ORIGINAL ARTICLE Systemic Toll-Like Receptor Ligands Modify B-cell Responses in Human Inflammatory Bowel Disease Marie McDonnell, MD,* YanMei Liang, RN, † Ansu Noronha, MD, ‡ Jennifer Coukos, BS, ‡ Dennis L. Kasper, MD, § Francis A. Farraye, MD, ‡ and Lisa M. Ganley-Leal, PhD † Background: Bacteria have a central, although poorly under- stood, role in inflammatory bowel disease (IBD). Host–bacteria interactions primarily take place in the gastrointestinal tract, but cells may also encounter translocated bacteria in the bloodstream. IBD is associated with activated, circulating Toll-like receptor (TLR)2 and TLR4-expressing B cells suggesting that blood-borne microbial TLR ligands modulate B cell responses. Methods: Serum levels of lipopolysaccharide (LPS)/endotoxin and high mobility group box 1 (HMGB1), an endogenous TLR ligand, were quantified in Crohn’s disease (CD) and ulcerative co- litis (UC). Responses of purified B cells to LPS and HMGB1 were correlated with levels of systemic TLR ligands and clinical parameters of disease. Results: While IBD patients have increased levels of blood LPS, the net effect of endotoxemia has unexpected characteristics illus- trating that LPS has both pro- and antiinflammatory roles through TLR4þ B cells. Experimental treatment of B cells demonstrates that the antiinflammatory effect of LPS is due to its hypo-acyla- tion of lipid A suggesting an increased prevalence of systemic, hypo-acylated LPS in CD. In contrast, high levels of LPS are associated with disease activity in UC. HMGB1 activates B cells through TLR2 and CD36. Serum levels of HMGB1 correlate with spontaneous IL-8 production by B cells suggesting that blood- borne TLR2 ligands increase B-cell activation in vivo. Conclusions: Systemic TLR ligands modulate B cells towards either proinflammatory or antiinflammatory activity depending on the predominant ligand(s). Further, the circulating B cell may rep- resent an important proxy for quantifying the LPS lipid A acyla- tion burden in patients with IBD. (Inflamm Bowel Dis 2011;17:298–307) Key Words: inflammatory bowel disease, B cell, endotoxin, Toll- like receptor, human, HMGB1 I nflammatory bowel disease (IBD) is a relapsing/remitting chronic, progressive disease and despite several available treatments, debilitating relapses remain unpredictable and require therapies that have dangerous side effects, particu- larly for children. Many of these chronic therapies rou- tinely fail and the more potent and toxic antiinflammatory therapies are often used in the setting of flares. Surgical management is the disappointing and often life-changing result of ineffective treatment of IBD, illustrating the need to better understand the pathogenesis of this disease. 1 Bacteria undoubtedly play a central role in the initia- tion and/or pathogenesis of IBD. 2 However, there are many factors about the host–microbial interactions that remain unclear in human disease. For example, it has not been defined whether certain species of bacteria are pathogenic or whether commensals induce, or exacerbate, disease by inappropriate stimulation of the host immune system. 1 The normal microbes of the gut are not pathogenic if they remain within their gut micro-niche, but they can induce an inflammatory response similar to that seen to pathogenic bacteria if they cross the epithelial barrier. 3 Bacterial trans- location from the lumen to the submucosa is thought to play a role in IBD. 1 While host–bacteria interactions are generally thought to take place in the gastrointestinal mucosal, cells may also encounter translocated bacteria in the bloodstream. 4–6 Although intermittent exposure to blood-borne bacteria is a common occurrence, such as after tooth-brushing, immune mechanisms are in place to prevent clinical infection in healthy individuals. 7 In IBD the pres- ence of bacteria and their products may persist in the blood due to an altered immune balance or other uncharacterized factors. Identified microorganisms in the blood of IBD patients include Escherichia coli, Staphylococcus aureus, Klebsiella spp., Streptococcus pneumoniae, and Enterococ- cus faecalis, suggesting the presence of a myriad of Received for publication June 3, 2010; Accepted June 14, 2010. From the *Section of Endocrinology, Evans Biomedical Research Center, Boston Medical Center, Boston, Massachusetts, † Section of Infectious Disease, Department of Medicine, Boston University School of Medicine, Boston Medical Center, Boston, Massachusetts, ‡ Section of Gastroenterology, Boston Medical Center, Boston, Massachusetts, § Channing Laboratory, Department of Medicine, Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts. Supported by Broad Medical Research Program of the Broad Foundation (to L.G.-L.), BD Grant Award 2007 (to L.G.-L.), and a Department of Medicine Pilot Award (to M.M.). Reprints: Lisa M. Ganley-Leal, PhD, Section of Infectious Diseases, Boston University School of Medicine, 650 Albany St., Room 630, Boston, MA 02118 (e-mail: lisa.ganleyleal@bmc.org) Copyright V C 2010 Crohn’s & Colitis Foundation of America, Inc. DOI 10.1002/ibd.21424 Published online 30 August 2010 in Wiley Online Library (wileyonlinelibrary.com). Inflamm Bowel Dis  Volume 17, Number 1, January 2011 298