Phage-induced diversification improves host evolvability.

Bacteriophage (viruses that infect bacteria) are of key importance in ecological processes at scales from biofilms to biogeochemical cycles. Close interaction can lead to antagonistic coevolution of phage and their hosts. Selection pressures imposed by phage are often frequency-dependent, such that rare phenotypes are favoured; this occurs when infection depends on some form of genetic matching. Also, resistance to phage often affects host fitness by pleiotropy (whereby mutations conferring resistance affect the function of other traits) and/or direct costs of resistance mechanisms. Here a simple model of bacteria and bacteriophage coevolving in a resource-limited chemostat is used to study the effect of coevolving phage on the evolution of bacterial hosts. Density-dependent mortality from phage predation limits the density of any single bacterial strain, preventing competitive exclusion by faster-growing strains. Thus multiple strains can coexist by partitioning resources and stable high diversity is created by negative frequency-dependent selection from phage. Standing bacterial diversity promotes adaptation in dynamic environments, since it increases the likelihood of a pre-existing genotype being suited to altered conditions. In addition, frequency-dependent selection for resistance creates transient local trade-offs between growth rate and resistance that allow bacterial strains to adapt across fitness valleys. Thus bacterial populations that (in the absence of phage) would have been trapped at sub-optimal local peaks in the adaptive landscape are able (in the presence of phage) to reach alternate higher peaks than could have been reached by mutation alone. This study shows that reasonable assumptions for coevolution of bacteria and phage create conditions in which phage increase the evolutionary potential of their hosts. Thus phage, in contrast to their deleterious effects on individual host cells, can confer an evolutionary benefit to bacterial populations. These findings have implications for the role of phage in ecosystem processes, where they have mainly been considered as a mortality factor; these results suggest that on long timescales phage may actually increase bacterial productivity by aiding the evolution of faster-growing strains. Furthermore, these results suggest that the therapeutic use of phage to treat bacterial infections (phage therapy) could have unintended negative side-effects.