Novel Reassortment of Eurasian Avian-Like and Pandemic/2009 Influenza Viruses in Swine: Infectious Potential for Humans ^{▿ †}

Surveillance.

On an approximately weekly basis from December 2009 to June 2010, a total of 3,600 tracheal swabs and 1,020 serum samples were collected from slaughtered pigs at abattoirs in the Guangdong (GD) (swabs, n = 2,240; sera, n = 625; sampling occasions, n = 21) and Guangxi (GX) (swabs, n = 1,360; sera, n = 395; sampling occasions, n = 13) provinces of the People's Republic of China. The samples collected in Guangdong represent pigs introduced from many neighboring provinces, while those collected in Guangxi are mainly from local pigs.

HI assays.

Sera were pretreated with a receptor-destroying enzyme (RDE; Denka Seiken Co. Ltd., Tokyo, Japan) to destroy nonspecific inhibitors, followed by heat inactivation at 56°C for 30 min. RDE-treated sera were then absorbed with turkey red blood cells (TRBC) to remove nonspecific agglutination substances. Antibody titers were determined by testing serial 2-fold dilutions (1:10 to 1:2,560) of each serum in duplicate. Hemagglutination inhibition (HI) assays were performed in 96-well microtiter plates (Corning Costar Co.) with 0.5% turkey erythrocytes using four hemagglutination units of virus.

Serological survey.

HI assays were performed on each of the 1,020 serum samples collected, with a contemporary human H3N2 virus (A/Shantou/1328/2008) and five representative swine H1 influenza virus strains: A/Sw/HK/294/09 (CS H1N2), A/Sw/HK/915/04 (TR H1N2), A/Sw/HK/NS1583/09 (pdm/09 H1N1), A/Sw/HK/2433/09 (EA H1N1), and A/Sw/HK/1532/09 (EA-like H1N1 variant) (24, 26, 27).

Virus isolation and sequencing.

Isolation of viruses from tracheal swabs in Madin-Darby canine kidney (MDCK) cells, viral RNA extraction, cDNA synthesis, PCR, and sequencing were carried out as previously described (24, 26, 27). All 288 nucleotide sequences of the segments of the 36 influenza virus isolates detected in this study have been deposited in GenBank under accession numbers JN374994 to JN375281. The virus A/Swine/Guangdong/1361/2010 (H1N1) was plaque purified and resequenced to confirm its identity before use in subsequent experiments.

Phylogenetic analysis.

For each gene segment identified here and representative influenza virus sequences from GenBank, maximum-likelihood (ML) phylogenies were inferred using the heuristic tree search method Garli 1.0 (34). Phylogenetic support for branch points was estimated by bootstrap analysis with 100 replicates using the same ML method.

Animals.

All pigs (n = 16; local domestic hybrid pigs, Putian White × Nianbian variants) used were confirmed to be free of influenza virus by virus isolation in MDCK cells and to be seronegative against circulating swine and human influenza viruses (CS, TR, EA, pdm/09, and seasonal H1N1 and H3N2), as well as avian H5N1 and H9N2 viruses, by HI assays prior to the start of the study. Ferrets (n = 9) negative for both influenza virus and antibodies were obtained from the experimental ferret breeding program at Sangosho Pet Park Co., Ltd. Animal experiments were approved by the Shantou University Medical College in compliance with the University Policies “Animal Ethics and Welfare” and “Use of Animals in Research” and the guidelines of the World Health Organization and the International Council for Laboratory Animal Science.

Viral infectivity studies.

Infection and transmission studies were carried out in biosafety level 3 (BSL-3) containment laboratories at 20 to 21°C and 76.5% ± 2.1% relative humidity. All animals were moved to the BSL-3 lab at least 4 days prior to the experiment for acclimatization. A microchip (implantable programmable temperature transponder IPTT-300; BioMedic Data Systems) was implanted in the skin of each animal, between the shoulder blades, to measure subcutaneous temperatures.

Immunization of ferrets with the CA4 virus.

Twenty-two-week-old male ferrets (n = 3) were inoculated with 10⁶ PFU of A/California/04/2009 (CA4, the prototype pdm/09 virus) and gained high titers of antibody (HI > 1,280) to this virus on the 14th day post-primary infection (dpi). These immunized ferrets were used for the transmission experiment at 80 dpi.

Infection and transmission experiments with Sw/GD1361 virus.

Four-week-old piglets (n = 7) were intranasally inoculated with 2 × 10⁶ PFU, i.e., 2.56 × 10⁶ TCID₅₀ (median tissue culture infective dose) of Sw/GD1361 virus in 2 ml of minimum essential medium (MEM; Gibco), delivered with a mucosal atomization device (MAD nasal drug delivery device; Wolfe Tory Medical, Inc.) to mimic aerogenous infection (2). Two naive pigs and three naive ferrets inoculated with phosphate-buffered saline (PBS) were used as negative-control animals.

At 24 h post-inoculation (hpi) of the pigs, immunologically naive pigs (n = 7; 4 weeks old) and ferrets (naive ferrets, n = 3; ferrets with antibodies against CA4, n = 3; 33.5 weeks old) were introduced into the animal infection and transmission facility holding the inoculated pigs (n = 7) (see Fig. S3 in the supplemental material). Four immunologically naive pigs were used as physical contact animals, and three were used for aerosol contact. The aerosol contact pigs and ferrets were housed in adjacent double-layer steel wired cages with a distance of at least 10 cm from those holding the directly inoculated pigs. Airflow inside and between each pair of cages was <0.1 m/s. To avoid inadvertent physical contact and artificial transmission, aerosol contact animals were always handled first; drinking water and food stock, gloves, and any other items in contact with the animals or their bedding were kept sterile by decontamination or reserved for the exclusive use of each individual animal.

Animal monitoring.

Body weights and temperatures were recorded daily around the same time (9:30 to 10:30 a.m.) for each animal. Clinical signs were observed twice daily. Nasal swabs from each piglet were collected daily and placed into 0.6 ml of cold sterile PBS with antibiotics. Nasal washes from each ferret were collected daily into 1 ml of PBS with antibiotics. The endpoint infectivity titration (TCID₅₀) was determined for all swabs and nasal washes in MDCK cells.

Animal serology and histology.

Blood was collected via venipuncture of the anterior vena cava of the pigs or from the ferret tail artery. Seroconversion was monitored by determining the HI titers of pre- and postexposure sera.

Four directly inoculated pigs were euthanized for postmortem examination (2 pigs at 4 dpi and 2 pigs at 6 dpi) by intracardiac injection of pentobarbital sodium (100 to 200 mg/kg of body weight). One physical contact pig was similarly euthanized at 5 days postcontact (dpc). Freshly excised tissues of the trachea, lungs, nasal turbinates, and other major organs of euthanized animals were fixed in 10% phosphate-buffered formalin, dehydrated, embedded in paraffin, and cut into 5-μm-thick sections. Standard hematoxylin and eosin (H&E) (Sigma) staining, as well as immunohistochemistry (IHC) assays with a mouse anti-NP (nucleoprotein) monoclonal antibody, were performed as described previously (32).

Infection of human lung tissue in ex vivo culture.

Similar to the method described in an earlier report (31), fresh lung tissues were surgically removed from patients with lung carcinoma, in accordance with a protocol approved by the Ethical Review Board of Shantou University Medical College. Only normal nonmalignant tissue fragments that were in excess of the requirements of clinical diagnosis were used. Tissues were cut into ∼3-mm by 4-mm by 2-mm cubes and placed in an F-12K nutrient mixture (Gibco) with l-glutamine and antibiotics. Three viruses were used for inoculation: CA4 (pdm/09), Sw/GD1361 (the novel reassortant; this work), and A/Wuhan/359/1995 (seasonal H3N2). For each virus, infections were performed in triplicate, with 7 lung tissue cubes per replicate inoculated with 0.5 × 10⁶ PFU of virus in 0.5-ml inocula in 6-well plates, and allowed to absorb for 1 h at 37°C and 5% CO₂. The tissue cubes were then washed three times with culture medium and incubated with 0.5 ml of F-12K medium supplemented with 0.2% TPCK (N-tosyl-l-phenylalanine chloromethyl ketone)-trypsin, 1% BSA (bovine serum albumin), and antibiotics. At 18 and 36 hpi, lung tissue cubes (n = 5 each per virus) were individually rinsed with medium, homogenized in 0.5 ml of cold PBS, and then clarified by centrifugation. Viral titers of the homogenates were determined by TCID₅₀ assays with MDCK cells. The remaining cubes (n = 11 per virus) were fixed and examined for viral NP expression by immunohistochemical staining (31, 32).

Seroprevalence of influenza virus in pigs of southern China.

In the study period, based on hemagglutination inhibition (HI) assays, about 51% (521/1,020) of the sera were positive (HI titer ≥ 1:80) against at least one of five reference H1 strains, while all were negative to the contemporary human H3N2 influenza virus (HI titer ≤ 1:20) (Table 1; see also Fig. S1 in the supplemental material).

Of the 625 sera collected in GD (over the entire surveillance period), 104 (16.6%) were positive solely for the pdm/09-like virus, while 35 (8.9%) of the sera from GX were positive solely for this virus. Additionally, 295 (28.9%) sera (GD, n = 166; GX, n = 129) were positive for pdm/09 and one or more other viruses (Table 1), indicating a high frequency of coinfection or multiple virus exposure in the pigs from southern China, although cross-reaction cannot be completely excluded. Some of the sera had extremely high HI titers (≥1:2,560) for different H1 reference viruses (see Fig. S1), suggesting a recent infection.

Genetic characterization of swine influenza virus isolates.

Thirty-six H1N1 and H1N2 influenza viruses were isolated from tracheal swabs obtained in Guangdong, but none were isolated from those taken in Guangxi. Full-length sequences were obtained for each of the eight gene segments of all 36 swine virus isolates. Phylogenetic analyses of the H1 hemagglutinin (HA) gene revealed that these viruses clustered into three different lineages: the pdm/09, CS, and EA virus lineages (see Fig. S2 in the supplemental material).

For each of the remaining genes, phylogenetic analyses revealed that these viruses belonged to four distinct genotypes: pdm/09-like H1N1 (n = 12, four sampling occasions), a previously undescribed EA-like H1N2 variant with its HA gene from the CS lineage and NA (N2) gene from the human lineage (n = 5, single sampling occasion), a novel reassortant with EA-like H1N1 surface genes and pdm/09-like internal genes (n = 10, single sampling occasion), and an EA-like H1N1 variant with the nonstructural (NS) gene from the TR lineage (n = 9, single sampling occasion) (Fig. 1; see also Fig. S2 in the supplemental material).

It was noted that viruses isolated from the same sampling occasion were very closely related to each other and always clustered together in the phylogenetic trees (see Fig. S2 in the supplemental material). All pdm/09-like viruses were detected from late December 2009 to the end of January 2010 (from four sampling occasions), while the remaining viruses were isolated in February and May 2010 (Fig. 1).

To see if there is any evidence of early adaptation of pdm/09-like viruses in pigs, the sequences of the 22 viruses containing pdm/09-like genes were compared with all of the pdm/09 sequences in GenBank (as of 5 March 2011). Nineteen of these viruses contained a total of 11 substitutions in 6 internal genes that were absent in all human pdm/09 virus sequences (Table 2). Ten of the novel reassortant viruses, represented by A/swine/Guangdong/1361/2010 (Sw/GD1361), had 6 of these substitutions (in the PB2, PA, M1, and NP genes), while Sw/GD/286/2010 had 2 unique substitutions (in the PB2 and PA genes). The other 8 viruses contained only one substitution, in the PB2 gene (5 viruses, represented by Sw/GD/275/2010) and the NS1/NS2 gene (3 viruses, represented by Sw/GD/94/2009). Only the substitution in the NS1/NS2 gene was observed on more than one sampling occasion.

Assessment of a novel H1N1 EA-pdm/09 reassortant virus.

The infectivity, transmissibility, and pathogenicity of a representative isolate with EA-like HA and NA genes and pdm/09-like internal genes (Sw/GD1361) were tested with experimental animals. Interspecies transmissibility and the potential for cross-protection from prior pdm/09 infection were investigated.

Infectivity and transmissibility in pigs.

Seven uninfected pigs were cohoused, either in physical or aerosol contact, with seven pigs experimentally inoculated with the Sw/GD1361 virus (see Fig. S3 in the supplemental material). Virus shedding from each of the inoculated pigs was detected from the first day postinoculation (dpi) till the 7th day, with a peak at 4 dpi (6.0 ± 0.3 log TCID₅₀/ml swab material) (Fig. 2A). In the physical-contact pigs, virus replication in the nasal cavity lasted from 1 to 7 days postcontact (dpc), with peak titers of 6.0 ± 0.7 log TCID₅₀/ml (Fig. 2B). In the aerosol-contact group, pigs started to shed virus from the nasal route between 3 and 5 dpc, and virus was detected for at least 4 to 6 days, with peak titers of 5.9 ± 1.3 log TCID₅₀/ml (Fig. 2C). There were no statistically significant differences in viral shedding among the inoculated, physical-contact or aerosol-contact pigs (Fig. 2A to C), and the peak virus shedding titers were comparable to those of prototype EA-like (27) and pdm/09 viruses (2, 12, 28, 29, 33), as previously reported. Lethargy, lower activity levels, and reduced interest in food occurred in the pigs, but no major clinical symptoms of infection were observed. On 15 dpi or 14 dpc, all experimental pigs, either inoculated or exposed by physical or aerosol contact, had seroconverted, with HI titers ranging from 160 to 1,280 (Table 3), which were comparable to those of EA-like viruses (27) and higher than those of pdm/09-like viruses (33). Cross-reaction of swine sera was observed only for the EA viruses. Naive pigs inoculated with PBS did not show virus shedding or seroconversion (data not shown).

Interspecies transmissibility of Sw/GD1361 to ferrets.

Ferrets were used as sentinel animals for an aerosol contact, interspecies transmission test. Two of the three immunologically naive ferrets (F1 and F2), held in separate cages with a distance of 10 cm from the cage of the infected pigs, began to shed virus at 2 dpc, with the third (F3) shedding virus at 5 dpc. Large amounts of virus (peak titer >5 log TCID₅₀/ml) were secreted from the nasal discharges of the ferrets, lasting for 5 to 6 days (Fig. 2D). Sneezing, nasal discharge, inactivity, and slight fever (1.3 to 1.4°C increase in body temperature) were observed in the ferrets. On 14 dpc, all ferrets had seroconverted, with an HI titer of at least 640 against the homologous virus (Table 3). In general, ferret sera showed broad cross-reactivity to all H1 viruses tested. Naive ferret controls did not show virus shedding, clinical signs, or seroconversion (data not shown).

Ferrets (F1* to F3*) with seroconversion against the prototype pandemic virus (A/California/04/09 [CA4]) developed signs of infection when cohoused with the other ferrets. Symptoms similar to those of its naive counterpart (F1) developed in ferret F1* at 2 dpc, with virus shedding for at least 4 days (2 to 5 dpc). The other seroconverted ferrets (F2* and F3*) also shed virus, although the onset was delayed 5 to 6 days (Fig. 2E). Comparison of the pre- and postexposure HI titers from these ferrets showed that two of them (F1* and F2*) had 4- to 8-fold increases in antibodies against the Sw/GD1361 virus, whereas the third one (F3*) had a 2-fold increase (Table 3). Even though ferrets seropositive to pdm/09 could be infected by Sw/GD1361, the amount of virus shed and the duration of shedding were reduced from those for the immunologically naive ferrets (Fig. 2D and E). In the viruses recovered from the nasal washes of all six contact ferrets, no amino acid substitutions that related to the antigenic sites (Ca1, Ca2, Cb, Sa, and Sb) of the HA protein were observed (data not shown).

Infectivity in the pig respiratory tract and ex vivo human lung tissue.

Euthanized pigs (four inoculated animals and one contact animal) showed similar lung lesions and clear evidence of virus replication in the nasal turbinate, trachea, and lower respiratory tract. Pulmonary consolidation and extensive bronchioalveolitis were also observed, characterized by multiple foci of severe inflammatory infiltrates and necrosis of bronchus epithelia (Fig. 3), which is similar to the effect of pdm/09-like virus infections (2, 12, 28, 29).

Immunostaining of viral proteins and TCID₅₀ titers showed that Sw/GD1361 and the prototype pdm/09 virus (CA4) could infect human lung tissues, whereas only very limited infection was observed with human seasonal H3N2 influenza virus (A/Wuhan/359/1995) (Fig. 4).