The immune response in the intestine is a delicate balance between effector and regulatory T cell responses. Recent studies have shown that IL-23 plays a key role in this balance and is a necessary factor for the development of T cell-dependent and -independent chronic intestinal inflammation (Elson et al., 2007; Hue et al., 2006; Kullberg et al., 2006; Uhlig et al., 2006b; Yen et al., 2006
). The resistance of IL-23-deficient mice to colitis has been attributed to a reduction in pathogenic T cell responses, particularly those mediated by Th17 cells. Yet, in accordance with a previous report, we find that IL-17 production is not required by T cells to induce colitis (Noguchi et al., 2007
). In addition, the absence of IL-23 did not significantly alter the intestinal amounts of IL-17 or the relative expression of RORγt in this model, suggesting that IL-23 can promote intestinal inflammation independently of its role in promoting Th17 cells. We show here that colitogenic T cell responses are retained in the absence of IL-23 but are masked by dominant IL-10- and TGF-β-mediated suppression. Furthermore, transfer of naive T cells to Il23a−/−Rag1−/−
mice fails to elicit colitis and is associated with an increase in the frequency of CD4+
cells in the intestine. These Foxp3+
cells appear to play a functional role in protection from IL-23-independent inflammation because transfer of Foxp3-deficient T cells to Il23a−/−Rag1−/−
hosts induces severe colitis, indistinguishable from disease induced after T cell transfer into IL-23-sufficient Rag1−/−
recipients. These results newly identify an important role for IL-23 in restraining local Treg cell responses in order to permit development of tissue inflammation.
An important concept to emerge from these studies is that factors may promote inflammation not only via direct effects on inflammatory mediators but also indirectly by impeding regulatory mechanisms. Precedence for this idea comes from studies on the role of IL-6 in pathogenic Th17 cell responses in the central nervous system. Originally thought to be required for the development of Th17 cell responses, it was recently shown that IL-6 promotes Th17 cell responses in part by alleviating Treg cell-mediated control (Korn et al., 2007
). Our data strongly suggest that IL-23 could be playing a similar role in the intestine. The fact that Foxp3-deficient T cells can induce high levels of colitis in the absence of IL-23 provides direct evidence that IL-23 is not essential to the pathogenesis of intestinal inflammation, if regulation is absent. Similarly, a number of key proinflammatory cytokines, including IL-6, IFN-γ, and TNF-α, can be expressed in the intestine in the absence of IL-23. Currently, it is not known whether the effects of IL-23 are mediated directly on T cells or act via effects on non-T cells that produce factors that inhibit Treg induction. IL-23 has been suggested to act directly on T cells to inhibit Foxp3 expression (Zhou et al., 2007
). However, our in vitro experiments did not show a direct effect of IL-23 on TGF-β-mediated Foxp3 induction. It must be borne in mind that the effects of IL-23 seem specific to the intestine and could therefore require a particular environment different from conventional cell-culture conditions. Alternatively, IL-23 could have an indirect effect on Treg cell generation. Several cytokines have been described to directly inhibit Foxp3 induction, including IL-6, IL-21, and IL-27 (Bettelli et al., 2006; Korn et al., 2007; Mangan et al., 2006; Nurieva et al., 2007; Zhou et al., 2007
). Although these cytokines could be involved in the control of Foxp3 generation by IL-23, IL-23 is not essential for their expression, because they are upregulated in inflamed colons from IL-23-deficient recipients. This suggests that IL-23 represents a distinct tissue-specific pathway to control Treg cell induction. Experiments with cell type-specific deletions of the IL-23R are required to dissect the individual components of this pathway in vivo.
By controlling Foxp3-mediated regulation, IL-23 may affect other pathways than Th17 cell responses. Because of the role of IL-23 in promoting IL-17 production, Th17 cells have been considered to play an important role in IL-23-dependent pathologies. This, however, may not be true in all cases. Although our data do not exclude the possibility that IL-23 directly sustains Th17 cells, they offer an additional explanation for its proinflammatory effects. Defects in IL-23 or IL-23R could lead to an increase or decrease in immune suppression that could affect not only Th17 cells, but also Th1 cells, Th2 cells, and innate immune responses. Indeed, a role for IL-23 in promoting non-Th17 cell responses has already been suggested by others. Thus, MOG-specific T cells from Il23a−/−
mice have reduced amounts of IFN-γ, and systemic IL-23 can enhance Th1 cell antitumor responses (Kaiga et al., 2007; Thakker et al., 2007
). Interestingly, both IL-23-deficient and -sufficient colitic mice showed increased amounts of intestinal IFN-γ, indicating a strong local Th1 response. CD4+
T lymphocytes and non-T cells have been identified as sources of intestinal IFN-γ in different models of colitis (Hue et al., 2006; Uhlig et al., 2006a
). Indeed, the Th1 response has been shown to be involved in the pathogenesis of T cell transfer-mediated colitis and both IFN-γ and the Th1 cell-specific transcription factor T-bet play functional roles (Neurath et al., 2002; Powrie et al., 1994
). Similarly, CD in humans has been linked to exacerbated Th1 cell responses (Cobrin and Abreu, 2005
). IL-12 is known to play a pivotal role in the control of Th1 cell responses. However, its requirement for the development of inflammation varies depending on the model. By contrast with chronic models, acute inflammation linked to intestinal injury is IL-12-dependent and inhibited by IL-23 (Becker et al., 2006
). Clearly, further studies are required to fully characterize the contributions of these two cytokines to intestinal inflammation in different models and in particular in human disease. Nevertheless, by controlling Foxp3, IL-23 could be permissive for the development of both Th1 and Th17 cell responses.
Strikingly, IL-23 reduced the percentage of naive T cell-derived Foxp3+
but had little impact on the frequency of the progeny from already developed Treg cells. Foxp3+
Treg cells can develop in the thymus (the so-called naturally arising Treg cells), but also be induced in the periphery. Recent data from our group and others have shown that Foxp3+
Treg cells can be induced in the intestine by a mechanism depending on TGF-β and retinoic acid (Coombes et al., 2007; Mucida et al., 2007; Sun et al., 2007
). These Treg cells are induced by a specific subset of dendritic cells (DCs) expressing CD103, which is enriched in the GALT. In contrast to the CD103+
DCs isolated from the MLN do not induce Foxp3+
cells. Importantly, they express high amounts of IL-23p19 mRNA compared to the Treg cell-inducing CD103+
subset upon CD40 stimulation (Coombes et al., 2007
). It is tempting to speculate that IL-23, together with TGF-β and RA, is one of several factors that decide whether a naive T cell will become regulatory and induce dominant tolerance toward its cognate antigen. Other factors could include Th1 cell-related cytokines. A recent report signaled no increase of CD4+
cells in Il12b−/− Ifng−/−
mice, which lack both IL-12 and IL-23 (Wang et al., 2007
). The analysis was, however, performed on the spleen, not on intestinal cells, where we find the highest effect of IL-23. More research is required to elucidate the contribution of different cytokines to the homeostasis of Treg cells under steady-state conditions.
It should be noted that although immune regulatory pathways in the intestine prevent hyperreactivity toward dietary antigens and harmless commensal flora, they can prove a double-edged sword. It is imperative to the host to mount protective responses toward intestinal pathogens, and hence it is necessary to temporarily overcome dominant suppression and Treg cell activity. IL-6 has been identified as one such inflammatory mediator that desensitizes T cells to Treg cell-mediated suppression (Pasare and Medzhitov, 2003
). IL-23, via its ability to impede Treg cell responses in the intestine, may promote host-protective immunity at this site. In support of this, IL-23-deficient mice, unlike wild-type mice, do not develop severe colitis after infection with the intestinal pathogen C. rodentium
, but they fail to clear the bacteria and die within 2 weeks (Mangan et al., 2006
In the years since the identification of IL-23, evidence has accumulated indicating that IL-23 orchestrates different aspects of the immune response in tissues. In addition to its proinflammatory action on the innate immune system and its proposed role in sustaining Th17 cell responses, we have identified the overcoming of Foxp3-mediated regulation as another key function of IL-23 during immune responses. Although the relevance of this mechanism in other experimental models and human disease remains to be ascertained, this could have important implications for both the understanding of mucosal immunity and designing therapeutic approaches to IL-23-dependent diseases.