Epidemiological, phylogenetic, and pathogenicity analysis of H4 subtype avian influenza viruses in China, 2011–2022

Abstract

The H4 subtype of avian influenza virus (AIV) is prevalent worldwide, but only receives little attention due to its low pathogenicity in poultry. Consequently, it remains largely unclear whether H4 AIVs pose a potential threat to the poultry industry and public health. During the period from 2011 to 2022, we conducted an active surveillance programm. A total of 154,762 swab samples were collected across various provinces, and 427 H4 viruses were detected, resulting in a positivity rate of 0.28%. All H4 viruses were isolated from poultry, primarily from ducks in live poultry markets. We further investigated the genetic evolutionary characteristics and pathogenicity of 20 H4Nx viruses isolated in our program. Phylogenetic analysis revealed that the 20 H4Nx viruses belonged to the Eurasian lineage and exhibited significant genetic diversity, with 19 distinct genotypes identified. Molecular characterization indicated that these viruses were low-pathogenicity AIVs with limited binding affinity to human receptors, yet they contained mutations associated with enhanced viral replication and pathogenicity in mammals. Pathogenicity tests conducted in ducks demonstrated that H4 viruses were weakly pathogenic, exhibiting limited replication and transmission capabilities. However, some viruses were able to replicate effectively in mice and induce weight loss. For instance, DK/AH/AG61/11(H4N6) can replicate efficiently in MDCK cells, indicating a potential threat to mammals. These findings underscore the importance of ongoing surveillance of H4 AIVs to better understand their evolution and transmission dynamics and to prevent potential public health risks.

Highlights

• •154,762 swab samples were collected across Chinese provinces (2011-2022), while 427 H4 viruses were detected, with a 0.28% positivity rate.
• •The transmission ability of the H4 virus in duck flocks is limited, but it can induce tissue lesions in some organs.
• •Most H4 viruses can effectively replicate in the lungs and turbinates of mice, indicating a potential public risk.
• •An H4N6 virus carrying the PB2 A588V mutation causes significant weight loss in mice.

Introduction

Since the initial discovery of avian influenza in chicken flocks in Italy in 1878 (Lupiani and Reddy, 2009), the avian influenza virus (AIV) has spread extensively among wild birds and poultry worldwide. Over this period, it has contributed to the emergence of five influenza pandemic viruses (Sutton, 2018), posing a significant threat to the healthy development of the poultry industry and public health security. AIV, classified within the family Orthomyxoviridae and the genus Influenza A virus, is a single-stranded, negative-sense RNA virus enveloped by a lipid bilayer membrane (Chauhan and Gordon, 2022). Hemagglutinin (HA) and neuraminidase (NA) are critical viral surface glycoproteins that enable the classification of AIVs into various subtypes due to their distinct antigenic properties (Webster et al., 1992). To date, subtypes H1–H18 and N1–N11 have been identified in wild waterfowl and mammals (Olsen et al., 2006; Wu, Y. et al., 2014). Additionally, a new HA subtype has been detected in wild birds in Kazakhstan and California, the United States (Fereidouni et al., 2023; Karakus et al., 2024). Based on their pathogenicity in poultry and the number of basic amino acids at the HA protein cleavage site, AIVs can be categorized into two pathogenic categories: highly pathogenic AIV (HPAIV) and low pathogenic AIV (LPAIV) (De Bruin et al., 2022). Historically, the vast majority of LPAIVs only cause mild respiratory or gastrointestinal symptoms and limited clinical signs in infected birds (Naguib et al., 2019). However, some H5 and H7 viruses can gradually evolve into HPAIVs after spreading among terrestrial poultry, leading to systemic infections and severe pathological damage in chickens, with mortality rates as high as 100% (Alexander and J., 2007; Sutton, 2018).

The H3, H4, H6, H9, and H10 subtypes of AIVs are the most prevalent LPAIVs in birds (Luo et al., 2021). Although not as destructive as HPAIV, they can act as gene donors and undergo reassortment with other AIVs, leading to the continuous emergence of new viral strains. The research has shown that the H9N2 subtype can contribute gene segments to multiple AIV subtypes (Wang et al., 2023), thereby enhancing the adaptability of newly generated viruses in poultry. This enhanced adaptability increases the risk of epidemics and potential spillover infections in humans, which can result in severe outcomes, including death. Notable examples include the H5N1 virus outbreak in Hong Kong in 1997, the emergence of H7N9 virus in 2013, and the H3N8 virus newly detected in chickens in 2022 (Guan et al., 1999; Wu, A. et al., 2013; Cui et al., 2023). The ability of LPAIVs to cross species barriers and infect humans and other mammals is increasingly recognized, attributed in part to their capacity to bind to both α2,3-sialic acid (α2,3-SA) and α2,6-sialic acid (α2,6-SA) receptors (Kim et al., 2021). The occurrence of human infections caused by subtypes such as H3N8, H6N1, H7N4, H9N2, H10N3, H10N5, and H10N8 has raised substantial public health concerns regarding LPAIVs (Yuan, J. et al., 2013; To et al., 2014; Xiang et al., 2019; Song, W. and Qin, 2020; Jing et al., 2021; Yuan, Z. et al., 2024; Chen et al., 2025, Zhuang et al., 2024).

The H4 subtype of AIV was first isolated from ducks in Czechoslovakia in 1956 (Donis et al., 1989). Over the following decades, its host range expanded to include wild birds, poultry, and mammals such as pigs, seals, and muskrats, primarily in North America and Eurasia (Liu, J. et al., 2024). Infections with H4 AIVs can lead to respiratory illnesses, decreased egg production, and reduced immunity in poultry (Zhang et al., 2011), thereby increasing susceptibility to co-infections with other viruses or bacteria. Certain H4 viruses can replicate efficiently in mice without prior adaptation, resulting in weight loss and even mortality; for example, an H4N8 strain isolated from waterfowl in the Hokkaido region by Bui et al. (2012). Although there are currently no reports of human infections with H4 viruses, this subtype has occasionally been isolated from swine. At least five natural cases of H4 AIV infection in swine have been documented in Canada, the United States, and China, involving subtypes H4N1, H4N6, and H4N8 (Hu et al., 2012; Su et al., 2012; Abente et al., 2017). Serological studies among high-risk occupational groups have detected specific antibodies against the H4 subtype of AIV in poultry farm workers in the United States and Lebanon (Kayali et al., 2009; Smith et al., 2011), suggesting that H4 viruses have the potential to cross the species barrier and infect humans.

Despite their significance, H4 viruses have not received sufficient attention or surveillance, largely due to the absence of historical large-scale outbreaks and their low pathogenicity. In this study, we performed epidemiological, phylogenetic, and pathogenicity analyses of H4 subtype AIVs using data collected through routine surveillance in live poultry markets and poultry farms in China from 2011 to 2022. The findings will improve our understanding of the evolution and transmission of H4 viruses in China's poultry industry and provide valuable data to support further research and comprehensive prevention and control measures.

Results

Epidemiologic characteristics of H4 AIVs in China

To investigate the prevalence of H4 AIVs in China, this study genetically analyzed the HA gene sequences of 317 H4 viruses collected from the Global Initiative on Sharing All Influenza Data (GISAID) database, which were isolated in China between 2000 and 2024. Additionally, an analysis of three factors (time, host, and location) was conducted on the information regarding these strains. The results indicated that the endemic H4 AIVs in China formed three distinct clades (I–III) in the phylogenetic tree, with clades I and II belonging to the Eurasian lineage and clade III to the North American lineage (Fig. 1). In terms of temporal distribution, H4 viruses were primarily isolated during the periods of 2005–2006, 2009–2016, and 2019, followed by a downward trend since 2019 (Fig. 2A). Globally, nine NA subtypes of H4 AIVs (from H4N1 to H4N9) have been detected (Liu, J. et al., 2024). With the exception of the H4N4 and H4N5, all other subtypes have been identified in China. The H4N6 subtype accounted for approximately 43.22% of the isolates, making it the predominant epidemic subtype of the H4 virus. The H4N2 subtype was isolated more frequently in 2014, 2015, and 2019, with an overall proportion of about 30.28%. In contrast, other subtypes such as H4N8, H4N3, H4N1, H4N9, and H4N7 had lower proportions (Supplementary Fig. S1). In terms of host distribution, H4 AIVs in China were found in a diverse range of hosts, including poultry, wild birds, the environment, and mammals. The highest proportion of isolates was found in ducks (74.45%), followed by wild birds (14.20%), indicating that ducks serve as the primary hosts of H4 viruses (Fig. 2B). Geographically, H4 viruses were predominantly circulating in east and southwest China, with high numbers of isolates reported in provinces and municipalities such as Taiwan, Sichuan, Shanghai, Jiangsu, and Zhejiang. Additionally, the distribution of NA subtypes varied significantly across regions. For instance, Guangdong, Guangxi, Hunan, Jiangsu, Jiangxi, and Taiwan each reported more than four distinct subtypes of H4 viruses (Fig. 2C). Notably, two cases of H4 AIV infection in swine were reported in Hubei and Guangdong, suggesting a potential cross-species transmission from avian to mammalian hosts (Fig. 2C).

Fig. 1.

The phylogenetic tree of HA genes from H4 AIVs isolated in China. The tree was constructed using the neighbor-joining method in MEGA software (v11.0.13). Bootstrap analysis was conducted with 1000 replicates, and clade numbers are indicated on the right side of the panel.

Fig. 2.

Three-dimensional distribution of H4 AIVs isolated in China. A Temporal distribution of H4 viruses isolated in China from 2000 to 2024. B Intergroup distribution of H4 viruses isolated in China. C Spatial distribution of H4 viruses isolated in China. Different H4 subtypes of AIV are represented by specific colors. The two reported cases of swine infection in China are marked with a black swine icon. A larger circle on the map indicates a higher number of H4 viruses isolated in that province.

Isolation of H4 AIVs during 2011–2022 monitoring in China

In this study, a total of 154,762 swab samples were collected from live poultry markets, farms, and slaughterhouses across various regions of China between 2011 and 2022. A total of 427 strains of H4 subtype AIV were identified through virus isolation, resulting in an overall positive rate of 0.28%. The annual province-specific positive rates, calculated as the percentage of virus-positive samples relative to the total samples collected per province per year, are provided in Supplementary Table S1. Among the positive samples, 375 originated from waterfowl (ducks and geese) and 40 from land fowl (chickens and pigeons). Notably, ducks accounted for the highest proportion of isolates (86.18%), indicating that H4 viruses exhibit greater susceptibility in waterfowl compared to other poultry species (Supplementary Fig. S2A). H4N2 was the most prevalent subtype, comprising 120 strains (28.10%) of the total positive samples, followed by H4N6 with 100 strains (23.42%). Additionally, small quantities of H4N8, H4N1, H4N3, H4N4, and H4N9 viruses were detected. However, 163 samples could not be assigned to a specific NA subtype, resulting in unclassified NA data (Supplementary Fig. S2B). These findings indicate that H4N2 and H4N6 are the dominant subtypes circulating in poultry farms and live poultry markets in China. Compared to H4N6, the H4N2 subtype may exhibit higher transmission potential. Over time, the positivity rate for H4 viruses remained relatively stable from 2011 to 2013, increased briefly in 2014, dropped to the lowest level during the monitoring period in 2015, and then rose steadily from 2018 to 2020. After 2020, the positivity rate began to decline and gradually returned to baseline levels (Fig. 3A). From a geographical perspective, the surveillance covered 28 provinces and municipalities in China. The isolation of H4 viruses was primarily clustered in 16 provinces, mostly located in southern China, with the highest number of positive cases detected in Jiangxi, Henan, and Guangxi. This distribution suggests that these regions may represent high-risk areas for the epidemiological spread of H4 AIVs (Fig. 3B).

Fig. 3.

The isolation of H4 AIVs in various regions of China from 2011 to 2022 is presented. A This section illustrates the number of positive detections and the positivity rate of the H4 virus over time. B The geographical distribution of the H4 virus isolated in China during this period is depicted. Yellow areas represent sampling locations, while circles indicate the number of positive detections. The size of the circle correlates with the number of detections, with larger circles indicating a higher number and a greater virus isolation rate.

Genetic evolutionary analysis of H4 AIVs

Twenty representative H4 AIVs were screened based on host, location, and sampling time, with detailed information provided in Supplementary Table S2. Whole-genome sequencing of these viruses was performed, and phylogenetic trees of all eight gene segments were constructed to better understand the genetic makeup and evolution of H4 viruses. The HA gene of the 20 H4 viruses belonged to the Eurasian lineage, exhibiting nucleotide sequence identities ranging from 93.4% to 100%. Using a 95% nucleotide identity threshold, these viruses were classified into four groups (Group 1 to Group 4) (Fig. 4). Group 1 included the HA genes of 10 viral strains from Jiangxi, Jiangsu, Hubei, Guangdong, Anhui, and Guangxi provinces, reflecting geographically diverse origins. The two strains in Group 2 clustered on the same branch as A/swine/Guangdong/K4/2011(H4N8), indicating a close genetic relationship. Group 3 comprised the HA genes of one H4N1 and two H4N6 viruses, all isolated from ducks. Group 4 contained HA genes from five viral strains from Hubei, Jiangxi, and Guizhou, among which four were derived from ducks and one from chicken. Notably, one H4 virus detected in swine (A/swine/Hubei/06/2009(H4N1)) was genetically distinct from the viruses in this study and formed a separate branch.

Fig. 4.

The phylogenetic tree of HA genes of H4 AIVs detected from 2011 to 2022. Viruses included in this study are highlighted in bold red font. The two reported cases of swine infection in China are marked with a black swine icon. The tree was constructed using the neighbor-joining method in MEGA software (v11.0.13), and bootstrap analysis was conducted with 1000 replicates. Clade numbers are indicated on the right side of the panel.

The 20 H4 viruses in this study exhibited diverse NA subtypes, including one H4N1 virus, seven H4N2 viruses, two H4N3 viruses, nine H4N6 viruses, and one H4N8 virus. Phylogenetic analysis of the NA gene indicated that the N1 gene of DK/GX/E1424/20(H4N1) may have originated from the H6N1 virus and showed distant relatedness to A/swine/Hubei/06/2009(H4N1) (Supplementary Fig. S3A). In contrast, the NA gene of DK/HB/H1140/18(H4N8) was evolutionarily more distant from A/swine/Guangdong/K4/2011(H4N8) but was closely related to a human H10N8 virus (Supplementary Fig. S3E). The NA genes of the seven H4N2 viruses formed three distinct genetic groups, with nucleotide sequence identity ranging from 86.8% to 100%. Notably, two viruses in Group 3 were closely related to H9N2 viruses (Supplementary Fig. S3B). The two duck-origin H4N3 viruses shared 99.5 % nucleotide identity, were grouped together, and exhibited a close phylogenetic relationship with H5N3 viruses (Supplementary Fig. S3C). The nucleotide identity of the NA gene among the nine H4N6 viruses ranged from 90.1% to 99.9%, and they were classified into four groups, likely resulting from recombination among H4N6 viruses (Supplementary Fig. S3D).

The nucleotide sequence identities of the internal genes PB2, PB1, PA, NP, MP, and NS ranged from 86.1% to 100%, 89.6%–100%, 89.5%–100%, 90.0%–99.9%, 93.3%–99.9%, and 70.3%–100%, respectively. Phylogenetic analysis revealed the presence of eight, six, eight, five, three, and three distinct genotypes for these six internal genes, respectively, indicating substantial genetic diversity (Supplementary Fig. S4). By integrating phylogenetic analyses of all eight viral genes, the 20H4 viruses were classified into 19 distinct genotypes (G1-G19) (Fig. 5). These findings suggest that H4 AIVs in China underwent complex and frequent recombination events between 2011 and 2022, leading to genotypic diversification.

Fig. 5.

Genotypes of H4 viruses analyzed in this study. Genotypes were determined based on phylogenetic analyses of eight viral gene fragments, with groups categorized by 95% differences in nucleotide homology.

Molecular characterization analysis of H4 AIVs

To investigate the molecular characteristics of H4 AIVs, we analyzed key amino acid residues of the internal and external genes of 20 H4 viruses. Multiple molecular markers associated with receptor binding, viral replication, drug resistance, and pathogenicity were identified across various gene segments. The coding region of the HA gene in all 20 H4 viruses is 1695 bp in length and encodes 564 amino acids. The cleavage site of the HA protein contains the amino acid sequence PEKASR/GLF, confirming that these viruses are LPAIVs and exhibit low pathogenicity in chickens (De Bruin et al., 2022). No mutations were observed at the two key residues (226Q and 228G, H3 numbering) within the receptor-binding domain (Supplementary Table S3), indicating a maintained preference for binding to avian-type receptor (SA α-2,3 Gal) (Abubakar et al., 2023). However, the T160A mutation was identified in the HA protein, which is known to enhance viral binding to human-type receptors (SA α-2,6 Gal) (Gao et al., 2021). No amino acid deletions were found in the stem region of the NA protein in any of the H4 viruses. Nonetheless, some strains carried mutations associated with neuraminidase inhibitor resistance, such as I117V and D198N (N2 numbering) (Kode et al., 2019). Among internal genes, the PB2 mutations E627K and D701N, which are known to significantly enhance viral pathogenicity and transmissibility in mammals (Xu et al., 2022), were not detected in any of the H4 viruses of this study. Meanwhile, the A588V mutation was found in the PB2 protein of only one H4N6 virus (DK/AH/AG61/11(H4N6)). The PB1–F2 protein, 90 amino acids in length, contained the N66S mutation in six viral strains; this mutation has been shown to enhance viral replication in a cell type- and strain-specific manner (Kamal et al., 2017). Mutations T20A and N383D in the PA protein can increase viral virulence (Stevaert et al., 2013; Song, J. et al., 2015), while mutations N30D and T215A in the M1 protein can improve viral replication efficiency and pathogenicity in mice (Fan et al., 2009). These amino acid mutations were present in all strains. All H4 viruses, except one H4N1 and one H4N6 strain carried the NS1–P/A42S mutation, which has been reported to enhance viral virulence in mice (Jiao et al., 2008).

Pathogenicity of H4 AIVs in ducks

To investigate the pathogenicity of H4 subtype AIV in waterfowl, four representative strains were selected to assess viral virulence and transmissibility in specific pathogen-free (SPF) ducks. This evaluation was based on mutations at key amino acid sites and NA gene recombination specificity. At 3 days post-inoculation (dpi), three ducks from each infected group were euthanized, and organ samples were collected for virus titration. The results indicated that the replication and tissue tropism of these four H4 viruses were limited. DK/GX/E1424/20(H4N1) was detected only in the trachea, bursa of fabricius, and pancreas of ducks (Fig. 6A). CK/AH/A1402/14(H4N2) was found in the trachea, cecum, and bursa of one duck, as well as in the pancreas of another duck (Fig. 6B). In contrast, DK/AH/AG61/11(H4N6) showed low virus titers only in the pancreases of two ducks (Fig. 6C). Interestingly, DK/HB/H1140/18(H4N8) did not replicate in any of the sampled organ (Fig. 6D). Throughout the trial, no clinical signs or mortality were observed in any of the four infected groups. Additionally, serum antibody titers in both directly infected and contact-exposed ducks remained below 2 log₂, indicating an absence of seroconversion.

Fig. 6.

Pathogenicity and transmissibility of H4 AIVs in ducks. After inoculating ducks with H4 viruses at a dose of 10⁶ EID₅₀, we collected different organs (lung, trachea, liver, cecum, spleen, kidney, pancreas, and bursa of fabricius) at 3 dpi and titrated the virus in chicken embryos. A, B, C, and D indicate the replication of DK/GX/E1424/20(H4N1) (A), CK/AH/A1402/14(H4N2) (B), DK/AH/AG61/11(H4N6) (C), and DK/HB/H1140/18(H4N8) (D) in ducks, respectively. Oropharyngeal (OP) and cloacal (CL) swabs were collected from both infected and contact-exposed ducks on days 1, 3, 5, 7, and 9 post-exposure, and the virus was titrated in chicken embryos. E, F, G, and H indicate virus shedding in DK/GX/E1424/20(H4N1) (A), CK/AH/A1402/14(H4N2) (B), DK/AH/AG61/11(H4N6) (C), and DK/HB/H1140/18(H4N8) (D) infected and contact-exposed ducks, respectively. The dotted line indicates the lower limit of detection.

Viral titers from oropharyngeal and cloacal swabs indicated that all four viruses exhibited low infectivity in ducks, with minimal viral detection in the contact group. The DK/GX/E1424/20(H4N1) virus was detected only in the oropharynx or cloaca of one infected duck, with no viruses detected until 5 days post exposure (Fig. 6E). The CK/AH/A1402/14(H4N2) virus, which showed low viral titers, was found in the oropharynx of one infected duck on days 1 and 3 post-exposure (Fig. 6F). Viral shedding was later detected in the cloaca of the same infected duck on day 5 post-exposure. On days 7 and 9 post-exposure, CK/AH/A1402/14(H4N2) was found to be shedding virus in the cloaca of another infected duck. Ducks infected with DK/AH/AG61/11(H4N6) showed no viral detection (Fig. 6G). The DK/HB/H1140/18(H4N8) virus was identified in the oropharynx of only one infected duck on days 3 and 5 post-exposure (Fig. 6H). These results suggest that H4 viruses are weakly pathogenic in ducks and pose a low risk of epidemic transmission.

On day 3 post-inoculation, trachea and lung samples from infected ducks were collected and fixed in 4% formaldehyde for histopathological analysis. Abnormal tissue architecture was observed in the tracheas of all infected groups, including mucosal epithelial hyperplasia, submucosal hemorrhage or edema, and inflammatory cell infiltration (Fig. 7A). In lung tissues, pathological changes such as alveolar hemorrhage and congestion, epithelial cell desquamation, inflammatory cell infiltration, and airway wall thickening were observed (Fig. 7B). No histopathological changes were observed in the control ducks. In summary, although H4 virus infection resulted in low mortality in ducks, it can cause considerable pathological damage to the respiratory system, thereby adversely affecting overall health and growth.

Fig. 7.

Histopathological analysis of trachea and lungs of infected ducks. Tissue sections of the trachea and lungs from each infected duck, collected at 3 dpi, were stained with hematoxylin-eosin. The pathological sections of the trachea are presented in panel A. Notable lesions include mucosal epithelial hyperplasia (black arrows), submucosal hemorrhagic edema (yellow arrows), and inflammatory cell infiltration (red arrowheads). The pathological sections of the lungs are displayed in panel B, revealing lesions such as alveolar hemorrhage (black arrow), inflammatory cell infiltration (red arrow), thickening of the blood vessel walls (blue arrow), and shedding of bronchial epithelial cells (yellow arrow). Scale bar = 100 μm.

Replication of H4 AIVs in mice

To evaluate the infectivity and pathogenicity of H4 subtype AIV in mammals, we infected BALB/c mice with 20 H4 viruses at a dose of 10⁶ EID₅₀. Viral replication was assessed by measuring virus titers in selected organs. The results indicated that these viruses exhibited variable replication capabilities in the lungs and nasal turbinates of mice (Fig. 8). Seventeen viruses replicated in the lungs, with titers ranging from 0.92 to 5.92 log₁₀EID₅₀/mL. Eighteen viruses replicated in the turbinates, with titers ranging from 0.91 to 3.75 log₁₀EID₅₀/mL. Notably, DK/HB/H2091/17(H4N6) did not replicate effectively in either the nasal turbinates or lungs. Furthermore, no virus was detected in the spleen, kidneys, or brain of any infected mice (data not shown). All mice survived the 14-day observation period. All 20 strains affected the body weight of the mice. The most significant weight loss, reaching up to 19.17%, was observed in mice inoculated with DK/AH/AG61/11(H4N6). In contrast, mice inoculated with DK/HB/H2091/17(H4N6) gained 14.03% in body weight. In summary, H4 AIVs exhibit low pathogenicity in mice and typically cause transient infections. However, some viruses can replicate effectively even without prior adaptation and induce weight loss, suggesting a potential zoonotic risk to mammals.

Fig. 8.

Replication of H4 AIVs in BALB/c mice in this study. Viral titers in the lungs and nasal turbinates of mice following inoculation with 10⁶ EID₅₀ of various viruses were assessed. Three mice from each group were euthanized at 3 dpi, and the virus was titrated in chicken embryos. Values labeled with one or two red asterisks indicate that the virus was only detected in the organs of one or two mice. The weight of the infected mice was monitored for 14 days, and the maximum weight change rate for each group was calculated.

Growth kinetics of H4 AIVs in chicken embryos and mammalian cells

To evaluate the replication kinetics of H4 viruses isolated from poultry in avian and mammalian cells, two representative strains, DK/AH/AG61/11(H4N6) and DK/HB/H2091/17(H4N6), were selected based on their performance in the mouse infectivity assay. These viruses were inoculated into SPF chicken embryos at a dose of 10³ EID₅₀ and into Madin-Darby Canine Kidney (MDCK) cells at a multiplicity of infection (MOI) of 0.01. Both tested viruses replicated effectively in chicken embryos, with DK/AH/AG61/11(H4N6) achieving higher titers than DK/HB/H2091/17(H4N6) at all time points, reaching a peak titer of 4.53 log₁₀TCID₅₀/mL at 60 h post-inoculation (hpi). In contrast, DK/HB/H2091/17(H4N6) exhibited more moderate replication (Fig. 9A). In MDCK cells, DK/AH/AG61/11(H4N6) demonstrated significantly enhanced replication (∗∗∗∗ P < 0.0001), while DK/HB/H2091/17(H4N6) was detectable only after 36 hpi and at lower titers (Fig. 9B). These results indicate that H4 viruses from poultry can replicate efficiently in both chicken embryos and MDCK cells, though notable strain-dependent differences were observed. DK/HB/H2091/17(H4N6) showed comparatively limited replication in both systems.

Fig. 9.

The growth curves of the two H4 subtype representatives in chicken embryos (A) and MDCK cells (B). Chicken embryos (A) were infected with a virus at an EID₅₀ of 10³, and MDCK monolayer cells (B) were infected with the virus at an MOI of 0.01. Allantoic fluid and supernatant were collected at the specified time points, and viral titers were determined in MDCK cells using the TCID₅₀ method. Each sample was tested three times. The statistical significance was analyzed by two-way ANOVA test using GraphPad Prism V8.0.2 (∗, P < 0.05; ∗∗, P < 0.01; ∗∗∗∗, P < 0.0001).

Discussion

H4 subtype of AIV is one of the most prevalent LPAIVs, detected in wild birds and poultry worldwide. Although H4 infections are not as severe as those caused by H5 and H7 subtypes, they can be transmitted asymptomatically between flocks, making them difficult to detect. Furthermore, H4 AIVs frequently act as co-infecting subtypes (Luo et al., 2021), facilitating viral gene reassortment between donor strains and other subtypes, which may lead to the emergence of novel AIVs. Historical evidence has shown that the emergence of novel, highly pathogenic, and highly infectious viruses poses significant threats to the poultry industry and public health. Therefore, a comprehensive understanding of the epidemiological status and biological characteristics of H4 subtype AIV is crucial for effective prevention and control of AIV in China.

In recent years, H4 AIVs have been detected in wildfowl and poultry across several provinces in China. Studies indicate that these viruses have been circulating in live poultry markets in central, eastern, and southern China, where complex recombination events involving multiple subtypes of AIVs, including H3, H5, H6, H7, and H9, occurring frequently in domestic ducks, leading to increased genotypic diversity (Li et al., 2024; Lin et al., 2024). During active surveillance from 2011 to 2022, we isolated 427 H4 viruses from poultry in 16 provinces in China, resulting in an isolation rate of 0.28%. These viruses were predominantly detected in domestic ducks within live poultry markets, with eight NA subtypes identified, among which H4N2 and H4N6 were the most prevalent. These findings indicate that H4 viruses have been circulating persistently among poultry populations in China. Most of these provinces are located in southern China, where environmental conditions such as abundant water networks and a favorable climate support large-scale poultry farming, especially waterfowl production. Wild waterfowl, which are natural reservoirs of AIV, often carry the virus latently. The virus can subsequently infect nearby domesticated waterfowl or land birds through contaminated water or food (Simancas-Racines et al., 2023). A global analysis of H4 virus distribution revealed that approximately 81.0% of isolates were obtained from wild birds, followed by poultry and environmental samples highlighting the key role of wild birds in the worldwide dissemination of H4 AIVs (Song, X. et al., 2024a). The warm and humid climate in the southern region provides an ideal environment for wild birds, thereby increasing the risk of virus transmission to poultry through contact and promoting the persistence and spread of LPAIVs (Luo et al., 2017). Additionally, live poultry markets are recognized as hotspots for AIV transmission, offering a conducive environment for genetic recombination and mutation among different subtypes of AIVs (Parums, 2023). Collectively, these factors facilitate the emergence and spread of H4 AIVs, leading to continued diversification of NA subtypes.

The genome of AIV is composed of eight highly variable RNA segments (Blagodatski et al., 2021). To further investigate the formation and spread of these viruses, some studies define virus genotype by grouping individual gene fragments according to differences in nucleotide sequence homology (Verhagen et al., 2021). In this study, we classified 20 representative H4Nx viruses into 19 genotypes based on a threshold of 95% nucleotide sequence homology. This classification indicates frequent gene reassortment among H4 viruses circulating in China, highlighting their high level of diversity. Phylogenetic analysis revealed that all H4 AIVs belong to the Eurasian lineage, but their gene origins are complex, with most involving recombination with LPAIVs circulating in China and neighboring regions. Ge et al. reported that some internal genes of H4 AIVs exhibit high homology with those of H5N1 and H5N8 viruses, supporting the role of LPAIVs as potential gene donors in the emergence of reassortant HPAIVs (Ge et al., 2025). The research conducted by Song et al. showed that the PB2 and PB1 genes of H4N6 viruses of wild bird origin were derived from clade 2.3.4.4b H5 HPAIVs in Japan, highlighting ongoing reassortment between HPAIVs and LPAIVs (Song, X. et al., 2024b). Furthermore, we found that the NA gene of DK/HB/H1140/18(H4N8) is closely related to two human H10N8 virus strains within the same genetic clade. Although no human infections with H4 viruses have been reported to date, these viruses demonstrate a considerable capacity for cross-species transmission and can infect multiple mammalian species (He et al., 2024). Therefore, it is imperative to enhance surveillance and control measures for AIV to reduce the risk of viral spread to human populations.

Domestic ducks are not only the primary hosts of H4 AIVs in China but also serve as key vectors for the transmission of AIVs from wild waterfowl to terrestrial poultry (Murcia et al., 2022). The population of domestic ducks in China constitutes over 75% of the global total. Since most of these ducks inhabit open waters, they are highly susceptible to viral infections through direct contact with infected birds or indirect contact with contaminated water and food (Deng et al., 2013). Although domestic ducks typically remain asymptomatic after infection, they provide a conducive environment for the recombination of different subtypes of LPAIVs, potentially leading to the emergence of novel HPAIVs (Wu, H. et al., 2015). In this study, we demonstrated that four strains of H4 viruses infecting domestic ducks did not produce clinical symptoms and replicated only in specific organs, exhibiting limited transmission within the flock. Although H4 viruses are less pathogenic to domestic ducks, there are undetectable tissue lesions in the respiratory organs, which may result in some damage to these birds. Some studies have indicated that duck-derived H4 viruses can be transmitted among ducks through direct contact and are also capable of causing systemic infections in chickens, confirming that certain H4 viruses have acquired the ability to infect land birds (Shi et al., 2016). A chicken-derived H4N6 virus isolated from a live poultry market in central China has been reported to cause fatalities in chickens (Liu, M. et al., 2003). However, it remains to be further verified whether H4 viruses can cross species barriers via domestic ducks to infect land birds and establish stable transmission in chickens.

Under natural selective pressure, AIVs continuously undergo genetic mutations to adapt to new hosts (Brown et al., 2001). Song et al. demonstrated that the Q226L and G228S mutations in the HA protein are critical for enabling H4 AIVs to switch from avian-type to human-type receptor binding (Song, H. et al., 2017). This phenomenon has been observed in certain avian- or swine-origin H4 viruses. There were no mutations at positions 226 and 228 in the isolates examined in this study. Additionally, multiple adaptive mutations associated with enhanced pathogenicity in mammals were identified in the internal gene segments of these H4 viruses. In mouse infectivity assays, we found that none of the 20 H4 viruses resulted in mortality in mice. Most strains were isolated from the lungs and nasal turbinates of mice, indicating that H4 viruses have the potential to cross the species barrier and infect mammals. Notably, the DK/AH/AG61/11(H4N6) and DK/HB/H2091/17(H4N6) strains exhibited opposing effects on mouse body weight, which may be linked to the A588V mutation in the PB2 protein. Therefore, we compared the replication capacity of these two viral strains in chicken embryos and mammalian cells. The results showed that that DK/AH/AG61/11(H4N6) replicated more efficiently DK/HB/H2091/17(H4N6) in both chicken embryos and MDCK cells, with a particularly significant difference observed in MDCK cells. In addition to infecting mice, Liang et al. confirmed that the H4 subtype AIV can be transmitted between guinea pigs through direct contact and can also spread via inefficient respiratory droplets (Liang et al., 2016). Therefore, enhanced surveillance of H4 subtype AIV is essential for timely understanding and monitoring of its epidemiological dynamics and evolutionary trends, which is crucial for preventing and addressing potential pandemic risks.

Conclusions

In summary, this study conducted active surveillance of avian influenza in several provinces of China from 2011 to 2022. A total of 427 strains of H4 subtype AIVs were isolated, with domestic ducks identified as the primary infectious host. The 20 tested H4Nx viruses were classified into genotypes G1–G19 based on the homology differences of each gene fragment, all of which belong to the Eurasian lineage. Notably, the 226th and 228th positions of the HA protein in all strains did not exhibit mutations. Some strains contained drug resistance mutations, including NA I117V and D198N. The PB2 A588V mutation was observed only in the DK/AH/AG61/11(H4N6) strain. We assessed the pathogenicity of H4 viruses in waterfowl and mammals using SPF ducks and BALB/c mice. Our findings revealed that these H4 viruses caused relatively little harm to domestic ducks; however, they were capable of replicating in the lungs and nasal turbinates of mice, with certain strains leading to significant weight loss and demonstrating the ability to replicate efficiently in MDCK cells. H4 viruses remain prevalent among poultry in China. Although classified as a LPAIV, its potential zoonotic capability should not be ignored.

Materials and methods

Cells

MDCK cells, stored at China Animal Health and Epidemiology Center, were cultured in Dulbecco's Modified Eagle's Medium (DMEM; Gibco, USA) supplemented with 10% fetal bovine serum (FBS; Gibco, USA) and 1% penicillin/streptomycin, and incubated at 37 °C under 5% CO₂.

Sample collection

From 2011 to 2022, our laboratory conducted routine surveillance for avian influenza viruses in China, covering 28 provinces, including Heilongjiang, Jilin, Liaoning, Shanxi, Hebei, Beijing, Henan, Hubei, Hunan, Jiangsu, Zhejiang, Anhui, Shanghai, Fujian, Jiangxi, Shandong, Chongqing, Sichuan, Yunnan, Tibet, Shaanxi, Qinghai, Ningxia, Xinjiang, Guangdong, Guangxi, Guizhou, and Hainan. The samples were primarily collected from domesticated birds, such as chickens and domestic ducks, from poultry farms, live poultry markets, slaughterhouses, and wholesale markets. Due to variations in local policies and the impact of the epidemics, the number of samples collected varied from year to year, with a total exceeding 10,000 samples. Pharyngeal and anal swabs from poultry were placed in a preservation solution containing penicillin (2000 IU/mL), streptomycin (2 mg/mL), and 10% glycerol (v/v) in sterile phosphate-buffered saline (pH 7.2). These samples were transported to the laboratory at 4 °C within 72 h and stored at −70 °C.

Virus isolation and identification

The samples were vortexed and centrifuged, and the resulting supernatant was inoculated into 10-day-old SPF chicken embryos (Spafas Poultry Co., Ltd, Jinan, China) and incubated at 37 °C for 96 h. The allantoic fluid from all chicken embryos was aseptically harvested, and positive samples were identified by hemagglutination assay. Real-time fluorescent quantitative reverse transcription PCR (RT-qPCR) was employed to determine the subtypes of viruses in the positive samples (Forward primer: 5′-GGCCAATGGGACTATGGTAAA-3′; Reverse primer: 5′-CTTAGAGGRCTCGGGCATA-3′; Probe: FAM-TGCAGTGACCACTTCCACTTGRTRTCRTC-BHQ1; unpublished). The isolated H4 viruses were purified through three rounds of limiting dilution in SPF chicken embryos and subsequently stored at −80 °C.

Epidemiological analysis

A total of 427 strains of H4 viruses were isolated from swab samples collected from live poultry markets and farms. We analyzed the sample size, sampling locations, and positive rates. To investigate the epidemiology of the H4 subtype AIV in China, this study retrieved HA gene sequences of H4 viruses from the GISAID database and collected relevant information regarding hosts, subtypes, and regions. These data were further analyzed to explore the temporal, spatial, and host-based distributions of the viruses.

RNA extraction and genome sequencing

Representative H4 viruses were selected based on sampling time, host, subtype, and isolation rate by province. Then, RNA was extracted from these viruses using the FinePure Virus DNA/RNA Kit (Genfine Biotech Co., Ltd, Changzhou, China) according to the manufacturer's instructions. Subsequently, the full-length genome was amplified using the HiScript High Fidelity One Step RT-PCR Kit (Vazyme Biotech Co., Ltd). Primers sequences and amplification conditions followed the methodology described by Hoffmann et al. (2001). The RT-PCR products were purified using the SteadyPure Agarose Gel DNA Purification Kit (Accurate Biotechnology Co., Ltd, Hunan, China) and sequenced on an ABI 3730xl DNA Analyzer (Applied Biosystems, Carlsbad, CA, the United States). The sequencing results were complied, and the coding region of each gene segment were assembled using the SeqMan software within the DNASTAR® Lasergene® package (DNASTAR, Inc., Madison, WI, the United States). The complete gene sequences of the viruses were submitted to the GISAID EpiFlu database to obtain the accession number (EPI 3440428-EPI3440435, EPI3961979-EPI3962122).

Sequence analysis

Key amino acid sites on each gene segment of the virus were analyzed using Megalign software. Full-length gene sequences of the viruses were compared by BLAST against the sequences of published strains in the GISAID database. Reference sequences that reflect the genetic background of the viruses were selected for phylogenetic analysis. MAFFT software (v7.520) was employed to align the virus gene fragment sequences with the reference sequences. MEGA software (v11.0.13) and the tvBOT online platform were utilized to construct the Neighbor-Joining phylogenetic tree, with a Bootstrap value of 1000 (Xie et al., 2023). A 95% nucleotide sequence identity threshold was applied to classify clades in the phylogenetic trees of the eight gene segments.

Infection test on ducks

The test virus was titrated to 10⁶ EID₅₀/0.1 mL and used to infect eight 3-week-old SPF ducks (Vital River Laboratory Animal Technology Co., Ltd., Beijing, China) via nasal inoculation at a dose of 200 μL. On day 3 post-inoculation, three infected ducks were euthanized. Aseptically collected samples included 1 g each of lung, trachea, liver, cecum, spleen, kidney, pancreas, and bursa of Fabricius, which were separately placed in 1 mL of 1% PBS solution. Virus replication titers were determined through chicken embryo inoculation. Concurrently, tracheal and lung tissues were preserved in 4% neutral buffered formalin for subsequent histopathological evaluation using hematoxylin-eosin (HE) staining. Twenty-four hours after inoculation, five uninfected ducks were introduced into isolation cages with the five infected ducks to assess virus transmissibility among the ducks. Oropharyngeal and cloacal swabs were collected from both the infected group and the contact group on days 1, 3, 5, 7, and 9 post-exposure, and virus titration was performed using the chick embryo inoculation method. Clinical signs of the experimental ducks were monitored daily. At the end of the 14-day trial period, blood samples from surviving animals were collected to detect seroconversion, followed by euthanasia.

Infection test on mice

Six-week-old female BALB/c mice (Vital River Laboratory Animal Technology Co., Ltd., Beijing, China) were randomly divided into an experimental group and a control group, with eight mice in each group. After mild anesthesia using dry ice, the mice were inoculated with virus suspension containing 10⁶ EID₅₀ and sterile 0.01 M PBS buffer at a dose of 50 μL via nasal drops, respectively. On 3 days post-infection, three mice from each group were randomly selected for euthanasia, and their lungs, nasal turbinates, spleens, kidneys, and brains were aseptically collected. Each tissue samples was processed through homogenization and centrifugation, followed by inoculation into chicken embryos. The viral titer was calculated using the Reed-Muench method to assess the replication level of the virus in the mice. Clinical signs, weight changes, and mortality were monitored and recorded daily until the conclusion of the experimental period.

Viral growth kinetics

Representative H4 viruses were inoculated into SPF chicken embryos at an infectious dose of 10³ EID₅₀ and incubated at 37 °C in an incubator. Allantoic fluid from the chicken embryos was collected and stored at −80 °C at 24, 36, 48, 60, and 72 hpi. Concurrently, the virus was inoculated onto MDCK monolayer cells at an MOI of 0.01. After incubating at 37 °C for 1 h, the viral solution was discarded, and the cells were washed. A viral maintenance solution (DMEM medium containing 1 μg/mL TPCK-trypsin and 1% antibiotics) was then added, followed by incubation at 37 °C in a 5% CO₂ cell culture incubator. Cell supernatants were collected at 12, 24, 36, 48, 60, and 72 hpi and stored at −80 °C. All viral samples obtained were assayed for viral titers in MDCK cells using the TCID₅₀ method.

Data availability

The data supporting the findings of this study are available within the article and its supplementary materials. Genomes sequences of H4 AIVs isolated in this study are available from GISAID EpiFlu database (Isolate ID: EPI_ISL_209681, EPI_ISL_19267510, EPI_ISL_19731904 to EPI_ISL_19731921), and in Science Data Bank (https://doi.org/10.57760/sciencedb.27898).

Ethics statement

The study protocol received approval from the Animal Welfare Ethical Review Committee of the China Animal Health and Epidemiology Center in Qingdao, China (Approval No. DWFL-2023-04). All research involving H4 viruses was conducted in a biosafety level 2 laboratory (BSL-2), and animal experiments were carried out in a high-efficiency filtered isolator. All experimental animals were treated humanely in accordance with animal welfare principles.

Author contributions

Qiuyan Mao and Junfeng Zhu: methodology, investigation, visualization and writing-original draft. Shuo Liu and Peng Cheng: validation and formal analysis. Tiantian Wu, Jie Tian, Xiaoqi Li and Jizhe Yang, investigation and data curation. Jinping Li and Guangyu Hou, resources. Wenming Jiang and Hualei Liu: conceptualization, writing-review & editing, supervision, and funding acquisition.

Conflict of interest

No potential conflict of interest was reported by the author(s).

Appendix A. Supplementary data

The following are the Supplementary data to this article:

Supplementary Figure S1.

Distribution of various NA subtypes of H4 AIVs isolated in China.

Supplementary Figure S2.

The number of H4 AIVs isolated in China during the surveillance period from 2011 to 2022. A Distribution of host origin for 427H4 viruses. B Distribution of NA subtypes among the 427H4 viruses.

Supplementary Figure S3.

Phylogenetic trees of NA genes of H4N1 (A), H4N2 (B), H4N3 (C), H4N6 (D), and H4N8 (E) AIVs detected from 2011 to 2022. Viruses included in this study are highlighted in bold red font. The two reported cases of swine infection in China are marked with a black swine figure, while the two cases of H10N8 human infection are indicated with a red human figure. The trees were constructed using the neighbor-joining method in MEGA software (v11.0.13), and bootstrap analysis was conducted with 1000 replicates. Clade numbers are displayed on the right side of the panel.

Supplementary Figure S4.

. Phylogenetic trees of PB2 (A), PB1 (B), PA (C), NP (D), MP (E), and NS (F) genes of H4 AIVs detected from 2011 to 2022. Viruses included in this study are highlighted in bold red font. The trees were constructed using the neighbor-joining method in MEGA software (v11.0.13), and bootstrap analysis was conducted with 1000 replicates. Clade numbers are indicated on the right side of the panel.