Primary human myoblasts display only minor alternative polyadenylation compared to the transformed C2C12 model of muscle differentiation

Abstract

Alternative polyadenylation has been linked to multiple developmental and disease transitions. The prevailing hypothesis being that differentiated cells use longer 3 UTRs with expended regulatory capacity whereas undifferentiated cells use shorter 3 UTRs. Here, we describe the gene expression and alternative polyadenylation profiles of human primary myoblasts over a time course of differentiation. Contrary to expectations, only minor changes in the 3 end choice were observed. To reconcile this finding with published research, we devised a new bioinformatic method to compare the degree of alternative polyadenylation in the differentiation of primary human and immortalized murine (C2C12) myoblasts. Differentiated human primary myotubes display only half the alternative polyadenylation of the mouse model, with less than 1/10 of the genes undergoing alternative polyadenylation in C2C12 cells showing evidence of alternative processing in human primary muscle differentiation. A global reduction in the expression of cleavage and polyadenylation factors in C2C12, but not in primary human myotubes may explain the lack of alternative polyadenylation in this system. Looking more broadly at transcriptome changes across differentiation shows that less than half of the genes differentially expressed in the immortalized model were recapitulated in primary cells. Of these, important metabolic pathways, such as glycolysis and sterol biosynthesis, showed divergent regulation. Collectively, our data caution against using immortalized cell lines, which may not fully recapitulate human muscle development, and suggest that alternative polyadenylation in the differentiation of primary cells might be less pronounced than previously thought.