Dubosiella newyorkensis modulates immune tolerance in colitis via the L-lysine-activated AhR-IDO1-Kyn pathway

Abstract

Commensal bacteria generate immensely diverse active metabolites to maintain gut homeostasis, however their fundamental role in establishing an immunotolerogenic microenvironment in the intestinal tract remains obscure. Here, we demonstrate that an understudied murine commensal bacterium, Dubosiella newyorkensis, and its human homologue Clostridium innocuum, have a probiotic immunomodulatory effect on dextran sulfate sodium-induced colitis using conventional, antibiotic-treated and germ-free mouse models. We identify an important role for the D. newyorkensis in rebalancing Treg/Th17 responses and ameliorating mucosal barrier injury by producing short-chain fatty acids, especially propionate and L-Lysine (Lys). We further show that Lys induces the immune tolerance ability of dendritic cells (DCs) by enhancing Trp catabolism towards the kynurenine (Kyn) pathway through activation of the metabolic enzyme indoleamine-2,3-dioxygenase 1 (IDO1) in an aryl hydrocarbon receptor (AhR)-dependent manner. This study identifies a previously unrecognized metabolic communication by which Lys-producing commensal bacteria exert their immunoregulatory capacity to establish a Treg-mediated immunosuppressive microenvironment by activating AhR-IDO1-Kyn metabolic circuitry in DCs. This metabolic circuit represents a potential therapeutic target for the treatment of inflammatory bowel diseases.

Fig. 1. Manipulation of gut microbiota affects mouse susceptibility to DSS-induced colitis.

a Outline of treatment regimens. b–e Wild-type C57BL/6J mice (WT, n = 4) received vehicle (Veh) or oral antibiotics (single Amp, Van, Neo, or Metro or in combination [Abx], n = 6) daily for 5 days and treated with 2.5% DSS in drinking water for 7 days. At day 7 (D7) post-DSS treatment, colon length (b, c), relative mRNA expression of proinflammatory cytokines Il1b, Il6, and Tnfa by qRT-PCR (d) and IL-6 protein level by ELISA (e) were determined in each group. f–h Abx mice were given fecal microbiota transplantation (FMT) from Neo-treated mice twice [Abx-FMT(N)] or Veh with a 48 h interval and then treated with DSS for 7 days (n = 5). WT B6 mice treated with Neo (Neo) were served as control. Colon length (f) was measured, histopathological changes were scored by HE staining (g, h) and infiltration of IL-6-secreting macrophages (Mø) was determined by IFA (h). i–j Principal coordinate analysis (PCoA) (i) and heatmap of the relative abundance (j) of bacteria in fecal samples from mice treated with Abx or single antibiotics at D0 and D7 post-DSS administration. k Relative abundance of Dubosiella and Akkemansia at D0 and D7 among groups treated with different antibiotics (D7. Amp, D7. Van, n = 4; D0. Veh, D0. Abx, D0. Metro, D0. Amp. D0. Van, n = 5; D0. Neo, D7. Veh, D7. Neo, D7. Abx, D7. Metro, n = 6). Dashed lines at 1 indicate that the treatments have equal value as normalized controls. Results are representative of data generated in at least two independent experiments and are expressed as mean ± SEM, and 2-sided P-values were examined by the Student’s t-test. Source data are provided as a Source Data file.

Fig. 2. D. newyorkensis protects mice from DSS-induced colitis and attenuates resulting mucosal inflammation and barrier damages.

a–d Conventional wild-type C57BL/6J mice were colonized with 10⁹ CFU of D. newyorkensis (Dub), A. muciniphila (Akk), E. faecalis (EF) or vehicle (Veh) twice with 2-day break in between, then treated with DSS as previously described (n = 5). At D7 post-DSS administration, colon samples were collected to determine colon length (a), histopathological score (b) by HE staining, IL-6-secreting macrophage infiltration, Muc2 and Occludin expression by IFA, number of goblet cells by Alcian blue staining and microstructure of colonic epithelia by TEM (c). The expression of Occludin and ZO-1 in each group was determined by western blot (WB) (d). e–i Abx-treated WT mice were colonized with 10⁹ CFU Dub, Akk, EF, or Veh and exposed to DSS (n = 5). At D7 post-DSS treatment, colon length (e), histopathological score by HE staining (f, g), IL-6-secreting macrophage infiltration (g) were examined. Meanwhile, the expression of ZO-1, OCLN, and Muc2 in isolated colonic intestinal epithelial cells (h), and the expression of proinflammatory cytokines Il1b, Il6, and Tnfa in colonic lamina propria (cLP) (i) was detected by qRT-PCR. c, j–l Conventional WT mice (n = 6) were orally administered heat-inactivated (HI) Dub, or Dub supernatant (Dub.sup) and exposed to DSS. At D7 post-DSS administration, colon length (j) was measured, histopathological changes were scored by HE staining (c, k, n = 5), level of tight junction proteins Occludin and ZO-1, and expression of proinflammatory cytokines IL-1β, IL-6, and TNF-α were determined by WB (l). Dashed lines at 1 indicate that the treatments have equal value as normalized controls. Results are representative of data generated in at least two independent experiments and are expressed as the mean ± SEM, and 2-sided P-values were examined by the Student’s t-test. Source data are provided as a Source Data file.

Fig. 3. D. newyorkensis prevents DSS-induced colitis by regulating CD25⁺Foxp3⁺ Tregs partially through the SCFA-GPR43 signaling axis.

a Heat map showing mRNA expression determined by bulk RNA-Seq in colon samples from vehicle (Veh)-treated (n = 3) or D. newyorkensis (Dub)-colonized wild-type C57BL/6J mice (n = 4) at D7 post-DSS treatment. b, c CD25⁺Foxp3⁺Tregs (b) and IL-17⁺ CD4⁺ T cells (c) in mesenteric lymph nodes (MLN) and colonic lamina propria (cLP) of conventional or Abx-treated WT mice colonized with Dub, A. muciniphila (Akk), or E. faecalis (EF) (n = 6) were examined by flow cytometry at D7 post-DSS treatment. d–f Conventional WT or Foxp3-DTR mice (n = 5) were colonized with 10⁹ CFU of Dub, Akk, or EF and subjected to DSS treatment. At D7 post-DSS administration, colon length measurement (d) and histopathological evaluation (e, f) were performed. g SCFA concentration in Dub, Akk, or EF culture supernatant determined by GC/MS (n = 4). h–m Conventional WT and Gpr43^−/− mice (n = 4) were colonized with Dub (10⁹ CFU) twice or administrated propionate (Prop) in drinking water for 3 weeks before DSS exposure. At D7 post-DSS administration, CD25⁺Foxp3⁺Tregs (h) and IL-17⁺ CD4⁺ T cells (i) in MLNs and cLP of conventional WT or Gpr43^−/− mice (n = 4) were quantified by flow cytometry. Meanwhile, colon length (j), histopathological score by HE staining (k, l) (n = 4), and expression of ZO-1, OCLN, and Muc2 in isolated colonic intestinal epithelial cells (cIECs) (WT, n = 4; Gpr43^−/−, n = 3) (m) was determined. Fold of change in frequencies of CD25⁺Foxp3⁺Tregs or IL-17⁺ CD4⁺ T cells (b, c, h, i), colon length (d), or SCFA concentrations (g) between groups was calculated and presented with numbers in red. Dashed lines at 1 indicate that the treatments have equal value as normalized controls. Results are representative of data generated in at least two independent experiments and are expressed as mean ± SEM, and 2-sided P-values were examined by the Student’s t-test. Source data are provided as a Source Data file.

Fig. 4. D. newyorkensis promotes CD25⁺Foxp3⁺ Tregs through enhancement of IDO1-mediated Trp metabolism in dendritic cells.

a Heat map showing differential metabolites in colon at D7 post-DSS treatment from wild-type C57BL/6J mice (WT) treated with vehicle (Veh, n = 4) or colonized with D. newyorkensis (Dub), A. muciniphila (Akk), or E. faecalis (EF) (n = 5). Corresponding statistical annotations of differential metabolites represent the distinctions between Dub-colonized and noncolonized mice. b, c T cell subsets in mesenteric lymph nodes (MLN) and colonic lamina propria (cLP) from Veh- or Kyn-treated (10 mg/kg) conventional WT mice were examined by flow (n = 6). d–f WT or DT-treated Foxp3-DTR mice (n = 4) were treated with Kyn and subjected to DSS administration for 7 days, colon length (d) and histopathology (e, f) were examined. g Schematic of the Kyn catabolic pathway of Trp metabolism. h, i Kinetics of colonic IDO1 expression (h) and Kyn serum concentration (i) in Dub- or Veh-colonized conventional WT mice (n = 4) at indicated times post-DSS administration. j IDO1 expression in colonic intestinal epithelial cells (cIECs) or lamina propria mononuclear cells (cLPMCs) extracted from Abx-Dub, Abx-EF, Abx-Veh, or Abx-Dub.sup mice (n = 4) at D3 post-DSS administration. k Ido1 expression in murine bone marrow-derived macrophages (mBMDMs) and bone marrow-derived dendritic cells (mBMDCs) or intestinal organoids treated with Dub or EF culture supernatants for 18 h. l–s Germ-free (GF) mice receiving fecal microbiota transplantation (FMT, n = 6) from conventional WT mice or colonized with Dub (n = 4) were exposed to DSS treatment for 7 days. Ido1 expression in cLPMCs (l), serum concentration ratio of Kyn/Trp (n = 4) (m), T cell subsets (n = 5) (n, o) in MLN and cLP, colon length (p), pathological changes (q, r) (FMT, n = 6; Dub, n = 4) and proinflammatory cytokines in cLPMCs (n = 4) (s) were determined. Fold of change in frequencies of T cell subsets (b, c, n, o) or colon length (d) between groups was calculated and presented with numbers in red. Dashed lines at 1 indicate that the treatments have equal value as normalized controls. Results are representative of data generated in at least two independent experiments and are expressed as mean ± SEM, and 2-sided P-values were examined by the Student’s t-test. Source data are provided as a Source Data file.

Fig. 5. L-Lys enhances Trp metabolism in dendritic cells to induce Treg-dependent immunosuppression by promoting IDO1 expression.

a Heat map showing differentially expressed metabolites of D. newyorkensis (Dub), A. muciniphila (Akk), E. faecalis (EF) and respective cultured supernatants (chopped meat carbohydrate broth, CMC; brain heart infusion medium, BHI; Luria–Bertani medium, LB) at 48 h. b Ido1 expression in mouse bone marrow-derived dendritic cells (BMDCs) treated with N-acetyl-L-Asp (NAA; 1 mM), L-Lys (8 mM), L-lactic acid (Lac; 100 mM), L-α-aminobutyric acid (Abu; 0.1 mM), ketoleucine (4-MOV; 3 mM) and Kyn (0.2 mM) (n = 4). c, d IDO1 expression and colon length of Abx mice treated with metabolites were measured at D3 or D7 post-DSS administration, respectively (n = 5). e IDO1 activity in BMDCs treated with Lys or IFN-γ was determined by Kyn/Trp ratio of cell culture supernatant (n = 4). f–j Conventional WT, DT-treated Foxp3-DTR, or Ido1^−/− mice were colonized with Dub or treated with Lys (20 mg/kg) and subjected to DSS administration for 7 days. T cell subsets (f, g) in mesenteric lymph nodes (MLN) and colonic lamina propria (n = 5) (cLP) colon length (n = 4) (h) and histopathology (n = 5) (i–j) were determined. k Serum Kyn concentration of Lys-treated conventional WT, DT-treated Foxp3-DTR or Ido1^−/− mice before DSS exposure (WT, Foxp3-DTR, n = 4; Ido1^−/−, n = 3). l In vivo imaging of colon from vehicle (Veh)-, Dub-, Lys- or Kyn-treated transgenic IL-17-EGFP mice. m–r Lys- or Veh-treated germ-free mice were exposed to DSS treatment for 7 days and colonic Ido1 expression (n = 5) (m), T cell subsets (n = 6) in MLN and cLP (n, o), colon length (p), histopathology (q) and proinflammatory cytokines (n = 5) (r) were evaluated. s Ido1 expression in Lys- (8 mM) or Dub supernatant (Dub.sup)-treated bone marrow-derived dendritic cells (BMDCs) extracted from GF mice at 18 h post-treatment (n = 4). Fold of change in frequencies of T cell subsets (f, g, n, o) or colon length (h) was calculated and presented with numbers in red. Dashed lines at 1 indicate that the treatments have equal value as normalized controls. Results are representative of data generated in three independent experiments and are expressed as mean ± SEM, and 2-sided P-values were examined by the Student’s t-test. Source data are provided as a Source Data file.

Fig. 6. L-Lys activates the AhR-IDO1 axis in DCs to induce immunosuppressive Treg responses.

a Transcriptomic analysis of mock-, Lys-, IFN-γ-treated mouse BMDCs at indicated times post-treatment (mock, n = 3; Lys 6 h, 18 h, 24 h, IFN 18 h, n = 4). b Ahr expression in Lys- or IFN-γ-treated BMDCs extracted from conventional WT mice (n = 5). c Ahr expression in Lys- or Dub.sup-treated BMDCs extracted from GF mice (n = 4). d Translocation of AhR (red) into the nucleus (DAPI, blue) from cytoplasm (β-tubulin, green) of mouse BMDCs after 6 h treatment with Lys or the AhR agonist 6-formylindolo[3,2-b]carbazole (FICZ). e AhR protein analysis in nuclear and cytoplasmic fractions of mock-, Lys-, or FICZ-treated mouse BMDCs at 6 h post-treatment by western blot. f Expression of Ido1 in mouse BMDCs treated with Lys, FICZ, or the AhR antagonist CH-223191, or a combination of CH-223191 and Lys/FICZ (n = 4). g Expression of Ido1 in mouse BMDCs transfected with negative control (siNC) or AhR-specific small interfering RNA (siAhR) and treated with Lys or FICZ at 18 h post-treatment (n = 4). For the experiments shown in (a–g) above, the following were used: Lys (8 mM), FICZ (300 nM), and CH-223191 (10 μM). For experiments shown in (h–l) below, conventional WT mice (n = 6–8) were treated with Lys (20 mg/kg), CH-223191 (10 mg/kg), or Lys plus CH-223191 and subjected to DSS administration. h Expression of Ido1 in the cLPMCs from Veh- or Lys-treated WT mice at D3 post-DSS administration in the presence of CH-223191 (n = 3). Colon length (i), histopathological changes by HE staining (n = 4) (j), CD25⁺Foxp3⁺Tregs (k) and IL-17⁺ CD4⁺ T cells (n = 6) (l) in MLN and cLP of Veh- or Lys-treated conventional WT mice in the presence of CH-223191 were examined at D7 post-DSS treatment. Dashed lines at 1 indicate that the treatments have equal value as normalized controls. Results are representative of data generated in three independent experiments and are expressed as mean ± SEM, and 2-sided P-values were examined by the Student’s t-test. Source data are provided as a Source Data file.

Fig. 7. The human homolog of D. newyorkensis, Clostridium innocuum, generates L-Lys and protects against DSS-induced colitis by activating IDO1 expression in DCs.

a Phylogenic analysis of the Erysipelotrichaceae family and the Clostridium innocuum isolate. b, c Relative abundance of C. innocuum (Clos) in fecal samples from healthy individuals and patients with inflammatory bowel disease (IBD) (b), and specifically ulcerative colitis (UC) and Crohn’s disease (CD) (c). Cohort PRJNA389280 (healthy, n = 55; IBD, n = 221), cohort PRJNA400072 (healthy, n = 55; IBD, n = 73), cohort PRJNA398089 (healthy, n = 178; UC, n = 155; CD, n = 269). d Levels of Lys in the culture supernatants of Dub (Dub.sup) or Clos (Clos.sup) (n = 3). e qPCR of Ido1 mRNA in the Lys- or Clos.sup-treated human DCs extracted from peripheral blood mononuclear cells at 18 h post-treatment (n = 3). f qPCR of Ido1 mRNA at 18 h post-treatment in wild-type C57BL/6J mouse (WT) bone marrow-derived dendritic cells (BMDCs) treated with 8 mM Lys, 20 ng/mL IFN-γ, Dub.sup, Clos.sup or chopped meat carbohydrate broth (CMC) (n = 4). g Kyn/Trp ratio of Clos.sup-treated, or nontreated BMDCs extracted from WT or Ido1^−/− mice at 18 h post-treatment (n = 4). h Expression of Ido1 in mouse BMDCs treated with Clos.sup, CH-223191, or a combination of CH-223191 and Clos.sup (n = 4). i Expression of Ido1 in mouse BMDCs transfected with negative control (siNC) or AhR-specific small interfering RNA (siAhR) and treated with Clos.sup at 18 h post-transfection (n = 4). j–m Conventional WT or Ido1^−/− mice were colonized with 10⁹ CFU of Clos or treated with Clos.sup and subjected to DSS administration. At D7 post-DSS administration, CD25⁺Foxp3⁺Tregs (j) and IL-17⁺ CD4⁺ T cells (k) in mesenteric lymph nodes (MLN) and colonic lamina propria (cLP) (WT, n = 5; Ido1^-/−, n = 4), colon length (l) and histopathological changes by HE staining (m) were examined at D7 post-DSS treatment (n = 4). (n) Schematic illustration of the D. newyorkensis metabolite Lys protecting against DSS-induced colitis by rebalancing Treg/Th17 response through the activation of the AhR-IDO1-Kyn circuitry. Dashed lines at 1 indicate that the treatments have equal value as normalized controls. Results are representative of data generated in at least two independent experiments and are expressed as mean ± SEM, and 2-sided P-values were examined by the Student’s t-test. Source data are provided as a Source Data file.