Objectives: A growing body of evidence has revealed the role of innate immune cells in transplantation; however, the nature of natural killer cell involvement in rejection is still elusive. Here, we aimed to determine the impact of natural killer cell activities in acute and chronic renal transplant rejection.
Materials and Methods: This preliminary case-control study included 63 participants: 19 were patients with kidney allograft rejection (8 patients with acute rejection and 11 patients with chronic rejection) and 44 comprised the control group (22 patients who had well-functioning grafts posttransplant and 22 healthy subjects). In addition to natural killer cell frequency, we also measured intracellular interferon-γ production and surface expression of CD107a as cytotoxic activity using flow cytometry.
Results: We observed a significant increase in CD107a expression (P = .021) in patients with acute rejection versus those with well-functioning grafts. Moreover, production of interferon-γ in patients with chronic rejection was significantly increased compared with patients with well-functioning grafts (P = .003). Finally, natural killer cell frequency was decreased in patients with rejection versus control groups; however, this reduction was not statistically significant.
Conclusions: These findings suggest that the increase in natural killer cell cytotoxicity is correlated with rejection in kidney transplant recipients and might be considered as a predictive marker in prevalence of graft rejection.
Key words : CD107a, Cytotoxicity, Interferon-gamma, Kidney transplant rejection, NK cells
The criterion standard treatment for patients with end-stage renal disease is kidney allograft transplant.1,2 However, allograft rejection hinders this process. The dynamics of organ allograft outcomes is defined by donor and recipient incompatibility, recipient immune response, the state of engrafted tissue, and transplant location.3
It is now recognized that, apart from adaptive immunity, innate immune cells are also implicated in the establishment of transplant tolerance.4 Natural killer (NK) cells have currently been acknowledged as essential participants that influence the status of the transplanted solid organ and hematopoietic stem cells.5,6 Natural killer cells are large, granular cytotoxic lymphocytes that play a central role in the innate immune response.7 These cells are actively involved in early immune protection against virus-infected cells and transformed or hassled cells, including immune-related pathologic diseases without the need for prior sensitization.8-11 Nonetheless, their involvement in allograft rejection is not well understood.12
Natural killer cells, in contrast to T and B lymphocytes, cannot recognize the major histo-compatibility complex (MHC)-antigen complex due to lack of clonal rearrangement of their receptors. On the other hand, NK cells express activating and inhibiting cell surface receptors such as killer cell immunoglobulin-like receptors (KIR; inhibitory for self-HLA), NKG2A, and LILRB1/ILT-2 encoded from germ-line control cell stimulation.5,13 High expression of NKG2D in posttransplant acute and chronic kidney allograft nephropathy, including its possible use as a transplant rejection marker, has been reported.12,14 The equilibrium between associated ligands and expressed activating and inhibitory KIRs synchronize NK cell stimulation. For augmenting T-cytotoxic lymphocyte stimulation and allogeneic responses, NK cells react to HLA ligand incompatibility by either indirect secretion of cytokines or by direct contact via allograft receptor ligands.15,16
In our previous study, we showed that the KIR3DL1+HLA-Bw4*A allele combination has a protective role in patients with acute kidney allograft rejection compared with that shown in recipients with well-functioning grafts (WFG) (P = .004; odds ratio [OR] =0.34; 95% confidence interval [95% CI], 0.16-0.72) and healthy subjects (P = .019; OR = 0.47; 95% CI, 0.25-0.89).6 Furthermore, we analyzed the rate of NK cell subsets and interleukin 2 (IL-2), IL-15, and IL-18 gene expression levels in participants with WFG and in recipients with chronic allograft disease. We found that allorecipients with chronic allograft disease exhibited higher rates of NKCD56bright subsets.17
Natural killer cells link both innate and adaptive immune systems through a reversible interaction with antigen-presenting cells (APCs).8 In addition, NK cells have been reported to intensify T-cell-mediated transplant rejection alloreactivity in pre-clinical models.18 The adaptive immune responses and hematopoiesis are modified by NK cells through their use of chemokines (MIP-1α, MIP-1β, IL-8, and RANTES) and cytokines (interferon-gamma [IFN-γ] and granulocyte-macrophage colony-stimulating factor), which are innate immune effectors that directly disrupt target cells. Through licensing (NK cell education), these cells procure the ability to produce cytokines (IFN-γ) and improve their cytolytic activity via granule exocytosis (granzyme and perforin) during development.8,18,19 The cytotoxic function of NK cells may be through the identification of CD107a expressed on their surface after granule exocytosis.15,20 However, it is unknown whether cytokine production by NK cells or direct target cell disruption is the utmost essential physiologic process used to regulate primary immune reactions to posttransplant allografts.8
In this study, we investigated whether in vitro NK cell cytotoxic activity, as shown by IFN-γ secretion and the expression of CD107a, can be a functional marker for the identification of NK cell degranulation with regard to occurrence of acute and chronic renal transplant rejection versus that shown in control samples. We also evaluated the frequency of NK cells in peripheral blood mononuclear cells (PBMCs) in these patients. We hypothesized that the cytotoxic activity and amount of NK cells may fluctuate in peripheral blood of recipients with kidney transplant rejection compared with that shown in recipients with WFGs.
Materials and Methods
In the present study, PBMCs were obtained from whole blood of 8 patients with acute kidney allograft rejection (AR group), 11 patients with chronic kidney allograft rejection (CR group), and 44 control samples. Control individuals were divided into 2 groups, each consisting of 22 individuals: (1) recipients with WFGs and (2) unrelated healthy volunteers.
Available clinical data of recipients, including age, sex, and donor relationship, were collected from Shahid Labbafinejad Hospital (Tehran, Iran) records between September 2014 and July 2015 (Table 1). Patient immunosuppression included cyclosporine (or tacrolimus), prednisolone, and mycophenolate mofetil. The WFG group consisted of patients who had no clinical manifestations of acute or chronic rejection for at least 18 months after transplant. Acute and chronic rejection episodes were defined by clinical manifestations, including increased serum creatinine levels and low glomerular filtration rates, and were confirmed by biopsy based on Banff criteria.21 A documented consent form was signed by all participants, and the university ethics committee ratified the study.
We obtained MHC class I-deficient K562 cells, a homogenous human erythroleukemic cell line, from the Iranian Biological Resource Center (Tehran, Iran). Cells were maintained in RPMI-1640 medium (Gibco Laboratories, Grand Island, NY, USA), which was composed of 10% fetal bovine serum (Gibco Laboratories), 2 mM L-glutamine, and 50 IU/mL penicillin, at ×106 cells/mL. Cells were grown and maintained in density ranging from 1 × 105 to 1 × 106 viable cells/mL by addition or replacement of fresh medium at 37°C in a 5% CO2 humidified atmosphere.
Assessment of frequency, cytotoxic activity, and intracellular cytokine
secretion in natural killer cells
Frequency, cytotoxic activity, and intracellular cytokine secretion in NK cells were quantitated by flow cytometry. We isolated PBMCs from collected blood samples and suspended them at 106 cells/mL in RPMI-1640 medium, which contained 10% FBS (Gibco Laboratories), 2 mM L-glutamine, and 50 IU/mL penicillin. Frequency of NK cells was evaluated in PBMCs. The monoclonal antibodies anti-CD3-FITC (clone SK7) and anti-CD56-APC (clone NCAM 16.2) were used to phenotype NK cells (CD56+CD3−) and detected on a FACSCalibur flow cytometer (Becton-Dickinson Immunocytometry Systems, San Jose, CA, USA) already calibrated with appropriate fluorochrome-labeled beads.
Assay development for assessment of in vitro viability and cytotoxicity of NK cells involved the use of PBMCs to set up an initial subtle granule exocytosis and cytokine release test. Through cytotoxic granule exocytosis, NK cells lyse target cells, thereby exposing the internal lysosomal membrane protein (CD107a), which, through flow cytometric analysis, can be identified on their surface. Concurrent production of IFN-γ can be estimated by flow cytometry through the parallel measurement of released cytokines by NK cells. Functional analysis was performed by excitation of PBMCs using K562-deficient MHC cells, at a 10:1 effector-to-target ratio. The total PBMCs included in the assay constituted the “effector NK cells”; however, the NK cells interact with the K562 MHC class I-deficient cells. Moreover, cells were stimulated with phorbol-12-myristate-13-acetate (5 μg/mL; Sigma, St. Louis, MO, USA) and ionomycin (1 μg/mL; Sigma).
CD107a-APC antibody (BioLegend, San Diego, CA, USA) was added directly to the tubes at 2 μL/mL. Cells were incubated for 6 hours at 37ºC in 5% CO2 after which brefeldin A (BioLegend) and monensin (Golgi-Stop, BioLegend) were also added, each at 6 μL/well. Although brefeldin A could prevent vesicles that released cytokines from endocytosis after activation, monensin inhibited endocytic vesicles from acidifying, thus evading degeneration of reinternalized membrane-bound CD107a proteins, to produce imagery of the released cytokines and this biomarker. The cells were garnered, washed, and tainted with antihuman CD3 and CD56 monoclonal antibodies after incubation for flow cytometric analysis. For intracellular IFN-γ detection, cells were permeabilized and stained for IFN-γ APC antibody (BioLegend). The cells were then suspended in paraformaldehyde (Sigma)-PBS solution (1:60) until a 3-color flow cytometric analysis was performed using a FACSCalibur (BD Biosciences, San Jose, CA, USA). With this technique, we acquired 30 000 events, which were then analyzed using Flow-Jo software (The Scripps Institute, Flow Cytometry Core Facility, San Diego, CA, USA). The lymphocyte cell population was analyzed via gating on the CD3-/CD56+ population. Within this population of cells, we noted independent expression of CD107a and IFN-γ for each sample after stimulation.
Data are presented as median and interquartile range (IQR) for numerical variables and as frequencies or percentages for categorical variables. Kruskal-Wallis test was performed to compare mean changes of variables between groups. We used the nonparametric Mann-Whitney U test for pairwise comparison of differences between 2 independent groups. Receiver operating characteristic curve analysis was done to find cutoff points with highest sensitivity and specificity for diagnosis of graft dysfunction.
Table 1 presents demographics of participants and their clinical parameters. Histologic findings showed 5 patients with antibody-mediated rejection and 3 patients with cell-mediated rejection in the AR group and 6 patients classified as Banff grade III type 5, 1 patient as Banff grade II type 5, and 4 patients as Banff grade I type 5 in the CR group.
In terms of donor sources, we found no detectable differences between recipient groups. With respect to transplant time, no significant difference was shown between the AR and the CR patient groups. For this study, HLA matching was not considered, and all recipients received HLA-mismatched kidney transplants from living unrelated donors, deceased donors, and living related donors.
Assessment of frequency of natural killer cells
We assessed the frequency of NK cells in PBMCs using flow cytometry. A Kruskal-Wallis analysis indicated significant differences between groups (P < .001; Table 2). Our Mann-Whitney U test pairwise comparison (Figure 1, Table 2) also revealed a significant decrease in NK cell frequency in PBMCs, with AR patients showing median of 3.45 (IQR, 0.82-6.50) versus WFG patients showing median of 10.00 (IQR, 5.92-15.80) (P = .007) and healthy control patients showing median of 8.35 (IQR, 6.60-10.50) (P = .049) and CR patients showing median of 3.20 (IQR, 1.90-5.10) versus results shown in WFG patients (P = .006) and healthy control patients (P = .047). However, the frequency of NK cells in the AR group versus CR group, as well as versus the WFG and healthy control groups, was not statistically significant (P = 1.000 and P = 1.000, respectively).
Measurement of CD107a expression in natural killer cells and secretion of
intracellular interferon-gamma cytokine in peripheral blood mononuclear cells
Natural killer cell degranulation was determined by measuring the increase in surface expression of CD107a as a biomarker for NK cell cytolytic function and the secretion of intracellular IFN-γ using flow cytometry after stimulation with the target cell. The target cell populations were then gated to determine the percentage of CD3-/CD56+ cells in the PBMC samples (Figure 2).
We found a significant difference between groups in CD107a expression on the surface of NK cells (P = .017; Table 2). As shown in Figure 3 and Table 2, patients with AR (median = 12.60; IQR, 4.92-23.12) exhibited a significant increase in CD107a expression on NK cells versus patients in the WFG group (median = 3.00; IQR, 2.00-5.08; P = .021). There was a marginal difference between the AR and the healthy control group (P = .055). We observed no significant differences in CD107a expression on the surface of NK cells in the PBMCs in comparisons of the following groups: CR versus WFG, CR versus healthy control, AR versus CR, and healthy control versus WFG (P > .6).
Regarding intracellular IFN-γ secretion (Table 2), significant differences were observed between groups (P = .003) and also between the CR group (median = 18.70; IQR, 7.80-25.60) and the WFG group (median = 2.45; IQR, 1.57-5.32) (P = .003), with trend toward significance between the WFG group and healthy controls (P = .062). Pairwise comparisons of cytokine secretion showed no significant differences in AR versus WFG, AR versus healthy control, AR versus CR, and CR versus healthy control groups (P > 0.5) (Figure 3, Table 2).
Receiver operating characteristic curve analysis was performed to determine correlations between diagnosis of acute rejection (glomerular filtration rate < 60 mL/min/1.73 m2) and expression of CD107a, secretion of intracellular IFN-γ, and NK cell frequency (Table 3). The calculated area under the curve was 0.84 (range, 0.68-0.99) for CD107a expression (95% CI, P = .017); 0.71(0.50-0.92) for the secretion of intracellular IFN-γ (95% CI, P = .003); and 0.84 (0.70-0.98) for NK cell frequency (95% CI, P < .001). At a cutoff point of 6.8, CD107a exhibited 75% sensitivity and 87% specificity, whereas IFN-γ with a 4.5 cutoff point displayed 75% sensitivity and 73% specificity. On the other hand, NK cells with a 5.8 cutoff point displayed 75% sensitivity and 78% specificity (Table 3). Additional results on other defined classifications based on rejection type are presented in Table 3.
Acute rejection has been shown to be one of the strongest negative prognostic factors for long-term graft survival. Allograft transplantation is mainly impeded by rejection. Long-lasting, stable graft function is mostly inhibited by acute rejection, in association with chronic allograft rejection following organ transplant.22 Besides its prominent role in orchestrating innate immune reactions, NK cells connect both the adaptive and innate immune systems.20 Emergent reports affirm that fluctuations of NK cell frequency and cytotoxicity influence the pathogenesis of allograft rejection, including in kidney transplant.
In this study, we used a cytometric assay to assess the frequency and function of NK cells, including cytotoxicity and cytokine secretion, in defined categories. Studies have demonstrated that NK cell effector function assessment relies on the ability to detect intracellular IFN-γ production and expression of CD107a on the surface of NK cells, which correlates closely with NK cell-induced cytotoxic lysis of target cells.15,20
In our study, there was a significant difference in frequency of NK cells in PBMCs of our AR group versus the WFG group. We also demonstrated this difference in patients with acute rejection versus healthy controls, patients with chronic rejection versus those with WFGs, and patients with chronic rejection versus health controls. However, there was no significant difference between other groups.
When we assessed NK cell effector function based on intracellular IFN-γ secretion and expression of CD107a on the surface of NK cells, our data revealed that patients with acute rejection exhibited a significant increase in CD107a expression on NK cells versus that shown in the WFG group, although no significant differences were shown between other groups. We also observed significantly increased levels of intracellular IFN-γ in patients with chronic rejection versus those with WFGs. Recent reports actually hold that cell-mediated allograft rejection, including acute rejection, is promulgated by primary stimulation of NK cells after transplant.23,24 However, Game and associates reported a reduction in NK cell frequency and function in PBMCs from patients after primary kidney transplant. The group suggested that early posttransplant time is when these recipients receive the highest amount of immunosuppressant drugs; therefore, the subsequent reduced frequency and function of NK cells can contribute to increased opportunities for viral infection.25 Our study confirmed their findings in the WFG group, not only in the initial posttransplant period but also over time, as we observed a decline in frequency and function of NK cells in this group compared with healthy individuals.
A recent clinical study on healthy individuals who received 3 immunosuppressive regimens revealed decreased cytotoxicity of NK cells with expression of CD107a and intracellular IFN-γ secretion, in response to different stimuli, including K562 cells and phorbol-12-myristate-13-acetate plus ionomycin. Lung trans-plant recipients exhibited an intense decrease in CD107a expression and intracellular IFN-γ secretion compared with healthy individuals.15
Morteau and associates demonstrated the inability of antibodies and macrophages to destroy trans-planted allografts without intervention of NK cells.12 In an animal study, NK cell cytolytic effect was increased in allograft tissue from rats after heart transplant. According to the study, allograft tissue NK cells were activated and cytotoxic.26 Although we thought that NK cell frequency would be decreased in PBMC samples from patients after kidney transplant rejection, their activities were increased.
In another recent study, recipients of kidney allografts had high IFN-γ production in reaction to nonpositive HLA class I target cells. However, the secretion of other cytokines, like IL-13, IL-17, IL-22, and IL-31, was significantly reduced in transplant recipients compared with that shown in healthy donors. Natural killer cells may be used as markers for determining the suppression of the immune system and may prompt a customized change of immunosuppressive drugs in kidney transplant recipients.27
Recent observations raise arguments in support of NK cell frequency alterations in PBMCs as being equally important for allograft rejection pathogenesis. Present knowledge of transplant immunology im-plicates primary stimulated NK cell and allograft tissue penetration as responsible for these occurrences. Natural killer cells that penetrate the allograft have been shown to be liable for producing chemokines, which are major precipitators of acute and chronic allograft rejection.23,28
Stimulated NK cells after allorecognition become cytotoxic and can either directly and/or indirectly cause solid-organ rejection by increasing Th1-adaptive alloimmune response through the discharge of proinflammatory cytokines such as tumor necrosis factor alpha and IFN-γ. These cytokines prompt and escalate costimulatory and MHC molecular expressions on professional (dendritic cells and B cells) and nonprofessional (endothelial cells) APCs, including their development and efficiency, thus enabling NK cells to activate T-cell alloreactions both directly and indirectly.23,26 These results have been confirmed in preclinical studies, which implicated NK cells as aggravators of T-cell-mediated alloreaction-dependent organ rejection.26 Although NK cells may provide auxiliary support for Th1-induced chronic rejection, the mechanisms by which NK cells stimulate CD4-positive T cells remain elusive. The mechanisms may include associated bonds between NK and T cells or bystander T-cell stimulation through soluble mediators or cytokines.23
In conclusion, this preliminary study showed that the frequency of NK cells in patients with acute and chronic rejection was significantly decreased compared with that shown in patients with WFGs and healthy individuals. Nevertheless, NK cell degranulation with subsequent expression of CD107a and secretion of intracellular IFN-γ were increased. Together, circulating levels of NK cells in peripheral blood of patients with organ rejection decreased, whereas the cytotoxicity levels of these cells increased. In patients with allograft rejection, circulating levels of NK cells may be causative agents of rejection, during homing into the transplanted graft. Further studies using in vivo and ex vivo systems are needed to confirm our findings and to propose that NK cells could be considered as a potential target for the rational treatment of kidney allograft rejection.
DOI : 10.6002/ect.2018.0142
From the 1Department of Immunology, School of Medicine, Kerman
University of Medical Sciences, Kerman, Iran; the 2Department of
Immunology, School of Medicine, Tehran University of Medical Sciences, Tehran,
Iran; the 3Chronic Kidney Disease Research Center, Labbafinejad
Medical Center, Shahid Beheshti University of Medical Sciences, Tehran, Iran;
the 4Molecular Immunology Research Center, Tehran University of
Medical Sciences, Tehran, Iran; and the 5Department of Epidemiology
and Biostatistics, School of Public Health, Tehran University of Medical
Sciences, Tehran, Iran
Acknowledgements: This project (MSc thesis) was financially supported by Tehran University of Medical Sciences, Tehran, Iran (grant number 26354). The authors have no conflict of interest to declare.
Corresponding author: Ali Akbar Amirzargar, Department of Immunology, School of Medicine, Tehran University of Medical Sciences, Tehran, Iran
Phone: +98 21 88953009
Table 1. Demographic Characteristics of Study Patients
Table 2. Frequency of Expressed Surface CD107a and Secreted Intracellular Interferon-γ in Natural Killer Cells
Table 3. Receiver Operating Characteristic Curve Analysis of Sensitivity and Specificity of Parameters for Detecting Patients With Transplant Rejection
Figure 1. Frequency of Natural Killer Cells in the Study Groups
Figure 2. Natural Killer Cell Cytotoxic Activity and Intracellular Interferon-γ Production Following Cell Stimulation Through Target Cell
Figure 3. Measurement of CD107a Expression on Natural Killer Cell Surface and Secreting Intracellular Interferon-γ Cytokines in Peripheral Blood Mononuclear Cells