Differential DNA methylation and transcriptional signatures characterize impairment of muscle stem cells in pediatric human muscle contractures after brain injury.

Limb contractures are a debilitating and progressive consequence of a wide range of upper motor neuron injuries that affect skeletal muscle function. One type of perinatal brain injury causes cerebral palsy (CP), which affects a child's ability to move and is often painful. While several rehabilitation therapies are used to treat contractures, their long-term effectiveness is marginal since such therapies do not change muscle biological properties. Therefore, new therapies based on a biological understanding of contracture development are needed. Here, we show that myoblast progenitors from contractured muscle in children with CP are hyperproliferative. This phenotype is associated with DNA hypermethylation and specific gene expression patterns that favor cell proliferation over quiescence. Treatment of CP myoblasts with 5-azacytidine, a DNA hypomethylating agent, reduced this epigenetic imprint to TD levels, promoting exit from mitosis and molecular mechanisms of cellular quiescence. Together with previous studies demonstrating reduction in myoblast differentiation, this suggests a mechanism of contracture formation that is due to epigenetic modifications that alter the myogenic program of muscle-generating stem cells. We suggest that normalization of DNA methylation levels could rescue myogenesis and promote regulated muscle growth in muscle contracture and thus may represent a new nonsurgical approach to treating this devastating neuromuscular condition.

Intramuscular Anatomy Drives Collagen Content Variation Within and Between Muscles.

The passive load bearing properties of muscle are poorly understood partly due to challenges in identifying the connective tissue structures that bear loads. Prior attempts to correlate passive mechanical properties with collagen content (often expressed as a mass ratio and used as a surrogate for connective tissue quantity within muscle) have not been successful. This is likely a result of not accounting for variability in intramuscular connective tissue throughout a muscle such that a single collagen content value likely does not adequately represent the connective tissue load bearing capacity of a muscle. Therefore, the purpose of this study was to determine how intramuscular connective tissue distribution throughout a muscle impacts measured collagen content. For this analysis, four mouse hindlimb muscles were chosen because of their varying actions and anatomy; rectus femoris, semimembranosus, tibialis anterior, and lateral gastrocnemius. Collagen content throughout each muscle was determined biochemically using an optimized hydroxyproline assay. Dense connective tissue distribution throughout each muscle's length was quantified histologically. We found that collagen content varied widely within and between muscles, from 3.6 ± 0.40 SEM µg/mg wet weight to 15.6 ± 1.58 SEM µg/mg, which is dependent on both the specific location within a muscle and particular muscle studied. Both collagen content and connective tissue structures demonstrated stereotypically patterns with the highest quantity at the proximal and distal ends of the muscles. Additionally, using three independent approaches: (1) linear regression, (2) predictive modeling, and (3) non-linear optimization, we found complementary and corroborating evidence suggesting a causal relationship between a muscle's connective tissue distribution and collagen content. Specifically, we found that muscle collagen content is driven primarily by its dense connective tissue structures due to the extremely high collagen content of connective tissue (227.52-334.69 µg/mg) compared to muscle tissue (1.93-4.03 µg/mg). A consequence of these findings is that a single collagen content measurement does not accurately represent a muscle's complex distribution of connective tissue. Future studies should account for collagen content variations and connective tissue anatomy to establish more accurate relationships between collagen content measurements and whole muscle passive mechanics.