Differentiation in vitro of striated muscle from apparently non-muscle precursor cells has been reported in thymus reticulum1, a fibroblast-like mouse embryo line2 and in a neuronelike cell line derived from a rat brain tumour3. Also Tomozawa and Sueko reported the differentiation of a peripheral neurotumour clonal stem cell line into separate neuronal and glial cell types4. We report here the reproducible and stable phenotypic change of a well characterised rat glial cell line, B9 (refs 5-;7), into multinucleate contractile skeletal muscle. The B9 line was derived from a nitrosoethylurea-induced brain tumour, contains S-100 and 14-3-2 proteins, is electrically non-excitable 5 and has a putative glial-specific surface antigen, G2 (ref. 6). It has been phenotypically stable in its laboratory of origin at Salk Institute since 1973. After being transported to the Mayo Clinic in Minnesota the cells began to change from cuboidal to elongate shape, fuse into contractile multinucleate fibres and express nicotinic acetylcholine receptors (AChR) on their plasma membranes. Striated myofibrils appeared in their cytoplasm. The first sign of mesectodermal differentiation in B9 was the transitory appearance on its surface of Thy-1 which, in the rat, is a marker of thymocytes8, certain brain cell lines7, immature skeletal muscle9,10, mammary myoepithelial lines11 and fibro-blasts12. This antigen was not detected in previous studies at Salk Institute7. The inductive influence is not yet known. We propose that transformation of a neurectodermal line into muscle may be evidence of the capacity of mammalian neurectoderm to give rise to skeletal tissues. The concept that skeletal muscle in different anatomical regions might have different embryonic origins could be relevant to the distribution of muscles affected by certain diseases in man.
ASJC Scopus subject areas