Next-Generation HLA Sequence Analysis Uncovers Shared Risk Alleles Between Clinically Distinct Forms of Childhood Uveitis.

Purpose: Classical alleles of the human leukocyte antigen (HLA) complex have been linked to specific entities of pediatric noninfectious uveitis, yet genetic predisposition encoded by the HLA super-locus across the patient population remains understudied. Methods: We performed next-generation full-length sequencing of HLA-A, HLA-B, HLA-C, HLA-DPB1, HLA-DQB1, and HLA-DRB1 in 280 cases. Dense genotype data from 499 Dutch controls from Genome of the Netherlands were imputed using an HLA-specific reference panel (n = 5225 samples from European ancestry). Cases and controls were compared using logistic regression models adjusting for sex. Results: In total, 179 common and rare alleles were detected. Considering all cases and controls, HLA-DQB1*04:02 and HLA-DRB1*08:01 were identified as the principal HLA association, which was mainly driven by 92 cases with juvenile idiopathic arthritis-associated uveitis (JIA-U). The HLA-DQB1*04:02-HLA-DRB1*08:01 haplotype was also the primary association for the phenotypically similar idiopathic chronic anterior uveitis without arthritis (CAU). Also, HLA-DQB1*05:03 was an independent risk allele for CAU, but not in JIA-U. Analysis of 185 cases with other forms of uveitis revealed HLA-wide associations (P < 2.79 × 10-4) for HLA-DRB1*01:02, HLA-DRB1*04:03, and HLA-DQB1*05:03, which could be primarily attributed to cases with panuveitis. Finally, amino acid substitution modeling revealed that aspartic acid at position 57 that distinguishes the risk allele HLA-DQB1*05:03 (for CAU and panuveitis) from nonrisk alleles, significantly increased the binding capacity of naturally presented ligands to HLA-DQ. Conclusions: These results uncovered novel shared HLA associations among clinically distinct phenotypes of pediatric uveitis and highlight genetic predisposition affecting the antigen presentation pathway.

High-resolution mapping identifies HLA class II associations with multifocal motor neuropathy.

OBJECTIVE: To gain further insight in the immunopathology underlying multifocal motor neuropathy (MMN) by exploring the association between MMN and the human leukocyte antigen (HLA) class II DRB1, DQB1, and DQA loci in depth and by correlating associated haplotypes to detailed clinical and anti-ganglioside antibody data. METHODS: We performed high-resolution HLA-class II typing for the DRB1, DQB1, and DQA1 loci in 126 well-characterized MMN patients and assessed disease associations with haplotypes. We used a cohort of 1305 random individuals as a reference for haplotype distribution in the Dutch population. RESULTS: The DRB1*15:01-DQB1*06:02 haplotype (OR 1.6 [95% CI 1.1-2.2], p < 0.05) and the DRB1*12:01-DQB1*03:01 haplotype (OR 2.7 [95% CI 1.2-5.5], p < 0.05) were more frequent in patients with MMN than in controls. These haplotypes were not associated with disease course, response to treatment or anti-ganglioside antibodies. CONCLUSIONS: MMN is associated with the DRB1*15:01-DQB1*06:02 and DRB1*12:01-DQB1*03:01 haplotypes. These HLA molecules or gene variants in their immediate vicinity may promote the specific inflammatory processes underlying MMN.

Completion of HLA protein sequences by automated homology-based nearest-neighbor extrapolation of HLA database sequences.

The IMGT/HLA database contains every publicly available HLA sequence. However, most of these HLA protein sequences are restricted to the alpha-1/alpha-2 domain for HLA class-I and alpha-1/beta-1 domain for HLA class-II. Nevertheless, also polymorphism outside these domains may play a role in alloreactivity after transplantation. Several computational algorithms that aim for predicting alloreactivity after transplantation, HLAMatchmaker and the PIRCHE algorithm, require a major or the whole part of the HLA protein sequence as input for their prediction. In this study we describe an automated homology-based nearest-neighbor method to extrapolate incomplete HLA protein sequences. To get insight in the reliability of our extrapolation method, we performed a 10-fold cross-validation. The majority of the amino acid positions of the individual HLA class-I and -II proteins were correctly predicted. Eplets as defined by HLAMatchmaker were located both at correctly predicted as well as at lower reliably predicted amino acid positions. Moreover, six out of seven completely sequenced HLA alleles with previously unknown exons sequences were in agreement with the extrapolated amino acid sequences. In conclusion, incomplete HLA sequences can be completed by a homology-based nearest-neighbor principle. Since this method is automated, future submitted incomplete HLA sequences can easily be extrapolated.

Predicted Indirectly ReCognizable HLA Epitopes Class I Promote Antileukemia Responses after Cord Blood Transplantation: Indications for a Potential Novel Donor Selection Tool.

Unrelated cord blood transplantation (UCBT) provides a curative therapy for patients with hematological malignancies. The effect of HLA mismatches in UCBT is currently the subject of debate. HLA-mismatched UCBT may lead to improved leukemia control but also to graft-versus-host disease (GVHD), resulting in nonrelapse mortality (NRM). The aim of this study was to investigate whether indirect recognition of mismatched HLA provides an explanation for the graft-versus-tumor effect and risk of GVHD. The probability of indirect recognition was predicted by the Predicted Indirectly ReCognizable HLA Epitopes (PIRCHE) model. The effect of the numbers of PIRCHE presented on HLA class I and II (PIRCHE-I and -II) was studied in 134 pediatric patients. To study the effects of higher numbers of PIRCHE, patients were divided in 2 equally sized groups, using the median number of PIRCHE as cutoff values. Proportional hazard models and competing risk analyses were performed to study the effect of PIRCHE on the clinical outcomes relapse, acute and chronic GVHD, NRM, and disease-free and overall survival. Above median PIRCHE-I were associated with reduced relapse risk (HR, .26; 95% CI, .07 to .94; P = .04), evaluating the 50 patients transplanted for a malignancy. Both PIRCHE-I and -II were not associated with other clinical outcomes, including GVHD and NRM. These data suggest that high PIRCHE-I may lead to improved graft-versus-tumor effects after UCBT, without an accompanying GVHD risk. Inclusion of PIRCHE in UCB selection criteria may enhance UCBT outcome, which needs to be tested in prospective studies.

Refinement of the definition of permissible HLA-DPB1 mismatches with predicted indirectly recognizable HLA-DPB1 epitopes.

Hematopoietic stem cell transplantation with HLA-DPB1-mismatched donors leads to an increased risk of acute graft-versus-host disease (GVHD). Studies have indicated a prognostic value for classifying HLA-DPB1 mismatches based on T cell-epitope (TCE) groups. The aim of this study was to determine the contribution of indirect recognition of HLA-DP-derived epitopes, as determined with the Predicted Indirectly ReCognizable HLA Epitopes (PIRCHE) method. We therefore conducted a retrospective single-center analysis on 80 patients transplanted with a 10/10 matched unrelated donor that was HLA-DPB1 mismatched. HLA-DPB1 mismatches that were classified as GVH nonpermissive by the TCE algorithm correlated to higher numbers of HLA class I as well as HLA class II presented PIRCHE (PIRCHE-I and -II) compared with permissive or host-versus-graft nonpermissive mismatches. Patients with acute GVHD grades II to IV presented significantly higher numbers of PIRCHE-I compared with patients without acute GVHD (P < .05). Patients were divided into 2 groups based on the presence or absence of PIRCHE. Patients with PIRCHE-I or -II have an increased hazard of acute GVHD when compared with patients without PIRCHE-I or -II (hazard ratio [HR], 3.19; 95% confidence interval [CI], 1.10 to 9.19; P < .05; and HR, 4.07; 95% CI, .97 to 17.19; P = .06, respectively). Patients classified as having an HLA-DPB1 permissive mismatch by the TCE model had an increased risk of acute GVHD when comparing presence of PIRCHE-I with absence of PIRCHE-I (HR, 2.96; 95% CI, .84 to 10.39; P = .09). We therefore conclude that the data presented in this study describe an attractive and feasible possibility to better select permissible HLA-DPB1 mismatches by including both a direct and an indirect recognition model.

Complete donor chimerism is a prerequisite for the effect of Predicted Indirectly ReCognizable HLA Epitopes (PIRCHE) on acute graft-versus-host disease.

Predicted indirectly recognizable HLA epitopes (PIRCHE) computationally predict donor T-cell recognition of mismatched-HLA derived peptides following allogeneic haematopoietic stem-cell transplantation (allo-HSCT), as is evidenced by the correlation between presence of HLA-DPB1-derived PIRCHE and the occurrence of graft-vs.-host disease (GVHD). Complete donor T-cell chimerism associates with an increased GVHD risk compared to mixed patient and donor chimerism. If the correlation between the presence of PIRCHE and GVHD occurrence is indeed mediated by donor T cells, the presence of donor T cells should be required to observe such a correlation. This study was initiated to investigate whether the effect of PIRCHE is different in patients with complete chimerism compared to those with mixed chimerism. Indeed, the correlation between PIRCHE and GVHD is present in patients with complete chimerism, whereas it is absent in those with mixed chimerism. The data presented here suggest that chimerism status is important for the detection of potential GVHD epitopes.