The congenital myopathies are a diverse group of genetic skeletal muscle diseases, which typically present at birth or in early infancy. There are multiple modes of inheritance and degrees of severity (ranging from foetal akinesia, through lethality in the newborn period to milder early and later onset cases). Classically, the congenital myopathies are defined by skeletal muscle dysfunction and a non-dystrophic muscle biopsy with the presence of one or more characteristic histological features. However, mutations in multiple different genes can cause the same pathology and mutations in the same gene can cause multiple different pathologies. This is becoming ever more apparent now that, with the increasing use of next generation sequencing, a genetic diagnosis is achieved for a greater number of patients. Thus, considerable genetic and pathological overlap is emerging, blurring the classically established boundaries. At the same time, some of the pathophysiological concepts underlying the congenital myopathies are moving into sharper focus. Here we explore whether our emerging understanding of disease pathogenesis and underlying pathophysiological mechanisms, rather than a strictly gene-centric approach, will provide grounds for a different and perhaps complementary grouping of the congenital myopathies, that at the same time could help instil the development of shared potential therapeutic approaches. Stemming from recent advances in the congenital myopathy field, five key pathophysiology themes have emerged: defects in (i) sarcolemmal and intracellular membrane remodelling and excitation-contraction coupling; (ii) mitochondrial distribution and function; (iii) myofibrillar force generation; (iv) atrophy; and (v) autophagy. Based on numerous emerging lines of evidence from recent studies in cell lines and patient tissues, mouse models and zebrafish highlighting these unifying pathophysiological themes, here we review the congenital myopathies in relation to these emerging pathophysiological concepts, highlighting both areas of overlap between established entities, as well as areas of distinction within single gene disorders.
The congenital myopathies are clinically defined by hypotonia and skeletal muscle weakness and pathologically by the presence of one or more histopathological or ultrastructural hallmarks on muscle biopsy (North, 2011; Romero and Clarke, 2013). In extremely rare instances, congenital myopathy may also be associated with hypertonia as will be discussed below (Jain et al., 2012; C. G. Bönnemann et al., unpublished data). There are three broad histopathological subtypes within the classical congenital myopathies: core myopathies characterized by the presence of myofibres with foci devoid of oxidative enzymes (Jungbluth et al., 2011); centronuclear myopathies (CNMs) defined by the presence of internally located myonuclei (Romero and Bitoun, 2011) (which are now recognized as one of the most common type of congenital myopathy, due to the relatively high prevalence of RYR1 mutations that can also be associated with this histological phenotype) (Amburgey et al., 2011; Maggi et al., 2013); and nemaline myopathy, defined by the presence of electron dense nemaline bodies/rods within myofibres (Wallgren-Pettersson et al., 2011). In addition, there are a number of other structural abnormalities that can be observed in the congenital myopathies including actin aggregates, caps, cylindrical spirals, dystrophic-like features, fibre-type disproportion, intranuclear rods, lobulated myofibres, necklace myofibres and zebra bodies (Clarke, 2011; Goebel and Blaschek, 2011). There are also frequently cases that present with mixed morphologies, such as cores and nemaline bodies (Romero et al., 2009; Olive et al., 2010) or cores and internal nuclei (Wilmshurst et al., 2010).
It has been increasingly recognized that the histological features observed within a muscle biopsy may represent manifestations along spectra of possible abnormalities, rather than discrete pathological entities (Sewry et al., 2008; Nance et al., 2012). For example central nuclei are classically associated with mutations in BIN1, DNM2, MTM1, and SPEG however, more recently, prominent central nuclei have been reported in cases resulting from mutations in CFL2 (Ockeloen et al., 2012), KBTBD13 (Olive et al., 2010), RYR1 (Wilmshurst et al., 2010) and TTN (Ceyhan-Birsoy et al., 2013; Palmio et al., 2014). On the flip side, although congenital myopathies have traditionally been classified purely by their defining histopathology, with the discovery of the underlying genetic cause for an increasing number of patients, the field has also begun in parallel to classify patients based on the causative disease gene, such that there is now, for instance, reference to ACTA1-related myopathies (Nowak et al., 2013), RYR1-related myopathies (Klein et al., 2012; Amburgey et al., 2013; Bharucha-Goebel et al., 2013) and SEPN1-related myopathies (Schara et al., 2008; Scoto et al., 2011). Each of these gene-associated spectra has continued to expand with multiple possible morphological appearances per gene, which together with ever-greater genetic heterogeneity, through the continuous identification of novel disease genes, has led to an increased blurring of the established diagnostic boundaries between the subsets of the classic congenital myopathies. Such novel genes providing new mechanistic aspects have included CCDC78 in a family with autosomal dominant CNM with cores (Majczenko et al., 2012), HACD1 (now known as PTPLA) in a family with a non-specific autosomal recessive congenital myopathy (Muhammad et al., 2013), STAC3 in Native American myopathy (Horstick et al., 2013), STIM1 in tubular aggregate myopathy (although this is not strictly a congenital myopathy) (Bohm et al., 2013a) but also TTN mutation in patients presenting with myopathies characterized by internal nuclei and/or cores (Ceyhan-Birsoy et al., 2013; Chauveau et al., 2014). In this review we have therefore focused on the emerging pathophysiological themes across these entities, under the hypothesis that increased understanding of these mechanistic aspects of disease will not only help classification, but will ultimately facilitate the preclinical development of therapeutic approaches for these diseases (see Fig. 1 for a schematic guide to the cellular context of the pathophysiological concepts discussed in the review and table 1 for a correlation of individual genetic entities and associated pathophysiological themes). Given the focus of this review, we are not redescribing the clinical entities but refer the reader to recent excellent reviews (Nance et al., 2012; Maggi et al., 2013; Romero and Clarke, 2013).
Membrane remodelling defects
Membrane remodelling and the phosphoinositide phosphatases
Membrane remodelling is crucial to a number of biological processes including exocytosis, intracellular transport, synaptic vesicle function, autophagy (see separate discussion below), and membrane repair. These processes involve proteins that regulate lipids; they act as lipid adaptor molecules and function to regulate cytoskeletal organization. Strikingly, a number of genes found to underlie congenital myopathies characterized by the presence of internal myonuclei, are involved in such membrane remodelling [amphiphysin 2/bridging integrator 1, BIN1 (MIM 601248); dynamin 2, DNM2 (MIM 602378); myotubularin, MTM1 (MIM 300415) and myotubularin-related protein 14, MTMR14 (MIM 611089)]; (reviewed in De Matteis and Luini, 2011; Cowling et al., 2012) and most recently, striated preferentially expressed gene, SPEG (Agrawal et al., 2014).
Myotubularin (MTM1), found to be the cause of X-linked myotubular myopathy (XLMTM, MIM 310400), is the canonical member of a family of phosphoinositide phosphatases. The mammalian myotubularin family comprises MTM1 itself and 14 MTM1-related proteins (MTMR1–14). Phosphoinositide phosphatases are crucial to the maintenance of phosphoinositides (PIs) and via the regulation of these metabolites, are central to membrane trafficking and signalling (Di Paolo and De Camilli, 2006; Nicot and Laporte, 2008). The key metabolites regulated by myotubularins are phosphatidylinositol 3-phosphate (PI3P), which acts as a membrane address marker and functions to recruit proteins to the membrane, and PtdIns(3,5)P2 and PtdIns(5)P. MTM1 dephosphorylates phosphoinositides, including PI3P [produced by PI 3-kinase (PI3K) activity] and PtdIns(3,5)P2 [PI3,5P2; produced by PI 5-kinase (PI5K) activity]. Mutations in MTMR14 have been associated with myopathy in fish (Dowling et al., 2010) and possibly CNM in man (Tosch et al., 2006), although the evidence for the latter remains incomplete. It is worth noting that mutation of MTMR2 and MTMR13 are responsible for some cases of autosomal recessive Charcot–Marie–Tooth type hereditary motor and sensory neuropathies (CMT4B), which is of interest because mutations in DNM2 are also seen in both CNM and in autosomal dominant CMT (CMTDIB) (De Matteis and Luini, 2011). How disrupted phosphoinositide regulation leads to neuromuscular disease via possible disturbance of membrane metabolism is still not completely understood; however, studies in animal models are beginning to shed light in this area, highlighting several potential mechanisms that are addressed in the following sections.
Excitation-contraction coupling, transverse-tubule and sarcoplasmic reticulum defects
The precise regulation of intracellular Ca2+ is crucial for normal myogenesis and skeletal muscle function. Excitation-contraction (E–C) coupling exquisitely modulates sarcoplasmic Ca2+ levels during muscle contraction and depends on a number of specialized membranes and structural components including components of the triad, namely T tubules and the voltage-sensing dihydropyridine receptors (DHPR) in the T tubule membrane, and the sarcoplasmic reticulum (SR) and the Ca2+ release channels/ryanodine receptors (RYR1) in the sarcoplasmic reticulum membrane. Structural defects of triads and abnormal localization of both DHPR and RYR1 have been noted in myofibres from patients with CNMs due to MTM1, BIN1 and CCDC78 (MIM 614666) mutations (Al-Qusairi et al., 2009; Dowling et al., 2009; Toussaint et al., 2011; Majczenko et al., 2012). Similar defects have also been observed and more closely investigated in animal models of the various CNMs.
Studies of Drosophila myotubularin (mtm1) showed that mtm1 is central to endolysosomal function, cortical actin remodelling and is also involved in integrin-mediated myofibre attachment (Velichkova et al., 2010). Mtm1 ablation resulted in accumulation of integrin and PI3P on endosomes, suggesting that mtm1 is required for intracellular integrin trafficking and recycling to the sarcolemma (Ribeiro et al., 2011). The phenotype could be rescued by depletion of class II PI3K, confirming that the abnormalities observed are truly IP-related.
Knock-down of mtm1 in zebrafish resulted in delayed hatching and an impaired touch-evoked escape contraction (Dowling et al., 2009). Muscle from the mtm1 morphants displayed centrally located and abnormally enlarged nuclei and skeletal muscle hypotrophy, mimicking the human disease. Interestingly, the levels of PI3P were elevated in the skeletal musculature of mtm1 morphants relative to controls (similar to the Drosophila studies), but they were not elevated in whole embryo extracts, suggesting that myotubularin is the primary phosphoinositide phosphatase in skeletal muscle (Dowling et al., 2009). The skeletal muscle phenotype of the mtm1 morphants was rescued by delivery of MTMR1 and MTMR2, indicating that the primary defect in the morphants was due to a lack of compensation by other phosphoinositide phosphatases (Dowling et al., 2009). In zebrafish (Dowling et al., 2009) and mice (Amoasii et al., 2013), myotubularin localizes to the T tubules as evidenced by co-localization with antibodies to the DHPR. The perinuclear compartments of mtm1 morphant myofibres displayed abnormal tubule-like structures, ranging from dilated T tubule and sarcoplasmic reticulum networks, to severe disorganization of the T tubules and sarcoplasmic reticulum (Dowling et al., 2009).
The most well studied myotubularin model is the Mtm1 knock-out (KO) mouse. These mice develop a generalized and progressive myopathy from 4 weeks of age with increased internalized nuclei leading to death at 6–14 weeks of age (Buj-Bello et al., 2002). Analysis of Mtm1 null mouse muscle found that the sarcoplasmic reticulum membranes were abnormally located within myofibres and that the sarcoplasmic reticulum was dilated (Amoasii et al., 2013). In order to study the SR/ER (endoplasmic reticulum) in isolation, Amoasii and colleagues (2013) examined the endoplasmic reticulum network in myoblasts from Mtm1 null mice and MTM1 patients. The myoblasts lacked triads and T tubules and showed defects in the endoplasmic reticulum network, suggesting that an absence of MTM1 resulted in sarcoplasmic reticulum defects (as the sarcoplasmic reticulum is a specialization of the endoplasmic reticulum) independent of abnormalities of the T tubules and triads. This suggests a general defect in the maintenance of various membranous muscle organelles. In vivo modulation of MTM1 lipid phosphatase activity in normal mouse muscle using viral overexpression of wild-type or mutant (phosphatase inactive) MTM1, resulted in altered PI3P levels specifically at the junctional sarcoplasmic reticulum as well as altered sarcoplasmic reticulum remodelling. Importantly, this study suggests that dysregulation of the MTM1/PI3P pathway, leading to altered modulation of the sarcoplasmic reticulum structure, may represent a crucial mechanism underlying disease pathogenesis in myotubular myopathy.
This same group also showed that the phenotype of the Mtm1 null mouse could be partially rescued by adeno-associated viral delivery of the MTM1Cys375Ser or MTM1Ser376Asn phosphatase-dead mutants, including a significant increase in muscle force generation, suggesting that many features of the disease are not purely due to the enzymatic defect (muscle atrophy, myofibre hypotrophy, abnormal organelle placement) (Amoasii et al., 2012). Consistent with this is the consideration that MTM1 is quite ubiquitously expressed with not only skeletal muscle specific localization, whereas XL-CNM is largely (though not exclusively) a skeletal muscle disease suggesting that defects in interactions between MTM1 and muscle-specific proteins could play a major role in disease causation. The most evident extra-phosphatase activity for MTM1 might be in protein–protein interactions and functioning as a scaffold for the intermediate filament network (desmin in particular) (Hnia et al., 2011); which is involved in proper organelle placement (Amoasii et al., 2012).
The mouse model generated by knocking-in the human disease associated mutation p.Arg69Cys in Mtm1 is less severe compared to the Mtm1 KO model, likely due to some residual activity of the full length mutant myotubularin (Pierson et al., 2012). Mtm1 p.Arg69Cys knock-in (KI) mice are asymptomatic at birth and show normal initial skeletal muscle development. However, by 2 months of age muscle atrophy and weakness is detected. These mice had a median survival age of 66 weeks and died suddenly and unexpectedly, consistent with cardiorespiratory failure. Histologically, skeletal muscle showed many small myofibres, central nuclei, staining suggestive of necklace fibres similar to the human pathology, and altered staining patterns with antibodies to RYR1 and DHPR. Reduced numbers of, and poorly formed, T tubules and triads were identified at the ultrastructural level (Pierson et al., 2012). These morphological abnormalities result in excitation-contraction coupling defects, since for Mtm1 KO and p.Arg69Cys KI mice, electrically-mediated whole muscle-specific force was greatly reduced, whereas there was no change between specific force generated by wild-type and Mtm1 chemically-skinned (in which T tubules and triads are not functioning) myofibres (Lawlor et al., 2013). Most recently, adeno-associated virus-mediated knock-down of Mtm1 in adult mouse muscle resulted in abnormally oriented T tubules that were more frequently appearing as longitudinal (L-) tubules (Joubert et al., 2013) highlighting that myotubularin is continuously required for adult skeletal muscle organelle maintenance.
Beggs and colleagues (2010) reported a Labrador Retriever line with a missense mutation in MTM1; the affected male dogs displayed many of the clinical and histological hallmark features of human MTM1 related XL-CNM, presenting with progressive limb weakness leading to an inability to rise or walk without assistance between 10–19 weeks of age. Clinically, there was severe muscle atrophy. Histologically there were numerous small centrally-nucleated myofibres, resembling myotubes, necklace myofibres, abnormal localization of RYR1 and DHPR and diminished and disturbed T tubule structures (Beggs et al., 2010). Thus in both non-mammalian and mammalian models (small and large) the pathophysiology of MTM1-related disease is remarkably similar and consistent with functionally important abnormalities of triadic structure with consequences on excitation-contraction coupling.
MTMR14 is a phosphatase that shares similar enzymatic activity to MTM1. Although two human MTMR14 variants have been described, and were shown to reduce the enzymatic activity of MTMR14 (Tosch et al., 2006), they are still awaiting further confirmation as a direct cause of CNM or a disease modifier. Of the two variants described, one de novo MTMR14 missense variant (p.Y462C) was identified in an isolated proband in combination with a de novo known DNM2 mutation (p.E368K), the p.Y462C variant was also identified in a control sample. The second proband harboured a p.R336Q variant in MTMR14, this variant was also present in the unaffected father (Tosch et al., 2006). Inactivation of Mtmr14 in mice (Shen et al., 2009; Hnia et al., 2012) and fish causes a myopathy characterized by excitation-contraction coupling defects but without the presence of central or abnormally located myonuclei. In zebrafish, knock-down of mtmr14 resulted in a functional deficit (reduced touch-evoked escape responses and a lack of response to high frequency stimulation) without obvious histopathological abnormalities (Dowling et al., 2010). Hnia et al. (2012) showed that muscle from Mtmr14 KO mice displayed significantly swollen and elongated T tubules and suggested that these structural abnormalities might contribute to the defects in excitation-contraction coupling noted previously in Mtmr14 KO mice (Shen et al., 2009) and morphant fish (Dowling et al., 2010).
Amphiphysin 2 and dynamin 2
Altered T-tubular and sarcoplasmic reticulum systems have also been noted due to mutation in DNM2 and BIN1, which are both known to underlie forms of centronuclear myopathy. Early studies showed that amphiphysin 2 (BIN1) was concentrated at the T tubules and was required for T tubule biogenesis. Expression of BIN1 in non-muscle cells resulted in tubular invaginations of the plasma membrane (Lee et al., 2002).
The original study that showed that BIN1 mutations resulted in autosomal recessive-CNM, also showed that missense mutations resulted in impaired tubule formation and pull-down assays showed that partial truncation of BIN1 abrogated the interaction with DNM2 (Nicot et al., 2007). The authors concluded that BIN1 mutation caused CNM by interfering with remodelling of T tubules and that interaction between BIN1 and DNM2 is required for muscle function and myonuclei positioning. Most recently, BIN1 has also been shown to interact directly with MTM1 (Royer et al., 2013).
BIN1 is ubiquitously expressed and there are at least 12 different splice isoforms, those containing a phosphoinositide domain are found exclusively in skeletal muscle isoforms, which all contain exon 11 (Toussaint et al., 2011). Bohm et al. (2013b) identified the first mutation affecting the muscle-specific BIN1 exon 11 in a consanguineous family presenting with a rapidly progressive and fatal CNM. Mutation of the same acceptor-splice site in exon 11 was identified to underlie canine inherited myopathy of Great Danes (IMGD). Cell-culture studies showed that BIN1 constructs without exon 11 were unable to initiate membrane tubulation (Bohm et al., 2013b). Analysis of patient and inherited myopathy of Great Danes muscle showed central nuclei, fibre atrophy and lobulation, abnormal triads, accumulations of membranes and vacuoles, autophagic vacuoles and mitochondrial accumulations. Proteins of the triad and sarcoplasmic reticulum (RYR1, DHPR, SERCA1) and membrane-trafficking proteins (MTM1, CAV3, DYSF) were also mislocalized in inherited myopathy of Great Danes muscle.
DNM2-related dysfunction has been studied in zebrafish and mice. Zebrafish overexpressing the p.Ser619Leu substitution in DNM2 showed impaired motility, with smaller myofibres, and dilated sarcoplasmic reticulum (Gibbs et al., 2013). More recently it was shown that the DNM2 p.Ser619Leu zebrafish accumulate aberrant vesicle material and show defects of excitation-contraction coupling. Expression of the p.Ser619Leu mutant in non-muscle cells resulted in defective BIN1-dependent tubulation (Gibbs et al., 2014). Thus it seems that an underlying mechanism in the CNMs is defective membrane tubulation.
Studies of a KI mouse model of the most frequently occurring DNM2 mutation (p.Arg465Trp) revealed that muscle from heterozygous KI mice contains increased tubular structures within the intermyofibrillar space and slightly disorganized T tubule and sarcoplasmic reticulum structures, myonuclei were positioned normally (Durieux et al., 2010). Homozygous mice displayed neonatal lethality. Cowling et al. (2011) generated mouse models that overexpressed either wild-type or p.Arg465Trp DNM2 using intramuscular delivery of viral vectors. This study showed that muscle-specific expression of mutant DNM2 resulted in a substantial loss of muscle mass, increased internal and central nuclei, reduced muscle strength and triad defects (including abnormal T tubule morphology and swollen sarcoplasmic reticulum). Swollen sarcoplasmic reticulum were also observed in myofibres overexpressing wild-type DNM2.
Cowling et al. (2014) observed that XLMTM patient muscle and Mtm1−/− mouse muscle contained elevated levels of DNM2. Subsequently Mtm1−/− mice that were hemizygous for Dnm2 were generated. Reduction of DNM2 levels, restored the lifespan of Mtm1−/− mice, improved whole-body strength and diaphragm function (Cowling et al., 2014). Structural abnormalities including myofibre atrophy, nuclear positioning and sarcomere and triad disorganization, were also reduced in Mtm1−/− muscle when DNM2 levels were reduced. This study suggests that MTM1 and DNM2 regulate skeletal muscle organization and function in a finely balanced system in which MTM1 seems to be negative regulator of DNM2 so that reduction of DNM2 restores this balance in MTM1 deficiency. This approach could thus be beneficial in XLMTM.
Agrawal and colleagues (2014) identified SPEG as a novel binding partner of MTM1 in a yeast two-hybrid screen and subsequently showed that it likely localized to the terminal cisternae of the sarcoplasmic reticulum. Recessive mutations of SPEG were identified in patients with a severe form of CNM, two cases also had cardiomyopathy (Agrawal et al., 2014). Speg−/− mice display perinatal death due to dilated cardiomyopathy (Liu et al., 2009), evaluation of Speg−/− skeletal muscle revealed increased central nuclei (Agrawal et al., 2014). While attractive to consider, it remains to be seen whether loss of SPEG leads to defects in membrane remodelling and tubulation, as occurs in MTM1-, BIN1- and DNM2-related CNM.
Triadic Ca2+ release channel: ryanodine receptor
The so-called ryanodine receptor (RYR1) is the major Ca2+ release channel of the sarcoplasmic reticulum. It is located at the triad, where it directly interacts with the voltage-sensitive dihydropyridine receptor in the T tubule. Via this connection it mediates excitation initiated Ca2+ release into the sarcoplasm (Zorzato et al., 1990).
Myopathies resulting from autosomal dominant RYR1 (MIM 180901) mutations [including malignant hyperthermia susceptibility (MHS; MIM 145600), central core disease (MIM 117000), multi-minicore disease, congenital fibre-type disproportion (CFTD), some subtypes of CNM and King-Denborough Syndrome] are thought to arise due to ‘leaky’ RyR1 channel activity, channel instability or reduced Ca2+ conductance (Wilmshurst et al., 2010). Recessive mutations may cause misolocalized RyR1 channels, or reduced RYR1 abundance (Zhou et al., 2013). KI RYR1 p.Tyr522Ser (this mutation is responsible for MHS in patients) mice exhibit malignant hyperthermia susceptibility, (MHS)-like responses to halothane and heat stress. Histologically the muscle is characterized by the step-wise development of contracted sarcomeres and core-like regions, secondary to mitochondrial dislocation and triad disruption (Boncompagni et al., 2009) (structured cores), followed by sarcomeric disintegration in the core (unstructured cores). KI p.Tyr522Ser myofibres also display disorganized T tubules and triad positioning (Boncompagni et al., 2009). Another KI mouse model (KI p.Ile4898Thr) of an excitation-contraction coupling RYR1 mutation (a common central core disease mutation in patients) exhibits slowly progressive myopathy with skeletal muscle weakness, impaired mobility and hindlimb paralysis. The mice also developed age-dependent structural lesions within their muscle, sarcomeric misalignment and Z-disc streaming progressing to mini-cores, cores and then also nemaline bodies (Zvaritch et al., 2009). Thus, this mouse model also models the development of a ‘mixed’ congenital myopathy, core-rod disease.
The pathogenesis of recessive RYR1 mutations has only begun to be elucidated. Zhou et al. (2013) showed that RYR1 protein levels are reduced in autosomal recessive RYR1 patient muscle, as are DHPR levels, and that the localization of the alpha-subunit of the DHPR is also perturbed in patient muscle and in cultured myotubes targeted with siRNA to RYR1 (Zhou et al., 2013). These alterations resulted in impaired excitation-contraction coupling and Ca2+ handling. Unexpectedly, IP3 receptors (a downstream effector of PI3,5P2 which resides in the ER/SR and upon IP3 binding mediates Ca2+ release) were more abundant in patient muscle and cultured myotubes deficient for RyR1, likely representing a secondary upregulation of the IP3-Ca2+ axis. In cultured muscle cells, IP3 receptors are expressed at the nuclear envelope and give rise to slow nucleoplasmic Ca2+ transients that modulate gene expression (reviewed in Carrasco et al., 2004). While the IP3-Ca2+ signalling pathway does not play a substantial role in skeletal muscle excitation-contraction coupling under normal conditions, it's upregulation due to RyR1 deficiency and potential effects on gene expression, may contribute to the underlying pathophysiology of this condition and could represent a potential therapeutic target (Zhou et al., 2013).
Excitation-contraction coupling and the sarcoplasmic reticulum in other congenital myopathies
Defects in sarcoplasmic reticulum function have also been implicated in nebulin (NEB, MIM 161650) deficiency, usually associated with the histological phenotype of nemaline myopathy. The sarcoplasmic reticulum Ca2+-ATPase pump (SERCA) inhibitor, sarcolipin, is upregulated 70-fold in muscle from Neb KO mice (Witt et al., 2006). Speed of Ca2+ uptake by SERCA was significantly slower in sarcoplasmic reticulum vesicles isolated from Neb KO myofibres compared to wild-type, contributing to a slower speed of muscle relaxation (Ottenheijm et al., 2008). Thus it appears that nebulin may also be involved in modulating sarcoplasmic reticulum Ca2+ handling.
Most recently, mutation of the SH3 and cysteine-rich domain 3 gene (STAC3; MIM 615521), encoding a novel excitation-contraction coupling protein (Nelson et al., 2013; Reinholt et al., 2013), has been found to underlie Native American myopathy (Horstick et al., 2013). Native American myopathy (MIM 255995) is characterized by congenital muscle weakness, susceptibility to malignant hyperthermia, joint contractures, dysmorphic features including ptosis, and non-specific myopathic changes on muscle biopsy (no rods, cores or internally-placed nuclei were observed) (Stamm et al., 2008a, b). Thus both primary defects of excitation-contraction coupling (due to RYR1 and STAC3 mutations) and secondary defects of excitation-contraction coupling (arising from the structural triad abnormalities seen in the CNMs due to mutations in DNM2 and MTM1, but also NEB mutations) can be identified in the congenital myopathies. In this context, as noted above, it is also of interest that dominantly acting mutations in the sarcoplasmic Ca2+ sensor STIM1 have now been found to underlie a form of tubular aggregate myopathy (which although being associated with characteristic morphological findings in the biopsy is not strictly speaking a congenital myopathy) (Bohm et al., 2013a). Given the crucial roles of the triad and sarcoplasmic reticulum in muscle function and this apparently widespread perturbation of these entities in the congenital myopathies, there is an emerging view that impaired excitation-contraction coupling and Ca2+ homeostasis are central to disease pathogenesis and may thus represent an important therapeutic target across a number of the congenital myopathies.
Neuromuscular junction defects
The neuromuscular junction (NMJ) is a crucial, deeply in-folded, specialization of the muscle membrane. Aberrant NMJs and clinical overlap with the congenital myasthenic syndromes were reported already in the earlier literature, as a feature of then genetically uncharacterized CNMs (Sher et al., 1967; Elder et al., 1983; Fidzianska and Goebel, 1994). In light of this and the previous discussion of the role of membrane remodelling and shaping, the NMJ has recently come into focus in our understanding of the pathogenesis of the congenital myopathies, in particular for the CNMs.
Dowling and colleagues (2012b) re-examined two mouse models of MTM1-associated disease (Mtm1 KO and KI p.Arg69Cys) and found that both models showed features consistent with a defect in neuromuscular transmission. Mtm1 KI mice displayed reduced treadmill-running endurance and increased grip fatigue compared to wild-type mice, indicating exercise intolerance and fatigable muscle weakness: two clinical indicators of defective neuromuscular transmission. Grip fatigue was attenuated following administration of neostigmine [an acetylcholine esterase (AChE) inhibitor], supporting the presence of a NMJ defect. Evaluation of Mtm1 KO and KI p.Arg69Cys muscle highlighted an increase in the area of individual NMJs and a reduction in the complexity of the postsynaptic junctional folds (Dowling et al., 2012b). The authors showed that trafficking of AChR from the endosomal compartment to the sarcolemma was impaired in Mtm1 KO myotubes and propose that aberrant regulation of junctional membrane turnover and abnormal NMJ trafficking may underlie the altered ultrastructural appearance and consequently abnormal function of the NMJs (Dowling et al., 2012b). Treatment of Mtm1 KI p.Arg69Cys mice with pyridostigmine (another AChE inhibitor) also resulted in an improvement in motor function (Dowling et al., 2012b).
Similarly, in mtm1 morphant zebrafish (Dowling et al., 2009), disorganized NMJs were noted (Robb et al., 2011) and movement and provoked swimming behaviour were dramatically improved upon treatment with an AChE inhibitor (edrophonium) (Robb et al., 2011).
In a zebrafish model overexpressing the p.Ser619Leu mutation of DNM2, a human mutation associated with severe CNM presenting in the neonatal period, severe motor defects were observed (Gibbs et al., 2013). Disorganized NMJs, including abnormal clustering of acetylcholine receptors and a reduced number of endplates per myofibre, were a striking feature of this model. Treatment with edrophonium improved motor function (escape response) of p.Ser619Leu larvae, but not uninjected controls or wild-type larvae, indicating that AChE inhibition improved the phenotype caused by the presence of the mutant dynamin 2 (Gibbs et al., 2013).
Studies of patient biopsies showed abnormal endplate morphology (irregular and less complex post-junctional folds) (Fidzianska and Goebel, 1994) and abnormal AChE staining (Elder et al., 1983). Later, examination of five patients harbouring either the DNM2 p.Ser619Leu or p.Glu368Lys substitutions, found that two of the patients had electrophysiological recordings suggestive of impaired neuromuscular transmission (Gibbs et al., 2013). Importantly, Gibbs et al. (2013) and others have reported anecdotally that treatment of MTM1, DNM2 and genetically-unresolved CNM patients with pyridostigmine, resulted in improved muscle strength and function, and reduced fatigability (Liewluck et al., 2011; Robb et al., 2011). Thus, modulation of neuromuscular transmission through AChE inhibition or enhanced acetylcholine release from the presynaptic nerve terminal (as proposed by Dowling et al., 2012b) may represent a viable therapeutic approach to attenuate the disease severity in patients with CNMs, that may lead to an improvement in quality of life. This approach will now need to be evaluated in a controlled clinical trial in patients with CNM.
The neuromuscular junction in other congenital myopathies
NMJ abnormalities may also occur in other congenital myopathies. Munot et al. (2010) presented two unrelated cases of congenital fibre-type disproportion, due to two different heterozygous slow α-tropomyosin (TPM3, MIM 191030) mutations, which mimicked myasthenic syndromes. Single fibre EMG recordings for both cases showed a jitter consistent with impaired neuromuscular transmission, however detailed histological assessment of the NMJ was not performed. Pyridostigmine administration resulted in a modest functional improvement in the only patient that undertook treatment (Munot et al., 2010). Thus, the link between TPM3 mutations and confirmed abnormalities on neuromuscular transmission has to remain tentative for now.
Most recently, it was shown that pyridostigmine administration resulted in clinical improvement in two siblings with recessive RYR1-related congenital myopathy associated with striking fatigable ptosis (Illingworth et al., 2014). This study extended the phenotype of RYR1-related myopathy to include a myasthenic-like phenotype. Pyridostigmine treatment resulted in improved facial expression and stamina and decreased ptosis. Again, a detailed analysis of neuromuscular transmission was not possible in these patients leaving the final link open.
All in all, it seems to be worthwhile to investigate the NMJ, histologically, ultrastructurally, and physiologically, in models of other congenital myopathies, to assess whether there might be a role for aberrant NMJ function beyond the CNMs, and to what extent modulation of neuromuscular transmission may also represent a therapeutic option for these patients.
Abnormal distributions of mitochondria are a well-recognized hallmark feature of the congenital myopathies that are characterized by the presence of cores or core-like regions within myofibres. In fact, cores are defined by being devoid of mitochondria, recognizable by the appropriate stains. Abnormal accumulations, localization and/or ultrastructure of mitochondria have been reported in central core disease, multi-minicore disease, centronuclear myopathy, and nemaline myopathy (Sanoudou et al., 2006; Hnia et al., 2011; Dowling et al., 2012a; Mokbel et al., 2013) with the possibility that this abnormal localization will also lead to abnormal function regionally in muscle.
A recessive ryr1 zebrafish model (due to a spontaneous mutation in the fast-muscle specific ryanodine receptor gene, ryr1b, that results in a premature stop codon) showed reduced levels of ryr1 (5–10% of wild-type), defective excitation-contraction coupling and a severe motor phenotype (Hirata et al., 2007). The oxidative stress pathway was found to be impaired in this model and it was confirmed that ryr1 zebrafish had elevated basal oxidant activity arising from mitochondrial-derived reactive oxygen species, which were notably not present in dystrophic zebrafish models (Dowling et al., 2012a). Furthermore, cultured RYR1 patient-derived myotubes also displayed features consistent with increased oxidative stress (Dowling et al., 2012a). The same authors showed that treatment of RYR1 patient myotubes and ryr1 zebrafish with the antioxidant N-acetylcysteine (NAC) corrected the oxidative stress and restored zebrafish motor function and muscle pathology. These data implicate mitochondrially-mediated oxidative stress as a contributing factor in the pathophysiology of RYR1-related myopathies and provide important preclinical evidence in support of antioxidant therapy as a potential treatment intervention in these diseases.
Autosomal recessive mutations in selenoprotein N1 (SEPN1, MIM 606210) underlie some cases of multi-minicore disease and congenital fibre-type disproportion; in addition to congenital muscular dystrophy with rigid spine (MIM 602771) and myopathy with Mallory body-like inclusions (Scoto et al., 2011). SEPN1 is an endoplasmic reticulum glycoprotein that seems more abundant in foetal than adult tissues (Petit et al., 2003). SEPN1 is thought to be involved in Ca2+ homeostasis and protection from oxidative stress (Arbogast and Ferreiro, 2010; Castets et al., 2012).
Sepn1−/− mice show normal growth and lifespan and only minor defects of muscle morphology and function under basal conditions. However, under physical challenge (e.g. forced swimming) Sepn1−/− mice developed an obvious phenotype characterized by reduced mobility, body rigidity and curvature of the spine - thus recapitulating aspects of the human disease (Rederstorff et al., 2011). Despite this phenotype, no cores or core-like regions were observed in Sepn1−/− muscle, and mitochondria were ultrastructurally normal (Rederstorff et al., 2011).
Similar to RYR1 mutant myotubes, cultured SEPN1-deficient patient myotubes also have a higher basal oxidative activity and patient-derived fibroblasts are more susceptible to H2O2 (Arbogast et al., 2009). This elevated sensitivity to H2O2 could be mitigated by pretreatment with NAC (Arbogast et al., 2009). It has been suggested therefore that selenoprotein SEPN1 may be involved in buffering reactive oxygen species and/or repairing damaged proteins (Castets et al., 2012).
RYR1 activity is regulated by its oxidative status (Oba et al., 2002) and both RYR1 and SEPN1 are enriched in sarcoplasmic reticulum terminal cisternae fractions (Jurynec et al., 2008), which resulted in the hypothesis that SEPN1 might modulate RYR1 activity. In support of such a connection, Ca2+ handling was altered in the sepn1 morphant zebrafish (Jurynec et al., 2008) and patient myotubes (Arbogast et al., 2009).
However, it remains to be fully explained how SEPN1 might modulate RYR1 given that SEPN1 is barely detectable in adult myofibres (Castets et al., 2012). Nonetheless, there is emerging evidence that links SEPN1 to RYR1 regulation, and both RYR1 and SEPN1 to increased oxidative stress, perhaps via mitochondria.
In striated muscle, there are three principal tropomyosin isoforms: α-tropomyosinfast (TPM1), α-tropomyosinslow (TPM3) and β-tropomyosin (TPM2). Tropomyosins interact with the troponin complex to regulate actin-myosin interactions. Mutations of TPM2 and TPM3 cause congenital myopathies with nemaline bodies, caps, or fibre-type disproportion (reviewed in Romero and Clarke, 2013). Sanoudou et al. (2006) reported enlarged mitochondria with abnormal structures and elevated levels of lipofusin indicative of oxidative stress, in muscle from nemaline myopathy mice harbouring a p.Met9Arg mutation in TPM3. Most recently, subsarcolemmal mitochondrial accumulations and a greater variation in the abundance of respiratory chain complexes and porin proteins has been described in muscle biopsies from some patients harbouring the p.Lys7del mutation in TPM2 compared with healthy individuals (Davidson et al., 2013; Mokbel et al., 2013), indicating a contribution of mitochondrial dysregulation and cellular stress to the pathogenesis of some TPM2-related myopathies.
Abnormal mitochondrial localization has also been reported in MTM1 patient cells (Buj-Bello et al., 2008) and the mtm1 morphant zebrafish (Dowling et al., 2009). In-depth examination of MTM1 patient muscle and Mtm1 deficient mouse muscle and cells, found that there was a collapse of the mitochondrial network, with the presence of enlarged mitochondria containing fewer cristae (Hnia et al., 2011). Mtm1-deficient cells also had reduced cytochrome c oxidase activity and lower ATP levels than wild-type cells (Hnia et al., 2011), suggesting altered mitochondrial function as a contributing factor to the pathophysiology of MTM1-related CNM. Hnia et al. (2011) identified that in skeletal muscle, the intermediate filament protein desmin regulates mitochondrial dynamics and morphology through direct interaction with MTM1. These studies found that Mtm1 KO mouse muscle, Mtm1 siRNA-treated C2C12 and MTM1 patient myoblasts, contained elevated levels of the intermediate filament protein desmin, and abnormal desmin aggregates. MTM1 mutations that cause CNM and are located within the critical region for desmin binding abolished the MTM1-desmin interaction. These data suggest that in the absence of MTM1, desmin aggregates impair intermediate filament function and contribute to aberrant internal organelle positioning and functioning (including mitochondria and nuclei). In this context it is of note that deletion of desmin in mice also results in altered distribution, morphology and activity of mitochondria (Milner et al., 2000). Similar abnormalities are seen in patients with DES-related myofibrillar myopathy (MIM 601419) (Reimann et al., 2003; Bar et al., 2007).
It remains to be seen whether mitochondrial dysfunction and oxidative stress are more widely applicable to the disease process in multiple congenital myopathies and how much they are able to influence disease severity and whether they could represent a more widely applicable therapeutic target across various disorders.
Myofibrillar force generation
Defects of sarcomeric function and myofibrillar force generation are the most obvious consequence to consider in congenital myopathies arising due to defects in genes encoding thin filament proteins or proteins interacting with or regulating such proteins (Nowak et al., 1999, 2013; Pelin et al., 1999; Laing and Nowak, 2005; Ottenheijm et al., 2012). However, as we will review, the exact effects on myofibrillar force generation may differ from gene to gene and even between different mutations in the same gene.
Myofibrillar structure and function has been most extensively studied in patients harbouring mutations in NEB and the Neb KO (Ottenheijm and Granzier, 2010) and NebΔExon55 mouse models (Ottenheijm et al., 2013). Studies of nebulin-deficient mice (Neb), zebrafish (neb) and NEB patient muscle, have revealed a greatly impaired force-generating capacity (Bang et al., 2006; Witt et al., 2006; Lawlor et al., 2011a; Ottenheijm et al., 2012, 2013; Telfer et al., 2012). Nebulin is thought of as a molecular ruler determining the length of the actin thin filament, thus it is perhaps not surprising that a unifying feature of nebulin-deficient muscle is a greater range in thin filament length, with many sarcomeres containing shorter thin filaments, that result in reduced maximal active tension (Bang et al., 2006; Witt et al., 2006; Ottenheijm et al., 2009, 2013; Telfer et al., 2012). Nebulin-deficient myofibrils also have a reduced Ca2+ sensitivity of force production (Witt et al., 2006), a slower rate of tension redevelopment and an elevated tension cost, meaning more ATP has to be expended to generate a given tension. This suggests that there is a reduced fraction of force-generating cross-bridges, which, in turn, contributes to muscle weakness (Ottenheijm et al., 2010; Lawlor et al., 2011a). Recently studies have shown that in vitro treatment of nebulin-deficient myofibres with a fast skeletal muscle troponin activator (CK-2066260) enhanced force production at submaximal activating Ca2+ levels (de Winter et al., 2013; Lee et al., 2013; Ottenheijm et al., 2013). This compound belongs to a class of fast skeletal muscle troponin activators that were recently shown to augment force development by slowing the rate of Ca2+ dissociation from troponin C (TnC) and enhancing cross-bridge formation at a given Ca2+ concentration (Russell et al., 2012). These fast myofibre specific troponin activators may be applicable to a range of skeletal muscle disorders, as they have also been shown to be effective in myasthenic patients (Russell et al., 2012). In the context of the congenital myopathies this class of compounds might of course be particularly beneficial in patients with disorders of the thin filament with abnormal contractility such as is the case in some nemaline myopathies. In addition it has also been suggested that they would be applicable in disorders that manifest defects of excitation-contraction coupling, including patients with core and central nuclear myopathies, where it could be possible to maximize the force produced from the reduced Ca2+ signalling during muscle contraction. It would of course only strengthen fast twitch type fibres, which in congenital myopathies are frequently fewer in number, but larger (as seen in fibre-type disproportion, a common histological pattern across various congenital myopathies). However, as described later, such treatments could prove to be problematic in cases where mutations result in increased Ca2+ sensitivity of the myofibrillar apparatus, where such a treatment could lead to worsening of the condition.
Skeletal muscle alpha-actin
Over 200 different mutations have been identified in the skeletal muscle α-actin gene (ACTA1, MIM 102610), the major component of the thin filament. The majority of these disease-associated mutations act in a dominant manner, causing significant weakness and hypotonia (Laing et al., 2009), except for a case notably presenting with muscle stiffness and hypertonia (Jain et al., 2012). Rare recessive loss-of-function cases exist and are compatible with survival because of compensatory upregulation of cardiac α-actin in skeletal muscle (Nowak et al., 2007). A range of structural histological lesions in muscle are associated with ACTA1 mutations, including actin aggregates, nemaline bodies (sarcoplasmic and intranuclear), caps, core-like regions, cytoplasmic bodies, dystrophic features, fibre-type disproportion and zebra bodies, while some cases only show minimal changes within the muscle biopsy (reviewed in Nowak et al., 2013). It remains unclear how the presence of mutant ACTA1 protein (in dominant cases) resulting in hypotonic or hypertonic phenotypes and the absence of ACTA1 (in the recessive ‘null’ cases) (Nowak et al., 2007) can give rise to some of the same pathologies, such as nemaline bodies. In this context it is of note that TPM3-domaint as well as null mutations can both give rise to nemaline bodies (Lehtokari et al., 2008). This, together with the fact that nemaline bodies can also be seen in seemingly unrelated genetic entities such as in RYR1 mutations, suggests that the development of the nemaline bodies is a result of dysfunction of the sarcomere rather than due to a direct effect of the presence of mutant proteins or muscle weakness per se.
Studies of transgenic mice expressing the nemaline myopathy-associated ACTA1 mutation p.Asp286Gly have revealed that there is severe muscle weakness resulting in early lethality in some mice, mimicking the severity seen in patients (Ravenscroft et al., 2011). Studies of isolated myofibres and whole muscle found a reduced Ca2+ sensitivity compared to wild-type (Ravenscroft et al., 2011), likely due to impaired actin–actin interactions within the filamentous actin (Ochala et al., 2012). A similar decrease in the force-frequency relationship (indicative of decreased Ca2+ sensitivity) has been found in muscle from another mouse model of nemaline myopathy (Ravenscroft, unpublished data), the KI Acta1H40Y (p.His40Tyr) line (Nguyen et al., 2011). In vitro studies of other ACTA1 mutations, p.Met132Val and p.Phe352Ser, have shown an increase in steady-state isometric force production associated with the substitutions (Marston et al., 2004; Lindqvist et al., 2012). Drosophila with ACTA1 mutations, in the indirect flight muscles, were flightless and their muscle showed myofibrillar disorganization (Sevdali et al., 2013). One mutation, p.Asp292Val, which causes an almost complete absence of Ca2+ activated contraction (Clarke et al., 2007), resides within a region of actin likely to bind tropomyosin and was shown to rescue the hypercontractile phenotype of troponin mutant flies (Sevdali et al., 2013). Jain et al. (2012) reported a patient with nemaline myopathy that presented with congenital hypertonia and stiffness, due to a de novo p.Lys328Asn substitution. It was shown experimentally, using the in vitro motility assay, that the mutant actin extracted from the patient's muscle biopsy displayed a substantially increased Ca2+ sensitivity (Jain et al., 2012). Thus Ca2+ sensitizers, such as the troponin activators, may be applicable to some ACTA1 mutations where there is reduced Ca2+ sensitivity or reduced steady-state force production, but not others. While the Jain et al. (2012) report is the only published example of a hypertonic actin myopathy, this clinical entity is increasingly recognized and may also be associated with other nemaline myopathy associated gene mutations (C.G. Bönnemann et al., unpublished data), thus a congenital myopathy should be considered even in cases presenting with hypertonia and stiffness.
Given the central role of tropomyosin as the crucial Ca2+-regulated inhibitor of actin-myosin interactions, it was thought that tropomyosin mutations would alter sarcomeric force production and Ca2+ sensitivity. However, there does not seem to be a simple unifying mechanism associated with these mutations. The most widely studied tropomyosin mutation is the slow α-tropomyosin (α-Tm, TPM3) p.Met9Arg substitution (Laing et al., 1995; Corbett et al., 2001, 2005). Expression of α-Tmp.Met9Arg in cardiomyocytes resulted in a greatly reduced Ca2+ sensitivity compared with cardiomyocytes expressing α-TmWT (Michele et al., 1999). Analysis of muscle biopsies from individuals with nemaline myopathy, or congenital fibre-type disproportion, due to dominant or recessive mutations in TPM3, revealed that changes in cross-bridge cycling kinetics underlie the muscular weakness (Ottenheijm et al., 2011). Myofibres from all five TPM3 patients examined in this study, showed a significantly lower rate of force redevelopment and higher tensions costs (as measured by ATP consumption rates per force production) suggesting that the proportion of force-generating cross-bridges is diminished in patient myofibres. A striking feature of these TPM3 myofibres was a greatly increased Ca2+ sensitivity, resulting in significantly higher pCa50 values (Ottenheijm et al., 2011). Notably, this increased sensitivity does not lead to muscle stiffness but is rather thought to represent a compensatory mechanism counteracting some of the effect of the diminished force generated by the cross- bridges. The sum of this may then underlie the comparatively mild phenotype of some of these patients, since force generation at submaximal Ca2+ concentrations was comparable between patients and healthy controls.
A similar mechanism to the altered cross-bridge cycling kinetics described for the TPM3 mutations was also observed in muscle expressing the p.Arg133Trp substitution in β-tropomyosin (TPM2) (Ochala et al., 2007). Marttila et al. (2012) studied nemaline and cap myopathy-associated mutations in the β-tropomyosin gene (TPM2) and showed that some mutations resulted in altered affinity for actin (three showed reduced affinity, one showed increased affinity) while others showed reduced Ca2+ sensitivity of contractility (p.Glu117Lys and p.Glu41Lys). Previously, studies of the TPM2 mutant p.Glu41Lys also showed a reduced Ca2+ sensitivity of force production, which could be mitigated by treatment of myofibres with a Ca2+ sensitizer (EMD 57033) (Ochala et al., 2008). A reduced actin binding affinity has also previously been observed with the TPM2 p.Arg91Gly substitution associated with a distal arthrogryposis phenotype (Robinson et al., 2007). Most recently, studies of the recurrent TPM2 p.Lys7del mutation associated with a distal arthrogryposis and potentially worsening muscle contracture phenotype (Davidson et al., 2013; Mokbel et al., 2013) have shown that myofibres harbouring the mutant β-Tm produce comparable specific forces to control myofibres but that the Ca2+ sensitivity of contraction is greatly increased (Mokbel et al., 2013). Davidson et al. (2013) also showed, by expressing the p.Lys7del mutation in zebrafish, that head-to-tail oligomerization of β-tropomyosin polymers was likely to be impaired, that the mutant tropomyosin failed to localize properly within the thin filament and that sarcomere length was altered. Actin-β-Tm co-sedimentation assays found that the actin binding affinity of the mutant β-Tm was 6-fold lower than wild-type β-Tm (Mokbel et al., 2013). A study by Marston et al. (2013) showed that all tested TPM2 or TPM3 mutations residing within repeating structural motifs that bind actin caused a gain-of-function through increased Ca2+ sensitivity, while mutations outside of those regions resulted in reduced Ca2+ sensitivity. It will be important to also correlate the other mutations modelled in this study with clinical observations. Interestingly, some TPM2 p.Lys7del patients presented with contractures at birth or developed contractures postnatally (Davidson et al., 2013; Mokbel et al., 2013) and one case was noted to have generalized muscle hypertonia (Mokbel et al., 2013), thus increased Ca2+ sensitivity associated with specific TPM2 mutations may lead to muscle stiffness with resulting contractures, adding to the differential diagnosis of this phenomenon (Sung et al., 2003; Robinson et al., 2007; Jain et al., 2012; Mokbel et al., 2013). As alluded to previously, Ca2+ sensitizers (Ochala, 2010) are unlikely to be of therapeutic benefit in patients where there already is enhanced sarcomeric Ca2+ sensitivity and may even exacerbate the development of stiffness and contractures. Thus, knowledge of the exact physiological consequences of a mutation will be essential in order to select potential treatments for a given patient.
The first skeletal muscle-specific MHC gene associated with disease was MYH2. There are a handful of myopathy cases in the literature arising due to autosomal dominant or recessive mutation of the myosin heavy chain IIA gene (MYH2; MIM 160740) (Martinsson et al., 2000; Tajsharghi et al., 2005, 2010; D'Amico et al., 2013). The phenotype of MYH2 cases can vary greatly with some cases presenting with multiple joint contractures (Martinsson et al., 2000; de Winter et al., 2013) and severe respiratory distress at birth (D'Amico et al., 2013) to cases of childhood or adult onset of mild muscle weakness and myalgia (Tajsharghi et al., 2005, 2010). In all cases the disease was either non-progressive or was slowly progressive in the third to fifth decade of life, ophthalmoparesis from birth was a prominent feature in all cases. In the more rarely occurring autosomal recessive cases, an absence of MHC IIA myofibres has been demonstrated (Tajsharghi et al., 2010). The surprisingly mild skeletal muscle phenotype of the recessive MYH2 null patients is likely due to compensation by the other MHC isoforms and contrasts with the profound severity observed in patients with null mutations of other sarcomeric genes [e.g. ACTA1 (Nowak et al., 2007), CFL2 (Agrawal et al., 2007; Ockeloen et al., 2012) and NEB (Ottenheijm et al., 2009)]. The proportion of MHC IIA myofibres increases with age and it is thus hypothesized that the progression of disease later in life is a direct result of the increasing proportion of MYH2 mutant protein (p.E706K) in the autosomal dominant cases (Tajsharghi et al., 2002). Thus, similarly to the autosomal dominant ACTA1 mutations, the mutant protein load seems to be critical to disease severity.
Autosomal dominant mutations in the slow β-myosin heavy chain gene (MYH7; MIM 160760) are known to cause two congenital myopathies: hyaline body (myosin storage) myopathy (Bohlega et al., 2004) and multi-minicore disease (Cullup et al., 2012; Clarke et al., 2013), in addition to Laing distal myopathy (MIM 160500) and cardiomyopathies. A recent report suggests that MYH7 mutations may be more frequent in the congenital myopathies than previously suggested (Clarke et al., 2013). Typically mutations altering amino acid residues within the globular head domain cause a cardiomyopathy whereas those altering amino acids within the tail rod domain cause a skeletal muscle phenotype (Armel and Leinwand, 2009, 2010); however, in recent years this distinction has become less obvious (Muelas et al., 2010; Homayoun et al., 2011; Lamont et al., 2014). The pathophysiology of MYH7-related cardiomyopathies has begun to be elucidated, the maximum force generating capacity of single cardiac myocytes and isolated myofibrils harbouring MYH7 mutations were lower than controls, even when normalized to cross-sectional area. Cardiomyocytes from patients harbouring MYH7 mutations were hypocontractile at maximal and submaximal Ca2+ concentrations (Witjas-Paalberends et al., 2013). Furthermore, cardiomyocytes from MYH7 patients show increased Ca2+ sensitivity, likely due to hypophosphorylation of PKA and smaller length-dependent activation (Sequeira et al., 2013); however these features were observed in all cases of hypertrophic cardiomyopathy (HCM) and are thus likely to be a generalized response to the cellular remodelling that occurs in hypertrophic cardiomyopathy rather than a direct result of the MYH7 gene defect. Similar studies now need to be done for the MYH7 mutations associated with skeletal muscle disease.
Autosomal dominant mutations in two genes (MYH3 and MYH8), encoding developmental MHC isoforms, are responsible for some presentations of congenital distal arthrogryposis, which postnatally is not progressive, thus suggesting a largely developmental mode of action (reviewed in Tajsharghi and Oldfors, 2013).
Skeletal muscle bulk is determined by the net effect of hypertrophy and atrophy processes: protein synthesis and degradation (comprising atrophy and autophagy) (reviewed in Schiaffino and Mammucari, 2011). Atrophy and disordered autophagy are common in a wide variety of muscle disorders so that it can be challenging to determine how specific or non-specific these phenomena are in any given disorder. In the congenital myopathies, gross skeletal muscle atrophy is frequently observed upon clinical examination and, at the level of the myofibre, a common secondary or primary histological abnormality is type I (slow) myofibre atrophy (Romero and Clarke, 2013; Wang and Pessin, 2013). An emerging theme, therefore, is the role of skeletal muscle protein synthesis and degradation molecular pathways in the congenital myopathies; however, it must be cautioned again that there is little direct evidence to support a causative link between the congenital myopathies and skeletal muscle atrophy.
Protein degradation pathways
There are two main proteolytic systems that control protein turnover in skeletal muscle: the ubiquitin-proteasome pathway (atrophy) and the autophagy-lysosome machinery. Atrophy and autophagy, while distinct programmes, are both coordinately regulated via the transcription factors FoxO3 and mammalian target of rapamycin (mTOR).
The ubiquitin-proteasome pathway is responsible for degradation of short-lived soluble proteins and myofibrillar proteins. Major players in the ubiquitin-proteasome pathway are the E3 ubiquitin ligases (also known as atrogenes), these include: F-box protein atrogin-1 (also known as MAFbx, now known as FBXO32) and striated muscle specific tripartite motif (TRIM) proteins, also known as MURF (muscle ring finger) proteins (reviewed in Perera et al., 2012). Of particular interest to the congenital myopathies is the finding that the sarcomeric contractile apparatus, dysfunction of which is so central to many congenital myopathies, is directly involved in regulating atrophy signalling. For example, the mechanosensitive M-line titin kinase domain is known to recruit MURF2 to the sarcomere (reviewed in Voelkel and Linke, 2011). Thus it is conceivable, although not yet proven, that changes in the tension generation on the sarcomere will influence these atrophy pathways.
Atrophy in centronuclear myopathy
The role of atrophy in congenital myopathies has been best characterized in the centronuclear myopathies. In the Mtm1 p.Arg69Cys mouse, skeletal muscle atrophy preceded the onset of muscle weakness (Pierson et al., 2012), leading the authors to suggest that the atrophy in this model was not a consequence of muscle inactivity but was associated intrinsically with the genetic defect. Joubert and colleagues (2013) found that intramuscular delivery of an adeno-associated virus designed to deplete MTM1 resulted in a myotubular myopathy phenotype that included skeletal muscle atrophy. The pathway by which MTM1 deficiency leads to atrophy has begun to be elucidated. Elegant studies by Razidlo et al. (2011) showed that in MTM1-depleted HeLa cells and myotubes there were decreased levels of phosphorylated Akt (activated, Akt-P), and identified that this was mediated by an increase in PI3P levels. Decreased levels of Akt-P resulted in greater activity of FoxO1 and decreased activity of proteins (S6K and 4E-BP1) regulated by mammalian target of rapamycin complex 1 (mTORC1) of the growth axis (Razidlo et al., 2011). Thus aberrant mTORC1 and FOXO1 signalling may underlie some of the pathology of MTM1 deficiency. Indeed mTORC1-deficient mice show skeletal muscle atrophy with numerous centrally nucleated myofibres (Bentzinger et al., 2008), mimicking some of the hallmark features of MTM1 deficiency.
In a Mtmr14 KO mouse model, it was found that age-related muscle wasting, sarcopenia, occurred precociously (Romero-Suarez et al., 2010). It was also noted that Mtmr14 expression and MTMR14 protein levels were decreased in old wild-type muscle, suggesting a role for MTMR14 in sarcopenia (Romero-Suarez et al., 2010). The force deficits and altered Ca2+ handling were similar for mature Mtmr14−/− (12–14 months of age) and old wild-type muscle (22–24 months of age), suggesting that muscle atrophy in both conditions (Mtmr14 deficiency and sarcopenia) may be due to altered Ca2+ signalling and feedback from the sarcomere.
KI mice, heterozygous for the p.Arg465Trp DNM2 mutation, develop skeletal muscle weakness prior to the onset of structural abnormalities and muscle atrophy (Durieux et al., 2010); homozygous mice died within 24 h of birth. Expression of atrophy- and autophagy-related genes: FOXO3, MURF1 and Gabarap11 were elevated in 2-month-old heterozygous KI muscle relative to wild-type. This study also showed upregulation of FOXO3 and Gabarap11 in skeletal muscle from a single patient with a DNM2 mutation (Durieux et al., 2010). Skeletal muscle atrophy was also observed in mouse muscle in which wild-type or p.Arg465Trp mutant DNM2 was overexpressed using adeno-associated virus vectors (Cowling et al., 2011).
An interesting example of how a sarcomeric protein can directly influence muscle protein turnover is supplied by titin and its kinase domain, which regulates a signalling complex by interacting with Nbr, which regulates MURF2 via p62/SQSTM1, thus providing a link to translate sensing of mechanical load to regulation of muscle protein turnover (Lange et al., 2005). A mutation in one recessive titinopathy patient with arthrogryposis, cardiomyopathy and also muscle atrophy has been shown to completely abolish this kinase activity (Chauveau et al., 2014). It will be interesting to see how this pathway will be affected in other patients and animal models.
Atrophy in nemaline myopathy
Mutation of genes encoding BTB-Kelch proteins, known to bind cullin 3 to form E3 ubiquitin ligases (Sambuughin et al., 2012; Canning et al., 2013), has been shown to cause nemaline myopathies [KBTBD13 (Sambuughin et al., 2010), KLHL40 (Ravenscroft et al., 2013b), KLHL41 (Gupta et al., 2013)] and an early-onset distal myopathy (KLHL9) (Cirak et al., 2010); suggesting a direct link between impairment of the ubiquitin-proteasome system and skeletal muscle myopathies. Recently, Garg et al. (2014) showed that Klhl40 KO mice presented with a lethal nemaline myopathy-like disease. They also identified nebulin and leiomodin 3 (LMOD3) as KLHL40-binding proteins and showed that loss of KLHL40 resulted in reduced abundance of NEB and LMOD3 protein in Klhl40 KO muscle. NEB and LMOD3 were also decreased in muscle biopsies from some KLHL40 patients. Intriguingly, co-expression of KLHL40 with either NEB or LMOD3, resulted in increased abundance of these proteins, in the absence of a change in transcript levels. The increase in LMOD3 abundance (in the presence of KLHL40) was similar to that observed when a proteasome inhibitor was added and it was shown that KLHL40 decreased polyubiquitination of LMOD3 (Garg et al., 2014). Thus despite the widely-held view that BTB-Kelch proteins mediate protein degradation, it appears that KLHL40 actually stabilizes the thin-filament proteins NEB and LMOD3, by inhibiting proteasome-mediated degradation. The absence of NEB in some KLHL40 patient muscle may explain the severity of KLHL40 disease, as patients with two null alleles of NEB show a similarly severe phenotype (Wallgren-Pettersson et al., 2002).
These observations of the multifaceted involvement of atrophy have lead to preclinical investigations of treatment counteracting atrophy in a number of congenital myopathy models. In a mouse model of nemaline myopathy (KI Acta1H40Y), upregulation of two hypertrophic factors [IGF1 and four and a half LIM domains protein 1 (FHL1; Cowling et al., 2008)] or delivery of ι-tyrosine, improved skeletal muscle strength and pathology (Nguyen et al., 2011). Downregulation of components of the ubiquitin-proteasome pathway (based on microarray analysis of transcript levels) were observed in skeletal muscle from nemaline myopathy patients suggesting altered protein turnover (Sanoudou et al., 2003), the functional significance of these alterations remains unclear. There is anecdotal and some experimental evidence from the TPM3 Met9Arg mouse that exercise may be beneficial in nemaline myopathy (Joya et al., 2004; Wallgren-Pettersson et al., 2011); presumably by minimizing susceptibility to disuse atrophy (Wallgren-Pettersson, 1989; Wallgren-Pettersson et al., 1990; North et al., 1997).
The potential benefits of controlled exercise in individuals with congenital myopathies clearly warrant further investigation. Similarly it remains to be investigated more systematically whether targeted disruption of the ubiquitin-proteosomal pathway might represent a potential therapeutic target for the congenital myopathies.
Protein synthesis pathways
One of the most potent known activators of skeletal hypertrophy is IGF1 (insulin-like growth factor 1) (Musaro et al., 1999). This hypertrophy requires myofibrillogenesis. Until more recently the mechanism whereby IGF1 mediated myofibrillogenesis remained completely obscure. Takano et al. (2010) showed that IGF1 produces its hypertrophic effects via activation of PI3K-Akt signalling that results in complex formation of N-WASP and nebulin. Nebulin-N-WASP complexes were found to mediate actin filament nucleation from the Z-disc (Takano et al., 2010). Thus IGF1-induced actin filament formation (via nebulin-N-WASP complexes) is required for muscle maturation and hypertrophy.
The most potent negative regulator of skeletal muscle hypertrophy is myostatin, a member of the TGF beta-like growth factor family, signalling through activin receptors (Amthor and Hoogaars, 2012). Naturally-occurring and genetically-engineered animal models that are myostatin nulls (Mstn−/−) present with a ‘double-muscled’ phenotype. Counterintuitively, despite the plethora of naturally occuring animal models (most notably Belgium Blue and Piedmontese cattle, sheep, whippet and mouse) (Stinckens et al., 2011) there is only a single-reported isolated case of a MSTN partial loss-of-function patient (Schuelke et al., 2004). Myostatin is a negative regulator of Akt and thereby inhibits protein synthesis. In addition to this direct role on protein synthesis, myostatin also signals through SMAD2 and SMAD3 to activate FOXO3, an inducer of both atrophy (through upregulation of atrogenes and MURF proteins) and autophagy.
Approaches to therapy targeting hypertrophy/atrophy pathways
Given the prominence of myostatin in regulating myofibre atrophy, modulation of myostatin signalling has been vigorously pursued as a potential therapeutic avenue for skeletal muscle diseases (particularly the muscular dystrophies; reviewed in Amthor and Hoogaars, 2012); however, some studies suggest that myostatin inhibition may be detrimental to skeletal muscle function (Amthor et al., 2007; Giannesini et al., 2013), perhaps in particular in a muscular dystrophy in which the muscle is inherently fragile. This may suggest that non-dystrophic myopathies with prominent muscle atrophy may be a better target for myostatin modulation. Inhibition of the activin receptor IIb in Mtm1−/− mice improved muscle strength and lifespan (Lawlor et al., 2011b). Thus modulation of the myostatin pathway might be of benefit in some congenital myopathies. It seems that this approach should be considered on a disease-by-disease basis and should take into account all of the different mechanisms resulting in skeletal muscle atrophy.
Satellite cells defects
Although satellite cells play a major role during myogenesis and muscle regeneration, their role in postnatal skeletal muscle maintenance and hypertrophy is controversial. In a limited number of recent studies a potential role for satellite cells has been implicated in the pathobiology of some congenital myopathies, but again, the disease specificity of these phenomena will need to be further investigated.
Similar to MTM1 patients, the early stages of myogenesis in the Mtm1 KO mouse model seems normal - this has led to the suggestion that there is defective skeletal muscle maintenance or regeneration associated with MTM1 deficiency (Buj-Bello et al., 2002). Reduced satellite cell numbers were observed in muscle from Mtm1 KO mice, determined by measuring PAX7 abundance and by flow cytometry to quantify the relative number of myogenic cells (CD31−/CD45−/CD106+) isolated from the muscle. This study also showed that the number of resident satellite cells declined with age in the Mtm1 KO muscle. The Mtm1 KO progenitor cells showed reduced proliferative capacity in vitro and in in vivo transplantation experiments and were more susceptible to apoptosis (Lawlor et al., 2012). The authors hypothesized that the altered Ca2+ homeostasis due to MTM1 deficiency may cause mitochondrial stress and thereby activate the mitochondrial-apoptotic pathway. In mouse skeletal muscle depleted of MTM1 postnatally, a ∼90% reduction in the number of resident satellite cells was observed (Joubert et al., 2013), these studies suggest that exhaustion of the satellite cell population may contribute to the disease progression associated with MTM1 deficiency.
Exhaustion of the satellite cell population was also a feature of the Sepn1 null mouse (Castets et al., 2012); reduced satellite cell numbers have also been observed in SEPN1 patient muscle and likely contributes to the muscle atrophy observed (Castets et al., 2012) suggesting that SEPN1 plays an essential role in maintenance of these skeletal muscle progenitor cells.
Recessive mutations in MEGF10 (MIM 612453) were associated with early onset myopathy, areflexia, respiratory distress and dysphagia (EMARDD; MIM 614399) (Logan et al., 2011; Boyden et al., 2012). MEGF10 is expressed in quiescent and activated satellite cells and knock-down of Megf10 in mouse muscle resulted in satellite cell depletion (Holterman et al., 2007). EMARDD patient muscle showed reduced mean myofibre diameter and lacked PAX7+ nuclei (Logan et al., 2011).
Elevated satellite cell numbers were observed in skeletal muscle from patients with nemaline myopathy (Sanoudou et al., 2003) and TPM3 p.Met9Arg transgenic mice (Sanoudou et al., 2006), suggesting that satellite cell-activated muscle regeneration might be a feature in some subtypes of nemaline myopathy. The role of satellite cells in the pathophysiology of nemaline myopathy has not been investigated further.
Myotubes and skeletal myofibres are extremely resistant to apoptosis, and this is thought to be due to expression of survival proteins: Akt, PI3K and p21 while satellite cells express anti-apoptotic Bcl-2 as their primary survival factor (reviewed in Schwartz, 2008). Recently it was shown that myotubularin regulates Akt-mediated cell survival signalling via elevated levels of PI3P (Razidlo et al., 2011). siRNA silencing of MTM1, in HeLa cells, resulted in a robust activation of caspase-dependent pro-apoptotic signalling that resulted in DNA fragmentation and cell death. Knock-down of myotubularin in cultured human myotubes also resulted in inhibition of Akt phosphorylation, caspase-dependent pro-apoptotic signalling and DNA fragmentation (Razidlo et al., 2011). Thus, there may also be a role for aberrantly active apoptosis in the pathogenesis of congenital myopathies.
Autophagy effected by membrane targeting and turnover
The role of autophagy in skeletal muscle homeostasis (reviewed in Sandri, 2010) is less well understood than atrophy. Autophagy lysosomes regulate the turnover of long-lived proteins and organelles (Sandri, 2010) and involves BNIP3, Gabarap11, LC3 (now knwon as MAP1LC3A) and p62, as well as a host of other players. Autophagy and atrophy are inextricably linked; counterintuitively, autophagy blockade in Atg7−/− mouse muscle resulted in profound skeletal muscle atrophy and force deficits (Masiero et al., 2009; Masiero and Sandri, 2010).
Alterations to autophagy and/or lysosomal function are observed in many skeletal muscle diseases, including: collagen VI-related muscular dystrophies (in which autophagy seems to underperform), Danon disease, Duchenne muscular dystrophy, Pompe disease and X-linked myopathy with excessive autophagy (Bonaldo et al., 1998; Grumati et al., 2010; De Palma et al., 2012; Ramachandran et al., 2013) and have also recently been implicated in myofibrillar myopathy (Kley et al., 2012, 2013). Thus both impaired autophagy and excessive autophagy can result in skeletal muscle disease, highlighting the need for the right balance between protein synthesis and degradation processes.
Recently, Castets et al. (2013) have shown that sustained activation of mTORC1 results in impaired autophagy (via inhibition of Ulk1, despite FOXO3 activation) and produced a late onset myopathy in the mouse. Conversely, inhibition of mTORC1, via depletion of Raptor, induced autophagy regardless of FOXO inhibition. These novel data suggest that mTORC1 is, in fact, the major regulator of skeletal muscle autophagy (Castets et al., 2013).
Within the congenital myopathies, autophagy has been most studied in the CNMs. Given that autophagy is essentially a membrane-driven process and that the primary defect in the CNMs involves membrane remodelling, defective autophagy in the CNMs may be a direct consequence of the underlying genetic defect.
A clear link between MTM1 deficiency and autophagy is tenuous. PI3P is required for the initiation of autophagy and PI3,5P2 is required for its progression. Muscle-specific deletion of PIK3C3 (the primary kinase responsible for the production of PI3P) causes a lethal dystrophic phenotype in mice and a severe defect of the autophagolysosomal pathway (Reifler et al., 2014). Thus turnover of PI3P by myotubularins may be required for autophagy progression and it is possible that loss of MTM1 could directly impair autophagy. However, a study examining the role of myotubularins in autophagy showed that MTM1 was only a weak activator of the process (Vergne et al., 2009).
Levels of some autophagy-associated transcripts (Gabarap11 and LC3b) are increased in Mtm1 KO mouse muscle (Fetalvero et al., 2013; Joubert et al., 2013). In Cre-induced Mtm1 depletion in adult skeletal muscle, levels of LC3-I, LC3-II and p62 proteins were elevated (Joubert et al., 2013), furthermore elevated levels of phosphorylated S6, 4EBP1, Akt and phosphorylated Akt were observed. Similarly, LC3-I, LC3-II, p62 and LAMP2 proteins were more abundant in Mtm1−/− muscle than wild-type; levels of phosphorylated S6 and Akt were also elevated in Mtm1−/− muscle (Fetalvero et al., 2013). Elevated levels of LC3, p62, polyubiquitinated proteins and dysfunctional mitochondrial are also a feature of Atg7−/− null mice (Masiero et al., 2009; Masiero and Sandri, 2010). This suggests that mTORC1-mediated blockade of normal autophagic degradation may contribute to the impaired autophagy observed in Mtm1 depleted muscle and partially contributes to the muscle pathology associated with MTM1 deficiency (Fetalvero et al., 2013; Joubert et al., 2013). However, the autophagic defect in Mtm1 deficient mouse muscle may be secondary to the triad dysfunction, the sarcoplasmic reticulum is a major membrane donor to autophagy, or the general poor health of these animals.
The link between autophagy and MTMR14 is much clearer, where failed regulation likely of PI3,5P2 results in impaired regulation of autophagic flux. Vergne and colleagues (2009) showed that siRNA knock-down of MTMR14 and loss of its enzymatic activity, in C2C12 myoblasts, resulted in the formation of autophagosomes and also facilitated their maturation into autolysosomes. LC3-II levels were unchanged following MTMR14 siRNA treatment in the absence of Atg5 (autophagy-related 5), showing that MTMR14 negatively regulates autophagy in an Atg5-dependent manner. In support of this, overexpressing MTMR14 resulted in an increase of p62 aggregates in transfected C2C12 myoblasts suggesting impaired autophagy (Vergne et al., 2009). MTMR14 was found to co-localize with ATG12, ATG16 and LC3-II in early autophagic organelles and autophagasomes in C2C12 myoblasts. Furthermore knock-down of MTMR14 resulted in increased recruitment of the autophagic PI3P-binding protein WIPI1 (the mammalian homologue of yeast ATG18) (Xie and Klionsky, 2007) to autophagosomes. Finally, Vergne et al. (2009) showed that overexpression of a CNM MTMR14 variant (p.Arg336Gln) within the phosphatase domain did not promote accumulation of p62 puncta, whereas wild-type and p.Tyr462Cys (a mutant outside of the catalytic domain) MTMR14, did increase p62 aggregation (Vergne et al., 2009). These data suggest that the CNM p.Arg336Gln substitution renders MTMR14 unable to regulate autophagy.
Dowling et al. (2010) found elevated induction of autophagy in mtm1/mtmr14 double morphant zebrafish, as determined by increased LC-II levels and vacuolated membranes, with severe effects on gross embryogenesis and extensive autophagic defects. Although accumulation of abnormal membranous structures had previously been reported for mtm1 morphant fish, autophagy was not altered in these fish (Dowling et al., 2009). Thus, these authors hypothesize that the phenotype of the double morphants results from dysregulated organelle and membrane breakdown and overloading of the autophagic system primarily due to loss of mtmr14 (Dowling et al., 2010).
Skeletal muscle and cultured embryonic fibroblasts from knock-in Dnm2 p.Arg465Trp (KI-Dnm2R465W) mice show elevated levels of LC3, Gabarap11 and p62 transcripts and greater protein abundance (Durieux et al., 2012). The number of p62-positive accumulations and early phagosomes were also elevated in KI- Dnm2R465W embryonic fibroblasts. These data suggest that impaired autophagy contributes to the phenotype of KI-Dnm2R465W mice, as opposed to upregulated autophagy observed with MTMR14 mutation.
Other possible therapies for congenital myopathies
As we have pointed out throughout the discussion so far, understanding the pathomechanisms active in the congenital myopathies is important in defining targets against which to develop therapies, such as compounds that can ameliorate the downstream consequences of the underlying genetic defect/s that could modulate diseases. However, targeting the primary cause of disease, i.e. the genetic mutation and their immediate consequences, while challenging, obviously represents a major goal for arriving at effective treatments for these conditions.
Enzyme replacement therapy
Enzyme replacement therapy has been a therapeutic strategy that has been pursued in a number of diseases resulting from enzyme deficiencies (reviewed in Valayannopoulos, 2013). Notably, enzyme replacement therapy is now in clinical use for a skeletal muscle disease, glycogen storage disease II/Pompe disease resulting from deficiency of GAA (Thurberg et al., 2006; Gungor et al., 2013). Lawlor et al. (2013) have shown that intramuscular delivery of myotubularin-fused with a skeletal muscle penetrating-antibody protein (3E10Fv-MTM1) resulted in improved muscle contractility and pathology and improved gait and ambulation of Mtm1 KO mice. Recently, AAV8-mediated MTM1 gene therapy was explored in Mtm1 KO mice and dogs carrying a naturally occurring Mtm1 mutation (Childers et al., 2014). It was shown that a single intravascular injection was sufficient to increased muscle strength, reduce muscle pathology and extend survival (Childers et al., 2014). These studies demonstrate the efficacy of enzyme replacement therapy and pave the way for clinical trials for myotubular myopathy. Similarly, enzyme replacement therapy may also be expected to be of therapeutic benefit in models of MTMR14-related congenital myopathies although this has not been attempted.
Evidence from patients (Nowak et al., 2007; Ravenscroft et al., 2008) and experimental data, concur that mutant actin load determines disease severity in dominant ACTA1 disease (Drummond et al., 1991; Domazetovska et al., 2007; Ravenscroft et al., 2008; 2011, 2013a; Haigh et al., 2010; Sevdali et al., 2013) and thus modulation of the amount of mutant ACTA1 represents a target for dominant ACTA1 patients. Furthermore, upregulation of the foetal isoform of skeletal muscle α-actin (cardiac α-actin; ACTC) in patients with recessive loss of ACTA1 moderates the disease allowing survival of affected children. Upregulation of ACTC is thus a valid target for upregulation or even gene transfer (Nowak et al., 2009; Ravenscroft et al., 2013a) as it would be less likely to cause an immune response to the introduced protein as compared with re-introducing ACTA1.
Here we have attempted to highlight emerging pathophysiological concepts spanning across different genetic and structural entities that traditionally comprise the congenital myopathies, with the goal of pointing out avenues for therapeutic developments that are based on disease mechanisms. It remains to be more precisely determined which of the pathophysiological concepts explored herein are identified in which of the congenital myopathies, and to what degree. For instance, in the most well characterized models, those for CNMs, there appears to be a role for all five of the key themes in the pathobiology of the disease.
It is perhaps necessary, given the wave of animal models generated and studied for the congenital myopathies and the studies of patient biopsies and cell lines, that standard operating procedures be developed and implemented for the systematic characterization of these models and tissues, to allow for greater intermodel interpretation and for comparison of the efficacies of various therapeutic strategies. This has been recognized and implemented for models of muscular dystrophies (Grounds et al., 2008; Nagaraju and Willmann, 2009) and spinal muscular atrophy (Willmann et al., 2011) and is recognized by the presence of standard operating procedures on the TREAT-NMD webpage (http://www.treat-nmd.eu/research/preclinical/). It would be desirable to expand this approach to models of the congenital myopathies.
Ultimately it seems likely that treatment of the congenital myopathies will encompass multimodal regimes: e.g. gene transfer therapy, enzyme replacement therapy, upregulation of alternative genes, anti-atrophy approaches, in addition to perhaps AChE inhibitors, and/or Ca2+ sensitizers and/or anti-oxidants. The considerable overlap of pathobiological mechanisms between the various genetic and structural subgroups of the congenital myopathies might indicate that some of the potential therapies may be effective in a broad range of the congenital myopathies and beyond.
We thank our anonymous reviewers for most helpful commentary and advice.
G.R. is supported by a National Health and Medical Research Council of Australia (NHRMC) Early Career Researcher Fellowship (APP1035955) and N.G.L. by a NHMRC Research Fellowship (APP1002147). C.G.B. is supported by intramural funds of NINDS/NIH.