Characterization of H9N2 Avian Influenza Viruses Isolated from Poultry Products in a Mouse Model

Low pathogenic H9N2 avian influenza viruses have spread in wild birds and poultry worldwide. Recently, the number of human cases of H9N2 virus infection has increased in China and other countries, heightening pandemic concerns. In Japan, H9N2 viruses are not yet enzootic; however, avian influenza viruses, including H5N1, H7N9, H5N6, and H9N2, have been repeatedly detected in raw poultry meat carried by international flight passengers from Asian countries to Japan. Although H9N2 virus-contaminated poultry products intercepted by the animal quarantine service at the Japan border have been characterized in chickens and ducks, the biological properties of those H9N2 viruses in mammals remain unclear. Here, we characterized the biological features of two H9N2 virus isolates [A/chicken/Japan/AQ-HE28-50/2016 (Ck/HE28-50) and A/chicken/Japan/AQ-HE28-57/2016 (Ck/HE28-57)] in a mouse model. We found that these H9N2 viruses replicate well in the respiratory tract of infected mice without adaptation, and that Ck/HE28-57 caused body weight loss in the infected mice. Our results indicate that H9N2 avian influenza viruses isolated from raw chicken meat products illegally brought to Japan can potentially infect and cause disease in mammals.


Introduction
H9N2 avian influenza viruses are of low pathogenicity in chickens but still cause mild respiratory symptoms and a drop in egg production, leading to economic losses in countries that rely on the poultry industry [1]. The H9N2 virus were first isolated in turkeys in the United States in 1966 [2,3]. Since the late 1990s, H9N2 viruses have been widespread in various species of wild birds and domestic poultry across the world, and have occasionally expanded their host range to mammalian species, such as pigs and humans [4]. The first case of human infection with a H9N2 virus was reported in China in 1998 [5]; since then, as of 20 December 2021, a total of 95 laboratory-confirmed cases of human infection with H9N2 viruses have been reported mainly in China, but also in other countries including Egypt, Bangladesh, Cambodia, Oman, Pakistan, India, and Senegal [6,7]. In addition, recent genetic analyses have revealed that H9N2 avian influenza viruses have contributed to the genetic and geographic diversity of H5N1 avian influenza viruses through genetic reassortment [8,9]. Furthermore, novel H7N9 avian influenza viruses, which emerged in 2013 and have infected a substantial number of people in China, are reassortants whose six internal genes come from H9N2 viruses [10]. Other subtypes of avian influenza viruses, such as H5N6, H5N8, and H10N8, can also possess the internal genes of H9N2 viruses due to genetic reassortment [11]. The worldwide distribution of H9N2 viruses with the ability to infect mammals and humans is increasing concern regarding their pandemic potential.
The HA gene of the H9N2 virus are classified into the Eurasian and American lineages. The Eurasian lineage is further classified into several sub-lineages, including A/duck/Hong Kong/Y280/1997 (Y280), A/quail/Hong Kong/G1/1997 (G1), and A/duck/Hong Kong/Y439/1997 (Y439). Y280-like viruses represent the dominant lineage in China and Southeast Asian countries. G1-like viruses have mainly circulated in South China, Central Asia, the Middle East, and North Africa, whereas Y439-like viruses have circulated in South Korea [12,13]. In Japan, H9N2 viruses are not yet enzootic. Therefore, careful surveillance of avian influenza viruses in poultry products carried by international flight passengers from other countries at the border is important to prevent their spread in Japan. Indeed, avian influenza viruses, including H5N1, H7N9, H5N6, and H9N2, have been repeatedly detected in raw poultry meat brought by international flight passengers from Asian countries to Japan [14,15]. Previous reports showed that seven H9N2 avian influenza viruses were isolated from chicken and duck meat products carried by passengers on flights from China and Vietnam; these viruses were classified into the Y280 sub-lineage [14,15]. One of these H9N2 viruses was characterized in avian models and found to be of low pathogenicity in chickens and ducks [14]; however, the biological properties of these H9N2 viruses in mammals remain unknown. Therefore, we here characterized the biological features of two H9N2 viruses [A/chicken/Japan/AQ-HE28-50/2016 (Ck/HE28-50) and A/chicken/Japan/AQ-HE28-57/2016 (Ck/HE28-57)] isolated from chicken meat products transported to Japan by international passengers from Vietnam [14] in a mouse model.

Viral Genome Sequencing
Viral RNA was extracted from the supernatants of stock viruses by using a QIAamp viral RNA Mini Kit (Qiagen, Hilden, Germany). The first-strand cDNA was synthesized using Uni12 primer, Superscript III (Invitrogen, Carlsbad, CA, USA), and universal primers specific for influenza A virus genes [16]. The resulting products were PCR-amplified using Phusion DNA polymerase (NEW ENGLAND BioLabs, Tokyo, Japan) with specific primers for each virus gene. The amplified PCR products were purified and subjected to direct sequencing using the ABI PRISM 310 system.

Viral Replication Assay
Triplicate wells of confluent MDCK and A549 cells were infected with viruses at a multiplicity of infection (MOI) of 0.001 and incubated for 1 h at 37 • C. After the 1 h incubation, the MDCK and A549 cells were incubated in MEM and Ham's F-12K containing 0.3% bovine serum albumin (BSA) at 37 • C, respectively. Triplicate wells of confluent DF-1 cells were infected with viruses at an MOI of 0.001, incubated for 1 h at 39 • C, and then incubated in DMEM containing 0.3% BSA at 39 • C. For the MDCK and DF-1 cells, aliquots of supernatants were harvested at 12 h post-infection (hpi), 24 hpi, and 48 hpi. For the A549 cells, aliquots of supernatants were harvested at 24-h intervals. Virus titers in the culture supernatants at each time point were determined utilizing plaque assays in MDCK cells.

Mouse Experiments
Female, six-week-old C57BL/6J mice (Japan SLC) were used for these experiments. To observe body weight changes, 5 mice/group for each virus were anesthetized with isoflurane and inoculated intranasally with 10 6 PFU/mouse in a 50-µL volume. The mice were monitored daily for clinical signs of infection and checked for changes in body weight and mortality for 14 days post-infection (dpi). Mice were euthanized if they lost more than 25% of their initial body weight. The remaining mice were euthanized at 14 dpi. For virus replication assessment, groups (3 mice/group) were infected intranasally with 10 4 or 10 6 PFU/mouse in a 50-µL volume. Three mice in each group were euthanized on Days 3 and 6 post-infection. Organs (lungs and nasal turbinates) were collected for virus titration in plaque assays in MDCK cells. The data shown are the mean virus titers ± standard deviations (SD). All experiments with mice were performed in accordance with the University of Tokyo's Regulations for Animal Care and Use and were approved by the Animal Experiment Committee of the Institute of Medical Science, the University of Tokyo (PA20-6).

Statistical Analysis
For statistical analyses of growth kinetics data, the virus titer values were converted to the log10 scale, and a two-way ANOVA, followed by a Dunnett's test, was performed. The virus titers of Ck/HE28-50 were compared with those of Ck/HE28-57, and the difference was considered significant for p values of <0.05 in GraphPad Prism6.

Replicative Ability and Pathogenicity of Two H9N2 Avian Influenza Viruses in Mice
We next examined the replicative ability of these H9N2 viruses in vivo. C57BL/6 mice were intranasally infected with 10 4 PFU or 10 6 PFU of Ck/HE28-50 or Ck/HE28-57, and organ samples were collected at 3 and 6 dpi for virus titration. In mice infected with 10 6 PFU of the virus, both viruses replicated well in the lungs and nasal turbinates. The mean virus titers in the lungs of mice infected with Ck/HE28-50 or Ck/HE28-57 were 6.25 ± 0.42 and 6.70 ± 0.20 log 10 PFU/g at 3 dpi, respectively, and those in the nasal turbinates were 4.87 ± 0.40 and 5.31 ± 0.77 log 10 PFU/g at 3 dpi, respectively, for the Ck/HE28-50 and Ck/HE28-57 groups (Figure 2A). For the mice infected with 10 4 PFU of the virus, the mean lung titers of the Ck/HE28-50-or Ck/HE28-57-infected mice were 4.32 ± 0.56 and 5.25 ± 0.26 log 10 PFU/g, respectively, at 6 dpi. In the nasal turbinate, the mean virus titer of Ck/HE28-57-infected mice was 4.22 ± 0.38 log 10 PFU/g at 6 dpi, whereas virus (2.52 log 10 PFU/g) was recovered from only one of the three mice infected with Ck/HE28-50 ( Figure 2B).
To examine the pathogenicity of the two H9N2 viruses, five C57BL/6 mice per group were inoculated intranasally with 10 6 PFU of each virus, and their body weight changes were monitored for 14 days (Figure 3). The Ck/HE28-50-infected mice continued to gain body weight during the observation period, as did the control group, which was inoculated intranasally with 50 µL/mouse of PBS. In contrast, the body weight of the Ck/HE28-57infected mice declined by day 7 post-infection ( Figure 3); however, all mice recovered from their infection with Ck/HK28/57. inoculated intranasally with 50 μL/mouse of PBS. In contrast, the body weight of the Ck/HE28-57-infected mice declined by day 7 post-infection ( Figure 3); however, all mice recovered from their infection with Ck/HK28/57.

Discussion
In this study, we examined the biological features of two H9N2 viruses (Ck/HE28-50 and Ck/HE28-57) that were isolated from raw poultry meat that was illegally brought into Japan from Vietnam [14]. We found that these H9N2 viruses replicate reasonably well in mammalian cells and mice, although the replicative ability of Ck/HE28-50 was lower than that of Ck/HE28-57 in A549 cells (Figure 1). Concerning pathogenicity in mice, Ck/HE28-57 caused a temporary loss of body weight, whereas the body weight of Ck/HE28-50infected mice gradually increased similarly to the mock-infected group (Figure 3). These data demonstrate that Ck/HE28-50 and Ck/HE28-57 can replicate in mammals and that Ck/HE28-57 is slightly more virulent than Ck/HE28-50 in mammalian hosts.
Avian influenza viruses rarely infect humans due to host range restrictions. Several viral factors determine the viral interspecies transmission and pathogenicity of avian influenza viruses in mammals [17][18][19][20][21]. The receptor-binding specificity of the HA protein is a major determinant of the influenza viral host range [19,[22][23][24]. In general, human influenza viruses preferentially bind to sialic acid-α2,6-galactose (SAα2,6Gal), the predominant sialyloligosaccharide species on epithelial cells in the upper respiratory tract of humans [19,25], whereas avian influenza viruses preferentially recognize sialic acid-α2,3-galactose (SAα2,3Gal), the major sialyloligosaccharide species in the duck intestinal tract [26]. A change of binding preference from the avian-type receptor (i.e., SAα2,3Gal) to the human-like receptor (i.e., SAα2,6Gal) is considered to be one of the important steps for the adaptation of avian influenza viruses to mammalian hosts [22,24,25,27]. There are several key residues in HA that affect viral receptor-binding preference; for example, replacing glutamine (Q) with leucine (L) at position 226 (H3 numbering), which is found in most H9N2 poultry isolates, can alter the viral receptor-binding preference from the avian-type receptor to the human-type receptor [23]. Ck/HE28-50 and Ck/HE28-57 possess four amino acid substitutions in HA (i.e., HA-159N, HA-190V, HA-198T, and HA-226L) that are associated with a shift in the binding preference of avian influenza viruses to the human-type receptor (Supplementary Table S1) [23,24,28,29]. In addition, the viral polymerase complex, which comprises PB2, PB1, PA, and NP, has key roles in the adaptation of avian influenza viruses to mammalian hosts, and several amino acid substitutions have been shown to contribute to increased polymerase activity, replicative ability, and/or virulence in mammalian models [20,[30][31][32][33][34]. The change from glutamic acid (E) to lysine (K) at position 627 of the PB2 protein is one of the most important host range determinants, but neither Ck/HE28-50 nor Ck/HE28-57 possesses the E-to-K change at position 627 of PB2. Instead, they have several amino acid substitutions in the viral polymerase complex that are associated with the mammalian adaptation and pathogenicity of avian influenza viruses (i.e., PB2-89V and 309D, PB2-504V, PB2-588V, PB1-622G, PA-63I, PA-356R, and PA-383D) ( Supplementary Table S1) [20,21,31,[33][34][35][36]. Ck/HE28-57, but not Ck/HE28-50, possesses the isoleucine (I)-to-valine (V) mutation at position 292 of PB2 protein that is associated with increased H9N2 virus polymerase activity in mammalian cells and enhanced virulence in mice [37]. In our study, Ck/HE28-57 was more pathogenic than Ck/HE28-50 in mice (Figure 3), indicating that PB2-292V may contribute to the increased virulence of Ck/HE28-57 in the mouse model. Alternatively, an as-of-yet undetermined amino acid change(s) in Ck/HE28-57 may have a role in its pathogenicity in mice. Further studies are needed to clarify the factors that determine the differences between the properties of Ck/HE28-50 and Ck/HE28-57.
Compared with highly pathogenic avian influenza viruses, such as H5N1 and H7N9 viruses, less attention is paid to low pathogenic avian influenza viruses. However, many reports, including this study, have described the capability of H9N2 viruses to adapt to mammals [23,24,30,33,[37][38][39], raising concern about their pandemic potential. Even though H9N2 avian viruses have not yet been isolated in domestic poultry in Japan, enhanced surveillance at the border is essential to prevent their spread.