Final Report Summary - GENPHENBACT (Genetic and Phenotypic Modelling of Bacterial Evolution)
With the rise of antibiotic resistance, it is now clear that an evolutionary perspective is necessary to study infectious diseases. Escherichia coli, a commensal bacteria of the gut is of particular interest in that regard. E. coli is killing about a million person every year, is becoming more virulent, more resistant to antibiotics and is also the work horse of molecular biology allowing the use of many different technologies. In GenPhenBact, our goal was to use a quantitative biology approach combining theoretical models and experiments to enlighten E. coli evolution at four different levels: the gene, the network, the genome and the species. For that purpose we used a multidisciplinary approach encompassing microbiology, experimental evolution, molecular genetics, population genetics, whole genome sequencing, epidemiology, animal models, deep mutational scans, and modelling. At the gene and network level, the study of thousands to tens of thousands of mutants has allowed us to uncover the molecular determinants of mutation effects and their integration into mechanistic models. Working on the antibiotic resistance gene TEM-1, we showed that a generic mechanistic models of protein evolution centred on the impact of mutation on protein stability could largely but not exclusively explain mutation effects. These data also lead to better characterise at the metabolic level how cell are killed by the antibiotic’s action. At the network level, using the simplest possible system, we showed that the amazing diversity of mutation interaction that emerged could be fully explained in a mechanistic model taking into account metabolism, intermediate toxicity and expression costs. At the genome and species level, with the study of hundreds of genomes, we have characterised the dynamics of genome evolution when adapting to in vitro conditions, to the mice gut and in the gut a healthy human subject. We revealed how artificial conditions favoured a fast mutation accumulation at the genome level that gradually decreased but remained sustained, how adaptation to more natural conditions resulted in more limited traces of adaptation or to a lack of them in the wild. Finally the mutation effect predictions and the genome dynamics approaches were combined to show that asexual populations with high mutation rate could, despite a marked adaptation, accumulate slightly deleterious mutations at a rate almost similar to the one found in lineages decaying in the absence of selection. The contrasted patterns of genome evolution observed in the wild and the ones observed in vitro call for a deeper characterisation of the selective forces at play in natural conditions.