ORIGINAL ARTICLE Differential Role of Tumor Necrosis Factor Receptors in TNBS Colitis Dawn R. Ebach, MD,* Rodney Newberry, MD,† and William F. Stenson, MD† Background: Tumor necrosis factor a (TNFa) plays a central role in the pathology of T helper 1–mediated colitis such as Crohn’s disease; however, the role of its 2 receptors in mediating pathology has not been fully explored. Methods: Trinitrobenzene sulfonic acid colitis was used to induce colitis in mice lacking each of the TNF receptors (TNFRs) and in wild-type mice. TNFR1 2/2 mice lost more weight, became hypothermic, and had increased mortality compared with wild-type C57Bl/6 mice. TNFR2 2/2 mice, however, lost less weight, had normal temperatures, and had improved survival. Results: Despite the improved clinical outcomes in TNFR2 2/2 mice, TNFa levels were increased in these mice. Conclusions: TNFa signaling through TNFR1 is protective in the trinitrobenzene sulfonic acid mouse model of inflammatory bowel disease. Key Words: inflammatory bowel disease, trinitrobenzene sulfonic acid, tumor necrosis factor a, tumor necrosis factor receptors (Inflamm Bowel Dis 2005;11:533–540) T rinitrobenzene sulfonic acid (TNBS) induces colitis by haptenation of colonic proteins, leading to a delayed-type hypersensitivity reaction by causing a CD4 + T helper 1 (T H 1) reaction to self-antigens. 1 This reaction leads to colitis similar to Crohn’s disease, with transmural mononuclear cell infiltrate, abnormal crypt architecture, ulcerations, and occasional granu- lomas. 2 Elevated levels of interferon g (IFNg), tumor necrosis factor a (TNFa), and granulocyte macrophage-colony stim- ulating factor (GM-CSF) characterize the T H 1 reaction. The central role of TNFa in mediating pathology in human inflam- matory bowel disease (IBD) and several models of colitis is well-described. 3–6 Mice with TNBS-induced colitis have improvement in clinical and histologic disease when treated with antibodies to TNFa , and TNFa 2/2 mice do not develop disease when challenged with TNBS. 3 Treatment of Crohn’s disease with antibodies to TNFa often leads to histologic and endoscopic healing. 7,8 TNFa interacts with 2 receptors, tumor necrosis factor receptor 1 (TNFR1) and tumor necrosis factor receptor 2 (TNFR2), which can be membrane-bound or soluble. Both receptors are present on all nucleated cells and have distinct signal transduction pathways and mediate distinct cellular responses. 9–11 TNFR1 has a death domain that mediates apopto- sis through caspase activation and activates nuclear factor kB, leading to transcription of inflammatory cytokines, chemo- kines, and antiapoptotic proteins. Signals from TNFR1 initiate apoptosis in activated T lymphocytes, have antiviral activities, and stimulate growth. TNFR2 does not contain the death domain but can mediate apoptosis through the kinase receptor- interacting protein (RIP). 12 Signaling through TNFR2 pro- motes thymocyte differentiation and T-cell proliferation. 10 TNFR1 2/ 2 and TNFR2 2/2 mice have different phenotypes. TNFR1 2/ 2 mice survive lethal doses of lipopolysaccharid (LPS), D-galactosamine, and TNFa. 13 They have defective PeyerÕs patch and germinal center formation and a marked susceptibility to Listeria monocytogenes infection. 13,14 TNFR2 2/2 mice have decreased lymphocyte proliferation, decreased scab formation in response to intradermal TNFa , and decreased TNFa-induced lethality. 15–18 The divergent functional effects of signaling through TNFR1 and TNFR2 are further shown by the TNF D ARE/+ mouse, which has persistent accumulation and decreased rates of decay of TNFa mRNA. The TNF D ARE/+ mouse develops inflammatory arthritis and severe intestinal trans- mural inflammation of the ileum-containing granulomas. 19 TNF D ARE/+ TNFR1 2/2 mice have no disease and normal histo- pathology. TNF DARE/+ TNFR2 2/2 mice have attenuated disease with mild mucosal and submucosal inflammation, but they have aggressive joint pathology. 19 The marked functional differences resulting from signaling through these 2 receptors led us to postulate that Received for publication March 1, 2005; accepted March 15, 2005. From the *Division of Pediatric Gastroenterology and Nutrition, Department of Pediatrics and †Division of Gastroenterology, Department of Medicine, Washington University, St. Louis, Missouri. Supported by The Washington University Digestive Disease Research Center (P30DK052574), National Institutes of Health Institutional Training Grant T32 HD-07409 (D.R.E.), a Crohn’s and Colitis Foundation of America Research Fellowship Award (D.R.E.), and DK55753 and DK33165 (W.F.S). Reprints: William F. Stenson, 660 South Euclid Avenue, Box 8124, St. Louis, Missouri 63110 (e-mail: wstenson@im.wustl.edu) Copyright Ó 2005 by Lippincott Williams & Wilkins Inflamm Bowel Dis  Volume 11, Number 6, June 2005 533