• Arthritis
• Autoantibodies
• Ant-CCP

A key characteristic of rheumatoid arthritis (RA), is the occurrence of antibodies against post-translationally modified proteins. Citrullination and anti-citrullinated protein antibodies (ACPAs) have been studied extensively.1 Carbamylation is another type of post-translational modification. During carbamylation, isocyanic acid reacts with the amine group of an amino acid. This mostly results in the conversion of lysine into homocitrulline.2 A subset of patients with RA harbour anticarbamylated protein (anti-CarP) antibodies and the presence of these antibodies is predictive of worse disease progression in ACPA-negative patients.3 Anti-CarP antibodies can be present in patients with arthralgia and their presence predicts the development of RA.4

A cornerstone of biomedical research is the use of animal models to explore basic pathophysiological mechanisms. Therefore, it is important to know whether antibodies against post-translationally modified proteins are present in these models. Although ACPAs were initially reported to be present in collagen-induced arthritis (CIA),5 ,6 this topic is now debated. Little is known about ACPA in other frequently used arthritis models. Recently, we demonstrated that a large part of mice with CIA harbour anti-CarP antibodies. Interestingly, similar to anti-type II collagen (CII) antibodies, anti-CarP antibodies could be detected prior to disease onset, resembling the kinetics of the anti-CarP response in humans.3–4 ,7 ACPA was not detected in these mice.7

We have now studied the presence of anti-CarP antibodies and ACPA in other frequently used arthritis models. The ACPA and anti-CarP ELISAs were performed as previously published.7 In serum from mice with collagen antibody induced arthritis (induced following the antibody cocktail manufacturer's instructions) and in serum from human tumour necrosis factor (TNF) transgenic mice (described in8), we did not find a significant increase in anti-CarP antibody levels compared with control animals/littermates. We could not detect ACPA either (figure 1). In rats with pristane-induced arthritis, CIA or adjuvant-induced arthritis (induced as described in9 ,10) we could not detect ACPA. Interestingly, similar to what we observed for mice with CIA7 and for patients with RA,3 we did detect anti-CarP antibodies in a proportion of the arthritic rats (figure 1). Table 1 shows an overview of the presence of ACPA and anti-CarP antibodies in different arthritis models.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Anti-CarP and ACPA in different animal arthritis models. Serum of arthritic animals and healthy controls was harvested and anti-CarP and ACPA levels were determined by ELISA using carbamylated, citrullinated or non-modified FCS.7 The depicted value is the OD of the modified FCS coated well minus the OD of the FCS coated well. Every symbol represents an individual animal and the line indicates the median. Outliers were determined using the Grubbs’ test and excluded from the data analysis. Different groups of animals were compared using a Mann-Whitney U test (**p<0.005, ***p<0.0005). ACPA, anti-citrullinated protein antibodies; AIA, adjuvant-induced arthritis; CAIA, Collagen antibody induced arthritis; anti-CarP, anticarbamylated protein; CIA, collagen induced arthritis; FCS, fetal calf serum; HC, healthy control; hTNFtg, human TNF transgenic; OD, optical density; PIA, Pristane-induced arthritis.

Theoretically the possibility exists that the assay used to detect ACPA is not sensitive enough. However, this assay was successfully used to detect ACPA in sera of patients with RA7 (data not shown). Furthermore, in previous studies, sera from rats with pristane-induced arthritis or human TNF transgenic mice did not react with cyclic citrullinated peptide (CCP₂) or citrullinated fibrinogen either.8 ,9 Interestingly, we only detected anti-CarP antibodies in models that require immunisation with an adjuvant and active involvement of the adaptive immune system.

In conclusion, anti-CarP antibodies can be detected in models that require activation of the adaptive immune system, indicating that these models can be used to study the role of anti-CarP antibodies in arthritis. Furthermore, these data indicate that the B cell tolerance to carbamylated antigens is readily broken, whereas the tolerance to citrullinated antigens is not, despite the high homology between citrulline and homocitrulline.