Ursodeoxycholic acid inhibits colonic mucosal cytokine release and prevents colitis in a mouse model of disease Joseph BJ Ward 1, Orlaith Kelly 1,2, Siobhan Smith 3, Joan Ní Gabhann 3, Murtaza Tambuwala 4, Frank Murray 2, Caroline Jefferies 3, Cormac Taylor 4 and Stephen Keely 1. 1 RCSI, Beaumont Hospital. 2 Dept. of Gastroenterology, Beaumont Hospital. 3 RCSI, St. Stephens Green, 4 UCD. INTRODUCTION Toll-like receptors (TLRs) play a critical role in innate immune responses to intestinal pathogens. In inflammatory bowel disease (Ulcerative Colitis and Crohn’s Disease), epithelial TLR expression is increased. Activation of these TLRs is likely to play an integral role in the pro-inflammatory cytokine release responsible for the large-scale epithelial damage that is typical of active colitis. Ursodeoxycholic acid (UDCA), a bile acid, is a well-established therapy for inflammatory diseases of the liver, where it is known to act at least partly, through inhibition of cytokine release. However, the role of UDCA in regulating intestinal epithelial cytokine secretion is unknown. Here we sought to investigate a potential role for UDCA in regulating colonic epithelial cytokine release and to investigate the therapeutic potential of UDCA in a mouse model of disease. These studies reveal a novel role for UDCA in regulating colonic epithelial cytokine secretion. UDCA significantly attenuates TLR-3- dependent cytokine release from colonic epithelial cells through a pathway dependent on TRIF/TBK-1. UDCA also ameliorates the effects of colitis in a mouse model. Interestingly, metabolically stable analogues of UDCA do not prevent colitis in a mouse model, whereas the metabolic product of UDCA, LCA, practically abolishes DSS- induced colitis. In conclusion, our data suggest that, by virtue of their effects in preventing TLR-induced proinflammatory cytokine release, UDCA and its metabolic products are good targets for developing new approaches to treat IBD. Figure 7. LCA attenuates in vitro cytokine release and is protective in a murine model of colitis. A) The TLR3 agonist, Poly I:C (25 μg/ml), significantly increased TNF-  release compared to control. This response was reduced by co-treatment with LCA (10  M) (n = 7 p  0.001). B) Intraperitoneal administration of the UDCA metabolite, LCA (30 mg/kg), to male C57 BL/6 mice receiving DSS in their drinking water, significantly reduced the DAI from 11.2 ± 0.6 (DSS alone) to 5.4 ± 0.9 (n = 5, p  0.001). A Metabolically stable analogues of UDCA are ineffective against colitis in a mouse model of disease Figure 5. 6MUDCA attenuates in vitro cytokine release, but is not protective in a murine model of colitis. A) The TLR3 agonist, Poly I:C (25 μg/ml), significantly increased TNF-  release compared to control. This response was reduced by co- treatment with either UDCA (200  M) or a stable analogue, 6MUDCA (n = 5 p  0.001). B) However, in contrast to UDCA, intraperitoneal administration of 6MUDCA (30mg/kg) to male C57 BL/6 mice receiving DSS in their drinking water, did not reduce the DAI (n = 3 - 9). B Untreated DSS Untreated DSS UDCA PBS D E Biochemistry of UDCA metabolism Figure 4. UDCA and its derivatives can be readily interconverted. Liver enzymes along with bacterial enzymes in the colon can readily interconvert UDCA to LCA and 7-keto LCA. Figure 2. UDCA inhibits TLR signalling at TRIF/TBK1. Overexpression analysis of proteins in the TLR signalling pathway revealed that UDCA (200  M) exerts its inhibitory effect at the TIR-domain-containing adapter-inducing interferon- β (TRIF)/TANK-binding kinase 1 (TBK1) junction (n = 6, p  0.05). Figure 3. UDCA is protective in a murine model of colitis. A) Intraperitoneal administration of UDCA (30 mg/kg and 100 mg/kg) to male C57 BL/6 mice receiving DSS in their drinking water, significantly reduced the disease activity index (DAI) from 10 ± 0.3 (DSS alone) to 7.2 ± 0.7 (UDCA 30mg/kg) and 5.8 ± 0.5 (UDCA 100mg/kg) (n = , p  0.001). B) Mice treated with DSS alone had shorter colons and a lack of faecal pellet formation. In contrast, mice co-treated with UDCA had longer colons with clear evidence of faecal pellet formation. C) The average length of colons in DSS-treated mice was shorter than in mice co- treated with UDCA (30 mg/kg) (n = 6-12, p < 0.05). D) H and E staining was performed on colonic sections and E) UDCA (30 mg/kg) significantly reduced the inflammation score of the sections from 37.3 ± 0.3 in DSS alone, to 29.0 ± 3.5 (ANOVA, n = 5, p  0.05). A A UDCA attenuates TRIF/TBK1 signalling METHODS RESULTS UDCA attenuates TLR-3 driven cytokine release from colonic epithelial cells Figure 1. UDCA attenuates TLR-3 driven cytokine release from T84 cells. A) The TLR3 agonist, Poly I:C (25 μg/ml), significantly increased TNF-  release compared to control, which was significantly reduced in a concentration-dependent manner by co-treatment with UDCA (0-200  M). Poly I: C-stimulated B) TNF-  release, C) IL-8 release, D) IL-6 release and E) IL-1  release were all significantly reduced to by co-treatment with UDCA (200  M). Statistical analyses were performed using ANOVA and SNK or Tukey post hoc test. B 6MUDCA C D E C Vehicle UDCA (30 mg/kg) DSS (2.5%) + Vehicle DSS (2.5%) + UDCA (30 mg/kg) UDCA ameliorates colitis in a mouse model of disease SUMMARY & CONCLUSION B The UDCA metabolite, LCA prevents colitis in a mouse model of disease UDCA metabolites are elevated in a mouse model of disease Figure 6. UDCA administration alters the caecal bile acid pool. Intraperitoneal administration of UDCA (30 mg/kg) to male C57 BL/6 mice receiving DSS in their drinking water, significantly elevated levels of UDCA metabolites, LCA and 7keto LCA.