Leviviricetes: expanding and restructuring the taxonomy of bacteria-infecting single-stranded RNA viruses

Abstract

The vast majority of described prokaryotic viruses have double-stranded or single-stranded DNA or double-stranded RNA genomes. Until 2020, a mere four prokaryotic single-stranded, positive-sense RNA viruses have been classified in two genera (Riboviria; Lenarviricota; Allassoviricetes; Leviviridae). Several recent metagenomic and metatranscriptomic studies revealed a vastly greater diversity of these viruses in prokaryotic soil communities than ever anticipated. Phylogenetic analysis of these newly discovered viruses prompted the reorganization of class Allassoviricetes, now renamed Leviviricetes, to include two orders, Norzivirales and Timlovirales, and a total of six families, 428 genera and 882 species. Here we outline the new taxonomy of Leviviricetes, approved and ratified in 2021 by the International Committee on Taxonomy of Viruses, and describe open-access hidden Markov models to accommodate the anticipated identification and future classification of hundreds, if not thousands, of additional class members into this new taxonomic framework.

Keywords: Leviviricetes; levivirus; metatranscriptomics; phage; virus classification; virus taxonomy.

Fig. 1.

Taxon demarcation criteria for leviviricetes classification. Taxonomic ranks for positive-sense single-stranded RNA (+ssRNA) viruses are shown alongside the demarcation criterion for each of the taxon ranks. PAAI, pairwise amino-acid sequence identity; CP, coat protein; RdRP, RNA-directed RNA polymerase; HMM, hidden Markov model; ORF, open reading frame.

Fig. 2.

HMM taxonomic predictions of Leviviricetes. While expanding and restructuring positive-sense single-stranded RNA (+ssRNA) viruses, nine genera could not be assigned with confidence to a family or an order, as the RNA-directed RNA polymerase (RdRP) and coat protein (CP) genome-encoded combinations did not adhere to established combinations. Additionally, there were ten instances (out of 882 species representative sequences) for which the HMM predicted taxonomy of +ssRNA viruses did not align with their phylogeny-based assignment. The colour and shape aesthetics of the phylogenetic tree illustrates these taxonomic outliers.

Fig. 3.

Examples of Leviviricetes genus and species demarcation cut-offs of 50 and 80 %, respectively, applied to pairwise RNA-directed RNA polymerase (RdRP) amino-acid sequence comparisons for members of norziviral Atkinsviridae. Inset (i) shows a distinct species clustering (red colouring), whereas inset (ii) shows three species represented by multiple sequences, and a species representing a single sequence, clustered into a genus (yellow-green colouring). Pairwise comparisons in shades of blue do not meet the set genus or species clustering criteria.