Early priming minimizes the age-related immune compromise of CD8⁺ T cell diversity and function

Abstract

The elderly are particularly susceptible to influenza A virus infections, with increased occurrence, disease severity and reduced vaccine efficacy attributed to declining immunity. Experimentally, the age-dependent decline in influenza-specific CD8(+) T cell responsiveness reflects both functional compromise and the emergence of 'repertoire holes' arising from the loss of low frequency clonotypes. In this study, we asked whether early priming limits the time-related attrition of immune competence. Though primary responses in aged mice were compromised, animals vaccinated at 6 weeks then challenged >20 months later had T-cell responses that were normal in magnitude. Both functional quality and the persistence of 'preferred' TCR clonotypes that expand in a characteristic immunodominance hierarchy were maintained following early priming. Similar to the early priming, vaccination at 22 months followed by challenge retained a response magnitude equivalent to young mice. However, late priming resulted in reduced TCRβ diversity in comparison with vaccination earlier in life. Thus, early priming was critical to maintaining individual and population-wide TCRβ diversity. In summary, early exposure leads to the long-term maintenance of memory T cells and thus preserves optimal, influenza-specific CD8(+) T-cell responsiveness and protects against the age-related attrition of naïve T-cell precursors. Our study supports development of vaccines that prime CD8(+) T-cells early in life to elicit the broadest possible spectrum of CD8(+) T-cell memory and preserve the magnitude, functionality and TCR usage of responding populations. In addition, our study provides the most comprehensive analysis of the aged (primary, secondary primed-early and secondary primed-late) TCR repertoires published to date.

Figure 1. Effect of age and early priming on 1⁰ and 2⁰ CD8⁺ T cell responses.

(A) For the primary responses, naïve mice were infected i.n. with 1×10⁴ pfu of the HK (H3N2) influenza A virus either at a young (<3 months; mo) or extreme (22 mo) age. Analysis of CD8⁺ T cell responses was performed on d10 after the primary infection. (B) For the secondary responses of the early-primed mice, animals were primed at <2 mo i.p. with 1.5×10⁷ pfu the PR8 (H1N1) influenza A virus, then challenged 6 weeks (young) or >22 mo (aged) later i.n. with 1×10⁴ pfu of the HK virus. Analysis of CD8⁺ T cell responses was performed on d8 after the secondary infection. (C, D) Numbers of epitope-specific CD8⁺ T cells in the spleens recovered from young (filled symbols) or aged (>22 month, open symbols) B6 mice on d10 (1⁰, C) or d8 (2^o, D) following primary (1⁰) or secondary (primed young) (2⁰) i.n. infection with the HK (H3N2) influenza A virus. Memory mice had been injected i.p. with the PR8 (H1N1) influenza A virus at <2 mo and were challenged. Lymphocyte populations were stimulated with the NP₃₆₆ or PA₂₂₄ peptides in the presence of Brefeldin A for 5 hrs, then stained with the anti-CD8PerCPCy5.5 mAb, fixed/permeabilised and stained with anti-IFN-γ-FITC mAb. Cytokine (IFN-γ) production was calculated by subtracting background fluorescence for the no-peptide controls, and the numbers of IFN-γ⁺CD8⁺ D^bNP₃₆₆- and D^bPA₂₂₄-specifc CD8⁺ T cells were determined from the % cells staining and the total cell counts. Data represent individual mice (symbols) and the mean (line). Experiments were performed at least twice. * = p<0.05.

Figure 2. Immunodominance hierarchies in aged mice after 1⁰ infection or 2⁰ challenge of primed-early mice.

The relative prevalence of the immunodominant D^bNP₃₆₆ ⁺CD8⁺ and D^bPA₂₂₄ ⁺CD8⁺ T cell population over the subdominant D^bPB1₇₀₃ ⁺CD8⁺ and K^bPB1-F2₆₂ ⁺CD8⁺ sets. Results are shown for (A, B) 1⁰ and (C, D) 2⁰ HK infection in (A, C) young and (B, D) aged mice. The relative contributions of particular antigen-specific CD8⁺ T cells were analysed based on total cell responses (Figure 1 for D^bNP₃₆₆ ⁺CD8⁺ and D^bPA₂₂₄ ⁺CD8⁺ and data not shown for D^bPB1₇₀₃ ⁺CD8⁺ and K^bPB1-F2₆₂ ⁺CD8⁺). Data represent the mean proportion of a particular peptide-specific CD8⁺ population. * = p<0.01 shows a difference between young and aged animals. Experimental outline as in Figure 1AB.

Figure 3. Cytokine polyfunctionality following 1⁰ or 2⁰ challenge.

Epitope-specific CD8⁺ T cells generated following 1⁰ (A, B) or 2⁰ (C, D) i.n HK challenge (see legend to Figure 1) of young (black bar) and aged (white bar) mice were assessed for the simultaneous production of IFN-γ, TNF-α (A, C) and IL-2 (B, D) using the ICS assay. The % values (A–F) were compared for spleens from groups of 3–5 mice and representative dot plots are shown (E, F). * = p<0.05. Experimental outline as in Figure 1AB.

Figure 4. Impaired polyfunctionality of D^bPA₂₂₄-specific CD8⁺ T cells in the aged mice during primary but not secondary influenza infection.

(A) Primary or (B) secondary (primed young) influenza-specific CD8⁺ T cell responses were assessed for simultaneous production of IFN-γ, TNF-α and IL-2 in the spleen of aged (22 months old) and young (6–8 weeks) mice. Compiled data (n = 3–5, mean±SD) are shown for the mean fluorescence intensity (MFI) of IFN-γ, IFN-γ and TNF-α as well as IFN-γ and IL-2 staining. * = p<0.05. Experimental outline as in Figure 1AB.

Figure 5. Profiles of Vβ usage for tetramer⁺ CD8⁺ T cells.

Profiles of TCR Vβ usage are shown for d10 (1⁰, A–D) or d8 (2⁰, early priming EF) CD8⁺ T cells from young (AB) or aged (C–F) mice. The splenocytes were stained with D^bNP₃₆₆ (ACE) and D^bPA₂₂₄ (BDF) PE tetramers, anti-CD8-APC and a panel of anti-Vβ mAbs conjugated with FITC. Results represent individual mice of 4 per group. Experimental outline as in Figure 1AB.

Figure 6. Priming at an extreme age leads to normal secondary influenza-specific CD8⁺ T cell responses.

(A) For the secondary responses of the old-primed mice, naïve B6 mice were i.p. primed with 1.5×10⁷ pfu of the PR8 virus either at 6 weeks of age (young mice) or at 22 months (primed late aged mice), followed by a secondary i.n. challenge with 1×10⁴ pfu of the HK influenza strain 6 weeks later. (B) The magnitude of CD8⁺ T cell responses in the spleen at the peak (d8) of secondary phase following influenza virus infection are shown for young (6–8 weeks) and aged (22 months old) B6 mice. Immunodominant D^bNP₃₆₆ ⁺ and D^bPA₂₂₄ ⁺ influenza-specific CD8⁺ T cell responses were assessed by IFN-γ production in an ex vivo ICS assay. (C, D) Polyfunctionality of influenza-specific CD8⁺ T cell responses was assessed by simultaneous production of IFN-γ, TNF-α and IL-2 in the spleen and of young and aged mice. (E) The contribution of immunodominant D^bNP₃₆₆ ⁺CD8⁺ and D^bPA₂₂₄ ⁺CD8⁺ T cell responses in comparison to subdominant D^bPB1₇₀₃ ⁺CD8⁺ and K^bPB1-F2₆₂ ⁺CD8⁺ sets was calculated based on the proportions of IFN-γ⁺CD8⁺ populations depicted in (B for D^bNP₃₆₆ ⁺CD8⁺ and D^bPA₂₂₄ ⁺CD8⁺ and data not shown for D^bPB1₇₀₃ ⁺CD8⁺ and K^bPB1-F2₆₂ ⁺CD8⁺). TCR Vβ usage for the (F) D^bNP₃₆₆ and (G) D^bPA₂₂₄ CD8⁺ sets in the spleen of recall responses of mice primed late. TCR Vβ results represent individual mice of 3 per group. * = p<0.05.

Figure 7. Comparison of TCRβ diversity, and inter-individual sharing and similarity for the D^bNP₃₆₆ ⁺Vβ8.3⁺CD8⁺ and D^bPA₂₂₄ ⁺Vβ7⁺CD8⁺ repertoires.

Shown are the relative measures of TCRβ repertoire diversity, (A–D) the number of different clonotypes and (E–H) Simpson's diversity index, and (I–L) % of repertoire comprised of shared clonotypes, and (M–P) inter-individual TCR repertoire similarity, as measured by the Morisita-Horn similarity index. The Simpson's diversity and Morisita-Horn similarity indices account for the clonal dominance hierarchy among the different clonotypes and vary between 0 (minimum diversity/similarity) and 1 (maximum diversity/similarity). Each of the diversity, inter-individual sharing and similarity measures were estimated for a standard sample size of 22 TCR sequences per individual mouse repertoire. The repertoire diversities were calculated for each mouse per age/priming group for primary (A, E) and secondary (B, F) D^bNP₃₆₆ ⁺Vβ8.3⁺CD8⁺ TCR repertoires and for primary (C, G) and secondary (D, H) D^bPA₂₂₄ ⁺Vβ7⁺CD8⁺ TCR repertoires. The repertoire similarities were assessed between pairs of primary (M) and secondary (N) D^bNP₃₆₆ ⁺Vβ8.3⁺CD8⁺ TCR repertoires and between pairs of primary (O) and secondary (P) D^bPA₂₂₄ ⁺Vβ7⁺CD8⁺ TCR repertoires within the same age/priming group. To evaluate TCR sharing, clonotypes were first defined as shared or non-shared across all D^bNP₃₆₆-specific or D^bPA₂₂₄-specific TCRβ repertoires. The proportions of the 22 TCRβ sequences per D^bNP₃₆₆ ⁺Vβ8.3⁺CD8⁺ TCR repertoire (I, J) or D^bPA₂₂₄ ⁺Vβ7⁺CD8⁺ TCR repertoire (K, L) that were comprised of shared clonotypes were then estimated. A Mann-Whitney test was used to compare between young and aged mice for the primary responses and between young mice, aged mice primed young and aged mice primed old for the secondary responses. For the comparison between age/priming groups for the secondary responses, the statistical significance for each pairwise comparison was determined at p<0.0167 (*), using Bonferroni correction for multiple pairwise comparisons.

Figure 8. Comparison between aged and young mice of the dominance of shared D^bNP₃₆₆ ⁺CD8⁺ Vβ8.3⁺ and D^bPA₂₂₄ ⁺CD8⁺ Vβ7⁺ TCR clonotypes during primary and secondary (primed-young and primed-old) infections.

Shown are the percentages of the D^bNP₃₆₆ ⁺CD8⁺Vβ8.3⁺ (A, B) and D^bPA₂₂₄ ⁺CD8⁺Vβ7⁺ (C, D) TCR repertoires per mouse that are comprised of aa clonotypes shared between a particular number of mice (indicated by colour-coding) during primary (A, C) and secondary (B, D) infections. The number of mice sharing a TCR clonotype was determined across young and aged mice and both primary and secondary challenges. For example, the public D^bNP₃₆₆ ⁺CD8⁺Vβ8.3⁺ TCR clonotype SGGANTGQL (red) was observed in 23 out of 30 mice (A, B). This clonotype contributed to 95% of the primary D^bNP₃₆₆ ⁺CD8⁺Vβ8.3⁺ TCR repertoire of aged mouse M4 (A). There were five D^bNP₃₆₆ ⁺CD8⁺Vβ8.3⁺ TCR clonotypes that were each observed in a number of mice ranging between 4 and 7 mice (beige). Three of these five clonotypes contributed to the D^bNP₃₆₆ ⁺CD8⁺Vβ8.3⁺ TCR repertoire responding to secondary infection in young mouse M7 (B, as indicated the three beige segments). Multiple unshared clonotypes (light grey), which were observed in only one mouse, contributed to the TCR repertoires of many of the mice (as indicated by multiple light grey segments per column).