|Home | About | Journals | Submit | Contact Us | Français|
Natural killer (NK) cells are important in the immune response against tumors and virally infected cells. A balance of inhibitory and activating receptors controls the effector functions of NK cells. We examined the fate of circulating NK cells and the expression of the NK cell activating receptors in pediatric liver transplant recipients. Blood specimens were collected from 38 pediatric liver transplant recipients before transplant, and at 1 week, 1, 3, 6, and 9 months, and 1 year post-transplant. PBMC were isolated and analyzed for the levels of NK cell activation receptors NKp30, NKp46, and NKG2D in the CD56dimCD16+ and CD56brightCD16+/− subsets of NK cells. We demonstrated that there is a significant decrease in the number of circulating NK cells post-transplant (pre-transplant 7.69 ± 1.54 vs. one week post-transplant 1.73 ± 0.44) in pediatric liver transplant recipients. Interestingly, NKp30 expression is significantly increased while NKp46 and NKG2D levels remain stable on the NK cells that persist at one-week post-transplant. These data indicate that the numbers and subsets of circulating NK cells are altered in children after liver transplantation.
Natural killer (NK) cells are granular lymphocytes that have been established as effectors involved in both the innate and adaptive immune responses. The cytotoxicity of NK cells allows the targeting of tumor and virally-infected cells without prior sensitization. The phenotype of NK cells can specify the type of response as either cytotoxic or cytolytic and this is defined by expression of cell surface markers CD16 and CD56 (1). CD16 is a low affinity Fc receptor expressed on NK cells and other cell types that is important in NK cell activation and antibody-dependent cellular cytotoxicity (2). CD56 (NCAM) is a membrane glycoprotein adhesion molecule expressed on NK cells, CD4 and CD8 T cells, neuronal cells and some tumor cells. High levels of CD56 expression is detected on about 10% of NK cells (CD56bright) in the periphery while the majority of NK cells express low levels of CD56 (CD56dim) (1, 3).
The CD56dim and CD56bright NK cell subsets demonstrate a capacity for high cytotoxicity and high cytokine production, respectively. The CD56dim subset preferentially expresses killer-cell immunoglobulin-like receptors (KIRs) and possesses potent cytolytic function in vitro, though it has a low proliferation and cytokine production potential. This high level of cytolytic activity in the CD56dim subset contrasts with the low cytotoxic function, but robust proliferation and high levels of inflammatory cytokine secretion seen in the CD56bright subset (4–5).
NK cell functionality is further defined through the expression of activating (or stimulatory) and inhibitory receptors. Inhibitory receptors recognize MHC class I alleles and prevent NK cell activation. Conversely, signals from specific stimulatory receptors on NK cells are necessary for NK cell activation (6). Thus, a balance of signals from inhibitory and activating receptors controls NK cell activity. Along these lines, the density of activating receptors has been demonstrated to be important for NK-mediated cytolysis (7).
Ligands for activation receptors that have been identified include viral proteins, stress or transformation induced molecules, and a tumor-specific B7 family member (8). The activating receptors NKp30 and NKp46, members of the natural cytotoxicity receptor (NCR) family, are type I glycoproteins that belong to the Ig superfamily. NCR have been implicated in the recognition and control of many viral infections including influenza, hepatitis C, and HIV (9–10). NKG2D is a type II transmembrane-anchored C-type lectin receptor found on NK cells, some NKT cells and on some populations of T cells. NKG2D interactions with its ligands can activate NK cell effector function including proliferation, cytotoxicity, and cytokine secretion. In humans, two families of NKG2D ligands (NKG2DL) have been reported, the MHC class I related antigen A (MICA), MICB, and members of the UL-16-binding protein (ULBP) family. NKG2D is important in mediating tumor clearance and the regulation of viral infections (11–14). Moreover, aberrant NKG2D signaling has been associated with numerous autoimmune diseases and graft rejection (12, 15–17). Since NK cells are important for an appropriate immune response and altered expression of NK cell receptors has been associated with disease states, we performed an analysis of NK cell activation receptors in pediatric liver transplant recipients in the critical early period post-transplant.
Thirty-eight pediatric liver transplant recipients at Lucile Packard Children’s Hospital at Stanford were consented and enrolled with Stanford Institutional Review Board approval as per the guidelines outlined by the Administrative Panels on Human Subjects in Medical Research. The demographics of enrolled patients are detailed in Table I. An equal distribution of male and female patients participated in the study, with a majority of the patients being under 5 years of age (Groups 1 and 2) and of Caucasian, Hispanic, or Asian descent. As would be expected, biliary atresia was the most common primary diagnosis for liver transplantation. Routine histocompability typing was not performed. Blood specimens were obtained in heparinized collection tubes immediately pre-transplant, 1 week and 1, 3, 6, 9, 12 months post-transplant although it was not possible to obtain all these time points from all the patients. Induction immunosuppression consisted of one dose of hydrocortisone (10 mg/kg) on the day of transplant, followed by methylprednisolone, 5 mg/kg, tapered to 1 mg/kg by day five post-transplant. Patients transplanted before August 2009 received Daclizumab: one dose of 2 mg/kg on the day of transplant, and one dose of 1 mg/kg three days post-transplant except for one patient who received thymoglobulin because of a positive cross-match. Maintenance immunosuppression consisted of tacrolimus with a trough level maintained at 12–15 ng/ml for the first month post-transplant, and at 6–8 ng/ml thereafter. One patient was treated with sirolimus (1 mg every other day) instead of tacrolimus. One patient was not on any immunosuppression at the time of blood draw (>1 year) because of a prior history of post transplant lymphoproliferative disease.
Peripheral blood mononuclear cells (PBMCs) were isolated over a Ficoll-Hypaque PLUS (GE Healthcare, Piscataway, NJ, USA) density gradient in RPMI 1640 containing L-glutamine and 25 mM HEPES (cellgro, Manassas, VA, USA). Red blood cells were lysed in ACK (0.15 M NH4Cl, 10 mM KHCO3, 0.1 mM EDTA, pH 7.2) buffer and cells were washed twice prior to storage or labeling. PBMCs were stained immediately or stored in freezing medium (fetal calf serum and 10% DMSO) in liquid nitrogen.
PBMCs (1 × 106) were labeled on ice for 30 minutes with monoclonal antibodies CD16-FITC (eBioscience, 11–0168), CD56-Cy5 (BD Pharmingen, 555517), NKp30-PE (Beckman Coulter, PN IM3709), NKp46-PE (BD Pharmingen, 557991), and NKG2D-PE (eBioscience, 12–5878). IgG1-FITC (Dako, X0927), IgG1-RPE (Dako, 0928) and IgG1-PE-Cy5 (BD Pharmingen, 55750) served as isotype controls. Post-labeling, PBMCs were washed twice and resuspended prior to analysis by flow cytometry using CellQuest version 3.3 software (Becton Dickinson) on a BD FACScan™ Flow Cytometer. Data were analyzed in FlowJo 8.1.1 (Tree Star, Inc). Live lymphocytes were gated based on size and granularity (intermediate FSC and low SSC, respectively).
GraphPad Prism 5 was used for analysis of flow cytometry data. The Mann-Whitney statistical test was used for comparison of the means at each time point with a significance level of p < 0.05.
Three distinct subsets of NK cells were observed in blood specimens from transplant recipients, CD56bright CD16- and CD56bright CD16+ and CD56dim CD16+ (Fig 1). The total percentage of circulating NK cells was defined to include both the CD56dim and CD56bright populations. Since we did not observe any major differences between the CD56bright CD16− and CD56bright CD16+ populations in our study, they have been combined as CD56bright and subsequent analyses refers to two populations, CD56bright and CD56dim. Before transplant, 7.69 ± 1.54% NK cells were detected in the circulation (Figure 2A). Uniformly, however, there was a dramatic decrease (p < 0.005) in the numbers of NK cells in the circulation at one-week post-transplant (pre-transplant 7.69 ± 1.54 vs. one week post-transplant 1.73% ± 0.44). However, at subsequent time points, the percentage of circulating NK cells stabilized to pre-transplant levels (one month - midyear: 4.61 ± 1.84, > midyear - one year: 7.52 ± 2.21, > one year: 5.34 ± 0.81).
In healthy people, the majority of NK cells in the circulation are CD56dim as are NK cells from patients pre-transplant, CD56dim: 5.86 ± 1.35 vs. CD56bright: 2.09 ± 0.58 (Figure 2B). As previously discussed, there is a dramatic decrease in NK cells at one week post-transplant; interestingly, the CD56dim subset is preferentially decreased. At one week post-transplant, the percentage of the CD56dim and CD56bright subsets is similar (0.93% ± 0.34 vs. 0.80% ± 0.12). Although the percentage of the CD56dim subset rebounds by one month post –transplant (one month-midyear: 3.51 ± 1.68, >midyear - one year: 6.34 ± 1.98, > one year: 3.82 ± 0.58), it still does not reach the pre-transplant level beyond one year post-transplant.
The pre-transplant ratio of CD56dim to CD56bright NK cells was between 4–5 (4.39 ± 1.26) (Figure 2C). In contrast, the ratio of CD56dim to CD56bright NK cells at one-week post-transplant was 1.07 ± 0.23. These data indicate that the percentage of circulating NK cells, particularly those belonging to the CD56dim subset, is dramatically decreased early after transplant.
The expression of NKp30, NKp46, and NKG2D was examined in 38 patients pre- and post-transplant in both the CD56dim (Fig 3A) and CD56bright (Fig 3B) NK cell populations. NKp46, which was expressed on the vast majority of cells pre-transplant, remained highly expressed on the NK cells of most pediatric liver transplant recipients post-transplant (Fig 3AB, center). In contrast, expression of NKp30 increased significantly in both the CD56dim and CD56bright (p<0.05) NK cell subsets one-week post-transplant (CD56dim: pre-transplant 14.25% ± 4.76 vs. 34.87% ± 7.45 one-week post-transplant; CD56bright: pre-transplant 24.54% ± 4.84 vs. 48.22% ± 6.76 one-week post-transplant) (Fig 3AB, left). Subsequent time points demonstrated NKp30 expression close to the pre-transplant levels. There was a similar trend towards increased NKG2D expression in both the CD56dim and CD56bright subsets one-week post-transplant with a return to baseline levels by one month post-transplant (Fig 3AB, right).
The pediatric patient population examined had minimal complications post-transplant thus the collection of samples did not generally overlap with rejection or infection events. However, we did examine the expression of NKp30, NKp46, and NKG2D in the CD56dim and CD56bright population in a patient with an acute rejection episode at one week post-transplant (Fig. 4). There was a modest increase in NKp30 expression in both the CD56dim and CD56bright populations and a dramatic increase in NKG2D expression in the CD56dim population (a pre-transplant level of 42.5% to one-week post-transplant level of 95.9%). The marked increase in NKG2D expression at the time of allograft rejection agrees with previous reports from our lab and others demonstrating a role for NKG2D in rejection (16, 18–19). Taken together, our data indicate that the NK cells that remain in the circulation in the early transplant period retain robust expression of NK cell receptors capable of inducing cytotoxicity and cytolytic effector functions.
Our data demonstrate a significant decrease in circulating NK cells early post-transplant in pediatric liver transplant recipients. We suggest that two plausible reasons for this decrease are the direct or indirect effects of immunosuppression or the migration of these circulating NK cells to the graft. The effects of immunosuppressive agents on NK cell numbers and function remains controversial. We have previously demonstrated both in vitro and in vivo that cyclosporine and tacrolimus do not effect NK cell proliferation or cytokine production although treatment with sirolimus does impair NK cell numbers and function (20). Corticosteriods are thought to impair NK cell function and have been reported to decrease expression of the activating receptors NKp30 and NKp46 (21). However, a recent report suggests that glucocorticoids, including methylprednisolone, in combination with IL-15 expand NK cells and retain functional capacities (22). Our results suggest that the activation receptors, NKp46 and NKp30 are actually increased after transplant. It has been reported, in a model of allogeneic hematopoeietic stem cell transplantation, that the CD56bright subset showed greater resistance to the effects of immunosuppressive agents as compared to the CD56dim subset (23). Indeed, daclizumab has been reported to expand the CD56bright NK cell subset however we did not detect any differences in the percentage of CD56bright NK between the children who received daclizumab and those that did not. In our study NK cells were significantly decreased at one week post-transplant in all children.
NK cells may also leave the circulation and traffic to the allograft in the early weeks post-transplant. It is well established that NK cells constitute a large proportion of the lymphocytes within the liver (24). Our results in an experimental model of liver transplant demonstrate that recipient-derived NK cells can be detected in the allograft as early as six hours post–transplant. Furthermore, there is a marked increase in NK cells in the graft and a corresponding decrease of NK cells in the circulation early post-transplant (25). Finally, it has been shown that hepatic NK cells are enriched in the CD56bright NK cell subset and that these cells can recirculate for two weeks after transplant thus it is possible that some of the CD56bright NK cells in the circulation are actually of donor origin in the first week post-transplant (26). Since the pediatric liver transplant recipients in the current study had minimal adverse events early post-transplant and our center does not perform protocol biopsies, tissue was not available to quantitate the numbers and subsets of NK cells in the liver allograft. It is important to note that the levels of NK cells stabilize and return to pre-transplant levels, by six months post-transplant, supporting a homeostatic interaction between the graft and the periphery.
The significant decrease in NK cells in early post-transplant is noteworthy since NK cells are important in the anti-viral immune response and activation receptors are essential in the recognition of virally-infected cells. Both NKp30 and NKG2D have been demonstrated to be important in the immune response to cytomegalovirus (CMV). Human NKp30 recognizes the pp65 protein of CMV and human CMV has been shown to affect expression of NKG2D ligands (27–28). However, when we examined the levels of NK cell activation receptors in pediatric liver transplant recipients, we found that NKp46 and NKG2D were generally expressed at pre-transplant levels in both the CD56dim and CD56bright NK cell subsets after transplant. In contrast, NKp30 was significantly increased in both NK cell subsets one week after transplant and gradually returned to pre-transplant levels over the next 12 months. Determination of NK cell function in the CD56dim and CD56bright subsets is clearly of interest, however sufficient numbers of NK cells could not be isolated from the small amount of blood obtained from these children. It should be noted that the levels of NKp30, NKp46 and NKG2D were all lower at one-year post-transplant compared to pre-transplant levels. These values may represent the true “normal” levels in children and that the expression levels in the pre-transplant samples were somewhat elevated as a result of underlying disease or end-stage liver disease. Examination of samples from normal healthy, age-matched children would clarify this minor point.
In the patient experiencing acute rejection, an elevated level of NKG2D was observed in the CD56dim subset. Numerous studies in experimental models and patient samples have detected an increase in NKG2D ligands during graft rejection. (16,18–19). The role of NKG2D-expressing cells in mediating rejection remains unclear although Kang et al. reported that treatment with an antibody to NKG2D was effective at blocking CD28-independent rejection of cardiac allografts (16). However this issue remains unsettled as we found that cardiac allograft survival was not significantly prolonged in the absence of NKG2D (Klrk −/− mice) (unpublished data). Interestingly, NK cells have been shown to be essential for the induction of allograft tolerance by a mechanism involving interactions with antigen presenting cells and dendritic cells (29–30).
Our data clearly show a transient, yet significant, decrease in both the circulating cytotoxic and cytokine producing CD56dim and CD56bright NK cell subsets in children immediately after liver transplantation. The increased expression of activation receptors on the spared NK cells may be a compensatory mechanism to help control viral infection. Further studies examining NK cell phenotypes and functions will clarify the role of NK cells in outcomes post-transplant.
This work was supported in part by funding from the NIH (AI44095), the Transplant and Tissue Engineering Center of Excellence at Lucile Packard Children’s Hospital and the Arnold and Barbara Silverman Endowment.
|Betty Pham||Sample Collection/preparation, Experiments, Data Analysis, Drafting manuscript|
|Karine Piard-Ruster||Sample Collection/preparation, Experiments, Drafting manuscript|
|Richard Silva||Sample Collection/preparation, Experiments|
|Amy Gallo||Sample Procurement, Data Interpretation, Review of Manuscript|
|Carlos O. Esquivel||Data Interpretation, Review of Manuscript|
|Olivia M. Martinez||Data Analysis and Interpretation, Review of Manuscript|
|Sheri M. Krams||Concept/Design, Data Analysis, Drafting manuscript|