Myotubular myopathy is a severe, X-linked congenital myopathy with onset in infancy 
. It is characterized by profound neonatal hypotonia and skeletal muscle weakness. It is associated with substantial mortality, with approximately half of all affected boys dying in the first year of life 
. Surviving children have significant morbidity associated with respiratory compromise and difficulties with ambulation. Currently there are no treatments or disease modifying therapies available for this condition.
The condition is defined by characteristic changes observed on muscle biopsy 
. Biopsies show muscle fiber hypotrophy and an abundance of fibers with large, centralized nuclei of unusual appearance. These nuclei are distinct in appearance from those observed in degenerative conditions like Duchenne muscular dystrophy, and are the defining pathologic features of a group of congenital myopathies called centronuclear myopathies 
Myotubular myopathy is caused by mutations in the myotubularin gene 
. Over 200 mutations have been reported in the myotubularin gene, the majority of which result in loss of functional gene expression 
. Myotubularin is the only gene associated with myotubular myopathy. It is the canonical member of a large family of homologous proteins called the myotubularin related proteins (MTMRs) 
. Of interest is the fact that several MTMRs are mutated in human neurologic diseases, including mutation of MTMR14 in an autosomal form of centronuclear myopathy 
Myotubularin was originally characterized as a protein tyrosine phosphatase, but was subsequently found instead to function primarily as a lipid phosphatase 
. It acts specifically to remove phosphates from the 3-position of phosphoinositide rings. As demonstrated in cell free biochemical assays 
and with forced exogenous expression 
, myotubularin converts phosphoinositide-3-phosphate (PI3P) to phosphoinositide phosphate (PIP) and phosphoinositide-3,5-bisphosphate (PI3,5P2) to phosphoinositide-5-phosphate (PI5P). Most recently, Cao and colleagues have demonstrated using RNAi in A431 cells that knockdown of myotubularin results in a 60–120% increase in PI3P levels, thus substantiating the requirement for myotubularin in the regulation of endogenous PI3P 
. Increased PI3P levels have also been observed in yeast lacking the myotubularin homolog ymr1 
. As yet, however, this activity has not been directly examined in whole vertebrates or in specific organ systems, including muscle. The functional importance of myotubularin's phosphatase activity is assumed from the fact that missense mutations that alter critical amino acids in the phosphatase domain without affecting protein stability result in myotubular myopathy 
Phosphoinositides are implicated in myriad cellular functions, chief among them the regulation of membrane traffic and vesicle/organelle movement 
. Because it acts to modify certain PI residues, myotubularin is assumed to function as a regulator of membrane traffic and in particular the movements of vesicles between endosomal compartments 
. Overexpression of myotubularin in cell culture delays traffic out of the endosomal compartment and causes vacuole accumulation. However, as with myotubularin phosphatase activity, a role for myotubularin in the regulation of membrane traffic in vivo
and specifically in skeletal muscle has yet to be determined. In addition, unlike with other myopathies due to altered membrane traffic (examples include Danon Disease due to LAMP2 mutation 
), myotubular myopathy is not characterized by the pathologic accumulation of vesicles.
Many critical questions remain unanswered concerning myotubularin function and myotubular myopathy pathogenesis. These include whether myotubularin truly functions as a lipid phosphatase and regulator of membrane traffic in vivo. Furthermore, the relationship between the proposed functions of myotubularin and disease pathogenesis is not clear. The same is true with the association between the unique histologic appearance of the muscle in myotubular myopathy patient biopsies and the etiology of muscle weakness and hypotonia. The lack of knowledge concerning these fundamental issues is a significant barrier in the development of therapeutic strategies for the disease.
A murine model of myotubular myopathy exists, generated by targeted mutagenesis 
. It recapitulates the clinical and histopathologic features of the disease, thus confirming the association between myotubularin and myotubular myopathy. However, due in part to technical limitations with the murine system, it does not address many of the fundamental questions mentioned above. To begin answering these questions, and to develop a model system amenable to rapid testing of therapeutic strategies, we report here the development of a zebrafish model of myotubular myopathy. Using antisense morpholino technology, we generated zebrafish embryos with reduced myotubularin protein expression. These embryos have severely impaired motor function, muscle fiber atrophy and the presence of large, abnormally located nuclei. These findings are reminiscent of those seen in myotubular myopathy. We also demonstrate that loss of myotubularin causes increased PI3P levels in muscle, thus confirming for the first time that myotubularin functions as a lipid phosphatase in a vertebrate model system. Using RNA-mediated rescue experiments, we show that the homologous myotubularin-related genes MTMR1 and MTMR2 are able to functionally compensate for the loss of myotubularin. Lastly, and most significantly, we identify alterations in the T-tubule and sarcoplasmic reticular networks in morphant zebrafish muscle. We confirm that similar disorganization of the tubulo-reticular network is present in biopsy samples from patients with myotubular myopathy. In all, we have successfully created a zebrafish model of myotubular myopathy, and have used this model to both answer fundamental questions concerning myotubularin function and to uncover a novel mechanism to explain the pathogenesis of the disorder.