Abstract <p>The azoxymethane/dextran sodium sulfate (AOM/DSS) murine modelis widely used to study the pathogenetic mechanisms of colitis-associatedcolorectal cancer (CRC). However, the reproducibility of this modelcan be challenged in the long run due to several factors, specifically,the variability of the gut microbiota composition in laboratorymice. Here, we report the results of our three cohort study of C57BL/6male mice with the AOM/DSS model of CRC, bred in the same vivariumfacility. The experiments with cohorts 1 and 2 from different roomswere performed in parallel, while the experiment with cohort 3 wasconducted 18 months later. The results demonstrate that mice ofthese cohorts developed colonic tumors with different efficiency:the percentage of animals with visible tumors was 37.5, 84.6, and96.7% respectively. Notably, cohorts 2 and 3 also differed by theaverage number of tumors per animal and the area of affected colonictissue. Gut microbiota analysis via 16S rRNA amplicon sequencingrevealed significant cross-cohort differences in the species diversity,which might account for the variability in AOM/DSS model performance.Our experience shows that long-term working with murine models mayengender problems with their reproducibility, specifically, withdata heterogeneity. To minimize such a heterogeneity, special considerationshould be given to mouse housing conditions, particularly to thosefactors that affect gut microbiota. A regular monitoring of specificpathogens according to FELASA recommendations, as well as the screeningof the commensal microbiota using next generation sequencing techniquesbefore the onset of experiments, are also highly reasonable.</p>