Avian paramyxovirus serotype‐1 isolation from migratory birds and environmental water in southern Japan: An epidemiological survey during the 2018/19–2021/2022 winter seasons

Newcastle disease caused by highly pathogenic viruses of avian paramyxovirus serotype‐1 (APMV‐1) is a highly contagious poultry disease. Although a large‐scale epidemic of Newcastle disease had occurred in Japan between the 1950s and the 2000s, there have been no outbreaks anywhere since 2010. In addition, there are no reports of epidemiological surveys of APMV‐1 in wild birds in Japan in the last 10 years. We conducted the first epidemiological survey of APMV‐1 in the Izumi plain, Kagoshima prefecture of southern Japan from the winter of 2018 to 2022. A total of 15 APMV‐1 strains were isolated, and isolation rates from roosting water and duck fecal samples were 2.51% and 0.10%, respectively. These results indicate that the isolation method from environmental water may be useful for efficient surveillance of APMV‐1 in wild birds. Furthermore, this is the first report on the success of APMV‐1 isolation from environmental water samples. Genetic analysis of the Fusion (F) gene showed that all APMV‐1 isolates were closely related to virus strains circulating among waterfowl in Far East Asian countries. All isolates have avirulent motifs in their cleavage site of F genes, all of which were presumed to be low pathogenic viruses in poultry. However, pathogenicity test using embryonated chicken eggs demonstrated that some isolates killed all chicken embryos regardless of viral doses inoculated (102–106 50% egg infectious dose). These results indicated that APMV‐1 strains, which are potentially pathogenic to chickens, are continuously brought into the Izumi plain by migrating wild birds.


INTRODUCTION
Newcastle disease (ND) is a highly contagious avian disease with primarily respiratory, gastrointestinal, and neurological symptoms. It is an important veterinary disease that causes serious global economic loss to the poultry industry. Newcastle disease virus (NDV) that causes ND is a highly pathogenic virus of avian paramyxovirus serotype-1 (APMV-1), which belongs to the Orthoavulavirus genus of the Paramyxoviridae family. 1 In addition to enhanced biosecurity, ND vaccines are gold standards for preventing ND epidemics in the poultry industry. However, recurrent global outbreaks suggest that differences in genotype and antigenicity of viral proteins between currently used ND vaccine strains and epidemic viruses in the field may reduce vaccine efficacy. [2][3][4][5] Although a large-scale epidemic has been observed in Japan between the 1950s and the 2000s, there have been no reports of ND outbreaks since 2010 owing to the success of preventive measures using the ND vaccine.
Although APMV-1 has a single serotype, its genome is genetically divided into two subtypes, namely, class I and class II. Class I viruses comprise a single genotype, while class II viruses are divided into 18 genotypes. [6][7][8] Class I viruses are generally nonpathogenic for chickens. 9 The virulence of class II viruses varies by genotype, with most highly pathogenic isolates reported as genotypes V, VI, VII, VIII, and XI-XVIII. [6][7][8]10 The commonly used vaccine strains, La Sota, B1, and VG/GA, which belong to class II genotype II, are lentogenic viruses causing mild infections in the chicken respiratory tract. Recently, virulence prediction of APMV-1 has become possible based on subtype and genotype, owing to the increasing number of studies on genomic analysis of isolated viruses around the world. 1,11 Wild aquatic birds, which are the natural hosts of APMV-1, show asymptomatic infection or mild clinical signs, even when infected with strains that are highly pathogenic in chickens. 12,13 Therefore, wild birds play an important role in the circulation and spread of APMV-1 across countries. 14, 15 Mase et al. 16 have reported that APMV-1 isolated from Japanese poultry between 2001 and 2005 is genetically related to the virus belonging to the class II genotype VII, which is prevalent in South Korea and China. In a concurrent epidemiological survey in Japan, APMV-1 of the class II genotype VII has also been isolated from wild cormorants, suggesting that the virus prevailing in East Asia may have invaded Japan via the migration of wild birds. 16 However, there are no reports of epidemiological surveys of APMV-1 in wild birds in Japan around the last 10 years.
The Izumi plain in Kagoshima Prefecture of southern Japan is one of the largest overwintering areas for a wide variety of aquatic birds that migrate from Alaska, the Russian Far East, Siberia, Mongolia, China, and other Far East Asian regions during winter season. The area around the overwintering site is also a region where the poultry industry is thriving. Therefore, extensive control measures are required in poultry farms located in the Izumi plain for various infectious diseases transmitted from wild birds. As several outbreaks of highly pathogenic avian influenza have occurred in chicken farms in the Izumi plain, epidemiological surveillance of avian influenza in wild birds at the overwintering site has been conducted and a large number of avian influenza viruses have been isolated. [17][18][19] However, because APMV-1 isolation from wild birds has not been conducted, the prevalence of the virus in the Izumi plain remains unclear. In this study, to understand the status of APMV-1 in the overwintering area of the Izumi plain, APMV-1 isolation and genetic analysis were conducted from November 2018 to March 2022.

Virus isolation
Water samples from roosts of wild birds and duck fecal samples were collected in the Izumi plain from November to March of 2018/19, 2019/20, 2020/21, and 2021/22. We collected water samples (50 mL/sample, 14 sampling points) weekly from wet paddy areas that were created as roosting sites for wild birds. As shown in Table 1 seasons, respectively. Virus isolation was conducted according to the influenza virus isolation procedure from water samples using embryonated chicken eggs. 18 In brief, water samples were centrifuged at 2000g for 5 min at 4°C to remove debris and a 1/10 volume of 10-fold concentrated PBS and 100 μL of chicken red blood cells (RBCs) were added and mixed. The mixture was incubated on ice for 1 h to adsorb potential hemagglutinating viruses to RBCs. After incubation, RBCs were collected by centrifugation at 2000g for 5 min at 4°C, and resuspended in 1 mL of viral transport medium containing 0.5% BSA, 10,000 U/mL penicillin, 10 mg/mL streptomycin, 0.3 mg/mL gentamicin, and 2.5 μg/mL amphotericin B. The RBC resuspension was directly inoculated into 9-to 11-day-old embryonated chicken eggs. As shown in Table 1, the total number of fecal samples was 1021; 96, 475, and 450 samples were collected in the 2018/19, 2019/20, and 2021/22 seasons, respectively. No samples were collected in the 2020/21 season. Virus isolation from fecal samples was conducted using embryonated chicken eggs as previously described. 20 The eggs inoculated with each sample were cultured at 37°C for 3 days. Following cooling of eggs at 4°C for several hours, the allantoic fluid was collected and a hemagglutination assay was performed. Allantoic fluid with hemagglutination activity was tested to confirm the presence of avian influenza virus using ESPLINE INFLUENZA A&B-N (FUJIREBIO Inc., Tokyo, Japan). When the rapid influenza viral antigen test was negative, the harvested allantoic fluid was stored at −80°C as a viral fluid that may contain APMV-1 for subsequent experiments.

APMV-1 identification using genetic analysis
Nucleic acid extraction from the allantoic fluid containing hemagglutinating virus was performed using the innuPREP Virus DNA/RNA Kit (Analytic Jena, Jena, Germany). To synthesize a DNA strand complementary to the viral genomic RNA, a reverse transcriptase reaction was performed using the PrimeScript II 1st strand cDNA Synthesis Kit (Takara Bio Inc., Shiga, Japan) according to the manufacturer's recommended protocol. Subsequently, PCR was performed using Takara Ex Taq DNA Polymerase (Takara Bio) with the reverse transcription product and specific primers to distinguish between class I and class II APMV-1 (class I: NDV-C1-F and NDV-C1-R; class II: NDV-C2-F and NDV-C2-R). 21 Following agarose electrophoresis of PCR products, amplified complementary DNA (cDNA) was purified using the Monarch DNA Gel Extraction Kit (New England Biolabs, Ipswich, MA). To determine the sequence of the amplified cDNA, sequencing analysis was conducted with Big-dye terminator v3.1 cycle sequencing kit and ABI 3130 sequencer (Applied Biosystems, Wartham, MA). Sequence similarity searches were performed using the BLAST (from the NCBI) to identify the hemagglutinating virus isolate as an APMV-1. All isolates in this study showed more than 98.9% sequence homology to previously reported APMV-1 strains listed the NCBI database.
The nucleotide sequences determined in this study are available in the DDBJ (DNA Data Bank of Japan) database under accession numbers LC715129-LC715143. The strains obtained in this study were named as host or isolation source/Kagoshima/KU-strain number/year. In this study, the strains are referred to by corresponding the abbreviation listed in Table 2.

Phylogenetic analysis
Based on partial nucleotide sequences (class I: 368 bp and class II: 459 bp) in the Fusion (F) gene of APMV-1 isolates, a molecular phylogenetic tree was constructed using the neighbor-joining method 22 with Molecular Evolutionary Genetics Analysis version 10.1.6. 23 The robustness of the neighbor joining analysis grouping was assessed using resampling with 1000 bootstraps.

Pathogenicity of APMV-1 isolates in chicken embryos
To evaluate the virulence of APMV-1 isolates against chicken embryos, the mean death time (MDT) was determined as described by Alexander 24 with some modifications. About 100 μL of each virus solution adjusted to 10 2 , 10 4 , and 10 6 EID 50 (50% egg infectious dose) was inoculated into the allantoic cavity of five 9-day-old embryonated chicken eggs, followed by incubation at 37°C. The eggs were observed once every 12 h for 7 days, and the time of embryo death was recorded. The MDT was determined when all embryos inoculated with each dose of the virus were killed. Japanese laws on animal research state that experiments with fertilized eggs performed before hatching do not require animal ethics approval.

Isolation of APMV-1 from duck feces and environmental water
From November 2018 to March 2022, 1021 duck fecal samples and 558 environmental water samples were collected when wild waterfowl were staying in the overwintering grounds of the Izumi plain. A total of 15 APMV-1 strains were isolated from these specimens and the summary of the results is shown in Tables 1 and 2. Genetic analysis showed that the putative amino acid sequences of the cleavage site of F protein in all APMV-1 isolates were either 112 E-R-Q-E-R-L 117 or 112 G-K-Q-G-R-L 117 (Table 2), which are generally found in nonpathogenic viruses. According to the criteria for APMV-1 pathogenicity, all isolates in this study were likely to be low pathogenic or nonpathogenic viruses. 1

Phylogenetic analysis of APMV-1 isolates
Phylogenetic tree analysis and BLAST searches were performed on F genes of the 15 APMV-1 isolates. Phylogenetic analysis showed that six of the APMV-1 isolates, Env/195D, Env/195I, Env/198 G, Env/199E, Env/ 2111D, and Env/21106, were classified as subtype class I (Figure 1). Nine strains, Dk/1930, Env/1934, Env/1936, Env/1995, Env/202E, Env/2032, Env/203D, Env/2011G, and Env/218 G, were classified as subtype class II genotype Ib (Figures 1 and 2). BLAST search based on sequences of the F gene revealed that all the 15 isolates showed high homology to several isolates from Anseriformes in Taiwan, China, and Russia (data not shown). These results revealed that all APMV-1 isolates in this study were genetically related to viruses that were maintained in waterfowl in the East Asian region, suggesting that APMV-1 has been brought from broad areas of the East Asian countries into Izumi plain by migration of wild birds.
All viruses isolated from ND outbreaks in Japan to date are classified into class II genotypes II, III, VI, VII, and VIII. 16,25 The phylogenetic analysis also showed that all APMV-1 isolates in this study were not genetically related to any virulent Japanese NDV isolates ( Figure 2).

Pathogenicity of APMV-1 isolates in chicken embryos
To evaluate the virulence of APMV-1 isolates in chicken embryos, pathogenicity tests were conducted (Table 3). A total of 14 strains, 12 strains isolated in the present study and 2 vaccine strains B1 and VG/GA, were tested. Alexander 24 defined APMV-1 to be low pathogenic or nonpathogenic (>90 h), intermediate virulence (60-90 h), and velogenic (<60 h), according to the MDT of embryonated chicken eggs inoculated with a stock virus of a tenfold series (10 −5 to 10 −9 ) dilution. In this study, adjusted stock viruses with infective titers of 10 2 , 10 4 , or 10 6 EID 50 , which correspond to approximately each stock virus of 10 6 -, 10 4 -10 2 -fold dilutions, respectively, were inoculated into chicken eggs. In the inoculum dose of 10 2 EID 50 , most APMV-1 isolates including vaccine strains did not kill all inoculated embryos and the MDTs of all embryos with fatal infection of Env/198G and Env/1995 were over 90 h, indicating that these viruses were not highly pathogenic  24 In the inoculum dose of 10 4 and 10 6 EID 50 , all embryos infected with Env/195D, Env/198G, and Env/1995 were dead within 5 days postinoculation regardless of viral inoculated dose and their MDT ranged from 29 to 91 h ( Table 3).

DISCUSSION
In this study, an epidemiological survey for APMV-1 was conducted in the Izumi plain, Kagoshima prefecture of southern Japan, and 15 APMV-1 strains were isolated from duck fecal samples and wild bird roost water (environmental water) samples collected from 2018 to 2022 (Table 2). To the best of our knowledge, this is the first report on APMV-1 isolation from environmental water. Horizontal distances are proportional to the minimum number of nucleotide differences required to join nodes and sequences. The isolates described in this study are shown in underlined bold font. The tree was generated using the neighbor-joining algorithm and alignments were bootstrapped 1000 times.
In epidemiological studies of APMV-1 in Japan, the isolation rates of the virus from wild waterfowl feces have been in the range of 0.15%-1.4% in the San-in district of western Japan from 1997 to 2012, 0.31%-1.06% in the Tohoku district of northeastern Japan from 2006 to 2009, and 0.91% in a national survey from 2011 to 2013. 20,[26][27][28][29][30] The APMV-1 isolation rate from duck feces in this study was 0.10% (Table 1) with a slightly lower rate compared with isolation rates in previous studies in Japan. Alternatively, the total isolation rate of APMV-1 from environmental water collected in the same area of Izumi plain was 2.51% (Table 1), which was higher than previously reported rates from waterfowl feces. 20,[26][27][28][29][30] As the virus isolation method from environmental water samples is easier than that of fecal samples and enables more efficient virus isolation, this method may be useful for efficient surveillance of APMV-1 in overwintering area of wild birds. A month-by-month comparison of isolation rates from environmental water samples showed that APMV-1 were isolated between November and January (2.34%-3.82% isolation rate), and not in February and March (Table 1). In previous surveillance of avian influenza virus in the Izumi plain, the most frequent isolation rate of the virus from environmental water was observed between November and December. 18 It has been known that young birds flying from nesting sites located in Eurasia are susceptible to many pathogens because of their undeveloped immune systems. 18,31 It is possible that pathogens introduced by migrating infected young birds contaminate the environment (i.e., waterfronts), leading to waterborne transmission to other birds. This may be the reason for the high prevalence of APMV-1 and avian influenza viruses in the wild bird population that migrates to the Izumi plain in early winter. We proposed that enhanced control measure in early winter season is required to prevent the transmission of wild bird-derived pathogens into poultry farms in this region.
Three types of in ovo or in vivo tests, (1) MDT, (2) intracerebral pathogenicity index, and (3) intravenous pathogenicity index, have been commonly used to evaluate the pathogenicity of APMV-1. 32 It has been postulated that the cleavage site motif of the F protein of APMV-1 is a major determinant of viral virulence. The typical cleavage site sequences for highly pathogenic viruses are 112 (R/K)-R-Q-(R/K)-R-F 117 , and in contrast, those of low pathogenic viruses were 112 (G/E)-(K/R)-Q-(G/E)-R-L 117 . 11 Although all APMV-1 isolated in this study have typical motifs for avirulent viruses in cleavage site of F protein (Table 2), the Env/195D isolate were classified as intermediate virulence in in ovo test: the pathogenicity test on chicken embryos (Table 3). These data suggest that there may be diversity of the APMV-1 isolates in susceptibility and virulence to chickens. To prevent APMV-1 from invading poultry farms, it is necessary to confirm the efficacy of currently used vaccines against field isolates. Our preliminary studies using sera from chickens vaccinated with VG/GA and B1 strains on commercial farms in the Kagoshima prefecture showed that the neutralizing antibody titers against our isolates (Env/199E and Dk/1930) were lower than those against the vaccine strain (Supporting Information). This result may be T A B L E 3 Pathogenicity of APMV-1 isolates in chicken embryos.

Isolate
Subtype-genotype Viruses belonging to class I or class II genotype I are often isolated from wild birds, and furthermore most of them showed low or nonpathogenicity to chickens. 9,12,33 However, several cases of ND outbreaks caused by highly pathogenic viruses classified into class I and class II genotype I have previously been reported. An outbreak of ND on Irish poultry farms in 1990 was caused by a highly pathogenic virus classified as class I, which was extremely similar in antigenic and genetic properties to nonpathogenic viruses from wild waterfowl. 34,35 In addition, highly pathogenic virus isolates during outbreaks at Australian poultry farms from 1998 to 2000 were genetically similar to a nonpathogenic waterfowl-origin virus classified as class II genotype Ia. 36 These reports suggest that nonpathogenic APMV-1 maintained in wild bird populations changed into highly pathogenic viruses, resulting in ND outbreaks in chicken populations. It has been reported that experimental passages in chicks lead nonpathogenic waterfowl-originated APMV-1 isolates belonging to class I to become highly virulent with 100% mortality in infected chickens. 37,38 Notably, since 2007, APMV-1 isolates from wild birds in Japan, Alaska, and Russia have a faster rate of base substitutions in the F gene than earlier isolates. 33 Based on amino acid sequence of cleavage site of F proteins, all APMV-1 isolates in this study were classified as nonpathogenic viruses; however, some of them showed enhanced pathogenicity to chicken embryos (Table 3). Therefore, measures to prevent APMV-1 maintained in the wild bird population from invading surrounding poultry farms in the Izumi plain are important to avoid the emergence of highly pathogenic mutant viruses.
In this study, the first epidemiological survey of APMV-1 in the overwintering site of wild birds in the Izumi plain, Kagoshima prefecture, southern Japan, was conducted from 2018 to 2022. Genetic analyses indicate that low pathogenic and/or nonpathogenic APMV-1 circulating in the East Asian countries invaded into Japan via wild bird migration, and a few isolates have the potential to be pathogenic in chickens. In addition, isolation events of highly pathogenic APMV-1 (NDV) from wild birds have been continued in East Asia. Therefore, it is important to take enhanced control measures for ND in poultry farms located in the Izumi plain and continuous monitoring of APMV-1 invading the area is necessary.