Introduction

Influenza A viruses, belonging to the family Orthomyxoviridae, contain a genome composed of eight segments of single-stranded, negative-sense RNA that each encode one or two proteins. Based on the antigenicity of their haemagglutinin (HA) and neuraminidase (NA), they are classified into 18 HA subtypes and 11 NA subtypes.

Influenza A viruses cause recurrent outbreaks in humans and domestic animals. Seasonal outbreaks occur annually, and pandemics occur with the emergence of new strains for which there is no immunity in humans [8]. Mechanistically, pandemic influenza can originate either by direct transmission from animals to humans or through reassortment between an avian influenza virus with a human influenza virus [7]. Since the influenza virus has a segmented genome, reassortment is an important mechanism for producing “novel” viruses [40]. Since the influenza global pandemic of 1918, reassortment of influenza virus genes among human influenza viruses with different subtypes, between human and avian viruses, and among animal influenza viruses [11] have resulted in the global pandemics of 1957 (H2N2) and 1968 (H3N2) [33, 38].

Early in 2009, a triple-reassortant swine-origin H1N1 influenza A virus (H1N1/2009), containing genes from avian, human, and swine influenza viruses, emerged among humans and spread globally through human-to-human transmission, causing the first influenza pandemic of the 21st century [6, 35]. The pandemic H1N1/2009 is generally associated with mild disease and a relatively low mortality rate comparable to that of seasonal influenza virus outbreaks [12, 29]. However, the possibility of the low-mortality pandemic H1N1 strain becoming more virulent due to reassortment with other high-mortality viruses is a global health concern.

Here, we report on a novel swine-origin H1N1 influenza virus that acquired a non-structural (NS) gene by reassortment from a highly pathogenic H5N1 virus. We characterized the virus by sequencing the complete coding regions of all eight segments and evaluated the effect of the acquired NS protein on viral virulence in both in vitro and in vivo systems. Our results indicate that the swine-origin H1N1 influenza virus can reassort with other co-circulating influenza viruses, and the novel NS-reassorted influenza virus replicated well in tissue culture and several organs of BALB/c mice but did not exhibit increased virulence. These results highlight the need for continuous surveillance of influenza virus transmission.

Material and methods

Ethics statement

The study protocol was approved by the ethics committee of Beijing Institute of Microbiology and Epidemiology (No. 2009100246). All participants signed an informed consent statement.

All animal studies were approved by the Committee for Laboratory Animal Care and Use of Beijing Institute of Microbiology and Epidemiology (No. 2009120067). All surgery was performed under sodium pentobarbital anesthesia, and all efforts were made to minimize suffering. All experiments were performed under animal biosafety level 2.

Cell lines and virus isolation and propagation

Madin-Darby canine kidney (MDCK) cells, HEp-2 cells, Vero cells (maintained in DMEM with 10% fetal bovine serum [FBS], Hyclone, Logan, USA) and A549 cells (maintained in F12 with 10% FBS, Hyclone) were obtained from ATCC and cultured at 37 °C in a 5% CO2 incubator.

Thirty throat swab specimens obtained from humans were inoculated onto MDCK cells. Virus isolates were identified and subtyped by reverse transcription PCR (RT-PCR) and with reference antisera.

Nucleotide sequencing and phylogenetic analysis

All eight gene segments of the viruses were amplified by RT-PCR as described [15]. In brief, viral RNA was extracted from the cell culture supernatant using an E.Z.N.A. Viral RNA Kit (Omega Bio-tek, Norcross, USA), and cDNA was prepared using the primer Uni12 (5’-AGCAAAAGCAGG-3’) with AMV reverse transcriptase (Promega, Madison, USA). The cDNA was amplified by using Pfu Ultra II fusion HS DNA polymerase (Stratagene, La Jolla, USA) with universal primers for influenza A virus. The PCR products were gel purified and cloned into pMD18-T. Five independent clones from each gene segment were sequenced by Beijing AuGCT Co. Ltd.

To investigate the relatedness of the viruses, we aligned the nucleotide sequences using the program Clustal X 2.0 [23]. Phylogenetic trees were constructed using the complete open reading frames of all eight gene segments, using the neighbor-joining (NJ) method in Molecular Evolutionary Genetics Analysis (MEGA, version 7) [22], and the reliability of the trees was evaluated by the bootstrap method with 1,000 replications. Reference virus sequences (S-OIV, published in 2009) were downloaded from the Influenza Virus Resource [1].

Virus titrations

Virus titers in cell culture supernatants or homogenized organ samples were determined by plaque assay or endpoint titration in MDCK cells, respectively. For plaque assay, MDCK cells were inoculated with a tenfold serial dilution of the sample, washed 1-2 h post-inoculation with phosphate-buffered saline (PBS), and covered by 0.5% agar in the maintenance medium (DMEM supplemented with 100 U of penicillin per ml, 100 μg of streptomycin per ml, 4 mM glutamine, and 1 μg of TPCK-treated trypsin) per ml. Three days after inoculation, the agar was removed and the cells were stained with 0.2% crystal violet. For endpoint titration, MDCK cells were inoculated with a tenfold serial dilution of each sample, washed 1-2 h postinfection (PI) with PBS, and grown in 200 μl of the maintenance medium. Five days after inoculation, the supernatants of inoculated cell cultures were tested for agglutinating activity using chicken erythrocytes as an indicator of virus replication in the cells. The titers were calculated by the method of Reed and Muench [34].

Replication curves

Cells were inoculated at a multiplicity of infection (MOI) of 0.001 plaque-forming units (PFU) per cell and washed once with PBS 1-2 h after inoculation, and the medium was replaced with fresh maintenance medium. Supernatants were collected at 24, 48, and 72 h PI, and the virus titers were determined by endpoint titration as described above.

Mouse experiments

Female 6-week-old BALB/c mice, obtained from the Institute of Jingfeng Medical Laboratory Animals, were inoculated intranasally with 500 PFU of infectious virus. Bodyweights were recorded daily as an indicator of disease.

Replication of the virus in nasal turbinates, lung, liver, brain and kidney tissues of mice (5 per group) was determined 3, 6 and 9 days PI. The nasal turbinates were washed with PBS, and the excised organs (brain, lung, liver, spleen, and kidney) were homogenized in 1 ml of PBS. Virus titers in the clarified homogenates were determined by endpoint titration in MDCK cells.

Cytokine quantitation

The in vivo levels of cytokines were determined for five individual mice per group. Lung tissues were collected from naive and infected mice on days 3, 6, and 9 PI and homogenized in 1 ml of cold PBS. The cytokines levels (TNF-α, IL-1β and IFN-γ) in the clarified lung homogenates we determined using enzyme-linked immunosorbent assay kits (Dakewe Biotech, Shenzhen, China) according to the manufacturer’s instructions.

Statistical analysis

The experimental data were analyzed using GraphPad Prism 6.0 and expressed as means ± SD. Comparisons between individual data were made using one-way ANOVA, and levels of significance (p-value) were determined. A p-value <0.05 was considered statistically significant.

Results

Virus isolation and sequence analysis

Thirty throat swab specimens obtained from H1N1/2009-infected humans from October 2009 to January 2010 were inoculated onto MDCK cells for primary isolation of H1N1/2009 influenza virus. The isolates were classified by hemagglutinin inhibition (HI) and RT-PCR according to the CDC protocol for real-time RT-PCR for influenza A (H1N1) virus published by WHO (http://www.who.int/csr/resources/publications/swineflu/realtimeptpcr/en). The viruses were propagated in MDCK cells.

Viral RNAs were extracted from cell culture supernatant and amplified by PCR with universal primers for influenza A virus. Complete full-length sequencing of all eight gene segments was performed on at least five cDNA clones for each segment, and a consensus sequence was produced for each virus. A nucleotide BLAST analysis indicated that one of our isolates shared the highest sequence similarity within the NS gene with an H5N1 influenza virus and A/Hong Kong/483/1997 (Table 1). We named this virus A/Guangdong/5301/2010(GD5301). The other seven gene segments showed the highest nucleotide sequence similarity to those of the pandemic H1N1/2009 virus (Table 1).

Table 1 Nucleotide sequence identity between A/Guangdong/5301/2009 and reference strains available in GenBank
Full size table

Phylogenetic analysis

To investigate the evolutionary origin of GD5301, phylogenetic analysis was conducted for all eight gene segments by the neighbor-joining method, using MEGA (version 7) [22] with reference virus sequences downloaded from the Influenza Virus Resource [1].

The topology of the HA gene tree revealed two distinct clades. The H1N1/2009 virus isolated from a human (GD5301) clustered with S-OIV isolates from humans and H1N1/2009 isolates from swine (Fig. 1A). The NA tree showed a similar phylogenetic relationship (Fig. 1B).

Fig. 1
figure 1figure 1figure 1

Phylogenetic trees of the HA, NA, and NS genes of the GD5301 influenza virus. The nucleotide sequences were analyzed by the neighbor-joining (NJ) method in Molecular Evolutionary Genetics Analysis (MEGA, version 7), and bootstrap values >50% are shown. The analysis was based on the complete open reading frames of all gene segments

Full size image

Three lineages of human, avian, and Goose/Guangdong/1/96-like viruses were observed in the phylogenetic tree of the NS gene, and GD5301 grouped with the human H5N1 cluster (Fig. 1C). Therefore, phylogenetic analysis confirmed that the NS gene of the GD5301 virus was acquired through reassortment with one H5N1 virus.

Molecular characterization

In order to evaluate the pathogenic properties of the GD5301 virus, we compared the amino acid sequences of the gene segments of the GD5301 virus with the recognized pathogenic determinants of influenza viruses. As shown in Table 2, sequence analysis showed the absence of markers normally associated with high pathogenicity, such as a multibasic hemagglutinin cleavage site [20], a lysine residue at position 627 [13], or an asparagine residue at position 701 [10, 24] in the PB2 protein. Moreover, there was no evidence of a serine residue at position 66 in the PB1-F2 [5] or a deletion of the stalk site in the NA protein [2]. However, the acquired NS protein may contribute to the pathogenicity of the GD5301 influenza virus, as it contains a glutamic acid residue at position 92 [37] and four C-terminal residues [18, 31]. The differences in NS1 between the GD5301 and A/Beijing/501/2009(H1N1) (BJ501) was showed in Fig. 2. The predicted amino acid sequences are 76.5% identical, with a total of 86.1% similar or identical amino acids.

Table 2 Determinants of viral pathogenicity
Full size table
Fig. 2
figure 2

Differences in the NS1 genes of isolates GD5301 and BJ501. The alignment of the sequences of the GD5301 and the BJ501 was analyzed using BioEdit 7.2.6. The similarity matrix was BLOSUM62. Light gray shading shows identical amino acids, and darker gray shading shows similar amino acids. The number of identical amino acids is 76.5%, and the number of similar of identical amino acids is 86.1%

Full size image

In vitro replication kinetics of the GD5301 virus

The replication kinetics of the GD5301 virus were studied in vitro. To do this, the virus was titrated in MDCK cells by plaque assay (Fig. 3A). Then, MDCK, Vero, A549 and Hep-2 cells were inoculated at a multiplicity of infection (MOI) of 0.001, the supernatants were harvested at fixed time points and the viral titers were determined in MDCK cells by TCID50 assay. As shown in Fig. 3B, the virus could replicate in all four cell lines, and the highest virus titer was obtained with MDCK cells.

Fig. 3
figure 3

Growth properties of the GD5301 virus in vitro. (A) Plaque phenotype of the GD5301 virus assayed on MDCK cells. Plaques were stained with crystal violent after incubation for three days at 37 °C. (B) Growth curves of the GD5301 virus on different cells. Cells were inoculated at an MOI of 0.001, and virus growth was assayed at the time points indicated. The figure is representative of three experiments

Full size image

In vivo viral growth and virulence of the GD5301 virus

The pathogenicity and viral growth of the GD5301 virus was evaluated in vivo in BALB/c mice that were inoculated intranasally with 500 PFU of virus. For comparison, we also challenged groups of mice with BJ501. Body weights were monitored daily as an indicator of disease. The GD5301-infected mice experienced a more than 20% peak body weight loss that started on day 3 PI and continued until day 7 Pi, after which the mice started to gain weight (Fig. 4). Similarly, mice challenged with BJ501 followed the same weight loss pattern; however, their weight loss was less drastic on day 3 PI than that of GD5301-challenged mice.

Fig. 4
figure 4

In vivo characterization of the GD5301 influenza virus. Weight loss of BALB/c mice inoculated with the GD5301 or BJ501 virus. Mice (n = 5) were inoculated with 500 PFU of virus or PBS, and their bodyweights were recorded each day after infection

Full size image

The presence of virus in the nasal washes and lungs and its ability to replicate in the respiratory tract was determined at multiple time points (Table 3). The GD5301 virus replicated in the respiratory tract without adaptation. The peak virus titers in the nasal washes and lungs were detected at day 6 and 3 PI, respectively. Virus was present up to day 9 PI in both nasal washes and lungs. GD5301 replicated more efficiently and reached a higher viral titer in the lungs than in the nasal cavity.

Table 3 Growth properties of the GD5301 virus in vivo a
Full size table

The ability of the virus to replicate in organs outside of the respiratory tract was also examined. GD5301 virus could replicate in almost all of the examined organs (Table 3). The viral titers in the brain and liver increased gradually and peaked on day 9 PI. The GD5301 virus titer in the kidney had the same pattern as in the nasal cavity.

Cytokine levels in the lungs of the infected mice

To better understand the pathogenicity of the GD5301 virus, the levels of TNF-α, IL-1β and IFN-γ in the lungs of the infected mice were examined (Fig. 5). The production of cytokines induced by the infection of the GD5301 virus could be detected as early as day 3 PI. IL-1β and IFN-γ reached peak levels at day 6 PI, which may have resulted from the highest virus load in the lung at day 3 PI.

Fig. 5
figure 5

Cytokine levels in lungs of infected mice. Lung homogenates collected on days 3, 6, and 9 after primary challenge were used to measure the levels of TNF-α, IL-1β, and IFN-γ, using enzyme-linked immunosorbent assays. Error bars show standard deviation

Full size image

Discussion

Early in 2009, the triple-reassortant S-OIV caused the first influenza pandemic of the 21st century, with mild disease and a relatively low mortality rate, but its presence posed a continued risk for further reassortment of the S-OIV with influenza virulence factors, considering that multiple lineages of avian H5N1 cocirculate with pH1N1 viruses [4]. Previous studies have indicated that reassortment between different influenza virus strains is possible in the event of a coinfection in a susceptible host [3, 32]. Here, we isolated a novel reassortant swine-origin H1N1 influenza virus that had acquired an NS gene from a highly pathogenic H5N1 virus.

We isolated 30 throat swab specimens from H1N1-infected humans, collected between October 2009 and January 2010, and determined the complete sequences of all eight segments. Nucleotide BLAST and phylogenetic analyses indicated that one of the isolated viruses had reassorted to acquire the NS gene of the H5N1 influenza virus, A/Hong Kong/483/1997. The molecular characteristics of the HA, NA, PB1-F2, and PB2 genes of the novel virus GD5301 are those of a low-pathogenicity influenza virus, but the presence of the nonstructural gene (NS) H5N1 from an influenza A virus could potentially contribute to high virulence [16, 21, 25, 40].

GD5301 replicated in the human airway epithelial cell lines Hep-2 and A549 at low titres in the absence of TPCK-trypsin, and more efficiently in MDCK cells with TPCK-trypsin. The pathogenicity of the virus was evaluated further in vivo by intranasal challenge of BALB/c mice. The virus caused disease in BALB/c mice, as indicated by weight loss, and it was able to replicate in the respiratory tract without adaptation. Moreover, GD5301 was maintained for a longer time than BJ501 and other pandemic S-OIVs, which were cleared from the lung at day 7 PI [17, 28]. The novel virus spread more systemically than BJ501 to other organs in mice. The pandemic S-OIV could not be detected in the organs outside the respiratory tract [14, 28], except in the intestinal tract [27].

Highly pathogenic H5N1 viruses, including HK483, were previously detected in brain, thymus, spleen and heart tissues [26, 30]. The introduction of the NS segment from an H5N1 influenza virus could alter cell tropism [25], and therefore the observed sustained viral load in the lung and systemic infection reported here could be due to the NS acquired from H5N1, and this certainly warrants further investigation.

It has been shown that the NS proteins of H5N1 viruses are associated with high levels of pro-inflammatory cytokines, which probably contributes to the high mortality of H5N1 virus infections [29, 37, 39]. In this study, the novel virus induced high levels of cytokines, and the influenza virus NS1 protein might contribute to viral replication and virulence by suppressing interferon-induced antiviral response [9], Rac1 signaling [19], or inhibition of PKR [36].

Our results indicate that further gene reassortment has been occurring and that the reassorted virus is maintained for a longer time in the lung than common H1N1/2009 and induces a cytokine imbalance in the lung. The present study enhances our understanding of the NS protein as a virulence factor and demonstrates that there is a continued risk for further reassortment. Therefore, systematic surveillance should be enhanced to prepare for the next possible pandemic, as several reassortant influenza viruses have been isolated in China recently [42, 43].