Natural T Cell-mediated Protection against Seasonal and Pandemic Influenza. Results of the Flu Watch Cohort Study

Abstract

Rationale: A high proportion of influenza infections are asymptomatic. Animal and human challenge studies and observational studies suggest T cells protect against disease among those infected, but the impact of T-cell immunity at the population level is unknown.

Objectives: To investigate whether naturally preexisting T-cell responses targeting highly conserved internal influenza proteins could provide cross-protective immunity against pandemic and seasonal influenza.

Methods: We quantified influenza A(H3N2) virus-specific T cells in a population cohort during seasonal and pandemic periods between 2006 and 2010. Follow-up included paired serology, symptom reporting, and polymerase chain reaction (PCR) investigation of symptomatic cases.

Measurements and main results: A total of 1,414 unvaccinated individuals had baseline T-cell measurements (1,703 participant observation sets). T-cell responses to A(H3N2) virus nucleoprotein (NP) dominated and strongly cross-reacted with A(H1N1)pdm09 NP (P < 0.001) in participants lacking antibody to A(H1N1)pdm09. Comparison of paired preseason and post-season sera (1,431 sets) showed 205 (14%) had evidence of infection based on fourfold influenza antibody titer rises. The presence of NP-specific T cells before exposure to virus correlated with less symptomatic, PCR-positive influenza A (overall adjusted odds ratio, 0.27; 95% confidence interval, 0.11-0.68; P = 0.005, during pandemic [P = 0.047] and seasonal [P = 0.049] periods). Protection was independent of baseline antibodies. Influenza-specific T-cell responses were detected in 43%, indicating a substantial population impact.

Conclusions: Naturally occurring cross-protective T-cell immunity protects against symptomatic PCR-confirmed disease in those with evidence of infection and helps to explain why many infections do not cause symptoms. Vaccines stimulating T cells may provide important cross-protective immunity.

Keywords: T lymphocytes; cellular immunity; cohort studies.

Figure 1.

Preexisting ex vivo influenza-specific T-cell response. The preseason ex vivo frequency of influenza (H3N2)-specific T-cell responses from 1,703 baseline measurements from 1,414 Flu Watch participants were quantified by IFN-γ enzyme-linked immunospot assay. Some participants contributed to more than one season, but no participant had more than one baseline measurement per season. (A) Each column represents a different baseline sample. The height of each column represents the ex vivo frequency (spot-forming units per million [SFU/M] peripheral blood mononuclear cells [PBMC]) of H3N2-specific T-cell response; each color represents the T-cell responses targeted at each H3N2 viral antigen as indicated below the bar chart. (B) For the same samples as in A, the ex vivo nucleoprotein (NP)-specific T-cell responses (y-axis) are plotted on a log10 scale and ordered on the x-axis by the strength of the response. (C) For the total ex vivo T-cell responses in the whole study, the proportional contribution of each H3N2 viral antigen specificity is shown, indicating for instance that NP-specific T cells constitute 25% of the whole response. (D) Percentage of Flu Watch baseline samples that had ex vivo detectable memory T-cell responses. For each influenza protein response, we subtracted the negative control well result from the test well result and accepted values of greater than or equal to 20 SFU/M PBMC as reactors if the original result was also significantly higher than that of the negative control wells (P < 0.05 based on the negative binomial distribution). HA = hemagglutinin; M = matrix; NA = neuraminidase; NS = nonstructural; PA = polymerase A; PB1 = polymerase B1; PB2 = polymerase B2.

Figure 2.

Preexisting influenza nucleoprotein (NP)- and matrix (M)-specific T cells cross-react to pandemic H1N1 2009 (H1N1pdm2009). The frequencies of T-cell responses to pandemic H1N1 2009 NP and M peptides were quantified ex vivo from Flu Watch participants (n = 222) who had no detectable antibody against pandemic H1N1 2009 HA at baseline. Each dot represents one Flu Watch participant. (A) The log10-transformed ex vivo frequency of H3N2 NP-specific T-cell response is shown on the x-axis, and the log10-transformed ex vivo frequency of the T cells targeted at pandemic H1N1 2009 NP is shown on the y-axis. The NP T-cell responses were quantified by IFN-γ enzyme-linked immunospot assay, and backgrounds were subtracted in the data presented in this figure. Linear regression line coefficient = 0.79 (95% confidence interval, 0.77–0.81); P < 0.001. (B) The equivalent for M responses. Linear regression line coefficient = 0.68 (95% confidence interval, 0.66–0.70); P < 0.0001. PBMC = peripheral blood mononuclear cells; SFU/M = spot-forming units per million.

Figure 3.

CD4 and CD8 phenotype of preexisting influenza-specific T cells. The preseason influenza (H3N2)-specific T-cell responses from 174 randomly selected Flu Watch participants were expanded by culture with peptide and IL-2, and the frequency of CD4⁺ and CD8⁺ influenza-specific T cells were measured by intracellular IFN-γ staining (y-axis). The height of each column indicates the percentage of all influenza-specific T cells (CD3⁺ IFN-γ⁺) responding to H3N2 virus peptides representing each protein. The gray and red indicate the relative CD4⁺ and CD8⁺ T-cell responses. HA = hemagglutinin; M = matrix; NA = neuraminidase; NP = nucleoprotein; NS = nonstructural; PA = polymerase A; PB1 = polymerase B1; PB2 = polymerase B2.