Distribution of the phenotypic effects of random homologous recombination between two virus species

UMR BGPI Equipe 2 et Equipe 6;International audienceAuthor Summary : Recombination creates new genome combinations by joining genome fragments of distinct “parental” origin. This phenomenon, frequent in viral populations, combines mutations originally present on distinct parental genomes, increasing genetic diversity and creating “offspring” with altered biological properties. Consistently, recombination is often associated with the emergence of economically important viruses, with modified host range and virulence. In fact, recombination events can be lethal, deleterious or beneficial, but the respective frequency of these phenotypic effects is unknown and unpredictable. A generally accepted view, which we formally challenge in the present paper, is that most viral recombination events are deleterious or lethal when the parental sequences diverge by more than 10%. However, at present, no dedicated data set supports this supposition. We generated hundreds of “mosaic” genomes randomly from two plant virus species diverging by 18%, and tested a subset of 47 of these recombinants for viability and within-host accumulation. Surprisingly, all were viable, replicated, and accumulated at a pace comparable to that of the parents. Our results are in striking contrast to the current view, and show that viral recombination can have little phenotypic effect, at least in some cases, even when the parental sequences diverge by far more than 10%