Microbial Evolution at Multiple Scales

Microbes live in complex environments, in which their evolution is shaped by many different selection pressures. These selection pressures may act at different levels of organization. The studies presented in this thesis use mathematical and computational modelling to study microbial evolution as the multilevel process it often is. It covers examples of different microbes (human viruses, bacteriophages, and bacteria) that are exposed to selection pressures at different scales in time and in space. Firstly, the evolution of HIV is addressed under the – possibly opposing – selection pressures it experiences within a single host (short-term adaptation to fast within-host replication), and at the level of the population (long-term adaptation to transmission). It is shown that long-lived latent reservoirs of cells, in which the viral DNA is integrated but does not replicate, can strongly affect the multilevel evolution of the virus by providing an archive of old viral variants that promotes population level evolution of between-host transmission. Secondly, the evolution of regulation by local density cues is studied in two different model systems: the small-peptide communication system that was recently discovered to control the lysis-lysogeny decision in a range of phages, and the regulation of toxin production in bacteria. In the first case, the evolution of phage-phage communication is shown to occur only under specific conditions, namely if the phages cause repeated outbreaks in large pools of susceptible bacteria. In the second case, the distribution of bacteria over space is found to be crucial for the evolution of density-based regulation, and bacteria are selected on the structure of the colonies they produce. Lastly, the final research chapter of this thesis describes an extension to the Price equation that formalises the effects of spatial structure on evolution. This mathematical analysis shows how selection can be decomposed into components that act within and among local environments for any length scale, and thus allows us to quantify the effects of different scales of spatial organisation on natural selection. Taken together, the work presented in this thesis increases our understanding of microbial evolution as the multiscale process it is.