Healthy adult dogs were studied for a defect in proteoglycan aggregation by immobilizing one limb for varying periods of time. Immobilization for 6 days resulted in a 41% reduction in proteoglycan synthesis by articular cartilage from the restrained knee compared with the contralateral control knee. After 3 weeks of immobilization, proteoglycan aggregation was no longer demonstrable in cartilage from the constrained limb. The aggregation defect was rapidly reversible and aggregates were again normal size 2 weeks after removal of a cast that had been worn for 6 weeks.In normal articular cartilage most of the proteoglycans (PG) exist in large aggregates that are noncovalently linked to hyaluronic acid (HA) (1). In osteoarthritis (OA) an aggregation defect exists within the tissue, since a greater than normal proportion of the PGs is not aggregated and aggregates which are present tend to be smaller than normal (2-5).Aggregation defects similar to those of OA in

Proteoglycan aggregates isolated from normal bovine knee cartilage were larger than those from osteoarthritic cartilage of the same joints and appeared relatively more resistant to digestion with leech hyaluronidase. Incubation of proteoglycan subunits from the arthritic cartilage with hyaluronic acid resulted in marked aggregation, comparable in magnitude to that shown by subunits from normal cartilage. The results indicate that the hyaluronate-binding region of these proteoglycans was functionally intact and suggest that diminished aggregation of proteoglycans in osteoarthritic cartilage may be due to an abnormality in some other constituent of the aggregates.Cartilage matrix consists to a considerable extent of large proteoglycan aggregates, in which many proteoglycan molecules are noncovalently linked to highly polymerized hyaluronic acid (1). At least one tissue glycoprotein (2) is also present in these macromolecular complexes, and appears to stalbilize the proteoglycan-hyaluronic acid interaction (3). The complexes, which have a sedimentation coefficient (So20,w) of 60-70 S (4), by virtue of their size alone must exercise a significant effect on the physical integrity and biomechanical properties of the tissue.In osteoarthritis, softening and ultimate loss of cartilage matrix are apparent, although the basis for these morphologic changes is not clear. Recent studies have indicated that a proportion of the proteoglycans of osteoarthritic cartilage are much more readily extracted than normal by brief agitation of the tissue in iso-osmotic neutral salt ( 5 ) . Because this extraction procedure does not dissociate large aggregates, the results imply that in osteoarthritis some proteoglycans are less aggregated than normal. Furthermore, aggregates obtained after extraction of bovine osteoarthritic cartilage with 4.0 M guanidinium chloride were smaller in hydrodynamic size than aggregates from normal cartilage of the same joints (6).Proteoglycans of osteoarthritic cartilage have been shown to contain less chondroitin sulfate (7) and keratan sulfate (8) and shorter chains of chondroitin sulfate (9,lO) than normal, but whether these differences in structure relate to differences in proteoglycan aggregation is not known. In the current view, proteoglycans consist of a ) an inhomogeneous protein core, on which are situated a polysaccharide attachment region, to which most of the chondroitin sulfate and keratan sul-

The macromolecular organization of proteoglycans in morphologically and histochemically normal hip cartilage from aged humans has been studied. In contrast to findings in articular and nonarticular cartilage from was diminished because of a defect in the core protein of the proteoglycans resulting in an impaired ability of these molecules to interact with hyaluronic acid.other sources, most of the proteoglycans in these tissues did not exist in large aggregates. Treatment with hyaluronic acid pl 4 3 hydrolase failed to diminish the size of proteoglycans prepared under conditions favoring aggregation, a finding suggesting that they were not complexed with hyaluronic acid. Polyacrylamide gel electrophoresis failed to demonstrate the presence of link glycoproteins associated with the proteoglycans. After incubation in vitro witb hyaluronic acid, minimal augmentation of hydrodynamic size of the preparation occurred, an indication that hyaluronate-proteoglycan interaction had not taken place. These results suggest that proteoglycan aggregation

Proteoglycan aggregates (PGC) facilitated dissolution of 2% times more sodium urate crystals than nonaggregated proteoglycans. This effect of aggregates on urate solubility was abolished by digestion of the aggregates with hyaluronic acid B1 -+ 3 hydrolase. PGC, however, did not sustain urate concentrations in supersaturated solutions. Potassium urate was severalfold more soluble than sodium urate. In vivo, where they exist predominantly as their sodium salts in the extracellular connective tissue, proteoglycans may not markedly influence the solubility of sodium urate.Most patients with gout have serum and interstititial fluid urate concentrations greater than normal (1,2). Most hyperuricemic individuals, however, neither experience attacks of gouty arthritis nor develop deposits of monosodium urate (NaU) in their connective tissue (for example, tophi) (2). The factors involved in the

SUMMARY The urate concentration of the supernatant was greater after supersaturated solutions of sodium urate were incubated in a suspension of CM-Sephadex C-25 than in one of Sephadex G-25 Fungizone (125 pcg) were added to the resin suspension. The tubes were allowed to stand at 250C for 16 hours, after which the samples were centrifuged for 20 minutes at 3000 r.p.m. The supernatants were removed by pipetting and, after an aliquot was taken for bacteriological culture the urate content was determined. Each experiment was performed in triplicate. Results and discussionThe volume of the supernatant recovered after centrifugation of the tubes containing Sephadex G-25 was 1 * 6 ml (Table 1). Based on the bed volume of the gel (4-6 ml/g) (Pharmacia Fine Chemicals, 1969) all of the buffer not recovered in the supernatant (0.4 ml) could have been contained within the resin. This fraction (total solvent volume -supernatant volume) is hereafter designated Vi (while Vi conventionally denotes the inner volume of the gel, under the conditions of the present study Vi will include also the small volume of solvent between the centrifuged resin beads). No urate crystals were observed within or between the beads after a portion of the pelleted resin was frozen, sectioned (4pm) with a cryostat, mounted by air drying on microscope slides, and viewed by polarisation microscopy.Based on the measured urate concentrations (0-.74-0 76 mg/ml) and the volume (1 -6 ml) of the Sephadex G-25 supernatants their urate contents represented 73 % and 75 %, respectively, of the total urate when the resin had been equilibrated with the sodium and potassium buffers (Table 1). On the assumption that the urate concentration of the Sehadex G-25 V1 was identical to that of the supernatant (Fig. 1) it may be calculated that 18% of the total urate content was contained in Vi. A small proportion (7-9 %) of the total urate cannot be