Inflammatory bowel disease and inflammatory molecules: Chickens, eggs and therapeutic targets

See article in volume 16, issue 11, pages 1228–1234 The gastrointestinal immune system is the largest collection of immune cells in the body. It is juxtaposed between the hostile external environment of the gut and the benign internal milieu that it protects. It deals with a vast array of luminal antigens in the form of food and bacteria in a unique, underplayed fashion. The immunity thereby generated is distinct in its nature and outcome from the systemic variety that we usually think of when we consider antigen-driven responses. The characteristic appearance of the normal intestine, unlike other unperturbed tissues, is one of controlled inflammation.1 What is most remarkable about this, however, is that full-blown inflammatory bowel disease (IBD) is as uncommon as it is. This is related to constitutive, active downregulation of both the innate immune system2 and antigen-specific immunity3 in the gut. Our understanding of the immunopathogenesis of IBD has advanced immensely over the last decade. A large number of inflammatory mediators present in excess in IBD tissues have been characterized. This has made it possible to successfully introduce specific, biologically directed therapies, such as monoclonal antibodies against tumor necrosis factor-α in Crohn's disease.4 An immunologist's dream of novel therapies is in the wings, in phase II and III trials. Indeed, so-called non-specific therapies including corticosteroids and aminosalicylates also target different molecules and pathways in the inflammatory cascade. Most of these inflammatory mediators are probably upregulated in response to the phlogistic environment in which they are expressed. Nevertheless, the efficacy of therapies directed against them indicates that, analagous to peptic ulcer disease before the discovery of Helicobacter pylori, targeting a final common pathway can be an effective approach to treatment of diseases of unknown etiology. A causative inflammatory defect in IBD is suggested by the nature of the newly described gene mutated in a proportion of familial and sporadic Crohn's disease patients.5–7Nod2 is a monocyte-specific signaling molecule in the nuclear factor-κB activation pathway that binds lipopolysaccharide, an essential constituent of the cell wall of Gram-negative bacteria, as well as other bacterial products.8 The mechanism by which this defect causes Crohn's disease is uncertain but it is presumably related to either a defect in innate immune response to gut bacteria,9 or to an abnormality in apoptosis regulation.5,10 This is an exciting discovery because it will allow dissection of interactions between bacteria, innate immunity and IBD, as well as direct exploration of other candidate genes in both familial and sporadic IBD. In addition to defects in innate immunity, there is also powerful evidence that IBD is dependent on antigen-driven immunity gone awry. The immune outcome of tolerance, which distinguishes the usual response to gut bacteria, is lost in IBD and is typically replaced by a polarized, T-helper type 1 (Th1) immune response and accumulation of activated CD4+ T cells. Th1 cytokines including interleukin (IL)-12 and γ-interferon are prominent in human IBD tissues11 and are the focus of therapeutic trials of anti-IL-12 antibodies currently underway. Animal models of IBD provide the best evidence for an antigen-dependent pathway to inflammation: virtually all of the many models of IBD rely on population of the intestine by normal flora, with no inflammation seen in germ-free mice. In fact, reintroduction of isolated strains of bacteria to the intestine of germ-free mice recapitulates the inflammation seen in normally colonized animals. Further evidence for a requirement for T-cell-dependent antigen recognition comes from a number of genetically altered mice which spontaneously develop colitis. These include animals that express defective T cell receptors (TCR)12 or have aberrant post-TCR signaling,13 and immunodeficient mice reconstituted with effector T cells.14 For productive T-cell activation to occur in response to antigen recognition, TCR signaling must be accompanied by ligation of costimulatory molecules on the surface of antigen-presenting cells and T cells. One such group of costimulatory molecules is the B7 family of peptides, found on antigen-presenting cells.15 These costimulatory molecules most commonly ‘meet’ CD28 on the T cell, although other interactions can also occur. Grose et al. in a previous issue of the Journal, highlight the potential involvement of the B7 molecules, CD80 and CD86, in the colitis lesions induced in mice that have ingested dextran sulfate sodium (DSS).16 Dextran sulfate sodium induces an acute colitis with epithelial ulceration, reminiscent of ulcerative colitis in humans. It is a model that has been used to study acute inflammatory events as well as later, more chronic changes such as dysplasia and carcinoma.17 It is probably dependent on macrophage ingestion of DSS, followed by macrophage activation, tissue destruction and subsequent T-cell activation. The paper by Grose et al. examines the expression of CD80 and CD86, using immunofluorescence staining, in the DSS-induced lesions.16 It demonstrates a markedly heightened display that occurs several days after the onset of severe inflammation and it implies that this expression is by upregulation on infiltrating macrophages and, to a lesser degree, on B cells. Somewhat surprisingly, dendritic cells, the classical antigen-presenting cells most capable of activating T cells, were not seen in these tissues and Grose et al. speculate that infiltrating macrophages are induced to become the chief antigen-presenting cells in this colitis model. The findings presented bear some similarities to earlier demonstrations of B7 family expression in human IBD,18,19 and in another murine model of colitis.20 The authors speculate that CD80 and CD86 may therefore become novel therapeutic targets in the search for specific anti-IBD agents, although recent data indicate that interfering with costimulatory molecules might simply shift the target tissue for autoimmune processes.21 There are, of course, unanswered questions, for example: how important is antigen presentation in the lamina propria of the colon? How much occurs in regional (mesenteric) lymph nodes and are different antigen-presenting cells responsible in that site? What of ‘immunosuppressive’ B7 interactions with the alternative ligand, the T-cell molecule CTLA-4, that might already be applying a brake on the inflammatory process? What is the likely pathogenetic role of B7 molecules that appear to be expressed as a secondary phenomenon? How likely to reduce inflammation is inhibition of molecules that are only seen at the height of inflammation? How accurately does DSS colitis mirror human ulcerative colitis anyway? For each incremental step in elucidating the chicken and egg pathways to inflammation in IBD, there are numerous unanswered questions. Some of these questions will be answered by grasping the nettle, and establishing biologically dictated therapeutic trials. Grose et al. have highlighted a few more questions; one way or another, the answers will arrive.