Ortervirales: New Virus Order Unifying Five Families of Reverse-Transcribing Viruses

LETTER

Reverse-transcribing viruses, which synthesize a copy of genomic DNA from an RNA template, are widespread in animals, plants, algae, and fungi (1, 2). This broad distribution suggests the ancient origin(s) of these viruses, possibly concomitant with the emergence of eukaryotes (3). Reverse-transcribing viruses include prominent human pathogens, such as human immunodeficiency viruses 1 and 2 (HIV-1/2) and hepatitis B virus, as well as plant pathogens that cause considerable economic losses (4).

The International Committee on Taxonomy of Viruses (ICTV) traditionally classified reverse-transcribing viruses into five families: Caulimoviridae, Hepadnaviridae, Metaviridae, Pseudoviridae, and Retroviridae (5). In 2018, the ICTV recognized an additional family, Belpaoviridae, which contains the genus Semotivirus (previously included in the Metaviridae family [6]). The infection cycles, nucleic acid types, genome organizations, and virion morphologies of these viruses are very diverse. Indeed, reverse-transcribing viruses are distributed between two Baltimore classes of viruses. Belpaoviruses, metaviruses, pseudoviruses (better known as Bel/Pao, Ty3/Gypsy, and Ty1/Copia retrotransposons, respectively [1, 7]), and retroviruses typically encapsidate single-stranded RNA (ssRNA) genomes (Table 1) and frequently integrate into the host genomes as part of their replication cycles (Baltimore class VI). In contrast, members of the families Caulimoviridae and Hepadnaviridae, often referred to as pararetroviruses (8), package circular double-stranded DNA (dsDNA) genomes and do not actively integrate into host chromosomes (Baltimore class VII). However, capture of pararetroviral DNA in host genomes, presumably by illegitimate recombination, is commonplace, particularly in plants, giving rise to the corresponding endogenous elements (9, 10).

Mechanistic studies of the replication cycles of reverse-transcribing viruses of different families have revealed many similarities, which have been reinforced by comparative genomics of the viral reverse transcriptases (RTs), the hallmark enzymes encoded by all reverse-transcribing viruses. Indeed, phylogenetic analyses support the monophyly of all viral RTs, excluding those encoded by nonviral retroelements from both eukaryotes and prokaryotes (11, 12). In addition to sharing RT phylogeny, belpaoviruses, caulimoviruses, metaviruses, pseudoviruses, and retroviruses share several conserved features that hepadnaviruses lack (Table 1). In particular, the polymerase (Pol) polyproteins of belpaoviruses, metaviruses, pseudoviruses, and retroviruses possess similar domain architectures. These Pol polyproteins contain an aspartate protease, which is responsible for the processing of viral polyproteins, and an integrase of the DDE recombinase superfamily. The genomes of these viruses also share long terminal repeats (LTRs) (13). Within certain clades, Pol polyproteins of retroviruses and metaviruses share additional features, such as a dUTPase domain (14–16) and the GPY/F subdomain of the integrase (17, 18). Caulimoviruses also possess a homologous aspartate protease domain in their Pol polyprotein (19) but lack an integrase and LTRs. However, RT-based phylogenies consistently classify these plant-infecting viruses as a sister clade of the metaviruses (Fig. 1). This phylogenetic position suggests that among pararetroviruses, encapsidation of a DNA genome is a homoplasious trait and, therefore, is not a reliable criterion for classification. The basal branches of the RT tree are not resolved and are presented as a multifurcation in Fig. 1. This topology is at least compatible with placing the Hepadnaviridae clade outside the viral group that includes belpaoviruses, caulimoviruses, metaviruses, pseudoviruses, and retroviruses.

Belpaoviruses, caulimoviruses, metaviruses, pseudoviruses, and retroviruses share not only homologous proteins involved in genome replication and polyprotein processing but also the two principal protein components of the virions, namely, the capsid and nucleocapsid proteins/domains (20–22). However, the nucleocapsid domain appears to be absent in spumaretroviruses (family Retroviridae) (Table 1). In contrast, hepadnaviruses encode an unrelated capsid protein (23). These findings suggest that belpaoviruses, caulimoviruses, metaviruses, pseudoviruses, and retroviruses have evolved from a common viral ancestor, rather than from distinct capsid-less retrotransposons (20).

Finally, similarities between belpaoviruses, caulimoviruses, metaviruses, pseudoviruses, and retroviruses extend to the mechanism of replication priming. All these viruses utilize host tRNA molecules as primers for genome replication by reverse transcription (24), whereas hepadnaviruses use a specific protein priming mechanism mediated by the polymerase terminal protein domain (25).

Taken together, the common complement of proteins required for genome replication, polyprotein processing, and virion formation, the topology of the RT phylogenetic tree, and mechanistic similarities in genome replication present strong evidence that belpaoviruses, caulimoviruses, metaviruses, pseudoviruses, and retroviruses share an evolutionary origin. The hepadnaviruses, which typically (i) branch out at the base of the viral RT clade (Fig. 1), (ii) possess a unique capsid protein, and (iii) employ a distinct replication mechanism, appear to be more distantly related to all these virus families. In recognition of these relationships, the ICTV has recently regrouped the families Belpaoviridae, Caulimoviridae, Metaviridae, Pseudoviridae, and Retroviridae into an order, Ortervirales (Orter, an inversion of “retro,” which was derived from reverse transcription; -virales, suffix for an order). This change in taxonomy acknowledges and formalizes the long-proposed evolutionary relationship among most groups of reverse-transcribing viruses (26). We note that although hepadnaviruses are not included in the order, they might be unified with other reverse-transcribing viruses at a higher taxonomic level in the future.