PMC

Immunohistochemical analysis of cartilage oligomeric matrix protein (COMP)-specific mAbs binding to cartilage in vitro. Immunohistochemistry was performed using 7 μm cryo-sections from naive neonatal mouse limbs, using 5 μg/ml of biotinylated COMP-specific mAbs. Binding was detected by extravidin-peroxidase and DAB substrate. Sections from COMP^+/+ mouse limbs stained with 12H8, 15A11, 1D10, 3F8, 1B8, 16B5 are depicted in the pictures. Sections from COMP^-/- mouse limbs stained with 12H8, 15A11, 1D10, 3F8, 1B8, 16B5 were used as negative controls. Anti-collagen type II (CII) monoclonal M2139 served as the positive control. The positive monoclonal staining on cartilage and tendon is indicated with an arrow.

Epitope mapping of 4 representative COMP-specific mAbs by using full-length cartilage oligomeric matrix protein (COMP) and recombinant fragments. B cell hybridomas were produced from COMP^+/+Ncf1^*/* or COMP^-/-Ncf1^*/* immunized with rat COMP. Monoclonal antibodies were obtained from culture supernatants after 14 days and tested by ELISA. Plates were individually coated with 10 μg/ml of recombinant full-length mouse COMP or COMP fragments. Serum samples from naïve mice served as negative controls and serum samples from COMP immunized mice were used as positive controls.

Immunohistochemical analysis of cartilage oligomeric matrix protein (COMP)-specific mAbs binding to joint cartilage in vivo. COMP specific-mAbs were purified, biotinylated and injected into neonatal COMP^+/+ mice (100 μg of antibody in 100 μl of PBS/mice) and limbs were collected after 24 h. Seven μm cryo-sections were stained with DAB using extravidin-peroxidase as the detection system. Anti-CII monoclonal antibody (M2139)-injected animals served as positive controls, anti-COMP mAb injected COMP^-/- mice were used as negative controls. Positive mAb binding to joint surfaces is marked by arrows.

Fine mapping of selected cartilage oligomeric matrix protein (COMP)-specific mAbs. To identify possible linear peptide epitopes for anti-COMP mAbs, synthetic peptides from rat COMP were produced. Each peptide spans amino acid residues differently in the rat and mouse sequence. Antibodies were obtained from culture supernatants after 14 days and tested by ELISA in wells individually coated with 10 μg/ml of synthetic peptide. Pooled sera from COMP-immunized mice were used as the positive control. Monoclonal Ab 15A11 bound peptide 6, which is derived from EGF repeat 4.

Cartilage oligomeric matrix protein (COMP)-specific mAbs mediate arthritis. (A) Accumulated incidence and severity of arthritis. Two-month-old naïve male QB F1 mice were injected intravenously (i.v.) with 9 mg of an equal combination of M2139 with a single COMP-mAb. All the mice received lipopolysaccharide (25 μg per mouse) intraperitoneally on day 5 and were scored for arthritis up to 16 days. The error bars in the severity graph indicate standard error of the mean (SEM), n = 6 per combination group. (B) Histology of tarsal joint sections. Paws of QB F1 mice on day 16 after the antibody transfer were collected, fixed, decalcified, sectioned and stained with hematoxylin/eosin (left panel) or toluidine blue (right panel). Top panel, M2139 plus PBS-injected control mice; Bottom panel, M2139 plus 15A11-injected mice. Note thickened synovial lining, inflammatory cellular infiltrate extending over the cartilage surface and erosions of cartilage and bone in animals injected with M2139 plus 15A11 (bottom). Results shown are representative of those obtained from three to four mice in each group. Infiltration cells, glycosaminoglycan loss and joint surface erosions are indicated with arrows. (C) Accumulated incidence and severity of arthritis. Two-month-old naïve male QB F1 mice were injected i.v. with 9 mg of an equal combination of 1B8, 1D10, 3F8, 15A11 and 16B5 anti-COMP monoclonal antibodies (filled circles). The same amount of a combination of isotype-matched control antibodies (L243 + G11) was injected into a separate group of animals (open circles). Results shown are pooled values from two similar experiments with balanced groups. The error bars in (A) and (B) indicate the SEM.

Relative level and kinetics of the antibody response to each truncated cartilage oligomeric matrix protein (COMP) fragment. (A) Graphical representation of the domain composition of the recombinant mouse COMP proteins used in this study. pCOMP: pentameric COMP; mCOMP: monomeric COMP; EGF: epidermal growth factor; TIII: thrombospondin type 3; CG: C-terminal globular domain. (B) Medium was collected from each transfected cell line and purified using Ni-NTA affinity resin followed by a concentration step using ion-exchange chromatography as described in Materials and methods, and 10 μl of the purified proteins were subjected to 4 to 10% gradient SDS-PAGE under reducing conditions. (C) Relative level of antibody response to different recombinant COMP fragments. Full length COMP (pCOMP) and COMP fragments (10 μg/ml in 100 μl) were tested by ELISA for COMP autoantibody response. Serum samples from COMP^+/+Ncf1*/*mice were taken at day 50 after immunization with rat COMP and diluted 1:1000 for the assay. Background levels were determined for each test using antigen, coated, but without primary antibodies in the wells. Results are expressed as absorbance reading at 405 nm and represent the mean and SEM of 13 mice. The lower binding to the TIII1-CG fragment compared to pCOMP, mCOMP or EGF1-TIII8 was confirmed to be statistically significant (P ≤ 0.05). (D) Kinetics of antibody response to rat COMP, full length mouse COMP (pCOMP) and mouse COMP fragments. Serum samples from COMP^+/+Ncf1*/* mice were taken at days 15, 30, 50 and 85 after immunization with rat COMP and were subjected to ELISA. Sera were diluted 1:1000 for the assays. Data represent the mean values from 13 mice.