Molecular basis of paradoxical muscle hypertrophy in dystrophin-deficient muscle

Duchenne muscular dystrophy and the genetically analogous canine condition, golden retriever muscular dystrophy (GRMD), show variable involvement of specific muscle groups within an affected individual. The sartorius muscle in Duchenne boys and the orthologous cranial sartorius (CS) in GRMD dogs shows relative sparing and muscle hypertrophy. The molecular mechanisms for this hypertrophy have not been defined to date. The goal of this dissertation was to define the molecular networks associated with CS muscle sparing and hypertrophy using muscle samples and phenotype measures from GRMD dogs at different ages. We noted that GRMD dogs with the largest CS muscles also appeared more severely affected from a phenotype standpoint. This was confirmed by hierarchical clustering of CS muscle size, tibiotarsal joint angle, and flexor and extensor force in 49 variably affected dogs, suggesting that imbalance across the hip joint may drive disability in this cohort of dogs. To define the molecular basis of differential muscle involvement, specifically CS hypertrophy, serial biopsies from the CS and atrophied vastus lateralis and long digital extensor muscles of variably affected GRMD and normal dogs were assessed at 4-9 weeks (mild clinical features) and 6 months (advanced disease). We initially studied a candidate protein, osteopontin, due to its association with both dystrophy and hypertrophy in previous studies. Osteopontin showed an age-dependent increase in GRMD muscles and was inversely correlated with the well-studied negative growth regulator, myostatin, in the CS muscle at 4-9 weeks. To test a cause-effect relationship between osteopontin and myostatin, myoblast and myotube murine cell cultures were treated with recombinant mouse osteopontin protein, which caused a dose-dependent activation of AKT1, phosphorylation of FoxO1, and a decrease in myostatin mRNA and protein expression. Osteopontin-treated myotubes also showed decreased myostatin protein. Previous studies have shown that the arginine-glycine-aspartic acid- (RGD) residues in OPN can bind several integrin receptors and a heparin-binding domain can bind the CD44 receptor. Normal, full-length and RGD mutant OPN proteins decreased myostatin expression, but the latter showed a reduced effect. This suggests that both RGD- and non-RGD-dependent receptors may be involved in the inhibition of myostatin. To carry out a discovery analysis of the molecular basis of hypertrophic remodeling, microRNA profiles were performed on GRMD CS muscles. Increased levels of miR-539 and 208b were found in the hypertrophied CS, both of which are predicted to down-regulate myostatin. Consistent with the microRNA findings, myostatin mRNA showed an age-dependent decrease in the CS compared to the VL and LDE, and also showed a significant inverse correlation with CS hypertrophy in variably affected GRMD dogs (r = -0.73; p < 0.05). To determine if additional compensatory or hypertrophy pathways were active in CS vs. VL and LDE, we carried out both mRNA expression profiling (72 samples) and proteomic profiling (12 samples) in GRMD and normal dogs. The top-ranked molecular network specific to the hypertrophied CS at 6 months of age included alpha-dystroglycan (DAG1), and like-acetylglucosylaminyltransferase (LARGE). Immuno-staining and blotting studies confirmed that the CS showed upregulation of DAG1 relative to more severely affected muscles, suggesting that compensatory increased expression of the dystrophin-associated DAG1/LARGE pathway could mediate some of the sparing of the CS. Proteomic studies identified upregulation of alpha- and beta- spectrin and myotrophin as additional proteins that could enhance cytoskeletal stability and promote hypertrophy, respectively. The studies in this dissertation provide new knowledge of compensatory and hypertrophy pathways that are activated in dystrophin-deficient, hypertrophied muscle. We conclude that efforts to pharmacologically manipulate single pathways, such as myostatin, may influence muscle groups differently, with muscle-specific interactions among several compensatory and hypertrophic networks.