Background/Purpose: In Rheumatoid Arthritis (RA), autoantibodies develop against endogenous proteins containing the amino acid citrulline, which results from a post-translational modification of arginine catalyzed by a family of peptidyl arginine deiminase (PAD) enzymes. Although citrullination is a ubiquitous physiological process, its impact on the structure and function of specific proteins is not well understood. Neutrophils contain 2 isoforms of PAD (2/4) and readily form citrullinated RA antigens during neutrophil extracellular trap (NETs) formation. However, the effects of citrullination on neutrophil protease activity have not been studied. We hypothesized that citrullination influences the activity of neutrophil serine proteases, namely Neutrophil Elastase (NE), Proteinase-3 (PR3) and Cathepsin G (CTG), which in turn may contribute to the pathogenesis of RA.

Methods: The activity of proteases and various biospecimens were analyzed using fluorogenic substrate after citrullination with PAD2/4 (confirmed by a cit-specific probe). Active protease sites were labelled and visualized using a probe (TAMRA-FP). Degradation of aggrecan, a proteoglycan found in cartilage, was determined. Cultured fibroblast-like synoviocytes (FLS) were treated with cit-PR3. IL-6/TNF and ACPA targeting cit-proteases were measured with ELISA.

Results: Using both recombinant proteases and neutrophil supernatant, proteolytic activity of NE, PR3 and CTG increased markedly after citrullination with PAD2, and to a lesser extent, PAD4 (Fig1A). For example, the activity of PAD2 cit-PR3 increased by 13-fold compared to native PR3. PAD2 also lead to enhanced stability of proteases, with detectable activity after 72 hours in-vitro, and evidence of protection from autoproteolysis (Fig1B-D) and exogenous proteases (trypsin, Fig 2A) in PR3 and NE. In the recombinant proteases, supernatant, and NETs, PAD2 citrullination enhanced the allosteric capacity of the protease multimers by opening new catalytic sites (Fig2A/B). This is suggestive of a conformational change induced by PAD which was confirmed by in-silico modelling (Fig2C). Models also suggested a higher binding energy between cit-PR3 and a peptide substrate. Aggrecan degradation was enhanced by PAD2-citrullination of NE and NETs. Using MMP and serine protease inhibitors, we found that degradation by NETs was undertaken by serine proteases, rather than aggrecanases (Fig2D). Serum cit-PR3 levels were elevated in RA than controls, and synovial fluid PR3, NE, and CTG activity were higher in RA compared to OA. FLS treated with cit-PR3 significantly increased the release of IL-6/TNF, and this was mediated by a protease activity receptor, PAR2 (Fig3C). Autoantibodies to PAD4 cit-proteases, but not PAD2, were higher in RA than controls (Fig3D).

Conclusion: PAD citrullination results in a gain of function for neutrophil serine proteases through conformational changes that result in enhanced proteolytic activity and stability. Although this process may provide a biological advantage for host defense, in the context of RA, this leads to degradation of cartilage components, inflammation and autoantibody formation.

Figure 1: PAD2-citrullination enhances neutrophil serine protease activity and stability. (A) The effect of PAD2 and PAD4 on recombinant serine protease activity (upper) and neutrophil supernatant protease activity (lower) measured with a fluorogenic substrate (increasing concentrations x-axis) and enzyme velocity. (B) Proteinase-3 activity after citrullination with PAD2 (orange) and PAD4 (green) up to 72 hours in-vitro. Only PAD2 citrullinated PR3 had detectable activity at 48 and 72h. (C) Protein staining (sypro-ruby) or citrullinated PR3 by PAD2 or PAD4 compared to buffer (Ca/DTT), measured every 24 hours, revealing protection of autoproteolysis mediated by PAD2 citrullination. (D) A 24-h analysis comparing PAD2/4 citrullination in neutrophil elastase (NE) and PR3. PR3: Proteinase 3, NE: Neutrophil Elastase, CTG: Cathepsin G, DTT: Dithiothreitol (reducing agent).

Figure 2: Citrullination induces a conformational change in neutrophil serine proteases which opens new catalytic sites to enhance activity, allowing for the degradation of aggrecan a proteoglycan in cartilage. (A) PAD2 citrullination of PR3 reveals new catalytic sites as evidenced by increased signal from a specific serine protease probe called TAMRA-FP. PR3 is protected from degradation by Trypsin, a common serine protease. (B) Evidence of new catalytic sites in neutrophil supernatants and NETs using TAMRA-FP in PAD2/PAD4 treated samples. (C) In-silico models of multimeric (dimer) proteinase-3 (upper structures) and a peptide ligand (VADCADQ, purple) which increases the distance between the 2 active sites in citrullinated PR3 but not native PR3 (lower left bar graph). Monomeric model (lower structure) revealing the peptide ligand (orange), citrulline (ball-stick) and the active serine (green sphere). SMINA affinity measure using a docking algorithm reveals a higher binding energy between the peptide substrate and cit-PR3 (lower right violin plot, P < 0.01). (D) Degradation of aggrecan and cartilage proteoglycan by recombinant NE (upper), NETs (middle) and NETs treated with an aggrecanase (matrix metalloproteinase) inhibitors called GM6001 or a biotinylated form of the FP probe which inhibits protease function (lower).

Figure 3: Proteases in RA serum, synovial fluid which serve as autoantigens. (A) Serum citrullinated proteinase-3 is higher in RA compared to controls (P < 0.0001) (B) Detection of PR3, NE, and CTG in RA synovial fluid compared to OA. (C) Treatment of fibroblast-like synoviocytes (FLS) with PR3 and cit-PR3 (PAD2) leads to increased IL-6 production, which is reduced with pre-treatment with a protease activated receptor (PAR2) inhibitor. (D) Autoantibodies targeting various citrullinated neutrophil proteases via PAD4 (upper) or PAD2 (lower) (* P < 0.05, ** P < 0.01, *** P < 0.001).

« Back to ACR Convergence 2025

ACR Meeting Abstracts - https://acrabstracts.org/abstract/citrullination-of-neutrophil-serine-proteases-enhances-proteolytic-activity-stability-and-autoantigenicity-in-rheumatoid-arthritis/