Figure 5.

CDKi affects the transcription of eRNA from Enhancers and Super-enhancers. (A) Venn diagram showing the overlap between PRO-seq called enhancers and H3K27ac ChIP-exo peaks from Kasumi-1 cells indicates that the majority of the PRO-seq called enhancers were marked by H3K27ac. See also Supplemental Figure S5A (B, C) Meta-analysis plots for the PROseq signal around enhancer centers are shown for commonly overlapped enhancers from CDK9i (B). (C) Similar analysis as in (B) but using 124 enhancers with increased polymerase around the enhancer center (top panel) after THZ1 treatment or 71 enhancers with decreased polymerase around enhancer center (bottom panel). See also Supplemental Figure S5B for the genes associated with these enhancers. (D) Meta-analysis of long eRNA from THZ1 treatments that had either an increase (top panel) or a decrease in gene body (bottom panel) showed a concomitant increase in polymerase density around the start site of the long eRNA. See also Supplemental Figure S5C and S5D for the general distribution of these long eRNA. (E) Hockey stick plot of PROseq identified intergenic enhancers ranked according to the increasing amounts of H3K27ac signal. Those enhancers that had high H3K27ac signal and above the slope of 1 were termed ‘super-enhancers’ (SE) and the ones below the slope of 1 were considered typical enhancers (TE). (F) Metaplot of PROseq signal at the super enhancer regions after THZ1 treatment showed a slight accumulation of RNA polymerases at the enhancer center in comparison to DMSO. Metaplots were generated using HOMER. (G) Box plots of nearest neighbor genes associated with SE and TE against all genes (All) were plotted for the log₂FC of the normalized PROseq read counts in the gene body. *P = 0.02, **P = 6.96e–05. Comparison testing was done by Kolmogorov–Smirnov (KS) test. (H) IGV browser screenshots of genome tracks for a ‘super-enhancer’ located downstream of SPI1 that was affected by THZ1 treatment.