Background: Ulcerative colitis (UC) is associated with dysbiosis and an increased risk of colon cancer over time. Non-invasive methods of identifying UC patients at risk of developing CAC do not exist. We have recently shown that patients with Crohn's disease have altered serum levels of quorum sensing molecules (QSMs) that reflect inflammatory state. QSMs regulate bacterial metabolism and may activate host signaling pathways. We hypothesize that changes in QSMs may reflect UC dysplasia risk progression and participate in the development of CAC.
Methods: We collected serum from healthy individuals and three UC patient subgroups stratified by risk of CAC: those with active inflammation, low risk of CAC (<5 years UC diagnosis/quiescent disease), and high risk of CAC (>10 years diagnosis/active disease). We used biosensors to determine levels of the QSMs autoinducer-2 (AI-2), long chain N-acyl homoserine lactones (lcAHL), and short chain AHL (scAHL) through bioluminescent intensity measurement. Wildtype (WT) mice underwent a CAC model of repeated azoxymethane (AOM) injections and 3% dextran sulfate sodium (DSS) administration, and blood was collected for QSM level determination. In a separate experiment, both specific-pathogen free (SPF) and germ-free (GF) mice underwent the same CAC model with concomitant intraperitoneal scAHL administration. Tumor burden was measured, and stool was collected for metabolomic analysis (SPF mice). Colonoids from WT mice were incubated with scAHLs, and supernatants were analyzed by ELISA.
Results: Patients with active UC had elevated levels of scAHLs and lcAHLs compared to healthy subjects. Patients at high risk for CAC had elevated levels of scAHLs and AI-2 compared with those at low risk of CAC. In AOM-DSS treated mice, tumor development was accompanied by lower levels of AI-2 and increased levels of scAHL in plasma. Administration of scAHL led to an increase in tumor size and number in SPF mice. Metabolomic analysis revealed significant differences in the metabolome of mice that were given scAHLs compared to controls. In GF mice, scAHL administration also led to an increase in tumor burden, suggesting that the mechanism by which scAHLs promote tumorigenesis is not completely dependent on the microbiota. In vitro, scAHL incubation led to an increase in the pro-inflammatory markers CXCL5, LIF, M-CSF.
Conclusion: We show that serum levels of bacterial QSMs increase with inflammation, particularly in patients with the highest risk of CAC. Our data in murine models establishes a link between altered QSM levels and increased tumor burden. At least part of the QSM effect may occur through host inflammatory signaling. We posit that QSMs may be useful as biomarkers of CAC risk in IBD patients. Longitudinal studies of QSM levels in patients that develop dysplasia will be needed to confirm its value.

Gregory O'Connor

Appearances