Multiple groups of endogenous betaretroviruses in mice, rats, and other mammals

Abstract

Betaretroviruses exist in endogenous and exogenous forms in hosts that are widely distributed and evolutionarily distantly related. Here we report the discovery and characterization of several previously unknown betaretrovirus groups in the genomes of Mus musculus and Rattus norvegicus. Each group contains both mouse and rat elements, and several of the groups are more closely related to previously known betaretroviruses from nonmurine hosts. Some of the groups also include members from hosts which were not previously known to harbor betaretroviruses, such as the gray mouse lemur (Microcebus murinus) and Seba's short-tailed bat (Carollia perspicillata). Some of the mouse and rat elements possess intact open reading frames for gag, pro, pol, and/or env genes and display characteristics of having retrotransposed recently. We propose a model whereby betaretroviruses have been evolving within the genomes of murid rodents for at least the last 20 million years and, subsequent to (or concomitant with) the global spread of their murid hosts, have occasionally been transmitted to other species.

FIG. 1.

Neighbor-joining trees based on alignments of betaretrovirus pol nucleotide (a) and Pol amino acid (b) sequences. The aligned sequences are from the same ∼738-bp or 246-amino-acid region spanning the eight conserved regions of retroviral RT proteins described by Jacobo-Molina and Arnold (15). The trees are rooted on the β1 group of retroviruses. Names of previously known betaretroviruses are surrounded by dashed boxes, and those of newly discovered nonmurine retroviruses are surrounded by solid boxes. Bootstrap values (out of 1,000 replicates) are shown at the nodes. Shapes correspond to those in Fig. 3 and indicate groups determined based on the Env tree.

FIG. 2.

ORF structures of selected murine betaretroviruses. The three forward reading frames are shown for each element, with vertical black lines indicating stop codons. The gag, pro, pol, and/or env ORFs from each element are indicated, with Δ indicating the presence of premature termination and/or frameshift mutations.

FIG. 3.

Neighbor-joining tree based on alignment of TM amino acid sequences. The aligned sequences are from the region spanning the cleavage site and the TM domain of the Env protein, as described by Bénit et al. (4). Names of previously known retroviruses are surrounded by dashed boxes, and those of newly discovered nonmurine retroviruses are surrounded by solid boxes. Shapes correspond to those in Fig. 1 and indicate groups determined based on the Env tree.

FIG. 4.

Dot plots of aligned common insertions in mouse and rat betaretroviral elements. Alignments and dot plots were generated as described in Materials and Methods. In all cases, the mouse and rat elements are on the horizontal and vertical axes, respectively. The grey boxes represent the retroviral regions. (a) Flanking sequences of 7.5 kb on either side of the pol region of MmERV-β2_NT_039339 and RnERV-β2_NW_0433361. (b) Flanking sequences of 5 kb on either side of RMER16 LTRs in mouse (accession number AC115914; horizontal axis) and rat (accession number AC095664; vertical axis) genomes. (c) Data in this panel are the same as described for panel b with mouse (accession number AC131039; horizontal axis) and rat (accession number AC125654; vertical axis) RMER16 LTRs. Notations along axes are accession numbers followed by the coordinates of the first and last nucleotides of the aligned sequence in the contig or clone and the orientation of the sequence relative to the contig or clone. +, element and contig or clone are in the same orientation; −, element and contig or clone are in opposite orientations.