ORIGINAL ARTICLE Genes Associated with Intestinal Permeability in Ulcerative Colitis: Changes in Expression Following Infliximab Therapy Gary Toedter, PhD,* Katherine Li, MS, † Sarah Sague, PhD,* Keying Ma, PhD,* Colleen Marano, PhD, ‡ Michael Macoritto, PhD, § Jennifer Park, PhD, § Ren ee Deehan, PhD, § Andrea Matthews, BA, § Gary D. Wu, MD, ¶ James D. Lewis, MD, MSCE, ¶ Ingrid Arijs, PhD, k Paul Rutgeerts, MD, PhD, k and Fr ed eric Baribaud, PhD* Background: Alterations in intestinal permeability have been implicated in ulcerative colitis (UC). Infliximab, a monoclonal anti-tumor necro- sis factor alpha (TNFa) antibody, can induce clinical response in UC. Gene expression in colonic biopsies taken from responders and nonres- ponders to infliximab can provide insight into the mechanisms of the altered intestinal permeability at a molecular level. Methods: Colonic biopsies (n ¼ 18 anti-TNFa naı ¨ve UC patients; n ¼ 8 normal controls; n ¼ 80 Active Ulcerative Colitis Trial [ACT] 1 patients) were analyzed for mRNA expression using gene expression microarrays. Computational reverse causal reasoning was applied to build causal network models of UC and response and nonresponse of UC to treatment. Quantitative reverse-transcription polymerase chain reaction (qPCR) was used to confirm differentially expressed genes. Results: Reverse causal reasoning on mRNA expression data from anti-TNFa-naı ¨ve UC and normal samples provided a mechanistic disease model of the biology of gene expression observed in UC. mRNA expression data from the ACT 1 study enabled construction of a mechanistic model describing the biology of nonresponders to infliximab, including evidence for increased intestinal permeability compared with normal and responder samples. Gene expression changes identified as central to intestinal permeability dysregulation were confirmed in normal, UC, and infliximab-treated patients by qPCR analysis. Gene expression returned toward normal levels in infliximab responders, but not in nonresponders. Conclusion: Gene expression analysis and causal network modeling in combination showed that aberrant mRNA expression of genes involved in intestinal epithelial permeability for infliximab responders was restored toward levels observed in normal samples. Infliximab nonresponders showed no equivalent restoration in the expression of these genes. (Inflamm Bowel Dis 2012;000:000–000) Key Words: ulcerative colitis, infliximab, tight junction, epithelial-mesenchymal transition U lcerative colitis (UC) is a chronic inflammatory bowel disease (IBD) of the colon. The pathogenesis of UC depends on the immune response to microbial infection, multiple genetic factors, and environmental factors. The result is an abnormal response to bacterial antigens nor- mally found in the gastrointestinal tract. 1,2 Increased intes- tinal permeability has been implicated in the pathology of UC. 3,4 The increase in permeability may be caused by changes in the function of tight junctions (TJ) and influ- enced by epithelial-mesenchymal transition (EMT) processes. 5 Current treatments for UC include immunomodula- tors, anti-inflammatory agents, and biologics such as inflixi- mab, a tumor necrosis factor-alpha (TNFa) antagonist. TNFa is a component of the inflammatory response and therapy with infliximab improves the clinical presentation of patients with UC and induces mucosal healing. 6,7 Clini- cal remission can be achieved with infliximab in over one- third of patients, with clinical response occurring in approximately two-thirds of patients in the Active Ulcera- tive Colitis Trial (ACT) 1 and 2 trials. 6 Additional Supporting Information may be found in the online version of this article. Received for publication November 10, 2011; Accepted November 15, 2011. From the *Biomarkers, Centocor Research & Development, Malvern, Pennsylvania; and † In Silico Immunology, Centocor Research & Development, Malvern, Pennsylvania, ‡ Clinical Research & Development, Inc., Malvern, Pennsylvania, § Selventa, Cambridge, Massachusetts, k Division of Gastroenterology, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania, ¶ Department of Gastroenterology, University Hospital Gasthuisberg, Leuven, Belgium. Funded by Centocor Research & Development, Role of the funding source: Centocor Research & Development, Inc. is the commercial entity selling infliximab under the trade name of Remicade. A steering committee of academic investigators and Centocor Research and Development, Inc. contributors designed the ACT 1 study. Centocor staff created the clinical database, performed the mRNA expression and qPCR analysis, and performed statistical analysis on the data. All authors interpreted the data and prepared and approved the article for submission. Reprints: Gary Toedter, PhD, Centocor, Research & Development, 145 King of Prussia Road, Radnor, PA 19087 (e-mail: gtoedter@its.jnj.com). Copyright V C 201 Crohn’s & Colitis Foundation of America, Inc. DOI 10.1002/ibd.22853 Published online in Wiley Online Library (wileyonlinelibrary.com). Inflamm Bowel Dis 1 2