Pathogenicity and genetic characterisation of a novel reassortant, highly pathogenic avian influenza (HPAI) H5N6 virus isolated in Korea, 2017

We investigated influenza A(H5N6) viruses from migratory birds in Chungnam and Gyeonggi Provinces, South Korea following a reported die-off of poultry in nearby provinces in November 2017. Genetic analysis and virulence studies in chickens and ducks identified our isolate from December 2017 as a novel highly pathogenic avian influenza virus. It resulted from reassortment between the highly virulent H5N8 strain from Korea with the N6 gene from a low-pathogenic H3N6 virus from the Netherlands.

In connection with an influenza A(H5N6) outbreak in poultry in November 2017 in JeollabukDo Province, South Korea [1], we collected influenza virus isolates investigated from wild migratory birds in the neighbouring Chungnam and Gyeonggi Provinces. Four novel reassortant highly pathogenic avian influenza (HPAI) H5N6 viruses were isolated on 13 December 2017 and continued to cause outbreaks in domestic poultry [personal communication: Dr Youn-Jeong Lee, Animal and Plant Quarantine Agency, South Korea, January 2018]. Genetic characterisation revealed that the haemagglutinin (HA) gene of the novel H5N6 viruses was closely associated with clade 2.3.4.4 influenza A(H5N8) viruses. Although genetically similar to H5N6 viruses that caused outbreaks in Japan (first reported on 10 November 2017) [2] and the Netherlands (7 December 2017) [3], there has not been a report about its pathogenic potential in poultry species. Therefore, we report here the genetic characterisation of the novel influenza A(H5N6) virus and the investigation of its pathogenic potential in chickens and ducks.

Genetic characterisation of novel influenza A(H5N6) viruses
Four influenza A(H5N6) viruses were isolated from faecal samples obtained from migratory bird habitats in Gyeonggi Province during a surveillance study conducted on 13 December 2017. Mitochondrial DNA sequence analysis of the faecal specimens revealed Anas platyrhynchos to be the viral host. In a full-length genomic sequence analysis, the viruses showed 99.9-100% nucleotide homology to one another but 97.  (Figure 3).

Figure 2
Phylogenetic trees comparing the neuraminidase nucleotide sequences of novel influenza A(H5N6) viruses, South Korea, December 2017(n = 4) 19801985199019952000200520102020 Group C

Virulence in chickens and ducks
One representative virus, A/AP/Korea/ W612/2017(H5N6), was selected for further study. The intravenous pathogenicity index (IVPI) was measured in accordance with World Organisation for Animal Health (OIE) standards. In chickens, the IVPI score was 2.76, resulting in classification of the A/AP/Korea/ W612/2017(H5N6) virus as HPAI [6]. To measure the chicken and duck lethal dose 50% (CLD 50 and DLD 50 ), specific pathogen-free chickens and ducks were infected with 10 7 -10 2 egg infectious dose (EID) 50 /mL by intranasal inoculation. The CLD 50 was 2.83 log 10 EID 50 / mL, with which the chickens died within 3 to 5 days after infection. In contrast, none of the infected ducks died during the experimental period (14 days) suggesting a DLD 50 of more than 10 7 EID 50 /mL.   Titres are indicated with standard deviation. Swabs from chickens were collected daily for the infected and contact group. Swabs from ducks were collected every 2 days for the infected group and daily for the contact group.

Table 2
Viral titres in chickens and ducks experimentally inoculated with influenza A/AP/Korea/W612/2017(H5N6), South Korea, December 2017 (n = 17) To investigate the pathogenicity and horizontal transmission ability of the A/AP/Korea/W612/2017(H5N6) virus in chickens and ducks, we proceeded with a group of intranasally inoculated (10 6 EID 50 /mL) animals (n = 14) and added a direct contact group (n = 3) one day after infection. Oropharyngeal and cloacal swab samples were collected for virus titration for 14 days post infection (dpi). Chickens in the infection and direct contact groups all succumbed by 4 dpi and 5 dpi, respectively. Moreover, virus was detected at all time points of swab collection and peaked at 2 dpi at 4.8 log 10 EID 50 /mL in oropharyngeal swabs ( Table 2). In the direct contact group, the virus was detected from the second day after contact and the highest cloacal swab viral titre (5.6 log 10 EID 50 /mL) was measured on the fourth day. To determine virus tissue distribution of the A/AP/Korea/W612/2017(H5N6) virus in infected chickens, we collected the lung, brain, kidney, spleen, heart, liver and colon from three birds each at 3 and 5 dpi using individual sterile equipment to avoid crosscontamination between them. The virus was detected in all organs collected from the inoculated chicken group at 3 dpi, the last point at which samples could be collected due to the lethality of the virus ( Table 2).
In ducks, we measured the viral replication and transmission efficacy although no mortalities were observed. However, high viral titres were detected at all time points in oropharyngeal swabs of both groups, peaking at 3 dpi (5.1 log 10 EID 50 /mL) for the infection group (Table 2). In contrast, low titres were observed in the cloacal swabs, which peaked at only 2.4 log 10 EID 50 /mL at 4 dpi in the direct contact group. Furthermore, the virus was not detectable in cloacal swabs at 7 dpi or in cloacal or oropharyngeal swabs at 9 dpi for the inoculated group. For the contact group, the virus was not detected in oropharyngeal or cloacal swabs at 7 days post contact. Seroconversion was observed in all ducks inoculated with the A/AP/Korea/ W612/2017(H5N6) virus and in their contact ducks, and haemagglutination inhibition geometric mean titres were 640 and 320 in the inoculated and direct contact group, respectively (Table 2). To investigate the tissue distribution of A/AP/Korea/W612/2017(H5N6) virus in ducks, we collected the lung, brain, kidney, spleen, heart, liver and colon from three birds each at 3 and 5 dpi. The A/AP/Korea/W612/2017(H5N6) virus was detected in all organs harvested from ducks in this experiment, except for the brain (Table 2).
In this study, we report the identification of a novel reassortant HPAI H5N6 virus that caused major outbreaks in domestic poultry during the 2017/18 winter season. This H5N6 virus is a reassortment of multiple subtypes (H5N8, H7N7, H5N1, H10N7 and H3N6) in the Eurasian gene pool of avian influenza viruses. Animal studies revealed that this novel influenza A(H5N6) virus is highly pathogenic in chickens but has attenuated virulence in ducks, which may result in their function as a virus reservoir. With the high susceptibility but attenuated virulence of HPAI infections in ducks, the duck can also be used as a sentinel for avian influenza virus surveillances [13]. Since the first avian influenza A(H5N8) (clade 2.3.4.4) virus was reported in Korean poultry in 2014 [14,15] and has continued to spread in South Korea, migratory birds have spread the virus to wild birds worldwide, including in Europe [16,17] and North America [18]. This rapid and widespread proliferation and reassortment underscores the need for continuous monitoring of avian influenza viruses in wild migratory birds, including virulence and pathogenicity studies using laboratory animals.