A westernized diet changed the colonic bacterial composition and metabolite concentration in a dextran sulfate sodium pig model for ulcerative colitis

Abstract

Introduction: Ulcerative colitis (UC) is characterized by chronic inflammation in the colonic epithelium and has a blurred etiology. A western diet and microbial dysbiosis in the colon were reported to play a role in UC development. In this study, we investigated the effect of a westernized diet, i.e., increasing fat and protein content by including ground beef, on the colonic bacterial composition in a dextran sulfate sodium (DexSS) challenged pig study.

Methods: The experiment was carried out in three complete blocks following a 2×2 factorial design including 24 six-week old pigs, fed either a standard diet (CT) or the standard diet substituted with 15% ground beef to simulate a typical westernized diet (WD). Colitis was induced in half of the pigs on each dietary treatment by oral administration of DexSS (DSS and WD+DSS, respectively). Samples from proximal and distal colon and feces were collected.

Results and discussion: Bacterial alpha diversity was unaffected by experimental block, and sample type. In proximal colon, WD group had similar alpha diversity to CT group and the WD+DSS group showed the lowest alpha diversity compared to the other treatment groups. There was a significant interaction between western diet and DexSS for beta diversity, based on Bray-Curtis dissimilarly. The westernized diet and DexSS resulted in three and seven differentially abundant phyla, 21 and 65 species, respectively, mainly associated with the Firmicutes and Bacteroidota phyla followed by Spirochaetota, Desulfobacterota, and Proteobacteria. The concentration of short-chain fatty acids (SCFA) was lowest in the distal colon. Treatment had a slight effect on the estimates for microbial metabolites that might have valuable biological relevance for future studies. The concentration of putrescine in the colon and feces and that of total biogenic amines was highest in the WD+DSS group. We conclude that a westernized diet could be a potential risk factor and an exacerbating agent for UC by reducing the abundance of SCFA-producing bacteria, increasing the abundance of pathogens such as Helicobacter trogontum, and by increasing the concentration of microbial proteolytic-derived metabolites in the colon.

Keywords: 16S rRNA gut metagenomics; colonic inflammation; dextran sulfate sodium; inflammatory bowel disease; meat consumption; porcine model; ulcerative colitis.

Figure 1

Non-metric multidimensional scaling (NMDS) plot based on Bray–Curtis dissimilarity index (A), and dbRDA MDS plot (B) of sample scores from Bray–Curtis dissimilarity for different treatments (CT: control, WD: westernized diet, DSS: group treated with dextran sulfate sodium, and WDDSS: WD + DSS) extracted from dbRDA model (variance explained by our treatments = 31.0%, p < 0.01). On dbRDA plot, the treatments (constrained factors) explained 70.1 and 16.9% of the fitted variance and 21.7 and 5.20% of total variance for Bray–Curtis dissimilarity on dbRDA1 and dbRDA2 axis, respectively. The shape of the points represents the origin of the samples, i.e., digesta from proximal and distal colon, and fecal samples and the color of the points represent different treatments. Graph-based analysis of the distributions in bacterial composition for different treatments (C), based on Bray–Curtis dissimilarity matrix with maximum distance of 0.35. The histogram of permutation test based on MST for Bray-Curtis is presented (p < 0.01).

Figure 2

Absolute abundances of different phyla with total abundance across all sequence reads in different treatments (A). The number of different phyla present in the treatment groups was 16, 16, 11, and 17 for CT, WD, DSS, and WDDSS (WD + DSS), respectively. Venn diagram of shared taxa between different treatments at Phylum (B) and Species (C) levels.

Figure 3

Differentially abundant Phyla for the main effect of diet (westernized diet groups vs. non-westernized group; A) and DexSS (DexSS groups vs. non-DexSS groups; B), and pairwise comparison for phyla with significant interaction for WD vs. CT (C), DSS vs. CT (D), WDDSS (WD + DSS) vs. CT (E), and WDDSS (WD + DSS) vs. WD (F) with FDR < 0.05.

Figure 4

Waterfall plot for differentially abundant species of the main effect of diet (westernized diet groups vs. non-westernized diet groups; A) and DexSS (DexSS groups vs. non-DexSS groups; B). Different colors represent different phyla. Only species with |LFC| > 2 and FDR < 0.01 are presented.

Figure 5

Heatmap of Spearman’s rank correlation coefficient (r_s) between the top 100 abundant species and SCFA and biogenic amines in all segments. Columns are clustered based on the standard deviations between species abundances, i.e., those with lower deviation are clustered together. Species are color-labeled with their correspondent phyla and significant associations are labeled with stars.