Immunological Interaction of HLA-DPB1 and Proteinase 3 in ANCA Vasculitis is Associated with Clinical Disease Activity

Abstract

Background: PR3-ANCA vasculitis has a genetic association with HLA-DPB1. We explored immunologic and clinical features related to the interaction of HLA-DPB1*04:01 with a strongly binding PR3 peptide epitope (PR3_225-239).

Methods: Patients with ANCA vasculitis with active disease and disease in remission were followed longitudinally. Peripheral blood mononuclear cells from patients and healthy controls with HLA-DPB1*04:01 were tested for HLA-DPB1*04:01 expression and interaction with a PR3 peptide identified via in silico and in vitro assays. Tetramers (HLA/peptide multimers) identified autoreactive T cells in vitro. RESULTS: The HLA-DPB1*04:01 genotype was associated with risk of relapse in PR3-ANCA (HR for relapse 2.06; 95% CI, 1.01 to 4.20) but not in myeloperoxidase (MPO)-ANCA or the combined cohort. In silico predictions of HLA and PR3 peptide interactions demonstrated strong affinity between ATRLFPDFFTRVALY (PR3_225-239) and HLA-DPB1*04:01 that was confirmed by in vitro competitive binding studies. The interaction was tested in ex vivo flow cytometry studies of labeled peptide and HLA-DPB1*04:01-expressing cells. We demonstrated PR3_225-239 specific autoreactive T cells using synthetic HLA multimers (tetramers). Patients in long-term remission off therapy had autoantigenic peptide and HLA interaction comparable to that of healthy volunteers.

Conclusions: The risk allele HLA-DPB1*04:01 has been associated with PR3-ANCA, but its immunopathologic role was unclear. These studies demonstrate that HLA-DPB1*04:01 and PR3_225-239 initiate an immune response. Autoreactive T cells specifically recognized PR3_225-239 presented by HLA-DPB1*04:01. Although larger studies should validate these findings, the pathobiology may explain the observed increased risk of relapse in our cohort. Moreover, lack of HLA and autoantigen interaction observed during long-term remission signals immunologic nonresponsiveness.

Keywords: ANCA; HLA; HLA-DPB1 antigen; PR3; T cell; maintenance; remission.

Graphical abstract

Figure 1.

Patients with PR3-ANCA carrying HLA-DPB1*04:01 are less likely to maintain remission. (A) Kaplan–Meier curve of relapse by HLA-DBP1*04:01 carrier status for all patients achieving remission (n=267). (B) Kaplan–Meier curve of relapse for patients with PR3-ANCA alone n=130. (C) Kaplan–Meier curve of relapse for patients with MPO-ANCA alone (n=137). Adjusted hazard ratio for relapse in HLA-DPB1*04:01 carriers with PR3-ANCA, 2.06 (95% confidence interval, 1.01 to 4.20).

Figure 2.

PR3-ANCA is dominant among homozygous HLA-DPB1*04:01 carriers who relapse. In subgroup analysis of patients who were relapsing, ANCA serotypes differ by HLA-DPB1*04:01 carriage, with PR3-ANCA dominant among homozygous carriers (P<0.001, Fisher’s exact test).

Figure 3.

Patients with PR3-ANCA in long-term remission and healthy controls have low level of interaction of PR3 peptide with HLA-DPB1–expressing cells. FITC-labeled PR3 peptide (PR3_225–239) was incubated with PBMCs and subsequently stained with flow cytometry antibodies to identify DPB1⁺antigen-presenting cells that were also positive for FITC-labeled PR3 peptide. Light gray indicates one copy of DPB1, dark gray indicates two copies of DPB1. Black outline indicates data from validation cohort.

Figure 4.

PR3_225–239 loaded tetramers, but not control peptides, identify autoreactive T cells. (A) DPB1*04:01 tetramers were loaded with either PR3_225–239 or a scrambled peptide to detect PR3-specific, autoreactive peripheral CD4⁺ T cells within a 14-day expanded population from patients and healthy controls (all were HLA-DPB*04:01 carriers). (B) Paired active-remission samples from the same patients were tested for tetramer positivity and demonstrated dynamic changes in autoreactive T cell populations in two of four paired samples.