Extended Data Fig. 3: Consequences of NIPBL and RAD21 depletion on genome organization.
a, Gating strategy used for fluorescence activated cell sorting of 2 N RAD21-B1/B3 cells for Hi-C. Cells first underwent initial steps for Hi-C, including crosslinking, restriction digestion, and proximity ligation, before being stained with propidium iodide for sorting. Post-sorting, cells were decrosslinked and sequencing libraries prepared. b, Relationship between distance and contact frequency at a 10 kb resolution based on Hi-C performed in cells depleted of NIPBL (top) or RAD21 (bottom; G1 sorted) for 4 or 24 hours. For each timepoint, Hi-C experiments performed on independent clonal cell lines for NIPBL (NIPBL-A2 and NIPBL-D7) and RAD21 (RAD21-B1 and RAD21-B3) were treated as replicates, with the resulting datasets merged for plotting. c, Pile-up analysis showing log-transformed observed/expected contact frequency from Hi-C at chromatin loops (left) and stripes (right) called from DMSO-treated FKBP12F36V-NIPBL cells. d, Saddleplots showing genome-wide changes in compartments with depletion of NIPBL (left) or RAD21 (right) for 4 or 24 hours. e, Quantification of saddle strength based on average signal strength in the saddleplot corners (top left: BB; bottom right: AA; top right: BA; bottom left: AB). f, Example heatmaps showing the consequences of NIPBL (top) and RAD21 (bottom) depletion for 4 and 24 hours. Beneath the NIPBL heatmaps are scale factor-normalized RAD21 ChIP-seq tracks for the corresponding treatment in NIPBL-D7 cells. A chromatin loop of interest is highlighted with an arrow in the heatmaps, while the loop anchors in the ChIP-seq tracks are marked a grey box.
