Abstract
Adherent-invasive Escherichia coli (AIEC) strains are overrepresented in the dysbiotic microbiota of Crohn's disease (CD) patients, and contribute to the onset of the chronic inflammation typical of the disease. However, the effects of anti-inflammatory drugs used for CD treatment on AIEC virulence have not yet been investigated. In this report, we show that exposure of AIEC LF82 strain to amino-6-mercaptopurine (6-MP) riboside, one of the most widely used anti-inflammatory drugs in CD, impairs its ability to adhere to, and consequently to invade, human epithelial cells. Notably, phagocytosis of LF82 treated with 6-MP by human macrophages is also reduced, suggesting that 6-MP affects AIEC cell surface determinants involved both in interaction with epithelial cells and in uptake by macrophages. Since a main target of 6-MP in bacterial cells is the inhibition of the important signal molecule c-di-GMP, we also tested whether perturbations in cAMP, another major signaling pathway in E. coli, might have similar effects on interactions with human cells. To this aim, we grew LF82 in the presence of glucose, which leads to inhibition of cAMP synthesis. Growth in glucose-supplemented medium resulted in a reduction in AIEC adhesion to epithelial cells and uptake by macrophages. Consistent with these results, both 6-MP and glucose can affect expression of cell adhesion-related genes, such as the csg genes, encoding thin aggregative fimbriae (curli). In addition, glucose strongly inhibits expression of the fim operon, encoding type 1 pili, a known AIEC determinant for adhesion to human cells. To further investigate whether 6-MP can indeed inhibit c-di-GMP signaling in AIEC, we performed biofilm and motility assays and determination of extracellular polysaccharides. 6-MP clearly affected biofilm formation and cellulose production, but also, unexpectedly, reduced cell motility, itself an important virulence factor for AIEC. Our results provide strong evidence that 6-MP can affect AIEC-host cell interaction by acting on the bacterial cell, thus strengthening the hypothesis that mercaptopurines might promote CD remission also by affecting gut microbiota composition and/or physiology, and suggesting that novel drugs targeting bacterial virulence and signaling might be effective in preventing chronic inflammation in CD.