Heterosubtypic cross-protection correlates with cross-reactive interferon-gamma-secreting lymphocytes in the ferret model of influenza

Abstract

An effective universal vaccine for influenza will likely need to induce virus-specific T-cells, which are the major mediator of heterosubtypic cross-protection between different subtypes of influenza A virus. In this study we characterise the cell-mediated immune response in ferrets during heterosubtypic protection induced by low-dose H1N1 virus infection against an H3N2 virus challenge, given 4 weeks later. Although the ferrets were not protected against the infection by H3N2 virus, the duration of virus shedding was shortened, and clinical disease was markedly reduced. No cross-reactive neutralizing antibodies were detected, but cross-reactive interferon-gamma-secreting T cells were detected in the circulation prior to H3N2 challenge. These T-cells peaked at 11 days post-H1N1 infection, and were strongly induced in blood and in lung following H3N2 infection. The rapid induction of interferon-gamma-secreting cells in ferrets previously infected with H1N1 virus, but not in naïve ferrets, suggests induction of memory T-cells. These results are in accord with the observations that pre-existing cross-reactive T-cells correlate with protection in humans and have implications for outbreak modelling and universal vaccine design.

Figure 1

Study design. Each group comprised 6 ferrets. Groups H1/H3, H1/H1 and H1/cull received identical doses of 100 pfu H1N1 virus on day 0. Groups H1/H3 and PBS/H3 received identical doses of 100 pfu H3N2 virus on day 28, whereas group H1/H1 received a second 100 pfu dose of H1N1 virus (homologous challenge). Nasal washes, heparinised blood samples and sera were collected from all animals at time-points specified in Methods. Red triangles indicate collection of PBMC and lung lymphocyte samples. i.n., intra-nasal inoculation.

Figure 2

Signs of clinical disease following second virus challenge. Groups H1/H3 and PBS/H3 were challenged with H3N2 virus, group H1/H1 with H1N1 virus. (a) frequency of sneezing following challenge, sum for each group; (b) AUC for temperature between 50 and 82 hours post-challenge; (c) AUC for weight between days 0 and 14 post-challenge. Weights were normalised to weight on day of challenge. In panels b and c, points represent individual ferrets and bars represent group means and standard deviations. Groups were compared by 1-way ANOVA with Holm-Sidak’s multiple comparisons test. *p < 0.05; ***p < 0.001.

Figure 3

Virus titres in nasal wash following second virus challenge. Points show log mean with standard deviation for each group. Time 0 samples were collected 2 days prior to challenge.

Figure 4

Nasal wash live cell counts. Lines show mean and standard deviation for each group. Ferrets were infected on days 0 and 28 (arrows).

Figure 5

Antibody response to initial infection with H1N1 virus (or PBS control). (a) geometric mean HAI titres, (b) geometric mean neutralization titres. Limit of detection in a is 4, values < 4 were scored as 2. Limit of detection in b is 10, values < 10 were scored as 5. Error bars show standard deviation. Solid bars show response to H1N1 virus, hatched bars show response to H3N2 virus. Sera were collected 24 days post-infection.

Figure 6

ELISpot titres in PBMC. (a) stimulation with H1N1 virus, (b) stimulation with H3N2 virus. Points show individual ferrets, with mean and standard deviation overlaid. Groups were compared by Kruskall-Wallis test, using Dunn’s multiple comparisons test. *p < 0.05. The values marked with a circle in group PBS/H3 are from the same ferret.

Figure 7

ELISpot titres in lung MNC. (a) stimulation with H1N1 virus, (b) stimulation with H3N2 virus. Points show individual ferrets, with mean and standard deviation overlaid. Groups were compared by Kruskall-Wallis test, using Dunn’s multiple comparisons test. *p < 0.05, **p < 0.01. The values marked with a circle in group PBS/H3 are from the same ferret as highlighted in Fig. 6.

Figure 8

Whole blood IFN-γ ELISA. Bars show group mean values in pg/ml following stimulation with (a) H1N1 virus and (b) H3N2 virus. Arrows indicate the timing of the first and second infections, at days 0 and 28. Error bars show standard error of the mean. Upper limit of detection was 12000 pg/ml.