Fig 2: DNA damage induces EREG expression in stromal cells via regulation by the NF-κB complex, C/EBP axis and epigenetic modifications. a Schematic of putative NF-κB binding sites in the proximal region of EREG promoter. A set of reporter constructs was generated by sequential cloning of the promoter fragments into a pGL4.22 vector (pGL-EREG-P01 to P05) that expresses firefly luciferase. Numeric numbers on the top denote the core site of each putative NF-κB binding motif, while numbers at the left mark the length of each segmental promoter clone. TSS transcription start site. Lower-left inlet, consensus binding motif of the NF-κB subunit p65. b Assessment of luciferase activities upon exposure of 293F cells pre-transfected with the individual EREG promoter constructs to TNF-α at 40 ng/ml in culture. The empty vector was used as a negative control, while a construct NAT11-Luc2CP encoding multiple copies of typical NF-κB binding sequences and an optimized IL-2 minimal promoter served as a positive control. Signals were presented as relative ratios of firefly/renilla luciferase activities. c Luciferase activity assay with lysates of PSC27 cells pre-transfected with each of the constructs used in b prior to treatment by 50 μg/ml bleomycin (BLEO) in culture. d Chromatin immunoprecipitation (ChIP) was performed to identify potential NF-κB binding sites in the proximal promoter of EREG. Left, EREG-p2/p3/p4/p5 denotes four representative genomic sites in EREG promoter region, while selective NF-κB binding sites from IL6 and CXCL8 served as positive controls. e EREG and MMP3 transcript expression in PSC27 cells exposed to BLEO, MIT (mitoxantrone) or DOX (doxorubicin), with or without the NF-κB inhibitor BAY (Bay 11-7982, 5 μM). Signals were normalized to untreated cells, with MMP3 expression analyzed as positive control. f The reporter construct pGL-EREG-P05 was transiently transfected into PSC27 cells before treatment by BLEO. BAY (5 μM), BA (betulinic acid, 10 μM), T-5224 (10 μM) were applied with BLEO as small molecule inhibitors against NF-κB, C/EBP family and AP-1, respectively. SR (SR 11302, 3 μM), a positive control inhibitor against AP-1. Cells were lysed 7 days after treatment, with lysates subject to luciferase activity assay. g PSC27 cells were treated in the same conditions as described in f, with lysates collected for total RNA preparation and subject to quantitative RT-PCR analysis. Expression of EREG (left), IL6 (mid) or CXCL8 (right) was compared between CTLR (untreated), Mock (PBS-treated), BAY, BA, T-5224 and SR treatment groups. Cells were damaged by BLEO (50 μg/ml) or VBL (vinblastine, 20 nM) treatment. h Immunoblot analysis of DDR signaling (ATM), p38MAPK activation, cellular senescence (p16, p21) and NF-κB activation (p65) in PSC27 cells treated by various chemotherapeutic agents as indicated. GAPDH, loading control. i Immunoblot analysis Expression assay of p65 nuclear translocation in PSC27 cells treated by VBL, PTX, BLEO or MIT, individually. C cytoplasmic, N Nuclear. Histone H3, loading control for nuclear proteins. Note, the relative signal intensities (RSI, presented as percentage) of p65 were quantified as the virtual intensity of an individual sample after scanning, and calculated in relative to that of the strongest signal (BLEO, C for the p65 blot). j Presentation of p65-specific ChIP-seq tracks of the gene locus of several SASP hallmarks and senescence-associated factors. Illustrations were prepared from datasets deposited in the GEO (accession number GSE141992), with raw data available at publicly released sources [29]. Data are representative of three independent experiments. All p values were calculated by Student’s t tests. ^p > 0.05, *p < 0.05, **p < 0.01, ***p < 0.001.

Fig 3: EREG induces profound changes of PCa cell expression profile and promotes phenotypic reprogramming.a Heatmap depicting differentially expressed human transcripts in PC3 cells after a 3-days culture with EREG-containing CM collected from PSC27 cells. In contrast to cancer cells cultured with control CM (CTRL), 970 and 1362 genes were upregulated and downregulated, respectively, in those treated with the CM from EREG-expressing PSC27 cells (EREG). b Graphical visualization of pathways by GO profiling. Those significantly enriched genes in the upregulated list were sorted according to their fold change in PC3 cells exposed to the CM of EREG-expressing PSC27 cells. c Venn diagram displaying the overlap of 39 transcripts upregulated in PC3 and DU145 cells upon treatment with EREG-containing CM from stromal cells (970 and 309 genes with unique annotations for PC3 and DU145, respectively). d Statistics of transcripts differentially expressed (fold change either ≥2 or ≤0.5, with p < 0.05) in PC3 and DU145 upon EREG stimulation, and classified into typical categories according to functional annotations mapped by Genecode (V27). e Heatmap showing the top 39 transcripts upregulated by both PC3 and DU145 cells, sorted according to their expression fold change in PC3. f Pie chart depicting the biological processes associated with transcripts upregulated by EREG after GO analysis of the 39 transcripts in PC3. g Quantitative RT-PCR measurement of the expression of KIF20A, MARCHF4 and SPNS2 in both PCa lines upon exposure to CM of stromal cells expressing EREG. Signals normalized to those of cells exposed to PSC27 cells transduced with vector. h Dose-response curves (non-linear regression/curve fit) plotted from drug-based survival assays of PC3 cells transduced with vector or MARCHF4 construct and treated by a range of concentrations of MIT. i Dose-response curves (non-linear regression/curve fit) plotted from drug-based survival assays of DU145 cells treated in a manner similar to that of PC3 cells. j Immunoblot assessment of protein expression of EMT-associated molecules. CD81, a downstream target of MARCHF4. β-actin, loading control. k Immunoblot profiling of apoptosis-related factors of self-cleavage activity in both PCa cell lines pre-transduced with vector or MARCHF4 construct and exposed to MIT for 72 h. β-actin, loading control. Data in g–k are representative of three independent experiments. All p values were calculated by Student’s t tests. ^p > 0.05, *p < 0.05, **p < 0.01, ***p < 0.001.

Fig 4: EREG is a novel circulating biomarker indicative of the SASP in vivo and predicts adverse therapeutic outcome in cancer clinics.a Abundance of EREG protein in the serum of untreated and chemo (MIT)-treated PCa patients. Data were derived from ELISA measurement and shown as mean ± SD. N = 20. b Abundance of CXCL8 protein in patient serum analyzed in a. Data from ELISA assays and presented as mean ± SD. N = 20. c Scatterplot showing correlation between EREG and CXCL8 in the serum of individual patients studied in a and b. Pearson’s correlation coefficient, p value and confidence interval are indicated. d Immunoblot examination of EREG and CXCL8 circulating in the serum of randomly selected PCa patients from untreated and chemo (MIT)-treated groups, respectively (n = 6 per group). Albumin, loading control for patient serum protein. e Heatmap depicting the overall correlation between stromal EREG, serum EREG, stromal CXCL8 and serum CXCL8 in chemo (MIT)-treated patients (n = 10). The raw scores of stromal EREG and CXCL8 were derived from independent pathological reading of primary tumor tissues of PCa patients, with those of serum EREG and CXCL8 obtained from ELISA assays. Color key, relative expression of these two factors in stromal tissue or patient serum. f Heatmap showing the relative expression of a panel of SASP signature factors in the tumor stroma of PCa patients, and the correlation of EREG/CXCL8 with these factors (n = 20). Stromal cells in the tumor tissues were isolated via LCM and expression of each target factor was measured by qRT-PCR, with signals per factor group normalized to that of the sample showing the lowest expression value. A subset of inflammatory factors typically not considered as SASP components was examined as random control (IL-2/3/5/12). Correlations of these factors are presented as dendrogram organized by hierarchical clustering. Trace lines indicate the trends of signal main streams. g Kaplan–Meier survival analysis of chemo (MIT)-treated PCa patients. Disease-free survival (DFS) stratified according to EREG expression in tumor stroma (low, average score <2, blue line; high, average score ≥2, yellow line). DFS represents the length (months) of period calculated from the date of chemotherapy to the point of first time disease relapse. Survival curves generated according to the Kaplan–Meier method, with p value calculated using a log-rank (Mantel–Cox) test. N = 10 per group. h TCGA data show alterations of EREG in human prostate cancer patients at genomic level, including mutation, amplification and deep deletion. Alteration frequency is displayed in percentage. Data in a–c are representative of three independent experiments. **p < 0.01, ***p < 0.001. p values were calculated by Student’s t test (a, b), Pearson test (c) and log-rank (Mantel–Cox) test (g). ^p > 0.05, **p < 0.01, ***p < 0.001.

Fig 5: Stromal EREG alters multiple phenotypes of prostate cancer cells in vitro.a Immunoblot analysis of EGFR-associated pathways in PC3 and DU145 cells treated by the CM from PSC27 cells transduced with the empty vector or EREG construct, or alongside the EGFR inhibitor AG-1478 (2 μM). Antibodies of p-EGFR (Y845), p-Akt (S473), p-mTOR (S2448), p-MEK (S217/S221) and p-ERK (T202/Y204) were applied to probe the individual molecules. Total protein per molecule and GAPDH were used as loading control. b Schematic diagram of the construct encoding the mature chain of EREG (upper) and immunoprecipitation (IP, lower) followed by immunoblot assay of EGFR and His-EREG (fusion protein) in the whole lysates of PC3 cells. PC3 was treated by the CM of PSC27Vector and PSC27His-EREG for 3 days. Antibodies including IgG and anti-EGFR were used for IP, with both EGFR and His-EREG in inputs analyzed. c Measurement of cellular senescence by quantification of SA-β-Gal staining positivity. Stromal cells were pre-transduced with shRNAs and treated by BLEO. Upper, statistics. Lower, representative images. Scale bar, 20 μm. d PCa cells were treated with the CM from PSC27 sublines for 3 days, and subject to cell proliferation assay. Native and shRNA-transduced PSC27 cells as indicated were treated by bleomycin (BLEO), with the conditioned media (CM) collected 7 days after drug treatment and used for PCa cell culture. The CM were collected from equal number of cells per condition, with a starting DMEM that contains 0.5% FBS to make the CM. e Migration assay of PCa cells seeded within transwells in 6-well plates, with cells cultured for 3 days in the CM from PSC27 sublines depicted in d. f Invasiveness appraisal of PCa cells across the transwell membrane upon culture with the CM from PSC27 sublines described in d. g Chemoresistance assay of PCa cells cultured with the CM from PSC27 sublines described in d. MIT (mitoxantrone) was applied at the concentration of IC50 value pre-determined per cell line. AG-1478 (2 μM), cetuximab (50 μg/ml) or EREG mAb (1 μg/ml) were applied alongside with PSC27 CM. h Dose-response curves (non-linear regression/curve fit) plotted from drug-based survival assays of PC3 cells cultured with the CM of PSC27 native or damaged by bleomycin (PSC27-BLEO), and concurrently treated by a wide range of concentrations MIT. AG-1478, cetuximab or EREG mAb (1 μg/ml) were applied with PSC27 CM. Data are representative of three independent experiments. All p values were calculated by Student’s t tests. ^p > 0.05, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.