Age related human T cell subset evolution and senescence

Immunosenescence is a multifactorial phenomenon that affects all compartments of the immune system. T cells are dramatically affected by ageing [3]. Based on age, we divided samples into 9 age groups with a span of 10 years. The results demonstrated that the absolute numbers of CD3+ (P = 0.0189) and CD8+ (P = 0.004) but not CD4+ (P = 0.1699) T cells linearly decline with age, and differences between adjacent groups are not significant (Fig. 1a), indicating that the peripheral T cell reservoir gradually decreases with age, which is particularly obvious for the CD8+ T cell subset.

Due to the rapid decrease in CD8+ T cells with age, the ratio of CD4 to CD8 (CD4/CD8) cells increased in the cohorts (Fig. 1b red line box and C). Vasson et al. performed a comparative analysis of 300 healthy individuals from France, Austria, and Spain and found that France and Spain had a decreased CD4 to CD8 ratio with increasing age, and our data were most similar to the results from Austria. These authors thought it was possibly related to diet [17]. However, there was no information from Asia to compare relative differences. The distribution of the CD4/CD8 ratio in our study is shown in Fig. 1c, and there is a higher percentage of persons in the over 60 age group with a CD4/CD8 ratio in excess of two. Whether the increased CD4/CD8 ratio was related to the proliferation of regulatory T cells (Treg) and/or Th2 cells in older people remains unknown. As previously defined, an inverted CD4/CD8 ratio (< 1:1) is an immune risk factor for almost any age, and an inverted CD4:CD8 ratio is related to fewer B cells, expansion of late-differentiated or senescent T cells (CD8 + CD28-), and higher human cytomegalovirus (HCMV) seropositivity [1, 18]. J Strindhall et al. reported that 8.0% of people aged 20–59 years have an inverted CD4/CD8 ratio, while 15.6% of those aged 60–90 years old have an inverted ratio [19]. These numbers are similar to our results where we also found a higher percentage of CD28- T cells in people with an inverted CD4/CD8 ratio in comparable age groups (Fig. 1c). In terms of HCMV infection, it was reported that the latent infection (IgG Seropositivity) rate in the Chinese population is more than 90%, and even reaches 97.03% in Shanghai province (One of the first-tier cities of China) over age 25 [20]. We did not test for the human cytomegalovirus (HCMV) infection in these cohorts, but data from another project from our team showed that 36 out of 37 healthy individuals were seropositive for HCMV IgG (unpublished data). Based on these data, it is hard for us to analyze whether there are differences between HCMV infected and noninfected cohorts.

The gating strategy for T cells ranging from naïve to T_EF cells is shown in Fig. 2a. Consistent with a previous study [4], our data also verified that the proportion of naïve T cells in the CD4+ and CD8+ T cell subsets sharply decrease with ageing, particularly for the CD8 population (Fig. 2b). With the exception of CD4+ T_EF cells, the percentage of T_CM, T_EM, and T_EF cells in the CD4+ and CD8+ T cell subsets accumulated with age (Fig. 2b left, highlighted by red line box). For the absolute numbers in the different T cell subsets, with the exception of a decrease in CD4+ and CD8+ naïve T cells with age, there was no difference in the other subsets (Fig. 2b right, highlighted by red line box). Moreover, the decrease in naïve T cells could be found both in the percentage and absolute number level, indicating that ageing has quite a large impact on the homeostasis of naïve T cells. Due to the erosion of the thymus beginning at approximately age 20, less naïve T cells can be produced, and the increasing antigens that have been encountered and infections that have occurred during the lifespan contribute to the differentiation of naïve T cells into more differentiated T cell subsets. When examining different memory T cell subsets, only the percentage of changes can be detected without apparent changes in absolute cell number, and this is partially due to the diversion in the subsets analyzed. In addition, the decrease in total CD8 T cell number further diluted the change in cell number for each subset. These results indicate that both the percentage and number of naïve T cells are physiologically related to ageing, while the composition of CD4 and CD8 T cell subsets also reflect the immune situation in individuals to different degrees.

T_SCM is a newly defined T cell subset with the capacities of self-renewal and differentiation into various memory/effector subsets, and this cell type provides a potential reservoir for T cell memory throughout life [21, 22]. In our previous studies, we found a lower proportion of T_SCM in patients with acute or chronic myeloid leukemia compared with age-matched healthy individuals [7, 8]; however, little is known about the T_SCM changes with age. Here, we found that T_SCM cells account for 1.27 ± 0.55% of the CD4 and 0.98 ± 0.53% of the CD8 population, and the absolute numbers of CD4+ and CD8+ T_SCM cells in peripheral blood were 7.97 ± 7.32/μl and 3.44 ± 3.14/μl, respectively. The absolute number of CD8+ T_SCM decrease with age (Fig. 2b and c). What is interesting is that although the absolute number of naïve T cells sharply decrease with age, the ratio of CD4+ T_SCM/CD4+ T_N and CD8+ T_SCM /CD8+ T_N cells linearly increase, particularly for CD8+ T_SCM (Fig. 2c). These results indicate that CD4+ T_SCM are more stable than CD8+ T_SCM, which may explain the slower senescence of the CD4+ T cell subset compared with the CD8 population. Furthermore, although the number of CD8+ T_SCM decreased with age, the rate of decrease was slower than that of the naïve population. Considering thymus erosion and naïve T cell contraction together with ageing [23], it is reasonable to suspect that T_SCM may provide a large contribution to maintaining the homeostasis of peripheral T cell subsets during ageing due to their strong self-renewal and differentiation ability. It appears that the homeostasis of the CD4 population is more stable than that of the CD8 population due to the stable number of CD4+ T_SCM in one’s lifetime.

We also analyzed changes in the co-stimulatory molecules CD28 and FAS (CD95) on the T cell subsets with age. As shown in Fig. 2b left (highlighted by a blue line box), the percentages of the CD28- and CD95+ T cell subsets in the CD4 and CD8 populations accumulated with age, but only the CD4 + CD28- T cells had an increase in absolute number. In addition, the increase in different sub-populations demonstrated different characteristics where it appears that the CD4 T_EM subset has more profound accumulation of CD28- T cells, while the CD8 T_EF subset has a more apparent increase in CD95+ T cells (Fig. 2b left and right, highlighted by a blue line box), which may be due to the change in both percentage and absolute number. Considering that CD28 plays an important role in T cell proliferation and activation, the accumulation of CD28- CD4+ T cells alone with ageing may be an adverse indication for older individuals. In addition, CD95 (Fas) can lead to apoptosis in target cells when it binds to Fas-ligand (Fas-L); thus, the accumulation of the Fas + CD8 + T_EF population may partially explain the sharp decrease in CD8+ T cells with ageing compared with CD4+ T cells.

Studies examining gender-dependent changes demonstrated that females produce higher antibody levels and have a higher number of CD4+ T cells than males [24, 25]. Here, we also analyzed the differences in the T cell subset distribution of three age groups (< 20 years; 20–60 years; > 60 years) between females and males. The results show that males in the 20–60 years group have a higher CD8 naïve T cell proportion and absolute count compared with females (p < 0.05), while males in the > 60 years group have a higher proportion of CD4 naïve T cells and a lower proportion of CD95 + CD4 T cells compared with females. We then further compared the differences in all of the subjects based on sex and found similar results where a higher CD8 naïve T cell proportion and absolute count were found for males in comparison with females (p < 0.05), while males had a lower proportion of CD8+ T_EF and CD28-CD8+ T cells compared with females (Additional file 1: Table S1 and Additional file 2: Table S2). These results are interesting and may indicate slower thymus erosion in males and a higher antigen resistant capacity for males compared with females over age 20; however, larger sample analysis is required in the future to confirm this finding.

Based on changes in memory T cell frequency, pathogen susceptibility and mortality throughout human life, Farber et al. divided an individual’s lifetime into three phases: memory generation (ages 0–20 years), memory homeostasis (ages 20–65 years), and immunosenescence (age > 65 years) [26]. We first summarized the CD4+ and CD8+ T cell changing characteristics in Chinese cohorts based on the three phases (Fig. 3). The total number of CD8+ T cells were more apparently reduced compared with CD4+ T cells across the human lifespan in peripheral blood, and in the sub-population, naïve T cells sharply decreased while T_CM, T_EM, and T_EF accumulated with age (Fig. 3a and c). Although the absolute number of CD8+ T_SCM decreased with age, the percentage of CD4+ and CD8+ T_SCM could maintain a stable level throughout the lifespan. At the same time, CD28- and CD95+ T cells also accumulated, which could result in a loss of activation and proliferation potential in T cells (Fig. 3b). These findings indicate that although T cell reservoirs and function decrease with age, the stable level of CD8+ T_SCM frequency may be quite important in older people for maintaining the homeostasis of peripheral T cell memory; however, much work remains to be done in the future to clearly understand the importance of T_SCM in humans, such as whether T_SCM cells can divide in a self-renewal manner at the clonal level or not?

Peripheral blood (PB) samples were obtained from the Department of the clinical laboratory, First Affiliated Hospital of Jinan University. Subjects had a current or recent acute infection, and those with autoimmune disease or diabetes mellitus were excluded. Ninety-two healthy volunteers (44 females and 48 males) ranging from 3 to 88 years old were enrolled. There were 9 different age groups with a span of 10 years. The characteristics of the healthy volunteers are listed in Table 1. The age ranges for each of the groups were as follows: 3–9 years (n = 9); 10–19 years (n = 10); 20–29 years (n = 11); 30–39 years (n = 9); 40–49 years (n = 10); 50–59 years (n = 11); 60–69 years (n = 11); 70–79 years (n = 10); and 80–88 years (n = 11). All procedures were conducted according to the guidelines of the Medical Ethics Committees of the Health Bureau of the Guangdong Province in China, and ethical approval was obtained from the Ethics Committee of the Medical School of Jinan University.

Cell surface staining for flow cytometry was performed using the following antibodies: CD45-APC, CD3-FITC, CD4-APC-H7, CD8-Percp-Cy5.5, CD28-PE, CD95-PE-Cy7, CCR7-BV421, CD45RO-BV510, a BV510 isotype Control, and a BV421 isotype Control. Extracellular staining was performed according to the manufacturer’s instructions. The CCR7-BV421 fluorescent antibody was stained independently. Twenty microliters of absolute count microsphere (Thermo; Cat: C36950) was added to samples for absolute number analysis. Cells were analyzed with a BD Verse flow cytometer (BD, Biosciences, USA), and data analysis was performed using FlowJo software.

All data are represented as median, and statistically significant differences between the different T cell populations and between CD28- and CD95+ T cells were analyzed by the Mann-Whitney U test for nonparametric values. Calculations were performed using GraphPad Prism version 7.00 software and SPSS 23.