Anti–citrullinated protein antibodies (ACPAs) are the serologic hallmark of rheumatoid arthritis. Functional studies on the role of ACPAs in experimental arthritis have yielded conflicting results, and therefore the present study was undertaken to assess systematically whether citrullinated proteins can really induce ACPAs and modulate arthritis in mice.


Balb/c, SJL, and DBA/1 mice were immunized with either native or citrullinated fibrinogen, myelin basic protein (MBP), and type II collagen (CII). ACPAs were detected with a peptide-based enzyme-linked immunosorbent assay (ELISA) and with Western blotting using fibrinogen as substrate. Arthritis was induced in mice by immunization with CII in Freund's complete adjuvant or by injection of anticollagen antibodies.


Analysis of the sera of mice immunized with citrullinated proteins revealed false-positive results with the citrulline peptide–based ELISA. In contrast, Western blot analysis using either citrullinated or native fibrinogen as substrate reliably detected ACPAs in Balb/c mice immunized with citrullinated fibrinogen, MBP, and CII. However, these ACPAs failed to induce or aggravate disease in Balb/c mice in the anticollagen antibody–induced arthritis model. Immunization with citrullinated fibrinogen induced ACPAs but did not lead to arthritis development in SJL and DBA/1 mice. In contrast, immunization with citrullinated CII failed to induce ACPAs or enhance disease in these strains in the collagen-induced arthritis model.


Mice can develop genuine ACPAs, but detection of ACPAs is highly dependent on strain, immunogen, immunization protocol, and detection assay. Murine ACPAs are not overtly pathogenic, since neither preexisting ACPAs nor the use of citrullinated collagen as immunogen modulates the clinical course of arthritis.