Analysis of antibodies to cartilage oligomeric matrix protein in rheumatoid arthritis and in mouse models

Abstract

Rheumatoid arthritis (RA), a common autoimmune disease, affects 0.5-1% of the world’s population. It is not only an important health problem but also a socioeconomic problem considering that it can lead to joint destruction and disability. It is well known that the disease can actually start many years ahead of the clinical onset. The presence of autoantibodies observed in serum is a hallmark. These autoantibodies included antibodies reactive with immunoglobulin (rheumatoid factors, RF) and antibodies reactive to citrullinated proteins (ACPA). With time antibodies to joint proteins also appear. The functional importance of these autoantibodies is however not yet clear, and they could have both regulatory and functional importance in RA’s development.

Type II collagen (CII), as the most abundant protein in joint cartilage, is a most common and well-studied target in RA. Thus, collagen induced arthritis (CIA) is the most commonly used model in which RA-like arthritis can be induced in susceptible mouse strains following immunization with heterogeneous CII emulsified in an adjuvant. Despite CIA’s gold standard status, other models are needed to capture the entirety of this complex disease. Cartilage oligomeric matrix protein (COMP), as a major glycoprotein in the extracellular matrix (ECM) of cartilage and synovium, has recently become a candidate autoantigen in RA. COMP induced arthritis (COMPIA) is thus also developed to mimic RA conditions in susceptible mouse strains.

In study I, COMPIA was established in C57BL/6 mice and the major T and B cell epitopes involved in the development of this arthritis model were defined. In study II, a library of native and citrullinated triple helical peptides (THPs) of CII was established. This library also contained some previously defined epitopes. Some monoclonal antibodies reactive against these defined epitopes, sera from RA patients, and sera from experimental mouse models showed a unique reactivity toward THPs as compared to cyclic ones. This suggested that the binding of epitopes to their autoantibody might be triple-helical-conformation-dependent. In addition, the responses of three identified THPs were observed at elevated levels in sera from RA patients when compared to healthy controls. These data gave support for the potential use of these epitopes as new biomarkers in RA.

In study III, arthritis was induced after immunization with an anti-COMP antibody, 15A, in naïve mice. Levels of antibodies specific to 15A’s epitope, P6 peptide, and its citrullinated variants were increased significantly in RA patients compared to population controls, also correlating with a greater disease activity. In study IV, despite an attempt to establish COMPIA in mouse strains that express the human major histocompatibility complex type II (MHCII) alleles DRB1*0401, DRB1*0402, and DRB1*0405, only mild arthritis developed in DR0401 mice with low incidence. All strains developed a strong IgG response to COMP after immunization though. Major B cell epitopes were identified in both sera from these immunized mice and from a RA cohort. One citrullinated COMP peptide showed significantly elevated response in both RA and mice immunized with citrullinated COMP and was associated with disease activity in RA patients.