A hemagglutinating agent, UPO216, was isolated from fresh wild bird droppings collected at the UPO wetlands (35°33′13.9”N 128°24′50.0”E) in the Gyeongnam province of Korea during an AI surveillance program in 2014 by inoculation into the allantoic cavities of 9–11-day-old specific pathogen-free (SPF) embryonated chicken eggs (ECEs, Valo BioMedia, USA) as previously described (Kang et al., 2010). The isolate was propagated in ECEs and UPO216 was named after the geological location of virus isolation. HA-positive allantoic fluids tested negative for AIV and APMV-1 by reverse-transcription polymerase chain reaction (RT-PCR) (Choi et al., 2012; Lee et al., 2014). The virus titer was measured by end-point titration in ECEs and expressed as the 50% egg infective dose (EID₅₀), as calculated by the Reed and Muench method (Reed and Muench, 1938).

For electron microscopy, UPO216 virus contained in allantoic fluid was concentrated by 20–60% (w/v) continuous sucrose gradient centrifugation at 90,000 × g for 1.5 h at 4°C in a Beckman SW41 rotor (Beckman, Germany). Each fraction was collected and tested for HA activity. HA-positive fractions were diluted in phosphate buffered saline (PBS, pH 7.4) and precipitated by centrifugation at 26,000 × g for 6 h at 4°C. The pellet was resuspended in PBS and kept at 4°C until use. Purified virus was adsorbed to 300 mesh copper grids at room temperature. The grids were negatively stained with 2% uranyl acetate and examined under a Hitachi 7100 electron microscope (Hitachi, Japan) at a magnification of × 100,000.

To serotype APMVs, a cross-HI test was performed using a reference panel comprising antigens and chicken anti-sera against representatives of APMV-1 to APMV-9 (except APMV-5; National Veterinary Service Laboratories, USA) using the method outlined in OIE (world organization for animal health) terrestrial manuals (OIE, 2012). Chicken antiserum homologous for the UPO216 virus was prepared by intravenously injecting 3-week-old SPF chickens with purified UPO216 virus (10^9.0 EID₅₀ per dose) as previously described (Miller et al., 2010a). The antigenic relatedness (R) between the two viruses was calculated using the method of Archetti and Horsfall (1951): R = (r₁ × r₂)^1/2, where r₁ is the ratio of the heterologous HI titer to the homologous HI titer for virus 1 and r₂ is the ratio of the heterologous HI titer to the homologous HI titer for virus 2.

The virulence of the virus was determined by the intracerebral pathogenicity index (ICPI) in day-old SPF chicks and the mean death time (MDT) in 9–11-day-old embryonated SPF ECEs, as previously described (Choi et al., 2012). All procedures were approved and supervised by the Institutional Animal Care & Use Committee of the Animal and Plant Quarantine Agency (QIA).

Viral RNA was extracted from the allantoic fluid using a QIAmp viral RNA mini kit (Qiagen, USA) according to the manufacturer′s instructions. Complementary DNA synthesis was performed using a LaboPass™ cDNA synthesis kit (Cosmogenetech, Korea). A combination of F gene-specific primers and primer walking was used to generate PCR amplicons covering the genome (except for both ends). The sequences of the 3′ and 5′ ends of the genome were amplified using the rapid amplification of cDNA ends (RACE) method (Li et al., 2005). All primer sets are described in Table 1. All PCR amplicons were sequenced using fluorescent dideoxynucleotide terminators and an automated sequencer (ABI 3730XL automated sequencer; Applied Biosystems Inc., USA).

Genome sequences of representative viruses belonging to the family Paramyxoviridae and sequences of each APMV serotype (all available in GenBank) were used for phylogenetic comparison. Editing and sequence analyses were performed using the BioEdit sequence alignment editor (Hall, 1999). Alignment of multiple nt and amino acid (aa) sequence alignments for the complete genome, the F gene, and the HN gene was performed using CLC Genomic Workbench 6.7.2 (CLC bio Aarhus, Denmark). Phylogenetic analysis was performed in MEGA 7.0 (Kumar et al., 2016) using both the maximum parsimony and maximum likelihood methods, with 1,000 bootstrap replicates. The evolutionary distance between and within APMV serotypes was determined using MEGA 7.0 (over 1,000 bootstrap replicates; Kumar et al., 2016).

Genome sequences of representative APMV serotypes were retrieved from GenBank public databases and used for the alignments. The GenBank accession numbers are as follows: APMV-1, accession nos. JF950510, AJ88027, and JF893453; APMV-2, accession nos. HM159993, HM159995, and HQ896023; APMV-3, accession nos. EU403085 and EU782025; APMV-4, accession nos. FJ177514 and EU877976; APMV-5, accession no. GU206351; APMV-6, accession nos. JN571486 and GQ406232; APMV-7, accession no. FJ231524; APMV-8, accession no. FJ619036; APMV-9, accession no. EU910942; APMV-10, accession no. HM159995; APMV-11, accession no. JQ886184; APMV-12, accession no. KC333050; APMV-13, accession nos. LC041132 and KX119151; and APMV-14, accession no. KX258200.

A total of 880 serum samples from wild birds in Korea were examined. The serum samples were originally taken from wild birds captured for an AI surveillance program; samples were kept at the AI laboratory, QIA, Korea. The serum samples were obtained from five Orders: Anseriformes, Charadriiformes, Ciconiiformes, Columbiformes, and Passeriformes covering 15 species. Sera were heat-inactivated at 56°C for 30 min before use in the serologic assay. Serological HI tests employing four HA units of test antigen in V-bottom microtiter plates were performed as described previously (OIE, 2012). The serum samples were first screened for the UPO216 virus isolated in the present study, and positive samples were further tested for cross-reaction with other APMV serotypes. The HI antibody titer was calculated as the reciprocal of the highest serum dilution that completely inhibited 4 HA units of antigens. HI titers ≥8 (3log₂) were considered positive. All tests were repeated twice.

The UPO216 virus was successfully propagated in ECEs, and the harvested infective allantoic fluid had a titer of 10^9.3 EID₅₀ per ml and a HA titer of 1,024–2,048 per 25 μl. Electron microscopy revealed that the UPO216 virus particles showed pleomorphic morphology, with a diameter of 80–300 nm. Glycoprotein projections of approximately 10 nm in length were evenly distributed across the envelope of the virus particle. Herringbone-like structures (nucleocapsids) were packaged inside the virion (Figure 1). These characteristics are typical of a paramyxovirus. In vivo pathogenicity testing revealed that UPO216 had a MDT index of >120 h and an ICPI-value of 0.00 (Table 2), indicating that it is avirulent in chickens. Serotyping revealed that UPO216 had a high HI titer (1:1,024) for homologous chicken antiserum (as expected), but low cross-reactivity (HI titers of ≤ 1:128) with several APMV serotypes, including APMV-1 and -9. The R between UPO216 virus and APMV serotypes was then calculated based on the HI titers obtained from the cross-HI test (Table 3). The UPO216 virus showed an R-value of <0.05 for all APMV serotypes except APMV-1 (R = 0.088) and APMV-9 (R = 0.125). The higher R-values for APMV-1 and APMV-9 were due to weak cross-reactions in the cross-HI test.

The nt sequence of the UPO216 virus was compiled from the sequences of 28 overlapping cDNA clones covering the entire genome. The UPO216 genome comprises 15,180 nts (accession no. KY511044). This length conforms to the “rule of six,” which plays an important role in the replication of paramyxoviruses (Kolakofsky et al., 1998).

A genome-wide BLASTN search (http://blast.ncbi.nlm.nih.gov/Blast.cgi) showed that UPO216 was closely related with known APMVs, especially with the highest identity with NDV strain 08-004 (accession no FJ794269.1; 75% identity) for N protein gene (positions 1–1,326) and with NDV strain BHG/Sweden/94 (accession no GQ918280; 72% identity) for L gene (positions 9,415–12,945). This indicates that the UPO216 virus might be an APMV.

The genome of UPO216 contains six non-overlapping transcriptional units in the following order: 3′-leader-NP-P-M-F-HN-L-trailer-5′ (Figure 2). Two additional proteins, V (245 aa) and W (140 aa), may arise during transcription of the P gene due to a putative RNA editing site (nt positions 2,292–2,300) in the viral genome. This occurs via addition of a single G residue to the editing site to yield a predicted V protein and the addition of two G residues to yield a predicted W protein, as is the case for APMV-1 (Steward et al., 1993). However, SH gene, that is found in APMV-6 (Chang et al., 2001), is not present between the F and HN genes. The 3′ leader sequence is 55 nt in length; this length is conserved among most APMV serotypes (Samuel et al., 2010; Samal, 2011). The length of the trailer at the 5′ end is 47 nt, the same as that in APMV-9 (Samuel et al., 2009; Figure 2). The first 12 nt of the leader sequence (3′-UGGUUU GUCUCU-5′) and the last eight nt of the 5′ trailer sequence (5′-ACCAAA CAAAGA-3′) show a high degree of homology (91.7%). The conserved sequences for the gene start (GS) and gene end (GE) of the UPO216 virus are UGC₃A/CUCUU and AAUNC/UU₅₋₆, respectively. The length of the intergenic region sequences ranges from 0 to 14 nt. The F protein cleavage sites within UPO216 possess the following unique aa sequence: ¹¹⁰L-V-Q-A-R-L¹¹⁵ (Table 2).

Comparison of the complete consensus sequences for the UPO216 genome with those of known APMV serotypes revealed that UPO216 is most closely related to APMV-1 (64.0% identity), followed by APMV-9 (56.4%), APMV-12 (52.9%), and APMV-13 (50.9%) (Table 4). UPO216 shows low homology with other APMV serotypes (37.0–41.1%).

A phylogenetic tree was constructed based on alignment of the complete nt sequences of the UPO216 genome and F gene with those of representative APMV serotypes (Figure 3). Phylogenetic analysis based on the complete sequences revealed that UPO216 formed a separate phylogenetic group along with APMV-1, -9, -12, and -13 (Figure 3A). Within this group, UPO216 is more closely related to APMV-1 and APMV-9 than to APMV-12 and APMV-13. APMV-3 viruses are closely related to APMV-4 viruses, APMV-5 viruses to APMV-6 and APMV-14 viruses, and APMV-8 to APMV-10 and APMV-2 viruses. Similar results were observed for phylogenetic analyses based on F gene sequences (Figure 3B).

To determine the genetic distance between the UPO216 and known APMV serotypes, we calculated the inter-serotype and intra-serotype evolutionary distances among APMV serotypes based on the sequences of complete genome (Table 5). At the genome level, estimates of the average inter-serotypic distance ranged from 0.43 to 1.62. The lowest inter-serotypic distance was between APMV-1 and UPO216 (0.43), followed by APMV-12 and APMV-13 (0.54), APMV-9 and UPO216 (0.64), APMV-2 and -10(0.69), and APMV-8 and APMV-10(0.69). When evolutionary distance was calculated with known serotypes of APMV, UPO216 was relatively close to APMV-1(0.43), APMV-9(0.64), APMV-12(0.78) and APMV-13(0.77), compared to other APMV serotypes (1.35–1.55). Estimates of intra-serotypic distances for APMV-1, -2, -3, and -6 were 0.21, 0.21, 0.30, and 0.26, respectively.

The 880 serum samples examined herein were first tested for antibodies specific for UPO216 in the HI test. Of these, 20 sera reacted with UPO216 (Table 6). These seropositive results were observed in sera obtained from wild ducks (Eurasian teal, European wigeon, mallard, Spot-Billed duck, and mandarin duck) belonging to the Order Anseriformes, but not in sera obtained from birds belonging to the Orders Charadriiformes, Ciconiiformes, Columbiformes, and Passeriformes. When seropositive sera (n = 20) were further tested in HI tests based on antigens derived from other serotypes of APMV, seven [Eurasian teal (n = 1), mallard (n = 4), European wigeon (n = 1), and Spot-Billed duck (n = 1)] did not cross-react with other APMV serotypes tested, and HI titers for UPO216 ranged from 3log₂ to >6log₂. The HI titers of the other 13 sera were the same (or higher) for APMV-1 as for UPO216.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.