Soluble Human Leukocyte Antigen-G in Pediatric Renal Transplantation

Abstract

Human leukocyte antigen (HLA)-G is a nonclassical major histocompatibility complex class Ib antigen that displays immunotolerogenic properties toward the main effector cells involved in graft rejection. In vitro, HLA-G molecules inhibit natural killer- and CD8+ T-cell-mediated cytolysis and CD4+ T-cell alloproliferative response through its interactions with various inhibitory receptors and favor the development of tolerogenic dendritic cells and anergic/immunosuppressive CD4+ and CD8+ T cells (1). Clinical data indicate that high levels of soluble HLA-G contribute to better graft acceptance after adult hematopoietic stem cell, heart, liver, and combined liver/kidney transplantation (2). We have recently reported that elevated serum HLA-G levels track with operational tolerance of liver grafts and support favorable outcomes in pediatric and young adult recipients (3). Limited data have been reported regarding the association of HLA-G and adult kidney allograft acceptance. Thus, Creput et al. (4) detected high concentration of HLA-G in serum from liver/kidney transplant patients but not in kidney transplant patients; however, HLA-G expression in the liver graft was associated with better kidney graft outcome in liver/kidney recipients. In kidney transplantation alone, HLA-G expression in the allograft was reported to be associated with a lower frequency of acute rejection (AR) and chronic allograft nephropathy (5), and serum HLA-G was detected in 50% of patients with functioning graft vs. 20.5% of patients who rejected their transplants (6). To date, there are no data reported regarding the role of serum HLA-G in pediatric kidney transplant recipients. In a pilot investigation of the clinical relevance of soluble serum HLA-G molecules in pediatric kidney recipients (mean age 10.8±6.7 years), we analyzed serum HLA-G levels in 90 serum samples, collected longitudinally before transplantation and at 3 and 6 months posttransplantation, from each of the 30 pediatric renal transplant recipients. Each serum sample was accompanied by a protocol biopsy at the serum collection time points, at each of 0, 3, and 6 months posttransplantation per center protocol, using our previously published approach of steroid-free immunosuppression (7). Patients were stratified into two groups: 20 pediatric patients (9.3±7.6 years of age at transplant) with stable graft function (STA) and normal protocol biopsies and 10 patients (13.5±3.03 years of age at transplant) with biopsy proven (8) AR within the first 6 months (mean time 4.2±2.9 months) posttransplantation. HLA-G levels were measured in all 90 serum samples by a commercial enzyme-linked immunosorbent assay (USBiological, Swampscott, MA, cat. no. H6098-71). Results were correlated with recipient and donor demographic data (patient age, weight, gender, and donor type) and graft function (Schwartz calculated creatinine clearance [CrCl], [9]). T test, chi-square test, and Pearson correlation coefficients were calculated for parametric clinical parameters, and Spearman correlation coefficients and Wilcoxon signed-ranked test were calculated for nonparametric clinical parameters (SAS 9.1.2, SAS Institute Inc., Cary, NC). There was no correlation between serum HLA-G levels and demographics, immunosuppression, and time posttransplantation that is similar to our observations in pediatric liver recipients (3). Within the STA and the AR groups, we observed no differences in baseline serum HLA-G levels (11.8±13.1 U/mL in STA vs. 13.8±14.8 U/mL in AR, P=0.8). Serum HLA-G levels at 3 and 6 months posttransplantation were similar to the baseline levels in 20 STA (delta=−5.03, P=0.2 at 3 months; delta=1.9, P=0.8 at 6 months) and in 10 AR recipients (delta=−3.24, P=0.7 at 3 months; delta=−1.6, P=0.8 at 6 months). Thus, no significant differences in HLA-G levels were detected between STA and AR groups within 6 months posttransplantation (Fig. 1). Within the AR group, baseline serum HLA-G levels trended to positively correlate with CrCl at 3 months (r=0.87, P=0.05) and 6 months (r=0.91, P=0.09) posttransplant with no significant changes in HLA-G levels at the time of AR compared with 3 months prior-rejection levels (delta=−3.7, P=0.7), and no correlations between HLA-G levels and number or severity of AR episodes were detected. There was no correlation between serum HLA-G levels and CrCl in the STA recipients. Previous observations in adult renal recipients have suggested an immunoregulatory role of soluble (3, 6) and intragraft (4, 5) HLA-G expression in allograft accommodation. Our current study of soluble HLA-G molecules in pediatric kidney transplantation did not detect significant differences in HLA-G levels in STA recipients and patients with early AR episodes within 6 months posttransplantation. This observation is different from the results in our recent study of soluble HLA-G levels measured at 1.1 to 18 years posttransplantation in 42 pediatric and young adult liver allograft recipients (Fig. 1), where we detected significantly higher serum HLA-G levels in patients with operational tolerance and stable allograft function compared with recipients with AR after first year posttransplantation (3). This discrepancy in the results between liver and kidney pediatric transplant studies may be because of longer time posttransplantation and slightly older patients' age in the liver transplant study. The other possible explanation is early AR episodes (within first 6 months) in kidney pediatric recipients with fast recovery of graft function as confirmed by CrCl and follow-up biopsies, whereas liver recipients were considered rejectors if they had AR episodes after 1 year posttransplantation and thus may have more sever tissue damage. This hypothesis is supported by the analysis of intragraft HLA-G expression performed by Racca et al., (10) which revealed that patients with kidney allograft AR, who overcame rejection had a tendency toward higher HLA-G1 levels than those with nephrotoxic acute tubular necrosis. Also, different soluble (HLA-G5–HLA-G7) and membrane bound (HLA-G1–HLA-G4) isoforms may be involved in regulation of accommodation of different allograft types. Recently, only HLA-G1 isoform in peripheral blood mononuclear cells and biopsy samples was reported to have association with kidney allograft state (10).FIGURE 1.: Serum human leukocyte antigen (HLA)-G levels in pediatric kidney and liver transplant recipients. Soluble HLA-G levels were measured in the serum samples of pediatric kidney or liver transplant recipients in U/mL and plotted as a log10 scale on y-axis. As shown, serum HLA-G levels at 3 months posttransplantation were similar between AR and STA pediatric kidney recipients (P=0.5), whereas in liver allograft recipients at 1.1 to 18 years posttransplantation (recent study, [3]) significant serum HLA-G increases were detected in STA (P=0.01) and TOL (P=0.009) patients when compared with AR group. Serum HLA-G levels in both AR and STA kidney recipient groups were significantly lower than in STA (P=0.004) and TOL (P=0.005) liver transplant patients and close to that in AR liver recipients (P=0.2). Data are presented as average ± standard error. HV, healthy volunteers without transplant; AR, acute rejection group; STA, patients with stable allograft function on immunosuppression; TOL, patients with operational tolerance.In summary, our results suggest that, unlike in liver transplant pediatric recipients, serological monitoring for HLA-G levels in pediatric kidney transplant recipients may have limited benefit and poor correlation with the underlying phenotype of transplant injury, at least early after transplantation. Further studies should be targeted at serial analysis of soluble and membrane-bound HLA-G molecules later after transplantation to assess whether undulating levels of serum HLA-G can provide some direction for causality for the immunologically diverse outcomes after pediatric kidney transplantation and to evaluate and investigate the specific regulatory properties of HLA-G in organ transplantation. Valeriya Zarkhin Maria Bezchinsky Li Li Minnie M. Sarwal Pediatric Nephrology Stanford University Stanford, CA