We characterize the genome of the first reported deer parvovirus, Ungulate tetraparvovirus 5, which we detected by PCR in multiple tissues from 2/9 California mule deer (Odocoileus hemionus californicus) with hair loss syndrome (HLS) and in 4/12 deer without HLS, suggesting this common infection does not cause HLS.
Hair loss syndrome (HLS) has been observed in native populations of Columbian black-tailed deer (Odocoileus hemionus columbianus) throughout western Washington and Oregon, US, since the mid-1990s (Bildfell et al. 2004; Foreyt et al. 2004). Deer affected with HLS exhibit progressive hair loss over the thorax, flanks, and hindquarters, and progressive wasting and weakness often culminating in death, particularly in 6–12-mo-old fawns (Bender and Hall 2004; Bildfell et al. 2004; Foreyt et al. 2004). Postmortem examinations of affected deer have demonstrated heavy endoparasitism as well as heavy infestation with ectoparasites including lice, keds, and mites (Bender and Hall 2004; Bildfell et al. 2004; Foreyt et al. 2004). Investigators have postulated that multiple underlying conditions are associated with massive infestation with biting lice, either Damalinia (Cervicola) sp. or Bovicola tibialis (Bender and Hall 2004; Bildfell et al. 2004; Foreyt et al. 2004; Mertins et al. 2011), or with both. In Northern California, HLS was recognized in 2004 and, in 2009, was associated with high morbidity and mortality in a California mule deer (Odocoileus hemionus californicus) herd outside Yosemite National Park. Hair loss syndrome became epidemic in California during 2009–12. The movement of HLS down the western US, and the excessive endo- and ectoparasitism associated with the condition, made a search for an infectious immunosuppressive virus a necessary part of the investigation to elucidate the etiology of HLS.
We used metagenomics and deep sequencing to investigate a potential viral etiology for HLS. We analyzed the virome in multiple tissues (retropharyngeal lymph node, spleen, liver, lung, and kidney) from a California mule deer with HLS submitted to the California Animal Health and Food Safety Laboratory (CAHFS), School of Veterinary Medicine, University of California, Davis (UCD), California for routine diagnostic testing in 2012.
Viral nucleic acids were enriched and extracted from tissues as previously described (Victoria et al. 2008; Li et al. 2015). The nucleic acids from each tissue were pooled, a library was constructed using the TruSeq RNA library preparation kit (Illumina, San Diego, California, USA), and sequencing was done using a MiSeq platform (Illumina) with paired-end 250-base reads. Clonal reads and low sequencing quality tails and adaptors were removed, and remaining unique sequence reads were de-novo assembled (Deng et al. 2015). The assembled contigs, along with singlets, were aligned to an in-house viral proteome database using BLASTx (National Center for Biotechnology Information 2016) and using an E-value cutoff of <1×10−5. Four sequence reads of 112–186 bases showing similarity to tetraparvoviruses (BLASTx E<10−10) were identified. No other virus-related sequences were detected. The parvovirus sequence was extended by PCR to bridge the known regions and by modified 5′ and 3′ rapid amplification of cDNA ends for the extremities, yielding a near-complete genome sequence of 5,050 nucleotides. The retropharyngeal lymph node, spleen, and lung tissues of the index animal were positive by PCR. To our knowledge, this is the first parvovirus infection reported in deer (Cervidae).
The genome organization of deer parvovirus was typical of parvoviruses (Lau et al. 2008; Tse et al. 2011; Cotmore et al. 2014), with two large, nonoverlapping open reading frames (ORFs), and with ORF1 encoding a nonstructural polyprotein NS1 and ORF2 encoding overlapping VP1/VP2 capsid proteins (Fig. 1A). A small noncoding gap of 141 nucleotides was found between the two ORFs. For deer parvovirus, the predicted NS1 protein consists of 650 amino acids (aa) and contains conserved sequence features including helicase and ATPase domains, in agreement with the nonstructural functional role of NS1 in parvovirus replication (Lau et al. 2008; Tse et al. 2011; Cotmore et al. 2014). The VP1 protein is predicted to contain 933 aa, comparable to that of previously identified tetraparvoviruses. Conserved functional motif phospholipase A2 was identified in the VP1 region (Lau et al. 2008; Tse et al. 2011; Cotmore et al. 2014). Similar to other tetraparvoviruses, there exists a putative third ORF of 84 aa in a different reading frame within the VP1 region (Simmonds et al. 2008). Pairwise comparison of the NS1 and VP1 proteins showed deer parvovirus was most related to Ungulate tetraparvovirus 4, also known as ovine hokovirus from sheep (JF504699, 70% and 75% aa identity to NS1 and VP1, respectively).
The phylogenetic trees of NS1 (Fig. 1B) and VP1 (Fig. 1C) showed similar topologies and placed the deer parvovirus in the same clade as ovine and bovine hokoviruses (Lau et al. 2008; Tse et al. 2011; Cotmore et al. 2014). The current criteria for the family Parvoviridae proposed by the International Committee on Taxonomy of Viruses (ICTV) requires that the NS1 of distinct virus species in the same genus should be <85% identical at the aa sequence level (Cotmore et al. 2014). The genetic distance between the deer parvovirus and ovine and bovine hokoviruses (68.4–70.2% aa identity) indicated that it qualifies as a distinct species within the genus. Deer parvovirus is therefore proposed as the prototype for a new species (Ungulate tetraparvovirus 5) in the genus Tetraparvoviridae pending review by the ICTV (GenBank accession: KT878837). The phylogeny showed deer, goat (Capra hircus), and cattle (Bos taurus) tetraparvoviruses to cluster relative to those from nonruminant hosts including pigs (Sus scrofa), a bat (Eidolon helvum), and humans (Homo sapiens).
We conducted prevalence studies using nested PCR for the detection of viral DNA in extracts from spleen or lymph node tissues of another eight deer with, and 12 deer without, HLS signs, which were California deer samples archived by CAHFS, UCD. The oligonucleotide primer set was Deerhoko-F1: 5′- GTCTGGTTTGGTGGGAAGAAG −3′, Deerhoko-R1: 5′- GATGTTGGGAATGGCTGTGAC −3′; Deerhoko-F2: 5′- CGCTCCTGTTAGATTGGATGT −3′, Deerhoko-R2: 5′- AATAGCCGTAGAGGACTGATG −3′. The PCR cycling profile for both PCR rounds was 95 C for 5 min, 40 cycles with 95 C for 1 min, 57 C for 30 s, and 72 C for 1 min, and a final incubation for 10 min at 72 C. Amplicons with the expected size of 338 base pairs were confirmed by direct sequencing, showing sequence conservation >99%. Viral DNA was detected in tissues of 1/8 of deer with hair loss and 4/12 deer without HLS. Including the index case, the prevalence of deer parvovirus was higher in deer without than in deer with HLS (4/12 vs. 2/9), suggesting that Ungulate tetraparvovirus 5 detection was not associated with HLS.
Parvovirus DNA was detected in retropharyngeal lymph node, spleen, and lung tissues, indicating that different organs were likely infected or contaminated with viremic blood. Detection of parvoviral DNA in two animals without HLS but from herds with undiagnosed mortality events may indicate a possible role in other diseases, but further studies will be required to elucidate whether and under what circumstances this virus may affect the health of deer.
We thank the Blood Systems Research Institute and National Institute of Health (R01-HL-105770) for financial support as well as the School of Veterinary Medicine, University of California, Davis for guidance and assistance with sample collection. We also thank Pam Swift from the California Department of Fish and Wildlife for submitting valuable animal samples.