Identification of diverse full-length endogenous betaretroviruses in megabats and microbats

Abstract

Background: Betaretroviruses infect a wide range of species including primates, rodents, ruminants, and marsupials. They exist in both endogenous and exogenous forms and are implicated in animal diseases such as lung cancer in sheep, and in human disease, with members of the human endogenous retrovirus-K (HERV-K) group of endogenous betaretroviruses (βERVs) associated with human cancers and autoimmune diseases. To improve our understanding of betaretroviruses in an evolutionarily distinct host species, we characterized βERVs present in the genomes and transcriptomes of mega- and microbats, which are an important reservoir of emerging viruses.

Results: A diverse range of full-length βERVs were discovered in mega- and microbat genomes and transcriptomes including the first identified intact endogenous retrovirus in a bat. Our analysis revealed that the genus Betaretrovirus can be divided into eight distinct sub-groups with evidence of cross-species transmission. Betaretroviruses are revealed to be a complex retrovirus group, within which one sub-group has evolved from complex to simple genomic organization through the acquisition of an env gene from the genus Gammaretrovirus. Molecular dating suggests that bats have contended with betaretroviral infections for over 30 million years.

Conclusions: Our study reveals that a diverse range of betaretroviruses have circulated in bats for most of their evolutionary history, and cluster with extant betaretroviruses of divergent mammalian lineages suggesting that their distribution may be largely unrestricted by host species barriers. The presence of βERVs with the ability to transcribe active viral elements in a major animal reservoir for viral pathogens has potential implications for public health.

Figure 1

A schematic representation of PaERV-βA. Two transcripts were identified in the P. alecto Illumina sequenced transcriptome that overlapped by 3,152 nt with 100% sequence identity which were used to assemble the PaERV-βA genomic sequence. Indicated are the retroviral genes gag, pro, pol, and env, which have been rendered defective by random mutation since integration. Also shown are the key enzymatic active sites of the viral protease (D×G), reverse transcriptase (DDD), and integrase (DDE); the betaretroviral dUTPase domain in pro; two unique open reading frames (ORFs); the polypurine tract (PPT); and the (Unique 3') (U3) region. ORF* does not appear to be genuine, but rather has arisen as a result of an insertion mutation that has disrupted a stop codon.

Figure 2

Phylogenetic relationships of bat and non-bat betaretroviruses. Maximum likelihood phylogenetic trees of (A) Gag, (B) Pol, and (C) Env amino acid sequences. Bootstrap values <70% are not shown, and branch lengths are drawn to a scale of amino acid substitutions per site. Bootstrap values are denoted as ** >90%; * >70% and < 90%. The trees are midpoint rooted for purposes of clarity only. βERV proteins of P. vampyrus and P. alecto are highlighted in red text. βERVs of M. lucifugus are highlighted in blue text. The clades within the Gag and Pol trees highlighted with a grey background (γ-Env) contain betaretroviruses whose Env sequence is not sufficiently closely related to the Env of other betaretroviruses to be included in the Env tree.

Figure 3

Phylogenetic comparison of the envelope (Env) protein sequence of betaretroviruses and gammaretroviruses. Bootstrap values <70% are not shown, and branch lengths are drawn to a scale of amino acid substitutions per site. Bootstrap values are denoted as ** >90%; * >70% and <90%. βERV proteins of P. vampyrus and P. alecto are highlighted in red text. βERVs of M. lucifugus and R. ferrumequinum are highlighted in blue text. Gammaretroviruses are highlighted in teal text.

Figure 4

Eight sub-groups of the betaretrovirus genus. A schematic diagram for a single representative of each group is depicted. Core retroviral genes gag, pro, pol, and env are bordered by the proviral long terminal repeats (LTRs). Also shown are other major genetic features such as open reading frames (ORFs) greater than 300nt in length and the rec gene of HERV-K113, enzymatic active site motifs of protease (D×G), reverse transcriptase (DDD), and integrase (DDE); the primer binding site (PBS) and polypurine tracts (PPT); and the characteristic betaretroviral dUTPase domain. NSR: non-sequenced region. ORF* is part of foreign nucleotide insertion within MlERV-βA and does not appear to be a retroviral element.

Figure 5

A proposed series of events leading to the current diversity in the genus Betaretrovirus. The proposed series of evolutionary events leading to eight distinct sub-groups of betaretroviruses based on a combination of the phylogenetic analyses of Gag, Pol, and Env protein sequences, and the genomic features and organizations of individual betaretroviruses.