Transcriptional regulatory networks vs. epigenetic constraints

9 October 2020

Ellen Rothenberg
Division of Biology & Biological Engineering
California Institute of Technology

ZOOM RECORDING

Abstract

Embryonic development is famously programmed in the genome based on the mobilization of canonical transcriptional regulatory networks. Embryonic patterning-associated gene regulatory networks, in turn, have revealed many of the principles of gene regulation as it is mediated by combinations of transcription factors acting on enhancers. In the well-studied embryonic cases, however, not only do target genes respond rapidly to changing combinations of active transcription factors, but also the regulatory states achieved (e.g. pair-rule gene stripes in Drosophila) are often transient. A very different picture is presented when one deals with gene regulation from the perspective of reprogramming with "Yamanaka factors". There, a mature cell type can be converted to a different cell type, but even when a cocktail of multiple collaborating transcription factors is introduced simultaneously, the cells often take weeks to convert to the new cell type. There, epigenetic barriers clearly contribute to the resistance to reprogramming. But when are these epigenetic barriers established, and what is their relationship to normal developmental pathways?

In this talk, I will look at this problem by considering the case of the T lymphocyte developmental pathway from hematopoietic stem and progenitor cells. These cells end up with a very robustly established gene expression identity, autonomously maintained at the single cell level and stable over months to decades. But their precursors, stem cells, are also a long-lived cell type that happens to be multipotent. What changes in between these two stable regulatory states? Much is known about the answer, and this system offers ways to approach really general questions about the acquisition of cell-type identity in long-lived, multicellular animals. What is the role of transcription factor-mediated gene networks, establishing dynamic attractor states, as opposed the role of stepwise epigenomic changes favoring path dependence? The talk will cover the partial answers that are already available and the opportunities that lie ahead to achieve deeper understanding.

current theory lunch schedule