1918 Influenza virus hemagglutinin (HA) and the viral RNA polymerase complex enhance viral pathogenicity, but only HA induces aberrant host responses in mice

Abstract

The 1918 pandemic influenza virus was the most devastating infectious agent in human history, causing fatal pneumonia and an estimated 20 to 50 million deaths worldwide. Previous studies indicated a prominent role of the hemagglutinin (HA) gene in efficient replication and high virulence of the 1918 virus in mice. It is, however, still unclear whether the high replication ability or the 1918 influenza virus HA gene is required for 1918 virus to exhibit high virulence in mice. Here, we examined the biological properties of reassortant viruses between the 1918 virus and a contemporary human H1N1 virus (A/Kawasaki/173/2001 [K173]) in a mouse model. In addition to the 1918 influenza virus HA, we demonstrated the role of the viral RNA replication complex in efficient replication of viruses in mouse lungs, whereas only the HA gene is responsible for lethality in mice. Global gene expression profiling of infected mouse lungs revealed that the 1918 influenza virus HA was sufficient to induce transcriptional changes similar to those induced by the 1918 virus, despite difference in lymphocyte gene expression. Increased expression of genes associated with the acute-phase response and the protein ubiquitination pathway were enriched during infections with the 1918 and 1918HA/K173 viruses, whereas reassortant viruses bearing the 1918 viral RNA polymerase complex induced transcriptional changes similar to those seen with the K173 virus. Taken together, these data suggest that HA and the viral RNA polymerase complex are critical determinants of Spanish influenza pathogenesis, but only HA, and not the viral RNA polymerase complex and NP, is responsible for extreme host responses observed in mice infected with the 1918 influenza virus.

Fig 1

Pathogenicity of reassortant viruses in mice. Mice were intranasally infected with 10⁶ PFU of each reassortant, 1918 virus, or K173 virus. (A) Lung tissues were collected from mice (n = 3) on the indicated days postinfection, and titers were determined in MDCK cells. Bars show the mean titers in lungs of the infected mice for the respective virus. Asterisks indicate that the mean titers of the respective virus were significantly lower than those of 1918 virus (*, P < 0.001; **, P < 0.01). (B) The body weights of the infected mice were monitored for 14 days. Mean changes in body weight are shown.

Fig 2

Pathological examination of lungs of mice infected with the indicated reassortant viruses at day 5 postinfection. Representative findings are shown to depict the distribution of lesions in the lung tissue sections (shown as sections placed next to the illustrations of lung), as follows: gray, inflammatory lesions containing viral-antigen-positive cells; violet, accumulation of neutrophils. The left and middle panels show lung sections with H&E staining, whereas the right panels show immunostaining with a rabbit anti-H1N1 influenza virus polyclonal antibody (anti-A/WSN/33) that was characterized previously (9). Global severe inflammatory reactions were prominent in the lungs of mice infected with 1918 or 1918HA/K173 virus. Viral antigens were distributed throughout the lung of mice infected with those viruses. Numerous neutrophilic infiltrations were prominent in the peribronchial area and alveolar wall. Hemorrhage and edematous lesions were observed (B and C, left and middle). In contrast, infection with K173, 1918PB1/K173, or 1918(3P+NP)/K173 virus produced only small, irregularly distributed inflammatory lesions in the mouse lungs (A, D, and E, left and middle). Lungs from the mice infected with 1918, 1918HA/K173, and 1918(3P+NP)/K173 viruses had viral antigen commonly found in bronchial lumina and alveoli (B, C, and E, right). Viral antigen in necrotic cellular debris was observed in the lungs of mice infected with 1918HA/K173 virus (C, right).

Fig 3

Global analysis of DE genes distinguishing high-pathogenicity and low-pathogenicity virus groups from mock infections. (A) Venn diagram representing the overlap of DE genes in a direct comparison between high-pathogenicity, low-pathogenicity, and mock infections at early (day 1 p.i.) and late (days 3 and 5 p.i.) time points. (B) SVD-MDS representation of the samples on correspondence space with the gene signature applied (6,197 DE genes differentiating low-pathogenicity and high-pathogenicity virus groups from mock infection in at least one group). Each data point represents the RNA expression profile of a single mouse. The convex hull for each set of biological replicates per virus group is shown. The quality of the representation is provided by the Kruskal stress criteria. The x and y axes represent the relative distance in correspondence space and therefore are unit-free. They correspond to the two first-principle dimensions of the singular value decomposition analysis, which is also the underlying mathematical operation in principle component analysis (PCA). (C) Hierarchical clustering of the signature gene intensities using correspondence metrics. D, day; A, animal; S, signal intensity data.

Fig 4

1918 and 1918HA/K173 viruses differentially regulate lymphocyte gene expression in lungs of the virus-infected mice. (A) The heatmap shows a total of 118 DE genes that were available for analysis in IPA. Average log₂(ratio) gene expression in infected lungs relative to that in mock-infected lungs is shown. Saturation is 3-fold. Red indicates that the gene expression is higher than that in the uninfected reference; green indicates that gene expression is lower than that in the uninfected reference. LPV, low-pathogenicity virus group; HPV, high-pathogenicity virus group. (B) Pathway analysis diagram showing the functional relationships among DE genes differentially regulated between 1918 and 1918HA/K173 viruses. Average lung gene expression for 1918 virus relative to 1918HA/K173 virus is overlaid onto the network. Nodes shaded red indicates that 1918 virus-induced gene expression is higher than the gene expression induced by the 1918HA/K173 reference; green indicates that 1918 virus-induced gene expression is lower than that induced by the 1918HA/K173 reference. R, log ratio data.

Fig 5

Mice show differential expression of genes associated with acute-phase responses during infection. (A) Differentially expressed IL-6 signaling pathway molecules. Molecules shaded in red were differentially upregulated between low-pathogenicity and high-pathogenicity virus groups. Red indicates upregulation relative to mock infection. Open circles signify group complexes. Transcriptional target genes encoding plasma molecules are indicated with diamonds. Genes depicting cellular pathway molecules or transcriptional target genes are depicted by circles. (B) Transcriptional changes of 134 inflammatory response genes in response to infection. Average log₂(ratio) gene expression of infected lung relative to mock is shown. Saturation is 3-fold. Red indicates that the gene expression is higher than that in the uninfected reference; green indicates that gene expression is lower than that in the uninfected reference. LPV, low-pathogenicity virus group; HPV, high-pathogenicity virus group; R, log ratio data.

Fig 6

The protein ubiquitination pathway is similarly activated by 1918 and 1918HA/K173. (A) Ingenuity pathway analysis (IPA) of the 56 DE genes. Direct functional relationships between pathway nodes were examined, and orphan molecules were eliminated. The remaining 52 DE genes are shown in the diagram with an overlay of the average log₂(ratio) gene expression of the high-pathogenicity virus group relative to that seen in mock infection. Red signifies upregulation in the high-pathogenicity virus group relative to mock infection, and green signifies downregulation. Genes outlined in blue exhibited increased expression in 1918(3P+NP)/K173-infected lungs relative to K173-infected lungs. (B) Heatmap of average log₂(ratio) gene expression of infected lung relative to mock infection. Saturation is 3-fold. Red indicates that the gene expression is higher than in the mock-infected reference; green indicates that gene expression is lower than in the mock-infected reference. LPV, low-pathogenicity virus group; HPV, high-pathogenicity virus group; R, log ratio data.