Short-term sulindac administration induced apoptosis in the intestinal stem cells and suppressed adenoma formation in APC^Min/+ mice. Ten-week-old APC^Min/+ mice were fed with control or sulindac-containing (20 mg/kg/d) AIN93G diet for 1 or 2 wk and killed immediately after treatment. Intestinal polyp phenotypes, β-catenin localization, and apoptosis were analyzed. (A) Small intestinal sections from the treated mice were analyzed for apoptosis by TUNEL staining. The fractions of crypts containing at least one TUNEL-positive cell were determined. (B) Numbers of small intestinal polyps (≥0.5 mm in diameter) were counted following sulindac treatment for 1 wk. (C) Staining of indicated makers in APC^Min/+ mice treated with sulindac for 1 wk. For Lgr5 (EGFP) staining, APC^Min/+ mice containing the Lgr5-EGFP lineage marking allele (Lgr5-EGFP/APC^Min/+ mice) were analyzed. Lgr5 marks CBC cells and occasionally +4 cells, whereas MMP7 labels Paneth cells. DAPI (blue) was used for nuclear counter staining. Arrows indicate example TUNEL-positive CBCs (Lgr5-positive or MMP7-negative). (D) Quantification of TUNEL-positive cells based on locations in the crypts. Apoptotic index represents the fraction of crypts containing one or more TUNEL-positive cells. (E) Quantification of Lgr5-positive crypts containing one or more TUNEL-positive cells in Lgr5-EGFP/APC^Min/+ mice treated with control or sulindac diet for 1 wk. (F) Left: Staining of Lgr5 (red) and active caspase 3 (green) in APC^Min/+ mice treated with sulindac for 1 wk, with arrows indicating double positive cells. Right: Quantification of crypts containing one or more active caspase 3-positive cells. Values in A, B, and D–F are means ± SD (n = 6 in each group). At least 500 crypts from each animal were analyzed. (Scale bars: 15 μm.)

NSAIDs induced apoptosis in human colonic polyps and removed cells with activated Wnt signaling. (A) TUNEL staining (brown) of intestinal polyps from patients taking or not taking NSAIDs. Arrows indicate TUNEL-positive apoptotic cells. (B) Quantification of crypts containing TUNEL-positive cells. Apoptotic index represents the percentage of intestinal crypts containing one or more TUNEL-positive cells. (C) Sections of intestinal polyps from four patients taking or not taking NSAIDs were stained for TUNEL (green), OLFM4 (red), and DAPI (blue). Arrows indicate TUNEL and OLFM4 double-positive cells. (D) Sections of intestinal polyps as in C were stained for p-β-catenin and quantified. Values in B and D are means ± SD (n = 4 in each group). At least 200 crypts from each patient were analyzed. (Scale bars: 15 μm.)

Sulindac treatment removed the intestinal cells with nuclear or phospho-β-catenin via apoptosis. WT and APC^Min/+ mice were fed with control or sulindac-containing (20 mg/kg/d) diet for 1 or 2 wk and killed immediately after treatment. Small intestinal sections from the mice were analyzed for β-catenin localization, β-catenin Ser552 phosphorylation (p-β-catenin), and apoptosis (TUNEL) by immunostaining. (A) Analysis of β-catenin localization. Left: Staining of β-catenin (green or white) and DAPI (blue) in APC^Min/+ mice treated with sulindac for 1 wk. Circles mark representative CBCs with nuclear β-catenin. Right: Quantification of crypts with nuclear β-catenin in WT or APC^Min/+ mice treated with control or sulindac diet for 1 wk. (B) Analysis of β-catenin localization and apoptosis. Left: Staining of β-catenin (green), TUNEL (red), and DAPI (blue) in APC^Min/+ mice treated with sulindac for 1 wk. Circles mark example CBCs with nuclear β-catenin that were undergoing apoptosis. Right: Quantification of crypts positive for both nuclear β-catenin and TUNEL in WT and APC^Min/+ mice treated with control or sulindac diet for 1 wk. (C) Analysis of β-catenin phosphorylation. Left: Staining of p-β-catenin (red) and DAPI (blue) in APC^Min/+ mice treated with control or sulindac diet for 1 or 2 wk. Right: Quantification of crypts containing p-β-catenin-positive cells. (D) Analysis of β-catenin phosphorylation and apoptosis. Upper: Staining of p-β-catenin (red), TUNEL (green), and DAPI (blue) in APC^Min/+ mice treated with sulindac for 1 wk. Arrows indicate TUNEL and p-β-catenin double-positive cells. Lower: Quantification of crypts containing apoptotic cells in WT or APC^Min/+ mice treated with control or sulindac diet for 1 or 2 wk. Values in A–D are means ± SD (n = 6 in each group). At least 500 crypts from each animal were analyzed. (Scale bars: 15 μm.)

SMAC deficiency attenuated the chemopreventive effect of sulindac in APC^Min/+ mice by blocking apoptosis in the intestinal stem cells. Age- and sex-matched parental (APC^Min/+) and SMAC-deficient APC^Min/+ mice (SMAC^−/−/APC^Min/+) were fed with control or sulindac-containing (20 mg/kg/d) diet for 1 or 2 wk and killed immediately after treatment. Intestinal polyp phenotypes, β-catenin localization, β-catenin Ser552 phosphorylation (p-β-catenin), and apoptosis were analyzed and compared. (A) Polyp (≥0.5 mm in diameter) number reduction in the treated mice. (B) Distribution of polyp number reduction in three different regions in the small intestine of the treated mice. (C) Upper: Staining of TUNEL (brown) and hematoxylin (blue) in the mice treated with sulindac for 1 wk. Arrows indicate example TUNEL-positive cells. (Scale bar: 15 μm.) Lower: Quantification of crypts containing one or more TUNEL-positive cells in the treated mice. (D) Fractions of crypts containing one or more CBC cells with nuclear β-catenin. (E) Fractions of crypts containing one or more TUNEL-positive cells with nuclear β-catenin. (F) Fractions of crypts containing one or more p-β-catenin and TUNEL double-positive cells. (G) Fractions of crypts containing one or more p-β-catenin–positive cells. Values are means ± SD; n = 6 in each group in A–D; n = 4 in each group in E–G. At least 500 crypts from each animal were analyzed.