Fig. 3: Distinctive CPD damage signatures at SRF and ETS factor binding in the EGR1 promoter.
a Overview of the EGR1 promoter (−2000 to +1000 bp) showing how Capture CPD-seq UV footprinting signals coincide with regulatory features. Predicted binding sites (sequence motif matches) for SRF (CArG boxes, CC[A/T]6GG) and ETS (TTCC[T/G]) are shown together with ENCODE DNase I and ChIP signals for SRF and ETS (ELK4/ELK1). Per-base CPD damage changes in cellular vs. naked or serum-stimulated vs. starved samples are shown as log2 ratios (+ 4 = 16-fold increase; −4 = 16-fold decrease) together with associated P-values (uncorrected from two-sided negative binomial test). CPD-based prediction of bound/unbound status for motif sites is indicated in black/gray, respectively (see Supplementary Fig. 10 for details). b Closeup view of region “1” indicated in Fig. 3a. Cell/naked differential CPD signals (up to 15.6-fold) stem primarily from ChIP-supported SRF and ETS motif matches in previously established serum response elements (SREs)19,28. The sites exhibited distinctive repeating CPD damage signatures. Each bar refers to one diPy (only one strand can contain a diPy at each position). c Closeup view of a second SRE cluster closer to the EGR1 TSS (region “2” in Fig. 3a), containing one informative SRF/CArG element (at −115 bp), and ETS site, and a second CArG lacking diPys in key informative positions. d Altered CPD formation in serum-stimulated compared to starved samples at the EGR1 −115 bp CArG element (see Supplementary Fig. 8 for details). e, f Close-up view of unbound (non-ChIP-supported) ETS and SRF/CArG sites in EGR1 (region “3”) and MYC, respectively, both lacking notable damage modulation signals. DNase I footprints from Vierstra, et al.31 are indicated as green lines in b–f. Source data are provided as a Source Data file.
