Latency for cytomegalovirus impacts T cell ageing significantly in elderly end-stage renal disease patients

Summary: The number of elderly patients with end‐stage renal disease (ESRD) has increased significantly during the last decade. Elderly ESRD patients are vulnerable to infectious complications because of an aged immune system. Additional immunological ageing effects may be derived from the uraemic environment and cytomegalovirus (CMV) latency. Elderly patients may be affected by these factors in particular, but data in this age group are limited. To assess the degree of immunological ageing and proliferative capacity of T lymphocytes, 49 elderly ESRD patients (defined as aged ≥ 65 years) on the renal transplantation waiting list were recruited and compared to 44 elderly healthy individuals (HI), matched for age and CMV serostatus. CMV latency was associated with more highly differentiated CD4+ and CD8+ T cells in both elderly HI and patients. Elderly CMV seropositive ESRD patients showed a substantial reduction in the number of naive CD4+ and CD8+ T cells compared with age‐ and CMV serostatus‐matched HI. Elderly ESRD patients also showed significantly decreased numbers of central memory CD4+ and CD8+ T cells compared with HI, independently of CMV serostatus. In addition, thymic output and relative telomere length of both CD4+ and CD8+ T cells were decreased in CMV seropositive ESRD patients compared with HI. The proliferative capacity of T cells was similar for patients and HI. Elderly ESRD patients have an advanced aged T cell compartment when compared to age‐matched healthy controls, which is driven mainly by CMV latency.

Elderly ESRD patients have an advanced aged T cell compartment when compared to age‐matched healthy controls, which is mainly driven by CMV latency. T cells from elderly CMV‐seropositive ESRD patients showed a decrease of thymic output, less naïve T cells and central memory phenotyping, in addition with shorter relative telomere length compared to CMV‐seropositive healthy individuals. This may be related to their vulnerability to infectious complications.

ageing; cytomegalovirus; end stage renal disease; T cells

The number of elderly patients (defined as aged ≥ 65 years) suffering from end‐stage renal disease (ESRD) keeps growing rapidly [1] . In the United States, during 1994–2004 patients aged more than 75 years increased by 67% compared to 24% for those aged between 5 and 74 years [2] . According to recent data from the Dutch renal replacement system (REgistratie NIerfunktievervanging NEderland, RENINE), the number of elderly ESRD patients (aged > 65 years) receiving renal replacement therapy (RRT) almost doubled from 2005 to 2015 (https://www.renine.nl/). Importantly, elderly ESRD patients are at high risk of developing serious infections [2] , [3] , [4] and show a poor response to vaccination [5] , [6] . Also, after successful kidney transplantation, elderly ESRD patients are more susceptible to infectious complications [7] , [8] . T cells are key players in the immune response to foreign antigens, such as those encountered during an infection and after vaccination.

With advanced ageing, the T cell‐mediated immune system undergoes dramatic changes [9] and loss of renal function is associated with a defective T cell‐mediated immune system [10] . We have demonstrated previously that ESRD‐related defects in T cell‐mediated immunity may be related to premature T cell ageing, as assessment of T cell receptor excision circle (TREC) content, T cell differentiation status and relative telomere length revealed a discrepancy of 15–20 years between the immunological age of the patients' T cells and their chronological age [11] , [12] .

Cytomegalovirus (CMV) may have a substantial impact on the composition and function of circulating T cells. Recent studies have shown that CMV latency expands the number of circulating CD8 T cells significantly by almost twofold [13] , promotes the emergence of highly differentiated T cell subsets [14] and may decrease T cell telomere length [15] in immune competent individuals. Depending on ethnicity, 65–100% of all elderly ESRD patients are CMV seropositive [16] , [17] . In young to middle‐aged ESRD patients, the additional effects of CMV latency on T cell ageing parameters are modest and confined mainly to the CD8+ T cells [18] .

However, little is known with respect to the impact of ESRD and CMV latency on the immunological age of the peripheral T cell compartment in elderly (≥ 65 years of age) ESRD patients. In this study, we show that CMV latency appears to be a dominant factor for the observed advanced immunological ageing of T cells from elderly ESRD patients compared to healthy age‐matched individuals.

Forty‐nine stable elderly (defined as ≥ 65 years) ESRD patients, defined as having a glomerular filtration rate of ≤ 15 ml/min with or without renal replacement therapy, and 44 elderly healthy individuals (HI) were included (study population characteristics are described in Table [NaN] ) from 1 November 2010 to 1 October 2013 at the out‐patient clinic. Patients with any clinical or laboratory evidence of acute bacterial or viral infection, malignancy or immunosuppressive drugs treatment within 28 days prior to transplantation (except glucocorticoids) were excluded. Lithium‐heparinized blood was drawn from ESRD patients and healthy kidney donors. All individuals included gave informed consent and the local medical ethical committee approved the study (METC number: 2012–022), which was conducted according to the principles of Declaration of Helsinki and in compliance with International Conference on Harmonization/Good Clinical Practice regulations.

Clinical and demographic characteristics of patients with end stage renal disease (ESRD) and healthy individuals

1 CMV = cytomegalovirus; Ig = immunoglobulin; HI = healthy individuals; n.s. = not significant.

Freshly drawn peripheral blood samples from 49 ESRD patients and 44 HI were stained and acquired on a fluorescence activated cell sorter (FACS)Canto II flow cytometer (BD Biosciences, Erembodegem, Belgium) to determine both absolute numbers and frequencies of the different T cell subsets, as described previously [11] , [19] . Data were analysed using FACS Diva software version 6.1.2 (BD Biosciences).

PBMCs from 11 elderly ESRD patients and 11 elderly CMV serostatus‐matched HI were isolated from peripheral blood, as described previously [20] . These PBMCs (responder cells) were labelled with carboxyfluorescein diacetate succinimidyl ester (CFSE), according to the manufacturer's instructions (Thermo Fisher Scientific, Waltham, MA, USA), and then co‐cultured in triplicate at 5 × 104/well with allogeneic PBMCs, autologous PBMCs (both irradiated at 40 gray) at a 1 : 1 ratio or with 5 µg/ml phytohaemagglutinin (PHA) (Sigma‐Aldrich, St Louis, MO, USA) as a positive control for 6 days. Culture medium consisted of RPMI‐1640 with GlutaMAX, 10% heat‐inactivated pooled human serum and 1% penicillin and streptomycin. After 6 days, PBMCs were harvested, pooled, washed and stained with AmCyan‐labelled CD3 (BD Pharmingen, Erembodegem, Belgium), Pacific Blue‐labelled CD4 (BD), allophycocyanin‐cyanin 7 (APC‐Cy7)‐labelled CD8 (BD Pharmingen); phycoerythin (PE)‐labelled CD28 (BD Pharmingen), APC‐labelled CD45RO (BD Pharmingen) and PE‐Cy7‐labelled CCR7 (R&D Systems, Uithoorn, the Netherlands) antibodies and a live–dead marker ViaProbe (7‐aminoactinomycin D; BD Pharmingen). Data were acquired on a FACSCanto II flow cytometer (BD Biosciences, Erembodegem, Belgium). Percentages of proliferating cells were analysed by Kaluza® software (Beckman Coulter, Brea, CA, USA). Kinetics of proliferation and precursor frequencies (PF), the latter defined as the proportion of cells present in the original sample being able to respond to the stimulus, were analysed by Modfit LT® software (Verity Software House, Topsham, ME, USA).

DNA was isolated from PMBCs using the QIAamp DNA Mini QIAcube kit, according to the manufacturer's instructions (Qiagen, Hilden, Germany). TREC content was measured by TaqMan quantitative polymerase chain reaction (PCR), as described previously [11] , [21] . The ΔCt was calculated by subtracting the Ct value for the albumin PCR from that of the TREC PCR. One/ΔCt was used to describe the TREC content of a sample. A Ct value greater than 41 for the TREC PCR was interpreted as the sample having an undetectable TREC content.

Flow fluorescence in‐situ hybridization was performed to determine the relative telomere length (RTL) of CD4+ and CD8+ T cells, as described previously [11] , [19] .

Statistical analyses were performed using spss version 20 (IBM, Chicago, IL, USA) and GraphPad Prism version 6 (GraphPad Software, La Jolla, CA, USA). Categorical variables were compared using the χ2 test or Fisher's exact test. Continuous variables were compared using the t‐test or the Mann–Whitney U‐test. All reported P‐values are two‐sided and were considered statistically significant when P < 0·05.

The demographic and clinical characteristics of the study population are given in Table [NaN] . Forty‐nine ESRD patients (aged 65–79 years) and 44 age‐matched HI (aged 65–89 years) were recruited into this study. The elderly ESRD patients consisted of a higher proportion (69.4%, P = 0·02) of males compared to the HI (45·5%). Approximately half the ESRD patients received RRT. CMV‐immunoglobulin (Ig)G seropositivity was present in 59·2% of elderly ESRD patients and in 63·6% of HI in this study.

The effect of CMV on absolute T cell numbers and composition was confined mainly to the memory compartment. Elderly CMV seropositive ESRD patients, but not HI, had significantly more total memory CD4+ T cells compared to their CMV seronegative counterparts (Fig. [NaN] a). Within the CD4+ memory compartment, CMV seropositivity was associated with increased numbers of EM (Fig. [NaN] e) and CD4+CD28– T cells in elderly ESRD patients (Fig. [NaN] g). The association of CMV seropositivity with higher numbers of CD4+CD28– T cells was also observed in elderly HI (Fig. [NaN] g). Elderly CMV seropositive ESRD patients had lower numbers of total (Fig. [NaN] a) and naive (Fig. [NaN] b) CD4+ T cells than CMV seropositive HI. Moreover, central memory (CM) CD4+ T cells were lower in elderly ESRD patients compared to HI, irrespective of their CMV serostatus (Fig. [NaN] d). Frequencies of T cell subsets also indicated CMV latency and ESRD to induce a more differentiated T cell compartment. CMV seropositivity was associated with lower percentages of CD4+ T cells (Supporting information, Fig. S1a) and higher percentages of CD4+CD28– T cells (Supporting information, Fig. S1g). Percentages of naive CD4+ T cells were lower in CMV seropositive compared with CMV seronegative ESRD patients (Supporting information, Fig. S1b) and within the memory compartment, higher percentages of EM were observed in CMV seropositive ESRD patients compared with CMV seronegative patients (Supporting information, Fig. S1e). In agreement with the absolute number of T cells, the differences of T cell differentiation analysed as percentage between elderly ESRD patients and HI were observed in the CMV seropositive group. Within the CMV seropositive group, ESRD patients had lower percentages of naive CD4+ T cells (Supporting information, Fig. S1b) and higher percentages of EM CD4+ T cells compared to HI (Supporting information, Fig. S1e).

Within the CD8+ T cell compartment, CMV latency induced a strong increase in total numbers of CD8+ T cells of both elderly ESRD patients as well as HI. The median CD8+ T cell numbers in CMV seropositive ESRD patients amounted to 373 cells/µl, which was a > 1·5‐fold increase compared to the median of CD8+ in CMV seronegative patients (214 cells/µl, P < 0·001) (Fig. [NaN] a). A more than twofold increase in numbers of CD8+ T cells was noted for CMV seropositive HI compared to CMV seronegative HI (476 cells/µl versus 217 cells/µl, P < 0·001) (Fig. [NaN] a). The increase in CD8+ T cells induced by CMV was due mainly to an increase in memory CD8+ T cells in both elderly ESRD patients and HI (Fig. [NaN] c). Within the memory compartment, the number of EM was significantly higher in CMV seropositive HI compared to CMV seronegative HI, and a similar trend was found in patients (Fig. [NaN] e). Increased numbers of highly differentiated T cell subsets including EMRA and CD8+CD28– were observed in both elderly ESRD patients and HI (Fig. [NaN] f,g). Similar to the CD4+ T cell compartment, elderly CMV seropositive ESRD patients also had significantly lower numbers of naive CD8+ T cells when compared to CMV serostatus‐matched HI. In addition, also within the CD8+ T cell compartment, lower numbers of CM CD8+ T cells were observed, irrespective of CMV serostatus when comparing elderly ESRD patients to HI. Comparison of frequencies of CD8+ T cell subsets revealed CMV seropositivity to be associated with higher frequencies of total CD8+, EMRA and CD28–CD8+ T cells (Supporting information, Fig. S2a,f,g, respectively) and lower frequencies of CM CD8+ T cells (Supporting information, Fig. S2d). In addition, percentages of naive CD8+ T cells were lower in CMV seropositive compared with CMV seronegative ESRD patients (Supporting information, Fig. S2b). Furthermore, higher frequencies of total memory (Supporting information, Fig. S2c), and within this lower frequencies of EM CD8+ T cells (Supporting information, Fig. S2e) were noted. Within the CMV seropositive group, ESRD patients had higher percentages of total, EMRA and CD28–CD8+ T cells than HI (Supporting information, Fig. 2a,f,g, respectively) and lower frequencies of CM CD8+ T cells (Supporting information, Fig. S2d). Both RRT (Supporting information, Table S1) and gender (data not shown) did not affect numbers of circulating T cell subsets in these elderly ESRD patients.

TREC content was comparable for CMV seronegative and CMV seropositive ESRD patients or HI (Fig. [NaN] a), confirming the idea that CMV effects are more limited to the memory compartment. Interestingly, in 4·6% and 10·2% of the elderly ESRD patients and healthy individuals, respectively, no DNA encoding for TREC was detected in the PCR assay, which could be related to the lower contribution of the thymus to the naive T cell pool at this age. Of note, in those cases in which DNA encoding for TRECs was detected, a significant (P = 0·046) decrease was observed for TREC content in elderly CMV seropositive but not CMV seronegative ESRD patients compared to CMV serostatus‐matched HI (Fig. [NaN] a). In agreement with this, a lower number of recent thymic emigrants, defined as CD31‐expressing naive CD4+ (Fig. [NaN] b: 88 cells/µl versus 146 cells/µl) and CD8+ T cells (Fig. [NaN] c: 35 cells/µl versus 48 cells/µl), was observed in elderly CMV seropositive ESRD patients compared to CMV serostatus‐matched HI. Thymic output as measured by TREC content and CD31‐expressing naive T cells was not influenced by RRT (Supporting information, Table S1) and gender (data not shown).

RTL was not significantly different comparing CMV seropositive ESRD patients or HI to their CMV seronegative counterparts. CMV seropositive, but not seronegative, elderly ESRD patients had shorter telomeres within CD4+ (Fig. [NaN] a, P < 0·001) and CD8+ (Fig. [NaN] b, P < 0·001) T cells than CMV serostatus‐matched HI. The median RTL of CMV seropositive ESRD patients amounted to 9·0 and 9·1% for CD4+ and CD8+ T cells and values observed in CMV serostatus‐matched HI were 16·8 and 14·9% for CD4+ and CD8+ T cells, respectively. No significant difference in RTL of CD4+ or CD8+ was observed between patients with RRT and without RRT (Supporting information, Table S1). The RTL of CD4+ or CD8+ was not influenced significantly by gender in our elderly population (data not shown).

The proliferative capacity as in percentages of proliferating CD4+ and CD8+ T cells in response to an allogeneic stimulus (Fig. [NaN] a,b), as well as a polyclonal stimulus (PHA; Supporting information, Fig. S3a,b) was equal between elderly ESRD patients and elderly HI. In addition, a similar precursor frequency (PF) of CD4+ and CD8+ T cells able to respond to alloantigen (Fig. [NaN] c,d) or a polyclonal stimulus (PHA; Supporting information, Fig. S3c,d) was observed between elderly patients and HI. Moreover, no differences were observed with respect to proliferation kinetics of CD4+ (Fig. [NaN] e) and CD8+ (Fig. [NaN] f) T cells in response to alloantigen‐stimulation. CMV did not influence significantly the capacity of T cells to respond to allogeneic or polyclonal stimulation in both elderly ESRD patients and HI (Supporting information, Table S2).

The main observation of this study is that CMV latency is a dominant factor for increased ageing of the peripheral T cells in elderly ESRD patients, outweighing the known premature T cell ageing effects of renal failure itself. In our previous study, premature ageing of peripheral T cells was demonstrated in ESRD patients but did not consider CMV latency and did not focus on elderly (≥ 65 years of age) patients [11] . The influence of CMV latency versus uraemia on T cell ageing was investigated in another cohort of young to middle‐aged ESRD patients and showed a modest effect consisting of increased T cell differentiation status, in particular higher percentages of CD28‐negative T cells, and reduced telomere length of CD8‐positive T cells [18] . The current study focused on elderly ESRD patients and identified specific additive effects of ESRD and in particular CMV latency on the ageing of the T cell system in the elderly population.

CMV latency is recognized increasingly as a significant factor for accelerated T cell ageing [22] , and as such may add to the increased risk for infections [23] as well as cardiovascular disease [24] in healthy elderly people. In elderly ESRD patients, the risk of cardiovascular disease events and death [16] , [25] , [26] , [27] or infections [28] is even more increased. Studies in very healthy elderly people demonstrated an immune risk phenotype (IRP) for increased mortality, defined by an inverted CD4/CD8 ratio and increased number of CD28–CD8+ T cells [29] , which was associated with CMV seropositivity [13] , [30] .

Our data indicate that CMV latency in combination with ESRD in elderly people is particularly harmful to the T cell system, as numbers of naive T cells are also affected negatively, as well as the known ageing effects on memory T cells. The decline in the number of naive T cells is a key feature associated with loss of renal function, and in particular ESRD [11] , [31] . Naive T cells that have recently left the thymus contain TRECs and express mainly CD31 [Platelet and Endothelial Cell Adhesion Molecule 1 (PECAM‐1)] [32] . TRECs were not detectable in several elderly healthy individuals or ESRD patients, suggestive of a low thymic output in the elderly population. This is in agreement with the observation generated in healthy individuals that a large part of the functional thymic tissue has been lost by the age of 50 years [33] . Aside from the thymus contributing to the naive T cell pool, homeostatic proliferation of the remaining naive T cells is able to maintain the naive T cell pool [34] . Homeostatic proliferation of naive T cells may occur in response to homeostatic cytokines such as, for example, IL‐7 [35] or in response to low‐affinity self‐antigens [36] , [37] , [38] . The decline in naive T cells induced by ESRD in elderly people might also be the result of defects in homeostatic proliferation, as plasma levels of IL‐7 were lower in ESRD patients compared to healthy individuals [31] . Moreover, the decline in naive T cells could also result from differentiation towards memory T cells. The memory compartment in the ESRD patients is more differentiated, i.e. containing fewer CM T cells [31] , [39] . Naive, but also CM, T cells are essential for generating a robust immune response [3] , [4] and naive T cells contain a more diverse T cell receptor (TCR) Vβ repertoire compared to memory T cells [40] , allowing for a better response to newly encountered antigens such as vaccination antigens. Low in‐vivo numbers of naive CD4+ recent thymic emigrants correlated well with reduced acute responsiveness and altered long‐term persistence of human cellular immunity to yellow fever vaccination in the elderly population [41] . The underlying mechanism for the reduction in naive and CM T cells in the peripheral blood is not yet clear, but may involve increased apoptosis [39] , [42] , [43] , [44] or enhanced proliferation to more differentiated T cell subsets [11] . During ageing, naive and CM T cells have been linked to increased sensitivity of tumour necrosis factor (TNF)‐α‐induced apoptosis [45] , [46] . Elevated concentrations of serum or plasma TNF‐α is associated strongly with progressive loss of renal function [47] , [48] . TNF‐α‐induced apoptosis may be an explanation of the loss of naive and CM in elderly ESRD patients. CMV may reactivate in healthy individuals and more frequently in the elderly [49] . This may be caused by an age‐related decrease in T cell‐mediated control of CMV reactivation as, e.g. IFN‐γ secretion in response to CMV peptide stimulation is decreased in very elderly individuals [50] . Data in ESRD patients on CMV reactivation are largely absent, but the frequent presence of anti‐CMV IgM titres in dialysis patients suggests that CMV reactivation is not a rare event [51] , and may even lead to CMV disease [52] , [53] . In addition, anti‐CMV IgG titres are increased in elderly ESRD patients but not in healthy elderly individuals, which may also be interpreted as the result of frequent CMV reactivation [54] .

A plausible hypothesis could be that ESRD‐related premature T cell ageing contributes to a decrease in anti‐viral T cell immunity, which allows for more frequent CMV reactivation. Reactivation is controlled at the expense of expanded populations of CD28null T cells with reduced telomere length [18] , fewer naive T cells and a narrowed TCR repertoire [55] . The expansion of CD4CD28null T cells adds to the increased risk for atherosclerotic disease, while the overall antigen‐specific T cell response is weakened further by the loss of T cell diversity.

Unexpectedly, we were unable to attribute the phenotypical defects to specific or more general functional deficits using CFSE dilution as a read‐out for the proliferation of PBMC in response to an alloantigen or polyclonal stimulus. This indicates that the overall proliferative potential of T cells is not affected severely in elderly ESRD patients. However, the stimuli used in our proliferation assay do not allow for measuring the potential of T cells to initiate responses to newly encountered antigens that are more dependent on a diverse T cell repertoire. In addition, the use of CMV seropositive donors as allogeneic stimuli might result in activation of CMV‐specific T cells in CMV seropositive responders in addition to the alloreactive T cells [56] , [57] . This might be an explanation for the trend in higher frequencies of proliferating T cells, as well as precursor frequencies comparing CMV seropositive responders to their negative counterparts (Supporting information, Table S2). By using high‐throughput sequencing of the diversity of the TCR repertoire [40] , [58] this might provide a more supportive functional read‐out.

In conclusion, CMV latency is a dominant factor for accelerated T cell ageing in elderly ESRD patients, and therefore should be taken into consideration to evaluate the risk of mortality, infection and response to vaccination in this patient population.

L. H. performed the experiments, analysed the data and wrote the manuscript. A. L. participated in the design of the study, interpretation of the data and revision of the manuscript. C. B. revised the manuscript. N. L. and M. B. designed the study, interpreted the data and revised the manuscript. All authors read and approved the final manuscript. The research was supported by the China Scholarship Council for funding PhD fellowship to Ling Huang (File no. 201307720043).

The authors have no financial or commercial conflicts of interest to disclose.

Additional Supporting information may be found in the online version of this article at the publisher's web‐site:

Graph: Absolute numbers of circulating CD4+ T cell subsets in elderly healthy individuals (HI) and elderly end stage renal disease (ESRD) patients. Numbers of (a) CD4+, (b) CD4+ naive, (c) CD4+ memory, (d) CD4+ central memory (CM), (e) CD4+ effector memory (EM) and (f) CD4+ highly differentiated effector memory (EMRA) and (g) CD28–CD4+ T cells in HI (n = 45; n = 16 cytomegalovirus (CMV)‐seronegative and n = 29 CMV seropositive) and ESRD patients (n = 49; n = 20 CMV seronegative and n = 29 CMV seropositive) was determined and dissected for CMV serostatus. Data are given as median with interquartile range. The open bars represent the CMV seronegative individuals and the closed bars that of CMV seropositive ones. P‐value: *< 0·05; n.s.: not significant.

Graph: Absolute numbers of circulating of CD8+ T cell subsets in elderly healthy individuals (HI) and elderly end stage renal disease (ESRD) patients. Numbers of (a) CD8+, (b) CD8+ naive, (c) CD8+ memory, (d) CD8+ central memory (CM), (e) CD8+ effector memory (EM) and (f) CD8+ highly differentiated effector memory (EMRA) and (g) CD28–CD8+ T cells in HI [n = 45; n = 16 cytomegalovirus (CMV) seronegative and n = 29 CMV seropositive] and ESRD patients (n = 49; n = 20 CMV seronegative and n = 29 CMV seropositive) were determined and dissected for CMV serostatus. Data are given as median with interquartile range. The open bars represent the CMV seronegative individuals and the closed bars that of CMV seropositive ones. P‐value: *< 0·05; **< 0·01; n.s.: not significant.

Graph: T cell receptor excision circle (TREC) content and CD31‐expressing naive CD4+ and CD8+ T cells in elderly healthy individuals (HI) and end stage renal disease (ESRD) patients. The (a) TREC content (HI: n = 39; ESRD patients: n = 43) and absolute number of CD31‐expressing naive (b) CD4+ and (c) CD8+ T cells (HI: n = 44; ESRD patients: n = 49) in elderly HI and ESRD patients was determined and dissected for cytomegalovirus (CMV) serostatus (open bars represent CMV seronegative and closed bars CMV seropositive individuals). Data are given as median with interquartile range. P‐value: *< 0·05; n.s.: not significant.

Graph: Relative telomere length (RTL) of CD4+ and CD8+ T cells in elderly healthy individuals (HI) and elderly end stage renal disease (ESRD) patients. The RTL of (a) CD4+ and (b) CD8+ T cells was determined in circulating T cells of HI [n = 36; n = 14 cytomegalovirus (CMV) seronegative and n = 22 CMV seropositive] and ESRD patients (n = 28; n = 12 CMV seronegative and n = 16 CMV seropositive). The open bars represent the CMV seronegative individuals and the closed bars that of CMV seropositive ones. Data are given as median with interquartile range. P‐value: ***< 0·001; n.s.: not significant.

Graph: Proliferation of CD4+ and CD8+ T cells in response to allogeneic stimulation in elderly healthy individuals (HI) and elderly end stage renal disease (ESRD) patients. Percentage of (a) dividing CD4+ and (b) CD8+ T cells in response to alloantigens. Precursor frequency (%) of (c) CD4+ and (d) CD8+ T cells in response to alloantigens. Proliferation kinetics of (e) CD4+ and (f) CD8+ T cells in response to alloantigens. Peripheral blood mononuclear cells (PBMCs) isolated from 11 HI and 11 ESRD patients [five cytomegalovirus (CMV) seronegativity and six CMV seropositivity] were used as response cells and the irradiated PBMCs from the third part were used as the allostimulation cells. Data are given as individual values (a–d) and median with interquartile range (a–f); open symbols/bars represent HI and closed symbols/bars the ESRD patients; n.s.: not significant.

Graph: Fig. S1. Frequencies of CD4+ T cell subsets in elderly healthy individuals (HI) and elderly end stage renal disease (ESRD) patients. Percentages of (a) CD4+ within CD3+ T cells, (b) naive, (c) memory, (d) central memory (CM), (e) effector memory (EM), (f) highly differentiated effector memory (EMRA) and (g) CD28– T cell subsets within CD4+ T cells, and (h) CD31+ T cells within CD4+ naive T cells in HI [n = 45; n = 16 cytomegalovirus (CMV) seronegative and n = 29 cytomegalovirus (CMV) seropositive] and ESRD patients (n = 49; n = 20 CMV seronegative and n = 29 CMV seropositive) was determined and dissected for CMV serostatus. Data are given as median with interquartile range. The open bars represent the CMV seronegative individuals and the closed bars that of CMV seropositive ones. P‐value: *< 0·05; **< 0·01; ***< 0·001; n.s.: not significant. Fig. S2. Frequencies of CD8+ T cell subsets in elderly healthy individuals (HI) and elderly end stage renal disease (ESRD) patients. Percentages of (a) CD8+ within CD3+ T cells, (b) naive, (c) memory, (d) central memory (CM), (e) effector memory (EM), (f) highly differentiated effector memory (EMRA) and (g) CD28– T cell subsets within CD8+ T cells, and (h) CD31+ T cells within CD8+ naive T cells in HI [n = 45; n = 16 cytomegalovirus (CMV) seronegative and n = 29 CMV seropositive] and ESRD patients (n = 49; n = 20 CMV seronegative and n = 29 CMV seropositive) was determined and dissected for CMV serostatus. Data are given as median with interquartile range. The open bars represent the CMV seronegative individuals and the closed bars that of CMV seropositive ones. P‐value: *< 0·05; **< 0·01; ***< 0·001; n.s.: not significant. Fig. S3. Proliferation of CD4+ and CD8+ T cells in response to phytohaemagglutinin (PHA) stimulation in elderly healthy individuals (HI) and elderly end stage renal disease (ESRD) patients. Percentage of (a) dividing CD4+ and (b) CD8+ T cells in response to PHA. Precursor frequency (%) of (c) CD4+ and (d) CD8+ T cells in response to PHA. Peripheral blood mononuclear cells (PBMCs) isolated from 11 HI and 11 ESRD patients [five cytomegalovirus (CMV) seronegativity and six CMV seropositivity] were used as response cells. Data are given as individual values; open symbols/bars represent HI and closed symbols/bars the ESRD patients; n.s.: not significant. Table S1. T cell ageing parameters of elderly elderly end stage renal disease (ESRD) patients with and without renal replacement therapy Table S2. Proliferation of CD4+ and CD8+ T cells in response to allogeneic and phytohaemagglutinin (PHA) stimulation in elderly healthy individuals (HI) and elderly end stage renal disease (ESRD) patients dissected for cytomegalovirus (CMV) serostatus