This study was performed in strict accordance with local ordinance and the recommendations by the Committee on the Use of Live Animals in Teaching and Research (CULATR) at The University of Hong Kong. All specimens of bats, monkeys, cats, and dogs were collected with the assistance of the Department of Agriculture, Fisheries and Conservation, Hong Kong Special Administrative Region (HKSAR); those of pigs and cattle were collected with the assistance of the Department of Food, Environmental and Hygiene, HKSAR, from various locations in HKSAR; and those of horses were collected with the assistance of the Hong Kong Jockey Club. All specimens of rabbits were collected from live food animal markets in Guangzhou, China, in October 2007. Rectal swabs were collected using procedures described previously (Lau et al., 2005). All fecal samples of marine mammals, including Indo-Pacific bottlenose dolphins, California sea lions and harbor seals, were collected by veterinary surgeons of the Ocean Park in HKSAR (Woo et al., 2014a). A total of 791 samples collected over a 75-month period (October 2007 to December 2013) from 157 bats, 52 monkeys, 100 pigs, 58 cats, 58 dogs, 50 cattle, 106 rabbits, 95 horses, 46 Indo-Pacific bottlenose dolphins, 54 California sea lions, and 15 harbor seals were tested (Table 1).

Viral RNA was extracted from rectal and cloacal swabs and fecal samples using EZ1 Virus Mini Kit v2.0 (Qiagen, Germany). RNA was eluted in 60 μl of AVE buffer (Qiagen, Germany) and about 200 ng of RNA was used as template for RT-PCR.

Genogroup I PBV screening was performed by PCR amplification of a 205-bp fragment of the RdRp gene of genogroup I PBVs using conserved primers (5′-CAAARTTYGACCARCACTT-3′ and 5′-TCRTCDGCRTTGGTACCACC-3′) designed by multiple alignments of the available RdRp genes of PBVs. Reverse transcription was performed using the SuperScript III kit (Invitrogen, USA) and the reaction mixture (10 μl) contained RNA, first-strand buffer (50 mM Tris-HCl pH 8.3, 75 mM KCl, 3 mM MgCl₂), 5 mM DTT, 50 ng random hexamers, 500 μM of each dNTPs and 100 U Superscript III reverse transcriptase. The mixtures were incubated at 25°C for 5 min, followed by 50°C for 60 min and 70°C for 15 min. The PCR mixture (25 μl) contained cDNA, PCR buffer (10 mM Tris-HCl pH 8.3, 50 mM KCl, 2 mM MgCl₂), 200 μM of each dNTPs and 1.0 U Taq polymerase (Applied Biosystems, USA). The mixtures were amplified in 60 cycles of 94°C for 1 min, 50°C for 1 min and 72°C for 1 min and a final extension at 72°C for 10 min in an automated thermal cycler (Applied Biosystems, USA).

All PCR products were gel-purified using the QIAquick gel extraction kit (Qiagen, Germany). Both strands of the PCR products were sequenced twice with an ABI Prism 3730xl DNA Analyzer (Applied Biosystems, USA), using the two PCR primers. As multiple nucleotide peaks were observed in most sequencing results, it was suggested that more than one type of PBV were present in each sample, thus the purified PCR products were cloned into the pCR-II-TOPO TA cloning vector (Invitrogen, USA) according to manufacturer’s instructions. Both strands of 10 clones for each sample were sequenced, using primers 5′-TAATACGACTCACTATAGGG-3′ and 5′-CGGCTCGTATGTTGTGTGGA-3′. The sequences of the clones were compared with known sequences of the RdRp of PBVs in the GenBank database.

Nine complete segments 1 and 15 complete segment 2 of otarine PBVs were amplified and sequenced using published strategies for double-stranded RNA viruses (Attoui et al., 2000), using RNA extracted from the original specimens of sea lions positive for PBV as template. Viral RNA was extracted using the EZ1 virus mini kit (Qiagen, Germany). Adaptor primer, with 3′ NH₂ blocking group, was ligated to the 3′ termini of the viral RNA and subjected to reverse transcription using complementary primer. After RNA hydrolysis, reannealing and end-filling, single-primer amplification of viral genomic segments was performed using complementary primer and genome specific primers. The 5′ and 3′ ends of the viral genomes were confirmed by rapid amplification of cDNA ends using the 5′/3′ RACE kit (Roche, Germany). The PCR products were gel purified and sequenced using an ABI Prism 3700 DNA analyzer (Applied Biosystems, USA). Sequences were assembled and manually edited to produce the final sequences of the viral genomes.

The nucleotide sequences of the genomes and the deduced amino acid sequences of the ORFs were compared to those of other PBVs. Novel genes were further predicted by FGENESV (SoftBerry, Inc.¹), a trained pattern/Markov chain-based viral gene prediction program. Phylogenetic tree construction was performed using the maximum likelihood method and MEGA7 (Kumar et al., 2016), with bootstrap values being calculated from 1,000 trees. The optimal substitution model for each ORF was selected by MEGA7. Protein domain, family and functional site analyses were performed using ScanProsite (De Castro et al., 2006). Transmembrane and coiled-coil domains were predicted by TMHMM Server v 2.0 (Krogh et al., 2001) and COILS (Lupas et al., 1991) respectively.

For real-time quantitative PCR assays, cDNA were amplified in SYBR Green I fluorescence reactions (Roche, Germany). Briefly, 10 μl of reaction mixtures containing 1 μl cDNA, 10 μl FastStart DNA master SYBR green I mix reagent (Roche) and 5 mM each of forward and reverse specific primers were thermal-cycled at 95°C for 10 min followed by 45 cycles of 95°C for 10 s, 60°C for 10 s and 72°C for 20 s using a Roche LightCycler 96 real time PCR system (Roche, Germany). Specific primers for each PBV segment were designed based on the sequences of all the segment 1 and segment 2 detected in the positive samples (Table 3). Plasmids with the corresponding target sequences were used for generating the standard curve. At the end of the assay, PCR products were subjected to a melting curve analysis (65–95°C, 0.1°C/s) to confirm the specificity of the assay.

A total of 791 fecal specimens from 676 terrestrial mammals and 115 marine mammals were obtained (Table 1). RT-PCR for a 205-bp fragment in the RdRp gene of genogroup I PBVs was positive in specimens from six cattle, six monkeys, 17 horses, nine pigs, one rabbit, one dog, and 12 sea lions. Marked nucleotide polymorphisms were observed in most of the RdRp sequences, suggesting the possible existence of multiple strains in the same specimen. Therefore, the PCR products were cloned and 10 clones from each specimen were sequenced. Multiple sequence types were confirmed to be present in most samples. Sequence analysis of these clones revealed that there were 11, 9, 23, 17, 1, 1, and 15 sequence types in the positive specimens from the five cattle, six monkeys, 17 horses, nine pigs, one rabbit, one dog, and 12 sea lions respectively, and 47.4–89.0% nucleotide identities were observed between these clones and the corresponding sequences of other genogroup I PBV strains available in the GenBank database (Figure 1). The PBV sequences from each kind of animal were widely distributed in the whole phylogenetic tree with high diversity (Figure 1). No PBV was detected in the specimens obtained from the 58 cats, 157 bats, 46 Indo-Pacific bottlenose dolphins, and 15 harbor seals (Table 1).

Nine complete segment 1 of otarine PBVs from sea lions were sequenced and assembled (Table 2; Figure 2). These segment 1 sequences ranged from 2,158 to 2,522 bases in length with overall G+C contents of 41.1–46.0%. The 5′ non-coding regions (44–169 bases) were AU-rich (G+C contents of 22.7–37.9%) with five conserved bases, GUAAA, located at the 5′ end. A predicted highly stable stem loop structure found in other known PBVs was observed in five segment 1 sequences as a result of the pairing of 5′-GUAAA-3′ and 5′-UUUAC-3′ in the 5′ non-coding region (Nates et al., 2011). The 3′ non-coding regions (19–32 bases) contained G+C contents ranging from 53.1 to 71.4% and end with 3–4 conserved bases (CTC, CTTC, or CACC). All the nine segment 1 sequences possess one long ORF (1,590–1,728 bp) encoding the capsid protein of 529–575 amino acids. These capsid proteins shared low (19.3–37.2%) amino acid identities with those of other PBV strains, being most closely related to turkey PBV TK/MN/2011 (GenBank number KJ495689), fox PBV F5-1 (GenBank number KC692367) and human PBV Hy005102 (GenBank number NC_007026). Upstream to the ORF for the capsid protein, there were one to two short ORFs in the nine segment 1, consistent with the organization of the segment 1 in other known PBVs (Wakuda et al., 2005; Bodewes et al., 2013; Banyai et al., 2014; Verma et al., 2015). The protein encoded by the ORF2 of segment 1 from the nine otarine PBVs possessed different numbers of repetitions of the same motif, ExxRxNxxxE, that was also observed in the corresponding protein in other known PBVs (Da Costa et al., 2011).

Fifteen complete segment 2 from the otarine PBVs were sequenced and assembled (Table 2; Figure 2). These segment 2 sequences ranged from 1,679 to 1,943 bases in length with overall G+C contents of 39.5–48.1%. The 5′ non-coding regions (38–295 bases) were also AU-rich (G+C contents of 18.4–35.9%) with the same conserved bases, GUAAA, located at the 5′ end. The stable stem loop structure observed in segment 1 was also observed in the 5′ non-coding regions of 13 segment 2 sequences. The 3′ non-coding regions (38–46 bases) have G+C contents ranging from 31.6 to 56.1% and end with four conserved bases (CUGC) in most of the genomes. All the 15 segment 2 sequences possess one long ORF (1,590–1,620 bp) encoding the RdRp of 529–539 amino acids. These RdRp shared 44.5–70.6% amino acid identities with those of other genogroup I PBV strains, being most closely related to fox PBV F5-1 (GenBank number KC692366), human PBV 1-CHN-97 (GenBank number AF246939), human PBV GPBV6C1 (GenBank number AB517731), human PBV HuPBV-E-CDC16 (GenBank number KJ663816) and human PBV VS10 (GenBank number GU968924). They possess three conserved motifs (D-T/S-D, SG-T, GDD) commonly found in the RdRp sequences of other dsRNA viruses. Conserved cysteine and proline residues present in other genogroup I PBVs were also observed in all 15 segment 2 sequences. In contrast to the segment 2 sequences of other known PBVs, six of our 15 sequenced segment 2 possess a previously undescribed ORF 48–71 amino acids upstream to the RdRp ORF. Multiple alignments of the sequences of these six ORFs showed that they formed three groups that were not homologous to each other, with three ORFs belonging to the first group, two ORFs to the second group and one ORF to the third group (Figure 2). In the first group, the three ORFs showed 97.6–99.4% nucleotide identities among each other. In the second group, the two ORFs showed 100% nucleotide identity. Sequence analysis of these ORFs did not reveal any significant sequence homology to other proteins in the GenBank database. Protein sequence analysis also did not reveal any significant matches to other known protein domains, families or functional sites in the PROSITE database. Moreover, there are no transmembrane and coiled-coil domains predicted by sequence analyses in these protein sequences.

In the phylogenetic trees constructed using ORF2 (Figure 3A) and ORF3 (capsid protein) (Figure 3B), the nine sequenced segment 1 of the otarine PBVs were clustered into four distinct clades (C1–C4) (Figures 3A,B). The sequence types that belonged to each clade were identical for both trees, suggesting that there was no recombination between different PBVs. In the phylogenetic tree constructed using the RdRp gene, the 15 sequenced segment 2 of the otarine PBVs were clustered into nine distinct clades (R1–R9) (Figure 3C). Although, R1, R6, R7, and R8 as well as R3, R4, and R5 seemed to be further clustered in the tree, there were >14% amino acid difference between any two clades.

Quantitative RT-PCR showed that the amounts of otarine PBV RNA ranged from 1.7 × 10⁴ to 5.9 × 10⁶ (for segment 1) and 4.1 × 10³ to 1.3 × 10⁶ (for segment 2) copies/ml in the fecal samples (Table 3).

Fecal samples were serially obtained from 18 sea lions over 6 years (Table 4). In the 12 samples that were positive for PBV, segment 2 could be detected and sequenced in all 12 samples, but segment 1 could only be detected and sequenced in seven samples due to difficulties in designing PCR primers as a result of the limited number of segment 1 PBV sequences available in GenBank. Overall, samples positive for PBV were collected mainly in 2008 and 2009. Among these positive results, most clades of segment 1 and segment 2 were observed in samples from either 2008 or 2009, while only segment 1 clades C3 and C4 and segment 2 clade R1 were present in samples from both 2008 and 2009. More than one clade of segment 2 was present in three of the 12 samples and more than one clade of segment 1 was present in two of the seven segment 1 positive samples. In four sea lions (Victory, Camy, Caddy, and GiGi), PBVs were detected in samples collected from two different years (Table 4). Among these four sea lions, the same segment 1 clade (C3) was present in two consecutive years in one sea lion (GiGi), with a total of 18 nucleotide changes in the capsid coding region of the same clade (C3) detected between 2008 and 2009. In the other three sea lions (Victory, Camy and Caddy), different clades were present in the fecal samples obtained in different years.

The genome sequences of otarine PBVs obtained from the present study were deposited in GenBank with accession numbers KU729746–KU729769.