Antibiotic use can lead to significant shifts in community composition of the gut microbiota followed by predictable successional dynamics after administration has ceased. It is unknown, however, whether abiotic conditions change during treatment, and, if so, how such changes interface with community dynamics. We hypothesized that during treatment resources would accumulate due to reduced consumption, and that during recovery resources would reach an intermediate minimum before returning to healthy levels. These altered resource conditions were predicted to both reflect and shape community composition.
We sought to test this hypothesis by exposing in mice to a broad-spectrum antibiotic cocktail for five days and then allowing for 10 days of recovery. We measured oxygen tension in vivo, and redox potential and metabolite abundance in fecal pellets daily during treatment and recovery. Microbial community composition from feces was evaluated by high-throughput sequencing of bacterial 16S rRNA gene. Total bacterial load was estimated with qPCR.
Results/Conclusions
Antibiotic application produced a two-fold decrease in bacterial load that persisted throughout treatment, with a significant reduction apparent within 12 hours. We observed an increase in reduction potential within 16 hours of the first dose of antibiotics, and reduction potential remained elevated (258±72mV) throughout treatment. Following treatment, reduction potential overshoots, becoming negative (-162±137mV) for 2-3 days, before returning to control values (17±87mV). In contrast, we observed only a transient increase in oxygen levels (to 23±23Torr) roughly one day after the initial antibiotic dose. Otherwise, oxygen remained near zero (2±2Torr) throughout treatment and recovery. Metabolite abundances varied greatly throughout treatment and recovery, reflecting changes in consumption and production by a hugely dynamic community. These results indicate that the abiotic conditions in the gut are highly dynamic and are, in part, controlled by microbial activity.
These environmental changes coincide with a compositional shift to a community dominated by gamma-Proteobacteria during treatment. Community recovery was reproducible and hysteretic. Complete recovery was possible within a week for animals that were cohoused with control mice, whereas those that were kept in isolation were unable to recover completely. These results indicate that following cessation of antibiotics, the major limit to community recovery is dispersal.