- Email: [email protected]
Quantitation of chondroitin-sulfates, disaccharides and hyaluronan in normal, early and advanced osteoarthritic sheep temporomandibular joints
Osteoarthritis and Cartilage (2001) 9, 365–370 © 2001 OsteoArthritis Research Society International doi:10.1053/joca.2000.0397, available online at http://www.idealibrary.com on
1063–4584/01/040365+06 $35.00/0
Quantitation of chondroitin-sulfates, disaccharides and hyaluronan in normal, early and advanced osteoarthritic sheep temporomandibular joints J.-I. Ishimaru*, N. Ogi†, S. Mizuno‡ and A. N. Goss§ *Department of Oral and Maxillofacial Surgery, Gifu Prefectural Gifu Hospital, Japan †The Second-Department of Oral and Maxillofacial Surgery, Aichi-Gakuin University, School of Dentistry, Japan ‡Medical Affairs Department, Seikagaku Corporation, Japan §Oral and Maxillofacial Surgery Unit, Dental School, The University of Adelaide, Australia Summary Objective: To determine the relationship between synovial fluid, chondroitin sulfate disaccharide and hyaluronic acid to differing degrees of experimental temporomandibular joint (TMJ) osteoarthritis (OA). Design: Twenty-four merino sheep were divided into three groups and had different TMJ surgical procedures to produce OA. Group I; control (six sheep), Group II; disc perforation (nine sheep) and Group III; disc perforation and articular damage (nine sheep). Synovial fluid was collected initially and at sacrifice at 3 months. Chondroitin 4-sulfate, chondroitin 6-sulfate and hyaluronic acid were measured and correlated to the OA histologic score. Results: The chondroitin-sulfate levels were significantly increased (Group I to Group II P<0.001; Group I to Group III P<0.001), the hyaluronic acid levels decreased (Group I to Group II P<0.01; Group I to Group III P<0.01) with the increasing OA score. Conclusion: Chondroitin-sulfate and hyaluronic acid show a correlation with surgically created TMJ osteoarthritis in sheep model. © 2001 OsteoArthritis Research Society International Key words: Temporomandibular joint, Osteoarthritis, Chondroitin sulfate, Hyaluronic acid.
degradative events will result in alteration of the biochemical composition of the synovial fluid (SF)6–8. Thus, interest has focused on the early diagnosis and related markers in SF for TMD7. Chondroitin-sulfate (CS) isomers are useful markers in SF of knee and hip OA9–16, rheumatoid arthritis (RA)15 and traumatic arthritis15,16. Yoshihara et al.13 have shown the increased levels of chondroitin 4-sulfate (C4S) and chondroitin 6-sulfate (C6S) in knee OA and following rupture of the anterior cruciate ligament. Generally it is accepted that CS levels increase and hyaluronic acid (HA) levels decrease in OA and acute trauma11,17,18. Whether or not this applies to the TMJ is unconfirmed. Although the TMJ has many similarities to the knee and hip, articular cartilage is different, with the knee and hip being hyaline cartilage and the TMJ fibrocartilage. There are also technical problems associated with the small size of the TMJ and hence the difficulty in obtaining adequate samples for accurate quantitation. Several reports show increased chondroitinsulfate levels in TMJ SF for patients with TMD19–21. The Japan–Australia Temporomanibular Joint Disorder (TMD) Research Group has produced a number of standardized models for systematic study of intraarticular TMJ pathology2,23–24. These include a model of advanced OA22, which involved articular damage to induce OA and has been used to study the effects of different types of surgical reconstruction25–27. It was also found that the OA
Introduction Temporomandibular joint disorders (TMD) cover a range of muscular and articular disorders1,2. In the last decade, there has been considerable interest in characterizing the different types of intraarticular TMJ pathologies. Plain radiographs do not correlate well with intraarticular pathology, but represent past remodeling of the condyle3. Arthrograms and magnetic resonance imaging demonstrate disc, condyle and fossa relationships and marked pathology such as disc perforation and osteoarthritis (OA). However, these findings can also be commonly demonstrated in asymptomatic joints, which limits the diagnostic value of those imaging techniques. Arthroscopy allows direct imaging, primarily of the superior joint space, which has allowed the evaluation of the disc and the articular surface by the presence of adhesions and inflammation4. More recently interest has been directed at the molecular level5. In the early stage, cartilage microfibrillation or degradation is not shown by any arthroscopic examination. These initial Received 20 September 1999; revision requested 3 December 1999; revision received 11 April 2000; accepted 7 November 2000; published online 2 April 2001. Address correspondence to: Dr Jun-Ichi Ishimaru, Department of Oral and Maxillofacial Surgery, Gifu Prefectural Gifu Hospital, 4-6-1 Noishiki Gifu 500-8717, Japan. Tel: +81 58 246 1111; Fax: +81 58 248 3805; E-mail: [email protected]
365
366
J.-I. Ishimaru et al.: Normal, early and advanced osteoarthritic sheep temporomandibular joints
process can be reversed if HA is injected repeatedly into the joint space shortly after OA induction28. The early stages of OA without advanced progression can be induced by disc perforation24. The aim of this paper is to investigate the changes in CS and HA levels in control, early and late OA in the sheep TMJ.
Materials and methods Twenty-four adult Merino sheep, each weighing approximately 60 kg, were used in this study and their management followed our previously described technique22,23. Six sheep (Group I) were unoperated and acted as controls. Nine sheep (Group II) had a surgically created disc perforation24. Nine sheep (Group III) had surgically induced OA22. In Groups II and III, anesthesia was induced with an injection of thiopentone sodium into the external jugular vein. Intubation was performed and anesthesia was maintained with isoflurane/N2O/O2. The right unilateral preauricular region was shaved and prepared with antiseptic solution and draped with a sterilized sheet. A vertical skin incision was made to the pre-auricular region and underlying tissue was dissected to expose the lateral capsule of the TMJ. Both the superior and the inferior joint space were opened by horizontal incisions to expose the condylar surface. In the central part of the disc a 3.8 mm diameter of perforation was then surgically created, using a round punch. In Group III, additional surgery was performed by gentle removal of both the articular fibrous cover and the underlying superficial fibrocartilage with a scalpel and periosteal elevator according to the method used by Ishimaru and Goss22. Then the wound was rinsed with normal saline and closed in layers. The sheep were returned to field conditions after 1 week’s observation in a covered animal house. At 3 months all sheep were reanesthetized and SF was collected by aspiration through the capsule; they were then killed by anesthetic overdose into the external jugular vein. The TMJ was removed ‘en bloc’ with a band saw and fixed in neutral buffered formalin. In Group I, SF was collected from the superior joint space in the TMJ under intravenous sedation with thiopentone. SF was placed in individual containers, and immediately centrifuged at 3000 rpm for 5 min to get rid of cell debris, then deeply frozen and stored at −70°C before analysis. HISTOLOGICAL PREPARATION OF THE TMJ
The fixed blocks were decalcified with 9.5% hydrochloric acid and 1% sodium acetate over saturated EDTA. Each specimen was sectioned in a parasagittal plane into lateral, central and medial specimens. These specimens were sectioned at 5 m thickness and stained with hematoxylin and eosin. Each specimen was assessed at its anterior, central and posterior aspects, thus dividing the joint into nine zones. Each zone was evaluated, both descriptively and by adding the index of osteoarthritic score according to the modified method by Ishimaru and Goss (Table I)22. A double determination was performed 1 week from the first observation. The OA score was then determined for each joint by the mean of the scores from the nine zones. BIOCHEMICAL ANALYSIS OF THE SYNOVIAL FLUID
The synovial fluid specimen was diluted 10-fold with distilled water and digested with chondroitinase ABC
Table I Osteoarthritic score The temporal bone No change Progressive remodeling Fibrillation Fibrous adhesion Complete loss of trabecular pattern Bony destruction The disc No change Disc thinning Disc tearing or torn Disc perforation Fibrous adhesion Complete loss of the disc The condyle (surface morphology) No change Fibrillation Denudation, eburnation Osteophyte, flattening, sclerosis Fibrous repair Fibrous ankylosis The condyle (bone structure) No change Bone remodeling Downgrowth of articular soft tissue Sub-cortical cyst formation Marrow fibrosis Complete loss of trabecular pattern
0 1 2 3 4 5 0 1 2 3 4 5 0 1 2 3 4 5 0 1 2 3 4 5
(0.625 U/ml) and hyaluronidase SD (33 mU/ml) derived from streptococcus dysgalactiae (HAase SD, Seikagaku corporation, Japan). To 200 l of diluted joint fluid, 50 l of chondroitinase ABC solution in distilled water was added 80 l of 10 mM of sodium acetate buffer (pH 8.0), and 70 l of distilled water. The mixture was incubated at 37°C for 2 h and then ultrafiltered. The filtrate obtained was analysed by high performance liquid chromatography (HPLC).
HPLC ANALYSIS
HPLC analysis of the unsaturated disaccharides derived from chondroitin-sulfate in the joint fluid was performed according to the method of Toyoda et al.29 and Shinmei et al.15. Briefly, the HPLC analysis system consists of two pumps (model 880-PU, Japan Spectroscopic Co. Ltd., Tokyo, Japan), an autosampling injector, a stainless steel column packed with silica gel (YMC gel PA-120), a spectrofluorometer (model 820-FP, Japan Spectoroscopic), and an integrator (model 805-GI, Japan Spectroscopic). The unsaturated disaccharides in each sample were eluted with a gradient of 0–100 sodium sulfate for 60 min at a flow rate of 0.5 ml/min. To the eluant from the column was added 100 mM sodium tetraborate buffer containing 10 mg/ml of cyanoacetamide. Then the mixture was passed through a polyetheletherketone (PEEK) reaction coil set in the dry reaction bath at 137°C. The effluent was monitored by spectrofluorometer set at an excitation of 331 nm and emission of 383 nm. Each peak corresponding to unsaturated dissacharide was calculated by the integrator.
Osteoarthritis and Cartilage Vol. 9, No. 4
Fig. 1. Histological appearance of normal sheep TMJ of Group I. (T: temporal bone, D: disc, C: condyle, HE×2.)
367
Fig. 2. Histological appearance of sheep TMJ of disc perforation in Group II. (T: temporal bone, D: disc, C: condyle, HE×2.)
MORGAN–ELSON’S METHOD
Colorimetric determination of the disaccharide derived from HA was performed by Morgan–Elson’s method. Briefly, the unsaturated disaccharide in each sample was reacted with Ehrlich reagent. After the reaction, absorbance was measured at 585 nm. The concentration of HA of sample was calculated from the analytical value of standard HA solution. STATISTICAL ANALYSIS
Data were stored and analysed with Statview 4.5 by a Macintosh computer. ANOVA and Fisher’s PLSD were applied for the analysis of CS and HA levels in SF of the three groups. Kruskal–Wallis test and Mann–Whitney U-analysis were applied for the comparison of OA scores among the groups. Regression analysis was performed and correlations were analysed by Spearman’s rank correlation coefficient between OA score and CS, HA levels in SF. P-values less than 0.05 were considered significant.
Fig. 3. Histological appearance of osteoarthritic sheep TMJ of Group III. (T: temporal bone, D: disc, C: condyle, HE×2.2.)
Results HISTOLOGICAL FINDINGS
For Group I, control TMJs (Fig. 1), osteoarthritic score was 0. Generally, Group II showed adaptive remodeling and early OA (Fig. 2). There was a consistency in histological changes in Group II. Group III showed typical OA (Fig. 3) and there was a consistency in histological changes. Generally the osteoarthritic stage was low in Group II and higher in Group III (Fig. 4). There was a significant difference of OA scores between Group I and II (*P<0.0001), Group I and Group III (*P<0.0001), and Group II and Group III (**P<0.0001, Fig. 4). SYNOVIAL FLUID ANALYSES
The measured CS isomers levels and HA concentrations are presented in Table II. Generally, GAG isomers such as chondroitin 4-sulfate (Di4S) and chondroitin 6-sulfate (Di6S) increased in Group II and, more markedly, in Group III. There were statistically significant differences both between Group I and Group III (P<0.001), and Group II and Group III (P<0.001). Di6S/Di4 increased accord-
Fig. 4. Comparison of OA stages (mean and S.D.) among Group I, II and III. Statistical analysis shows that OA scores are significantly high in Group II and III (*P<0.0001) compared to Group I, and significantly high in Group III (**P<0.0001) compared to Group II. (P<0.0001, Kruskal–Wallis test, and Mann–Whitney U-analysis as post-hoc.)
ing the group number with significant differences among any groups (P<0.01, P<0.001). On the other hand, HA concentration decreased in Groups II and III (P<0.01) compared with Group I.
368
J.-I. Ishimaru et al.: Normal, early and advanced osteoarthritic sheep temporomandibular joints Table II Released chondroitin-sulfate isomers and hyaluronic acid concentrations in sheep TMJ synovial fluid by experimental group
Group I (Control) Group II (Early OA) Group III (Advanced OA) GI vs GII GI vs GIII GII vs GIII
Di6S
Di4S
Di6S+Di4S
Di6S/Di4S
HA
5.9±0.4 (3.7–8.2) 11.5±1.4 (5.4–16.3) 49.3±7.8 (25.2–93.8) n.s. P<0.001 P<0.001
4.9±0.3 (3.9–6.5) 6.8±0.6 (4.1–8.4) 23.6±2.7 (13.0–38.5) n.s. P<0.001 P<0.001
10.8±0.5 (8.5–14.5) 18.4±2.0 (9.5–24.5) 72.9±10.5 (40.8–132.3) n.s. P<0.001 P<0.001
1.2±0.3 (0.8–1.9) 1.7±0.2 (1.1–2.2) 2.0±0.4 (1.2–2.5) P<0.01 P<0.001 P<0.001
2.8±0.2 (1.8–4.6) 2.5±0.2 (1.5–3.3) 1.9±0.2 (0.9–3.6) n.s. P<0.01 P<0.01
Di6S, Di4S (g/ml), HA (mg/ml) mean± S.D., with ranges. ANOVA and Fisher’s PLSD; n.s.: not significant.
Fig. 5. Correlation between level of chondroitin 6-sulfate in TMJ synovial fluid and OA stage. A significant positive correlation was observed (Spearman’s rank correlation coefficient, r=0.866, P<0.0001).
THE RELATIONSHIP BETWEEN OA SCORE AND SEVERAL PARAMETERS IN SF
A significant positive correlation was observed between OA scores and C6S (r=0.866, P<0.0001, Fig. 5), C4S (r=0.870, P<0.0001, Fig. 6), C4S+C6S(r=0.068, P<0.0001, Fig. 7) and C6S/C4S(r=0.828, P<0.0001, Fig. 8). However, a significant negative correlation was observed between OA scores and HA(r= −0.692, P<0.01, Fig. 9).
Discussion In this study a strong correlation was demonstrated between the severity of osteoarthritic changes and the released CS and HA levels in SF of the sheep TMJ. Histologically, Group II with the surgical disc perforation and minor osteoarthritic changes was consistent with early OA. Group III showed typical OA. C6S, C4S, the sum of C6S and C4S and the ratio of C6S to C4S showed a strong positive correlation to osteoarthritic changes and those parameters increased well in accordance with the intraarticular osteoarthritic changes. In contrast, HA showed a strong negative correlation to the OA changes. These
Fig. 6. Correlation between level of chondroitin 4-sulfate in TMJ synovial fluid and OA stage. A significant positive correlation was observed (Spearman’s rank correlation coefficient, r=0.870, P<0.0001).
results strongly suggest that intraarticular pathology is reflected in SF, and those markers have a diagnostic potential in SF analysis. Clinically, SF is the only material readily obtainable from human TMJ, meaning that this biochemical analysis may be useful for real time diagnosis for clinical patients. This study is consistent with the findings of orthopedic research, which have shown that CS level increases and HA level decreases in SF of joints with OA. However, such clinical and experimental studies were done in knee and hip joints. There have been limited studies to evaluate CS and HA levels in human TMJ19–21. Those studies were done using SF lavage, and this remains a problem for a precise calibration of each parameter. Thus, an experimental study was indispensable in order to confirm the correlationship between CS and HA levels in TMJ SF and histological changes. Previous studies reported that C6S is the main CS of articular cartilage30. Thus, the C6S level in SF has been considered to represent the degree of matrix degradation. This coincides with the results of this study. However, recent studies suggest that biosynthesis of CS may increase in the early stage of OA in order to attempt repair, but if this attempt fails then subsequent OA may result11,18. Therefore it is still controversial whether the increased level
Osteoarthritis and Cartilage Vol. 9, No. 4
Fig. 7. Correlation between sum level of chondroitin 6-sulfate and chondroitin 4-sulfate in TMJ synovial fluid and OA stage. A significant positive correlation was observed (Spearman’s rank correlation coefficient, r=0.868, P<0.0001).
369
Fig. 9. Correlation between HA level of in TMJ synovial fluid and OA stage. A significant negative correlation was observed (Spearman’s rank correlation coefficient, r= −0.692, P<0.01).
limitation to the simulation of human OA in experimental models. However, it is important to find the several evidences of articular pathology for the research of human OA. Certainly further investigation is necessary to define the precise correlationship between the levels of several synovial markers and ongoing intraarticular pathology in the TMJ. However, from this study it can be concluded that C4S, C6S and HA are strongly correlated with intraarticular pathology in the TMJ.
References
Fig. 8. Correlation between ratio of C6S to C4S in TMJ synovial fluid and OA stage. A significant positive correlation was observed (Spearman’s rank correlation coefficient, r=0.828, P<0.0001).
of CS represents the ongoing articular cartilage degradation or the result of hypersynthesis to an attempt at repair17. On the other hand, HA is mainly secreted by synoviocytes in synovial membrane. In this study HA concentration decreased in SF in accordance with the OA progress. This finding coincides with previous reports31,32. In this pathological condition, however, it is still controversial whether biosynthesis of HA from synovial membrane decreased or body fluid infiltration reduces HA concentration. Recently, a clinical cross-sectional study showed decreased levels of HA and GAG in OA compared with normal11. Further studies are necessary to examine the calibrated CS and HA levels by total GAG or total protein and to evaluate the correlationship between those levels in TMJ SF and histology. Thus, at present, clinical data from TMJ SF must be considered with caution. However, in OA intraarticular friction may occur, causing cartilage damage. Intraarticular matrix degradation then gradually develops, and a vicious circle is formed. This is an essential difference of human OA from experimentally induced OA. Thus, there is a
1. Ishimaru J-I, Kurita K, Ogi N, Handa Y, Toida M. Effect of tizanidine for a primary care of patients with myofacial pain. A preliminary study. Hosp Dent Oral Maxillofac Surg 1995;7:2–7. 2. Ishimaru J-I, Kurita K, Handa Y, Goss AN. Temporomandibular joint osteoarthritis. Literature review and experimental animal models. Hosp Dent Oral Maxillofac Surg 1991;3:64–8. 3. Muir C, Goss AN. The radiological morphology of asymptomatic temporomandibular joints. Oral Surg 1990;70:355–59. 4. Goss AN, Bosanquet AG. The arthroscopic appearance of acute temporomandibular joint trauma. J Oral Maxillofac Surg 1990;48:780–83. 5. Milan SB, Schmitz JP. Molecular biology of temporomandibular joint disorders: Proposed mechanism of disease. J Oral Maxillofac Surg 1995;53:1448–54. 6. Kubota E, Imamura H, Kubota T, Shibata T, Murakami K. Interleukin 1 and stromelysin(MMP-3) activity of synovial fluid as possible markers of osteoarthritis in the temporomandibular joint. J Oral Maxillofac Surg 1997;55:20–7. 7. Israel HA, Saed-Nejad F, Ratcliffe A. Early diagnosis of osteoarthrosis of the temporomandibular joint: correlation between arthroscopic diagnosis and keratan sulfate levels in the synovial fluid. J Oral Maxillofac Surg 1991;49:708–11. 8. Miyamoto K, Ishimaru J-I, Tatematsu N, Kurita K, Ogi N, Handa Y, et al. Biochemical analysis of temporomandibular joint synovial fluid. Literature review and
370
9.
10. 11.
12. 13.
14.
15.
16.
17. 18.
19.
20.
J.-I. Ishimaru et al.: Normal, early and advanced osteoarthritic sheep temporomandibular joints experimental animal models. Hosp Dent (Tokyo) 1995;7:44–50. Ratcliffe A, Shurety W, Caterson B. The quantitation of a native chondroitin sulfate epitope in synovial fluid lavages and articular cartilage from canine experimental osteoarthritis and disuse atrophy. Arthritis Rheum 1993;36:543–51. Plaas AH, West LA, Wong-Palms S, Nelson FRT. Glycosaminoglycan sulfation in human osteoarthritis. J Biol Chem 1988;15:12642–9. Belcher C, Yaqub R, Fawthrop F, Bayliss M, Doherty M. Synovial fluid chondroitin and keratan sulfate epitopes, glycosaminoglycans, and hyaluronan in arthritic and normal knees. Ann Rheum Dis 1997;56:299–307. Hardingham T. Changes in chondroitin sulfate structure induced by joint disease. Acta Orthop Scand 1995;266:107–10. Yoshihara Y, Yamada H, Miyauchi S, Ito S, Samura A, Shinmei M. Levels of chondroitin 4-sulfate, chondroitin 6-sulfate and carboxy-terminal type II procollagen peptide in knee synovial fluid after injury to the anterior cruciate ligament. Ryumachi 1996;36: 734–40 (in Japanese). Yamada H, Miyauchi S, Hotta H, Morita M, Yoshihara Y, Kikuchi T, et al. Levels of chondroitin sulfate isomers in synovial fluid of patients with hip osteoarthritis. J Orthop Sci 1999;4:250–4. Shinmei M, Miyauchi S, Machida A, Miyazaki K. Quantitation of chondroitin 4 sulfate and chondroitin 6-sulfate in pathologic joint fluid. Arthritis Rheum 1992;35:1304–8. Shinmei M, Kobayashi T, Yoshihara Y, Samura A. Significance of the levels of carboxy terminal type II collagen peptides, chondroitin sulfate isomers, tissue inhibitor of metalloproteinases, and metalloproteinases in osteoarthritis joint fluid. J Rheumatol Suppl 1995;43:78–81. Hardingham T. Changes in chondroitin sulfate and joint disease. Osteoarthritis Cart 1998;6(Suppl A):3–5. Uebelhart D, Thonar EJMA, Zhang J, Williams JM. Protective effect of exogenous chondroitin 4,6-sulfate in the acute degradation of articular cartilage in the rabbit. Osteoarthritis Cart 1998;6(Suppl A):6–13. Murakami KI, Shibata T, Kubota E, Maeda H. Intraarticular levels of prostaglandin E2, hyaluronic acid, and chondroitin-4 and-6 sulfates in the temporomandibular joint synovial fluid of patients with internal derangement. J Oral Maxillofac Surg 1998;56:199–203. Ratcliffe A, Israel HA, Saed-Nejad F, Diamond B. Proteoglycans in the synovial fluid of the temporo-
21.
22. 23. 24.
25.
26.
27.
28.
29.
30. 31.
32.
mandibular joint as an indicator of changes in cartilage metabolism during primary and secondary osteoarthritis. J Oral Maxillofac Surg 1998;56:204–8. Shibata T, Murakami KI, Kubota E, Maeda H. Glycosaminoglycan components in temporomandibular joint synovial fluid as markers of joint pathology. J Oral Maxillofac Surg 1998;56:209–13. Ishimaru J-I, Goss AN. A model for osteoarthrosis of the temporomandibular joint. J Oral Maxillofac Surg 1992;50:1191–5. Ishimaru J-I, Handa Y, Kurita K, Goss AN. The effect of marrow perforation in sheep temporomandibular joint. Int J Oral Maxillofac Surg 1992;21:239–42. Bosanquet AN, Ishimaru J-I, Goss AN. The effect of experimental disc perforation in sheep temporomandibular joints. Int J Oral Maxillofac Surg 1991;20:177–81. Ogi N, Kurita K, Handa Y, Goss AN. The short-term effect of discectomy on the osteoarthrotic temporomandibular joint in sheep. Int J Oral Maxillofac Surg 1996;25:319–24. Ogi N, Kurita K, Handa Y, Goss AN. Short-term effect of fresh disk allograft on the osteoarthrotic sheep temporomandibular joint. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 1997;84:259–64. Ogi N, Kurita K, Handa Y, Goss AN. The short-term effect of autogenous auricular cartilage graft following discectomy on the osteoarthrotic temporomandibular joint in sheep. Int J Oral Maxillofac Surg 1997;26:217–22. Neo H, Ishimaru J-I, Kurita K, Goss AN. The effect of hyaluronic acid on experimental temporomandibular joint osteoarthrosis in the sheep. J Oral Maxillofac Surg 1997;55:1114–9. Toyoda H, Shinomiya K, Yamanashi S, Koshiishi I, Imanari T. Microdetermination of unsaturated disaccharides produced from chondroitin sulfates in rabbit plasma by high performance liquid chromatography with fluorometric detection. Anal Sci 1988;4:381–4. Mourao PAS. Distribution of chondroitin 4-sulfate and chondroitin 6-sulfate in human articular and growth cartilage. Arthritis Rheum 1988;31:1028–33. Balazs EA, Watson D, Duff IF, Roseman S. Hyaluronic acid in synovial fluid. Molecular parameters of hyaluronic acid in normal and arthritic human fluids. Arthritis Rheum 1967;10:357–76. Dahl LB, Dahl IMS, Engstrom-Laurent A, Granath K. Concentration and molecular weight of sodium hyaluronate in synovial fluid from patients with rheumatoid arthritis and other arthropathies. Ann Rheum Dis 1985;44:817–22.
1063–4584/01/040365+06 $35.00/0
Quantitation of chondroitin-sulfates, disaccharides and hyaluronan in normal, early and advanced osteoarthritic sheep temporomandibular joints J.-I. Ishimaru*, N. Ogi†, S. Mizuno‡ and A. N. Goss§ *Department of Oral and Maxillofacial Surgery, Gifu Prefectural Gifu Hospital, Japan †The Second-Department of Oral and Maxillofacial Surgery, Aichi-Gakuin University, School of Dentistry, Japan ‡Medical Affairs Department, Seikagaku Corporation, Japan §Oral and Maxillofacial Surgery Unit, Dental School, The University of Adelaide, Australia Summary Objective: To determine the relationship between synovial fluid, chondroitin sulfate disaccharide and hyaluronic acid to differing degrees of experimental temporomandibular joint (TMJ) osteoarthritis (OA). Design: Twenty-four merino sheep were divided into three groups and had different TMJ surgical procedures to produce OA. Group I; control (six sheep), Group II; disc perforation (nine sheep) and Group III; disc perforation and articular damage (nine sheep). Synovial fluid was collected initially and at sacrifice at 3 months. Chondroitin 4-sulfate, chondroitin 6-sulfate and hyaluronic acid were measured and correlated to the OA histologic score. Results: The chondroitin-sulfate levels were significantly increased (Group I to Group II P<0.001; Group I to Group III P<0.001), the hyaluronic acid levels decreased (Group I to Group II P<0.01; Group I to Group III P<0.01) with the increasing OA score. Conclusion: Chondroitin-sulfate and hyaluronic acid show a correlation with surgically created TMJ osteoarthritis in sheep model. © 2001 OsteoArthritis Research Society International Key words: Temporomandibular joint, Osteoarthritis, Chondroitin sulfate, Hyaluronic acid.
degradative events will result in alteration of the biochemical composition of the synovial fluid (SF)6–8. Thus, interest has focused on the early diagnosis and related markers in SF for TMD7. Chondroitin-sulfate (CS) isomers are useful markers in SF of knee and hip OA9–16, rheumatoid arthritis (RA)15 and traumatic arthritis15,16. Yoshihara et al.13 have shown the increased levels of chondroitin 4-sulfate (C4S) and chondroitin 6-sulfate (C6S) in knee OA and following rupture of the anterior cruciate ligament. Generally it is accepted that CS levels increase and hyaluronic acid (HA) levels decrease in OA and acute trauma11,17,18. Whether or not this applies to the TMJ is unconfirmed. Although the TMJ has many similarities to the knee and hip, articular cartilage is different, with the knee and hip being hyaline cartilage and the TMJ fibrocartilage. There are also technical problems associated with the small size of the TMJ and hence the difficulty in obtaining adequate samples for accurate quantitation. Several reports show increased chondroitinsulfate levels in TMJ SF for patients with TMD19–21. The Japan–Australia Temporomanibular Joint Disorder (TMD) Research Group has produced a number of standardized models for systematic study of intraarticular TMJ pathology2,23–24. These include a model of advanced OA22, which involved articular damage to induce OA and has been used to study the effects of different types of surgical reconstruction25–27. It was also found that the OA
Introduction Temporomandibular joint disorders (TMD) cover a range of muscular and articular disorders1,2. In the last decade, there has been considerable interest in characterizing the different types of intraarticular TMJ pathologies. Plain radiographs do not correlate well with intraarticular pathology, but represent past remodeling of the condyle3. Arthrograms and magnetic resonance imaging demonstrate disc, condyle and fossa relationships and marked pathology such as disc perforation and osteoarthritis (OA). However, these findings can also be commonly demonstrated in asymptomatic joints, which limits the diagnostic value of those imaging techniques. Arthroscopy allows direct imaging, primarily of the superior joint space, which has allowed the evaluation of the disc and the articular surface by the presence of adhesions and inflammation4. More recently interest has been directed at the molecular level5. In the early stage, cartilage microfibrillation or degradation is not shown by any arthroscopic examination. These initial Received 20 September 1999; revision requested 3 December 1999; revision received 11 April 2000; accepted 7 November 2000; published online 2 April 2001. Address correspondence to: Dr Jun-Ichi Ishimaru, Department of Oral and Maxillofacial Surgery, Gifu Prefectural Gifu Hospital, 4-6-1 Noishiki Gifu 500-8717, Japan. Tel: +81 58 246 1111; Fax: +81 58 248 3805; E-mail: [email protected]
365
366
J.-I. Ishimaru et al.: Normal, early and advanced osteoarthritic sheep temporomandibular joints
process can be reversed if HA is injected repeatedly into the joint space shortly after OA induction28. The early stages of OA without advanced progression can be induced by disc perforation24. The aim of this paper is to investigate the changes in CS and HA levels in control, early and late OA in the sheep TMJ.
Materials and methods Twenty-four adult Merino sheep, each weighing approximately 60 kg, were used in this study and their management followed our previously described technique22,23. Six sheep (Group I) were unoperated and acted as controls. Nine sheep (Group II) had a surgically created disc perforation24. Nine sheep (Group III) had surgically induced OA22. In Groups II and III, anesthesia was induced with an injection of thiopentone sodium into the external jugular vein. Intubation was performed and anesthesia was maintained with isoflurane/N2O/O2. The right unilateral preauricular region was shaved and prepared with antiseptic solution and draped with a sterilized sheet. A vertical skin incision was made to the pre-auricular region and underlying tissue was dissected to expose the lateral capsule of the TMJ. Both the superior and the inferior joint space were opened by horizontal incisions to expose the condylar surface. In the central part of the disc a 3.8 mm diameter of perforation was then surgically created, using a round punch. In Group III, additional surgery was performed by gentle removal of both the articular fibrous cover and the underlying superficial fibrocartilage with a scalpel and periosteal elevator according to the method used by Ishimaru and Goss22. Then the wound was rinsed with normal saline and closed in layers. The sheep were returned to field conditions after 1 week’s observation in a covered animal house. At 3 months all sheep were reanesthetized and SF was collected by aspiration through the capsule; they were then killed by anesthetic overdose into the external jugular vein. The TMJ was removed ‘en bloc’ with a band saw and fixed in neutral buffered formalin. In Group I, SF was collected from the superior joint space in the TMJ under intravenous sedation with thiopentone. SF was placed in individual containers, and immediately centrifuged at 3000 rpm for 5 min to get rid of cell debris, then deeply frozen and stored at −70°C before analysis. HISTOLOGICAL PREPARATION OF THE TMJ
The fixed blocks were decalcified with 9.5% hydrochloric acid and 1% sodium acetate over saturated EDTA. Each specimen was sectioned in a parasagittal plane into lateral, central and medial specimens. These specimens were sectioned at 5 m thickness and stained with hematoxylin and eosin. Each specimen was assessed at its anterior, central and posterior aspects, thus dividing the joint into nine zones. Each zone was evaluated, both descriptively and by adding the index of osteoarthritic score according to the modified method by Ishimaru and Goss (Table I)22. A double determination was performed 1 week from the first observation. The OA score was then determined for each joint by the mean of the scores from the nine zones. BIOCHEMICAL ANALYSIS OF THE SYNOVIAL FLUID
The synovial fluid specimen was diluted 10-fold with distilled water and digested with chondroitinase ABC
Table I Osteoarthritic score The temporal bone No change Progressive remodeling Fibrillation Fibrous adhesion Complete loss of trabecular pattern Bony destruction The disc No change Disc thinning Disc tearing or torn Disc perforation Fibrous adhesion Complete loss of the disc The condyle (surface morphology) No change Fibrillation Denudation, eburnation Osteophyte, flattening, sclerosis Fibrous repair Fibrous ankylosis The condyle (bone structure) No change Bone remodeling Downgrowth of articular soft tissue Sub-cortical cyst formation Marrow fibrosis Complete loss of trabecular pattern
0 1 2 3 4 5 0 1 2 3 4 5 0 1 2 3 4 5 0 1 2 3 4 5
(0.625 U/ml) and hyaluronidase SD (33 mU/ml) derived from streptococcus dysgalactiae (HAase SD, Seikagaku corporation, Japan). To 200 l of diluted joint fluid, 50 l of chondroitinase ABC solution in distilled water was added 80 l of 10 mM of sodium acetate buffer (pH 8.0), and 70 l of distilled water. The mixture was incubated at 37°C for 2 h and then ultrafiltered. The filtrate obtained was analysed by high performance liquid chromatography (HPLC).
HPLC ANALYSIS
HPLC analysis of the unsaturated disaccharides derived from chondroitin-sulfate in the joint fluid was performed according to the method of Toyoda et al.29 and Shinmei et al.15. Briefly, the HPLC analysis system consists of two pumps (model 880-PU, Japan Spectroscopic Co. Ltd., Tokyo, Japan), an autosampling injector, a stainless steel column packed with silica gel (YMC gel PA-120), a spectrofluorometer (model 820-FP, Japan Spectoroscopic), and an integrator (model 805-GI, Japan Spectroscopic). The unsaturated disaccharides in each sample were eluted with a gradient of 0–100 sodium sulfate for 60 min at a flow rate of 0.5 ml/min. To the eluant from the column was added 100 mM sodium tetraborate buffer containing 10 mg/ml of cyanoacetamide. Then the mixture was passed through a polyetheletherketone (PEEK) reaction coil set in the dry reaction bath at 137°C. The effluent was monitored by spectrofluorometer set at an excitation of 331 nm and emission of 383 nm. Each peak corresponding to unsaturated dissacharide was calculated by the integrator.
Osteoarthritis and Cartilage Vol. 9, No. 4
Fig. 1. Histological appearance of normal sheep TMJ of Group I. (T: temporal bone, D: disc, C: condyle, HE×2.)
367
Fig. 2. Histological appearance of sheep TMJ of disc perforation in Group II. (T: temporal bone, D: disc, C: condyle, HE×2.)
MORGAN–ELSON’S METHOD
Colorimetric determination of the disaccharide derived from HA was performed by Morgan–Elson’s method. Briefly, the unsaturated disaccharide in each sample was reacted with Ehrlich reagent. After the reaction, absorbance was measured at 585 nm. The concentration of HA of sample was calculated from the analytical value of standard HA solution. STATISTICAL ANALYSIS
Data were stored and analysed with Statview 4.5 by a Macintosh computer. ANOVA and Fisher’s PLSD were applied for the analysis of CS and HA levels in SF of the three groups. Kruskal–Wallis test and Mann–Whitney U-analysis were applied for the comparison of OA scores among the groups. Regression analysis was performed and correlations were analysed by Spearman’s rank correlation coefficient between OA score and CS, HA levels in SF. P-values less than 0.05 were considered significant.
Fig. 3. Histological appearance of osteoarthritic sheep TMJ of Group III. (T: temporal bone, D: disc, C: condyle, HE×2.2.)
Results HISTOLOGICAL FINDINGS
For Group I, control TMJs (Fig. 1), osteoarthritic score was 0. Generally, Group II showed adaptive remodeling and early OA (Fig. 2). There was a consistency in histological changes in Group II. Group III showed typical OA (Fig. 3) and there was a consistency in histological changes. Generally the osteoarthritic stage was low in Group II and higher in Group III (Fig. 4). There was a significant difference of OA scores between Group I and II (*P<0.0001), Group I and Group III (*P<0.0001), and Group II and Group III (**P<0.0001, Fig. 4). SYNOVIAL FLUID ANALYSES
The measured CS isomers levels and HA concentrations are presented in Table II. Generally, GAG isomers such as chondroitin 4-sulfate (Di4S) and chondroitin 6-sulfate (Di6S) increased in Group II and, more markedly, in Group III. There were statistically significant differences both between Group I and Group III (P<0.001), and Group II and Group III (P<0.001). Di6S/Di4 increased accord-
Fig. 4. Comparison of OA stages (mean and S.D.) among Group I, II and III. Statistical analysis shows that OA scores are significantly high in Group II and III (*P<0.0001) compared to Group I, and significantly high in Group III (**P<0.0001) compared to Group II. (P<0.0001, Kruskal–Wallis test, and Mann–Whitney U-analysis as post-hoc.)
ing the group number with significant differences among any groups (P<0.01, P<0.001). On the other hand, HA concentration decreased in Groups II and III (P<0.01) compared with Group I.
368
J.-I. Ishimaru et al.: Normal, early and advanced osteoarthritic sheep temporomandibular joints Table II Released chondroitin-sulfate isomers and hyaluronic acid concentrations in sheep TMJ synovial fluid by experimental group
Group I (Control) Group II (Early OA) Group III (Advanced OA) GI vs GII GI vs GIII GII vs GIII
Di6S
Di4S
Di6S+Di4S
Di6S/Di4S
HA
5.9±0.4 (3.7–8.2) 11.5±1.4 (5.4–16.3) 49.3±7.8 (25.2–93.8) n.s. P<0.001 P<0.001
4.9±0.3 (3.9–6.5) 6.8±0.6 (4.1–8.4) 23.6±2.7 (13.0–38.5) n.s. P<0.001 P<0.001
10.8±0.5 (8.5–14.5) 18.4±2.0 (9.5–24.5) 72.9±10.5 (40.8–132.3) n.s. P<0.001 P<0.001
1.2±0.3 (0.8–1.9) 1.7±0.2 (1.1–2.2) 2.0±0.4 (1.2–2.5) P<0.01 P<0.001 P<0.001
2.8±0.2 (1.8–4.6) 2.5±0.2 (1.5–3.3) 1.9±0.2 (0.9–3.6) n.s. P<0.01 P<0.01
Di6S, Di4S (g/ml), HA (mg/ml) mean± S.D., with ranges. ANOVA and Fisher’s PLSD; n.s.: not significant.
Fig. 5. Correlation between level of chondroitin 6-sulfate in TMJ synovial fluid and OA stage. A significant positive correlation was observed (Spearman’s rank correlation coefficient, r=0.866, P<0.0001).
THE RELATIONSHIP BETWEEN OA SCORE AND SEVERAL PARAMETERS IN SF
A significant positive correlation was observed between OA scores and C6S (r=0.866, P<0.0001, Fig. 5), C4S (r=0.870, P<0.0001, Fig. 6), C4S+C6S(r=0.068, P<0.0001, Fig. 7) and C6S/C4S(r=0.828, P<0.0001, Fig. 8). However, a significant negative correlation was observed between OA scores and HA(r= −0.692, P<0.01, Fig. 9).
Discussion In this study a strong correlation was demonstrated between the severity of osteoarthritic changes and the released CS and HA levels in SF of the sheep TMJ. Histologically, Group II with the surgical disc perforation and minor osteoarthritic changes was consistent with early OA. Group III showed typical OA. C6S, C4S, the sum of C6S and C4S and the ratio of C6S to C4S showed a strong positive correlation to osteoarthritic changes and those parameters increased well in accordance with the intraarticular osteoarthritic changes. In contrast, HA showed a strong negative correlation to the OA changes. These
Fig. 6. Correlation between level of chondroitin 4-sulfate in TMJ synovial fluid and OA stage. A significant positive correlation was observed (Spearman’s rank correlation coefficient, r=0.870, P<0.0001).
results strongly suggest that intraarticular pathology is reflected in SF, and those markers have a diagnostic potential in SF analysis. Clinically, SF is the only material readily obtainable from human TMJ, meaning that this biochemical analysis may be useful for real time diagnosis for clinical patients. This study is consistent with the findings of orthopedic research, which have shown that CS level increases and HA level decreases in SF of joints with OA. However, such clinical and experimental studies were done in knee and hip joints. There have been limited studies to evaluate CS and HA levels in human TMJ19–21. Those studies were done using SF lavage, and this remains a problem for a precise calibration of each parameter. Thus, an experimental study was indispensable in order to confirm the correlationship between CS and HA levels in TMJ SF and histological changes. Previous studies reported that C6S is the main CS of articular cartilage30. Thus, the C6S level in SF has been considered to represent the degree of matrix degradation. This coincides with the results of this study. However, recent studies suggest that biosynthesis of CS may increase in the early stage of OA in order to attempt repair, but if this attempt fails then subsequent OA may result11,18. Therefore it is still controversial whether the increased level
Osteoarthritis and Cartilage Vol. 9, No. 4
Fig. 7. Correlation between sum level of chondroitin 6-sulfate and chondroitin 4-sulfate in TMJ synovial fluid and OA stage. A significant positive correlation was observed (Spearman’s rank correlation coefficient, r=0.868, P<0.0001).
369
Fig. 9. Correlation between HA level of in TMJ synovial fluid and OA stage. A significant negative correlation was observed (Spearman’s rank correlation coefficient, r= −0.692, P<0.01).
limitation to the simulation of human OA in experimental models. However, it is important to find the several evidences of articular pathology for the research of human OA. Certainly further investigation is necessary to define the precise correlationship between the levels of several synovial markers and ongoing intraarticular pathology in the TMJ. However, from this study it can be concluded that C4S, C6S and HA are strongly correlated with intraarticular pathology in the TMJ.
References
Fig. 8. Correlation between ratio of C6S to C4S in TMJ synovial fluid and OA stage. A significant positive correlation was observed (Spearman’s rank correlation coefficient, r=0.828, P<0.0001).
of CS represents the ongoing articular cartilage degradation or the result of hypersynthesis to an attempt at repair17. On the other hand, HA is mainly secreted by synoviocytes in synovial membrane. In this study HA concentration decreased in SF in accordance with the OA progress. This finding coincides with previous reports31,32. In this pathological condition, however, it is still controversial whether biosynthesis of HA from synovial membrane decreased or body fluid infiltration reduces HA concentration. Recently, a clinical cross-sectional study showed decreased levels of HA and GAG in OA compared with normal11. Further studies are necessary to examine the calibrated CS and HA levels by total GAG or total protein and to evaluate the correlationship between those levels in TMJ SF and histology. Thus, at present, clinical data from TMJ SF must be considered with caution. However, in OA intraarticular friction may occur, causing cartilage damage. Intraarticular matrix degradation then gradually develops, and a vicious circle is formed. This is an essential difference of human OA from experimentally induced OA. Thus, there is a
1. Ishimaru J-I, Kurita K, Ogi N, Handa Y, Toida M. Effect of tizanidine for a primary care of patients with myofacial pain. A preliminary study. Hosp Dent Oral Maxillofac Surg 1995;7:2–7. 2. Ishimaru J-I, Kurita K, Handa Y, Goss AN. Temporomandibular joint osteoarthritis. Literature review and experimental animal models. Hosp Dent Oral Maxillofac Surg 1991;3:64–8. 3. Muir C, Goss AN. The radiological morphology of asymptomatic temporomandibular joints. Oral Surg 1990;70:355–59. 4. Goss AN, Bosanquet AG. The arthroscopic appearance of acute temporomandibular joint trauma. J Oral Maxillofac Surg 1990;48:780–83. 5. Milan SB, Schmitz JP. Molecular biology of temporomandibular joint disorders: Proposed mechanism of disease. J Oral Maxillofac Surg 1995;53:1448–54. 6. Kubota E, Imamura H, Kubota T, Shibata T, Murakami K. Interleukin 1 and stromelysin(MMP-3) activity of synovial fluid as possible markers of osteoarthritis in the temporomandibular joint. J Oral Maxillofac Surg 1997;55:20–7. 7. Israel HA, Saed-Nejad F, Ratcliffe A. Early diagnosis of osteoarthrosis of the temporomandibular joint: correlation between arthroscopic diagnosis and keratan sulfate levels in the synovial fluid. J Oral Maxillofac Surg 1991;49:708–11. 8. Miyamoto K, Ishimaru J-I, Tatematsu N, Kurita K, Ogi N, Handa Y, et al. Biochemical analysis of temporomandibular joint synovial fluid. Literature review and
370
9.
10. 11.
12. 13.
14.
15.
16.
17. 18.
19.
20.
J.-I. Ishimaru et al.: Normal, early and advanced osteoarthritic sheep temporomandibular joints experimental animal models. Hosp Dent (Tokyo) 1995;7:44–50. Ratcliffe A, Shurety W, Caterson B. The quantitation of a native chondroitin sulfate epitope in synovial fluid lavages and articular cartilage from canine experimental osteoarthritis and disuse atrophy. Arthritis Rheum 1993;36:543–51. Plaas AH, West LA, Wong-Palms S, Nelson FRT. Glycosaminoglycan sulfation in human osteoarthritis. J Biol Chem 1988;15:12642–9. Belcher C, Yaqub R, Fawthrop F, Bayliss M, Doherty M. Synovial fluid chondroitin and keratan sulfate epitopes, glycosaminoglycans, and hyaluronan in arthritic and normal knees. Ann Rheum Dis 1997;56:299–307. Hardingham T. Changes in chondroitin sulfate structure induced by joint disease. Acta Orthop Scand 1995;266:107–10. Yoshihara Y, Yamada H, Miyauchi S, Ito S, Samura A, Shinmei M. Levels of chondroitin 4-sulfate, chondroitin 6-sulfate and carboxy-terminal type II procollagen peptide in knee synovial fluid after injury to the anterior cruciate ligament. Ryumachi 1996;36: 734–40 (in Japanese). Yamada H, Miyauchi S, Hotta H, Morita M, Yoshihara Y, Kikuchi T, et al. Levels of chondroitin sulfate isomers in synovial fluid of patients with hip osteoarthritis. J Orthop Sci 1999;4:250–4. Shinmei M, Miyauchi S, Machida A, Miyazaki K. Quantitation of chondroitin 4 sulfate and chondroitin 6-sulfate in pathologic joint fluid. Arthritis Rheum 1992;35:1304–8. Shinmei M, Kobayashi T, Yoshihara Y, Samura A. Significance of the levels of carboxy terminal type II collagen peptides, chondroitin sulfate isomers, tissue inhibitor of metalloproteinases, and metalloproteinases in osteoarthritis joint fluid. J Rheumatol Suppl 1995;43:78–81. Hardingham T. Changes in chondroitin sulfate and joint disease. Osteoarthritis Cart 1998;6(Suppl A):3–5. Uebelhart D, Thonar EJMA, Zhang J, Williams JM. Protective effect of exogenous chondroitin 4,6-sulfate in the acute degradation of articular cartilage in the rabbit. Osteoarthritis Cart 1998;6(Suppl A):6–13. Murakami KI, Shibata T, Kubota E, Maeda H. Intraarticular levels of prostaglandin E2, hyaluronic acid, and chondroitin-4 and-6 sulfates in the temporomandibular joint synovial fluid of patients with internal derangement. J Oral Maxillofac Surg 1998;56:199–203. Ratcliffe A, Israel HA, Saed-Nejad F, Diamond B. Proteoglycans in the synovial fluid of the temporo-
21.
22. 23. 24.
25.
26.
27.
28.
29.
30. 31.
32.
mandibular joint as an indicator of changes in cartilage metabolism during primary and secondary osteoarthritis. J Oral Maxillofac Surg 1998;56:204–8. Shibata T, Murakami KI, Kubota E, Maeda H. Glycosaminoglycan components in temporomandibular joint synovial fluid as markers of joint pathology. J Oral Maxillofac Surg 1998;56:209–13. Ishimaru J-I, Goss AN. A model for osteoarthrosis of the temporomandibular joint. J Oral Maxillofac Surg 1992;50:1191–5. Ishimaru J-I, Handa Y, Kurita K, Goss AN. The effect of marrow perforation in sheep temporomandibular joint. Int J Oral Maxillofac Surg 1992;21:239–42. Bosanquet AN, Ishimaru J-I, Goss AN. The effect of experimental disc perforation in sheep temporomandibular joints. Int J Oral Maxillofac Surg 1991;20:177–81. Ogi N, Kurita K, Handa Y, Goss AN. The short-term effect of discectomy on the osteoarthrotic temporomandibular joint in sheep. Int J Oral Maxillofac Surg 1996;25:319–24. Ogi N, Kurita K, Handa Y, Goss AN. Short-term effect of fresh disk allograft on the osteoarthrotic sheep temporomandibular joint. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 1997;84:259–64. Ogi N, Kurita K, Handa Y, Goss AN. The short-term effect of autogenous auricular cartilage graft following discectomy on the osteoarthrotic temporomandibular joint in sheep. Int J Oral Maxillofac Surg 1997;26:217–22. Neo H, Ishimaru J-I, Kurita K, Goss AN. The effect of hyaluronic acid on experimental temporomandibular joint osteoarthrosis in the sheep. J Oral Maxillofac Surg 1997;55:1114–9. Toyoda H, Shinomiya K, Yamanashi S, Koshiishi I, Imanari T. Microdetermination of unsaturated disaccharides produced from chondroitin sulfates in rabbit plasma by high performance liquid chromatography with fluorometric detection. Anal Sci 1988;4:381–4. Mourao PAS. Distribution of chondroitin 4-sulfate and chondroitin 6-sulfate in human articular and growth cartilage. Arthritis Rheum 1988;31:1028–33. Balazs EA, Watson D, Duff IF, Roseman S. Hyaluronic acid in synovial fluid. Molecular parameters of hyaluronic acid in normal and arthritic human fluids. Arthritis Rheum 1967;10:357–76. Dahl LB, Dahl IMS, Engstrom-Laurent A, Granath K. Concentration and molecular weight of sodium hyaluronate in synovial fluid from patients with rheumatoid arthritis and other arthropathies. Ann Rheum Dis 1985;44:817–22.