1025 - A Quantitative Perspective of Transcription in Pluripotent and Differentiated Cells

Abstract

The differentiation of embryonic stem cells (ESC) requires three fundamentally distinct transitions in transcriptional state: 1. the activation of genes that are fully silent in ESC but transcribed in the differentiated state, 2. the silencing of genes that are transcribed in ESC but must be fully inactive in the differentiated cells, and 3. The increased or decreased transcription of genes that are actively transcribed in both populations. Because distinct mechanisms are likely to regulate each type of transition, the genes included in each category must be carefully delineated before the mechanisms can be elucidated and fully appreciated. However, genes in categories 1 and 2 have not been separated adequately from genes in category 3 because studies of differential gene expression typically use low stringency thresholds to define differential expression; they therefore group genes that are moderately upregulated or downregulated with those that transition between active and silent states. A quantitative analysis of nascent transcript RNA-seq data sets revealed that the vast majority of differentially expressed genes in mouse ESC in comparison to three primary differentiated cell types exhibit dynamic ranges of expression of less than 10-fold, with only a small number of genes exhibiting a dynamic range of 100-fold or more. This finding raises a number of questions, one of which is the following: What are the functions of the thousands of binding sites for pluripotency transcription factors, such as Oct4 and Sox2, if the vast majority of differentially expressed genes change expression level by only a small magnitude as cells leave the pluripotent state? As expected, a small number of these binding sites appear to be involved in the regulation of the small number of genes that exhibit large dynamic ranges of expression in ESC versus differentiated cells. However, do the thousands of other binding sites support small (e.g. 2-fold) changes in expression of genes that are efficiently expressed in both the pluripotent and differentiated states? If so, do they regulate these genes by binding constitutively active promoters and enhancers or, alternatively, by activating dedicated enhancers that support small changes in gene expression? Or, do they maintain the active state of constitutively expressed genes in ESC without playing a critical role in boosting or suppressing transcription? By combining nascent transcript RNA-seq with pluripotency factor ChIP-seq data sets and CRISPR/Cas9 disruption of individual Oct4:Sox2 composite binding sites, insights into these and other unanswered questions are beginning to emerge.