Dissection of BET Protein Function in a Hematopoietic Differentiation Model

Abstract

The mammalian Bromodomain and Extra-Terminal motif (BET) proteins associate with acetylated histones and transcription factors including the master erythroid regulator GATA1. Pharmacologic inhibitors that broadly target BET family proteins are being evaluated in clinical trials for hematologic malignancies. Additionally, a rapidly growing number of studies in model organisms and in vitro systems suggests BET modulation is a therapeutic avenue in diverse disorders. However the mechanisms through which BET proteins act as well as the contributions of individual BET family members to biological processes remain mostly uncharacterized.Previously we showed that BET family members Brd3 and Brd4 can bind to GATA1 in an acetylation-dependent manner and that pharmacologic BET inhibition impairs erythroid maturation (Lamonica et al PNAS 2011). Using a combination of genome-wide occupancy analysis, pharmacologic inhibition, genome editing and knockdown we characterized the mechanism and function of BET proteins in the context of GATA1-driven erythroid differentiation. BET inhibitors prevented GATA1-mediated transcriptional activation, but not repression, genome-wide. Mechanistically, GATA1 required BETs both for initial chromatin association and for transcriptional activation following the establishment of GATA1 occupancy. As BET proteins associate with chromatin during mitosis when transcription is globally disrupted, they have been implicated as mitotic “bookmarks”. BET inhibition specifically during mitosis failed to elicit a measurable impact on post-mitotic gene activation, calling into question a role of BET proteins as mitotic bookmarks in this system.To determine the BET protein(s) most relevant to GATA1-activated transcription, we genomically disrupted BET family members Brd2, Brd3 and Brd4 using the CRISPR/Cas9 tool. Despite being present at nearly all GATA1-occupied sites genome-wide, Brd3 disruption had little impact on GATA1-dependent transcription. In contrast, both Brd2 and Brd4 were individually required for efficient gene activation by GATA1. A Brd3 requirement was only revealed in the setting of Brd2 deficiency, indicating functional compensation among select BET proteins. These results provide a comprehensive definition of the functions of BET proteins in a model of cellular differentiation. They further suggest that pharmacologic BET inhibition should be viewed in the context of distinct steps in transcriptional activation and overlapping functions among BET family members. DisclosuresNo relevant conflicts of interest to declare.