Abstract
Most living organisms on the tree of life can be infected by viruses. The ubiquity of viruses is driven by different factors including high mutation rates, high population sizes and low generation times, which allow for quick adaptation to very different host species. The dynamics of adaptation - the rate of change of the mean fitness of the viral population - results from the interplay between multiple evolutionary forces that may promote or hamper viral adaptation. But the interactions between these different factors may often be difficult to understand. During this PhD we developed a combination of theoretical and experimental approaches to disentangle the influence of some of these factors on viral adaptation.First, we explored the dynamics of viral adaptation to a homogeneous host population. We used Fisher’s Geometric Model of adaptation and studied the joint evolutionary and epidemiological dynamics of a viral population spreading in a host population. This modeled allowed us to explore the lethal mutagenesis hypothesis: is it possible to treat viral infections with mutagenic drugs to increase the mutation load of the viral population beyond a threshold that may result in the extinction of the within-host population? We show which parameters affect the critical mutation rate leading to viral extinction and we show how epidemiology and evolution can affect the transient within-host dynamics of the viral population when a single virus life-history trait (transmission rate) is under selection. We extend this modeling framework to study the joint evolution of transmission and virulence during the adaptation of an emerging pathogen. At the beginning of an epidemic, these two traits are expected to evolve independently but a trade-off may build up with viral adaptation.Second, we studied viral adaptation in heterogeneous host populations when the virus spreads among a diversified population of resistance host. We studied the evolutionary emergence of viruses: can viruses avoid extinction by the acquisition of escape mutations allowing them to infect some of the resistant hosts in the population? We developed a simple birth-death process to predict the probability of evolutionary emergence as a function of the composition of the host population. In particular, we show how the proportion of multiple resistant hosts can reduce the risk of pathogen evolutionary emergence. We put some of these predictions to the test using bacteriophages spreading in bacterial populations. We manipulate the diversity of CRISPR immunity in Streptococcus thermophilus bacteria and we confirm the key influence of multiple resistance on the risk of viral adaptation.Third, we also studied viral adaptation in time-varying environments where the host population is allowed to coevolve with the virus. In this experimental project we monitored the adaptation of bacteriophages as they coevolved with the CRISPR immunity of S. thermophilus bacteria. We track reciprocal adaptive changes in which bacteria acquire new layers of resistance (new spacers in the CRISPR array) and phages acquire new escape mutations in the corresponding protospacers. This experiment allows us to monitor the dynamics of viral adaptation across time and space. Interestingly, we find a significant asymmetries in competitive abilities among different bacterial strain in the absence of phage predation. This asymmetric competition has dramatic consequences on the maintenance of diversity of host resistance and on the coevolutionary dynamics with the virus. This thesis demonstrates the possibility to use experimental evolution with microbial microcosms to explore the validity of some theoretical predictions on the dynamics of viral adaptation. This experimental validation is particularly important if one wants to use evolutionary models to make public-health recommendations.