A phylogeny of the tymoviruses, sensu stricto, and its global interpretation in space and time.

Abstract

Maximum likelihood (ML) phylogenies of 109 tymoviruses, including three obtained directly from metagenomes, were calculated from all three open reading frames separately, but the concatenated sequences of their replicase and coat protein genes gave the most representative trees. ML phylogenies were also calculated from all recorded tymomvirus coat protein genes, and from datasets of the turnip yellow mosaic virus cluster, and separately of tomato yellow blotch, Andean potato latent and Andean potato mild mosaic viruses. These phylogenies showed that the basal divergence of tymoviruses occurred in a population infecting Eurasian brassicas (rosids), and more recently, one of the basal lineages diversified and adapted to infect some solanaceous (asterid) plants and crops of Central and South America. Heterochronous dating of the phylogenies failed, but heuristic comparisons based on patristic distances, branching patterns and external events suggested that the 'most recent common ancestor' of all known tymoviruses existed before the last Ice Age. Some lineages reached the Americas about 15,000 years ago. However, most spread of the few tymoviruses now found on more than one continent occurred during the past two centuries. The only recombinants were two sequences of Chiltepin yellow mosaic virus both with Nemesia ring necrosis virus as minor parent. Population genetic analysis found significant evidence of population contraction in the tymovirus populations infecting asterid hosts in the Americas. It also found the replicase and coat protein genes were significantly negatively selected. By contrast, the overlapping movement protein genes were positively selected which may help them adapt to new host species, including infecting economically significant crops. This knowledge about tymoviruses is important to plant biosecurity authorities.