Une approche intégrative pour caractériser et prédire la mort cellulaire et l'échappement aux traitements antibiotiques par beta-lactamine ; An integrative approach to characterize and predict cell death and escape to beta-lactam antibiotic treatments

Resistance to first-line antimicrobial drugs is now commonly encountered. In particular, the increasing fraction of commensal and pathogenic Escherichia coli expressing extended-spectrum beta-lactamases and/or carbapenemases is alarming. E. coli is a major cause of common infections such as urinary tract infections, affecting over 150 million people worldwide. Importantly, many infections relapse. Therefore, an in-depth understanding of the susceptibility of E. coli clinical isolates to beta-lactams is essential for proposing effective treatments.Bacteria might escape treatments in many different ways. Resistant bacteria grow and divide normally in the presence of antibiotics. Their characterization is easy using standard diagnostic tests. Resilient bacteria merely survive in the presence of antibiotics and regrow when the antibiotic is removed or degraded. This biphasic behavior complicates the prediction of treatment outcomes. Resilience to treatment is notably observed in collective antibiotic tolerance, where dead cells release beta-lactamases degrading the antibiotic in the environment. Standard approaches are not adapted for quantifying and understanding the role of resistance and/or resilience.Our main objectives are to quantify the dynamics of cell death during repeated treatments and to quantify the impact of different growth conditions on cell death. First, we developed novel protocols to address variability issues in optical density measurements, and to perform colony forming unit assays in an efficient manner. Using these techniques, we generated an extensive dataset describing the impact of repeated treatments on different clinical isolates. We calibrated a previously developed in the team model of population response to antibiotic and evolution of the environment in the context of collective antibiotic tolerance. We calibrated the model to our dataset, and we showed that the model accounts for the temporal evolution of both biomass and live cell counts. Further, we demonstrated that using this ...