- Email: [email protected]
Production of collagenase by human osteoblasts and osteoclasts in vivo
Bone Vol. 19, No. 1 July 19963540
ELSEVIER
Production Osteoclasts S. BORD,’
of Collagenase In Vivo
A. HORNER,’ R. M. HEMBRY,*
by Human Osteoblasts
J. J. REYNOLDS,”
and J. E. COMPSTON’
’Department of Medicine, University of Cambridge, School of Clinical Medicine, Addenbrooke’s Hospital, Cambridge, ’Strangeways Research Laboratory, Warts Causeway, Cambridge, UK ’Department of Orthodontics and Paediatric Dentistry, Guy’s Tower, Guy’s Hospital, London, UK
Osteoblasts;
Materials and Methods Tissue
Os-
Heterotopic bone was surgically removed from three patients (mean age 29 years) who had previously sustained severe head injury. Osteophytes were removed from three patients (mean age 68 years) during routine joint replacement surgery. Samples approximately 1 cm3 were snap frozen in precooled n-hexane (Sigma) and maintained at -70°C or fixed in 70% alcohol for a minimum of 48 h and dehydrated in progressive concentrations of alcohol (70%-100%). The latter samples were then transferred to LR White medium resin (TAAB), left for 4-7 days on a roller to allow the resin to infiltrate the tissue and then placed in glass vials in fresh resin and cured in an oven at 60°C.
Introduction Matrix metalloproteinases (MMPs) are a family of zinc-dependent enzymes responsible for the extracellular degradation of connective tissue. The best known member of the MMP family is collagenase. Specific collagenases cleave interstitial collagen types I, II, and III into three-fourths and one-fourth fragments. The enzyme is secreted in a latent proform that requires activation, a process that may involve plasmin and stromelysin.23 The denatured collagen may then be further degraded by another MMP, gelatinase. Degradation of matrix occurs under both physiological and
Hisrology
Undecalcified 8 pm sections from LR White embedded tissue were stained with 1% toluidine blue (pH 4.2), Diff-Quik (Baxter Dade AG), or Von Kossa with a Van Giesen counterstain and examined by bright field microscopy. Alkaline phosphatase and tartrate resistant acid phosphatase (TRAP) staining were demon-
Address for correspondence and reprints: Dr. J. E. Compston, Department of Medicine, Level 5, Addenbrooke’s Hospital, Cambridge CB2
2QQ, UK. 0 1996 by Else&r Science Inc. All rights reserved.
UK
pathological conditions in many tissues throughout the body. In the musculoskeletal system, MMPs are known to be involved in endochondral bone formation6 and osteoclastic degradation of bone matrix,‘0.‘2,29 but their role in bone remodeling has not been firmly established. Animal studies have shown that osteoblasts secrete collagenase in response to a number of stimulators of bone resorption, including 1,25 dihydroxyvitamin D, [1,25 (OH), D3] and interleukin-1 (IL-1).‘5,22 It has been suggested that the enzyme may degrade the nonmineralized collagenous membrane which covers mineralized bone, prior to osteoclastic resorption.2 Collagenase has also been identified in rodent osteoclasts’s and in the subosteoclastic compartment of bone undergoing resorption.’ The activity of all MMPs is tightly regulated by their specific inhibitors, TIMPs (tissue inhibitor of metalloproteinases). We have recently reported that human osteoblasts in vitro synthesise TIMP-1 and can be induced with bone resorbing hormones to produce collagenase.*’ However, the presence of MMPs in human bone in vivo has not yet been demonstrated. In this study we have investigated the expression of collagenase in human bone using heterotopic and osteophytic bone as models of high turnover.
Studies in some animal species have demonstrated the production of metalloproteinases by hone cells, suggesting that they may play a role in hone modeling and remodeling. The aim of the present study was to investigate the expression of collagenase in human hone in situ, using heterotopic and osteophytic hone. Immunohistochemistry was performed on chilled sections of hone, using well characterized polyclonal antibodies to human collagenase. The heterotopic and osteophytic hone exhibited high turnover and both bone modeling and remodeling were evident. Collagenase expression by osteoblasts was demonstrated in cells synthesising matrix and in lining cells; the strongest signal was seen in areas of de novo matrix formation, where bridges of woven bone were being formed between areas of mineralized bone. Collagenase was also present in some osteoclasts associated with eroded bone surfaces and in some mononuclear cells that were present in resorption cavities and in the bone marrow. Our results provide the first demonstration, in situ, of collagenase in human bone and suggest that it may play a role in human bone modeling and remodeling. Production of collagenase by active osteoblasts and lining cells suggests that it may be involved both in matrix formation and activation of hone remodeling. The presence of collagenase in osteoclasts provides further evidence that metalloproteinases may play a role in bone resorption. (Bone I9:35-40; 2996) Key Words: Collagenase; Metalloproteinases; teoclasts; Bone remodeling.
and
35
8756.3282/96/$15.00 PI1 S8756-3282(96)00106-8
36
S. Bord et al. Expression of collagenase
strated histologically tions (Sigma).
Bone Vol. 19, No. 1 July 1996:3540
in human bone in viva
according
to the manufacturer’s
instruc-
Antisera Human interstitial collagenase was purified according to Whitham et al.,” injected into sheep and the resulting antiserum characterized by Western blotting, inhibition and immunolocalization as described previously for the antibody to rabbit collagenase. I7 Specificity was also confirmed by positive staining on immunolocalization of NSO mouse myeloma cells transfected with human collagenase,23 but showing no staining of cells transfected with either human stromelysin-lz5 and -223 or gelatinase A2h or gelatinase B.” This antiserum has been used successfully to immunolocalize collagenase in human synovium.‘6 Immunoglobulins were prepared by triple ammonium sulphate precipitation. Antibodies were also raised in rabbits immunized with collagenase purified from the conditioned medium of W138 foetal lung fibroblasts; immunoglobulins were purified from this antiserum by protein A-sepharose chromatography and monospecificity tested by Ouchterlony immunoelectrophoresis and immunoblotting. Both antibodies recognize pro and active collagenase as well as collagenase complexed to TIMP. Neither antiserum cross-reacts with human neutrophil collagenase. Pooled normal sheep IgG and normal rabbit IgG were used appropriately as negative controls at the corresponding immunoglobulin concentrations. Immunohistochemistry 8 pm sections were cut on a Bright cryostat (Huntingdon, England) with a cabinet temperature of -30°C and equipped with a slow drive high torque motor and automatic speed control, using a highly polished tungsten carbide knife. Sections were picked off onto glass slides coated with 2% APES (3 - aminoprophyl triethoxy silane), air-dried for 10 min, fixed in 4% paraformaldehyde for 30 min at room temperature, and washed in phosphate buffered saline (PBS). Sections were incubated in 10% normal serum (from the species in which the second antibody was made) prior to application of the primary antibody (10 kg/mL), which was left on overnight at 4°C in a humid chamber. After washing and a further step to block endogenous peroxidase enzymes (20 min) using ImmunoPure (Pierce), the sections were washed again and treated with a secondary biotinylated antibody (5 kg/mL, Vectalabs, 40 min at room temperature). The reaction was completed using an ABC reagent (avidin/biotin complex, Vectalabs, 40 min, room temperature). 3-amino-9-ethylcarbazole (AEC), 3,3’-diaminobenzidine (DAB), or 4- chloro-1-naphthol (CN) substrate was used to detect sites of antibody binding. Some sections were counterstained with haematoxylin while others were treated with propidium iodide (0.1 kg/mL for 4 min) to detect nuclei. Finally, the slides were cover-slipped with Apathy’s mounting medium (BDH) and observed under bright field microscopy on an Olympus BH-2 photomicroscope. Photographs were taken on Ektachrome 64 film. Results Histologic Features of Heterotopic
those of subchondral ossification with nonhypertrophied chondrocytes in their matrix surrounding bone. The bone exhibited high turnover, with multilayers of cuboidal osteoblasts on plump osteoid seams and many areas in which multinucleated osteoclasts were seen in association with resorption cavities (Figure 1). Both bone remodeling and modeling were evident and examination of bone using polarized light revealed a combination of woven and lamellar bone, the former predominating. Within the woven bone, “bridges” of matrix linking areas of mineralized bone were observed (Figures 2, 3A, 3B). Staining with Von Kossa (Figure 2) and birefringence on toluidine blue stained sections viewed under polarized light microscopy confirmed that these bridges consisted of osteoid and sections above and below showed no overlying or underlying mineral. Haematoxylin counterstained sections showed the highly cellular areas within these bridges (Figure 3A) as did sections treated with propidium iodide which fluorescently labels nucleic acids. Their osteoblastic nature was confirmed by positive alkaline phosphatase staining. In some areas many blood vessels were present, particularly in fibrous tissue. The osteophytes demonstrated similar histological features as described in detail by Dodds et al.” Collagenase Both antibodies revealed very similar cell-associated staining for collagenase, with similar results observed in all samples of heterotopic and osteophytic bone. Multilayers of plump cuboidal osteoblasts within newly formed matrix in previously resorbed lacunae showed strong cytoplasmic staining in approximately half the cells (Figures 4 and 6C), with a stronger signal in the lining cells (Figure 4). In addition, osteoblasts associated with de novo bone formation exhibited intense collagenase staining, which was also observed both in the cells and to a lesser extent in the matrix of the bridges in the woven bone (Figure 3A). The intensity of collagenase staining in osteoclasts varied, some staining strongly (Figure 5) and others showing intermediate (Figures 6A, 6B) or low staining. In many osteoclasts juxtanuclear staining was apparent, probably in the Golgi apparatus. Positive tar&ate-resistant acid phosphatase (TRAP) expression by these cells in serial sections confirmed their osteoclastic nature. Mononuclear cells in the resorption cavities and in the marrow also showed collagenase positivity as did some of the osteocytes adjacent to the resorbed edges. There was negligible staining in all control sections. Alkuline Phosphatase Positive staining for alkaline phosphatase was most marked in areas of new bone formation. Large cuboidal osteoblasts associated with thick osteoid seams showed intense staining. Cartilage and fibrous tissue adjoining the bone showed high activity along their interfaces. TRAP TRAP, demonstrated histochemically, was particularly evident in the upper part of the hypertrophic zone. Intense staining appeared in the large spherical chondroblasts of the calcifying cartilage adjacent to the mineralizing edge. Within the bone tissue TRAP was confined to the large multinucleated and mononuclear cells associated with eroded surfaces.
and Osteophytic Bone
The heterotopic bone contained fibrous tissue, cartilage, and bone. Evidence of both intramembranous and endochondral ossification was seen, although in some areas the appearances were
Discussion Our results demonstrate, for the first time, the presence of collagenase in situ in human bone. These observations are consistent
Bone Vol. 19, No. 1 July 1996:35-40
Expression
of collagenase
S. Bord et al. in human hone in vivo
37
Figure 1. Heterotopic bone formation from patient 1. Undccalcified sections from LR White embedded tissue stained with Diff-Quik demonstrating high cellular activity. Key: b = bone; ob = osteoblasts; oc = osteoclasts; o = osteocytes. Bar = 80 km. Figure 2. Von Kossa stained undecalcified section of heterotopic woven bone from patient 1. The mineralized matrix stains black and the nonmineralized osteoid pink. “Bridges” of osteoid link one area of mineral with another (arrow). Bar = 40 pm. Figure 3. Indirect immunoperoxidase localization of collagenase in the de novo “bridge” formation in woven bone (arrow) of heterotopic bone from patient 1. The substrate, 4-chloronaphthol gives a black color reaction at sites of antigen binding. Bar = 20 pm. (A) Rabbit anticollagenase polyclonal antibody showing strong collagenase staining in the highly cellular “bridges” demonstrated by haematoxylin counter stain. (B) Normal rabbit serum shows absence of reaction (no counter stain).
38
S. Bord et al. Expression of collagenase
in human bone in viva
Bone Vol. 19, No. July 1996:3540
1
Figure 4. Indirect immunoperoxidase localization of collagenase in heterotopic bone from patient 3 using a sheep polyclonal antibody visualized with AEC as substrate shows positive staining of osteoblasts (ob) in a previously resorbed lacuna; note the distinct cytoplasmic staining (j) close to the nucleus (n). New matrix synthesis extends from the cement line (arrow heads). Note the intense signal of the lining cells (*). Bar = 20 km. Figure 5. Heterotopic bone from patient 3 showing immunolocalization of collagenase in osteoclasts (oc) within a resorption lacuna, and in osteocytes (0) in the bone matrix (b) using a rabbit antihuman polyclonal antibody with indirect immunoperoxidase. The positive cells are identified by the red color reaction of the AEC substrate, with the nucleus (n) and the juxtanuclear cytoplasmic staining (j) identified. The section was viewed under polarized light and shows random distribution of collagen fibrils in woven bone. Bar = 20 km. Figure 6. Immunolocalization in sections from an osteophyte from patient A, using sheep antihuman collagenase polyclonal antibody, visualized with the brown color reaction of DAB substrate. Key: b = bone. Bar = 20 pm. (A) Multinucleated osteoclast (arrow) close to the bone surface shows distinct nuclei (arrowheads). Mononuclear cells (m) within the marrow space also show strong collagenase staining. (B) An osteoclast (arrow) slightly pulled away from the bone surface during processing shows distinct nuclei (arrowheads): staining is evident in the perinuclear and cytoplasmic regions. (C) Osteoblasts (arrows) at sites of bone formation show cell-associated cytoplasmic collagenase staining. The cement line (arrowheads) of a previously resorbed cavity can be seen. (D) Normal sheep serum control shows absence of staining in the osteoblasts (arrows) at sites of formation.
Bone Vol. 19, No. 1 July 1996:3540
with the reported production of the enzyme by osteoblasts and osteoclasts in animals7S’5~‘8~22and in primary cultures of human osteoblasts,2’ and suggest that collagenase is involved in the regulation of bone turnover in man. Its localization in both cell types and at sites of bone resorption and formation (Figures 3A, 4, 5, 6A, 6B, and 6C) indicates that it may perform multiple functions in the process of bone remodeling. The use of heterotopic and osteophytic bone as a model of rapid bone turnover has obvious advantages related to the high levels of cellular activity and expression of cellular and matrix products. The formation of heterotopic bone or osteophytic bone is, however, a pathological event and factors which regulate its modeling and remodeling may differ qualitatively and quantitatively from those which operate under normal circumstances. Osteophytic bone’ ’and bone affected by Paget’s disease” have also been used as models of high bone turnover to study expression of cytokines and growth factors in man; however, presence of collagenase was not investigated in these studies. The production of collagenase by osteoblastic lining cells may be important in the activation of bone remodeling, in which the surface is “prepared’ for resorption by retraction of the lining cells and removal of the nonmineralized collagenous membrane covering the bone surface. Most studies indicate that this material, which is less than 1 km thick and separates mineralized bone from cells on its surface, forms a continuous layer over quiescent bone surfaces’~*” and it is generally believed that removal of this collagenous layer is required before resorption can take place, although resorption of nonmineralized matrix by osteoclasts may also occur.*’ The lining cells are ideally placed to initiate activation, being in communication with the strainsensitive osteocytic-canalicular network and in close approximation to the endosteal membrane. Collagenase was also observed in osteoblasts involved in matrix formation, both in the presence and absence of previous resorption at that site (Figures 3A, 4, and 6C). The strong expression at sites of de novo bone formation (Figure 3A) suggests a role distinct from that postulated in activation, possibly concerned with the organization and alignment of rapidly laid down collagen fibrils prior to mineralization. The production of collagenase by osteoclasts indicates that it may play a role in bone resorption, possibly in the degradation of collagen following demineralization of bone. Metalloproteinases are optimally active at a neutral pH and thus are unlikely to be involved in the initial stages of resorption, in which bone mineral is dissolved in a highly acidic environment. However, at a later stage when the subcellular zone approaches a neutral pH, collagenase may be secreted by the osteoclast in order to degrade the demineralized collagen. The presence of collagenase in osteoclasts is controversial; in one study” collagenase mRNA in rat osteoclasts could not be detected, although it was present in adjacent cells. Furthermore. collagenase inhibitors did not affect resorption by isolated rabbit osteoclasts. In contrast, collagenase mRNA has been demonstrated in bovine odontoclasts,2x and the presence of collagenase reported in rodent osteoclasts.’ In addition, evidence that osteoclasts contain two other metalloproteinases, stromelysin and gelatinase, has been reported in rats’ and rabbits,“,” respectively, and metalloproteinase inhibitors have been shown to inhibit osteoclastic resorption.s.“.” Hill et al.‘* reported that specific MMP inhibitors had a dose dependent effect on “Ca release from IL-1 stimulated rat calvarial cultures and also demonstrated by immunolocalization the presence of collagenase. gelatinases A and B, stromelysin and MMP inhibitor-l (TIMP- I) in IL-1 stimulated isolated rabbit osteoclasts. Confocal microscopy revealed that these MMPs were localized along the entire secretory pathway, including the juxtanuclear Golgi complexes. Thus, although it is possible that some colla-
Expression
of collagenase
S. Bord et al. in human bone in vivo
39
genase detected within osteoclasts may be ingested during the process of bone resorption, evidence from a number of studies now supports the hypothesis that MMPs are synthesized by osteoclasts. Further studies are now required to investigate the role of collagenase and other metalloproteinases in bone and to extend our findings to normal human bone and to disease states such as osteoporosis. However, the implications of a role for metalloproteinases in the process of activation are considerable, in view of the importance of increased activation frequency as a mechanism of bone loss in osteoporosis? and the potential for inhibitors of metalloproteinases to reverse this process. At present, activation cannot be directly demonstrated by histological techniques and activation frequency can only be calculated on the basis of several, possibly untenable, assumptions. The use of fluorochromes such as tetracycline provides direct information about the dynamics of bone formation;‘4 the potential of metalloproteinases and other cell and matrix products to act as markers of activation and active resorption in bone provides an important area for future research.
Acknowledgments: One of the authors (J.E.C.) is supported by the Wellcome Trust and Sharyn Bord by SmithKline Beecham Pharmaceuticals, UK. R.M.H. and J.J.R. are supported by the Medical Research Council, U.K. The authors are indebted to Dr. Roger Wolman and Chris Constant for the supply of heterotopic and osteophytic bone.
References 1. Case, .I. P., Sane, H., Lafyati?, R., Remmers, E. F., Kumkumian, G. K., and Wilder, R. L. Transin/stromelysin expression in the synovium of rats with experimental erosive arthritis. In situ localization and kinetics of expression of the transformation-associated metalloproteinase in euthymic and athymic Lewis rats. .J Clin Invest 84:1731-1740; 1989. 2. Chambers, T. 1. and Fuller, K. Bone cells predispose hone surfaces to resorption by exposure of mineral to osteoclastic attack. J Cell Sci 76:155-165; 1985. 3. Chow, J. and Chambers, T. J. An assessment of the prevalence material on bone surfaces. Calcif Tissue Int 50:118-122; 1992.
of organic
4. Clark, I. M., Wright, J. K., Cawston, T. E., and Hazleman, B. L. Polyclonal antibodies against human iibroblast collagenase and the design of an enrymelinked immunoabsorbent assay to measure TIMP-collagenax complex. Matrix 12:108-l 15; 1992. 5. Croucher, P. 1. and Compston, J. E. Histomorphometric assessment of trabecular bone remodeling in osteoporosis. Bone Miner l4:91-102: 1991. 6. Dean, D. D., Muniz, 0. E., and Howell, D. S. Association of collagenax and tissue inhibitor of metalloproteinases (TIMP) with hypertrophic cell enlargement in the growth plate. Matrix 9:36&375; 1989. 7. DelalssB, J. M., Eeckhout, Y., Neff, L., Francois-Gillet, Ch., Hen&, P., So, Y., Vaes, G., and Baron. R. (F’ro)collagenase (matrix metalloproteinase-1) is prevznt in rodent osteoclasts and in the underlying bone-resorbing compartment. J Cell Sci 106:1071-1082: 1993. 8. DelairsB. J. M., Eeckhout. Y., Sear, C., Gailoway, A., McCullagh, K., and Vaes, G. A new synthetic inhibitor of mammalian tissue collagenase inhibits bone resorption in culture. Biochem Biophys Res Commun 133:483-490; 1985. 9. DelaissC, J. M.. Terlain, B., Cartwright, T., Lefebvre, V., and Vaes, G. A collagenase inhibitor inhibits bone resorption in viva, as evaluated by % tetracycline retention in bones of prelabelled mice. Bone 12:289; 1991. 10. Delaisst, J. M., Eeckhout. Y., and Vaes, G. Bone resorbing agents affect the production and distribution of procollagenase as well as the activity of collagenase in bone tissue. Endocrinology 123:264-276; 1988. 11. Dodds, R. A. and Gowen, M. The growing osteophyte: A model system for the study of human bone development and remodeling in situ. J Histotech 17:3745; 1994. 12. Evens, V., DelaissC, J. M., Korper, W., Neohof, A., Vaes, G., and Beer&n, W. Degradation of collagen in the bone resorbing compartment underlying the
40
13.
14. 15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
S. Bord et al. Expression of collagenase
Bone Vol. 19, No. 1 July
in human bone in vivo
osteoclast involves both cysteine-proteinases and matrix metalloproteinases. J Cell Physiol 150:221-231; 1992. Fuller, K. and Chambers, T. J. Localization of mRNA in osteocytic, bone surface, and chondrocytic cells but not osteoclasts. J Cell Sci 108:2221-2230; 1995. Frost, H. M. Tetracycline-based histological analysis of bone remodeling. Calcif Tissue Int 3:21 l-237; 1969. Heath, J. K., Atkinson, S. J., Meikle, M. C., and Reynolds, J. J. Mouse osteoblasts synthesize collagenase in response to bone-resorbing agents. Biochim Biophys Acta 802:151-154; 1984. Hembty, R. M., Bagga, M. R., Reynolds, J. J., and Hamblen, D. L. Immunolocalization studies on six matrix metalloproteinases and their inhibitors, TIMP-1 and TIMP-2, in synovia from patients with osteo- and rheumatoid arthritis. Ann Rheum Dis 54:25-32; 1995. Hembry, R. M., Murphy, G., Cawston, T. E., Dingle, J. T., and Reynolds, J. J. Characterization of a specific antiserum for mammalian collagenase from several species: Immunolocalization of collagenase in rabbit chondrocytes and uterus. J Cell Sci 81:105-123: 1986. Hill, P. A., Murphy, G. M., Docherty, A. J. P., Hembry, R. M., Millican, T. A., and Reynolds, J. J. The effects of selective inhibitors of matrix metalloproteinases (MMPs) on bone resorption and the identification of MMPs and TIMP-I in isolated osteoclasts. J Cell Sci 107:3055-3064; 1994. Hoyland, J., Freemont, A., and Sharpe, P. Interleukin-6, IL-6 receptor and IL-6 nuclear factor gene expression in Paget’s disease. J Bone Miner Res 9:75-80: 1994. Jayasinghe, J. A. P., Jones, S. J., and Boyde, A. Scanning electron microscopy of human lumbar vertebral trabecular bone surfaces. Virchows Archiv A Pathol Anat 422125-34; 1993. Meikle, M. C., Bard, S., Hembry, R. M., Compston, J. E., Croucher, P. I., and Reynolds, J. J. Human osteoblasts in culture synthesize collagenase and other matrix metalloproteinases in response to osteotrophic hormones and cytokines. J Cell Sci 103:1093-1099; 1992. Meikle, M. C., McGanity, A.M., Thomson, B. M., and Reynolds, J. J. Bonederived growth factors modulate collagenase and TIMP (tissue inhibitor of metalloproteinases) activity and type I collagen degradation by mouse calvtial osteoblasts. Bone Miner 12:41-55; 1991. Murphy, G., Allan, J. A., Willenbrock, F., Cockett, M. I., O’Connell, J. P., and Dochetty, A. J. P. The role of the C-terminal domain in collagenase and stromelysin specificity. J Biol Chem 267:9612-9618; 1992. Murphy, G., Cockett, M. I., Stephens, P. E., Smith, B. J., and Docherty, A. J. P.
1996:3540
Stromelysin is an activator of procollagenase. A study with natural and recombinant enzymes. Biochem J 248:265-268; 1987. 25. Murphy, G., Cockett, M. I., Ward, R. V., and Docherty, A. J. P. Matrix metalloproteinase degradation of elastin, type IV collagen and proteoglycan. A quantitative comparison of the activities of 95 kDA and 72 kDA gelatinases, stromelysins-1 and -2 and punctuated metalloproteinase (PUMP). Biochem J 2771277-279; 1991. 26. Murphy, G., Willenbrock, F., Ward, R. V.. Cockett, M. I., Eaton, D., and Dochetty, A. J. P. The C-terminal domain of 72 kDA gelatinase A is not required for catalysis, but is essential for membrane activation and modulates interactions with tissue inhibitors of metalloproteinases. Biochem J 283:637641; 1992. 27. O’Connell, J. P., Willenbrock, F., Docherty, A. J. P., Eaton, D., and Murphy, G. Analysis of the role of the COOH- terminal domain m the activation, proteolytic activity, and tissue inhibitor of metalloproteinase interactions of gelatinase B. J Biol Chem 269: 14967-14973; 1994. 28. Okamura, T., Shimokawa, H., Takagi, Y., Ono, H., and Sasaki, S. Detection of collagenase messenger RNA in odontoclasts of bovine root-resorbing tiswe by in sifu hybridization. Calcif Tissue Int 52:325-330; 1993. 29. Raina, V. Normal osteoid tissue. J Clin Pathol 25:229-232;
1972.
30. Sakamoto, S. and Sakamoto, M. Bone collagenase, osteoblasts, and cell mediated bone resorption. In: Peck, W. A., Ed. Bone and Mineral Research. Vol. 4. Elsevier: Amsterdam; 1986; 49-102. 31. Tezuka, K., Nemoto, K., Tezuka, Y., Sate, T., Ikeda, Y., Kobori, M., Kawashima, H., Eguhi, H., Hakeda, Y., and Kumegawa, M. Identification of matrix metalloproteinase 9 in rabbit osteoclasts. J Biol Chem 269:1500615009; 1994. 32. Whitham, S. E., Murphy, G., Angel, P., Rahmsdorf, H. J., Smith, B. J., Lyons, A., Harris, T. I. R., Reynolds, J. J., Herrlich, P., and Docheny. A. J. P. Comparison of human stromelysin and collagenase by cloning and sequence analysis. Biochem J 240:913-916; 1986.
Date Received: September 20, 1995 Date Revised: February 15, 1996 Date Accepted: March 13, 1996
ELSEVIER
Production Osteoclasts S. BORD,’
of Collagenase In Vivo
A. HORNER,’ R. M. HEMBRY,*
by Human Osteoblasts
J. J. REYNOLDS,”
and J. E. COMPSTON’
’Department of Medicine, University of Cambridge, School of Clinical Medicine, Addenbrooke’s Hospital, Cambridge, ’Strangeways Research Laboratory, Warts Causeway, Cambridge, UK ’Department of Orthodontics and Paediatric Dentistry, Guy’s Tower, Guy’s Hospital, London, UK
Osteoblasts;
Materials and Methods Tissue
Os-
Heterotopic bone was surgically removed from three patients (mean age 29 years) who had previously sustained severe head injury. Osteophytes were removed from three patients (mean age 68 years) during routine joint replacement surgery. Samples approximately 1 cm3 were snap frozen in precooled n-hexane (Sigma) and maintained at -70°C or fixed in 70% alcohol for a minimum of 48 h and dehydrated in progressive concentrations of alcohol (70%-100%). The latter samples were then transferred to LR White medium resin (TAAB), left for 4-7 days on a roller to allow the resin to infiltrate the tissue and then placed in glass vials in fresh resin and cured in an oven at 60°C.
Introduction Matrix metalloproteinases (MMPs) are a family of zinc-dependent enzymes responsible for the extracellular degradation of connective tissue. The best known member of the MMP family is collagenase. Specific collagenases cleave interstitial collagen types I, II, and III into three-fourths and one-fourth fragments. The enzyme is secreted in a latent proform that requires activation, a process that may involve plasmin and stromelysin.23 The denatured collagen may then be further degraded by another MMP, gelatinase. Degradation of matrix occurs under both physiological and
Hisrology
Undecalcified 8 pm sections from LR White embedded tissue were stained with 1% toluidine blue (pH 4.2), Diff-Quik (Baxter Dade AG), or Von Kossa with a Van Giesen counterstain and examined by bright field microscopy. Alkaline phosphatase and tartrate resistant acid phosphatase (TRAP) staining were demon-
Address for correspondence and reprints: Dr. J. E. Compston, Department of Medicine, Level 5, Addenbrooke’s Hospital, Cambridge CB2
2QQ, UK. 0 1996 by Else&r Science Inc. All rights reserved.
UK
pathological conditions in many tissues throughout the body. In the musculoskeletal system, MMPs are known to be involved in endochondral bone formation6 and osteoclastic degradation of bone matrix,‘0.‘2,29 but their role in bone remodeling has not been firmly established. Animal studies have shown that osteoblasts secrete collagenase in response to a number of stimulators of bone resorption, including 1,25 dihydroxyvitamin D, [1,25 (OH), D3] and interleukin-1 (IL-1).‘5,22 It has been suggested that the enzyme may degrade the nonmineralized collagenous membrane which covers mineralized bone, prior to osteoclastic resorption.2 Collagenase has also been identified in rodent osteoclasts’s and in the subosteoclastic compartment of bone undergoing resorption.’ The activity of all MMPs is tightly regulated by their specific inhibitors, TIMPs (tissue inhibitor of metalloproteinases). We have recently reported that human osteoblasts in vitro synthesise TIMP-1 and can be induced with bone resorbing hormones to produce collagenase.*’ However, the presence of MMPs in human bone in vivo has not yet been demonstrated. In this study we have investigated the expression of collagenase in human bone using heterotopic and osteophytic bone as models of high turnover.
Studies in some animal species have demonstrated the production of metalloproteinases by hone cells, suggesting that they may play a role in hone modeling and remodeling. The aim of the present study was to investigate the expression of collagenase in human hone in situ, using heterotopic and osteophytic hone. Immunohistochemistry was performed on chilled sections of hone, using well characterized polyclonal antibodies to human collagenase. The heterotopic and osteophytic hone exhibited high turnover and both bone modeling and remodeling were evident. Collagenase expression by osteoblasts was demonstrated in cells synthesising matrix and in lining cells; the strongest signal was seen in areas of de novo matrix formation, where bridges of woven bone were being formed between areas of mineralized bone. Collagenase was also present in some osteoclasts associated with eroded bone surfaces and in some mononuclear cells that were present in resorption cavities and in the bone marrow. Our results provide the first demonstration, in situ, of collagenase in human bone and suggest that it may play a role in human bone modeling and remodeling. Production of collagenase by active osteoblasts and lining cells suggests that it may be involved both in matrix formation and activation of hone remodeling. The presence of collagenase in osteoclasts provides further evidence that metalloproteinases may play a role in bone resorption. (Bone I9:35-40; 2996) Key Words: Collagenase; Metalloproteinases; teoclasts; Bone remodeling.
and
35
8756.3282/96/$15.00 PI1 S8756-3282(96)00106-8
36
S. Bord et al. Expression of collagenase
strated histologically tions (Sigma).
Bone Vol. 19, No. 1 July 1996:3540
in human bone in viva
according
to the manufacturer’s
instruc-
Antisera Human interstitial collagenase was purified according to Whitham et al.,” injected into sheep and the resulting antiserum characterized by Western blotting, inhibition and immunolocalization as described previously for the antibody to rabbit collagenase. I7 Specificity was also confirmed by positive staining on immunolocalization of NSO mouse myeloma cells transfected with human collagenase,23 but showing no staining of cells transfected with either human stromelysin-lz5 and -223 or gelatinase A2h or gelatinase B.” This antiserum has been used successfully to immunolocalize collagenase in human synovium.‘6 Immunoglobulins were prepared by triple ammonium sulphate precipitation. Antibodies were also raised in rabbits immunized with collagenase purified from the conditioned medium of W138 foetal lung fibroblasts; immunoglobulins were purified from this antiserum by protein A-sepharose chromatography and monospecificity tested by Ouchterlony immunoelectrophoresis and immunoblotting. Both antibodies recognize pro and active collagenase as well as collagenase complexed to TIMP. Neither antiserum cross-reacts with human neutrophil collagenase. Pooled normal sheep IgG and normal rabbit IgG were used appropriately as negative controls at the corresponding immunoglobulin concentrations. Immunohistochemistry 8 pm sections were cut on a Bright cryostat (Huntingdon, England) with a cabinet temperature of -30°C and equipped with a slow drive high torque motor and automatic speed control, using a highly polished tungsten carbide knife. Sections were picked off onto glass slides coated with 2% APES (3 - aminoprophyl triethoxy silane), air-dried for 10 min, fixed in 4% paraformaldehyde for 30 min at room temperature, and washed in phosphate buffered saline (PBS). Sections were incubated in 10% normal serum (from the species in which the second antibody was made) prior to application of the primary antibody (10 kg/mL), which was left on overnight at 4°C in a humid chamber. After washing and a further step to block endogenous peroxidase enzymes (20 min) using ImmunoPure (Pierce), the sections were washed again and treated with a secondary biotinylated antibody (5 kg/mL, Vectalabs, 40 min at room temperature). The reaction was completed using an ABC reagent (avidin/biotin complex, Vectalabs, 40 min, room temperature). 3-amino-9-ethylcarbazole (AEC), 3,3’-diaminobenzidine (DAB), or 4- chloro-1-naphthol (CN) substrate was used to detect sites of antibody binding. Some sections were counterstained with haematoxylin while others were treated with propidium iodide (0.1 kg/mL for 4 min) to detect nuclei. Finally, the slides were cover-slipped with Apathy’s mounting medium (BDH) and observed under bright field microscopy on an Olympus BH-2 photomicroscope. Photographs were taken on Ektachrome 64 film. Results Histologic Features of Heterotopic
those of subchondral ossification with nonhypertrophied chondrocytes in their matrix surrounding bone. The bone exhibited high turnover, with multilayers of cuboidal osteoblasts on plump osteoid seams and many areas in which multinucleated osteoclasts were seen in association with resorption cavities (Figure 1). Both bone remodeling and modeling were evident and examination of bone using polarized light revealed a combination of woven and lamellar bone, the former predominating. Within the woven bone, “bridges” of matrix linking areas of mineralized bone were observed (Figures 2, 3A, 3B). Staining with Von Kossa (Figure 2) and birefringence on toluidine blue stained sections viewed under polarized light microscopy confirmed that these bridges consisted of osteoid and sections above and below showed no overlying or underlying mineral. Haematoxylin counterstained sections showed the highly cellular areas within these bridges (Figure 3A) as did sections treated with propidium iodide which fluorescently labels nucleic acids. Their osteoblastic nature was confirmed by positive alkaline phosphatase staining. In some areas many blood vessels were present, particularly in fibrous tissue. The osteophytes demonstrated similar histological features as described in detail by Dodds et al.” Collagenase Both antibodies revealed very similar cell-associated staining for collagenase, with similar results observed in all samples of heterotopic and osteophytic bone. Multilayers of plump cuboidal osteoblasts within newly formed matrix in previously resorbed lacunae showed strong cytoplasmic staining in approximately half the cells (Figures 4 and 6C), with a stronger signal in the lining cells (Figure 4). In addition, osteoblasts associated with de novo bone formation exhibited intense collagenase staining, which was also observed both in the cells and to a lesser extent in the matrix of the bridges in the woven bone (Figure 3A). The intensity of collagenase staining in osteoclasts varied, some staining strongly (Figure 5) and others showing intermediate (Figures 6A, 6B) or low staining. In many osteoclasts juxtanuclear staining was apparent, probably in the Golgi apparatus. Positive tar&ate-resistant acid phosphatase (TRAP) expression by these cells in serial sections confirmed their osteoclastic nature. Mononuclear cells in the resorption cavities and in the marrow also showed collagenase positivity as did some of the osteocytes adjacent to the resorbed edges. There was negligible staining in all control sections. Alkuline Phosphatase Positive staining for alkaline phosphatase was most marked in areas of new bone formation. Large cuboidal osteoblasts associated with thick osteoid seams showed intense staining. Cartilage and fibrous tissue adjoining the bone showed high activity along their interfaces. TRAP TRAP, demonstrated histochemically, was particularly evident in the upper part of the hypertrophic zone. Intense staining appeared in the large spherical chondroblasts of the calcifying cartilage adjacent to the mineralizing edge. Within the bone tissue TRAP was confined to the large multinucleated and mononuclear cells associated with eroded surfaces.
and Osteophytic Bone
The heterotopic bone contained fibrous tissue, cartilage, and bone. Evidence of both intramembranous and endochondral ossification was seen, although in some areas the appearances were
Discussion Our results demonstrate, for the first time, the presence of collagenase in situ in human bone. These observations are consistent
Bone Vol. 19, No. 1 July 1996:35-40
Expression
of collagenase
S. Bord et al. in human hone in vivo
37
Figure 1. Heterotopic bone formation from patient 1. Undccalcified sections from LR White embedded tissue stained with Diff-Quik demonstrating high cellular activity. Key: b = bone; ob = osteoblasts; oc = osteoclasts; o = osteocytes. Bar = 80 km. Figure 2. Von Kossa stained undecalcified section of heterotopic woven bone from patient 1. The mineralized matrix stains black and the nonmineralized osteoid pink. “Bridges” of osteoid link one area of mineral with another (arrow). Bar = 40 pm. Figure 3. Indirect immunoperoxidase localization of collagenase in the de novo “bridge” formation in woven bone (arrow) of heterotopic bone from patient 1. The substrate, 4-chloronaphthol gives a black color reaction at sites of antigen binding. Bar = 20 pm. (A) Rabbit anticollagenase polyclonal antibody showing strong collagenase staining in the highly cellular “bridges” demonstrated by haematoxylin counter stain. (B) Normal rabbit serum shows absence of reaction (no counter stain).
38
S. Bord et al. Expression of collagenase
in human bone in viva
Bone Vol. 19, No. July 1996:3540
1
Figure 4. Indirect immunoperoxidase localization of collagenase in heterotopic bone from patient 3 using a sheep polyclonal antibody visualized with AEC as substrate shows positive staining of osteoblasts (ob) in a previously resorbed lacuna; note the distinct cytoplasmic staining (j) close to the nucleus (n). New matrix synthesis extends from the cement line (arrow heads). Note the intense signal of the lining cells (*). Bar = 20 km. Figure 5. Heterotopic bone from patient 3 showing immunolocalization of collagenase in osteoclasts (oc) within a resorption lacuna, and in osteocytes (0) in the bone matrix (b) using a rabbit antihuman polyclonal antibody with indirect immunoperoxidase. The positive cells are identified by the red color reaction of the AEC substrate, with the nucleus (n) and the juxtanuclear cytoplasmic staining (j) identified. The section was viewed under polarized light and shows random distribution of collagen fibrils in woven bone. Bar = 20 km. Figure 6. Immunolocalization in sections from an osteophyte from patient A, using sheep antihuman collagenase polyclonal antibody, visualized with the brown color reaction of DAB substrate. Key: b = bone. Bar = 20 pm. (A) Multinucleated osteoclast (arrow) close to the bone surface shows distinct nuclei (arrowheads). Mononuclear cells (m) within the marrow space also show strong collagenase staining. (B) An osteoclast (arrow) slightly pulled away from the bone surface during processing shows distinct nuclei (arrowheads): staining is evident in the perinuclear and cytoplasmic regions. (C) Osteoblasts (arrows) at sites of bone formation show cell-associated cytoplasmic collagenase staining. The cement line (arrowheads) of a previously resorbed cavity can be seen. (D) Normal sheep serum control shows absence of staining in the osteoblasts (arrows) at sites of formation.
Bone Vol. 19, No. 1 July 1996:3540
with the reported production of the enzyme by osteoblasts and osteoclasts in animals7S’5~‘8~22and in primary cultures of human osteoblasts,2’ and suggest that collagenase is involved in the regulation of bone turnover in man. Its localization in both cell types and at sites of bone resorption and formation (Figures 3A, 4, 5, 6A, 6B, and 6C) indicates that it may perform multiple functions in the process of bone remodeling. The use of heterotopic and osteophytic bone as a model of rapid bone turnover has obvious advantages related to the high levels of cellular activity and expression of cellular and matrix products. The formation of heterotopic bone or osteophytic bone is, however, a pathological event and factors which regulate its modeling and remodeling may differ qualitatively and quantitatively from those which operate under normal circumstances. Osteophytic bone’ ’and bone affected by Paget’s disease” have also been used as models of high bone turnover to study expression of cytokines and growth factors in man; however, presence of collagenase was not investigated in these studies. The production of collagenase by osteoblastic lining cells may be important in the activation of bone remodeling, in which the surface is “prepared’ for resorption by retraction of the lining cells and removal of the nonmineralized collagenous membrane covering the bone surface. Most studies indicate that this material, which is less than 1 km thick and separates mineralized bone from cells on its surface, forms a continuous layer over quiescent bone surfaces’~*” and it is generally believed that removal of this collagenous layer is required before resorption can take place, although resorption of nonmineralized matrix by osteoclasts may also occur.*’ The lining cells are ideally placed to initiate activation, being in communication with the strainsensitive osteocytic-canalicular network and in close approximation to the endosteal membrane. Collagenase was also observed in osteoblasts involved in matrix formation, both in the presence and absence of previous resorption at that site (Figures 3A, 4, and 6C). The strong expression at sites of de novo bone formation (Figure 3A) suggests a role distinct from that postulated in activation, possibly concerned with the organization and alignment of rapidly laid down collagen fibrils prior to mineralization. The production of collagenase by osteoclasts indicates that it may play a role in bone resorption, possibly in the degradation of collagen following demineralization of bone. Metalloproteinases are optimally active at a neutral pH and thus are unlikely to be involved in the initial stages of resorption, in which bone mineral is dissolved in a highly acidic environment. However, at a later stage when the subcellular zone approaches a neutral pH, collagenase may be secreted by the osteoclast in order to degrade the demineralized collagen. The presence of collagenase in osteoclasts is controversial; in one study” collagenase mRNA in rat osteoclasts could not be detected, although it was present in adjacent cells. Furthermore. collagenase inhibitors did not affect resorption by isolated rabbit osteoclasts. In contrast, collagenase mRNA has been demonstrated in bovine odontoclasts,2x and the presence of collagenase reported in rodent osteoclasts.’ In addition, evidence that osteoclasts contain two other metalloproteinases, stromelysin and gelatinase, has been reported in rats’ and rabbits,“,” respectively, and metalloproteinase inhibitors have been shown to inhibit osteoclastic resorption.s.“.” Hill et al.‘* reported that specific MMP inhibitors had a dose dependent effect on “Ca release from IL-1 stimulated rat calvarial cultures and also demonstrated by immunolocalization the presence of collagenase. gelatinases A and B, stromelysin and MMP inhibitor-l (TIMP- I) in IL-1 stimulated isolated rabbit osteoclasts. Confocal microscopy revealed that these MMPs were localized along the entire secretory pathway, including the juxtanuclear Golgi complexes. Thus, although it is possible that some colla-
Expression
of collagenase
S. Bord et al. in human bone in vivo
39
genase detected within osteoclasts may be ingested during the process of bone resorption, evidence from a number of studies now supports the hypothesis that MMPs are synthesized by osteoclasts. Further studies are now required to investigate the role of collagenase and other metalloproteinases in bone and to extend our findings to normal human bone and to disease states such as osteoporosis. However, the implications of a role for metalloproteinases in the process of activation are considerable, in view of the importance of increased activation frequency as a mechanism of bone loss in osteoporosis? and the potential for inhibitors of metalloproteinases to reverse this process. At present, activation cannot be directly demonstrated by histological techniques and activation frequency can only be calculated on the basis of several, possibly untenable, assumptions. The use of fluorochromes such as tetracycline provides direct information about the dynamics of bone formation;‘4 the potential of metalloproteinases and other cell and matrix products to act as markers of activation and active resorption in bone provides an important area for future research.
Acknowledgments: One of the authors (J.E.C.) is supported by the Wellcome Trust and Sharyn Bord by SmithKline Beecham Pharmaceuticals, UK. R.M.H. and J.J.R. are supported by the Medical Research Council, U.K. The authors are indebted to Dr. Roger Wolman and Chris Constant for the supply of heterotopic and osteophytic bone.
References 1. Case, .I. P., Sane, H., Lafyati?, R., Remmers, E. F., Kumkumian, G. K., and Wilder, R. L. Transin/stromelysin expression in the synovium of rats with experimental erosive arthritis. In situ localization and kinetics of expression of the transformation-associated metalloproteinase in euthymic and athymic Lewis rats. .J Clin Invest 84:1731-1740; 1989. 2. Chambers, T. 1. and Fuller, K. Bone cells predispose hone surfaces to resorption by exposure of mineral to osteoclastic attack. J Cell Sci 76:155-165; 1985. 3. Chow, J. and Chambers, T. J. An assessment of the prevalence material on bone surfaces. Calcif Tissue Int 50:118-122; 1992.
of organic
4. Clark, I. M., Wright, J. K., Cawston, T. E., and Hazleman, B. L. Polyclonal antibodies against human iibroblast collagenase and the design of an enrymelinked immunoabsorbent assay to measure TIMP-collagenax complex. Matrix 12:108-l 15; 1992. 5. Croucher, P. 1. and Compston, J. E. Histomorphometric assessment of trabecular bone remodeling in osteoporosis. Bone Miner l4:91-102: 1991. 6. Dean, D. D., Muniz, 0. E., and Howell, D. S. Association of collagenax and tissue inhibitor of metalloproteinases (TIMP) with hypertrophic cell enlargement in the growth plate. Matrix 9:36&375; 1989. 7. DelalssB, J. M., Eeckhout, Y., Neff, L., Francois-Gillet, Ch., Hen&, P., So, Y., Vaes, G., and Baron. R. (F’ro)collagenase (matrix metalloproteinase-1) is prevznt in rodent osteoclasts and in the underlying bone-resorbing compartment. J Cell Sci 106:1071-1082: 1993. 8. DelairsB. J. M., Eeckhout. Y., Sear, C., Gailoway, A., McCullagh, K., and Vaes, G. A new synthetic inhibitor of mammalian tissue collagenase inhibits bone resorption in culture. Biochem Biophys Res Commun 133:483-490; 1985. 9. DelaissC, J. M.. Terlain, B., Cartwright, T., Lefebvre, V., and Vaes, G. A collagenase inhibitor inhibits bone resorption in viva, as evaluated by % tetracycline retention in bones of prelabelled mice. Bone 12:289; 1991. 10. Delaisst, J. M., Eeckhout. Y., and Vaes, G. Bone resorbing agents affect the production and distribution of procollagenase as well as the activity of collagenase in bone tissue. Endocrinology 123:264-276; 1988. 11. Dodds, R. A. and Gowen, M. The growing osteophyte: A model system for the study of human bone development and remodeling in situ. J Histotech 17:3745; 1994. 12. Evens, V., DelaissC, J. M., Korper, W., Neohof, A., Vaes, G., and Beer&n, W. Degradation of collagen in the bone resorbing compartment underlying the
40
13.
14. 15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
S. Bord et al. Expression of collagenase
Bone Vol. 19, No. 1 July
in human bone in vivo
osteoclast involves both cysteine-proteinases and matrix metalloproteinases. J Cell Physiol 150:221-231; 1992. Fuller, K. and Chambers, T. J. Localization of mRNA in osteocytic, bone surface, and chondrocytic cells but not osteoclasts. J Cell Sci 108:2221-2230; 1995. Frost, H. M. Tetracycline-based histological analysis of bone remodeling. Calcif Tissue Int 3:21 l-237; 1969. Heath, J. K., Atkinson, S. J., Meikle, M. C., and Reynolds, J. J. Mouse osteoblasts synthesize collagenase in response to bone-resorbing agents. Biochim Biophys Acta 802:151-154; 1984. Hembty, R. M., Bagga, M. R., Reynolds, J. J., and Hamblen, D. L. Immunolocalization studies on six matrix metalloproteinases and their inhibitors, TIMP-1 and TIMP-2, in synovia from patients with osteo- and rheumatoid arthritis. Ann Rheum Dis 54:25-32; 1995. Hembry, R. M., Murphy, G., Cawston, T. E., Dingle, J. T., and Reynolds, J. J. Characterization of a specific antiserum for mammalian collagenase from several species: Immunolocalization of collagenase in rabbit chondrocytes and uterus. J Cell Sci 81:105-123: 1986. Hill, P. A., Murphy, G. M., Docherty, A. J. P., Hembry, R. M., Millican, T. A., and Reynolds, J. J. The effects of selective inhibitors of matrix metalloproteinases (MMPs) on bone resorption and the identification of MMPs and TIMP-I in isolated osteoclasts. J Cell Sci 107:3055-3064; 1994. Hoyland, J., Freemont, A., and Sharpe, P. Interleukin-6, IL-6 receptor and IL-6 nuclear factor gene expression in Paget’s disease. J Bone Miner Res 9:75-80: 1994. Jayasinghe, J. A. P., Jones, S. J., and Boyde, A. Scanning electron microscopy of human lumbar vertebral trabecular bone surfaces. Virchows Archiv A Pathol Anat 422125-34; 1993. Meikle, M. C., Bard, S., Hembry, R. M., Compston, J. E., Croucher, P. I., and Reynolds, J. J. Human osteoblasts in culture synthesize collagenase and other matrix metalloproteinases in response to osteotrophic hormones and cytokines. J Cell Sci 103:1093-1099; 1992. Meikle, M. C., McGanity, A.M., Thomson, B. M., and Reynolds, J. J. Bonederived growth factors modulate collagenase and TIMP (tissue inhibitor of metalloproteinases) activity and type I collagen degradation by mouse calvtial osteoblasts. Bone Miner 12:41-55; 1991. Murphy, G., Allan, J. A., Willenbrock, F., Cockett, M. I., O’Connell, J. P., and Dochetty, A. J. P. The role of the C-terminal domain in collagenase and stromelysin specificity. J Biol Chem 267:9612-9618; 1992. Murphy, G., Cockett, M. I., Stephens, P. E., Smith, B. J., and Docherty, A. J. P.
1996:3540
Stromelysin is an activator of procollagenase. A study with natural and recombinant enzymes. Biochem J 248:265-268; 1987. 25. Murphy, G., Cockett, M. I., Ward, R. V., and Docherty, A. J. P. Matrix metalloproteinase degradation of elastin, type IV collagen and proteoglycan. A quantitative comparison of the activities of 95 kDA and 72 kDA gelatinases, stromelysins-1 and -2 and punctuated metalloproteinase (PUMP). Biochem J 2771277-279; 1991. 26. Murphy, G., Willenbrock, F., Ward, R. V.. Cockett, M. I., Eaton, D., and Dochetty, A. J. P. The C-terminal domain of 72 kDA gelatinase A is not required for catalysis, but is essential for membrane activation and modulates interactions with tissue inhibitors of metalloproteinases. Biochem J 283:637641; 1992. 27. O’Connell, J. P., Willenbrock, F., Docherty, A. J. P., Eaton, D., and Murphy, G. Analysis of the role of the COOH- terminal domain m the activation, proteolytic activity, and tissue inhibitor of metalloproteinase interactions of gelatinase B. J Biol Chem 269: 14967-14973; 1994. 28. Okamura, T., Shimokawa, H., Takagi, Y., Ono, H., and Sasaki, S. Detection of collagenase messenger RNA in odontoclasts of bovine root-resorbing tiswe by in sifu hybridization. Calcif Tissue Int 52:325-330; 1993. 29. Raina, V. Normal osteoid tissue. J Clin Pathol 25:229-232;
1972.
30. Sakamoto, S. and Sakamoto, M. Bone collagenase, osteoblasts, and cell mediated bone resorption. In: Peck, W. A., Ed. Bone and Mineral Research. Vol. 4. Elsevier: Amsterdam; 1986; 49-102. 31. Tezuka, K., Nemoto, K., Tezuka, Y., Sate, T., Ikeda, Y., Kobori, M., Kawashima, H., Eguhi, H., Hakeda, Y., and Kumegawa, M. Identification of matrix metalloproteinase 9 in rabbit osteoclasts. J Biol Chem 269:1500615009; 1994. 32. Whitham, S. E., Murphy, G., Angel, P., Rahmsdorf, H. J., Smith, B. J., Lyons, A., Harris, T. I. R., Reynolds, J. J., Herrlich, P., and Docheny. A. J. P. Comparison of human stromelysin and collagenase by cloning and sequence analysis. Biochem J 240:913-916; 1986.
Date Received: September 20, 1995 Date Revised: February 15, 1996 Date Accepted: March 13, 1996