Objectives In addition to anti-citrullinated protein antibodies (ACPAs), antibodies targeting carbamylated (i.e. homocitrullinated) proteins (anti-CarP antibodies) have been described in rheumatoid arthritis (RA). However, the extent to which  anti-CarP antibodies are truly distinct from ACPA remains unclear, and few studies have focused on specific autoantigens. Here, we examine cross-reactivity between ACPA and anti-CarP antibodies, in the context of the candidate autoantigen α-enolase.

Methods Cross-reactivity was examined by immunoblotting of citrullinated and carbamylated proteins using purified ACPA; and by peptide absorption experiments, using the citrullinated α-enolase peptide CEP-1 and a homocitrulline-containing version (carb-CEP-1) in ELISA. The population-based case-control cohort EIRA (n = 2,836 RA; 373 controls) was screened for reactivity with CEP-1 and carb-CEP-1, using the ISAC multiplex array. Associations of RA subsets with smoking and genetic risk alleles were calculated using unconditional logistic regression models. Differences in antibody levels were examined using Mann-Whitney U test.

Results Affinity-purified ACPA was found to bind carbamylated proteins and homocitrulline-containing peptides, demonstrating definitive cross-reactivity between ACPA and anti-CarP antibodies. Anti-carb-CEP-1 reactivity in EIRA was almost exclusively confined to the CEP-1-positive subset, and this group (21%) displayed a particularly strong ACPA response with marked epitope-spreading. The small RA subset (3%) with homocitrulline-reactivity in the absence of citrulline-reactivity did not associate with smoking or risk genes, and had significantly lower anti-carb-CEP-1 antibody levels.

Conclusion Our data presented herein cast doubt on the specificity of anti-CarP antibodies in RA, which we posit may be a subset of cross-reactive ACPA.