Discovery and characterization of mammalian endogenous parvoviruses

Abstract

Public databases of nucleotide sequences contain exponentially increasing amounts of sequence data from mammalian genomes. Through the use of large-scale bioinformatic screening for sequences homologous to exogenous mammalian viruses, we found several sequences related to human and animal parvoviruses (PVs) in the Parvovirus and Dependovirus genera within genomes of several mammals, including rats, wallabies, opossums, guinea pigs, hedgehogs, African elephants, and European rabbits. However, phylogenetic analysis of these endogenous parvovirus (EnPV) sequences demonstrated substantial genetic divergence from exogenous mammalian PVs characterized to date. Entire nonstructural and capsid gene sequences of a novel EnPV were amplified and genetically characterized from rat (Rattus norvegicus) genomic DNA. Rat EnPV sequences were most closely related to members of the genus Parvovirus, with >70% and 65% amino acid identities to nonstructural and capsid proteins of canine parvovirus, respectively. Integration of EnPV into chromosome 5 of rats was confirmed by PCR cloning and sequence analysis of the viral and chromosomal junctions. Using inverse PCR, we determined that the rat genome contains a single copy of rat EnPV. Considering mammalian phylogeny, we estimate that EnPV integrated into the rat genome less than 30 million years ago. Comparative phylogenetic analysis done using all known representative exogenous parvovirus (ExPV) and EnPV sequences showed two major genetic groups of EnPVs, one genetically more similar to genus Parvovirus and the other genetically more similar to the genus Dependovirus. The full extent of the genetic diversity of parvoviruses that have undergone endogenization during evolution of mammals and other vertebrates will be recognized only once complete genomic sequences from a wider range of classes, orders, and species of animals become available.

FIG. 1.

(A) Genetic map of EnPV sequences encoding proteins with significant similarity to the structural proteins of canine parvovirus (CPV) and adeno-associated virus (AAV). Identity of amino acid residues and hydrophobicity scores of each sequence are plotted (sliding window size of 5 amino acids) to show the relatedness between structural proteins of mammalian EnPV and ExPV. Blocks in the protein alignment represent aligned amino acid residues, and lines represent gaps or missing data. In the identity plot, green represents amino acid residues conserved in >50% of sequences. In the hydrophobicity plot, red indicates higher concentrations of hydrophobic amino acids. Chromosome numbers for EnPV sequences of wallabies, guinea pigs, and armadillos are not known. (B) Protein alignment of the complete structural protein of rat EnPV with the other members of the genus Parvovirus. Amino acid residues identical in all sequences are shown in blue, amino acid residues identical in >50% sequences are shown in green, and variable residues are shown in a white background. (C) Results of specific PCR and inverse PCR for rat EnPV. Approximate sizes of amplification products are shown on the gel. MW, molecular weight marker; Neg, nontemplate control; K, kidney; Li, liver; Br, brain. L2 is a lung epithelial cell line, AvrII, SpeI, BspHI, AflII, and KpnI are restriction enzymes used for DNA digestion.

FIG. 2.

Genomic organization and phylogenetic relationship of rat EnPV. (A) Map of chromosome 5 of rats (Rattus norvegicus) showing integration site of EnPV, open reading frame (ORF) length, and orientation (arrowheads of yellow boxes) and organization of capsid (VP) and nonstructural (NS) protein genes. The LINE-like element identified at the end of NS ORF is shown as a double-headed arrow. Sequence diversity plots for corresponding chromosome regions in genomes of mice, humans, dogs, and opossums showed the absence of rat EnPV sequences in other animals. (B) VP and NS protein sequences of known exogenous parvovirus reference strains available in GenBank were used to determine the genetic relatedness of rat EnPVs by neighbor-joining analysis of pairwise distances between translated amino acid sequences. Accession numbers of sequences used precede the names of corresponding parvovirus species. Bootstrap resampling was used to determine robustness of individual clades (values above 70% are shown above the branches).

FIG. 3.

Genetic diversity and phylogenetic analysis of the capsid and nonstructural proteins of EnPVs. Phylogenetic trees were constructed by neighbor-joining of pairwise protein distances based on alignments generated by ClustalW implemented in MEGA4.1. For clarity the accession numbers of sequences used precede the names of the corresponding parvovirus species. Bootstrap resampling was used to determine robustness of individual clades (values above 70% are shown above the branches). All trees are condensed to show the genetic relatedness not the distances. (A) VP gene sequences of different parvovirus species representing all genera of subfamily Parvovirinae, unclassified parvoviruses (human parvovirus 4 and turkey and chicken parvoviruses), and one species of Densovirinae were phylogenetically compared to the VP gene-like sequences of EnPVs from different animals. The two different genetic groups of EnPVs based on VP protein genetic relatedness are designated A and B (details in Results). (B) NS protein sequences of different parvovirus species representing all genera of subfamily Parvovirinae, unclassified parvoviruses (human parvovirus 4 and turkey and chicken parvoviruses), and one species of Densovirinae were compared with the NS protein-like sequences of EnPVs from different animals. The four different genetic groups of EnPVs based on NS protein genetic relatedness are designated A to D and are described in detail in Results. CH, chromosome (CH-, unknown chromosome number); EST, expressed sequence tags; BoPV, bovine parvovirus; CnMV, canine minute virus; HBoV, human bocavirus.