Background/Purpose:

Rheumatoid arthritis (RA) associated anti-citrullinated protein autoantibodies (ACPA) bind a wide range of citrullinated proteins with high cross-reactivity. Recent findings suggest that certain, but not all, ACPA may have pathogenic properties. Citrullination occurs during physiological processes such as apoptosis, neutrophil extracellular trap (NET) formation and histone modifications, yet little is known about the interaction of ACPA with these antigens. Since uncleared apoptotic cells and NET products have been postulated to be central sources of autoantigen and immunostimulation in autoimmune disease, we sought to determine the anti-nuclear and anti-neutrophil ACPA reactivity.

Methods:

We screened a total of 11 recombinant single B-cell isolated RA monoclonal human ACPA-IgG, derived from different cellular origin and compartments, for binding to full-length citrullinated nuclear antigens including histones and hnRNPs by ELISA. All ACPA mAbs have previously been confirmed anti-CCP2 reactive, with specific and extensive cit-peptide reactivity demonstrated by ELISA and antigen-arrays. ACPA apoptotic cell binding was determined by flow cytometry and immunoprecipitation. Immunofluorescence was utilized for primary human neutrophil and the Ecom-G murine neutrophil systems. ACPA binding to Ecom-G cells was also evaluated with flow cytometry. Anti-nuclear antibody reactivity (ANA) was evaluated with standard HEp-2 tests. CRISPR-Cas9 technology was used to generate peptidylarginine deiminase type 4 (PAD4) KO neutrophils and chlor-amidine (Cl-A) was used as pharmacological inhibition of PAD enzymes.

Results:

We could observe that a subset of three monoclonal ACPA (37CEPT2C04, 37CEPT1G09, 1325:01B09) had high ANA reactivity and high binding to apoptotic cells. This did to some extent correlate with ELISA reactivity patterns to full-length citrullinated histones, but less so to citrullinated hnRNPs. Indeed, immunoprecipitations from apoptotic cell lysates revealed citrullinated histones as primary targets. One of the ACPA (1325:01B09) had a high capacity to induce IL-8 release from human peripheral mononuclear cells in plate-bound immune complexes with citrullinated histones. Importantly, all the anti-nuclear ACPA mAbs bound strongly to activated murine and human neutrophils and NETs. However, we also found two mAbs (1325:01C03 and BVCA1) in a second NET-reactive ACPA subset with contrasting perinuclear neutrophil binding and cross-reactivity patterns more towards the cytoplasmic antigens vimentin and alpha-enolase. Notably, CRISPR-Cas9 KO studies and pharmacological PAD inhibition showed that the cytoplasmic NET binding was fully dependent on PAD4 while the nuclear and histone mediated NET-reactivity was independent of PAD4.

Conclusion:

When investigating monoclonal ACPA, we identified distinct subsets based on binding to neutrophils and NETs with either a nuclear pattern or a cytoplasmic perinuclear pattern. Importantly, only the cytoplasmic anti-neutrophil ACPA binding was PAD4 dependent. This may have important functional impact and provide insights in RA pathogenesis.

« Back to 2018 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/pad4-independent-interaction-of-acpa-with-nuclear-antigens-in-apoptotic-cells-and-neutrophil-extracellular-traps-nets-defines-a-subset-of-autoantibodies/