Switching to a Healthy Diet Prevents the Detrimental Effects of Western Diet in a Colitis-Associated Colorectal Cancer Model

Abstract

Inflammatory bowel disease increases the odds of developing colitis-associated cancer. We hypothesized that Western-style diet (WD) aggravates azoxymethane (AOM)/dextran sulfate sodium salt (DSS)-induced colitis-associated tumorigenesis and that switching to the standard AIN93G diet will ameliorate disease symptoms even after cancer initiation. Female BALB/c mice received either WD (WD group) or standard AIN93G diet (AIN group) for the whole experimental period. After five weeks, the mice received 12.5 mg/kg AOM intraperitoneally, followed by three DSS cycles. In one group of mice, the WD was switched to AIN93G the day before starting the first DSS cycle (WD/AIN group). Feeding the WD during the whole experimental period aggravated colitis symptoms, shortened the colon (p < 0.05), changed microbiota composition and increased tumor promotion. On molecular level, the WD reduced proliferation (p < 0.05) and increased expression of the vitamin D catabolizing enzyme Cyp24a1 (p < 0.001). The switch to the AIN93G diet ameliorated this effect, reflected by longer colons, fewer (p < 0.05) and smaller (p < 0.01) aberrant colonic crypt foci, comparable with the AIN group. Our results show that switching to a healthy diet, even after cancer initiation is able to revert the deleterious effect of the WD and could be an effective preventive strategy to reduce colitis symptoms and prevent tumorigenesis.

Keywords: CYP24A1; Wnt pathway; aberrant crypt foci; colitis-associated cancer; inflammatory bowel disease; microbiome; mucosal regeneration; non-alcoholic fatty liver disease; vitamin D; western diet.

Figure 1

Graphical representation of the treatment protocol. Three groups of female BALB/c mice received either the AIN93G diet or the Western-style diet (WD) for 35 days, before they received azoxymethane (AOM) intraperitoneal (i.p.) one day before the start of the first cycle of dextran sulfate sodium salt (DSS), in one group, the WD was switched to the AIN93G diet. The mice were sacrificed 28 days after the third cycle of DSS.

Figure 2

Effect of the WD on macroscopic changes (A) Body weight loss during the period of AOM/DSS administration (grey-shaded area). Day 0: Start of AIN93G or WD feeding, day 35: AOM i.p., day 49: End of first DSS cycle, day 70: End of second DSS cycle, day 89: End of third DSS cycle, and day 117: Day of sacrifice. Body weight of each mouse was set relative to day 0. Mean ± SEM, (B) Body weight changes based on the analysis of the area under the curve (AUC) during the whole experimental period of 117 days. (C) Effect of the diet on colon length, (D) liver, and (E) spleen weight measured on the day of sacrifice. n = 8, median, interquartile range, and whiskers (min to max); ANOVA with Tukey post-hoc test; * p < 0.05, ** p < 0.01, and *** p < 0.001.

Figure 3

Effect of the diet on the (A) number (n = 7–8), and (B) size of colonic lymph follicles (n = 4–5), ANOVA with Tukey post-hoc test. Effect of the diet on (C) number of aberrant crypt foci (ACFs), ANOVA with Tukey post-hoc test on log transformed data (n = 7–8). Effect of the diet on the (D) number of aberrant crypts (n = 7–8), Kruskal Wallis with Dunn post-hoc test; median, interquartile range, and whiskers (min to max); * p < 0.05, ** p < 0.01.

Figure 4

Effect of the diet on mRNA expression of the hepatic (A) Cyp3a11, and (B) Cyp2r1 and on the colonic expression of Cyp24a1 in (C) the ascending and (D) descending colon (n = 4–7). Median, interquartile range, and whiskers (minimum to max). In the WD group of (A,C) one outlier, as determined by Grubbs’ test, was excluded from the analysis. ANOVA with Tukey post-hoc test; *** p < 0.001.

Figure 5

Effect of the diets on the microbiome. (A) Multi-dimensional scaling plot of fecal samples from day 43 (first day after diet switch) and from day 89 (47th day after diet switch), (B,C) pairwise UniFrac distances between the WD/AIN group and the AIN and WD groups, respectively. Effect of the diets on day 89 on the: (D) Firmicutes/Bacteroidetes ratio, (E) relative abundance of Faecalibaculum, and (F) relative abundance of Ruminoclostridium. * p < 0.05, *** p < 0.001.

Figure 6

Effect of the diet on the mRNA expression of genes involved in apoptosis, survival, inflammation (A,B) and of genes of the Wnt pathway (C–F) in colon ascendens. n = 4–8 (A–E), n = 5–8 (F) median, interquartile range, and whiskers (min to max); ANOVA with Tukey post-hoc test (A–E) or Kruskal Wallis with Dunn post-hoc test (F), * p < 0.05, ** p < 0.01.

Figure 7

Effect of the diet on proliferation of colonic crypts. Ki67-positive cells in untransformed crypts of (A) colon ascendens and (B) descendens. For each animal, three regions consisting of two to five crypts were chosen and the total number of Ki67-positive and of Ki67-negative cells were determined. n = 4–6; median, interquartile range, and whiskers (min to max); in the WD group of (A) one outlier, as determined by Grubbs’ test, was excluded from analysis; and ANOVA with Tukey post-hoc test, * p < 0.05. Representative pictures of Ki67 (brown) staining (C) in the untransformed mucosa of the WD/AIN group and (D) in a tumor of the WD group. Color and contrast of the images (C,D) were non-linearly enhanced for purpose of presentation only, scale bars = 100 µM.

Figure 8

Effect of a Western Diet in the chemically-induced colorectal cancer model.