Background/Purpose: Rheumatoid arthritis (RA) is an autoimmune disease in which citrullinated self-antigens are recognized by anti-citrullinated protein antibodies (ACPA) and T cells. ACPA are a key diagnostic indicator of disease and have been shown to bind to arrays of citrullinated self-proteins that are expressed in synovial tissue. To date, the majority of T cell studies have focused on well-documented ACPA targets. Although considerable antigenic overlap between self-reactive CD4+ T cells and antibodies can be reasonably anticipated, T cell targets that are not known to be targeted by ACPA are likely to comprise an important (but as yet, understudied) component of the antigen specific responses that underlie disease. Furthermore, although CD8+ T cells comprise a significant proportion of T cells within synovial tissue and fluid, knowledge about their specificity remains extremely limited. Therefore, we sought to generate relevant HLA peptidomics data sets to inform T cell antigen discovery efforts.

Methods: To characterize the disease proximal HLA peptidome, synovial tissue was obtained from 7 subjects with seropositive RA and 4 subjects with seronegative RA who had undergone arthroplasty procedures. Synovial cells were isolated from tissue after mincing and digesting in collagenase I. After confirming HLA expression by flow cytometry, samples were solubilized in lysis buffer and human leukocyte antigen (HLA) class I and HLA-DR complexes were captured (separately) on affinity columns. HLA-Class I- or HLA-DR-bound peptides were eluted in acetic acid, concentrated, and peptide spectra were identified by LC-MS/MS analysis. The resulting datasets were assigned sequences by searching against a human protein database. Mass shifts associated with each assigned sequence were utilized to identify post-translational modifications – most notably citrullination of native arginine residues. Comprehensive libraries of the HLA-Class I- and HLA-DR-bound peptides from each individual were imported into a custom database, which was then used to catalogue the most prevalent self-proteins for each patient type.

Results: The most prevalent self-proteins in the HLA-DR-bound peptidome from the synovial tissue of RA subjects included expected targets such as vimentin, alpha-enolase, fibrinogen, collagen, histones, and BIP but also contained more novel targets such as fibronectin, gelsolin, and proteoglycan 4. Prevalent self-proteins in the HLA-Class I-bound peptidome included well-studied CD4+ T cell targets such as vimentin, alpha-enolase, and collagen but also contained more novel targets such as caspase-14, stromelysin-1, and filamin-A. Among these candidates, we focused on gelsolin as a putative antigen, identifying multiple peptides with apparent T cell immunogenicity.

Conclusion: These findings demonstrate that HLA peptidomic analysis of synovial tissue provides a rich source of candidate antigens for T cell studies. Assessing T cell responses toward such novel antigens has the potential to provide important new insights about the specificity and character of antigen specific T cell responses that promote the development of RA.

« Back to ACR Convergence 2022

ACR Meeting Abstracts - https://acrabstracts.org/abstract/mass-spectrometry-based-hla-peptidomics-analysis-reveals-both-expected-and-novel-citrullinated-antigen-targets-in-rheumatoid-arthritis/