Do antibodies have a role in IBD pathogenesis?

Abstract

The presence of several types of antibodies in the sera of inflammatory bowel disease (IBD) patients is 1 of several findings suggesting the immune-mediated nature of both Crohn’s disease (CD) and ulcerative colitis (UC). The first detection of such a reactivity emerged in the late 1950s with the report of autoantibodies against colonic epithelial cells in UC patients.1 Since then, several types of antibodies have been shown to react with a vast array of antigens, including microbial as well as autoantigens from intestinal or extraintestinal structures. Many studies have been performed that were mainly aimed at evaluating the possible use of these antibodies as a noninvasive diagnostic tool or as a means to detect distinct clinical subsets of patients with particular phenotypic or prognostic profiles. Fewer efforts have been devoted to understanding the possible role of such antibodies in the pathogenesis of IBD. Among the circulating antibodies detected in IBD, the most studied are those directed toward some specific intestinal epithelial antigens (ECAC and TM-5), a yet undefined nuclear antigen of neutrophils (p-ANCA), and mannose residues from the cell wall mannans of Saccharomyces cerevisiae (ASCA). More recently, circulating antibodies reacting with different glycans and other microbial antigens have also been reported. ECAC are intestinal goblet cell glycoproteins recognized by both circulating antibodies and mononuclear cells from IBD patients, suggesting autosensitization to these potential autoantigens. Mucosal CD3 lymphocytes inducing antibody-mediated cytotoxicity against ECAC have also been reported.2 However, further characterization of this reactivity has not been carried out, so its importance in IBD pathogenesis remains undefined. TM-5 is a colonic isoform of the cytoskeletal protein tropomyosin. Circulating and lamina propria mononuclear cell-secreted antibodies against this protein have been reported in UC but not in CD subjects. An intriguing finding related to this antigen is the observation that similar epitopes are present in skin, eye, joint, and bile ducts, all common sites of extraintestinal complications of IBD.3 This finding suggests that a crossreactive antigenic recognition might play a role in the pathogenesis of the disease and its complications. Although the observation of an activated complement and of IgG1 deposits colocalized with anti-TM-5 monoclonal antibody in intestinal epithelium would suggest a direct pathogenic role of this autoantigen–antibody reaction, its relevance in the pathogenesis of IBD is still undefined. p-ANCA (perinuclear antineutrophil cytoplasmic antibodies) are directed toward a yet undefined nuclear lamina antigen(s) of human neutrophils, different from the antigens recognized by sera of patients with vasculitis or nephritis.4 Several studies have shown that this antibody is associated with UC, where it is detected in 40%–80% of the patients,5 whereas in CD it is observed in a smaller proportion of patients (5%–15%), particularly those with a UC-like disease pattern. However, very few studies have addressed their possible pathogenic role, and most have produced negative results. In particular, contrary to what is observed with vasculitis associated p-ANCA, IgG from p-ANCA-positive UC patients were not able to activate neutrophil respiratory burst.6 This, together with the observation that 20%–50% of UC patients develop in the absence of the antibody, suggests that p-ANCA does not play a prominent role in the pathogenesis of the disease. Anti-Saccharomyces cerevisiae antibodies (ASCA) are associated with CD, where they are observed in up to 68% of the patients.5 Since they are directed against the cell wall mannan of S. cerevisiae, they cannot be considered a true autoantigen. Although they may be observed in association with increased intestinal permeability, a putative pathogenetic mechanism of CD, their direct role in the mechanism of disease appears unlikely. The sensitivity of this antibody is quite poor, in fact, 30%–50% of CD patients do not have ASCA. Thus, despite both pANCA and ASCA being specific markers of UC and CD, respectively,5 and seemingly predicting the development of the disease years before the onset, as shown in a Jewish cohort of patients,7 their low prevalence in affected subjects does not support the hypothesis of their primary involvement in IBD pathogenesis. Interestingly, the seroreactivity against oligosaccharides appears to be a preeminent feature of CD. In fact, recently each component of a wide panel of anti-glycan antibodies has been tested for its association with IBD; among these, antibodies anti-laminaribioside (ALCA), antichitobioside (ACCA), anti-mannobioside (AMCA) and against a mannan epitope of S. cerevisiae (gASCA) displayed a high specificity for CD; furthermore, some data suggest that From the IRCCS Policlinico San Donato Hospital & University of Milan, Milan, Italy. Copyright © 2008 Crohn’s & Colitis Foundation of America, Inc. DOI 10.1002/ibd.20699 Published online in Wiley InterScience (www.interscience.wiley.com).