Extended Data Fig. 8: Reversibility of HDAC inhibition-induced architectural changes and response to sequential TSA treatment.
a, Ratio of cis versus trans contacts in Micro-C datasets. b, Contact frequency in Micro-C plotted against genomic separation by compartment. c, Micro-C maps at the Nkx2-2 locus showing incomplete architectural recovery. ChIP–seq tracks of the corresponding condition are shown above. d, Diverging bar plot showing the number of differentially expressed genes at the indicated conditions. e, Differential ChIP–seq peaks identified by DESeq2 between the first (TSA) and second TSA treatments (reTSA), with 24 h or recovery in between. f, PCA showing increasing distance between DMSO and recovery samples following a second TSA treatment with increasing recovery time. g, Heatmaps showing mean expression z-scores of recovered and not recovered genes through the TSA–recovery treatment course. h, Scatterplot showing correlation between transcriptomic changes induced by sequential TSA treatment following 24-h and 48-h recovery. Fitted line represents linear regression. i, Differential H3K9ac ChIP–seq peak frequency distribution relative to TSSs of up- (left) and downregulated (right) genes that recover or do not recover.
