Aim 1. We identified five clinical isolates of ESBL E. coli from different STs able to colonize the gut of conventional mice harboring a fully protective microbiota. We were able to generate libraries of mutants in these strains using transposon insertion mutagenesis. We are currently using these libraries in mice in order to identify genes that favor colonization in the presence of a competitive gut microbiota.
Aim 2. In parallel to genetic screening, we monitored the within-host evolution of the five clinical isolates in conventional mice. The experiment has been pursued over a year. We have discovered that the isolates behave differently regarding the long-term stability of the colonization, certain strains being more stable than the others. Interestingly, our model recapitulates the colonization dynamics observed in humans with a highly variable duration of carriage spanning from few weeks to several months, for yet unknown reasons that we can now investigate in a tractable animal model. Isolated E. coli clones from colonized mice were isolated from fecal samples and sequenced on a monthly basis. We found that genetic drift (i.e. non-adaptive evolution) dominates in this selection regime. This will be further confirmed by evaluating the fitness effect of non-synonymous mutations detected in evolved clones with competition experiments.
Aim 3. Mutation profiling of isolated clones from long-term colonization experiments in different mice indicated transmission events. We observed that ESBL E. coli clones isolated from co-housed mice sometimes shared the exact same mutations demonstrating that transmission occurred in this model despite the low colonization levels, high colonization resistance from the microbiota and the presence of the same strain in the recipient mouse. Interestingly, we also observed that re-infection can occur via transmission, since mice in which E. coli was cleared were then re-infected with bacteria from another mouse. Our experimental system therefore mimics transmission within a household also documented in humans. We are currently testing if strain replacement events were caused by a set of adaptive mutations by performing controlled transmission experiments with evolved clones compared to ancestors.
In addition, we isolated virulent bacteriophages able to infect and kill the five E. coli isolates. We are assessing the impact of these phages on the population of E. coli in the gut and on the transmission dynamics of these strains. Our goal is to use these phages to accelerate the exclusion of ESBL E. coli from the gut of colonized hosts.
