Enhanced B Cell Alloantigen Presentation and Its Epigenetic Dysregulation in Liver Transplant Rejection

Abstract

T cell suppression prevents acute cellular rejection but causes life-threatening infections and malignancies. Previously, liver transplant (LTx) rejection in children was associated with the single-nucleotide polymorphism (SNP) rs9296068 upstream of the HLA-DOA gene. HLA-DOA inhibits B cell presentation of antigen, a potentially novel antirejection drug target. Using archived samples from 122 white pediatric LTx patients (including 77 described previously), we confirmed the association between rs9296068 and LTx rejection (p = 0.001, odds ratio [OR] 2.55). Next-generation sequencing revealed that the putative transcription factor (CCCTC binding factor [CTCF]) binding SNP locus rs2395304, in linkage disequilibrium with rs9296068 (D' 0.578, r(2) = 0.4), is also associated with LTx rejection (p = 0.008, OR 2.34). Furthermore, LTx rejection is associated with enhanced B cell presentation of donor antigen relative to HLA-nonidentical antigen in a novel cell-based assay and with a downregulated HLA-DOA gene in a subset of these children. In lymphoblastoid B (Raji) cells, rs2395304 coimmunoprecipitates with CTCF, and CTCF knockdown with morpholino antisense oligonucleotides enhances alloantigen presentation and downregulates the HLA-DOA gene, reproducing observations made with HLA-DOA knockdown and clinical rejection. Alloantigen presentation is suppressed by inhibitors of methylation and histone deacetylation, reproducing observations made during resolution of rejection. Enhanced donor antigen presentation by B cells and its epigenetic dysregulation via the HLA-DOA gene represent novel opportunities for surveillance and treatment of transplant rejection.

Keywords: B cell biology; flow cytometry; genomics; immune; immunobiology; immunosuppression/immune modulation; liver transplantation/hepatology; monitoring; rejection:acute; translational research/science.

Conflict of interest statement

The authors of this manuscript have conflicts of interest to disclose as described by the American Journal of Transplantation. Disclosed conflicts of interest have been managed in accordance with the University of Pittsburgh’s policies and procedures. The other authors have no conflicts of interest to disclose.

© Copyright 2015 The American Society of Transplantation and the American Society of Transplant Surgeons.

Figures

Figure 1

Flow chart demonstrates steps to identify SNPs in linkage disequilibrium with rs9296068, which was associated with liver transplant rejection in our previous study. R=Rejector, NR=Non-rejector, MAF=minor allele frequency, dB=database.

Figure 2

Upper panel shows the HLA-DOA and its upstream SNPs. Arrows denote positions of SNPs and black bars denote transcription factor binding sites from ChIP-seq data of ENCODE. rs9296068 was associated with rejection previously. rs2305304, located closest to (179 bp) a well-characterized regulatory CTCF binding site termed C13, was selected for confirmatory sequencing and an evaluation of its association with rejection. Also shown is rs6906636, one of the nine putative regulatory SNPs identified with next generation sequencing (NGS). Lower panel shows chromatogram traces from Sanger sequencing for the reference allele, minor allele and the ambiguous allele at SNP rs2395304.

Figure 3

Panel A. Three scatterplots each from a rejector (upper three plots) and a non-rejector (lower three plots) show that recipient B-cells present more donor antigen or less donor antigen respectively, compared with HLA-non-identical reference alloantigen. Background=recipient B-cells cultured without antigen. Panel B. B-cell antigen presentation Index (API) in single samples from Rejectors (R, n=12) and Non-rejectors (NR, n=16). Rejectors show increased antigen presentation or API>1 when compared with Non-rejectors who show API

Figure 4

Serial changes in B-cell antigen presentation index (upper bar diagram) dCt of the HLA-DOA gene (middle bar diagram) and dCT of the CTCF gene (lower bar diagram) in four rejectors (black) and four non-rejectors (grey). Serial samples were obtained pre transplant (Pre-Tx) and at 1–60 days and >61 days after transplantation. p-values for R vs NR comparisons are obtained using Wilcoxon rank-sum exact test with continuity correction. p≤0.05 is significant. Samples were collected at a median interval of −1 (−13 to 0 days) before transplantation, and at 18 (8 to 60 days) and 209.5 (120 to 730 days) after transplantation. Among rejectors, samples collected during the 1–60-day period were collected at a median interval of 0 (0 to 41 days) before rejection

Figure 5

Scatterplots on left show Raji cells which express HLA-DOA (HLA-DOA+Raji cells, upper three scatterplots) and Raji cells which express alloantigen (alloantigen+Raji cells, lower three scatterplots) among untreated (left and middle columns) and MO-CTCF-FITC-transfected (right) Raji cells. Untreated Raji cells are shown without (left column, negative control) and with (middle column) HLA-DOA staining. Scatterplots depicting HLA-DOA staining after transfection (right column) have been used to obtain frequencies of HLA-DOA+cells and alloantigen+Rajic cells among non-transfected (b) and transfected (c) Raji cells. Bar diagrams on right summarize HLA-DOA+Raji cells (upper) and alloantigen+Raji cells (bar) seen in three replicates of this experiment. P values are shown in figure when non-transfected (b) and transfected cells (c) are compared. Knockdown of CTCF and HLA-DOA is confirmed with western blot (far right) as a second technique. FITC=fluorescein isothiocyanate, MO=morpholino antisense oligonucleotides.

Figure 6

Raji cells expressing HLA-DOA increase after treatment with 5-Aza or trichostatin A compared with untreated cells (upper four histograms). Alloantigen presentation by Raji cells decreases with 5-Aza and tricho in the same experiment (lower four scatterplots). Bar diagrams on right summarize results from three replicates of these experiments. P values are shown in figure when treated cells are compared with untreated cells. Aza=5-azacytidine, Tricho=trichostatin-A.