(Wight, T. N. & Mechem. K. P.. eds.). pp. 59-104. Academic. Press, New York. Turnbull, J. E. & Gallagher, J. T. (1988) Biochem. J. 251,. 597-608. Turnbull, J. E.
PROTEOGLYCANS
209
I I . Laurie, G. W., Bing, J. T., Kleinemann, H. Y., Hassell, J. R., 24. Cheifetz. S. & Massague, J. (1989) J. Hiol. C'hem. 264, 12025- 12028 Aumailley. M., Martin, G. K. & Feldman, R. J. (1986) J . Mol. 25. Heremans, A,, Cassiman, J.-J., Van Den Berghe. H. & David. G. Biol. 189,205-216 12. Kanwar, Y. S., Hascall, V. C. & Farquhar, M. G. (1980) J. Cell ( 1988) J . Biol. (%rm. 263, 473 1-4739 Riol. 90,527-532 26. Lindblom. A,, Carlstedt. I . & Fransson. L. A . ( 19x9) Riochem. J. in the press 13. Hayashi, K. M., Hayashi. M., Jalkanen, M., Firestonc. J. H.. Trelstad, K. L. & Bernfield. M. ( 1987) J . Hislochem. C'yiochem. 27. Schmidtchen, A,, Carlstedt. I.. Malmstrom. A . & Fransson. 35, 1079-1088 L.A.(1989)Hiot.hem..I. 261, 145-153 14. Saunders, S. & Bernfield, M. ( 1988) J. ('ell Bid. 106,423-430 28. Lorics, V., De Roeck. H., David, G., Casseman. J.-J. & Van den 15. Saunders, S., Jalkanen, M., OFarrell, S. & Bernfield, M. ( 1989) Berghe, H. (1987)J . Hiol. f'/?c,m.262,854-859 J. Cell Bid. 108, 1547- I556 29. Ishihara. M.. Fedarko. N. & Conrad. H. E. ( I 986) J. Hiol. ('/?em.261, 13575-135x0 16. Bourdon, M. A,, Krusius. T.. Campbell, S.. Schwartz. N. B. 62 Ruoslahti, E. ( 1987) /'roc. N u / / . Acntl'. Sci. U.S.A. 84, 30. Gallagher, J. T. & Walker, A. ( 1 985) Biochem. J . 230,665-674 31. Lindahl. U. & Kjellin, L. ( 1987) in Biology ff I'roteoglwrrzs 3 194-3 I98 17. Ledbetter, S. R., Tyree, B.. Hashell. J. R. & Horrigan. E. A. (Wight, T . N. & Mechem. K. P.. eds.). pp. 59-104. Academic Press, New York ( 1985) J . Biol. C'hrm. 260,8106-8 I 13 18. Hassell, J. R., Noonan, D. M., Ledbetter, S. R. & Laurie. G. W. 32. Turnbull, J. E. & Gallagher, J. T. (1988) Biochem. J. 251, 597-608 (1986) C'ihri Found. Symp. 124,204-222 19. Noonan. D. M.. Horrigan. E. A.. Ledbetter. S. R.. Vogeli. G., 33. Turnbull, J. E. & Gallagher. J. T. ( 1990) H i o c . l i m i . J. 265, in the presh Sasaki. M.. Yamada. Y. & Hassell. J . R. ( 1988) J . Riol. ('hem. 263, 16379-16387 34. Merchant, Z. M., Kim, Y. S., k c e , K. G . & Linhardt, R. J. 20. Paulsson. M.. Yurchenko. P. D.. Ruben. G. C.. Engel. J. & ( 1985) Biochem. J. 299,369-377 3 5 . Casu. B., Petitou. M., Provasoli, M. & Sinay, P. ( 1988) Trc.nc1.c Timpl, R. ( 1987) J . Mol. Biol. 197,297-3 I3 21. Clement, B.. Seyui-Real, B., Hassell, J. R.. Martin, G. R. &i Hiochem. Sci. 1 3 , 2 2 1-225 Yamada. Y. ( 19891 J. Biol. C'hem. 264. 12467- 1247 1 36. Sanderson. P. N.. Huckerbv. T. M. & Nieduszvnstki. I . A. 22. Segarini, P. R. &'Seyedin, S. M. (1988) J. Biol. ('hem. 263, ( 1987) Biochem. J . 243, 175: I8 1 8366-8370 23. Cheifetz. S., Andres, J. L. & Massague. J. ( 1988) J . Hiol. C'liem. 263.16984-16991 Received I7 October I989 I
/
Structural variability of large and small chondroitin sulphate/dermatan sulphate proteoglycans DICK HEINEGARD, ERIK HEDBOM. PER ANTONSSON and AKE OLDBERG Depurtment of Medical and P ~ y s i o l o ~ i cChemistry, al Lirtrd Utriversity, P.O. Box 94, S-221 M>Litrid, Sweden Extracellular matrices contain several types of proteoglycans, having side-chains of chondroitin sulphate/ dermatan sulphate in common. There are two major groups of these proteoglycans. In one group, the molecules have molecular masses in the order of millions and have extended core proteins of about 4000 nm [ l ] with a high M, of about 200000 [2, 31. The proteoglycans in the other group have molecular masses of about 100000 and the globular core protein has an apparent M, of about 45 000 [4].
Large uggregatirrg proteoglycatrs There appear to exist at least three types of the large core proteins. The one in cartilage proteoglycans appears unique for this tissue. Fibrous tissues contain one type of similar large proteoglycans and the intima or aorta and muscle contains a third, somewhat more distinct, large proteoglycan. The different core proteins of the three proteoglycans have some structural features in common [ 11. They all contain one globular domain in their C-terminal end [ I ] , which differs somewhat in character between proteoglycans of this class. The one in cartilage proteoglycans appear to be rather conserved between molecules from chicken, rat and bovine sources and shows sequence similarity with a liver lectin (see IS]) and also with the C-terminal globular domain in fibroblast proteoglycans [6].All three types of large proteoglycans have a globular protein domain in the N-terminus that has the capacity to specifically interact with hyaluronate. This allows for formation of large aggregates containing several monomers - schematically depicted in Fig. 1. While the proteoglycans from aorta and muscle contain only one detectable globular domain in each end [ 1, 71, those VOl. 18
from cartilage and those from fibrous tissues contain a second globular domain close to the N-terminus [ I ] . The sequence of this domain in cartilage proteoglycans shows extensive sequence similarity with the hyaluronate-binding N-terminal globular structure [2], although it does not bind to hyaluronate [8]. The remaining part of the core protein of the proteoglycans represents the extended portion containing the glycosaminoglycan chains. This portion shows differences between the molecules in the three groups 141. Thus, spacing of glycosaminoglycan chains, their distribution to different regions along the core protein and the size and character o f the side-chains differ between the three groups [ 1, 41, and also within the groups between species. Thus the protcoglycans from cartilage, although apparently specific for this tissue, show differences between species. Typically, those from rat contain no, or very small amounts, of keratan sulphate compared with those from larger mammals like cow and man (see 191).At the protein level this depends on the presence of a region of a hexapeptide repeated 23 times in the proteoglycans [ 101. This domain is not present in the proteoglycans from rat, apparently a result of differences at the gene level [lo]. The 23 repeat structures are shown in Fig. 2. The major part of the core protein carrying the chondroitin sulphate side-chains contains two types of repeat structures 12, 5 , 101 and altogether constitutes about half of the above 2200 amino acids [2]. The chondroitin sulphate side-chains are attached to Ser-Gly dipeptides. Thus, about half of the chains, in the C-terminal-half of the core protein are distributed in clusters each made up of seven consecutive units of a ten amino acid repeat structure IS].The other half of the chains are located in longer repeat structures, showing some difference between species with some 60n/n conserved residues [2, 101. The side-chains in the other two types of large aggregating proteoglycans are considerably more sparsely distributed along the core protein as viewed by electron microscopy [ 11.
BIOCHEMICAL SOCIETY TRANSACTIONS
210 Domain 1 CS-chains
Domain 2 CS-chains
Fig. 1. Schematic illustration of the structural features of the large aggregating proteoglycan from bovine cartilage The protein region containing the keratan sulphate (KS)-rich region is not present in rat. CS, chondroitin sulphate.
- _
2
E G P S A T
212-217
E V P
2 18-22 3
E K P E E P E E P
224-229 2 30-235 236-241
E K P F P P
242-247
E E L F P S
248-253
1;
E K P F P S
254-259
E K P F P S
260-265
E E P F P S
266-271
E K P F P P
272-277
E E L F P S
278-283
E K P I P S
284-289
E E P F P S
290-295
E E P F P S
296-301
E E K PP F P P
302-307 308-313
E K P
314-319
E E P
320-325
E K P F P S
326-331
E E P F P S
332-337 I
E E P S T L
338-343
P V
344-349
:A
s ]
Fig. 2. The 23 repeats of 6 amino acid residues representing the keratan sulphate-rich region
Only limited sequence data on one of these structural proteoglycans has presently been published [ 61. Although available data indicate that the large proteoglycans in cartilage are secreted as one class of molecules [lo],at least two rather polydisperse populations can be isolated from the tissue 131. One population appears to be formed by partial proteolytic degradation from the Cterminal end of the proteoglycans. The fragments carrying the hyaluronate binding region are retained in the tissue bound in proteoglycan aggregates, while the C-terminal fragments appear to be lost from the tissue. Thus, the proteoglycans in the tissue are polydisperse primarily as a result of a slowly progressive proteolysis in the cartilage. Small proteoglycans A group of interstitial proteoglycans with distinctly different character are much smaller, with core proteins of M , around 30 000-40 000. These core proteins share structural features, but do represent different gene products [4]. They are substituted near their N-terminus with either one or two side-chains of dermatan sulphate or chondroitin sulphate, depending on the type of proteoglycan and the tissue of origin [4, 11, 121. Two types of these proteoglycans from a bovine source have been cloned and sequenced [ 12- 141. They show extensive sequence similarity, containing the same number of conserved cysteine residues. One of these proteoglycans (PG-S1) contains two side-chains in the Nterminal region attached at typical Ser-Gly sequences, while the other proteoglycan (decorin)[ 151 contains only one sidechain attached close to the N-terminus. A related and very similar ubiquitous matrix protein, fibromodulin, initially isolated from articular cartilage [I 61, contains keratan sulphate attached at site(s) different from the N-terminal glycosaminoglycan-binding sites of the other two members of this group of molecules [14]. This protein contains no Ser-Gly sequence in its N-terminal portion. It is interesting to note, however, that this region in fibromodulin contains a tyrosinerich structure characteristic of sulphate tyrosine residues. In preliminary experiments, we could indeed show that the protein isolated by immunoprecipitation from fibroblasts labelled with [%]sulphate contained the radiolabel in a position apparently different from keratan sulphate, probably representing a tyrosine sulphate domain (P. Antonsson, A. Oldberg & D. Heinegard, unpublished work). Thus the data
1990
PROTEOGLYCA NS
21 1
can be taken to indicate that fibromodulin, similarly to the two small proteoglycans, contains a polyanionic domain close to the N-terminus. The major part of all three proteins contain similar approximately 25 amino acid residues long repeats showing extensive sequence similarity. Although all proteins contain cysteine residues, and probably disulphide bond at conserved locations [ 141, they appear to be functionally different. Decorin and fibromodulin both bind to collagens I and I1 and interfere with fibrillogenesis it1 wifro [ 171, while the third protein/proteoglycan (PG-S 1) appears not to bind. This binding does not depend on the presence of the glycosaminoglycan side-chains and interestingly
appears to be directed to different sites on the collagen fibre ~71. The two small proteoglycans are present in most tissues. Both proteoglycans have been sequenced from bone and cartilage [ 12-14]. Furthermore, these proteoglycans have been isolated from other tissues like sclera, intima of aorta, cornea and tendon [4]. Patterns of tryptic peptides can be used to identify the proteoglycans (Fig. 3 ) and also to demonstrate minor modifications of the two types of patterns, probably a result of different substitution with carbohydrate side-chain substituents. In particular, the decorin proteoglycan from bone and cornea, although showing a tryptic
II
Sclera
Nasal cartilage w, 1
-40
-30
8 ;
.z .-
0
L
-20 - 10
L1 1
r
j j
111111 1
L I i i L L l i U >L l L L L l U 1 1 1 1 1 J 1 L1111 I12 1 1 1 I L I J I I U
Retention time ( m i n )
"3
2
~ , , ~ l U l , U l ' L L L L l L U U l L l l i l i l i l i l i l i l IU, , , ,
II'IULLII
Retention time (mini
Fig. 3. 7 j p f i cpeptide puttems of.f'smuNproteoglycam~from wirioiis .soiirces Proteoglycans (30 pg) o f the two side-chain (PG-S 1 ) type ( u )and of the decorin (PG-S2)type ( h )were digested with N-tosyl-i*phenylalanine chloromethyl ketone-treated trypsin under conditions ensuring limited digestion and chromatographcd o n Lichrosorb C- 18 reversed phase columns (Merck, Darmstadt, B.D.R.) eluted with a gradient of acetonitrile in 0.1% (v/v) aqueous trifluoroacetic acid. Absorbancc at 225 nm was recorded. Baseline obtained in a separate run has been subtracted using a Nelson computer program for data handling. Vol. 18
212
BIOCHEMICAL SOCIETY TRANSACTIONS
peptide pattern similar to the other molecules of this class, also exhibits some differences in cleavage patterns (Fig. 3). There are other differences also between the side-chain substituents of the small proteoglycans. Decorin isolated from adult bovine bone and bovine nasal cartilage is substituted with a chondroitin sulphate chain. The molecule from cornea contains a dermatan sulphate chain with a low degree of epimerization with iduronate (see [4]),while the proteoglycans from sclera, tendon, articular cartilage and skin contain a dermatan sulphate chain with a high degree of epirnerization with iduronate. The two side-chain class of small proteoglycans, when isolated from nasal cartilage, contain chondroitin sulphate chains, while the one isolated from the intima of aorta contains highly epimerized, iduronaterich dermatan sulphate chains 141. Thus, one type of protein core of these small proteoglycans may be substituted with chondroitin sulphate or dermatan sulphate side-chains depending on the tissue of origin. From these results, it appears that the signal for a specific type of side-chain does not depend on the core protein. It is possible that this regulation depends on availability of the epimerase in different cells.
C‘onclnding remarks During the last few years our knowledge of the structure of the protein core of proteoglycans in the extracellular matrix has rapidly expanded. Thus, the primary sequence of several distinct, typical core proteins has been determined. Future work should be focused on establishing the functions of the proteoglycans. Although it is quite clear that a major function of the large proteoglycans is to provide resistance to load, the presence of keratan sulphate chains remains an enigma. It is of particular interest that the domain of the protein core containing the major proportion of these chains is not present in rat, while proteoglycans from human and bovine cartilage contains this domain. The functions of the small proteoglycans are not clear, although decorin appears to have a role in regulating collagen fibrillogenesis. Several attempts to identify an interaction of the structurally related two side-chain small proteoglycans have been negative. This is surprising in view of the fact that fibromodulin, a protein structurally closely related to these proteoglycans, like decorin binds to collagen.
At present our understanding of the function of the sidechain(s) is limited. It is possible that an improved knowledge of the regulation of the synthesis of a specific type of these glycosaminoglycan side-chains will provide some clues to and in understanding their function. This study was supported by grants from the Swedish Medical Research Council. Folksarns Stiftelse. Kocks Stiftelser, b t e r l u n d s Stiftelse, Konung Gustaf V s 80-arsfond and the Medical Faculty, Lund University. 1. Miirgelin, M., Paulsson, M., Malmstrijm, A. & Heincgird, D. ( 1989) J. Rid. Chern. 264, 12080- I2090 2. Doege, K., Sasaki, M.. Horigan, E., Hassell. J. R. & Yamada. Y. (1987)J. B i d . Chem. 262, 17757-17767 3. Heinegard, D., Sheehan, J., Sommarin, Y., Wieslander, J. & Paulsson, M. ( 1985) Biochem. 1. 2 2 5 9 5 - 106 4. Heinegard, D.. Bjorne-Persson, A., Ciister, L., Franzen, A,, Gardell, S., Malmstrorn. A,, Paulsson, M.. Sandfalk, R. & Vogel, K. ( 1985) Biochem. J. 230, 18 1- I94 5 . Oldberg, A., Antonsson, P. & Heinegard, D. ( 1 987) Biochem. J.
243,255-259 6. Krusius, T., Gehlsen, K. & Ruoslathi, E. ( 1 987) J. Hiol. C’hrm. 262,13 120-131 25 7. Haynesworth, S., Carrino, D., Drushel, R. & Caplan, A. ( 1 989) Connect Tissue Rex in the press 8. Morgelin. M.. Paulsson. M., Hardingham, T., Heinegbrd. D. & Engel, J. ( 1988) Biochem. J. 253, 175- I85 9. Heinegbrd, D. & Oldberg, A. ( 1989) FASEB J. 3, 204200205 1 10. Antonsson, P.. Oldberg. A. & Heinegard, D. ( 1 989) J. Hiol. [’hem. 264, 16 1 70- 16 1 73 1 I . Rosenberg, L., Choi, H., Tang, L.-H., Johnson, I‘.. Pal, S., Webber, C., Reiner. A. & Poole, R. ( I 985) J. H i d . (‘hem. 264, 457 1-4576 12. Fischer. L.,Termine. J . & Young. M. ( 1989) J . Riol. (’hem. 264, 457 1-4576 13. Nearne, P.. Choi. H. & Roenberg. L. (1989) J. Hiol. C % e m 264,8653-866 1 14. Oldberg, A., Antonsson. P., Lindblom. K. & Heinegird, D. ( I9891 EMBO J. 8. 260 1-2604 15. Ruoslahti, E. ( 1988) Annu. Re\.. (‘dl Riol. 4, 229-255 16. Heinegard, D., Larsson, T., Sommarin, Y., FranzCn. A., Paulsson, M. & Hedbom, E. (1986) J. Brol. C’hvrn. 261. 13866-13872 17. Hedbom, E. & Heinegard, D. (1989) J. Riol. C’hern. 264, 6898-6905 Received 17 October 1989
Osteoarthritis - clinical aspects PATRICK J. CASHIN Deplrrtmerit of Rheumutology. South lrijirrnuty, Cork, Repiihlic of’lrelurid Hyaline articular cartilage, the joint bearing, is one of the least expendable components of diarthrodial joints. It is not surprising, therefore, that osteoarthritis - a condition where changes in articular cartilage are one o f its striking features is a common cause of impairment of joint function. Prior concepts of the condition, as primarily an age-related degenerative process, have had to be modified to a more dynamic perception in the light of increasing data on the structure and function and reparative capacity of hyaline articular cartilage. As the pathological process. especially during its incipient phase, remains unclear, all definitions to date must be regarded as somewhat incomplete. A recent definition of the condition by the American Rheumatism Association is as follows ‘a heterogenous (sic)group of conditions that lead to joint symptoms and signs which are associated with defective integrity of cartilage. in addition to the
related changes in underlying bone and at the joint margin’ [ I ] . The presence o f a number o f factors acting singly o r in concert is presently the accepted hypothesis of the osteoarthritis disease process (see Fig. 1 ). Clinically. the patient presents with joint pains, stiffness and impairment of function. As hyaline articular cartilagc is aneural. the source of the joint pain is likely t o be a reaction in the innervated subchondral bone and the pcriarticular tissues. Considerable disparity exists between clinical signs and radiological evidence of structural joint disease. Radiographic evidence of osteoarthritis in weight-bearing joints in the population over 40 years. but with relatively few symptoms, is not an uncommon finding [ 21. Osteochondral proliferation at the joint margin (e.g. Herberden’s Nodes). a finding considered typical of osteoarthritis, develops in the absence of symptoms in up to 50% of patients. Reliable noninvasive techniques that will identify and permit study of joint changes during the early phase of the disease should advance our understanding of the pathophysiology and natural history of the disease process. 1990
