EGCG drives gut microbial remodeling-induced epithelial GPR43 activation to lessen Th1 polarization in colitis

Abstract

Modulation of immune microenvironment is critical for inflammatory bowel disease (IBD) intervention. Epigallocatechin gallate (EGCG), as a natural low toxicity product, has shown promise in treating IBD. However, whether and how EGCG regulates the intestinal microenvironment is not fully understood. Here we report that EGCG lessens colitis by orchestrating Th1 polarization and self-amplification in a novel manner that required multilevel-regulated intestinal microecosystem. Mechanistically, EGCG activates GPR43 on IEC to inhibit Th1 polarization dependently of short chain fatty acid (SCFA)-producing gut microbiota. Inhibition of GPR43 activity weakens the protective effects of EGCG on colitis development. Moreover, we confirm that fecal SCFAs and/or intestinal GPR43 are limited in patients with colitis and are correlated with Th1 cell number. Taken together, our study reveals an intestinal microenvironment-dependent immunoregulatory effects of EGCG in treating IBD and provides insight into mechanisms of EGCG-based novel immunotherapeutic strategies for IBD.

Keywords: EGCG; GPR43; Inflammatory bowel disease; Short chain fatty acids; Th1 cells.

Fig. 1

EGCG reduces gut Th1 cell numbers in DSS-induced colitis. (A) Experimental scheme for (B–H). C57BL/6 mice were administrated 2.5 % (w/v) DSS in drinking water for 5 days after they were treated with 80 mg/kg/day EGCG via oral gavage for 21 days, then allowed to recover for 2 days using normal drinking water. (B) Body weight, gross bleeding, and stool consistency were recorded daily to calculate disease activity index (DAI). n = 8–12. (C) Representative images of colonic histopathology by hematoxylin and eosin staining and pathological scoring. Scale bar, 100 μm, n = 6 (D) Colon length was measured. n = 8–12. (E and F) Colonic T cell subtypes were detected by flow cytometry after colon immune cells were isolated, and CD4⁺ and CD8⁺ T cells were quantified using FlowJo, gating strategy (E) and cell subpopulation ratios were obtained (F). Two to three mice were pooled for one sample, data are one representative of two independent experiments. n = 3. (G) RNA-seq analysis for colon tissue included Th cell differentiation-related transcriptome heatmap. n = 6–10. (H) mRNA expression of factors that promote Th1 differentiation in colon tissue were detected by qRT-PCR. n = 6–8. Data are presented as mean ± SEM. Statistical significance was determined using one-way ANOVA, followed by Tukey's multiple-comparison tests or Student's t-test. *P ≤ 0.05, **P ≤ 0.01. See also Figure S1.

Fig. 2

EGCG indirectly regulates the Th1 subtype via microbes. (A) Experimental scheme of EGCG directly treatment for activated T and Th1 cells in (B). Naïve CD4⁺ T cells from mouse spleens were isolated, activated in vitro using anti-CD28 and anti-CD3, and induced Th1 polarization. Cells were then stimulated with 0.5 % (w/v) DSS after incubation of EGCG with different concentrations (0.5 μM, 1.0 μM, and 2.0 μM) for 24 h. (B) Effects of direct EGCG treatment on cell function-related factors of activated T and polarized Th1 cells were determined by flow cytometry. Data are one representative of two independent experiments. n = 4. (C–E) Mice were orally gavaged with 80 mg/kg/day EGCG for 21 days, then given drinking water containing 2.5 % (w/v) DSS for 5 days, followed by 2 days of recovery. 16S rDNA sequencing for colon feces was performed via the 454 FLXtitanium system and sequences were analyzed using QIIME software. Heatmap showed differences of top 35 microbial gene enrichment for different treatments (C). Bacterial difference at the phylum (D) and genus (E) levels among groups were analyzed, bacteria with significant differences were listed. n = 8–10. (F) Experimental scheme for antibiotic treatment in (F–I). Mice were administered EGCG at 80 mg/kg/day orally for 21 days, then treated with an antibiotic cocktail (ABX, a volume of 10 μL/g body weight of drinking water supplemented with 0.1 mg/mL Amphotericin-B, 10 mg/mL Ampicillin, 10 mg/mL Neomycin Trisulfate salt hydrate, 10 mg/mL Metronidazole, and 5 mg/mL Vancomycin hydrochloride) for 14 days. The mice were then given water containing 2.5 % (w/v) DSS for 5 days, allowed for 2 days recovery with normal water (F). (G and H) Daily DAI (G) during the experiment were recorded. n = 8–10. At the end of the experiment, the mice were sacrificed, and colon length (H) were measured. n = 6. (I) Th1 cell subsets were detected and quantified by flow cytometry after colonic immune cells were isolated. Data are one representative of two independent experiments. n = 6. Data are shown as mean ± SEM. Statistical significance was determined using one-way ANOVA, followed by Tukey's multiple-comparison tests or Student's t-test. *P ≤ 0.05, **P ≤ 0.01. See also Figure S2–S4.

Fig. 3

EGCG activates colonic SCFA receptor GPR43 in mice. Mice were administrated 2.5 % (w/v) DSS in drinking water for 5 days after they were treated with 80 mg/kg/day EGCG via oral gavage for 21 days, then allowed to recover for 2 days using normal drinking water. (A) The abundance of SCFA-producing bacteria in colon feces. n = 8–10. (B) Fecal concentrations of SCFA. n = 6–8. (C–E) RNA-Seq analysis for colon tissues revealed the DEGs (C) and signaling pathways (D) in the EGCG treated colitis mice compared with the non-treated, as well as KEGG of Ffar2 related genes (E). n = 8–10. (F) GSEA from RNA-seq analysis for colon tissues showed the significant downregulation of NF-κB in EGCG treated colitis mice. n = 8–10. (G) qRT-PCR results for colon tissues showed gene levels of GPR43 downstream signaling pathways. n = 4–6. (H) Western blotting analysis of NF-κB signaling and GPR43 were determined and quantified. Data are one representative of two independent experiments. n = 3. Data are presented as mean ± SEM. Statistical significance was determined using one-way ANOVA, followed by Tukey's multiple-comparison tests or Student's t-test. *P ≤ 0.05, **P ≤ 0.01. See also Fig. S5.

Fig. 4

GPR43 mediates inhibition of Th1 polarization by EGCG. (A) Experimental scheme for inhibitor treatment in (B–G). Mice were treated with 80 mg/kg/day EGCG for 24 days, and orally gavaged GLPG0974 (GPR43 inhibitor) at 10 mg/kg for every 3 days, then mice were administrated 2.5 % (w/v) DSS in drinking water for 5 days, and allowed to recover for 2 days using normal drinking water. (B–D) Body weight changes (B), disease activity index (DAI) (C), and colon length (D). n = 6. (E and F) T cell subtypes were detected by flow cytometry after colonic immune cells were isolated and stained. CD4⁺IFN- γ⁺ T cells (Th1) were showed (E), Major effector T cells including CD4⁺IL-17A⁺ T cells (Th17), CD4⁺IL-4⁺ T cells (Th1) were quantified after flow cytometry, and Th1/Th2 ratio was calculated (F). n = 6. (G) Expressions of NF-κB signals, IL12, and GPR43 in colon were determined by Western blot. Data are one representative of two independent experiments. n = 3. (H) IFN-γ protein concentration in colon was detected by ELISA. n = 6. (I) Dynamic molecular docking was employed to study interactions and binding affinity between acetate/propionate and GPR43, amino acid residues were predicted. Data are presented as mean ± SEM. Statistical significance was determined using one-way ANOVA, followed by Tukey's multiple-comparison tests or Student's t-test. *P ≤ 0.05, **P ≤ 0.01.

Fig. 5

Epithelial GPR43 is required for the inhibition of EGCG fermentation on Th1 polarization. (A) Experimental scheme for (B–D). EGCG and mouse colonic contents were fermented, and the fermentations was centrifuged, then concentrated at low temperature and diluted with cell culture medium. The resulting supernatant was used to treat mouse colonic epithelial cells (MCoEpiCs) for 24 h after epithelial cells were treated with GLPG for GPR43 inhibition. Then culture medium was refreshed and followed by 2.5 % (w/v) DSS treatment. The culture medium was refreshed to exclude previous treatment subjects, and cells were cultured for another 24 h. Subsequently, the resulting medium was collected to treat the activated mouse T cells, which were isolated from spleen and stimulated with anti-CD28 and anti-CD3. (B) GPR43 expression in MCoEpiCs by flow cytometry. n = 4. (C) The resulting Medium was collected to determine IFN-γ concentration. n = 4. (D) Treated CD4⁺ T cells were analyzed for Th1 markers and the ratio of Th1 cells out of CD4⁺ T cells was calculated. n = 4. (E) Experimental scheme for (F and G). Naïve CD4⁺ T cells were isolated from mouse spleen and activated with anti-CD3, anti-CD28, and Th1 polarization proteins, then treated with GLPG. Supernatant from fermentations was used to treat the resulting T cells above, followed by 0.5 % (w/v) DSS treatment. (F and G) Th1, Th2, Th17, and Treg markers were determined both in mRNA and protein levels by qRT-PCR and flow cytometer to detect the effect of the fermentation on Th1 functions including cytokine secretions and differentiation potentials to other effector T cells. Data are representative from one of two independent experiments, n = 4. Data are presented as mean ± SEM. Statistical significance was determined using one-way ANOVA, followed by Tukey's multiple-comparison tests or Student's t-test. *P ≤ 0.05, **P ≤ 0.01. See also Fig. S6.

Fig. 6

EGCG fermentation inhibits Th1 differentiation via colonic GPR43 in human. Data resources for (A–D). SCFA concentrations in human stools (including non-IBD, Crohn's disease (CD), and Ulcerative colitis (UC) patients) were shown. Untargeted metabolic data were obtained from Project PR000639 in the Metabolomics Workbench (http://www.metabolomicsworkbench.org). (A) Concentrations of SCFA were from peak area and shown as abundance relative to non-IBD. Data are shown in median. n = 9–22. (B) Bulk RNA-sequencing data for biopsies collected 10 cm from rectum are is accessed through GEO Series accession number GSE111889. FFAR2 expressions among non-IBD, UC, and CD patients were shown. n = 21–48. (C) Correlation of FFAR2 gene expression with propionic acid abundance in corresponding UC and CD patients was shown (r = 0.4080, p = 0.0135; Two-tailed Pearson correlation coefficient; Linear regression with 95 % confidence). Th1-related genes in UC and CD patients were scored based on gene set variation analysis (GSVA) enrichment score. The resulting scores were shown. (D) Correlation between the resulting scores for Th1-related genes and propionic acid abundance in corresponding UC and CD patients were analyzed (r = −0.6523, p < 0.0001; Two-tailed Pearson correlation coefficient; Linear regression with 95 % confidence). n = 30. (E and F) EGCG and colon contents were fermented, and the fermentations was centrifuged, then concentrated at low temperature and diluted with cell culture medium. The resulting supernatant was used to treat human colonic epithelial cells (HCoEpiCs) for 24 h after HCoEpiCs were treated with GLPG. Then culture medium was refreshed and followed by 2.5 % (w/v) DSS treatment. The culture medium was refreshed and epithelial cells were cultured for another 24 h. Subsequently, the above Medium was collected to treat the activated human CD4⁺ T cells, which were isolated from PBMC and stimulated with PHA-L. Th1, Th2, Th17, and Treg markers were determined by flow cytometer after cells were stained. Data are representative from one of two independent experiments, n = 3–9. Data are presented as mean ± SEM. Statistical significance was determined using one-way ANOVA, followed by Tukey's multiple-comparison tests or Student's t-test. *P ≤ 0.05, **P ≤ 0.01. See also Fig. S7.

Fig. 7

EGCG inhibits Th1 cell polarization and self-amplification in a manner that required multilevel regulated intestinal microecosystem. In the colitis condition, pathogenic bacteria dominated in the intestinal lumen damage the epithelial lining, causing increased permeability, which allows damaged epithelial cells and activated immune cells to produce pro-inflammatory cytokines, including IFN-γ and IL-12, leading to a promotion of Th1 cell polarization and self-amplification, and contributing to the chronic inflammation in IBD. Upon EGCG treatment, a variety of probiotics and environmental acetic acid/propionic acid are enriched, and a direct binding of acetic acid/propionic acid to the epithelial GPR43 activates down-stream pathways, and subsequently lead to a decrease in IFN-γ release, which promotes the repair of damaged epithelial cells through reducing Th1 cell polarization and numbers.