CMV-Specific T-cell Responses at Older Ages: Broad Responses With a Large Central Memory Component May Be Key to Long-term Survival

Abstract

Cytomegalovirus (CMV) infection sometimes causes large expansions of CMV-specific T cells, particularly in older people. This is believed to undermine immunity to other pathogens and to accelerate immunosenescence. While multiple different CMV proteins are recognized, most publications on age-related T-cell expansions have focused on dominant target proteins UL83 or UL123, and the T-cell activation marker interferon-γ (IFN-γ). We were concerned that this narrow approach might have skewed our understanding of CMV-specific immunity at older ages. We have, therefore, widened the scope of analysis to include in vitro-induced T-cell responses to 19 frequently recognized CMV proteins in "young" and "older" healthy volunteers and a group of "oldest old" long-term survivors (>85 years of age). Polychromatic flow cytometry was used to analyze T-cell activation markers (CD107, CD154, interleukin-2 [IL-2], tumor necrosis factor [TNF], and IFN-γ) and memory phenotypes (CD27, CD45RA). The older group had, on average, larger T-cell responses than the young, but, interestingly, response size differences were relatively smaller when all activation markers were considered rather than IFN-γ or TNF alone. The oldest old group recognized more proteins on average than the other groups, and had even bigger T-cell responses than the older group with a significantly larger central memory CD4 T-cell component.

Keywords: Ageing; Central memory T-cells; Cytomegalovirus; Response breadth.; T-cell memory inflation; T-cells.

Figure 1.

The frequency of target protein recognition is unrelated to T-cell response size. PBMCs from CMV⁺ participants were stimulated overnight with 19 CMV protein-derived overlapping peptide-pools. Activated T cells were identified by flow cytometry. A, Bars represent all age groups and indicate the fraction of individuals recognizing individual proteins with respect to CD4 and CD8 T cells. Proteins are ordered by decreasing frequency of recognition. B, The sizes of CD4 and CD8 T-cell responses (log₁₀-transformed fractions) across all age groups are shown for all proteins in the same order as under (A).

Figure 2.

The breadth of the CMV-specific T-cell response is not significantly different between the young and older groups, but strongly increased in the oldest old group. PBMCs from CMV⁺ participants were stimulated overnight with 19 CMV protein-derived overlapping peptide pools. Activated T cells were identified by flow cytometry. A, A comparison of response breadth between the young (white bars) and older (gray bars) groups revealed no significant differences in terms of protein recognition frequencies (CMV proteins are ordered by decreasing frequency of recognition in the older group); however, there were several significant differences between the older and the oldest old groups (dark gray bars) (Bonferroni multiple endpoint correction, significance threshold set to P = .003, significant differences indicated by asterisks). B, The number of recognized CMV target proteins (between 1 and 15) was computed separately for CD4 and CD8 T cells in the young (left), older (middle), and oldest old (right) groups, suggesting a mild (nonsignificant) trend for higher response counts in the older compared to the young groups, but also shows a significant difference between the older and oldest old groups. Cross bars indicate median and interquartile ranges.

Figure 3.

Age-related increases in T-cell response size depend on target protein–specificity and functional response readout. PBMCs from CMV⁺ participants were stimulated overnight with 19 CMV protein-derived overlapping peptide pools. Activated T cells were identified by flow cytometry. While our study focused on “average” aging, (ie, differences between young and older participants, the oldest old group participants are shown as examples of unusually successful aging). A–B, The fractions of all cells displaying at least 1 activation marker (“combined readout”), IFN-γ, or TNF are shown. Diagrams show the CMV-specific T-cell response size (log-transformed fractions of CD4 or CD8 T cells) for all 19 proteins combined (left panels) and the most frequently recognized CMV proteins in the UK cohort for CD4 T cells with respect to UL83 (middle) and UL55 (right) (A), for CD8 T-cells with respect to UL83 (middle) and UL123 (right) (B). Statistical significance levels are indicated. The main study endpoint was the increase in CMV-specific T-cell response size between the young and older groups (combined readout in connection with all 19 proteins); the significance level of P ≤ .05 was not adjusted. C–D, T-cell memory compartment distributions defined by the expression of CD27 and CD45RA (CD45RA⁺/CD27⁺ = “naive” or T_NA; CD45RA^–/CD27⁺ = “central memory” or T_CM; CD45RA^–/CD27^– = “effector memory” or T_EM; CD45RA⁺/CD27^– = “revertant” or T_EMRA) showed significant differences between the young and older groups among CMV-specific T_CM CD4 T cells and CD8 revertant (T_EMRA) T cells. Compared with the older group, the oldest old group displayed a striking and significant increase of the CD4 central memory (T_CM) compartment (Mann–Whitney test, significance threshold set to at P ≤ .0125, Bonferroni correction for 4 endpoints). No direct comparison was made between the young and oldest old group participants. Boxplots show minimum, maximum, median, interquartile range, and outliers (“o”).