Resveratrol modulates the gut microbiota to prevent murine colitis development through induction of Tregs and suppression of Th17 cells

Abstract

Inflammatory diseases of the gastrointestinal tract are often associated with microbial dysbiosis. Thus, dietary interactions with intestinal microbiota, to maintain homeostasis, play a crucial role in regulation of clinical disorders such as colitis. In the current study, we investigated if resveratrol, a polyphenol found in a variety of foods and beverages, would reverse microbial dysbiosis induced during colitis. Administration of resveratrol attenuated colonic inflammation and clinical symptoms in the murine model of 2,4,6-trinitrobenzenesulfonic acid (TNBS)-induced colitis. Resveratrol treatment in mice with colitis led to an increase in CD4⁺ FOXP3⁺ and CD4⁺ IL-10⁺ T cells, and a decrease in CD4⁺ IFN-γ⁺ and CD4⁺ IL-17⁺ T cells. 16S rRNA gene sequencing to investigate alterations in the gut microbiota revealed that TNBS caused significant dysbiosis, which was reversed following resveratrol treatment. Analysis of cecal flush revealed that TNBS administration led to an increase in species such as Bacteroides acidifaciens, but decrease in species such as Ruminococcus gnavus and Akkermansia mucinphilia, as well as a decrease in SCFA i-butyric acid. However, resveratrol treatment restored the gut bacteria back to homeostatic levels, and increased production of i-butyric acid. Fecal transfer experiments confirmed the protective role of resveratrol-induced microbiota against colitis inasmuch as such recipient mice were more resistant to TNBS-colitis and exhibited polarization toward CD4⁺ FOXP3⁺ T cells and decreases in CD4⁺ IFN-γ⁺ and CD4⁺ IL-17⁺ T cells. Collectively, these data demonstrate that resveratrol-mediated attenuation of colitis results from reversal of microbial dysbiosis induced during colitis and such microbiota protect the host from colonic inflammation by inducing Tregs while suppressing inflammatory Th1/Th17 cells.

Keywords: 2,4,6-trinitrobenzenesulfonic acid; T helper cells; butyrate; fecal transfer; inflammation; short-chain fatty acid.

Figure 1:

Treatment with resveratrol reduces clinical symptoms associated with TNBS-induced colitis murine model. Balb/c mice were administered intrarectally with 1mg of TNBS to induce colitis. Four groups of mice were used: Vehicle, Resveratrol, TNBS+Vehicle and TNBS+Resveratrol. The percent weight loss (A) was determined over the course of the study. (B) Survival curve of mice up to day 6 with colitis and those treated with RES. Colon lengths (C-D) were measured upon sacrifice (Day 5). Serum levels of SAA (E), MPO (F) and Lcn2 (G) were evaluated by ELISA. Endoscopy (H) was performed on mice on days 0, 3, and 5. Colonoscopy scores are provided (I). Significance (p-value: *<0.05, **<0.01, ***<0.005, ****<0.001) was determined by using one-way ANOVA and post-hoc Tukey’s test. In all data presented in the figure, 5 mice were used in each group. Data presented is representative of at least 3 independent experiments.

Figure 2:

Treatment with resveratrol prevents cellular infiltration and mucin degradation and maintains colon gut structural architecture in TNBS model. The study was designed as described in Fig 1 legend. . Colons (n=5) were excised from experimental mice at the endpoint of experiment, fixed in 10% formaldehyde, and embedded in paraffin blocks. Cross-section slides containing colons from experimental groups were stained using H&E (A) or PAS (B) for histological evaluation. Images of stained tissue were taken using both 4x and 20x objectives, and histological scores were provided (C). Scale bars (white) depicted are at 100 μM. Data is representative of at least 3 independent experiments.

Figure 3:

Resveratrol alters T cell subsets during TNBS colitis. The study was designed as described in Fig 1 legend. Flow cytometry histograms/dot plots are shown for the following T cell subsets: CD3⁺ (A), CD4⁺ or CD8⁺ cells (B), CD4⁺FOXP3⁺ (C), CD4⁺IL10⁺ (D) and CD4⁺IFNγ⁺ (E), and CD4⁺IL-17⁺ (F) expressing cells. For Figures C-F, cells were gated on the CD4⁺ population. The gating strategy for the CD4+ populations is detailed in Supplemental Figure 2. Quantitative bar graphs depicting absolute cell numbers of the T cell subsets is provided (G) Each experimental group had at least 5 mice included, and significance (p-value: *<0.05, **<0.01, ***<0.005, ****<0.001) was determined for absolute cell numbers by using one-way ANOVA followed by Tukey’s post-hoc multiple comparisons test. Data is representative of at least 3 independent experiments.

Figure 4:

16S rRNA gene sequencing analysis. The study was designed as described in Fig 1 legend. Gut microbiome samples were collected from experimental groups by performing cecal flushes. Genomic DNA was isolated and V3-V4 regions of 16S rRNA gene subunit were sequenced. Three randomly selected mice from each group (n=3) were used for these experiments. All sequencing samples were analyzed using Nephele software 16S metagenomics provided at Nephele website (nephele.niaid.nih.gov). Alpha diversity (A), and Beta diversity (B) are depicted. LeFSe analysis of the Nephele OTU output files generated the cladogram (C) and LDA score bar graph (D) depicting microbial biomarkers among TNBS+Vehicle vs. TNBS+RES groups. OTU percent abundances are shown for the species Bacteroides acidifaciens (E) Ruminococcus gnavus (F) and Akkermansia muciphila (G). Validation of these significantly-altered bacterial species were performed using PCR and the fold changes are calculated using the delta-delta CT method with comparison to Vehicle controls (H-J). For 16S rRNA gene sequencing, 3 representative cecal flushes from each experimental group were processed and sequenced. For PCR validation, 10 mice were used in each group and fold changes were calculated using the delta-delta CT method compared to Vehicle control. Significance (p-value: *<0.05, **<0.01, ***<0.005, ****<0.001) was determined by using one-way ANOVA followed by Tukey’s post-hoc multiple comparisons test. Experiments are representative of 3 independent experiments.

Figure 5:

Resveratrol treatment alters SCFA production in TNBS colitis. The study was designed as described in Fig 1 legend. SCFA were isolated from cecal contents of experimental groups through acidification using metaphosphoric acid. GC-FID analysis was performed to determine the concentrations of acetic (A), propionic (B), i-butyric (C), n-buytric (D), i-valeric (E), and n-valeric (F) acids. SCFAs were identified using standard compounds purchased from Sigma-Aldrich. Representative data from two independent experiments with 5 mice in each group is depicted. Significance (p-value: *<0.05, **<0.01, ***<0.005, ****<0.001) was determined using one-way ANOVA followed by Tukey’s post-hoc multiple comparisons test.

Figure 6:

Transfer of resveratrol-treated fecal contents leads to amelioration of colitis. Female Balb/c mice were treated for 4 weeks with streptomycin and ampicillin (1g/L) prior to being injected intrarectally with 1mg of TNBS to induce colitis. Antibiotic-treated mice were weighed (A) and PCR performed on colonic flush samples to determine abundance of bacteria in the gut compared to naïve mice (B-C). These mice received fecal transfer (FT) from either colitis disease groups, (FT) TNBS+Vehicle, or from TNBS+Resveratrol-treatment groups, (FT) TNBS+RES. The percent weight loss (D) was determined over the course of the study. Colon lengths were measured upon sacrifice (E-F). Serum biomarkers for SAA (G), MPO (H), LCN2 (I) were detected using ELISA kits. Endoscopic images (J, top panel) and H&E stains of colons (J, bottom panel) are depicted (n=5 per group). Colonoscopy scores (K, top) and histological scores (K, bottom) are provided. Cells were isolated from mesenteric lymph nodes of experimental groups and absolute cell numbers from fecal transfer experiments were determined for CD4⁺FOXP3⁺ (L), CD4⁺IL10⁺ (M), CD4⁺IL17⁺ (N) and CD4⁺IFNγ⁺ (O). PCR validation from colonic flushes was performed after fecal transfer to confirm alterations in Akkermansia muciphila (P) Ruminococcus gnavus (Q) and Bacteroides acidifaciens (R), using delta-delta CT method with comparison to Vehicle controls. Each group had 10 recipient mice in this experiment and significance (p-value: *<0.05, **<0.01, ***<0.005, ****<0.001) was determined using one-way ANOVA followed by Tukey’s post-hoc multiple comparisons test.