Integration of Transcriptomic and Metabolomic Profiling to Identify Mechanisms and Biomarkers of Statin-Induced Myopathy.

摘要

Despite extensive research efforts indicating statin-induced myopathy is an on-target effect mediated by HMG-CoA reductase inhibition, the underlying pathophysiology remains unknown. The purpose of this dissertation was to use transcriptomic and metabolomic profiling to identify mechanisms and biomarkers of statin-induced myopathy. Investigative studies were conducted with cerivastatin (0.5 and 1 mg/kg; 14 days) in the rat because this species is predictive of the clinical response. Integration of transcriptomic and metabolomic profiles of skeletal muscles revealed alterations in cholesterol homeostasis and energy metabolism associated with myopathy in fast-twitch muscles, whereas slow-twitch muscles showed evidence of altered energy metabolism without myotoxicity. To further investigate these potential mechanisms of myopathy, a time course study was conducted with cerivastatin (1 mg/kg; 1, 6, 8, 10, and 14 days) in female rats which indicated that alterations in cholesterol homeostasis, and specifically upregulation of the 25-hydroxycholesterol pathway is likely mediated by pathology, removal of excess cholesterol and regeneration. This time course study also indicated an early and sustained transcriptional induction of pyruvate dehydrogenase kinase 4 in fast-twitch muscles, suggesting a metabolic switch in energy metabolism which precedes induction of PGC-1alpha resulting in a more oxidative phenotype. However, evidence of myofibrillar protein degradation and myopathy ensued indicating impaired energy metabolism may be causally related to myopathy. In contrast, this transcriptional profile was not observed in slow-twitch muscles. Similar time course experiments were also conducted in primary fast-twitch myofibers. However, the in vivo findings were not always recapitulated in this model despite dose- and time-dependent cerivastatin-induced toxicity which was prevented with mevalonate supplementation. Accordingly, the value and limitations of the in vitro model for investigating mechanisms of toxicity was demonstrated in this work. To identify biomarkers of myopathy, metabolomic profiling was conducted on urine from cerivastatin treated rats which revealed an increase in 1- and 3-methylhistidine (MH) with myopathy. These metabolites were most abundant in skeletal muscle and originated from the dipeptide anserine (1-MH) and myofibrillar proteins (3-MH: actin and myosin). Experimental paradigms resulting in myofibrillar protein catabolism indicated transcriptional induction of the muscle specific ubiquitin ligases (atrogin-1 and MuRF-1) were generally correlated with increased serum 3-MH. Furthermore, serum 3-MH was more sensitive than urine 3-MH, and traditional biomarkers of myopathy, including creatine kinase (CK) and aspartate aminotransferase (AST), suggesting it may be a useful preclinical biomarker of myopathy in non-fasted rats. In contrast, urine 3-MH provides an advantage over CK and AST as it detects skeletal muscle atrophy and hypertrophy. Collectively, this dissertation represents a body of work demonstrating the utility of transcriptomic and metabolomic profiling in identifying potential mechanisms and biomarkers of statin-induced myopathy.