Studies on the role of autoantibodies and autoantigens in rheumatoid arthritis and myositis

Abstract

A major population of patients affected with rheumatoid arthritis (RA) and inflammatory idiopathic myopathies (IIM, collectively called myositis) is characterized by the presence of autoantibodies. The pathogenic impact of anti-citrullinated protein/peptide (ACPA) or anti-histidyl transfer RNA synthetase (HisRS; Jo1) autoantibodies remains largely unknown. The aims of the thesis were to understand the molecular mechanisms underlying the autoimmune component (specifically, the auto-antibody and -antigen dynamics) in RA and myositis; and to develop ACPA neutralizing compounds.

Paper I: Purified anti-cyclic citrullinated peptide antibodies (aCCP2 IgG) were estimated to represent ~2% of the total RA IgG pool, and demonstrated to display distinct and individual reactivities against citrullinated epitopes from fibrinogen, collagen type II, α-enolase and vimentin. In-vivo generated autoantigens in synovial tissue and fluid were recognized by aCCP2 IgG.

Paper II: Anti-CCP2 IgG purified according to the methodology described in paper I were injected in mice and shown to induce pain-like behaviour. The underlying molecular mechanism appears to be chemokine-dependent because 1) aCCP2 IgG promoted activation of murine osteoclasts with generation of CXCL1 (human IL-8); 2) aCCP2 IgG stained CD68-bearing macrophages/osteoclasts, closely localized to the calcitonin gene related peptide (CGRP)-positive sensory nerve fibers; 3) reparixin, a CXCL1/2 receptor antagonist, blocked aCCP2 IgG induced pain.

Papers III and IV: Endogenously citrullinated fibrinogen peptides were found to be recognized by a large portion of sera from the Epidemiological Investigation of RA (EIRA) cohort. CCP2+ RA sera displayed 65% and 15 % immune reactivity against peptides from the fibrinogen α chain Cit573 (563-583) and Cit591 (580-600), respectively. Peptides from the β chain, Cit72 (62-81) and Cit74 (62-81), were recognized by 35% and 53% of the CCP2+ RA patients. The same fibrinogen peptides were shown to in vitro individually bind and block purified aCCP2 IgG in a dose-dependent manner, displaying a maximum of 83% blockade. Approximately 90% autoantibodies were neutralized by Cit573 and Cit591 combined, and further improvement of the blocking capacity was registered (>90%) when incubating aCCP2 IgG with a cyclic version of Cit573.

Since the peptide Cit573 (also termed [Cit573]fib(563-583)) showed the best inhibition percentage, a truncated version was inserted into the stable framework sunflower trypsin inhibitor-1 (SFTI-1, and denoted s1[Cit573]fib(566-580)). This compound showed enhanced blocking capacity, 79% antibodies were neutralized with an estimated IC50 of 20 μM, in comparison to the linear counterparts (73% maximum inhibition with the IC50:s ranging from 59 to 123 μM). Stability in blood was also improved, with above 90% of s1[Cit573]fib(566-580) remaining after five hours, whereas the cyclic and linear counterparts were degraded after one hour. Using a mutant of s1, s2[Cit573,Arg575]fib(566-580), anti-[Cit573]fib(563-583) IgG (aCit573 IgG) were purified from RA plasma and estimated to comprise 0.33% of the total RA IgG pool. Binding affinity measurements demonstrated that aCit573 IgG bind the mutant scaffold peptide s3[Cit573,Arg575]fib(566-580) with high affinity (Kd = 2 nM). Thus, subsets of ACPA recognizing citrullinated fibrinogen epitopes appear to be of high affinity.

Paper V: The cytoplasmic autoantigen HisRS was found in extracellular compartments (bronchoalveolar lavage, BAL, sera and plasma). High circulating levels of HisRS were found in myositis, and further increased in anti-HisRS+ patients; significant levels of HisRS were also detected in healthy individuals, whereas anti-HisRS IgG, IgA and IgM autoantibodies were only detected in sera and BAL from patients with myositis. In addition, anti-TRIM21 IgG were also identified in myositis BAL, positively correlating with the presence of anti-HisRS IgG. A so far uncharacterized factor in BAL of myositis patients was found to bind exogenous HisRS. In experiments addressing platelet activation, HisRS was found to trigger platelet activation in a dose-response dependent manner at low picomolar concentrations.

In conclusion, the development of a methodology to isolate autoantibodies from RA patients’ plasma, serum and synovial fluid, provided opportunities to address the pathogenic role of ACPA. Purified aCCP2 IgG induced pain-like behaviour, raising a possible mechanism for the pain RA patients occasionally feel before clinical onset of disease or after being medicated and in remission. A stable, high affinity anti-citrullinated fibrinogen autoantibody blocking compound was designed and we propose ACPA neutralization with fibrinogen-derived peptides as a complimentary treatment strategy for ACPA+RA. ACPA pain mediated effect may also be a functional target, amenable for blocking. The presence of HisRS extracellularly suggests novel and so far undescribed functions, which merits further investigations. Finally, a possible coupled immune response among HisRS and TRIM21 in lungs of myositis patients provides new clues for the development of autoimmunity in myositis and the associated anti-synthetase syndrome.