CXCR4 is a novel target of cancer chemopreventative isothiocyanates in prostate cancer cells

Abstract

Isothiocyanates (ITCs) derived from cruciferous vegetables, including phenethyl isothiocyanate (PEITC) and sulforaphane (SFN), exhibit in vivo activity against prostate cancer in a xenograft and transgenic mouse model, and thus are appealing for chemoprevention of this disease. Watercress constituent PEITC and SFN-rich broccoli sprout extract are under clinical investigations but the molecular mechanisms underlying their cancer chemopreventive effects are not fully understood. The present study demonstrates that chemokine receptor CXCR4 is a novel target of ITCs in prostate cancer cells. Exposure of prostate cancer cells (LNCaP, 22Rv1, C4-2, and PC-3) to pharmacologically applicable concentrations of PEITC, benzyl isothiocyanate (BITC), and SFN (2.5 and 5 μmol/L) resulted in downregulation of CXCR4 expression. None of the ITCs affected secretion of CXCR4 ligand (stromal-derived factor-1). In vivo inhibition of PC-3 xenograft growth upon PEITC treatment was associated with a significant decrease in CXCR4 protein level. A similar trend was discernible in the tumors from SFN-treated TRAMP mice compared with those of control mice, but the difference was not significant. Stable overexpression of CXCR4 in PC-3 cells conferred significant protection against wound healing, cell migration, and cell viability inhibition by ITCs. Inhibition of cell migration resulting from PEITC and BITC exposure was significantly augmented by RNAi of CXCR4. This study demonstrates, for the first time, that cancer chemopreventive ITCs suppress CXCR4 expression in prostate cancer cells in vitro as well as in vivo. These results suggest that CXCR4 downregulation may be an important pharmacodynamic biomarker of cancer chemopreventative ITCs in prostate adenocarcinoma.

Figure 1

ITCs downregulated CXCR4 protein level in prostate cancer cells. A, structures of PEITC, BITC, and SFN. B, western blots showing effect of ITC treatment on CXCR4 protein level in LNCaP cells, 22Rv1, and C4-2 cells. GAPDH was probed as a loading control. C, immunofluorescence microscopy for effect of ITC treatment (5 µmol/L, 24 hour) on CXCR4 protein level in LNCaP cells. Western blotting was performed 2-4 times using independently prepared lysates. Data on effect of SFN in 22Rv1 cell was inconsistent.

Figure 2

Effect of ITCs on CXCR4 protein and mRNA expression in PC-3 cells. A, effect of ITC treatment on protein levels of CXCR4 by western blotting. Samples after 6 hour treatment were also used for western blotting for PEITC, but these data are not shown because of inconsistency. B, immunofluorescence microscopy for effect of ITCs (5 µmol/L) on CXCR4 protein level in PC-3 cells after 24 hour treatment with DMSO or specified ITC compound. C, RT-PCR for analysis of CXCR4 mRNA after treatment with DMSO or specified ITC compound D, quantitation of CXCL12 secretion in culture media of PC-3 cells after 12 or 24 hour treatment with DMSO or ITCs (PEITC or BITC). The results shown (mean ± SD) for PEITC are combined from two independent experiments (n = 5). Analysis of CXCL12 secretion in culture media of BITC-treated cells was done once (n = 3). ^aSignificantly different compared with control by one-way ANOVA with Dunnett’s adjustment.

Figure 3

PEITC administration downregulated CXCR4 protein expression in vivo in PC-3 xenografts. A, western blotting for CXCR4 expression in tumor lysates from PC-3 xenografts from control and PEITC-treated mice. B, densitometric quantitation of CXCR4 protein in PC-3 tumors; control (n = 7) and PEITC (n = 8). C, western blotting for CXCR4 protein using tumor lysates from control and SFN-treated TRAMP mice. D, densitometric quantitation of CXCR4 protein in TRAMP tumors (n = 3). The results shown are mean ± SD. Statistical significance was determined by unpaired Student’s t-test.

Figure 4

Stable overexpression of CXCR4 conferred protection against wound healing inhibition by PEITC and BITC. A, western blots show overexpression (left panel) or knockdown (right panel) of CXCR4 protein in PC-3 cells. B, results of scratch assay showing the effect of PEITC and BITC treatments (10 hour treatment) on wound healing in PC-3 cells transfected with CXCR4 plasmid or empty vector. C, quantitation of wound healing. The results shown (mean ± SD) are representative of two independent experiments (n = 3). Significantly different compared with ^acorresponding control, and ^bbetween Neo_PC-3 and CXCR4_PC-3 cells by one-way ANOVA followed by Bonferroni’s test. D, bar graphs showing effect of CXCR4 knockdown using a siRNA on wound healing inhibition by PEITC or BITC. The results shown are mean ± SD (n = 2). Significantly different compared with ^acorresponding control, and ^bbetween control siRNA and CXCR4 siRNA by one-way ANOVA followed by Bonferroni’s test.

Figure 5

Cell migration inhibition by PEITC and BITC was augmented by CXCR4 knockdown in PC-3 cells. A, representative images showing the effect of PEITC treatment (24 hours) on PC-3 cell migration. A decrease in cell migration by PEITC treatment as well as CXCR4 knockdown was clearly visible. B, bar graphs show quantitation of cell migration by PC-3 cells transfected with control siRNA or CXCR4-specific siRNA after 24 hour treatment with DMSO or the specified ITC compound. C, microscopic images depicting migration by Neo_PC-3 and CXCR4_PC-3 cells after 24 hour treatment with DMSO or PEITC. D, bar graphs show quantitation of cell migration. The results shown are mean ± SD (n = 2-3). Significantly different compared with ^acorresponding control, and ^bbetween Neo_PC-3 and CXCR4_PC-3 by one-way ANOVA followed by Bonferroni’s test. Each experiment was repeated for 2-3 times and representative data from one such experiment are shown.

Figure 6

Cell viability inhibition by ITCs was significantly attenuated by CXCR4 overexpression in PC-3 cells. A, western blot showing effect of PEITC treatment (24 hours) on CXCR4 protein level in Neo_PC-3 and CXCR4_PC-3 cells. B, Effects of PEITC, BITC, and SFN treatments (24 hours) on viability of Neo_PC-3 and CXCR4_PC-3 cells. The results shown are mean ± SD (n = 3). Significantly different compared with ^acorresponding control, and ^bbetween Neo_PC-3 and CXCR4_PC-3 cells by one-way ANOVA followed by Bonferroni’s test. C, western blotting for S473 phosphorylated AKT using lysates from Neo_PC-3 or CXCR4_PC-4 after 24 hour treatment with DMSO or specified ITC compound. D, western blotting for phospho-ERK and total ERK using lysates from Neo_PC-3 or CXCR4_PC-4 after 4 hour treatment with DMSO or specified ITC compound. The blots were stripped and re-probed with GAPDH to ensure equal protein loading. Each experiment was repeated at least twice.