There is an influenza host barrier between many kinds of animals species[6, 25]. AIV is widely detected in wild waterfowl and can infect domestic poultry under certain conditions. There can be highly pathogenic influenza viruses and low pathogenic influenza viruses on the basis of pathogenicity. Currently, we know that the AIV species associated with AIV polymerase contains the PB2, PB1, PA and NP genes[26–27]. When AIV develops the ability to infect one species effectively, some adaptive mutations on these proteins can often be observed [7, 20, 24]. Another study showed that the HA protein plays a key role in the receptor binding process[28–30]. However, how the influenza virus causes epidemics is less well known.
Compared to the seasonal influenza virus, the AIV H13 subtype rarely causes infection.. This subtype is a low-pathogenic influenza viral subtype and is prevalent in gulls and ducks[31–32]. In this study, we obtained two AIV H13 subtypes isolated from waterfowl in China. Previous study results showed that these isolates could not replicate in 6-week-old adult chickens or 20-day-old chicks (Table S2); therefore, their host range may be restricted to waterfowl. When AIV crosses the host barrier from waterfowl to domestic poultry, the virus usually contains genetic modifications. The haemagglutinin protein is an important factor in host specificity[28, 33], and the NA deletions are often found in the stalk of the NA protein[34–36]. NS1 is a multifunctional protein in the influenza virus and may have a function in host-specific adaptation[37–38]. Our results showed that the surface glycoproteins of the two AIV subtypes belonged to different lineages : A/mallard/Dalian/DZ-137/2013 (H13N6) belonged to Group I, while A/Eurasian Curlew/Liaoning/ZH-385/ 2014 (H13N8) belonged to Group III (Fig. 1). In addition, two AIV H13 subtypes of wild bird origin were reported in 2016 in China. We found that A/black-tailed_gull/Weihai/115/2016 (H13N2) and A/mallard/Dalian/DZ-137/2013 (H13N6) belonged to the same group, while both A/black-tailed_gull/Weihai/17/2016 (H13N8) and A/Eurasian Curlew/Liaoning/ZH-385/ 2014 (H13N8) belonged to Group III. Although we just compared and analysed parts of H13 strains, we suspected that these AIV H13 subtypes were circulating frequently.
The two tested influenza viruses showed high homology to sequences in GenBank (Table 1). We sequenced the whole viral genomes and analysed key amino acids in the PB2, HA, NA and NS genes (Tables 2 and 3), and we found no deletions in the NA and NS genes. Moreover, we aligned the HA and NA sequences of the DZ137 and the ZH385 strains and found low homology. These results suggested differences at the molecular level were present. Additionally, the molecular markers of host adaptation (such as E627K and D701N) were not detected in the two H13 viruses. We also blasted the H13 subtypes in GISAID and NCBI. However, the PB2 genes of these strains showed a result of 627E and 701D. Therefore, these H13 subtypes have not obtained the ability to adapted to mammals in general.
As these AIV H13 subtypes could hardly replicate in 20-day-old chicks or adult chickens, we chose 1-day-old chicks as an experimental model. Chicks are commonly used to study vaccine efficacy and susceptibility to AIVs. In this study, we divided chicks into a 1-day-old chick group and a 10-day-old chick group. The DZ137 virus replicated in only the 1-day-old chicks. Compared to the ZH385 strain, the DZ137 strain had a lower replication rate. ZH385 efficiently replicated in 1-day-old chicks. We detected the virus titres in the lung and colon tissues of 1-day-old chicks (Fig. 2). The titres in the tissues of 10-day-old chicks were evaluated in embryonated chicken eggs and presented high values (Fig. 3). These results suggested a difference in infectivity of the two AIV H13 subtypes by challenge experiments. Further study to determine the reason for these differences in infectivity is necessary. More H13 influenza viruses need to be tested in animal models. Additionally, we evaluated the transmission of A/black-tailed_gull/Weihai/17/2016 (H13N8) in 4-week-old chickens. The results showed no detectable virus shedding in oropharyngeal and cloacal swabs from the donor, direct contact and airborne contact groups (Table S3-S5) .
In a serological survey from 2015 to 2016 in China, an HI assay detected chicken serum antibody titres against the H13 subtypes ranging from 40 to 320 (Tables 4–6). Among all 328 chicken specimens, we found detectable HI titres against DZ137 in 15 (4.6%) samples and detectable HI titres against ZH385 in 34 (10.4%) samples. We did not detect the other influenza subtypes(H1, H3, H5, and so on) because of the low serum volume available. However, there has already been a report of poultry infection with H13 strains of wild bird origin. This serological evidence suggests that the AIV H13 subtype may overcome the host barrier and circulate in domestic chickens in the future.
We evaluated the differences in viral growth in mammal cells, and the virus titres in MDCK cells were higher than those in other cells (Fig. 4). The peak virus titres occurred at different time points in the three mammal cell cultures, suggesting that the two H13 subtypes can replicate in both mammal cells and chicken cells. In addition,the two H13 subtypes could infect not only CEF cells but also MDCK and 293T cells. These results show that H13 maybe infect domestic poultry and other mammals.