Age-associated Epstein–Barr virus-specific T cell responses in seropositive healthy adults

Abstract

Epstein–Barr virus (EBV) is present in 95% of the world's adult population. The immune response participates in immune vigilance and persistent infection control, and this condition is maintained by both a good quality (functionality) and quantity of specific T cells throughout life. In the present study, we evaluated EBV-specific CD4⁺ and CD8⁺ T lymphocyte responses in seropositive healthy individuals younger and older than 50 years of age. The assessment comprised the frequency, phenotype, functionality and clonotypic distribution of T lymphocytes. We found that in both age groups a similar EBV-specific T cell response was found, with overlapping numbers of tumour necrosis factor (TNF)-α⁺ T lymphocytes (CD4⁺ and CD8⁺) within the memory and effector cell compartments, in addition to monofunctional and multi-functional T cells producing interleukin (IL)-2 and/or interferon (IFN)-γ. However, individuals aged more than 50 years showed significantly higher frequencies of IL-2-producing CD4⁺ T lymphocytes in association with greater production of soluble IFN-γ, TNF-α and IL-6 than subjects younger than 50 years. A polyclonal T cell receptor (TCR)-variable beta region (Vβ) repertoire exists in both age groups under basal conditions and in response to EBV; the major TCR families found in TNF-α⁺/CD4⁺ T lymphocytes were Vβ1, Vβ2, Vβ17 and Vβ22 in both age groups, and the major TCR family in TNF-α⁺/CD8⁺ T cells was Vβ13·1 for individuals younger than 50 years and Vβ9 for individuals aged more than 50 years. Our findings suggest that the EBV-specific T cell response (using a polyclonal stimulation model) is distributed throughout several T cell differentiation compartments in an age-independent manner and includes both monofunctional and multi-functional T lymphocytes.

Introduction

Epstein–Barr virus (EBV) is a genetically stable virus with 95% prevalence among the adult population worldwide 1. The primary infection is typically asymptomatic and occurs during childhood; none the less, primary infection by EBV during adolescence or early adulthood can induce lymphoproliferative diseases such as infectious mononucleosis (IM), characterized by an exacerbated cytotoxic CD8⁺ T lymphocyte response and a reduced CD4⁺ T lymphocyte response to specific lytic and/or latency-stage viral antigens 2. These cell types are also found in healthy carriers with no prior history of IM, who display a low viral load (1–50 infected B cells/10⁶ circulating B cells) 3 that is most probably linked to immunosurveillance mechanisms and the control of possible viral reactivation due to persistent infection 4–6.

The expansion of T lymphocytes displaying (oligo) monoclonal and polyclonal T cell receptor (TCR)-αβ repertoires is typically identified in IM patients 7,8. Such expansions emerge because of the dominant proliferation of specific clonotypes, which disappear during the latent phase despite viral persistence, and are subsequently replaced by the appearance of subdominant clonotypes. This phenomenon has been linked to functional T cell exhaustion, as reflected at later stages by the deletion of some TCR-Vβ families 9–11, that may also be absent in healthy carriers with no prior history of IM and whose EBV-specific CD8⁺ T lymphocytes show a polyclonal 12 and stable TCR repertoire over time 4. EBV is an oncogenic virus involved in the development and progression of B, T and natural killer (NK) cell lymphoproliferative disorders as well as several epithelial tumours 13–16, which are frequently associated with immunodeficiency and/or immunosuppression and escape from immune surveillance mechanisms 17–19. However, in 2008, the World Health Organization (WHO) established a new category of aggressive B cell non-Hodgkin's lymphoma (B-NHL) associated with EBV infection, which typically affects individuals ≥ 50 years of age with no prior history of immunodeficiency 14,20. Although the pathogenesis of this EBV-associated B-NHL remains largely unknown, it has been associated with immune impairment or senescence linked to ageing 14,21 and a decrease in naive T cells in parallel with the accumulation of memory T lymphocytes with a narrow immune repertoire 22–24.

Most studies evaluating EBV-specific T cell responses have used lytic and latent-stage epitopes loaded in, for example, tetramers, peptide–human leucocyte antigen (HLA) multimers 4–6,25–29, B lymphoblastoid cell lines (LCLs) 6,12 and/or EBV lysate 30–33; however, in most of these studies, healthy carriers aged < 50 years have typically been analysed. Overall, based on these assays, the reported frequency of CD8⁺ T lymphocytes against EBV latent and lytic-stage antigens is < 2% 25,34–36, and that of CD4⁺ T lymphocytes is typically ∼1% 2,31,32,37. EBV-specific T cells are described as monofunctional populations secreting interferon (IFN)-γ 4,34,38–44 and multi-functional cells producing tumour necrosis factor (TNF)-α, IFN-γ and interleukin (IL)-2 [12,30,31,33,37,38,45] in addition to perforin, CD107a and macrophage inflammatory protein (MIP)-1α (CCL3) 45,46. In contrast, among individuals aged > 50 years, the results regarding anti-EBV T-cell responses in vitro are controversial. Thus, it is unclear whether CD4⁺ and CD8⁺ T lymphocytes behave as monofunctional and multi-functional cells in response to lytic and latent-stage antigens or LCLs 6,47, or whether there are CD8⁺ T lymphocytes dysfunctional in IFN-γ production in individuals > 60 years versus subjects younger than 40 years 26,39. Independent of age, EBV-specific T cell responses are typically ascribed to the central memory and effector memory T cell compartments 4,25–27,31,37,47,48.

Here, we evaluated the TCR-Vβ repertoire of different maturation-associated compartments of EBV-specific memory CD4⁺ and CD8⁺ T lymphocytes, which show cell membrane TNF-α⁺ expression after short-term in-vitro stimulation; our major goal was to investigate potential differences in the EBV-specific T cell repertoire of healthy adults grouped according to age (<50 versus ≥50 years), as in individuals aged ≥50 years there is a higher incidence of aggressive lymphomas associated with EBV. In parallel, we also measured the soluble and intracellular cytokine profiles in both groups of individuals.

Materials and methods

Subjects and samples included in this study

Heparin anti-coagulated peripheral blood (PB) samples were collected from 27 EBV-seropositive healthy adult volunteers. Of these subjects, 20 individuals were aged < 50 years (seven males and 13 females; median age 31 years; range 21–47 years) and seven were aged > 50 years (two males and five females; median age 64 years; range 52–83 years). In all cases, PB samples were obtained after written informed consent was provided by each individual, and the study was reviewed and approved by the Ethics Committee of the Pontificia Universidad Javeriana (Bogotá, Colombia). EBV serostatus was determined by an anti-virus capsid antigen (VCA)-specific immunoglobulin (Ig)G and IgM enzyme-linked immunosorbent assay (ELISA) assay (Vircell S.L., Granada, Spain). EBV plasma viral loads were determined using a commercial real-time polymerase chain reaction (PCR) technique (TIbMolBiol; Roche Diagnostics, Mannheim, Germany); plasma samples were negative for EBV DNA. Absolute leucocyte and lymphocyte counts were determined using a Coulter LH-750 haematology analyser, and lymphocyte subpopulations were determined by flow cytometry (FACSARIA II; Becton Dickinson Biosciences, San José, CA, USA). All samples were processed and measured within the first 12 h after they were collected.

Immunophenotypic identification and characterization of EBV-specific T lymphocytes

To evaluate T cell responses in vitro, PB stimulation with an EBV lysate was used, as described in detail previously 33,49. Briefly, two aliquots of PB samples in RPMI-1640 (1:1 vol : vol) adjusted to a cell concentration of 1 × 10⁶ leucocytes/μl were incubated for 6 h at 37°C in a 5% CO₂-humidified atmosphere with 5 μg/ml of an EBV lysate (strain B95·8; Advanced Biotechnologies, Columbia, MD, USA) together with 20 μM TAPI-2, a TNF-α protease inhibitor that induces TNF-α accumulation on the cell surface (Cytognos SL, Salamanca, Spain), and 1 μg/ml of both anti-CD28 (clone L293; Becton Dickinson Biosciences) and anti-CD49d (clone L25; Becton Dickinson Biosciences) as co-stimulatory monoclonal antibodies (BD Pharmingen, San Diego, CA, USA) 50. Another PB aliquot was processed under the same conditions but in the absence of the viral lysate (unstimulated sample), and used as a negative control. As a positive control, a third aliquot of PB was cultured with 25 ng/ml phorbol myristate acetate (PMA) (Sigma, St Louis, MO, USA), 1 μg/ml ionomycin (Sigma) and 20 μM of TAPI-2 49,51.

After stimulation all PB aliquots, including the controls, were incubated at room temperature and labelled with CD3-phycoerythrin cyanin 7 (PE-Cy7) (clone SK7), CD4-peridinin chlorophyll (PerCP) (clone SK3), CD8-allophycocyanin (APC)-Cy7 (SK1 clone), CD45RA-fluorescein isothiocyanate (FITC) (clone HI100), CCR7-PE (clone 3d12) and TNF-α-APC (clone 64,011,111) antibody combination (all purchased from Becton Dickinson Biosciences) or 2 ml/tube of fluorescence activated cell sorter (FACS) lysing solution ×1 (Becton Dickinson Biosciences) for 15 and 10 min, respectively.

Immediately after sample preparation, data acquisition was performed in a FACSAria II flow cytometer (Becton Dickinson Biosciences) using the FACSDiva software program (Becton Dickinson Biosciences) and a double-step data acquisition procedure. In the first step, approximately 5 × 10⁴ events from the entire cell population in the PB were stored. In the second step, the events were acquired through a live-cell gate containing only CD3⁺ cells (1–2 × 10⁵) and specifically gated for TNF-α⁺CD4⁺ and TNF-α⁺CD8⁺ T lymphocyte identification and characterization. For data analysis, paint-a-gate-pro software program (BDB) and FlowJo software (Tree Star, Ashland, OR, USA) were used. Instrument set-up, calibration and quality control were performed throughout the study using commercial standard reagents (BD cytometer set-up and tracking beads and the BD comp beads; Becton Dickinson Biosciences) according to the manufacturer's instructions.

TCR-Vβ repertoire of EBV-specific TNF-α⁺ T lymphocytes

Quantitative analysis of 24 TCR-Vβ families (approximately 70% coverage of the normal human TCR-Vβ repertoire) was conducted using PB samples from basal conditions and in EBV-specific TNF-α⁺CD4⁺ and TNF-α⁺CD8⁺ T lymphocytes following culture in the presence or absence of EBV lysate stimulation. Sample staining and analysis using the TCR-Vβ Repertoire Kit (Immunotech, Monrovia, CA, USA) were performed according to the manufacturer's recommendations 8,52,53 and the protocol described by Rodriguez-Caballero et al. 49. Briefly, PB samples were stained with anti-CD3-PE-Cy7, anti-CD4-Per-CP, anti-TNF-α-APC and anti-CD8-APC-Cy7 (Becton Dickinson Biosciences) monoclonal antibodies (mAbs) in combination in addition to the TCR-Vβ-FITC and TCR-Vβ-PE markers from the TCR-Vβ repertoire kit. Approximately 1–2 × 10⁵ CD3⁺ events/sample were measured using a FACSAria II flow cytometer and FACSDiva software. The paint-a-gate-pro software program was used for data analysis.

Measurement of soluble and intracellular cytokines produced in vitro by PB lymphocytes

Soluble cytokine levels were measured in the culture supernatants of in-vitro-cultured PB samples after short-term (6 h) stimulation with EBV or PMA+ionomycin or no stimulation, using the BD cytometric bead array (CBA) human T helper type 1 (Th1)/Th2 cytokine kit II according to the manufacturer's recommendations. The detection limits for each cytokine were as follows: IL-2 and IL-4, 2·6 pg/ml; IL-6, 3·0 pg/ml; IL-10 and TNF-α, 2·8 pg/ml; and IFN-γ, 7·1 pg/ml. The concentrations of individual cytokines were calculated from the corresponding (cytokine) calibration curve determined using the kit standards (range 0–5000 pg/ml) 49, and data were analysed using FCAP Array (Becton Dickinson Biosciences) software.

The intracellular production of TNF-α, IL-2 and IFN-γ was measured after short-term (6 h) incubation, as described. In-vitro culture of PB samples (1 × 10⁶/well) were incubated with or without EBV lysate in the presence of anti-CD28 and anti-CD49d. After 2 h of culture, brefeldin A (BFA, 1 μg/ml; Becton Dickinson Biosciences) was added to the culture. As a positive control, 1 × 10⁶ PB white blood cells (WBCs) stimulated with PMA (10 ng/ml) plus ionomycin (1 μg/ml) in the presence of BFA (1 μg/ml) were cultured in parallel under the same conditions. The cells were then incubated for 15 min in the dark at room temperature (RT) with CD3-PE-CY7 (clone SK7), CD4-PerCP (clone SK3) and CD8-APC-Cy7 (SK1 clone) (Becton Dickinson Biosciences). The cells were washed once (5 min at 540 g), fixed and permeabilized using the InstaStain reagent kit (Dako, Glostrup, Denmark) and subsequently stained with IFN-γ-FITC (clone 4515; Miltenyi Biotec, Bergisch Gladbach, Germany), TNF-α-PE (clone 6401·1111; Becton Dickinson Biosciences) and IL-2-APC (clone N7·48 A, Miltenyi Biotec) for 30 min (RT). Data were acquired in a FACSAria II flow cytometer and analysed using FACSDiva software (Becton Dickinson Biosciences).

Statistical analysis

The median values, mean, standard deviation (s.d.) and range were calculated for all variables using the spss software program (spss version 19; SPSS, Inc., Chicago, IL, USA). To establish the statistical significance of differences observed among different groups, the Wilcoxon and Mann–Whitney U-tests were used. P-values < 0·05 were considered statistically significant.

Results

Distribution and maturation compartments of EBV-specific T cells according to age

The overall number of PB CD3⁺ T lymphocytes was significantly lower in seropositive healthy individuals aged > 50 years versus subjects aged < 50 years (median 1119 versus 1786 cells/μl; range 688–1592 versus 1162–4975 cells/μl; P < 0·01); such decreases involved CD8⁺ T cells (median 401 versus 577 cells/μl) and were particularly prevalent among CD4⁺ T lymphocytes (641 versus 1079 cells/μl) (P < 0·01).

However, from a functional viewpoint, a similar polyclonal response in vitro to PMA+ionomycin was observed in both groups of subjects (P > 0·05), with similar percentages of both TNF-α⁺CD4⁺ T lymphocytes (median of 47·2 versus 51·7% for cases < 50 versus > 50 years, respectively) and CD8⁺ T lymphocytes (median 44·7 versus 49·7% for subjects < 50 years versus > 50 years, respectively) (Fig. 1a).

Fig. 1.

Frequency of tumour necrosis factor (TNF)-α⁺ T cells after in-vitro stimulation with phorbol myristate acetate (PMA)+ionomycin or an Epstein–Barr virus (EBV) lysate. (a) Frequency of CD4⁺ and CD8⁺ T cells (TNF-α⁺) after polyclonal stimulation with PMA+ionomycin in the peripheral blood (PB) of healthy subjects grouped according to age. (b) Relative frequencies of activated CD4⁺ and CD8⁺ T cells (TNF-α⁺) in the absence and presence of EBV lysate in individuals from the two age groups. (c) Comparison of the specific T cell response after stimulus with EBV lysate among healthy individuals. In all figures, the median and range are shown. *P < 0·05 compared with the condition without stimulus, as calculated by the Wilcoxon test, and between the age groups younger and older than 50 years, as calculated by the Mann–Whitney U-test.

The overall proportion of EBV-activated T lymphocytes (TNF-α⁺) 49 was also identified in the PB of healthy subjects after short-term in-vitro stimulation with EBV. The frequency of TNF-α⁺ T cells was significantly higher (P < 0·05) in EBV-stimulated versus unstimulated samples (Fig. 1b). Of note, similar numbers of EBV-specific CD4⁺ T lymphocytes (0·07 versus 0·06%; P > 0·05) and CD8⁺ T lymphocytes (0·2 versus 0·22%; P > 0·05) were found in cases aged younger and older than 50 years (Fig. 1c).

Based on CD45RA and CCR7 expression, EBV-specific CD4⁺ and CD8⁺ T lymphocytes were subdivided into four maturation-associated functional compartments 47,54: naive (CD45RA⁺/CCR7⁺), central memory (CD45RA⁻/CCR7⁺), effector memory (CD45RA⁻/CCR7⁻) and terminal effector (CD45RA⁺/CCR7⁻). For each age group, the EBV-specific T cell response was investigated in 10 representative cases (five per group). Regarding CD4⁺ T cells, we found that the memory compartments were distributed heterogeneously in both groups. However, while two of five of the < 50-year-old individuals had central memory cells, a heterogeneous distribution was observed for those aged > 50 years. It can be interpreted as a high maturation stage of CD4⁺ T cells, which can be evaluated in a larger cohort. In contrast, no difference was observed for CD8+ T cells, which showed a heterogeneous distribution in both groups (Fig. 2a). Notably, the interindividual variability was higher than the variation between the two age groups regarding the distribution of TNF-α⁺CD4⁺ or CD8⁺ T lymphocytes (P > 0·05), suggesting different immunological backgrounds for each individual (Fig. 2b).

Fig. 2.

Distribution of Epstein–Barr virus (EBV)-specific T cells [tumour necrosis factor (TNF)-α⁺] after in-vitro stimulation with an EBV lysate. (a) EBV-specific CD4⁺ (upper) and CD8⁺ (down) T lymphocyte distribution in the memory compartments: naive cells (CD45RA⁺ CCR7⁺), central memory (CD45RA⁺ CCR7⁺), effector memory (CD45RA⁻ CCR7⁻) and effector T cells (CD45RA⁺ CCR7⁻). The pie charts represent the distribution of the specific T cell response in the differentiation compartments after stimulation with EBV, in each of five individuals representative of the two age groups. (b) Frequency of TNF-α⁺ CD4⁺ and TNF-α⁺ CD8⁺ T lymphocytes in the differentiation compartments after EBV lysate stimulation in individuals younger and older than 50 years of age. The middle line and the vertical lines represent the median and ranges, respectively.

PB cell response to polyclonal stimulation according to age

Polyclonal PBMC stimulation with PMA+ionomycin revealed a significant response, characterized by the secretion of IFN-γ, TNF-α, IL-4 and IL-2 (P < 0·05), in all subjects analysed (data not shown), while no changes were observed in IL-10 production. However, significant secretion of IL-6 was restricted to subjects older than 50 years. Similar to PMA+ionomycin stimulation, EBV-specific responses included IFN-γ and TNF-α secretion but also IL-10 and IL-6 secretion (P < 0·05) in both age groups. In contrast, IL-4 and IL-2 secretion was found only in individuals aged < 50 years (data not shown).

Comparative analysis between the two age groups revealed that individuals aged > 50 years produced significantly higher levels of IFN-γ (P < 0·05) and more IL-6 (not significant) after polyclonal stimulation with PMA+ionomycin, which suggests the existence of an age-dependent proinflammatory profile (Fig. 3a). Similarly, after EBV stimulation, individuals older than 50 years produced more IFN-γ (median 50 versus 11 pg/ml in those younger than 50 years), TNF-α (64 versus 39 pg/ml in those younger than 50 years) and IL-6 (4131 versus 1655 pg/ml in those younger than 50 years) (P < 0·05), while younger subjects (< 50 years) showed greater IL-2 and IL-4 production (3 pg/ml versus undetectable in subjects older than 50 years) after in-vitro culture. No differences in the levels of secreted IL-10 were found between the two age groups (Fig. 3b).

Fig. 3.

Analysis of soluble cytokines after polyclonal stimulation or stimulation with an Epstein–Barr virus (EBV) lysate. (a) Comparative analysis of soluble cytokines (pg/ml) among age groups younger and older than 50 years of age after stimulation with PMA+ionomycin. In control samples, only dimethyl sulphoxide and ethanol were used. (b) Comparative analysis of soluble cytokines among age groups after stimulation with EBV lysate. In control samples, only the co-stimulating molecules anti-CD49 and anti-CD28 were used. The boxes represent the median and range data. *P < 0·05 and **P < 0·01 between age groups younger and older than 50 years of age, as calculated by the Mann–Whitney U-test.

Cytokine production by individual EBV-specific T lymphocytes according to age

T lymphocyte functionality was estimated according to their ability to produce one, two or three cytokines (intracellular detection) in response to PMA+ionomycin. Overall, no significant differences were found between younger and older subjects, except for a greater frequency of IFN-γ⁺ CD8⁺ T lymphocytes (P < 0·05) in older subjects (median 55·9 versus 30·2% in cases < 50 years) (Fig. 4a). The latter results are consistent with the higher levels of soluble IFN-γ secreted by T cells from individuals older than 50 years.

Fig. 4.

Percentage of cytokine-producing T cells after short-term stimulation with phorbol myristate acetate (PMA)+ionomycin or an Epstein–Barr virus (EBV) lysate. (a) Comparative analysis of the frequencies of monofunctional and multi-functional CD4⁺ and CD8⁺ T lymphocytes for tumour necrosis factor (TNF)-α, interleukin (IL)-2 and interferon (IFN)-γ after stimulation with PMA+ionomycin among individuals younger and older than 50 years of age. (b) Comparative analysis of T lymphocyte functionality between age groups after stimulation with EBV lysate. The middle line and the vertical lines represent the median and ranges, respectively. *P < 0·05 between age groups younger and older than 50 years, as calculated by the Mann–Whitney U-test.

After EBV stimulation, CD4⁺ and CD8⁺ T lymphocytes proved to be primarily monofunctional for the analysed cytokines, with frequencies of cytokine-positive cells below 1%. Of note, a greater frequency of IL-2⁺ EBV-specific CD4⁺ T lymphocytes (P < 0·05) was found among older subjects versus subjects aged < 50 years (median 0·08 versus 0·03% in younger individuals) (Fig. 4b). Interestingly, the presence of minor multi-functional EBV-specific T cell subpopulations, e.g. TNF-α⁺/IL-2⁺CD8⁺ T lymphocytes and both CD4⁺ and CD8⁺TNF-α⁺/IFN-γ⁺ double-positive T lymphocytes, was detected in subjects younger and older than 50 years, respectively (Fig. 4b).

The TCR-Vβ repertoire is polyclonal in both age groups, with some differences in family usage in response to EBV

Under basal conditions (without stimulus), the T cell repertoire was broad and diverse in both age groups. The average usage of 24 TCR-Vβ families evaluated in CD4⁺ T lymphocytes was 69·6% for individuals aged < 50 years and 71·2% for subjects older than 50 years. For CD8⁺ T lymphocytes, the TCR-Vβ repertoire for the same 24 TCR-Vβ families tested was 63·8 and 63·1% for cases younger and older than 50 years, respectively. The relative mean frequency of each family did not exceed 10% among either the CD4⁺ or CD8⁺ T lymphocyte subpopulations. However, subjects younger than 50 years showed higher expression of the TCR-Vβ18 and Vβ22 families (P < 0·05 versus older subjects) on CD4⁺ T lymphocytes (Fig. 5a); in contrast, a higher frequency of CD4⁺ and CD8⁺ T lymphocytes expressing the Vβ7·2 family (P < 0·05) was detected in individuals aged more than 50 years (Fig. 5b). No significant differences were observed between the two age groups for the other TCR-Vβ families evaluated, suggesting a somewhat stable TCR repertoire during adult life.

Fig. 5.

Basal and specific (TNF-α⁺) T cell receptor (TCR)-variable beta region (Vβ) repertoires of T cells after stimulation with an Epstein–Barr virus (EBV) lysate. (a and b) Frequencies of CD4⁺ T and CD8⁺ T cells in each of 24 TCR-Vβ families under basal conditions (no stimulation) in individuals younger and older than 50 years of age. In all figures, the mean and range are shown. *P < 0·05 between age groups younger and older than 50 years of age, as calculated by the Mann–Whitney U-test. (c,d) Percentage of individuals with EBV-specific CD4⁺ and CD8⁺ T cells (TNF-α⁺) in each TCR-Vβ family among individuals younger and older than 50 years of age.

After EBV stimulation, both age groups showed a polyclonal and diverse TCR-Vβ repertoire on EBV-specific T cells, which varied from individual to individual. Overall, EBV-specific TNF-α⁺CD4⁺ T cells from 19 and 23 of the 24 TCR-Vβ families evaluated were identified in individuals younger and older than 50 years, respectively. In turn, 16 and 23 of the 24 TCR-Vβ families were found on TNF-α⁺CD8⁺ T lymphocytes from cases younger and older than 50 years, respectively (Fig. 5c,d).

In both age groups, the TCR-Vβ families that were most frequently present on EBV-specific CD4⁺ T lymphocytes were Vβ1, Vβ2, Vβ17 and Vβ22 (Fig. 5c). In contrast, a more variable repertoire was found regarding the most represented TCR-Vβ families on EBV-specific CD8⁺ T lymphocytes from the two age groups. The Vβ13·1 family was found in nearly 80% of subjects younger than 50 years, followed by Vβ5·2, Vβ9, Vβ11, Vβ13·2, Vβ18 and Vβ23 (Fig. 5d); by contrast, the Vβ9 family was present in almost 86% of subjects older than 50 years, followed by Vβ5·1, Vβ8, Vβ18 and Vβ23 in 71·4% of the subjects, Vβ1, Vβ5·2 and Vβ13·2 in 60% and Vβ11 and Vβ13·1 in 43%.

Interestingly, our results showed that the TCR-Vβ families that were frequently represented within the EBV-specific CD4⁺ and CD8⁺ T cell subsets corresponded to those that were the most represented under basal conditions (Vβ1 in CD4⁺ T lymphocytes from subjects < 50 years and Vβ5·1 and Vβ13·2 within the CD8⁺ T lymphocytes from subjects > 50 years). However, some of the relevant EBV-specific TCR-Vβ families found on CD8⁺ T lymphocytes, e.g. Vβ13·1 (age group < 50 years) and Vβ9 (age group > 50 years), were not among the most frequent TCR-Vβ families observed under basal conditions.

Discussion

EBV is detected in more than 95% of the healthy population worldwide 1 and has been implicated in diseases such as lymphoma 14–16, which are more common in individuals older than 50 years. Given that both immunodeficiency and immunosenescence are phenomena that are potentially involved in viral reactivation, the characterization of the magnitude, duration and quality of EBV-specific T cell responses in healthy subjects is essential to determine whether an alteration of the immune response underlies viral reactivation or if, instead, it is a consequence of EBV reactivation.

In this regard, we characterized the T cell repertoire and EBV-specific in-vitro T cell response of healthy adults grouped according to age. Overall, our results showed a lower number of total CD3⁺ lymphocytes (both CD4⁺ and CD8⁺ T cells) among individuals aged more than 50 years. Such a decrease has been reported previously and has been associated with increased immunosenescence, particularly among individuals aged > 65 years, and may be related to thymic involution 22,23,55,56.

Despite these differences, similar percentages of TNF-α⁺ T lymphocytes and levels of secreted TNF-α, IL-2 and IL-4 were found in the two age groups after short-term PMA+ionomycin stimulation in vitro. By contrast, older subjects showed higher levels of secreted IFN-γ and IL-6 and a greater frequency of IFN-γ⁺CD8⁺ T lymphocytes (versus younger subjects) after in-vitro PMA+ionomycin stimulation 22,24,57. These findings could reflect more pronounced persistent antigenic stimulation associated with age leading to a chronic inflammatory state with progressive activation of organ- and tissue-resident macrophages and monocytes producing proinflammatory cytokines such as TNF-α, IL-1β and IL-6 22,24. Of note, the differences in the cytokine profile between the two age groups could have resulted from EBV reactivation or the presence of other infectious persistent agents such as cytomegalovirus (CMV) 22, which was not evaluated in our study.

In both age groups, a similar EBV-specific T cell response was found, with overlapping numbers of TNF-α⁺ T lymphocytes (CD4⁺ and CD8⁺), in addition to similar numbers of monofunctional and multi-functional T cells producing IL-2 and/or IFN-γ. However, individuals aged more than 50 years showed significantly higher frequencies of IL-2-producing CD4⁺ T lymphocytes in association with greater production of soluble IFN-γ, TNF-α and IL-6 compared with subjects younger than 50 years.

Jabs et al. 58 have demonstrated that the IFN-γ and IL-10 levels produced by PBMCs depleted of T lymphocytes are decreased compared with whole PBMCs exposed to LCL supernatants, indicating the direct involvement of T lymphocytes in EBV responses. However, the presence of these cytokines in the culture supernatant cannot be attributed only to T lymphocytes, because other mononuclear cells such as monocytes may also be involved, as shown in the evaluation of the response to herpesviruses such as CMV 49.

Overall, our results are consistent with previous reports for individuals aged < 50 years, regarding the secretion of soluble IFN-γ, IL-6 and IL-10 in the presence of LCLs or lytic and latent-stage antigens (e.g. BZLF1, BMLF1, EBNA-1, EBNA-3C, LMP1, LMP2) 4,6,34,40–44,58–60. In contrast, the higher levels of TNF-α, IFN-γ and IL-6 found among individuals aged more than 50 years is a new observation related to EBV responses; these results suggest that individuals aged more than 50 years have an EBV-competent immune response alongside low-grade inflammation that may emerge with age (e.g. naturally) 22.

CD4⁺ and CD8⁺ T lymphocytes capable of producing at least one cytokine and TNF-α⁺/IL-2⁺ and TNF-α⁺/IFN-γ⁺ double-positive T lymphocytes have been detected in both age groups in response to EBV 12,45,47. These results do not match fully with previous observations noting the age-dependent presence CD8⁺ T lymphocytes that were unable to produce IFN-γ 26,39, suggesting that immune exhaustion may be restricted to specific EBV epitopes, not the EBV response as a whole. Moreover, the presence of IL-2-producing T lymphocytes in the two age groups demonstrates the ability of T cells to respond to a second stimulus and to participate in antigen removal 61,62. However, it should be noted that individuals older than 50 years, who showed higher frequencies of CD4⁺/IL-2⁺ cells, had a lower overall cytokine content per cell than subjects younger than 50 years (mean fluorescence intensity of 15 versus 28; P > 0·05). These findings can explain, at least in part, the higher concentrations of soluble IL-2 found among the younger subjects after in-vitro EBV stimulation.

In the majority of individuals in both groups, EBV-specific CD4⁺ and CD8⁺ T lymphocytes (TNF-α⁺) included central and effector memory T cells as well as effector T lymphocytes, although great interindividual variability was observed in the absence of significantly different distributions according to age. Such variability, particularly among CD4⁺ T lymphocytes in a significant proportion of subjects (TNF-α⁺CD4⁺ EBV-specific T cells were restricted to a single but variable compartment) may partially explain the differences observed in cytokine production between individuals at the same time 62; it might also be associated with the individual's EBV replication history 30. The heterogeneous distribution of specific T cells directed against EBV in the memory compartment has been shown previously 4. In fact, distinct patterns of CD8⁺ T cell differentiation and expansion may be primed by different routes of antigen exposure or particular homeostatic environments occurring in different viral infections, as well as whether the cells respond to epitopes from lytic or latent EBV proteins 63.

Previous reports, in which TNF-α, IL-2 or IFN-γ production has been analysed in adults of different age (< 45 years or > 60 years) 4,25,26,47,48, have demonstrated major EBV-specific responses in the effector memory T cell compartment after the EBV-specific stimulation of both CD4⁺ and CD8⁺ T lymphocytes 31,32,37. Additional data indicate that even among nonagenarians, only a minority of EBV-specific CD8⁺ T lymphocytes are CD28⁻ (effector-suppressor compartment) 26,27. In persistent infections, including EBV, CMV and HSV-1 (herpes simplex virus, type 1), virus-specific CD4⁺ T lymphocytes are also concentrated at the effector stage 64, including cells in the central memory to effector T cell stages 30. In addition, the presence of these specific T cell populations has been associated with the control of persistent HIV infection among non-progressors 48,62,65.

Our findings suggest that EBV-specific T cells (using a model of polyclonal stimulation with a viral lysate) are distributed throughout several differentiation compartments in an age-independent manner, with the presence of monofunctional and multi-functional T lymphocytes.

Under basal conditions and in the presence of EBV, a polyclonal and highly overlapping TCR-Vβ repertoire of both CD4⁺ and CD8⁺ T lymphocytes was found in both age groups. These findings confirm that clonal diversity (in the absence of a stimulus) is maintained in healthy individuals aged between 0 and 86 years, as also found previously 66. However, the predominant TCR-Vβ family expressed by EBV-specific cells varied from individual to individual, although the TCR-Vβ1, Vβ2, Vβ17 and Vβ22 TCR-Vβ families were the most common form to be expressed among TNF-α⁺ CD4⁺ T lymphocytes in both age groups. Similarly, TNF-α⁺ CD8⁺ T lymphocytes in the two age groups also showed common expression of the Vβ5·2, Vβ11, Vβ13·2, Vβ18 and Vβ23 TCR-Vβ families, although greater frequencies of Vβ13·1 and Vβ9 were found in younger and older subjects, respectively. Altogether these results suggest that a common selection of TCR-Vβ families could be relevant in controlling EBV infection in healthy seropositive individuals.

Several TCR-Vβ families have been reported previously to be associated with EBV-specific T cell repertoires 4,7,12,67–70; in particular, for CD8⁺ T lymphocytes, the TCR-Vβ2, Vβ4, Vβ16, Vβ18 and Vβ22 families have been shown to recognize GLC (BMLF1) 4,7,70,71. However, the relative frequency of individual TCR-Vβ families in healthy individuals aged < 50 years has been shown to fluctuate over time 4. Specific epitope-targeted responses vary from a single TCR, e.g. TCR-Vβ13 (TRVB6) to FLR (EBNA-3A) 67,68, to multiple families (e.g. 20 families for RAK; BZLF1) 71, which may also explain, at least in part, the multiple families involved. In this regard, a polyclonal/oligoclonal repertoire has been reported for CD4⁺ T lymphocytes with expansion of the Vβ2·1, Vβ8·1, Vβ12·2, Vβ13·6, Vβ17·1 and Vβ22·1 TCR families during IM; afterward, the oligoclonality against ex-vivo LCLs is retained 8,43,72. In the present study, we found that Vβ5·2, Vβ5·3, Vβ7·1, Vβ7·2 and Vβ13·2 were expressed by EBV-specific CD8⁺ T cells and that Vβ13·2 and Vβ5·2 were present in approximately 50% of individuals of both age groups.

This considerable interindividual variability might be due to genetic polymorphisms 28, fluctuations in epitope-specific T cell frequencies 4–6, associations between some HLA class II and class I haplotypes and the latent or lytic epitopes recognized by specifically expanded TCR-Vβ clones [52], and differences in the EBV-specific CD8⁺ T cell clones migrating preferentially to the bone marrow rather than to lymphoid tissues or the peripheral blood 29.

Regarding senescence of the immune system, in our study we found that individuals aged more than 50 years had a reduced number of CD3⁺ cells, with diminutions of both CD8 and, to a greater extent, CD4 T cells. Regardless of this diminution, CD8 cells produce a high quantity of IFN-γ after stimulation with PMA/ionomycin or EBV lysate (Fig. 3), suggesting that it is an important effector response both for bulk T cells and the EBV-specific population. Other studies of immunosenescence have focused on very elderly subjects (mean age 90 years), and they have typically not considered the whole T cell population, only those cells specific for the immunodominant epitope 39. Analysis of the immune response to subdominant epitopes has shown that exhaustion of the immune response occurs principally in cells directed against immunodominant epitopes and that responses to subdominant epitopes are highly protective 73. Our stimulation protocol using the virus lysate allowed us to reveal the entire repertoire of T cells specific for EBV, not just the epitope-specific T cells, which may explain why we detected functional T cells able to produce IFN-γ after a specific antigenic stimulus.

We can also compare our results with those found for other chronic viral diseases, particularly those for the CMV-specific CD8⁺ T cell population, which is more studied. It is well known that the CMV response is significantly more differentiated than that against influenza 74, EBV 26, hepatitis C and HIV 75, and a large proportion of the responding cells are CD45RA, making them resistant to apoptosis 76. In fact, most T cell expansion in elderly subjects is CMV-specific 77, and these cells are impaired in their ability to secrete IFN-γ.

The origin of the differences in T cell responses between CMV and EBV is unknown, but the findings of Appay et al. 75 indicate that substantial enrichment of a particular phenotype occurs according to the viral specificity (EBV, CMV, HIV-1 and HCV). Their findings also indicated that these phenotypical enrichments are not related to cellular activation status (generally linked to viral load) or to the age of the patients. The authors suggest that CD8⁺ T cell responses with distinct characteristics arise in different chronic virus infections and that it is inappropriate to make generalizations about virus-specific CD8⁺ T cell responses without considering the particular infecting virus 75.

In summary, our results show that EBV-specific T cell responses are similar in healthy seropositive subjects aged < 50 and > 50 years, with a relatively high variety of functional memory and effector T lymphocytes persisting over time. However, a proinflammatory environment appears to be more evident among older subjects.

The stimulation pattern used herein will enable the evaluation of the EBV-specific T cell response in terms of the quantity, functionality, phenotype and clonotypic distribution in patients with EBV-associated diseases. Immune evaluation could be used as prognostic factor and monitoring tool.

Author contributions

S. F., A. O. and S. M. Q. conceived and designed the work. D. C., G. V. and S. M. Q. acquired and analysed the data. D. C., G. V., S. F., A. O. and S. M. Q. interpreted the data. D. C., G. V., S. F., A. O. and S. M. Q. drafted and critically revised the paper. D. C., G. V., S. F., A. O. and S. M. Q. gave final approval of the version to be published.

Disclosures

The authors declare no conflicts of interest.