Innate lymphoid cells regulate CD4+ T-cell responses to intestinal commensal bacteria

Abstract

Innate lymphoid cells (ILCs) are a recently characterized family of immune cells that have critical roles in cytokine-mediated regulation of intestinal epithelial cell barrier integrity. Alterations in ILC responses are associated with multiple chronic human diseases, including inflammatory bowel disease, implicating a role for ILCs in disease pathogenesis. Owing to an inability to target ILCs selectively, experimental studies assessing ILC function have predominantly used mice lacking adaptive immune cells. However, in lymphocyte-sufficient hosts ILCs are vastly outnumbered by CD4(+) T cells, which express similar profiles of effector cytokines. Therefore, the function of ILCs in the presence of adaptive immunity and their potential to influence adaptive immune cell responses remain unknown. To test this, we used genetic or antibody-mediated depletion strategies to target murine ILCs in the presence of an adaptive immune system. We show that loss of retinoic-acid-receptor-related orphan receptor-γt-positive (RORγt(+)) ILCs was associated with dysregulated adaptive immune cell responses against commensal bacteria and low-grade systemic inflammation. Remarkably, ILC-mediated regulation of adaptive immune cells occurred independently of interleukin (IL)-17A, IL-22 or IL-23. Genome-wide transcriptional profiling and functional analyses revealed that RORγt(+) ILCs express major histocompatibility complex class II (MHCII) and can process and present antigen. However, rather than inducing T-cell proliferation, ILCs acted to limit commensal bacteria-specific CD4(+) T-cell responses. Consistent with this, selective deletion of MHCII in murine RORγt(+) ILCs resulted in dysregulated commensal bacteria-dependent CD4(+) T-cell responses that promoted spontaneous intestinal inflammation. These data identify that ILCs maintain intestinal homeostasis through MHCII-dependent interactions with CD4(+) T cells that limit pathological adaptive immune cell responses to commensal bacteria.

Figure 1. RORγt⁺ ILCs regulate adaptive immune cell responses to commensal bacteria and are enriched in MHCII-associated genes

a–d, Defined age- and sex-matched mouse strains were examined for the frequency of splenic Ki-67⁺ CD4⁺ T cells (top) and CD44^high CD62L^low CD4⁺ T cells (bottom) (a), spleen size (b), spleen weight (c) and relative optical density (OD) values of serum IgG specific to commensal bacteria (d). Antibiotics (ABX) were administered in the drinking water from weaning until 6–8 weeks of age. Scale bar represents 0.5 cm (b). Flow cytometry plots are gated on live CD4⁺ CD3⁺ T cells (a). e, f, DAVID pathway analysis of GO terms enriched in the transcriptional profiles of naïve CD4⁺ T cells and group 3 RORγt⁺ ILCs (e) and heat map of selected lymphoid-associated and MHCII-associated gene transcripts (f). g, h, Gating strategy for ILCs and expression of RORγt and MHCII in ILCs from the mesenteric lymph node of naïve RORγt-eGFP reporter mice (g) and the small intestine of healthy humans (h). Blue line = ILCs, Grey fill = negative control population. Data are representative of 3 independent experiments containing 3–5 mice per group or 4 human donors. Results are shown as the means ± s.e.m. ** p < 0.01, *** p < 0.001 (two-tailed students t-test).

Figure 2. T-bet⁻ NKp46⁻ RORγt⁺ ILCs express MHCII and process and present antigen

a, b, Gating strategy (a) and expression of MHCII (b) in group 3 ILC subsets in the small intestine of naïve mice. c, d Sorted cell populations were cultured in the absence (thin line) or presence of DQ-OVA at 4°C (shaded) or 37°C (thick line) (c) or cultured in the absence (media) or presence of E-alpha-GFP protein and stained with Y-Ae antibody (d). e, Sort-purified CFSE-labeled CD4⁺ T cells from OTII mice were cultured in the presence of media alone or with OVA-pulsed DCs or OVA-pulsed ILCs. f, Expression of co-stimulatory molecules on DCs (black line), ILCs (blue line) or T cells (shaded) from the mLNs of naïve mice. Data are representative of 2–3 independent experiments containing 2–5 mice per group or 2–3 in vitro replicates.

Figure 3. Loss of RORγt⁺ ILC-intrinsic MHCII results in commensal bacteria-dependent intestinal inflammation

a–d, Age- and sex-matched mice were examined for the frequency of splenic Ki-67⁺ CD4⁺ T cells (top) and CD44^high CD62L^low CD4⁺ T cells (bottom) (a), spleen size (b), spleen weight (c) and relative serum IgG specific to commensal bacteria (d). Scale bar represents 0.5 cm (b). Flow cytometry plots are gated on live CD4⁺CD3⁺ T cells (a). e–g, Mice were examined for incidence of rectal prolapse (e), histological changes in H&E stained sections of the terminal colon (f) and colon length (g). Scale bars represent 25 μm (f). Antibiotics (ABX) were continuously administered in the drinking water of selected mice at weaning until 8–18 weeks of age. h–j, Frequency of total IFN-γ⁺, IL-17A⁺ and TNF-α⁺ CD4⁺ T cells (h) and IL-17A⁺ IFN-γ⁺ CD4⁺ T cells (i) in the colons following a brief ex vivo stimulation and frequency of CD11b⁺ Ly6G⁺ neutrophils in colonic lamina propria (j). Data are representative of 3 independent experiments containing 3–5 mice per group. Results are shown as the means ± s.e.m. * p < 0.05, ** p < 0.01, *** p < 0.001 (two-tailed students t-test).

Figure 4. RORγt⁺ ILC-intrinsic MHCII regulates pathologic CD4⁺ T cell responses to commensal bacteria

a–d, Rag1^−/− mice received CD4⁺ T cells sort-purified from defined donor mouse strains or e–g conventional (CNV) and germ-free (GF) Rag1^−/− mice received sort-purified CD4⁺ T cells from MHCII^ΔILC mice (+/−Tx). Recipients were examined for changes in weight (a, e), macroscopic colon pathology (b), colon length (c, f) and histological changes in the terminal colon (d, g). Scale bars represent 0.5 cm (b) or 25 μm (d, g). In some experiments antibiotics (ABX) were administered in the drinking water of donor mice from weaning (a–d). Data are representative of 2 independent experiments containing 3–5 mice per group. Results are shown as the means ± s.e.m. * p < 0.05, ** p < 0.01 (two-tailed students t-test).