Global repression and variable compensation following SWI/SNF inhibition

Abstract

SWI/SNF chromatin remodelers are multi-subunit complexes that use energy from ATP to evict or reposition nucleosomes, modulating chromatin accessibility. While both activating and repressive roles have been described for the complex, recent work with pharmacological inhibitors suggests that the direct role of SWI/SNF is to promote accessibility, particularly at enhancers. However, the scope of this requirement remains unclear, and the transcriptional consequences directly associated with SWI/SNF inhibition have yet to be fully characterized. Therefore, we performed ATAC-seq and PRO-seq in mouse embryonic stem cells following two-hour treatment with BRM014, a pharmacological inhibitor of SWI/SNF’s ATPase activity, demonstrating that virtually all promoters and enhancers exhibit reduced accessibility and repressed transcription in the absence of functional SWI/SNF. Given the apparent discrepancy between this widespread repression and the limited, variable effects associated with long-term depletion of SWI/SNF, we hypothesized that the initial effects of SWI/SNF inhibition might be mitigated over longer intervals through the action of compensatory factors. Upon extending our initial BRM014 treatment experiments through eight hours, we discovered that the majority of promoters can readily recover both accessibility and transcription, while some promoters and most enhancers remain persistently repressed. We identified EP400/TIP60 as the key factor underlying promoter compensation; in the absence of this alternative chromatin remodeler, promoters cannot effectively recover from BRM014 treatment. We also defined a subset of enhancers that recover using a distinct mechanism involving CTCF and SNF2H, suggesting that compensation is a heterogeneous process. Finally, to better understand the interplay between accessibility and gene transcription, we characterized the chromatin landscape after inhibiting transcription initiation. We find that accessibility at both promoters and enhancers is reduced in the absence of transcription, providing a solid foundation for future experimental work. As a whole, the findings presented herein provide significant insight into the direct and indirect effects of SWI/SNF inhibition, highlighting the scope of SWI/SNF’s role genome-wide and reconciling existing conflicts between short-term and long-term depletion studies.