Tumor-derived CXCL8 signaling augments stroma-derived CCL2-promoted proliferation and CXCL12-mediated invasion of PTEN-deficient prostate cancer cells

Abstract

Impaired PTEN function is a genetic hallmark of aggressive prostate cancers (CaP) and is associated with increased CXCL8 expression and signaling. The current aim was to further characterize biological responses and mechanisms underpinning CXCL8-promoted progression of PTEN-depleted prostate cancer, focusing on characterizing the potential interplay between CXCL8 and other disease-promoting chemokines resident within the prostate tumor microenvironment. Autocrine CXCL8-stimulation (i) increased expression of CXCR1 and CXCR2 in PTEN-deficient CaP cells suggesting a self-potentiating signaling axis and (ii) induced expression of CXCR4 and CCR2 in PTEN-wild-type and PTEN-depleted CaP cells. In contrast, paracrine CXCL8 signaling induced expression and secretion of the chemokines CCL2 and CXCL12 from prostate stromal WPMY-1 fibroblasts and monocytic macrophage-like THP-1 cells. In vitro studies demonstrated functional co-operation of tumor-derived CXCL8 with stromal-derived chemokines. CXCL12-induced migration of PC3 cells and CCL2-induced proliferation of prostate cancer cells were dependent upon intrinsic CXCL8 signaling within the prostate cancer cells. For example, in co-culture experiments, CXCL12/CXCR4 signaling but not CCL2/CCR2 signaling supported fibroblast-mediated migration of PC3 cells while CXCL12/CXCR4 and CCL2/CCR2 signaling underpinned monocyte-enhanced migration of PC3 cells. Combined inhibition of both CXCL8 and CXCL12 signaling was more effective in inhibiting fibroblast-promoted cell motility while repression of CXCL8 attenuated CCL2-promoted proliferation of prostate cancer cells. We conclude that tumor-derived CXCL8 signaling from PTEN-deficient tumor cells increases the sensitivity and responsiveness of CaP cells to stromal chemokines by concurrently upregulating receptor expression in cancer cells and inducing stromal chemokine synthesis. Combined chemokine targeting may be required to inhibit their multi-faceted actions in promoting the invasion and proliferation of aggressive CaP.

Figure 1. Autocrine CXCL8 signaling increases chemokine receptor expression in prostate cancer cells

(A) Bar graph illustrating qPCR validation of CXCR1 (left panel) and CXCR2 (right panel) gene expression in multiple prostate cancer cell lines, subjected to stimulation with 3nM rh-CXCL8. (B) Immunoblots demonstrating increased protein levels of chemokine receptors in PC3 (left panel) and LNCaP cells (right panel) following exposure to 3nM rh-CXCL8. (C) Bar graph illustrating qPCR validation of CCR2 (left panel) and CXCR4 (right panel) gene expression in multiple prostate cancer cell lines, subjected to stimulation with 3nM rh-CXCL8. Data shown in (A) and (C) is the mean ± S.E.M value, determined from a minimum of 4 replicate experiments. Statically significant differences in expression were determined by performing a two-tailed Mann-Whitney U-test (*p<0.05; ** p<0.01; ***p<0.001).

Figure 2. Paracrine CXCL8 signaling induces CCL2 and CXCL12 synthesis and secretion in stromal cells

(A) Bar graph illustrating qPCR validation of CCL2 (left panel) and CXCL12 (right panel) gene expression in multiple prostate cancer and stromal cell lines. (B) Bar graph illustrating qPCR validation of CCL2 (left panel) and CXCL12 (right panel) gene expression in THP-1 macrophage-like cells, 293T and WPMY-1 stromal fibroblasts, subjected to stimulation with 3nM rh-CXCL8. (C) Bar graph illustrating the levels of CCL2 (left panel) and CXCL12 (right panel) secreted by stromal cells, in the absence and presence of stimulation with 3nM rh-CXCL8. Data shown represents the mean ± S.E.M. value determined by repetitive ELISAs. (D) Bar graph illustrating qPCR validation of CCR2 (left panel) and CXCR4 (right panel) gene expression in stromal THP-1 cells, 293T cells and WPMY-1 cells subjected to stimulation with 3nM rh-CXCL8. Data shown in (A), (B) and (D) is the mean ± S.E.M value, determined from a minimum of 4 replicate experiments. Statically significant differences in expression were determined by performing a two-tailed Mann-Whitney U-test (*p<0.05; ** p<0.01; ***p<0.001).

Figure 3. CXCL12 signaling potentiates the chemotactic migration of PC3 cells

(A) Representative images of wound scratch assays conducted using PC3 monolayers, subjected to treatment with relevant concentrations of CXCL8 and CXCL12, or treatment with the CXCR4 inhibitor AMD3100. Images shown depict the uniformity of the wound scratch at time of initiation (t=0) and the resulting closure of the wound after 8h stimulation. (B) Bar graph presenting the quantitation of wound closure of a PC3 monolayer resulting from various chemokine treatments. Data shown is the mean ± S.E.M. value of three independent experiments, each performed in triplicate. (C) Representative images of wound scratch assays conducted using PC3 monolayers, subjected to treatment with relevant concentrations of CXCL8 and CCL2. Images shown depict the uniformity of the wound scratch at time of initiation (t=0) and the resulting closure of the wound after 6 h stimulation. (D) Bar graphs illustrating the extent of wound closure of the PC3 monolayers promoted by stimulation with CXCL8 or CCL2, in isolation or in combination (left panel), and the impact of administering a CCR2 antagonist RS102895 upon CCL2-induced wound closure (right panel). Data shown is the mean ± S.E.M value, determined from a minimum of 3 replicate experiments. Statistically significant differences in expression were determined by performing a two-tailed Students t-test (*p<0.05; ** p<0.01; ***p<0.001).

Figure 4. Fibroblasts accelerate PC3 cell motility through a CXCR4-dependent mechanism

(A) Left; Representative images of wound scratch assays conducted using PC3 monolayers in the presence of WPMY-1 prostate stromal fibroblasts and the impact of administering the CXCR4 antagonist AMD3100. Right; Bar graphs presenting the quantification of wound closure observed under various treatments in wound scratch assays. Data shown is the mean ± S.E.M. value calculated from 4 independent experiments, each performed in duplicate/triplicate. Images shown depict the uniformity of the wound scratch at time of initiation (t=0) and the resulting closure of the wound after 6h stimulation. (B) Representative images of wound scratch assays conducted using PC3 monolayers in the presence of WPMY-1 prostate stromal fibroblasts and the impact of administering the x1/2pal-i3 pepducin to the co-culture and corresponding bar graphs presenting the quantification of wound closure. Data shown is the mean ± S.E.M. value calculated from 4 independent experiments, each performed in duplicate/triplicate. (C) Representative images of wound scratch assays conducted using PC3 monolayers in the presence of WPMY-1 prostate stromal fibroblasts and corresponding bar graphs presenting the quantification of wound closure. Data shown is the mean ± S.E.M. value calculated from 4 independent experiments, each performed in duplicate/triplicate. Images show the effect of administering the CXCR4 inhibitor AMD3100 in the absence and presence of the CXCR1/CXCR2-targeted x1/2pal-i3pepducin upon the WPMY-1 fibroblast accelerated wound closure of PC3 cells. Statistically significant differences in expression were determined by performing a two-tailed Students t-test (*p<0.05; ** p<0.01; ***p<0.001).

Figure 5. Macrophage-promoted acceleration of prostate cancer cell motility is sensitive to CXCR1/CXCR2 inhibition

(A) Representative images of wound scratch assays conducted using PC3 monolayers, examining the impact of THP-1 cells upon the acceleration of prostate cancer cell motility, in the absence and presence of the CXCR1/CXCR2-targeting x1/2i3-pal and the CXCR4 inhibitor AMD3100. Images shown depict the uniformity of the wound scratch at time of initiation (t=0) and the resulting closure of the wound after 6 h stimulation. (B) Bar graph presenting the quantification of wound closure effected by the addition of THP-1 cells to PC3 cell monolayers and the impact of administering the CXCR1/CXCR2-targeting x1/2pal-i3pepducin or the CXCR4 antagonist AMD3100 to the co-culture. Data shown is the mean ± S.E.M. value calculated from 4 independent experiments, each performed in duplicate/triplicate. (C) Bar graphs presenting the quantification of wound scratch assays examining the effect of the CCR2 inhibitor RS102895 on THP-1 (left panel) or WPMY-1 (right panel)-induced migration of PC3 cells. Data shown is the mean ± S.E.M. value calculated from 4 independent experiments, each performed in duplicate/triplicate. Statistically significant differences in expression were determined by performing a two-tailed Students t-test (*p<0.05; ** p<0.01; ***p<0.001).

Figure 6. Characterization of the effects of CCL2 and CXCL12 on the proliferation and viability of prostate cancer cells

(A) Bar graph presenting cell count data illustrating the effect of administering the CXCR1/CXCR2-targeting inhibitor x1/2pal-i3or the CCR2 antagonist RS102895, independently or in combination upon the proliferation of PTEN-deficient PC3 cells. (B) Flow cytometry data demonstrating the effect of administering RS102895 upon cell cycle progression (left panel) and cell viability (right panel). (C) Bar graph presenting cell count data illustrating the effect of the CXCR4 inhibitor AMD3100 on the proliferation of PC3 cells. (D) Left panel; Bar graph presenting the effect of CCL2 (100ng/mL) upon the viability of PC3 cells, transfection-control PC3 cells (PC3-NT) or low CXCL8-expressing PC3 cells, assessed by MTT assay. Right panel; bar graph illustrating the effect on cell proliferation affected by administering the CCR2 antagonist RS102895 to PC3-NT or low CXCL8-expressing PC3-120 cells in the absence and presence of CCL2 (100 ng/mL). Statistically significant differences in expression were determined by performing a two-tailed Students t-test or two-tailed Mann-Whitney U test (*p<0.05; ** p<0.01; ***p<0.001).

Figure 7. Schematic diagram representing chemokine crosstalk within the tumor microenvironment

1. Autocrine CXCL8 signaling results in the up-regulation of CXCR1, CXCR2, CXCR4 and CCL2 on the tumor cells. 2. Paracrine CXCL8 signaling leading to the up-regulation of CXCR1, CXCR2, CXCR4 and CCR2 expression by tumor-associated stromal cells. Paracrine CXCL8 signaling also induces secretion of CCL2 by macrophages and fibroblasts, as well as CXCL12 secretion by tumor-associated macrophages.