Jurrrrial of Orthopaedic Research 1014-22 Raven Prew, Ltd New York 0 1992 Orthopaedic Research Society
.
Effects of Free and Bound Insulin-Like Growth Factors on Proteoglycan Metabolism in Articular Cartilage Explants Gregory H. Tesch, "Christopher J. Handley, Hugh J. Cornell, and ?Adrian C. Herington Department of Applied Chemistry, Royal Melbourne Institute of Technology, Melbourne, Victoria; *Department of Biochemistty, Monash University, Clayton, Victoria; and iPrince Henry's Institute of Medical Research, Melbourne, Victoria, Australia
Summary: This article describes the effects of bound forms of insulin-like growth factors (IGFs) on proteoglycan metabolism by bovine articular cartilage in explant culture. When these growth factors were added to articular cartilage explants complexed with their native serum binding proteins (BPs), both IGF-1-BP complex and IGF-11-BP complex stimulated proteoglycan synthesis to different degrees over a 3-day pcriod. When added to the medium of cultures of articular cartilage over 5 days. IGF-11-BP complex induced high rates of synthesis and low rates of catabolism of proteoglycans, giving rise to tissue levels of proteoglycan similar to those observed in fresh tissue. When articular cartilage was maintained in culture with the same concentration of ICF-I-BP complex, tissue levels of proteoglycans fell over the culture period because of lower rates of proteoglycan synthesis. Analysis of the proteoglycans synthesized by articular cartilage in the presence of free or bound IGF-I or TGF-TI showed that these growth factors stimulated the rate of synthesis of the large proteoglycan species present in cartilage but did not affect the synthesis of the small proteoglycans. Key Words: Insulin-like growth factorArticular cartilage-Proteoglycan-Synthesis-Catabolism.
Articular cartilage has the unique mechanical property of being able to withstand compressive loads. This ability is attributed to the presence of highly hydrated proteoglycans contained within the collagen network of the extracellular matrix (for review see 13,20). These proteoglycans are continuously being turned over within the matrix, and so the regulation of proteoglycan metabolism is essential for the maintenance and function of articular cartilage. Studies have suggested that insulin-like growth
factors (IGFs) present in serum are responsible for the maintenance of proteoglycan synthesis by explant cultures of cartilage (17,19,27). It is therefore apparent that IGFs may play a major role in the regulation of growth and maintenance of articular cartilage and in the repair of damaged tissue. In normal adult human serum, IGF-I1 is present in higher quantities (three tq five times) than IGF-I (28). In addition, both IGF-I and IGF-I1 are found in serum almost entirely bound to binding proteins (BPs) to form a circulating complex with a molecular weight of -150,000 daltons (8). These BPs are believed to regulate the action of the IGFs toward their target tissues (1). Where no measurable free IGF is present, the ability of whole serum to stimulate proteoglycan synthesis suggests that the
Received November 1 I , 1990; accepted June 20, 1991. Address correspondenceand reprint requests to C. J . Handley at Department of Biochemistry, Monash University, Clayton, Victoria 3168, Australia.
14
INSULIN-LIKE GROWTH FACTORS AND CARTILAGE METABOLiSM
15
bound forms of IGFs are active in proteoglycan stimulation or that the free forms are released in situ (14). This article describes the effects of the bound forms of IGF-I and IGF-I1 on proteoglycan metabolism in bovine articular cartilage in explant culture.
contaminant with NSILA in this fraction (7). NSILA-p was not present in the IGF-I-BP complex fraction. The IGF species and NSILA-p are the only substances in serum that possess NSILA (15).
MATERIALS AND METHODS
p, and the IGF-RP complexes was determined in a
Determination of IGF Activity The biological activity of IGF-I, IGF-11, NSILA-
Materials Recombinant human IGF-I was generously supplied by Dr. Anna Skottner, Kabi Vitrum, Stockholm, Sweden. Purified human IGF-I1 was kindly donated by Dr. R. Baxter, Royal Prince Alfred Hospital, Camperdown, N.S.W., Australia. Human plasma was obtained from healthy adult blood bank donors. All other materials were obtained as described previously (7,14,18). Purification of IGF Complexes Purification of the IGF-BP complexes was carried out as previously described by Cornell and associates (7). Briefly, IGF-I-BP complexes and IGF11-BP complexes were purified from human plasma by ion-exchange chromatography on DEAESephadex A-50 using a stepwise NaCl gradient to elute each complex preferentially. The IGF-I-BP complex and IGF-11-BP complex fractions were further purified by affinity chromatography on Con-A Sepharose. This procedure produced an IGF-I-BP complex fraction with a 74-101-fold purification and an IGF-11-BP complex fraction with 46-54-fold purification relative to plasma-specific activities of IGF-I and IGF-11, respectively, based on the nonsuppressible insulin-like activity (NSILA) bioassay. Analytical chromatography on Sephacryl S-200 was used to confirm the presence of high Mr IGF-BP complexes in each purified fraction. Both the IGFI-BP and the IGF-11-BP complex fractions contained complexes with Mr 50,000-160,000. No free IGF was detected in either of these fractions, indicating that all the IGFs present were in a bound form. The IGF-11-BP complex contained no detectable free or bound immunoreactive IGF-I, whereas the partially purified IGF-I-BP complex contained 14% (w/w) IGF-11-BP complex. The IGF-11-BP complex contained NSILA-p. a protein with nonsuppressible insulin-like activity that precipitates with acid-ethanol solution (22), which was the only
-
standard NSILA bioassay by monitoring the incorporation of [U-'4Cjglucose into lipid by isolated rat adipocytes in the presence of excess anti-insulin serum (9). The activity of each species assayed was subsequently expressed in terms of units of insulinlike activity based on an insulin standard. In the case of the IGF-BP complexes, the NSILA activity was determined by assaying the amount of low Mr-free IGF liberated after each of the IGF complex fractions had been subjected to gel filtration on a Sephadex G-75 column (1.2 cm X 54 cm) equilibrated and elutcd with 1% (viv) formic acid (19).
Cartilage Cultures Bovine articular cartilage explants were obtained from the metacarpalphalangeal joints of 1-2-yearold steers and prepared as described previously (14). Samples of sliced cartilage (10G150 mgivial) were maintained in culture for 5 days in a modification of Dulbecco's Modified Eagle Medium (11) or the same medium supplemented with fetal calf serum, bovine serum albumin, IGFs, IGF-BP complexes, or NSILA-p. Unless stated, NSILA-p, IGFs, and IGF-BP complexes were added to medium containing 0.01% (wiv) bovine serum albumin to prevent nonspecific absorption. The batch of bovine serum albumin used was pretested to ensure that it had no stimulatory activity on proteoglycan synthesis by articular cartilage explants. The medium containing NSILA-p and growth factors was sterilized by filtration through microporous membranes. In experiments determining the rate of loss of proteoglycans from cartilage explants, cartilage from the same joint (-3 g) was maintained in culture for 3 days in 30 ml of medium containing 20% (vh) fetal calf serum before use. We have previously noted variation in the rates of proteoglycan metabolism by different samples of bovine articular cartilage when placed in explant culture, primarily due to animal variation (5,9). All experiments were therefore carried out using tissue
J Orthop Rrs, Vol. 10, N o . I , 1992
G . H . TESCH ET AL.
16
obtained from a single metacarpalphalangeal joint; further experiments were carried using tissue from different animals to confirm any one observation. Measurement of Proteoglycan Synthesis Cartilage cultures were preincubated for 1 h in fresh medium and then incubated for 2 h in 2 ml of medium alone, containing 20 FCi of [35S]sulfate/ml (19). The labeled proteoglycans were extracted with 4 M guanidinium chloride/0.5 M sodium acetateacetic acid, pH 6.0, in the presence of proteinase inhibitors for 2 days at 4°C (21). The tissue was then extracted with 2 ml of 0.5 M NaOH for 16 h at room temperature. Some of the cartilage explants were also extracted directly with 2 ml of 0.5 M NaOH. The rate of [35S]sulfateincorporation into total proteoglycans (dpmi100 mg wet weight of cartilage per 2 hj was determined by size exclusion chromatography on Sephadex G-25 (PD-10) columns (14). The rate of [35S]~~lfate incorporation (dpm/lOO mg wet weight per 2 hj into individual species of proteoglycan was calculated by multiplying the rate of incorporation of [ 3 5 ~ ~ s u l f ainto t e total proteoglycans (dpm/100 mg wet weight per 2 h) by the percentage of the radiolabel appearing in the the particular species of proteoglycan. The latter was determined by size exclusion chromatography on Sepharose CL-2B and is described below. Measurement of Proteoglycan Catabolism Proteoglycan turnover in cartilage explant cultures was determined as described by Handley and Campbell (12). After 5 days in culture in medium containing 20'30 (viv) fetal calf serum, articular cartilage (-3 g) from a single joint was incubated for 6 h with 10 ml of medium containing fetal calf serum The tissue was then and 300 pCi of [35S]~~lfate. washed well in medium not containing the isotopic precursor, then replaced in culture for a further 6 days in 100 mg aliquots of tissue in 3 ml of culture medium containing 20% (v/v) fetal calf serum, 0.01% (w/v) bovine serum albumin, IGF-I-BP complex, of IGF-11-BP complex. The medium was changed every 24 h and stored in the presence of proteinase inhibitors at -22°C. At the end of the experiment the tissue was treated with 0.5 M NaOH for 24 h at room temperature to extract the remaining [35S]-labeledglycosaminoglycans. To determine the amount of [35S]-labeled macromolecules released into the medium on each day or present in
J Orthup Res, Vul. 10, No. 1 , 1992
the NaOH extract, aliquots of these fractions were subjected to size exclusion chromatography on Sephadex G-25 columns as described above. The logarithm of the percentage of the [35S]-labeledproteoglycans remaining in the matrix of the explants on each day was plotted as a function of time in culture, and the time taken for 50% of the [35S]labeled proteoglycans to be lost from the tissue was taken as the half-life. Analytical Gel Chromatography Proteoglycans labeled with [35S]sulfatewere applied to a Sepharose CL-2B column (0.8 cm X 108 cm) and eluted with 4 M guanidinium chloridei0.1 M sodium sulphate/O.1% (w/v) Triton X-100/0.05 M sodium acetateiacetic acid, pH 6.0, at a flow rate of 5 mUh. Fractions of 1.4 ml were collected and assayed for radioactivity. Determination of Proteoglycan Content of Cartilage Explant Cultures Sodium hydroxide extracts of cartilage explant cultures were analyzed for hexuronate as described by Bitter and Muir (2).
RESULTS Effects of IGFs and IGF-BP Complexes on the Rate of Proteoglycan Synthesis by Articular Cartilage Explants A series of experiments were carried out to investigate the effects of free and bound IGFs on proteoglycan metabolism by adult bovine cartilage in explant culture. Articular cartilage was maintained for 5 days in medium alone to obtain a basal level of proteoglycan synthesis. The tissue was then cultured for another 3 days in 20% (v/v) fetal calf serum, medium with 0.01% (w/v) bovine albumin, or the same medium supplemented with different amounts of IGF-I, IGF-11, IGF-I-BP complex, IGF11-BP complex, or NSILA-p. The rate of proteoglycan synthesis was then determined by incubation of tissue with [35S]sulfatefor 2 h followed by extraction and quantification of the [35S1-labeledmacromolecules. Figure 1A shows that IGF-1 and IGF-I1 at concentrations of 10 ng/ml and 20 ng/ml stimulated proteoglycan synthesis by explant cultures of bovine cartilage in a dose-dependent manner. The level of
17
INSULIN-LIKE GROWTH FACTORS AND CARTILAGE METABOLISM
IGF-11. Figure 1B shows that stimulation of proteoglycan synthesis by IGF-I approached a maximum by 20 ng/ml, which was slightly less than that of 20% (vh) fetal calf serum. The maximum level of stimulation produced by the combination of both IGFs was greater than that of fetal calf serum but less than the value predicted from the addition of the individual effects of IGF-I and IGF-11. In a separate experiment, articular cartilage was maintained in culture in medium containing the IGF-I-BP and IGF-11-BP complexes. Figure 2 shows the rate of proteoglycan synthesis produced by each IGF-BP complex fraction at a concentration of 1.0 and 2.0 pU NSILA/ml. The stimulation produced by LGF-I and IGF-I1 at the same dose levels was included for comparison. In the NSILA assay, 1.0 pU NSILA was equivalent to -14 ng IGF-I or 11 ng IGF-11. At 1.0 p U NSILA/ml the stimulation produced by the IGF-I-BP complex fraction was -70% of the stimulation produced by IGF-I. From 1.0 p U NSILNml to 2.0 p U NSILAI ml the stimulation by the IGF-I continued to rise, whereas the stimulation by the IGF-I-BP complex fraction did not rise any more. A different result
300 FCS ICF-I
A
c
N
200 Q &
B
IGF-II
.u
f
f 100 0
0, \
E
Q
I
P
v
I
10 20 Concentration (ng/ml)
0' 0
I
I
I
I
I
10 20 30 40 50 Concentration (ng/ml)
I
FIG. 1. Concentration dependence (A) and additive effects (B) of IGF-I and IGF-It on the stimulation of proteoglycan synthesis in articular cartilage explants. Cartilage from a single animal was maintained in explant culture for 5 days in medium alone and then in medium containing 0.01% (wiv) BSA (O),20% (viv) fetal calf serum (O),and varying concentrations of IGF-I (El) or IGF-II (A)for another 3 days before the rate of proteoglycan synthesis was determined (A). In a separate experiment, after 5 days in explant culture in medium alone, articular cartilage was incubated in medium containing 0.01% (wiv) BSA,).( 20% (viv) fetal calf serum (O), IGF-I (a), IGF-I + 50 ng/ml IGF-II ( + ) , or 50 nglml IGF-II (A)for 3 more days before the rate of proteoglycan synthesis was determined (B). Each point is a mean of three determinations, and the error bars represent the standard deviation.
io
500
ICF-I
w-II ICF-II-BP
complex
0
stimulation produced by IGF-I was approximately twofold greater than that of IGF-I1 at these doses. At a concentration of 20 ng/ml, the stimulation produced by IGF-I was similar to that produced by 20% fetal calf serum. To determine if these effects were additive, bovine articular cartilage was maintained in culture with IGF-1 at concentrations of 10, 20, and 50 ng/ ml. Tissue was also maintained in culture with the same concentrations of IGF-I but with 50 ng/ml of
1 p U NSIL4/ml
2
FIG. 2. Concentration dependence of stimulation of proteoglycan synthesis in articular cartilage explants by IGF-I-BP complex and IGF-II-BP complex. Cartilage was maintained in explant culture for 5 days in medium alone. Tissue was then incubated in medium containing 0.01% (wiv) BSA (O),20% (viv) fetal calf serum (0),IGF-I (O), IGF-II (A), IGF-I-BP complex or IGF-II-BP complex (A) for another 3 days before the rate of proteoglycan synthesis was determined. The doses of IGF-BP complexes are expressed on the basis of the NSILA (pU)of the IGF content of the complex, determined following acid gel chromatography. Each point is a mean of three determinations, and the error bars represent the standard deviation.
(m),
J Orfhop Res, Voi. 10, N O . 1 , 1992
18
G . H . TESCH ET AL.
occurred with the IGF-11-BP complex fraction, which saw a maximal stimulation at 1.0 pU NSILAI ml that was nearly twice that of an equivalent dose of IGF-11. The level of stimulation was equal to an equivalent dose of IGF-I. Like IGF-I-BP complex, IGF-11-BP complex at 2.0 pU NSILA/ml did not further stimulate proteoglycan synthesis in the explant cultures, but the unbound IGFs caused additional stimulation at this concentration. It has been previously shown by Cornell and colleagues (7) that the IGF-11-BP complex fraction used in these cartilage experiments also contained NSILA-p. The relative contributions of NSILA-p and the IGF-11-BP complex to the total NSILA of this fraction were determined using the adipocyte assay after separation of NSILA-p (the acid-stable, high Mr NSILA fraction) and low Mr IGF-I1 on acidified Sephadex G-75 (3). It was found that NSILA-p accounted for -40% of the total NSILA in this fraction, while the IGF-11-BP complex accounted for -60%. Figure 3 shows that when preparations of NSILA-p were assayed at the same NSILA dose levels as a fraction that contained both IGF-11-BP complex and NSILA-p [fraction C (7)1, no stimulation was observed with the NSILA-p, whereas fraction C showed a dose-response stimulation. These results indicate that the NSILA-p found in fractions of the IGF-11-BP complex does
350 325 120
40
50 30
1200
a 00
04
1 .o pU NSILA/ml
1.5
2.0
FIG. 3. Effect of NSILA-p on proteoglycan synthesis by articular cartilage explants. Articular cartilage was maintained in culture for 5 days in medium alone. Tissue was then incubated in medium containing either 0.01% (wiv) BSA (O),20% (vh) fetal calf serum (0),NSILA-p (V),or fraction C (A) for another 3 days before the rate of proteoglycan synthesis was determined. Each point is a mean of three determinations, and the error bars represent the standard deviation.
J Orthop Res, Vol. 10, No. I , 1992
08
12
00
04
08
12
Kav FIG. 4. Elution profiles from Sepharose CL-2B of proteoglycans synthesized by cartilage explants cultured i n the presence of IGF-I, IGF-II, IGF-I-BP complex, or IGF-II-BP complex Samples of 4 M guanidinium chloride extracts indicated by (A+ F) in Fig 2 were applied to Sepharose CL-2B eluted with 4 M guanidinium chloride Profiles are shown for proteoglycans from explant cultures maintained i n culture with 20% (viv) fetal calf serum (A); IGF-I (2 pU NSILNml) (B); IGF-I-BP complex (2 klJ NSILNml) (C); 0.01% (wiv) BSA (D); IGF-II (2 kU NSILNml) (E); or IGF-II-BP complex (2 p U NSILNml) (F). The doses of IGF-BP complexes are as expressed in Fig 2 The percentage values indicate the proportion of 35Slabel in each proteoglycan species
Characterization of the Proteoglycans Synthesized by Articular Cartilage Explants
7
0.5
w
300
The hydrodynamic size of the proteoglycans from articular cartilage explants described in Fig. 2, extracted by 4 M guanidinium chloride, was determined by gel filtration on Sepharose CL-2B (Figure 4A-F). The elution profile of the proteoglycans extracted from all the cultures showed the presence of two distinct proteoglycan species characteristic of this tissue-a large proteoglycan (Kav 0.21-0.25) and a small proteoglycan (Kav 0.70-0.75), as previously described by Hascall and co-workers (14). Previous work has shown that [35S]sulfateis incorporated into proteoglycans only by bovine articular cartilage in explant culture and that the small proteoglycan species present in the matrix of the cultures is a synthetic product and not derived from the catabolism of the large proteoglycan species (4, 14,18). It was evident that under basal conditions,
11
0.0
t It
not contribute to the stimulation of proteoglycan synthesis.
T
) 5
6oo
INSULIN-LIKE GROWTH FACTORS AND CARTILAGE METABOLISM
in medium containing 0.01% (wiv) BSA, the large proteoglycan made up -86% of newly synthesized proteoglycans and the smaller proteoglycan 14% (Fig. 4D). In tissue that had been cultured with fetal calf serum, IGFs, or IGF-BP complexes, the large proteoglycan species represented 95% of the newly synthesized proteoglycans. The rate of synthesis of each species of proteoglycan was calculated from the percentage of the radiolabel appearing in the two proteoglycan populations (as shown in Fig. 4A-F) and the rate of incorporation of [35S]sulfateinto total proteoglycans by explants of articular cartilage maintained in explant culture with IGFs, IGF-BP complexes, or fetal calf serum (Fig. 2). Table 1 shows that when IGFs, IGF-BP complexes, or fetal calf serum were included in the medium of articular cartilage explants, there was a stimulation of the large proteoglycan species and little change in the rate of synthesis of small proteoglycans.
-
Effect of IGFs and IGF Complexes on Proteoglycan Metabolism in Long-Term Articular Cartilage Explant Cultures The ability of IGFs and IGF complexes to affect proteoglycan synthesis and tissue levels of proteoTABLE 1. Relutive rutes uf synthesis of proteoglycans by curtiluge explants cultured in the presence of IGb-I, IGF-II, IGF-I-BP complex and IGF-II-BP complex [35S]Sulfateincorporation into proteog~ycans( X dpmil00 mg wet weight per 2 h) ____
Addition to medium 20% (vh) Fetal calf serum (A) IGF-I (2 FU NSILA/ml) (B) IGF-I-BP complex (2 KU NSILNml) (C) 0 01% (wlv) BSA (D) IGF-II(2 (*.UNSILAld) (E) IGF-11-BP complex (2 FU NSILA/ml) (F)
Large proteoglycans
Small proteoglycans
564.2 (3.1) 430.8 (2.4)
28.4 (1 .0) 24.6 (0.9)
279.1 (1.5) 181.5 (1.0) 395.0 (2.2)
17.2 (0.6) 29.0 (1.0) 19.9 (0.7)
371.6 (2.1)
22.5 (0.8)
Data from the elution profiles of proteoglycans shown in Fig. 4A-F and that given in Fig. 2 on the rate of synthesis of proteoglycans were used to calculatc the rate of incorporation of [35S]sulfate into the two species of proteoglycans present in explant cultures of cartilage maintained in medium containing different IGF fractions. Values in parentheses represent the increase in synthesis expressed relative to the rate measured in cultures incubated in medium containing 0.01% (wh) bovine serum albumin.
19
glycans over a 5-day period in explant cultures of cartilage was investigated. Fresh articular cartilage was maintained for =3days in medium containing 20% (viv) fetal calf serum, 0.01% (wiv) bovine serum albumin or in medium containing 0.01% (wiv) bovine albumin and IGF-1 (20 ngiml = 1.4 pU NSILAlml) or IGF-I1 (SO ngiml = 4.5 p,U NSILAi ml) or 1 pU NSTLAiml of either IGF-I-BP complex or IGF-11-BP complex. The data shown in Table 2 cover two experiments using separate batches of tissue and shows that the rate of proteoglycan synthesis by cartilage maintained in culture in medium containing fetal calf serum, IGF-I, IGF-11, and each of the IGF-BP complexes was equal to or greater than the rate observed on day 0. Tissue cultured throughout with medium containing 0.01% (wiv) bovine serum albumin showed a decline in the rate of synthesis. Cultures maintained in medium containing fetal calf serum, IGF-I, and IGF-I1 showed no or little decline in the proteoglycan content of the tissue (measured as hexuronate content) during the 5-day culture period. Cartilage maintained in medium containing 0.01% (wiv) bovine serum albumin showed a decline in proteoglycan levels. When tissue was maintained in medium containing IGF-IIBP complex, a slight fall in tissue proteoglycan level was evident, similar to that observed in the same tissue maintained in medium containing IGF-I or fetal calf Serum. However, when the same tissue was maintained in medium containing IGF-I-BP complex, a greater fall in proteoglycan tissue levels was evident, similar to that obtained for tissue cultured in medium containing 0.01% (wh) bovine serum albumin. The effect of IGF-I-BP and IGF-11-BP complexes on the rate of proteoglycan catabolism in articular cartilage explants was also determined. Articular cartilage was maintained in culture for 5 days and then incubated with [35S]sulfatefor 6 h, washed to remove unincorporated label, and placed in culture in medium supplemented with 0.01% bovine serum albumin and either IGF-I-BP complex ( 1 pU NSILAlml) or IGF-11-BP complex (1 pU NSILA/ ml) or in medium containing 0.01% (wiv) bovine serum albumin or 20% (viv) fetal calf serum. Figure 5 shows the rate of loss of [35S]-labeledproteoglycans from the cartilage explants. The half-lives of proteoglycans were determined and found to be 15 and 18 days for tissue maintained with IGF-I-BP complex or IGF-11-BP complex, respectively. For cultures maintained in medium containing 0.01%
3 Orthop Res, Vol. 10, N o . 1 , 1992
G . H . TESCH ET AL.
20
TABLE 2. EHect of free and bound lGFs on proteoglycan synthesis and hexuronute content of cartilage explants Days in culture
Addition to medium
Experiment 1 0 5 5 5 5 Experiment 2 0 5 5 5 5
0.01% (wlv) BSA 20% (viv) Fetal calf serum IGF-I (1.4 pU NSILAlml) IGF-I1 (4.5 pU NSILAlmI)
[35S]Sulfateincorporation into proteoglycans (X lo-’; dpm/ 100 mg wet weight per 2 h)
Hexuronate content (pg hexwonate/ mg wet weight)
107.7 & 6.4 (1.0) 48.3 -+ 8.6 (0.5) 304.7 2 35.7 (2.8) 160.9 2 21.1 (1.5) 132.4 ? 0.8 (1.2)
5.4 -+ 0.1 (1.0) 4.7 k 0.2 (0.9) 5.5 & 0.5 (1.0) 5.6 f 0.1 (1.0) 5.5 & 0.4 (1.0)
15.8 -+ 13.2 48.6 2 20.2 -t 32.9 ?
-
1.4 (1.0) (0.8) 3.7 (3.1) 0.6 (1.3) 0.2 (2.1)
* 1.4
0.1% (wlv) BSA 20% ( d v ) Fetal calf serum IGF-I-BP complex (1 pU NSILAlml) IGF-II-BP complex (1 pU NSILAlml)
6.9 2 0.3 (1.0) 4.9 2 0.2 (0.7) 6.4 2 0.1 (0.9) 5.3 ? 0.2 (0.8) 6.1 ? 0.1 (0.9)
Cartilage was maintained in explant culture for 5 days in medium containing 0.01% (wiv) BSA supplemented with either IGFs or IGF-BP complexes. On day 0 or day 5 of culture the rate of synthesis and hexuronate content of the explant cultures was determined. Each point is the mean of three determinations + standard deviation. The value in parenthesis represents the synthesis or hexuronate content expressed relative to the values obtained for articular cartilage measured before it was placed in culture.
(w/v) bovine serum albumin or 20% (v/v) fetal calf serum, the half-lives were found to be 10 and 15 days, respectively. This suggests that IGF-I-BP and IGF-II-BP complexes suppress the rate of proteoglycan catabolism in articular cartilage explant cultures.
bQ
I
I
I
I
6
7
8
I
1
9 10 Days in culture
I
I
11
12
I
FIG. 5. Effect of IGF-I-BP complex and IGF-II-BP complex on the rate of loss of [35S]-labeledproteoglycans from articular cartilage explants. Cartilage from a single animal was maintained in culture for 5 days in medium containing 20% (v/v) fetal calf serum, incubated for 6 h with [35S]sulfate,then replaced in culture in medium containing IGF-I-BP complex;.( 1 pU NSILNml), IGF-II-BP complex (A;1 pU NSILNml), 20% (vlv) fetal calf serum (0),or 0.01% (wlv) BSA (0). The logarithm of the percentage of the [35S]-labeled proteoglycans remaining in the matrix on each day was plotted against time in culture. Each point is a mean of two determinations.
J Orthop Res, Vol. 10, N O . I , 1992
DISCUSSION
In this study it was shown that free IGF-I is more potent in stimulating proteoglycan synthesis than free IGF-IT. regardless of whether doses are compared on a weight basis or on a NSILA basis. This is comparable to previous findings that show that recombinant human IGF-I is a more potent stimulator of proteoglycan synthesis by articular cartilage explants than purified rat IGF-II(l7). However, the magnitude of stimulation by TGF-I over IGF-I1 reported in the current experiments is much lower than that reported by Luyten and associates (17) and may reflect either the difference in the age of the animals used or the use of human IGF-I1 in this study. When IGF-I1 was added to IGF-I, the stimulation was further increased but was less than additive. This suggests that stimulation occurs, at least in part, through different mechanisms, e.g., both type 1 IGF and type I1 IGF receptors. Luyten and collaborators (17) suggested that IGF-I and IGF-I1 stimulate proteoglycan synthesis in articular cartilage from young animals through type I IGF receptors only. A number of groups (16,24,25) have shown the existence of both type I and type I1 IGF receptors on the chondrocytes of 4-6-week-old rats and 18-day-old rabbits. The less than additive effect of IGF-I and IGF-I1 seen in our study could be explained by competitive binding of IGF-I and IGF-I1 for both the type 1 and type I1 receptors. The affinity of IGF-I and IGF-I1 for each other’s receptors has been well documented (see review, 23). IGF-I-
INSULIN-LIKE GROWTH FACTORS AND CARTILAGE METABOLISM
BP complex and IGF-11-BP complex were both shown to stimulate proteoglycan synthesis, but to different degrees. Unlike the IGF-I-BP complex fraction, the IGF-IT-BP complex fraction showed a greater stimulation of proteoglycan synthesis compared with free IGF-I1 at low concentrations (1.0 pU NSILAlml). At higher concentrations (i.e., 2.0 pU NSILAlml), the stimulation by free and bound IGF-11 was similar. The data presented here shows that when IGF-IBP and TGF-11-BP complexes were included in the medium of explant cultures of adult articular cartilage, there was a decrease in the rate of catabolism of proteoglycans compared with control cultures maintained in medium containing 0.01% (wiv) bovine serum albumin. Both free and complexed IGFs were shown to maintain proteoglycan synthesis by explant cultures to varying degrees. Since the levels of proteoglycan present in the extracellular matrix of the cartilage explants must reflect the sum of the rates of synthesis and catabolism of these macromolecules by the tissue, the inability of IGF-I-BP complex to maintain proteoglycan levels in cartilage explants may be due in part to the inability of this growth factor complex to stimulate proteoglycan synthesis. Since our work shows that the maximal stimulation of proteoglycan synthesis is achieved with bound IGF-I and IGF-I1 at a concentration of -I pU NSILAlml , higher concentrations of bound IGF-I 118 ng = 12 pU NSILAlml, as determined in the rat adipocyte assay; (1 I)] and bound IGF-II(600 ng = 50 pU NSILNml), such as those found in plasma, are unlikely to result in further stimulation of proteoglycan synthesis by articular cartilage. Throughout this work a variation was observed in the magnitude of the response of articular cartilage from different animals when exposed to free or bound IGFs, but nevertheless the growth factors always stimulated proteoglycan synthesis. The variation in stimulation of proteoglycan synthesis observed between tissue from different animals expressed relative to values obtained for the same tissue maintained in culture in medium containing 0.01% (wh) bovine serum albumin were: for cultures maintained in medium containing IGF-I (1 $J NSJLNml), 2.6 t 0.5 (mean t SD) (tissue from seven animals); for IGF-II(1 pU NSILA/ml), 2.1 2 0.3 (tissue from five animals); for IGF-I-BP complex (1 pU NSILA/ml), 1.4 e 0.1 (tissue from six animals); for IGF-11-BP complex (1 pU NSII.A/ml),
21
2.4 2 0.6 (tissue from five animals); and for 20% (vlv) fetal calf serum, 3.8 t 0.8 (tissue from eight animals). There is increasing evidence to suggest that BPs can modify the actions of IGFs on various tissue types in a stimulatory and inhibitory manner (1). For instance, the insulin-like activity of circulating IGFs is inhibited by the formation of the 150,000 dalton complex, which is incapable of binding to the insulin receptor (10). Furthermore, it has been pointed out that IGF-BP complexes may regulate their release of IGFs, thereby controlling the target tissue receptor occupancy (3). IGFs have been reported to be present in synovial fluid from humans (26) and steers (C. J. Handley, unpublished information). In the latter case the IGF activity was associated with a high-molecular-weight complex that was dissociated under acid conditions. It is not known whether the BP complexes of both IGF-I and IGF-I1 can penetrate the extracellular matrix of cartilage and interact with receptors on chondrocytes. Work by Clernmons and associates (6) has shown that when samples containing IGF-BP complexes are incubated for 24 h with highly negatively charged macromolecules, such as heparin, or with proteinases, the BPs are stripped from the complex, leaving the free IGF. Since the extracellular matrix of cartilage contains highly negatively charged proteoglycan and proteinases, it is possible that IGFs and their BPs may be separated before the IGFs bind to their receptors. The activity of the IGF-BP complexes observed in cartilage may therefore be dependent on the ability of components of the extracellular matrix to separate out the BPs. If the IGF-BP complexes were to reach the cells, then the difference between the stimulation of proteoglycan synthesis observed for the IGF-I-BP and the IGF11-BP complexes might be explained by a differing ability to bind to their respective receptors on the chondrocyte, as previously noted (23). IGF-I and IGF-I1 in both free and bound forms and fetal calf serum were all shown to stimulate preferentially the synthesis by cartilage explants of the large proteoglycans capable of forming aggregates with hyaluronate and link protein without affecting the synthesis of the small proteoglycans. This finding is in agreement with the observations of McQuillan and colleagues (19), which suggested that fetal calf serum and IGF-I stimulate the synthesis of the large proteoglycans of articular cartilage. These proteoglycans are characteristic of car-
J Orthop Res, Vol. 10, No. 1, 1992
22
G. H . TESCIY ET AL.
tilage and are responsible for the biochemical properties of this tissue. This work indicates that the lGFs may be responsible for the maintenance of the phenotype of cartilage. This finding is important not only in the development of the tissue but also in later life, for the maintenance of important components of the extracellular matrix of the tissue. If there is a defect in the expression of the IGFs or their receptors, it may lead to degeneration of the tissue. Acknowledgment: The authors thank Mrs. M. T. Scott and Ms. M. Bistrin for expert technical assistance and Castricum Bros. Pty. Ltd. (Ferntree Gully Abattoirs) for the donation of bovine hocks. This work was supported by the National Health and Medical Research Council of Australia and the Arthritis Foundation of Australia.
REFERENCES 1. Baxter RC, Martin JL: Binding proteins for the insulin-like growth factors: Structure, regulation and functlon. Prog Growth F a d o r Res 1:49-68, 1989 2. Bitter T, Muir HI: A modified uronic acid carbazole reaction. Anallt Biochem 4:330-334, 1962 3. Blum WF, Jenne EW, Reppin F, Kietzman K . Ranke MB, Bierich JK: Insulin-like growth factor I protein complex is a better mitogen than free IGF-I. Endocrinology 125:76&772, 1989 4. Campbell MA, Handley CJ: The effect of retinoic acid on the turnover of proteoglycans in cultured articular cartilage. Arch Biorhem Biophys 258:143-155, 1987 5. Campbell MA, Handley CJ, Hascall VC. Campbell KA, Lowther DA: Turnover of proteoglycans in cultures of bovine articular cartilage. Arch Biochem Biophys 234:275-281, 1984 6. Clemmons DR, Underwood LE, Chatelain PG, van Wyk JJ: Liberation of immunoreactive somatomedin-C from its binding proteins by proteolytic enzymes and heparin. J Chi1 Endocrinnl Metab 56:384-389, 1983 7. Cornell HJ, Enberg G , Herington AC: Preferential association of the insulin-like growth factors I and I1 with metabolically inactive and active carrier bound complexes in serum. Biochern J 241:745-750, 1987 8. Daughaday WH, Kipnis DM: The growth promoting and anti-insulin actions of stromatotrophin Rec Prog Horm Res 22:49-93, 1966 9. Franklin RC. Renme GC, Burger HG, Cameron DP: A bioassay for NSILA-S In individual serum samples and its relationship to somatotropin. J Clin Endocrinol M e t a b 43:1164-1169, 1976 10. Guler HP, Zapf J, Froesch ER: Short-term metabolic effects of recombinant human insulin-like growth factor-I in healthy adults. N e n Engl J Med 317:137-140, 1987 11. Handley CJ, Lowther DA: Extracellular matnx metabolism by chondrocytes. 111. Modulation of proteoglycan synthesis by extracellular levels of proteoglycans in cartilage cells in culture. Biochim Biophyr Actn 500.132-139, 1977 12. Handley CJ, Campbell MA: Catabolism and turnover of proteoglycans. Meth Enzymol 144:412419, 1987
I Orthop Res, Voi. IO, No. 1 , 1992
13. Handley CJ, Lowther IDA, McQuillan DJ: 'The structure and synthesis of proteoglycans of articular cartilage. Cell B i d Int RPU9~753-782, 1985 14. Hascall VC, Handley CJ, McQuillan DJ, Hascall GK, Robinson HC. Lowther DA: The effect of serum on biosynthesis of proteoglycans by bovine articular cartilage in culture. Arch Biochem Biophys 224:20&223, 1983 15. Herington AC, Cornell HJ, Kuffer AD: Recent advances in the biochemistry and physiology of the insulin-like growth factorisomatomedin family. Int J Biochem IS: 1201-1210, 1983 16. Jansen J. van Buul-Offers SC, Hoogerbrugge CM, de Porter TL, Corvol MT, van den Brande JL: Characterization of specific insulin-like growth factor-I and -11 receptors on cultured rabbit articular chondrocyte membranes. J Endocrinol 1 20:245-249, 1989 17. Luyten FP, Hascall V C , Nissley SP, Morales TI, Reddi AH: Insulin-like growth factors maintain steady-state metabolism of proteoglycans in bovine articular cartilage explants. Arch Biochem Biophys 267:416-425, 1988 18. McQuillan DJ, Handley CJ, Robinson HC, Ng K, Tzaicos C, Brooks PR, Lowther DA: Effects of cycloheximide on chondroitin sulphate synthesis. Biochem J 2241977-988; 1984 19. McQuillan DJ, Handley CJ, Campbell MA, Bolis S, Milway VE, Herington, AC: The stimulation of proteoglycan biosynthesis by serum and insulin-like growth factor-I in cultured bovine articular cartilage. Biochem J 240:423-430, 1986 20. Muir H: The chemistry of the ground substance ofjoint cartilage. In: Sokoloff L , ed. The Joinfs and Synovinl Fluid, vol. 2, New York: Academic Press, 1980, pp 27-94. 21. Oegema TR. Hascall VC. Eisenstein K: Characterization of bovine aorta proteoglycan extracted with guanidine hydrochloride in the presence of protease inhibitors. J Riol Chem 259: 172G1726, 1979 22. Poffenbarger PL: The purification and partial characterization of an insulin-like protein from human serum. J Clin Invest 56:1455-1463, 1975 23. Rechler MM, Nissley SP: The nature and regulation of the receptors for insulin-like growth factors. Ann Rev Physiol 47:425-442, 1985 24, Schalch DS, Sessions CM, Fdrley AC, Masakawa A, Emler CA. Dills DG: Interaction of insulin-like growth factor I/ somatomedin C with cultured rat chondrocytes: Receptor binding and internalization. Endocrinology 118: 1590-1597, 1986 25. Sessions CM. Emler CA, Schalch DS: Interaction of insulinlike growth factor-I1 with rat chondrocytes: Receptor binding, internalization and degradation. EndocrinnloRv 120:2108-2116, 1987 26. Schalkwijk J , Joosten LAB, van den Berg WB, van Wyk JJ, van de Putte LBA: Insulin-like growth factor stimulation of chondrocyte proteoglycan synthesis by human synovial fluid. Arth Rheum 32:66-71, 1989 27. Zapf J, Schoenle E, Froesch ER: Insulin-like growth factors I and 11: Some biological actions and receptor binding characteristics of two purified constituents of nonsuppressible insulin-like activity of human serum. Eur J Biochem 87:285296, 1978 28. Zapf J, Walter H, Froesch ER: Kadioimmunological determination of insulin-like growth factors 1 and I1 in normal subjects and in patients with growth disorders and extrapancreatic tumor hypoglycemia. J Clin Invest 68: 1321-1330, 1981
.
Effects of Free and Bound Insulin-Like Growth Factors on Proteoglycan Metabolism in Articular Cartilage Explants Gregory H. Tesch, "Christopher J. Handley, Hugh J. Cornell, and ?Adrian C. Herington Department of Applied Chemistry, Royal Melbourne Institute of Technology, Melbourne, Victoria; *Department of Biochemistty, Monash University, Clayton, Victoria; and iPrince Henry's Institute of Medical Research, Melbourne, Victoria, Australia
Summary: This article describes the effects of bound forms of insulin-like growth factors (IGFs) on proteoglycan metabolism by bovine articular cartilage in explant culture. When these growth factors were added to articular cartilage explants complexed with their native serum binding proteins (BPs), both IGF-1-BP complex and IGF-11-BP complex stimulated proteoglycan synthesis to different degrees over a 3-day pcriod. When added to the medium of cultures of articular cartilage over 5 days. IGF-11-BP complex induced high rates of synthesis and low rates of catabolism of proteoglycans, giving rise to tissue levels of proteoglycan similar to those observed in fresh tissue. When articular cartilage was maintained in culture with the same concentration of ICF-I-BP complex, tissue levels of proteoglycans fell over the culture period because of lower rates of proteoglycan synthesis. Analysis of the proteoglycans synthesized by articular cartilage in the presence of free or bound IGF-I or TGF-TI showed that these growth factors stimulated the rate of synthesis of the large proteoglycan species present in cartilage but did not affect the synthesis of the small proteoglycans. Key Words: Insulin-like growth factorArticular cartilage-Proteoglycan-Synthesis-Catabolism.
Articular cartilage has the unique mechanical property of being able to withstand compressive loads. This ability is attributed to the presence of highly hydrated proteoglycans contained within the collagen network of the extracellular matrix (for review see 13,20). These proteoglycans are continuously being turned over within the matrix, and so the regulation of proteoglycan metabolism is essential for the maintenance and function of articular cartilage. Studies have suggested that insulin-like growth
factors (IGFs) present in serum are responsible for the maintenance of proteoglycan synthesis by explant cultures of cartilage (17,19,27). It is therefore apparent that IGFs may play a major role in the regulation of growth and maintenance of articular cartilage and in the repair of damaged tissue. In normal adult human serum, IGF-I1 is present in higher quantities (three tq five times) than IGF-I (28). In addition, both IGF-I and IGF-I1 are found in serum almost entirely bound to binding proteins (BPs) to form a circulating complex with a molecular weight of -150,000 daltons (8). These BPs are believed to regulate the action of the IGFs toward their target tissues (1). Where no measurable free IGF is present, the ability of whole serum to stimulate proteoglycan synthesis suggests that the
Received November 1 I , 1990; accepted June 20, 1991. Address correspondenceand reprint requests to C. J . Handley at Department of Biochemistry, Monash University, Clayton, Victoria 3168, Australia.
14
INSULIN-LIKE GROWTH FACTORS AND CARTILAGE METABOLiSM
15
bound forms of IGFs are active in proteoglycan stimulation or that the free forms are released in situ (14). This article describes the effects of the bound forms of IGF-I and IGF-I1 on proteoglycan metabolism in bovine articular cartilage in explant culture.
contaminant with NSILA in this fraction (7). NSILA-p was not present in the IGF-I-BP complex fraction. The IGF species and NSILA-p are the only substances in serum that possess NSILA (15).
MATERIALS AND METHODS
p, and the IGF-RP complexes was determined in a
Determination of IGF Activity The biological activity of IGF-I, IGF-11, NSILA-
Materials Recombinant human IGF-I was generously supplied by Dr. Anna Skottner, Kabi Vitrum, Stockholm, Sweden. Purified human IGF-I1 was kindly donated by Dr. R. Baxter, Royal Prince Alfred Hospital, Camperdown, N.S.W., Australia. Human plasma was obtained from healthy adult blood bank donors. All other materials were obtained as described previously (7,14,18). Purification of IGF Complexes Purification of the IGF-BP complexes was carried out as previously described by Cornell and associates (7). Briefly, IGF-I-BP complexes and IGF11-BP complexes were purified from human plasma by ion-exchange chromatography on DEAESephadex A-50 using a stepwise NaCl gradient to elute each complex preferentially. The IGF-I-BP complex and IGF-11-BP complex fractions were further purified by affinity chromatography on Con-A Sepharose. This procedure produced an IGF-I-BP complex fraction with a 74-101-fold purification and an IGF-11-BP complex fraction with 46-54-fold purification relative to plasma-specific activities of IGF-I and IGF-11, respectively, based on the nonsuppressible insulin-like activity (NSILA) bioassay. Analytical chromatography on Sephacryl S-200 was used to confirm the presence of high Mr IGF-BP complexes in each purified fraction. Both the IGFI-BP and the IGF-11-BP complex fractions contained complexes with Mr 50,000-160,000. No free IGF was detected in either of these fractions, indicating that all the IGFs present were in a bound form. The IGF-11-BP complex contained no detectable free or bound immunoreactive IGF-I, whereas the partially purified IGF-I-BP complex contained 14% (w/w) IGF-11-BP complex. The IGF-11-BP complex contained NSILA-p. a protein with nonsuppressible insulin-like activity that precipitates with acid-ethanol solution (22), which was the only
-
standard NSILA bioassay by monitoring the incorporation of [U-'4Cjglucose into lipid by isolated rat adipocytes in the presence of excess anti-insulin serum (9). The activity of each species assayed was subsequently expressed in terms of units of insulinlike activity based on an insulin standard. In the case of the IGF-BP complexes, the NSILA activity was determined by assaying the amount of low Mr-free IGF liberated after each of the IGF complex fractions had been subjected to gel filtration on a Sephadex G-75 column (1.2 cm X 54 cm) equilibrated and elutcd with 1% (viv) formic acid (19).
Cartilage Cultures Bovine articular cartilage explants were obtained from the metacarpalphalangeal joints of 1-2-yearold steers and prepared as described previously (14). Samples of sliced cartilage (10G150 mgivial) were maintained in culture for 5 days in a modification of Dulbecco's Modified Eagle Medium (11) or the same medium supplemented with fetal calf serum, bovine serum albumin, IGFs, IGF-BP complexes, or NSILA-p. Unless stated, NSILA-p, IGFs, and IGF-BP complexes were added to medium containing 0.01% (wiv) bovine serum albumin to prevent nonspecific absorption. The batch of bovine serum albumin used was pretested to ensure that it had no stimulatory activity on proteoglycan synthesis by articular cartilage explants. The medium containing NSILA-p and growth factors was sterilized by filtration through microporous membranes. In experiments determining the rate of loss of proteoglycans from cartilage explants, cartilage from the same joint (-3 g) was maintained in culture for 3 days in 30 ml of medium containing 20% (vh) fetal calf serum before use. We have previously noted variation in the rates of proteoglycan metabolism by different samples of bovine articular cartilage when placed in explant culture, primarily due to animal variation (5,9). All experiments were therefore carried out using tissue
J Orthop Rrs, Vol. 10, N o . I , 1992
G . H . TESCH ET AL.
16
obtained from a single metacarpalphalangeal joint; further experiments were carried using tissue from different animals to confirm any one observation. Measurement of Proteoglycan Synthesis Cartilage cultures were preincubated for 1 h in fresh medium and then incubated for 2 h in 2 ml of medium alone, containing 20 FCi of [35S]sulfate/ml (19). The labeled proteoglycans were extracted with 4 M guanidinium chloride/0.5 M sodium acetateacetic acid, pH 6.0, in the presence of proteinase inhibitors for 2 days at 4°C (21). The tissue was then extracted with 2 ml of 0.5 M NaOH for 16 h at room temperature. Some of the cartilage explants were also extracted directly with 2 ml of 0.5 M NaOH. The rate of [35S]sulfateincorporation into total proteoglycans (dpmi100 mg wet weight of cartilage per 2 hj was determined by size exclusion chromatography on Sephadex G-25 (PD-10) columns (14). The rate of [35S]~~lfate incorporation (dpm/lOO mg wet weight per 2 hj into individual species of proteoglycan was calculated by multiplying the rate of incorporation of [ 3 5 ~ ~ s u l f ainto t e total proteoglycans (dpm/100 mg wet weight per 2 h) by the percentage of the radiolabel appearing in the the particular species of proteoglycan. The latter was determined by size exclusion chromatography on Sepharose CL-2B and is described below. Measurement of Proteoglycan Catabolism Proteoglycan turnover in cartilage explant cultures was determined as described by Handley and Campbell (12). After 5 days in culture in medium containing 20'30 (viv) fetal calf serum, articular cartilage (-3 g) from a single joint was incubated for 6 h with 10 ml of medium containing fetal calf serum The tissue was then and 300 pCi of [35S]~~lfate. washed well in medium not containing the isotopic precursor, then replaced in culture for a further 6 days in 100 mg aliquots of tissue in 3 ml of culture medium containing 20% (v/v) fetal calf serum, 0.01% (w/v) bovine serum albumin, IGF-I-BP complex, of IGF-11-BP complex. The medium was changed every 24 h and stored in the presence of proteinase inhibitors at -22°C. At the end of the experiment the tissue was treated with 0.5 M NaOH for 24 h at room temperature to extract the remaining [35S]-labeledglycosaminoglycans. To determine the amount of [35S]-labeled macromolecules released into the medium on each day or present in
J Orthup Res, Vul. 10, No. 1 , 1992
the NaOH extract, aliquots of these fractions were subjected to size exclusion chromatography on Sephadex G-25 columns as described above. The logarithm of the percentage of the [35S]-labeledproteoglycans remaining in the matrix of the explants on each day was plotted as a function of time in culture, and the time taken for 50% of the [35S]labeled proteoglycans to be lost from the tissue was taken as the half-life. Analytical Gel Chromatography Proteoglycans labeled with [35S]sulfatewere applied to a Sepharose CL-2B column (0.8 cm X 108 cm) and eluted with 4 M guanidinium chloridei0.1 M sodium sulphate/O.1% (w/v) Triton X-100/0.05 M sodium acetateiacetic acid, pH 6.0, at a flow rate of 5 mUh. Fractions of 1.4 ml were collected and assayed for radioactivity. Determination of Proteoglycan Content of Cartilage Explant Cultures Sodium hydroxide extracts of cartilage explant cultures were analyzed for hexuronate as described by Bitter and Muir (2).
RESULTS Effects of IGFs and IGF-BP Complexes on the Rate of Proteoglycan Synthesis by Articular Cartilage Explants A series of experiments were carried out to investigate the effects of free and bound IGFs on proteoglycan metabolism by adult bovine cartilage in explant culture. Articular cartilage was maintained for 5 days in medium alone to obtain a basal level of proteoglycan synthesis. The tissue was then cultured for another 3 days in 20% (v/v) fetal calf serum, medium with 0.01% (w/v) bovine albumin, or the same medium supplemented with different amounts of IGF-I, IGF-11, IGF-I-BP complex, IGF11-BP complex, or NSILA-p. The rate of proteoglycan synthesis was then determined by incubation of tissue with [35S]sulfatefor 2 h followed by extraction and quantification of the [35S1-labeledmacromolecules. Figure 1A shows that IGF-1 and IGF-I1 at concentrations of 10 ng/ml and 20 ng/ml stimulated proteoglycan synthesis by explant cultures of bovine cartilage in a dose-dependent manner. The level of
17
INSULIN-LIKE GROWTH FACTORS AND CARTILAGE METABOLISM
IGF-11. Figure 1B shows that stimulation of proteoglycan synthesis by IGF-I approached a maximum by 20 ng/ml, which was slightly less than that of 20% (vh) fetal calf serum. The maximum level of stimulation produced by the combination of both IGFs was greater than that of fetal calf serum but less than the value predicted from the addition of the individual effects of IGF-I and IGF-11. In a separate experiment, articular cartilage was maintained in culture in medium containing the IGF-I-BP and IGF-11-BP complexes. Figure 2 shows the rate of proteoglycan synthesis produced by each IGF-BP complex fraction at a concentration of 1.0 and 2.0 pU NSILA/ml. The stimulation produced by LGF-I and IGF-I1 at the same dose levels was included for comparison. In the NSILA assay, 1.0 pU NSILA was equivalent to -14 ng IGF-I or 11 ng IGF-11. At 1.0 p U NSILA/ml the stimulation produced by the IGF-I-BP complex fraction was -70% of the stimulation produced by IGF-I. From 1.0 p U NSILNml to 2.0 p U NSILAI ml the stimulation by the IGF-I continued to rise, whereas the stimulation by the IGF-I-BP complex fraction did not rise any more. A different result
300 FCS ICF-I
A
c
N
200 Q &
B
IGF-II
.u
f
f 100 0
0, \
E
Q
I
P
v
I
10 20 Concentration (ng/ml)
0' 0
I
I
I
I
I
10 20 30 40 50 Concentration (ng/ml)
I
FIG. 1. Concentration dependence (A) and additive effects (B) of IGF-I and IGF-It on the stimulation of proteoglycan synthesis in articular cartilage explants. Cartilage from a single animal was maintained in explant culture for 5 days in medium alone and then in medium containing 0.01% (wiv) BSA (O),20% (viv) fetal calf serum (O),and varying concentrations of IGF-I (El) or IGF-II (A)for another 3 days before the rate of proteoglycan synthesis was determined (A). In a separate experiment, after 5 days in explant culture in medium alone, articular cartilage was incubated in medium containing 0.01% (wiv) BSA,).( 20% (viv) fetal calf serum (O), IGF-I (a), IGF-I + 50 ng/ml IGF-II ( + ) , or 50 nglml IGF-II (A)for 3 more days before the rate of proteoglycan synthesis was determined (B). Each point is a mean of three determinations, and the error bars represent the standard deviation.
io
500
ICF-I
w-II ICF-II-BP
complex
0
stimulation produced by IGF-I was approximately twofold greater than that of IGF-I1 at these doses. At a concentration of 20 ng/ml, the stimulation produced by IGF-I was similar to that produced by 20% fetal calf serum. To determine if these effects were additive, bovine articular cartilage was maintained in culture with IGF-1 at concentrations of 10, 20, and 50 ng/ ml. Tissue was also maintained in culture with the same concentrations of IGF-I but with 50 ng/ml of
1 p U NSIL4/ml
2
FIG. 2. Concentration dependence of stimulation of proteoglycan synthesis in articular cartilage explants by IGF-I-BP complex and IGF-II-BP complex. Cartilage was maintained in explant culture for 5 days in medium alone. Tissue was then incubated in medium containing 0.01% (wiv) BSA (O),20% (viv) fetal calf serum (0),IGF-I (O), IGF-II (A), IGF-I-BP complex or IGF-II-BP complex (A) for another 3 days before the rate of proteoglycan synthesis was determined. The doses of IGF-BP complexes are expressed on the basis of the NSILA (pU)of the IGF content of the complex, determined following acid gel chromatography. Each point is a mean of three determinations, and the error bars represent the standard deviation.
(m),
J Orfhop Res, Voi. 10, N O . 1 , 1992
18
G . H . TESCH ET AL.
occurred with the IGF-11-BP complex fraction, which saw a maximal stimulation at 1.0 pU NSILAI ml that was nearly twice that of an equivalent dose of IGF-11. The level of stimulation was equal to an equivalent dose of IGF-I. Like IGF-I-BP complex, IGF-11-BP complex at 2.0 pU NSILA/ml did not further stimulate proteoglycan synthesis in the explant cultures, but the unbound IGFs caused additional stimulation at this concentration. It has been previously shown by Cornell and colleagues (7) that the IGF-11-BP complex fraction used in these cartilage experiments also contained NSILA-p. The relative contributions of NSILA-p and the IGF-11-BP complex to the total NSILA of this fraction were determined using the adipocyte assay after separation of NSILA-p (the acid-stable, high Mr NSILA fraction) and low Mr IGF-I1 on acidified Sephadex G-75 (3). It was found that NSILA-p accounted for -40% of the total NSILA in this fraction, while the IGF-11-BP complex accounted for -60%. Figure 3 shows that when preparations of NSILA-p were assayed at the same NSILA dose levels as a fraction that contained both IGF-11-BP complex and NSILA-p [fraction C (7)1, no stimulation was observed with the NSILA-p, whereas fraction C showed a dose-response stimulation. These results indicate that the NSILA-p found in fractions of the IGF-11-BP complex does
350 325 120
40
50 30
1200
a 00
04
1 .o pU NSILA/ml
1.5
2.0
FIG. 3. Effect of NSILA-p on proteoglycan synthesis by articular cartilage explants. Articular cartilage was maintained in culture for 5 days in medium alone. Tissue was then incubated in medium containing either 0.01% (wiv) BSA (O),20% (vh) fetal calf serum (0),NSILA-p (V),or fraction C (A) for another 3 days before the rate of proteoglycan synthesis was determined. Each point is a mean of three determinations, and the error bars represent the standard deviation.
J Orthop Res, Vol. 10, No. I , 1992
08
12
00
04
08
12
Kav FIG. 4. Elution profiles from Sepharose CL-2B of proteoglycans synthesized by cartilage explants cultured i n the presence of IGF-I, IGF-II, IGF-I-BP complex, or IGF-II-BP complex Samples of 4 M guanidinium chloride extracts indicated by (A+ F) in Fig 2 were applied to Sepharose CL-2B eluted with 4 M guanidinium chloride Profiles are shown for proteoglycans from explant cultures maintained i n culture with 20% (viv) fetal calf serum (A); IGF-I (2 pU NSILNml) (B); IGF-I-BP complex (2 klJ NSILNml) (C); 0.01% (wiv) BSA (D); IGF-II (2 kU NSILNml) (E); or IGF-II-BP complex (2 p U NSILNml) (F). The doses of IGF-BP complexes are as expressed in Fig 2 The percentage values indicate the proportion of 35Slabel in each proteoglycan species
Characterization of the Proteoglycans Synthesized by Articular Cartilage Explants
7
0.5
w
300
The hydrodynamic size of the proteoglycans from articular cartilage explants described in Fig. 2, extracted by 4 M guanidinium chloride, was determined by gel filtration on Sepharose CL-2B (Figure 4A-F). The elution profile of the proteoglycans extracted from all the cultures showed the presence of two distinct proteoglycan species characteristic of this tissue-a large proteoglycan (Kav 0.21-0.25) and a small proteoglycan (Kav 0.70-0.75), as previously described by Hascall and co-workers (14). Previous work has shown that [35S]sulfateis incorporated into proteoglycans only by bovine articular cartilage in explant culture and that the small proteoglycan species present in the matrix of the cultures is a synthetic product and not derived from the catabolism of the large proteoglycan species (4, 14,18). It was evident that under basal conditions,
11
0.0
t It
not contribute to the stimulation of proteoglycan synthesis.
T
) 5
6oo
INSULIN-LIKE GROWTH FACTORS AND CARTILAGE METABOLISM
in medium containing 0.01% (wiv) BSA, the large proteoglycan made up -86% of newly synthesized proteoglycans and the smaller proteoglycan 14% (Fig. 4D). In tissue that had been cultured with fetal calf serum, IGFs, or IGF-BP complexes, the large proteoglycan species represented 95% of the newly synthesized proteoglycans. The rate of synthesis of each species of proteoglycan was calculated from the percentage of the radiolabel appearing in the two proteoglycan populations (as shown in Fig. 4A-F) and the rate of incorporation of [35S]sulfateinto total proteoglycans by explants of articular cartilage maintained in explant culture with IGFs, IGF-BP complexes, or fetal calf serum (Fig. 2). Table 1 shows that when IGFs, IGF-BP complexes, or fetal calf serum were included in the medium of articular cartilage explants, there was a stimulation of the large proteoglycan species and little change in the rate of synthesis of small proteoglycans.
-
Effect of IGFs and IGF Complexes on Proteoglycan Metabolism in Long-Term Articular Cartilage Explant Cultures The ability of IGFs and IGF complexes to affect proteoglycan synthesis and tissue levels of proteoTABLE 1. Relutive rutes uf synthesis of proteoglycans by curtiluge explants cultured in the presence of IGb-I, IGF-II, IGF-I-BP complex and IGF-II-BP complex [35S]Sulfateincorporation into proteog~ycans( X dpmil00 mg wet weight per 2 h) ____
Addition to medium 20% (vh) Fetal calf serum (A) IGF-I (2 FU NSILA/ml) (B) IGF-I-BP complex (2 KU NSILNml) (C) 0 01% (wlv) BSA (D) IGF-II(2 (*.UNSILAld) (E) IGF-11-BP complex (2 FU NSILA/ml) (F)
Large proteoglycans
Small proteoglycans
564.2 (3.1) 430.8 (2.4)
28.4 (1 .0) 24.6 (0.9)
279.1 (1.5) 181.5 (1.0) 395.0 (2.2)
17.2 (0.6) 29.0 (1.0) 19.9 (0.7)
371.6 (2.1)
22.5 (0.8)
Data from the elution profiles of proteoglycans shown in Fig. 4A-F and that given in Fig. 2 on the rate of synthesis of proteoglycans were used to calculatc the rate of incorporation of [35S]sulfate into the two species of proteoglycans present in explant cultures of cartilage maintained in medium containing different IGF fractions. Values in parentheses represent the increase in synthesis expressed relative to the rate measured in cultures incubated in medium containing 0.01% (wh) bovine serum albumin.
19
glycans over a 5-day period in explant cultures of cartilage was investigated. Fresh articular cartilage was maintained for =3days in medium containing 20% (viv) fetal calf serum, 0.01% (wiv) bovine serum albumin or in medium containing 0.01% (wiv) bovine albumin and IGF-1 (20 ngiml = 1.4 pU NSILAlml) or IGF-I1 (SO ngiml = 4.5 p,U NSILAi ml) or 1 pU NSTLAiml of either IGF-I-BP complex or IGF-11-BP complex. The data shown in Table 2 cover two experiments using separate batches of tissue and shows that the rate of proteoglycan synthesis by cartilage maintained in culture in medium containing fetal calf serum, IGF-I, IGF-11, and each of the IGF-BP complexes was equal to or greater than the rate observed on day 0. Tissue cultured throughout with medium containing 0.01% (wiv) bovine serum albumin showed a decline in the rate of synthesis. Cultures maintained in medium containing fetal calf serum, IGF-I, and IGF-I1 showed no or little decline in the proteoglycan content of the tissue (measured as hexuronate content) during the 5-day culture period. Cartilage maintained in medium containing 0.01% (wiv) bovine serum albumin showed a decline in proteoglycan levels. When tissue was maintained in medium containing IGF-IIBP complex, a slight fall in tissue proteoglycan level was evident, similar to that observed in the same tissue maintained in medium containing IGF-I or fetal calf Serum. However, when the same tissue was maintained in medium containing IGF-I-BP complex, a greater fall in proteoglycan tissue levels was evident, similar to that obtained for tissue cultured in medium containing 0.01% (wh) bovine serum albumin. The effect of IGF-I-BP and IGF-11-BP complexes on the rate of proteoglycan catabolism in articular cartilage explants was also determined. Articular cartilage was maintained in culture for 5 days and then incubated with [35S]sulfatefor 6 h, washed to remove unincorporated label, and placed in culture in medium supplemented with 0.01% bovine serum albumin and either IGF-I-BP complex ( 1 pU NSILAlml) or IGF-11-BP complex (1 pU NSILA/ ml) or in medium containing 0.01% (wiv) bovine serum albumin or 20% (viv) fetal calf serum. Figure 5 shows the rate of loss of [35S]-labeledproteoglycans from the cartilage explants. The half-lives of proteoglycans were determined and found to be 15 and 18 days for tissue maintained with IGF-I-BP complex or IGF-11-BP complex, respectively. For cultures maintained in medium containing 0.01%
3 Orthop Res, Vol. 10, N o . 1 , 1992
G . H . TESCH ET AL.
20
TABLE 2. EHect of free and bound lGFs on proteoglycan synthesis and hexuronute content of cartilage explants Days in culture
Addition to medium
Experiment 1 0 5 5 5 5 Experiment 2 0 5 5 5 5
0.01% (wlv) BSA 20% (viv) Fetal calf serum IGF-I (1.4 pU NSILAlml) IGF-I1 (4.5 pU NSILAlmI)
[35S]Sulfateincorporation into proteoglycans (X lo-’; dpm/ 100 mg wet weight per 2 h)
Hexuronate content (pg hexwonate/ mg wet weight)
107.7 & 6.4 (1.0) 48.3 -+ 8.6 (0.5) 304.7 2 35.7 (2.8) 160.9 2 21.1 (1.5) 132.4 ? 0.8 (1.2)
5.4 -+ 0.1 (1.0) 4.7 k 0.2 (0.9) 5.5 & 0.5 (1.0) 5.6 f 0.1 (1.0) 5.5 & 0.4 (1.0)
15.8 -+ 13.2 48.6 2 20.2 -t 32.9 ?
-
1.4 (1.0) (0.8) 3.7 (3.1) 0.6 (1.3) 0.2 (2.1)
* 1.4
0.1% (wlv) BSA 20% ( d v ) Fetal calf serum IGF-I-BP complex (1 pU NSILAlml) IGF-II-BP complex (1 pU NSILAlml)
6.9 2 0.3 (1.0) 4.9 2 0.2 (0.7) 6.4 2 0.1 (0.9) 5.3 ? 0.2 (0.8) 6.1 ? 0.1 (0.9)
Cartilage was maintained in explant culture for 5 days in medium containing 0.01% (wiv) BSA supplemented with either IGFs or IGF-BP complexes. On day 0 or day 5 of culture the rate of synthesis and hexuronate content of the explant cultures was determined. Each point is the mean of three determinations + standard deviation. The value in parenthesis represents the synthesis or hexuronate content expressed relative to the values obtained for articular cartilage measured before it was placed in culture.
(w/v) bovine serum albumin or 20% (v/v) fetal calf serum, the half-lives were found to be 10 and 15 days, respectively. This suggests that IGF-I-BP and IGF-II-BP complexes suppress the rate of proteoglycan catabolism in articular cartilage explant cultures.
bQ
I
I
I
I
6
7
8
I
1
9 10 Days in culture
I
I
11
12
I
FIG. 5. Effect of IGF-I-BP complex and IGF-II-BP complex on the rate of loss of [35S]-labeledproteoglycans from articular cartilage explants. Cartilage from a single animal was maintained in culture for 5 days in medium containing 20% (v/v) fetal calf serum, incubated for 6 h with [35S]sulfate,then replaced in culture in medium containing IGF-I-BP complex;.( 1 pU NSILNml), IGF-II-BP complex (A;1 pU NSILNml), 20% (vlv) fetal calf serum (0),or 0.01% (wlv) BSA (0). The logarithm of the percentage of the [35S]-labeled proteoglycans remaining in the matrix on each day was plotted against time in culture. Each point is a mean of two determinations.
J Orthop Res, Vol. 10, N O . I , 1992
DISCUSSION
In this study it was shown that free IGF-I is more potent in stimulating proteoglycan synthesis than free IGF-IT. regardless of whether doses are compared on a weight basis or on a NSILA basis. This is comparable to previous findings that show that recombinant human IGF-I is a more potent stimulator of proteoglycan synthesis by articular cartilage explants than purified rat IGF-II(l7). However, the magnitude of stimulation by TGF-I over IGF-I1 reported in the current experiments is much lower than that reported by Luyten and associates (17) and may reflect either the difference in the age of the animals used or the use of human IGF-I1 in this study. When IGF-I1 was added to IGF-I, the stimulation was further increased but was less than additive. This suggests that stimulation occurs, at least in part, through different mechanisms, e.g., both type 1 IGF and type I1 IGF receptors. Luyten and collaborators (17) suggested that IGF-I and IGF-I1 stimulate proteoglycan synthesis in articular cartilage from young animals through type I IGF receptors only. A number of groups (16,24,25) have shown the existence of both type I and type I1 IGF receptors on the chondrocytes of 4-6-week-old rats and 18-day-old rabbits. The less than additive effect of IGF-I and IGF-I1 seen in our study could be explained by competitive binding of IGF-I and IGF-I1 for both the type 1 and type I1 receptors. The affinity of IGF-I and IGF-I1 for each other’s receptors has been well documented (see review, 23). IGF-I-
INSULIN-LIKE GROWTH FACTORS AND CARTILAGE METABOLISM
BP complex and IGF-11-BP complex were both shown to stimulate proteoglycan synthesis, but to different degrees. Unlike the IGF-I-BP complex fraction, the IGF-IT-BP complex fraction showed a greater stimulation of proteoglycan synthesis compared with free IGF-I1 at low concentrations (1.0 pU NSILAlml). At higher concentrations (i.e., 2.0 pU NSILAlml), the stimulation by free and bound IGF-11 was similar. The data presented here shows that when IGF-IBP and TGF-11-BP complexes were included in the medium of explant cultures of adult articular cartilage, there was a decrease in the rate of catabolism of proteoglycans compared with control cultures maintained in medium containing 0.01% (wiv) bovine serum albumin. Both free and complexed IGFs were shown to maintain proteoglycan synthesis by explant cultures to varying degrees. Since the levels of proteoglycan present in the extracellular matrix of the cartilage explants must reflect the sum of the rates of synthesis and catabolism of these macromolecules by the tissue, the inability of IGF-I-BP complex to maintain proteoglycan levels in cartilage explants may be due in part to the inability of this growth factor complex to stimulate proteoglycan synthesis. Since our work shows that the maximal stimulation of proteoglycan synthesis is achieved with bound IGF-I and IGF-I1 at a concentration of -I pU NSILAlml , higher concentrations of bound IGF-I 118 ng = 12 pU NSILAlml, as determined in the rat adipocyte assay; (1 I)] and bound IGF-II(600 ng = 50 pU NSILNml), such as those found in plasma, are unlikely to result in further stimulation of proteoglycan synthesis by articular cartilage. Throughout this work a variation was observed in the magnitude of the response of articular cartilage from different animals when exposed to free or bound IGFs, but nevertheless the growth factors always stimulated proteoglycan synthesis. The variation in stimulation of proteoglycan synthesis observed between tissue from different animals expressed relative to values obtained for the same tissue maintained in culture in medium containing 0.01% (wh) bovine serum albumin were: for cultures maintained in medium containing IGF-I (1 $J NSJLNml), 2.6 t 0.5 (mean t SD) (tissue from seven animals); for IGF-II(1 pU NSILA/ml), 2.1 2 0.3 (tissue from five animals); for IGF-I-BP complex (1 pU NSILA/ml), 1.4 e 0.1 (tissue from six animals); for IGF-11-BP complex (1 pU NSII.A/ml),
21
2.4 2 0.6 (tissue from five animals); and for 20% (vlv) fetal calf serum, 3.8 t 0.8 (tissue from eight animals). There is increasing evidence to suggest that BPs can modify the actions of IGFs on various tissue types in a stimulatory and inhibitory manner (1). For instance, the insulin-like activity of circulating IGFs is inhibited by the formation of the 150,000 dalton complex, which is incapable of binding to the insulin receptor (10). Furthermore, it has been pointed out that IGF-BP complexes may regulate their release of IGFs, thereby controlling the target tissue receptor occupancy (3). IGFs have been reported to be present in synovial fluid from humans (26) and steers (C. J. Handley, unpublished information). In the latter case the IGF activity was associated with a high-molecular-weight complex that was dissociated under acid conditions. It is not known whether the BP complexes of both IGF-I and IGF-I1 can penetrate the extracellular matrix of cartilage and interact with receptors on chondrocytes. Work by Clernmons and associates (6) has shown that when samples containing IGF-BP complexes are incubated for 24 h with highly negatively charged macromolecules, such as heparin, or with proteinases, the BPs are stripped from the complex, leaving the free IGF. Since the extracellular matrix of cartilage contains highly negatively charged proteoglycan and proteinases, it is possible that IGFs and their BPs may be separated before the IGFs bind to their receptors. The activity of the IGF-BP complexes observed in cartilage may therefore be dependent on the ability of components of the extracellular matrix to separate out the BPs. If the IGF-BP complexes were to reach the cells, then the difference between the stimulation of proteoglycan synthesis observed for the IGF-I-BP and the IGF11-BP complexes might be explained by a differing ability to bind to their respective receptors on the chondrocyte, as previously noted (23). IGF-I and IGF-I1 in both free and bound forms and fetal calf serum were all shown to stimulate preferentially the synthesis by cartilage explants of the large proteoglycans capable of forming aggregates with hyaluronate and link protein without affecting the synthesis of the small proteoglycans. This finding is in agreement with the observations of McQuillan and colleagues (19), which suggested that fetal calf serum and IGF-I stimulate the synthesis of the large proteoglycans of articular cartilage. These proteoglycans are characteristic of car-
J Orthop Res, Vol. 10, No. 1, 1992
22
G. H . TESCIY ET AL.
tilage and are responsible for the biochemical properties of this tissue. This work indicates that the lGFs may be responsible for the maintenance of the phenotype of cartilage. This finding is important not only in the development of the tissue but also in later life, for the maintenance of important components of the extracellular matrix of the tissue. If there is a defect in the expression of the IGFs or their receptors, it may lead to degeneration of the tissue. Acknowledgment: The authors thank Mrs. M. T. Scott and Ms. M. Bistrin for expert technical assistance and Castricum Bros. Pty. Ltd. (Ferntree Gully Abattoirs) for the donation of bovine hocks. This work was supported by the National Health and Medical Research Council of Australia and the Arthritis Foundation of Australia.
REFERENCES 1. Baxter RC, Martin JL: Binding proteins for the insulin-like growth factors: Structure, regulation and functlon. Prog Growth F a d o r Res 1:49-68, 1989 2. Bitter T, Muir HI: A modified uronic acid carbazole reaction. Anallt Biochem 4:330-334, 1962 3. Blum WF, Jenne EW, Reppin F, Kietzman K . Ranke MB, Bierich JK: Insulin-like growth factor I protein complex is a better mitogen than free IGF-I. Endocrinology 125:76&772, 1989 4. Campbell MA, Handley CJ: The effect of retinoic acid on the turnover of proteoglycans in cultured articular cartilage. Arch Biorhem Biophys 258:143-155, 1987 5. Campbell MA, Handley CJ, Hascall VC. Campbell KA, Lowther DA: Turnover of proteoglycans in cultures of bovine articular cartilage. Arch Biochem Biophys 234:275-281, 1984 6. Clemmons DR, Underwood LE, Chatelain PG, van Wyk JJ: Liberation of immunoreactive somatomedin-C from its binding proteins by proteolytic enzymes and heparin. J Chi1 Endocrinnl Metab 56:384-389, 1983 7. Cornell HJ, Enberg G , Herington AC: Preferential association of the insulin-like growth factors I and I1 with metabolically inactive and active carrier bound complexes in serum. Biochern J 241:745-750, 1987 8. Daughaday WH, Kipnis DM: The growth promoting and anti-insulin actions of stromatotrophin Rec Prog Horm Res 22:49-93, 1966 9. Franklin RC. Renme GC, Burger HG, Cameron DP: A bioassay for NSILA-S In individual serum samples and its relationship to somatotropin. J Clin Endocrinol M e t a b 43:1164-1169, 1976 10. Guler HP, Zapf J, Froesch ER: Short-term metabolic effects of recombinant human insulin-like growth factor-I in healthy adults. N e n Engl J Med 317:137-140, 1987 11. Handley CJ, Lowther DA: Extracellular matnx metabolism by chondrocytes. 111. Modulation of proteoglycan synthesis by extracellular levels of proteoglycans in cartilage cells in culture. Biochim Biophyr Actn 500.132-139, 1977 12. Handley CJ, Campbell MA: Catabolism and turnover of proteoglycans. Meth Enzymol 144:412419, 1987
I Orthop Res, Voi. IO, No. 1 , 1992
13. Handley CJ, Lowther IDA, McQuillan DJ: 'The structure and synthesis of proteoglycans of articular cartilage. Cell B i d Int RPU9~753-782, 1985 14. Hascall VC, Handley CJ, McQuillan DJ, Hascall GK, Robinson HC. Lowther DA: The effect of serum on biosynthesis of proteoglycans by bovine articular cartilage in culture. Arch Biochem Biophys 224:20&223, 1983 15. Herington AC, Cornell HJ, Kuffer AD: Recent advances in the biochemistry and physiology of the insulin-like growth factorisomatomedin family. Int J Biochem IS: 1201-1210, 1983 16. Jansen J. van Buul-Offers SC, Hoogerbrugge CM, de Porter TL, Corvol MT, van den Brande JL: Characterization of specific insulin-like growth factor-I and -11 receptors on cultured rabbit articular chondrocyte membranes. J Endocrinol 1 20:245-249, 1989 17. Luyten FP, Hascall V C , Nissley SP, Morales TI, Reddi AH: Insulin-like growth factors maintain steady-state metabolism of proteoglycans in bovine articular cartilage explants. Arch Biochem Biophys 267:416-425, 1988 18. McQuillan DJ, Handley CJ, Robinson HC, Ng K, Tzaicos C, Brooks PR, Lowther DA: Effects of cycloheximide on chondroitin sulphate synthesis. Biochem J 2241977-988; 1984 19. McQuillan DJ, Handley CJ, Campbell MA, Bolis S, Milway VE, Herington, AC: The stimulation of proteoglycan biosynthesis by serum and insulin-like growth factor-I in cultured bovine articular cartilage. Biochem J 240:423-430, 1986 20. Muir H: The chemistry of the ground substance ofjoint cartilage. In: Sokoloff L , ed. The Joinfs and Synovinl Fluid, vol. 2, New York: Academic Press, 1980, pp 27-94. 21. Oegema TR. Hascall VC. Eisenstein K: Characterization of bovine aorta proteoglycan extracted with guanidine hydrochloride in the presence of protease inhibitors. J Riol Chem 259: 172G1726, 1979 22. Poffenbarger PL: The purification and partial characterization of an insulin-like protein from human serum. J Clin Invest 56:1455-1463, 1975 23. Rechler MM, Nissley SP: The nature and regulation of the receptors for insulin-like growth factors. Ann Rev Physiol 47:425-442, 1985 24, Schalch DS, Sessions CM, Fdrley AC, Masakawa A, Emler CA. Dills DG: Interaction of insulin-like growth factor I/ somatomedin C with cultured rat chondrocytes: Receptor binding and internalization. Endocrinology 118: 1590-1597, 1986 25. Sessions CM. Emler CA, Schalch DS: Interaction of insulinlike growth factor-I1 with rat chondrocytes: Receptor binding, internalization and degradation. EndocrinnloRv 120:2108-2116, 1987 26. Schalkwijk J , Joosten LAB, van den Berg WB, van Wyk JJ, van de Putte LBA: Insulin-like growth factor stimulation of chondrocyte proteoglycan synthesis by human synovial fluid. Arth Rheum 32:66-71, 1989 27. Zapf J, Schoenle E, Froesch ER: Insulin-like growth factors I and 11: Some biological actions and receptor binding characteristics of two purified constituents of nonsuppressible insulin-like activity of human serum. Eur J Biochem 87:285296, 1978 28. Zapf J, Walter H, Froesch ER: Kadioimmunological determination of insulin-like growth factors 1 and I1 in normal subjects and in patients with growth disorders and extrapancreatic tumor hypoglycemia. J Clin Invest 68: 1321-1330, 1981
