Cytomegalovirus Exposure in the Elderly Does Not Reduce CD8 T Cell Repertoire Diversity

Abstract

With age, the immune system becomes less effective, causing increased susceptibility to infection. Chronic CMV infection further impairs immune function and is associated with increased mortality in the elderly. CMV exposure elicits massive CD8⁺ T cell clonal expansions and diminishes the cytotoxic T cell response to subsequent infections, leading to the hypothesis that to maintain homeostasis, T cell clones are expelled from the repertoire, reducing T cell repertoire diversity and diminishing the ability to combat new infections. However, in humans, the impact of CMV infection on the structure and diversity of the underlying T cell repertoire remains uncharacterized. Using TCR β-chain immunosequencing, we observed that the proportion of the peripheral blood T cell repertoire composed of the most numerous 0.1% of clones is larger in the CMV seropositive and gradually increases with age. We found that the T cell repertoire in the elderly grows to accommodate CMV-driven clonal expansions while preserving its underlying diversity and clonal structure. Our observations suggest that the maintenance of large CMV-reactive T cell clones throughout life does not compromise the underlying repertoire. Alternatively, we propose that the diminished immunity in elderly individuals with CMV is due to alterations in cellular function rather than a reduction in CD8⁺ T cell repertoire diversity.

FIGURE 1.

Impact of CMV on the proportion of large clones with age. Boxplot comparing the proportion of the most numerous 0.1% of clones in the peripheral blood T cell repertoire of CMV⁺ (blue) and CMV⁻ (orange) subjects. TCRβ sequencing was performed on the PBMCs of each subject. For each subject, the cumulative abundance of the most numerous 0.1% of clones was divided by the total sample abundance to yield a proportion. Nonproductive TCRβ rearrangements were excluded from this calculation. The band inside each box represents the median, and the whiskers extend to values that are within 1.5 times the interquartile range. Outliers are represented by dots. The N beneath each box represents the number of samples in each group. *p < 0.0005, **p < 0.0001.

FIGURE 2.

Distribution of large peripheral blood T cell clones in the elderly. (A) Scatterplot comparing the fraction of the most numerous 0.1% of peripheral blood T cell clones present in each sorted T cell subset in CMV⁺ (blue) and CMV⁻ (orange) subjects. TCRβ sequencing was performed on the PBMCs and specified T cell subsets of eight subjects >70 y old. Nonproductive TCRβ rearrangements were excluded from these analyses. The fraction of the most numerous 0.1% of peripheral blood T cell clones present in each T cell subset is depicted. (B) The proportion of each sorted T cell subset composed of the most numerous 0.1% of peripheral blood T cell clones in CMV⁺ and CMV⁻ subjects. The cumulative abundance of the most numerous 0.1% of peripheral blood T cell clones present in each T cell subset were divided by the total abundance of each subset to yield a proportion. The most numerous 0.1% of peripheral blood T cell clones found in both naive and memory or T_EMRA subsets were bioinformatically removed from the naive repertoire. *p < 0.05. CD4, Bulk CD4⁺ T cells; CD8 Memory, CD8⁺ central and effector memory T cells; CD8 Naive, CD8⁺ naive T cells.

FIGURE 3.

Identification of CMV-reactive T cell clones in the elderly. (A and B) Scatterplot comparing clone frequencies in CMV-stimulated CD137⁺ and resting memory (A) or T_EMRA (B) subsets from CMV⁺ subjects four and five. CD8⁺CD45RO⁺ memory and CD8⁺CD45RA⁺CD62L^lo/− T_EMRA cells were sorted and stimulated with autologous CMV-infected fibroblasts for 24 h. CD137⁺ T cells were then sorted and TCRβ sequencing was performed. The frequency of productive TCRβ sequences from unstimulated memory and T_EMRA samples are plotted against the frequency of productive TCRβ sequences from the corresponding CD137⁺ sample. Each point represents a unique clone. Points along the axis represent clones present in one sample. Points colored blue in (A) represent clones present in the CD137⁺ T_EMRA sample and in (B) represent clones also present in the CD137⁺ memory T cell sample. Logarithmic scale, base 10. (C) Comparison of the fraction of each T cell subset as unique rearrangements (pink) or reads (green) composed of CD137⁺ CMV-reactive clones. Memory and T_EMRA CD137⁺ TCRβ sequences were combined to capture all CMV-reactive clones present in each PBMC T cell repertoire. Pink represents the fraction of unique CMV-reactive clones found in each corresponding unstimulated sample. Green represents the cumulative abundance CMV-reactive clones found in each corresponding unstimulated sample divided by the total abundance of the sample. Memory, CD8⁺ central and effector memory; PB, peripheral blood.

FIGURE 4.

Effect of CMV on the underlying T cell repertoire. (A and B) Scatterplot comparing the Shannon entropy (A) or clonality (B) of each T cell subset in CMV⁺ (blue) and CMV⁻ (orange) subjects. The most numerous 0.1% of peripheral blood T cell clones found in both naive and memory or T_EMRA repertoires were bioinformatically removed from naive T cell entropy and clonality calculations. Nonproductive TCRβ rearrangements were excluded from these calculations. (C) Histogram comparing the frequency distribution of naive T cell clones in CMV⁺ and CMV⁻ subjects. Each bar represents the total number of unique clones present at a particular frequency. Naive T cell clones with frequencies greater than 10⁻⁴ are not displayed. Logarithmic scale, base 10. *p < 0.05. Memory, CD8⁺ central and effector memory; PB, peripheral blood. S, subject.

FIGURE 5.

Accommodation of large clonotypes in the T cell repertoire. (A) Scatterplot comparing the Shannon entropy in CMV⁺ (blue) and CMV⁻ (orange) subjects after removal of the most numerous 0.1% of peripheral blood T cell clones from each T cell subset. The most numerous 0.1% of peripheral blood T cell clones were bioinformatically removed from each T cell subset and the Shannon entropy values were recalculated. Nonproductive TCRβ rearrangements were excluded from these calculations. (B) Stacked bar chart comparing the total number of T cells in each subset in four CMV⁺ and two CMV⁻ subjects. Clinical laboratory CD8⁺ T cell counts were performed on blood samples from six to eight study subjects. The relative frequency of each CD8⁺ T cell subset obtained by flow cytometry is used to identify the number of T cells in each subset. Light gray, Naive T cells; Medium gray, Memory T cells; and Dark gray, T_EMRA cells. (C–E) A model of the T cell repertoire with subordinate clones (orange) and CMV-reactive clones (blue) depicting the potential impact of CMV infection in (C) T cell repertoire at baseline prior to CMV exposure. All clones share similar frequency distributions. (D and E) T cell repertoire after CMV exposure with massive clonal expansions that supplant rare T cell clones from the repertoire (D) or that does not modify the underlying T cell repertoire (E). Sphere size corresponds to clone abundance. Memory, CD8⁺ central and effector memory; PB, peripheral blood.