Age-associated decline in T cell repertoire diversity leads to holes in the repertoire and impaired immunity to influenza virus

Abstract

A diverse T cell repertoire is essential for a vigorous immune response to new infections, and decreasing repertoire diversity has been implicated in the age-associated decline in CD8 T cell immunity. In this study, using the well-characterized mouse influenza virus model, we show that although comparable numbers of CD8 T cells are elicited in the lung and lung airways of young and aged mice after de novo infection, a majority of aged mice exhibit profound shifts in epitope immunodominance and restricted diversity in the TCR repertoire of responding cells. A preferential decline in reactivity to viral epitopes with a low naive precursor frequency was observed, in some cases leading to "holes" in the T cell repertoire. These effects were also seen in young thymectomized mice, consistent with the role of the thymus in maintaining naive repertoire diversity. Furthermore, a decline in repertoire diversity generally correlated with impaired responses to heterosubtypic challenge. This study formally demonstrates in a mouse infection model that naturally occurring contraction of the naive T cell repertoire can result in impaired CD8 T cell responses to known immunodominant epitopes and decline in heterosubtypic immunity. These observations have important implications for the design of vaccine strategies for the elderly.

Figure 1.

In vivo limiting dilution analysis to determine functional precursor frequencies of NP-, PA-, and PB1-specific CD8 T cells. The indicated numbers of CD8 T cells isolated from young naive C57BL/6 donor mice were adoptively transferred into young TCRβ/δ^−/− recipient mice 1 d before infection with influenza virus. Lung tissue was harvested from individual recipient mice 14 d after infection. Each FACS dot plot shows the frequency of donor lymphocytes present within the lung tissue of individual recipient mice that stained positive for both CD8 and either influenza NP (NP_366-374/D^b), PA (PA_224-233/D^b), or PB1 (PB1_703-711/K^b) MHC class I tetramer. The data are representative examples of analysis of 3–4 mice per dilution, shown in Fig. 2.

Figure 2.

The functional precursor frequency of CD8 T cells in young C57BL/6 mice is lower for NP than for PA or PB1. The indicated numbers of purified CD8 T cells isolated from young C57BL/6 mice were adoptively transferred into young TCR β/δ^−/− recipient mice via i.v. injection 1 d before infection with influenza virus. Donor CD8 T cells specific for NP (NP_366-374/D^b), PA (PA_224-233/D^b), or PB1 (PB1_703-711/K^b) present in lung tissue of individual mice on day 14 after infection were identified by flow cytometric analyses. The relative proportion of total measured donor cell response was calculated by dividing the percentage of each epitope-specific CD8 T cell population by the total percentage of CD8 T cells specific for all three epitopes examined (e.g., percentage of NP/[percentage of NP + percentage of PA + percentage of PB1]). Each bar represents data calculated from an individual infected recipient mouse.

Figure 3.

A majority of aged C57BL/6 mice show reduced CD8 T cell responses to influenza NP, but not to PA or PB1. BAL and lung tissue were harvested from individual young (closed symbols) and aged (open symbols) C57BL/6 mice 10 d after influenza virus infection. Young and aged mice presented with comparable frequencies (A) and absolute numbers (B) of CD8 T cells within the BAL and lung tissue. The frequencies (C) and total numbers (D) of CD8 T cells positive for NP (NP_366-374/D^b), PA (PA_224-233/D^b), or PB1 (PB1_703-711/K^b), assessed by tetramer staining of the BAL cells from infected mice, are shown. Bars indicate the medians calculated from data compiled from four independent experiments (11 young and 19 aged mice for NP and PA, and 10 young and 14 aged mice for PB1). Significance was assessed using the Mann-Whitney rank test (two-tailed, 95% confidence).

Figure 4.

Aged C57BL/6 mice show reduced NP responses when infected with an influenza virus lacking the PA epitope. Young and aged C57BL/6 mice were intranasally infected with PA-deficient influenza virus (600 EID₅₀) and lung tissue was harvested 10 d later. The frequencies (A) and total numbers (B) of CD8 T cells positive for NP (NP_366-374/D^b) in the lungs of individual infected mice are shown. Bars indicate the medians calculated from data compiled from two independent experiments (total of 8 young and 15 aged mice). Significance was assessed using the Mann-Whitney rank test (two-tailed, 95% confidence).

Figure 5.

NP-specific CD8 T cells from aged mice exhibit altered Vβ usage. Young and aged C57BL/6 were intranasally infected with influenza virus and BAL was harvested on day 10 after infection. Panels on the left indicate the frequencies of CD8 T cells detected in the BAL of each individual mouse specific for the (A) NP (NP_366-374/D^b), (B) PA (PA_224-233/D^b), or (C) PB1 (PB1_703-711/K^b) epitopes. Panels on the right indicate TCR Vβ chain usage of epitope-specific CD8 T cells determined using the indicated anti-Vβ antibodies. Each symbol identifies an individual mouse. Closed symbols represent individual young mice (n = 3), and open symbols represent individual aged mice (n = 5).

Figure 6.

Spectratype profiles for selected Vβ gene families of NP- and PA-specific CD8 T cells from young and aged influenza virus–infected mice. Representative spectratype analysis of CD8 T cells from young, naive mice (top row) or purified NP- (middle rows) and PA-specific (bottom rows) T cells isolated by FACS-sorting from representative individual young and aged mice 10 d after influenza virus infection. Relative intensity is plotted along the y axis and nucleotide size is plotted along the x axis. The numbers refer to the size, in base pairs, of the individual expanded peaks. Young mouse 5, aged mouse 5, and aged mouse 4 had 28, 8, and 1% NP-specific cells among total CD8 T cells. Young mouse 4, aged mouse 2, and aged mouse 4 had 9, 19, and 4% PA-specific cells among total CD8 T cells, respectively. The complete spectratype analysis for all mice examined is presented in Tables I and II. The y axis is the same for all spectratype profiles, with the maximal height at 6,400 arbitrary units.

Figure 7.

Thymectomy results in the perturbation of the NP-specific repertoire of CD8 T cells in young C57BL/6 mice. Young naive C57BL/6 mice were thymectomized, rested for 7 mo, and then intranasally infected with influenza virus. Lung tissue was harvested from both groups of mice 11 d after infection. The frequencies and absolute numbers of CD8 T cells detected in the lungs of individual infected mice are shown in A and B. The frequencies and absolute numbers of CD8 T cells positive for the influenza virus MHC class I tetramers NP (NP_366-374/D^b), PA (PA_224-233/D^b), or PB1 (PB1_703-711/K^b) detected in the lungs of individual infected mice are shown in C and D. Open and closed symbols represent individual thymectomized (ThyX) and control mice, respectively (n = 5 for each group). Bars indicate the medians calculated from the groups of control and thymectomized mice. Significance was assessed using the Mann-Whitney rank test (two-tailed, 95% confidence).

Figure 8.

Protective heterologous immunity in aged mice correlates with the ability to develop a primary response to the NP epitope. Young and aged C57BL/6 mice were first intranasally infected with 300 EID₅₀ PR8 (H1N1), and then were intranasally challenged with 30,000 EID₅₀ x31 (H2N3) 30 d later. On day 7 after challenge, lung tissue was harvested from infected mice for measurement of viral titers, and splenic tissue was harvested for tetramer analyses, as described in the Materials and methods. Each symbol represents the frequency of splenic CD8 T cells specific for NP as determined for individual young (open circles) and aged (closed circles) mice that had either cleared or not cleared virus. The means of 18 young mice that cleared virus, 14 aged mice that cleared virus, and 7 aged mice that failed to clear virus are indicated.