Protease activated receptor-1 inhibits the Maspin tumor-suppressor gene to determine the melanoma metastatic phenotype

Abstract

The thrombin receptor protease activated receptor-1 (PAR-1) is overexpressed in metastatic melanoma cell lines and tumor specimens. Previously, we demonstrated a significant reduction in tumor growth and experimental lung metastasis after PAR-1 silencing via systemic delivery of siRNA encapsulated into nanoliposomes. Gene expression profiling identified a 40-fold increase in expression of Maspin in PAR-1-silenced metastatic melanoma cell lines. Maspin promoter activity was significantly increased after PAR-1 silencing, suggesting that PAR1 negatively regulates Maspin at the transcriptional level. ChIP analyses revealed that PAR-1 decreases binding of Ets-1 and c-Jun transcription factors to the Maspin promoter, both known to activate Maspin transcription. PAR-1 silencing did not affect Ets-1 or c-Jun expression; rather it resulted in increased expression of the chromatin remodeling complex CBP/p300, as well as decreased activity of the CBP/p300 inhibitor p38, resulting in increased binding of Ets-1 and c-Jun to the Maspin promoter and higher Maspin expression. Functionally, Maspin expression reduced the invasive capability of melanoma cells after PAR-1 silencing, which was abrogated after rescuing with PAR-1. Furthermore, tumor growth and experimental lung metastasis was significantly decreased after expressing Maspin in a metastatic melanoma cell line. Moreover, silencing Maspin in PAR-1-silenced cells reverted the inhibition of tumor growth and experimental lung metastasis. Herein, we demonstrate a mechanism by which PAR-1 negatively regulates the expression of the Maspin tumor-suppressor gene in the acquisition of the metastatic melanoma phenotype, thus attributing an alternative function to PAR-1 other than coagulation.

Fig. 1.

Maspin expression after PAR-1 silencing in A375SM and C8161 melanoma cell lines. FACS analyses reveal a significant decrease in PAR-1 expression in (A) A375SM and (B) C8161 cells after transduction with PAR-1 shRNA. Mouse IgG was used as an isotype control. As a negative control, secondary phycoerythrin (PE) ab was used without PAR-1 ab. Bar colors in histograms correspond to data from the representative FACS analysis image (Insets). PE intensity indicates PAR-1 expression. (C) Western blot analysis of PAR-1–silenced cells depicts a significant increase in Maspin expression compared with NT-transduced cells. For all panels, data were obtained in triplicate and are expressed as mean values ± SD (*P < 0.001).

Fig. 2.

Differential binding of Ets-1 and c-Jun transcription factors to the Maspin promoter. (A) Illustration of two AP-1 transcription factor binding sites (−510 and 54 from TIS) and two Ets-1 binding sites (−476 and −115 from TIS) within the first 550 bp of the Maspin promoter. (B) ChIP studies depict increased binding of Ets-1 and (C) c-Jun to the promoter of Maspin in both PAR-1–silenced cell lines compared with NT-transduced cells.

Fig. 3.

PAR-1 regulates Maspin expression by modulating binding of Ets-1 and c-Jun transcription factors via p38 and CBP/p300. (A) Phospho-p38 expression was significantly decreased after silencing PAR-1 in metastatic melanoma cells, whereas total levels of p38 remain unaffected. (B) Phospho-CBP/p300 was increased in PAR-1–silenced cells compared with NT-transduced cells. A decrease in phospho-p38, using 10 μM of the p38 inhibitor (SB203580), results in a significant increase in CBP/p300 expression in NT shRNA-transduced cells (i.e., low CBP/p300 expressors). (C) Silencing CBP/p300 with siRNA in PAR-1–silenced cells, results in decreased Maspin expression similar to levels from NT shRNA-transduced cells. (D) Inhibition of p38 results in increased Maspin expression in NT shRNA-transduced cells. Data are presented as means ± SD from three independent experiments. (E) ChIP shows increased binding of Ets-1 and c-Jun to the Maspin promoter in PAR-1–silenced cells. Silencing CBP/p300 with siRNA in PAR-1–silenced cells decreased binding of Ets-1 and c-Jun to the Maspin promoter.

Fig. 4.

Effects of altering PAR-1 levels on Maspin expression and invasion. (A) Invasion assays demonstrate a significant decrease in the number of invasive PAR-1–silenced cells. Data are presented as means ± SD (*P < 0.001). (B) Western blots revealed that Maspin is significantly reduced in PAR-1–rescued cells similar to Maspin protein levels from NT-transduced cells. (C) Rescuing PAR-1 significantly increased cell invasiveness compared with PAR-1–silenced cells transduced with EV (*P < 0.001). Data are presented as means ± SD. (D) Zymography assay depicts decreased MMP-2 activity in PAR-1–silenced cells. Rescuing PAR-1 results in increased MMP-2 activity compared with PAR-1-silenced cells transduced with EV. FBS 1% is used as a positive control. Serum-free medium served as a negative control.

Fig. 5.

In vivo effects of Maspin in melanoma. Maspin was stably transduced into A375SM cells (i.e., low Maspin expression) and injected s.c. and i.v. into nude mice. Significant decreases in (A) tumor growth and (B) experimental lung metastasis were seen in mice injected with A375SM expressing Maspin. (C) Silencing Maspin in PAR-1–silenced cells significantly increased tumor growth similar to levels from NT-transduced cells. Data are presented as means ± SEM. (D) Silencing Maspin in PAR-1–silenced cells significantly increased lung colony formation compared with PAR1-silenced cells alone. (*P < 0.01; n = 10 per group.)