Lower proportion of naïve peripheral CD8+ T cells and an unopposed pro-inflammatory response to human Cytomegalovirus proteins in vitro are associated with longer survival in very elderly people

Abstract

The low percentages of naïve T cells commonly observed in elderly people are thought to be causally associated with mortality, primarily from infectious disease, and are taken as a hallmark of "immunosenescence". Whether low levels of naive cells actually do associate with mortality has, however, not been tested in longitudinal studies. Here, we present correlations between peripheral T-cell phenotypes and 8-year survival in individuals from the population-based prospective Leiden 85-plus Study. Counter-intuitively, we found that a lower frequency of naïve CD8+ T cells (characterized as CD45RA+CCR7+CD27+CD28+) at baseline (>88 years) correlated with significantly better survival, while there was a tendency for the reciprocal accumulation of late-differentiated effector memory cells (CD45RA-CCR7-CD27-CD28-) also to associate with better survival. These findings suggest that better retention of memory cells specific for previously encountered antigens may provide a survival advantage in this particular population. Given the prevalence of Cytomegalovirus (CMV) and its reported association with immunosenescence, we tested whether memory for this potential pathogen was relevant to survival. We found that individuals mounting an exclusively pro-inflammatory ex vivo response (TNF, IFN-γ, IL-17) to the major CMV target molecules pp65 and IE1 had a significant survival advantage over those also having anti-inflammatory responses (IL-10). These findings suggest that higher levels of naïve T cells may not necessarily be associated with a survival advantage and imply that the nature of immunosurveillance against CMV may be crucial for remaining longevity, at least in the very elderly.

Fig. 1

CMV seropositivity and distribution of T-cell phenotypes. Frequency of naïve and late-differentiated memory T cells within CD8+ (a) and CD4+ T cells (b) in seven CMV-seronegative (CMV−) and 41 CMV-seropositive (CMV+) individuals. Frequency of T cells carrying putative senescence markers CD57+ (c) and KLRG-1 (d) on CD4+ and CD8+ T cells were compared in the same individuals. Horizontal bars represent the median of each group

Fig. 2

FACS plots demonstrating production of INFγ (upper row), TNF (middle row), and IL-10 (lower row) in the CD3+ gated population. The unstimulated control (shown in the left-hand panel) was used to set the cut-off for cytokine production

Fig. 3

Correlation analysis between the in vitro immune response against immunodominant CMV antigens and mortality in individuals 88 years old at baseline. Kaplan–Meier survival curves for 21 CMV-seropositive individuals according to response to either pp65 or IE1 (a), or pp65 or IE1 (b) are shown. Donors were grouped according to the type of cytokine response; only pro-inflammatory—IFNγ, TNF, and/or IL-17 in the absence of any IL-10 production (black line), pro-/anti-inflammatory—IL-10 production by CD4+ and/or CD8+ T cells together or in the absence of pro-inflammatory cytokines (gray line)