Skeletal muscle cells possess a ‘memory’ of acute early life TNF-α exposure: role of epigenetic adaptation

Adam P. Sharples, Ioanna Polydorou, David C. Hughes, Daniel J. Owens, Thomas M. Hughes, Claire E. Stewart

Sufficient quantity and quality of skeletal muscle is required to maintain lifespan and healthspan into older age. The concept of skeletal muscle programming/memory has been suggested to contribute to accelerated muscle decline in the elderly in association with early life stress such as fetal malnutrition. Further, muscle cells in vitro appear to remember the in vivo environments from which they are derived (e.g. cancer, obesity, type II diabetes, physical inactivity and nutrient restriction). Tumour-necrosis factor alpha (TNF-α) is a pleiotropic cytokine that is chronically elevated in sarcopenia and cancer cachexia. Higher TNF-α levels are strongly correlated with muscle loss, reduced strength and therefore morbidity and earlier mortality. We have extensively shown that TNF-α impairs regenerative capacity in mouse and human muscle derived stem cells [Meadows et al. (J Cell Physiol 183(3):330-337, 2000); Foulstone et al. (J Cell Physiol 189(2):207-215, 2001); Foulstone et al. (Exp Cell Res 294(1):223-235, 2004); Stewart et al. (J Cell Physiol 198(2):237-247, 2004); Al-Shanti et al. (Growth factors (Chur, Switzerland) 26(2):61-73, 2008); Saini et al. (Growth factors (Chur, Switzerland) 26(5):239-253, 2008); Sharples et al. (J Cell Physiol 225(1):240-250, 2010)]. We have also recently established an epigenetically mediated mechanism (SIRT1-histone deacetylase) regulating survival of myoblasts in the presence of TNF-α [Saini et al. (Exp Physiol 97(3):400-418, 2012)]. We therefore wished to extend this work in relation to muscle memory of catabolic stimuli and the potential underlying epigenetic modulation of muscle loss. To enable this aim; C2C12 myoblasts were cultured in the absence or presence of early TNF-α (early proliferative lifespan) followed by 30 population doublings in the absence of TNF-α, prior to the induction of differentiation in low serum media (LSM) in the absence or presence of late TNF-α (late proliferative lifespan). The cells that received an early plus late lifespan dose of TNF-α exhibited reduced morphological (myotube number) and biochemical (creatine kinase activity) differentiation vs. control cells that underwent the same number of proliferative divisions but only a later life encounter with TNF-α. This suggested that muscle cells had a morphological memory of the acute early lifespan TNF-α encounter. Importantly, methylation of myoD CpG islands were increased in the early TNF-α cells, 30 population doublings later, suggesting that even after an acute encounter with TNF-α, the cells have the capability of retaining elevated methylation for at least 30 cellular divisions. Despite these fascinating findings, there were no further increases in myoD methylation or changes in its gene expression when these cells were exposed to a later TNF-α dose suggesting that this was not directly responsible for the decline in differentiation observed. In conclusion, data suggest that elevated myoD methylation is retained throughout muscle cells proliferative lifespan as result of early life TNF-α treatment and has implications for the epigenetic control of muscle loss.