Sulindac selectively inhibits colon tumor cell growth by activating the cGMP/PKG pathway to suppress Wnt/β-catenin signaling

Abstract

Nonsteroidal anti-inflammatory drugs (NSAID) display promising antineoplastic activity for colorectal and other cancers, but toxicity from COX inhibition limits their long-term use for chemoprevention. Previous studies have concluded that the basis for their tumor cell growth inhibitory activity does not require COX inhibition, although the underlying mechanism is poorly understood. Here, we report that the NSAID sulindac sulfide inhibits cyclic guanosine 3',5'-monophosphate phosphodiesterase (cGMP PDE) activity to increase intracellular cGMP levels and activate cGMP-dependent protein kinase (PKG) at concentrations that inhibit proliferation and induce apoptosis of colon tumor cells. Sulindac sulfide did not activate the cGMP/PKG pathway, nor affect proliferation or apoptosis in normal colonocytes. Knockdown of the cGMP-specific PDE5 isozyme by siRNA and PDE5-specific inhibitors tadalafil and sildenafil also selectively inhibited the growth of colon tumor cells that expressed high levels of PDE5 compared with colonocytes. The mechanism by which sulindac sulfide and the cGMP/PKG pathway inhibits colon tumor cell growth involves the transcriptional suppression of β-catenin to inhibit Wnt/β-catenin T-cell factor transcriptional activity, leading to downregulation of cyclin D1 and survivin. These observations suggest that safer and more efficacious sulindac derivatives can be developed for colorectal cancer chemoprevention by targeting PDE5 and possibly other cGMP-degrading isozymes.

Figure 1

Colon tumor cell growth and cGMP PDE inhibitory activity of SS. A, chemical structure of sulindac sulfide. B, dose-dependent decrease of cell viability following 72 hours of SS treatment. C, apoptosis induction of HCT116 cells by SS after 6 hours of treatment. D, inhibition of proliferation by SS treatment after 24 hours in NCM460 colonocytes and HCT116 colon tumor cells. E and F, dose-dependent inhibition of cGMP (E) and cAMP (F) PDE activity in whole cell lysates from NCM460 colonocytes and HCT116, and HT29 colon tumor cells.

Figure 2

SS activation of cGMP signaling in HCT116 human colon tumor cells. A and B, dose-dependent increase in intracellular cGMP levels after 45 min of SS treatment in HCT116 colon tumor cells (A), but not NCM460 colonocytes (B). C, time-dependent (left) and dose-dependent (right) increase in VASP phosphorylation at Ser²³⁹ after SS treatment in HCT116 cells.

Figure 3

Sensitivity to PDE5 inhibitors and expression of PDE5 in colon tumor cells and colonocytes. A and B, inhibitory effect of sildenafil and tadalafil on viability of HCT116 colon tumor cells (A), but not NCM460 colonocytes (B) after 72h treatment. C, expression of PDE5 in normal colonocytes and colon tumor cell lines.

Figure 4

Suppression of PDE5 with siRNA is sufficient to selectively inhibit colon tumor cell growth. A, siRNA knockdown of PDE5 protein expression in HCT116 and HT29 colon tumor cells and NCM460 colonocytes (Western Blotting). B, selective suppression of cGMP hydrolysis by PDE5 siRNA in HCT116 and HT29 colon tumor cells, but not NCM460 colonocytes. C, no effect of PDE5 siRNA on cAMP hydrolysis in HCT116, HT29 and NCM460 cells. D, selective inhibition of tumor cell viability by PDE5 siRNA. E, tumor cell specific induction of apoptosis by PDE5 siRNA. F, selective inhibition of tumor cell proliferation by PDE5 siRNA. All treatment effects (A–F) were measured 72 hours post transfection.

Figure 5

Activation of cGMP signaling correlates with β-catenin signaling. A, time-dependent decrease in the expression of β-catenin, cyclin D1 and survivin with 80 μmol/L SS treatment in HCT116 cells. B, time-dependent decrease in the expression of β-catenin, cyclin D1 and survivin with 1 mmol/L 8-bromo-cGMP in HCT116 cells. C, suppression of β-catenin, cyclin D1 and survivin with PDE5 siRNA after 72 hours transfection in HCT116 cells. The numbers indicated below the blots derive from band quantification normalized to GAPDH and are expressed as ratio of scrambled control siRNA treatment. D, inhibition of TCF transcriptional activity after 72 hours transfection of PDE5 siRNA in HCT116 and HT29 colon tumor cells. Data obtained were normalized for β-galactosidase activity, and normalized FOP-FLASH values were then subtracted from the normalized TOP-FLASH values.

Figure 6

Downregulation of β-catenin by PKG activation is associated with inhibition of β-catenin transcription. A, dose-dependent decrease in β-catenin mRNA levels after 6 hours of treatment with SS in HCT116 cells as measured by semi-quantitative RT-PCR. The band intensities were quantified by ImageJ software and the β-catenin mRNA values were normalized against GAPDH levels. Fold change values under each lane indicate the normalized β-catenin intensities as compared to vehicle control. B, dose-dependent inhibition of CTNNB1 transcriptional activity after 6 hours of SS treatment in HCT116 cells. C, suppression of CTNNB1 transcriptional activity by PDE5 siRNA after 72 hours transfection in HCT116 and HT29 colon tumor cells. D, proposed model showing the COX-independent anticancer mechanism of SS involving activation of cGMP/PKG pathway and suppression of β-catenin signaling in colon cancer cells. Inhibition of PDE5 by SS/PDE5 inhibitors results in accumulation of cGMP, activation of PKG, and suppression of β-catenin mediated gene transcription.