Diversity and evolution of the ribovirian class Stelpaviricetes

Metatranscriptome mining has dramatically expanded the known diversity of ribovirians at all taxonomic levels. We explored the class Stelpaviricetes in the phylum Pisuviricota that currently includes three orders and six recognized virus families: order Stellavirales with family Astroviridae infecting vertebrates; Patatavirales with families Potyviridae, the largest known family of plant ribovirians, and Potyliviridae; and Hypofuvirales consisting of three families of viruses associated largely with phytopathogenic fungi. Using an order-specific collection of Hidden Markov Model profiles for the RNA-dependent RNA polymerase (RdRP), the hallmark protein of ribovirians, we identified 103 putative families within Stelpaviricetes, most of which share an astrovirus-like genome architecture, encoding three protein domains: the RdRP, a chymotrypsin-like protease, and a single jellyroll capsid protein. The topology of the phylogenetic tree of the RdRP implies that the common ancestor of Stelpaviricetes had an astrovirus-like genome. The ancestor of the order Patatavirales acquired two additional domains, a papain-like protease and a superfamily 2 helicase, whereas the capsid protein was replaced with an unrelated one, forming filamentous capsids. The ancestor of Hypofuvirales also encoded a superfamily 2 helicase and a papain-like protease, but had lost the capsid protein. Hypofuvirales might have derived from Patatavirales or evolved independently from an astro-like ancestor. For the vast majority of Stelpaviricetes identified in metatranscriptomes, host assignment remains elusive. Nevertheless, analysis of endogenous virus elements and isolated viruses provides some clues suggesting that the common ancestor of Potyviridae and Potyliviridae had a fungal host, whereas the common ancestor of Stelpaviricetes might have been a protist virus. IMPORTANCE The enormous diversity of viruses on Earth is only now coming to light through extensive metagenome and metatranscriptome mining. Using vast databases of ribovirus genomes, we performed phylogenomic analysis of the class Stelpaviricetes , which includes the families Astroviridae and Potyviridae, the former comprising animal viruses and the latter comprising plant viruses. We identified more than 100 family-level groups of viruses, most of which have small, astrovirus-like genomes, likely resembling the common ancestor of Stelpaviricetes . In contrast, the putative common ancestor of Potyviridae and its sister family Potyliviridae exhibits higher genome complexity that apparently emerged in viruses of fungi. The most plausible scenario for the origin of Potyviridae involves horizontal transfer of a virus across kingdoms between plant-associated fungi and plants.