Escherichia coli mutators: selection criteria and migration effect.

In silico, it has been shown that mutator alleles that increase mutation rate can be selected for by generating adaptive mutations. In vitro and in vivo, competition between wild-type bacteria and isogenic mutator mutants is consistent with this view. However, in vivo, the gain of the mutator seems to be reduced when migration is allowed. In vitro, the advantage of mutators has been described as frequency-dependent, leading to mutator advantage only when they are sufficiently frequent. Using an in vitro system, it is demonstrated that (i) the selection of mutators is frequency-independent, yet depends on at least one mutator bacterium bearing an adaptive mutation (its presence depends on chance, mutation rates and population size of mutator bacteria); (ii) on average, the mutator gain is always equal to the ratio of the adaptive mutation frequency of the mutator versus wild-type; (iii) when migration into an empty niche is allowed, the mutator benefit is reduced if migration occurs after fixation of the adaptive mutation into the wild-type population. It is concluded that in all cases, mutator gain depends directly on the ratio of bacteria carrying a beneficial mutation in mutator versus wild-type lineages.