Carbamylation of vimentin is inducible by smoking and represents an independent autoantigen in rheumatoid arthritis

Abstract

Objectives: Smoking has been connected to citrullination of antigens and formation of anti-citrullinated peptide antibodies (ACPAs) in rheumatoid arthritis (RA). Since smoking can modify proteins by carbamylation (formation of homocitrulline), this study was conducted to investigate these effects on vimentin in animal models and RA.

Methods: The efficiency of enzymatic carbamylation of vimentin was characterised. B-cell response was investigated after immunisation of rabbits with different vimentin isoforms. Effects of tobacco smoke exposure on carbamylation of vimentin and formation of autoantibodies were analysed in mice. The antibody responses against isoforms of vimentin were characterised with respect to disease duration and smoking status of patients with RA.

Results: Enzymatic carbamylation of vimentin was efficiently achieved. Subsequent citrullination of vimentin was not disturbed by homocitrullination. Sera from rabbits immunised with carbamylated vimentin (carbVim), in addition to carbVim also recognised human IgG-Fc showing rheumatoid factor-like reactivity. Smoke-exposed mice contained detectable amounts of carbVim and developed a broad immune response against carbamylated antigens. Although the prevalence of anti-carbamylated antibodies in smokers and non-smokers was similar, the titres of carbamylated antibodies were significantly increased in sera of smoking compared with non-smoking RA. CarbVim antibodies were observed independently of ACPAs in early phases of disease and double-positive patients for anti-mutated citrullinated vimentin (MCV) and anti-carbVim antibodies showed an extended epitope recognition pattern towards MCV.

Conclusions: Carbamylation of vimentin is inducible by cigarette smoke exposure. The polyclonal immune response against modified antigens in patients with RA is not exclusively citrulline-specific and carbamylation of antigens could be involved in the pathogenesis of disease.

Trial registration number: ISRCTN36745608; EudraCT Number: 2006-003146-41.

Keywords: DAS28; Rheumatoid Arthritis; Smoking.

Figure 1

Images showing the rapid citrullination and carbamylation of vimentin under optimal conditions detected by immunoblotting. Induction of both citrullination and carbamylation generates relevant amounts of antigen after 30 min of exposure. Citrullination can occur after carbamylation without affecting antigenic properties. (A) Upper lane: Confirmation of carbamylation using a specific serum. Lower lane: Control with commercial H-84 antibodies, which shows a reduced signal against vimentin after 10 min and disappearance after 30 min due to complete conversion. (B) Upper lane: Confirmation of citrullination using a positive anti-citrullinated peptide antibody (ACPA) serum from a patient with rheumatoid arthritis (RA). Lower lane: Control with commercial H-84 antibodies, which shows reactivities against vimentin as well as against citrullinated isoforms of vimentin. (C) Upper lane: Confirmation of citrullination after carbamylation using a positive ACPA serum from a patient with RA. Lower lane: Confirmation of carbamylation using specific serum. CarbVim, carbamylated vimentin; MCV, mutated citrullinated vimentin; MM, molecular weight marker.

Figure 2

Representations showing the development of antibody after immunisation of rabbits with carbamylated vimentin (carbVim). (A) Kinetics of antibody development after immunisation of rabbits with carbVim, vimentin and mutated citrullinated vimentin (MCV). Only immunisations with carbVim induced the formation of anti-carbVim antibodies. Human IgG-Fc from serum samples of healthy controls was isolated employing protein-A-affinity chromatography, separated under denatured conditions in SDS-PAGE and transferred onto nitrocellulose membrane for immunoblotting. IgG heavy chains were recognised by rabbit serum immunised with carbVim. (B) Differences in antibody specificities (profiles) after immunisation of rabbits with either carbVim (K1-3) or citrullinated vimentin (citrVim) (K4-6). Rabbits were immunised with 20 µg of the indicated antigen in Freund's complete adjuvant. Sera from the indicated animals were 1:200 diluted and analysed with the indicated antigens using enzyme immune assay (observed optical absorption are shown from a representative assay). RF, rheumatoid factor.

Figure 3

Representations showing carbamylated vimentin (carbVim) and antibodies against carbamylated antigens detectable in mouse sera after smoke exposure. (A) Serum samples derived from mice exposed to smoke contained detectable amounts of carbVim using anti-carb IgY in immunoblotting. Of note, the observed differences in the SDS-PAGE are well-known phenomena for the different isoforms of vimentin depending on their degree of citrullination and phosphorylation. (B) Antibodies with different recognition patterns to carbamylated proteins are detectable in sera of mice exposed to tobacco smoke. Aliquots of unmodified, citrullinated vimentin or carbVim together with modified and untreated HeLa cell fractions (F1 cytosolic proteins, F3 nucleic proteins and F4 components of the cytoskeleton) were subjected to SDS-PAGE (4%–12% NuPAGE, Invitrogen) for immunoblotting with the indicated sera and anti-vimentin antibodies as a reference. (C) Levels of anti-carbVim antibodies in sera of smoke exposed and control mice are shown. Mean of titres was shown. Mann-Whitney U test was used and statistical significance is indicated for p<0.05 (*). MCV, mutated citrullinated vimentin.

Figure 4

Plots showing the autoantibody levels in sera of patients with rheumatoid arthritis (RA) classified according to the smoking status. Comparison of autoantibody reactivities of anti-mutated citrullinated vimentin IgG (A MCV IgG), anti-citrullinated peptide of vimentin IgG (A P18 IgG), anti-carbamylated peptide of vimentin IgG (A HC52 IgG) and rheumatoid factor (RF) IgM in sera of patients with early onset RA having known smoking status. Smokers (n=42), non-smokers (n=68) and ex-smokers (n=34). Cut-off was identified as 20 U/mL. Values of mean with SEM are shown. Mann-Whitney U test was used and statistical significance is indicated for p<0.05 (*) and p<0.01 (**).