The Presence of Activated T Cell Subsets prior to Transplantation Is Associated with Increased Rejection Risk in Pancreas Transplant Recipients

Type 1 diabetes (T1D) is an autoimmune disease caused by the destruction of pancreatic β cells by T cells of the immune system (1). Whole organ pancreas transplantation (PTx) or transplantation of isolated pancreatic islets are the current clinical options to achieve insulin independence or improved glycemic control in T1D patients (2–4). Some T1D patients develop life-threatening complications, including hypoglycemia unawareness and end-stage renal disease (nephropathy), which warrant whole organ PTx and which are performed either as an isolated procedure (pancreas transplantation alone [PTA]) or combined with kidney transplantation (simultaneous pancreas–kidney [SPK]) in patients with diabetic nephropathy (5). Overall, PTx reduces the progressive secondary complications in T1D patients, which include retinopathy, neuropathy, nephropathy, and cardiovascular disease (2, 5).

Despite the advantages of PTx, a gradual decrease in graft function makes most patients revert to insulin therapy within two to five years posttransplantation (4). The major causes of a decline in graft survival are conventional T cell–mediated allograft rejection, recurrence of autoimmune assault by T cells, and the adverse effects of the immunosuppressive treatment (6–10). The conventional T cell–mediated allograft rejection is primarily caused by the interaction of recipient T cells with donor MHC molecules through direct or indirect allorecognition mechanisms (11). Overexpression of MHC class I at baseline and MHC class II upon T cell infiltration in islet cells indicates a key role for both CD4⁺T cells and CD8⁺T cells in allograft rejection (12). During the natural course of the disease, there is a massive infiltration of immune cells, particularly CD4⁺ T cells and CD8⁺ T cells within islets, suggesting a role in immune-mediated destruction and disease progression (12–15). Accumulating evidence also suggests that recurrent autoimmunity after transplantation impairs graft survival (4, 7, 8). The activated CD4⁺ T cells are instrumental in enhancing CD8⁺ T cell responses and stimulate islet-resident macrophages and provide help to B cells in secreting autoantibodies (6). The presence of autoreactive T cells in T1D patients prior to islet transplantation is also associated with lower graft survival (9). Similarly, patients that tested positive for either one or two of the T1D-associated autoantibodies (glutamic acid decarboxylase 65 and islet cell autoantigens IA2 and 512) are associated with reduced graft function (4). However, the relative contributions of allorecognition and recurrence of autoimmunity in organ rejection are largely unknown.

The majority of the preclinical and clinical studies of T cells have been performed in islet transplantation, and their role in the whole organ PTx recipients is poorly understood. In this context, in-depth analyses of T cell subsets are necessary to advance our understanding of PTx rejection. In this study, we aimed to characterize the circulating T cell subsets based on their phenotype, activation, differentiation, proliferation, and effector mechanisms from PTx patients prior to and 1 y post-PTx. Our findings suggest that the presence of activated CD4⁺ and CD8⁺ T cells prior to transplantation is strongly associated with PTx rejection and CD4⁺ and CD8⁺T cell activity 1 y post-PTx.

The Regional Ethics Committee of the Southern and Eastern Norway Regional Health Authority approved the study (Regional Ethics Committee for South-Eastern Norway approval number 012/2278-33). All patients were recruited at Oslo University Hospital (Oslo, Norway) and gave written informed consent to participate in this study. Clinical characteristics of all patients are described in Table I. Patient blood samples were collected at baseline prior to transplantation and at 3 wk, 6 wk, and 1 y post-PTx. Baseline samples were collected before starting immunosuppression on the day of transplantation. Additional samples were taken at any time on indication (suspected rejection). Only patients with biopsy-proven rejections classified according to Banff criteria were deemed “rejectors” (16–18). For two patients, we included blood samples from week 62 and week 112 as the time point closest to when rejection was diagnosed. For the other patients, samples from week 52 were included as time point closest to assessment of rejection status. All patients received induction therapy with a combination of anti-thymoglobulin, prednisolone, tacrolimus (TAC) and mycophenolate mofetil. The combination of prednisolone, TAC, and mycophenolate mofetil was used as maintenance treatment. Because of severe depletion of CD3⁺ T cells following induction therapy (19), the 3 and 6 wk samples post-PTx were not included in this study.

Peripheral blood samples were collected in EDTA tubes (BD Vacutainer), and PBMCs were separated by Ficoll density gradient centrifugation (Axis-Shield Diagnostics). PBMC was cryopreserved in FCS with 10% DMSO at −180°C in the vapor phase of liquid nitrogen. Blood samples from healthy donors (n = 6) were obtained from Oslo University Hospital Blood Center (Oslo, Norway).

Multicolor flow cytometry was used for immunophenotyping. PBMC (1 × 10⁶ cells) were stained to characterize lymphocytes, myeloid cells, and innate-lymphoid cells in multiple 14-color fluorochrome panels. Fluorescence minus one was used as a gating control. Before Ab staining, thawed PBMC were incubated with DNase I (Roche) to prevent cell clumping and followed by Human TruStain FcX (BioLegend) to reduce nonspecific binding. Cells were stained with anti-CD3 PerCP-Cy5.5 (UCHT1), anti-CD3 Brilliant Violet 421 (UCHT1), anti-CD8 PerCP-Cy5.5 (RPA-T8), anti-CD8 Brilliant Violet 510 (RPA-T8), anti-CD4 allophycocyanin-H7 (RPA-T4), anti-CD45RA Brilliant Violet 510 (HI100), anti-CD127 Brilliant Violet 786 (HIL-7R-M21), anti-IgD allophycocyanin-H7 (IA6-2), anti-FOXP3 PE-CF594 (259D/C7), anti-ICOS PE (DX29), anti-CD25 Brilliant Violet 421 (M-A251), anti-integrin β7 Brilliant Violet 421 (FIB504), anti-granzyme B PE (GB11), anti-TCRγδ Brilliant Violet 650 (B1), anti-HLA-DR PE (TU36), anti–IL-2 PE (MQ1-17H12), anti–INF-γ PE-Cy7 (B27), and anti–TNF-α allophycocyanin (Mab11) were from BD Biosciences. Anti-CD8 FITC (SK1), anti-CD27 Brilliant Violet 650 (O323), anti-CD19 PerCP-Cy5.5 (SJ25C1), anti–Ki-67 Brilliant Violet 711 (Ki-67), anti–PD-1 PE-Cy7 (EH12.2H7), anti-CXCR5 allophycocyanin (J252D4), anti-CD49d PE (9F10), anti–CTLA-4 allophycocyanin (L3D10), anti-CCR9 allophycocyanin (L053E8), anti-CCR7 Brilliant Violet 421 (G043H7), anti-CCR7 allophycocyanin (G043H7), anti-CD11c Brilliant Violet 510 (3.9), anti-CD16 FITC (3G8), anti-CD56 FITC (HCD56), anti-CD14 PE-Cy7 (HCD14), anti-CD123 allophycocyanin (6H6), anti-CD1a FITC (HI149), anti-CD11c FITC (Bu15), anti-CD19 FITC (HIB19), anti-CD3 FITC (UCHT1), anti-CD4 FITC (RPA-T4), anti-CD14 FITC (HCD14), anti-CD34 FITC (581), anti-CD123 FITC (6H6), anti-TCRαβ FITC (IP26), anti-TCRγδ FITC (B1), anti-CD16 FITC (B73.1), and anti-CD94 FITC (DX22) were from BioLegend. Anti-perforin PE-Cy7 (dG9) was from eBioscience. All surface markers were stained first and followed by fixable viability dye 700 (BD Biosciences). The intracellular markers were then stained using the Human FoxP3 Buffer Set (BD Biosciences) according to the manufacturer’s instructions. All multicolor flow cytometry data were acquired with a BD LSRFortessa and analyzed using FlowJo version 10 (Tree Star). Expression of markers was gated using fluorescence minus one controls, as shown in Supplemental Fig. 3D and 3E.

PBMCs were thawed and rested overnight at 37°C with 5% CO₂ in RPMI 1640 with GlutaMAX supplemented with 10% FCS, 1% penicillin–streptomycin (Life Technologies), 1% sodium pyruvate, and 1% minimum nonessential amino acids (unless otherwise specified). CFSE (Invitrogen)–loaded PBMC (2 × 10⁵ cells per well in 96-well plates) were incubated with anti–CD2/CD3/CD28–coated beads (T Cell Activation/Expansion Kit; Miltenyi Biotec) in a 1:5 bead/cell ratio or TAC for 96 h. For intracellular cytokine staining, PBMC from 96-h incubations were stimulated for 4 h with PMA (50 ng/ml) and ionomycin (1 mg/ml) and brefeldin A (10 mg/ml). Dead cells were excluded with Fixable Viability Stain 700 (BD Biosciences). All multicolor flow cytometry data were acquired with a BD LSRFortessa and analyzed using FlowJo version 10 (Tree Star).

The frequencies of INF-γ– and IL-10–secreting CD4⁺ T cells, CD8⁺ T cells, and regulatory T cells (Tregs) following PMA (20 ng/ml) and ionomycin (500 ng/ml) stimulation were measured using Dual-Color ELISPOT assay (Cellular Technology Europe Limited). Briefly, PBMC were thawed and rested overnight at 37°C with 5% CO₂ in complete RPMI 1640 supplemented with 10% human Ab serum (Sigma-Aldrich) followed by staining with anti-CD3, CD4, CD8, CD127, CD25, and FACSorting. Next, the sorted cells were incubated in precoated ELISPOT 96-well plates as duplicates (1000 cells per well) and stimulated (PMA/ionomycin) overnight in complete RPMI 1640 supplemented with 10% human Ab serum. On the next day, cells were washed, and spots were developed according to the manufacturer’s instructions. Spots were analyzed as spot forming units using CTL ImmunoSpot S6 Analyzer (Cellular Technology Europe Limited). The working concentration of cells per well and PMA/ionomycin were titrated using frozen PBMCs (data not shown).

Data were analyzed using GraphPad Prism version 8. The results were compared using nonparametric (Mann–Whitney or Wilcoxon signed-rank) two-tailed tests. Correlations between parameters were calculated using the Spearman correlation and linear regression analysis. The p values < 0.05 were considered statistically significant.

The clinical characteristics of the PTx recipients are shown in Table I (n = 22). The median recipient age was 41 y, and the median donor age was 33 y. All the patients were diagnosed with T1D. Seventeen patients underwent PTA transplantation. Five patients had end-stage kidney failure due to diabetic nephropathy and received SPK transplantation. Rejection risk was not different between these two transplantation types. Most of the rejections after 1 y post-PTx were classified as cell mediated (90%), and the rest were characterized as Ab mediated. Immunological risk factors, such as CMV serostatus and HLA mismatches, and serum markers, such as amylase, lipase, creatinine, glucose, HbA1c, C-peptide ratio (creatinine/glucose/C-peptide) were not significantly different between the rejector and nonrejector groups.

(Fig. 1 shows the frequencies of CD3⁺ T cells and the CD4⁺, CD8⁺, and Treg subsets in the peripheral blood of healthy controls and PTx recipients at baseline (prior to PTx) and at 1 y post-PTx. The median frequency of CD3⁺T cells was significantly lower in the PTx recipients at both baseline and 1 y post-PTx recipients compared with healthy controls. However, no difference was noted in the PTx recipients between the baseline and 1 y post-PTx (Fig. 1A). The median frequency of CD4⁺ T cells was significantly reduced, and the median frequency of CD8⁺ T cells was significantly increased after 1 y compared with the baseline of PTx recipients and healthy controls (Fig. 1B). This resulted in significantly lower CD4/CD8 ratio after 1 y compared with the baseline of the PTx recipients and the healthy controls (Fig. 1D). Next, we analyzed the frequency of Tregs based on the phenotype of CD4⁺CD25⁺CD127⁻FOXP3⁺ T cells (20). The median frequency of Tregs was significantly increased 1 y post-PTx compared with baseline (Fig. 1C). This led to a significantly increased Treg/CD3⁺ T cell ratio and Treg/CD4⁺ T cell ratio compared with the baseline and compared with healthy controls (Fig. 1E, 1F). The observed changes in the frequency of CD3⁺, CD4⁺ and CD8⁺ T cells and the low CD4/CD8 ratio were not different between rejectors and nonrejectors (Supplemental Fig. 2A, 2B). We also analyzed the frequency of other circulating lymphocyte subsets in PTx recipients. The gating strategy is presented in Supplemental Fig. 1A and 1B. We found that the median frequency of B cells was significantly higher in rejectors compared with nonrejectors 1 y post-PTx (Supplemental Fig. 2C). In contrast, no difference was found in γδ T cells, NK T cells, NK cells, myeloid dendritic cells, plasmacytoid dendritic cells, and innate-lymphoid cells (Supplemental Fig. 2C). In summary, a lower CD4/CD8 ratio and a higher Treg/CD3⁺ T cell ratio and Treg/CD4⁺ T cell ratio was observed in the circulating lymphocyte compartment in PTx recipients 1 y post-PTx.

We analyzed the activation status of CD4⁺ and CD8⁺ T cells based on the expression of the proliferation marker Ki-67 in CD4⁺ and CD8⁺ T cells between rejectors and nonrejectors at baseline and after 1 y post-PTx and compared them with healthy controls (Fig. 2). The median frequencies of Ki-67–expressing CD4⁺ T cells and CD8⁺ T cells were significantly increased both at baseline and 1 y post-PTx in rejectors compared with nonrejectors (Fig. 2A, 2B). Furthermore, we analyzed the frequency of activated CD4⁺ T cells and CD8⁺ T cells in the gut-homing T cell compartment as defined by the phenotype of α4β7⁺⁺ CCR9⁺ Ki-67⁺ T cells (Fig. 2C) (21). This extended the above findings to show that the activated gut-homing CD4⁺ T cells and CD8⁺ T cells were significantly increased in baseline and 1 y post-PTx in rejectors compared with nonrejectors (Fig. 2D). Moreover, we assessed the CD4⁺ and CD8⁺ T cell differentiation states based on the differential expression of CCR7 and CD45RA to phenotypically define the naive (N), central memory (CM), effector memory (EM), and end-stage effector (ES) cells. We did not notice any difference in the memory CD4⁺ and CD8⁺ T cell distributions between the groups (Table II). We also analyzed the expression of the activation-induced coinhibitory receptors CTLA-4 and PD-1 (22). The PD-1 expression was dominant in CD4⁺ and CD8⁺ T cells from both rejectors and nonrejectors. However, the expression of CTLA-4 was found to be significantly increased in CD4⁺ and CD8⁺ T cells in rejectors compared with nonrejectors (Table II). Interactions between CD4⁺ T cells, CD8⁺ T cells, and B cells are essential for their effector functions (23). In line with this, the activated Ki-67–expressing CD4⁺ and CD8⁺ T cells correlated with B cell frequencies in PTx recipients, both at baseline and 1 y post-PTx (Supplemental Fig. 3A). Together, the frequencies of activated CD4⁺ and CD8⁺ T cells were elevated in PTx recipients both prior to and during rejection, highlighting that rejectors may also have an activated immune system before transplantation.

Tregs have been shown to reduce the risk of rejection (24, 25). This is consistent with our finding that the median frequency of Tregs was significantly increased post-PTx in nonrejectors. However, we also observed a similar trend in rejectors (Fig. 3A). No significant difference was found in Treg frequencies between rejectors and nonrejectors 1 y post-PTx indicating that the effect was transient in the months following transplantation. Intriguingly, the median frequency of Ki-67–expressing Tregs was significantly higher in rejectors compared with nonrejectors 1 y post-PTx (Fig. 3B). Assessment of Treg differentiation status by expression of CCR7 and CD45RA (Fig. 3C) showed a significant decrease in N Tregs and subsequent increase in Treg EM cells post-PTx in rejectors (Fig. 3D). Treg CM cells were not significantly different between rejectors and nonrejectors. Treg ES cells were neither detectable in healthy controls nor PTx recipients (Fig. 3D). To see how Treg differentiation and activation aligned, we examined Ki-67 expression in N/CM/EM Tregs (Supplemental Fig. 3B). In rejectors, the percentage of Ki-67⁺ N/CM/EM Tregs was similar at baseline and 1 y post-PTx; however, it was significantly higher than nonrejectors at both time points. However, Treg activation and memory status did not correlate in a significant manner (Supplemental Fig. 3C). Next, we analyzed the expression of the Treg differentiation and the functional markers CTLA-4 and PD-1 (26, 27) (Fig. 3E). Comparing the median frequency of CTLA-4 and PD-1 in the Tregs showed a significantly increased expression in rejectors and nonrejectors compared with their respective baselines (Fig. 3F). Particularly, CTLA-4 expression was significantly increased both at baseline and 1 y post-PTx in rejectors compared with nonrejectors. No such difference was observed for PD-1 expression between the groups. Treg secretion of IFN-γ and IL-10 was also examined by ELISPOT based on sample availability (n = 4 in each group). In rejectors, significantly more Tregs secreted IFN-γ compared with their baseline, whereas Tregs from nonrejectors tended to secrete more IL-10 compared with rejectors post-PTx (Fig. 3G). In summary, rejectors show an activated Treg phenotype post-PTx; however, these Tregs were unable to secrete the suppressive cytokine IL-10 in vitro.

Intracellular staining for granzyme B (GrzB) and perforin in ex vivo–isolated PBMC showed a significantly increased frequency of GrzB and perforin expression in CD4⁺ T cells in rejectors post-PTx compared with their baseline. In contrast, no significant difference was observed in nonrejectors. (Fig. 4A, 4B). Increased median frequency of GrzB and perforin expression was found in CD8⁺ T cells in both in rejectors and nonrejectors post-PTx compared with their respective baselines (Fig. 4A, 4C). Also, significantly increased expression of perforin was found in rejectors (Fig. 4A, 4C). Stimulation of ex vivo–isolated PBMC with PMA/ionomycin showed a significantly increased production of INF-γ by CD4⁺ T cells post-PTx in rejectors compared with their baseline (Fig. 4D, 4E), whereas no significant difference was observed in nonrejectors. No significant differences were observed between the groups in the CD8⁺ T cells (Fig. 4D, 4E). IFN-γ and IL-10 secretion by ELISPOT corroborated the intracellular cytokine data (Supplemental Fig. 4A). Both CD4⁺ and CD8⁺ T cells showed a tendency toward increased secretion of IFN-γ post-PTx in rejectors. Although very few cells secreted IL-10, corresponding CD4⁺ and CD8⁺ T cells showed decreased secretion of IL-10 post-PTx in rejectors. We found that less than 2% of CD4⁺ and CD8⁺ T cells produced IL-4 and IL-17A (data not shown). Next, we analyzed the expression of the T-bet transcription factor that induces the production of INF-γ (23). The frequencies of T-bet–expressing CD4⁺ and CD8⁺ T cells were significantly increased in both rejectors and nonrejectors compared with their respective baselines (Fig. 4F, 4G). However, the expression of T-bet was more dominant in CD8⁺ T cells compared with CD4⁺ T cells and was significantly increased in rejectors compared with nonrejectors in CD8⁺ T cells. In conclusion, the expression of cytotoxic effector molecules and INF-γ secretion from CD4⁺ and CD8⁺ T cells were increased in PTx recipients during rejection.

We analyzed the proliferation potential of the CD4⁺ and CD8⁺ T cells and their cytokine profile in vitro upon anti–CD2/CD3/CD28 polyclonal stimulation (Fig. 5). Comparing the median frequency of proliferating CD4⁺ and CD8⁺ T cells revealed no difference between baseline and 1 y post-PTx in nonrejectors. However, a trend toward hypoproliferation was observed in the rejectors (Fig. 5A, 5B). The median frequency of INF-γ–producing CD4⁺ T cells (Figure 5C, left panel) and CD8⁺ T cells (Fig. 5D, left panel) was not altered in baseline versus 1 y post-PTx except for a trend toward an increase in CD8⁺ T cells between rejectors and nonrejectors. IL-2– and TNF-α–producing CD4⁺ T cells (Fig. 5C, middle and right panel) and CD8⁺ T cells (Fig. 5D, middle and right panel) were significantly reduced in rejectors compared with nonrejectors, whereas no difference was observed between the baseline and 1 y post-PTx groups. To mimic the in vivo function of PTx recipients’ CD4⁺ and CD8⁺ T cells, similar (as above) in vitro experiments were set up using healthy donor PBMCs and TAC, which is one of the drugs used post-PTx. A concentration-response curve for the proliferation of CD3⁺ T cells was established (Supplemental Fig. 4B, 4C). At a TAC concentration of 3.7 nM, the proliferation of CD3 T cells was reduced to 50% (IC50). Hence, for the subsequent experiments, 5 nM TAC was used to examine CD4⁺ and CD8⁺ T cell functionality. Treatment with TAC significantly reduced the proliferation of both CD4⁺ and CD8⁺ T cells (Fig. 5E). In addition, the median frequency of INF-γ–, IL-2–, and TNF-α–producing CD4⁺ T cells (Fig. 5F, top panel) and CD8+ T cells (Fig. 5F, bottom panel) was significantly reduced after treatment with TAC. In conclusion, polyclonal stimulation of CD4⁺ and CD8⁺ T cells from PTx rejectors revealed that the T cells appeared to have a trend toward reduced proliferation capacity and significantly reduced IL-2 and TNF-α secretion, potentially indicating a current hyporesponsive T cell phenotype associated with postactivation or post-PTx treatment.

In this study, we performed comprehensive analyses of the immune phenotype of circulating CD3⁺ T cell subsets from whole organ PTx recipients and correlated the findings with the histologically confirmed rejection status. Overall, the frequency of CD3⁺ T cells was significantly reduced 1 y post-PTx, and this observation can most likely be ascribed to T cell depletion by the induction therapy using anti-thymocyte globulin (ATG) and continuous immune suppression. ATG is widely used in solid organ transplantation for the prevention and treatment of acute rejections and by inducing long-lasting lymphocytopenia (28, 29). However, repopulation of CD8 and CD4 T cell subsets has been shown to occur 1 y after treatment (30). In our study, CD4⁺ T cells remained strongly depleted until 1 y post-PTx, whereas CD8⁺ T cells were fully replenished and thus resulted in a lower CD4/CD8 ratio. Studies have shown that CD4⁺ and CD8⁺ T cells play a key role in allograft rejection (31, 32). Also, pancreas sections from T1D patients, in contrast to T2D patients, have shown a strong infiltration of CD4⁺ and CD8⁺ T cells, corroborating a key role for CD4⁺ and CD8⁺ T cells in β cell destruction (10, 13, 14, 33). Thus, in PTx recipients, the lower CD4/CD8 ratio suggests a possible risk of cytotoxic responses and allograft rejection.

Allograft rejection by alloimmunity and recurrence of autoimmunity are the major concerns in T1D patients (156). A significant finding in this study was that the fractions of proliferating Ki-67⁺ CD4⁺, and CD8⁺ T cells were strongly elevated in rejectors, both prior to and 1 y post-PTx. This indicates that the presence of activated CD4⁺ and CD8⁺ T cells prior to PTx in T1D patients may predispose the patients to rejection. Activated CD4⁺ T cells stimulate the function of CD8⁺ T cells and provide help to B cells stimulating Ab production and Ag presentation (23). We demonstrated that activated Ki-67–expressing CD4⁺ and CD8⁺ T cells positively correlated with the frequency of activated B cells, indicating possible cooperation between these cell types in PTx recipients. In pancreatic islet cell transplant recipients, it has previously been shown that a high frequency of autoreactive T cells toward the islet Ags and high B cell counts were associated with lower graft function (9). Most of the PTx rejections analyzed in this study were subclinical, and the patients were treated with methylprednisolone and ATG in combination or alone before the development of clinical rejection with loss of β cell function and hyperglycemia (19). Whether the activated T cells directly contributed to concurrent autoimmunity and alloimmunity, however, still needs to be addressed.

Tregs represent 5–10% of the total CD4⁺ T cell population and play a pivotal role in transplantation tolerance (25, 34, 35). Several studies have shown an association between high number and proportions of Tregs and graft survival (36–38). ATG has also been shown to induce Tregs in vivo in renal transplant recipients (28). We found an elevated frequency of Tregs in nonrejectors compared with rejectors. These results are in line with findings after kidney (39), liver (40), and heart (41) transplantation, in which elevated Treg frequencies have been shown to be associated with reduced risk for rejection reactions. Several studies have shown that Tregs mirror the ongoing immune response (35, 42), and similarly, we found that activated Ki-67–expressing Tregs were specifically elevated in rejectors. A transient increase of activated Tregs early has also been reported after kidney transplantation (43); however, it does not appear to be associated with acute rejection (39).

Tregs from rejectors appeared to differentiate into and EM-type cells (CCR7⁻CD45RA⁻). Furthermore, proliferating Ki-67⁺ EM Tregs were significantly increased in rejectors (baseline) compared with nonrejectors. To further substantiate these results, we analyzed the expression of CTLA-4 and PD-1. We found, particularly, that the expression of CTLA-4 was elevated in Tregs from rejectors both prior to and 1 y post-PTx. The possible explanation for elevated numbers of activated Tregs in rejectors could be that Tregs are triggered to restrict the activated CD4⁺ and CD8⁺ T cells and the ongoing rejection-associated inflammation. Different reports have shown that the suppressive ability of Tregs is diminished in T1D patients and contributes to disease progression (44–48). In the current study, we found by ELISPOT assays that Tregs in rejectors produced significantly higher IFN-γ compared with baseline, whereas production of IL-10 tended to be higher in nonrejectors.

Our study suggests that CD4⁺ T cells are mostly CM type, whereas CD8⁺ T cells are predominantly differentiated into ES memory type in both rejectors and nonrejectors. It is well documented that ES-CD8 T cells are associated with other solid organ transplant rejections, such as kidney (49, 50), and characterized by secreting cytotoxic granules (51). The activated memory T cells produced cytokines (e.g., INF-γ) and cytolytic molecules (e.g., GrzB and perforin) to mediate effector functions (23). Previous studies have shown that type 1 inflammatory and cytotoxic components from T cells are strongly associated with T1D progression (13, 14) and graft rejection, including islet allografts (24, 52–56). In the context of PTx, we found similar evidence showing that the frequencies of INF-γ–, GrzB–, and perforin-expressing CD4⁺ and CD8⁺ T cells were elevated only in rejectors following ex vivo isolation and analysis by intracellular flow cytometry. These results were also substantiated by functional ELISPOT assays, in which both CD4⁺ and CD8⁺ T cells were shown to secrete elevated levels of IFN-γ. We further confirmed the type 1 response related to transcription factor T-bet (23) expression was strongly upregulated, particularly in CD8⁺ T cells from the PTx rejectors.

We found that the CD4⁺ and CD8⁺ T cells from rejectors appeared to have reduced proliferation capacity and produced significantly lower IL-2 and TNF-α upon polyclonal stimulation in vitro. This suggests CD4⁺ and CD8⁺ T cells from rejectors may have a hyporesponsive phenotype, which could be due to the effect of steroid and induction therapy used to restrict inflammation in PTx recipients. In kidney transplant recipients, it has been shown that T cell exhaustion after lymphocyte-depleting induction therapy correlates with improved allograft function (57). The effect of post-PTx treatment was further substantiated by in vitro treatment of healthy CD4⁺ and CD8⁺ T cells with TAC, which at 5 nM abrogated both proliferation and cytokine secretion capabilities (IFN-γ, IL-2, and TNF-α). However, it is intriguing that the hyporesponsive CD4⁺ and CD8⁺ T cells from rejectors retained the production of INF-γ. Human anergic T cells that retain INF-γ production have been reported earlier (58), and paradoxical type 1 inflammatory responses were shown to both promote (59, 60) and hinder (24, 52, 55) long-term allograft survival. Further investigations in the graft for the expression of type 1 inflammatory CD4⁺ and CD8⁺ T cells are required to confirm their pathological role in PTx recipients. However, a recent study hints that the stable PTx recipients with low-grade rejection episodes at 1 y post-PTx showed infiltration of CD8⁺ T cells (61).

In conclusion, our data provide a detailed phenotypic characterization of circulating Tregs, CD4⁺, and CD8⁺ T cells from stable and rejecting PTx recipients. Activated and differentiated CD4⁺ and CD8⁺ T cells with the capability of producing type 1 inflammatory and cytotoxic responses appear to be the key factors associated with rejection in PTx recipients, and our findings suggest that the presence of activated CD4⁺ and CD8⁺ T cells prior to transplantation is strongly associated with PTx rejection 1 y post-PTx. These findings may be used to identify patients with a higher risk for rejection and patients that may benefit from being more closely monitored posttransplant.

The authors have no financial conflicts of interest.