Regulated expression of muscle-specific genes introduced into mouse embryonal stem cells: inverse correlation with DNA methylation

Abstract

Abstract Pluripotent embryonal stem cell lines (ES) were isolated from cultured normal mouse blastocysts. These cells retained their capacity to differentiate into a great variety of cell types in cell cultures or in tumors formed after subcutaneous injection of the cells into nude mice. A chimeric actin/globin gene containing about two-thirds of the rat skeletal muscle actin gene and 730 bp of its 5′flanking region fused to the 3′end of the human embryonic ε-globin gene, was inserted into a plasmid containing a neomycin resistance gene ( neo r) whose transcription is regulated by the SV40 early control elements. The prokaryotic vector DNA sequences of this plasmid ( pAG-Neo) were deleted and the two linked genes were introduced into the ES cells by electroporation. G418-resistant clones were isolated, amplified and injected subcutaneously into nude mice. From the teratocarcinoma-like tumors which developed we isolated myogenic as well as nonmyogenic cell lines. In cell lines derived from three independent transfected ES clones, expression of the actin/globin gene was developmentally regulated in myogenic cells. In contrast, in a number of experiments in which the actin/globin gene or other musclespecific genes were introduced into the ES cells without the removal of the pBR sequences, no expression could be detected at any stage. Moreover, in the differentiated lines derived from these clones, G418 resistance was lost, and no neo r transcripts could be detected. Southern-blot analysis of MSPI- or HpaII-digested DNA revealed extensive methylation in the clones that did not express the foreign DNA, whereas no significant methylation of the inserted DNA was observed in clones which expressed the transfected genes. Examination of the DNA extracted from transgenic mice carrying the same actin/globin gene revealed an inverse correlation between methylation of the exogenous gene and its potential to be expressed in the transgenic strain. However, no tissue-specific differences in methylation, related to the tissue specificity of expression of the exogenous gene, could be detected in these experiments. These results suggest that the process of methylation reported here is causally related to constitutive inactivation of the exogenous genes.