Measurement of cartilage oligomeric matrix protein (COMP) in normal and diseased equine synovial fluids

Osteoarthritis and Cartilage (2001) 9, 119–127 © 2001 OsteoArthritis Research Society International doi:10.1053/joca.2000.0367, available online at http://www.idealibrary.com on

1063–4584/01/020119+09 $35.00/0

Measurement of cartilage oligomeric matrix protein (COMP) in normal and diseased equine synovial fluids K. Misumi*, V. Vilim†, P. D. Clegg‡, C. C. M. Thompson‡ and S. D. Carter‡ *Department of Veterinary Surgery, Kagoshima University, 21–24 Korimoto 1-chome, Kagoshima 890-0065, Japan †Institute of Rheumatology, Na slupi 4, 128 50 Praha 2, Czech Republic ‡Departments of Veterinary Pathology and Veterinary Clinical Science, University of Liverpool, PO Box 147, Liverpool L69 3BX, U.K. Summary Objective: This study was designed to assay cartilage oligomeric matrix protein (COMP) in equine synovial fluids and to compare the concentration in synovial fluids from normal horses with joint diseased horses. The relationship between the COMP degradation and the matrix metalloproteinase activity in synovial fluids was also investigated. Design: Using COMP antigen prepared from equine articular cartilage and murine monoclonal antibody (12C4) raised against human COMP, an inhibition ELISA was developed. COMP in equine synovial fluids from normal and diseased joints was quantified. Metalloproteinase activities were evaluated in the same synovial fluids by a gelatin degradation ELISA. COMP fragments were evaluated qualitatively by Western blotting. Results: The COMP inhibition ELISA was reliable at concentrations of equine COMP between 62.5 and 2000 ng/ml. COMP values in joint fluids in both aseptic and septic joint disease (19.7±15.3 and 16.1±11.2
g/ml, respectively) were significantly (P<0.001) lower than normal (53.2±29.0
g/ml). The molecular sizes of COMP on immunoblots were different between normal and diseased synovial fluids; more fragments were seen in diseased fluids. The aseptic (26.6±20.6%) and septic joint disease synovial fluids (36.1±37.5%) had significantly higher (P<0.02 and 0.002, respectively) gelatinolytic activities than normal (13.6±13.7%). There was a negative correlation (R= −0.31, P<0.002) between COMP level and gelatinase activity. Conclusions: We conclude that the fragment pattern and the absolute COMP concentration maybe useful for monitoring joint disease, and that COMP degradation in synovial fluids from progressed joint disease may be due to MMP gelatinolytic activity. © 2001 OsteoArthritis Research Society International Key words: COMP, ELISA, Horse, MMP.

organization of collagen fibrils through its C-terminal globular domains8. The use of COMP as a molecular marker for joint diseases by measurement of levels in synovial fluid9 and serum10 has been advocated. In studies of arthropathies, COMP has usually been measured using either monoclonal antibodies or polyclonal antisera. The available reagents appear to have different specificities. For example, one described monoclonal antibody appears to be specific for cartilage-derived COMP11, whereas polyclonal antisera can bind COMP from various cell sources including synovial fibroblasts11,12. This may be important, because COMP is not only cartilage-derived but is found widely in other tissues, including synovium13,14 and tendon15–17. Previous studies have demonstrated that COMP levels detected with polyclonal antisera in synovial fluids and serum were increased in the early stages of OA9,18,19 and rheumatoid arthritis (RA)20,21, and subsequently decreased during disease progression. However, while the use of particular antibodies may favor detection of COMP from one particular source, it is impossible to be confident that this is the case for all assay conditions and all tissue samples. Thus, detected COMP must be considered potentially to be derived from any source.

Introduction Cartilage oligomeric matrix protein (COMP), a macromolecule distributed abundantly through cartilage1,2, may have potential for predicting progression of cartilage destruction in diseases such as osteoarthritis (OA). Intact COMP is pentameric, with five globular domains connected to a central assembly domain by flexible strands3,4. Based on its structural characteristics, COMP belongs to the thrombospondin family4,5. While the precise role of COMP in cartilage is not known, it appears to have a structural function. COMP has been shown to mediate primary chondrocyte attachment to matrix component6. Also, since it was reported that mutations of the COMP gene cause pseudochondroplasia7, it has been proposed that COMP has a role in organizing matrix assembly in cartilage. In a recent study, it was suggested that COMP influences the

Received 17 August 1999; accepted 19 May 2000. Address correspondence to: Stuart D. Carter, Department of Veterinary Pathology and Veterinary Clinical Science, University of Liverpool, PO Box 147, Liverpool L69 3BX, U.K. Tel: +44-(0)151-794-4206; Fax: +44-(0)151-794-4219; E-mail: [email protected]

119

120 It has been demonstrated that various sizes of COMP fragments are contained in synovial fluid22 and cartilage extracts23. The fragmentation is supposed to be the result of degradation by proteinases, in which matrix metalloproteinases (MMPs) play an important role to maintain the normal turnover of cartilage matrix molecules. Previous clinical studies demonstrated that the release and activities of MMPs in synovial fluid increase in OA24,25 and RA26,27, and that there is a significant correlation between MMPs and matrix turnover in OA28. While the involvement of MMPs with COMP fragmentation has been suggested in vitro and ex vivo29, this area needs further investigation. In studies of joint disease, control samples from normal joints are often difficult to obtain in man. For this reason, it is valid to consider naturally occurring joint disease in other species in which joint samples such as synovial fluids are more readily available. Equine joint diseases are a serious veterinary problem and serve as models for human arthropathies. The most common type is osteoarthritis, which is radiographically detected in advanced disease as bone destruction or production (osteophytosis), joint space narrowing and deformity30, and is strikingly similar to the disease in man19,31. As in man, when radiographical lesions are demonstrated, the disease is in an advanced stage. Limited studies have been made to identify biochemical markers indicating equine joint destruction before radiographical alterations become evident. So far, glycosaminoglycans including keratan sulfate32–35 in synovial fluids and serum has been reported to have the potential to indicate ongoing joint destruction. However, as in human OA, such markers are not necessarily specific to joint disease, or can be affected by a range of joint diseases35. Thus, the information available for clinical practice is restricted and joint markers more specific to cartilage catabolism in OA should be investigated. This study was performed to develop an assay system for COMP in equine synovial fluids, using a monoclonal antibody to COMP, and to compare the concentration in synovial fluids from normal horses to those with joint diseases. In addition, the relationship between the COMP degradation and catabolic enzyme activities in synovial fluids was investigated.

Methods

K. Misumi et al.: COMP in equine synovial fluids N-ethylmaleimide. This solution was applied to a DEAE anion-exchange column (Mono Q HR5/5, Pharmacia). The material bound on the column was eluted with a linear gradient of 0–1 M sodium chloride in the same buffer. The protein concentration of the eluent was determined by the absorbance at 280 nm. Aliquots from the eluted fractions were analysed by sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE). Based on its mobility under reducing and nonreducing SDS-PAGE conditions36, the fractions eluted between 0.2–0.3 M sodium chloride were judged as the COMP-rich fractions, pooled, and dialysed against phosphate-buffered saline (PBS). The pooled solution was further analysed by SDS-PAGE and Western blotting with a monoclonal antibody to human COMP, and the protein concentration determined37. This preparation (150
g/ml) was subsequently used as an equine COMP standard.

ANTI-COMP ANTIBODY

A murine monoclonal antibody (12C4) has been raised to reduced and alkylated human COMP subunit as previously described36. This antibody binds to native oligomeric human COMP as well as its reduced subunit and cross-reacts with bovine COMP (unpublished data).

SDS-PAGE AND WESTERN BLOTTING

SDS-PAGE on 4–15% gradient gels was performed according to the protocol of Laemmli38. After the electrophoresis, the gels were stained with Coomassie brilliant blue R-250, or electrotransferred onto polyvinyl difluoride (PVDF) membranes in Tris-glycine buffer (25 mM Tris/HCl, 192 mM glycine, pH 8.3, 20% methanol) at 70 V overnight at 4°C. The membranes were blocked with 5% skimmed milk in Tris buffered saline, pH 7.4, including 0.05% Tween 20 (TBS/Tween). Antigenic COMP was detected with the murine monoclonal antibody (diluted 1:4000 in 2% skimmed milk in TBS/Tween). Positive binding to alkaline phosphatase conjugated goat anti-mouse IgG antibody (A3688, Sigma) diluted 1:10,000 in 2% skimmed milk in TBS/Tween was demonstrated by the development of phosphatase reaction using 5-bromo-4-chloro-3-indolyl phosphate nitro blue tetrazolium (BCIP/NBT, Sigma Fast/ B-5655) as a substrate.

PREPARATION OF EQUINE COMP

Fresh cartilage slices were obtained from normal femoropatellar joints of an adult horse, which was euthanazed for a reason other than orthopedic disease and had no history of locomotor diseases. The tissues were washed with distilled water and then kept frozen at −70°C until processed. COMP was isolated according to the method of Smith et al. 17 with some modifications. Briefly, 25 g wet weight of frozen cartilage slices was milled in liquid nitrogen to obtain a fine particulate. The powdered tissue was suspended in 10 volumes of 4 M guanidine hydrochloride in 50 mM sodium acetate buffer (pH 5.8) containing proteinase inhibitors (10 mM EDTA, 100 mM 6-aminohexanoic acid, 5 mM benzamidine-HCl, and 5 mM N-ethylmaleimide), homogenized, and extracted at 4°C for 24 h with gentle stirring. After centrifugation at 17 000 g at 4°C for 30 min, the supernatant was separated, filtered, concentrated, and dialysed against 7 M urea, 20 mM Tris-HCl, pH 8.0, containing, 5 mM EDTA, 1 mM benzamidine, and 2 mM

QUANTIFICATION OF ANTIGENIC COMP IN EQUINE SYNOVIAL FLUIDS WITH INHIBITION ELISA

Synovial fluid samples Equine synovial fluids (N=123) were provided by veterinary surgeons at the Large Animal Hospital, University of Liverpool. Normal synovial fluids (N=40) were sampled either from joints judged to be free of any articular pathology on clinical, radiological or arthroscopic examination, or from horses killed for medical problems unrelated to joint disease that had no articular abnormality in the sampled joints clinically and macroscopically at post mortem. Synovial fluids from horses with clinical joint problems were obtained either during diagnostic procedures or prior to surgery. These synovial fluids were sorted into two categories, namely aseptic joint disease (AJD, N=62) and septic joint disease (SJD, N=21). The AJD group consisted of OA (N=25), osteochondritis dissecans (OCD, N=16),

Osteoarthritis and Cartilage Vol. 9, No. 2 articular fractures (N=14), and synovitis (N=7). The sampled synovial fluids were centrifuged at 3500 g for 15 min at 4°C, the clear supernatant collected, aliquoted and stored at −70°C until assay.

COMP inhibition ELISA An inhibition ELISA to quantify antigenic COMP was designed according to the method reported previously36,39 with some modifications.

Plate coating Fifty microliters of purified horse COMP antigen, in a coating buffer (20 mM sodium carbonate, 20 mM sodium bicarbonate, 0.002% sodium azide, pH 10), was placed into each well of 96-well microtitre ELISA plates (Dynatek) at 0.6
g/ml, incubated for 2 h at room temperature and then overnight at 4°C.

121 in PBS) per well and incubated for 1 h at 37°C and overnight at 4°C in a humidified chamber. After washing with PBS, 100
l of synovial fluid diluted 1/40 in MMP buffer (50 mM Tris/HCl, 5 mM CaCl2, 0.05% Brij 35, pH 7.6) was added to each well and the plate incubated at 37°C for 2 h in a humidified chamber. Following three washes with PBS/Tween (0.05%), a rabbit anti-gelatin antisera diluted at 1:1000 in PBS/Tween was added and incubated at 37°C for 1 h in a humidified chamber. After removing the unbound antibodies by three washes with PBS/Tween, the bound primary antibodies were detected using anti-rabbit IgGalkaline phosphatase conjugate (1/20 000) and Sigma 104 phosphatase substrate, as a secondary antibody and substrate, respectively. The absorbance at 405 nm (A405) was read by ELISA reader. The amount of gelatin remaining on the plate following incubation with a test sample, which was obtained from the percentage reduction in A405 readings, was calculated according to the following formula: [(A405 of MMP buffer only)−(A405 of sample)] ×100/(A405 of MMP buffer only)

Inhibition step IMMUNOBLOTTING OF SYNOVIAL FLUIDS

Equine COMP standards were prepared to give a range of 8000–3.9 ng/ml by doubling dilutions in PBS, pH 7.0, containing 0.05% Tween 20 (PBS/Tween, pH 7.0). Synovial fluids were diluted 1/50 and 1/100 in the same buffer. Diluted standards and synovial fluids were mixed with monoclonal antibody 12C4 (ascites fluid diluted 1/40,000 in PBS/Tween, pH 7.0, 1% BSA) and incubated overnight at 4°C.

The synovial fluid samples for western blotting were selected based on both the COMP concentration and the gelatinolytic activity. The samples were pre-treated with hyaluronidase from bovine testis according to the method described by Neidhart et al. 22 prior to the assay. The tested samples were applied to 4–15% SDS-polyacrylamide gel. Following SDS-PAGE, the gels were transferred to blotting membranes to detect antigenic COMP fragments with 12C4.

ELISA Coated wells were washed three times (5 min each) with PBS, pH 7.0 and blocked with 100
l/well of 5% BSA in PBS, pH 7.0 for 1 h at room temperature. Blocked plates were washed three times (2 min each) with PBS/Tween, pH 7.0, and incubated with 100
l/well of the inhibition mixture (monoclonal antibody plus COMP standard or synovial fluid) for 1 h at room temperature and then for 1 h at 4°C. The plates were washed three times (2 min each) with PBS/Tween, pH 7.0 and incubated with 100
l/well of alkaline phosphatase conjugated goat anti-mouse IgG (A3688, Sigma) diluted 1/10,000 in PBS/Tween, pH 7.0, 1% BSA. After three 2-min washes with PBS/Tween, pH 7.0, 100
l of glycine buffer (0.1 M glycine, 1 mM MgCl2, 1 mM ZnCl2, pH 10.4) containing 1 mg/ml of p-nitrophenyl phosphate (104-0, Sigma) was placed into each well and the plates were left for color to develop. The absorbance at 405 nm was read after 1 h using an ELISA plate reader (Titertek Multiscan® PLUS). A semi-log standard graph, where log10 (concentration of COMP standards) was plotted against the readings, was constructed, and the concentration of antigenic COMP in the synovial fluids was calculated using the linear portion of the standard curve corresponding to the reliable range of COMP concentration in this assay.

STATISTICS

All quantitative data were expressed as mean (S.D.). The statistical differences in COMP concentration among some categories were analysed by factorial ANOVA, and Scheffe´ ’s method was used for simultaneous multiple comparisons. A P-value less than 0.05 was considered significant.

Results ANALYSIS OF EXTRACTED EQUINE CARTILAGE COMP

Fractions eluted from the DEAE column separation of the equine cartilage extract between 0.2–0.3 M sodium chloride contained abundant protein which migrated at high molecular weight (512 kDa) in 4–15% PAGE gel under non-reducing condition (Fig. 1). Under reducing conditions, a protein band with the estimated molecular weight (MW) of 101 kDa stained prominently with Coomassie blue (Fig. 1). Western blotting of equine COMP showed that both the non-reduced oligomeric and reduced monomeric COMP bands on the immunoblots were recognized by the monoclonal antibody to human COMP (12C4) (Fig. 2). Antibody 12C4 also identified fainter protein bands (MW 85, 69, 64, and 61 kDa) under reducing conditions (Fig. 2).

GELATIN DEGRADATION ELISA MEASUREMENT OF ANTIGENIC COMP IN SYNOVIAL FLUIDS WITH

The degradation ELISA (to quantify gelatinase MMP activity) was designed according to the protocol described by Clegg et al. 25 with some modifications. Briefly, 96-well ELISA plates were coated with 100
l of gelatin (0.5
g/ml

INHIBITION ELISA

To quantify COMP in equine synovial fluids, an inhibition ELISA was designed using the purified equine cartilage

122

K. Misumi et al.: COMP in equine synovial fluids

Fig. 1. Fractions eluted from the DEAE column separation of the equine cartilage extract between 0–1.0 M sodium chloride. COMP containing fractions (bar) pooled between 0.2–0.3 M sodium chloride and analysed by SDS-PAGE. On 4–15% gradient gel, COMP pentamer can be seen close to stacking level (estimated 512 kDa) under nonreducing condition. The monomer (101 kDa) was detected under reducing condition.

COMP as the coating antigen and standard (Fig. 3). Based on the inhibition curve produced by the standard COMP from equine cartilage, the reliability of the ELISA assay was confirmed at the concentration between 62.5 and 2000 ng/ ml. The linear portion of the curves, created by diluting equine synovial fluids between 1/3 and 1/192, was parallel to that of the standard curve derived from isolated horse cartilage COMP. The curve produced by synovial fluid from an osteoarthritic joint shifted to the right of the profile from normal synovial fluid. The results for COMP measurements in horse synovial fluids are shown in Fig. 4. The mean COMP value in the normal joint fluids was 53.2±29.0
g/ml, and this was significantly (P<0.001) higher than both of AJD and SJD joints (19.7±15.3 and 16.1±11.2
g/ml, respectively). There were no significant differences between the values for AJD and SJD. The COMP concentration in AJD joints were analysed further by sorting the samples according to the clinical diagnosis. The mean COMP values in OA, OCD and intraarticular fracture were higher than in synovitis and SJD, but no statistically significant differences were shown. RELATIONSHIP BETWEEN COMP CONCENTRATION AND GELATINOLYTIC ACTIVITY IN SYNOVIAL FLUIDS

We evaluated the gelatinase activity in 117 samples of synovial fluid tested in the COMP ELISA. The aseptic (26.6±20.6%) and septic synovial fluids (36.1±37.5%) had significantly higher (P<0.02 and P<0.002, respectively) gelatinolytic activities than normal joint fluids (13.6±13.7%), as shown in Fig. 5. As the percentage

Fig. 2. Western blot analysis of COMP antigen separated by SDS-PAGE with monoclonal antibody 12C4. Under non-reducing condition the COMP pentamer is demonstrated as the strong band at the start of the separating gel. Under reducing condition, an intense band consistent with the migration of COMP monomer and some fainter bands (85, 69, 64, and 61 kDa) are produced.

Osteoarthritis and Cartilage Vol. 9, No. 2

123

COMP concentration (µg/ml)

100

53.2 ± 29.0

(A)

75 * * 21.8 * 21.3 19.7 ± 19.2 ± 17.5 ± 12.7

50

25

* 16.1 * 11.7 ± 11.2 ± 6.8

Septic disease

Synovitis (n = 7)

Fracture (n = 14)

OCD (n = 16)

Fig. 3. The inhibition curve obtained using the standard COMP from equine cartilage, and the synovial fluids from normal and diseased joints. The linear portion is shown between 62.5 and 2000 ng of COMP standard/ml. The linear portion of the curves from equine synovial fluids was parallel to the standard curve. The curve of synovial fluid from diseased joints shifts to the left of the normal synovial fluid.

OA (n = 25)

Normal

0

IMMUNOREACTIVE COMP FRAGMENTS IN SYNOVIAL FLUIDS (FIG. 7)

To identify COMP fragments in synovial fluids, 12 samples each from normal and diseased joints were selected. These were selected for a range of COMP and gelatinase activities. In Western blots of both normal and diseased synovial fluids, some COMP fragments were seen. Fragments with molecular weights of less than 60 kDa were not found. The fragment sizes in normal synovial fluids were 520, 424, 330, 191, 62 kDa. For diseased samples these were 516, 421, 280, 177, 136, 101, and 67 kDa. The relative proportion of the fragments were clearly different between normal and diseased synovial fluids. While in the normal samples the highmolecular-weight molecules (the size of intact COMP) were apparent, they were less readily seen in diseased synovial fluids. On the other hand, in diseased samples there were more bands, which were prominent, apparent in the range of 100 to 250 kDa. In septic samples, particularly, the high-molecular-weight COMP (516 kDa) was not seen; only fragments were detected.

Discussion This work identified COMP in chromatographic fractions of guanidine hydrochloride extracts of equine articular cartilage. Identity was based on electrophoretic mobility in SDS-PAGE and the specificity of a murine monoclonal antibody to human COMP. Pentameric equine COMP was seen as a 512 kDa band in SDS-PAGE under non-reducing conditions, while reduced COMP had a molecular mass of

COMP concentration (µg/ml)

100

degradation of gelatin in AJD joints was sorted further according to the clinical diagnosis, the averages in OCD and synovitis were higher than others but there were no statistically significant differences. Between the COMP concentration and the percentage degradation of gelatin (converted logarithmic scale) in the samples, a negative correlation (R= −0.31) was seen, which was significant (P<0.002) (Fig. 6).

(B)

53.2 ± 29.0 75

50

* 19.7 ± 15.3

* 16.1 ± 11.2

25

0

Normal synovial fluid (n = 40)

Aseptic joint disease synovial fluid (n = 62)

Septic joint disease synovial fluid (n = 21)

Fig. 4. COMP concentration (mean± S.D.) in horse synovial fluids. The COMP concentrations in aseptic joint disease fluids were analysed further by sorting the samples according to the clinical diagnosis (A). The COMP value in the normal joint fluids is significantly higher than either aseptic or septic joints (B). * shows statistical difference (P<0.001) vs normal synovial fluids. OA; osteoarthritis, OCD; osteochondritis dissecans.

101 kDa. These sizes are consistent with those identified for COMP in a previous study in other equine tissues, including tendon, ligament, scarred skin and granulation tissue17. A murine anti-human COMP monoclonal antibody (12C4) recognized both non-reduced and reduced COMP protein. The sizes of some smaller and fainter reduced bands detected by 12C4 are similar to previously reported degradation products of purified COMP22. The concentration of COMP, measured by inhibition ELISA, was significantly lower in synovial fluids from diseased joints compared with normal joints. However, no significance was shown for the differences between septic and aseptic joint disease synovial fluids. Thus, while decreased synovial fluid COMP concentrations may indicate pathological changes, it is not possible to differentiate

124

K. Misumi et al.: COMP in equine synovial fluids y = –19.5LOG(x) + 54.6

100 36.1* 75

40.0 ± 20.4

32.5 ± 24.2 25.2 ± 14.7

19.2 ± 18.8

50

COMP concentration (µg/ml)

Degradation of gelatin (%)

(A)

13.6 25

Septic disease

Synovitis (n = 7)

Fracture (n = 13)

OCD (n = 16)

OA (n = 23)

Normal

0

50

1

10

100

Degradation of gelatin (%)

(B)

Fig. 6. COMP concentration and percentage degradation of gelatin in equine synovial fluids. A negative correlation is shown (N=117, R= −0.31, P<0.002).

36.1 ± 37.5*

75 †

26.6 ± 20.6

50 13.6 ± 13.7 25

0

100

0 0.1

100 Degradation of gelatin (%)

r = –0.31 (P < 0.002)

150

Normal synovial fluid (n = 40)

Aseptic joint disease synovial fluid (n = 59)

Septic joint disease synovial fluid (n = 18)

Fig. 5. Degradation (%) of gelatin (mean± S.D.) in horse synovial fluids. The percentage in AJD joints were analysed further by sorting the samples according to the clinical diagnosis (A). The percentage gelatin degradation in either aseptic or septic joints was significantly higher than the normal joints (B). Statistical difference was seen (P<0.002* and 0.02†, respectively) vs normal synovial fluids. OA; osteoarthritis, OCD; osteochondritis dissecans.

between different disease conditions. Immunoblotting did identify qualitative differences between breakdown product patterns in aseptic and septic joint disease fluids. However, these differences were not consistent within or between the aseptic and septic groups. In previous studies relating the progression of joint disease with COMP levels in synovial fluid31,40 or serum18,21,41,42, the COMP level in synovial fluid and serum from patients with joint diseases has been shown to be higher than in healthy joints and normal sera, respectively. It has also been suggested that COMP in synovial fluids increases in the early stage of joint disease and then decreases in patients with advanced joint damage9,39. The discrepancy between previous reports and our results may be dependent on the ELISA protocol, particularly in the

choice of primary antibodies used to detect COMP. While in previous clinical studies COMP has been measured using predominantly rabbit polyclonal antisera, we used a murine monoclonal antibody 12C4 for the assay. It has been reported that there is a difference in molecular weight of COMP-derived antigens recognized by polyclonal and some particular monoclonal antibodies in human synovial fluid11. For example, in synovial fluid from a healthy person, polyclonal antibody recognized low-molecular-weight fragments as well as oligomeric COMP, while a monoclonal antibody (18G3) showed a high affinity for oligomeric forms of COMP exclusively, with no binding of low-molecularweight COMP fragments11. Similarly, on Western blots of nonreduced healthy synovial fluids22, polyclonal antibody detected predominantly smear fractions (100–150 kDa) and low-molecular-weight fragments (70–100 kDa) but no oligomers. Interestingly, our immunoblot of normal synovial fluids from equine joints showed that the murine monoclonal anti-COMP antibody (12C4) can recognize intact COMP and partially degraded COMP oligomer, but not the very low-molecular-weight fragments. This antibody also detected, in samples from diseased joints, degraded lowmolecular-weight fragments, while high molecular weight bands were fainter or not detected. Therefore, we propose that the COMP concentration measured using ELISA with monoclonal antibody 12C4 predominantly reflects the level of intact or less degraded high-molecular-weight COMP antigen in synovial fluids, and does not detect small fragments of COMP which lose their antigenicity as COMP degradation progresses to a certain level. This is consistent with the fact that the epitope recognized by monoclonal antibody 12C4 is located near the C-terminus of COMP subunit, i.e. either in a C-terminal globular domain or in a domain of calmodulin-like repeats and, hence, can be removed by proteolytic cleavage of the COMP subunit which would trim the molecule from its C-terminus. An alternative explanation may result from differences in antibody specificity. Hummel et al. who used another

Osteoarthritis and Cartilage Vol. 9, No. 2

125

Fig. 7. Immunoblots of synovial fluids from normal and diseased joints. The similar sizes of COMP fragments are found in both of normal and diseased samples. While the high-molecular-weight fragments (>400 kDa) predominated in normal, more degraded fragments at the range of 100 to 250 kDa were shown in diseased samples. Interestingly, one sample (lane 8) of normal synovial fluids shows similar fragment pattern to the diseased sample (lane 1). Lane 13 in blots of normal synovial fluids; COMP standard extracted from horse cartilage. Figures for COMP concentration were determined using the inhibition ELISA; percentages gelatin digestion were measured by gelatin degradation ELISA.

monoclonal antibody to human COMP (18G3), have suggested that particular monoclonal antibodies may only detect cartilage-derived COMP and may be unable to detect COMP from other sources11. Additionally, polyclonal antibodies bind synovial fibroblast- as well as chondrocytederived COMP11,12,14. While we have no data as to the antigenicity of different tissue-specific COMPs to our antibody, it is plausible that 12C4 detects predominantly cartilage-derived COMP.

Increases in the activities of several MMPs, including MMP-1, 3, 9, and 13, have been suggested as a potential causes of COMP degradation29. In particular, MMP-9 can efficiently degrade cartilage-derived COMP into more fragments than other MMPs29. This study showed significantly higher gelatinolytic activities in diseased, compared to normal, joint fluids. A slight negative regression (R-squared=0.31; P<0.002) between COMP concentration and gelatinase activity suggests that increased digestion by

126 one or more gelatinases may account for reduced detection of COMP. In conclusion, our findings indicate that the inhibition ELISA for equine COMP, in its present form, is a potentially valuable tool in the study and possibly the diagnosis of equine joint disease. There are clearly issues which remain to be resolved. In particular, it will be important to determine the specificity of our antibody to COMP from different sources. This will help lead to an explanation for the differences between our findings of a reduction in COMP in disease and those of others who identified an increase. Finally, although we have shown a negative correlation between COMP and gelatinase activity, further and extensive studies will be required to establish or reject the hypothesis that gelatinases are actually responsible for COMP degradation within cartilage.

K. Misumi et al.: COMP in equine synovial fluids

10.

11.

12.

13.

Acknowledgments We would like to thank Dr E. J-M. A. Thonar and Dr M. E. Lenz for recommendation and assistance in establishing this collaboration, and Dr J. Crean, Dr A. Bee, Dr I. Demirkan and Mr F. Al-Hizab for expert assistance. This study was financially supported by The Home of Rest for Horses. VV was supported by a grant no. 4887-3 from IGA MZ, Czech Republic. KM was supported by Sakamoto & Takemoto Funds for Research, Japan.

14.

15.

16.

References 1. Hedbom E, Antonsson P, Hjerpe A, Aeschlimann D, Paulsson M, Rosa-Pimental E, et al. Cartilage matrix proteins. An acidic oligomeric protein (COMP) detected only in cartilage. J Biol Chem 1992;267: 6132–6. 2. DiCesare PE, Mo¨ rgelin M, Carlson CS, Pasumarti S, Paulsson M. Cartilage oligomeric matrix protein: isolation and characterization from human articular cartilage. J Orthop Res 1995;13:422–8. 3. Mo¨ rgelin M, Heinega˚ rd D, Engel J, Paulsson M. Electron microscopy of native cartilage oligomeric matrix protein purified from the swarm rat chondrosarcoma reveals a five-armed structure. J Biol Chem 1992;267:6137–41. 4. DiCesare PE, Mo¨ rgelin M, Mann K, Paulsson M. Cartilage oligomeric matrix protein and thrombospondin 1. Purification from articular cartilage, electron microscopic structure, and chondrocyte binding. Eur J Biochem 1994;223:927–37. 5. Oldberg A r , Antonsson P, Lindblom K, Heinega˚ rd D. COMP (cartilage oligomeric matrix protein) is structurally related to the thrombospondins. J Biol Chem 1992;267:22346–50. 6. DiCesare PE, Paulsson M. COMP mediates chondrocyte attachment. Experientia 1992;40:A60. 7. Hecht JT, Nelson LD, Crowder E, Wang Y, Elder FF, Harrison WR, et al. Mutations in exon 17B of cartilage oligomeric matrix protein (COMP) cause pseudoachondroplasia. Nat Genet 1995;10:325–9. 8. Rosenberg K, Olsson H, Mo¨ rgelin M, Heinega˚ rd D. Cartilage oligomeric matrix protein shows high affinity zinc-dependent interaction with triple helical collagen. J Biol Chem 1998;273:20397–403. 9. Lohmander LS, Saxne T, Heinega˚ rd DK. Release of cartilage oligomeric matrix protein (COMP) into joint

17.

18.

19.

20.

21.

22.

23.

24.

fluid after knee injury and in osteoarthritis. Ann Rheum Dis 1994;53:8–13. Clark AG, Jordan JM, Vilim V, Renner JB, Dragomir AD, Luta G, et al. Serum cartilage oligomeric matrix protein reflects osteoarthritis presence and severity: the Johnston County Osteoarthritis Project. J Biol Chem 1998;273:20397–403. Hummel KM, Neidhart M, Vilim V, Hauser N, Aicher WK, Gay RE, et al. Analysis of cartilage oligomeric matrix protein (COMP) in synovial fibroblasts and synovial fluids. Br J Rheum 1998;37:721–8. Dodge GR, Hawkins D, Boesler E, Sakai L, Jimenez SA. Production of cartilage matrix protein (COMP) by cultured human dermal and synovial fibroblasts. Osteoarthritis Cart 1998;6:435–40. DiCesare PE, Carlson CS, Stollerman ES, Chen FS, Leslie M, Perris R. Expression of cartilage oligomeric matrix protein by human synovium. FEBS Lett 1997;412:249–52. Recklies AD, Baillargeon L, White C. Regulation of cartilage oligomeric matrix protein synthesis in human synovial cells and articular chondrocytes. Arthritis Rheum 1998;41:997–1006. DiCesare P, Hauser N, Lehman D, Pasumarti S, Paulsson M. Cartilage oligomeric matrix protein (COMP) is an abundant component of tendon. FEBS Lett 1994;354:237–40. Hauser N, Paulsson M, Kale AA, DiCesare PE. Tendon extracellular matrix contains pentameric thrombospondin-4 (TSP-4). FEBS Lett 1995;368: 307–10. Smith RKW, Zunino L, Webbon PM, Heinega˚ rd D. The distribution of cartilage oligomeric matrix protein (COMP) in tendon and its variation with tendon site, age, and load. Matrix Biol 1997;16:255–71. Sharif M, Saxne T, Shepstone L, Kirwan JR, Elson CJ, Heinega˚ rd D, Dieppe PA. Relationship between serum cartilage oligomeric matrix protein levels and disease progression in osteoarthritis of the knee joint. Br J Rheum 1995;34:306–10. Petersson IF, Boegard T, Dahlstrom J, Svensson B, Heinga˚ rd D, Saxne T. Bone scan and serum markers of bone and cartilage in patients with knee pain and osteoarthritis. Osteoarthritis Cart 1998;6:33–9. Saxne T, Heinega˚ rd D. Cartilage oligomeric matrix protein: a novel marker of cartilage turnover detectable in synovial fluid and blood. Br J Rheum 1992; 31:583–91. Forslind K, Eberhardt K, Jonsson A, Saxne T. Increased serum concentrations of cartilage oligomeric matrix protein. A prognostic marker in early rheumatoid arthritis. Br J Rheum 1992;31:593–8. Neidhart M, Hauser N, Paulsson M, DiCesare PE, Michel BA, Ha¨ uselmann HJ. Small fragments of cartilage oligomeric matrix protein in synovial fluid and serum as markers for cartilage degeneration. Br J Rheum 1997;36:1151–60. DiCesare PE, Carlson CS, Stolerman ES, Hauser N, Tulli H, Paulsson M. Increased degradation and altered tissue distribution of cartilage oligomeric matrix protein in human rheumatoid and osteoarthritic cartilage. J Orthop Res 1996;14:946–55. Lohmander LS, Hoerrner LA, Lark MW. Metalloproteinases, tissue inhibitor, and proteoglycan fragments in knee synovial fluid in human osteoarthritis. Arthritis Rheum 1993;36:181–9.

Osteoarthritis and Cartilage Vol. 9, No. 2 25. Clegg PD, Coughlan AR, Riggs CM, Carter SD. Matrix metalloproteinases 2 and 9 in equine synovial fluids. Equine Vet J 1997;29:343–8. 26. Hirose T, Reife RA, Smith GN, Stevens RM, Mainardi CL, Hasty KA. Characterization of type V collagenase (gelatinase) in synovial fluid of patients with inflammatory arthritis. J Rheumatol 1992;19:593–9. 27. Koolwijk P, Miltenburg AMM, van Erck MGM, Oudshoom M, Neidbala MJ, Breedveld FC, et al. Activated gelatinase-B (MMP-9) and urokinase-type plasminogen activator in synovial fluids of patients with arthritis. Correlation with clinical and experimental variables of inflammation. J Rheumatol 1995;22: 385–93. 28. Lohmander LS, Ionescu M, Jugessur H, Poole AR. Changes in joint cartilage aggrecan after knee injury and in osteoarthritis. Arthritis Rheum 1999;42: 534–44. 29. Ganu V, Goldberg R, Peppard J, Rediske J, Melton R, Hu SI, et al. Inhibition of interleukin-1alpha-induced cartilage oligomeric matrix protein degradation in bovine articular cartilage by matrix metalloproteinase inhibitors. Potential role for matrix metalloproteinases in the generation of cartilage oligomeric matrix protein fragments in arthritic synovial fluid. Arthritis Rheum 1998;41:2143–51. 30. Park RD, Steyn PF, Wrigley RH. Imaging techniques in the diagnosis of equine joint disease. In: Mcllwraith CW, Trotter CW, Eds. Joint Disease in the Horse. Philadelphia, PA: W.B. Saunders 1996:145–64. 31. Petersson I, Sandqvist L, Svensson B, Saxne T. Cartilage markers in synovial fluid in symptomatic knee osteoarthritis. Ann Rheum Dis 1997;56:64–7. 32. Alwan WH, Carter SD, Bennett D, Edwards GB. Glycosaminoglycans in horses with osteoarthritis. Equine Vet J 1991;23:44–7. 33. Fuller CJ, Barr ARS, Dieppe PA, Sharif M. Variation of an epitope of keratan sulphate and total gly-

127

34.

35.

36.

37. 38. 39.

40.

41.

42.

cosaminoglycans in normal equine joints. Equine Vet J 1996;28:490–3. Okumura M, Fujinaga T, Urakawa E, Tagami M, Tsukiyama K. Evaluation of the catabolic activity of cartilage by measurement of serum keratan sulphate concentration in foals. Am J Vet Res 1997;58:925–9. Todhunter RJ, Fubini SL, Freeman KP, Lust G. Concentrations of keratan sulphate in plasma and synovial fluid from clinically normal horses and horses with joint disease. J Am Vet Med Assoc 1997;210:369–74. Vilim V, Lenz ME, Vytasek R, Masuda K, Pavelka K, Kuittner KE, et al. Characterization of monoclonal antibodies recognising different fragments of cartilage oligomeric matrix protein in human body fluids. Arch Biochem Biophys 1997;341:8–16. Lowry OH, Rosebrough NJ, Farr AL, Randall RJ. Protein measurement with the folin phenol reagent. J Biol Chem 1951;193:265–75. Laemmli UK. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 1970;227:680–5. Thonar EJ-MA, Lenz ME, Klintworth GK, Caterson B, Pachman LM, Glickman P, et al. Quantification of keratan sulphate in blood as a marker of cartilage catabolism. Arthritis Rheum 1985;28:1367–76. Saxne T, Heinega˚ rd D. Matrix proteins: potentials as body fluid markers od changes in the metabolism of cartilage and bone in arthritis. J Rheumatol 1995;22(suppl 43):71–4. Ma˚ nsson B, Geborek B, Saxne T. Cartilage and bone macromolecules in knee joint synovial fluid in rheumatoid arthritis: relation to development of knee or hip joint destruction. Ann Rheum Dis 1997;56:91–6. Georges C, Vigneron H, Ayral X. Serum biologic markers as predictors of disease progression in osteoarthritis of the knee. Arthritis Rheum 1997;40: 590–1.