Volume 138, Issue 2, Pages 595-605.e3 (February 2010) Mesalamine Inhibits Epithelial β-Catenin Activation in Chronic Ulcerative Colitis  Jeffrey B. Brown, Goo Lee, Elizabeth Managlia, Gery R. Grimm, Ramanarao Dirisina, Tatiana Goretsky, Paul Cheresh, Nichole R. Blatner, Khashayarsha Khazaie, Guang–Yu Yang, Linheng Li, Terrence A. Barrett  Gastroenterology  Volume 138, Issue 2, Pages 595-605.e3 (February 2010) DOI: 10.1053/j.gastro.2009.10.038 Copyright © 2010 AGA Institute Terms and Conditions

Figure 1 Mesalamine (5-ASA) treatment reduces P-β-catenin signaling in human UC. Human biopsy samples from control and UC patients obtained during colonoscopy were processed for assessment of histologic inflammation scores and immunohistochemical analysis with P-β-catenin antibody. Data are presented as means ± SEM; *P < .01 (A) Representative sigmoid colon tissue stained for nuclear P-β-catenin (arrows) from control, chronic inflamed mucosa obtained from adjacent untreated, mild and severely inflamed UC (top panels), and mesalamine-treated (≥2.4 g/day) mild and severely inflamed UC (see Materials and Methods section; original magnification, 20×). (B) Enumeration of P-β-catenin+ cells/40× field vs inflammatory scores and (C) mean P-β-catenin+ cells/40× field with or without mesalamine treatment. (D) Histologic scores demonstrate comparable levels of inflammation between treatment groups segregated into mild or severe inflammation based on tissue histology (see Materials and Methods section). (E) Mean number of P-β-catenin+ cells/hpf in mild and severely inflamed UC biopsy specimens ± mesalamine treatment. (F) Increased β-catenin target gene mRNA expression in untreated colitis is attenuated by mesalamine, as measured by real-time RT-PCR. Results are shown as fold induction in relation to mRNA expression of Gapdh. Gastroenterology 2010 138, 595-605.e3DOI: (10.1053/j.gastro.2009.10.038) Copyright © 2010 AGA Institute Terms and Conditions

Figure 2 BrdU incorporation, P-Akt, and P-β-catenin levels increase as IL-10−/− colitis progresses to dysplasia. (A–C and E–M) Sections from baseline (day 0) mice and areas of colitis or dysplasia from piroxicam-treated IL-10−/− mice (day 70) are shown stained for (A–C) H&E, (E–G) BrdU incorporation, (H–J) P-Akt, and (K–M) P-β-catenin. (D) Evolution of histologic inflammation (solid squares) and prevalence of dysplasia (open squares) in IL-10−/− colitis. (N) P-Akt+ and (O) P-β-catenin+ cells/100 epithelial cells for groups indicated. Representative results from 3 independent experiments are shown. For all figures, arrows indicate positive crypt-based staining; arrowheads indicate middle/upper crypt staining. Original magnification, 20×; with 40× insets; scale bar, 200 μm. All values represent the mean ± SEM. *P < .05 vs day 0. Gastroenterology 2010 138, 595-605.e3DOI: (10.1053/j.gastro.2009.10.038) Copyright © 2010 AGA Institute Terms and Conditions

Figure 3 Mesalamine attenuates epithelial proliferation and PI3K signaling in chronic IL-10−/− colitis. Sections from day 28 and day 70 colitic mice compared with LD and HD mesalamine-treated IL-10−/− mice showing staining for (A–D) H&E, (E–H) BrdU, (I–L) P-Akt, and (M–P) P-β-catenin. For all panels, arrows indicate positive crypt-based staining; arrowheads indicate middle/upper crypt staining. LD and HD mesalamine were given from day 28 to day 70 to IL-10−/− mice as 500 mg and 1650 mg mesalamine/kg chow (equivalent to 600 mg and 2 g/day human dose respectively39). Stains showed LD and HD mesalamine reduced BrdU+, P-Akt+, and P-β-catenin+ cells in upper crypt regions of day 70 IL-10−/− mice. Scale bar, 200 μm. (Q and R) Cell counting shows mean ± SEM for P-Akt and P-β-catenin+/100 epithelial cells for groups indicated. *P < .05 vs day 70 no treatment. Gastroenterology 2010 138, 595-605.e3DOI: (10.1053/j.gastro.2009.10.038) Copyright © 2010 AGA Institute Terms and Conditions

Figure 4 Mesalamine impairs epithelial β-catenin and Akt activation in colitis. Epithelial cells were isolated and purified from IL-10−/− mice at baseline (day 0) and at day 70 as described in the Materials and Methods section. (A) Confirmation of purity of subcellular fractions (nucleus, histone H3 positive/α-tubulin negative; cytosolic, histone H3 negative/α-tubulin positive) by Western blotting. (B) Assessment of nuclear P-β-catenin and cytosolic P-Akt in colonic epithelial cells from baseline and colitic day 70 IL-10−/− colitis mice with or without LD and HD mesalamine treatment, as in Figure 3. β-actin was used to control for protein loading. (C and D) Densitometric comparisons of data in B normalized to β-actin relative to day 0. Values represent mean ± SEM. *P < .01. Gastroenterology 2010 138, 595-605.e3DOI: (10.1053/j.gastro.2009.10.038) Copyright © 2010 AGA Institute Terms and Conditions

Figure 5 Mesalamine attenuates progression of dysplasia in IL-10−/− colitis. Colons were removed from piroxicam-treated IL-10−/− mice on day 28 (pretreatment baseline) and day 70, then processed for H&E staining. Mice received standard, LD mesalamine, or HD mesalamine chow from day 28 through day 70. (A) Histologic inflammation scores and (B) prevalence (percent mice with any detectable dysplasia) of dysplasia were determined. (C) Dysplastic burden was measured by enumerating the average number of dysplastic lesions per colon and average length of dysplastic lesions per colon (D) for groups indicated. Values represent mean ± SEM. *P < .01, **P < .05, vs no treatment. Gastroenterology 2010 138, 595-605.e3DOI: (10.1053/j.gastro.2009.10.038) Copyright © 2010 AGA Institute Terms and Conditions

Figure 6 Mesalamine reduces epithelial proliferation and PI3K signaling in colitis-associated dysplasia. From left to right: early reactive epithelial changes adjacent to ulceration (day 28) compared with dysplasia at day 70 in mesalamine-treated IL-10−/− mice. Panels showing staining for (A–D) H&E, (E–H) BrdU incorporation, (I–L) P-Akt (arrows), and (M–P) P-β-catenin (arrows) within areas of dysplasia demonstrate little difference at day 28 vs day 70 with or without mesalamine. (Q and R) Mean ± SEM of positively stained cells/100 epithelial cells. For all panels, original magnification, 20×; with 40× insets; scale bar, 200 μm. *P < .05, vs no treatment, day 70. Data suggest that regions of dysplasia that persist after mesalamine treatment maintained active Akt and β-catenin signaling. Gastroenterology 2010 138, 595-605.e3DOI: (10.1053/j.gastro.2009.10.038) Copyright © 2010 AGA Institute Terms and Conditions

Supplementary Figure 1 Mesalamine treatment reduces COX-2 expression in human UC. Human biopsy samples from control and UC patients obtained during colonoscopy were immediately placed into RNeasy (whole tissue analysis) or fixed and processed for isolation of epithelial and stromal fractions by laser capture microdissection (LCM). Data are presented as mean ± SEM. (A) Whole tissue analysis demonstrates increased COX-2 mRNA expression is unaffected in mild colitis but reduced in severe mesalamine-treated UC as measured by real-time RT-PCR. (B) LCM isolated fractions showed proportionate reductions in stromal and epithelial COX -2 gene expression. Results are shown as fold induction in relation to mRNA expression of Gapdh. Gastroenterology 2010 138, 595-605.e3DOI: (10.1053/j.gastro.2009.10.038) Copyright © 2010 AGA Institute Terms and Conditions

Supplementary Figure 2 Mesalamine treatment reduces polyposis and P-β-catenin staining in APCΔ468 mice. Intestinal sections from 4-month-old control (untreated) APCΔ468 mice and mesalamine-treated APCΔ468 mice isolated after 2 weeks of HD mesalamine therapy are shown stained for (A) P-β-catenin. Left panels represent normal intestinal tissue with similar P-β-catenin staining isolated to the crypts. Right panels represent areas of polyposis. (B) From left to right: prevalence of intestinal polyps, size of polyps, total P-β-catenin+ epithelial cells in all polyps/intestine, and density of P-β-catenin+ epithelial cells within individual polyps. The latter figure was determined by dividing the total number of polyp-associated P-β-catenin+ in each intestine by the size of the polyps to generate the density of P-β-catenin+ intestinal epithelial cells/unit area. Original magnification, 20×; scale bar, 200 μm. All values represent the mean ± SEM. *P < .01 vs untreated. Gastroenterology 2010 138, 595-605.e3DOI: (10.1053/j.gastro.2009.10.038) Copyright © 2010 AGA Institute Terms and Conditions