Functional Evolution of the 2009 Pandemic H1N1 Influenza Virus NS1 and PA in Humans

Abstract

In 2009, a pandemic H1N1 influenza A virus (IAV) (pH1N1) emerged in the human population from swine causing a pandemic. Importantly, this virus is still circulating in humans seasonally. To analyze the evolution of pH1N1 in humans, we sequenced viral genes encoding proteins inhibiting general gene expression (nonstructural protein 1 [NS1] and PA-X) from circulating seasonal viruses and compared them to the viruses isolated at the origin of the pandemic. Recent pH1N1 viruses contain amino acid changes in the NS1 protein (E55K, L90I, I123V, E125D, K131E, and N205S), as previously described (A. M. Clark, A. Nogales, L. Martinez-Sobrido, D. J. Topham, and M. L. DeDiego, J Virol 91:e00721-17, 2017, https://doi.org/10.1128/JVI.00721-17), and amino acid changes in the PA-X protein (V100I, N204S, R221Q, and L229S). These amino acid differences between early and more recent pH1N1 isolates are responsible for increased NS1-mediated inhibition of host gene expression and decreased PA-X-mediated shutoff, including innate immune response genes. In addition, currently circulating pH1N1 viruses have acquired amino acid changes in the PA protein (V100I, P224S, N321K, I330V, and R362K). A recombinant pH1N1 virus containing PA, PA-X, and NS1 genes from currently circulating viruses is fitter in replication in cultured cells and in mice and is slightly more pathogenic than the original ancestor pH1N1 virus. These results demonstrate the need to monitor the evolution of pH1N1 in humans for mutations in the viral genome that could result in enhanced virulence. Importantly, these results further support our previous findings suggesting that inhibition of global gene expression mediated by NS1 and PA-X proteins is subject to a balance which can determine virus pathogenesis and fitness.IMPORTANCE IAVs emerge in humans from animal reservoirs, causing unpredictable pandemics. One of these pandemics was caused by an H1N1 virus in 2009, and this virus is still circulating seasonally. To analyze host-virus adaptations likely affecting influenza virus pathogenesis, protein amino acid sequences from viruses circulating at the beginning of the pandemic and those circulating currently were compared. Currently circulating viruses have incorporated amino acid changes in two viral proteins (NS1 and PA-X), affecting innate immune responses, and in the PA gene. These amino acid differences led to increased NS1-mediated and decreased PA-X-mediated inhibition of host gene expression. A recombinant pH1N1 virus containing PA, PA-X, and NS1 genes from recently circulating viruses is fitter in replication in tissue culture cells and in mice, and the virus is more pathogenic in vivo Importantly, these results suggest that a balance in the abilities of NS1 and PA-X to induce host shutoff is beneficial for IAVs.

Keywords: NS1; PA-X; gene expression inhibition; inflammatory responses; influenza virus; innate immunity; interferon responses.

FIG 1

Frequencies of identified mutations in currently circulating pH1N1 viruses over time. Publicly available sequences in the Influenza Research Database were downloaded, and the frequencies of PA-X (A) and PA (B) sequences containing the original amino acids present in A/California/04/2009 (black) and the mutant amino acids (white) are represented according to the month and season of virus isolation. Mutation V100I affecting both PA-X and PA amino acid sequences is represented only in panel A for simplicity. The numbers of sequences available in the Influenza Research Database (https://www.fludb.org/) for the periods April to September 2009, October 2009 to September 2010, October 2010 to September 2011, October 2011 to September 2012, October 2012 to September 2013, October 2013 to September 2014, October 2014 to September 2015, October 2015 to September 2016, and October 2016 to the present were 2,654, 1,744, 599, 44, 611, 238, 129, 1,766, and 960, respectively.

FIG 2

Effect of pH1N1 NS1 and PA-X amino acid changes on inhibition of host gene expression. HEK293T cells were transiently cotransfected with pCAGGS plasmids expressing the indicated HA epitope-tagged NS1 and PA-X proteins, along with Gluc- and GFP-expressing plasmids. An empty pCAGGS plasmid was included as an internal control. NS1_MUT encodes the amino acid changes E55K, L90I, I123V, E125D, K131E, and N205S. PA-X_MUT encodes the amino acid changes V100I, N204S, R221Q, and L229S. At 24 hpt, Gluc (A), GFP (B), and NS1 (C) protein expression levels were analyzed. (A) Gluc expression was quantified in a Lumicount luminometer. Error bars represent the standard deviations for triplicates. *, P < 0.05 (NS1_WT versus NS1_MUT, PA-X_WT versus PA-X_MUT, NS1_MUT/PA-X_MUT versus NS1_MUT/PA-X_WT, and NS1_WT/PA-X_MUT versus NS1_WT/PA-X_WT) using Student's t test (n = 3 per time point). (B) GFP was visualized using a fluorescence microscope, and representative images obtained with a 20× objective are shown. Bars, 100 μm. (C) pH1N1 NS1 and PA-X and cellular actin protein expression levels were analyzed by Western blotting using cell extracts and antibodies specific to the HA epitope tag (to detect PA-X and NS1 proteins [top and bottom bands in the top blot, respectively]) and actin (loading control) (bottom blot). Western blots were quantified by densitometry using ImageJ software (v1.46), and the amounts of PA-X and NS1 proteins were normalized to the amounts of actin (top and bottom numbers, respectively, between the two blots). Protein expression levels in cells transfected with the pCAGGS plasmid expressing NS1_WT and PA-X_WT were considered 100%. ND, not detected. Molecular mass markers (in kilodaltons) are indicated on the left. The experiments were repeated three times, with similar results.

FIG 3

Effect of pH1N1 NS1 and PA-X amino acid changes on IFN responses induced by SeV infection. (A and B) HEK293T cells were transiently cotransfected, using calcium phosphate, with the indicated HA-tagged NS1- and PA-X-expressing pCAGGS plasmids, together with plasmids expressing Fluc under the control of an IFN-β (A) or an ISRE (B) promoter. NS1_MUT encodes the amino acid changes E55K, L90I, I123V, E125D, K131E, and N205S. PA-X_MUT encodes the amino acid changes V100I, N204S, R221Q, and L229S. An empty (E) pCAGGS plasmid was included as an internal control. At 24 hpt, cells were mock infected (M) or infected (MOI of 3) with the SeV Cantell strain (+SeV) to induce the activation of the IFN-β (A) or the ISRE (B) promoters. At 16 hpi, cell lysates were prepared for reporter gene expression. Fluc expression was measured by luminescence. Data show the means and standard deviations of the results determined for triplicate wells. Experiments were repeated 3 times in triplicate, with similar results. *, P < 0.05 (NS1_WT versus NS1_MUT, PA-X_WT versus PA-X_MUT, NS1_MUT/PA-X_MUT versus NS1_MUT/PA-X_WT, and NS1_WT/PA-X_MUT versus NS1_WT/PA-X_WT) using Student's t test (n = 3 per time point). (C) pH1N1 PA-X and NS1 and cellular actin protein expression levels were analyzed by Western blotting using cell extracts and antibodies specific to the HA tag (to detect PA-X and NS1 proteins [top and bottom bands in the top blot, respectively]) and actin (loading control) (bottom blot). Western blots were quantified by densitometry using ImageJ software (v1.46), and the amounts of PA-X and NS1 proteins were normalized to the amounts of actin (top and bottom numbers, respectively, between the two blots). Protein expression in cells transfected with the pCAGGS plasmid expressing NS1_WT and PA-X_WT was considered 100%. ND, not detected. Molecular mass markers (in kilodaltons) are indicated on the left. The experiments were repeated 3 times, with similar results.

FIG 4

Schematic representation of the recombinant pH1N1 viruses. (A) PA (left) and PA-X (right) WT (gray) and MUT (black) viral proteins. The 5 amino acid changes in pH1N1 PA (V100I, P224S, N321K, I330V, and R362K) and the 4 amino acid changes in pH1N1 PA-X (V100I, N204S, R221Q, and L229S) selected in currently circulating pH1N1 viruses are indicated. Numbers on the top indicate the amino acid lengths of the PA and PA-X proteins. The PA and PA-X proteins share the N-terminal 191 amino acids. The +1 frameshift is indicated with a striped bar. (B) pH1N1 NS1 WT (gray) (top) and MUT (black) (bottom) proteins containing 6 amino acid differences (E55K, L90I, I123V, E125D, K131E, and N205S) in currently circulating strains (38) are indicated. The number on the top indicates the amino acid length of WT and MUT NS1 proteins.

FIG 5

Recombinant pH1N1 mutant virus growth kinetics in MDCK cells. (A and B) Canine MDCK cells were infected (MOI of 0.001) in triplicates with the recombinant viruses encoding different variants of NS1 and PA proteins (NS1_MUT/PA_MUT, NS1_MUT/PA_WT, NS1_WT/PA_MUT, and NS1_WT/PA_WT) and incubated at 33°C (A) or 37°C (B). Virus titers in infected cell TCS were determined at the indicated hours postinfection by an immunofocus assay. ns, not significant. (C) MDCK cells were infected with the indicated pH1N1 viruses and incubated at 33°C (top) or 37°C (bottom) for 3 days. Plaque phenotypes were assessed by immunostaining with the anti-NP monoclonal antibody HB-65.

FIG 6

pH1N1 mutant virus growth kinetics in human A549 cells. (A and B) Human A549 cells were infected (MOI of 0.1) in triplicates with the recombinant viruses encoding different variants of NS1 and PA proteins (NS1_MUT/PA_MUT, NS1_MUT/PA_WT, NS1_WT/PA_MUT, and NS1_WT/PA_WT) and incubated at 33°C (A) or 37°C (B). Virus titers in infected cell TCS were determined at the indicated hours postinfection by an immunofocus assay. *, P < 0.05 (NS1_MUT/PA_MUT versus NS1_MUT/PA_WT, NS1_MUT/PA_MUT versus NS1_MUT/PA_WT, and NS1_MUT/PA_MUT versus NS1_WT/PA_WT) using Student's t test (n = 3 per time point); ns, not significant (P > 0.05). (C and D) Human A549 cells were infected (MOI of 0.1) in triplicates with viruses isolated from two patients (ACU009 and ACU022) infected during the 2015–2016 season (43) and with the A/California/04/2009 strain and incubated at 33°C (C) or 37°C (D). Virus titers in infected cell TCS were determined at the indicated hours postinfection by an immunofocus assay. *, P < 0.05 (virus isolates ACU009 and ACU022 versus A/California/04/2009).

FIG 7

Virulence of recombinant pH1N1 viruses containing amino acid changes in NS1 and PA. Groups of 7- to 8-week-old C57BL/6 female mice (n = 5/group) were infected with 100 FFU/mouse (A), 1,000 FFU/mouse (B), or 10,000 FFU/mouse (C) of the recombinant viruses encoding different variants of NS1 and PA proteins (NS1_MUT/PA_MUT, NS1_MUT/PA_WT, NS1_WT/PA_MUT, and NS1_WT/PA_WT). Weight loss (left) and survival (right) were evaluated daily for 2 weeks.

FIG 8

Growth and induction of innate immune responses in vivo of recombinant pH1N1 viruses encoding NS1 and/or PA variants. Groups of 7- to 8-week-old C57BL/6 female mice (n = 6/group) were infected with 1,000 FFU/mouse of NS1_MUT/PA_MUT, NS1_MUT/PA_WT, NS1_WT/PA_MUT, and NS1_WT/PA_WT recombinant pH1N1 viruses. (A and B) Mice were sacrificed at 2 dpi (n = 3) and 4 dpi (n = 3), and right lungs (A) and nasal turbinates (B) were harvested, homogenized, and used to quantify viral titers by an immunofocus assay (FFU per milliliter). (C) Left lungs were collected (n = 3), and total RNA was extracted to quantify levels of CCL2, TNF, IFN-β, and viral gene M1 mRNAs by RT-qPCR. *, P < 0.05 (NS1_MUT/PA_MUT versus NS1_MUT/PA_WT, NS1_MUT/PA_MUT versus NS1_MUT/PA_WT, and NS1_MUT/PA_MUT versus NS1_WT/PA_WT) using Student's t test (n = 3 per time point); ns, not significant (P > 0.05). r.u., relative units.