(A) Experimental overview and summary of overlaps in enriched Reactome gene sets between comparisons of mouse ccRCC with normal primary cells and human ccRCC cell lines with RPTECs. UpSetR plot depicts the number of significantly enriched terms (Benjamini-Hochberg adj. P < 0.05) within each individual comparison (x axis) and of overlapping terms (y axis) between the indicated comparisons (dots). Box lists DNA damage–relevant terms that overlap between at least 5 individual comparisons. (B–E) Gene set enrichment plots for Reactome term “DNA repair” for mouse ccRCC cell line versus normal primary cells (B), human ccRCC cell lines versus RPTECs (C), mouse ccRCC tumors versus normal mouse renal cortex (D), and human ccRCC tumors (TCGA-KIRC) versus matched normal (E). Numbers in brackets represent number of samples. Colored lines show the running enrichment score (RES). Ranked list metric (RLM) is shown (bottom). P values were adjusted using Benjamini-Hochberg correction. (F) Immunohistochemical staining for γ-H2AX in normal mouse renal cortex and ccRCC in Vhl/Trp53/Rb1 mutant mice and quantification of positively stained nuclei in normal (n = 10) and ccRCC (n = 92 tumors from 16 mice). Mean ± standard deviation (std. dev.). Two-sided P value calculated by Student’s unpaired t test with Welch’s correction. Scale bar = 50 μm. (G) Immunohistochemical staining for γ-H2AX in normal human renal cortex and ccRCC and quantification of positively stained nuclei in paired samples of normal and ccRCC (n = 10). Two-sided P value calculated by Student’s paired t test. Scale bar = 50 μm. (H) Distribution of TCGA-KIRC sample-specific normalized enrichment scores (NESs) for Reactome pathways. Violins show the kernel probability density of the NES deriving from ssGSEA; boxes indicate median and interquartile range (between 25th and 75th percentiles). Groups were compared using 2-sided Mann-Whitney U tests without multiple comparisons (****P < 0.0001). VpR, Vhl/Trp53/Rb1 mutant cell lines.