 months. The mice were suspended from a cage lid for maximum 60 s, and the test was repeated three times. The average time each mouse fall from the grid is presented in the graph. The grip test was conducted each month from 3 to 12 months. The four-paw strength was measured using a dynamometer, and the test was repeated three times for each animal in each time point. The average of the three repetitions is reported in the graph. Results are represented as force relative to body weight in grams. The rotarod test was conducted at 5 weeks and 5 months of age. The mice performed the test for 5 days. During day 1, mice were trained to run in acceleration mode on the rotarod. From days 2 to 5, mice ran for a maximum of 5 min with increasing speed (4 to 40 rpm). Each mouse performed the test three times, and the average of three repetitions was represented. We did not use the same cohort of mice at 5 weeks and 5 months. The data reported in the graph correspond to the amount of time the animal run on the rotarod. The bar test was performed placing the mice in a suspended bar. The time to walk along the bar was measured, and the experiment was performed only at 5 weeks of age. In this experiment, only WT, TgBIN1, and Mtm1−/y TgBIN1 mice were tested as Mtm1−/y mice could not walk on a suspended bar. The footprint test was performed at 5 weeks and 5 months of age. For this test, the mice hindpaw placement was analyzed as previously described (31). Hindlimbs were colored with ink, and the placement of mouse hindlimbs was recorded. The angle between the hindlimb position was measured using ImageJ.

Muscle force measurement

Mice were anesthetized using pentobarbital (50 mg/kg) by intraperitoneal injection, and the force of TA was measured using a force transducer (Aurora Scientific) as described previously (39). The absolute maximal force of the TA was measured after tetanic stimulation of the sciatic nerve with a pulse frequency from 1 to 125 Hz. The specific maximal force was determined dividing the absolute maximal force with the TA weight. The fatigue was measured stimulating continuously the sciatic nerve with a frequency of 50 Hz.

AAV transduction of TA muscle

In intramuscular injection, 3-week-old male WT or Mtm1−/y mice were anesthetized by intraperitoneal injection of ketamine (20 mg/ml) and 0.4% xylazine (5 μl/g of body weight). The TA muscle was injected with 20 μl of AAV9 (7 × 1011 viral genome/ml) cytomegalovirus (CMV) human BIN1 isoform 8 without exon 17 or empty AAV9 control diluted in phosphate-buffered saline (PBS). In systemic injection, pups were injected intraperitoneally in the first 24 hours after birth. A volume of 50 μl of AAV9 (1013 viral genome/ml) CMV human BIN1 isoform 8 without exon 17 or with empty AAV Ctrl was used (49). The pups were immediately housed in the cage with their mother after the injection.

Tissue collection

Cervical dislocation was used to euthanize mice after carbon dioxide suffocation. TA and GAS muscles were extracted and then frozen in isopentane cooled in liquid nitrogen. The diaphragm was collected and directly frozen in Optimal Cutting Temperature compound (OCT) on dry ice. The heart, liver, and brain were collected and directly frozen in liquid nitrogen. All tissues were then stored at −80°C.


Eight-micrometer cryosections of TA, GAS, and diaphragm muscles were cut and stained with H&E and SDH for histological analysis. After staining, the images were acquired with the Hamamatsu Nano Zoomer 2HT slide scanner. The percentage of internalized nuclei was counted using cell counter plugin in Fiji software. A macro was used to measure the TA fiber diameter for each genotype. The TA fiber diameter was calculated (>100 fibers per mouse) from three to five mice per group. The percentage of TA muscle fibers with centralized or internalized nuclei was counted in >350 fibers using the cell counter plugin in ImageJ.


Transversal 8-μm cryosections were prepared from TA frozen in isopentane. For longitudinal staining, the TA was incubated overnight in PFA at 4°C and, after three 1× PBS washing, transferred into 30% sucrose overnight at 4°C. After removing the sucrose with PBS washes, muscles were kept at −80°C. The sections were permeabilized with 0.5% PBS-Triton X-100 and then saturated with 5% bovine serum albumin (BSA) in PBS. The primary antibodies diluted in 1% BSA used were as follows: laminin (1:200; ab11575), EEA1 (1:50; sc-137130), α7 integrin (1:50; ab195959), β1 integrin (1:50; MAB1997), vinculin (1:200; V9131), DHPR (1:50; abcam2862), and BIN1 (1:50; C99D). The secondary antibodies were anti-mouse, -rabbit, or -rat, and Alexa Fluor 488, Alexa Fluor 594, and Alexa Fluor 647 were diluted 1:250 in 1% BSA.

Electron microscopy

TA was stored in 2.5% PFA and 2.5% glutaraldehyde in 0.1 M cacodylate buffer. Sections were observed by electron microscopy. To observe T-tubules, potassium ferrocyanide was added to the buffer [0.8% K3Fe(CN)6, 2% osmium, and 0.1 M cacodylate] (45). The triad number per sarcomere and T-tubule direction were measured manually.

RNA extraction and BIN1 isoform 8 detection

TA muscles were lysed in TRIzol reagent (Invitrogen) to extract RNA, and the reverse transcriptase (Thermo Fisher Scientific) was used to obtain complementary DNA (cDNA). To identify human BIN1 isoform overexpressed in Mtm1−/y TgBIN1 mice, BIN1 cDNA was amplified using human BIN1 primers (5′-ACGGCGGGAAAGATCGCCAG and 3′-TTGTGCTGGTTCCAGTCGCT). Human BIN1 cDNAs were cloned into pENTR1A vector and then sequenced using Eurofins Genomics Europe Sequencing laboratory (Eurofins GATC) service.

Protein extraction and Western blotting

TA muscle was lysed in radioimmunoprecipitation assay buffer with 1 mM dimethyl sulfoxide, 1 mM phenylmethylsulfonyl fluoride, and mini EDTA-free protease inhibitor cocktail tablets (Roche Diagnostics) on ice. The protein concentration was measured using the Bio-Rad Protein Assay Kit (Bio-Rad). Loading buffer (50 mM tris-HCl, 2% SDS, and 10% glycerol) was added to protein lysates, and proteins were separated by 8 or 10% in SDS–polyacrylamide gel electrophoresis containing TCE to visualize all tryptophan-containing proteins. After transfer to nitrocellulose, saturation was performed with 3% BSA or 5% milk, and primary and secondary antibodies were added as follows: β1 integrin (1:500; MAB1997), vinculin (1:1000; V9131), BIN1 (1:1000; 2405, IGBMC), MTM1 (1:1000; 2827, IGBMC), and GAPDH (1:100,000; MAB374).

Statistical analysis

The data are expressed as means ± SEM. Graph and curves were made using GraphPad Prism software versions 5 and 6. The unpaired Student’s t test was used to compare two groups. Nonparametric and Kruskal-Wallis or Dunn’s post hoc tests were used to compare multiple groups. One-way ANOVA and Bonferroni or Tukey’s post hoc test were used to compare different groups if the data followed a normal distribution and if the samples analyzed had the same genetic background. P values smaller than 0.05 were considered significant. The number of mice and the tests used for each experiment are listed for each experiment in the figure legends.


Fig. S1. Creation of transgenic mice expressing human BIN1 (TgBIN1 mouse).

Fig. S2. Increased BIN1 expression rescues the motor function and muscle force of Mtm1−/y mice.

Fig. S3. Increased BIN1 expression rescues muscle histology of Mtm1−/y mice at 2 and 7 months old.

Fig. S4. Postnatal intramuscular overexpression of BIN1 rescues muscle force and myofiber organization in Mtm1−/y mice.

Fig. S5. Organ weights and muscle fatigue in the Mtm1−/y mice expressing BIN1 after systemic AAV delivery.

Fig. S6. Extracellular matrix defects in Mtm1−/y muscle.

Fig. S7. Focal adhesion defects in Mtm1−/y myofibers.

Table S1. Breeding strategy and outcome for Mtm1−/y × Bin1−/+ with expected mice and obtained at E18.5 and 10 days after birth.

Table S2. Breeding strategy and outcome for Bin1−/+ × Bin1−/+ TgBIN1 with expected mice and obtained at E18.5 and 10 days after birth.

Table S3. Raw data (Excel file).

Movie S1. Expression of human BIN1 did not generate any obvious clinical phenotypes in mice.

Movie S2. Increased BIN1 expression rescues Mtm1−/y phenotype.

Movie S3. Postnatal systemic injection of AAV-BIN1 rescued Mtm1−/y mice phenotypes.


Acknowledgments: We thank the imaging, animal, and histological platforms of the IGBMC for help and F. Saudou for support. Funding: This study was supported by INSERM, CNRS, University of Strasbourg, ANR-10-LABX-0030-INRT, a French State fund managed by the ANR under the frame program Investissements d’Avenir (10-IDEX-0002), Myotubular Trust, and Sparks the Children’s Medical Research Charity. Author contributions: B.S.C., A.-S.N., and J.L. designed and supervised the research. V.M.L., A.-S.N., M.S., C.K., S.B., S.D., I.P., N.M., and N.B.R. performed the research. V.M.L. performed the animal characterization, injections, protein immunofluorescence, and primary myoblast experiments. A.-S.N performed the protein quantity analysis by Western blot. M.S. and I.P. characterized the Bin1–/– TgBIN1 mice. C.K. and S.B. performed the mice genotyping and helped with animal phenotyping tests. S.D. contributed to primary myoblasts extraction. N.M. acquired electron microscopy images. N.B.R. provided the Masson staining on human skeletal muscle. P. Kessler, P. Koebel, and Y.H. provided materials and analysis tools. P. Kessler designed a macro to analyze fiber size. P. Koebel generated the AAVs to inject in mice. Y.H. provided the TgBIN1 mice. V.M.L., A.-S.N., B.S.C., and J.L. analyzed data. V.M.L., B.S.C., and J.L. wrote the paper. Competing interests: V.M.L., B.S.C., and J.L. submitted a patent on the rescue of XLCNM by BIN1 modulation, which is entitled “Compositions and methods for the treatment of X-linked centronuclear myopathy,” the European patent filing number is 17 306 566.5, and the initial submission date was 11 November 2017. B.S.C. and J.L are cofounders of Dynacure. B.S.C. is now a Dynacure employee (Dynacure, 67400 Illkirch, France). J.L. is a scientific advisor for Dynacure. All other authors declare that they have no competing interests. Data and materials availability: All the data used for this study are included in the main text or in the Supplementary Materials. The AAV serotype 9 used in this study was received from the University of Pennsylvania under a material transfer agreement.
View Abstract

Navigate This Article