32 Impact of colibactin-producing escherichia coli on immune microenvironment in preclinical colorectal cancer models

Abstract

Introduction Increasing evidence link the immune microenvironment, microbiota and colorectal cancer (CRC). E. coli pks+ are more frequently detected on CRC patient mucosa and exhibit procarcinogenic properties in CRC murine models. Aim of this work was to evaluate the impact of chronical infection by colibactin-producing E. coli on immune cells in APCmin/+ mice colon. Material and methods Min mice were per os inoculated with a CRC E. coli pks+ strain (11 G5), non-pathogenic E. coli (K-12) or PBS. Induction of oxidative stress and inflammation during the infection was evaluated using optical in vivo imaging (IVIS spectrum). After 7 weeks, number and volume of polyps were evaluated and colonic samples were histologically analysed. Immunofluorescent staining of immune cells (neutrophils and T cell populations) was performed to quantify positive cells in three interesting colon areas: mucosa, lymphoid follicle and tumour using an innovative algorithm created with Tissue Studio software. In parallel, quantification of immune cells (lymphoid populations) was performed in mesenteric lymph node (MLNs) by flow cytometry. Results and discussions Using optical imaging, we detected a significant increase of oxidative stress and inflammation in the gut after thirty days of 11 G5 infection. However, using our specific algorithm and CD45 immunostaining, we observed a significant increase of lymphoid follicle size in the gut of mice infected with the 11 G5 strains. Interestingly, follicle size was positively correlated with tumour volume suggesting an association between pro-carcinogenic properties of 11 G5 strain and gut immune response. In addition, we observed an increase of neutrophils on the mucosa and lymphoid follicle of 11 G5 infected mice. These results can be linked to our in vivo optical imaging observations. Moreover we observed a trend to decrease of CD3+ T cells on mucosa and tumour for the 11 G5 group, suggesting a possible link between E. coli pks +and anti-tumour T cells. Finally we noticed a significant decrease of CD4+ CD25+ anti-inflammatory T cells in MLNs of 11 G5 infected mice, negatively correlated with gut bacteria colonisation. Conclusion Here we report a potential immunomodulatory effect on gut microenvironment by pks+E. coli, which could be associated with its carcinogenic effect. Our results suggest that gut lymphoid follicle and immune cells such as neutrophils and T cells could be involved in this process. This work shows a link between immune microenvironment, pathogenic E. coli and tumour development.