Whole-genome sequences of Odocoileus hemionus deer adenovirus isolates from deer, moose and elk are highly conserved and support a new species in the genus Atadenovirus

Abstract

We present the first complete genome sequence of Odocoileus hemionus deer adenovirus 1 (OdAdV-1). This virus can cause sporadic haemorrhagic disease in cervids, although epizootics with high mortality have occurred in California. OdAdV-1 has been placed in the genus Atadenovirus, based on partial hexon, pVIII and fibre genes. Ten field isolates recovered from naturally infected mule deer (Odocoileus hemionus), white-tailed deer (Odocoileus virginiana) and moose (Alces alces) from Wyoming, black-tailed deer (Odocoileus hemionus columbianus) from California, and Rocky Mountain elk (Cervus elaphus nelsoni) from Colorado and Washington state were sequenced. The genome lengths ranged from 30 620 to 30 699 bp, contained the predicted proteins and gene organization typical of members of genus Atadenovirus, and had a high percentage of A/T nucleotides (66.7 %). Phylogenic analysis found that the closest ancestry was with ruminant atadenoviruses, while a divergence of the hexon, polymerase and penton base proteins of more than 15 % supports classification as a new species. Genetic global comparison between the 10 isolates found an overall 99 % identity, but greater divergence was found between those recovered from moose and elk as compared to deer, and a single variable region contained most of these differences. Our findings demonstrate that OdAdV-1 is highly conserved between 10 isolates recovered from multiple related cervid species, but genotypic differences, largely localized to a variable region, define two strains. We propose that the virus type name be changed to cervid adenovirus 1, with the species name Cervid atadenovirus A. Sequence data were used to develop molecular assays for improved detection and genotyping.

Fig. 1.

The genome organization of OdAdV-1 is depicted schematically (a). Potential ORFs were identified using blastx search to identify homologous proteins in GenBank. The genome sequences between the 10 isolates were highly identical (99 %), but had a variable region within the A/T-rich noncoding region (NCR). Multiple sequence alignment of the 10 isolates from the variable region is shown (b). This region contained two deletions totalling 54 nt (22 nt+32 nt, O.h. and O.v.) or three deletions totaling 75 nt (O.h.c.) compared to moose and elk (A.a., C.e.).

Fig. 2.

Maximum-likelihood phylogenic trees depicting the relationship between the 10 OdAdV-1 isolates in this study based on whole-genome sequences (a). Maximum-likelihood trees estimating the relationship between OdAdV-1 and the other atadenoviruses based on full amino acid sequences of the penton base (b), hexon (c) and polymerase proteins (d). Trees were constructed using PhyML in SeaView version 3.2 with a BioNJ starting tree and bootstrap values calculated from 1000 pseudo-replicates. A single OdAdV-1 sequence (99 A.a.) was used in (b)–(d) due to isolate identity in these proteins. Atadenoviruses: ovine adenovirus 7 (OAdV^T, type species), duck adenovirus 1 (DAdV), psittacine adenovirus 3 (PsAdV), possum adenovirus 1 (PossAdV), lizard adenovirus 2 (LzAdV), snake adenovirus 1 (SnAdV), and bovine adenovirus 4 and 6 (BAdV-4 and 6). Representative ruminant viruses from the genus Mastadenovirus were included as outliers, bovine adenoviruses 1 and 3 (BAdV-1 and 3). Bootstrap values and branch lengths (b)–(d) are given. The accession numbers for the protein reference sequences retrieved from GenBank are listed in Table 2.

Fig. 3.

Primers specific for OdAdV-1 were designed for a diagnostic PCR and a genotyping PCR. Assays were run using DNA extracted from additional diagnostic cases that were positive for OdAdV-1 or other viral agents to test for specificity. (a) The diagnostic PCR amplified a 502 bp amplicon, a 100 bp molecular weight marker, and OdAdV-1 from mule deer (lanes 1, 3), elk (lane 2), pronghorn (lanes 4, 8) and moose (lane 11). No amplification product was produced with samples positive for bovine, canine, or ovine adenovirus of undetermined type (lanes 5–7), chlamydophila from a mule deer (lane 9), or the no template control (lane 10). (b) The genotyping PCR was designed to span part of the variable region and produced amplicons of 680, 701, or 733 bp, depending on the presence or absence of deletions: lanes 1–4, California black-tailed deer (680 bp due to 53 bp deletion); lanes 5–8, pronghorn, white-tailed deer, mule deer, (701 bp due to 32 bp deletion); and lane 8–9, mule deer with the moose/elk variation (733 bp with no deletions). Traditional Sanger sequencing using forward and reverse primers confirmed the expected sequence changes.