Copycat innate lymphoid cells dampen gut inflammation

The mechanisms whereby the gut mucosa tolerates trillions of commensal bacteria without developing inflammation remain poorly understood. A recent Science article reveals that gut innate lymphoid cells constrain inflammatory T cell responses to commensal bacteria by adopting a strategy usually deployed by thymic epithelial cells to negatively select self-reactive T cells.

Innate lymphoid cells (ILCs) promote intestinal immunity by releasing distinct sets of cytokines that mirror those produced by CD4⁺ T effector (Teff) cells¹. Unlike ILCs, Teff cells express somatically recombined receptors that recognize antigen in the context of major histocompatibility class-II (MHC-II) molecules expressed on dendritic cells (DCs)¹. The immunostimulating properties of Teff cells are constrained by CD4⁺ T regulatory (Treg) cells expressing the transcription factor Foxp3². These Treg cells maintain gut homeostasis in cooperation with various cell types, including ILCs expressing the transcription factor RORγt and known as ILC3.

ILC3 maintain epithelial integrity and stimulate mucus secretion by releasing interleukin-22 (IL-22) and lymphotoxin^1,3,4, whereas Treg cells mitigate the pro-inflammatory activity of Teff cells by releasing transforming growth factor-β (TGF-β) and IL-10². Of note, these cytokines also stimulate B cell production of immunoglobulin A (IgA), a mucosal antibody class that controls commensal bacteria inhabiting the lumen of the gut⁵.

Unlike ILC3, Treg cells require instructive signals from DCs to exert their homeostatic function. After capturing intraluminal antigen through transepithelial dendrites, CX3CR1⁺ macrophages transfer antigen to CD103⁺ DCs, which migrate to mesenteric lymph nodes (MLNs) to serve as antigen-presenting cells (APCs) for the generation of Foxp3⁺ Treg cells^6,7. Besides MHC-II-restricted cognate T-DC interactions, this process involves release of TGF-β by DCs and conversion of dietary vitamin A into a tolerogenic metabolite known as retinoic acid (RA) in DCs⁷.

Recent studies show that gut ILC3 induce Treg cells by releasing granulocyte monocyte-colony stimulating factor (GM-CSF)⁸. This cytokine elicits expression of RA in CD103⁺ DCs in response to IL-1β from macrophages⁸. Besides inducing Treg cell expansion, ILC3 constrain Teff cell expansion by establishing MHC-II-dependent cognate interactions⁹. However, how do MHC-II⁺ ILC3 exert this suppressive activity? An elegant study by Hepworth et al.¹⁰ shows that MHC-II⁺ ILC3 eliminate commensal-reactive CD4⁺ Teff cells by adopting a negative selection strategy usually deployed by thymic epithelial cells (TECs) to eliminate self-reactive CD4⁺ T cells.

Hepworth et al.¹⁰ crossed mice expressing a TCR transgene specific for either the commensal antigen CBir1 or the food protein ovalbumin (OVA) with mice selectively lacking MHC-II in ILC3. In these mice, intestinal CBir1-specific but not OVA-specific Teff cells undergo spontaneous expansion, indicating that MHC-II⁺ ILC3 regulate Teff cell responses to commensals under steady-state conditions¹⁰. Of note, mice selectively expressing MHC-II on ILC3 only show reduced expansion of adoptively transferred CBir1-specific Teff cells but not Treg cells when transgenic CD4⁺ T cells are activated by CBir1 prior to adoptive transfer¹⁰. These findings indicate that MHC-II⁺ ILC3 constrain Teff cells following their priming by commensal antigens (Figure 1). It remains to be verified whether MHC-II⁺ ILC3 also regulate Teff cell responses to food antigens, including OVA.

Figure 1

ILC3 eliminate commensal-reactive CD4⁺ T cells in healthy but not IBD individuals. Healthy subjects. MHC-II⁺B7− ILC3 from T-B areas of MLNs establish MHC-II-restricted interactions that activate a Nur77-dependent Bim-mediated apoptotic program in commensal-specific CD4⁺ Teff cells. In addition, MHC-II⁺B7− ILC3 “starve” CD4⁺ Teff cells by removing IL-2 from the extracellular environment through surface CD25. Unlike Teff cells, Treg cells express abundant surface CD25 and thus are resistant to IL-2 starvation. IBD patients. ILC3 fail to negatively select commensal-specific CD4⁺ Teff cells due to defects of the molecular machinery supporting MHC-II expression. These defects induce expansion of Th17 cells.

Next, Hepworth et al.¹⁰ determined that MHC-II⁺ ILC3 inhabit the B-T border of MLNs, an area specialized in T cell priming by antigen. Considering that ILC3 crosstalk with CD103⁺ DCs via GM-CSF, MHC-II⁺ ILC3 may capture antigen from CD103⁺ DCs in MLNs⁸. Alternatively, MHC-II⁺ ILC3 may migrate to MLNs after acquiring antigen from CX3CR1⁺ macrophages or CD103⁺ DCs from the gut lamina propria⁶. Accordingly, ILC3 can home to mucosal draining lymph nodes by following chemotactive gradients established by CCR7 ligands¹¹.

Hepworth et al.¹⁰ further found that MHC-II expression by ILC3 is uncoupled from expression of co-stimulatory molecules (Figure 1), even following activation by microbial or inflammatory stimuli. Studies from another group show that splenic ILC3 initiate productive CD4⁺ T cell responses after upregulating both MHC-II and co-stimulatory molecules upon exposure to inflammatory stimuli¹². One possibility that may reconcile these discrepant findings is that ILC3 inhibit co-stimulatory molecule expression in response to tissue-specific conditioning signals emerging from gut but not splenic microenvironments.

Further transcriptional studies revealed that gut ILC3 express MHC-II through a mechanism involving the pIV promoter of class II transactivator (CIITA), a co-activator of MHC-II genes also used by TECs¹⁰. By setting up elegant in vitro ILC3-CD4⁺ T cell co-cultures, Hepworth et al.¹⁰ demonstrated that ILC3 activate a CD4⁺ T cell death program similar to that activated by TECs for the negative selection of developing CD4⁺ T cells. In the thymus, TECs eliminate autoreactive CD4⁺ T cells by presenting self-antigen in the context of MHC-II. The resulting TCR signals activate a Nur77-controlled pathway that upregulates the pro-apoptotic protein Bim. A similar mechanism is also adopted by MHC-II⁺ ILC3 to remove commensal-reactive Teff cells (Figure 1).

However, this is not all. Indeed, Hepworth et al.¹⁰ found that MHC-II⁺ ILC3 express elevated levels of CD25, a high-affinity receptor for the T cell growth factor IL-2. By extracting IL-2 from the extracellular environment via CD25, MHC-II⁺ ILC3 starve CD4⁺ Teff cells, thereby further restraining their expansion¹⁰. This finding may explain why MHC-II⁺ ILC3 spare Treg cells (Figure 1), which express elevated levels of CD25.

Finally, Hepworth et al.¹⁰ found that the gut ILC3 expressed less MHC-II in patients with inflammatory bowel disease (IBD). Concomitantly, these patients also showed more pro-inflammatory gut Teff cells, including T helper 17 (Th17) cells¹⁰. Thus, perturbations of MHC-II expression by ILC3 may cause gut inflammation by impairing the negative selection of CD4⁺ Teff cells reactive to commensals (Figure 1). This raises the possibility that IBD may benefit from drugs capable to restore MHC-II expression by ILC3.