Hypermethylated Myoblasts Specifically Deficient in MyoD Autoactivation as a Consequence of Instability of MyoD

Abstract

MyoD is one of a family of basic helix loop helix (bHLH) transcription factors acting as master switches of skeletal muscle differentiation. In addition to transcriptionally activating differentiation-specific genes, it autoactivates its own expression through a positive feedback loop. It was cloned following the observation that treatment with the DNA methylation inhibitor 5-azacytidine converts cultured fibroblasts into muscle, presumably through the activation of a transcriptionally silenced locus. In an attempt to experimentally recapitulate this phenomenon, I have stably transfected mouse C2C12 myoblasts with a selectable marker fused to the MyoD promoter/enhancer, treated the cells with methyldeoxycytidine triphosphate to promote genomic hypermethylation, and negatively selected for loss of activity from the MyoD promoter/enhancer. Several clones were recovered that had lost expression of MyoD and the ability to differentiate to myotubes, but could be reverted by treatment with 5-azacytidine. One clone (“C2G2”) was studied in detail. C2G2 cells resume differentiation through the forced expression of exogenous MyoD, but paradoxically fail to autoactivate the endogenous MyoD, suggesting that they are deficient in the MyoD positive feedback circuit but otherwise competent in downstream events in myogenesis. The myogenic bHLH proteins in this clone exhibit nuclear instability. Both the nuclear stability of MyoD and resumption of the autoactivation circuit may be restored either through cell fusion with fibroblasts or by treatment of the cells with protease inhibitors. This suggests that the transcriptional silencing of a factor, not in itself necessary for muscle differentiation but governing the proteolytic stability of MyoD, is responsible for the decay of the autoactivation circuit in these cells. A theoretical analysis offers an explanation for how increased rates of MyoD turnover may result in the kinetic dissociation of autoactivation from the “cross-activation” of downstream genes and suggests a mechanism for the phenomenon of myogenic “memory.”