Older (≥55 years) and young (≤45 years) KT recipients were recruited from the Erasmus Medical Center and prospectively followed up for 3 years with blood sampling between August 2018 and May 2023. To minimize inter-assay variation, only complete series (i.e., prior to transplantation and 6, 12, and 36 months after transplantation) of recipient material were assayed within one experiment. The inclusion criteria for analysis were the absence of T-cell depletion therapy and stable graft function without evidence of recurrent disease and chronic rejection. For this study, heparinized peripheral blood samples at all designated timepoints from a total of 24 older recipients were present, with one participant being excluded due to chronic TCMR and one due to poor sample quality ( Table 1 ). Similarly, samples from 25 young recipients were collected, but one participant was excluded due to having received T-cell depletion therapy and four due to poor sample quality ( Table 1 ). All recipients received similar non-depleting induction therapy with basiliximab (Simulect®, Novartis Pharma, Amsterdam, the Netherlands). Maintenance therapy for the majority of patients consisted of tacrolimus (Prograf®, Astellas Pharma, Leiden, the Netherlands) and mycophenolate mofetil (MMF) (CellCept®, Roche, Woerden, the Netherlands). This study was approved by the Medical Ethical Committee of the Erasmus Medical Center, and all participating KT recipients gave written informed consent to participate in this study (MEC No. 2018–048). This study was conducted in accordance with the Declaration of Helsinki and the Declaration of Istanbul and in compliance with the International Conference on Harmonization Good Clinical Practice regulations.

Peripheral blood mononuclear cells (PBMCs) were isolated from heparinized peripheral blood samples on the day of blood sampling as described previously (19, 20) and stored until further use.

Stimulator PBMCs from a donor or a third-party control were depleted of CD3+ T cells (>98% depleted) using CD3 microbeads (Miltenyi Biotec, Bergisch Gladbach, Germany), according to the manufacturer’s instructions. Third-party stimulator cells had an equal number of, but different, Human leukocyte antigen (HLA) mismatches with the tested recipient as the donor. In Supplementary Table 1 , the HLA typing for each kidney transplant recipient and source of stimulator cells (donor and third party) are listed. In addition, the assays for which the specific samples were used are listed as well including a reason for exclusion, when appropriate.

The CD137 multiparameter flow cytometric assay was used to determine the frequencies of cytokine-producing cells as described previously (8). Following 15 hours of stimulation, cells were stained at the cell surface and intracellular as described before (8) using antibodies listed in Supplementary Table 2 . Samples were measured on the Symphony A3 light (BD Biosciences, San Jose, CA, USA). During measurement, 0.5–1 million viable T cells were stored for subsequent analysis of frequencies of cytokine-producing cells and phenotypic characteristics using Kaluza version 2.1 software (Beckman Coulter, Woerden, the Netherlands). A representative example of a gating strategy for T-cell phenotype and cytokine expression is illustrated in Supplementary Figure 1 . Allo-stimulated samples with poor-quality donor or third-party spleen cells (reflected by the absence of proliferation above the background when compared to the unstimulated condition) were excluded. This resulted in a total of N = 21 older and N = 19 young KT recipients analyzed for cytokine data.

PBMCs of KT recipients were measured for proliferative capacity, as described previously (8, 21). Following 6 days of stimulation, supernatants were collected and stored until further use, cells were harvested, and extracellular staining was performed with the exclusion of CD137-directed antibody ( Supplementary Table 2 ). Proportions of proliferating [carboxyfluorescein succinimidyl ester (CFSE)-negative] T cells were determined, and subsequent analysis was performed using Kaluza version 2.1 software (Beckman Coulter, Woerden, the Netherlands). A representative example of a gating strategy to identify proliferating T cells is included in Supplementary Figure 2 . Allo-stimulated samples with poor-quality donor or third-party spleen cells (reflected by the absence of proliferation above the background when compared to the unstimulated condition) were excluded. This resulted in a total of N = 21 older and N = 20 young KT recipients analyzed for proliferation. In a separate set of experiments, the effect of exogenous IL2 on the proliferation of CD8+ T cells to alloantigen stimulation was explored. For this purpose, CD8+ T cells were isolated from CFSE-labeled PBMCs from seven young and 12 older KT recipients using the untouched CD8+ T-cell isolation kit (Miltenyi Biotec), according to the manufacturer’s instructions. Obtained CD8+ T-cell fractions had a purity above 94%. CD8+ T cells were stimulated with alloantigen (irradiated CD3-depleted PBMCs) in the absence and presence of increasing concentrations of human recombinant IL2 (Proleukin®, Clinigen Healthcare Ltd., Burton, UK). For a smaller group of KT recipients (three young and five older KT recipients), recombinant human IL15 (10 ng/mL; PeproTech EC Ltd., London, UK) was also added. In addition, isolated CD8+ T cells were stimulated with different concentrations ranging from 0.16 to 5 μg/mL of) plate-bound anti-CD3 (clone UCHT-1, BD). Following 6 days of stimulation, the proliferation of CD8+ T cells was measured via flow cytometry and analyzed as described above. Moreover, percentages of low- (IL2Rα), intermediate- (IL2Rβγ), and high-affinity (IL2Rαβγ) IL2 receptors were determined on day 0 and day 6 following staining of CD8+ T cells with antibodies directed to IL2Rα (CD25), IL2Rβ (CD122), and IL2Rγ (CD132) ( Supplementary Table 2 ) and measured by flow cytometry.

After 6 days of proliferation in a mixed lymphocyte reaction, the concentration of IL2 in the supernatant was measured using the BD™ CBA Flex Set (BD, Erembodegem, Belgium) in conjunction with the BD CBA Human Soluble Protein Master Buffer Kit (BD), following the manufacturer’s instructions. Concentrations were calculated using linear regression analysis in GraphPad Prism 9.0.0 (GraphPad Software Inc., San Diego, CA, USA).

Statistical analyses and graphs were created using GraphPad Prism 9.0.0 (GraphPad Software Inc., San Diego, CA, USA). Due to missing values, mixed-effects analysis instead of two-way ANOVA was used to compare the time course of cytokine-producing and proliferating T cells for young and older KT recipients. Post-hoc analyses were used to evaluate where significant effects were noted in time or between age groups. Appropriate paired and unpaired parametric tests were used to compare different timepoints within an age category and between different age categories per timepoint, respectively. Normally distributed data are depicted as the mean and standard error of the mean (SEM), and non-normally distributed data are depicted as the median and interquartile range (IQR). A p-value <0.05 indicates a statistically significant difference.

Prior to transplantation, older KT recipients had a higher percentage (0.75% ± 0.07%) of alloreactive CD137+ CD4+ T cells compared to younger KT recipients (0.48% ± 0.06%) (p < 0.01) ( Figure 1A ). The percentage of alloreactive CD137+ CD8+ T cells was also higher (0.36% ± 0.04%) in older KT recipients compared to 0.23% ± 0.03% in younger KT recipients (p = 0.02) ( Figure 1B ). Alloreactive CD4+ T cells were primarily of the memory phenotype in both young and older KT recipients. Of the alloreactive CD4+ T cells, young KT recipients had 25.5% ± 2.2% with a central memory (CM) and 37.6% ± 2.5% with an effector memory (EM) phenotype compared to 28.9% ± 1.8% CM and 47.5% ± 2.3% EM (p < 0.01) in older KT recipients ( Figure 1A ). Alloreactive CD8+ T cells were primarily of the naïve T-cell phenotype in young KT recipients where percentages amounted to 59.2% ± 4.1% compared to 33.6% ± 3.5% in older KT recipients (p < 0.01). In older KT recipients, the majority of alloreactive CD8+ T cells had a terminally differentiated effector memory (TEMRA) phenotype, which was significantly higher (p < 0.01) at percentages of 36.3% ± 3.4% compared to 20.9% ± 2.8% in younger KT recipients ( Figure 1B ).

Although the total frequency of alloreactive CD4+ T cells was higher in older KT recipients, the percentage of alloreactive CD4+ T cells expressing pro-inflammatory cytokines (IFNγ, IL2, and TNFα) was lower compared to that in younger KT recipients prior to transplantation, reaching significance (p = 0.02) for the subset expressing IL2 and TNFα (IL2+TNFα+) ( Figure 1C ). The percentage of alloreactive CD4+ T cells expressing one, two, and three pro-inflammatory cytokines in older KT recipients amounted to 22.0% ± 1.6%, 7.9% ± 0.7%, and 2.1% ± 0.4%, respectively, compared to those in younger KT recipients with 22.2% ± 1.5% (p = 0.95), 9.3% ± 0.8% (p = 0.17), and 2.2% ± 0.3% (p = 0.83), respectively ( Figure 1C , top). Older KT recipients had significantly lower frequencies (5.6% ± 0.6%) of alloreactive CD4+ T cells expressing both IL2 and TNFα compared to 7.8% ± 0.8% in younger KT recipients (p = 0.02) ( Figure 1C , bottom). No significant difference in frequency of alloreactive CD8+ T cells expressing multiple cytokines was identified between age groups prior to transplantation apart from a lower frequency of triple cytokine-producing cells amounting to 0.14% ± 0.03% for older compared to 0.27% ± 0.06% for young KT recipients (p = 0.04) ( Figure 1D ).

Following transplantation, a significant effect of time on the frequency of donor-reactive CD4+ T cells expressing more than one pro-inflammatory cytokine was observed (p < 0.01) ( Table 2 ). The effect of time on the frequency of polyfunctional donor-reactive CD4+ T cells varied across age groups with a significant interaction between time and age observed for donor-reactive CD4+ T cells expressing multiple pro-inflammatory cytokines including those expressing two (p = 0.04) and in particular the combination IL2+TNFα+ (p = 0.03) ( Table 2 ). This finding indicates a more rapid decrease in polyfunctional donor-reactive CD4+ T cells in the older KT recipients.

In general, no effect of time or age was observed for samples stimulated by third-party antigens as well as for frequencies of donor-reactive CD8+ T cells producing multiple pro-inflammatory cytokines ( Table 2 ).

A significant decline in polyfunctional donor-reactive CD4+ T cells was observed 6 months post-transplant for both age groups ( Figure 2A , left panel). The average ± SEM percentage of polyfunctional CD4+ T cells declined from 9.7% ± 1.3% to 5.1% ± 1.3% in older (p = 0.02) and from 12% ± 1.3% to 5.8% ± 1.4% in young (p < 0.01) KT recipients ( Figure 2A , left panel). Following 6 months, the frequency of polyfunctional donor-reactive CD4+ T cells remained more or less at the same level for both age groups ( Figure 2A , left panel). The observed decline was donor-specific, as third party-reactive CD4+ T cells did not decline over time ( Supplementary Figure 3A ).

Within the triple+ donor-reactive CD4+ T-cell population, a decline was observed for both age groups, reaching significance 12 months after transplantation. The proportion of triple+ declined from 1.7% ± 0.4% to 0.6% ± 0.2% in older (p < 0.05) and from 2.1% ± 0.4% to 0.9% ± 0.3% in young (p < 0.05) KT recipients ( Figure 2B , left panel). The frequency of triple+ donor-reactive CD4+ T cells remained low 36 months post-transplant at 0.7% ± 0.2% (p = 0.02) and 0.7% ± 0.2% (p < 0.01) in older and young KT recipients, respectively ( Figure 2B , left panel).

Proportions of double+ and IL2+TNFα+ donor-reactive CD4+ T cells of both young and older KT recipients declined significantly 6 months after transplantation ( Figures 2C, D , left panel). Older KT recipients reached a lower frequency of IL2+TNFα+ donor-reactive CD4+ T cells at 6 months post-transplant declining from an average percentage pre-transplant of 5.6% ± 0.8% to 2.2% ± 0.4% (p < 0.01) compared to 8.5% ± 1.1% to 3.4% ± 0.9% in younger (p < 0.01) KT recipients ( Figure 2D , left panel). Younger KT recipients reached this level after 12 months; i.e., the average proportion amounted to 2.4% ± 0.4% ( Figure 2D , left panel).

No significant differences were observed for cytokine-producing donor-reactive CD8+ T cells ( Figures 2A–D , right panel) for stable young and older kidney transplant recipients after transplantation.

Prior to transplantation, the percentage of proliferating alloreactive T cells in response to stimulation with alloantigen was significantly different between age groups for CD8+, but not CD4+, T cells ( Figure 3A ). Older KT recipients had a similar percentage of proliferating CD4+ T cells (23.1% ± 2.9%) compared to 27.4% ± 3.0% in young KT recipients (p = 0.31) and significantly fewer proliferating CD8+ T cells (26.0% ± 3.4%) compared to 37.3% ± 3.5% in younger KT recipients (p = 0.02) ( Figure 3A ). Of note, polyclonal stimulation of PBMCs with phytohemagglutinin (PHA) leads to substantial proliferation of CD8+ T cells, irrespective of age; i.e., average (SEM) % of proliferating CD8+ T cells amounted to 78.4% ± 3.9% versus 86.8% ± 3.0% (p = 0.09) for older and young KT recipients, respectively.

There was a significant effect of time after transplantation for the percentage of proliferating CD4+ T cells (p < 0.01) and CD8+ T cells (p < 0.01) ( Table 3 ). In addition, a significant effect of the interaction between time and age (p < 0.01) was observed for the percentage of CD8+ T cells that proliferated in response to donor antigens ( Table 3 ). No effect of time or age was observed for samples stimulated by third-party antigens ( Table 3 ).

The average of proliferating CD4+ T cells declined in both age groups but remained fairly stable after 6 months post-transplant ( Figure 3B , top). A similar but more pronounced pattern was seen for the CD8+ T cells. The average ± SEM of proliferating CD8+ T cells in older KT recipients declined to 19.1% ± 4.6% after 6 months and then decreased further to 15.2% ± 4.5% after 36 months ( Figure 3B , bottom). Younger KT recipients had a significant decline in proliferating CD8+ T cells over time from 37.6% ± 5.0% at pre-transplant to 21.3% ± 4.1% at 6 months post-transplant (p < 0.01) ( Figure 3B , bottom). One year after transplantation, the percentage of proliferating CD8+ T cells in young KT recipients declined further to 19.9% ± 4.2% and then to 15.7% ± 3.1% at 36 months post-transplant, reaching a similar level as older KT recipients ( Figure 3B , bottom). No change in the percentage of proliferating T cells in response to third-party was observed over time for both age groups ( Supplementary Figure 3B ). Frequencies of alloreactive CD137+ CD4+ T cells capable of producing the combination of IL2 and TNFα tended to positively associate with the alloreactive proliferative potential of CD8+ T cells for KT transplant recipients (N = 75, Pearson’s R = 0.20, p = 0.08) ( Figure 3C ). However, older KT recipients (N = 22) had significantly lower concentrations of IL2 in cell culture supernatants. Values amounted to 10 pg/mL (IQR 8.0–16.2) compared to 15.7 pg/mL (IQR 13.3–18.3) (p = 0.01) for the younger group (N = 20) ( Figure 3D ).

The results suggested that decreased IL2 production by alloreactive CD4+ T cells mediated the decreased proliferation of alloreactive CD8+ T cells in the older KT group. To test this hypothesis, CD8+ T cells were isolated from PBMCs of both young and older KT recipients prior to transplantation and evaluated the effect of increasing concentrations of exogenous IL2 on alloreactive CD8+ T-cell proliferation. First, the proliferation of isolated CD8+ T cells to an allogeneic stimulus was almost absent in the older KT recipients, and average ± SEM % amounted to 5.3% ± 1.1% as compared to 19.7% ± 1.8% in young KT recipients (p < 0.01). A dose–response effect was observed for increasing concentrations of IL2 added to alloantigen-stimulated CD8+ T cells in all young KT recipients but not in the majority of older KT recipients ( Figures 4A, B ). Only two out of 12 older KT recipients reached levels of the young KT recipients at 200 U/mL IL2, and average proliferation amounted to 17.3% ± 5.3% versus 44.7% ± 5.3% for the older versus young KT recipients, respectively. Basal expression of the low-, intermediate-, and high-affinity IL2 receptors was similar for young and elderly KT patients ( Supplementary Figure 4A ), but in the young exogenous IL2, it leads to induction of the high-affinity IL2 receptor (p < 0.01, Supplementary Figure 4B ). Percentages amounted to 12.9% ± 1.1% versus 2.8% ± 1.6% for young and older KT recipients, respectively. Next, whether alloantigen-stimulated CD8+ T cells of older KT respond to IL15 was evaluated, as this cytokine shares both IL2Rβ and IL2Rγ with IL2 ( Figure 4B ). At 10 ng/mL, the average alloantigen-induced proliferation of CD8+ T cells of older KT recipients amounted to 18.6% ± 4.6% versus 56.3% ± 6.5% for younger KT recipients (p < 0.01). Strong proliferative CD8+ T-cell responses were observed after polyclonal stimulation with plate-bound anti-CD3, independent of age and the presence of exogenous IL2 or IL15 ( Figure 4C ). A dose–response curve of CD8+ T-cell stimulation with decreasing plate-bound anti-CD3 was performed to investigate whether the sensitivity of T-cell receptor (TCR) signaling differed for CD8+ T cells between young and old KT recipients. The results showed that proliferative response curves were practically overlapping, indicating a similar threshold for TCR-mediated induction of CD8+ T-cell proliferation ( Supplementary Figure 5 ).

Given these results, exhaustion of CD8+ T cells of older KT recipients is not likely, and optimal stimulation of the TCR can overcome the lack of IL2 and IL15 responsiveness.