Marked suppression by salicylate of the augmented proteoglycan synthesis in osteoarthritic cartilage.код для вставкиСкачать
83 MARKED SUPPRESSION BY SALICYLATE OF THE AUGMENTED PROTEOGLYCAN SYNTHESIS IN OSTEOARTHRITIC CARTILAGE MARSHALL J. PALMOSKI. ROBERT A. COLYER, and KENNETH D. BRANDT In osteoarthritis a net increase in proteoglycan synthesis has been noted until the disease is far advanced and presumably reflects an attempt by the chondrocyte to repair the defect in the cartilage matrix. Because salicylates are the agents most commonly employed in treatment of osteoarthritis and because we recently showed that IO-’M sodium salicylate (i.e., approximately 20 mg%) suppresses proteoglycan synthesis in normal canine knee cartilage in vitro, we have studied the effects of this compound on osteoarthritic knee cartilage from dogs whose anterior cruciate ligament had been transected 9 weeks previously. The data indicated that the augmented synthesis of glycosaminoglycans in the degenerating cartilage was suppressed to a much greater degree by IO-’M sodium salicylate than the lower level of glycosaminoglycan synthesis in control cartilage from the contralateral knee of the same animal. Uptake of ‘‘C-acetylsalicylic acid was increased about 35% in osteoarthritic cartilage, suggesting that the drug permeated it more readily than normal cartilage. The salicylate-induced suppression of proteoglycan synthesis in the osteoarthritic cartilage was From the Departments of Medicine and Orthopaedic Surgery, Indiana University School of Medicine, Indianapolis, Indiana. Supported in part by N I H grant #5S07 R R 05371, NIAMDD grant #AM 20582, and awards from the Arthritis Foundation and the Grace M. Showalter Trust. Marshall J. Palmoski, PhD: Assistant Professor of Medicine; Rohert A. Colyer, MD: Assistant Professor of Orthopaedic Surgery; Kenneth D. Brandt, MD: Professor of Medicine and Chief, Rheumatology Division, Indiana University School of Medicine. Address reprint requests to Marshall J. Palmoski, PhD, Rheumatology Division, Indiana University School of Medicine, 1 100 W. Michigan Street, Indianapolis, Indiana 46223. Submitted for publication May 7, 1979; accepted in revised form August 10, 1979. Arthritis and Rheumatism, Vol. 23, No. 1 (January 1980) not accompanied by reversal of the defect in proteoglycan aggregation or by improvement in the (presumed) defect in proteoglycan-collagen interaction in the matrix, as reflected by the abnormally high proportion of 3sS-proteoglycanspresent in the culture medium. The extracellular matrix of normal articular cartilage is constituted principally of two species of macromolecules--collagen and proteoglycans (PGs). The latter, which have average molecular weights of 1-2 x lo6 daltons (I), possess a high fixed negative charge density, which renders them highly hydrophilic (2). There is evidence that the PG content of joint cartilage accounts for its elasticity and ability to resist compression (3), while the collagen confers tensile strength (4).In normal articular cartilage the proteoglycans exist chiefly in very large aggregates, in which many individual PGs are noncovalently associated with hyaluronic acid in a linkage stabilized by tissue glycoproteins (5). The role of aggregation is unknown but it may, by facilitating formation of macromolecular structures of enormous size (Swzoo= 60-70) (6), help constrain the PGs within the collagen network of the tissue. In osteoarthritis (OA) the articular cartilage matrix is deficient. Its content of proteoglycans is subnormal and its compressive stiffness is diminished (7). This loss of PGs, at least in the earlier stages of the disease, occurs despite an increase in net synthesis of PGs (which presumably reflects an attempt by the chondrocytes to “repair” the lesion) (S), suggesting an accelerated rate of PG breakdown. A defect in PG aggregation (9,lO) and an increase in water content (1 1) of the articular cartilage commonly occur in the initial stages of osteoarthritis. The increased water content, which may PALMOSKI ET AL 84 be of fundamental importance, can be accounted for by enlargement of the molecular domains of the PGs, which are presumably less constrained than normal by the collagen network (12). It is notable that salicylates, the drugs most commonly employed in treatment of this disease, inhibit synthesis of glycosaminoglycans (GAGS) (13,14). We recently reported that sodium salicylate (SodSal), when introduced into the culture medium at a concentration of 10-3M(which is readily achieved in the serum of patients treated with salicylates for rheumatic disease), reduced PG synthesis in cultures of normal adult canine articular cartilage by about 30%, but had no effect on PG catabolism or aggregation (15). The symptomatic relief produced by salicylates in many patients with osteoarthritis may be attributed to the analgesic and/or antiinflammatory properties of these agents. Do these drugs have other effects on the disease, which might result from a direct action on articular cartilage, e.g., on PG catabolism or aggregation, or on the stability of PG-collagen interactions? The answer is not clear at this time. Aspirin has been reported to inhibit the cartilage degeneration occurring from superficial scarification (16) or prolonged joint compression (17) in rabbits, and to reduce the prevalence of degenerative changes in cartilage of human beings suffering recurrent dislocation of the patella (1 8). However, the cartilage degeneration in C57 black mice, which are genetically predisposed to OA, has been shown to be aggravated by administration of salicylates (19). Clearly, more information is needed about the effects of salicylates on normal and osteoarthritic cartilage. The present study examines the effect of SodSal on articular cartilage from dogs who had undergone transection of their anterior cruciate ligament. This surgical procedure produces instability of the knee and leads predictably to the morphologic and biochemical changes of OA (20). Our data indicate that the inhibition of net PG synthesis produced by SodSal in the degenerating cartilage is more marked than that produced by the drug in normal cartilage, while the PG aggregation defect and the greater than normal extractability of PGs from the matrix, which are integral features of the disease, are not improved by the drug. MATERIALS AND METHODS Surgical procedure. Five adult mongrel dogs were anesthetized with intravenous sodium pentothal, after which the left knee was shaved and scrubbed with Betadine antiseptic soap. Its joint capsule was opened through a short oblique anteromedial incision and the anterior cruciate ligament was transected with a scalpel, with care taken to avoid damage to the articular cartilage. The capsule was closed with 2-0 chromic sutures and the skin with interrupted 4-0 nylon. Two dogs underwent sham operations i.e., the joint capsule was opened, the anterior cruciate ligament was identified but not transected, and the capsule and skin were closed as above. Postoperatively all animals were allowed to ambulate freely in pens large enough to allow walking. All dogs bore weight on the operated limb by the third day after surgery. None developed a wound infection. Tissues and culture conditions. Nine weeks postoperatively the dogs whose anterior cruciate ligaments had been transected (dogs 1-5) were killed with an overdose of sodium pentothal. The sham-operated dogs were killed in the same fashion 7 weeks (dog S1) and 12 weeks (dog S2) after surgery. Immediately after killing, both knees were opened aseptically. The distal femurs were removed with a bone saw and representative full-thickness samples of the articular cartilage, including a portion of the underlying subchondral bone, were obtained with a Craig biopsy needle for histologic study. Other portions (approximately 10 mg) were taken for determination of dry weight and uronic acid content, while the remainder of the cartilage from the distal weight-bearing portions of the femoral condyles of each joint (approximately 200 mg) was shaved into slices less than 0.5 mm thick, with the tissue from the operated knee and contralateral control knee of each dog cultured separately. Approximately 50 mg of cartilage were incubated for 24 hours at 37OC under 5% C02:95%air in 10 ml of medium consisting of Ham's F-12 nutrient mixture, pH 7.4, (containing approximately 5 X 10-6M inorganic SO,); fetal calf serum (10%); streptomycin (50 pg/ ml); penicillin (50 units/ml) and no SodSal or 10-3M SodSal. At the end of the incubation period the medium was removed with a pasteur pipette and replaced with fresh medium containing Na25S0, (20 pCi/ml) (New England Nuclear Corp., Boston, Massachusetts) and the same concentration of SodSal. The incubation was then continued for an additional 8 hours. The medium was then decanted and the cartilage washed with cold saline. Medium and wash were combined and dialyzed for 48 hours at 4°C against 200 volumes of 0.05M sodium acetate, pH 6.8, in Spectrapor No. 3 dialysis tubing (Spectrum Medical Industries, Inc.; Los Angeles, California), which has a molecular weight cutoff of approximately 3,500 daltons. The tubing was rinsed with distilled water and the retentate and rinses were combined. Extraction of PGs from the cartilage. The washed cartilage slices were stirred for 48 hours at 4OC in 10 ml of 4M guanidinium hydrochloride (GuHC1) in 0.05M sodium acetate, pH 5.8, containing the protease inhibitors ethylenediaminetetraacetic acid (O.OlM), 6-aminohexanoic acid (0. lM), and benzamidine hydrochloride (0.OSM) (21). The extract was filtered, the residue was washed with fresh 4M GuHC1, and filtrate and washings were combined and dialyzed against 0.05M sodium acetate, pH 6.8, as above. After extraction with 4M GuHCl the cartilage residue was suspended in 0.2M borate, pH 8.0, containing 0.02M calcium chloride and pronase (Calbiochem; San Diego, California) (1 mg enzyme/50 mg cartilage wet weight). The tissue SALICYLATE SUPPRESSION OF PG SYNTHESIS was digested under toluene for 24 hours at 55"C, after which the small amount of tissue remaining was removed by centrifugation. The supernatant was dialyzed overnight at 4°C against distilled water and aliquots (0.1 ml) of the dialyzed samples of the medium recovered at the end of the incubation period, the 4M GuHCl extract and the pronase digest were added to Ready-Solve HP 60 (Beckman Instruments, Inc., Irvine, California) and counted in a Beckman liquid scintillation spectrometer. Results (cpm) were adjusted for differences in wet weight of the cartilage. Isolation of PGs. After dialysis of the 4M GuHCl extract, PG aggregates were isolated and purified by cesium chloride equilibrium density gradient centrifugation under associative conditions in the presence of 0.4M GuHCl, according to the method of Hascall and Sajdera (22). Fractions from the bottom 2/5 of the density gradient (h 2 1.76 gm/ml) were combined and dialyzed exhaustively at 4°C against 0.05M sodium acetate, pH 6.9, to yield fraction AGU. Digestion of fraction AGuwith hyaluronic acid /31-+3 hydrolase. Aliquots (2 ml) of fraction AGuwere made 0.1M with respect to citric acid a n d 0.2M with respect to Na,HPO, ' 7 H,O. After the pH was adjusted to 5.6, 1.5 pg of hyaluronic acid /?l -+ 3 hydrolase (Biotrics, Inc., Arlington, Massachusetts) were added in 5 equal portions over a 5 hour incubation period at 37°C (23). Sepharose 2B chromatographs of the proteoglycans before and after incubation were compared. The specificity of the enzyme for hyaluronic acid was confirmed by its failure to alter the elution profile of disaggregated 35Sproteoglycans after they had been treated as described in the previous section. Determination of net GAG synthesis. To assess the effect of SodSal on GAG synthesis, triplicate samples (approximately 15 mg each) of cartilage from control and operated limbs of each dog were incubated for 16 hours in 5 ml of medium (see above) containing no SodSal or 10-3M SodSal. Na,35S0, (20 pCi/ml) was then added and the incubation was continued for an additional 8 hours after which the medium was decanted and the tissues were washed with cold saline. The medium and wash were combined and dialyzed against distilled water at 4"C, and the tissue was digested with pronase as above. Net GAG synthesis was determined from the sum of the nondialyzable '%04 radioactivities of the medium and the pronase digest. ''C-acetylsalicylic acid uptake by the cartilage. In one set of experiments triplicate samples of cartilage from the unstable and control knees of Dog 4 and Dog 5 were each incubated as above in medium containing lOP3Macetylsalicylic acid plus 2 pCi/ml of acetylsalicylic acid, ~arboxyl-'~C (New England Nuclear Corp., Boston, Massachusetts). At the end of the incubation period the medium was removed with a pasteur pipette and the tissues were washed with three 5 ml aliquots of fresh unlabeled medium. The cartilage was digested with pronase as above and 0. Iml aliquots of the digests were counted in a liquid scintillation spectrometer, with the results (cpm) adjusted for differences in the wet weight of the cartilage. Gel chromatography. Samples (0.5 ml) of proteoglycans in 0.5M sodium acetate, pH 6.9, were applied to a column (95 X 1.0 cm) of Sepharose 2B (Pharmacia Fine Chemicals, Piscataway, New Jersey) and eluted with the ace- 85 tate buffer at a rate of 2 ml/hour. The radioactivity ( 3 5 S 0 , ) of 1 ml effluent fractions was determined and the partition coefficients (Kav)of PG samples were calculated from the formula: K,, = (V, - V,)/(V, - VJ. V, was taken as the peak fraction in the elution diagram, V, as the void volume, and V, as the total column volume. Analytical methods. To determine water content, portions of the cartilage (approximately 10 mg) were placed in two changes of acetone for 48 hours and dried to constant weight in vacuo at 80°C. The dried cartilage was digested with pronase as above, after which the GAGS were isolated by precipitation with 9-aminoacridine hydrochloride and converted to their sodium salts with Bio-Rad AG-50 (Na+) (24). The resin was removed by filtration and the uronic acid content of the filtrate was determined by the method of Bitter and Muir (25). Histologic examination. Histologic sections (6p) of the samples of articular cartilage and subchondral bone obtained with the Craig biopsy needles from the central portion of the medial condyle were examined microscopically after staining with Safranin-0-fast green (26). RESULTS Gross and histologic observations. Articular cartilage from the control knee of every animal was white, glistening, and smooth, with an intact surface. No osteophytes were noted. Safranin-0 staining of the matrix was normal and intense, and cell distribution was normal, with only one chondrocyte per lacuna. The tidemark was intact in every case. The femoral condyles of the knees, which had been rendered unstable by previous cruciate ligament transection, exhibited apparently normal cartilage over the weight-bearing areas, with no evidence of softening, pitting, or discoloration. Surface fibrillation was noted only in the sample from Dog 3 (Table 1). In contrast to the controls, marginal osteophytes were apparent in every case. All samples of articular cartilage from the operated knees showed a diminution in intensity of Safranin-0 staining when compared to cartilage from the control knees (Table 1). Chondrocyte clusters, containing up to 5 cells per lacuna, were seen near the tidemark in all sections obtained from the unstable knees, although the tidemark was intact in all cases. Water and uronic acid content of the cartilage. Water constituted 66.2-79.3% of the total weight of the articular cartilage from the control knees (mean, 69.5%), while uronic acid represented 5.0-6.8% of the dry weight (mean, 5.7%). In every case cartilage from the unstable knee showed a higher water content and lower uronic acid content than that from the control knee of the same animal (Table 1). The increase in water content ranged from 3.34.1% (mean, 3.6%) of the tissue PALMOSKI ET AL 86 Table 1. Osteoarthritic changes in knee cartilage of dogs whose anterior cruciate ligament had been transected 9 weeks previously Knee from which cartilage was obtained Safranin-0 staining Fibrillation Dog 1 Control Operated Normal Moderate reduction - - + + - 66.2 10.3 5.2 3.4 Dog 2 Control Operated Normal Slight reduction - - - 13.5 16.9 5.0 4.1 Dog 3 Control Operated Normal Severe reduction - - - 63.8 61.4 5.8 2.9 Dog 4 Control Operated Normal Slight reduction - - - 64.4 68.2 5.1 4.1 Dog 5 Control Operated Normal Slight reduction - - - 19.3 82.6 6.8 5.9 Animal + Chondrocyte cluster + + + + Osteophytes + + + + Uronic Water acid content* content? * Percent of tissue wet weight. t Percent of tissue dry weight. wet weight. The uronic acid content of cartilage from the unstable knees ranged from 2.9-5.9% of the tissue dry weight (mean, 4.2%) which represented an average decrease in uronic acid content of about 25%, with a maximum decrease of 50% (Dog 3), in comparison with the control cartilage (Table 1). GAG metabolism. Based on nondialyzable "S radioactivity measured after pronase digestion of the cartilage, plus that contained in the culture medium, the amount of newly synthesized GAGS in articular cartilage from the unstable knees was greater (164-308%) than that in cartilage from control knees (Table 2). SodSal lO-'M invariably reduced GAG synthesis in control cartilage (range, 21-55%; mean, 40.4%) (Table 2). In osteoarthritic cartilage the augmented (baseline) level of GAG synthesis was reduced by SodSal in every case. Notably, except for the experiment with cartilage from Dog 1, the SodSal-induced inhibition of GAG synthesis (40-99%) in the osteoarthritic cartilage was much greater than that in control cartilage from the same animal (Table 2). Salicylate uptake by the tissue. In experiments in which cartilage from Dogs 4 and 5 was incubated with 14C-acetylsalicylic acid, uptake of the radionuclide by degenerating cartilage from the unstable knee was 35% and 37% greater, respectively, than that by cartilage from control knees of these animals (Table 3). Extractability of PGs from the cartilage. There was no effect from lO-'M SodSal on the distribution of "S PGs between the culture medium, 4M GuHCl ex- tract, and pronase digest of the cartilage (Table 4). In every case the proportion of total 3sSPGs contained in the medium after incubation of cartilage from the unstable knee was greater than that in the medium of cultures of control cartilage. Thus, some 7.9-13% of the total nondialyzable 3sS radioactivity was recovered in medium of the control cultures and 13.2-23.7% in medium of cultures of the degenerating cartilage. In all cases the majority of the '%O, proteoglycans were extracted with 4M GuHCl (Table 4), and fraction AGuaccounted for about 70-80% of the total nondialyzable radioactivity in the 4M GuHCl extracts of control and osteoarthritic cartilage. SodSal did not alter the distribution of 35SPGs between the medium, 4M GuHCl extract, and the pronase digest in cultures of normal or osteoarthritic cartilage, nor the proportion of total "S PG radioactivity in the 4M GuHCl extract which was recovered in fraction AGu. PG aggregation. The proteoglycans in fraction AGu from control cartilage were markedly heterogeneous in hydrodynamic size and some 23-38% eluted in the void volume on Sepharose 2B chromatography. The K,, of those PGs from fraction A,, of control cartilage which were small enough to be retarded by the gel ranged from 0.20-0.49 (Table 5). In every case the retarded peak was symmetrical, broad, and smooth. In contrast to this finding, proteoglycans in fraction AGufrom the degenerating cartilage contained no material sufficiently large in hydrodynamic size to elute in the Sepharose 2B void volume. The K,,'s of the PGs SALICYLATE SUPPRESSION OF PG SYNTHESIS 87 Table 2. Effect of sodium salicylate (SodSal) on %04incorporation into glycosaminoglycans (GAGS)* in knee cartilage of dogs 9 weeks after transection of anterior cruciate ligament % inhibition ’3 incorporation 35Scpm x 10-~/10 Source of cartilage Dog I Dog 2 Dog 3 Culture conditions mg wet weight of cartilage Control Control SodSal Operated Operated SodSal + + 229.5 f 15.6 129.2 f 3.0 377.2 f 17.0 226.6 f 10.5 164 Control Control + SodSal Operated Operated + SodSal 193.9 f 10.0 87.2 f 6.6 403.3 f 24.3 100.8 f 7.1 208 Control Control + SodSal Operated Operated SodSal 221.6 f 12.5 106.4 f 2.3 683.6 f 21.9 7.4 f 0.3 308 Control Control + SodSal Operated Operated SodSal 264.7 f 17.5 209.8 f 7.2 481.1 f 19.8 182.1 f 5.8 182 Control Control + SodSal Operated Operated SodSal 272.7 f 8.7 191.7 f 8.4 479.0 f 24.5 198.6 f 9.5 176 + Dog 4 + Dog 5 into GAGs in cartilage from unstable kneet + of 35s incorporation into GAGs produced by IO-’M SodSal 44 40 55 7s 52 99 21 62 30 59 * Nondialyzable 35Scpms in culture medium plus pronase digests of triplicate samples f one standard error. t Percent of central knee, in absence of SodSal. retarded by the gel were somewhat larger than the K,,’s of analagous PGs from control cartilage of these animals, indicating that they tended to be smaller than normal in hydrodynamic size. Sepharose 2B elution profiles of fraction AGufrom the osteoarthritic and control cartilage were wholly unaffected by the presence of SodSal in the culture medium (Table 5). Treatment with hyaluronic acid pl+3 hydrolase abolished the Sepharose 2B void volume peaks in all Table 3. Uptake of I4C acetylsalicylic acid by normal and osteoarthritic canine knee cartilage Animal Dog 4 Dog 5 Knee from which cartilage was obtained I4C cpm/lo mg wet weight of cartilage* 14Ccpm incorporation, % of control Control Operated 11,184 f 671 15,088 -+ 603 135 Control Operated 9,624 f 485 13,209 f 748 137 - - * Results represent the mean f one standard error of triplicate samples. A,, samples from control cartilage but had no effect on the chromatographs of AGufrom osteoarthritic cartilage, suggesting that the PGs in the latter fraction did not exist as aggregates in association with hyaluronic acid. Characterization of knee cartilage from dogs undergoing sham operation. Cartilage from the control and sham-operated joints of Dogs S1 and S2 appeared grossly normal, and histologic sections stained with Safranin-0 showed no abnormality (Table 6). The water content of the cartilage from the sham-operated limbs was similar to that from the control knees (Dog S1 73.3%, 12.9% respectively; Dog S2 74.9%, 75.2% respectively, of tissue wet weight), and no appreciable difference was observed between the uronic acid contents of the cartilage from the control and sham-operated knees (Table 6). incorporation into GAGs by cartilage from the operated joints of these dogs was 103% and 98% of controls, respectively. No difference was noted between the average hydrodynamic size of the PGs in fraction AGufrom the sham-operated knees and that of the AGu PALMOSKI ET AL 88 Table 4. Effect of sodium salicylate (SodSal) on distribution of 35Sproteoglycans* between culture medium, 4M GuHCl extract, and pronase digest of normal and osteoarthritic canine knee cartilage No SodSal Animal Knee from which cartilage was obtained 10-3M SodSal % distribution of total cpm Medium 4M GuHCl Pronase digest % distribution of total cpm Medium 4M GuHCl Pronase digest Dog I Control Operated 13.0 23.7 56.0 59.8 31.0 16.5 13.0 21.3 60.2 59.0 26.8 19.7 Dog 2 Control Operated 8.3 19.3 65.7 61.5 26.0 19.2 10.8 20.8 58.3 62.2 30.9 17.0 Dog 3 Control Operated 9.1 22.2 66.4 44.8 24.5 33.0 7.0 21.1 60.2 47.9 32.8 31.0 Dog 4 Control Operated 11.1 21.1 66.3 56.5 22.6 22.4 9.2 20.2 69.3 57.9 21.5 19.9 Dog 5 Control Operated 7.9 13.2 56.7 51.4 35.4 35.4 8.4 14.5 57.9 48.2 33.7 37.3 * Nondialyzable "S radioactivity. from the controls. In all cases about 30% of the "S-PGs eluted in the Sepharose 2B void volume. Digestion with hyaluronic acid pl+3 hydrolase eliminated the void volume peaks of fraction A,". the cartilage of the operated knees in the present study can be attributed to the consequences of knee instability, rather than to the trauma of the operative procedure. The cartilage from the knee that did not undergo DISCUSSION The experimental model of osteoarthritis emplayed here has been studied extensively by us and produces the structural. biochemical, and metabolic changes that occur in the spontaneously occurring form of the disease (20). The model is similar to that described by Pond and Nuki (27), except that in our studies an arthrotomy is performed and the anterior cruciate ligament is transected under direct vision, whereas in the Pond-Nuki model the ligament is cut by a blind stab wound. Our own attempts at blind sectioning of the ligament have occasionally resulted in laceration of articular cartilage and/or incomplete transection of the ligament. Since arthrotomy itself could conceivably produce some degenerative changes in cartilage, we have examined knee cartilage from dogs undergoing sham operations, in which arthrotomy was performed and the joint capsule closed without transection of the ligament. At a postoperative interval relevant to the present study (7-12 weeks) the cartilage in these shamoperated dogs was morphologically, metabolically, and biochemically normal (Table 6), so that the changes in Table 5. Effect of sodium salicylate (SodSal) on the Sepharose 2B elution profiles of proteoglycans extracted with 4M GuHCl from canine knee cartilage 9 weeks after sectioning of anterior cruciate ligament No SodSal Animal Knee from which cartilage was obtajned Fraction A,, 10-3M SodSal Fraction A,, Dog I Control Operated 23 0 0.20 0.30 22 0 0.24 0.30 Dog2 Control Operated 32 0 0.35 0.56 35 0 0.38 0.57 Dog3 Control Operated 30 0 0.45 0.58 28 0 0.42 0.55 Dog4 Control Operated 38 0 0.49 0.44 40 0 0.47 0.47 Control Operated 25 0 0.38 0.49 28 0 0.35 0.47 Dog 5 * Percent of 35S-proteoglycans eluting in the Sepharose 2B void volume. t Partition coefficient of 35S-proteoglycansretarded by Sepharose 2B. SALICYLATE SUPPRESSION OF PG SYNTHESIS 89 Table 6. Characterization of knee cartilage from dogs undergoing sham operation Animal Dog S I Dog S2 Interval between shamKnee from operation which cartilage and was obtained sacrifice Safranin-0 staining Fibrillation Osteophytes - - - - - - - - Control Operated 7 Normal Normal Control Operated 12 Normal Normal Water Uronic acid content, content, 35S cpm,* % of tissue % of % of wet weight dry weight control 12.9 13.3 4.5 4.1 75.2 74.9 5.3 5. I - I03 - 98 Sepharose 2B chromatography of fraction AGu %Vat k,,, Vr* 29 30 0.20 0.20 25 28 0.32 0.34 * Nondialyzable ”S in culture medium plus pronase digests. Results represent analyses of triplicate samples. Standard error was less than 6% in each case. t Percent of 35S-proteoglycanseluting in the Sepharose 2B void volume. Partition coefficient of 35S-proteoglycanswhich were retarded by Sepharose 2B. * surgery may not have been “normal,” since it was presumably subjected to excessive loading because of the instability induced in the opposite limb. Indeed, we found that some decrease in PG aggregation occurred in cartilage from the left knee of dogs when the right hind limb was immobilized in a cast for several weeks (28). However, in the present study, PGs from the unoperated limb were aggregated and the tissue appeared histologically and histochemically normal. The histologic changes in the cartilage of the unstable knee, i.e. the decrease in Safranin-0 staining, clustering of chondrocytes and, in one case, fibrillation (Table l), as well as the osteophytes, are typical of osteoarthritis (29). The increase in cartilage water (1 l), decrease in uronic acid content (8,11,26), and the decrease in PG aggregation, which was indicated by Sepharose 2B chromatography (9, lo), are also characteristic of this disease. A decrease in PG aggregation has also recently been demonstrated by differential ultracentrifugation in this animal model of OA (30,31). The observed increase in net GAG synthesis in cartilage from the unstable knee is also a well recognized feature of the earlier stages of osteoarthritis (8) and may reflect attempts by the cartilage cells to replenish the deficit in matrix proteoglycans. The concentration of SodSal chosen in the present study, 10-3M, corresponds to a therapeutic concentration of approximately 20 mg% and is readily achieved in human serum and synovial fluid (32). Our results are significant insofar as they show that at this reasonable concentration the drug-induced inhibition of GAG synthesis in OA Cartilage was appreciably greater than that in normal cartilage. It is possible that this difference resulted from a greater penetration of salicylate into the diseased cartilage, whose matrix integrity had been altered (as indicated, e.g., by the increased loss of PGs into the medium). This possibility is strengthened by the evidence that uptake of ’‘C-acetylsalicylic acid by osteoarthritic cartilage was greater than that by control cartilage from the same animal. Indeed, cartilage from the unstable knee of Dog 3, which showed the most severe histologic and histochemical abnormalities in the series (Table l), exhibited the greatest degree of salicylate-induced inhibition of GAG synthesis (99% suppression). Although SodSal inhibited net GAG synthesis, it did not reverse the defect in PG aggregation in the osteoarthritic cartilage or improve the organizational integrity of the matrix. Thus, the abnormally high proportion of proteoglycans present in the culture medium after incubation of cartilage from the unstable joints was not diminished by inclusion of SodSal in the cultures. The present studies provide no information concerning whether cartilage breakdown in osteoarthritis is altered by SodSal. The cartilage activity of acid and neutral proteases, which degrade PGs, is greater in OA cartilage than in normal (33,34) so that if the drug inhibited release of enzymes or suppressed their activity, it could have a beneficial effect that might counterbalance its suppression of GAG synthesis. The present studies also do not deal with the basis for the salicylate-induced suppression of GAG synthesis in normal and osteoarthritic cartilage. This may, however, be due to inhibition of enzymes involved in the early stages of GAG synthesis, which has been demonstrated at the salicylate concentration employed in the present study (35). Finally, it must be emphasized that the results reported here represent data from short- PALMOSKI ET AL 90 term in vitro experiments. T h e effects of salicylate on normal and osteoarthritic cartilage in vivo, if any, remain to be clarified. Studies aimed at clarifying this point are in progress in this laboratory. 14. ACKNOWLEDGMENTS We are grateful to Janet Anderson and Jeffrey Wilson for technical assistance and to Margaret Britner for secretarial support. 15. REFERENCES 16. 1. Rosenberg L, Hellman W, Kleinschmidt AK: Macromo- 17. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. lecular models of protein polysaccharides from bovine nasal cartilage based on electron microscopic studies. J Biol Chem 245:41234130, 1970 Maroudas A: Physico-chemical properties of articular cartilage, Adult Articular Cartilage. Edited by MAR Freeman. London, Pitman Medical, 1973, pp 131-170 Kempson GE, Muir H, Swanson SAV, Freeman MAR: Correlations between the compressive stiffness and chemical constituents of human articular cartilage. Biochim Biophys Acta 215:70-77, 1970 Kempson GE, Muir H, Pollard C, Tuke M: The tensile properties of the human femoral condyles related to the content of collagen and glycosaminoglycans. Biochim Biophys Acta 297:456472, 1973 Hascall VC, Heinegard D: Aggregation of cartilage proteoglycans. I. The role of hyaluronic acid. J Biol Chem 249:4232-424 1, 1974 Rosenberg LC, Pal S, Beale RJ: Proteoglycans from bovine proximal humeral articular cartilage. J Biol Chem 248:3681-3690,1973 Kempson GE, Spivey CJ, Swanson SAV, Freeman MAR: Patterns of cartilage stiffness on normal and degenerate human femoral heads. J Biomechanics 4597-609, 1971 Mankin HJ: The reaction of articular cartilage to injury and osteoarthritis. N Engl J Med 291:1335-1310, 1974 Brandt KD, Palmoski MJ, Perricone E: Aggregation of cartilage proteoglycans. 11. Evidence for the presence of a hyaluronate binding region in proteoglycans from osteoarthritic cartilage. Arthritis Rheum 19:1308-13 14, 1976 Palmoski MJ, Brandt KD: Hyaluronate-binding by proteoglycans. Comparison of mildly and severely osteoarthritic regions of human femoral cartilage. Clin Chim Acta 70:87-95, 1976 Bollet AJ, Nance JL: Biochemical findings in normal and osteoarthritic articular cartilage. 11. Chondroitin sulfate concentration and chain length, water, and ash contents. J Clin Invest 45:1170-I 177, 1966 Maroudas A: Balance between swelling pressure and collagen tension in normal and degenerate cartilage. Nature 260:808-809, 1976 McKenzie LS, Horsburgh BA, Ghosh P, Taylor TKF: Ef- 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. fect of anti-inflammatory drugs on sulphated glycosaminoglycan synthesis in aged human articular cartilage. Ann Rheum Dis 35:487497, 1976 Bostrom H, Bersten K, Whitehouse MW: Biochemical properties of antiinflammatory drugs. 11. Some effects on s~lphate-’~S metabolism in vivo. Biochem Pharmacol 13:413420, 1964 Palmoski MJ, Brandt KD: Effects of salicylate on proteoglycan metabolism in normal canine articular cartilage in vitro. Arthritis Rheum 22:746-754, 1979 Simmons DP, Chrisman OD: Salicylate inhibition of cartilage degeneration. Arthritis Rheum 8:960-969, 1965 Roach JE, Tomblin W, Eyring EJ: Comparison of the effects of steroid, aspirin and sodium salicylates on articular cartilage. Clin Orthop 106:350-356, 1975 Chrisman OD, Snook GA, Wilson TC: The protective effect of aspirin against degeneration of human articular cartilage. Clin Orthop 84:193-196, 1972 Wilhelmi VG: Fordernde und hemmende Einflusse von Tribenosid und Acetylsalicylsaure auf die spontane Arthrose der Maus. Arzneim-Forsch 28: 1724-1726, 1978 McDevitt CA, Muir H: Biochemical changes in the cartilage of the knee in experimental and natural osteoarthritis in the dog. J Bone Joint Surg 58-B:94-101, 1976 Oegema TR, Hascall VC, Dziewiatkowski DD: Isolation and characterization of proteoglycans from the swarm rat chondrosarcoma. J Biol Chem 250:6151-6159, 1975 Hascall VC, Sajdera SW: Protein polysaccharide complex from bovine nasal cartilage: the function of glycoprotein in the formation of aggregates. J Biol Chem 244:23842396, 1969 Perricone E, Palmoski MJ, Brandt KD: Failure of proteoglycans to form aggregates in morphologically normal aged human hip cartilage. Arthritis Rheum 20: 1372-1380, 1977 Tsiganos CD, Muir H: Studies on protein-polysaccharides from pig laryngeal cartilage: extraction and purification. Biochem J 113:879-884, 1969 Bitter T, Muir H: A modified uronic acid carbazole reaction. Anal Biochem 4:330-340, 1962 Mankin JH, Dorfman H, Lippiello L: Biochemical and metabolic abnormalities in articular cartilage from osteoarthritic human hips. J Bone Joint Surg 53-A:523-537, 1971 Pond MJ, Nuki G: Experimentally induced osteoarthritis in the dog. Ann Rheum Dis 32:387-388, 1973 Palmoski M, Perricone E, Brandt K: Development and reversal of a proteoglycan aggregation defect in normal canine knee cartilage after immobilization. Arthritis Rheum 22:508-517, 1979 Sokoloff L: The pathology and pathogenesis of osteoarthritis, Arthritis and Allied Conditions. Eighth edition. Edited by JL Hollander, DJ McCarty. Philadelphia, Lea and Febiger, 1972, pp 1009-1031 SALICYLATE SUPPRESSION OF PG SYNTHESIS 30. McDevitt CA, Billingham MEJ, Muir H: A defect in proteoglycan aggregates in early experimental osteoarthrosis. Ann Rheum Dis 37:484, 1978 31. Tenenbaum J, Altman RD, Riskin W, Latta L, Howell DS: Biochemical and biomechanical properties of osteoarthritic dog cartilage. Arthritis Rheum 22:666, 1979 32. Rosenthal RK, Bayles TB, Fremont-Smith K: Simultaneous salicylate concentration in synovial fluid and plasma in rheumatoid arthritis. Arthritis Rheum 7: 103-109, 1964 33. Sapolsky AI, Altman RD, Woessner JF, Howell DS: The 91 action of cathepsin D in human articular cartilage on proteoglycans. J Clin Invest 52:624-633, 1973 34. Lahey P, Ehrlich MG, Treadwell BV, Mankin HJ: Neutral proteoglycanase activity vs. severity of osteoarthritis. Proceedings 25th Annual Orthopaedic Research Society Meeting, San Francisco, California, 1979, p 46 35. Lee KH, Spencer MR: Studies on the mechanics of action of salicylates. V. Effect of salicylic acid on enzymes involved in mucopolysaccharide synthesis. J Pharm Sci 58:464468. 1969 Rheumatoid Arthritis: Advances in Therapy and Pathophysiology The New Mexico Chapter of the Arthritis Foundation will hold its Thirteenth Annual Arthritis Symposium, “Rheumatoid Arthritis: Advances in Therapy and Pathophysiology,” March 14 and 15, 1980 at the Airport Marina Hotel in Albuquerque. Registration fee is $50 advance, $60 after March 1. For more information write Arthritis Foundation, PO Box 8022, Albuquerque, New Mexico 87198, or call (505) 265-1545. Romanian Conference of Rheumatology The Third Romanian Conference of Rheumatology will be held in Bucharest May 2 2 and 23, 1980. lmmunopathologic mechanisms, antiinflammatory drugs, and medico-surgical rehabilitation will be some of the subjects covered. The official languages of the conference will be Romanian, English, French, and Russian. Registration fee for U.S. participants is $50.00. For more information, contact Dr. I. Stroescu, General Secretary, The Union Societies of Medical Sciences, Romanian Society of Rheumatology, Progresului Str. 8, Bucharest 70754, Romania.