Background and Objectives The complement system is a major effector mechanism of both innate and adaptive immunity. It is activated in rheumatoid arthritis (RA) patients but the pathway involved remains unclear. RA is associated with the production of autoantibodies with some of them, anti-citrullinated peptide antibodies (ACPA), representing a diagnosis and even a prognosis marker. Antibodies are known to activate the complement system via the classical pathway in a C1q-dependent manner. But some models suggest C1q-independent pathways in arthritis. Therefore, we have investigated whether C1q is necessary for disease development in a mouse model of RA.

Materials and Methods Disease development was followed in the collagen-induced arthritis (CIA) mouse model. The impact of C1q was evaluated by comparing wild-type (WT) mice and C1q-knockout (KO) true littermates. Arthritis was followed by clinical score evaluation. Inflammation and bone destruction were estimated by histology. Anti-collagen antibody, blood cytokine and C3a levels were measured by ELISA. B and T lymphocytes as well as neutrophils were analysed by flow cytometry. Osteoclastogenesis was analysed by culturing bone marrow cells with M-CSF and RANKL and then counting TRAP-positive multinucleated cells by microscopy.

Results C1q is absolutely required for arthritis development in CIA as C1q-KO mice did not develop clinical signs of arthritis, although both WT and KO mice developed anti-collagen antibodies. Moreover, neither inflammation nor joint destruction were observed in C1q-KO mice. Surprisingly, no statistically significant difference was observed between WT and KO mice regarding the percentage or the activation of lymphocytes and neutrophils in the blood, spleen and lymph nodes. As a recent study suggested that C1q might influence osteoclastogenesis in vitro, we have shown that C1q deficiency does not alter osteoclast development in vivo in CIA mice. The impact of C1q on cytokine secretion and C3a production in vivo is currently being analysed.

Conclusions C1q-KO mice are nearly completely protected from arthritis development and thus C1q plays a key role in the CIA model. This suggests that the classical pathway cannot be compensated by the alternative (or the lectin) pathway in this model. Our results strongly differ from those reported in the serum-transfer model where C1q is not necessary for disease development, suggesting that C1q might be involved at different steps, maybe before antibodies are produced, for example in the response to damage-associated molecular patterns (DAMP).