Search

Type II collagen is a major component of cartilage where it is present at a high concentration, which is essential for the functional maintenance of the tissue. Therefore, any fundamental understanding of the physiology of cartilage tissue must include an understanding of the mechanism that allows the high level of expression of type II collagen gene, Col2a1, by chondrocytes. To this end, we developed a new reporter assay system based on the co-transfection of candidate enhancer elements and reporter construct into Swarm rat chondrosarcoma chondrocytes that allowed their stable expression. Using this system, we screened more than 70 kb of the Col2a1 gene and found an enhancer domain that is responsible for the regulation of its expression level. The domain is localized in intron 7, and consists of an 800-bp region that contains within it a previously unidentified domain, ∼140 bp in size.

Follicular dendritic cell-secreted protein (FDC-SP) is a small secretory protein having structural similarities to statherin, a protein in saliva thought to play a role in calcium retention in saliva. In contrast, FDC-SP is thought to play a role in the immune system associated with germinal centers. We report here the very specific expression of FDC-SP in junctional epithelium at the gingival crevice. This region is very important for the host defense against pathogens and for periodontal protection. To be able to better understand the function of FDC-SP, we developed a novel gene expression system that exploited gene trapping and site-specific gene integration to introduce the protein into a mammalian cell culture system. Using this system we were able to express FDC-SP as a fusion protein with green fluorescent protein in an osteogenic progenitor cell line with long term stability, which we then used to find that the fusion protein specifically adsorbs onto mineral deposits and the surface of hydroxyapatite particles exogenously added to the culture. This adsorption was highly dependent on the structural integrity of FDC-SP. These results suggest that FDC-SP may play an important role, adsorbing onto the surface of cementum and alveolar bone adjacent to periodontal ligament and onto tooth surface at the gingival crevice.

This study reports a new system that is very effective in identifying genes closely related to the early phase of chondrogenic differentiation. While studying chondrogenesis in a progenitor cell line, ATDC5, we found that the amount of culture media overlying an ATDC5 monolayer affected the extent to which differentiation occurred. Therefore, to gain insight into the molecular mechanisms of chondrogenic differentiation, differential gene expression between differentiating and non-differentiating ATDC5 cultures was examined by suppression subtractive hybridization analysis. In this study, we focused on transcription factors that were identified in differentiating cultures, and found that activating transcription factor 5, ATF5, exhibited a conspicuous activation pattern using two methods to induce chondrogenesis of ATDC5 cells. Furthermore, ATF5 was found to be elevated in the developing limb bud by in situ hybridization in a pattern that was highly restricted to the cartilaginous anlagen, suggesting a positive association with ATF5 expression and chondrogenesis.

The Swarm rat chondrosarcoma is a tumor tissue line derived from a tumor that arose spontaneously in a Sprague-Dawley rat. The original tissue has given rise to several tissue lines and cell lines that have been prepared in different laboratories. It has been observed that these lines differed in their growth rates and biochemical characteristics. We have characterized our Swarm rat chondrosarcoma tissue and cell lines currently in use in terms of their cytogenetic profiles and their tumorigenic properties in vivo. We found a wide variety of chromosomal abnormalities among cell lines, including translocations, deletions and polyploidy. There were also significant differences in their growth properties in vivo, giving rise to tumors of a few milligrams in the case of Ng cells, to 35 grams in the tissue line JWS. The cytogenetic complexity of the Swarm rat chondrosarcoma between and among different lines makes it very suitable to address questions about the changes that occur as a result of karyotypic abnormalities and to provide links between cytogenetic abnormalities and the dynamic oncogenic machinery.

We have developed a simple and unique strategy for identifying cell surface and extracellular matrix molecules produced by chondrocytes. Our strategy comprises two methods, retrovirus-based signal sequence gene trapping and culturing of Swarm rat chondrosarcoma chondrocytes. After infection with a retrovirus vector and isolation of hygromycin-resistant clones, trapped genes could be easily identified by the 5' rapid amplification of cDNA ends (5' RACE) method. Furthermore, the treatment of isolated clones with gadolinium chloride enabled us to determine whether the trapped gene expression was dependent on the state of chondrogenic differentiation.

A cell line was developed from the transplantable Swarm rat chondrosarcoma (RCS) and has been maintained in continuous monolayer tissue culture for a number of years. This long term-cultured (LTC) cell line exhibits the morphological and biochemical characteristics of chondrocytes and resembles the RCS tumor by electron and light microscopy. The cell line differs from the original tumor cells in that about 90% of the sulfated macromolecules are retained in the LTC extracellular matrix as compared to 30% by primary cultures of cells from the RCS tumor. An interesting and useful feature of this cell line is that it contains clonal populations of cells which differ in the quality and quantity of matrix produced. Two such clones serve to illustrate the diversity of cell types within the LTC cell line. One termed Rex accumulates an intensely staining matrix around it, while the other, Ng, accumulates a matrix, that remains virtually unstained. The chondrocytic nature and ease of cloning make these cells ideal for biochemical analysis of the chondrocyte and its extracellular matrix., (Copyright 2003 Wiley-Liss, Inc.)

The unilateral canine model is the most commonly used model of experimental osteoarthritis (OA). In this model, the anterior cruciate ligament (ACL) of one knee is transected and the contralateral joint is usually used as a control. However, dogs, similar to humans, can develop OA spontaneously with old age. Additionally, certain breeds of dogs are genetically predisposed to OA and can develop symptoms at a young age. The goal of this study was to compare the pathological changes of proteoglycans in OA cartilage from dogs that developed OA spontaneously to those that underwent ACL transection. For this reason, biglycan, decorin and fibromodulin levels and degradation patterns were compared by Western blot hybridization, and aggrecan contents were quantified by dimethylmethylene blue assay. The changes in proteoglycan levels in the cartilage of dogs with spontaneous OA, regardless of their age, were very similar to those published for human OA cartilage. However, when OA developed as a result of ACL-surgery, the changes in proteoglycans were different from those of slowly developing spontaneous OA. Therefore, these differences should be taken into consideration when the ACL-transection model is used.

As part of a program of research aimed at determining the role of mechanical forces in connective tissue differentiation, we have developed a model for investigating the effects of dynamic compressive loading on chondrocyte differentiation in vitro. In the current study, we examined the influence of cyclic compressive loading of chick limb bud mesenchymal cells to a constant peak stress of 9.25 kPa during each of the first 3 days in culture. Cells embedded in agarose gel were subjected to uniaxial, cyclic compression at 0.03, 0.15, or 0.33 Hz for 2 h. In addition, load durations of 12, 54, or 120 min were evaluated while holding frequency constant at 0.33 Hz. For a 2 h duration, there was no response to loading at 0.03 Hz. A significant increase in chondrocyte differentiation was associated with loading at 0.15 Hz, and an even greater increase with loading at 0.33 Hz. Holding frequency constant at 0.33 Hz, a loading duration of 12 min elicited no response, whereas chondrocyte differentiation was enhanced by loading for either 54 or 120 min. Although not statistically significant from the 120 min response, average cartilage nodule density and glycosaminoglycan synthesis rate were highest in the 54 min duration group. This result suggests that cells may be sensitive to the level of cumulative (nonrecoverable) compressive strain, as well as to the dynamic strain history.

The isolated mouse tail tendon fascicle, a functional and homogenous volume of tendon extracellular matrix, was utilized as an experimental system to examine the structure function relationships in tendon. Our previous work using this model system demonstrated relationships between mean collagen fibril diameter and fascicle mechanical properties in isolated tail tendon fascicles from three different groups of mice (3-week and 8-week control and 8-week Mov13 transgenic) K.A. Derwin, L.J. Soslowsky, J. Biomech. Eng. 121 (1999) 598-604. These groups of mice were chosen to obtain tendon tissues with varying collagen fibril structure and/or biochemistry, such that relationships with material properties could be investigated. To further investigate the molecular details of matrix composition and organization underlying tendon function, we report now on the preparation, characterization, and quantitation of fascicle PGs (proteoglycans) from these three groups. The chondroitin sulfate/dermatan sulfate (CS/DS)-substituted PGs, biglycan and decorin, which are the abundant proteoglycans of whole tendons, were also shown to be the predominant PGs in isolated fascicles. Furthermore, similar to the postnatal maturation changes in matrix composition previously reported for whole tendons, isolated fascicles from 8-week mice had lower CS/DS PG contents (both decorin and biglycan) and a higher collagen content than 3-week mice. In addition, CS/DS chains substituted on PGs from 8-week fascicles were shorter (based on a number average) and richer in disulfated disaccharide residues than chains from 3-week mice. Fascicles from 8-week Mov13 transgenic mice were found to contain similar amounts of total collagen and total CS/DS PG as age-matched controls, and CS/DS chain lengths and sulfation also appeared normal. However, both decorin and biglycan in Mov13 tissue migrated slightly faster on sodium dodecyl sulfate polyacrylamide gel electorphoresis (SDS-PAGE) than the corresponding species from 8-week control, and biglycan from the 8-week Mov 13 fascicles appeared to migrate as a more polydisperse band, suggesting the presence of a unique PG population in the transgenic tissue. These observations, together with our biomechanical data [Derwin and Soslowsky, 1999] suggest that compensatory pathways of extracellular matrix assembly and maturation may exist, and that tissue mechanical properties may not be simply determined by the contents of individual matrix components or collagen fibril size.

It is well established that mechanical loading is important to homeostasis of cartilage tissue, and growing evidence suggests that it influences cartilage differentiation as well. Whereas the effect of mechanical forces on chondrocyte biosynthesis and gene expression has been vigorously investigated, the effect of the mechanical environment on chondrocyte differentiation has received little attention. The long-term objective of this research is to investigate the regulatory role of mechanical loading in cell differentiation. The goal of this study was to determine if mechanical compression could modulate chondrocyte differentiation in vitro. Stage 23/24 chick limb-bud cells, embedded in agarose gel, were subjected to either static (constant 4.5-kPa stress) or cyclic (9.0-kPa peak stress at 0.33 Hz) loading in unconfined compression during the initial phase of commitment to a phenotypic lineage. Compared with nonloaded controls, cyclic compressive loading roughly doubled the number of cartilage nodules and the amount of sulfate incorporation on day 8, whereas static compression had little effect on these two measures. Neither compression protocol significantly affected overall cell viability or the proliferation of cells within nodules. Since limb-bud mesenchymal cells were seeded directly into agarose, an assessment of cartilage nodules in the agarose reflects the proportion of the original cells that had given rise to chondrocytes. Thus, the results indicate that about twice as many mesenchymal cells were induced to enter the chondrogenic pathway by cyclic mechanical compression. The coincidence of the increase in sulfate incorporation and nodule density indicates that the primary effect of mechanical compression on mesenchymal cells was on cellular differentiation and not on their subsequent metabolism. Further studies are needed to identify the primary chondrogenic signal associated with cyclic compressive loading and to determine the mechanism by which it influences commitment to or progression through the chondrogenic lineage, or both.

Chondroitin 4-sulfotransferase, which transfers sulfate from 3'-phosphoadenosine 5'-phosphosulfate to position 4 of N-acetylgalactosamine in chondroitin, was purified 1900-fold to apparent homogeneity with 6.1% yield from the serum-free culture medium of rat chondrosarcoma cells by affinity chromatography on heparin-Sepharose CL-6B, Matrex gel red A-agarose, 3',5'-ADP-agarose, and the second heparin-Sepharose CL-6B. SDS-polyacrylamide gel electrophoresis of the purified enzyme showed two protein bands. Molecular masses of these protein were 60 and 64 kDa under reducing conditions and 50 and 54 kDa under nonreducing conditions. Both the protein bands coeluted with chondroitin 4-sulfotransferase activity from Toyopearl HW-55 around the position of 50 kDa, indicating that the active form of chondroitin 4-sulfotransferase is a monomer. Dithiothreitol activated the purified chondroitin 4-sulfotransferase. The purified enzyme transferred sulfate to chondroitin and desulfated dermatan sulfate. Chondroitin sulfate A and chondroitin sulfate C were poor acceptors. Chondroitin sulfate E from squid cartilage, dermatan sulfate, heparan sulfate, and completely desulfated N-resulfated heparin hardly served as acceptors of the sulfotransferase. The transfer of sulfate to the desulfated dermatan sulfate occurred preferentially at position 4 of the N-acetylgalactosamine residues flanked with glucuronic acid residues on both reducing and nonreducing sides.

Type XI collagen, a heterotrimer specific to cartilage matrix, plays an important role in cartilage morphogenesis. We analyzed various alpha2(XI) collagen promoter-lacZ reporter gene constructs in transgenic mice to understand tissue-specific transcriptional regulation. The -530 promoter sequence was sufficient to direct reporter gene expression specifically to cartilage. Further deletion to -500 abolished reporter gene expression in cartilage but activated the expression specific to neural tissues such as brain and neural tube. An additional 47-base pair deletion resulted in random tissue expression patterns. A 24-base pair sequence from -530 to -507 of the alpha2(XI) promoter was able to switch the activity of the heterologous neurofilament light gene promoter from neural tissues to cartilage. These results suggest that the alpha2(XI) collagen gene is regulated by at least three modular elements: a basal promoter sequence distal to -453, a neural tissue-specific element (-454 to -500), and a cartilage-specific element (-501 to -530), which inhibits expression in neural tissues and induces expression in cartilage.

Mechanical stress is an important regulator of chondrocyte functions but the mechanisms by which chondrocytes sense mechanical signals are unknown. Receptors for matrix molecules are likely involved in the mechanical signaling. In the first part of this study we identified integrins with affinity for the cartilage-specific collagen type II. We report that the collagen-binding integrins alpha 1 beta 1 and alpha 2 beta 1 isolated from bovine chondrocytes or human chondrosarcoma cells bound collagen type II as judged from affinity chromatography. The integrins alpha 3 beta 1 or alpha 9 beta 1 did not bind collagen type II-Sepharose. In the second part of the study we investigated the effect of mechanical stress on expression of matrix molecules and integrin subunits. Chondrocytes and chondrosarcoma cells, cultured on uncoated flexible silicone membranes in the presence of serum, were exposed to mechanical stress by the Flexercell system. Dynamic stimulation of chondrocytes for 3 h increased the mRNA expression of collagen type II and aggrecan as judged by Northern blotting, while the beta 1-integrin subunit was not changed. When chondrosarcoma cells were exposed to mechanical stimulation under the same conditions, mRNA expression of alpha 5 was found to increase while beta 1, alpha 2, and alpha v did not increase to significant levels. In another study the effect of mechanical stress on integrins was investigated when the cells were cultured on collagen type II-coated flex-dishes. Three hours of dynamic stress increased the mRNA expression of alpha 2-integrin subunit while the level of mRNA for integrin subunits beta 1, alpha 1, alpha 5, and alpha v showed no or small changes, indicating that matrix components may modulate the expression of integrins during mechanical stress.

We show that a new rat chondrosarcoma (RCS) cell line established in long-term culture from the Swarm tumor displayed a stable differentiated chondrocyte-like phenotype. Indeed, these cells produced the collagen types II, IX, and XI and alcian blue-stainable cartilage-specific proteoglycans, but no type I or type III collagen. To functionally characterize their chondrocytic nature, the cells were stably transfected with a type II collagen/beta geo chimeric gene which confers essentially perfect chondrocyte-specific expression in transgenic mice. RCS cells expressed both beta-galactosidase and G418 resistance, in comparison with similarly transfected 10T1/2 and NIH/3T3 fibroblasts which did not. These cells were then used to perform a systematic deletion analysis of the first intron of the mouse type II collagen gene (Col2a1) using transient expression experiments to determine which segments stimulated expression of a luciferase reporter gene in RCS cells but not in 10T1/2 fibroblasts. Cloning of two tandem copies of a 156-base pair (bp) intron 1 fragment (+2188 to +2343) in a construction containing a 314-bp Col2a1 promoter caused an almost 200-fold increase in promoter activity in RCS cells but no increase in 10T1/2 cells. DNase I footprint analysis over this 156-bp fragment revealed two adjacent protected regions, FP1 and FP2, located in the 3'-half of this segment, but no differences were seen with nuclear extracts of RCS cells and 10T1/2 fibroblasts. Deletion of FP2 to leave a 119-bp segment decreased enhancer activity by severalfold, but RCS cell specificity was maintained. Further deletions indicated that sequences both in the 5' part of the 119-bp fragment and in FP1 were needed simultaneously for RCS cell-specific enhancer activity. A series of deletions in the promoter region of the mouse Col2a1 gene progressively reduced activity when these promoters were tested by themselves in transient expression experiments. However, these promoter deletions were all activated to a similar level in RCS cells by a 231-bp intron 1 fragment that included the 156-bp enhancer. The RCS cell-specific activity persisted even if the Col2a1 promoter was replaced by a minimal adenovirus major late promoter. This 231-bp intron 1 fragment also had strong enhancing activity in transiently transfected mouse primary chondrocytes. Our experiments establish the usefulness of RCS cells as an experimental system for studies of the control of chondrocyte-specific genes, provide an extensive delineation of segments in the Col2a1 first intron involved in chondrocyte-specific activity, and show that promoter sequences are dispensable for chondrocyte specificity.

Endochondral resorption is most clearly recognized at the metaphyseal interface of the growth plate with the adjacent vasculature; however, apparently identical processes of endochondral resorption are seen in sites of primary and secondary ossification of the cartilaginous anlage of bones and in ossifying fracture callus. Recent evidence of the expression of the hypertrophic phenotype in osteoarthrotic articular cartilage suggests that endochondral resorption also may be a factor in the loss of articular cartilage in this condition. To investigate the mechanism of endochondral resorption, a model culture system was developed and characterized. The two primary centers of ossification with surrounding cartilage were dissected from embryonic chick sterna prior to (18-day-old embryos) or after (20-day-old embryos) the initiation of resorption. They were cultured either in plastic culture dishes or on chorioallantoic membranes, and resorption was detected by analysis of the loss of types II and X collagen and by histological characterization. Only sterna showing active resorption in vivo were resorbed when cultured on chorioallantoic membrane. The histological appearance of the resorption site and the specificity of resorption to the primary ossification center, seen from both the analysis of loss of collagen and histological observation, suggested that the resorption of sterna cultured on chorioallantoic membrane was similar to that observed in vivo. These studies further indicated that both vascular cells and viable chondrocytes were required for resorption. Susceptibility to resorption could be induced in resistant primary ossification centers by prior culture in the absence of vascular cells, and it is suggested that it results from the accumulation of resorption-susceptible cells and matrix as a result of continued chondrocyte development.

A rat chondrosarcoma cell line and primary bovine chondrocytes have been used to study cell-mediated aggrecan catabolism. Addition of 1 microM retinoic acid to chondrosarcoma cultures resulted in aggrecan proteolysis with the release of greater than 90% of the cell layer aggrecan into the medium within 4 days. NH2-terminal sequencing of chondroitin sulfate-substituted catabolic products gave a single major NH2-terminal sequence of ARGNVILTXK, initiating at Ala374. This showed that the proteinase, commonly referred to as "aggrecanase," which cleaves the Glu373-Ala374 bond of the interglobular domain of aggrecan (Sandy, J. D., Neame, P. J., Boynton, R. E., and Flannery, C. R. (1990) J. Biol. Chem. 266, 8683-8685), is active in this cell system. Aggrecan G1 domain, generated by cleavage of the interglobular domain, was also liberated during catabolism and this was characterized with three antipeptide antisera. Anti-CDAGWL was used as a general probe for G1 domain. Anti-FVDIPEN was used to specifically detect G1 domain with COOH terminus of Asn341, the form which is readily generated by cleavage of aggrecan by a wide range of matrix metalloproteinases. Anti-NITEGE antiserum was used to specifically detect G1 domain with COOH terminus of Gln373, the form which is the expected product of "aggrecanase"-mediated cleavage of aggrecan. Western blot analysis indicated that a single form of G1 domain of about 60 kDa was formed. G1 domain of this size reacted with both anti-CDAGWL and anti-NITEGE but not with anti-FVDIPEN. Similar experiments with primary bovine chondrocyte cultures, treated with either retinoic acid or interleukin 1, showed that two forms of catabolic G1 domain, of about 62 and 66 kDa, were formed. Both of these forms reacted on Western blots with anti-CDAGWL and also with anti-NITEGE. It is suggested that cell-mediated catabolism of the aggrecan interglobular domain in these culture systems, whether promoted by retinoic acid or interleukin 1, primarily involves cleavage of the Glu373-Ala374 bond by aggrecanase. The accumulation of G1 domain with a COOH-terminal of Glu373 shows that such aggrecanase-mediated cleavage can occur independent of the cleavage of the Asn341-Phe342 bond by matrix metalloproteinases.

Many studies have illustrated the effect of mechanical loading on articular cartilage and the corresponding changes in chondrocyte metabolism, yet the mechanism through which the cells respond to loading still is unclear. The purpose of this study was to evaluate the change in shape of chondrocytes under a statically applied uniaxial compressive load. Isolated chondrocytes from rat chondrosarcoma were embedded in 2% agarose gel. Strains of 5, 10, and 15% were applied, and images of the cell were recorded from initial loading to equilibrium (15 minutes). A finite-element model was used to model the experimental setup and to estimate the mechanical properties of the chondrocyte at equilibrium. The transient behavior of the composite in the experiment was analyzed with use of a standard linear viscoelastic model. We found that all cells decreased in cross-sectional area under each of the applied compressive strains. In the finite-element model, the elasticity of the chondrocyte was similar to that of the surrounding agarose gel (4.0 kPa) and had a Poisson's ratio of 0.4. Viscoelastic analysis showed that the chondrocytes contributed a significant viscoelastic component to the behavior of the composite in comparison with the agarose gel alone. If a decrease in cell volume proportional to the decrease in cross-sectional area is assumed, the decrease observed was greater than would be predicted by a passive cellular response due to an equivalent osmotic pressure. This indicates that the chondrocyte may be altering its intracellular composition by cellular processes in response to mechanical loading.

The ability of chondrocytes from calf articular cartilage to synthesize and assemble a mechanically functional cartilage-like extracellular matrix was quantified in high cell density (approximately 10(7) cells/ml) agarose gel culture. The time evolution of chondrocyte proliferation, proteoglycan synthesis and loss to the media, and total deposition of glycosaminoglycan (GAG)-containing matrix within agarose gels was characterized during 10 weeks in culture. To assess whether the matrix deposited within the agarose gel was mechanically and electromechanically functional, we measured in parallel cultures the time evolution of dynamic mechanical stiffness and oscillatory streaming potential in uniaxial confined compression, and determined the intrinsic equilibrium modulus, hydraulic permeability, and electrokinetic coupling coefficient of the developing cultures. Biosynthetic rates were initially high, but by 1 month had fallen to a level similar to that found in the parent calf articular cartilage from which the cells were extracted. The majority of the newly synthesized proteoglycans remained in the gel. Histological sections showed matrix rich in proteoglycans and collagen fibrils developing around individual cells. The equilibrium modulus, dynamic stiffness, and oscillatory streaming potential rose to many times (>5x) their initial values at the start of the culture; the hydraulic permeability decreased to a fraction (approximately 1/10) that of the cell-laden porous agarose at the beginning of the culture. By day 35 of culture, DNA concentration (cell density), GAG concentration, stiffness, and streaming potential were all approximately 25% that of calf articular cartilage. The frequency dependence of the dynamic stiffness and potential was similar to that of calf articular cartilage. Together, these results suggested the formation of a mechanically functional matrix.

Newly synthesized 35S-labeled chondrocytic keratan sulfate chains were generated by chondrocytes of human chondrosarcoma cell line 105KC and were analyzed for heterogeneity of regional substitution, hydrodynamic size, and charge density. After isolation of the high density large chondrocytic proteoglycans and sequential digestions with chondroitinase ABC, L-1-tosylamido-2-phenylethyl chloromethyl ketone-treated trypsin, and alpha-chymotrypsin, followed by Superose 6 chromatography, two populations of keratan sulfate-containing proteoglycan fragments were identified and pooled separately. Keratan sulfate chains from each of the regions were compared after release by Pronase digestion, and differences in substitution patterns were observed; keratan sulfate chains of greater polydispersity, as well as a population of larger hydrodynamic size, were present in only one of the two regions. Alkaline/borohydride treatment confirmed both the existence of a population of uniquely large keratan sulfate chains and its restriction to a single region of proteoglycan fragments. In addition to heterogeneity of hydrodynamic size, the keratan sulfate chains exhibited regional heterogeneity of charge density and hence, of sulfation patterns. Analysis by Mono Q chromatography identified distinct groups of keratan sulfate that segregated by charge density and whose proportionate composition differed between the proteoglycan regions. Furthermore, the most highly charged species were unique to a single region and encompassed the chains of larger hydrodynamic size. This suggests that there may be regional heterogeneity of keratan sulfate chains substituted along a single class of proteoglycans and identifies a novel population of large, highly sulfated chondrocytic keratan sulfate chains.

Keratan sulphate is an integral component of the large aggregating proteoglycans of mature human articular cartilage. The keratan sulphate content of chondrocytic proteoglycans increases during maturation, and it is a useful marker of mature-type chondrocytic proteoglycans. Ordinarily, in cell culture, chondrocytes from non-neoplastic tissues dedifferentiate, diminish or cease to synthesize aggregating proteoglycans with the same amount of keratan sulphate as those formed in vivo, and do not maintain their in vivo phenotype. In tissue culture, this down-regulation of synthesis of keratan sulphate is irreversible. The study of the metabolism of mature human chondrocytes has been hampered by the absence of stable models. We report a cell-line, 105KC, derived from a human chondrosarcoma, that has maintained a stable proteoglycan phenotype during more than three years of culture. Analysis with immunofluorescence suggested that 105KC cells continued to synthesize keratan sulphate in long-term culture. Biochemical analysis demonstrated that 105KC cells maintained the production of chondrocytic large-aggregating proteoglycans and that keratan sulphate composed 13 per cent of their glycosaminoglycan content. To our knowledge, 105KC represents the first model to have maintained the post-fetal chondrocytic proteoglycan phenotype in stable culture. This study documents the feasibility of the development of mature chondrocytic cell-lines and sheds light on the biological characteristics of chondrosarcomas.

Articular cartilage chondrocytes maintain biosynthetic heterogeneity in cell culture, but undergo irreversible dedifferentiation of their proteoglycan phenotype, as defined by keratan sulfate content. A recently described cell line of malignant human chondrocytes, 105KC, was the first to maintain a differentiated keratan sulfate-proteoglycan phenotype in long-term culture. A clone of 105KC, labeled KC2H3, is currently described and represents a distinct and metabolically more homogeneous population of mature chondrocytes than 105KC. KC2H3 cells universally express keratan sulfate biosynthesis, as defined by indirect immunofluorescence. In addition, KC2H3 expresses a more mature proteoglycan phenotype than 105KC, as demonstrated by the keratan sulfate content: 24% of glycosaminoglycan content of the aggregating proteoglycans of KC2H3 versus 13% for 105KC. Further reported are the effects of long term monolayer culture on the proteoglycan phenotype expressed by KC2H3. After more than 16 months in continuous monolayer, KC2H3 cells remained morphologically indistinguishable from those maintained in suspension alternating with monolayer. In addition, the proteoglycan phenotype remained mature, without a tendency towards dedifferentiation. The flattened morphology adopted by chondrocytes while in monolayer has been considered a stimulus of dedifferentiation; the present study is the first to examine the direct effects of physical state on a homogeneous and stable population of chondrocytes.

Difficulty in maintaining phenotypic stability of the Swarm rat chondrosarcoma in long-term monolayer cultures has prompted investigation of alternative conditions that would enable extended maintenance of these cells, permitting use of the tumor as a model system for the long-term study of proteoglycan metabolism. Morphological analysis of the growth of the chondrosarcoma chondrocytes in agarose has shown stability of the culture over a 20 day period with respect to the ability of the cells to proliferate and synthesize an Alcian blue-positive extracellular matrix. The present study confirms these findings through analysis of the growth characteristics of the culture and the pattern of proteoglycan and collagen synthesis. The chondrocytes actively synthesize a proteoglycan-rich matrix at a rate dependent on the initial plating density and concentration of serum in the culture medium. These factors similarly affect the proliferative capabilities of the culture as demonstrated by the growth curves obtained at different culture conditions. During 20 days in culture, the cells synthesize an aggregating chondroitin sulfate proteoglycan and collagen type II, typical of cartilage and this chondrosarcoma. In addition, small molecular weight proteoglycans were found to be present at concentrations of up to 10% of the total proteoglycan population. Degradative rates are slow, the proteoglycan half-life is about 30 days, but can be enhanced with retinol, reducing the half-life to 2 days.

When chondrocytes from the Swarm rat chondrosarcoma are isolated by trypsin and collagenase digestion and cultured in Petri dishes, they form a new extracellular matrix within 24 hours, with proteoglycan matrix granules and collagen fibrils. This rapid synthesis of new matrix, together with the biochemical and morphological characterization of the proteoglycans as typical of cartilage, demonstrates the value of these cultures as a model system for studies of synthesis, secretion, and organization of extracellular matrix.

After incubating cultured chondrocytes from the Swarm rat chondrosarcoma for 30 min with [3H]serine, a labeled macromolecule was found predominantly as Mr = approximately 370,000 species which was subsequently identified as a core protein precursor to cartilage proteoglycan from the following properties: (a) it was immunoprecipitated along with completed proteoglycan from cell extracts by an antiserum to the complex of hyaluronic acid-binding region, link protein, and hyaluronic acid. Its immunoprecipitation could be inhibited completely by the addition of purified hyaluronic acid-binding region to the extracts, indicating the presence of common antigenic determinants with this region of the proteoglycan core protein. (b) the core protein precursor was able to interact with the hyaluronic acid and link protein in proteoglycan aggregates added as carrier to extracts to form mixed aggregates of high buoyant density in associative CsCl density gradients. Labeled core protein precursor and link protein were subsequently isolated from the mixed aggregates from the top of dissociative CsCl density gradients. (c) radioactivity in core protein precursor after a 30-min pulse of [3H]serine disappeared after inhibiting further protein synthesis with cycloheximide concurrent with the appearance of label in completed proteoglycan molecules.

An enzyme-linked immunosorbent assay, in combination with an independent inhibition step, was established to quantitate two components of the proteoglycan aggregate, namely link protein and hyaluronate-binding region, at concentrations below 100 ng/ml. The presence of other components of the aggregate in the samples to be tested influenced quantitation in a specific manner. The apparent antigenicity of link protein increased 2-5 times when either purified proteoglycan monomer, purified hyaluronate-binding region, or purified hyaluronate (macromolecular or oligomers) were present in the link protein samples. These findings are interpreted as showing different states of conformation or degree of association of the link protein with other components of aggregate in solution. In separate experiments, a 2-4-fold increase in the apparent antigenicity of purified hyaluronate-binding region was observed when hyaluronate molecules with at least 20 disaccharides were present in the samples. Co-incubation of the hyaluronate-binding region or proteoglycan monomer with either purified link protein or with smaller hyaluronate oligomers did not change its antigenicity in the assay. However, when hyaluronate oligomers with 8 disaccharides were included in a mixture of macromolecular hyaluronate with hyaluronate-binding region, the increase in apparent antigenicity was blocked. The results illustrate the inherent difficulties in using the enzyme-linked immunosorbent assay for the quantitation of link protein or proteoglycan monomers in samples where these macromolecules can associate with themselves or other components of proteoglycan aggregates.

Preformed Cs2SO4 zonal gradients were used to purify aggregates from proteoglycan preparations derived from associative extracts of the Swarm rat chondrosarcoma. Zonal gradients were used in solvents with different concentrations of guanidine HCl and different solvent pH values to study the mechanisms for dissociating the aggregates. Aggregates are stable in concentrations of guanidine HCl up to 1.5 M at pH 6.6. At 2 M guanidine HCl, partial dissociation occurs over 20 h in which a link protein is completely dissociated for every monomer proteoglycan dissociated from the aggregate structure. This suggests that in this solvent disaggregation occurs concurrent with complete separation of link protein from monomer. At solvent pH 2.7 to 3.3 in ionic conditions which normally promote aggregation, dissociation occurs by a mechanism in which the link protein remains associated with monomer. Thus, link protein-monomer complexes dissociate as bimolecular units from hyaluronic acid; such complexes then exhibit physical properties indistinguishable from pure monomers. The link protein-monomer complexes reassociate with hyaluronic acid to form link-stabilized aggregates when the solvent pH is raised to pH 7, i.e. to associative conditions. The study provides additional evidence for the role that link protein-monomer interactions have in proteoglycan aggregate structures.

A range of structurally related zwitterionic detergents, Zwittergents 3-06, 3-08, 3-10, and 3-12, and a derivative of cholic acid (Chaps) were examined for their ability to enhance the extraction of newly synthesized, intracellular proteoglycans and for their effect on the functional properties of cartilage proteoglycan. Although none of the detergents could extract greater than 4% of the intracellular proteoglycans when used alone, Zwittergents 3-10, 3-12, and Chaps proved equally as effective when used in combination with 4 M guanidine HCl extracting greater than 90% of newly synthesized proteoglycans. Rate zonal centrifugation of aggregates containing either 3H-link protein or 3H-monomer, which had been incubated with 2% (w/v) detergent indicated that none of the test detergents caused a disassembly of intact aggregates. However, both Zwittergents 3-10 and 3-12 prevented the reaggregation of components dissociated with 4 M guanidine HCl. Similar to the finding with aggregate, none of the detergents caused a disassembly of monomer-link protein complexes prepared from purified 3H-link protein and proteoglycan monomer, while Zwittergents 3-10 and 3-12 prevented their assembly from free link protein and monomer. However, monomer-link protein complexes once formed were able to associate with hyaluronic acid to form link-stable ternary complexes in the presence of all detergents tested including Zwittergents 3-10 and 3-12.

Rat chondrosarcoma chondrocytes were cultured in the presence of puromycin to induce premature termination of core protein precursor. The structure and function of intracellular and extracellular proteoglycans were assessed by molecular sieve chromatography and polyacrylamide gel electrophoresis. [3H]Serine incorporation was maximally inhibited by 3 X 10(-4) M puromycin but unaffected by 10(-5) M puromycin. Proteoglycans synthesized in the presence of puromycin exhibited increased monomer size due to increased chondroitin sulfate chain size, typical of proteoglycans synthesized in the presence of protein synthesis inhibitors, but no loss in ability to bind to hyaluronic acid; and no loss in core protein size was observed after treatment with chondroitinase. These data suggest that chondrocytes select only completed or nearly completed core protein molecules to process into proteoglycans.

Biosynthesis of cartilage proteoglycan was examined in a model system of cultured chondrocytes from a transplantable rat chondrosarcoma. Extensive modification with the addition of chondroitin sulfate glycosaminoglycan, N-linked oligosaccharide, and O-linked oligosaccharide is required to convert a newly synthesized core protein precursor into a proteoglycan. Kinetic analyses revealed the presence of a large pool of core protein precursor (t 1/2 approximately 90 min) awaiting completion into proteoglycan. The large t 1/2 of this pool allowed kinetic labeling experiments with a variety of radioactive precursors to distinguish between early biosynthetic events associated primarily with the rough endoplasmic reticulum from late events associated primarily with the Golgi apparatus. The results of a series of experiments indicated that the addition of N-linked oligosaccharide chains occurs early in the biosynthetic process in association with the rough endoplasmic reticulum, whereas the initiation and completion of O-linked oligosaccharides occurs much later, at about the same time as chondroitin sulfate synthesis. This also indicated that keratan sulfate chains, when present in the completed molecule, are added in the Golgi apparatus, as they are probably built on oligosaccharide primers closely related to the O-oligosaccharide chains. Furthermore, when 3H-glucose was used as the precursor, the entry of label into xylose, the linkage sugar between the core protein and the chondroitin sulfate chain, was found to occur within 5 min of the entry of label into galactose and galactosamine in the remainder of the chondroitin sulfate chain. This indicated that the initiation and completion of the chondroitin sulfate chain occurs late in the pathway probably entirely in the Golgi apparatus. Thus, proteoglycan synthesis can be described as occurring in two stages in this system, translation and N-glycosylation of a core protein precursor which has a long half-life in the rough endoplasmic reticulum, followed by extensive rapid modification in the Golgi complex in which the majority of glycosaminoglycan and oligosaccharide chains are added to the core protein precursor with subsequent rapid secretion into the extracellular matrix.

Proteoglycan synthesis by cultured chondrocytes from the Swarm rat chondrosarcoma was examined after treatment with 0.1 mg/ml of cycloheximide which inhibited [3H]serine incorporation into total protein by greater than 90%. Incorporation of [35S]sulfate into proteoglycans decreased with nearly first order kinetics (t 1/2 = 96 +/- 6 min) with an accompanying increase in the size of the proteoglycan molecules, primary due to an increase in chondroitin sulfate chain sizes. After 5 h of cycloheximide treatment, when [35S]sulfate incorporation was inhibited by about 90%, addition of 1 mM beta-D-xyloside restored 76% of the incorporation into chondroitin sulfate observed in cultures treated only with xyloside. This suggests that the biochemical pathways for the affected by cycloheximide treatment. Cultures were prelabeled for 15 min with either [3H]serine or [35S]-methionine, and then cycloheximide was added to block further protein synthesis. Both precursors appeared in completed proteoglycan molecules with nearly first order kinetics with t 1/2 values of 92 +/- 8 and 101 +/- 11 min for [3H]serine and [35S]methionine, respectively, values in close agreement with the t 1/2 from the [35S]sulfate data. These results suggest that after cycloheximide treatment, the rate of [35S]sulfate incorporation into proteoglycan, after a correction for increases in chondroitin sulfate chain size, was directly proportional to the size of the intracellular pool of core protein. From the steady state rate of proteoglycan synthesis (estimated to be about 80 ng/min/10(6) cells in separate experiments) and a corrected t 1/2 value of 60 min, the amount of precursor core protein can be calculated to be about 500 ng/10(6) cells in these experiments.

The synthesis and subsequent assembly of proteoglycan monomer and link protein into aggregates were examined using cultured chondrocytes from the Swarm rat chondrosarcoma. Proteoglycans purified directly from culture medium after a 6-h label with [14C]serine contained up to 80% of the radioactivity in aggregates, 8% of which was in link protein. The addition of hyaluronic acid oligosaccharides (averaging 50 monosaccharides) to the culture medium during labeling inhibited aggregate formation by 94%. More than 90% of the labeled link molecules (about 7% of the total radioactivity in the proteoglycan fraction) was associated in ternary complexes with monomeric proteoglycan and HA approximately 50. The stoichiometry of labeled link protein to labeled proteoglycan core protein aggregated with either macromolecular hyaluronic acid or hyaluronic acid oligosaccharides was calculated to be 1:1. The effect of exogenous proteoglycan on aggregate formation was examined by incubating cultures for 4 h with up to 5 mg of exogenous monomer. Over this time, about 85 micrograms of newly synthesized proteoglycan accumulated in the culture medium. With increasing amounts of exogenous monomer, up to 80 micrograms of 35S-labeled exogenous monomer was incorporated into link-stabilized aggregate, while the aggregation of endogenous proteoglycan decreased from 73 to 44%. The total endogenous plus exogenous monomer in link-stabilized aggregate increased to about 130 micrograms, approximately 1 1/2 times the total amount of endogenous monomer in the culture medium, indicating that at least 1 1/2 times more link molecules than endogenous proteoglycan molecules were present. Further, the molar ratio of labeled link to labeled monomer molecules associated with the nonaggregated proteoglycan fraction remained constant at approximately 1:1, indicating that the exogenous proteoglycans were ineffective in displacing link molecules from endogenous monomer.link complexes. The data suggest that the formation of proteoglycan aggregates may be a sequential process in which a link.monomer complex forms and this complex subsequently binds to hyaluronic acid.

Chondrocytes from the Swarm chondrosarcoma, a transplantable rat tumor, have been difficult to maintain in tissue culture for extended periods due to a time-dependent alteration of the culture to a more fibroblastic phenotype. This feature precluded the use of these cultures to examine chronic conditions that may affect cell metabolism, and the homogeneity or heterogeneity of the tumor cells within the culture population could not be examined. Use of suspension culture in agarose stabilized the chondrocyte phenotype, permitting long-term culture. Clones of tumor chondrocytes were established in agarose and were examined over 2-3 weeks for evidence that the cells were accumulating a proteoglycan-rich extracellular matrix, as determined by positive staining by Alcian blue, and were undergoing cell division. Nearly 90% of the cloned cells exhibited a prominent extracellular matrix by day 7 of culture and greater than 99% did so by day 14. Cell division did not occur to any great extent until days 6-7 of culture. After this lag, the cells appeared to undergo logarithmic growth, with a cell generation time of about 12 days. By 20 days of culture, between 80 and 90% of the initial clones contained multiple cells, indicating that nearly all the cells were in, or had entered, the cell cycle. These results suggest that the chondrocytes from the rat chondrosarcoma form a homogeneous cell population with respect to their ability to synthesize an extra-cellular matrix and divide.

Chondrocytes from immature and mature rabbits have been compared in biosynthetic studies with [3H] leucine and [35S]sulfate as precursors. The time course of incorporation of [3H]leucine into general protein, proteoglycan monomer core protein, and link protein and of [35S]sulfate into proteoglycan monomer has been examined. Proteoglycan monomer was isolated from the high buoyant density (p greater than 1.60) fractions of dissociative CsCl gradients and link protein by immunoprecipitation with antibody 8A4 followed by gel electrophoresis. Results based on the period of linear isotope incorporation showed that mature cells synthesize protein at about 40% of the rate of immature cells and both proteoglycan and link protein at about 20% of the rate of immature cells. The labeling rates obtained suggest that immature cells synthesize an approximate 1:1 molar ratio of link protein to proteoglycan monomer, and for mature cells this ratio is about 0.8:1. While cell layer retention of newly synthesized proteoglycan was markedly lower in mature relative to immature cell cultures, link protein retention was high in both immature and mature cultures; this finding provides an explanation for our previous observation (Plaas, A. H. K., and Sandy, J. D. (1984) Biochem, J. 220, 337-340) that link-free monomer accumulates in the medium of mature but not immature cultures. The link protein synthesized by both ages of cells and isolated from cell layer or medium was a single major species of apparent molecular mass 48-51 kDa. The results suggest that mature chondrocytes are less efficient than immature chondrocytes in the coordinated assembly of link-stabilized proteoglycan aggregates in this culture system.

Native and clostripain-derived link proteins from proteoglycan aggregates were separated by gel chromatography in 4 M guanidine HCl from low-buoyant-density proteoglycan and proteoglycan hyaluronic acid-binding region after extraction from adult bovine nasal cartilage and the Swarm rat chondrosarcoma. Separations were monitored with Laurell immunoelectrophoresis using precipitating antibodies to link protein, hyaluronic acid-binding region, and low-buoyant-density proteoglycan. These immunoanalyses, with sodium dodecyl sulfate-polyacrylamide gel analyses, revealed the usefulness of this combined approach when assessing the purities and identities of these molecules. Using these isolated link proteins we provide data suggesting the presence of two immunologically detectable forms of link proteins of the same molecular weight. On immunoelectrophoresis both rat and bovine link proteins of a single molecular weight each produced two precipitin reactions with antibody to link protein. They were of similar intensity for native molecules but of different intensity for clostripain-isolated link protein. Isoelectric focusing in 6 M urea revealed that the 48,000 and 44,000 molecular weight link proteins of bovine nasal cartilage together produce a complex pattern of many bands. Link proteins of a single molecular size produced a much simpler, predominantly five-banded focusing pattern. Immunoelectrophoresis of the electrofocused clostripain rat link protein (42,000) revealed that the three major central bands each produced double precipitin reactions. Mixing of native and clostripain-derived link protein in 4 M guanidine HCl followed by dialysis to 6 M urea prior to isoelectric focusing did not change the focusing position of the individual bands. This suggests that the heterogeneity of focusing forms was due to the existence of different isoforms. The double precipitin reactions may be due to the existence of two different conformations expressing different epitopes of the kind reported previously by Thonar et al. (Thonar, E. J.-M. A., Kimura, J. H., Hascall, V. C., and Poole, A. R. (1982) J. Biol. Chem. 257, 14173-14180).

We purified and characterized the bone proteoglycans from fetal calves, growing rats, and human fetuses. The major proteoglycan is part of the mineralized tissue matrix and only 10-20% can be extracted prior to demineralization. This bone proteoglycan is a small glycoconjugate (Mr = 80,000-120,000) containing approximately 20-30% protein and either one or two chondroitin sulfate chains (Mr = 40,000) attached to a relatively monodisperse protein core (Mr = 38,000). "O"-linked and "N"-linked oligosaccharide units are also present. Antibodies directed against the protein core of calf bone proteoglycan do not cross-react with cartilage, skin, corneal, or basement membrane proteoglycans in immunoassays and have minimal cross-reactivity with scleral proteoglycans. Quantitative immunoassays and indirect immunofluorescence were used to show that the molecule is localized to forming bone trabeculae and dentin, but not to any other tissue. Osteoblasts and osteoprogenitor cells adjacent to areas undergoing rapid osteogenesis also contain this small proteoglycan. A second proteoglycan (Mr approximately equal to 1,000,000) was extracted from newly forming bone prior to demineralization. This large proteoglycan, which was isolated from the cartilage-free areas of developing intramembranous bone, has a protein core similar to that of the cartilage aggregating proteoglycan and cross-reacts with antisera raised against these cartilage proteoglycans but not with the small mineral-entrapped proteoglycan. It contains larger (Mr = 40,000) and fewer chondroitin sulfate chains than its cartilage-derived analogue, and is localized to the soft connective tissue mesenchyme lying between growing bone trabeculae. More fully formed compact bone did not contain detectable quantities of this proteoglycan.

1. Primary cultures of chondrocytes from the Swarm rat chondrosarcoma were labelled with either [3H]glucosamine or [14C]glucosamine, and hyaluronate synthesized by the cells was isolated from the cell layer. Parallel cultures were labelled with either [3H]serine or [3H]lysine, and identical fractions were isolated from the cell layer. Some cultures were dual-labelled. 2. In cultures labelled with [3H]serine for between 30 min and 24 h and extracted with 4.0 M-guanidine, a procedure that solubilizes predominantly extracellular macromolecules, small amounts of [3H]serine-labelled molecules were found associated with the hyaluronate fraction purified from the extract by dissociative CsCl-density-gradient centrifugation and dissociative Sepharose CL-2B chromatography. About 75% of the [3H]serine-labelled molecules in the fraction were specifically associated with hyaluronate, since they could be removed by prior treatment with proteinase-free Streptomyces hyaluronidase. The association of the [3H]serine-labelled molecules with hyaluronate was non-covalent, since they could be separated from it by further centrifugation in CsCl density gradients containing 4 M-guanidinium chloride and a zwitterionic detergent. 3. In other experiments the cultures were extracted with a sequential zwitterionic-detergent/guanidinium chloride procedure that completely solubilized the cell layer and enabled fractions containing newly synthesized cell-associated hyaluronate to be isolated. Zwitterionic detergent was present throughout. No [3H]lysine was incorporated into these fractions, irrespective of whether the cultures were pulsed concurrently with [3H]lysine and [14C]glucosamine or sequentially with [3H]lysine to prelabel the protein pool (24 h) followed by [14C]-glucosamine to label hyaluronate (1 h). 4. The results show that newly synthesized hyaluronate is not associated with covalently bound protein, and suggest that chain synthesis is initiated by a mechanism other than on to a core protein. Small amounts of [3H]serine-labelled molecules are, however, non-covalently associated with extracellular hyaluronate. Their identity is at present unknown, but they are probably of low molecular weight.

Proteoglycan monomer and link protein isolated from the Swarm rat chondrosarcoma both contain glycosylamine-linked oligosaccharides. In monomer, these N-linked oligosaccharides are concentrated in a region of the protein core which interacts specifically with both hyaluronate and link protein to form proteoglycan aggregates present in cartilage matrix. Chondrocyte cultures were treated with tunicamycin to inhibit synthesis of the N-linked oligosaccharides, and the ability of the deficient proteoglycan and link protein to form aggregates was studied. Cultures were pretreated with tunicamycin for 3 h and then labeled with either [3H]mannose, [3H]glucosamine, [3H]serine, or with [35S]sulfate for 6 h in the presence of tunicamycin. Formation of link protein-stabilized proteoglycan aggregates in the culture medium was inhibited by up to 40% when the cells were treated with 3 micrograms of tunicamycin/ml, a concentration which inhibited 3H incorporation with mannose as a precursor by about 90%, but by only 15% with glucosamine as a precursor. When exogenous proteoglycan aggregate was added to the culture medium, however, it was found that both endogenous monomer and link protein synthesized in the presence of tunicamycin were fully able to form link-stabilized aggregates. This suggests that glycosylamine-linked oligosaccharides on monomer and on link protein are not necessary for their specific interactions with hyaluronate and with each other. Further, although tunicamycin did not inhibit net synthesis of hyaluronate, transfer of hyaluronate from the cell layer to the culture medium was retarded. This phenomenon accounted for most if not all of the decrease in the amount of proteoglycan which formed aggregates in the medium of cultures treated with tunicamycin.

Tunicamycin, an inhibitor of dolichol-diphospho-N-acetylglucosamine formation and hence an inhibitor of N-linked oligosaccharide biosynthesis, suppressed total proteoglycan synthesis by Swarm rat chondrosarcoma chondrocytes without affecting the size of the proteoglycan molecule, its secretion from the cell, or its ability to be retained in the extracellular matrix. In addition, tunicamycin did not substantially alter the ability of the chondrocytes to polymerize glycosaminoglycan onto an exogenous beta-D-xyloside acceptor. A secondary effect of tunicamycin suppression of proteoglycan synthesis was that a lesser amount of newly synthesized proteoglycan diffused from the extracellular matrix into the culture medium. The ability of exogenous hyaluronic acid and proteoglycan to increase the percentage of newly synthesized 35S-proteoglycan in the medium in tunicamycin-treated cultures indicates that matrix retention of 35S-proteoglycan is related to the total extracellular uronic acid content rather than to the presence or absence of mannose oligosaccharides bound to the proteoglycan molecule. These noncytotoxic concentrations of tunicamycin (33-333 ng/ml) decreased [3H]mannose incorporation to the same extent that they decreased total [35S]sulfate and [3H]serine incorporation, and caused the chondrocyte to synthesize and secrete a species of beta-hexosaminidase that was mannose-deficient as assessed by its failure to bind to concanavalin A. The additional finding of decreased insulin binding to tunicamycin-treated chondrosarcoma chondrocytes suggested that the inhibition of proteoglycan synthesis was due to diminution of receptors which respond to stimulatory hormones.

Cultured chondrocytes from the Swarm rat chondrosarcoma incorporate [35S]sulfate into proteoglycans typical of hyaline cartilage. The movement of newly synthesized proteoglycans from inside the cells into the extracellular matrix and, finally, into the culture medium was examined by measuring the distribution of 35S-labeled proteoglycans in the medium, a 4 M guanidine HCl extract of the cell layer, and in the remaining residue for a number of chase times following a 5-min pulse with [35S]sulfate. When hyaluronate oligosaccharides containing greater than or equal to 10 monosaccharides were included in the chase media, a proportion of newly synthesized proteoglycans were displaced from the matrix (4 M extract) into the culture medium. This displacement was greatest when oligomers were in the chase media between 10 and 20 min after the pulse, approximately the time when the molecules are being secreted from the cells. The proportion of link-stabilized aggregate in the medium was examined by Sepharose 2B chromatography after adding an excess of unlabeled monomer which displaces labeled monomer from complexes with hyaluronate which are not link-stabilized. The proportion of link-stabilized aggregate increased from 12% to about 70% between 12 and 120 min of chase. The presence of 40 micron hyaluronate oligosaccharides of 16 monosaccharides in the chase media retarded but did not prevent aggregate formation. Oligomers of about 50 monosaccharides, which are large enough to bind both a monomer proteoglycan and a link protein, almost completely prevented the formation of the large link-stabilized aggregates. The results suggest: (a) newly synthesized proteoglycans are not bound into link-stabilized aggregates at the time of secretion; (b) hyaluronic acid oligomers which are long enough to interact only with the hyaluronic acid-binding site of proteoglycans will retard but not prevent link-stabilized aggregation; and (c) hyaluronic acid oligomers long enough to accommodate additionally a link protein form a link-stabilized ternary complex and prevent aggregation with larger hyaluronic acid molecules.