The pathogenesis of Crohn's disease (CD), one of the entities that comprise inflammatory bowel disease (IBD), is ascribed to a multifactorial process in which genetic and environmental factors in concert with immune abnormalities give rise to intestinal inflammation. In that the latter factor (immune abnormalities) is most amenable to treatment, a better understanding of its role in the activation and maintenance of the inflammatory process may give rise to better therapeutic modalities.

In initial studies of human Crohn's disease and corresponding animal models, a T helper type 1 (Th1) cytokine-mediated disease characterized by increased interferon-γ (IFN-γ) production under the guidance of interleukin (IL)-12 was observed (reviewed in ref. 1). More recently, however, the concept that a Th17 response (IL-17-producing cells matured in the presence of IL-23) is sufficient and necessary to support the development of this type of intestinal inflammation has been at the forefront. Nonetheless, the relevance of IL-17 as the proximal effector response of human Crohn's disease requires closer examination, for Certum est quod certum reddi potest (“that is certain which may be rendered certain”).

The initial studies examining the relevance of Th17 responses as a proximal cause of inflammation were in animal models—in particular, the severe combined immunodeficiency (SCID) transfer murine colitis model.2 This model involves the intestinal inflammation that develops in immunodeficient mice after the transfer of naive CD4+ T cells (CD45RBhigh) that differentiate into proinflammatory effector cells in the absence of a mature CD45RBlow regulatory cell population2 or upon injection with an agonist CD40 mAb.3 In initial studies, RAG immunodeficient mice backcrossed with IL-23p19 deficient mice (thus deficient in IL-23) demonstrated no signs of intestinal inflammation upon CD40 mAb administration.3 Recipient mice displayed greatly reduced levels of IL-17 but also manifested decreased production of tumor necrosis factor-α as well as of IFN-γ.

In a recent study along these lines, transfer of naive T cells that lacked IL-23R also failed to induce colitis.4 These recipient mice lacked the accumulation of effector cell populations at the intestinal level—the most significant reduction in a cell population that produced both IL-17 and IFN-γ. However, it should be noted that in mice that failed to develop colitis, a concomitant decrease in a T-cell population producing IFN-γ alone was also observed. It must also be considered that in the earliest SCID transfer colitis studies, complete amelioration of inflammation occurred following the administration of neutralizing antibodies to IFN-γ. Taken together, these studies suggest that, even in the presence of an IL-17 component, concomitant increases in IFN-γ can be observed that may have a more proximal etiology in inflammation, leaving uncertain the role of IL-17 as an effector response.

Further questions concerning the role of IL-17 as an effector response have been raised by recent cell transfer colitis studies in which transfer of T-cell populations from mice deficient in IL-17 led to earlier-onset and more severe disease in recipients, marked by significant increases in IFN-γ-producing cells.5 In a similar vein, colitis induced by intrarectal administration of the hapenating reagent trinitrobenzenesulfonic acid in IL-23p19-deficient mice led to exacerbated disease despite decreased IL-17 levels but accompanying increases in IFN-γ levels.6 These studies demonstrated that a lack of cells producing IL-17 led to increased, not decreased, disease induction. A possible explanation is that Th1 cells may be modulated by IL-17 through IL-17 receptors on their surface and IL-17 may be playing a more regulatory role in limiting the occurrence of inflammation.

Turning to human studies, IL-17A has been expressed in the sera of Crohn's disease patients and both T cells producing IL-17A or IFN-γ and those producing both IL-17A and IFN-γ have been demonstrated.7 In more functional studies, resident mesenteric lymph node dendritic cell populations isolated from Crohn's disease patients, when cocultured with naive T cells, were able to induce IFN-γ secretion to a much larger degree than IL-17 (which was barely detectable).8 Thus, studies in humans are similar to those of the animal models, which suggests that both IL-17 and IFN-γ responses occur in the setting of this intestinal inflammation, whereas IFN-γ may be the more quantitative effector cytokine produced. However, despite these possibilities surrounding IFN-γ, there remains the question as to whether IL-17 is indeed an effector or a protector cytokine in human disease.

The recent results of a therapeutic trial of a neutralizing antibody to IL-17A may shed light on the question. In this double-blinded trial, administration of anti-IL-17A monoclonal antibody to patients with moderate to severe Crohn's disease had no therapeutic effect, and in some patients acute exacerbation of disease was noted, suggesting a protective role of IL-17A (ref. 9). However, it would be beneficial to look back at these patients and assess whether inhibition of IL-17A led to the above-mentioned increases in Th1 responses (or possible effects on other inflammatory IL-17 components, e.g., IL-17F). By contrast, as reviewed in ref. 10, changes in other components of the Th17 responses were observed (e.g., IL-22, a cytokine secreted in concert with IL-17A and thought to have positive effects on the epithelial barrier), giving credence to a more regulatory-like role for modulation of IL-17A responses.

Clearly, all the pieces of the dynamic puzzle have not been assembled. The effects of IL-17A appear to be quite complex and additional studies are needed before the role of this cytokine is assigned with certainty. In doing so, Respice adspice prospice (“examine the past, examine the present, examine the future”), for these answers will come with an open mind.

Ivan J Fuss, Associate Editor