Fig. 4

P53-occupied C2 enhancers gain H3K27ac upon treatment with Nocadazol and transiently during MEF to iPSC reprogramming. a Luciferase signal in WT and Trp53^−/− ESCs at selected P53-bound STARR-seq peaks. **p = 0.007, *p = 0.02, paired t test, n = 5. Error bars denote the standard error of the mean for technical triplicates. b Number of P53-C1 and P53-C2 ChIP-seq peaks with STARR-seq enrichment (FC ≥ 3 and p < 0.05 in both replicates). The percentage of P53-bound loci with STARR activity is similar between 2iL and SL. c Riverplot showing the number of P53-C1 and P53-C2 STARR-seq peaks. Ninety percent of these peaks do not change from C1 to C2 class between 2iL and SL. 249 peaks change from a C1 in 2iL to a C2 in SL and 134 peaks change from a C2 peak in 2iL to a C1 peak in SL. d Average ATAC-seq-derived nucleosome occupancy of P53-C1, P53-C2, and OC4T-SOX2-NANOG (OSN) peaks relative to randomly selected regions. P53-C1 and P53-C2 regions show a similar nucleosome occupancy pattern, but the lower average nucleosome occupancy of P53-C1 regions indicate that these loci are more frequently depleted of nucleosomes compared to P53-C2 regions. OSN bound regions are even more frequently depleted of nucleosomes. Nucleosome occupancy was computed with NucleoATAC for each nucleotide and smoothed using a running median over 11 adjacent nucleotides (see the “Methods” section) [PD]. e Number of differentially expressed genes (DEGs; p < 0.05 and |FC| ≥ 2.5) in Trp53^−/− ESCs compared to WT. f Mutual overlap between P53-C2 loci (n = 1551) and H3K27ac peaks for n = 2032 murine H3K27ac ChiP-seq libraries as deposited in the Cistrome database. The 23 H3K27ac libraries that had the highest overlap with our P53-C2 loci are shown. H3K27ac peaks in Nocodazole-treated ESCs overlap more than 80% of the P53-C2 loci. At the same time, the 1551 P53-C2 loci also overlap 8% of all the H3K27ac peaks detected in the Nocodazole-treated ESCs. This high mutual overlap indicates that Nocodazole treatment induces a specific deposition of H3K27ac at P53-C2 loci. The red dashed lines show that 24% of the P53-C2 loci intersect a H3K27ac peak in WT ESCs, which comprises 0.7% of the total number of H3K27ac peaks detected in WT ESCs (union 2iL and SL). See Table S6 for accession numbers and descriptions of these samples. g Left: Heatmap of H3K7ac, STARR-seq, and P53 occupancy in ESCs. Loci are ordered by difference (descending) in Nocodazole-treated (Noc) vs untreated ESCs. Right. Same loci showing H3K27ac in Nocodazole-treated and untreated erythrocyte progenitors [PD]. h Mouse embryonic fibroblasts (MEFs) that are being reprogrammed to iPSCs transiently gain H3K27ac at P53-C2 loci [PD]. i Examples of P53-C2 loci that transiently gain H3K27ac during iPSC reprogramming [PD]. Some of the panels in these figures contain public data. These panels are annotated with [PD]. The accession numbers of public data and their corresponding panels are annotated in Additional file 2: Table S1