ABSTRACT
The recent outbreak of H7N9 influenza in China has resulted in many human cases with a high fatality rate. Poultry are the likely source of infection for humans on the basis of sequence analysis and virus isolations from live bird markets, but it is not clear which species of birds are most likely to be infected and shedding levels of virus sufficient to infect humans. Intranasal inoculation of chickens, Japanese quail, pigeons, Pekin ducks, Mallard ducks, Muscovy ducks, and Embden geese with 106 50% egg infective doses of the A/Anhui/1/2013 virus resulted in infection but no clinical disease signs. Virus shedding was much higher and prolonged in quail and chickens than in the other species. Quail effectively transmitted the virus to direct contacts, but pigeons and Pekin ducks did not. In all species, virus was detected at much higher titers from oropharyngeal swabs than cloacal swabs. The hemagglutinin gene from samples collected from selected experimentally infected birds was sequenced, and three amino acid differences were commonly observed when the sequence was compared to the sequence of A/Anhui/1/2013: N123D, N149D, and L217Q. Leucine at position 217 is highly conserved for human isolates and is associated with α2,6-sialic acid binding. Different amino acid combinations were observed, suggesting that the inoculum had viral subpopulations that were selected after passage in birds. These experimental studies corroborate the finding that certain poultry species are reservoirs of the H7N9 influenza virus and that the virus is highly tropic for the upper respiratory tract, so testing of bird species should preferentially be conducted with oropharyngeal swabs for the best sensitivity.
IMPORTANCE The recent outbreak of H7N9 influenza in China has resulted in a number of human infections with a high case fatality rate. The source of the viral outbreak is suspected to be poultry, but definitive data on the source of the infection are not available. This study provides experimental data to show that quail and chickens are susceptible to infection, shed large amounts of virus, and are likely important in the spread of the virus to humans. Other poultry species can be infected and shed virus but are less likely to play a role of transmitting the virus to humans. Pigeons were previously suggested to be a possible source of the virus because of isolation of the virus from several pigeons in poultry markets in China, but experimental studies show that they are generally resistant to infection and are unlikely to play a role in the spread of the virus.
INTRODUCTION
On 1 April 2013, the People's Republic of China reported the first 3 human cases of a novel H7N9 subtype of influenza A virus. Over the subsequent days, the number of confirmed cases ballooned to over 82, with over 17 fatalities occurring in 6 different provinces (1). Sequence analysis of the virus showed the H7 and N9 genes to be those of avian influenza viruses (AIVs) of Eurasian lineage, but at only 95% similarity to other isolates in the public sequence databases, the viruses were uniquely different from what had previously been described (2). However, the internal genes were all closely related to the well-established H9N2 lineage of avian influenza virus circulating in poultry in the region since at least 1997 (2, 3). This H9N2 lineage virus has also been associated with human disease (4, 5).
The epidemiology of H7N9 virus showed that human cases were widely distributed in the affected provinces and there was no strong evidence of human-to-human transmission. Because the genome sequences of the isolates showed that they were genetically related to AIVs, Chinese veterinary officials quickly started testing poultry associated with live bird markets and commercial poultry operations and wild birds in the regions where human infections were being reported. The H7N9 viruses were detected at relatively low rates in avian species in the live bird markets, including chickens, pigeons, and ducks, and the environment (6, 7). Additional evidence of an epidemiological link of exposure to birds in markets has been found in some human cases (8, 9). Therefore, live bird markets were suspected of being a source of human infections, and Chinese officials required temporary closure of live poultry markets in the affected provinces, resulting in an immediate reduction of human cases, providing further evidence of a role of live poultry markets in the spread of the virus and that closure of the markets is an effective control strategy (10). However, it is not clear which species of birds are most likely to be infected and are shedding levels of virus sufficient to infect humans. The lack of understanding of the virus ecology in birds has recently resulted in an additional number of human cases, demonstrating that the virus still circulates in China (11). Based on the initial reports of this virus and previous experience with avian influenza, we evaluated the potential role of different poultry species in the epidemiology of H7N9 influenza.
MATERIALS AND METHODS
Virus.The virus used in this study was A/Anhui/1/2013 (H7N9), which was kindly provided by the Centers for Disease Control and Prevention, Atlanta, GA. The virus was propagated in specific-pathogen-free (SPF) embryonating chicken eggs (ECEs) according to standard procedures (12). Allantoic fluid was diluted in brain heart infusion (BHI) medium (BD Bioscience, Sparks, MD) in order to prepare an inoculum with 102, 104, or 106 50% egg infective doses (EID50) per 0.1 ml per bird. All challenge doses were confirmed by back-titration in ECEs. All experiments using the H7N9 influenza virus, including work with animals, were reviewed by the institutional biosecurity committee and were performed in biosecurity level 3 enhanced (BSL-3E) and animal biosecurity level 3 enhanced (ABSL-3E) facilities at the Southeast Poultry Research Laboratory (SEPRL), Agricultural Research Service, United States Department of Agriculture (USDA), and all personnel were required to wear a powered air-purifying respirator with high-efficiency particulate air (HEPA) filtration (3M, St. Paul, MN).
Birds.Fifty-nine-week-old SPF White Leghorn chickens (Gallus gallus domesticus) (egg layer type) were obtained from SEPRL in-house flocks. Four-week-old Japanese quail (Coturnix japonica), 6- to 12-month-old rock pigeons (Columbia livia domestica), 2-week-old Pekin ducks (Anas platyrhynchos var. domestica), 2-week-old Mallard ducks (Anas platyrhynchos), 2-week-old Muscovy ducks (Cairina moschata), and 2-week-old Embden geese (Anser anser domesticus) were obtained from commercial farms. Birds were housed in self-contained isolation units that were ventilated under negative pressure with HEPA-filtered air and maintained under continuous lighting. Serum samples were collected from all birds immediately prior to challenge and found to be negative for antibodies to the H7 subtype of influenza virus by hemagglutination inhibition assay, as described below. Feed and water were provided with ad libitum access. All bird experiments were approved by and performed under the regulations of the SEPRL Institutional Animal Care and Use Committee.
Pathogenicity and virus transmission studies.Eleven chickens, 11 Japanese quail, 11 pigeons, 11 Pekin ducks, 11 Mallard ducks, 7 Muscovy ducks, and 9 Embden geese were intranasally (i.n.) inoculated through the choanal cleft with an inoculum containing 106.0 EID50 of A/Anhui/1/2013 (H7N9) in 0.1 ml. At 3 days postinoculation (dpi), 2 to 3 birds from each group were euthanized and gross lesions were recorded. The following tissues were collected in 10% neutral buffered formalin solution to determine microscopic lesions and the extent of virus replication in tissues: nasal turbinates, trachea, lung, air sac, comb, eyelid, heart, brain, esophagus, proventriculus, ventriculus, duodenum, jejunum, cecal tonsils, pancreas, liver, spleen, bursa, thymus, Harderian gland, kidney, gonads, adrenal gland, and muscle from the breast and left thigh. Lung, spleen, intestine, kidney, and thigh and breast muscle tissues, as well as allantoic fluid from eggs laid by inoculated chickens, were collected separately in BHI and kept frozen at −70°C for subsequent virus detection. The remaining birds were observed for clinical signs over an 11-day period, during which time any clinical signs were recorded.
To evaluate the susceptibility of quail, pigeons, and Pekin ducks, three doses of virus (102, 104, or 106 EID50) were administered i.n. to groups of five birds. At 2 dpi, three uninfected birds were placed in the same cage with the directly inoculated birds in each dose group to determine the transmission potential of the virus by contact exposure.
Oropharyngeal (OP) and cloacal (C) swab specimens were collected at 2, 4, 6, 8, and 11 dpi from directly inoculated birds and at 2, 4, 6, and 9 days after birds infected through contact exposure to determine virus shedding. Swab specimens were collected in 2 ml of BHI medium with 1× antibiotic-antimycotic. All birds remaining at the end of the experiment were euthanized by the intravenous (i.v.) administration of sodium pentobarbital (100 mg/kg of body weight).
RNA extraction and quantitative rRT-PCR.RNA was extracted from swab and tissue specimens using a MagMAX 96 AI/ND viral RNA isolation kit (Ambion, Inc. Austin, TX) with a KingFisher magnetic particle processor (Thermo Scientific, Waltham, MA). Quantitative real-time reverse transcription-PCR (rRT-PCR) was performed using a SmartCycler (version 2.0) apparatus and an AgPath-ID OneStep RT-PCR kit (Ambion, Inc.). The processing of the samples from the chicken, quail, and pigeon experiment was performed at the same time, and an H7 assay was used for quantitation (13). For the duck and goose samples, the matrix gene assay was used for quantitation (14). A standard curve for virus quantification was established with RNA extracted from dilutions of the same titrated stock of virus used to inoculate the birds. Viral titers were extrapolated from the standard curve. Data were analyzed using Prism (v.5.01) software (GraphPad Software Inc.). Two-way analysis of variance with Tukey's posttest was used to the compare virus titers in oropharyngeal swab specimens. For statistical purposes, all oropharyngeal swab specimens from which viruses were not detected were given a numeric value of 101.0 EID50/ml. Statistical significance was set at a P value of <0.05.
Virus replication in tissues.Virus replication in lung, spleen, intestine, kidney, and muscle tissues from 2 to 3 birds was examined at 3 days following intranasal infection with the H7N9 virus. Titers of infectious virus were determined in ECEs or by rRT-PCR, as described above. Allantoic fluid collected from virus-inoculated laying chickens was also examined by rRT-PCR for the presence of virus.
HI assays.Hemagglutination inhibition (HI) assays were used to evaluate antibody to H7 influenza virus prior to challenge and to confirm exposure and infection with serum collected at 11 dpi and preadsorbed with 10% chicken red blood cells (15). The HI assay was conducted in accordance with standard procedures. Briefly, 2-fold serial dilutions of 25 μl of serum were made in 25 μl of phosphate-buffered saline (PBS). Diluted sera were incubated for 30 min at 4°C with 4 hemagglutination units (HAU)/25 μl of A/Anhui/1/2013 (H7N9) virus which had been treated with 0.1% beta-propiolactone (the pH was adjusted to 7.0 with sodium bicarbonate), and then 50 μl of 0.1% chicken red blood cells was added. The test result was evaluated after 30 min of incubation at room temperature. Titers were calculated as the reciprocal of the last HI-positive serum dilution, and samples with HI titers of 8 or below were considered negative.
Histopathology and IHC.Tissues were prepared for histopathology and immunohistochemistry (IHC) as previously described (16). Briefly, collected tissues were fixed by submersion in 10% neutral buffered formalin and embedded in paraffin. In addition, the nasal cavity was decalcified for 2 days. Sections were made at 5 μm and were stained with hematoxylin-eosin (HE). A duplicate section was immunohistochemically stained by first microwaving the sections in Citra antigen retrieval solution (BioGenex, San Ramon, CA) for antigen exposure. A 1:2,000 dilution of a mouse-derived monoclonal antibody (P13C11) (17) specific for type A influenza virus nucleoprotein (NP) was applied, and the mixture was allowed to incubate overnight at 4°C (18). The primary antibody was then detected by the application of biotinylated goat anti-mouse IgG secondary antibody using a biotin-streptavidin detection system (supersensitive multilink immunodetection system; BioGenex). Fast Red TR (BioGenex) served as the substrate chromogen, and hematoxylin was used as a counterstain. All tissues were systematically screened for microscopic lesions and virus antigen staining.
Sequencing.Viral RNA from selected OP and cloacal swab samples collected at 2 to 11 dpi from birds of each species directly inoculated with virus and from contact-exposed birds was directly amplified by one-step reverse transcription-PCR (RT-PCR) for the HA1 gene and the region of the PB2 protein around amino acid position 627. Selected viruses isolated from the experimentally inoculated birds and propagated in ECEs were also sequenced for comparison. A commercial one-step RT-PCR kit (Qiagen Inc., Valencia, CA) and primers which matched the sequence of A/Anhui/1/2013 (H7N9) virus and which were directed to the conserved sequences at the ends of the gene segments were used. Primer sequences are available upon request. Templates were then purified by agarose gel extraction with a QIAquick gel extraction kit (Qiagen Inc., Valencia CA) and quantified by UV spectroscopy. A BigDye Terminator kit (Applied Biosystems, Foster City, CA) was used for cycle sequencing, and the samples were subsequently run on an ABI 3730 DNA analyzer (Applied Biosystems, Foster City, CA). Bioinformatics analysis looking at the amino acid differences of the hemagglutinin (HA) gene was performed using the Influenza Research Database, which was used to look for single nucleotide polymorphisms (19).
RESULTS
Pathogenicity of H7N9 influenza virus in different poultry species.No clinical disease signs were observed in any of the directly inoculated or contact-exposed birds during the 11-day observation period. Results for virus shedding are presented in Table 1. Virus was detected in OP swabs taken at 2 dpi from all the quail and Muscovy ducks and from 10 of 11 chickens, 7 of 11 pigeons, 9 of 11 Pekin ducks, 10 of 11 Mallard ducks, and 7 of 9 Embden geese inoculated with 106.0 EID50 of the virus. At this time point, virus was detected in cloacal swabs from 4 of 11 chickens, 10 of 11 quail, 5 of 11 pigeons, 2 of 7 Muscovy ducks, and in none of the Pekin ducks, Mallard ducks, and geese. At 4 dpi, virus was detected in OP swabs of all chickens, quail, and Muscovy ducks but in smaller numbers of pigeons, Pekin ducks, Mallard ducks, and geese. Cloacal swab specimens from all quail, 3 of 5 Muscovy ducks, and 4 of 8 the chickens were virus positive, but swab specimens from the rest of the birds were negative. By 6 dpi, pigeons, Pekin ducks, and Mallard ducks had stop shedding by both the OP and C routes. Two of 6 geese had positive OP swabs at this time but were negative for both OP and C swabs on the rest of the days. Limited numbers of chicken and quail continued shedding virus until 11 dpi, and some Muscovy ducks were still shedding virus at 8 dpi.
Susceptibility of poultry to influenza A (H7N9) virus
High virus titers were present in OP swabs, with quail shedding an average of 107.5 EID50 on days 2 and 4 dpi and chickens shedding 106.2 EID50, on average, on the same days. Most chickens and quail continued shedding virus at 6 and 8 dpi, although the titers had decreased (Fig. 1). The titers in cloacal swabs averaged 3 to 4 log units lower than those in OP swabs. Only 7 of 11 directly exposed pigeons in the pathogenesis experimental group shed detectable virus at 2 dpi, with most shedding virus at low titers close to the assay limit of detection. One pigeon with the highest titer at day 2 continued to shed virus on day 4, but no other pigeons had detectable viral shedding for the remainder of the study. In the waterfowl study, Muscovy ducks had the highest numbers of birds shedding virus and the highest titers, with all birds shedding virus by the oropharyngeal route on days 2, 4, and 6 and some shedding virus by the cloaca until day 8. The Pekin ducks, Mallard ducks, and Embden geese had similar patterns of infection, with most directly inoculated birds shedding virus by the oropharyngeal route on day 2 and with decreasing numbers of positive birds being detected on days 4 and 6 and no virus being detected from the cloacal swabs (Table 1).
Comparison of oropharyngeal virus shedding after experimental challenge. Oropharyngeal shedding was detected at 2, 4, 6, 8, and 11 dpi with 106 EID50 of A/Anhui/1/2013 (H7N9) virus. The rRT-PCR values were interpolated by quantitative real-time RT-PCR using a standard curve generated with the challenge isolate. The estimated lower limit of sensitivity of the rRT-PCR test was 102, as determined by the limit of detection on the standard curve. The results are presented as the average viral shedding for each day, and samples negative by rRT-PCR were given a value 1 log unit lower than the limit of detection.
Susceptibility and transmission of H7N9 influenza virus in quail, pigeons, and Pekin ducks.Quail, pigeons, and Pekin ducks were chosen on the basis of the differences in pathogenicity observed as described above. Results are presented in Table 2. All directly inoculated quail in all three dose groups eventually became infected, and virus was transmitted to all contacts. Only a single quail receiving 102 EID50 was infected at 2 dpi, but it shed enough virus to infect its cage mates. In contrast, although some pigeons in the groups inoculated with 104 and 106 EID50 were shedding virus at 2 dpi, none of the contact-exposed pigeons became infected. No virus was detected from the pigeons that received 102 EID50 of the virus. Similarly, 2 of 5 and 4 of 5 Pekin ducks in the groups receiving 104 and 106 EID50, respectively, became infected, but the virus was transmitted to only 2 contact ducks, which shed virus for only a short period of time.
Transmission of influenza A (H7N9) virus in quail, pigeons, and ducks
Gross and microscopic lesions and virus antigen staining in tissues.Very mild gross lesions were observed at necropsy and mainly consisted of mild sinusitis. Microscopic lesions were consistent with low-pathogenic AIV (LPAIV) infection. Most of the lesions were confined to the upper respiratory tract. In the chickens and quail, the virus caused moderate to severe catarrhal and/or lymphocytic rhinitis and sinusitis, with mucocellular exudates containing sloughed epithelial cells, submucosal edema, and glandular hyperplasia (Fig. 2). The trachea presented mild degenerative changes of the overlying epithelium, mild lymphocytic infiltration in the submucosa, and mild edema (Fig. 2). In the lung, mild congestion, mild interstitial inflammation with mixed mononuclear cells, and mild catarrhal bronchitis were observed. Lesions in the gastrointestinal tract consisted of mild proliferation of gut-associated lymphoid tissues (GALTs). The remaining organs lacked significant histopathologic lesions. In ducks and geese, similar to the chickens and quail, most of the microscopic lesions were found in the upper respiratory tract (nasal turbinates, trachea); however, no lesions were observed in any other tissues, including the enteric tract. Mild to severe catarrhal rhinitis with congestion and loss of epithelial cells lining the nasal cavity was present in some ducks. In others, lymphoplasmacytic inflammation of the nasal submucosa was observed. Tracheitis with exudates and epithelial loss was common. In geese, mild to moderate lymphocytic rhinitis was the only lesion observed. No significant lesions were found in noninoculated birds.
Histopathology and immunohistochemical staining for avian influenza virus antigen in tissues of quail intranasally infected with the A/Anhui/1/2013 (H7N9) virus 3 dpi. Virus is stained in red. (A and B) Nasal epithelium. Severe necrotizing rhinitis with submucosal congestion and edema, glandular hyperplasia, and lymphoplasmacytic infiltration (A) and viral antigen in epithelial cells (B) are shown (magnification, ×200). (C and D) Nasal epithelium. Necrosis of epithelial cells and lymphocytic infiltration (arrowhead) (C) and viral antigen in epithelial cells (D) are shown (magnification, ×400). (E and F) Nasal gland. Lymphocytic infiltration in submucosa (arrowheads) (E) and viral antigen in epithelial cells (F) are shown (magnification, ×200). (G and H) Trachea. Necrosis of epithelial cells (arrowhead) (G) and viral antigen staining in epithelial cells (H) are shown (magnification, ×400).
In order to determine sites of virus replication, immunohistochemical staining for AIV antigen with an antibody to NP was conducted. Common viral staining was present in the epithelial cells and macrophages of the nasal cavity and adjacent glands in all of the quail and chickens examined (Fig. 2). Similar viral antigen staining was present in the nasal epithelium of 2 of 3 Pekin ducks, 3 of 3 Mallards, and 1 of 2 Muscovy ducks but not in the geese or pigeons. Viral staining was also present in epithelial cells, macrophages, and desquamated cells of the trachea of one quail and one Muscovy duck and in enterocytes and submucosal macrophages in the intestine of one quail and two Muscovy ducks.
Virus detection in tissue samples.Virus isolation and/or virus detection by rRT-PCR from tissues collected at 3 dpi from birds infected with 106 EID50 of the H7N9 virus was attempted. Low virus titers (100.97 to 101.23 EID50) were detected in the intestine of two Muscovy ducks, one quail, and one chicken; in the spleen of three quail, one chicken, and one Mallard duck; and in the kidney of one Pekin duck, two Mallard ducks, one Muscovy duck, and two geese. However, the lungs and muscle tissues and the contents of eggs laid by infected chickens were virus negative.
Serology.When examined at 11 dpi, all quail, most chickens and Muscovy ducks, and some Pekin and Mallard ducks infected with 106 EID50 of the H7N9 virus had detectable titers of antibodies against the virus, but no pigeons or geese seroconverted (Table 1). All quail given 104 to 106 EID50 of the virus and 3 of 5 quail given 102 EID50 also seroconverted (Table 2), and antibodies were also detected in contact-exposed quail. However, most pigeons and Pekin ducks had undetectable antibody titers regardless of the virus dose given.
Sequencing.The HA1 gene and part of the PB2 gene were RT-PCR amplified and sequenced from selected virus-positive OP swabs collected at 2 to 11 dpi from inoculated chickens (n = 3), quail (n = 2), a pigeon (n = 1), a Pekin duck (n = 1), and a Muscovy duck (n = 1) and from contact control birds chickens (n = 3) and quail (n = 6). Similar numbers of sequences were obtained from virus isolated in embryonating chicken eggs from OP swabs. All samples had lysine (K) at position 627 of the PB2 gene, which is the same as the sequence of the parent human isolate. Lysine at this position is associated with virulence in mammals (20). However, three amino acid differences were commonly observed in the HA1 gene sequence compared to the A/Anhui/1/2013 sequence: N123D, N149D, and L217Q (H7 numbering) (Table 3). Both the virus that was originally received from China and the virus that was used as the inoculum for these experiments and that had been passaged once in embryonating chicken eggs were sequenced. While the virus received from China had asparagine (N) residues at both positions 123 and 149, the egg-passaged inoculum virus had aspartic acid (D) at both positions. Both isolates maintained the leucine (L) at position 217. Four different combinations of amino acids were observed in the infected birds. Most birds in this study continued to have aspartic acid at positions 123 and 149, but some isolates from chickens and Mallard and Pekin ducks had asparagine at both positions. The most consistent difference observed was the change at position 217 from leucine (L) to glutamine (Q). Position 217 correlates to position 226 in H3 human influenza viruses, and position 226 is critical for determining specificity to either α2,6 human-like sialic acid receptors or α2,3 avian-like sialic acid receptors (21). Most of the Chinese H7N9 sequences have leucine at this position, but passage in poultry seems to provide selection pressure for glutamine. Leucine was still found in a minority of isolates from both challenged and contact-infected birds. All the viruses that were passaged in eggs had glutamine at position 217. Examination of GenBank for sequence variation of Eurasian H7 influenza viruses at position 217 showed that over 99% of AIVs have glutamine at this position, and none are reported to have leucine. At position 123, asparagine is found in over 97% of viruses and aspartic acid is found in less than 3% of viruses, and at position 149, asparagine is found in 99% of isolates, with no viruses having aspartic acid.
Comparison of common sequence polymorphisms in H7 HA1 and PB2 proteins between parental strain A/Anhui/1/2013 (H7N9) and derivative viruses recovered from birds inoculated with A/Anhui/1/2013a
DISCUSSION
Quail are experimentally susceptible to many subtypes of both mammalian and avian influenza viruses (22–24) and have been proposed to be a bridging species or disease amplifiers between wild waterfowl and domestic gallinaceous poultry (25–29). In this study, we demonstrate that quail are susceptible to even a low-dose challenge of the Chinese H7N9 virus. The virus replicated to high titers in the upper respiratory tract for at least a week, and the virus transmitted easily by direct contact to cage mates. Although quail are proportionally a minor poultry species, they have the potential to be a major reservoir of the H7N9 virus for transmission to other poultry and to humans. The adult chickens in this study also shed high levels of virus, indicating that chickens are also an important source of virus which could be infecting humans either through direct contact or by aerosolization of virus, which occurs in particular during the slaughter process in live bird markets (30).
Our data support the suggestion that pigeons are generally resistant to AIV infection, with only 1 of 26 pigeons shedding moderate titers of virus. Historically, isolation of AIV from pigeons is rare, with less than 40 sequences appearing in GenBank, including the sequences of 8 H9N2 viruses and no H7 viruses (19). Experimentally, pigeons have generally been resistant to H5N1 highly pathogenic AIV (HPAIV) infection (31–34); however, inoculation of pigeons with high virus doses or with specific strains resulted in infrequent morbidity and mortality (35–37). Experimental data suggest that pigeons are unlikely to play a major role in the maintenance and transmission of the Chinese H7N9 virus. Exposure to high levels of virus from chickens or other species in live bird markets could explain the reported H7N9 isolations from pigeons (6).
In this study, we also examined the pathogenesis of the H7N9 influenza virus in three different types of ducks and one type of goose. Pekin and Muscovy ducks and Embden geese are domestic waterfowl frequently present in live bird markets in China. Although closely related to Pekin ducks, we chose to also include Mallard ducks to address the possibility that wild birds could become infected with this virus by contact with domestic poultry and possibly spread the virus to other areas. Wild Mallard ducks also have one of the highest isolation rates for AIV and are a primary reservoir in the wild (38). All four species could be infected by high-dose challenge with the virus, but the birds did not show any clinically observable disease and most species shed relatively small amounts of virus for shorter periods of time than quail and chickens. There was some evidence of contact transmission in Pekin ducks, but the ducks infected through contact shed little virus and shed virus for only short periods of time. Of the four species, Muscovy ducks shed the most virus. This is not surprising, since Muscovy ducks have been shown to be more susceptible to infection with highly pathogenic H5N1 AIV strains, show more severe disease, and shed larger amounts of virus than other domestic duck species (39–41). Muscovy ducks, it must be remembered, are a different species (Cairina moschata) than Pekin and Mallard ducks (Anas platyrhynchos) and should not be expected to have a similar response to infection (42). These differences in response to AIV infection in different waterfowl species should be taken into account when determining which species are involved in the transmission of emerging viruses. In this study, Muscovy ducks appeared to play a more important role as a possible biological vector of H7N9 AIV than Pekin ducks, Mallard ducks, and Embden geese.
AIV is maintained in wild birds, but occasionally, the virus can spread from its natural reservoir to poultry. Wild-bird AIVs are generally poorly adapted to domestic galliformes (chickens, quail, partridge), but as conditions permit, the virus can be transmitted and adapt to the new host. Wild aquatic birds do not typically show clinical signs of infection with AIVs, and although AIVs can replicate in cells of both the respiratory and intestinal tracts, in ducks they are reported to favor the intestinal tract (43, 44). The results of these studies are consistent with those of previous studies indicating that chicken-adapted AIVs replicate better in chickens than in ducks (45, 46). The underlining mechanism is not clear, but a shorter neuraminidase protein due to a deletion in the stalk region may be linked to this feature (47–49).
Control of H7N9 influenza is complicated by the lack of disease signs in poultry. Detection of LPAIVs is more difficult than detection of HPAIVs, like H5N1, because testing cannot target sick or dead birds like syndromic surveillance can. Critically, the data from both quail and chickens show high levels of viral replication in the upper respiratory tract and the shedding of much less virus in cloacal swabs, findings which are not unexpected, because poultry-adapted AIVs are typically shed at much higher levels in the respiratory tract in gallinaceous poultry (45, 46, 50–52). The disease pathogenesis of the H7N9 virus was unusual, in that virus replication was primarily restricted to the upper respiratory tract for all the species examined and the virus did not replicate well in lungs. Testing of gallinaceous and waterfowl bird species should preferentially be conducted with OP swabs for the best sensitivity.
One of the unusual features of this H7N9 virus was the presence of leucine at position 217 in the HA1 protein, a sequence which many have speculated increased the ability of the virus to infect humans. This position, analogous to position 226 of H3 viruses, forms part of the receptor binding site. The presence of leucine at this position is associated with binding to α2,6-sialic acid, which is the primary type of sialic acid found in the human upper respiratory tract. The presence of glutamine at this position is associated with binding to α2,3-sialic acid, the most common type of sialic acid found in avian species (53, 54). Receptor binding studies have confirmed that the human H7N9 virus with leucine at position 226 binds more strongly to α2,6-sialic acid than an avian H7 strain, but the human virus also had strong binding to α2,3-sialic acid (55). Examining published Eurasian H7 sequences, glutamine is almost exclusively found at position 217 in avian influenza viruses, so the presence of leucine in the human isolate appears to support a change for adaptation to humans. The change of leucine back to glutamine in most of the challenged birds shows positive selection for glutamine in poultry and because of the mixed results suggests that the challenge virus was a mixed population. The change of L217Q in egg-passaged virus is also not surprising, as egg adaptation of human viruses is a common occurrence (56). Experimental studies examining a European H7 virus showed that when the double mutation of Q226L and G228S was introduced into the hemagglutinin protein, the tropism of the hemagglutinin changed from α2,3-sialic acid to α2,6-sialic acid (14), further supporting our observations. The single mutation Q226L was not examined, but because most Chinese H7N9 viruses already have leucine, it suggests that a single additional change could result in a major shift in viral tropism. Two other amino acid changes at positions 123 and 149 were also commonly observed. The amino acid at position 123 is part of the 120 loop that is projected to form part of the receptor binding site, but it is not clear how this amino acid might affect viral binding (57). The role of aspartic acid at position 149 also remains unclear. The data from this study and the sequence data for H7 viruses in general suggest a conundrum about the H7N9 reservoir, because although the H7N9 virus with leucine at position 226 was transmitted in poultry, the majority of transmitted virus in this study had glutamine. Are there other animal species in the Chinese wet markets that may be a bridge species that supports leucine at position 226, or does the Q226L mutation occur after the virus jumps the species barrier to humans?
The epidemiology of the H7N9 virus outbreak clearly has an important poultry component, on the basis of the sequence analysis and the epidemiology of the virus, and these experimental studies have shown that chickens and quail are likely important reservoirs of the virus and pigeons are not. The role of domestic waterfowl is less clear, with relatively high levels of virus being shed by Muscovy ducks and smaller amounts being shed by Pekin ducks and Embden geese. Efforts in China have targeted live bird markets in an attempt to control the virus, and understanding the contribution of poultry farms, wholesale markets, and the markets themselves in the maintenance of the virus is critical. In the live bird market system in the United States, poultry farms are generally free of infected birds, but with H7N2 LPAIV, wholesale markets seemed to be the major sources of the virus (58).
The H7 virus has infected humans on numerous occasions, although the clinical disease has usually been mild. The Chinese H7N9 virus is unusual in the manifestation of severe disease in humans with a high case fatality rate. The viral sequence suggests that the virus is poised to mutate to a form that has α2,6-sialic acid receptor specificity, which is likely a prerequisite for human-to-human transmission, which could lead to a highly virulent pandemic. An understanding of viral pathogenesis and key reservoir species can potentially allow interventions in animals to stop a future human pandemic.
ACKNOWLEDGMENTS
We thank Tim Olivier, Suzanne Deblois, Diane Smith, Kira Moresco, Mark Freeman, Michelle Edenfield, Scott Lee, and Raul Otalora for technical assistance and Ruben Donis and Todd Davis at the Centers for Disease Control and Prevention for providing the challenge virus.
This research was supported by U.S. Department of Agriculture CRIS Project 6612-32000-063-00D and federal funds from the National Institute of Allergy and Infectious Diseases, National Institutes of Health (NIH), under IAA no. AAI12004-001-00001.
The contents of this article are solely the responsibility of the authors and do not necessarily represent the official views of the NIH.
FOOTNOTES
- Received 12 December 2013.
- Accepted 16 February 2014.
- Accepted manuscript posted online 26 February 2014.
- Copyright © 2014, American Society for Microbiology. All Rights Reserved.
