Fig 2: NRF2 cooperates with CEBPB for enhancer formation in NRF2-activated NSCLCs.a Genome-wide colocalization of NRF2 and other transcription factors at active enhancers marked with H3K27ac deposition in A549 cells. Colocalization tendency was analyzed by Jaccard test using ChIP-seq data deposited in ENCODE. Higher rank indicates better colocalization with NRF2. b Aggregation plots and heat maps of ChIP-seq data of NRF2, H3K27ac, and CEBPB surrounding NRF2 binding sites obtained from the ENCODE database. c Affinity purification of the NRF2 complex from A549 cells. A representative result of two independent experiments is shown. d Immunoblot analysis of CEBPB protein levels in nuclear extracts of four NRF2-activated and five NRF2-normal NSCLC cells. An arrowhead indicates the CEBPB isoform that corresponds to the one detected in the immunoprecipitation experiment with the NRF2 antibody shown in c. Lamin B expression was used as a loading control. The results shown are representative of 2 independent experiments. e Immunoblot analysis detecting KEAP1, NRF2, and CEBPB in three NRF2-activated NSCLC cells with GFP expression (control) or wild-type KEAP1 reconstitution. Tubulin expression was used as a loading control. The results shown are representative of three independent experiments. f ChIP assay using the H3K27ac antibody in A549 cells treated with control or CEBPB siRNA. H3K27ac deposition at NRF2-activated NSCLC-specific NRF2 target loci were examined. The average and SD of three independent experiments are shown. Two-sided confidence interval estimation was conducted for knockdown samples incubated with H3K27ac antibody. **a < 0.01, n.s.: not significant.

Fig 3: Evaluation of potential chromatin-modifying TFs. a Analysis of PFS using the TCGA dataset. Samples were divided into low- and high-expression groups based on the median expression of each gene. A survival curve of CEBPB is representatively shown. b Analysis of patient outcomes (differences in median survival times) vs. differences in gene expression. Left, progress-free survival; right, overall survival. c Gene expression meta-analysis of CEBPB in ovarian cancer. Horizontal lines represent 95% CI; diamonds represent summary estimates with the corresponding 95% CI. d IHC analysis of C/EBPß protein in HG-SOC and normal control samples. Bar, 25 µm. e, f Kaplan–Meier analysis of PFS (e) and OS (f) in patients with ovarian cancer classified based on tumor C/EBPß protein levels. PFS progress-free survival, OS overall survival

Fig 4: CEBPB-NRF2 cooperation is required for NOTCH3 enhancer formation in NRF2-activated NSCLC cells.a ChIP-seq profiles of the NOTCH3 locus in A549 cells. b Sequence comparison of flanking regions surrounding NRF2 binding sites in the NOTCH3 enhancer between human and mouse. AREs (NRF2:sMAF binding sites) are shown in red, and three tandem human AREs are enclosed by green frames. Consensus binding motifs of other transcription factors are shown in blue. c RT-PCR of NOTCH3 in A549 cells treated with siRNAs. Fold changes of the normalized values were calculated in comparison to A549 cells treated with control siRNA. The average and SD of the fold changes from three independent experiments are shown. Comparison was made between two cell groups; control siRNA and each test siRNA. Two-sided confidence interval estimation was conducted to evaluate statistical significance. *a < 0.05, **a < 0.01, n.s.: not significant. d, e ChIP assay using the H3K27ac (d) and NRF2 (e) antibodies in A549 cells treated with control siRNA or CEBPB siRNA. Fold changes of %input values were calculated in comparison to the control samples incubated with H3K27ac or NRF2 antibody. The average and SD of three independent experiments are shown except for the experiment with CEBPB siRNA2 in d, which was independently conducted for four times. Two-sided confidence interval estimation was conducted for knockdown samples incubated with H3K27ac or NRF2 antibody. *a < 0.05, **a < 0.01, n.s.: not significant. f ChIP assay using the CEBPB antibody in A549 cells. Fold changes of %input values were calculated in comparison to the control samples reacted with CEBPB antibody. The average and SD of 3 independent experiments are shown. Two-sided confidence interval estimation was conducted for knockdown samples reacted with CEBPB antibody. **a < 0.01. g DNA sequence of ?N3U A549 cells. A CEBPB binding site, shown in blue, is preserved in both ?N3U A549 clones. Deletion and substitution by genome editing are shown in red and green, respectively. h ChIP assay using the CEBPB antibody in ?N3U and wild-type (WT) A549 cells. Fold changes of %input values were calculated in comparison to WT A549 cells incubated with CEBPB antibody. The average and SD of 3 (Clone 2–4) and 4 (WT and Clone 1–1) independent experiments are shown. Comparison was made between two cell groups; WT A549 and each mutant clone. Two-sided confidence interval estimation was conducted for Clones 1–1 and 2–4 incubated with CEBPB antibody. *a < 0.05.

Fig 5: C/EBPß promotes cisplatin resistance in ovarian cancer. a The probability of cisplatin resistance in patients with ovarian cancer in the CEBPB-low and CEBPB-high groups (Pearson’s chi-squared test). mRNA and protein levels were analyzed using the TCGA dataset and IHC, respectively. b C/EBPß expression was detected by western blotting. c Colony formation of the indicated cells after exposure to cisplatin (50 µM) for 12 h. d–g One week after orthotopic inoculation with the indicated cells, mice were treated with cisplatin (5 mg/kg) or vehicle (PBS) intraperitoneally every 4 days (n = 10 per group). Tumor size was monitored by three-dimensional reconstruction of in vivo bioluminescence images (d). Six weeks after tumor inoculation, tumors were excised and weighed (e). The cisplatin-induced tumor reduction rate was calculated as described in the Methods using the following formula: (1 - Vcisplatin/VPBS) × 100% (f). IHC analysis of C/EBPß protein levels in xenograft tumor sections collected from mice treated with or without cisplatin. Bar, 25 µm (g). h IHC analysis of C/EBPß protein in paired tissue sections collected from initial operation and reoperation performed for management of tumor recurrence in the same patient. i Tumor weights from mice treated as described in d–g. Six weeks after tumor inoculation, tumors were excised and weighed. Uncropped images of blots are shown in Supplementary Figure 25. *P < 0.05; **P < 0.01; ***P < 0.001; n.s. not significant