Diversity and function of anti-modified protein autoantibodies in rheumatoid arthritis

Abstract

Rheumatoid arthritis (RA) is a chronic autoimmune disease particularly affecting synovial joints. Anti-citrullinated protein autoantibodies (ACPA) are detected in the serum of about 2/3 of RA patients and are being used to classify the disease. These autoantibodies may occur years before any signs of arthritis which implies that they are a cause rather than a consequence of disease. By using modified autoantigens to detect ACPAs in large patient cohorts and by purifying polyclonal autoantibodies from patients, the hunt for disease promoting autoantigens has been going on for decades. However, the relationship between ACPA specificity and any functional effects remains unclear. In order to understand the evolution, specificity, and function of autoreactive B cells in RA, the focus of this thesis is the generation of monoclonal antibodies (mAbs) from paired variable heavy- and light-chain immunoglobulin (Ig) sequences from identified single B cells from various RA tissue compartments.

We have identified citrulline-reactive autoantibody producing plasma cells in the synovium of RA patients with established disease. By generating single plasma cell derived mAbs, we learned that such autoantibodies may be directly involved in the pathogenesis of RA by promoting bone degrading osteoclasts. The mAbs were all multireactive to citrulline-peptides and citrullinated proteins, but with unique distinct binding patterns. We recognized glycine in +1 position to the citrulline and a fraction of the citrulline-reactive mAbs cross-reacted with carbamyl-peptides. In addition, the identified plasma cells displayed features of high somatic mutations and fragment antigen binding (Fab) variable N-glycosylation sites introduced by affinity maturation. By analyzing a selection of RA patient B cell derived mAbs for reactivity against apoptotic cells and activated neutrophils, we learned that a subset of the citrulline reactive mAbs bound nuclear antigens. Interestingly, a fraction of these mAbs could target nuclear histones independently of the citrullinating enzyme PAD by binding to acetylated histones. We explored the extent of the ACPA mAb multireactivity and cross-reactivity by acknowledging the importance of neighboring amino acids in addition to glycine in +1 to the citrulline. By analyzing the bone marrow plasma cell repertoire of RA patients, we observed differences in Ig-frequencies and variable Fab N-linked glycosylation sites between ACPA+ and ACPA- patients. We also found RA patient bone marrow plasma cell clonotypes. In addition, we identified citrulline-reactive bone marrow plasma cells that could bind activated neutrophils which strengthen previous reports of citrullinated histones as ACPA targets. Lastly, we identified autoantibodies against the oxidation-induced post translational protein modification adducts malondialdehyde (MDA) and malondialdehyde-acetaldehyde (MAA) in the bone marrow and lung of RA patients and individuals that also harbor ACPAs. A majority of these autoantibodies, that can promote osteoclastogenesis, need cross-linked MAA-protein for recognition independent of protein backbone.

Taken together, the generation of RA patient single B cell derived mAbs, have revealed remarkable features of the autoreactivity that increases the understanding of B cell involvement in the pathogenesis of RA.