Society: AGA

Perturbations in the gut mucosal immune response contributes to IBD. Non-immune cell popopulations including epithelial and stromal cells also play an important role in intestinal inflammation.

Background: In the colon, FoxP3⁺RORγt⁺ regulatory T cells (Tregs) comprise 65% of all Tregs. This population of immunosuppressive cells is absent in germ free conditions, requires microbial antigens, and is involved in controlling intestinal inflammation. It is conceivable that under homeostatic conditions, FoxP3⁺RORγt⁺ Tregs and RORγt⁺ Th17 cells populations in represent phenotypically intermediate or transient subpopulations which are subjected to environmental cues. While certain commensal bacteria and bile acids have been shown to regulate these two T cell populations in the gut, the role of epithelial-derived metabolites in their control is less known. Methods: We developed an epithelial cell-specific knockout of Aldehyde dehydrogenase 1 family, member A1 (Aldh1a1^ΔIEC), a key enzyme in the retinoic acid synthesis pathway. Lamina propria (LP) T cell and DC phenotype was assessed by flow cytometry directly after isolation or after anti-CD3/CD28 activation. Colonoids from Aldh1a1^WT and Aldh1a1^ΔIEC mice were incubated with or without retinol, and conditioned media was then used to treat bone marrow-derived DCs (BMDCs) in the presence/absence of cecal antigen extract (CAE) prepared from naïve WT mice. In some cases, pre-treated BMDCs were washed, and co-cultured with naïve T cells to analyze Treg differentiation. Results: Aldh1a1^ΔIEC mice were healthy and showed no histological signs of spontaneous intestinal inflammation. Compared to Aldh1a1^WT, Aldh1a1^ΔIEC mice had reduced frequencies FoxP3⁺RORγt⁺ Tregs in the colonic lamina propria (7.7 vs 2.9% of CD4⁺ T cells; p<0.001, n=15), a statistical trend towards increase in the IL17⁺FoxP3^-RORγt⁺ Th17 cell population (3.6 vs 9.12% of CD4⁺ T cells; p<0.1, n=3), and CD4⁺IFNγ⁺ Th1 cells (4.5 vs 8.1% of CD4⁺ T cells; p<0.1, n=3). Frequency of tolerogenic CD11c⁺11b⁺CD103⁺ dendritic cells (DCs) was also lower in Aldh1a1^ΔIEC mice (31.7 vs. 14.2%; p=0.03). Organoid conditioned media and CAE additively increased the frequencies of CD103⁺ DCs and FoxP3⁺RORγt⁺ Treg differentiation in vitro. Compared to Aldh1a1^WTcolonoids, conditioned media from Aldh1a1^ΔIEC colonoids (with or without CAE) were less potent in imprinting tolerogenic CD103⁺DCs and downstream differentiation of naïve T cells into FoxP3⁺RORγt⁺ Tregs. Conclusions: Epithelial Aldh1a1 activity is a key regulator of mucosal tolerogenic CD11c⁺11b⁺CD103⁺ DC and FoxP3⁺RORγt⁺Treg populations in synergy with the gut luminal antigens.

BACKGROUND: Innate immune and stromal cells contribute to the pathogenesis of IBD. To better understand how these cells contribute to IBD, we have generated a model of spontaneous colitis that occurs in the absence of adaptive immune cells. This colitis is 100% penetrant, driven by microbes, resistant to anti-TNF therapy and prevented by anti-Thy1.2 treatment. We therefore investigated the contributions of innate lymphoid cells, TNF and stromal cells to the colitis observed in these mice.
METHODS: villin-TNFAIP3 x RAG1^-/- (TRAG) mice that develop 100% penetrant colitis by 6 weeks of age were used to investigate the roles of cytokines and innate cell types in innate colitis. To investigate innate cell types, TRAG mice were crossed with mice lacking ILC3 (Rorγt^-/-), NK cells (IL15Rα^-/-), or all ILC types (IL2Rγ^-/-). To investigate cytokine requirements, TRAG mice were crossed to mice lacking IFNγ (IFNγ^-/-), IL12 signaling (STAT4^-/-), type-I interferon signaling (INFαR^-/-), or TNF (TNF^-/-). The role of stromal cells was investigated by treating TRAG mice lacking ILCs (TRAG x IL2Rγ^-/-) mice with anti-Thy1.2 antibody. Cell death and related signaling pathways were evaluated by quantitative immunohistochemistry. Cell populations were investigated by flow cytometry of LPLs.
RESULTS: The colitis in TRAG mice persisted in TRAG x Rorγt^-/- and TRAG x IL15Rα^-/- mice, indicating that ILC3 and NK cells are dispensable for this innate colitis. Similarly, colitis persisted in TRAG x IFNγ^-/-, TRAG x STAT4^-/- OR TRAG INFαR^-/-mice, indicating the the innate colitis in this model does not require IFNγ or either IL12 signaling or type-I interferon signaling. Colitis was prevented in TRAG x TNF^-/- mice, indicating that TNF is required for colitis in this model. The TNF-dependent colitis was independent of cell death but associated with increased activation of phospho-MLKL, the executioner of the necroptosis pathway. Although necroptosis was not correlated to colitis, the release of IL1, which is mediated by phospho-MLKL, was associated with TNF-dependent innate colitis. Colitis persisted in the absence of adaptive immunity and all ILCs in TRAG x IL2Rγ^-/- mice. Treatment of TRAG x IL2Rγ^-/- mice with anti-Thy1.2 antibody prevented colitis, suggesting that Thy1.2^+ve stromal cells contribute to the colitis in this model.
CONCLUSION: Our model suggests that TNF drives this innate colitis through a non-cell death mechanism mediated by release of IL1. This results in colitis that is refractory to anti-TNF therapy and promoted by stromal cell remodeling involving expression of Thy1.2, which is known to interact with myeloid cells expressing αVβ3. We are therefore investigating whether Thy1.2 interactions with innate immune cells may be a therapeutic avenue for colitis treatment that is refractory to anti-TNF therapy.

Background: IL-22 is a critical cytokine in maintaining epithelial homeostasis. Lack of IL-22 impedes mucosal healing and promotes metabolic syndrome. Fermentable fibers, including inulin, drive IL-22 production by an undefined microbiota-dependent mechanism. Bacterial flagellin is one of few known bacterial metabolites capable of inducing IL-22 but inulin-induced IL-22 is independent of this flagellin recognition suggesting existence of others.
Aim: Identify microbiota metabolites from fiber-fed mice capable of activating IL-22 production.
Results: We developed a tractable ex vivo model, using freshly isolated Peyer’s patch cells. Such cells produced copious IL-22 in response to fecal supernatant (FS) of mice fed diets enriched with inulin but not cellulose mimicking established in vivo findings. FS from germ-free or antibiotic-treated did not induce IL-22 irrespective of diet. In vitro interrogation of FS revealed that its IL-22 inducing activity was sensitive to heat and proteinase. These results suggested that inulin elicitation of IL-22 was mediated by bacterial protein(s). To reduce the protein complexity of FS, we generated FS from altered Schaedler flora (ASF) mice, whose microbiome consists of only eight strains but was nonetheless capable of inducting IL-22. Hence, ASF FS was subjected to ion exchange chromatography. Eluate fractions were screened for IL-22 inducing bioactivity and then assayed by SDS-PAGE. A 34kDa that associated with IL-22 induction was extracted, subjected to proteomic analysis, which indicated it was likely outer membrane protein A (ompA). 16S analysis revealed the microbiome of inulin-fed ASF mice was dominated by Parabacteroides goldsteinii, whose genome encodes 7 OmpA proteins, 4 of which we generated recombinant his-tagged proteins 2 of which strongly induced IL-22 production in our ex vivo model and in vivo when administered to mice. OmpA-induced IL-22 production was independent of IL-23 indicating it is mediated by a novel signaling pathway.
Conclusion: OmpA is one common bacterial protein capable of inducing robust IL-22 and thus may be an important link between diet, mucosal immune homeostasis, and subsequently intestinal and metabolic health.

Jun Zou

Appearances