Mouse Models of Influenza Infection with Circulating Strains to Test Seasonal Vaccine Efficacy

Abstract

Influenza virus infection is a significant cause of morbidity and mortality worldwide. The surface antigens of influenza virus change over time blunting both naturally acquired and vaccine induced adaptive immune protection. Viral antigenic drift is a major contributing factor to both the spread and disease burden of influenza. The aim of this study was to develop better infection models using clinically relevant, influenza strains to test vaccine induced protection. CB6F1 mice were infected with a range of influenza viruses and disease, inflammation, cell influx, and viral load were characterized after infection. Infection with circulating H1N1 and representative influenza B viruses induced a dose-dependent disease response; however, a recent seasonal H3N2 virus did not cause any disease in mice, even at high titers. Viral infection led to recoverable virus, detectable both by plaque assay and RNA quantification after infection, and increased upper airway inflammation on day 7 after infection comprised largely of CD8 T cells. Having established seasonal infection models, mice were immunized with seasonal inactivated vaccine and responses were compared to matched and mismatched challenge strains. While the H1N1 subtype strain recommended for vaccine use has remained constant in the seven seasons between 2010 and 2016, the circulating strain of H1N1 influenza (2009 pandemic subtype) has drifted both genetically and antigenically since 2009. To investigate the effect of this observed drift on vaccine induced protection, mice were immunized with antigens from A/California/7/2009 (H1N1) and challenged with H1N1 subtype viruses recovered from 2009, 2010, or 2015. Vaccination with A/California/7/2009 antigens protected against infection with either the 2009 or 2010 strains, but was less effective against the 2015 strain. This observed reduction in protection suggests that mouse models of influenza virus vaccination and infection can be used as an additional tool to predict vaccine efficacy against drift strains.

Keywords: Antibodies; Infection; Influenza Vaccines; Viral; mouse models; vaccine drift.

Figure 1

Seasonal influenza H1 and B but not seasonal H3 cause disease in mice. Mice were infected with increasing doses of different influenza viruses intranasally in 100 μl volumes. Weight loss was measured after infection with A/England/195/2009 (A), A/Puerto Rico/8/1934 (B), A/England/691/2010 (C), A/X-31 (D), B/Florida/04/06 (Yamagata) (E) or B/Brisbane/60/2008 (Victoria) (F). Points represent mean of n ≥ 4 animals ± SEM.

Figure 2

Characterization of pathogenic response to seasonal influenza infections. Mice were infected with H1N1, Flu B, or X31. Weight (A) and temperature (B) were measured daily after infection. Viral load was measured by plaque assay (C) or rt-PCR (D) on day 4 after infection. Inflammation in the upper (E) and lower (F) airways were measured on day 7 after infection. Cell numbers in the lung was assessed at day 4 were counted (G) and compared to inflammation score (H). NK (I), CD4 (J), and CD8 (K) cells in lungs assessed by flow cytometry. Points represent means of n ≥ 4 animals ± SEM (A,B) or individual animals (C–K).

Figure 3

Immunization protects against homologous challenge. CB6F1 mice were immunized once with varying doses of H1N1 antigens. Antibody was measured prior to challenge (A) with A/California/7/2009 (H1N1pdm09), weight loss measured daily (B). Mice were immunized twice with varying doses of H1N1 antigens. Antibody was measured prior to challenge (C) with A/California/7/2009 (H1N1pdm09), weight loss measured daily (D). Points represent individual animals (A,C) or means of n ≥ 4 animals ± SEM (B,D). Panel (B) *p < 0.05 between 1.5 µg and naïve group and #p < 0.05 between 1.5 µg and 0.02 µg group. Panel (C) **p < 0.01.

Figure 4

Genetic and antigenic drift of H1N1pdm09 strains between 2009 and 2016. Integrated phylogenetic and antigenic cartography analysis of 61 viruses using Bayesian multidimensional scaling (14). (A) Time-resolved phylogenetic tree of viruses circulating in England (black circles) and WHO reference viruses (red circles) with branches colored by inferred antigenic difference from A/California/07/2009 vaccine strain in antigenic map units. Study viruses are shown with text and vaccine virus names in bold. (B) Two-dimensional antigenic map showing antigenic (circles, colored by virus collection date) and phylogenetic relationships (lines) of viruses. Distances for antigenic dimensions are measured in antigenic units, where one unit represents a twofold dilution in heterologous hemagglutination inhibition titer. Study viruses (gray outline) and vaccine viruses (red outline) are highlighted.

Figure 5

Antigenic drift in H1N1 strains is seen in mouse models. Mice were immunized with A/Cal/7/2009 (black squares) and challenged with A/England/195/2009 (A), A/Eng/672/2010 (B), A/England/336/2015 (C), responses were compared to naïve animals (white circles). Antibody response to Cal/09 antigen prior to challenge (D). Points mean of n = 4 mice ± SEM (A–C), or individual animals (D). *p < 0.05, **p < 0.01.