Accumulation of human-adapting mutations during circulation of A(H1N1)pdm09 influenza virus in humans in the United Kingdom

Abstract

The influenza pandemic that emerged in 2009 provided an unprecedented opportunity to study adaptation of a virus recently acquired from an animal source during human transmission. In the United Kingdom, the novel virus spread in three temporally distinct waves between 2009 and 2011. Phylogenetic analysis of complete viral genomes showed that mutations accumulated over time. Second- and third-wave viruses replicated more rapidly in human airway epithelial (HAE) cells than did the first-wave virus. In infected mice, weight loss varied between viral isolates from the same wave but showed no distinct pattern with wave and did not correlate with viral load in the mouse lungs or severity of disease in the human donor. However, second- and third-wave viruses induced less alpha interferon in the infected mouse lungs. NS1 protein, an interferon antagonist, had accumulated several mutations in second- and third-wave viruses. Recombinant viruses with the third-wave NS gene induced less interferon in human cells, but this alone did not account for increased virus fitness in HAE cells. Mutations in HA and NA genes in third-wave viruses caused increased binding to α-2,6-sialic acid and enhanced infectivity in human mucus. A recombinant virus with these two segments replicated more efficiently in HAE cells. A mutation in PA (N321K) enhanced polymerase activity of third-wave viruses and also provided a replicative advantage in HAE cells. Therefore, multiple mutations allowed incremental changes in viral fitness, which together may have contributed to the apparent increase in severity of A(H1N1)pdm09 influenza virus during successive waves.

Importance: Although most people infected with the 2009 pandemic influenza virus had mild or unapparent symptoms, some suffered severe and devastating disease. The reasons for this variability were unknown, but the numbers of severe cases increased during successive waves of human infection in the United Kingdom. To determine the causes of this variation, we studied genetic changes in virus isolates from individual hospitalized patients. There were no consistent differences between these viruses and those circulating in the community, but we found multiple evolutionary changes that in combination over time increased the virus's ability to infect human cells. These adaptations may explain the remarkable ability of A(H1N1)pdm09 virus to continue to circulate despite widespread immunity and the apparent increase in severity of influenza over successive waves of infection.

FIG 1

Phylogenetic relationship of complete influenza A(H1N1)pdm09 virus genomes. The tree is rooted on A/California/04/2009, shown as a blue-filled circle. United Kingdom first-wave isolates are highlighted as blue circles, while isolates sequenced by MOSAIC are shown as red circles for second-wave isolates from community and hospitalized patients, and green circles show isolates from third-wave community and hospitalized patients. Isolates characterized in this study are indicated with colored arrows, while the inferred ancestral location of the asparagine-lysine mutation in PA is indicated with a black arrow. Nodes with bootstrap support of >75% are highlighted with asterisks. Bar, 0.002 substitutions/site.

FIG 2

Replication of A(H1N1)pdm09 viruses in cell culture. (a and b) Viral growth of 10 clinical isolates from first-wave (blue), second-wave (red), or third-wave (green) viruses in MDCK cells (a) and human nasal MucilAir cell cultures (HAE; Epithelix) (b). Cells were infected at an MOI of 0.001 (MDCK cells) or 0.01 (MucilAir) and incubated at 34°C. The dashed line represents the mean for A/195. Statistics in the tables adjoining the chart keys were calculated using unpaired t tests with Holm-Sidak corrections. (c) Replication of a representative first-wave (A/195; blue circles) and third-wave (A/687; green squares) virus pair was assessed in HAE cells (left panel), MDCK cells (left middle panel), CALU3 cells (right middle panel), and pig tracheal cells (right panel). Cells were infected at an MOI of 0.001. Statistics were calculated using unpaired t tests.

FIG 3

Infection of mice with A(H1N1)pdm09 viruses. (a to c) Weight loss was followed after infection of 15 BALB/c mice inoculated intranasally with 2 × 10⁵ PFU of virus isolates: (a) one A/195 first-wave isolate (blue); (b) three second-wave isolates (red); (c) six third-wave isolates (green). Weights in mice inoculated with PBS are shown in black. (d and e) Virus titers (d) and interferon levels (e) in lung homogenates at day 2. (f) Mouse mortality data for virus infections where weight loss necessitated culling.

FIG 4

Cytokine induction by first- and third-wave viruses and the role of NS gene segments in virus replication in HAE cells. (a) Human nasal MucilAir cell cultures (HAE; Epithelix) were infected in triplicate with RG A/195 and A/687 virus at an MOI of 1 for 16 h. The basal medium was harvested, and levels of the cytokine IP-10 were measured using mesoscale discovery (MSD) plates. Statistical analysis was performed with an unpaired t test. (b) 293T cells transiently transfected with a beta interferon promoter luciferase reporter plasmid were infected with RG viruses A/195 (dark blue), A/195 with A/687 segment 8 (light green), A/687 (dark green), and A/687 with the A/195 segment 8 (light blue) at an MOI of 3. Infection with NDV was used as a positive control. Statistical analysis entailed a one-way ANOVA, with Tukey's multiple-comparison test (and associated adjusted P values). (c) 293T cells transiently transfected with an beta interferon promoter luciferase reporter and pCAGGs NS1 plasmids A/195 (blue), A/687 (green), and H3N2 (purple). Positive controls were stimulated with NDV. Statistical analysis was performed with a one-way ANOVA and Tukey's multiple-comparison test.

FIG 5

Variation in surface genes HA and NA in third-wave virus isolates leads to altered receptor binding, enhanced replication in HAE cells, and enhanced infectivity in mucus. (a) Hemagglutination assay with 11 clinical isolates from first-wave (blue), second-wave (red), and third-wave (green) isolates, assessed for binding to 0.5% chicken, turkey, or guinea pig red blood cells. The dashed line represents the A195 HA score. (b) Hemagglutination assay with equal PFU of A/195 first-wave (blue) and A/687 third-wave (green) RG viruses with 0.5% chicken or turkey red blood cells. (c) Viral replication in human nasal MucilAir cell cultures (HAE; Epithelix) of RG viruses based on A/195 with HA and NA from A/195 first-wave (blue) or HA and NA of A/687 third-wave (green triangle) isolates. (d) A/687 (green square) or A/687 with A/687 with HA and NA from A/195 (blue triangle). Cells were infected at an MOI of 0.01. *, P < 0.05 based on an unpaired t test. (e) Mucus inhibition assay. An equal PFU of A/195 (blue) or A/195 with A/687 third-wave isolate NA (green) RG virus was incubated with diluted human mucus for 1 h prior to infection of MDCK cells. Infectivity remaining was plotted as the percentage of the titer in the absence of mucus. *, P = 0.044 by unpaired t test.

FIG 6

PA of third-wave virus confers enhanced polymerase activity and a fitness advantage in HAE cells. (a) Activities of polymerase reconstituted from plasmids expressing polymerase components of A/195 first-wave (dark blue) and A/687 third-wave (dark green) viruses. 293T cells were transiently transfected with a plasmid that directs in situ synthesis of a minigenome in which a luciferase reporter gene is flanked by the influenza A virus promoter. Cotransfection of a Renilla expression plasmid was used to normalize for transfection efficiency. Combinations of the PB1, PB2, PA, and NP expression plasmids of A/195 (blue) or A/687 (green) or lacking the PB2 polymerase (white) were transfected. At 24 h posttransfection, cells were harvested. The results were normalized based on Renilla results (transfection control) and are from three separate set of experiments, each with triplicate wells (n = 9). Differences were analyzed using a one-way ANOVA test with Tukey's multiple-comparison test. (b) Human nasal MucilAir cell cultures (HAE; Epithelix) were infected in triplicate with RG viruses based on A/195 that differed only in PA at N321K. Virus released was harvested at 24, 48, 72, and 96 h and sequenced using the Illumina system. The percentage of N allele (first wave) is represented in blue and K (third wave) is shown in green.