Treatment-induced damage to the tumor microenvironment promotes prostate cancer therapy resistance through WNT16B

Abstract

Acquired resistance to anticancer treatments is a substantial barrier to reducing the morbidity and mortality that is attributable to malignant tumors. Components of tissue microenvironments are recognized to profoundly influence cellular phenotypes, including susceptibilities to toxic insults. Using a genome-wide analysis of transcriptional responses to genotoxic stress induced by cancer therapeutics, we identified a spectrum of secreted proteins derived from the tumor microenvironment that includes the Wnt family member wingless-type MMTV integration site family member 16B (WNT16B). We determined that WNT16B expression is regulated by nuclear factor of κ light polypeptide gene enhancer in B cells 1 (NF-κB) after DNA damage and subsequently signals in a paracrine manner to activate the canonical Wnt program in tumor cells. The expression of WNT16B in the prostate tumor microenvironment attenuated the effects of cytotoxic chemotherapy in vivo, promoting tumor cell survival and disease progression. These results delineate a mechanism by which genotoxic therapies given in a cyclical manner can enhance subsequent treatment resistance through cell nonautonomous effects that are contributed by the tumor microenvironment.

Figure 1

Genotoxic damage to primary prostate fibroblasts induces expression of a spectrum of secreted proteins that includes WNT16B. (a) Schematic of the prostate cancer treatment regimen comprising a pretreatment prostate biopsy and four cycles of neoadjuvant DOC and MIT chemotherapy followed by radical prostatectomy. (b) DNA damage foci in human prostate tissues collected before and after chemotherapy. Tissue sections were probed with antibodies recognizing γ-H2AX (red and pink signals), and nuclei were counterstained with Hoechst 33342 (blue). Gl, gland lumen; e, epithelium; s, stroma. Scale bars, 50 µm. (c) Analysis of gene expression changes in prostate fibroblasts by transcript microarray quantification. The heatmap depicts the relative mRNA levels after exposure to H₂O₂, BLEO or RAD compared to vehicle-treated cells. Columns are replicate experiments. WNT16B is highlighted in bold for emphasis. (d) Measurements of WNT16B expression by qRT-PCR in prostate fibroblasts. Shown are the log2 transcript measurements before (Pre) or after exposure to the indicated factors relative to vehicle-treated control cells. Data are mean ± s.e.m. of triplicates. The P value was calculated by analysis of variance (ANOVA) followed by t test. (e) WNT16B protein expression in prostate PSC27 fibroblast extracellular conditioned medium (CM) or in cell lysates (IC) after genotoxic exposures. β-actin is a loading control. (f) Immunohistochemical analysis of WNT16B expression in prostate fibroblasts before (Pre) and after exposure to MIT or RAD. Brown chromogen indicates WNT16B expression. Scale bars, 50 µm. (g) Expression of Wnt family members in prostate fibroblasts after exposure to DNA-damaging agents. Transcript quantification was determined by microarray hybridization. Columns represent independent replicate experiments. WNT16B is listed in bold for emphasis. (h) WNT16B expression by qRT-PCR in PSC27 fibroblasts and prostate cancer cell lines after the indicated genotoxic exposure relative to pretreatment transcript amounts. Data are mean ± s.e.m.

Figure 2

Cytotoxic chemotherapy induces WNT16B expression in the tumor microenvironment. (a) Chemotherapy-induced gene expression changes in human prostate-cancer–associated stroma measured by qRT-PCR of microdissected cells. The amounts of transcript before treatment (x axis) are plotted against the amounts of transcript after chemotherapy (y axis) from the same individual. Each data point represents the measurements from an individual patient. The results are shown as PCR cycle number relative to ribosomal protein L13 (RPL13), which served as the reference control. The P values were calculated by Student’s t test. (b) IHC assessment of prostate stromal WNT16B expression in prostatectomy tissue samples from men with prostate cancer who were either untreated (n = 30) or treated with chemotherapy (n = 50). Patients were assigned to four categories based on their stromal WNT16B staining: 0, no expression; 1, faint or equivocal expression; 2, moderate expression; 3, intense reactivity. P < 0.0001 by ANOVA. (c) Representative example of intense WNT16B expression in prostate stroma after in vivo exposure to MIT and DOC. The black arrows denote areas of the stroma with fibroblasts and smooth muscle. Note the minimal WNT16B reactivity in the epithelium (gray arrows). Scale bars, 50 µm. (d) Kaplan-Meier plot of biochemical (prostate-specific antigen) relapse-free survival based on the expression of WNT16B in prostate stroma after exposure to MIT and DOC chemotherapy (P = 0.04 by log-rank test comparing WNT16B < 1 with WNT16B ≥ 2 survival distributions). DFI, disease-free interval from surgery. (e,f) WNT16B staining of breast (e) and ovarian (f) carcinoma from patients receiving neoadjuvant chemotherapy or no treatment before surgical resection. Staining is recorded on a 4-point scale: 0, no expression; 1, faint or equivocal expression; 2, moderate expression; 3, intense reactivity. Scale bars, 50 µm. The P values were calculated by ANOVA.

Figure 3

WNT16B is a major effector of the full DDSP and promotes the growth and invasion of prostate carcinoma. (a) Conditioned medium from WNT16B-expressing prostate fibroblasts (PSC27^WNT16B) promotes the proliferation of neoplastic prostate epithelial cells. shRNA^C, control shRNA; shRNA^WNT16B#1 and shRNA^WNT16B#2, WNT16B-specific shRNAs. (b) Scratch assay showing the enhanced motility of PC3 cells exposed to conditioned medium from prostate fibroblasts expressing a control vector (PSC27^C) or fibroblasts expressing WNT16B (PSC27^WNT16B). Scale bars, 100 µm. (c) The full fibroblast DDSP induced by radiation (PSC27-RAD) promotes the proliferation of tumorigenic prostate epithelial cells. The proliferative effect is significantly attenuated by the suppression of damage-induced expression of WNT16B (PSC27-RAD+shRNA^WNT16B). (d) The full paracrine-acting fibroblast DDSP induced by radiation (PSC27-RAD) promotes the invasion of neoplastic epithelial cells. Invasion is significantly attenuated by the suppression of damage-induced expression of WNT16B (PSC27-RAD+shRNA^WNT16B). Data in a, c and d are mean ± s.e.m. of triplicates, with P values calculated by ANOVA followed by t test. (e) Schematic of the xenograft cell recombination experiment to assess the ability of fibroblasts expressing WNT16B to influence prostate tumorigenesis in vivo. PFC, prostate fibroblast cells; PC, prostate cancer cells. (f) Prostate fibroblasts engineered to express WNT16B promote the growth of prostate carcinoma in vivo. Subrenal capsule grafts comprised of PC3 prostate epithelial cells alone, PC3 cells in combination with PSC27^C control fibroblasts or PC3 cells in combination with PSC27^WNT16B fibroblasts are shown. The green dashed lines denote the size of the tumor outgrowth from the kidney capsule. (g) Irradiated prostate fibroblasts (PSC27-RAD) promote the growth of prostate carcinoma cells in vivo, and this effect is significantly attenuated by the suppression of fibroblast WNT16B using WNT16B-specific shRNAs (PSC27-RAD+shRNA^WNT16B) (P < 0.05). Shown are tumor volumes 8 weeks after renal capsule implantation of PC3 and PSC27 cell grafts. In f and g, horizontal lines denote the mean of each group of eight tumors, and P values were determined by ANOVA followed by t test.

Figure 4

Genotoxic stress upregulates WNT16B through NF-κB and signals through the canonical Wnt–β-catenin pathway to promote tumor cell proliferation and the acquisition of mesenchymal characteristics. (a) Assay of canonical Wnt pathway signaling through activation of a TCF/LEF luciferase reporter construct (TOPflash) or a control reporter (FOPflash). Epithelial cells were exposed to conditioned medium (CM) from PSC27 prostate fibroblasts expressing WNT16B (PSC27^WNT16B) or control vector (PSC27^C). Data are mean ± s.e.m. of triplicates, and P values were determined by ANOVA followed by t test. (b) qRT-PCR assessment of the expression of β-catenin target genes in prostate cancer cell lines (BPH1, M12 and PC3) before and 72 h after exposure to PSC27^WNT16B-conditioned medium. Data represent the mean ± s.e.m. fold change after as compared to before exposure for three replicates. (c) Expression of β-catenin target genes in human prostate cancers in vivo after neoadjuvant treatment with MIT and DOC. Log2 transcript amounts in carcinoma cells after and before chemotherapy are shown in relation to low (blue) or high (red) WNT16B expression in the prostate stroma. Each data point represents an individual patient; n = 8 patients. Horizontal bars are group means. *P < 0.05, **P < 0.01 by ANOVA followed by t test. (d) The β-catenin pathway inhibitor XAV939 suppresses the proliferation of prostate cancer cells in response to PSC27^WNT16B CM and attenuates the response to the full DDSP in PSC27-RAD–conditioned medium. Cell numbers were determined 72 h after treatment. (e) Quantification of transcripts in neoplastic prostate epithelial cells encoding proteins associated with a phenotype of EMT. Measurements are from epithelial cells exposed to control media (DMEM), media conditioned by PSC27 fibroblasts (PSC27^C) or PSC27 fibroblasts expressing WNT16B (PSC27^WNT16B). E-cad, E-cadherin; N-cad, N-cadherin. (f) Analysis of EMT-associated protein expression by western blot. M12 or PC3 cells were exposed to control media, media conditioned by PSC27 fibroblasts (PSC27^C), PSC27 fibroblasts expressing WNT16B (PSC27^WNT16B) or PSC27 fibroblasts expressing WNT16B and shRNA targeting WNT16B (PSC27^shRNA-WNT16B). (g) Chromatin immunoprecipitation assays identified NF-κB binding sites within the proximal promoter of the WNT16 gene. PCR reaction products from Mock (no DNA loading), NF-κB immunoprecipitation, input control DNA and no antibody (Ab) control before treatment (Pre) and after irradiation (Rad). p1, p2 and p3 indicate primer pairs corresponding to putative NF-κB binding regions in WNT16, IL-6 and IL-8, respectively (see the Supplementary Methods for the primer sequences). (h) Analysis of WNT16B transcript expression by qRT-PCR in PSC27 prostate fibroblasts with (PSC27^IκBα) or without (PSC27^C) inhibition of NF-κB signaling before and after DNA-damaging exposures. (i) Inhibition of NF-κB signaling in fibroblasts responding to DNA damage attenuates the effect of the DDSP on tumor cell proliferation. Cell numbers were determined 72 h after RAD exposure to conditioned medium from fibroblasts with (PSC27^IκBα) or without (PSC27^C) inhibition of NF-κB signaling. Data in d, e, h and i are mean ± s.e.m. of triplicates, and P values were determined by ANOVA followed by t test.

Figure 5

Paracrine-acting WNT16B promotes the resistance of prostate carcinoma to cytotoxic chemotherapy. (a) Viability of prostate cancer cells 3 d after treatment with a half-maximal inhibitory concentration (IC₅₀) of MIT and medium conditioned by fibroblasts with (PSC27^WNT16B) or without (PSC27^C) WNT16B. (b) Bright field microscopic view of PC3 cells cultured with control or PSC27^WNT16B-conditioned medium photographed 24 h after exposure to vehicle or the IC₅₀ of MIT. Arrowheads denote apoptotic cell bodies. Scale bars, 50 µm. (c) Acute tumor cell responses to chemotherapy in vitro. Quantification of apoptosis by assays reflecting combined caspase 3 and 7 activity measured 24 h after the exposure of PC3 cells to vehicle or the IC₅₀ of MIT. Data in a and c are mean ± s.e.m. of triplicate experiments, and P values were determined by ANOVA followed by t test. RLU, relative luciferase unit. (d) In vivo responses of PC3 tumors to MIT chemotherapy. Grafts were comprised of PC3 cells alone or PC3 cells combined with either PSC27 prostate fibroblasts expressing a control vector (PC3+PSC27^C) or PSC27 prostate fibroblasts expressing WNT16B (PC3+PSC27^WNT16B). MIT was administered every 2 weeks for three cycles, and grafts were harvested and tumor volumes determined 1 week after the final MIT treatment. Each data point represents an individual xenograft. Horizontal lines are group means of ten tumors, with P values determined by ANOVA followed by t test. (e) Acute tumor cell responses to chemotherapy in vivo. Quantification of apoptosis by cleaved caspase 3 (C-caspase 3) IHC and of DNA damage by γ-H2AX immunofluorescence in PC3 and fibroblast xenografts measured 24 h after in vivo treatment with vehicle (C) or MIT. Values represent a minimum of 100 cells counted from each of 3–5 tumors per group. Data are mean ± s.e.m., and P values were determined by ANOVA followed by t test.

Figure 6

Chemotherapy resistance promoted by damaged fibroblasts is attenuated by blocking WNT16B, β-catenin or NF-κB signaling. (a) Viability of prostate cancer cells across a range of MIT concentrations with (PSC27-RAD+shRNA^WNT16B) or without (PSC27-RAD+shRNA^C) the suppression of WNT16B in irradiated-fibroblast–conditioned medium or with the addition of the β-catenin pathway inhibitor XAV939. Data are mean ± s.e.m. of triplicates. (b) Viability of prostate cancer cells 3 d after treatment with two times the IC₅₀ of MIT in the context of conditioned medium from irradiated prostate fibroblasts (PSC27-RAD) expressing shRNAs targeting and suppressing WNT16B (shRNA^WNT16B), a vector control (shRNA^C) or combined with the β-catenin pathway inhibitor XAV939. (c) Viability of prostate cancer cells 3 d after treatment with the IC₅₀ of MIT in the context of conditioned medium from prostate fibroblasts pretreated with radiation (PSC27-RAD) or MIT (PSC27-MIT) and with (PSC27^IκBα) or without (PSC27^C) the suppression of NF-κB signaling. (d) Acute tumor cell responses to chemotherapy in vitro. Quantification of apoptosis by caspase 3 and 7 activity measured 24 h after the exposure of PC3 cells to vehicle or the IC₅₀ of MIT. Data for b, c and d are mean ± s.e.m. of triplicates, and P values were determined by ANOVA followed by t test. (e,f) In vivo effects of MIT chemotherapy in the context of suppressing the induction of the expression of fibroblast WNT16B. Tumors comprised PC3 cells in combination with irradiated (PSC27-RAD) fibroblasts (e) or unirradiated (PSC27^C) (f) prostate fibroblasts expressing shRNAs targeting WNT16B (shRNA^WNT16B) or a vector control (shRNA^C). MIT was administered every 2 weeks for three cycles, and grafts were harvested and tumor volumes determined 1 week after the final treatment. Each data point represents an individual xenograft. Tumor volumes of PSC27^C+shRNA^C grafts in f averaged 20 mm³, and tumor volumes of PSC27^C+shRNA^WNT16B grafts averaged 12 mm³ (P < 0.001). Horizontal lines are group means, with n = 10 in e and n = 8 in f. P values were determined by ANOVA followed by t test. The bracket boundaries in f are the group means for PSC27^C+shRNA^C grafts compared to PSC27^C+shRNA^WNT16B grafts showing a 40% difference in size. Asterisks, as for the previous panel. (g) Model for cell nonautonomous therapy-resistance effects originating in the tumor microenvironment in response to genotoxic cancer therapeutics. The initial round of therapy engages an apoptotic or senescence response in subsets of tumor cells and activates a DNA damage response (DDR) in DDR-competent benign cells (+DDR) comprising the tumor microenvironment. The DDR includes a spectrum of autocrine- and paracrine-acting proteins that are capable of reinforcing a senescent phenotype in benign cells and promoting tumor repopulation through progrowth signaling pathways in neoplastic cells. Paracrine-acting secretory components such as WNT16B also promote resistance to subsequent cycles of cytotoxic therapy. CEC, cancer epithelial cell; BEC, benign epithelial cell; FC, fibroblast cell; –DDR, DDR-incompetent benign cells.