Sildenafil Suppresses Inflammation-Driven Colorectal Cancer in Mice

Abstract

Intestinal cyclic guanosine monophosphate (cGMP) signaling regulates epithelial homeostasis and has been implicated in the suppression of colitis and colon cancer. In this study, we investigated the cGMP-elevating ability of the phosphodiesterase-5 (PDE5) inhibitor sildenafil to prevent disease in the azoxymethane/dextran sulfate sodium (AOM/DSS) inflammation-driven colorectal cancer model. Treatment of mice with sildenafil activated cGMP signaling in the colon mucosa and protected against dextran-sulfate sodium (DSS)-induced barrier dysfunction. In mice treated with AOM/DSS, oral administration of sildenafil throughout the disease course reduced polyp multiplicity by 50% compared with untreated controls. Polyps that did form in sildenafil treated mice were less proliferative and more differentiated compared with polyps from untreated mice, but apoptosis was unaffected. Polyps in sildenafil treated mice were also less inflamed; they exhibited reduced myeloid-cell infiltration and reduced expression of iNOS, IFNγ, and IL6 compared with untreated controls. Most of the protection conferred by sildenafil was during the initiation stage of carcinogenesis (38% reduction in multiplicity). Administration of sildenafil during the later promotion stages did not affect multiplicity but had a similar effect on the polyp phenotype, including increased mucus production, and reduced proliferation and inflammation. In summary, the results demonstrate that oral administration of sildenafil suppresses polyp formation and inflammation in mice treated with AOM/DSS. This validation of PDE5 as a target highlights the potential therapeutic value of PDE5 inhibitors for the prevention of colitis-driven colon cancer in humans. Cancer Prev Res; 10(7); 377-88. ©2017 AACRSee related editorial by Piazza, p. 373.

Figure 1. Sildenafil protects the intestinal epithelium from DSS-induced damage

A, Immunoblot shows phosphorylation of the PKG substrate vasodilator stimulated phosphoprotein (VASP-P) in colon mucosa from control and sildenafil treated groups. β-actin is a loading control. Right panel, quantitation of densitometric data from blots shown in the left panel. B, Mucus staining (AB/PAS) of colon sections from untreated (Ctrl) and sildenafil treated mice (Sild) for 7 days. Histogram showing percentage of total AB/PAS-positive staining area. C, Ki67 staining for proliferation in colons from Ctrl and 7 day Sild-treated mice. Histogram shows percent Ki67-positive staining cells per crypt. D, Barrier permeability assessed with FITC-dextran in mice treated with 3% DSS for 5 days with or without sildenafil treatment. E, Disease activity index (DAI) in mice treated with 3% DSS for 5 days with or without sildenafil treatment. A-E, n=3 mice per group. Scale bars in panels B,C are 50 μm.

Figure 2. Sildenafil suppresses polyp formation in the AOM/DSS model of colon cancer

A, Experimental scheme for the AOM/DSS-induced colon carcinogenesis model. C57Bl/6J mice provided water (Control) or water containing sildenafil (5.7mg/kg/day). B, Change in body weight from the initial weight. AOM was administered on day 0, and grey panels depict DSS treatment cycles. C, Representative colons from AOM/DSS treated mice, top-bottom orientation is proximal-rectal and representative polyps are indicated in the first panels by arrows. D, Plot shows the effect of sildenafil on median polyp number per animal (n=14 per group). Asterisks show difference between weight loss between treated and control groups at each day, p < 0.0001; Panel D, p-value generated with Mann–Whitney test.

Figure 3. Polyps from sildenafil treated mice are less inflamed

A, Representative image of H&E stained colon section (left panel) and polyp (right panel). Scale bar is 500 μm in the left panel, and 50 μm in the right panel. B, Histogram shows average polyp number in different size classes (small, medium, and large). C, Average polyp size in untreated (Ctrl) and treated (Sild) mice. D, Quantification of CD11b⁺Gr1⁺ cells in colorectums of untreated and sildenafil treated mice. E, Real-time qPCR analysis of inflammatory cytokine gene expression in polyps derived from AOM/DSS treated mice. For panels B & C, n=14 mice group; D, n=10 mice per group, E, n=8 polyps per group.

Figure 4. Polyps from sildenafil treated mice are less proliferative and more differentiated

A, IHC staining of Ki67 (left panel) and quantitation (right panel) in polyps of untreated (Ctrl) and treated (Sild) mice. B, IHC staining of cleaved caspase 3 (CC3) (left panel) and quantitation (right panel) in polyps of untreated (Ctrl) and treated (Sild) mice. C, AB/PAS staining of mucus (left panel) and quantitation (right panel) in polyps of untreated (Ctrl) and treated (Sild) mice. In all panels, the scale bar is 100 μm and n=10 mice per group.

Figure 5. Expression of cGMP signaling components in polyps

A, Expression of cGMP generators in colon mucosa and polyps measured by real-time qPCR. B, Expression of cGMP effectors in colon mucosa and polyps assayed by real-time CPCR. C, Immunoblot analysis of cGMP effectors in colon polyps of untreated (Control) and treated (Sildenafil) mice, β-actin is a loading control. For panels A–B, n=8 polyps per group, and panel C, n=5 per group. Asterisks, *p < 0.05, ***p < 0.001.

Figure 6. Sildenafil primarily blocks early carcinogenesis in the AOM/DSS model

A, Experimental scheme for the AOM/DSS-induced colon carcinogenesis model. Horizontal arrows show control (water only) and early or late administration of sildenafil. B, Plot shows the effect of sildenafil added early on median polyp number per animal. C, Plot shows the effect of sildenafil added later on median polyp number per animal. D, Histogram quantification of proliferative index in polyps of untreated (Ctrl) and late-treated (Sild) mice. E, Histogram quantitation of polyp mucus density. F, Expression of inflammatory cytokine genes in colon tissues assayed by real-time qPCR. For panels B–C, n ≥ 8 mice group and p-values for panel B–C generated with Mann–Whitney test. For Panels D–E, n=11 polyps per group. For panel F, n=8 polyps per group.