Regulation of cardiomyocyte differentiation from human embryonic stem cells

National Heart, Lung, And Blood Institute

Investigators

Abstract

The post-translational covalent modifications of the histone proteins that comprise the nucleosome have been shown to be associated with either transcriptional activation or repression. Previous studies have shown that the distribution of such histone marks can be used to predict cis-regulatory element activities in different biological contexts. Building on these findings, in the present project we have used predictive gene expression signatures within a multi-species framework to identify candidate genes that comprise the gene regulatory networks that govern cardiac cell fate decisions in differentiating mammalian embryonic stem cells (ESCs). Moreover, large numbers of predicted cardiogenic genes were validated for their inferred cell-specific functions using a rapid and cost-effective RNAi-based screen in Drosophila embryos followed by phenotypic characterization and transcriptome profiling of human ESCs subjected to shRNA knockdown of a limited number of the corresponding human orthologs. To identify human cardiogenic genes from published chromatin immunoprecipitation (ChIP) data derived from human ESCs as they differentiate along the cardiac lineage, such genes should show: (1) histone marks characteristic of repression in the ESC state, (2) chromatin modifications consistent with activation in the cardiomyocyte (CM) state, and (3) upregulation of gene expression as the pluripotent cells become restricted during CM differentiation. The genes possessing such a combined epigenetic and gene expression signature exhibited the expected characteristics of cardiogenic genes upon gene ontology analysis. A parallel study using published histone mark data from mouse ESCs yielded complementary results, with those overlapping with the human cardiogenic gene candidates having the highest probability of being correctly assigned to promoting the CM cell state. Moreover, the majority of the latter genes were conserved in Drosophila, enabling a rapid fly-based RNAi screen of the fly orthologs of the candidate mammalian cardiogenic genes to confirm their anticipated heart developmental functions. These results for the mammalian heart gene candidates were next validated by shRNA screening of human ESCs which resulted in the prevention of these cells acquiring heart cell identities. In addition, transcriptome profiling using RNA-seq of human ESCs treated with shRNA directed against the newly identified human cardiogenic genes further demonstrated that these heart promoting functional predictions were correct. Of note, instead of these shRNA-treated ESCs differentiating along the cardiac lineage, we found that the pluripotent stem cells instead assumed alternative cellular identities such as hematopoietic or hepatic fates, as reflected in their transcriptome profiles. These findings suggest that not only do certain genes promote CM differentiation, but they also repress the formation of alternative cellular lineages. In summary, the present results demonstrate the utility of computational analysis of combined orthologous genomic datasets with empirical testing by RNA knockdown and transcriptome profiling of shRNA-treated ESCs under CM differentiation conditions to identify the individual genes and their association in regulatory networks that govern formation of the human cardiac lineage. Moreover, the strategy employed here establishes an integrated platform for conducting similar studies in other experimental systems.