Histochem Cell Biol (1996) 106:495501

Springer-Verlag 1996

O R I G I N A L PA P E R

&roles:Elisabeth M. Aarden Anne-Marie M. Wassenaar Marcel J. Alblas Peter J. Nijweide

Immunocytochemical demonstration of extracellular matrix proteins in isolated osteocytes

&misc:Accepted: 9 July 1996

&p.1:Abstract Cultures of isolated osteocytes may offer an appropriate system to study osteocyte function, since isolated osteocytes in culture behave very much like osteocytes in vivo. In this paper we studied the capacity of osteocytes to change their surrounding extracellular matrix by production of matrix proteins. With an immunocytochemical method we determined the presence of collagen type I, fibronectin, osteocalcin, osteopontin and osteonectin in cultures of isolated chicken osteocytes, osteoblasts and periosteal fibroblasts. In osteoblast and periosteal fibroblast cultures, large extracellular networks of collagen type I and fibronectin were formed, but in osteocyte populations, extracellular threads of collagen or fibronectin were only rarely found. The percentage of cells positive for osteocalcin, osteonectin and osteopontin in the Golgi apparatus, on the other hand, was highest in the osteocyte population. These results show that osteocytes have the ability to alter the composition of their surrounding extracellular matrix by producing matrix proteins. We suggest this property is of importance for the regulation of the calcification of the bone matrix immediately surrounding the cells. More importantly, as osteocytes depend for their role as mechanosensor cells on their interaction with matrix proteins, the adaptation of the surrounding matrix offers a way to regulate their response to mechanical loading.&bdy:

Introduction
The development of a monoclonal antibody (mAb) directed to antigenic sites only present on the cell surface of osteocytes (mAb OB7.3; Nijweide and Mulder 1986) has made it possible to identify osteocytes in mixed populations of cells isolated from chicken bone. The antiE.M. Aarden (u) A.-M.M. Wassenaar M.J. Alblas P.J. Nijweide Laboratory of Cell Biology and Histology, Department of Cell Biology, Faculty of Medicine, Leiden University, PO Box 2156, NL-2301 CD Leiden, The Netherlands Tel. +31 71-5276415; Fax +31 71-5276437&/fn-block:

body has also been used to isolate osteocytes from mixed calvarial cell populations by applying an immunomagnetic isolation procedure (van der Plas and Nijweide 1992). More recently this method has been used to study some characteristics and properties of osteocytes in culture (van der Plas et al. 1994): isolated osteocytes when seeded on a culture substratum formed networks of cells interconnected with each other via long, slender cell processes; osteocytes appeared not to be able to proliferate; alkaline phosphatase activity per cell was variable between the individual cells but their mean activity was low compared to that of osteoblasts. The conclusion from these experiments was that isolated osteocytes behave in culture, as far as we know at the moment, very much like osteocytes in vivo, and that therefore cultures of isolated osteocytes may offer an appropiate system to study osteocyte function. At present little is known about the function of osteocytes in bone metabolism (Aarden et al. 1994). A number of decades ago it was thought that osteocytes were involved in blood calcium homeostasis and were capable of local bone resorption, osteocytic osteolysis (Blanger 1969). Later, this idea of osteocytic osteolysis was refuted (Parfitt 1977; Boyde 1980; Marotti et al. 1990). Indeed, we observed no sign at all of any resorptive activity when isolated osteocytes were seeded onto dentine slices (van der Plas et al. 1994). The possibility remains, however, that osteocytes are involved in facilitation of calcium and phosphate exchange between blood and bone, thereby contributing to the fine regulation of blood calcium homeostasis (Bonucci 1990; Aarden et al. 1994). A second possible function of osteocytes is a role in the maturation of osteoid matrix and regulation of its calcification. Matrix mineralisation generally takes place some distance from the osteoblastic layer, i.e. in the matrix between recently incorporated, osteoid osteocytes. It is feasible that osteocytes, by secreting (non-collagenous) matrix proteins into the matrix around them, prepare or adapt that matrix for mineralisation (Mikuni-Takagaki et al. 1995). A third possible role for osteocytes is as mechanosensory cells in the regulation of the process of functional adapta-

496

tion to mechanical stress (Lanyon 1993; Aarden et al. 1994). The notion that the mechanical function of bone determines to a large extent skeletal design and tissue organisation has become widely accepted (Rubin et al. 1990). Osteoblasts and osteoclasts are clearly responsible for the execution of the necessary adaptation to mechanical stress, but the identity of the sensor cell, the cell that translates mechanical stimuli into chemical messages which activate the (re)modelling system (osteoblasts, osteoclasts), is not known. It is a very attractive idea that osteocytes play such a role as mechanosensory cells and supporting evidence for this notion is accumulating (see for review Aarden et al. 1994). Indeed, we have recently acquired experimental evidence that isolated osteocytes are extremely sensitive to mechanical stress (Klein Nulend et al. 1995). If osteocytes are the mechanosensory cells of bone, the composition of the matrix around the osteocytes is of extreme importance. Osteocytes have to be anchored to the matrix in order to perceive strain. This anchorage probably takes place via bone extracellular matrix (ECM) proteins. The nature and the number of anchoring sites may determine the sensitivity of the cells for strain and the signal transduction system that translates strain into intra- or extracellular signal molecules. Finally, whatever the function of the osteocytes, they have to be actively involved in the adaptation of the matrix directly around them for a more general reason. Osteocytes are dependent on the diffusion of oxygen, ions, nutrients, waste products and probably chemical signal molecules through the lacunar and canalicular spaces. They have to be able to stop the mineralisation front at some distance from the cell membrane in order to keep an open space in which diffusion is possible. Involvement of osteocytes in matrix adaptation by production of ECM proteins is feasible. From immunohistochemical and in situ hybridisation studies on human, rat and porcine bone sections, it is known that ECM proteins and/or ECM protein mRNA are present in osteocytes, such as osteocalcin protein (Bronckers et al. 1985; Vermeulen et al. 1989; Boivin et al. 1990) and osteocalcin mRNA (Ikeda et al. 1992), osteopontin protein (Mark et al. 1987; Chen et al. 1991; 1993a) and osteopontin mRNA (Arai et al. 1993; Chen et al. 1993b, 1994), and osteonectin protein (Chen et al. 1991, 1993a) and osteonectin mRNA (Bianco et al. 1985; Metsranta et al. 1989). Of special interest is that osteopontin and bone sialoprotein, besides being concentrated in cement lines and lamina limitans, are also particularly present in the perilacunar matrix of some osteocytes (Ingram et al. 1993; McKee et al. 1993). In the present paper we have studied the production of ECM proteins by isolated chicken osteocytes and compared isolated osteocytes with osteoblasts and fibroblasts in this respect. With immunocytochemical methods we have determined the presence of several proteins in and around osteocytes: the predominant bone matrix protein collagen type I (COLL I), proteins that may be involved in the regulation of mineralisation, such as osteocalcin (OC; Hauschka et al. 1975; Price et al. 1976) and osteonectin (ON; Termine et al. 1981) and proteins that are

probably involved in cell attachment, such as fibronectin (FN; Hynes and Yamada 1982) and osteopontin (OP; Oldberg et al. 1986). A difference in the matrix around osteocytes, when compared to that surrounding osteoblasts and periosteal fibroblasts, may reflect a different relationship and/or adhesion of osteocytes to their surrounding matrix and hence be important in the light of their proposed mechanosensory function.

Materials and methods

Isolation of bone cells Bone cell populations were isolated from 18-day-old foetal chicken calvariae according to van der Plas and Nijweide (1992). In short, calvariae were dissected and the outer fibrous layers were stripped off; from these outer layers periosteal fibroblasts were isolated with collagenase (245 min; 1 mg/ml crude collagenase type 1, Sigma, St. Louis, USA) in HEPES buffer (Hefley 1987); the calvariae were submitted to subsequent treatments with 1 mg/ml collagenase (10 and 45 min), 4 mM EDTA in phosphatebuffered saline (PBS; 10 min) and collagenase (45 min); the three last fractions were combined, washed and seeded for culture, this population, a mixture of osteocytes (stellate, mAb OB7.3-positive cells), osteoblasts (roundish or oval, mAb OB7.3-negative cells) and a small number of fibroblasts (elongated, spindle shaped, mAb OB7.3-negative cells) was designated OBmix. From the OBmix population the osteocytes were isolated by an immunomagnetic procedure (van der Plas and Nijweide 1992). Briefly, 1-daycultured OBmix were harvested with a mixture of 0.05% trypsin and 0.01% EDTA in PBS; the suspension of cells was filtered through a 30-m pore nylon filter; after centrifugation the pellet was resuspended in PBS and subsequently immunomagnetic beads (Dynabeads M-450 coated with sheep anti-mouse IgG1; Dynal, N0212, Oslo, Norway), coupled to the osteocyte-specific mAb OB7.3 (3 g/ml IgG to 3 mg/ml beads), were added; osteocytes were allowed to couple to immunolabelled beads for 5 min; osteocytes bound to the beads were removed from the cell suspension with a magnet, and subsequently detached from the beads with an

Fig. 1ah The presence of extracellular matrix proteins (ECM) proteins in cultures of osteocytes, osteoblasts and periosteal fibroblasts. a Extracellular collagen type I network (red) formed between periosteal fibroblasts after 3 days of culture (magnification 365). b Intracellular (orange) collagen type I in an osteocyte and extracellular (red) fibres of collagen after 1 day of culture. The osteocyte is identified with monoclonal antibody (mAb) OB7.3 (green). The Golgi apparatus and the secretory vesicles are stained orange because of the green fluorescence on top of the red fluorescence (magnification 1400). c Fibronection network (green) produced by osteoblasts after 3 days of culturing of mixed cells (OBmix) (magnification 300). d Fibronectin staining (yellow when on the red, mAb OB7.3-positive osteocyte and green outside the cell profile) on and around an osteocyte and an osteoblast (OBmix after 1 day; magnification 1300). e Identification of osteopontin (yellow) in osteocytes (green) within the Golgi apparatus and secretory vesicles (culture of osteocyte population (OCY) for 3 days; magnification 1300). f Extracellularly, osteopontin (red) shows a collagen-like fibre orientation in OBmix culture (1 day; magnification 520). g Osteocalcin (yellow/red) in the Golgi apparatus and secretory vesicles of two osteocytes (green) and one osteoblast (culture of OCY after 4 days; magnification 570). h Intracellular osteonectin (red) distribution in the Golgi apparatus and secretory vesicles of OBmix cultured for 3 days (magnification 520). Nuclei are stained blue (4,6-diamidino-2-phenylindole) in all cell cultures. (Asterisk Golgi apparatus, arrow secretory vesicle)ig.c:&/f

497

498 excess of mAb OB7.3 (250 g/ml) in PBS; this final cell population consisted of more than 95% osteocytes (van der Plas and Nijweide 1992) and was called the osteocyte population (OCY). Cell culture and immunocytochemistry The cell populations periosteal fibroblasts, OBmix and OCY were cultured on glass coverslips in MEM (minimal essential medium; Gibco Life Technologies, Gaithersburg, USA) supplemented with 2% foetal calf serum (Gibco). After 16 days of culture, the cells were washed, twice with Hanks Balanced Salt Solution (HBSS) and once with 0.25% bovine serum albumin (BSA) in HBSS (HBSA). For immunocytochemistry with the antibody against OC, 1% foetal calf serum in HBSS was used instead of HBSA because of the cross-reactivity of the antibody with some batches of BSA. After removal of the culture medium and washing, the cells were incubated with mAb OB7.3 for 30 min. Again, the cells were washed with HBSA and HBSS and subsequently fixed in 4% formaldehyde in HBSS for 10 min at 4 C. Washing followed, with PBS for extracellular labelling of proteins or with 0.1% Triton X-100 in PBS (10 min at 0 C) for both extra- and intracellular staining. For the identification of the mAb OB7.3-labelled osteocytes, cell preparations were incubated with biotinylated horse anti-mouse antibody (15 g/ml; Vector Laboratories, Burlingame, USA) and subsequently with fluoresceince isothiocyanate (FITC)conjugated extravidine (green fluorescence; 20 g/ml, Sigma) or with Cy3-conjugated donkey anti-mouse antibody (red fluorescence; 3 g/ml, Jackson Immuno Research Laboratories, West Grove, USA). The antibodies used for the demonstration of the various matrix proteins were: rabbit anti-chicken COLL I (Chemicon, Temecula, USA; working concentration 1/500); rabbit antichicken OC (gift of J. Lian; see Hauschka et al. 1983; concentration 1/200); rabbit anti-chicken OP (gift of Y. Gotoh; see Gerstenfeld et al. 1990; Gotoh et al. 1990; concentration 1/300); rabbit anti-chicken bone ON (LF8; gift of L. Fisher; see Pacifici et al. 1990; concentration 1/300); goat anti-chicken FN (gift of D.R. Garrod; see Roach 1992; concentration 1/1000). Incubation with these primary antibodies was followed, after thorough washing, by incubation with an appropriate secondary antibody, either donkey anti-rabbit conjugated to Cy3 (3 g/ml, Jackson) or FITC-conjugated rabbit anti-goat (5 g/ml, Sigma). Finally, nuclei were counterstained with 4,6-diamidino-2-phenylindole.2 HCl (DAPI; 60 ng/ml; Serva, Heidelberg, Germany). As controls, cell populations were stained according to the procedures described above, leaving out the primary antibody. Additionally, the number of cells positive for OP, OC and ON intracellularly was counted. Osteocytes, osteoblasts and periosteal fibroblasts were identified as described above in OCY, OBmix and periosteal fibroblast populations, respectively, and the percentage of positive cells was determined.

Fig. 2 Simultaneous intra- and extracellular identification of collagen type I. After 6 days of culture of OBmix, an extracellular network can be stained around osteoblasts together with an intracellular localisation in the Golgi apparatus and secretory vesicles of the cells (magnification 480)ig.c:&/f

Results
Immunocytochemically, the presence of COLL I, FN, OC, OP and ON could be demonstrated in the three cell
Table 1 Percentage of osteocytes, osteoblasts and periosteal fibroblasts intracellularly positive for osteopontin (OP), osteocalcin (OC) and osteonectin (ON). Percentages are expressed as meansSD. The numbers of cultures are in parentheses. Osteocytes were identified as monoclonal antibody (mAb) OB7.3-poECM protein identified OP OC ON &/tbl.: Osteocytes 65.912.8 (8) 83.87.4 (5) 76.55.8 (6)

types investigated: osteocytes (i.e. mAb OB7.3-positive cells in OBmix and OCY populations), osteoblasts (i.e. roundish or oval mAb OB7.3-negative cells in OBmix) and periosteal fibroblasts. Reactivity was heterogeneous, meaning that not all cells of each type were positive and, when positive, not all to the same extent. COLL I immunoreactivity was seen as an extracellular network of woven fibres between osteoblasts and periosteal fibroblasts (Fig. 1a). In OCY populations, such a network was only visible in the immediate vicinity of an occasional, contaminating osteoblast. Around osteocytes, without neighbouring non-osteocytes, only a few collagen fibres could be observed (Fig. 1b). Intracellularly, collagen production was demonstrated as an intensive staining of the Golgi apparatus and of secretory vesicles in osteoblasts and periosteal fibroblasts (Fig. 2), and also in osteocytes
sitive cells in mixed cell (OBmix) and osteocyte populations, osteoblasts as roundish or oval mAb OB7.3-negative cells in OBmix, and periosteal fibroblasts as elongated, spindle shaped, mAb OB7.3negative cells in periosteal fibroblasts. (ECM) Extracellular matrix protein)&/tbl.c: Periosteal fibroblasts 12.43.9 (4) 0.81.1 (3) 33.09.8 (3)

Osteoblasts 17.86.7 (8) 24.85.8 (5) 23.412.7 (6)

499

(Fig. 1b). Osteoblasts (Fig. 1c) and fibroblasts in culture also formed a web of FN, very similar to the COLL I network. A typical staining pattern could be observed on and around individual osteoblasts (Fig. 1d) and fibroblasts. On osteocytes, only little dots or very short fibronectin fibres could be detected (Fig. 1d). In all populations only occasionally was the Golgi apparatus found to be positive for FN (not shown). OP, OC and ON staining showed very similar patterns. OP was found in the Golgi apparatus within osteocytes (Fig. 1e), osteoblasts and periosteal fibroblasts. Additionally, extracellular threads of anti-OPpositive material were found in OBmix (Fig. 1f) and periosteal fibroblasts, resembling the COLL I and FN network orientation. Distinct reactivity for OC was also found intracellularly in the Golgi apparatus and vesicles of osteocytes, osteoblasts (Fig. 1g) and fibroblasts. ON too could be identified intracellularly in all types of cells, i.e. osteocytes, osteoblasts (Fig. 1h) and periosteal fibroblasts. The number of osteocytes positive for OP, OC and ON intracellularly was higher than the number of osteoblasts or periosteal fibroblasts (Table 1).

Discussion
The extracellular matrix of bone consists of about 90% COLL I. The collagen fibres form by their many interand intramolecular cross-links the basic structural network in which mineralisation takes place. Non-collagenous proteins comprise about 10% of the total bone protein content. A small fraction of these proteins are of extra-osseous origin, but the majority is produced and secreted by bone cells. Most of these ECM proteins are able to bind to COLL I, such as FN (Hynes and Yamada 1982; Majmudar et al. 1991), OC (Bianco et al. 1985; Vermeulen et al. 1989; Boivin et al. 1990), ON (Termine et al. 1981) and OP (Chen et al. 1992). The precise functions of most of the non-collagenous proteins are not known. The fact that they are deposited into bone at different stages of bone formation may reflect their various functions. FN is produced during various phases of bone formation (Weiss and Reddi 1981) and clearly has a function in cell-matrix adhesion (Yamada 1983; Grzesik and Gehron Robey 1994). OP is only incorporated in the bone matrix prior to mineralisation (Mark et al. 1988; McKee et al. 1992; Sodek et al. 1992), and is probably involved both in cell-matrix adhesion and in the initiation of mineralisation (Roach 1994). ON is found in newly laid down osteoid (Bianco et al. 1988) and, together with OC, it probably prevents excessive mineralisation (Roach 1994). OC is not present in osteoid, only in calcified matrix (Groot et al. 1986). Its synthesis has been reported not to take place until after mineralisation has occurred (McKee et al. 1992; Kasai et al. 1994). The production and calcification of bone matrix has long been considered to be a primary function of the os-

teoblast. However, over the last few years, an increasing number of reports have also indicated that other cells in bone, notably osteocytes, may contribute to bone matrix synthesis (see Introduction). In this paper we studied the presence of the matrix proteins COLL I, FN, OP, ON and OC in and around isolated and cultured chicken osteocytes. Osteocytes were shown to produce only occasional COLL I fibres or FN threads, while osteoblasts and periosteal fibroblasts strongly secreted COLL I, as is shown by the large networks of collagen fibres that are formed around these cells during culture. The anti-FN-positive networks around the osteoblasts and periosteal fibroblasts, which are so similar to the COLL I networks, probably represent FN adherence to the collagen fibres. Therefore, the small amount of FN seen around osteocytes does not have to signify that osteocytes secrete little FN, but may be due to the small number of collagen fibres. Intracellularly, FN was less demonstrable in all three cell types than COLL I, perhaps due to a faster secretion of FN compared to COLL I. Besides COLL I (and to a lesser extent FN) OP, ON and OC could also be immunocytochemically demonstrated in the Golgi apparatus and secretory vesicles of periosteal fibroblasts, osteoblasts and osteocytes in culture. OP, and to a lesser extent ON, were also found extracellularly, in osteoblast and fibroblast, but not in osteocyte cultures. On the other hand, the percentage of osteocytes positive for OP, ON and OC intracellularly was much higher than that of osteoblasts or periosteal fibroblasts. This probably relates to the fact that the OCY population is a homogeneous population of mature, differentiated cells, while the OBmix and periosteal fibroblast populations contain a large variety of differention stages, from actively secreting osteoblasts and fibroblasts to precursor stages that probably do not (yet) produce many matrix proteins. These results suggest that osteocytes are capable of producing various matrix proteins, although probably to a limited extent as was explicitly shown for insoluble COLL I. OCY cultures, such as those described here, allow study of the possible effects of external factors such as hormones, cytokines and mechanical loading on the matrix production profile of osteocytes in a systematic way. In particular, the study of the effects of mechanical loading on matrix production by osteocytes will be of extreme interest considering their proposed role as mechanosensor cells (see Introduction). The recent papers of Klein Nulend et al. (1995) and Sun et al. (1995) have demonstrated that isolated osteocytes are extremely sensitive to mechanical loading as they respond (in vivo) to mechanical stress with an increased COLL I mRNA content. The mechanism proposed here, by which osteocytes may regulate their response to mechanical loading through regulation of matrix production and, maybe, the regulation of the attachment to these proteins, is now open for investigation.

500

References
Aarden EM, Burger EH, Nijweide PJ (1994) Function of osteocytes in bone. J Cell Biochem 55:287299 Arai N, Ohya K, Ogura H (1993) Osteopontin mRNA expression during bone resorption: an in situ hybridization study of induced ectopic bone in the rat. Bone Miner 22:129145 Blanger LF (1969) Osteocytic osteolysis. Calcif Tissue Res 4:112 Bianco P, Hayashi Y, Silvestrini G, Termine JD, Bonucci E (1985) Osteonectin and GLA-protein in calf bone: ultrastructural immunohistochemical localization using the protein A-gold method. Calcif Tissue Int 37:684686 Bianco P, Silvestrini G, Termine JD, Bonucci E (1988) Immunohistochemical localization of osteonectin in developing human and calf bone using monoclonal antibodies. Calcif Tissue Int 43:155161 Boivin G, Morel G, Lian JB, Anthoine-Terrier C, Dubois PM, Meunier PJ (1990) Localization of endogenous osteocalcin in neonatal rat bone and its absence in articular cartilage: effect of warfarin treatment. Virchows Archiv 417:505512 Bonucci E (1990) The ultrastructure of the osteocyte. In: Bonucci E, Motta PM (eds) Ultrastructure of skeletal tissues. Kluwer, Amsterdam, pp 223237 Boyde A (1980) Evidence against osteocytic osteolysis. Bone 2S:239255 Bronckers ALJJ, Gay S, Dimuzio MT, Butler WT (1985) Immunolocalization of gamma-carboxyglutamic acid containing proteins in developing rat bones. Collagen Rel Res 5:273281 Chen J, Zhang Q, McCulloch CAG, Sodek J (1991) Immunohistochemical localization of bone sialoprotein in fetal porcine bone tissues: comparisons with secreted phosphoprotein 1 (SPP-1, osteopontin) and SPARC (osteonectin). Histochem J 23:281289 Chen J, McCulloch CAG, Sodek J (1993a) Bone sialoprotein in developing porcine dental tissues: cellular expression and comparison of tissue localization with osteopontin and osteonectin. Arch Oral Biol 38:241249 Chen J, Singh K, Mukherjee BB, Sodek J (1993b) Developmental expression of osteopontin (OPN) mRNA in rat tissues: evidence for a role for OPN in bone formation and resorption. Matrix 13:113123 Chen J, McKee MD, Nanci A, Sodek J (1994) Bone sialoprotein mRNA expression and ultrastructural localization in fetal porcine calvarial bone: comparisons with osteopontin. Biochem J 26:6778 Chen Y, Bal BS, Gorski JP (1992) Calcium and collagen binding properties of osteopontin, bone sialoprotein, and bone acidic glycoprotein-75 from bone. J Biol chem 267:2487124878 Gerstenfeld LC, Gotoh Y, McKee MD, Nanci A, Landis WJ, Glimcher MJ (1990) Expression and ultrastructural immunolocalization of a major 66 kDa phosphoprotein synthesized by chicken osteoblasts during mineralization in vitro. Anat Rec 228:93103 Gotoh Y, Gerstenfeld LC, Glimcher MJ (1990) Identification and characterization of the major chicken bone phosphoprotein. Analysis of its synthesis by cultured embryonic chick osteoblasts. Eur J Biochem 187:4958 Groot CG, Danes JK, Blok J, Hoogendijk A, Hauschka PV (1986) Light and electron microscopic demonstration of osteocalcin antigenicity in embryonic and adult rat bone. Bone 7:379385 Grzesik WJ, Gehron Robey P (1994) Bone matrix RGD glycoproteins: immunolocalization and interaction with human primary osteoblastic bone cells in vitro. J Bone Miner Res 9:487496 Hauschka PV, Lian JB, Gallop PM (1975) Direct identification of the calcium-binding amino acid, gamma-carboxyglutamate, in mineralized tissue. Proc Natl Acad Sci USA 72:39253929 Hauschka PV, Frenkel J, DeMuth R, Gundberg CM (1983) Presence of osteocalcin and related higher molecular weight 4-carboxyglutamic acid-containing proteins in developing bone. J Biol Chem 258:176180

Hefley TJ (1987) Utilization of FPLC-purified bacterial collagenase for the isolation of cells from bone. J Bone Miner Res 2:502516 Hynes RO, Yamada KM (1982) Fibronectins: multifunctional modular glycoproteins. J Cell Biol 95:369377 Ikeda T, Nomura S, Yamaguchi A, Suda T, Yoshiki S (1992) In situ hybridization of bone matrix proteins in undecalcified adult rat bone sections. J Histochem Cytochem 40:10791088 Ingram RT, Clarke BL, Fisher LW, Fitzpatrick LA (1993) Distribution of noncollagenous proteins in the matrix of adult human bone: evidence of anatomic and functional heterogeneity. J Bone Miner Res 8:10191029 Kasai R, Bianco P, Gehron Robey P, Kahn AJ (1994) Production and characterization of an antibody against the human bone GLA protein (BGP/osteocalcin) propeptide and its use in immunocytochemistry of bone cells. Bone Miner 25:167182 Klein Nulend J, Plas A van der, Semeins CM, Ajubi NE, Frangos JA, Nijweide PJ, Burger EH (1995) Sensitivity of osteocytes to biomechanical stress in vitro. FASEB J 9:441445 Lanyon LE (1993) Osteocytes, strain detection, bone modeling and remodeling. Calcif Tissue Int 53(S1):S102S107 Majmudar G, Bole D, Goldstein SA, Bonadio J (1991) Bone cell culture in a three-dimensional polymer bead stabilizes the differentiated phenotype and provides evidence that osteoblastic cells synthesize type III collagen and fibronectin. J Bone Miner Res 6:869881 Mark MP, Prince CW, Oosawa T, Gay S, Bronckers ALJJ, Butler WT (1987) Immunohistochemical demonstration of a 44 kDa phosphoprotein in developing rat bones. J Histochem Cytochem 35:707715 Mark MP, Butler WT, Prince CW, Finkelman RD, Ruch J-V (1988) Developmental expression of 44-kDa bone phosphoprotein (osteospontin) and bone gamma-carboxyglutamic acid (Gla)-containing protein (osteocalcin) in calcifying tissues of rat. Differentiation 37:123136 Marotti G, Cane V, Palazzini S, Palumbo C (1990) Structure-function relationships in the osteocyte. Ital J Miner Electrolyte Metab 4:93106 McKee MD, Glimcher MJ, Nanci A (1992) High-resolution immunolocalization of osteopontin and osteocalcin in bone and cartilage during endochondral ossification in the chicken tibia. Anat Rec 234:479492 McKee MD, Farach-Carson MC, Butler WT, Hauschka PV, Nanci A (1993) Ultrastructural immunolocalization of noncollagenous (osteopontin and osteocalcin) and plasma (albumin and 2HS-glycoprotein) proteins in rat bone. J Bone Miner Res 8:485496 Metsranta M, Young MF, Sandberg M, Termine J, Vuorio E (1989) Localization of osteonectin expression in human skeletal tissues by in situ hybridization. Calcif Tissue Int 45:146152 Mikuni-Takagaki Y, Kakai Y, Satoyoshi M, Kawano E, Suzuki Y, Kawase T, Saito S (1995) Matrix mineralization and the differentiation of osteocyte-like cells in culture. J Bone Miner Res 10:231242 Nijweide PJ, Mulder RJP (1986) Identification of osteocytes in osteoblast-like cell cultures using a monoclonal antibody specifically directed against osteocytes. Histochemistry 84:342347 Oldberg , Franzn A, Heinegrd D (1986) Cloning and sequence analysis of rat bone sialoprotein (osteopontin) cDNA reveals an Arg-Gly-Asp cell-binding sequence. Proc Natl Acad Sci USA 83:88198823 Pacifici M, Oshima O, Fisher LW, Young MF, Shapiro IM, Leboy PS (1990) Changes in osteonectin distribution and levels are associated with mineralization of the chicken tibial growth cartilage. Calcif Tissue Int 47:5161 Parfitt AM (1977) The cellular basis of bone turnover and bone loss. Clin Orhtop 127:236247 Plas A van der, Nijweide PJ (1992) Isolation and purification of osteocytes. J Bone Miner Res 7:389396 Plas A van der, Aarden EM, Feijen JHM, Boer AH de, Wiltink A, Alblas MJ, Leij L de, Nijweide PJ (1994) Characteristics and

501 properties of osteocytes in culture. J Bone Miner Res 9:16971704 Price PA, Otsuka AS, Poser JW, Kristaponis J, Raman N (1976) Characterization of a gamma-carboxyglutamic acid-containing protein from bone. Proc Natl Acad Sci USA 73:14471451 Roach HI (1992) Trans-differentiation of hypertrophic chondrocytes into cells capable of producing a mineralized bone matrix. Bone Miner 19:120 Roach HI (1994) Why does bone matrix contain non-collagenous proteins? The possible roles of osteocalcin, osteonectin, osteopontin and bone sialoprotein in bone mineralisation and resorption. Cell Biol Int 18:617628 Rubin GT, McLeod KJ, Bain SD (1990) Functional strains and cortical bone adaptation: epigenetic assurance of skeletal integrity. J Biomech 23:4354 Sodek J, Chen J, Kasugai S, Nagata T, Zhang Q, Shapiro HS, Wrana JL, Goldberg HA (1992) Sialoproteins in bone remodelling. In: Davidovitch Z (ed) The biological mechanisms of tooth movement and craniofacial adaptation. The Ohio State University, College of Dentistry, Columbus; pp 127136 Sun Y-Q, McLeod KJ, Rubin CT (1995) Mechanically induced periosteal bone formation is paralleled by the upregulation of collagen type one mRNA in osteocytes as measured by in situ reverse transcript-polymerase chain reaction. Calcif Tissue Int 57:456462 Termine JD, Kleinman HK, Whitson SW, Conn KM, McGarvey ML, Martin GR (1981) Osteonectin, a bone-specific protein linking mineral to collagen. Cell 26:99105 Vermeulen AHM, Vermeer C, Bosman FT (1989) Histochemical detection of osteocalcin in normal and pathological bone. J Histochem Cytochem 37:15031508 Weiss RE, Reddi AH (1981) Appearance of fibronectin during the differentiation of cartilage, bone, and bone marrow. J Cell Biol 88:630636 Yamada KM (1983) Cell surface interactions with extracellular materials. Annu Rev Biochem 52:761799