(A) The expression of mRNAs for 37 chemokines was studied by RT-PCR in xenografts (CWR22, WISH PC14) and prostate tumor cell lines (PC3, DU145, LNCaP). Systematic and non-systematic names for chemokines are shown. Amplicons were analyzed by agarose gel electrophoresis and visual inspection after staining with ethidium bromide. Sequences of primers are available on request. The levels of expression were evaluated as follows: no expression (−), low expression (−+), moderate expression (+), and high expression (++). Depending on the chemokine, the analysis was done from one to four times. (B) mRNA expression of chemokines in prostate cancer cell lines. The expression of mRNAs for 14 chemokines was studied by real-time RT-PCR on four prostate tumor cell lines (PC3, DU145, LNCaP, 22Rv1). The value of 2^-ΔCT was calculated and then normalized to the highest number for a given cell line, which was given the value of 100, where ΔCT = CT (chemokine)-CT (GAPDH). Numbers are averages of two determinations from one representative experiment of two performed. (C) Flow cytometric analysis of surface expression of CXCL16 on prostate cancer cell lines. PC3, 22Rv1 and DU145 cells were incubated without primary antibody (black shading), with goat anti-human CCL4 as an additional control (light gray line), or with goat anti-human CXCL16 (dark gray line) and then stained with rabbit anti-goat IgG-FITC. Data are from one representative experiment of three performed. (D) Staining for CXCL16 and CXCR6 in prostatectomy specimens using DAB (brown) for detection is shown for: normal prostate and prostate cancer. Panels are representative of over 80 prostate cancers cases examined. Each panel includes an H&E stain on the left. (E) Arrays containing prostate cancer tissue from 40 patients were stained for CXCL16 and CXCR6. (I) Samples were analyzed as described in , grouped according to scores for CXCL16 on the X-axis (n = samples in each CXCL16 group), and mean scores for CXCR6 expression were calculated. Error bars show SEM. Cross-bars indicate comparisons that are significant. *, p<0.05 and **, p<0.01. (II) Immunofluorescence and confocal imaging of a section showing prostate cancer cells displays anti-CXCL16 (488 tyramide) as green, anti-CXCR6 (594 tyramide) as red, and nuclear staining with Hoechst 33258 as blue. Double staining in yellow shows co-localization. Panels are representative of more than 15 cases of prostate cancer. Original magnifications are noted on these and all subsequent photomicrographs.

(A) Arrays containing prostate cancer tissue from 40 patients were stained for CXCL16 and CXCR6. Samples were scored for CXCL16 and CXCR6 expression as described in . Samples were grouped according to stage or Gleason score (X-axes) as determined by the supplier, and mean scores for CXCL16 and CXCR6 expression were calculated. The number of samples (n) in each stage or Gleason score group is shown. Error bars show SEM. Cross-bars indicate comparisons that are significant. *, p<0.05 and ***, p<0.001. (B) PC3 cells were transfected with the CXCR6-YFP plasmid, cultured for 48 hours without chemokine (control, ctrl) or with 0.05–5 µg/ml CXCL16 and numbers of CXCR6-YFP⁺ (R6YFP⁺) vs. CXCR6YFP⁻ (R6YFP⁻) cells were counted as described in . Values were normalized to the control, untreated sample to yield the fold difference in cell number. Bars show means±SEM combined from three experiments. **, p<0.01 and ***, p<0.001 vs. control values. (C) Following transfection with the CXCR6-YFP plasmid, PC3 cells were cultured as in (B), without or with 5 µg/ml CXCL16. Left panel, gating is shown for separating untreated and CXCL16-treated cells into “low” (L) or “high” (H) for CXCR6-YFP. Percentages of cells in the high gates are shown. One representative experiment is shown out of three performed. Right panel, CXCR6-YFP⁺low (CXCR6+low) and CXCR6-YFP⁺high (CXCR6+high) cells from CXCL16-treated and control cultures were counted after 48 hours as in (B). Data were combined from three experiments. *, p<0.05 and ***, p<0.001 vs. untreated cells. (D) Confocal imaging displays anti-CXCL16 or anti-CXCR6 (488 tyramide) as green, anti-Ki67 (594 tyramide) as red, and nuclear staining with Hoechst 33258 as blue. Panel is representative of more than 15 cases.

(A) Staining for CXCL16 and CXCR6 in prostatectomy specimens using DAB (brown) for detection is shown for: chronic prostatitis, PIA and PIN. Panels are representative of over 80 prostate cancer cases examined. Each panel includes an H&E stain on the left. (B) CXCL16 secreted by primary normal epithelial cells (PrEC) and stromal cells (PrSC), and the DU145 and PC3 cell lines was measured 48 hours after no treatment (control), or treatment with 5 ng/ml IFN-γ and 0.5 µg/ml TNF-α alone and in combination. Bars show means±SEM from duplicate wells, of one representative experiment of three performed for each of three donors (D1–D3). *, p<0.05 and **, p<0.01 vs. untreated cells. (C) CXCL16 mRNA expression in PrEC from three donors and two prostate tumor cell lines (DU145 and PC3) treated with 5 ng/ml IFN-γ and 0.5 µg/ml TNF-α alone and in combination were analyzed using real-time RT-PCR. Values were normalized by setting the control sample with lowest expression equal to 1. Each bar represents the mean±SEM obtained from duplicate wells, from one representative experiment of three performed. **, p<0.01 vs. untreated cells. (D) CXCR6 mRNA expression in PrEC from three donors treated with 5 ng/ml IFN-γ and 0.5 µg/ml TNF-α alone and in combination were analyzed using real-time RT-PCR. Values were normalized by setting the sample with lowest expression equal to 1. **, p<0.01 vs. untreated cells. (E) Each row contains serial sections from the same regions of prostatectomy specimens stained using DAB (brown) for αSMA, CXCL16, and CXCR6 including an H&E stain on the left. Panels are representative of over 120 prostate specimens. (F) Separate slides of arrays containing prostate tissue from 40 patients were stained for αSMA, CXCL16 and CXCR6 and scored as described in for αSMA in stroma and CXCL16 and CXCR6 in cancer cells. Samples were grouped according to scores for CXCL16 or CXCR6 (n = samples in each group), and mean scores were calculated for αSMA in stroma. Data are displayed and analyzed as in .I. *, p<0.05.

(A, B) Confocal imaging shows anti-CXCL16 (A, left panel) or anti-CXCR6 (A, right panel) or anti-Ki67 (B) as green (488 tyramide), anti-CD3 (594 tyramide) as red, and nuclear staining with Hoechst 33258 as blue. Double staining in yellow shows co-localization. In (A), arrows point to cancer cells and T cells. Panels are representative of more than 15 cases for each double-stain. (C) Staining for Ki67 and CXCR6 in a region of prostatitis using DAB (brown, two left panels) or immunofluorescence (right panel) for detection. Confocal imaging shows anti-CXCR6 as green (488 tyramide) and anti-Ki67 as red (594 tyramide), and nuclear staining with Hoechst 33258 as blue. Panel is representative of more than 15 cases.

(A) CD4⁺ T cells were sorted into naïve (CD45RO-), total effector/memory (CD45RO+), CD45RO⁺CXCR6⁻ (CXCR6-) and CD45RO⁺CXCR6⁺ (CXCR6+) subsets and stimulated with PMA and ionomycin for 4 hours before measuring levels of CXCL16 mRNA by real-time RT-PCR. Values were normalized by setting the sample with lowest expression equal to 1. The data are means from duplicate samples from one representative experiment out of three performed. (B) Jurkat E6.1 T cells transfected with the CXCR6-YFP plasmid were analyzed for migration to CXCL16. Some samples were pre-incubated for two hours with 400 ng/ml pertussis toxin as indicated. Numbers of YFP+ and YFP- migrating cells were analyzed by flow cytometry. Each bar represents the mean±SEM obtained from triplicate wells, from one representative experiment of three performed. ***, p<0.001 vs. the corresponding control value. (C) Following transfection with the CXCR6-YFP plasmid, Jurkat E6.1 T cells were cultured for 48 hours without chemokine (control, ctrl) or with 1 µg/ml CXCL8 as an additional control, or with 0.05–5 µg/ml CXCL16 before counting numbers of CXCR6-YFP⁺ vs. CXCR6-YFP⁻ cells. Bars show means±SEM combined from three different experiments. *, p<0.05 and ***, p<0.001. (D) Transfected cells were treated as in (C) and stained with Annexin V and propidium iodide (PI) before counting numbers of CXCR6-YFP⁺ cells within the subgroups as noted. Bars show means±SEM combined from three experiments. **, p<0.01 and ***, p<0.001 vs. control untreated cells. (E) Following transfections with CXCR6-YFP plasmid, Jurkat E6.1 T cells were loaded with 2 µM Far Red DDAO dye and cultured without chemokine (control, ctrl) or with 0.05–5 µg/ml CXCL16 or 1 µg/ml CXCL8. Numbers of CXCR6YFP⁺ vs. CXCR6YFP⁺DDAO⁻ (proliferated) and CXCR6YFP⁺DDAO⁺ (non-proliferated) cells were determined after 48 hours using counting beads and flow cytometry. Bars show means±SEM from triplicate wells, from one representative experiment of three performed. ***, p<0.001 and *, p<0.05 vs. control untreated cells.

(A) CD4⁺ T cells purified from elutriated lymphocytes from healthy donors were stimulated for 3 days with plate-bound anti-CD3 (OKT3, 10 µg/ml) with or without 5 µg/ml plate-bound CXCL16 (bL16). Treatments of anti-CD3-activated cells with PTX, anti-CXCR6 (aR6) and anti-CXCL16 (aL16) without bL16 were done as controls. Mouse IgG (mIgG) and rat IgG (rIgG) were used as controls for anti-CXCR6 antibody and anti-CXCL16 antibody, respectively. Bars show means±SEM from one representative experiment out of five, using five donors. Anti-CXCL16 was used in a total of four, and anti-CXCR6 in two experiments. ns, not significant and ***, p<0.001. (B) CXCR6 and CXCL16 mRNAs were measured by real-time RT-PCR in CD3-activated and non-activated CD4⁺ T cells after 3 days. Values were normalized by setting the non-activated sample with lowest expression equal to 1. Bars show means±SEM combined from duplicate wells from seven different donors. **, p<0.01 and ***, p<0.001 vs. non-activated cells. (C) Anti-CD3-activated or nonactivated CD4⁺ T cells were analyzed for migration to CXCL16, expressed as percentages of input cells migrating. Each bar represents the mean±SEM obtained from triplicate wells from a total of three experiments performed. ***, p<0.001 on cross bars are indicated for comparisons between activated cells - 0 vs. 50 ng/ml CXCL16 and 50 ng/ml vs. 500 ng/ml CXCL16.

(A) * The intensity of CXCL16 staining was scored as described in . The % of cases staining high is shown for each cancer type. ** RCC, renal cell carcinoma; GBM, glioblastoma multiforme. (B) CXCL16 expression in arrays of multiple human cancers by immunohistochemistry using AEC (red) or DAB (brown) for detection. Panels are representative of the numbers of specimens of the various cancers as listed in (A).

Secretion of CXCL16 from pre-cancer or cancer cells leads to CXCR6-mediated recruitment of leukocytes. Cytokines such as TNF-α and IFN-γ produced by these cells in turn induce expression of CXCL16 and CXCR6, creating a positive feedback loop that enhances tumorigenesis – directly, through effects on the growth of pre-cancer and cancer cells, and indirectly, by increasing the migration and proliferation of the inflammatory cells, together leading to the increased growth of inflammation-associated cancers.