Humoral and cell-mediated immunity to pandemic H1N1 influenza in a Canadian cohort one year post-pandemic: implications for vaccination

Abstract

We evaluated a cohort of Canadian donors for T cell and antibody responses against influenza A/California/7/2009 (pH1N1) at 8-10 months after the 2nd pandemic wave by flow cytometry and microneutralization assays. Memory CD8 T cell responses to pH1N1 were detectable in 58% (61/105) of donors. These responses were largely due to cross-reactive CD8 T cell epitopes as, for those donors tested, similar recall responses were obtained to A/California 2009 and A/PR8 1934 H1N1 Hviruses. Longitudinal analysis of a single infected individual showed only a small and transient increase in neutralizing antibody levels, but a robust CD8 T cell response that rose rapidly post symptom onset, peaking at 3 weeks, followed by a gradual decline to the baseline levels seen in a seroprevalence cohort post-pandemic. The magnitude of the influenza-specific CD8 T cell memory response at one year post-pandemic was similar in cases and controls as well as in vaccinated and unvaccinated donors, suggesting that any T cell boosting from infection was transient. Pandemic H1-specific antibodies were only detectable in approximately half of vaccinated donors. However, those who were vaccinated within a few months following infection had the highest persisting antibody titers, suggesting that vaccination shortly after influenza infection can boost or sustain antibody levels. For the most part the circulating influenza-specific T cell and serum antibody levels in the population at one year post-pandemic were not different between cases and controls, suggesting that natural infection does not lead to higher long term T cell and antibody responses in donors with pre-existing immunity to influenza. However, based on the responses of one longitudinal donor, it is possible for a small population of pre-existing cross-reactive memory CD8 T cells to expand rapidly following infection and this response may aid in viral clearance and contribute to a lessening of disease severity.

Figure 1. Detection of influenza-responsive CD8 T cells by multicolour flow cytometry.

Total PBMC were stimulated for 18 hours with pH1N1 influenza, or as a control, with LCMV Armstrong, or left unstimulated and then assessed for IFNγ production by intracellular cytokine staining and flow cytometry. Gates are based on fluorescence minus one controls. (A) Representative gating used to identify IFNγ⁺ CD8 T cells from total PBMC. (B) Sample non-responder, weak responder, and strong responder to pH1N1 identified in the Toronto cohort 8-10 months post-pandemic; positive versus non-responder is defined in the results. A representative “weak” responder was arbitrarily chosen from the bottom third of positive responses whereas the “strong” responder was from the top third of responders.

Figure 2. T cell analysis in the Toronto seroprevalence and case/control cohorts.

(A) Bin separation of IFNγ responses in CD8 and CD4 T cells specific to pH1N1 stimulation. Frequencies have been corrected for background IFNγ production in LCMV and unstimulated control cultures. (B) Spearman correlation between pH1N1-responding CD8 T cells and donor age. (C) Combinations of effector molecule expression of IFNγ⁺ CD8 T cells from the responder subset. P values above the bars indicate the level of statistical significance compared to all other bars as determined by ANOVA and Tukey test. (D) Spearman correlation between the CD8 T cell response to pH1N1 and the frequency of responding cells with multiple effector functions. (E) CD8 T cell response in case and control subjects. Groups were compared using a nonparametic Mann-Whitney test. (F) Spearman correlation for pH1N1 response and frequency of CD8 T cells with multiple effector functions in cases and controls.

Figure 3. Acute and persisting antibody and memory T cell responses to pandemic H1N1 infection in one PCR case-confirmed donor.

Longitudinal samples of unfractionated PBMC were challenged with influenza virus or controls for 18h. (A) Frequency of pandemic H1N1-responsive CD8 T cells out of total CD8 T cells as measured by IFNγ staining. IFNγ responsive CD8 T cells were also sub-divided by expression of other effector markers, granzyme B and CD107a. (B) Memory phenotypes of influenza-responsive CD8 T cells at various times post-onset of influenza symptoms. (C) Frequency and phenotypes of IFNγ⁺ CD4 T cells after pandemic H1N1 challenge. (D) Antibody titers in serum as detected by microneutralization (MN), hemagglutination inhibition (HAI), and a pandemic H1-specific ELISA assay. BLD  =  below the limits of detection.

Figure 4. CD8 T cell IFNγ responses to pH1N1 may be cross-reactive with other influenza strains.

IFNγ response to A/California/7/2009 and A/PR8 in a set of 25 donors. A paired t test was used to compare the differentially stimulated cultures on a per-donor basis.

Figure 5. Infection followed by vaccination boosts antibody but not T cell responses to pandemic H1N1.

(A) Antibody titers against pH1N1 for vaccinated and unvaccinated donors in the entire cohort 8-10 months post-pandemic. Vaccinations were self-reported from October 2009 to January 2010. A non-parametric Mann-Whitney test was used for statistical significance. (B) CD8 and CD4 responses to pH1N1 for vaccinated and unvaccinated donors in the total Toronto cohort, measured 8-10 months post-pandemic. Groups were compared using a Mann-Whitney test. (C) IFNγ⁺ CD8 T cell responses in donors with both antibody and CD8 T cell responses, T cell responses only, antibodies only, or no antibody or T cell response to pH1N1. Data has been normalized using log transformation to represent Gaussian distribution; groups were compared using ANOVA and Tukey test. (D) Normalized CD8 T cell response in cases and controls with differing vaccination history for pH1N1. Groups were compared by ANOVA and Tukey test. PCR-confirmed infections were reported from April-November 2009; vaccination was self-reported from October 2009-January 2010. (E) Pandemic-specific antibody responses as measured by microneutralization in the case/control cohort, separated by self-reported vaccination history for the monovalent pH1N1 vaccine. PCR-confirmed infections were reported from April-November 2009; vaccination was self-reported from October 2009-January 2010. Nonparametric Kruskal-Wallis and Mann-Whitney tests were performed to determine statistical significance.