Context-dependent activity of a pioneer transcription factor, ASCL1, in early development

Abstract

Transitions from one cell type to another are fundamental to developmental biology and they require the concerted activity of transcription factors (TFs) to ensure the expression of the right genes at the right time. Some TFs, so-called pioneer TFs, direct cell identity by opening previously closed chromatin to drive gene expression. However, there are mixed reports in the literature about the sensitivity of pioneer TFs to the epigenetic landscape. Moreover, the epigenome is remodelled during the exit from pluripotency, and it is not well understood how these changes in turn influence the activity of lineage-specifying pioneer TFs. This project sets out to address these unanswered questions, using the proneural ASCL1 as a model pioneer TF. Using a mouse embryonic stem cell line with inducible ASCL1 activity, I showed that cells undergo ASCL1-directed neuronal differentiation after priming towards a neuroectoderm identity. However, pluripotent cells do not undergo neurogenesis for two reasons: Failure to activate key neuronal target genes, and inappropriate expression of many non-neuronal genes. Next, I demonstrated that these divergent transcriptional responses to ASCL1 are caused by cell type-specific ASCL1 binding. Comparison with publicly available data indicated that the favourability of binding sites in a given cell is partly determined by the epigenetic landscape. Consistently, most ASCL1 binding sites were in accessible chromatin. ASCL1 pioneer activity is relevant at a subset of sites, but this activity is cell type-specific and its contribution to neurogenesis remains undetermined. Taken together, the data in this dissertation support the central thesis that ASCL1 is more sensitive to developmental changes in chromatin than reported previously. I propose that prior priming of the epigenome directs ASCL1 to a set of sites that are required for differentiation. Pluripotent cells that have not yet been primed present a different set of possible binding sites in open chromatin and therefore ASCL1 binding is biased towards non-neuronal enhancers. I also speculate, based on enrichment of additional motifs at a subset of ASCL1 binding sites, that cooperative interaction with other TFs confers an additional layer cell type-specific regulation. This work raises fundamental questions about how the epigenetic landscape is in turn remodelled to facilitate the activity of ASCL1 and pioneer TFs more generally.