I succesfully engineered laboratory strains of E. coli to contain a combination of three characteristics: an amino acid auxotrophy (i.e. inability to growht unless it is externally provided), an elevated (and biased) mutation rate, and a fluorescent marker. These characteristics ensured that we had pairwise mutualistic communities that contained mutator strains, and that the composition of these communities could be tracked over time. After pairing these engineered strains into either single mutator or double mutator communities, I conducted two separate evolution experiments.
In the first evolution experiment, communities were exposed to increasing concentrations of the beta-lactam antibiotic cefotaxime. In this adaptive evolution experiment, we found large differences in the survivability of each community to antibiotic stress. We found that only communities in which both members had elevated mutation rates were able to survive increased antibiotic concentrations across all replicates. Following this evolution experiment, we extensively characterised the type of interaction each community contained, categorising each community as either ‘revertant’, ‘commensal’ and ‘mutualistic’, depending on if each strain needed the other strain to survive. We found similar rates of mainenance of mutualism in our double mutator communities as ‘wild type’ communities that evolved without antibiotic, indicating that adaptation to environmental stress does not need to come at the expense of mutualistic interactions (as previously suggested).
In the second evolution experiment, we serially bottlenecked our communities through individual clones of each constituent species, before growing both species together again. During this non-adaptive portion of this evolution experiment, mutations were able to accumulate in each species’ genome, potentially destabilising the mutualistic interaction: as mutator strains have increased mutation rates, we should expect to see a higher degree of metabolic reversions and extinctions in these communities. However, so far we have observed the opposite, with slightly more extinctions witnessed in the non-mutator communities. We are continuing this work with more bottlenecks to get a stronger signal of mutualism breakdown.
Overall, this work shows that communities can be protected from environmental stress without compromising on community structure and function, and that this is predictable based on the presence of mutator strains in the populations. Furthermore, it reiterates the importance of incorporating knowledge of mutator strains into evolutionary predictions.
