Kevin J. Maloy, Sir William Dunn School of Pathology, University of Oxford, South Parks Road, Oxford OX1 3RE, UK. (fax: 44 1865 275591; e-mail: firstname.lastname@example.org).
Although the precise aetiology of inflammatory bowel disease (IBD) remains unclear, recent discoveries have led to an improved understanding of disease pathogenesis. Whilst these findings have underscored the central role of innate and adaptive immune responses in intestinal inflammation, they have also precipitated a paradigm shift in the key cytokine pathways that drive disease. The prevailing dogma that IBD was mediated by interleukin (IL)-12-driven T-helper (Th)1 CD4 T cell responses towards the bacterial flora has been largely dispelled by findings that the closely related cytokine IL-23 appears to be the key mediator of intestinal inflammation. IL-23 is associated with a novel subset of IL-17-secreting CD4 T cells termed Th17 cells and rodent studies have implicated the IL-23/IL-17 axis in autoimmune inflammation. Genome-wide association studies in IBD patients have confirmed the predominant role of the IL-23 pathway, indicating that this could represent an important future therapeutic target.
Human inflammatory bowel diseases (IBD) encompassing Crohn’s disease (CD) and ulcerative colitis are chronic inflammatory disorders of the intestinal tract of unknown aetiology for which no current cure exists. Although multiple host and environmental factors may contribute to the development of IBD, it is widely accepted that the intestinal pathology is mediated by aberrant immune responses directed against components of the intestinal microflora [1, 2]. This hypothesis was strengthened by the discovery that mutations in NOD2, a host innate immune receptor involved in the detection of bacterial peptidoglycan were associated greatly with an increased risk of CD [3, 4]. However, despite intense investigation, the precise mechanism through which NOD2 mutations may predispose to CD remains unresolved [5, 6].
A variety of inflammatory cytokines have been associated with IBD, including interleukin (IL)-1, IL-6, tumour necrosis factor (TNF)-α and interferon (IFN)-γ, suggesting that the disease may be amenable to treatment with biological therapies to target these host cytokines. In fact, anti-TNFα antibody therapy has proven to be an efficacious treatment for steroid-resistant CD and is now widely used clinically . Although at present there is little evidence on the long-term consequences of blocking these cytokines, one potential problem is that their role in a wide range of protective immune responses suggests that there may be an increased risk of infections or tumours. In fact, there is currently a great deal of controversy over reports of possible increased incidences of infection or lymphoma development in patients treated with anti-TNFα [8–10]. There is therefore a continued need to gain a better understanding of the pathogenesis of IBD, so that future therapies can be improved to selectively target the intestinal inflammation whilst preserving systemic immune protective functions.
IL-23 plays a key role in chronic intestinal inflammation
Until relatively recently, it was widely believed that the chronic intestinal inflammation characteristic of human IBD was the consequence of pathogenic T helper (Th)1 CD4 cell responses against the bacterial microflora, which were in turn driven by the pro-inflammatory cytokine IL-12 . Animal models of IBD supported this hypothesis as intestinal inflammation could be blocked by treatment with monoclonal antibodies specific for the IL-12p40 subunit [12, 13]. This work led to clinical trials, using anti-IL-12 antibodies in CD patients, that have yielded promising results . In parallel, other investigations revealed that the IL-12 family of heterodimeric cytokines contained at least two additional members, IL-23 and IL-27 . IL-23 comprises the IL-12p40 subunit complexed with a novel IL-23p19 subunit and signals through a distinct receptor [16, 17]. Like IL-12, IL-23 is primarily produced by activated dendritic cells in response to microbial ligands or activated through CD40 . The realization that anti-IL-12p40 antibodies would neutralize both IL-12 and IL-23 led to re-evaluation of their relative roles in many inflammatory disorders, including IBD.
Subsequent studies in animal models of IBD have revealed that IL-23 plays a key role in chronic intestinal inflammation. Selective depletion of IL-23 using monoclonal antibodies specific for the p19 subunit or by genetic ablation of the p19 gene greatly attenuated T cell-dependent colitis in T cell transfer models of colitis [18, 19] and also inhibited spontaneous colitis development in IL-10−/− mice . In addition, IL-23 was also shown to play an essential role in the induction of intestinal inflammation by innate immune mechanisms. Thus, IL-23 was required for development of typhlitis and colitis in lymphocyte-deficient Rag−/− mice either following infection with the pathogenic bacterium Helicobacter hepaticus , or following treatment with agonistic anti-CD40 monoclonal antibodies . By contrast, targeted depletion of IL-12 had no effect on the development of either innate immune-mediated or T cell-dependent intestinal inflammation [18–21].
Another important finding from these IBD models was that IL-23 expression was highly increased in the inflamed intestine but not in systemic sites such as the spleen and liver and that systemic immune pathology was not abrogated in the absence of IL-23 [18, 21]. The selective role of IL-23 in mucosal inflammation makes it an attractive therapeutic target from the perspective of ablating intestinal immune pathology whilst sparing systemic immune responses. Recent studies indicated that treatment with anti-IL-23p19 antibodies was able to cure established colitis in mice  and also that the beneficial effects of anti-IL-12p40 antibody therapy in human CD patients were associated with a reduction in both IL-12 and IL-23 . Importantly, the central role of IL-23 in chronic intestinal inflammation revealed in the mouse IBD models has been extrapolated into human IBD patients by recent extensive genome-wide association studies of large cohorts of CD patients and healthy controls. These studies revealed that several polymorphisms in the IL-23R gene locus were associated with either susceptibility or resistance to CD [24, 25]. Although the mechanisms through which IL-23R polymorphisms influence susceptibility to CD remain to be elucidated, these findings highlight the future potential for therapeutic manipulation of the IL-23 pathway in CD.
IL-23 / IL-17 axis in mucosal immunity
Early studies of IL-23 indicated that it was closely associated with a novel subset of pro-inflammatory CD4 T cells that secreted high levels of IL-17 which have been subsequently termed as T helper (Th)17 cells [26–28]. Subsequent work has demonstrated that IL-23 is not required for the differentiation of Th17 cells, but that it’s role appears to be in the maintenance or expansion of Th17 cells [29–31]. The demonstration that their development was directed by a unique set of transcription factors [32, 33] confirmed that Th17 cells represent a distinct lineage of pro-inflammatory CD4 T cells that have been strongly associated with autoimmune tissue pathology [27, 28]. Studies in murine models of autoimmune disease, such as experimental autoimmune encephalitis and collagen-induced arthritis indicated that IL-23 played a critical role in autoimmune disease and that this correlated with an increased accumulation of Th17 cells in the inflamed tissues [34–37]. Consequently, there is also a great deal of current interest in manipulation of this inflammatory axis for the treatment of human autoimmune diseases.
In terms of mucosal immunity, the IL-23/IL-17 axis has been shown to play an important role in defence against pulmonary infections . This was first demonstrated for extracellular bacterial infections, such as Klebsiella pneumoniae, where IL-23−/− mice and IL-17RA−/− mice greatly exhibited increased susceptibility and mortality compared with wild-type mice [39, 40]. It is thought that this protective response is related to the potent ability of IL-17 to promote granulopoeisis and neutrophil recruitment to the lung in response to infection . Additional studies have indicated important roles for IL-23 and/or IL-17 in protection against a range of pulmonary infections, including Mycoplasma pneumoniae  Mycobacterium tuberculosis  and Bordetella pertussis ; as well as in systemic protection against Candida albicans . In these cases, the protective effects of IL-17 may also be related to the elaboration of chemokines that help recruit Th1 effector cells .
With respect to the gastrointestinal tract, there is emerging evidence that the IL-23/IL-17 axis plays a role in normal intestinal homeostasis, although the precise actions of these cytokines in the gut remain to be fully delineated. For example, in healthy normal mice, IL-23 is expressed by lamina propria dendritic cells in the terminal ileum  and the frequency of Th17 cells in the intestinal lamina propria is markedly higher than their frequency in peripheral lymphoid tissues [30, 32]. Despite its pathogenic role in IBD, IL-23 may have important immune protective effects in the intestine. Thus, IL-23−/− mice exhibited enhanced susceptibility and mortality following infection with the intestinal bacterial pathogen Citrobacter rodentium . Interestingly, C. rodentium infected IL-23-deficient mice still generated potent mucosal Th17 responses, suggesting that IL-23-mediated protective responses need not necessarily involve IL-17 . Similarly, although the studies in various murine IBD models noted that IL-23-driven intestinal pathology was associated with increased IL-17 production, a plethora of other inflammatory cytokines were also elevated in the inflamed colon, including IL-1β, IL-6, IFN-γ and TNF-α [18, 19, 21]. Furthermore, in many of the T cell-dependent IBD models, Th1 cells clearly predominated in inflamed lamina propria and inhibition of Th1 responses attenuated disease [18, 19, 46]. Lastly, the observations that IL-23 also drives chronic colitis mediated by cells of the innate immune system are also consistent with the hypothesis that IL-23-mediated intestinal inflammation need not necessarily involve Th17 cells [18, 21]. The distinction between IL-23 and Th17 responses in intestinal pathology may have important therapeutic implications.
The exact role of IL-17 and Th17 cells in intestinal pathology and homeostasis is currently not well understood. IL-17 may have some protective functions in the epithelial layer, as it has been shown to fortify tight junction formation between epithelial cells in vitro  and treatment of mice with anti-IL-17 neutralizing antibodies actually enhanced the severity of colitis induced by administration of dextran sodium sulphate . By contrast, a recent study comparing the ability of Th1 and Th17 cells to induce colitis in mice found that the Th17 cells were significantly more pathogenic than their Th1 counterparts . Furthermore, it was recently reported that IL-17 secreting cells could be isolated from the gut of patients with CD . In this respect, it is vital to bear in mind the important distinction between Th17 cells and IL-17. In addition to secreting IL-17 (IL-17A) and the closely related IL-17F, Th17 cells may produce a range of other pro-inflammatory cytokines including IL-22 and IL-21 [27, 28]. Thus, whilst IL-17 may be a good marker for the Th17 cell lineage, other factors may contribute to their pathogenic or protective effects. For example, although IL-22 with IL-17 has been shown to synergize the production of anti-microbial peptides by keratinocytes , Th17 cell-derived IL-22 has also been implicated in driving the dermal pathologies that characterize autoimmune psoriasis .
So far, there have been few studies that have employed selective blockade of IL-17 during intestinal inflammation. However, in IL-10−/− mice, treatment with anti-IL-17 specific antibodies had little impact on colitis unless anti-IL-6 antibodies were also co-administered , suggesting that IL-17 may synergize with other inflammatory mediators in the gut. Recent studies have highlighted further potential heterogeneity within Th17 cell populations by demonstrating that some may even secrete IL-10 , a factor known to inhibit intestinal inflammation. Thus, it is possible that the actions of Th17 cells may differ depending on other factors that may be present in the local environment. In the normal intestine, the primary function of Th17 cells may be like sentinels who contribute in maintaining epithelial barrier function, whereas in sites of chronic intestinal inflammation, high levels of IL-23 may activate their full pathogenic and anti-bacterial functions. Similar paradigms of context-dependent divergent activities have also been observed for other cytokines. For example, whilst transforming growth factor (TGF)-β has been demonstrated to drive the development of regulatory T cells (Treg) [53, 54] and to contribute in their suppressive functions in the intestine , the presence of pro-inflammatory cytokines such as IL-6 can override Treg suppression  and act together with TGF-β to direct the development of Th17 cells [29–31].
Another important point is that Th17 cells are not the only sources of IL-17, as a range of other cells have been shown to be able to produce IL-17. For example, CD8+ T cells, γδ T cells and natural killer T cells have all been identified as important sources of IL-17 during infection with different pulmonary pathogens [57–59]. Furthermore, during innate immune-mediated intestinal inflammation, increased IL-17 expression was observed in neutrophils and monocytic cells isolated from the intestinal lamina propria . Lastly, in human IBD patients, the increased expression of IL-17 in the inflamed mucosa was attributed to both T cells and CD68+ monocytes/macrophages . Together, these results indicate that a variety of haematopoetic cells may act as sources of IL-17 and thus that IL-17 may have important functions in inflamed tissues that are independent of Th17 cells.
Outlook and translational perspectives
The recent discoveries indicating a key role for IL-23 in chronic intestinal inflammation in mouse IBD models, together with the identification of IL-23R polymorphisms that influence susceptibility to CD, make it an attractive target for future IBD therapies. This represents a paradigm shift away from thinking of IBD as primarily being mediated by IL-12-driven Th1 responses. The selective expression of IL-23 in inflamed intestine suggests that targeting the IL-23 pathway may have the potential benefit of sparing host systemic immune protective pathways, rather than acting as a global immune suppressant. Recent studies have indicated that IL-23 is a potent tumour-promoting factor and have also described elevated IL-23 expression in human colorectal tumours , suggesting that inhibition of IL-23 may also have potential as a prophylactic treatment for inflammation-driven colon cancer in IBD patients. As with all therapies, this optimistic outlook must be tempered by consideration of possible disadvantages of IL-23 blockade such as the potential for increased infections at mucosal surfaces. In addition, some murine IBD models have identified IL-23-independent pathways of intestinal inflammation  which suggests that targeting IL-23 may best be employed as part of specific patient-tailored therapies.
Further studies in both mouse models and human IBD patients are required to more clearly define the roles of IL-17 and Th17 cells in intestinal homeostasis. In addition, findings in mouse IBD models suggest that the IL-23/IL-17 axis of pathology that has been strongly implicated in mouse-models of tissue-specific autoimmune disease may not operate in exactly the same manner in the intestine. The effects of IL-23 clearly extend further than simply enhancing Th17/IL-17 responses in the gut and instead appear to stimulate a range of innate and adaptive immune responses. Emerging evidence suggests that the effects of IL-17 and other Th17 cell derived factors may be context-dependent, promoting either protective or pathogenic responses depending on additional factors present in the local tissue (Fig. 1). An increased knowledge of these cytokine networks in normal and inflamed intestine may identify additional opportunities for attenuation of chronic inflammation and for stimulation of protective immune responses for mucosal vaccination.
Conflict of interest statement
The author declares no conflict of interest.
KM is supported by a Wellcome Trust Research Career Development Fellowship.