Large aggregating and small leucine-rich proteoglycans are degraded by different pathways and at different rates in tendon Tom Samiric, Mirna Z. Ilic and Christopher J. Handley School of Human Biosciences, La Trobe University, Melbourne, Victoria, Australia This work investigated the kinetics of catabolism and the catabolic fate of the newly synthesized 35 S-labelled proteo- glycans present in explant cultures of tendon. Tissue from the p roximal r egion o f bovine deep flexor tendon was incu- bated with [ 35 S]sulfate for 6 h and then placed in explant cultures for periods of up to 15 days. The amount of radi- olabel associated with proteoglycans and free [ 35 S]sulfate lost to the medium a nd retained in the matrix w as determined for each d ay i n c ulture. I t w as shown that t he rate of catabolis m of radiolabelled small proteoglycans (decorin and biglycan) was s ignificantly slower (T ½ > 20 days) c ompared w ith the radiolabelled large proteoglycans (aggrecan and versican) that were rapidly lost from the tissue (T ½  2 days). Both the small and large newly synthesized proteoglycans were lost from the matrix with either intact or proteolytically modified core proteins. When explant cultures of tendon were maintained either at 4 °C or in the presence of the lysosomotrophic 2 agent ammonium chloride, inhibition of the cellular catabolic pathway for small proteoglycans was demonstrated indicating the involvement of cellular activity and lysosomes in the catabolism of small proteoglycans. It was e stimated from these studies that approximately 60% of the radiolabelled s mall proteoglycans t hat were l ost from t he tissue were degraded by the intracellular pathway present in tendon cells. This work shows that the pathways of catabolism for large aggregating and small leucine-rich proteoglycans are different in tendon and this may reflect the roles that these two populations of proteoglycans play in the maintenance of the extracellular matrix of t endon. Keywords: catabolism; proteoglycan; tendon. The extracellular matrix of tendon is composed of parallel bundles of collagen, which endows the tissue with tensile strength and its ability t o transmit f orce generated by muscle to bone. A lso present within the e xtracellular matrix o f tendon are two groups of proteoglycans that can be distinguished on the basis of their size. The small leucine- rich proteoglycans m ake u p a pproximately 8 0% of the total proteoglycans present in the tendon with decorin being the predominant species and biglycan being present at lower levels [1–3]. The remaining 20% of proteoglycans present i n tendon are t he large a ggregating proteoglycans, versican and aggrecan, which are present in similar levels [1–3]. Tendon cells are responsible for the synthesis and degradation of extracellular proteoglycans. Studies investi- gating the catabolism of the chemical pool of aggrecan and versican by tendon in explant culture have revealed that this process i nvolves t he proteolytic cleavage of the c ore p roteins of these proteoglycans by aggrecanase activity as well as other proteinases [1–3]. The catabolism of the chemical pool of decorin and biglycan involves the loss of intact c ore proteins from the tendon matrix as well as limited proteo- lytic cleavage [1–3]. We have previously studied the kinetics of catabo lism of newly synthesized proteoglycans in bovine collateral liga- ment [4] and demonstrated that the catabolism of newly synthesized 35 S-labelled large proteoglycans was rapid (T ½  2 days) and involved p roteolytic cleavage of the core protein. O n the other h and, 35 S-labelled small proteoglycans were lost from the tissue at a slower rate (T ½  20 days) and were either lost from the tissue with a n intact or partially degraded core protein, or were internalized by the c ells and c ompletely degraded w ithin the lys osomes of the cells [4]. Indeed, i t has been shown t hat the cellular uptake of small proteoglycans is mediated b y recep tor proteins present in the plasma membranes of fibroblasts [5,6]. This study was undertaken to determine the metabolic fate of newly synthesized 35 S-labelled proteoglycans by tendon in order to elucidate the specific processes and pathways that are involved in the catabolism of newly synthesized proteoglycans present in a dense collagenous connective tissue and to compare the resulting r adiolabelled catabolic products with those reported by us for the chemical pool present in the tissue [1]. Experimental procedures Materials Dulbecco’s modified Eagle’s medium (DMEM), Eagle’s nonessential amino ac ids, penicillin and streptomycin were purchased from CSL (Melbourne, Victoria, Australia). Sephadex G-25 (as prepacked PD-10 columns) was from Pharmacia (Uppsala, Sweden). Aqueous solution of Correspondence to C. J. Hand ley, School of Hum an Biosciences, L a Trobe University, 3086, Vict oria, A ustralia. Fax: +61 39479 5784, Tel.: +61 39479 5800, E-mail: C.Handley@latrobe.edu.au Abbreviation: GdnHCl, guanidine h ydrochloride. Enzymes: chondroitinase ABC from Proteus vulgaris (EC 4.2.2.4); keratanase from Pseudomonas sp. (EC 3.2.1.103). (Received 16 June 2004, revised 21 July 2004, accepted 27 July 2004) Eur. J. Biochem. 271, 3612–3620 (2004) Ó FEBS 2004 doi:10.1046/j.1432-1033.2003.04307.x [ 35 S]sulfate (carrier free) was from DuPont New England Nuclear (Boston, MA, USA). Keratanase (from Pseudo- monas sp.; EC 3.2.1.103) was obtained f rom Sigma Chem ical Co. (St. Louis, MO, USA) and chondroitinase ABC (protease free; from Proteus vulgaris; EC 4.2.2.4) from ICN Biochemicals (Costa Mesa, CA, USA). Adult bovine meta- carpophalangeal joints were obtained from a local abattoir. Tendon explant cultures Deep flexor tendon proximal to the bifurcation was dissected from a s ingle m etacarpophalangeal j oint of aone-to-two year old steer. The tendon was then chopped into small pieces and incubated in sulfate-free medium (1 g tissue per 10 mL medium) containing 200 lCiÆmL )1 [ 35 S]sulfate at 37 °Cfor 6 h. The sulfate-free medium contained 0.13 M sodium chloride, 4.74 m M potassium chloride, 2.54 m M calcium chloride, 1.9 m M magnesium chloride, 10 m M glucose, 1.0 m ML -glutamine, 1.19 m M potassium dihydrogen phos- phate, 0.02 gÆmL )1 Phenol Red and was buffered with 25 m M HEPES adjusted to pH 7 .4 with sodium hydroxide [7]. Sulfate-free medium was used in order to increase the incorporation of [ 35 S]sulfate into proteoglycans as it has previously been reported that the rate of incorporation of [ 35 S]sulfate i nto proteoglycans by tendon is considerably less than other connective tissues such as articular cartilage, and this necessitated the use of high specific radioactivity [ 35 S]sulfate [8]. It was also shown that the rate of incorpor- ation of [ 35 S]sulfate into proteoglycans was linear over the 6 h incubation period. After washing the tissue e xtensively in DMEM to remove most of the unincorporated radioisotope, duplicate samples containing 100 ± 20 mg of tissue were distributed into individual sterile preweighed plastic vials containing 4 mL of DMEM. DMEM contains sufficient chemical levels of sulfate (0.8 M magnesium sulfate) so that the specific radioactivity o f radiolabelled sulfate present i n or produced by the explant cultures will be considerably reduced, thereby decreasing the level of re-use of [ 35 S]sulfate by the cells of the cultures. The c ulture medium was collected and r eplaced daily with 4 mL of fresh DMEM. The collected medium was stored at )20 °C in the presence of proteinase inhibitors [9]. At the end of the culture p eriod, the tissue was extracted with 4 M guanidine hydrochloride (GdnHCl) in the presence of proteinase inhibitors at 4 °C for 72 h, followed by 0.5 M NaOH at 21 °C for 24 h. Determination of the percentage of 35 S-labelled proteoglycans remaining in the matrix of tendon explant cultures To determine the percentage of 35 S-labelled proteoglycans remaining i n t he matrix of tendon cultures on each day after incubation with [ 35 S]sulfate, 0.5 mL a liquots of the medium fractions, G dnHCl and NaOH extracts were applied to columns of Sephadex G-25 (PD-10 columns) equilibrated andelutedwith4 M GdnHCl, 0.1 M Na 2 SO 4 ,0.05 M sodium acetate, 0.1% (v/v) Triton X-100, pH 6.1. The 35 S-labelled material that eluted in the excluded volume of the column was attributable to 35 S-labelled macromolecules in the medium that were originally derived from proteo- glycans present in tendon matrix. The 35 S-labelled material which e luted in the total volume was shown to r epresent free [ 35 S]sulfate. The rate of loss of 35 S-labelled proteoglycans from the matrix of explant cultures was calculated from the amount of 35 S-labelled macromolecules in the medium on each day o f culture and t hat remained in the matrix at the end of t he culture period. From these d ata t he percentage of 35 S-labelled proteoglycans remaining in the matrix was plotted as a function of time in culture [10]. Separation of 35 S-labelled proteoglycans remaining in the matrix of tendon explant cultures by size exclusion chromatography Aliquots (1 mL) of the G dnHCl extracts obtained f rom tissue after predetermined times in culture were applied to a column of Sepharose CL-4B (1.3 · 87.0 cm) equilibrated andelutedwith4 M GdnHCl, 5 0 m M sodium acetate buffer, 0.1% (v/v) Triton X-100 pH 5.8. Fractions of 1 mL were collected at a flow rate of 6 mLÆh )1 and assayed for 35 S-radioactivity. From these data the percentage of 35 S-labelled large and small proteoglycan species remaining in the matrix at d ifferent times i n culture was determin ed [10]. Detection of 35 S-labelled proteoglycan core proteins remaining in the matrix or released into the medium of tendon explant cultures by fluorography Tissue was dissected from a single metacarpophalangeal joint and incubated with [ 35 S]sulfate for 6 h as described above. The tissue was maintained in D MEM alone for up t o 10 days. The culture medium was collected and replaced daily. After predetermined times in culture, the tissue was extractedwith4 M GdnHCl as described above. Proteo- glycans were isolated from tissue extracts and medium samples by ion-exchange chromatography on Q-Sepharose as described previously [1]. The s amples were then dialysed again st distilled H 2 O containing proteinase inhibitors, lyophilized, and dissolved in 1 mL of 0.1 M Tris/0.1 M sodium acetate pH 7.0 containing proteinase inhibitors [10]. The dried samples were then digested with chondroitinase ABC (0.0375 U) and keratanase (0.075 U) at 37 °C for 24 h in the presence of proteinase inhibitors [11]. S amples were subjected t o electrophoresis on a 4–15% gradient polyacrylamide/SDS slab gel. The gel was then fixed in a solution of 30% (v/v) methanol and 10% (v/v) acetic acid for 30 min, soaked in Amplify for 30 min, dried and exposed to X-ray film at )20 °C for approximately 40 days 3 [10]. Intracellular degradation of 35 S-labelled small proteoglycans In or der t o determine the rate o f i ntracellular degradation of 35 S-labelled small proteoglycans, bovine tendon from a single metacarpophalangeal joint was incubated with [ 35 S]sulfate for 6 h as described above and then maintained in DMEM alone for 5 days to allow time for loss of the 35 S-labelled large proteoglycans from the tissue cultures. For the subsequent days (days 6–15) the tissue was maintained in culture under the conditions described below. The rate of intracellular proteoglycan catabolism by tendon explants was determined from the amount of radiolabelled Ó FEBS 2004 Catabolism of proteoglycans in tendon 1 (Eur. J. Biochem. 271) 3613 sulfate appearing in the medium on each day. For this, aliquots of the culture medium were applied to Sephadex G-25 (PD-10) columns and the amount of [ 35 S]sulfate determined fr om the amount of 35 S-radioactivity i n t he total volume of the columns. The rate of release of 35 S-labelled proteoglycans into the culture medium in these experiments was determined by the 35 S-radioactivity that eluted in the excluded volume on Sephadex G-25 columns. The percent- age of 35 S-labelled proteoglycans remaining in the matrix was determined as described above. Treatment of data Previous work has shown that there is variation between animals in the absolute rates of metabolism of macro- molecular components o f the extracelullar matrix o f synovial connective tissues. Because the amount of tissue was limiting, this only e nabled points to be r epeated in duplicate. Therefore, individual experiments were repeated at least three times using tissue from different animals. Results Determination of the loss of 35 S-labelled proteoglycans from the extracellular matrix of tendon explant cultures Explants of bovine tendon were in cubated with [ 35 S]sulfate for 6 h and then maintained in culture in DMEM for up to 10 days. Medium fractions w ere collected daily and the tissue extracted at the end of the culture period with 4 M GdnHCl followed b y 0 .5 M NaOH as des cribed above. Approximately 75% of 35 S-labelled proteoglycans were extracted from the tendon matrix with GdnHCl (data not shown). T he medium fractions and the GdnHCl and NaOH extracts were analyzed by size exclusion chromatography on Sephadex G-25 as described above. All the radiolabelled material appearing in the total v olume of the columns was showntobefree[ 35 S]sulfate because it was all precipitated by barium acetate (data not shown). Figure 1 shows the rate at which 35 S-labelled proteoglycans and free [ 35 S]sulfate appeared in medium samples on e ach d ay in culture. During the first two days of culture the appearance of free [ 35 S]sulfate in the culture medium was attributable to unincorporated [ 35 S]sulfate following incubation of the tissue with [ 35 S]sulfate on day 0. Over the subsequent days (days 3–10), both the free [ 35 S]sulfate and 35 S-labelled proteoglycans appeared in the culture medium at a similar rate. Any re-use of free [ 35 S]sulfate during the first two days of explant culture would be minimal as the specific radioactivity of the radiolabelled s ulfate would b e markedly reduced by the sulfate content of D MEM. Figure 2A shows that there was a faster rate of loss of 35 S-labelled proteo- glycans from the matrix in the first four days o f culture and approximately 60% of 35 S-labelled proteoglycans remained in the matrix after 10 days in culture. Kinetics of loss of large and small 35 S-labelled proteoglycans from the extracellular matrix of tendon explant cultures The a mount of 35 S-radiolabel ass ociated with large and small proteoglycans remaining in the matrix of tendon explants described in Fig. 2A was determined from tissue extracts on days 0, 2, 4, 6, 8 and 10. These extracts were subjected to gel filtration on a column of Sepharose CL-4B eluted under dissociative conditions (Fig. 3). It is evident that there are two 35 S-labelled proteoglycan p eaks, a minor peak (K av  0.05) representing t he large p rote- oglycans and a major peak (K av  0.5) representing the small p roteoglycans. T he proportion of 35 S-radioactivity associated with the large proteoglycans decreased from 16.6% on day 0 (the day of incubation with [ 35 S]sulfate) to 2.1% on day 10, whereas that f or the small proteo- glycans showed an apparent increase from 83.4% on day 0 to 97.9% by day 10. This indicates that there is a preferential loss of the newly synthesized 35 S-labelled large proteoglycans from the extracellular matrix of tendon. Although 40% of the 35 S-labelled proteoglycans was lost from the extracellular matrix of tendon over the 10 day culture period, the hydrodynamic size of each 35 S-labelled proteoglycan species extracted from the matrix of the tissue immediately after incubation of the tissue with [ 35 S]sulfate and at various time points in culture remained constant. The presence of unincorporated [ 35 S]sulfate early in the culture period (days 0 and 2 ) is i ndicated in t he elution profiles in the total volume of the column 4 . The percentage of 35 S-labelled large and small proteo- glycans remaining in the matrix at various times after incubation with [ 35 S]sulfate was determined by multiplying the percentage of each proteoglycan species present in the tissueextractsondays0,2,4,6,8and10ofthe culture period (Fig. 3) by the percentage of 35 S-labelled proteoglycans r emaining in the m atrix o n t he corresponding day of culture (Fig. 2A). This was then expressed as the percentage of the amount of each proteoglycan species present i n the tissue on d ay 0 (Fig. 3; top) in order to determine the kinetics of loss of the large and small Fig. 1. Rate of appearance of 35 S-labelled proteoglycans and [ 35 S]sul- fate in to the c ulture medium of explant c ultures of tendon. The p roximal region of bovine de ep flexor tendon was incubated with [ 35 S]sulfate as described in Experimental procedures, and cultured in DMEM alone for10days.Therateofappearanceof 35 S-labelled proteoglycans (d) and [ 35 S]sulfate (s) into the culture medium from bovine tendon explant cultures was determined by analysis of medium samples from each day of culture period on columns of Sephadex G-25. The error bars represent the range of duplicate samples. 3614 T. Samiric et al.(Eur. J. Biochem. 271) Ó FEBS 2004 35 S-labelled proteoglycans in tendon. Figure 2B shows that the loss of 35 S-labelled small proteoglycans from the extracellular matrix was much slower (T ½ > 20 days) compared to the 35 S-labelled large proteoglycans (T ½  2 days). It is evident from Fig. 2B that over 85% of large proteoglycans were lost from the tissue within the first six days after incubation of the tissue with [ 35 S]sulfate, whilst only 20% of small proteoglycans were lost over this time period. Characterization of 35 S-labelled proteoglycans remaining in the matrix and released into the medium of tendon explant cultures by fluorography To analyze the 35 S-labelled proteoglycan core proteins isolated from either the tissue or released into the culture medium, t endon was incubated with [ 35 S]sulfate f or 6 h prior to being maintained in culture in DMEM for up to 10 days. Radiolabelled proteoglycans present in the matrix Fig. 2. Percentage of 35 S-labelled proteoglycans remaining in the extracellular matrix of tendon explants cultures. (A) The proximal region o f bovine deep flexor tendon was inc ubated with [ 35 S]sulfate and m aintained in D MEM for 10 days. The percentage of 35 S-labelled proteoglycans remaining i n the matrix of tendon c ultures o n each d ay after incubation with [ 35 S]sulfate was determined as described in Experimental procedures. The error b ar represents the r ange of duplicate samples. (B) The percentage of 35 S-labelled large proteogly- cans (d)and 35 S-labelled small proteoglycans (s)remaininginthe tissue at each time after incubation of bovine tendon with [ 35 S]sulfate was dete rmined as described in Results. The error bars represent the range of duplicate samples. Fig. 3. Elution profiles on Sepharose CL-4B of the 35 S-labelled pro- teoglycans remaining in the matrix of tendon cultures maintained in DMEM. On the days indicated, tissue samples from the experiment described in Fig. 2 were extracted with 4 M GdnHCl and aliquots of the 35 S-labelled proteoglycans we re app lied to a co lu mn of Se pharo se CL-4B eluted with a buffer containing 4 M GdnHCl. In each profile, the amount of 35 S-labelled proteoglyc ans extracted fro m th e tissue o n each day is expressed a s a pe rcen tage of the 35 S-labelled p roteoglycans extracted on day 0. The values in parentheses refer to the relative percentage of large and sm all p ro teoglyca n s pecie s p resent 7 on the day of extraction. Ó FEBS 2004 Catabolism of proteoglycans in tendon 1 (Eur. J. Biochem. 271) 3615 and culture medium were digested with chondroitinase ABC and keratanase, which results in the removal of most of the glycosaminoglycan chains but leaves 35 S-radio- labelled glycosaminoglycan stubs associated with the core protein. The partially deglycosylated core proteins were then subjected to electrophoresis on a 4–15% gradient polyacrylamide/SDS large gel followed by fluorography as described in Experimental procedures. Figure 4 (lane i) shows that three distinct high molecular mass bands above 300 kDa were present in tendon matrix immediately after incubation with [ 35 S]sulfate. Based on our previous work it is likely that these bands represent intact core protein of aggrecan and V 0 and/or V 1 splice-variants of versican [1]. With time in culture, a distinct band at  300 kDa (indicated by asterisk) appeared, and remained in the matrix over the culture period of 10 days (lanes ii and iii). The precise identity of this band is not known but it is likely to be a product of the proteolytic processing of the core protein of aggrecan or versican. A series of weak bands ranging b etween 80 to above 250 kDa were a lso present and these are likely to represent degradation products of the large proteoglycans that are retained in the m atrix. The majority of rad iolabelled material present in the matrix of fresh tendon was associated with the band ranging between 37 and 45 kDa. This band corresponds to the decorin core protein; also present are small levels o f b iglycan core protein [1]. A number of bands were also observed at 33 kDa and below, which we have shown to be degradation products of decorin [1]. A diffuse band at 50 kDa, which is likely to represent intact fibromodulin or degradation products of large proteoglycans, was also evident [10]. It was apparent that the decorin core protein of 43 kDa (lane i) present in tissue immediately after incubation with [ 35 S]sulfate, decreased in size with time in culture (lanes ii and iii) indicating extracellular p rocessing of d ecorin core p rotein. It is possible that newly synthesized decorin contains an intact amino- terminal propeptide which is remove d with time in culture by the action of proteinases present in the e xtracellular matrix of the tissue [12,13]. Further proteolytic processing of decorin core protein was shown by the presence of additional distinct bands at 25 kDa and below (lanes ii and iii). These observations indicate that degradation of core proteins of newly synthesized small proteoglycans occurs and that fragments are retained within the matrix. Figure 4 (lanes iv and v) shows the proteoglycan core proteins released i nto the medium of explant cultures after 3 and 6 days in culture, respectively. A number of distinct bands of over 250 kDa and a series of bands ranging between 75 and 160 kDa are present in t he medium and we have previously shown that they represent catabolic prod- ucts of aggrecan a nd versican [1]. It m ust b e pointed out that the amount of 35 S-radioactivity associated with these high molecular mass peptides is directly attributable to the high density of sulfate groups associated with these large proteoglycans. Intracellular catabolism of 35 S-labelled small proteoglycans by tendon explant cultures Because it was shown that [ 35 S]sulfate appeared in the culture medium throughout the culture period (Fig. 1), experiments were performed to determine if this was due to intracellular d egradation of 35 S-labelled s mall proteogly- cans. Bovine deep flexor tendon was maintained in culture in DMEM for 5 days after incubation with [ 35 S]sulfate to allow for the loss of the majority of the radiolabelled large proteoglycans (Fig. 2B). Cultures were then maintained in DMEM at 37 °Cor4°C for a subsequent 10 days to determine the effect of reduced cellular activity on the appearance of free [ 35 S]sulfate in the medium. In some cultures, the temperature was switched at the mid-point of the culture period to determine whether the effect of low temperature on the generation of free [ 35 S]sulfate was reversible. Figure 5A shows that in cultures maintained at 37 °C, there was a continuous rate of formation of free [ 35 S]sulfate in the culture medium. The r ate of generation o f [ 35 S]sulfate in t he medium was reduced by over 90% in cultures maintained at 4 °C, suggesting that metabolically active cells were required for this process. This reduction was also demonstrated when cultures were switched from 37 °Cto4°C on day 10 o f the culture period, whereas in cultures that we re initially maintained at 4 °C, there was an Fig. 4. Analysis of 35 S-labelled proteoglycan core proteins p resent in th e matrix or medium of explant cultures of tendon. Newly synthesized 35 S-labelled p roteoglycans remaining in t he matrix or released into the medium of tendon explant cultures after 10 days in culture were iso- lated as described in Experimental procedures and d ige sted with chondroitinase ABC and keratanase, prior to electrophoresis on a 4–15% polyacrylamide/SDS large gel. The gel was subjected to fluo- rography as described in Experimental procedures. Lanes show pep- tides present in (i) fresh tendon t issue, (ii) t issue after 6 days in culture, (iii) tissue after 10 days in culture, (iv) days 1–3 pooled medium, and (v) days 4–6 pooled medium. Approximate molecular mass of observed pept ides are given. 3616 T. Samiric et al.(Eur. J. Biochem. 271) Ó FEBS 2004 apparent increase in the rate of [ 35 S]sulfate appearance when these cultures were switched to 37 °C, demonstrating that this effect was reversible. In contrast to the rate of formation of [ 35 S]sulfate, Fig. 5B shows that there was an increase by 40% of 35 S-labelled proteoglycans appearing in the culture medium of tendon explants maintained at 4 °C. The percentage of 35 S-labelled proteoglycans remaining in the matrix of cu ltures maintained at 37 °C (calculated from both the release of 35 S-labelled proteoglycans and the appearance of [ 35 S]sulfate with t ime in culture) w as approximately 80% by the end of the culture period on day 15 as shown in Fig. 5C. However, the loss of 35 S-labelled proteoglycans was reduced in cultures main- tained at 4 °C, where about 95% of 35 S-labelled proteo- glycans remained in the matrix by the end of the culture period on day 15. The work described above suggests that small proteo- glycans a re tak en up by the cells and digested intracellularly. To demonstrate that the lysosomal s ystem is involved i n the appearance of free [ 35 S]sulfate in the culture medium, tendon cultures were maintained in DMEM contain ing 10 m M ammonium chloride following 5 days in culture in DMEM alone. Ammonium chloride is a lysosomotropic amine and acts by raisin g the intralysos omal pH which inhibits the activity of lysosomal enzymes, and is known to be an effective reversible inhibitor of lysosomal function at low concentration [14]. Figure 6A shows that in cultures maintained in DMEM containing 10 m M ammonium chloride, the rate of [ 35 S]sulfate appearing in the medium was suppressed by approximately 78% compared with control c ultures. This suppression was further demonstrated when cultures were switched on day 10 from DMEM alone to DMEM containing 10 m M ammonium chloride. When cultures were switched on day 10 from DMEM containing 10 m M ammonium chloride to DMEM alone, the rate of [ 35 S]sulfate a ppearance was restored, demonstrating that this effect was reversible. The rate of release of 35 S-labelled proteoglycans into the culture medium was increased by approximately 107% in cultures maintained in the presence of ammonium chloride (Fig. 6B). The percentage of 35 S-labelled proteoglycans remaining in the matrix in DMEM alone was approximately 80% by the end of the cultureperiodonday15asshowninFig.6C.However,the loss of 35 S-labelled proteoglycans was reduced in cultures maintained in the presence of ammonium chloride, where about 85% of 35 S-labelled proteoglycans remained in the matrix by the end of the culture period on day 15. Discussion This study showed that the loss o f the large aggregating proteoglycans (aggrecan and V 0 and/or V 1 splice-variants of versican) that make up approximately 17% of the 35 S-labelled pool of newly synthesized proteoglycans was rapid, with a half-life of about 2 days (Fig. 4). These findings are consistent with studies using other joint connective tissues such as articular cartilage [15] and collateral ligament [10]. In the case of articular cartilage , it has been shown that t he majority of newly s ynthesized aggrecan remains closely associated with the chondrocytes [16,17]. However, the majority of the chemical pool of aggrecan resides in the interterritorial matrix and it is this Fig. 5. Effect of r educ ed temperature on the rate of formation of [ 35 S]sulfate and release of 35 S-labelled proteoglyc ans from tend on explant cultures. Explant cultures of deep flexor tendon were incubated with [ 35 S]sulfate as described in Experimental procedures and then maintained in DMEM for 5 days prior to analysis. Tissue was sub- sequently cultured for a further 10 days in DMEM at 37 °C(d), DMEM at 4 °C(s), DM EM at 37 °C which was switched t o 4 °Con day 10 ( ,), or D MEM at 4 °C which was switched t o 37 °C on day 10 (.). The culture medium was collected daily and analyzed for the presence o f [ 35 S]sulfate and 35 S-labelled proteoglycans. F rom this data, (A) the rate of appearan ce of [ 35 S]sulfate, (B) the rate of release of 35 S-labelled p roteoglycans, and (C) t he percentage of 35 S-labelled proteoglycans remaining in the matrix were determined, as described in Experimental p rocedur es. T he error bars represent the r ange of duplicate samples over the remaining 10 days. Ó FEBS 2004 Catabolism of proteoglycans in tendon 1 (Eur. J. Biochem. 271) 3617 population that is responsible for the biomechanical prop- erties of cartilage. Work has shown that this population of aggrecan turns over very slowly, with a half-life in excess of 3.5 years [18]. If this i s applied to the present s tudy, it is likely that newly synthesized agg recan and v ersican may be closely associated with tendon cells where the turnover is mediated by proteolytic enzymes originating from tendon cells. Indeed, we have shown that the catabolism of aggrecan in tendon appears to be exclusively attributed to aggrecanase proteinases whereas the catabolism of versican may involve aggrecanase as well as othe r proteinases [1]. It is likely that these enzymes are responsible for the rapid turnover of the newly synthesized pool of large proteoglycans, as the resulting radiolabelled co re protein fragments are of similar size to those previously reported by our laboratory for the chemical pools of large aggregating proteoglycans present in tendon [1]. Furthermore, it has been reported that the aggrecanase proteinases A DAMTS-4 and ADAMTS-5 are expressed i n bovine tendon cells [2], bu t a t d ifferent stages of development of the animal. We have observed the expres- sion of both ADAMTS-4 and ADAMTS-5 in bovine tendon cells from mature cattle (T. Samiric, M.Z. Ilic & C.J. Handley, unpublished data) 5 . In contrast to the rapid rate of loss of newly synthesized large proteoglycans, the n ewly synthesized small proteo- glycans were lost slowly from the matrix of tendon cultures with a h alf-life of greater than 20 days, which is consistent with fin dings f rom e arlier s tudies in explant cultures o f tendon [19], articular cartilage [15] and ligament [4,10]. This slow loss of newly synthesized small proteoglycans may be indicative of their association with other matrix molecules, particularly Type I collagen fibres [20], and it is possible that the turnover of this group of proteoglycans may be coordinated with the turnover of other matrix macromolecules. However, some of t he radiolabelled decorin undergoes proteolytic cleavage and these products are either retained within the matrix or lost to the culture medium in a similar manner to that observed for the chemical pool [1]. Approximately 60% of the 35 S-labelled decorin that was lost from the matrix was taken up by t he tendon cells and degraded within the lysosomal system. This was shown by the gene ration o f f ree [ 35 S]sulfate by t endon e xplant cultures throughout the culture period. This finding is supported by similar studies using ligament explant cultures [4]. The cellular uptake and subsequent degradation of decorin has been observed in a variety of cells of mesenchymal origin [4,21]. It has been shown that the leucine-rich repeat region of decorin binds to specific receptors present in the plasma membrane and endo somes of s kin fibroblasts, osteosarcoma cells and c hondrocytes [5,22]. Upon entering the c ell by endocytosis, decorin is subsequently transported to the lysosomes. Previous work has shown that at least two intracellular pathways are involved in the c atabolism of endogenously radiolabelled proteoglycans associated with the c ell s urface in rat ovarian granulosa cells [23]. O ne pathway leads to a rapid and complete intralysosomal degradation resulting in th e release of [ 35 S]sulfate. In the second pathway, the rate of degradation is slower and commences with extensive p roteolysis, generating glycos- aminoglycan chains bound to peptides before final hydro- lysis takes p lace [23]. Fig. 6. Effect of ammonium chlo ride on the r ate of f orma tion of [ 35 S]sulfate and release of 35 S-labelled proteoglycans from tendon explant cultures. Explant cultures of deep fl exor tendon were incubated with [ 35 S]sulfate as described in Experimental procedures and then maintained in DMEM for 5 days prior to analysis. Tissue was sub- sequently cultured f or a further 10 days in DMEM alone (d), DMEM containing 10 m M ammonium c hloride (s), DMEM alone which was switched to DMEM containing 10 m M ammonium chloride on day 10 (,), or DMEM containing 10 m M ammonium chloride which was switched to DMEM alone on day 10 (.). The culture medium was collected daily and analyzed for the p rese nce of [ 35 S]sulfate and 35 S-labelled proteoglycans. F rom t his d ata, (A) the rate of ap pearance of [ 35 S]sulfate, (B) t he rate of rele ase of 35 S-labelled pr oteoglycans, and (C) the percentage of 35 S-labelled proteoglycans remaining in the matrix were determined, as described in Experimental procedures. The error b ars represent the range o f duplicate samples over the remaining 10 days. 3618 T. Samiric et al.(Eur. J. Biochem. 271) Ó FEBS 2004 This study indicates that the intracellular degradation of decorin requires metabolically active cells including a functional lysosomal system because this process was inhibited at 4 °C a nd in the p resence o f a mmonium chloride (Figs 5 and 6). In addition, when thes e treatments w ere applied t o tendon explant c ultures there was an inhibition of the intracellular degradation of decorin and a simultaneous increase in the appearance of 35 S-labelledofdecorininthe medium throughout the culture period (Figs 5 and 6), al beit to different degrees. This enhanced loss of decorin by the pathway that results in the loss of decorin from the extracellular matrix has also been observed in ligament explant cultures [4]. It has previously been reported that decorin is only taken up by cells if it is not bound to other extracellular matrix molecules [24]. However, little is known about the nature of interactions of newly synthesized decorin with other extracellular components and its distri- bution within the matrix of fibrous connective tissues. It is possible t hat a proportion of newly synthesized decorin remains located close to the cell and may be loosely associated with the cell membrane and/or extracellular matrix. This pool of decorin is likely to be subjected to intracellular degradation. The inhibition of the cellular uptake of newly synthesized decorin appears to result in more of this pool of decorin being lost from the tissue into the culture medium. This may involve displacement of newly synthesized decorin that is further away from the cell and subsequent release to the medium, thus reducing the accumulation of this proteoglycan within the extra- cellular matrix of tendon. The low level of loss of radio- labelled decorin suggests that a significant proportion of the newly synthesized decorin is retained in the extra- cellular m atrix i n strong interactions with other extracellular matrix macromolecules where the core protein of this proteoglycan can undergo p roteolytic processing (F ig. 4; lanes ii and iii). The work presented in this paper supports previous observations which show that the catabolism of large and small proteoglycans follow distinct separate pathways. Furthermore, it is evident that in both tendon and ligament [4] t he processes involved in the catabolism of proteoglycans are similar and this is not unexpected considering the similarity in the structure and organization of these two dense connective tissues. In both tissues the intracellular degradation pathway plays a significant role in the catabo- lism of newly synthesized s mall proteoglycans. In the case of tendon this pathway represents about 60% of the radio- labelled pool of small proteoglycans and in t he case of ligament represents 30% of this pool [4]. This raises the question of whether this pathway is also involved in the catabolism of the chemical pool of small proteoglycans that are present in the extracellular matrix of these tissues. Furthermore, the contribution of this intracellular pathway of catabolism of small proteoglycans needs to be taken into account in stu dies investigating the c atabolism o f small proteoglycans in d ense connective tissues in pathological conditions. 6 Acknowledgements We wish to thank the Arthritis Foundation of Australia and the Faculty of Health Sciences, La Trobe University for s upport. References 1. Samiric, T., Ilic, M.Z. & Handley, C.J. (200 4) Characterization of proteoglycans and th eir c atabolic p rodu cts i n t endon a nd e xp lant cultures of tendon. Matrix Biol. 23, 127–140. 2. Rees, S .G., Flannery, C .R., Little, C.B., Hughes, C.E., Caterson, B. & Dent, C.M. (2000) Catabolism of aggrecan, dec- orin and biglycan in tendon. Biochem. J. 350, 181–188. 3. Vogel, K.G. & Meyers, A.B. (1999) Proteins in the tensile region of adult b ovine d eep fle xor tend on. Clin. Orthop. 367, S344– S355. 4. Winter, A.D., Campbell, M.A., Robinson, H.C. & Handley, C.J. (2000) Catabolism of newly synthesized decorin by explant cultures of bovine ligament. Matrix Biol. 19, 129–138. 5. Hausser, H., Ober, B., Quentin-Hoffmann, E., Schmidt, B. & Kresse, H. (1992) Endocytosis of different members of the small chondroitin/dermatan su lfate proteo glycan family. J. Biol. Chem . 267, 11559–11564. 6. Hausser, H., S chonherr, E., Muller, M., Liszio, C., Bin, Z., Fis her, L.W. & K resse, H. ( 1998) Receptor-mediated endoc ytosis of decorin: involvement of leucine-rich repeat structures. Arch. Bio- chem. Biophys. 349, 363–370. 7. Robinson, H.C. & Lindahl, U. (1981) Effect of cyclohexamide, beta- D -xylosides and beta- D -galactosides on heparin biosynthesis in mouse mastocytoma. Biochem. J. 94, 575–586. 8. Hey, N.J., Handley, C.J., Ng, C.K. & Oakes, B.W. (1990) Char- acterisation and synthesis of macromolecules by adult collateral ligament. Biochim. Biophys. Acta 1034, 73–80. 9. Oegema, T.R. Jr, Hascall, V.C. & Eisenstein, R. (1979) Char- acterization of bovine aorta proteoglycan extracted with guanidine hydrochloride in the presence of protease inhibito rs. J. Biol. Chem. 254, 1312–1318. 10. Campbell, M.A., Winter, A.D., Ilic, M.Z. & Handley, C.J. (1996) Catabolism and loss of proteoglycans from cultures of bovine collateral ligament. Arch. Biochem. Biophys. 328, 64–72. 11. Oike, Y., Kimata, K., Shinomura, T. & Suzuki, S. (1980) Proteinase activity in chondroitin l yase (chondroitinase) an d endo- beta- D -galactosidase (keratanase) preparations and a method to abolish their proteolytic effect on proteoglycan. Biochem. J. 191, 203–207. 12. Oldberg, A., Antonsson, P., Moses, J. & Fransson, L.A. (1996) Amino-terminal deletions in the deco rin core protein leads to t he biosynthesis of proteoglycans w ith shorter glycosaminoglycan chains. FEBS Lett. 386, 29–32. 13. Roughley, P.J., White, R.J. & Mort, J.S. (1996) Presence of pro- forms of decorin and b iglycan in human articular cartilage. Biochem. J. 318, 385–397. 14. Seglen, P.O. & Gordon, P.B. (1980) Effects of lysosomotropic monoamines, diamines, amino a lcohols, and other amino com- pounds on pro tein degrad ation an d protein synthesis in isolated rat hepatocytes. Mol. Pharm. 18, 468–475. 15. Campbell, M.A., Handley, C.J., Hascall, V.C., C ampbell, R.A. & Lowther, D .A. (1984) Turnover of proteoglycans in cultures of bovine articular cartilage. Arch. Biochem. Biophys. 234, 275– 289. 16. Winter, G.M., Poole, C.A., Ilic, M.Z., Ross, J.M., Robinson, H.C. & Handley, C.J. (1998) Identification o f distinct metabolic pools of aggrecan and their relationship to type VI collagen in the chondrons of m ature bovine a rticu lar cartilage explants. Connect. Tissue Res. 37, 277–293. 17. Handley, C.J., Winter, G.M., Ilic, M.Z., Ross, J.M., Poole, C.A. & Robinson, H.C. (2002) Distribution of newly synthesised aggrecan in explant cultures of bovine cartilage trea ted with retinoic acid. Matrix Biol. 21, 579–592. 18. Maroudas, A., Bayliss, M.T., Uchitel-Kaushansky, N., Schneir- derman, R. & Gilav, E. (1998) Aggrecan turnover in human Ó FEBS 2004 Catabolism of proteoglycans in tendon 1 (Eur. J. Biochem. 271) 3619 articular cartilage: use of as partic acid racemization as a m arker of molecular age. Arch. Biochem. Biophys. 350, 61–71. 19. Koob, T.J. & Vog el, K.G . (19 87) Proteogly can s ynthesis i n o rgan cultures from regions of bovine tendon subjected to different mechanical forces . Biochem. J. 24 6, 589–598. 20. Scott, J.E. & Orford, C.R. (1981) Dermatan sulphate-rich pro- teoglycanassociateswithrattail-tendoncollagenatthedbandin the gap region. Biochem. J. 197, 213–216. 21. Gotte, M., Kresse, H. & Hausse r, H. (1995) Endocytosis of decorin by bovine aortic endothelial cells. Eur. J. Cell Biol. 66 , 226–233. 22.Hausser,H.,Hoppe,W.,Rauch,U.&Kresse,H.(1989) Endocytosis of a small dermatan sulphate proteoglycan. Identifi- cation of binding proteins. Biochem. J. 263, 137–142. 23. Yanagishita, M. & H ascall, V.C. (1984) Metabolism of pro- teoglycans in rat ovarian granulosa cell culture. Multiple intracellular degradative pathways and the effe ct of chloroquine. J. Biol. Chem. 25, 10270–10283. 24. Schmidt, G. , Hausser, H. & Kresse, H. (1 990) Extracellular accumulation o f small d ermatan sulphate proteoglycan II by interference with the secretion-recap ture pathway. B ioc hem . J. 266, 591–595. 3620 T. Samiric et al.(Eur. J. Biochem. 271) Ó FEBS 2004 . Large aggregating and small leucine-rich proteoglycans are degraded by different pathways and at different rates in tendon Tom Samiric, Mirna Z. Ilic and. were degraded by the intracellular pathway present in tendon cells. This work shows that the pathways of catabolism for large aggregating and small leucine-rich proteoglycans