Aberrant Regulation of MyoD1 Contributes to the Partially Defective Myogenic Phenotype of BC3H1 Cells

Two skeletal muscle-specific regulatory factors, myogenin and MyoD1, share extensive homology within a myc similarity region and have each been shown to activate the morphologic and molecular events associated with myogenesis after transfection into nonmyogenic cells. The BC3H1 muscle cell line expresses myogenin and other muscle-specific genes, but does not express MyoDl during differentiation. BC3H1 cells also do not upregulate or-cardiac actin or fast myosin light chain, nor do they form multinucleate myotubes during differentiation. In this study, we examined the basis for the lack of MyoDl expression in BC3H1 cells and investigated whether their failure to express MyoD1 is responsible for their defects in differentiation. We report that expression of an exogenous MyoD1 cDNA in BC3H1 cells was sufficient to elevate the expression of o~-cardiac actin and fast myosin light chain, and to convert these cells to a phenotype that forms multinucleate myotubes during differentiation. Whereas myogenin and MyoD1 positively regulated their own expression in transfected 1011/2 cells, they could not, either alone or in combination, activate MyoD1 expression in BC3H1 cells. Exposure of BC3HI cells to 5-azacytidine also failed to activate MyoD1 expression or to rescue the cell's ability to fuse. These results suggest that BC3H1 cells may possess a defect that prevents activation of the MyoDl gene by MyoD1 or myogenin. That an exogenous MyoD1 gene could rescue those aspects of the differentiation program that are defective in BC3H1 cells also suggests that the actions of MyoDl and myogenin are not entirely redundant and that MyoDl may be required for activation of the complete repertoire of events associated with myogenesis. D FFERENTIATION of skeletal myoblasts involves irreversible conversion to a postmitotic state, fusion to form multinucleate myotubes, and transcriptional activation of muscle-specific genes. Three muscle-specific regulatory factors, MyoDl (3), myogenin (5, 40, and myf-5 (2), have recently been identified and shown to share a high degree of homology within a domain related to the myc protein family. When transfected into fibroblasts under the transcriptional control of a constitutive promoter, each of these genes can activate the morphologic and molecular events of myogenesis. Initial evidence for a hierarchy of myogenic regulatory genes was obtained from studies that showed that exposure of the multipotential mesodermal stem cell line C3H10T1/2 to the demethylating agent 5-azacytidine resulted in the cells' conversion to myoblasts at high frequency (25-50%) (37). These observations led to the proposal that hypomethylation of a single gene or a few closely linked loci was responsible for conversion to the myogenic lineage (13). MyoD1 expression is activated in 5-azacytidine-derived myoblasts (3); however, it is not known whether the MyoD1 gene is the actual target for demethylation by 5-azacytidine. Nonetheless, the expression of MyoD1 as a result of 5-azacytidine treatment indicates that the MyoD1 gene itself or a gene that regulates MyoD1 is activated in 10TI/2 cells as a consequence of demethylation. In this regard, Emerson and co-workers have identified a gene referred to as myd, that efficiently converts 10"I"1/2 cells to myoblasts when it is transfected as demethylated cosmid DNA (27). MyoDl and myogenin are expressed by myd-transfected 10"I'1/2 cells, which led to the proposal that these genes might be regulated by myd in a dependent myogenic regulatory cascade (27). Recent studies by Thayer et al. (38) also showed that MyoD1 positively autoregulates its own expression and that myogenin and MyoD1 regulate each others' expression. Similarly, MyoD1 is expressed by myf-5-transfected 10T1/2 cells (2). The mouse muscle cell line BC3H1 has been used extensively as a model for studying the mechanisms through which growth factors and oncogenes regulate myogenesis (10, 12, 17, 21, 22, 25, 30, 31, 32, 39). When maintained in the presence of high concentrations of serum, fibroblast growth factor, or type-/~ transforming growth factor, BC3H1 cells exhibit a fibroblast-like morphology and do not express muscle-specific genes. Upon exposure to growth factordeficient medium, these cells exit the cell cycle, and an array of skeletal muscle gene products is induced. However, unlike normal skeletal muscle cells, BC3H1 cells do not fuse or commit to terminal differentiation, nor do these cells express © The Rockefeller University Press, 0021-9525/90/04/929/9 $2.00 The Journal of Cell Biology, Volume 110, April 1990 929-937 929 on Jauary 7, 2018 jcb.rress.org D ow nladed fom MyoDl at detectable levels (3). a-Cardiac actin (33) and fast myosin light chain (MLC-lf) ~ (36) also do not appear to be upregulated to normal levels in these cells. The lack of MyoD1 expression in BC3H1 cells suggests that the myogenic program can be activated, at least in part, through a MyoD1independent mechanism. That finding initially led us to search for regulatory genes related structurally or functionally to MyoD1. Using the myc similarity region of a MyoD1 cDNA as a probe to screen a BC3H1 cDNA library under conditions of reduced stringency, a MyoD-related cDNA was identified (5). This cDNA was found to be homologous to a rat cDNA referred to as myogenin, which was isolated by subtraction-hybridization (41). Subsequent studies showed that myogenin possessed the potential to induce musclespecific genes and fusion in transfected 10T1/2 cells (5). That BC3H1 cells express myogenin at high levels but do not form myotubes suggests that the actions of myogenin are more restricted in these cells than in 10T1/2 cells. To explore the mechanism that prevents myogenin from activating the complete myogenic program in BC3H1 cells, we examined whether the MyoD1 gene can be expressed in BC3H1 cells, and whether these cells possess the genetic potential to fully differentiate. We report here that expression of an exogenous MyoD1 cDNA in BC3H1 cells is sufficient to restore their ability to form multinucleate myotubes and to elevate the expression of a-cardiac actin and MLC-If during differentiation. Whereas myogenin and MyoDl positively regulate each others' expression in transfected 10"I"1/2 cells, this regulatory interaction is inactive in BC~H1 cells, such that neither myogenin nor MyoD1, alone or in combination, can activate the endogenous MyoD1 gene. These results demonstrate that the failure Of BC3HI cells to activate the complete repertoire of events associated with myogenesis is due to their inability to express MyoDl. The fact that neither myogenin nor MyoD1 are able to activate MyoD1 expression in BC3HI cells also suggests that the MyoD1 gene itself has undergone a stable alteration that prevents its activation, or that the regulatory pathway that activates the MyoD1 gene in 10TI/2 fibroblasts is nonfunctional in these cells. Materials and Methods