- Email: [email protected]
Expansion of cytolytic CD4+CD28− T cells in end-stage renal disease
original article
http://www.kidney-international.org & 2008 International Society of Nephrology
Expansion of cytolytic CD4 þ CD28 T cells in end-stage renal disease Michiel G.H. Betjes1, Martin Huisman1, Willem Weimar1 and Nicolle H.R. Litjens1 1
Division of Nephrology, Department of Internal Medicine, Erasmus Medical Center, Rotterdam, The Netherlands
Cytomegalovirus (CMV) seropositivity is associated with increased risk for atherosclerotic disease in patients with end-stage renal disease. This association is due to a unique peripheral blood CD4 þ T cell population which lack CD28 (CD4 þ CD28 T cells). Here we found that this patient population has a significant age-dependent increase of CD4 þ CD28 T cells that comprise over half of the circulating CD4 T cells in some. Patients over 50 years of age have a 50-fold higher percentage of CD4 þ CD28 T cells compared to seronegative patients and a 5-fold higher percentage when compared to seropositive healthy controls. Stimulation by CMV-antigen or by polyclonal stimulation using PMA and ionomycin showed that CD4 þ CD28 cells in patients with end stage renal disease degranulated and secreted interferon c thus indicating that they are cytolytic. The average anti-CMV IgG titer displayed a remarkable age-dependent increase only in patients with end stage renal disease. These findings are highly suggestive of repetitive antigenic stimulation of the immune system in patients with end stage disease by subclinical CMV reactivation which might contribute to increased atherosclerotic risk. Kidney International (2008) 74, 760–767; doi:10.1038/ki.2008.301; published online 9 July 2008 KEYWORDS: cytomegalovirus; end-stage renal disease; CD4 þ T cell; CD28; atherosclerosis
Correspondence: Michiel G.H. Betjes, Division of Nephrology, Department of Internal Medicine, Erasmus Medical Center, Dr Molewater plein 40, 3015 GD Rotterdam, The Netherlands. E-mail: [email protected] Received 23 January 2008; revised 27 March 2008; accepted 23 April 2008; published online 9 July 2008 760
Patients with end-stage renal disease (ESRD) carry a greatly increased risk for cardiovascular disease.1,2 Traditional risk factors such as smoking, hypertension, and hypercholesterolemia can be identified, but do not explain the full magnitude of the increment in risk.3 We have recently showed that latency for cytomegalovirus (CMV) is an unexpectedly high and independent risk factor for atherosclerosis in patients with ESRD.4 In addition, analysis of the T-cell subsets in patients with chronic kidney disease indicated that numbers of CMV-specific T cells increased as the stage of renal failure progressed.5 CMV infection is unique as it leaves a clear footprint in the T-cell repertoire. It has been estimated that on average, as high as 10% of the circulating CD4-positive T cells (CD4pos T cells) are CMV-specific in the latent situation.6 The phenotype of CD4-positive CMV-specific T cells is interesting with respect to atherosclerosis, as CMV infection increases the frequency of circulating CD4-positive CD28negative (CD4posCD28null) T cells.7–9 The circulating CD4posCD28null T-cell subset is enriched in the blood of patients with acute coronary syndromes and typically produces high levels of interferon-g (IFN-g).10 Plaque-infiltrating CD4 T cells, which are associated with plaque vulnerability for rupture, are likely recruited from this circulating CD4posCD28null T-cell subset.11 Apoptotic death of smooth muscle cells may be a mechanism of plaque destabilization, transforming stable plaques into ruptureprone lesions. Recently, it has been shown that CD4pos CD28null plaque-infiltrating T cells can indeed directly kill endothelial cells and vascular smooth muscle cells by a T-cell receptor-independent mechanism.12 These findings have given experimental support for the pathogenetic role of this specific T-cell subset in atherosclerotic plaque destabilization. We hypothesized that the CD4posCD28null T-cell population is increased in patients with ESRD compared with healthy controls, thereby imposing a risk for plaque rupture. Therefore, we quantified the circulating CD4posCD28null T cells and analyzed their phenotype, expression of cytolytic molecules, and degranulation capacity. The results show that the presence of circulating CD4posCD28null T cells is unique for CMV-seropositive healthy individuals and ESRD patients. However, elderly patients with ESRD have a massively expanded population of circulating cytolytic CD4posCD28null T cells. These findings may explain why CMV Kidney International (2008) 74, 760–767
original article
RESULTS CD4posCD28null T cells are greatly expanded in CMV-seropositive ESRD patients
In accordance with previous studies,5 the ESRD patients showed a lower number of CD4-positive T cells compared with healthy individuals (381 106/l versus 691 106/l, Po0.01). The frequency of the CD4posCD28null T cells was evidently associated with CMV infection. These cells were rarely observed (median frequency of 0.2% of the total CD4pos T cells) in CMV-seronegative individuals (Figure 1). The frequency and absolute number of CD4posCD28null T cells increased with age in CMV-seropositive individuals (Figure 2a and b). However, these cells were greatly expanded in CMV-seropositive ESRD patients, and comprised in some individuals up to 65% of the total CD4pos T-cell population. The expansion of the CD4posCD28null T-cell population was most evident in CMV-seropositive ESRD patients over 50 years of age. Compared to age-matched seronegative ESRD patients, they had on average a 50 times increased percentage of CD4posCD28null T cells. Compared to elderly CMVseropositive healthy controls, these patients had on average a five times increased percentage (3 versus 15% of total CD4pos T cells) and an almost fourfold increased absolute number of CD4posCD28null T cells (61 106/l versus 16 106/l, Po0.05). The expansion of CD4posCD28null T cells differed substantially among ESRD patients. However, except for age, no other demographic characteristics or clinical characteristics (listed in Table 1), for example type of renal replacement therapy or diabetes, were associated with the number of CD4posCD28null T cells. Remarkably, an age-dependent
CD4+ CD28null T lymphocytes (10 6/l)
seropositivity is associated with an increased risk factor for atherosclerotic disease, especially in ESRD patients.
CD4+ CD28null T lymphocytes (%)
MGH Betjes et al.: Cytolytic CD4 T cells and ESRD
70 60
0.010
50 0.003
40 0.015
30 20 10 0
⭐ 40
41 – 50 51 – 60 Age (years)
61 – 75
0.031
350 300 250
0.029
200 150
0.038
100 50 0
⭐ 40
41 – 50 51 – 60 Age (years)
61 – 75
CMV-seropositive healthy controls CMV-seropositive ESRD patients
Figure 2 | CD4posCD28null T cells and age. CMV-seropositive ESRD patients (closed symbols, N ¼ 70) and age-matched healthy controls (open symbols, N ¼ 70) were divided into age groups (p40, n ¼ 9; 41–50, n ¼ 17; 51–60, n ¼ 18; 60–75, n ¼ 26) and either percentages (a) or absolute numbers expressed as 106/l (b) of CD4posCD28null T cells were shown. The Bonferroni-corrected P-value for statistical significance is Po0.006.
P = 0.000
CD4 +CD28null (%)
70
Table 1 | Clinical and demographic characteristics of CMV-seropositive patients with end-stage renal disease and CMV-seropositive healthy controls
50 30 10 10 5 0 Healthy controls
ESRD patients
Closed symbols = CMV-seronegative individuals Open symbols = CMV-seropositive individuals
Figure 1 | CD4posCD28null T cells and CMV seropositivity. Percentages of CD4posCD28null T cells of the total number of circulating CD4pos T cells, comparing CMV-seronegative to CMV-seropositive individuals. The CMV-seronegative groups included 30 healthy controls and 30 patients with ESRD. The CMV-seropositive groups included 72 healthy controls and 72 ESRD patients. Kidney International (2008) 74, 760–767
Number of individuals Age Male Non-Western European origin Diabetes mellitus Hypertension History of atherosclerotic disease Underlying kidney disease Hypertensive nephropathy Primary glomerulopathy Diabetic nephropathy Polycystic kidney disease Other Unknown
Patients
Healthy controls
72 54 (21–75)a 68.0% 44.0% 30.0% 84.7% 38.5%
72 55 (23-71)a 27.8%** 30.5% 0.0%** 16.6%** 0.0%**
31.9% 20.8% 18.1% 11.1% 13.9% 4.2%
CMV, cytomegalovirus.[2]**P-value o0.001. a Data are given in median years with range in parentheses.
761
original article
MGH Betjes et al.: Cytolytic CD4 T cells and ESRD
increase of the average anti-CMV IgG titer was observed for ESRD patients, but not for healthy individuals (Figure 3). CD4posCD28null T cells are CCR7-negative and effector memory cells
CMV-specific IgG (AU/ml)
The expression of the CD28 molecule on the cell surface of CD4pos T cells yields a clear distinction between a CD28pos and CD28null population (Figure 4a). The circulating T cells can be subdivided into naive T cells and memory T cells by differential expression of the chemokine receptor CCR7 and CD45RO. The memory cells are able to either display immediate effector function in inflamed tissues (effector
Granzyme B- and perforin-positive CD4posCD28null T cells are significantly increased in ESRD patients
100
50
0
memory T cells (Tem)) or constitute a pool of long-lived memory cells that develop more quickly into effector cells following antigenic stimulation in lymphoid organs (central memory T cells (Tcm)). More than 97% of the CD4posCD28null T cells were CCR7 negative, with the majority of the cells (on average 75%) expressing CD45RO, and thus classified as classical effector memory T cells (Figure 4c and d). CD4posCD28null T cells without expression of CD45RO were observed only in CMV-seropositive individuals, similar to the previously reported CD45RO-negative, CMV-specific, CD8-positive effector T cells.13
⭐ 40
41 – 50 51 – 60 Age (years)
61 – 75
CMV-seropositive healthy controls CMV-seropositive ESRD patients
Figure 3 | Age-related increase in CMV-specific IgG titers for ESRD patients but not healthy controls. Mean (±s.e.m.) CMV-specific IgG titers are depicted on the Y axis and expressed as AU/ml, whereas on the X axis, the different age groups are displayed. Open symbols represent the healthy controls and the closed ones, the ESRD patients. A significant (Po0.01, Spearman’s rank correlation coefficient) age-dependent increase was noticed for ESRD patients (N ¼ 70) but not for healthy controls (N ¼ 66).
Expression of the cytotoxic molecules granzyme B and perforin was limited to the CD4posCD28null T cells (Table 2). Specifically, the percentage of granzyme B-positive cells within the CD4posCD28null T-cell population was higher for ESRD patients compared with healthy individuals (65 versus 52%, Po0.01). Next, we calculated the total number of granzyme B- and perforin-positive CD4posCD28null T cells and the percentage of these cells in the total CD4pos T-cell population. A significantly increased percentage and total number of granzyme B- and percentage of perforinpositive CD4posCD28null T cells was observed in ESRD patients 50 years and older (Figure 5). CD4posCD28null T cells are functional as they degranulate and produce IFN-c after CMV-antigen and polyclonal stimulation
Upon stimulation with CMV-antigen, the CD4posCD28null T cells degranulated, as shown by expression of membranebound CD107a, and produced IFN-g. This was observed at a
Table 2 | Percentages of T cells with cytotoxic potential and effector function in the CD4+CD28null and the CD4+CD28+ subsets CD4+CD28null T cells
CD4+CD28+ T cells
ESRD patients
Healthy controls
P-value1
ESRD patients
Healthy controls
P-value2
60 65.1±3.8b 7.8±1.3
60 51.6±4.3 10.8±1.8
0.009 0.152
60 4.8±1.3 0.6±0.1
60 1.7±0.3 0.2±0.04
0.000 0.006
CD107a CMV specific PMA/ionomycin
1.7±0.3 6.5±1.4
1.0±0.2 7.6±0.4
0.050 0.671
0.3±0.1 3.5±1.0
0.04±0.01 4.2±0.9
0.003 0.396
IFN-g CMV specific PMA/ionomycin
4.7±1.0 65.2±5.2
4.1±1.3 36.1±8.9
0.175 0.021
0.6±0.2 21.9±3.0
0.2±0.04 12.5±2.0
0.338 0.016
CD154 CMV specific PMA/ionomycin
2.7±0.7 3.1±0.4
2.7±0.9 3.1±0.8
0.319 0.790
0.4±0.2 11.5±2.1
0.2±0.03 12.5±2.5
0.496 0.657
Number Granzyme Ba Perforina
CMV, cytomegalovirus; ESRD, end-stage renal disease; IFN-g, interferon-g; PMA, phorbol 12-myristate 13-acetate.[2]Data of patients with end-stage renal disease and agematched healthy controls are shown. P-value1 and P-value2 for the differences between data of CD28null T cells and CD28positive cells, respectively, comparing ESRD patients with healthy controls. The Bonferroni-corrected level of statistical significance is P-value o0.003. a Expression determined using unstimulated PBMCs. b Data are expressed in mean percentages±standard error of the means.
762
Kidney International (2008) 74, 760–767
original article
MGH Betjes et al.: Cytolytic CD4 T cells and ESRD
b 35.4
51
CD28
88.4
CCR7
a 1.3
12.3
c 0.3
0
11.6 CD4 9.7
90 CD45RO
CD4+CD28null distribution
d
0.001
CD4+CD28null (%)
70 50 30 10 10 5 0 Tnaive
Tcm
Tem
Open symbols: CMV-seropositive healthy controls Closed symbols: CMV-seropositive ESRD patients
Figure 4 | CD28 expression within the different CD4pos T-cell subsets of a CMV-seropositive individual. A representative example of the flow cytometric analysis of CD4pos T cells for a CMV-seropositive individual is shown in (a–c). Panel a shows a typical example of CD28 expression within CD4-positive peripheral blood T cells. The CD28pos (b) and CD28null (c) CD4pos T-cell subsets are further characterized by their expression of CD45RO and CCR7. Using these markers, the following T-cell subsets can be discriminated: naive T cells (Tnaive, CCR7 positive and CD45RO negative), central memory T cells (Tcm, CCR7 positive and CD45RO positive), and effector memory T cells (Tem, CCR7 negative and CD45RO positive). The overall distribution of CD4posCD28null T cells of ESRD patients (n ¼ 70) and age-matched healthy controls (n ¼ 66) is displayed in (d). The Bonferroni-corrected P-value for statistical significance is Po0.02.
significantly lower level in the CD4posCD28pos T cells (Table 2). After a strong polyclonal stimulus such as phorbol 12-myristate 13-acetate (PMA)/ionomycin, IFN-g production was predominantly observed within the subset of CD4posCD28null T cells (Table 2). The percentages and number of CD4posCD28null T cells expressing CD107 and IFN-g after CMV (Figure 6) or polyclonal stimulation were significantly higher in elderly ESRD patients, compared with the age-matched healthy controls. The expression of CD40 ligand (CD154) after antigenic stimulation is an important part of the costimulatory pathway during antigen presentation. It can be Kidney International (2008) 74, 760–767
used as a sensitive and specific method for identifying antigen-specific T cells.14 However, the percentage of CD4posCD28null T cells expressing CD154 after stimulation was always significantly lower (on average 50%) than the percentage of IFN-g-positive cells (Table 2). This observation indicates that CD4posCD28null T cells have become largely independent of the CD40-CD40L co-stimulation pathway. DISCUSSION
The CD4posCD28null circulating T-cell population has been associated with both atherosclerotic disease and CMV infection.6–11,15 The results of this study show that CD4posCD28null T cells are rarely observed in CMVseronegative individuals, but may expand to as much as 65% of the total CD4pos T-cell population in elderly CMVseropositive ESRD patients. This expansion of CD4posCD28null T cells is in sharp contrast with the well-documented uremia and age-related lymphopenia of the CD4pos T-cell compartment.5,9 An age-related increased number of CMVspecific T cells has previously been reported in healthy adults, but only in the very elderly.9,16 The published clinical and experimental data suggest a role for CD4posCD28null T cells in atherosclerotic plaque destabilization,10–12 and in a previous study we found that CMV infection was significantly and strongly associated with atherosclerotic disease in ESRD patients.4 It is evident that the presence of atherosclerosis is not synonymous with plaque rupture, which may cause an acute vascular event. As such, the present data cannot be directly correlated with the findings of our previous clinical study.4 However, many patients with a positive history for atherosclerotic disease probably have experienced a plaque rupture, as they had a history of myocardial infarction or cerebrovascular event with known atherosclerotic plaques in the carotid artery. Therefore, the data of the previous clinical study are not in contradiction with the current results, but further research is needed before it can be established that the expansion of CD4 þ CD28 T cells in CMV-seropositive ESRD patients is truly associated with plaque rupture. This study was not designed and has not a sufficient number of patients to make such an analysis. The strong association between CMV infection and the presence of CD4posCD28null T cells is in accordance with several previous reports.7–9 Most likely the large majority of CD4posCD28null T cells are CMV-specific, as they emerge only after CMV infection and proliferate in response to CMV but not to other common antigens.7 However, it cannot be fully excluded that CD28 loss is induced on some T cells, which are not specific for CMV peptides, due to cytokinemediated effects.17 Paradoxically, after CMV-specific stimulation in vitro, only a minority of CD4posCD28null T cells respond by expression of IFN-g, CD154, or CD107. This may truly reflect a low responsiveness to antigenic stimulation, but most likely is a consequence of the assay used, which does not allow for detection of CMV antigens that are expressed only during viral replication. The CMV genome encodes 763
original article
MGH Betjes et al.: Cytolytic CD4 T cells and ESRD
b Perforin
Granzyme B
P = 0.15
0
0
1000
P = 0.82
P = 0.27
500
ESRD patient
0 Heathy volunteerl
ESRD patient
Healthy volunteer
P = 0.89
Perforin+CD4+ CD28null T cells (104/l)
P = 0.007
ESRD patient
7000 6000 5000 4000 3000 2000 1000 0
P = 0.05
1
d
Healthy volunteer
Granzyme B+CD4+ CD28null T cells (104/l)
c
P = 0.58
ESRD patient
10
2
Healthy volunteer
P = 0.000
20
Perforin+CD4+ CD28null T cells (%)
Granzyme B+CD4+ CD28null T cells (%)
a
< 50 years ⭓ 50 years
Figure 5 | Cytotoxic potential of CMV-seropositive individuals. The cytotoxic potential of CD4posCD28null T cells was determined by intracellular staining of unstimulated PBMCs with antibodies directed against granzyme B (a and c) or perforin (b and d). Percentages (a and b) or absolute numbers (c and d) of granzyme B- or perforin-positive CD4posCD28null T are shown. The ESRD patients (n ¼ 63) and the age-matched controls (n ¼ 63) were divided into a young (o50 years, n ¼ 20) and elderly (X50 years, n ¼ 43) group. The Bonferroni-corrected P-value for statistical significance is Po0.007.
many of these antigens, and by analysis of all open reading frames it has become evident that on average as high as 10% of all circulating CD4pos T cells are in fact CMV-specific.6 After a-specific polyclonal stimulation with PMA/ionomycin, a much larger part of the CD4posCD28null T cells responded with IFN-g production and degranulation, indicative of their high cytolytic capacity. Virtually all CD4posCD28null T cells lacked the expression of the chemokine receptor CCR7. This receptor enables Tcm to migrate into lymphoid structures. The Tem lack this receptor, migrate through the tissues, and are capable of a rapid cytokine and proliferative response. In accordance with this concept, we found that the presence of effector molecules such as IFN-g, granzyme B, and perforin was largely confined to the CD4posCD28null Tem population. These phenotypical and functional properties make CD4posCD28null T cells very well suited for tissue invasion and subsequent destruction of target cells. In addition, similar to CMV-specific CD8pos T cells, they may lose CD45RO expression.13 Therefore, the CD4posCD28null T cells have a striking similarity to classical cytolytic CD8 T cells, a situation that seems unique for CMV infection. ESRD patients showed a significantly higher percentage of granzyme positivity and activation-induced degranulation within the CD4posCD28null T-cell population. As the 764
expression of granzyme in CD4pos CMV-specific T cells increases with maturation,18 this finding indicates the emergence of more terminally differentiated cytolytic CD4posCD28null T cells in ESRD patients, which are fully functional. An important issue that arises from our data is the question of why the CD4posCD28null T-cell population is so expanded in elderly ESRD patients. We have put forward the hypothesis that the uremia-associated immune defect causes frequent subclinical CMV.4 There are several lines of evidence, although indirectly, in favor of this hypothesis. First, there are serological data that indicate frequent CMV reactivation in ESRD patients, and even CMV organ disease has been reported.4,19,20 In this study, we observed an agedependent increase of anti-CMV IgG titres in ESRD patients. This observation is all the more remarkable as humoral immune responses are generally depressed in ESRD patients and supports the concept of frequent boosting of the immune system with CMV antigens. Second, T cells specific for herpesviruses such as CMV are partly interleukin-7 receptor negative21 and probably depend on frequent virus reactivation as one of the mechanisms to sustain the CMV-specific T-cell population. Lastly, oligoclonic T-cell receptor usage was shown in the CD4posCD28null T-cell populations of patients with an increased Kidney International (2008) 74, 760–767
original article
MGH Betjes et al.: Cytolytic CD4 T cells and ESRD
CD107a
IFN- γ
b P = 0.006
2 IFN-γ+CD4+ CD28null T cells (%)
P = 0.05
0.30 0.25 0.20 0.15 0.10 0.05 0.00 P = 0.31
P = 0.18
d IFN-γ+CD4+ CD28null T cells (104/l)
P = 0.007
P = 0.26
P = 0.02
1
100 75 50 25
600 500 400 300 200 100 ESRD patient
Healthy volunteer
ESRD patient
Healthy volunteer
Healthy volunteer
0
0
ESRD patient
125
P = 0.08
0
Healthy volunteer
CD107a+CD4+ CD28null T cells (104/l)
c
0.35
ESRD patient
CD107a+CD4+ CD28null T cells (%)
a
< 50 years ⭓ 50 years
Figure 6 | Percentages and absolute numbers of CD107a- and IFN-c-expressing CD4posCD28null T cells following CMV-antigen stimulation. Effector functions of CD4posCD28null T cells were determined by stimulating the PBMCs for 6 h with CMV-antigen. Percentages and absolute numbers of CD4posCD28null T cells expressing CD107a (a and c) on their cell surface or IFN-g (b and d) intracellularly were determined using flow cytometry. The ESRD patients (n ¼ 62) and the age-matched controls (n ¼ 62) were divided into a young (o50 years, n ¼ 22) and elderly (X50 years, n ¼ 40) group. The Bonferroni-corrected P-value for statistical significance is Po0.007.
frequency of these cells, indicating selective outgrowth and propagation of CMV-specific T cells because of frequent antigenic challenge.7,10,17 One may argue that the age-related decrease of immunological defence22 adds to the uremia-associated immune defect, thereby increasing the incidence of subclinical CMV reactivation. Alternatively, the constant CMV antigen-driven proliferation in uremic patients may by itself lead to impaired function of the immune system.23 There is little evidence to support this concept, but it is interesting that CMV seropositivity and antibody titre were negatively associated with the serological response after influenza vaccination in healthy individuals.24 In conclusion, elderly CMV-seropositive ESRD patients are prone to a huge expansion of cytolytic CD4posCD28null T cells that profoundly alters the composition of the CD4pos T-cell compartment. This previously unrecognized phenomenon may contribute to the increased risk for atherosclerotic disease associated with CMV seropositivity in ESRD patients. MATERIALS AND METHODS Study population Patients were recruited for the study from our outpatient clinic for renal transplantation. All patients had ESRD, defined as an Kidney International (2008) 74, 760–767
estimated glomerular filtration rate of less than 15 ml/min with or without renal replacement therapy. Patients with a malignancy, active infection, or taking immunosuppressive drugs were excluded. The healthy controls were recruited from their relatives, thereby matching for socio-economic and ethnical background. One control per ESRD patient was chosen and in case there was the option of more than one donor, the relative who age-matched the best was selected for the study. The clinical and demographical data of the CMV-seropositive patients and healthy controls are shown in Table 1. A patient was considered to have diabetes if he or she was taking insulin or oral hypoglycemic agents, or had previously received such treatment and was currently using dietary modification to control the condition. A patient was considered to have hypertension if he or she had received the diagnosis, or was being treated with antihypertensive medications at the time of evaluation. All patients were carefully evaluated for the presence of atherosclerotic disease, using the criteria and imaging techniques as described in detail in a recent publication.4 Healthy controls were evaluated for the presence of atherosclerotic disease by taking their medical history and physical examination. All controls showed normal renal function. All patients including healthy controls gave informed consent and the local medical ethical committee approved the study. It was conducted according to the Declaration of Helsinki principles and in compliance with International Conference on Harmonization/Good Clinical Practice regulations. 765
original article
Flow cytometric analysis of CD28 expression within the different T-cell subsets To determine the percentages and amount of various T lymphocytes, we performed a whole-blood staining as described previously but with some minor adjustments.5 Briefly, to examine expression of CD28 within the various T-lymphocyte subsets, we used either peridinin chlorophyll protein-labelled anti-CD4 or CD8 in combination with allophycocyanin-labelled anti-CD45RO (BD, Erebodegem, Belgium), fluorescein isothiocyanate-labelled anti-CCR7 (R&D Systems Europe Ltd, Abingdon, UK), and phycoerythrin-labelled anti-CD28 (BD). The various subsets are defined as CD4 þ and either naive being CCR7 þ CD45RO, central memory being CCR7 þ CD45RO þ (Tcm), effector memory being CCR7 CD45RO þ (Tem), or Temra being CCR7 CD45RO. In addition, for each of these subsets, the percentage (and amount) of cells negative and positive for CD28 was determined. Analysis was performed using the FACS Canto II (BD) and FACSDIVA software (BD). Isolation of peripheral blood mononuclear cells Peripheral blood mononuclear cells (PBMCs) were isolated from 35 ml of heparinized venous blood samples.25 PBMCs were resuspended in RPMI-1640 containing 10% heat-inactivated pooled human serum and 10% dimethyl sulfoxide (Sigma-Aldrich, Schnelldorf, Germany) and stored at –801C until further use. Before use, PBMCs were quickly thawed in RPMI-1640 (GibcoBRL, Paisley, Scotland) supplemented with 100 IU/ml penicillin, 100 mg/ml streptomycin, 2 mM L-glutamine, and 20% heat-inactivated AB þ pooled human serum (standard culture medium) with DNase (0.1 mg/ml; Roche Diagnostics GmbH, Mannheim, Germany). Finally, the cells were resuspended in standard culture medium at 4 106/ml. Following an overnight recovery, the percentage of viable cells was determined (always 480%) and the PBMCs were used for the different assays. Cytotoxic potential Unstimulated PBMCs (2 106/ml) were analyzed for cytotoxic potential by staining the cell surface with peridinin chlorophyll protein-labelled anti-CD4 and fluorescein isothiocyanate-labelled anti-CD28. Following fixation and permeabilization, cells were subjected to intracellular staining with one of the following phycoerythrin-labelled antibodies directed against perforin (BD) or granzyme B (Sanquin, Amsterdam, The Netherlands). Percentages of granzyme B- or perforin-positive cells were determined by analyzing the samples using the FACSCanto II (BD), selecting cells that have a typical lymphocyte scatter pattern, are CD4 positive, and either lack or express CD28. CD107a degranulation assay This assay was described in detail previously and detects surface mobilization of CD107a, a marker of degranulation associated with cytolytic function.26 For this purpose, PBMCs were stimulated with CMV lysate (30 mg/ml; Microbix Biosystems Inc., Toronto, Canada) for 6 h, with the last 5 h in the presence of monensin (Golgistop, BD). As a control, PBMCs were stimulated in standard culture medium alone or using a combination of PMA (50 ng/ml) and ionomycin (1 mg/ml) (both from Sigma-Aldrich). Subsequently, CD107a surface expression was determined by staining the cells with peridinin chlorophyll protein-labelled anti-CD4, phycoerythrinlabelled anti-CD28, and fluorescein isothiocyanate-labelled antiCD107a (BD Pharmingen, Erembodegem, Belgium) and analyzing the samples using the FACSCanto II in a similar way as described 766
MGH Betjes et al.: Cytolytic CD4 T cells and ESRD
before. Net antigen-specific frequencies of CD107a-expressing CD4 þ T cells were determined by correcting the signal induced by CMV-antigen or PMA/ionomycin for the background induced by standard culture medium alone. Intracellular cytokine staining The intracellular cytokine assay was performed as described recently in detail.27 Briefly, PBMCs (2 106/ml) were stimulated with either CMV lysate for 6 h, with the last 5 h in the presence of brefeldin A (Golgiplug, BD Pharmingen; 1 ml/106 PBMCs). As a control, PBMCs were stimulated with standard culture medium or with a mixture of PMA (50 ng/ml) and ionomycin (1 mg/ml). To analyze IFN-gproducing cells, the cell surface of PBMCs was stained with peridinin chlorophyll protein-labelled anti-CD4 and phycoerythrin-labelled anti-CD28. Following fixation and permeabilization, cells were stained with fluorescein isothiocyanate-labelled anti-IFN-g and allophycocyanin-labelled anti-CD154. Percentages of cytokine-positive cells were determined by analyzing the samples using the FACSCanto II (BD) as described above. To determine the net antigen-specific frequencies of IFN-g-producing CD4 þ T cells, the signal induced by the antigen was corrected by subtracting the background induced by standard culture medium alone. CMV serology Serum IgG antibodies to CMV were measured with an enzyme immunoassay (Biomerieux, VIDAS, Lyon, France) and expressed as arbitrary units/ml (AU/ml). As per the manufacturer’s guidelines, a test result exceeding 6 AU/ml was considered positive for the presence of CMV-specific IgG antibodies. Statistical analysis Categorical data were analyzed by the w2-test. Differences between groups were analyzed with the Student’s t-test when the variable was normally distributed and otherwise by the Mann–Whitney test. The relation between clinical parameters and T-cell characteristics was analyzed using multivariate regression analysis. All P-values were calculated two-sided and a P-value less than 0.05 was considered significant. The level of statistical significance for P-values was adjusted by a Bonferroni correction (less than 0.05 divided by the number of comparisons) when multiple comparisons were made, which is stated in the legends of the figures. The SPSS software version 10.1 was used for all statistical tests. DISCLOSURE All the authors declared no competing interests. REFERENCES 1.
2.
3. 4.
5.
Go AS, Chertow GM, Fan D et al. Chronic kidney disease and the risks of death, cardiovascular events, and hospitalization. N Engl J Med 2004; 351: 1296–1305. Shlipak MG, Fried LF, Cushman M et al. Cardiovascular mortality risk in chronic kidney disease: comparison of traditional and novel risk factors. J Am Med Assoc 2005; 293: 1737–1745. Menon V, Gaul A, Sarnak MJ. Cardiovascular risk factors in chronic kidney disease. Kidney Int 2005; 68: 1413–1418. Betjes MGH, Litjens NHR, Zietse R. Seropositivity for cytomegalovirus in patients with end-stage renal disease is strongly associated with atherosclerotic disease. Nephrol Dial Transplant 2007; 22: 3298–3303. Litjens NHR, van Druningen CJ, Betjes MGH. Progressive loss of renal function is associated with activation and depletion of naive T lymphocytes. Clin Immunol 2006; 118: 83–91. Kidney International (2008) 74, 760–767
original article
MGH Betjes et al.: Cytolytic CD4 T cells and ESRD
6. Sylwester AW, Mitchell BL, Edgar JB et al. Broadly targeted human cytomegalovirus-specific CD4+ and CD8+ T cells dominate the memory compartments of exposed subjects. J Exp Med 2005; 202: 673–685. 7. Hooper M, Kallas EG, Coffin D et al. Cytomegalovirus seropositivity is associated with the expansion of CD4+CD28 and CD8+CD28 T cells in rheumatoid arthritis. J Rheumatol 1999; 26: 1452–1457. 8. van Leeuwen EM, Remmerswaal EB, Vossen MT et al. Emergence of a CD4posCD28null granzyme B+, cytomegalovirus-specific T cell subset after recovery of primary cytomegalovirus infection. J Immunol 2004; 173: 1834–1841. 9. Weinberger B, Lazuardi L, Weiskirchner I et al. Healthy aging and latent infection with CMV lead to distinct changes in CD8+ and CD4+ T-cell subsets in the elderly. Hum Immunol 2007; 68: 86–90. 10. Liuzzo G, Kopecky SL, Frye RL et al. Perturbation of the T-cell repertoire in patients with unstable angina. Circulation 1999; 100: 2135. 11. Liuzzo G, Goronzy JJ, Yang H et al. Monoclonal T-cell proliferation and plaque instability in acute coronary syndromes. Circulation 2000; 101: 2883. 12. Sato K, Niessner A, Kopecky S et al. TRAIL-expressing T cells induce apoptosis of vascular smooth muscle cells in the atherosclerotic plaque. J Exp Med 2006; 203: 239–250. 13. Kuijpers TW, Vossen MT, Gent MR et al. Frequencies of circulating cytolytic, CD45RA+CD27–, CD8+ T lymphocytes depend on infection with CMV. J Immunol 2003; 170: 4342. 14. Frentsch M, Arbach O, Kirchhoff D et al. Direct access to CD4+ T cells specific for defined antigens according to CD154 expression. Nat Med 2005; 11: 1118–1124. 15. Nowik M, Nowacki P, Grabarek J et al. Can we talk about CD4posCD28null lymphocytes as a risk factor for ischemic stroke? Eur Neurol 2007; 58: 26–33. 16. Vescovini R, Biasini C, Fagnoni FF et al. Massive load of functional effector CD4+ and CD8+ T cells against cytomegalovirus in very old subjects. J Immunol 2007; 179: 4283–4291. 17. Fletcher J, Vukmanovic-Stejic M, Dunne P et al. Cytomegalovirus-specific CD4+ T cells in healthy carriers are continuously driven to replicative exhaustion. J Immunol 2005; 175: 8218–8225.
Kidney International (2008) 74, 760–767
18.
19.
20. 21.
22.
23.
24.
25.
26.
27.
Casazza JP, Betts MR, Price DA et al. Acquisition of direct antiviral effector functions by CMV-specific CD4+ T lymphocytes with cellular maturation. J Exp Med 2006; 203: 2865–2877. Falagas ME, Griffiths J, Prekezes J et al. Cytomegalovirus colitis mimicking colon carcinoma in an HIV-negative patient with chronic renal failure. Am J Gastroenterol 1996; 91: 168–169. Hardiman AE, Butter KC, Roe CJ et al. Cytomegalovirus infection in dialysis patients. Clin Nephrol 1985; 23: 12–17. Van Leeuwen EM, de Bree GJ, Remmerswaal EB et al. IL-7 receptor alpha chain expression distinguishes functional subsets of virus-specific human CD8+ T cells. Blood 2005; 106: 2091–2098. Huppert FA, Solomou W, O’Connor S et al. Aging and lymphocyte subpopulations: whole-blood analysis of immune markers in a large population sample of healthy elderly individuals. Exp Gerontol 1996; 33: 593–600. Khan N, Hislop A, Gudgeon N et al. Herpesvirus-specific CD8 T cell immunity in old age: cytomegalovirus impairs the response to a coresident EBV infection. J Immunol 2004; 173: 7481–7489. Trzonkowski P, Mysliwska J, Szmit E et al. Association between cytomegalovirus infection, enhanced proinflammatory response and low level of anti-hemagglutinins during the anti-influenza vaccination—an impact of immunosenescence. Vaccine 2003; 21: 3826–3836. Verkade MA, van de Wetering J, Klepper M et al. GM-CSF reduces the number of circulating dendritic cells in hemodialysis patients while enhancing the efficacy of hepatitis B vaccination. Kidney Int 2004; 66: 614–621. Betts MR, Brenchley JM, Price DA et al. Sensitive and viable identification of antigen-specific CD8+ T cells by a flow cytometric assay for degranulation. J Immunol Methods 2003; 281: 65–78. Litjens NHR, Huisman M, Baan CC et al. Hepatitis B vaccine-specific CD4(+) T cells can be detected and characterised at the single cell level: limited usefulness of dendritic cells as signal enhancers. J Immunol Methods 2008; 330: 1–11.
767
http://www.kidney-international.org & 2008 International Society of Nephrology
Expansion of cytolytic CD4 þ CD28 T cells in end-stage renal disease Michiel G.H. Betjes1, Martin Huisman1, Willem Weimar1 and Nicolle H.R. Litjens1 1
Division of Nephrology, Department of Internal Medicine, Erasmus Medical Center, Rotterdam, The Netherlands
Cytomegalovirus (CMV) seropositivity is associated with increased risk for atherosclerotic disease in patients with end-stage renal disease. This association is due to a unique peripheral blood CD4 þ T cell population which lack CD28 (CD4 þ CD28 T cells). Here we found that this patient population has a significant age-dependent increase of CD4 þ CD28 T cells that comprise over half of the circulating CD4 T cells in some. Patients over 50 years of age have a 50-fold higher percentage of CD4 þ CD28 T cells compared to seronegative patients and a 5-fold higher percentage when compared to seropositive healthy controls. Stimulation by CMV-antigen or by polyclonal stimulation using PMA and ionomycin showed that CD4 þ CD28 cells in patients with end stage renal disease degranulated and secreted interferon c thus indicating that they are cytolytic. The average anti-CMV IgG titer displayed a remarkable age-dependent increase only in patients with end stage renal disease. These findings are highly suggestive of repetitive antigenic stimulation of the immune system in patients with end stage disease by subclinical CMV reactivation which might contribute to increased atherosclerotic risk. Kidney International (2008) 74, 760–767; doi:10.1038/ki.2008.301; published online 9 July 2008 KEYWORDS: cytomegalovirus; end-stage renal disease; CD4 þ T cell; CD28; atherosclerosis
Correspondence: Michiel G.H. Betjes, Division of Nephrology, Department of Internal Medicine, Erasmus Medical Center, Dr Molewater plein 40, 3015 GD Rotterdam, The Netherlands. E-mail: [email protected] Received 23 January 2008; revised 27 March 2008; accepted 23 April 2008; published online 9 July 2008 760
Patients with end-stage renal disease (ESRD) carry a greatly increased risk for cardiovascular disease.1,2 Traditional risk factors such as smoking, hypertension, and hypercholesterolemia can be identified, but do not explain the full magnitude of the increment in risk.3 We have recently showed that latency for cytomegalovirus (CMV) is an unexpectedly high and independent risk factor for atherosclerosis in patients with ESRD.4 In addition, analysis of the T-cell subsets in patients with chronic kidney disease indicated that numbers of CMV-specific T cells increased as the stage of renal failure progressed.5 CMV infection is unique as it leaves a clear footprint in the T-cell repertoire. It has been estimated that on average, as high as 10% of the circulating CD4-positive T cells (CD4pos T cells) are CMV-specific in the latent situation.6 The phenotype of CD4-positive CMV-specific T cells is interesting with respect to atherosclerosis, as CMV infection increases the frequency of circulating CD4-positive CD28negative (CD4posCD28null) T cells.7–9 The circulating CD4posCD28null T-cell subset is enriched in the blood of patients with acute coronary syndromes and typically produces high levels of interferon-g (IFN-g).10 Plaque-infiltrating CD4 T cells, which are associated with plaque vulnerability for rupture, are likely recruited from this circulating CD4posCD28null T-cell subset.11 Apoptotic death of smooth muscle cells may be a mechanism of plaque destabilization, transforming stable plaques into ruptureprone lesions. Recently, it has been shown that CD4pos CD28null plaque-infiltrating T cells can indeed directly kill endothelial cells and vascular smooth muscle cells by a T-cell receptor-independent mechanism.12 These findings have given experimental support for the pathogenetic role of this specific T-cell subset in atherosclerotic plaque destabilization. We hypothesized that the CD4posCD28null T-cell population is increased in patients with ESRD compared with healthy controls, thereby imposing a risk for plaque rupture. Therefore, we quantified the circulating CD4posCD28null T cells and analyzed their phenotype, expression of cytolytic molecules, and degranulation capacity. The results show that the presence of circulating CD4posCD28null T cells is unique for CMV-seropositive healthy individuals and ESRD patients. However, elderly patients with ESRD have a massively expanded population of circulating cytolytic CD4posCD28null T cells. These findings may explain why CMV Kidney International (2008) 74, 760–767
original article
RESULTS CD4posCD28null T cells are greatly expanded in CMV-seropositive ESRD patients
In accordance with previous studies,5 the ESRD patients showed a lower number of CD4-positive T cells compared with healthy individuals (381 106/l versus 691 106/l, Po0.01). The frequency of the CD4posCD28null T cells was evidently associated with CMV infection. These cells were rarely observed (median frequency of 0.2% of the total CD4pos T cells) in CMV-seronegative individuals (Figure 1). The frequency and absolute number of CD4posCD28null T cells increased with age in CMV-seropositive individuals (Figure 2a and b). However, these cells were greatly expanded in CMV-seropositive ESRD patients, and comprised in some individuals up to 65% of the total CD4pos T-cell population. The expansion of the CD4posCD28null T-cell population was most evident in CMV-seropositive ESRD patients over 50 years of age. Compared to age-matched seronegative ESRD patients, they had on average a 50 times increased percentage of CD4posCD28null T cells. Compared to elderly CMVseropositive healthy controls, these patients had on average a five times increased percentage (3 versus 15% of total CD4pos T cells) and an almost fourfold increased absolute number of CD4posCD28null T cells (61 106/l versus 16 106/l, Po0.05). The expansion of CD4posCD28null T cells differed substantially among ESRD patients. However, except for age, no other demographic characteristics or clinical characteristics (listed in Table 1), for example type of renal replacement therapy or diabetes, were associated with the number of CD4posCD28null T cells. Remarkably, an age-dependent
CD4+ CD28null T lymphocytes (10 6/l)
seropositivity is associated with an increased risk factor for atherosclerotic disease, especially in ESRD patients.
CD4+ CD28null T lymphocytes (%)
MGH Betjes et al.: Cytolytic CD4 T cells and ESRD
70 60
0.010
50 0.003
40 0.015
30 20 10 0
⭐ 40
41 – 50 51 – 60 Age (years)
61 – 75
0.031
350 300 250
0.029
200 150
0.038
100 50 0
⭐ 40
41 – 50 51 – 60 Age (years)
61 – 75
CMV-seropositive healthy controls CMV-seropositive ESRD patients
Figure 2 | CD4posCD28null T cells and age. CMV-seropositive ESRD patients (closed symbols, N ¼ 70) and age-matched healthy controls (open symbols, N ¼ 70) were divided into age groups (p40, n ¼ 9; 41–50, n ¼ 17; 51–60, n ¼ 18; 60–75, n ¼ 26) and either percentages (a) or absolute numbers expressed as 106/l (b) of CD4posCD28null T cells were shown. The Bonferroni-corrected P-value for statistical significance is Po0.006.
P = 0.000
CD4 +CD28null (%)
70
Table 1 | Clinical and demographic characteristics of CMV-seropositive patients with end-stage renal disease and CMV-seropositive healthy controls
50 30 10 10 5 0 Healthy controls
ESRD patients
Closed symbols = CMV-seronegative individuals Open symbols = CMV-seropositive individuals
Figure 1 | CD4posCD28null T cells and CMV seropositivity. Percentages of CD4posCD28null T cells of the total number of circulating CD4pos T cells, comparing CMV-seronegative to CMV-seropositive individuals. The CMV-seronegative groups included 30 healthy controls and 30 patients with ESRD. The CMV-seropositive groups included 72 healthy controls and 72 ESRD patients. Kidney International (2008) 74, 760–767
Number of individuals Age Male Non-Western European origin Diabetes mellitus Hypertension History of atherosclerotic disease Underlying kidney disease Hypertensive nephropathy Primary glomerulopathy Diabetic nephropathy Polycystic kidney disease Other Unknown
Patients
Healthy controls
72 54 (21–75)a 68.0% 44.0% 30.0% 84.7% 38.5%
72 55 (23-71)a 27.8%** 30.5% 0.0%** 16.6%** 0.0%**
31.9% 20.8% 18.1% 11.1% 13.9% 4.2%
CMV, cytomegalovirus.[2]**P-value o0.001. a Data are given in median years with range in parentheses.
761
original article
MGH Betjes et al.: Cytolytic CD4 T cells and ESRD
increase of the average anti-CMV IgG titer was observed for ESRD patients, but not for healthy individuals (Figure 3). CD4posCD28null T cells are CCR7-negative and effector memory cells
CMV-specific IgG (AU/ml)
The expression of the CD28 molecule on the cell surface of CD4pos T cells yields a clear distinction between a CD28pos and CD28null population (Figure 4a). The circulating T cells can be subdivided into naive T cells and memory T cells by differential expression of the chemokine receptor CCR7 and CD45RO. The memory cells are able to either display immediate effector function in inflamed tissues (effector
Granzyme B- and perforin-positive CD4posCD28null T cells are significantly increased in ESRD patients
100
50
0
memory T cells (Tem)) or constitute a pool of long-lived memory cells that develop more quickly into effector cells following antigenic stimulation in lymphoid organs (central memory T cells (Tcm)). More than 97% of the CD4posCD28null T cells were CCR7 negative, with the majority of the cells (on average 75%) expressing CD45RO, and thus classified as classical effector memory T cells (Figure 4c and d). CD4posCD28null T cells without expression of CD45RO were observed only in CMV-seropositive individuals, similar to the previously reported CD45RO-negative, CMV-specific, CD8-positive effector T cells.13
⭐ 40
41 – 50 51 – 60 Age (years)
61 – 75
CMV-seropositive healthy controls CMV-seropositive ESRD patients
Figure 3 | Age-related increase in CMV-specific IgG titers for ESRD patients but not healthy controls. Mean (±s.e.m.) CMV-specific IgG titers are depicted on the Y axis and expressed as AU/ml, whereas on the X axis, the different age groups are displayed. Open symbols represent the healthy controls and the closed ones, the ESRD patients. A significant (Po0.01, Spearman’s rank correlation coefficient) age-dependent increase was noticed for ESRD patients (N ¼ 70) but not for healthy controls (N ¼ 66).
Expression of the cytotoxic molecules granzyme B and perforin was limited to the CD4posCD28null T cells (Table 2). Specifically, the percentage of granzyme B-positive cells within the CD4posCD28null T-cell population was higher for ESRD patients compared with healthy individuals (65 versus 52%, Po0.01). Next, we calculated the total number of granzyme B- and perforin-positive CD4posCD28null T cells and the percentage of these cells in the total CD4pos T-cell population. A significantly increased percentage and total number of granzyme B- and percentage of perforinpositive CD4posCD28null T cells was observed in ESRD patients 50 years and older (Figure 5). CD4posCD28null T cells are functional as they degranulate and produce IFN-c after CMV-antigen and polyclonal stimulation
Upon stimulation with CMV-antigen, the CD4posCD28null T cells degranulated, as shown by expression of membranebound CD107a, and produced IFN-g. This was observed at a
Table 2 | Percentages of T cells with cytotoxic potential and effector function in the CD4+CD28null and the CD4+CD28+ subsets CD4+CD28null T cells
CD4+CD28+ T cells
ESRD patients
Healthy controls
P-value1
ESRD patients
Healthy controls
P-value2
60 65.1±3.8b 7.8±1.3
60 51.6±4.3 10.8±1.8
0.009 0.152
60 4.8±1.3 0.6±0.1
60 1.7±0.3 0.2±0.04
0.000 0.006
CD107a CMV specific PMA/ionomycin
1.7±0.3 6.5±1.4
1.0±0.2 7.6±0.4
0.050 0.671
0.3±0.1 3.5±1.0
0.04±0.01 4.2±0.9
0.003 0.396
IFN-g CMV specific PMA/ionomycin
4.7±1.0 65.2±5.2
4.1±1.3 36.1±8.9
0.175 0.021
0.6±0.2 21.9±3.0
0.2±0.04 12.5±2.0
0.338 0.016
CD154 CMV specific PMA/ionomycin
2.7±0.7 3.1±0.4
2.7±0.9 3.1±0.8
0.319 0.790
0.4±0.2 11.5±2.1
0.2±0.03 12.5±2.5
0.496 0.657
Number Granzyme Ba Perforina
CMV, cytomegalovirus; ESRD, end-stage renal disease; IFN-g, interferon-g; PMA, phorbol 12-myristate 13-acetate.[2]Data of patients with end-stage renal disease and agematched healthy controls are shown. P-value1 and P-value2 for the differences between data of CD28null T cells and CD28positive cells, respectively, comparing ESRD patients with healthy controls. The Bonferroni-corrected level of statistical significance is P-value o0.003. a Expression determined using unstimulated PBMCs. b Data are expressed in mean percentages±standard error of the means.
762
Kidney International (2008) 74, 760–767
original article
MGH Betjes et al.: Cytolytic CD4 T cells and ESRD
b 35.4
51
CD28
88.4
CCR7
a 1.3
12.3
c 0.3
0
11.6 CD4 9.7
90 CD45RO
CD4+CD28null distribution
d
0.001
CD4+CD28null (%)
70 50 30 10 10 5 0 Tnaive
Tcm
Tem
Open symbols: CMV-seropositive healthy controls Closed symbols: CMV-seropositive ESRD patients
Figure 4 | CD28 expression within the different CD4pos T-cell subsets of a CMV-seropositive individual. A representative example of the flow cytometric analysis of CD4pos T cells for a CMV-seropositive individual is shown in (a–c). Panel a shows a typical example of CD28 expression within CD4-positive peripheral blood T cells. The CD28pos (b) and CD28null (c) CD4pos T-cell subsets are further characterized by their expression of CD45RO and CCR7. Using these markers, the following T-cell subsets can be discriminated: naive T cells (Tnaive, CCR7 positive and CD45RO negative), central memory T cells (Tcm, CCR7 positive and CD45RO positive), and effector memory T cells (Tem, CCR7 negative and CD45RO positive). The overall distribution of CD4posCD28null T cells of ESRD patients (n ¼ 70) and age-matched healthy controls (n ¼ 66) is displayed in (d). The Bonferroni-corrected P-value for statistical significance is Po0.02.
significantly lower level in the CD4posCD28pos T cells (Table 2). After a strong polyclonal stimulus such as phorbol 12-myristate 13-acetate (PMA)/ionomycin, IFN-g production was predominantly observed within the subset of CD4posCD28null T cells (Table 2). The percentages and number of CD4posCD28null T cells expressing CD107 and IFN-g after CMV (Figure 6) or polyclonal stimulation were significantly higher in elderly ESRD patients, compared with the age-matched healthy controls. The expression of CD40 ligand (CD154) after antigenic stimulation is an important part of the costimulatory pathway during antigen presentation. It can be Kidney International (2008) 74, 760–767
used as a sensitive and specific method for identifying antigen-specific T cells.14 However, the percentage of CD4posCD28null T cells expressing CD154 after stimulation was always significantly lower (on average 50%) than the percentage of IFN-g-positive cells (Table 2). This observation indicates that CD4posCD28null T cells have become largely independent of the CD40-CD40L co-stimulation pathway. DISCUSSION
The CD4posCD28null circulating T-cell population has been associated with both atherosclerotic disease and CMV infection.6–11,15 The results of this study show that CD4posCD28null T cells are rarely observed in CMVseronegative individuals, but may expand to as much as 65% of the total CD4pos T-cell population in elderly CMVseropositive ESRD patients. This expansion of CD4posCD28null T cells is in sharp contrast with the well-documented uremia and age-related lymphopenia of the CD4pos T-cell compartment.5,9 An age-related increased number of CMVspecific T cells has previously been reported in healthy adults, but only in the very elderly.9,16 The published clinical and experimental data suggest a role for CD4posCD28null T cells in atherosclerotic plaque destabilization,10–12 and in a previous study we found that CMV infection was significantly and strongly associated with atherosclerotic disease in ESRD patients.4 It is evident that the presence of atherosclerosis is not synonymous with plaque rupture, which may cause an acute vascular event. As such, the present data cannot be directly correlated with the findings of our previous clinical study.4 However, many patients with a positive history for atherosclerotic disease probably have experienced a plaque rupture, as they had a history of myocardial infarction or cerebrovascular event with known atherosclerotic plaques in the carotid artery. Therefore, the data of the previous clinical study are not in contradiction with the current results, but further research is needed before it can be established that the expansion of CD4 þ CD28 T cells in CMV-seropositive ESRD patients is truly associated with plaque rupture. This study was not designed and has not a sufficient number of patients to make such an analysis. The strong association between CMV infection and the presence of CD4posCD28null T cells is in accordance with several previous reports.7–9 Most likely the large majority of CD4posCD28null T cells are CMV-specific, as they emerge only after CMV infection and proliferate in response to CMV but not to other common antigens.7 However, it cannot be fully excluded that CD28 loss is induced on some T cells, which are not specific for CMV peptides, due to cytokinemediated effects.17 Paradoxically, after CMV-specific stimulation in vitro, only a minority of CD4posCD28null T cells respond by expression of IFN-g, CD154, or CD107. This may truly reflect a low responsiveness to antigenic stimulation, but most likely is a consequence of the assay used, which does not allow for detection of CMV antigens that are expressed only during viral replication. The CMV genome encodes 763
original article
MGH Betjes et al.: Cytolytic CD4 T cells and ESRD
b Perforin
Granzyme B
P = 0.15
0
0
1000
P = 0.82
P = 0.27
500
ESRD patient
0 Heathy volunteerl
ESRD patient
Healthy volunteer
P = 0.89
Perforin+CD4+ CD28null T cells (104/l)
P = 0.007
ESRD patient
7000 6000 5000 4000 3000 2000 1000 0
P = 0.05
1
d
Healthy volunteer
Granzyme B+CD4+ CD28null T cells (104/l)
c
P = 0.58
ESRD patient
10
2
Healthy volunteer
P = 0.000
20
Perforin+CD4+ CD28null T cells (%)
Granzyme B+CD4+ CD28null T cells (%)
a
< 50 years ⭓ 50 years
Figure 5 | Cytotoxic potential of CMV-seropositive individuals. The cytotoxic potential of CD4posCD28null T cells was determined by intracellular staining of unstimulated PBMCs with antibodies directed against granzyme B (a and c) or perforin (b and d). Percentages (a and b) or absolute numbers (c and d) of granzyme B- or perforin-positive CD4posCD28null T are shown. The ESRD patients (n ¼ 63) and the age-matched controls (n ¼ 63) were divided into a young (o50 years, n ¼ 20) and elderly (X50 years, n ¼ 43) group. The Bonferroni-corrected P-value for statistical significance is Po0.007.
many of these antigens, and by analysis of all open reading frames it has become evident that on average as high as 10% of all circulating CD4pos T cells are in fact CMV-specific.6 After a-specific polyclonal stimulation with PMA/ionomycin, a much larger part of the CD4posCD28null T cells responded with IFN-g production and degranulation, indicative of their high cytolytic capacity. Virtually all CD4posCD28null T cells lacked the expression of the chemokine receptor CCR7. This receptor enables Tcm to migrate into lymphoid structures. The Tem lack this receptor, migrate through the tissues, and are capable of a rapid cytokine and proliferative response. In accordance with this concept, we found that the presence of effector molecules such as IFN-g, granzyme B, and perforin was largely confined to the CD4posCD28null Tem population. These phenotypical and functional properties make CD4posCD28null T cells very well suited for tissue invasion and subsequent destruction of target cells. In addition, similar to CMV-specific CD8pos T cells, they may lose CD45RO expression.13 Therefore, the CD4posCD28null T cells have a striking similarity to classical cytolytic CD8 T cells, a situation that seems unique for CMV infection. ESRD patients showed a significantly higher percentage of granzyme positivity and activation-induced degranulation within the CD4posCD28null T-cell population. As the 764
expression of granzyme in CD4pos CMV-specific T cells increases with maturation,18 this finding indicates the emergence of more terminally differentiated cytolytic CD4posCD28null T cells in ESRD patients, which are fully functional. An important issue that arises from our data is the question of why the CD4posCD28null T-cell population is so expanded in elderly ESRD patients. We have put forward the hypothesis that the uremia-associated immune defect causes frequent subclinical CMV.4 There are several lines of evidence, although indirectly, in favor of this hypothesis. First, there are serological data that indicate frequent CMV reactivation in ESRD patients, and even CMV organ disease has been reported.4,19,20 In this study, we observed an agedependent increase of anti-CMV IgG titres in ESRD patients. This observation is all the more remarkable as humoral immune responses are generally depressed in ESRD patients and supports the concept of frequent boosting of the immune system with CMV antigens. Second, T cells specific for herpesviruses such as CMV are partly interleukin-7 receptor negative21 and probably depend on frequent virus reactivation as one of the mechanisms to sustain the CMV-specific T-cell population. Lastly, oligoclonic T-cell receptor usage was shown in the CD4posCD28null T-cell populations of patients with an increased Kidney International (2008) 74, 760–767
original article
MGH Betjes et al.: Cytolytic CD4 T cells and ESRD
CD107a
IFN- γ
b P = 0.006
2 IFN-γ+CD4+ CD28null T cells (%)
P = 0.05
0.30 0.25 0.20 0.15 0.10 0.05 0.00 P = 0.31
P = 0.18
d IFN-γ+CD4+ CD28null T cells (104/l)
P = 0.007
P = 0.26
P = 0.02
1
100 75 50 25
600 500 400 300 200 100 ESRD patient
Healthy volunteer
ESRD patient
Healthy volunteer
Healthy volunteer
0
0
ESRD patient
125
P = 0.08
0
Healthy volunteer
CD107a+CD4+ CD28null T cells (104/l)
c
0.35
ESRD patient
CD107a+CD4+ CD28null T cells (%)
a
< 50 years ⭓ 50 years
Figure 6 | Percentages and absolute numbers of CD107a- and IFN-c-expressing CD4posCD28null T cells following CMV-antigen stimulation. Effector functions of CD4posCD28null T cells were determined by stimulating the PBMCs for 6 h with CMV-antigen. Percentages and absolute numbers of CD4posCD28null T cells expressing CD107a (a and c) on their cell surface or IFN-g (b and d) intracellularly were determined using flow cytometry. The ESRD patients (n ¼ 62) and the age-matched controls (n ¼ 62) were divided into a young (o50 years, n ¼ 22) and elderly (X50 years, n ¼ 40) group. The Bonferroni-corrected P-value for statistical significance is Po0.007.
frequency of these cells, indicating selective outgrowth and propagation of CMV-specific T cells because of frequent antigenic challenge.7,10,17 One may argue that the age-related decrease of immunological defence22 adds to the uremia-associated immune defect, thereby increasing the incidence of subclinical CMV reactivation. Alternatively, the constant CMV antigen-driven proliferation in uremic patients may by itself lead to impaired function of the immune system.23 There is little evidence to support this concept, but it is interesting that CMV seropositivity and antibody titre were negatively associated with the serological response after influenza vaccination in healthy individuals.24 In conclusion, elderly CMV-seropositive ESRD patients are prone to a huge expansion of cytolytic CD4posCD28null T cells that profoundly alters the composition of the CD4pos T-cell compartment. This previously unrecognized phenomenon may contribute to the increased risk for atherosclerotic disease associated with CMV seropositivity in ESRD patients. MATERIALS AND METHODS Study population Patients were recruited for the study from our outpatient clinic for renal transplantation. All patients had ESRD, defined as an Kidney International (2008) 74, 760–767
estimated glomerular filtration rate of less than 15 ml/min with or without renal replacement therapy. Patients with a malignancy, active infection, or taking immunosuppressive drugs were excluded. The healthy controls were recruited from their relatives, thereby matching for socio-economic and ethnical background. One control per ESRD patient was chosen and in case there was the option of more than one donor, the relative who age-matched the best was selected for the study. The clinical and demographical data of the CMV-seropositive patients and healthy controls are shown in Table 1. A patient was considered to have diabetes if he or she was taking insulin or oral hypoglycemic agents, or had previously received such treatment and was currently using dietary modification to control the condition. A patient was considered to have hypertension if he or she had received the diagnosis, or was being treated with antihypertensive medications at the time of evaluation. All patients were carefully evaluated for the presence of atherosclerotic disease, using the criteria and imaging techniques as described in detail in a recent publication.4 Healthy controls were evaluated for the presence of atherosclerotic disease by taking their medical history and physical examination. All controls showed normal renal function. All patients including healthy controls gave informed consent and the local medical ethical committee approved the study. It was conducted according to the Declaration of Helsinki principles and in compliance with International Conference on Harmonization/Good Clinical Practice regulations. 765
original article
Flow cytometric analysis of CD28 expression within the different T-cell subsets To determine the percentages and amount of various T lymphocytes, we performed a whole-blood staining as described previously but with some minor adjustments.5 Briefly, to examine expression of CD28 within the various T-lymphocyte subsets, we used either peridinin chlorophyll protein-labelled anti-CD4 or CD8 in combination with allophycocyanin-labelled anti-CD45RO (BD, Erebodegem, Belgium), fluorescein isothiocyanate-labelled anti-CCR7 (R&D Systems Europe Ltd, Abingdon, UK), and phycoerythrin-labelled anti-CD28 (BD). The various subsets are defined as CD4 þ and either naive being CCR7 þ CD45RO, central memory being CCR7 þ CD45RO þ (Tcm), effector memory being CCR7 CD45RO þ (Tem), or Temra being CCR7 CD45RO. In addition, for each of these subsets, the percentage (and amount) of cells negative and positive for CD28 was determined. Analysis was performed using the FACS Canto II (BD) and FACSDIVA software (BD). Isolation of peripheral blood mononuclear cells Peripheral blood mononuclear cells (PBMCs) were isolated from 35 ml of heparinized venous blood samples.25 PBMCs were resuspended in RPMI-1640 containing 10% heat-inactivated pooled human serum and 10% dimethyl sulfoxide (Sigma-Aldrich, Schnelldorf, Germany) and stored at –801C until further use. Before use, PBMCs were quickly thawed in RPMI-1640 (GibcoBRL, Paisley, Scotland) supplemented with 100 IU/ml penicillin, 100 mg/ml streptomycin, 2 mM L-glutamine, and 20% heat-inactivated AB þ pooled human serum (standard culture medium) with DNase (0.1 mg/ml; Roche Diagnostics GmbH, Mannheim, Germany). Finally, the cells were resuspended in standard culture medium at 4 106/ml. Following an overnight recovery, the percentage of viable cells was determined (always 480%) and the PBMCs were used for the different assays. Cytotoxic potential Unstimulated PBMCs (2 106/ml) were analyzed for cytotoxic potential by staining the cell surface with peridinin chlorophyll protein-labelled anti-CD4 and fluorescein isothiocyanate-labelled anti-CD28. Following fixation and permeabilization, cells were subjected to intracellular staining with one of the following phycoerythrin-labelled antibodies directed against perforin (BD) or granzyme B (Sanquin, Amsterdam, The Netherlands). Percentages of granzyme B- or perforin-positive cells were determined by analyzing the samples using the FACSCanto II (BD), selecting cells that have a typical lymphocyte scatter pattern, are CD4 positive, and either lack or express CD28. CD107a degranulation assay This assay was described in detail previously and detects surface mobilization of CD107a, a marker of degranulation associated with cytolytic function.26 For this purpose, PBMCs were stimulated with CMV lysate (30 mg/ml; Microbix Biosystems Inc., Toronto, Canada) for 6 h, with the last 5 h in the presence of monensin (Golgistop, BD). As a control, PBMCs were stimulated in standard culture medium alone or using a combination of PMA (50 ng/ml) and ionomycin (1 mg/ml) (both from Sigma-Aldrich). Subsequently, CD107a surface expression was determined by staining the cells with peridinin chlorophyll protein-labelled anti-CD4, phycoerythrinlabelled anti-CD28, and fluorescein isothiocyanate-labelled antiCD107a (BD Pharmingen, Erembodegem, Belgium) and analyzing the samples using the FACSCanto II in a similar way as described 766
MGH Betjes et al.: Cytolytic CD4 T cells and ESRD
before. Net antigen-specific frequencies of CD107a-expressing CD4 þ T cells were determined by correcting the signal induced by CMV-antigen or PMA/ionomycin for the background induced by standard culture medium alone. Intracellular cytokine staining The intracellular cytokine assay was performed as described recently in detail.27 Briefly, PBMCs (2 106/ml) were stimulated with either CMV lysate for 6 h, with the last 5 h in the presence of brefeldin A (Golgiplug, BD Pharmingen; 1 ml/106 PBMCs). As a control, PBMCs were stimulated with standard culture medium or with a mixture of PMA (50 ng/ml) and ionomycin (1 mg/ml). To analyze IFN-gproducing cells, the cell surface of PBMCs was stained with peridinin chlorophyll protein-labelled anti-CD4 and phycoerythrin-labelled anti-CD28. Following fixation and permeabilization, cells were stained with fluorescein isothiocyanate-labelled anti-IFN-g and allophycocyanin-labelled anti-CD154. Percentages of cytokine-positive cells were determined by analyzing the samples using the FACSCanto II (BD) as described above. To determine the net antigen-specific frequencies of IFN-g-producing CD4 þ T cells, the signal induced by the antigen was corrected by subtracting the background induced by standard culture medium alone. CMV serology Serum IgG antibodies to CMV were measured with an enzyme immunoassay (Biomerieux, VIDAS, Lyon, France) and expressed as arbitrary units/ml (AU/ml). As per the manufacturer’s guidelines, a test result exceeding 6 AU/ml was considered positive for the presence of CMV-specific IgG antibodies. Statistical analysis Categorical data were analyzed by the w2-test. Differences between groups were analyzed with the Student’s t-test when the variable was normally distributed and otherwise by the Mann–Whitney test. The relation between clinical parameters and T-cell characteristics was analyzed using multivariate regression analysis. All P-values were calculated two-sided and a P-value less than 0.05 was considered significant. The level of statistical significance for P-values was adjusted by a Bonferroni correction (less than 0.05 divided by the number of comparisons) when multiple comparisons were made, which is stated in the legends of the figures. The SPSS software version 10.1 was used for all statistical tests. DISCLOSURE All the authors declared no competing interests. REFERENCES 1.
2.
3. 4.
5.
Go AS, Chertow GM, Fan D et al. Chronic kidney disease and the risks of death, cardiovascular events, and hospitalization. N Engl J Med 2004; 351: 1296–1305. Shlipak MG, Fried LF, Cushman M et al. Cardiovascular mortality risk in chronic kidney disease: comparison of traditional and novel risk factors. J Am Med Assoc 2005; 293: 1737–1745. Menon V, Gaul A, Sarnak MJ. Cardiovascular risk factors in chronic kidney disease. Kidney Int 2005; 68: 1413–1418. Betjes MGH, Litjens NHR, Zietse R. Seropositivity for cytomegalovirus in patients with end-stage renal disease is strongly associated with atherosclerotic disease. Nephrol Dial Transplant 2007; 22: 3298–3303. Litjens NHR, van Druningen CJ, Betjes MGH. Progressive loss of renal function is associated with activation and depletion of naive T lymphocytes. Clin Immunol 2006; 118: 83–91. Kidney International (2008) 74, 760–767
original article
MGH Betjes et al.: Cytolytic CD4 T cells and ESRD
6. Sylwester AW, Mitchell BL, Edgar JB et al. Broadly targeted human cytomegalovirus-specific CD4+ and CD8+ T cells dominate the memory compartments of exposed subjects. J Exp Med 2005; 202: 673–685. 7. Hooper M, Kallas EG, Coffin D et al. Cytomegalovirus seropositivity is associated with the expansion of CD4+CD28 and CD8+CD28 T cells in rheumatoid arthritis. J Rheumatol 1999; 26: 1452–1457. 8. van Leeuwen EM, Remmerswaal EB, Vossen MT et al. Emergence of a CD4posCD28null granzyme B+, cytomegalovirus-specific T cell subset after recovery of primary cytomegalovirus infection. J Immunol 2004; 173: 1834–1841. 9. Weinberger B, Lazuardi L, Weiskirchner I et al. Healthy aging and latent infection with CMV lead to distinct changes in CD8+ and CD4+ T-cell subsets in the elderly. Hum Immunol 2007; 68: 86–90. 10. Liuzzo G, Kopecky SL, Frye RL et al. Perturbation of the T-cell repertoire in patients with unstable angina. Circulation 1999; 100: 2135. 11. Liuzzo G, Goronzy JJ, Yang H et al. Monoclonal T-cell proliferation and plaque instability in acute coronary syndromes. Circulation 2000; 101: 2883. 12. Sato K, Niessner A, Kopecky S et al. TRAIL-expressing T cells induce apoptosis of vascular smooth muscle cells in the atherosclerotic plaque. J Exp Med 2006; 203: 239–250. 13. Kuijpers TW, Vossen MT, Gent MR et al. Frequencies of circulating cytolytic, CD45RA+CD27–, CD8+ T lymphocytes depend on infection with CMV. J Immunol 2003; 170: 4342. 14. Frentsch M, Arbach O, Kirchhoff D et al. Direct access to CD4+ T cells specific for defined antigens according to CD154 expression. Nat Med 2005; 11: 1118–1124. 15. Nowik M, Nowacki P, Grabarek J et al. Can we talk about CD4posCD28null lymphocytes as a risk factor for ischemic stroke? Eur Neurol 2007; 58: 26–33. 16. Vescovini R, Biasini C, Fagnoni FF et al. Massive load of functional effector CD4+ and CD8+ T cells against cytomegalovirus in very old subjects. J Immunol 2007; 179: 4283–4291. 17. Fletcher J, Vukmanovic-Stejic M, Dunne P et al. Cytomegalovirus-specific CD4+ T cells in healthy carriers are continuously driven to replicative exhaustion. J Immunol 2005; 175: 8218–8225.
Kidney International (2008) 74, 760–767
18.
19.
20. 21.
22.
23.
24.
25.
26.
27.
Casazza JP, Betts MR, Price DA et al. Acquisition of direct antiviral effector functions by CMV-specific CD4+ T lymphocytes with cellular maturation. J Exp Med 2006; 203: 2865–2877. Falagas ME, Griffiths J, Prekezes J et al. Cytomegalovirus colitis mimicking colon carcinoma in an HIV-negative patient with chronic renal failure. Am J Gastroenterol 1996; 91: 168–169. Hardiman AE, Butter KC, Roe CJ et al. Cytomegalovirus infection in dialysis patients. Clin Nephrol 1985; 23: 12–17. Van Leeuwen EM, de Bree GJ, Remmerswaal EB et al. IL-7 receptor alpha chain expression distinguishes functional subsets of virus-specific human CD8+ T cells. Blood 2005; 106: 2091–2098. Huppert FA, Solomou W, O’Connor S et al. Aging and lymphocyte subpopulations: whole-blood analysis of immune markers in a large population sample of healthy elderly individuals. Exp Gerontol 1996; 33: 593–600. Khan N, Hislop A, Gudgeon N et al. Herpesvirus-specific CD8 T cell immunity in old age: cytomegalovirus impairs the response to a coresident EBV infection. J Immunol 2004; 173: 7481–7489. Trzonkowski P, Mysliwska J, Szmit E et al. Association between cytomegalovirus infection, enhanced proinflammatory response and low level of anti-hemagglutinins during the anti-influenza vaccination—an impact of immunosenescence. Vaccine 2003; 21: 3826–3836. Verkade MA, van de Wetering J, Klepper M et al. GM-CSF reduces the number of circulating dendritic cells in hemodialysis patients while enhancing the efficacy of hepatitis B vaccination. Kidney Int 2004; 66: 614–621. Betts MR, Brenchley JM, Price DA et al. Sensitive and viable identification of antigen-specific CD8+ T cells by a flow cytometric assay for degranulation. J Immunol Methods 2003; 281: 65–78. Litjens NHR, Huisman M, Baan CC et al. Hepatitis B vaccine-specific CD4(+) T cells can be detected and characterised at the single cell level: limited usefulness of dendritic cells as signal enhancers. J Immunol Methods 2008; 330: 1–11.
767