The esBAF and ISWI nucleosome remodeling complexes influence occupancy of overlapping dinucleosomes and fragile nucleosomes in murine embryonic stem cells

Nucleosome remodeling factors regulate the occupancy and positioning of nucleosomes genome-wide through ATP-driven DNA translocation. While many nucleosomes are well- and consistently positioned, some nucleosomes and nucleosome-like structures are more sensitive to nuclease digestion or transitory in nature. To better understand the role of nucleosome remodeling factors in generating and clearing these alternative nucleosome structures, we depleted murine embryonic stem cells of the remodeler ATPases BRG1 and SNF2H then performed MNase-seq in murine embryonic stem cells. We performed MNase-seq under high- and low-MNase conditions to assess the effects of nucleosome remodeling factors on nuclease-sensitive or “fragile” nucleosome occupancy. In parallel, we gel-extracted MNase-digested fragments to enrich for another alternative nucleosome structure, the overlapping dinucleosome. Overlapping dinucleosomes are composed of two canonical nucleosomes, asymmetrically lacking one H2A:H2B dimer and wrapped by ~250 bp of DNA. In vitro studies of nucleosome remodeling have suggested that the collision of adjacent nucleosomes by nucleosome sliding can stimulate formation of overlapping dinucleosomes. Using these methods, we were able to identify fragile nucleosomes and overlapping dinucleosomes near transcription start sites and gene-distal DNaseI hypersensitive sites in mouse embryonic stem cells, among other loci. We find that BRG1 consistently stimulates occupancy of fragile nucleosomes but represses occupancy of overlapping dinucleosomes through its nucleosome remodeling function, while SNF2H expression slightly increases fragile nucleosome occupancy. 36 MNase-seq samples are included, comprised of three independent replicates with knockdown of EGFP (control, unexpressed), Smarca4, and Smarca5. Independent knockdowns were split into gel-extracted libraries to enrich for overlapping dinucleosomes and unextracted libraries for traditional MNase-seq (no size selection). Each experiment was performed in triplicate under both high (60U) and low (10U) MNase digestion conditions. MNase-ChIP-seq samples. Two replicates are included for wildtype samples and one replicate is included for EGFP KD samples. Digestion was performed with high- and low-MNase treatments (30U and 5U, respectively). Input samples digested with low MNase are provided for all samples, as are IgG- and H3-ChIP samples.