Telomere Shortening, Inflammatory Cytokines, and Anti-Cytomegalovirus Antibody Follow Distinct Age-Associated Trajectories in Humans

Abstract

A number of biological parameters have been cited as hallmarks of immune aging. However, it is not clear whether these multiple biological changes are the result of common underlying aging processes and follow correlated trajectories, or whether the patterns of change for multiple parameters vary across individuals and reflect heterogeneity in the aging process. Here, we have studied parameters of immune system aging through longitudinal analysis of telomere length, inflammatory cytokines, and antibody titer to cytomegalovirus (CMV) in 465 subjects ranging in age from 21 to 88 years at the first visit, with an average of 13 years (7-19 years) follow-up. We observed a highly variable rate of change in telomere length of PBMCs with a relatively slow average rate of telomere shortening (-16 bp/year). Similarly, there were significant increases with age in vivo in three inflammation-related cytokines (interferon gamma, IL-6, and IL-10) and in anti-CMV IgG titer, which varied widely across individuals as well. We further observed positive correlative changes among different inflammatory cytokines. However, we did not find significant correlations among the rate of changes in telomere length, inflammatory cytokines, and anti-CMV IgG titers. Our findings thus reveal that age-related trajectories of telomere attrition, elevated circulating inflammatory cytokines, and anti-CMV IgG are independent and that aging individuals do not show a uniform pattern of change in these variables. Immune aging processes are complex and vary across individuals, and the use of multiple biomarkers is essential to evaluation of biological aging of the immune system.

Keywords: IL-10; IL-6; aging; anti-CMV IgG; interferon gamma; telomeres.

Figure 1

Reduction of telomere length in PBMCs with age in vivo. (A) Representative image of telomere measurement. DNA from PBMCs of eight subjects (age at each visit) was digested, separated on 0.6% agarose gel, and probed with ³²P-TTAGGG₄. The image was collected from a phosphorimager. (B) Graph of telomere length (kb) of PBMCs in each subject for the two visits connected by a line (N = 465). (C) Graph of telomere length (kb) of PBMC by six age groups (under 40, 40–49, 50–59, 60–69, 70–79, and 80 years and older) for male in blue and female in red. Two P values were presented: P_C refers to the initial evaluation and represents cross-sectional effect of age, and P_L refers to the longitudinal time change within subjects (for all figures). Models were calculated by linear mixed effect regression (33). P < 0.05 is considered as significant.

Figure 2

Increase of circulating interferon gamma (IFN-γ), IL-6, and IL-10 with age in vivo. (A) Line graphs of IFN-γ (pg/mL) by date adjusted raw data for the two visits of each subject (N = 432) (left) and by five age groups (under 50, 50–59, 60–69, 70–79, and 80 years and older) for male in blue and female in red (right). (B) Line graph of IL-6 (pg/mL) by date adjusted raw data for the two visits of each subject (N = 432) (left) and by five age groups (right). (C) Line graph of IL-10 (pg/mL) by date adjusted raw data for the two visits of each subject (N = 432) (left) and by five age groups (right). P_I refers to the age group difference using the mixed effects model.

Figure 3

Increase in titers of anti-cytomegalovirus (CMV)-IgG in serum with age in vivo. (A) Line graph of anti-CMV-IgG for the two visits of each subject (N = 435). (B) Graph of anti-CMV-IgG by five age groups (under 50, 50–59, 60–69, 70–79, and 80 years and older) for male in blue and female in red.

Figure 4

No correlation of the rate changes among telomere length attrition, increased inflammatory-related cytokines, and anti-cytomegalovirus (CMV) IgG. (A) Partial correlation coefficient among the rate changes of telomere length, eight inflammatory cytokines [IL-1β, IL-2, IL-4, IL-6, IL-10, IL-12p70, IL-13, and interferon gamma (IFN-γ)], and anti-CMV IgG titers. The correlation coefficients between pair comparison of the rate of change were analyzed with adjustment for age, sex, and date, and data presented as a clustered heat map. The scale is −0.5 to 1. (B) Selected significant correlations were presented with correlation coefficient (R) and P values. After adjusting by FDR, P value < 0.001 was significant at P < 0.05.