E. Larsson, H. Erlandsson Harris, J. C. Lorentzen, A. Larsson 1, B. Månsson 2, L. Klareskog and T. Saxne 2

Rheumatology 2002;41:996 1000 Serum concentrations of cartilage oligomeric matrix protein, fibrinogen and hyaluronan distinguish inflammation and cartilage destruction in experimental arthritis in rats E. Larsson, H. Erlandsson Harris, J. C. Lorentzen, A. Larsson 1, B. Månsson 2, L. Klareskog and T. Saxne 2 Department of Medicine, Rheumatology Unit, Karolinska Hospital, Stockholm, 1 Department of Clinical Chemistry, University Hospital Uppsala and 2 Department of Rheumatology and Department of Cell and Molecular Biology, Lund University, Sweden Abstract Objectives. We investigated if changes in serumuplasma fibrinogen (FIB), hyaluronan (HA) and cartilage oligomeric matrix protein (COMP) levels can be used to differentiate between inflammation and cartilage involvement during arthritis. Methods. Collagen-induced arthritis (CIA), oil-induced arthritis (OIA) and for comparison, experimental autoimmune encephalitis (EAE) induced in DA rats were investigated. Results. Elevations of FIB concentrations were apparent at days 4 7 post-immunization in both arthritis models reaching a maximum on day 20 21, i.e. before peak arthritis. Elevations of HA in both models were seen shortly before macroscopically apparent arthritis, and peaked at or just before maximal arthritis, i.e. later in CIA than in OIA. COMP levels increased only after onset of arthritis and peaked late in disease (days 34 37), being significantly higher in the more destructive CIA compared with the less destructive OIA. During EAE flares, only FIB levels increased. Conclusions. FIB is a general inflammation marker, HA appears to be a marker for synovitis and changes in COMP levels appear to reflect the cartilage destruction process. KEY WORDS: Collagen-induced arthritis, Oil-induced arthritis, Experimental autoimmune encephalomyelitis, Cartilage oligomeric matrix protein, Fibrinogen, Hyaluronan. Polyarthritis may vary in chronicity and severity. The magnitude of the overall inflammatory response, the intensity of local synovitis and the extent of cartilage destruction may vary independently of each other. For example, cartilage and bone destruction in equally inflamed joints is less pronounced in SLE as compared with rheumatoid arthritis (RA) w1x. Destruction of joint cartilage in RA may also occur in patients with little clinical sign of synovitis w2x. For evaluation of disease development and activity, for example when studying the effects of anti-arthritic treatment regimens, signs of inflammation as determined by quantification of serum levels of acute phase proteins are used. The magnitude of ongoing synovitis and cartilage and bone destruction is difficult to Submitted 19 March 2001; revised version accepted 13 March 2002. Correspondence to: E. Larsson, Rheumatology Unit, Department of Medicine, Karolinska Hospital, S-171 76 Stockholm, Sweden. estimate. Indirect measures such as physical examination and radiography are used. Easily accessible serum markers for monitoring synovitis and joint destruction would represent valuable tools for evaluating arthritis development and treatment. Several such potential markers are being investigated with promising results w3x. Cartilage oligomeric matrix protein (COMP) is a pentameric protein, originally purified from cartilage w4x. It has also been shown to be present in other pressure loaded tissues, e.g. tendon w5, 6x, and meniscus w7x (B. Månsson et al., unpublished results). COMP can also be produced by cells in the synovial membrane w8, 9x, and the relative amounts of the protein in different tissues vary, being highest in cartilage (T. Saxne and D. Heinegård, unpublished results). Both in inflammatory arthritis and osteoarthritis, as well as in experimental arthritis, COMP has shown promise as a potential biomarker for monitoring progression of cartilage destruction, for evaluating cartilage effects of 996 ß 2002 British Society for Rheumatology

Levels of COMP, fibrinogen and hyaluronan in arthritis 997 therapy, and as a prognostic tool reflecting cartilage damage wreviewed in 10x. Although being a promising cartilage marker, the lack of tissue specificity is a potential confounding factor for interpretations of changes in serum levels, especially in conditions with a marked inflammatory response in the synovium. It has previously been demonstrated that serum COMP increases during arthritis development in collagen II-induced arthritis (CIA) and pristane-induced arthritis in rats w10, 11x, and that this increase coincides with development of cartilage damage. Furthermore, therapeutic intervention which ameliorated cartilage destruction normalized serum COMP levels in murine CIA, whereas treatment which only reduced signs of inflammation did not affect COMP levels w12x. In the present study, we wanted to analyse further the relationship between changes in serum COMP concentrations and the process in cartilage and the inflammatory process both locally and systemically. To accomplish this we studied changes in serum concentrations of COMP as a potential cartilage marker, hyaluronan (HA) as a putative marker for synovial inflammation w13, 14x and fibrinogen (FIB) as a marker for generalized inflammation w15x in the rat. Fibrinogen was chosen as the marker of preference for general inflammation in the rat, as serum levels of FIB have shown a more dynamic response to inflammation than C-reactive protein (CRP), and as there were no available tests for amyloid A w15x. HA was chosen as a the marker of preference for synovitis on the basis of previous rat studies w13x as well as on human studies in RA where serum levels of HA correlated to synovitis as estimated by the Ritchie articular index w14x. We utilized two models of arthritis with different disease courses (CIA w16, 17x and oil-induced arthritis (OIA) w18x), and experimental autoimmune encephalitis (EAE) as a non-arthritic inflammatory control. CIA in the DA rat is a chronic and heavily destructive disease w17x, while OIA in the same rat strain is transient and has a less destructive disease course w18x. The serum concentrations of the different markers were related to clinical signs of disease. Materials and methods Animals Male DA rats aged 3.5 4 months at the start of the experiments were used. The animals were health monitored according to guidelines from the Swedish Veterinary Board, and found to be free of pathogens. The Ethical Board (Djurförsöks etisk nämnd) at the Karolinska Institute, Stockholm, approved all animal procedures performed. Induction and clinical monitoring of experimental diseases For CIA, collagen II was prepared from rat chondrosarcoma as previously described w19, 20x. The collagen was dissolved in 0.1 M acetic acid and emulsified 1:1 with Freund s incomplete adjuvant (FIA) (Difco, Detroit, MI, USA). A 150 mg quantity of collagen II in 200 ml emulsion, was injected intradermally at the base of the tail. OIA was induced by injection of 200 ml of FIA intradermally at the base of the tail. EAE was induced by intradermal injection at the base of the tail of 200 ml with homogenized DA rat spinal cord emulsified 1:1 in FIA. Arthritis was quantified by a clinical scoring system, scaled 0 16. Each paw was scored as follows: 0=no arthritis, 1=swelling in one type of joint, 2=swelling in two types of joints, 3=swelling in three types of joints and 4=swelling of the entire paw. A total score for an animal was calculated by summing up the scores for each of the four paws w18x. EAE was evaluated using a clinical scoring system scaled 0 3 where 0=no illness, 1=dropping tail, 2=unsteady walk and 3=inability to walk w21x. Blood samples were taken by retroorbital puncture before immunization and at selected time points after immunization. Immunoassays Serum concentrations of COMP were determined by ELISA, using similar conditions as described for the assay for human COMP w22x. The assay was modified by using rat COMP for coating microtitre plates and for the standard curve included in each plate as well as by using a polyclonal antiserum raised against rat COMP w10x. Plasma levels of FIB were measured with nephelometry as previously described by Larsson et al. w15x. Results are presented as per cents of a reference sample consisting of pooled plasma from healthy rats. Hyaluronan was analysed using a previously described radiometric technique according to the manufacturer s instructions (Pharmacia HA test, Pharmacia Diagnostica, Uppsala, Sweden) w23x. The feasibility of the assay technique for rat serum samples has previously been documented w13x. Statistical calculations Wilcoxon s matched pairs test (two-tailed) was used for comparing concentrations of FIB, HA, COMP and scores at different time points. The Mann Whitney U-test was used for comparing differences between groups. A P-value <0.05 was considered significant. Only animals developing disease after immunization were included in the calculations. Results Development of disease The disease scores are presented in Table 1 and Fig. 1A. In the CIA group 58% (7u12) and in the OIA group 80% (8u10) exhibited clinical signs of arthritis. Ninety per cent (9u10) of immunized animals in the EAE group exhibited clinical signs of encephalitis (and no signs of arthritis).

998 E. Larsson et al. TABLE 1. Clinical scores, levels of plasma FIB, serum HA levels and serum COMP levels in CIA and OIA wmedian (lower-upper quartile range)x Days p.i. Score FIB (values in % of reference sample) HA (nguml) COMP (mguml) CIA 0 0 109 (89 131) 97 (67 119) 2.6 (2.1 3) 7 0 194 (142 234) a 83 (80 123) 2.2 (1.8 3) 16 0 224 (158 341) a 159 (136 211) a 2.0 (1.6 3) 19 5 (4 6) a nd nd nd 21 6 (5 6) a 334 (280 436) a 480 (320 636) a 4.9 (3.9 6) a 27 8 (6 10) a nd nd nd 34 9 (6 10) a 150 (106 66) 536 (399 635) a 9 (8.6 10) a 42 8 (5 9) a 217 (211 268) a 341 (176 414) a 8.0 (6.9 9) a 56 6 (4 10) a 191 (46 233) 141 (117 213) 7.5 (6.2 8) a 68 5 (4 9) a 119 (62 156) 118 (91 217) 4.6 (4 5) a 77 6 (5 6) a 57 (28 75) 151 (120 232) 3.5 (3 4) a 93 6.5 (6 9) a 94 (54 170) 118 (112 143) 2.9 (2.3 3) 113 6 (5 6) a 95 (53 97) 90 (68 121) 2.2 (1.8 2.5) OIA 0 0 112 (61 153) 74 (74 88) 3.3 (2.7 3.7) 4 0 256 (201 285) a 97 (64 130) 3 (2.4 3.2) 8 0 128 (122 220) 101 (89 128) a 2.6 (2.2 2.8) a 12 0 276 (207 339) a 240 (148 423) a 2.6 (2.2 2.9) a 16 0 (0 4) 319 (160 339) a 283 (152 360) a 2.8 (2.5 3) 19 4 (1.5 7) a nd nd nd 20 6 (0 6.5) a 384 (240 411) a 433 (319 535) a 3.4 (2.8 4.2) 25 7 (4 7.5) a 243 (169 326) 111 (73 167) 5 (4.2 6.8) a 37 3.5 (0 4) 139 (112 237) 89 (59 148) 5.4 (4.7 6.0) a 46 1 (0 3) 124 (84 159) 92 (73 123) 3.8 (2.8 4.5) 54 0 164 (151 186) 63 (53 68) 3.8 (2.8 4.5) 68 0 102 (82 143) 80 (46 91) 2.4 (2.2 2.6) a 82 0 173 (120 197) 58 (48 81) 2.8 (2.4 3.1) a a P-values <0.05 as calculated by Wilcoxon matched pairs test (value compared with baseline value). nd, not done. The onset of disease demonstrated by the arthritis score occurred between days 13 and 19 post-immunization (p.i.) in CIA as well as in OIA. For CIA, the maximum arthritis score was found between days 27 and 34 p.i. The arthritis score remained elevated during the whole observation period (P<0.05 vs baseline day 16 113 p.i.). OIA was most pronounced at day 25 p.i. (P<0.05 vs baseline day 17 28 p.i.), hereafter the rats gradually improved and clinical signs of arthritis had disappeared completely at day 54 p.i. in all animals. The onset of EAE occurred between days 8 and 20 p.i. The rats were very disabled and were killed for ethical reasons at different time points after day 28 p.i. Serum concentrations of FIB, HA and COMP in CIA Serum concentrations of FIB were increased at day 7 p.i. and peaked at day 21 p.i. (P<0.05 vs baseline at the respective time points). The levels of FIB thereafter decreased, and were down to baseline after day 42 p.i. in both models. Elevations of HA levels appeared shortly before macroscopically apparent arthritis, and peaked at maximal arthritis, i.e. day 34 p.i. and decreased rather rapidly thereafter. In contrast, COMP levels started to increase after arthritis onset, i.e. at day 20 p.i., peaked at day 34 42 p.i. and remained elevated until day 77 p.i. (P<0.05 vs baseline at the respective time points) (Table 1 and Fig. 1B D). Serum concentrations of FIB, HA and COMP in OIA Serum levels of FIB were increased at day 4 p.i., peaked at day 20 p.i. (P<0.05 vs baseline), and then rapidly declined (P>0.05 vs baseline at day 25 p.i.). Levels of HA were increased before arthritis onset (days 8 and 12 p.i.), peaked at day 20, and then decreased rapidly. COMP levels were seen only after onset of arthritis (day 25 p.i.) and peaked at day 37 p.i. (P<0.05 vs baseline at the respective time points). At day 46 p.i. COMP had returned to baseline levels. The COMP increase was less pronounced (P<0.002 for peak values) and less prolonged in OIA as compared with CIA (Table 1 and Fig. 1B D). Serum concentrations of FIB, HA and COMP during development of EAE Serum concentrations of FIB increased and reached two peaks in EAE, at day 4 p.i. (P<0.05 vs baseline) and day 25 p.i. (P<0.05 vs baseline) Notably, the first peak was also observed in the animal that did not develop disease. Levels of HA and COMP did not increase significantly during the observation period (Fig. 1B D). Discussion The main findings in the present study are that serumuplasma levels of FIB, HA and COMP show different patterns of changes in CIA, a chronic and

Levels of COMP, fibrinogen and hyaluronan in arthritis 999 FIG. 1. (A) Disease scores for CIA, OIA and EAE. The maximal score in arthritic models is 16 points whereas the maximal score in EAE is 3 points. (B) Plasma levels of FIB in CIA, OIA and EAE. (C) Serum levels of HA in CIA, OIA and EAE. (D) Serum levels of COMP in CIA, OIA and EAE. All values in the figure are medians. severely destructive arthritis, in OIA, a transient and less destructive arthritis and in EAE, a demyelinating encephalomyelitis. In the arthritis models, FIB and HA increased before onset of clinical disease. This indicates that levels of these markers reflect inflammation. However, no significant change in HA levels could be detected in EAE. Thus, HA did not seem to reflect inflammation per se but rather inflammation in relation to arthritis. This is further emphasized by the observation that HA levels declined more rapidly in the transient OIA as compared with the more long-standing CIA. In contrast, COMP levels increased after onset of clinical arthritis. This observation indicates that the COMP levels reflect cartilage involvement. COMP did not only seem to reflect cartilage involvement but possibly also the extent of the involvement because the increase was more pronounced and more extended in time in the more chronic and more destructive CIA as compared with the transient, less destructive OIA. In conclusion, this study, which has investigated potential markers for inflammation, synovitis, and cartilage involvement in experimental arthritis, provides support for the discriminative value of these markers. Thus, FIB is a marker of inflammation in both the arthritis models and in EAE. We suggest that HA could be a marker preferentially reflecting local inflammation in the joint, i.e. synovitis. Taken together the experiments indicate that changes in serum COMP concentrations reflect the cartilage process. Thus, this study with experimental models supports the feasibility of COMP as a serum marker for cartilage involvement in arthritis. It also strengthens its potential as a tool, both in studies of mechanisms of cartilage damage in arthritis, and in studies examining effects of therapeutic interventions aimed at modifying the destructive process. Acknowledgements We thank Mette Lindell for skilful technical assistance and Dr R. A. Harris for a critical reading of the manuscript. This study was supported by grants from The Swedish Medical Research Council, The Österlund, Kock and Crafoord Foundations, the King Gustaf V 80-year Fund and Reumatikerförbundet. Börje Dahlin Foundation and Nanna Svartz Foundation. References 1. Alarcon-Segovia D, Abud-Mendoza C, Diaz-Jouanen E, Iglesias A, De los Reyes V, Hernandez-Ortiz J. Deforming

1000 E. Larsson et al. arthropathy of the hands in systemic lupus erythematosus. J Rheumatol 1988;15:65 9. 2. Kirwan JR. The relationship between synovitis and erosions in rheumatoid arthritis. Br J Rheumatol 1997;36:225 8. 3. Wollheim FA. Markers of disease in rheumatoid arthritis. Curr Opin Rheumatol 2000;12:200 4. 4. Hedbom E, Antonsson P, Hjerpe A et al. Cartilage matrix proteins. An acidic oligomeric protein (COMP) detected only in cartilage. J Biol Chem 1992;267:6132 6. 5. 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. 6. Smith RK, Zunino L, Webbon PM, Heinegard 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. 7. Neidhart M, Hauser N, Paulsson M, DiCesare PE, Michel BA, Hauselmann HJ. Small fragments of cartilage oligomeric matrix protein in synovial fluid and serum as markers for cartilage degradation. Br J Rheumatol 1997;36:1151 60. 8. 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. 9. Saxne TMB, Firestein G, Panayi G, Wollheim FA. Molecular markers for assessment of cartilage damage in rheumatoid arthritis. In: Firestein GS, Panayi GS, Wollheim F, eds. Rheumatoid arthritis new frontiers in pathogenesis and treatment, Chap. 21. Oxford: Oxford University Press, 2000:291 304. 10. Vingsbo-Lundberg C, Saxne T, Olsson H, Holmdahl R. Increased serum levels of cartilage oligomeric matrix protein in chronic erosive arthritis in rats. Arthritis Rheum 1998;41:544 50. 11. Larsson E, Mussener A, Heinegard D, Klareskog L, Saxne T. Increased serum levels of cartilage oligomeric matrix protein and bone sialoprotein in rats with collagen arthritis. Br J Rheumatol 1997;36:1258 61. 12. Joosten LAB, Helsen MA, Saxne T, Van de Loo FAJ, Heinegård D, van den Berg WB. IL-1 blockade prevents cartilage and bone destruction in murine type II collageninduced arthritis, whereas TNF blockade only ameliorates joint inflammation. J Immunol 1999;163:5049 55. 13. Bjork J, Kleinau S, Tengblad A, Smedegard G. Elevated levels of serum hyaluronate and correlation with disease activity in experimental models of arthritis. Arthritis Rheum 1989;32:306 11. 14. Engstrom-Laurent A, Hallgren R. Circulating hyaluronic acid levels vary with physical activity in healthy subjects and in rheumatoid arthritis patients. Relationship to synovitis mass and morning stiffness. Arthritis Rheum 1987;30:1333 8. 15. Larsson A, Bjork J, Lundberg C. Nephelometric determination of rat fibrinogen as a marker of inflammatory response. Vet Immunol Immunopathol 1997;59:163 9. 16. Trentham DE, Townes AS, Kang AH. Autoimmunity to type II collagen, an experimental model of arthritis. J Exp Med 1977;146:857 68. 17. Larsson P, Kleinau S, Holmdahl R, Klareskog L. Homologous type II collagen-induced arthritis in rats. Characterization of the disease and demonstration of clinically distinct forms of arthritis in two strains of rats after immunization with the same collagen preparation. Arthritis Rheum 1990;33:693 701. 18. Kleinau S, Erlandsson H, Holmdahl R, Klareskog L. Adjuvant oils induce arthritis in the DA rat. I. Characterization of the disease and evidence for an immunological involvement. J Autoimmun 1991;4:871 80. 19. Andersson M, Holmdahl R. Analysis of type II collagenreactive T cells in the mouse. I. Different regulation of autoreactive vs. non-autoreactive anti-type II collagen T cells in the DBAu1 mouse. Eur J Immunol 1990;20:1061 6. 20. Smith BD, Martin GR, Miller EJ, Dorfman A, Swarm R. Nature of the collagen synthesized by a transplanted chondrosarcoma. Arch Biochem Biophys 1975;166:181 6. 21. Lorentzen JC, Issazadeh S, Storch M et al. Protracted, relapsing and demyelinating experimental autoimmune encephalomyelitis in DA rats immunized with syngeneic spinal cord and incomplete Freund s adjuvant. J Neuroimmunol 1995;63:193 205. 22. Saxne T, Heinegård D. Cartilage oligomeric matrix protein: a novel marker of cartilage turnover detectable in synovial fluid and blood (published erratum appears in Br J Rheumatol 1993;32:247). Br J Rheumatol 1992; 31:583 91. 23. Brandt R, Hedlof E, Asman I, Bucht A, Tengblad A. A convenient radiometric assay for hyaluronan. Acta Otolaryngol Suppl. 1987;442:31 5.