Direct Measurement of T Cell Subset Kinetics In Vivo in Elderly Men and Women¹

The deterioration in immunoresponsiveness that accompanies ageing is thought to contribute to the greater susceptibility of the elderly to autoimmune disease, cancer, and infectious disease and to their impaired responses to vaccination (1, 2).

Although the absolute numbers of lymphocytes in the recirculating peripheral pool of T cells remain at a stable level over the life span of an individual (3, 4), an age-related decline in both humoral and cellular responses is observed (5, 6). This may be due in part to alterations in the functional properties of specific cell types, but could also be linked to changes in the distribution of subsets within the lymphocyte compartment. Understanding how the latter occurs depends on information about rates of cellular proliferation and death within the lymphocyte pool.

Development of the total T cell pool in early life is dependent on the output of naive T cells with a diverse αβ TCR repertoire from the thymus and on the establishment of an Ag-experienced memory cell pool with a more restricted repertoire (7, 8). With normal aging there is a decline in thymic output; this has been shown in both humans and experimental animals (9, 10). In humans it has been shown that thymectomy and thymic irradiation have little impact on absolute numbers of T cells in adults (11, 12), and there is an age-associated decline in the naive T cell regenerative capacity after bone marrow transplants (13). More recently, quantitation of thymic output by measuring TCR rearrangement excision circles has shown that the percentage of recent thymic emigrants decreases dramatically with age. Although TCR rearrangement excision circle numbers decrease 1–1.5 log from birth to age 73 years, there is only a 4-fold decrease in naive CD4⁺ and CD8⁺ T cell numbers (14). One explanation may be that as thymic output decreases, the adult naive T cell pool is maintained by changes in proliferation and survival rates of naive T cells.

The naive and memory T cell pools in the periphery seem to be independently regulated (15). Studies in both mice and humans have shown that naive T cells have a slow rate of turnover, without phenotype change, and have a relatively long life span compared with memory cells (16, 17). Their survival requires TCR ligation by self MHC molecules (18, 19) and the presence of particular cytokines, most notably IL-7 (20). In the memory pool, a fraction of cells is cycling (16). In the case of CD8 memory cells, this proliferation does not appear to require foreign Ag or MHC interactions (21) and seems to be driven predominantly by IL-15 and IL-7 (22, 23, 24). Whether there is active maintenance of the CD4 memory pool and how this is controlled are less clear (25, 26).

A number of studies have shown that one change observed with age is an increased proportion of cells, particularly within the T cell compartment, that have a memory/effector phenotype (27, 28). This may be due in part to altered kinetics of subpopulations within the T cell pool. Another consequence of ageing is the build up of very large clonal populations, particularly within the CD8 compartment (29, 30). These Ag-specific, clonally expanded cells exist not only in the CD45RO⁺ subpopulation, but also within the CD45RA⁺ subpopulation, indicating that such clonally expanded cells are phenotypically heterogeneous. Clonal expansion in combination with dwindling thymic output of naive cells leads to shrinkage of the immune repertoire. Recent evidence suggests that some large oligoclonal expansions may be driven by persistent viral infections such as CMV (31). In the case of CMV it has been shown that 55% of individuals aged 40–60 years are CMV⁺ and that by the age of 86 years this has risen to 87% (32). How such large clones persist and whether they are terminally differentiated is not clear, but their state of differentiation may correlate with the antigenic load associated with chronic viral infection. Increasing viral load would lead to exhaustion of T cell effector function and cessation of clonal expansion, resulting in accumulation of cells that have reached replicative senescence. The finite capacity of human T cells to divide is demonstrated in in vitro experiments where cultured T cells undergo limited rounds of divisions (33). That this may be important in the elderly is suggested by data indicating that T cells from elderly subjects have shorter telomeres (34) and lose the expression of CD27 (35) and of costimulatory molecules such as CD28 (36), but have increased expression of CD57 (37) and the inhibitory killer cell lectin-like receptor G1 (38), both of which are associated with the loss of replicative potential.

Recently methods have been developed that allow measurement of turnover in vivo by labeling dividing lymphocytes from humans using nonradioactive isotopes (39). Such studies have demonstrated that in young donors CD45RO⁺ T cells turnover more rapidly than CD45RA⁺ T cells (40). We have used this method to assess whether impaired T cell function in the elderly might be attributable to changes in turnover rates in healthy elderly compared with young controls, and whether oligoclonal expansions within the CD8 compartment occur as a consequence of dysregulation of proliferation and/or disappearance rates.

Elderly subjects were defined as being at least 65 years old and were in good general health. Subjects were screened before study entry by a physician and were excluded if they had any history of significant known medical problems, including infection, inflammatory conditions, or malignancy. Other specific exclusion criteria included current or recent active infection, febrile illness, surgery or vaccination (within the last 4 wk), diabetes mellitus, excess alcohol intake (>36 U/wk for men, >24 U/wk for women), or medication with glucocorticoids within the last 2 mo. Blood was checked for normal electrolyte concentrations; renal, hepatic, thyroid, and liver function; and C-reactive protein. CMV was tested by enzyme immunoassay (Vidar; bioMerieux, Marcy l’Etoile, France). Elderly subjects were compared with a cohort of healthy young individuals, who have been described previously (39). All subjects gave written informed consent to procedures, and the study was approved by the local research ethics committee.

To investigate the kinetic basis of homeostasis of lymphocyte subpopulations in elderly subjects, proliferation and disappearance rates of lymphocytes were measured by labeling dividing cells in vivo with deuterated glucose. Labeling consisted of a primed 24-h i.v. infusion of 1 g/kg [6,6⁻²H₂]glucose (Cambridge Isotopes, Cambridge, MA), during which diet was restricted to small, low energy meals. Blood for measurement of plasma glucose deuterium (²H) enrichment was taken after 1 h and approximately every 4 h thereafter during the infusion. Labeled glucose has a half-life of <1 h in the circulation. Blood samples for estimation of deuterium enrichment in DNA were taken 4, 10, and 21 days after infusion or on the closest day possible. In young controls, blood was also taken on day 3, but such data were disregarded in analysis to maintain comparability with the elderly; as a consequence, data from young controls differ slightly from previously published analyses, modeling of which did include day 3 data (39).

PBMC were isolated from 50 ml of heparinized fresh blood by Ficoll-Paque (Pharmacia Biotech, St. Albans, U.K.) density gradient centrifugation. Cells (1 × 10⁷/ml in PBS and 0.2% BSA) were stained with CD3-RPE (Serotec, Oxford, U.K.), then sorted into CD3⁺ and CD3⁻ fractions using a Moflo cytometer (DakoCytomation, Fort Collins, CO). CD3⁺ cells were further stained with biotinylated Ab against CD8 (Serotec) and streptavidin-allophycocyanin (BD Pharmingen, San Diego, CA) together with CD45RA-R-PE-CyChrome 5 (Serotec) and sorted into CD8⁺CD45RA⁺, CD8⁻CD45RA⁺, CD8⁺CD45RA⁻, and CD8⁻CD45RA⁻ subsets. In Results and Discussion, CD45RA⁻ cells are referred to as being CD45RO⁺, and CD8⁻CD3⁺ cells are referred to as being CD4⁺. For the V_β sort, cells were stained with CD45RA-FITC (BD Pharmingen), V_β5-R-PE (BD Pharmingen), and CD8-allophycocyanin (BD Pharmingen).

Enrichment of deuterium in DNA was assayed essentially as previously described (39). DNA from sorted subsets was extracted and digested enzymatically to deoxynucleosides. Deoxyadenosine, purified by C-18 solid phase extraction column chromatography, was converted to its aldononitrile triacetate derivative by reaction with hydroxylamine/pyridine (1%, w/v; 100°C, 45 min) and acetic anhydride (room temperature, 30 min) and was analyzed by gas chromatography-mass spectrometry, monitoring ions m/z 198 and 200 by PCI in SIM (HP-225 column, HP 6890/5973 gas chromatography-mass spectrometer; Hewlett Packard, Bracknell, U.K.). Abundance-matched samples were analyzed in triplicate alongside a standard curve derivatized concurrently. Plasma glucose enrichment was measured using the same derivatization (m/z 328 and 330). The typical precision of reproducibility of the M+2/M+0 ratio was ±0.02%.

The fraction of labeled cells (F) present on each day, is given by the ratio of the enrichment of label in DNA (E) and the precursor enrichment, b (mean glucose enrichment × 0.65), F = E/b, as previously described (39). The magnitude of the peak value for F represents a crude measure of the cellular proliferation rate. Because this estimate does not take into account either cell death between the end of labeling and sampling on day 3 or changes in pool sizes, appearance and disappearance of labeled cells were modeled as previously described (40), assuming no change in lymphocyte subpopulation pool sizes during the experiment.

The observed fraction of labeled cells, F, at a given time will depend upon the rate of proliferation during labeling, p,⁴ and subsequent loss of labeled cell loss, d*, according to the following equations, where t is time and τ is the length of the labeling period:

during the labeling period

after the labeling period.

Curves for fraction of labeled cells vs time were modeled using these equations and solved (Sigmaplot version 8.02; SPSS, Woking, U.K.) to yield the cellular proliferation and disappearance rate constants, p and d*. In this model no assumption of equality between p and d* has been made; p represents the average proliferation rate of the whole population, whereas d* refers only to labeled cells (i.e., cells that divided during the labeling period). For a kinetically heterogeneous population, even one at steady state, these two rates will not be the same, as discussed previously (39). Proliferation is expressed also as the average intermitotic time or estimated time for 50% of cells to divide (doubling time) and disappearance as the time for loss of 50% of the label (t_1/2, half-life equivalent), these were calculated as ln2/p and ln2/d, respectively. Data are expressed as means. Comparisons between groups were made by Student’s t test (two-tailed).

Total RNA was extracted from isolated CD8⁺CD45RA⁺V_β5⁺, CD8⁺CD45RA⁺V_β5⁻, and CD8⁺CD45RA⁻ T cells using TriReagent (Sigma-Aldrich, St. Louis, MO). First-strand cDNA was synthesized using Moloney murine leukemia virus reverse transcriptase (Promega, Madison, WI). Heteroduplex analysis was performed as previously described (41). Briefly, TCR V_β transcripts were amplified by PCR using primers specific for human V_β5.2 and V_β6. The PCR product was mixed 20:1 with V_β-matched DNA carrier, denatured, and allowed to reanneal at 50°C. The product was run on a 12% acrylamide gel, blotted, and hybridized with a probe to the external V_β region of the carrier. The probe was detected using anti-digoxigenin Fab Ab and the chemiluminescent substrate CDP-Star (Roche, Lewes, U.K.).

PBMC (2 × 10⁵) were stained with allophycocyanin-conjugated CD8 (BD Pharmingen), CD45RO-biotin (BD Pharmingen), and streptavidin-PerCP (BD Pharmingen) together with one of the following: FITC-conjugated CD27 (Serotec), CD28, CD95, CD11a, CD62L (BD Pharmingen), CCR7mAb (BD Pharmingen), and goat F(ab′)₂ anti-mouse IgM-RPE (Southern Biotechnology Associates, Birmingham, AL) or with a panel of FITC-conjugated TCR V_β mAbs (Serotec) at 4°C for 30 min. In total, 10,000 events were collected on a FACSCalibur flow cytometer (BD Biosciences, San Jose, CA) and were analyzed using CellQuest software.

Seven elderly subjects (mean age, 76 years) were recruited and compared with eight young controls (mean age, 26 years; Table I). All were healthy at the time of the study. Six of the seven elderly subjects were CMV IgG positive; one (E06) was negative. One positive (C06) was found among the young controls.

Comparison of lymphocyte phenotypes revealed very similar subset distributions for CD3⁺ and CD4⁺ (Table I). In common with published studies, there was a marked increase in the proportion of CD45RO⁺CD8⁺ cells in the elderly (55 vs 26% in young individuals; p < 0.01; Table I), with two individuals having a CD4:CD8 ratio in the range of 1:1 (Table I).

TCR V_β expansions within the CD8⁺ population have previously been described (29, 30). A panel of V_β mAbs was used to determine which subjects had such expansions. All elderly subjects had one or more expansions within the CD8⁺ subpopulation (data not shown).

Cell labeling kinetics were investigated after the infusion of deuterated glucose. The magnitude of peak labeling reflects the proliferation rate, and the decline after the peak reflects the rate of disappearance of label from the blood (Table II). Under steady state conditions, loss of label from blood must mean either death of cells or their irreversible migration into tissues, followed by cell death, although migration and death may be temporally separated. In sheep, fluorescent dye-labeled lymphocytes distribute evenly through several lymphoid compartments and recirculate so that sequestration does not appear to be a major factor in disappearance of label from blood, which mainly reflects cell death (42, 43).

When CD45RO⁺ T cells were analyzed, kinetic profiles were obtained as shown in Fig. 1. Kinetic profiles obtained for the elderly were comparable with those previously obtained for young control subjects (40). For CD4⁺ cells (Fig. 1, A and B), labeling rates between subjects were quite consistent. CD8⁺ populations showed greater heterogeneity between individual subjects in both groups (Fig. 1, C and D). In all subjects rates of labeling of both CD4⁺ and CD8⁺ CD45RO⁺ T cells exceeded those for equivalent CD45RA⁺ cells (Fig. 2). Thus in the elderly, as previously described in the young, CD45RA⁺ cell populations have a slower rate of turnover than their primed counterparts. CD4⁺CD45RA⁺ labeling kinetics were similar in elderly subjects and young controls (Fig. 2, A and B). By contrast, in the CD8⁺CD45RA⁺ population there was a different pattern in elderly subjects; specifically, longer persistence of labeled cells in this population (Fig. 2, C and D).

Values for appearance and disappearance rate constants, derived from modeling, were strikingly similar between elderly and young subjects for all cell types except the CD8⁺CD45RA⁺ population (Table II). In this population uniquely, elderly subjects showed a reduced rate of cell death (or disappearance) of ∼1 vs 14%/day (p < 0.005; Table II), equivalent to half-lives of ∼84 vs 5 days, respectively. Thus, there appeared to be a substantial lengthening of survival in CD8⁺CD45RA⁺ cells in the elderly subjects. In all the other subsets from both groups, disappearance rates generally exceeded proliferation rates. This is possible because disappearance rates take account only of labeled cells, suggesting that newly divided cells have a greater propensity to die.

To investigate the origin of such kinetic differences, we examined the expression of phenotypic markers used to define naive (CD45RA⁺, CD27⁺, CD28⁺, CD62L⁺, CCR7⁺, CD11a^low, CD95⁻), central memory (CD45RO⁺, CD27⁺, CD28⁺, CCR7⁺, CD62L⁺, CD11a^high, CD95⁺), T effector (CD45RO⁺, CD27⁻, CD28⁻, CD62L⁻, CCR7⁻, CD11a^high, CD95⁺), and CD45RA⁺ primed T cells (CD45RA⁺, CD27⁻, CD28⁻, CD62L⁻ CCR7⁻, CD11a^high, CD95⁺). Table III shows an analysis of the expression of these markers in our cohort of elderly subjects compared with young controls. The pattern of expression of the markers was similar in all groups except CD8⁺CD45RA⁺ cells, where there was a significantly lower percentage of cells in the elderly that were CD11a^low, CD27⁺, CD28⁺, CD62L⁺, CCR7⁺, or CD95⁻ (p < 0.002 for all markers). This observation is in keeping with previous reports that the elderly have increased proportions of primed cells in the CD8⁺CD45RA⁺ population (44). Detailed analysis of the phenotypic composition of the CD45RA⁺ population confirmed that the elderly cohort had a primed phenotype compared with the young controls (Fig. 3). These results suggest that alterations in proportions of naive and primed cells within the CD8⁺CD45RA⁺ population may account for the changed kinetics of this subset in the elderly (Fig. 2).

As stated above, all elderly subjects had TCR V_β expansions within the CD8⁺ population. It was therefore of interest to compare the behavior of such expanded clones with that of other T cells. In one subject (E01) we were able to investigate the kinetics of an expansion of V_β5.2 CD8⁺ T cells within the CD45RA⁺ compartment, which heteroduplex analysis showed to be oligoclonal (Fig. 4,a). Both V_β5⁻ CD45RA⁺ and CD45RO⁺ CD8⁺ cells showed many clonal bands when analyzed for other V_β genes (illustrated by V_β6 in Fig. 4,a). This expansion represented 28% of CD8⁺CD45RA⁺ cells, and the cells within this population had a primed phenotype (CD27⁻, CD28⁻, CD62L⁻, CCR7⁻, CD11a^high, CD95⁺; Fig. 5). Conversely, naive cells were found within the CD8⁺CD45RA⁺V_β5.2⁻ population, although they represented a small proportion within this pool, whereas the CD45RO⁺ population had a mixed T central memory/T effector memory distribution. Interestingly, the expression of CD45RB and Bcl-2 was high in the CD45RA⁺V_β5.2⁺ subset, suggesting that these cells are not terminally differentiated.

The kinetic behavior of cells within this expanded clonal population was investigated by sorting for CD8⁺CD45RA⁺V_β5.2⁺, CD8⁺CD45RA⁺V_β 5.2⁻, and CD8⁺CD45RA⁻ T cells and determining the proportion of cells labeled on day 3 postinfusion. This allowed us to compare the proliferation rates among the different populations, although limitations of sampling prevented analysis of the kinetics of disappearance of these cells. The results showed that labeling in the CD45RA⁺V_β5.2⁺ population was considerably higher than that in the CD45RA⁺V_β5.2⁻ cells, but remained lower than the high levels seen in the CD45RA⁻V_β5.2⁻ cells (Fig. 4 b). One explanation for this observation is that phenotypic reversion from CD45RO⁺ to CD45RA⁺ occurs for a proportion of rapidly dividing CD45RO⁺ cells. However, this possibility seems unlikely to account for our findings given the short 3-day time frame of the experiment. Rather, it appears that the clonally expanded CD45RA⁺V_β5.2⁺ cells have a high rate of proliferation relative to their V_β5.2⁻ counterparts, which consist of a mixture of primed and unprimed CD45RA⁺ cells.

In the elderly a higher incidence of certain infections, cancers, and autoimmune diseases are observed (2). At the same time the ability to mount immune responses, in particular to neo-Ags, has been shown to be impaired and may result in decreased or inconsistent vaccine responses (45, 46). Such impaired immune responsiveness is not due to reduced absolute numbers of lymphocytes, which are maintained at a remarkably constant level throughout life (3, 4). Rather, immune dysfunction may result from a number of factors, including 1) changes in the composition of subsets in the T cell pool, leading to a narrowing of the repertoire; 2) loss of costimulatory molecules, resulting in signaling defects; and 3) changes in levels of those cytokines and chemokines that maintain homeostasis. In this study we investigated whether there is an underlying dysfunction in the maintenance of peripheral T cell pools by comparing proliferation and disappearance rates of different T cell subpopulations in a group of healthy elderly with those in a group of young control subjects using in vivo labeling of dividing cells with deuterated glucose.

Previous kinetic studies performed in young subjects have shown that the turnover rates of CD4⁺ and CD8⁺CD45RO⁺ memory subpopulations are higher than those of the CD45RA⁺ subpopulations (41). We found that the kinetics of proliferation and disappearance of CD4⁺CD45RA⁺, CD4⁺CD45RO⁺, and CD8⁺CD45RO⁺ subpopulations were strikingly similar between healthy young and elderly donors, implying that the maintenance of these subsets is not impaired with ageing.

In contrast, when the kinetic behavior of CD8⁺CD45RA⁺ cells from the two groups was compared, it appeared that whereas proliferation rates did not differ significantly, the rate of death/disappearance of this subset was greatly decreased in the elderly, implying that it represents a long-lived pool of cells that persist in the circulation. Although the CD8⁺CD45RA⁺ subset in the young exhibited a typically naive phenotype, this subpopulation in the elderly had a primed phenotype, as demonstrated by a reduction in numbers of CD27⁺, CD28⁺, and CCR7⁺ cells and an increase in the number of CD95⁺ and CD11a^high cells. The dissociation of proliferation from death/disappearance could therefore contribute to the accumulation of primed cells in elderly individuals. The low disappearance rate of the CD8⁺CD45RA⁺ subset in the elderly may be due in part to the fact that such primed cells are known to be resistant to apoptosis (47, 48).

Several explanations might account for the difference in kinetics between CD8⁺ CD45RA⁺ cells of the young and the elderly. The first is that primed CD8⁺CD45RA⁺ cells from donors of any age differ in kinetics from naive CD8⁺CD45RA⁺ cells. Alternatively, the difference in kinetics is an age-related effect. Lastly, it may be that CMV infection, which is known to induce large oligoclonal expansions and affect T cell differentiation (49), contributes to the different kinetics of young and elderly CD8⁺CD45RA⁺ cells. In any case, uncoupling of proliferation and death or disappearance from the circulation may contribute to the development of the oligoclonal expansions commonly seen in the elderly. Experiments to compare the kinetics of naive and primed CD8⁺CD45RA⁺ cells from the same donor or of CD8⁺ subsets from more CMV⁺ and CMV⁻ donors should resolve these issues in the future.

All the elderly donors in this study showed V_β TCR expansions in the CD8⁺ subpopulation, which have been commonly reported to arise in the elderly (29, 30), most often in response to persistent viral infections such as EBV/CMV (31, 47, 48). In this study five of six elderly were CMV positive, and a high proportion of their CD8⁺CD45RA⁺ cells exhibited a primed phenotype.

Examination of the kinetics of turnover of an oligoclonal expansion within the CD8⁺CD45RA⁺ subset from one elderly subject showed that the V_β⁺ subpopulation had a more rapid proliferation rate than V_β⁻ cells from this subset. This in combination with the fact that although cells within this clone had a primed phenotype, they also expressed high levels of Bcl-2 and CD45RB suggest that these clonally expanded cells are not terminally differentiated.

In summary we have applied recently developed methodology of deuterated glucose labeling to provide a direct comparison of the in vivo kinetics of lymphocyte turnover in healthy elderly and young. Our data indicate that, in general, the kinetic basis of homeostasis is not impaired in healthy elderly, because proliferation and disappearance rates for CD4⁺CD45RA⁺, CD4⁺ CD45RO⁺ and CD8⁺ CD45RO⁺ populations were similar in both young and elderly. However an age-related dissociation does appear within the CD8⁺CD45RA⁺ subpopulation in the elderly, where normal levels of proliferation become decoupled from normal death/disappearance rates, resulting in accumulation of long-lived, primed, TCR repertoire-restricted cells in the elderly and a paucity of true naive phenotype cells. Such changes are likely to be functionally important and may contribute to the age-related impairment of immune function.

We thank Dr. Becca Asquith for help with modeling, and Nadine Clifford for assistance and care of the subjects throughout the study period.