Diet Rich in Animal Protein Promotes Pro-inflammatory Macrophage Response and Exacerbates Colitis in Mice

Abstract

Diet is a major factor determining gut microbiota composition and perturbances in this complex ecosystem are associated with the inflammatory bowel disease (IBD). Here, we used gnotobiotic approach to analyze, how interaction between diet rich in proteins and gut microbiota influences the sensitivity to intestinal inflammation in murine model of ulcerative colitis. We found that diet rich in animal protein (aHPD) exacerbates acute dextran sulfate sodium (DSS)-induced colitis while diet rich in plant protein (pHPD) does not. The deleterious effect of aHPD was also apparent in chronic DSS colitis and was associated with distinct changes in gut bacteria and fungi. Therefore, we induced acute DSS-colitis in germ-free mice and transferred gut microbiota from aCD or aHPD fed mice to find that this effect requires presence of microbes and aHPD at the same time. The aHPD did not change the number of regulatory T cells or Th17 cells and still worsened the colitis in immuno-deficient RAG2 knock-out mice suggesting that this effect was not dependent on adaptive immunity. The pro-inflammatory effect of aHPD was, however, abrogated when splenic macrophages were depleted with clodronate liposomes. This treatment prevented aHPD induced increase in colonic Ly-6C^high pro-inflammatory monocytes, but the ratio of resident Ly-6C^-/low macrophages was not changed. These data show that the interactions between dietary protein of animal origin and gut microbiota increase sensitivity to intestinal inflammation by promoting pro-inflammatory response of monocytes.

Keywords: colitis; dietary protein; germ-free; macrophage; microbiota.

Figure 1

In immune-competent BALB/c mice, HPD of animal origin significantly worsens the severity of acute DSS-colitis (A) while HPD of plant origin does not (B). Colitis severity was measured by disease activity index (weight loss, stool consistency, and rectal bleeding), colon shortening and damage to colon mucosa. Neither aHPD (C) nor pHPD (D) changed the proportions of regulatory T cells (CD3⁺CD4⁺FoxP3⁺) or Th17 cells (CD3⁺CD4⁺RorγT⁺) in mesenteric lymph nodes, while monocytes (CD11c^lowLy-6C^hi) were significantly increased and more activated (iNOS⁺) in colons of aHPD fed mice. The weight loss was analyzed by two-way ANOVA and other parameters by unpaired Student's t-test; ^*p < 0.05 ^**p < 0.01 ^***p < 0.001. Data are from one representative experiment out of 6 (A,C) or 3 (B,D) independent experiments (n = 5–8).

Figure 2

HPD of animal origin worsens the severity of chronic DSS-colitis. The weight loss was analyzed by two-way ANOVA and other parameters by unpaired Student's t-test; ^*p < 0.05 ^**p < 0.01 ^***p < 0.001. Data are from one representative experiment out of 2 independent experiments (n = 5–8).

Figure 3

Gut microbiota, both bacterial and fungal, is profoundly changed by diet and intestinal inflammation. Comparison of impact of diet and gut inflammation on bacterial (A) and fungal (D) alpha diversity using Chao1 diversity index. Principal coordinates analysis (PCoA) plot using the Bray-Curtis distance metric shows the compositional differences induced by diet and inflammation at the diet switch (day −21) and before (day 0) and after the DSS treatment (day 8) for bacteria (B) and fungi (E). Each colored orb represents the microbiota composition in feces of one mouse. Each color represents one group of mice at the day −21, 0, or 8. Linear discriminant analysis effect size (LEfSe) for bacteria (C) and fungi (F) shows taxonomic units typical for each group. The first taxonomic unit represents phylum (p) and the second represents the lowest taxonomic rank revealed by the analysis, it being class (c)—order (o)—family (f)—genus (g)—species (s) or variety (v). Heatmap of bacterial gene functional predictions using the PICRUSt algorithm showing the changes in KEGG level-3 pathways (columns) and experimental groups (row) (G). Fecal protein content, as measured by BCA assay and amounts of SCFA, phenylalanine and BCAA, as measured by NMR-based metabolomics (H). The changes in alpha diversity were analyzed by ANOVA with Tukey's multiple comparison test; †p < 0.05 for changes in aCD or pCD and ^*p < 0.05 ^**p < 0.01 for the other comparisons. The metabolomic data were analyzed by Wilcoxon-Mann-Whitney test; ^*p < 0.05 ^**p < 0.01 ^***p < 0.001 (n = 3). Boxplot legend: asterisk—mean value, black dots—metabolites' intensities in particular samples.

Figure 4

HPD of animal origin does not worsen the severity of acute DSS-colitis in germ-free animals (A), but microbiota transferred from aHPD fed mice mitigates colitis severity (B). The weight change shows percentage of weight gain/reduction between day 0 and 8 of DSS treatment. All parameters were analyzed by unpaired Student's t-test; ^***p < 0.001 (n = 5–8).

Figure 5

In immune-deficient RAG2^−/− mice, aHPD worsens the severity of acute DSS-colitis (A), which is accompanied with the increase in TNF-α, IL-6, IL-1β, and IL-33 in the colon (B). The weight loss was analyzed by two-way ANOVA and other parameters by unpaired Student's t-test; ^*p < 0.05 ^**p < 0.01 ^***p < 0.001. Data are from one representative experiment out of 3 independent experiments (n = 6–7).

Figure 6

Intraperitoneal administration of clodronate liposomes is sufficient to mitigate the colitis in RAG2^−/− (A) mice. Splenic (B) and colonic (C) populations of neutrophils, macrophages and monocytes as analyzed by flow cytometry in colitic RAG2^−/− mice treated either with empty (–LIPO) or clodronate-loaded (+LIPO) liposomes. The weight loss was analyzed by two-way ANOVA and other parameters by unpaired Student's t-test; ^*p < 0.05 ^**p < 0.01 ^***p < 0.001. Data are from one representative experiment out of 3 independent experiments (n = 6–7).