Increased thrombin generation and hypercoagulability are prominent features of inflammatory colitis and previous studies from our laboratory have suggested that thrombin-mediated proteolysis is a driver of both colitis and colitis-associated colon cancer (CAC). However, the downstream thrombin targets important in these disease processes have not been fully defined. Based on studies showing that the protease activated receptor-1 (PAR-1) can contribute to both inflammatory pathologies and cancer progression in other settings, we hypothesized that PAR-1 is a significant determinant of colitis and CAC. To test this hypothesis, we induced colitis in PAR-1−/− and control mice using intrarectal administration of 2,4,6-trinitrobenzenesulfonic acid (TNBS). Consistent with the concept that PAR-1 is a modifier of colitis pathobiology, PAR-1−/− mice lost significantly less weight than WT mice challenged in parallel. Furthermore, multiple inflammatory cytokines known to drive colitis pathology, including IL-6, TNFa, and MIP-1α, were significantly diminished in PAR-1−/− mice. However, histological analyses of colonic tissue revealed similar degrees of inflammatory cell infiltration, crypt abscesses, and mucosal hyperplasia in both genotypes. In order to explore the role of PAR-1 in the more complex process of inflammation-driven colon cancer pathogenesis, we induced CAC in PAR-1−/− and WT mice using a two step protocol consisting of azoxymethane (AOM) and dextran sodium sulfate (DSS) exposure. In contrast to findings in the setting of TNBS challenge, PAR-1−/− mice challenged with DSS developed, not less, but more severe clinical signs of colitis, including wasting and severe diarrhea. More detailed comparative studies of DSS-challenged PAR-1−/− and control mice established that PAR-1-deficient animals developed significantly greater immunological and histopathological evidence of colitis, including elevated IL-6 and MIP-1α levels in colonic tissue and increased edema, ulceration, crypt loss, and inflammatory cell infiltration. Consistent with the more severe antecedent colitis, PAR-1−/− mice challenged with AOM/DSS developed significantly larger adenomas than WT mice challenged and evaluated in parallel. Thus, the impact of PAR-1 on colitis appears to be context-dependent and the distinct outcomes in TNBS- and DSS-challenged mice are likely to stem from the different mechanisms by which these agents induce colitis. TNBS is thought to haptenate colonic mucosal proteins inducing a T cell-mediated colitis akin to human Crohn's disease. In contrast, DSS directly intoxicates colonic crypt epithelia, resulting in loss of barrier function and translocation of colonic microflora, leading to a primarily innate immune-driven colitis sharing many features with ulcerative colitis. A major challenge in dissecting the precise mechanisms coupling PAR-1 to colitis is the fact that PAR-1 is expressed on multiple cell types that can influence colitis and CAC in distinct ways, including immune cells, endothelial cells and colonic mucosa. Therefore, we recently generated mice carrying a conditional “floxed” PAR-1 allele. We interbred these animals with mice expressing Cre recombinase in either colonic epithelia or the hematopoietic/endothelial compartment. Preliminary studies revealed that loss of PAR-1 expression in the hematopoietic/endothelial compartments, but not the colonic epithelia, recapitulates the more severe DSS-induced weight loss and mucosal damage observed in constitutionally PAR-1-deficient mice. These results suggest that PAR-1 activation in either immune cells and/or endothelial cells limits colitis severity in this experimental context. Taken together, these data show that PAR-1 contributes to the pathogenesis of inflammatory colitis and CAC, but the precise contribution is dependent on the underlying insult and disease pathway. Analyses in mice carrying a conditional PAR-1 allele should prove invaluable for dissecting the precise mechanisms coupling PAR-1 to inflammatory bowel disease.
Palumbo:Novo Nordisk Corporation: Consultancy.
Asterisk with author names denotes non-ASH members.