Development of a new antibiotic-induced dysbiosis model of the canine colonic microbiota.

As in humans, antibiotics are widely used in dogs to treat gastrointestinal infections, contributing to the global burden of antimicrobial resistance on both human and animal health. Close contact between pets and their owners can lead to horizontal transfer of gut microbes, including transmission of antibiotic resistance. Nevertheless, until now, the impact of antibiotics on the canine gut microbiota has been poorly described. The aim of this study was to adapt the canine mucosal artificial colon (CANIM-ARCOL) model, reproducing the main nutritional, physicochemical and microbial parameters found in the large intestine of the dog to simulate an antibiotic-induced perturbation. Following initial investigation of five antibiotic cocktails at in-field doses, a 5-day regimen of metronidazole/enrofloxacin (ME) was selected for further model development. Two CANIM-ARCOL bioreactors were inoculated with a faecal sample (n=2 donors) and run in parallel for 26 days under control or antibiotic conditions. ME reduced microbial diversity and induced major shifts in bacterial populations, leading to a state of dysbiosis characterized by an increase in the relative abundance of Streptococcaceae, Lactobacillaceae and Enterobacteriaceae, and a decrease in the relative abundance of Bacteroidaceae, Fusobacteriota and Clostridiaceae. Overall, mucus-associated microbiota were less impacted by antibiotics than luminal microbes. Microbial alterations were associated with drastic decreases in gas production and short-chain fatty acid concentrations. Finally, the model was well validated through in-vitro-in-vivo comparisons in a study in dogs. The CANIM-ARCOL model provides a relevant platform as an alternative to in-vivo assays for an in-depth understanding of antibiotic-microbiota interactions and further testing of restoration strategies at individual level.