Novel Reassortant Influenza A(H5N8) Viruses in Domestic Ducks, Eastern China

Domestic ducks are natural reservoirs of avian influenza viruses and serve as reassortant hosts for new virus subtypes. We isolated 2 novel influenza A(H5N8) viruses from domestic ducks in eastern China, sequenced their genomes, and tested their pathogenicity in chickens and mice. Circulation of these viruses may pose health risks for humans.


The Study
During surveillance of poultry for avian influenza viruses in live poultry markets in Zhejiang Province in eastern China in 2013, we isolated 2 influenza A(H5N8) viruses, A/duck/Zhejiang/W24/2013(H5N8) (W24) and A/ duck/Zhejiang/6D18/2013(H5N8) (6D18), from domestic ducks. To better understand genetic relatedness between these viruses, we sequenced all gene segments of these 2 viruses and compared them with influenza virus sequences in GenBank. We also determined the virulence of the 2 isolates in chickens and mice.
For virus isolation, cloacal swab specimens from domestic ducks were inoculated into embryonated chicken eggs as described (5). All experiments with viruses were performed in a Biosafety Level 3 laboratory.
RNA was extracted by using the Viral RNA Mini Kit (QIAGEN, Hilden, Germany) according to the manufacturer's instructions. All gene segments were amplified with primers, and fragments were sequenced and analyzed as described (6)(7)(8) Figure  2). On the basis of analysis of phylogenetic relationships, we found that W24 was a reassortant virus that derived its genes from a virus of a different subtype from poultry in China. We also found that H5N8 subtype viruses had been present in eastern China for several years and these viruses might have been spread to other countries by wild birds in recent years. On the basis of deduced amino acid sequences of HA genes, we found that the HA cleavage site pattern (PLREKRRKR) of the 2 novel H5N8 subtype viruses indicated that these viruses were highly pathogenic. In this study, amino acid sequences of these H5N8 subtype viruses at positions 236-241 and 146-150 were NGQR-GR and GVSAA, respectively. Receptor-binding sites (Gln226 and Gly228) of H5N8 subtype viruses were similar to those of the 2 novel H5N8 subtype viruses, which suggested that these 2 viruses would preferentially bind to avian-like receptors (7). The PB2 protein Lys627Glu mutation has been reported to influence the host range and confer increased virulence for H5N1 subtype viruses in animal models (12). This mutation was not observed in PB2 of the 2 novel H5N8 subtype viruses analyzed in this study, which indicated that these 2 viruses had low levels of pathogenicity for mice.
Deletion of several amino acids (position 80-84) in NS1 proteins had been observed more frequently in H5N1 subtype viruses, which indicated possible adaptation of these viruses to avian species (13). This deletion was not observed in the 2 novel H5N8 subtype viruses. These 2 viruses contained the NS1 Pro42Ser mutation, which is associated with increased virulence in mice (14).
To evaluate pathogenicity of W24 and 6D18 in chickens, we inoculated groups of ten 6-week-old specific pathogen-free chickens intravenously with a 10 6 median egg infective dose of each virus in a 0.2-mL volume of phosphate-buffered saline; deaths were observed over a 10-day period (7). Animal studies were conducted according to the recommendation of the World Organisation for Animal Health (Paris, France). Characteristics of W24 and 6D18 viruses are shown in Table 2. Results showed that these viruses were highly pathogenic to chickens.
To determine the pathogenicity of these viruses in a mammalian host, we inoculated BALB/c mice intranasally with a 10 6 median egg infective dose for each virus. Over a 14-day period, we observed virus replication in various  organs and deaths. After intranasal administration of W24 and 6D18, we observed 5 mice per group for survival over a 14-day period. On day 9 postinfection, high titers of virus was detected in lung and liver but were not detected in brain and heart. Mice had signs of illness but had survival rates of 80% (4/5) and 100% (5/5), respectively, for each virus during 14 days postinfection (Table 2). These results suggested that the 2 novel H5N8 subtype viruses did not kill mice but that they could replicate in the lung. Results of our study are consistent with those of Zhao et al. (15), who reported that H5N8 subtype virus from chickens did not cause deaths in mice.

Conclusions
We isolated 2 influenza A(H5N8) viruses were isolated from domestic ducks in eastern China in 2013. Results of phylogenetic analysis showed that these viruses were reassortant viruses that derived their genes from different virus subtypes. These reassortant H5N8 subtype viruses and their 3 possible parent viruses, A/duck/Jiangsu/k1203/2010 (H5N8), A/environment/Jiangxi/28/2009 (H11N9), and A/ duck/Hunan/8-19/2009 (H4N2), were isolated in China. Both H5N8 subtype isolates were highly pathogenic for chickens and showed moderate pathogenicity for mice. Domestic ducks are considered the natural reservoir of avian influenza viruses and serve as reassortant hosts for creation of new virus subtypes. Continued circulation of these viruses may pose health threats for birds and humans.