Repeated Low-Dose Influenza Virus Infection Causes Severe Disease in Mice: a Model for Vaccine Evaluation

Abstract

Influenza infection causes severe disease and death in humans. In traditional vaccine research and development, a single high-dose virus challenge of animals is used to evaluate vaccine efficacy. This type of challenge model may have limitations. In the present study, we developed a novel challenge model by infecting mice repeatedly in short intervals with low doses of influenza A virus. Our results show that compared to a single high-dose infection, mice that received repeated low-dose challenges showed earlier morbidity and mortality and more severe disease. They developed higher vial loads, more severe lung pathology, and greater inflammatory responses and generated only limited influenza A virus-specific B and T cell responses. A commercial trivalent influenza vaccine protected mice against a single high and lethal dose of influenza A virus but was ineffective against repeated low-dose virus challenges. Overall, our data show that the repeated low-dose influenza A virus infection mouse model is more stringent and may thus be more suitable to select for highly efficacious influenza vaccines.

Importance: Influenza epidemics and pandemics pose serious threats to public health. Animal models are crucial for evaluating the efficacy of influenza vaccines. Traditional models based on a single high-dose virus challenge may have limitations. Here, we describe a new mouse model based on repeated low-dose influenza A virus challenges given within a short period. Repeated low-dose challenges caused more severe disease in mice, associated with higher viral loads and increased lung inflammation and reduced influenza A virus-specific B and T cell responses. A commercial influenza vaccine that was shown to protect mice from high-dose challenge was ineffective against repeated low-dose challenges. Overall, our results show that the low-dose repeated-challenge model is more stringent and may therefore be better suited for preclinical vaccine efficacy studies.

FIG 1

Repeated low-dose influenza virus infection causes earlier weight loss and mortality. (A) Virus infection experiment schedule. C57BL/6 mice were intranasally infected with 0.5, 1.0, and 2.0 LD₅₀ of A/PR/8 virus once or three times or with 10.0 LD₅₀ of virus once. Control mice were inoculated with PBS three times. (B to D) Weight loss of mice infected with different doses of A/PR/8 virus once or repeatedly. The weight of each mouse was monitored daily until day 21 postinfection. Data are shown as means ± standard errors of the means. (B) Weight loss of the 0.5 LD₅₀ single- and repeated-infection groups, the 10.0 LD₅₀ single-infection group, and the PBS group. (C) Weight loss of the 1.0 LD₅₀ single- and repeated-infection groups, the 10.0 LD₅₀ single-infection group, and the PBS group. (D) Weight loss of the 2.0 LD₅₀ single- and repeated-infection groups, the 10.0 LD₅₀ single-infection group, and the PBS group. (E) All infected mice were monitored daily for survival until day 14 postinfection. Each graph represents the combined results from 3 separate experiments, each with 6 mice per group per experiment.

FIG 2

Mice with repeated low-dose influenza virus infection have higher virus titers in lung tissues and more severe lung pathology. (A) Five days after infection, lung virus titers were determined by quantitative PCR. The graph shows titers of viral genomes (copies) per microgram of total RNA in lung tissue of individual mice (n = 6 to 8) in each group, with mean values ± standard errors of the means. (B) Viral titers lung tissues at days 3 and 5 after the first infection were determined by a TCID₅₀ assay. Data are shown as mean values ± standard errors of the means. dpi, days postinfection. (C) Correlation between viral genome copy numbers and TCID₅₀ titers in lung tissues on day 5 after the first infection. (D) Histological scoring for virus-infected mice (n = 6 to 8). Mean values for each group are shown. ***, P < 0.0001; n.s., no significance. (E) Representative hematoxylin- and eosin-stained sections were derived from mice (n = 6 to 8) from each group killed on day 5 after the first infection. ×1, mice were infected once; ×3, mice were infected three times. Original magnification, ×100.

FIG 3

Antibody responses in mice after virus infection. Mice (n = 8 to 10) from each group were bled at day 7 after the first infection, and sera were collected to determine antibody responses by a microneutralization assay. Graphs show titers of individual mice in each group, with mean values indicated by lines. Sera from PBS-inoculated mice served as negative controls. ***, P < 0.0001; n.s., no significance.

FIG 4

Virus-specific CD8⁺ T cell responses in mice after virus infection. Seven days after the first infection, PBMCs were isolated, and tetramer staining was performed to analyze the CD8⁺ T cell responses to the immunodominant epitope of NP in mice (n = 8 to 10). PBS-inoculated mice (n = 8) served as negative controls. tet⁺, tetramer positive.

FIG 5

Repeated low-dose infection elicits greater inflammasome responses in mice. Mice (n = 8 to 10) were killed on day 5 after the first infection. Lung homogenates and BALF were collected for inflammasome response analysis using an ELISA. (A) IL-18 levels in lung homogenates. (B) IL-18 levels in BALF. (C) IL-1β expression in lung homogenates. (D) IL-1β concentrations in BALF. Data are presented as means ± standard errors of the means. ***, P < 0.0001; n.s., no significance.

FIG 6

TIV cannot protect mice from repeated low-dose infection with influenza virus. (A) Immunization and challenge schedule. Mice (n = 12 to 13) were immunized intramuscularly with an inactivated split influenza vaccine (Fluarix; GSK) twice at a 2-week interval. Three weeks after the boost, mice were challenged with repeated low-dose infection and single high-dose infection with pandemic H1N1 virus. PBS-injected mice (n = 10) were used as controls. (B) Antibody titers of postprime and postboost sera were measured by an ELISA with serial serum dilutions. Sera of the PBS group served as negative controls. Data are presented as means ± standard errors of the means. OD450, optical density at 450 nm. (C) Sera of all mice from the vaccine groups and the control group were collected 2 weeks after the prime and boost. Neutralizing antibody titers in MDCK cells were determined by a microneutralization assay. Data are presented as means ± standard errors of the means. (D) Weight loss of all mice was monitored every day for 21 days. Data are presented as means ± standard errors of the means. (E) Survival of all mice was monitored for 2 weeks after challenge. ***, P < 0.0001; n.s., no significance.

FIG 7

TIV-immunized mice with repeated virus challenge have higher viral titers and more pathogenic damage in the lungs. (A) Viral titers were determined by a TCID₅₀ assay of lung tissues from immunized mice and controls at days 3 and 5 after the first challenge. Data are presented as mean values ± standard errors of the means. dpi, days postinfection. (B) Mouse lung tissue damage was analyzed and scored at day 5 after the first challenge. Data are shown as mean values ± standard errors of the means. ***, P < 0.0001; n.s., no significance. (C) Representative hematoxylin- and eosin-stained sections from each group of mice scarified on day 5 after the first infection. Original magnification, ×100.