Functional characterization of mSWI/SNF complexes in normal and malignant human hematopoiesis

Abstract

The mammalian SWI/SNF (mSWI/SNF) complexes are chromatin remodeling complexes responsible for modulating chromatin compaction and subsequent changes in gene expression. These complexes are broadly defined as three main subcomplexes: canonical BAF (cBAF), polybromo-associated BAF (PBAF) and non-canonical BAF (ncBAF). The recently discovered ncBAF complex is the most structurally divergent and the function of ncBAF is thus far poorly characterized. In the present work, we sought to understand ncBAF function by leveraging genome-wide fitness dropout screens across > 1,000 cell lines. We identified components of the Spt-Ada-GCN5 acetyltransferase (SAGA) complex as top correlators across almost all cell lines. We employed biochemical as well as bulk and single-cell genomic analyses to ascertain the mechanism underpinning this correlation. Our results indicate the histone acetyltransferase activity of the SAGA complex is responsible for directing genomic localization of the ncBAF complex. This relationship has profound implications in hematopoietic stem cells (HSCs), where loss of either the ncBAF or SAGA complex results in an upregulation of granulocyte gene expression programs. These results not only strongly suggest a role for the ncBAF and SAGA complexes in maintaining cell identity, but also reveal the underlying mechanism. In characterizing ncBAF-SAGA complex relationship, we also investigated the roles of the cBAF and PBAF complexes in both long-term HSC maintenance and myeloid differentiation. We uniquely report on the roles of each mSWI/SNF subcomplex in normal hematopoiesis and HSC maintenance in humans, where most previous reports are restricted to the mouse context. In addition to this effort, we sought to deepen our understanding of the importance of the mSWI/SNF complexes in malignant hematopoiesis. We once again leveraged genome-wide fitness dropout and small molecule screens across a wide range of cell lines to identify pan-mSWI/SNF disruption as a sensitivity in acute myeloid leukemia. In order to improve the therapeutic window of small molecule-mediated mSWI/SNF inhibition by FHD-286, we profiled the general chromatin occupancy of these complexes and chromatin accessibility in these cells. We also undertook an effort to identify agents that may synergistically inhibit AML cell proliferation when used in combination with FHD-286. We found that venetoclax uniquely exhibits a synergistic signature in several AML cell lines. Taken together, the results reported in this body of work signify an effort to characterize the roles of mSWI/SNF complexes in both normal and malignant hematopoiesis, with profound implications at both the basic science and clinical levels.