- Email: [email protected]
Type X Collagen Isolated from the Hypertrophic Cartilage of Embryonic Chick Tibiae Contains Both Hydroxylysyl- and Lysylpyridinoline Cross-Links
JOBNAME: BBRC 218#3 PAGE: 1 SESS: 15 OUTPUT: Thu Apr 11 23:46:00 1996 /xypage/worksmart/tsp000/68480f/7
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS ARTICLE NO.
219, 301–305 (1996)
0227
Type X Collagen Isolated from the Hypertrophic Cartilage of Embryonic Chick Tibiae Contains Both Hydroxylysyl- and Lysylpyridinoline Cross-Links M. W. Orth,1 L. J. Luchene, and T. M. Schmid Department of Biochemistry, Rush Medical College, Rush Presbyterian-St. Luke’s Medical Center, Chicago, IL 60612 Received December 27, 1995 Hypertrophic cartilage from the tibiotarsus of Day 20 chick embryonic tibiae was found to contain an unusually high concentration of lysylpyridinoline (LP), a nonreducible collagen cross-link normally found only in bone and dentin but not in cartilage. Since type X collagen is abundant in this cartilage, research was conducted to see if type X was the primary source of LP. The 45-kDa pepsin-resistant form of type X was purified by immunoaffinity chromatography. It contained a high concentration of the LP cross-link while type II contained primarily hydroxylysylpyridinoline (HP), the predominant cross-link in cartilage. This, to our knowledge is the first time that type X has been shown directly to form nonreducible cross-links and that a collagen other than type I has a high level of LP. Also, it is interesting that the HP and LP cross-links are found in a collagen that is degraded so rapidly. Possible explanations for these findings are discussed. © 1996 Academic Press, Inc.
During endochondral bone formation, chondrocytes in the proliferative zone of the growth plate synthesize primarily types II, IX, and XI collagen which form fibrils in the extracellular matrix (1). However, as the cells mature and hypertrophy, they decrease their expression of these collagens and begin to express type X collagen (2,3). Type X collagen, a non-fibril forming collagen consisting of three identical 59 kDa chains, is almost exclusively located in hypertrophic cartilage (4), although it has been found in the eggshell membrane (5). Even though its function is not well understood, some have shown its importance by using mice in which formation of the collagen X is perturbed by the insertion of a truncated avian a1(X) gene (6). The mice develop skeletal deformities including accentuated neck lordosis and thoracolumbar kyphosis. Histological analyses of the growth plates in the hind limb of the transgenic mice reveal a reduced thickness in the hypertrophic zone as well as an altered morphology of the hypertrophic chondrocytes (6). However, others have shown normal long bone growth in type X null mice suggesting that only abnormal type X deposition into the matrix will adversely effect bone growth (7). Type X collagen has two cleavage sites for vertebrate collagenase while collagens I, II, and III have only one (8). In solution assays it is degraded faster than type II collagen and in its native state can also be degraded by 72-kDa type IV collagenase (9). Thus, some have thought that it might facilitate matrix turnover and subsequent vascularization by metaphyseal arteries (8). Collagen X has also been thought to be involved in the calcification of the cartilage. It is synthesized by chondrocytes just prior to matrix calcification and persists in regions of calcified cartilage (8,10). Both in vivo and in vitro studies have shown apparently parallel increases in type X and alkaline phosphatase production preceding calcification (11,12). Some investigators have observed an association of type X with matrix vesicles (13,14) while others have not (15). Recently, others have shown that osteocalcin and osteopontin, two proteins associated with the mineralization of bone, are produced by hypertrophic chondrocytes in vitro and found at sites of mineralization (16). Thus, type X might not be directly involved in the calcification of cartilage, but instead provide a permissive environment for it to occur by altering the extracellular matrix. 1
Corresponding author Fax 312-942-3053. 301 0006-291X/96 $18.00 Copyright © 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.
JOBNAME: BBRC 218#3 PAGE: 2 SESS: 15 OUTPUT: Thu Apr 11 23:46:00 1996 /xypage/worksmart/tsp000/68480f/7
Vol. 219, No. 2, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
In the extracellular matrix type X is found in primarily two forms. One is as fine filaments (<5 nm in diameters) within mats which are in the pericellular matrix around hypertrophic chondrocytes (17,18). The second is associated with the surface of fibrils (between 10–20 nm in diameter) composed primarily of type II collagen (15,18). In an avian embryonic chick sternal cartilage culture system, type X is able to diffuse through the matrix and become associated to type II collagen fibrils (19). This association apparently involves lysine-derived covalent cross-links since b-aminoproprionitrile, a lysyl oxidase inhibitor, greatly increases the extractability of type X from cartilage and chondrocyte cultures (20,21). Also, in ovo injections of b-aminoproprionitrile inhibited the covalent interactions of type X with the extracellular matrix of hypertrophic cartilage from the tibiotarsi of embryonic chicks (22). Recently, it was shown that the hypertrophic cartilage from the tibiotarsi of newly hatched broiler chicks contained both of the nonreducible lysine-derived collagen cross-links, hydroxylysylpyridinoline (HP) and lysylpyridinoline (LP), with almost four times the concentration of LP as HP (23). Normally cartilage contains almost exclusively HP. However, one difference between hypertrophic cartilage from the chick and other cartilages is the presence of appreciable levels of collagen X. In avian dyschondroplastic cartilage, the amount of type X collagen is directly proportional to the concentration of LP in the cartilage (Unpublished observations). Thus, the purpose of this research was to determine if type X collagen was involved in the formation of HP and more interestingly LP in avian hypertrophic cartilage. MATERIALS AND METHODS Type X collagen purification. Between 200–600 mg wet weight of cartilage from the proximal tibiotarsus was isolated from the tibiotarsi of day 19 and 20 White Leghorn chick embryos. The cartilage was extracted with 4 M Guanidine, 50 mM Sodium acetate, pH 6.0 containing several protease inhibitors to remove the proteoglycans. After extensive rinsing with water, the cartilage was digested with 1 mg/ml pepsin in 0.5 M acetic acid and 0.2 M NaCl for two days at 4°C. The pepsinized sample was neutralized with NaOH and centrifuged at 10,000 rpm for 30 minutes at 4°C. The supernatant was then dialyzed against 1 mg/ml 3-[(3-cholamidopropyl)-dimethylammonio]-1-propanesulfonate (CHAPS), 50 mM Tris, 5 mM EDTA, and 1 M NaCl at pH 7.5. Type X was purified from the dialyzate using an immunoaffinity column consisting of the XAC-9 anti-type X antibody bound to sepharose beads. The specificity of this monoclonal antibody for type X has been previously demonstrated (4,24). Type X was eluted from the sepharose column with 2 M Guanidine, 50 mM Tris, pH 7.5 and then dialyzed either with 6.25 mM Tris, 5 mM EDTA for SDS-polyacrylamide gel electrophoresis (SDS-PAGE) or in 0.1 M acetic acid for storage and biochemical analysis. Partial purification of Type II Collagen. Pepsinized collagen (primarily types II, IX, and XI) which did not bind the type X immunoaffinity column was dialyzed against 0.5 M acetic acid and then 0.7M NaCl was added to precipitate out the type II collagen. The type II collagen was dialyzed against 0.1M acetic acid and stored at −20°C. Gel electrophoresis. 8% SDS-PAGE gels were performed to determine the purity of the type X preparations. A 45 kDa pepsinized standard of type X as well type II were included. The gels were stained with Coomassie Blue and scanned with a laser densitometer (LKB, Sweden) to measure the purity of the preparation. Collagen cross-link (HP and LP) analysis. Collagen samples were hydrolyzed and partially purified using CF1-cellulose chromatography as previously described (23). The samples were run on a C18 reverse phase column using high performance liquid chromatography (HPLC) according to the protocol of Uebelhart et al. (25). In order to determine the cross-link concentrations, collagen content of the samples was analyzed by a modified Steggman procedure (26) or by reverse phase HPLC using phenylisothiocyanate (PITC) as a precolumn derivatizing agent for the amino acids (27).
RESULTS AND DISCUSSION The immunoaffinity purified pepsin-resistant helical domain (45 kDa chain) of type X was found to contain the contain HP and LP cross-links (Table 1). The type X preparations contained a small portion (less than 2%) of type II collagen (Figure 1). Type II collagen was purified to determine if this might be contributing to the HP and LP found in type X. However, the type II had a completely different HP:LP ratio than the type X, showing that type II is not contributing significantly to the LP fraction of the purified type X preparation. Also, the type II collagen had 0.43 Moles of HP and LP/Mole of collagen while type X contained 0.81 Moles of HP and LP/Mole of Collagen X (based on 180 kDa M.W. for collagen X). Other collagens including types IX and XI 302
JOBNAME: BBRC 218#3 PAGE: 3 SESS: 17 OUTPUT: Thu Apr 11 23:46:00 1996 /xypage/worksmart/tsp000/68480f/7
Vol. 219, No. 2, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS TABLE 1 Collagen Cross-Linking Concentration for the Hypertrophic Cartilage from Days 19 and 20 Chick Embryos and Type II and X Isolated from That Source1 Collagen source Type X2 Type II3 Hyp. Cart.4
HP
LP
(nmoles/mg Collagen) 1.76 1.29 0.15
2.74 0.14 0.45
HP:LP ratio 0.64 9.21 0.33
1 The type II and X cross-link concentrations are not quantitative but qualitative. Both were only portions of the total collagen pool, leading to the higher concentrations relative to the whole tissue. 2 Average of three different preparations. 3 A representative value for type II from pepsinized collagen that had been run over an immunoaffinity column for type X twice to remove most of the type X. It was then partially purified by sale precipitation. d Average of three different hypertrophic cartilage comes from Day 20 embyronic chick tibias.
were not detected in the purified type X collagen preparations. Even if 10% of another collagen was found in the preparation it could not be the sole source of the HP and LP because the cross-link concentration would be out of the physiological range (highest found is about 3–4 Moles/Mole of collagen in nucleus pulposus; personal observations). Furthermore, the 45 kDa band was isolated from an 8% SDS-page gel and its HP:LP ratio (0.67) is consistent with the ratio presented in Table 1. Thus, we are confident that type X and not some contaminating collagen contains the cross-links. Nonreducible collagen cross-links are typically formed between the telopeptide region of one collagen molecule where specific lysine residues have been converted to peptidyl-a-aminoadipicd-semialdehydes by lysyl oxidase which react with the triple helical region of an adjacent collagen, such as the linkage between the COL2 portion of type IX and the N-telopeptide portion of type II (28,29). Further work is being undertaken to determine which collagen type provides the telopeptide region for the nonreducible cross-link with type X. Based on ultrastructural studies, type X or type II are the most likely candidates. The formation of covalent lysine-derived cross-links in collagen are important for the stability of the collagen fibrils. However, the physiological significance for the presence of relatively high quantities of LP relative to HP is currently not known. LP is typically found in bone and dentin
FIG. 1. An 8% SDS–PAGE gel showing the purity of a representative type X collagen preparation. Lane A, 45-kDa standard of type X from cultured chick chondroctyes (Small arrow). Lane B, Purified type X from hypertrophic cartilage. Lane C, Partially purified type II collagen from hypertrophic cartilage (Large arrow). 303
JOBNAME: BBRC 218#3 PAGE: 4 SESS: 17 OUTPUT: Thu Apr 11 23:46:00 1996 /xypage/worksmart/tsp000/68480f/7
Vol. 219, No. 2, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
(30). Recently, LP has also been shown to be a major cross-link in mineralized turkey tendon (31). The formation of LP in turkey tendon coincides with an overall decrease in the hydroxylation of lysine residues in type I collagen and LP does not alter the amino acid composition of the major fluorescent peptides, suggesting that the presence of LP in mineralizing tendon is dependent on the degree of lysine hydroxylation. Also, an HP:LP ratio of about 1:1 is found in the articular cartilage of humans with Ehlers-Danlos Syndrome VI due to a decrease in lysyl hydroxylase activity (30). Since type X is such a transient molecule, a shift in the degree of hydroxylysine formation is not likely involved in the formation of LP. Moreover, type X has a higher ratio of hydroxylysine:lysine than type I. The fact that it is involved primarily in the formation of LP could support its function in the mineralization process of hypertrophic cartilage, whether it is direct or indirect. Collagen cross-links have been shown to impede collagen degradation (29). The fact that type X is cross-linked into the matrix is quite intriguing since upon hatching the hypertrophic cone of cartilage in the tibiotarsus is quickly vascularized and resorbed, the exact opposite of what you would expect to happen in a relatively highly cross-linked collagen network. One possible explanation for this is that collagen cross-linking involving minor collagens provides structural adaptations to optimize the function of the fibril network rather than just providing mechanical stability. As an example, type V collagen is cross-linked into type I collagen fibrils (32). The fibrillar properties of these heterotypic fibrils are likely required for corneal transparency (33). In the case of type X, it could be cross-linking to the surface of type II fibrils leading fibrillar property changes that either aid in the eventual degradation of the matrix or allow mineralization to occur. Further work is needed to test this concept. Alternatively, indirect evidence has suggested type X collagen might provide mechanical stability for the hypertrophic zone (34). Thus, type X cross-linking to itself in the pericellular matrix could provide mechanical stability for the hypertrophic chondrocyte in an area where many molecules are being degraded. Since type X has been found in osteoarthritic cartilage (35,36), understanding how type X cross-links into the collagen network could help delineate its impact in arthritis. ACKNOWLEDGMENTS The authors thank Dan Pietryla for his excellent technical assistance with the HPLC analyses. The research was supported by NIH Grants AR39239 and HD 23681 and an Arthritis Foundation Illinois chapter grant (M.W.O).
REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19.
Mendler, M., Eich-Bender, S. G., Vaughan, L., Winterhalter, K. H., and Bruckner, P. (1989) J. Cell. Biol. 108, 191–198. Schmid, T. M., and Conrad, H. E. (1982) J. Biol. Chem. 257, 12444–12450. Schmid, T. M., and Conrad, H. E. (1982) J. Biol. Chem. 257, 12451–12457. Schmid, T. M., and Linsenmeyer, T. F. (1985) J. Cell. Biol. 100, 598–605. Arias, J. L., Fernandez, M. S., Dennis, J. E., and Caplan, A. I. (1991) Conn. Tiss. Res. 23, 37–45. Jacenko, O., LuValle, P. A., and Olsen, B. R. (1993) Nature 365, 56–61. Rosati, R., Horan, G. S., Pinero, G. J., Garofalo, S., Keene, D. R., Horton, W. A., Vuorio, E., de Crombrugghe, B., and Behringer, R. R. (1994) Nature Genetics 8, 129–135. Schmid, T. M., Maynes, R. M., Jeffrey, J. J., and Linsenmeyer, T. M. (1986) J. Biol. Chem. 261, 4184–4189. Welgus, H. G., Fliszer, C. J., Seltzer, J. L., Schmid, T. M., and Jeffrey, J. J. (1990) J. Biol. Chem. 265, 13521–13527. Kwan, A. P. L., Freemont, A. J., and Grant, M. E. (1986) Biosci. Rep. 6, 155–162. Gerstenfeld, L. C., and Lian, L. B. (1988) Trans. Orth. Res. Soc. 13, 290. Morris, J. B., and Balian, G. (1985) Ann. N.Y. Acad. Sci. 460, 475–477. Wu, L., Genge, B. R., and Wuthier, R. E. (1991) J. Biol. Chem. 266, 1195–1203. Habuchi, H., Conrad, H. E., and Glaser, J. H. (1985) J. Biol. Chem. 260, 13029– 13034. Poole, A. R., and Pidoux, I. (1989) J. Cell. Biol. 109, 2547–2554. Lian, J. B., McKee, M. D., Todd, A. M., and Gerstenfeld, L. C. (1993) J. Cell. Biochem. 52, 206–219. Kwan, A. P. L., Cummings, C. E., Chapman, J. A., and Grant, M. E. (1991) J. Cell. Biol. 114, 597–604. Schmid, T. M., and Linsenmeyer, T. F. (1990) Dev. Biol. 138, 53–62. Chen, Q., Gibney, E. P., Fitch, J. M., Linsenmeyer, C., Schmid, T. M., and Linsenmeyer, T. F. (1990) PNAS 87, 8046–8050. 304
JOBNAME: BBRC 218#3 PAGE: 5 SESS: 17 OUTPUT: Thu Apr 11 23:46:00 1996 /xypage/worksmart/tsp000/68480f/7
Vol. 219, No. 2, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33.
Reginato, A. M., and Jimenez, S. A. (1991) Biochem. J. 273, 333–338. Schmid, T. M., and Linsenmeyer, T. F. (1983) J. Biol. Chem. 258, 9504–9509. Chen, Q., Fitch, J. M., Linsenmeyer, C., and Linsenmeyer, T. F. (1992) Bone Miner. 17, 223–227. Orth, M. W., Martinez, D. A., Cook, M. E., and Vailas, A. C. (1993) Biochim. Biophys. Acta 1157, 229–232. Bonen, D. K., and Schmid, T. M. (1991) J. Cell Biol. 116, 1171–1178. Uebelhart, D., Thonar, E. J.-M. A., Pietryla, D. W., and Williams, J. M. (1993) Osteoarthritis Cart. 1, 185–192. Schwartz, D. E., Choi, Y., Sandell, L. J., and Hanson, W. R. (1985) Histochem. J. 17, 655–663. Orth, M. W., Martinez, D. A., Cook, M. E., and Vailas, A. C. (1991) Biochem. Biophys. Res. Comm. 179, 1582–1586. Nagan, N., and Kagan, H. M. (1994) J. Biol. Chem. 269, 22366–22371. Wu, J.-J., Woods, P. E., and Eyre, D. R. (1992) J. Biol. Chem. 267, 23007–23014. Eyre, D. R., Paz, M. A., and Gallop, P. M. (1984) Annual Review of Biochemistry 53, 717–748. Yamauchi, M., and Katz, E. P. (1993) Conn. Tiss. Res. 29, 81–98. Niyibizi, C., and Eyre, D. R. (1994) Eur. J. Biochem. 224, 943–950. Linsenmayer, T. F., Gibney, E., Igoe, F., Gordon, M. K., Fitch, J. M., Fessler, L. I., and Birk, D. E. (1993) J. Cell. Biol. 121, 1181–1189. 34. Warman, M. L., Abbott, M., Apte, S. S., Hefferon, T., McIntosh, I., Cohn, D. H., Hecht, J. T., Olsen, B. R., and Francomano, C. A. (1993) Nature Genetics 5, 79–82. 35. Aigner, T., Reichenberger, E., Bertling, W., Kirsch, T., Stöß, H., and von der Mark, K. (1993) Virchow Archiv B Cell Pathol. 63, 205–211. 36. Hoyland, J. A., Thomas, J. T., Donn, R., Marriott, A., Ayad, S., Boot-Handford, R. P., Grant, M. E., and Freemont, A. J. (1991) Bone Miner. 15, 151–164.
305
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS ARTICLE NO.
219, 301–305 (1996)
0227
Type X Collagen Isolated from the Hypertrophic Cartilage of Embryonic Chick Tibiae Contains Both Hydroxylysyl- and Lysylpyridinoline Cross-Links M. W. Orth,1 L. J. Luchene, and T. M. Schmid Department of Biochemistry, Rush Medical College, Rush Presbyterian-St. Luke’s Medical Center, Chicago, IL 60612 Received December 27, 1995 Hypertrophic cartilage from the tibiotarsus of Day 20 chick embryonic tibiae was found to contain an unusually high concentration of lysylpyridinoline (LP), a nonreducible collagen cross-link normally found only in bone and dentin but not in cartilage. Since type X collagen is abundant in this cartilage, research was conducted to see if type X was the primary source of LP. The 45-kDa pepsin-resistant form of type X was purified by immunoaffinity chromatography. It contained a high concentration of the LP cross-link while type II contained primarily hydroxylysylpyridinoline (HP), the predominant cross-link in cartilage. This, to our knowledge is the first time that type X has been shown directly to form nonreducible cross-links and that a collagen other than type I has a high level of LP. Also, it is interesting that the HP and LP cross-links are found in a collagen that is degraded so rapidly. Possible explanations for these findings are discussed. © 1996 Academic Press, Inc.
During endochondral bone formation, chondrocytes in the proliferative zone of the growth plate synthesize primarily types II, IX, and XI collagen which form fibrils in the extracellular matrix (1). However, as the cells mature and hypertrophy, they decrease their expression of these collagens and begin to express type X collagen (2,3). Type X collagen, a non-fibril forming collagen consisting of three identical 59 kDa chains, is almost exclusively located in hypertrophic cartilage (4), although it has been found in the eggshell membrane (5). Even though its function is not well understood, some have shown its importance by using mice in which formation of the collagen X is perturbed by the insertion of a truncated avian a1(X) gene (6). The mice develop skeletal deformities including accentuated neck lordosis and thoracolumbar kyphosis. Histological analyses of the growth plates in the hind limb of the transgenic mice reveal a reduced thickness in the hypertrophic zone as well as an altered morphology of the hypertrophic chondrocytes (6). However, others have shown normal long bone growth in type X null mice suggesting that only abnormal type X deposition into the matrix will adversely effect bone growth (7). Type X collagen has two cleavage sites for vertebrate collagenase while collagens I, II, and III have only one (8). In solution assays it is degraded faster than type II collagen and in its native state can also be degraded by 72-kDa type IV collagenase (9). Thus, some have thought that it might facilitate matrix turnover and subsequent vascularization by metaphyseal arteries (8). Collagen X has also been thought to be involved in the calcification of the cartilage. It is synthesized by chondrocytes just prior to matrix calcification and persists in regions of calcified cartilage (8,10). Both in vivo and in vitro studies have shown apparently parallel increases in type X and alkaline phosphatase production preceding calcification (11,12). Some investigators have observed an association of type X with matrix vesicles (13,14) while others have not (15). Recently, others have shown that osteocalcin and osteopontin, two proteins associated with the mineralization of bone, are produced by hypertrophic chondrocytes in vitro and found at sites of mineralization (16). Thus, type X might not be directly involved in the calcification of cartilage, but instead provide a permissive environment for it to occur by altering the extracellular matrix. 1
Corresponding author Fax 312-942-3053. 301 0006-291X/96 $18.00 Copyright © 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.
JOBNAME: BBRC 218#3 PAGE: 2 SESS: 15 OUTPUT: Thu Apr 11 23:46:00 1996 /xypage/worksmart/tsp000/68480f/7
Vol. 219, No. 2, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
In the extracellular matrix type X is found in primarily two forms. One is as fine filaments (<5 nm in diameters) within mats which are in the pericellular matrix around hypertrophic chondrocytes (17,18). The second is associated with the surface of fibrils (between 10–20 nm in diameter) composed primarily of type II collagen (15,18). In an avian embryonic chick sternal cartilage culture system, type X is able to diffuse through the matrix and become associated to type II collagen fibrils (19). This association apparently involves lysine-derived covalent cross-links since b-aminoproprionitrile, a lysyl oxidase inhibitor, greatly increases the extractability of type X from cartilage and chondrocyte cultures (20,21). Also, in ovo injections of b-aminoproprionitrile inhibited the covalent interactions of type X with the extracellular matrix of hypertrophic cartilage from the tibiotarsi of embryonic chicks (22). Recently, it was shown that the hypertrophic cartilage from the tibiotarsi of newly hatched broiler chicks contained both of the nonreducible lysine-derived collagen cross-links, hydroxylysylpyridinoline (HP) and lysylpyridinoline (LP), with almost four times the concentration of LP as HP (23). Normally cartilage contains almost exclusively HP. However, one difference between hypertrophic cartilage from the chick and other cartilages is the presence of appreciable levels of collagen X. In avian dyschondroplastic cartilage, the amount of type X collagen is directly proportional to the concentration of LP in the cartilage (Unpublished observations). Thus, the purpose of this research was to determine if type X collagen was involved in the formation of HP and more interestingly LP in avian hypertrophic cartilage. MATERIALS AND METHODS Type X collagen purification. Between 200–600 mg wet weight of cartilage from the proximal tibiotarsus was isolated from the tibiotarsi of day 19 and 20 White Leghorn chick embryos. The cartilage was extracted with 4 M Guanidine, 50 mM Sodium acetate, pH 6.0 containing several protease inhibitors to remove the proteoglycans. After extensive rinsing with water, the cartilage was digested with 1 mg/ml pepsin in 0.5 M acetic acid and 0.2 M NaCl for two days at 4°C. The pepsinized sample was neutralized with NaOH and centrifuged at 10,000 rpm for 30 minutes at 4°C. The supernatant was then dialyzed against 1 mg/ml 3-[(3-cholamidopropyl)-dimethylammonio]-1-propanesulfonate (CHAPS), 50 mM Tris, 5 mM EDTA, and 1 M NaCl at pH 7.5. Type X was purified from the dialyzate using an immunoaffinity column consisting of the XAC-9 anti-type X antibody bound to sepharose beads. The specificity of this monoclonal antibody for type X has been previously demonstrated (4,24). Type X was eluted from the sepharose column with 2 M Guanidine, 50 mM Tris, pH 7.5 and then dialyzed either with 6.25 mM Tris, 5 mM EDTA for SDS-polyacrylamide gel electrophoresis (SDS-PAGE) or in 0.1 M acetic acid for storage and biochemical analysis. Partial purification of Type II Collagen. Pepsinized collagen (primarily types II, IX, and XI) which did not bind the type X immunoaffinity column was dialyzed against 0.5 M acetic acid and then 0.7M NaCl was added to precipitate out the type II collagen. The type II collagen was dialyzed against 0.1M acetic acid and stored at −20°C. Gel electrophoresis. 8% SDS-PAGE gels were performed to determine the purity of the type X preparations. A 45 kDa pepsinized standard of type X as well type II were included. The gels were stained with Coomassie Blue and scanned with a laser densitometer (LKB, Sweden) to measure the purity of the preparation. Collagen cross-link (HP and LP) analysis. Collagen samples were hydrolyzed and partially purified using CF1-cellulose chromatography as previously described (23). The samples were run on a C18 reverse phase column using high performance liquid chromatography (HPLC) according to the protocol of Uebelhart et al. (25). In order to determine the cross-link concentrations, collagen content of the samples was analyzed by a modified Steggman procedure (26) or by reverse phase HPLC using phenylisothiocyanate (PITC) as a precolumn derivatizing agent for the amino acids (27).
RESULTS AND DISCUSSION The immunoaffinity purified pepsin-resistant helical domain (45 kDa chain) of type X was found to contain the contain HP and LP cross-links (Table 1). The type X preparations contained a small portion (less than 2%) of type II collagen (Figure 1). Type II collagen was purified to determine if this might be contributing to the HP and LP found in type X. However, the type II had a completely different HP:LP ratio than the type X, showing that type II is not contributing significantly to the LP fraction of the purified type X preparation. Also, the type II collagen had 0.43 Moles of HP and LP/Mole of collagen while type X contained 0.81 Moles of HP and LP/Mole of Collagen X (based on 180 kDa M.W. for collagen X). Other collagens including types IX and XI 302
JOBNAME: BBRC 218#3 PAGE: 3 SESS: 17 OUTPUT: Thu Apr 11 23:46:00 1996 /xypage/worksmart/tsp000/68480f/7
Vol. 219, No. 2, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS TABLE 1 Collagen Cross-Linking Concentration for the Hypertrophic Cartilage from Days 19 and 20 Chick Embryos and Type II and X Isolated from That Source1 Collagen source Type X2 Type II3 Hyp. Cart.4
HP
LP
(nmoles/mg Collagen) 1.76 1.29 0.15
2.74 0.14 0.45
HP:LP ratio 0.64 9.21 0.33
1 The type II and X cross-link concentrations are not quantitative but qualitative. Both were only portions of the total collagen pool, leading to the higher concentrations relative to the whole tissue. 2 Average of three different preparations. 3 A representative value for type II from pepsinized collagen that had been run over an immunoaffinity column for type X twice to remove most of the type X. It was then partially purified by sale precipitation. d Average of three different hypertrophic cartilage comes from Day 20 embyronic chick tibias.
were not detected in the purified type X collagen preparations. Even if 10% of another collagen was found in the preparation it could not be the sole source of the HP and LP because the cross-link concentration would be out of the physiological range (highest found is about 3–4 Moles/Mole of collagen in nucleus pulposus; personal observations). Furthermore, the 45 kDa band was isolated from an 8% SDS-page gel and its HP:LP ratio (0.67) is consistent with the ratio presented in Table 1. Thus, we are confident that type X and not some contaminating collagen contains the cross-links. Nonreducible collagen cross-links are typically formed between the telopeptide region of one collagen molecule where specific lysine residues have been converted to peptidyl-a-aminoadipicd-semialdehydes by lysyl oxidase which react with the triple helical region of an adjacent collagen, such as the linkage between the COL2 portion of type IX and the N-telopeptide portion of type II (28,29). Further work is being undertaken to determine which collagen type provides the telopeptide region for the nonreducible cross-link with type X. Based on ultrastructural studies, type X or type II are the most likely candidates. The formation of covalent lysine-derived cross-links in collagen are important for the stability of the collagen fibrils. However, the physiological significance for the presence of relatively high quantities of LP relative to HP is currently not known. LP is typically found in bone and dentin
FIG. 1. An 8% SDS–PAGE gel showing the purity of a representative type X collagen preparation. Lane A, 45-kDa standard of type X from cultured chick chondroctyes (Small arrow). Lane B, Purified type X from hypertrophic cartilage. Lane C, Partially purified type II collagen from hypertrophic cartilage (Large arrow). 303
JOBNAME: BBRC 218#3 PAGE: 4 SESS: 17 OUTPUT: Thu Apr 11 23:46:00 1996 /xypage/worksmart/tsp000/68480f/7
Vol. 219, No. 2, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
(30). Recently, LP has also been shown to be a major cross-link in mineralized turkey tendon (31). The formation of LP in turkey tendon coincides with an overall decrease in the hydroxylation of lysine residues in type I collagen and LP does not alter the amino acid composition of the major fluorescent peptides, suggesting that the presence of LP in mineralizing tendon is dependent on the degree of lysine hydroxylation. Also, an HP:LP ratio of about 1:1 is found in the articular cartilage of humans with Ehlers-Danlos Syndrome VI due to a decrease in lysyl hydroxylase activity (30). Since type X is such a transient molecule, a shift in the degree of hydroxylysine formation is not likely involved in the formation of LP. Moreover, type X has a higher ratio of hydroxylysine:lysine than type I. The fact that it is involved primarily in the formation of LP could support its function in the mineralization process of hypertrophic cartilage, whether it is direct or indirect. Collagen cross-links have been shown to impede collagen degradation (29). The fact that type X is cross-linked into the matrix is quite intriguing since upon hatching the hypertrophic cone of cartilage in the tibiotarsus is quickly vascularized and resorbed, the exact opposite of what you would expect to happen in a relatively highly cross-linked collagen network. One possible explanation for this is that collagen cross-linking involving minor collagens provides structural adaptations to optimize the function of the fibril network rather than just providing mechanical stability. As an example, type V collagen is cross-linked into type I collagen fibrils (32). The fibrillar properties of these heterotypic fibrils are likely required for corneal transparency (33). In the case of type X, it could be cross-linking to the surface of type II fibrils leading fibrillar property changes that either aid in the eventual degradation of the matrix or allow mineralization to occur. Further work is needed to test this concept. Alternatively, indirect evidence has suggested type X collagen might provide mechanical stability for the hypertrophic zone (34). Thus, type X cross-linking to itself in the pericellular matrix could provide mechanical stability for the hypertrophic chondrocyte in an area where many molecules are being degraded. Since type X has been found in osteoarthritic cartilage (35,36), understanding how type X cross-links into the collagen network could help delineate its impact in arthritis. ACKNOWLEDGMENTS The authors thank Dan Pietryla for his excellent technical assistance with the HPLC analyses. The research was supported by NIH Grants AR39239 and HD 23681 and an Arthritis Foundation Illinois chapter grant (M.W.O).
REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19.
Mendler, M., Eich-Bender, S. G., Vaughan, L., Winterhalter, K. H., and Bruckner, P. (1989) J. Cell. Biol. 108, 191–198. Schmid, T. M., and Conrad, H. E. (1982) J. Biol. Chem. 257, 12444–12450. Schmid, T. M., and Conrad, H. E. (1982) J. Biol. Chem. 257, 12451–12457. Schmid, T. M., and Linsenmeyer, T. F. (1985) J. Cell. Biol. 100, 598–605. Arias, J. L., Fernandez, M. S., Dennis, J. E., and Caplan, A. I. (1991) Conn. Tiss. Res. 23, 37–45. Jacenko, O., LuValle, P. A., and Olsen, B. R. (1993) Nature 365, 56–61. Rosati, R., Horan, G. S., Pinero, G. J., Garofalo, S., Keene, D. R., Horton, W. A., Vuorio, E., de Crombrugghe, B., and Behringer, R. R. (1994) Nature Genetics 8, 129–135. Schmid, T. M., Maynes, R. M., Jeffrey, J. J., and Linsenmeyer, T. M. (1986) J. Biol. Chem. 261, 4184–4189. Welgus, H. G., Fliszer, C. J., Seltzer, J. L., Schmid, T. M., and Jeffrey, J. J. (1990) J. Biol. Chem. 265, 13521–13527. Kwan, A. P. L., Freemont, A. J., and Grant, M. E. (1986) Biosci. Rep. 6, 155–162. Gerstenfeld, L. C., and Lian, L. B. (1988) Trans. Orth. Res. Soc. 13, 290. Morris, J. B., and Balian, G. (1985) Ann. N.Y. Acad. Sci. 460, 475–477. Wu, L., Genge, B. R., and Wuthier, R. E. (1991) J. Biol. Chem. 266, 1195–1203. Habuchi, H., Conrad, H. E., and Glaser, J. H. (1985) J. Biol. Chem. 260, 13029– 13034. Poole, A. R., and Pidoux, I. (1989) J. Cell. Biol. 109, 2547–2554. Lian, J. B., McKee, M. D., Todd, A. M., and Gerstenfeld, L. C. (1993) J. Cell. Biochem. 52, 206–219. Kwan, A. P. L., Cummings, C. E., Chapman, J. A., and Grant, M. E. (1991) J. Cell. Biol. 114, 597–604. Schmid, T. M., and Linsenmeyer, T. F. (1990) Dev. Biol. 138, 53–62. Chen, Q., Gibney, E. P., Fitch, J. M., Linsenmeyer, C., Schmid, T. M., and Linsenmeyer, T. F. (1990) PNAS 87, 8046–8050. 304
JOBNAME: BBRC 218#3 PAGE: 5 SESS: 17 OUTPUT: Thu Apr 11 23:46:00 1996 /xypage/worksmart/tsp000/68480f/7
Vol. 219, No. 2, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33.
Reginato, A. M., and Jimenez, S. A. (1991) Biochem. J. 273, 333–338. Schmid, T. M., and Linsenmeyer, T. F. (1983) J. Biol. Chem. 258, 9504–9509. Chen, Q., Fitch, J. M., Linsenmeyer, C., and Linsenmeyer, T. F. (1992) Bone Miner. 17, 223–227. Orth, M. W., Martinez, D. A., Cook, M. E., and Vailas, A. C. (1993) Biochim. Biophys. Acta 1157, 229–232. Bonen, D. K., and Schmid, T. M. (1991) J. Cell Biol. 116, 1171–1178. Uebelhart, D., Thonar, E. J.-M. A., Pietryla, D. W., and Williams, J. M. (1993) Osteoarthritis Cart. 1, 185–192. Schwartz, D. E., Choi, Y., Sandell, L. J., and Hanson, W. R. (1985) Histochem. J. 17, 655–663. Orth, M. W., Martinez, D. A., Cook, M. E., and Vailas, A. C. (1991) Biochem. Biophys. Res. Comm. 179, 1582–1586. Nagan, N., and Kagan, H. M. (1994) J. Biol. Chem. 269, 22366–22371. Wu, J.-J., Woods, P. E., and Eyre, D. R. (1992) J. Biol. Chem. 267, 23007–23014. Eyre, D. R., Paz, M. A., and Gallop, P. M. (1984) Annual Review of Biochemistry 53, 717–748. Yamauchi, M., and Katz, E. P. (1993) Conn. Tiss. Res. 29, 81–98. Niyibizi, C., and Eyre, D. R. (1994) Eur. J. Biochem. 224, 943–950. Linsenmayer, T. F., Gibney, E., Igoe, F., Gordon, M. K., Fitch, J. M., Fessler, L. I., and Birk, D. E. (1993) J. Cell. Biol. 121, 1181–1189. 34. Warman, M. L., Abbott, M., Apte, S. S., Hefferon, T., McIntosh, I., Cohn, D. H., Hecht, J. T., Olsen, B. R., and Francomano, C. A. (1993) Nature Genetics 5, 79–82. 35. Aigner, T., Reichenberger, E., Bertling, W., Kirsch, T., Stöß, H., and von der Mark, K. (1993) Virchow Archiv B Cell Pathol. 63, 205–211. 36. Hoyland, J. A., Thomas, J. T., Donn, R., Marriott, A., Ayad, S., Boot-Handford, R. P., Grant, M. E., and Freemont, A. J. (1991) Bone Miner. 15, 151–164.
305