- Email: [email protected]
Expression of cartilage-derived retinoic acid-sensitive protein during healing of the rat tooth-extraction socket
Archives of Oral Biology 44 (1999) 751±757
www.elsevier.com/locate/archoralbio
Expression of cartilage-derived retinoic acid-sensitive protein during healing of the rat tooth-extraction socket Y.C. Shyng a, H. Devlin a,*, D. Riccardi b, P. Sloan a a
The Department of Dental Medicine and Surgery, University Dental Hospital of Manchester, Higher Cambridge St., Manchester, M15 6FH, UK b The School of Biological Science, Stopford Building, University of Manchester, Manchester, UK Accepted 30 March 1999
Abstract Cartilage-derived retinoic acid sensitive protein (CD-RAP) is a recently described, cartilage-speci®c protein. During early healing of the tooth-extraction socket, cells express both chondrogenic and osteogenic cell markers, but no cartilage is formed. Cartilaginous collagen type II protein, a major constituent of hyaline cartilage, has not been detected in the healing socket, although type IX collagen, which coats these ®bres, has been detected transiently in early socket healing. This study investigated the spatial and temporal expression of CD-RAP and various osteoblast cell markers, i.e., alkaline phosphatase, osteopontin, osteonectin and osteocalcin, during healing. Immunolocalization of these proteins was determined in the rat tooth socket at 3, 4, 5, 6, 7, 8, 10 and 14 days after extraction. CD-RAP was expressed by preosteoblast cells maximally at 6, 7, and 8 days after extraction. Fully dierentiated osteoblasts expressed osteocalcin, a speci®c osteoblast marker. Preosteoblasts and ®broblasts did not express osteocalcin. On double immuno¯uorescent staining, some preosteoblasts coexpressed CD-RAP (indicative of chondrogenic dierentiation), and either alkaline phosphatase or osteopontin (markers of osteogenic stem-cell maturation). There was no colocalization between osteopontin and osteonectin. CD-RAP was unique amongst the cell markers used in that it was expressed by preosteoblasts, but not by osteoblasts lining the newly formed trabeculae. CD-RAP may have an important role in osteoblast cell dierentiation during bone healing. # 1999 Elsevier Science Ltd. All rights reserved. Keywords: Osteoblast; Dierentiation; Chondrogenic
1. Introduction CD-RAP is a recently described protein whose mRNA expression was isolated from cartilaginous tis-
Abbreviations: BSA, bovine serum albumin; CD-RAP, cartilage-derived retinoic acid-sensitive protein; NIH, National Institute of Health; USA; HS PBS, high-salt phosphate-buffered saline; SDS, sodium dodecyl sulphate; PBS, phosphatebuered saline. * Corresponding author. Fax: 0161-275-7822.
sue, and which was colocalized with a probe for type IIA procollagen mRNA (Dietz and Sandell, 1996). CD-RAP expression is restricted to cartilage in the adult animal (Bosserho et al., 1997) and is not present in fully dierentiated osteoblasts. No cartilage or chondrocytes are evident in the healing tooth-extraction socket of the rat (Devlin et al., 1997), but the dierentiating progenitor cells express both chondrogenic and osteogenic cell markers. Procollagen type II mRNA, normally expressed in hyaline cartilage, is detectable in socket preosteoblasts (Devlin et al., 1995). Collagen type IX, which normally
0003-9969/99/$ - see front matter # 1999 Elsevier Science Ltd. All rights reserved. PII: S 0 0 0 3 - 9 9 6 9 ( 9 9 ) 0 0 0 6 2 - X
752
Y.C. Shyng et al. / Archives of Oral Biology 44 (1999) 751±757
coats the collagen type II ®bre, has also been detected, but only in the early bone-formation stages of extraction socket healing (Jahangiri et al., 1998). This raises the intriguing possibility that proteins associated with cartilaginous dierentiation may be expressed transiently during early intramembranous healing. Using the rat tooth-extraction socket as a model allows the unique study of the dierentiation pathway of osteoprogenitor cells, in which cells committed fully to the chondrogenic pathway are not present. In early socket healing, osteogenic stem-cells may be capable of dierentiation along either an endochondral or intramembranous pathway depending on interactions with the extracellular matrix and growth factors, and may therefore express both cartilage and osteoblast markers. Osteo-progenitor cells populating the extraction socket may be derived from the remnants of periodontal ligament cells, pericytes or marrow-derived cells from the surrounding residual ridge. Osteoblast cells can express type I collagen, osteonectin, osteopontin, proteoglycan I and II, bone sialoprotein, osteocalcin, alkaline phosphatase and other proteins (Rodan and Noda, 1991). However, the most speci®c of the osteoblast markers, such as bone sialoprotein, are not expressed by a large proportion of osteoblasts during active bone deposition (Bianco et al., 1993). Alkaline phosphatase is expressed early, and osteopontin and osteocalcin expressions appear late during osteoblast dierentiation; therefore, in our experiments, the temporal expression of these markers was compared with CD-RAP. Our aim was to determine whether ``chondrocytespeci®c'' CD-RAP protein expression was evident in the rat extraction socket and, secondly, to compare its temporal expression with other markers of osteogenic stem-cell maturation. Alkaline phosphatase, osteonectin and osteopontin are present in osteogenic stem cells during development of rat endochondral and membranous skeletal tissues. They are expressed sequentially with osteogenic dierentiation in vivo (Nakase et al., 1994) and in vitro culture systems (Strauss et al., 1990). 2. Materials and methods Sixteen 29 day-old male Sprague±Dawley rats were used. Under general anaesthesia with Hypnorm1 intramuscular (Janssen Pharmaceutical Ltd, Oxford, UK), the upper right maxillary molars were extracted. The rats were killed humanely in pairs 3, 4, 5, 6, 7, 8, 10 and 14 days after the extraction; their craniomaxillary tissues were recovered and ®xed in formalin. The tissues were demineralized in EDTA (pH 7.0), dehydrated in graded alcohol solutions and embedded in paran wax. The second maxillary molar sockets were
serially sectioned in the coronal plane and stained using immunohistochemical techniques. 2.1. Immunoperoxidase technique A polyclonal osteonectin antibody (diluted 1:200), a monoclonal osteopontin antibody (diluted 1:100), a polyclonal rabbit CD-RAP antibody (diluted 1:400) and a monoclonal alkaline phosphatase (diluted 1:100) were used to detect these antigens in paran sections of the tooth socket. In addition, a polyclonal rabbit osteocalcin antibody (diluted 1:200) was used, a generous gift of Dr D. Modrowski (Institut National de la Sante et de la Recherche MeÂdicale, Paris, France). The CD-RAP antibody (1881) was a generous gift from the Roche/Boehringer Mannheim company (Penzberg, Germany). This antibody is chondrocyte-speci®c (Bosserho et al., 1997). The osteonectin antibody (LF-23) was raised in the rabbit and was generously donated by Dr L. Fisher, NIH (see Wewer et al., 1988). The alkaline phosphatase (B4-50) and osteopontin MPIIIB10(1) mouse monoclonal antibodies were developed by Lawson et al. (1985) and Gorski et al. (1990), respectively. They were both obtained from the Developmental Studies Hybridoma Bank, maintained by the University of Iowa, Department of Biological Sciences, Iowa City, IA 52242, under contract NO1HD-7-3263 from the National Institute of Child Health and Development. The sections were dewaxed in xylene and hydrated in graded alcohol solutions. Antigen was retrieved by using trypsin (diluted 1:20) for 30 min at 378C; the trypsin (Dako Ltd, Cambridge, UK) was derived from bovine pancreas (Code S2012, immunohistochemical grade; spec. act. 148 units/mg at 258C). Endogenous peroxidase activity was eliminated by incubation with 3% hydrogen peroxide for 15 min. The tissue sections were rinsed in PBS, treated with normal goat serum for 30 min, and then incubated with the primary antibody overnight at 48C. A biotinylated secondary antibody was applied for 15 min at room temperature and a DAKO LSAB1 2 kit, peroxidase (Dako Corp., Carpinteria, CA, USA) was used to produce a red precipitate at the antigen site. The sections were counterstained with haematoxylin for 5 sec and mounted with a water-based medium. Negative controls were performed by substituting normal goat serum in PBS for the primary antibodies. Positive controls with the CD-RAP antibody were performed by staining mature rat femur. These showed strong cytoplasmic expression in articular chondrocytes, with no expression in bone cells or soft tissue. 2.2. Double immuno¯uorescent staining On day 1, the sections were dewaxed, hydrated and trypsinized. They were then washed with PBS, 1%
Y.C. Shyng et al. / Archives of Oral Biology 44 (1999) 751±757
Fig. 1. Alkaline phosphatase was expressed throughout the 7day extraction socket (i.e., positively stained osteoprogenitor (Os), preosteoblast (P), and osteoblast (O) cells). Counterstained with haematoxylin. 243.
SDS in PBS for 5 min and again with PBS, all at room temperature. Auto¯uorescence was prevented by applying 100 mM glycine in 1% BSA/PBS for 10 min, two washes (5 min each) of HS PBS and a further two washes (5 min each) of PBS; 3% normal goat serum in 1% BSA/PBS was also applied to the sections for 30 min. The sections were incubated with the ®rst primary antibody (either osteopontin (1:50) or alkaline phosphatase (1:50) antibody) overnight in 48C. Then, after two washes of HS PBS and two of PBS (5 min each), the sections were incubated with the ®rst secondary antibody (goat antimouse antibody conjugated with Texas Red, diluted 1:100; Jackson Immunoresearch Ltd, USA). The incubation lasted for 1 h at room temperature. After washes of HS PBS and PBS, the sections were again blocked with 3% normal goat serum in 1% BSA/PBS for 30 min. The second primary antibody (CD-RAP, diluted 1:200) was applied and incubated overnight at 48C. On day 3, after washes of HS PBS and PBS, the sections were incubated with the second secondary antibody (goat antirabbit antibody conjugated with ¯uorescein isothiocyanate, 1:100) at room temperature for 1 h. The sections were washed with HS PBS and PBS as before and then mounted with Vectashield (Vector Ltd, CA, USA). The sections were observed under the immuno¯uorescent microscope with dierent ®lters to view the distribution of the two ¯uorescent secondary antibodies. The image was captured with a Zeiss Axioskop microscope using a cooled CCD camera and processed using Improvision Openlab and Adobe Photoshop software. Negative controls were performed by substituting normal goat serum in PBS for either primary or secondary antibodies.
753
Fig. 2. At 7 days after extraction, CD-RAP was not expressed by osteoblasts (O) lying against the bone trabeculae (counterstained with haematoxylin). Other preosteoblast (P) and osteoprogenitor (Os) cells stained positively (red stain). Counterstained with haematoxylin. 300.
3. Results 3.1. Alkaline phosphatase In 4-day-old extraction sockets, osteocytes, osteoblasts and many adjacent cells of ¯attened and more circular shape were stained positively. At 5, 6 and 7 days, preosteoblasts and osteoblasts were stained (Fig. 1). Thick collagen ®bres coursed into the socket between positively stained osteoblasts and preosteoblasts. These collagen bundles were inserted into the primary bone trabeculae. Small, circular osteoprogenitor cells surrounding the birefringent ®bres were stained positively for alkaline phosphatase. At 10 and 14 days after tooth extraction, staining was less obvious in the sockets, with a limited staining of a layer of osteoblasts against the trabeculae. 3.2. CD-RAP In 3-day-old sockets, CD-RAP did not stain the osteoblasts, ®broblasts or osteoclasts but stained the reversal lines in the bone. At 4, 5 and 6 days after extraction, small, positively stained, spherical osteoprogenitor cells surrounded the osteoblasts in the socket. Staining was at maximum intensity at 6 and 7 days (Fig. 2) and was less intense at 10 and 14 days after extraction. Many osteoblasts surrounding the newly formed trabeculae failed to express CD-RAP. CD-RAP was expressed in chondrocytes of the nasal cartilage and palatal suture. 3.3. Osteonectin At 4 days after extraction, preosteoblasts and osteoblasts lining the trabeculae and resorption cavities
754
Y.C. Shyng et al. / Archives of Oral Biology 44 (1999) 751±757
Fig. 3. At 7 days after tooth-extraction, osteoprogenitor cells throughout the socket (Os) expressed osteonectin (red stain). Counterstained with haematoxylin. 243.
were stained positively, but most of the osteoclasts were negatively stained. Cellular staining in the socket matrix was of maximum intensity at 6 days after extractions, but many ®broblasts, osteoblasts and osteocytes failed to stain. At time points after 7 days (Fig. 3), the socket had a declining number of positively stained osteoblast cells. Osteonectin was expressed in chondrocytes of the nasal cartilage and palatal suture. 3.4. Osteopontin In 3-day-old sockets, resorption was active, with many positively stained osteoclasts present. In 4-dayold sockets, ¯attened and rounded osteoblastic cells lining the new trabeculae of woven bone were strongly positively stained. About three layers of cells against the forming trabeculae were stained positively. The staining pattern at 5, 6 and 7 days (Fig. 4) after extraction was similar to that in 4-day-old sockets, but the
Fig. 4. A layer of osteoblasts strongly expressed osteopontin at 7 days after tooth-extraction. Red-stained osteoblasts (O). Counterstained with haematoxylin. 243.
Fig. 5. Osteopontin expression in socket 7 days after extraction (Texas Red-conjugated secondary antibody). 500.
positive cells lining the trabeculae had reduced in thickness to 1±2 layers. At 8 days after extraction, there was selective staining of some osteoblast and preosteoblast cells surrounding the trabeculae. At 10 and 14 days after extraction, the number of osteopontinpositive cells surrounding the trabeculae was much reduced. 3.4.1. Osteopontin and CD-RAP double immuno¯uorescent staining Some preosteoblast cells coexpressed CD-RAP (indicative of chondrogenic dierentiation) and osteopontin (a marker of osteogenic stem-cell maturation). Heterogeneous expression of either CD-RAP or osteopontin was seen in other morphologically identical cells (Figs. 5±8). When present, CD-RAP staining was expressed in the perinucleus and cytoplasm, but osteopontin was expressed only in the cytoplasm of cells.
Fig. 6. CD-RAP staining of same section of 7-day socket as Fig. 5 (¯uorescein isothiocyanate-conjugated secondary antibody). 500.
Y.C. Shyng et al. / Archives of Oral Biology 44 (1999) 751±757
755
4. Discussion
Fig. 7. Phase contrast of same tissue as shown in Figs. 5 and 6. Arrows indicate the border of the bone surface (T, bone trabeculum). 500.
3.4.2. Alkaline phosphatase and CD-RAP double immuno¯uorescent staining Some preosteoblasts expressed alkaline phosphatase and CD-RAP, whereas osteoblasts expressed only alkaline phosphatase. All controls were negative. 3.5. Osteocalcin At 6 and 8 days after tooth-extraction, osteocalcin was expressed maximally by fully dierentiated osteoblasts lying on the woven-bone trabecular surface (Figs. 9 and 10). None of the ®broblasts or preosteoblast cells in the surrounding ®brous connective tissue was stained.
Fig. 8. A heterogeneous mixture of cells expressing either osteopontin (red), CD-RAP (green) or both (orange); 7-day extraction socket (same tissue as shown in Figs. 5, 6 and 7). 500.
In the healing tooth-extraction socket, coarse collagen ®bres form a preliminary framework on which deposition of woven bone occurs (Devlin et al., 1997). No resting chondrocytes, hypertrophic chondrocytes or cartilaginous tissues are observed. Here preosteoblastic cells expressed CD-RAP from 4 days after extraction, a protein that has only previously been described in cartilage tissue. CD-RAP expression was thought to be restricted to cartilage in the adult animal (Bosserho et al., 1997). In the present study, preosteoblasts and immature osteoblasts surrounding the collagen ®bres strongly expressed alkaline phosphatase, osteonectin and osteopontin. Fully dierentiated osteoblasts expressed osteocalcin, which agreed with the in vitro ®ndings of Nefussi et al. (1997). During healing of the rat extraction socket in our experiments, the temporal and spatial distribution of CD-RAP more closely resembled that of osteonectin than osteopontin. In the nasal cartilage, chondrocytes expressed osteonectin and CD-RAP, but not osteopontin. Osteopontin is expressed at sites that become mineralized, such as the palatal suture and hypertrophic chondrocytes of the femoral epiphysis. Owen et al. (1990) observed sequential gene expression in cultures of fetal calvarial osteoblasts. In our experiments, strong expression of alkaline phosphatase and osteopontin was evident early in the healing process at 4 and 5 days after tooth-extraction. Mark et al. (1987) have shown that the young rat jaw bone contains many preosteoblasts that are immunopositive for osteopontin. The preosteoblast markers, osteonectin and CD-RAP, were expressed maximally in the socket at 6 or 7 days post-extraction. At later time points, all antigen expression became less apparent as healing progressed. The expression of osteopontin at 4±5 days after tooth-extraction is simultaneous with the onset of woven-bone formation (Hsieh et al., 1995), and may indicate a regulatory role for osteopontin in mineralization. The induction of osteopontin activity has recently been shown to be dependent on a minimal level of alkaline phosphatase activity (Beck et al., 1998), which might explain the similar temporal expression of these proteins in our experiments. In the present experiments, marked heterogeneity in the expression of osteoblast-associated markers was observed in morphologically identical cells. For example, some preosteoblasts expressed CD-RAP and osteopontin, whereas other similarly shaped cells expressed only one of these proteins. Rat calvaria cells grown in vitro express the same heterogeneity of osteoblast-associated marker proteins (Liu et al., 1994). Alternatively, the healing extraction socket may contain separate populations of chondrogenic and osteogenic precursors rather than bipotential progenitors for
756
Y.C. Shyng et al. / Archives of Oral Biology 44 (1999) 751±757
Fig. 9. Only fully dierentiated osteoblasts and osteocytes expressed osteocalcin (dark red). The area outlined by the black rectangle is shown at higher magni®cation in Fig. 10 (W, newly formed woven bone). 50.
bone and cartilage (osteochondroprogenitors; Hall, 1970). Membrane bones are certainly capable of forming cartilage, as seen in the repair of fractured rat parietal bones (Pritchard, 1946); and secondary cartilages are a normal feature of development of the mandible. Osteonectin and CD-RAP may function as growth-
regulating factors in wound-healing tissues. Lane and Sage (1990) have shown that osteonectin is a potent inhibitor of cell spreading and may locate the preosteoblasts to the bone-forming site. CD-RAP has been shown to inhibit tumour stem-cell colony formation, but whether it has any growth-regulatory function in the healing of bone tissues is unknown (Bogdahn et al., 1989). We show that CD-RAP is expressed transiently by a heterogeneous population of preosteoblasts in the healing tooth socket. It is surprising that ``chondrocytespeci®c'' proteins are expressed in the healing socket, as this tissue is derived from intramembranous ossi®cation. The function of these proteins is uncertain, but the presence of CD-RAP in cartilage and healing bone may indicate similarities at the molecular level between endochondral and intramembranous ossi®cation.
References
Fig. 10. Osteoblasts, but not preosteoblasts, expressed osteocalcin (P, osteoblasts). 500.
Beck Jr, G.R., Sullivan, E.C., Moran, E., Zerler, B., 1998. Relationship between alkaline phosphatase levels, osteopontin expression, and mineralization in dierentiating MC3T3-E1 osteoblasts. J. Cell Biochem 68, 269±280. Bianco, P., Riminucci, M., Bonucci, E., Termine, J.D., Robey, P.G., 1993. Bone sialoprotein (BSP) secretion and osteoblast dierentiation: relationship to bromodeoxyuridine incorporation, alkaline phosphatase, and matrix deposition. J. Histochem. Cytochem 41, 183±191. Bogdahn, U., Apfel, R., Hahn, M., Gerlach, M., Behl, C.,
Y.C. Shyng et al. / Archives of Oral Biology 44 (1999) 751±757 Hoppe, J., Martin, R., 1989. Autocrine tumour cell growth-inhibiting activities from human malignant melanoma. Cancer Res 49, 5358±5363. Bosserho, A.K., Kondo, S., Moser, M., Dietz, W., Copeland, N.G., Gilbert, D.J., Jenkins, N.A., Buettner, R., 1997. Mouse CD-RAP/MIA gene: structure, chromosomal localization, and expression in cartilage and chondrosarcoma. Dev. Dyn 208, 516±525. Devlin, H., Hoyland, J., Freemont, A.J., Sloan, P., 1995. Localization of procollagen type II mRNA and collagen type II in the healing tooth socket of the rat. Archs Oral Biol 40, 181±185. Devlin, H., Hoyland, J., Newall, J.F., Ayad, S., 1997. Trabecular bone formation in the healing of the rodent molar tooth-extraction socket. J. Bone Min. Res 12, 2061± 2067. Dietz, U.H., Sandell, L.J., 1996. Cloning of a novel retinoic acid-sensitive cDNA expressed in cartilage and during chondrogenesis. J. Biol. Chem 271, 3311±3316. Gorski, J.P., Grin, D., Dudley, G., Stanford, C., Thomas, R., Huang, C., Lai, E., Karr, B., Solursh, M., 1990. Bone acidic glycoprotein-75 is a major synthetic product of osteoblastic cells and localized as 75- and/or 50-kDa forms in mineralized phases of bone and growth plate and in serum. J. Biol. Chem 265, 14956±14963. Hall, B.K., 1970. Cellular dierentiation in skeletal tissues. Biol. Rev 45, 455±484. Hsieh, Y.D., Devlin, H., McCord, F., 1995. The eect of ovariectomy on the healing tooth socket of the rat. Archs Oral Biol 40, 529±531. Jahangiri, L., Devlin, H., Ting, K., Nishimura, I., 1998. Current perspectives in residual ridge remodeling and its clinical implications: A review. J. Prosthet. Dent 80, 224± 237. Lane, T., Sage, H., 1990. Functional mapping of SPARC: peptides from two distinct Ca2+-binding sites modulate cell shape. J. Cell Biol 111, 3065±3076. Lawson, G.M., Katzmann, J.A., Kimlinger, T.K., O'Brien, J.F., 1985. Isolation and preliminary characterization of a monoclonal antibody that interacts preferentially with the
757
liver isoenzyme of human alkaline phosphatase. Clin. Chem 31, 381±385. Liu, F., Malaval, L., Gupta, A.K., Aubin, J.E., 1994. Simultaneous detection of multiple bone-related mRNAs and protein expression during osteoblast dierentiation: polymerase chain reaction and immunocytochemical studies at the single cell level. Dev. Biol 166, 220±234. Mark, M.P., Prince, C.W., Oosawa, T., Gay, S., Bronckers, A.L.J.J., Butler, W., 1987. Immunohistochemical demonstration of a 44-kD phosphoprotein in developing rat bones. J. Histochem. Cytochem 35, 707±715. Nakase, T., Takaoka, K., Hirakawa, K., Hirota, S., Takemura, T., Onoue, H., 1994. Alterations in the expression of osteonectin, osteopontin and osteocalcin mRNAs during the development of skeletal tissues in vivo. Bone Min 26, 109±122. Nefussi, J.R., Brami, G., Modrowski, D., Oboeuf, M., Forest, N., 1997. Sequential expression of bone matrix proteins during rat calvaria osteoblast dierentiation and bone nodule formation in vitro. J. Histochem. Cytochem 45, 493±503. Owen, T.A., Aronow, M., Shalhoub, V., Barone, L.M., Wilming, L., Tassinari, M.S., Kennedy, M.B., Pockwinse, S., Lian, J.B., Stein, G.S., 1990. Progressive development of the rat osteoblast phenotype in vitro: reciprocal relationships in expression of genes associated with osteoblast proliferation and dierentiation during formation of the bone extracellular matrix. J. Cell Physiol 143, 420±430. Pritchard, J.J., 1946. Repair of fracture of the parietal bone in rats. J. Anat 80, 55±60. Rodan, G.A., Noda, M., 1991. Gene expression in osteoblastic cells. Crit. Rev. Eukaryotic Gene Expression 1 (2), 85±98. Strauss, P.G., Closs, E.I., Schmidt, J., Er¯e, V., 1990. Gene expression during osteogenic dierentiation in mandibular condyles in vitro. J. Cell Biol 110, 1369±1378. Wewer, U.M., Albrechtsen, R., Fisher, L.W., Young, M.F., Termine, J.T., 1988. Dierential deposition of osteonectin in basement membranes of human tissues undergoing profound changes in tissue architecture. Am. J. Path 132, 345±355.
www.elsevier.com/locate/archoralbio
Expression of cartilage-derived retinoic acid-sensitive protein during healing of the rat tooth-extraction socket Y.C. Shyng a, H. Devlin a,*, D. Riccardi b, P. Sloan a a
The Department of Dental Medicine and Surgery, University Dental Hospital of Manchester, Higher Cambridge St., Manchester, M15 6FH, UK b The School of Biological Science, Stopford Building, University of Manchester, Manchester, UK Accepted 30 March 1999
Abstract Cartilage-derived retinoic acid sensitive protein (CD-RAP) is a recently described, cartilage-speci®c protein. During early healing of the tooth-extraction socket, cells express both chondrogenic and osteogenic cell markers, but no cartilage is formed. Cartilaginous collagen type II protein, a major constituent of hyaline cartilage, has not been detected in the healing socket, although type IX collagen, which coats these ®bres, has been detected transiently in early socket healing. This study investigated the spatial and temporal expression of CD-RAP and various osteoblast cell markers, i.e., alkaline phosphatase, osteopontin, osteonectin and osteocalcin, during healing. Immunolocalization of these proteins was determined in the rat tooth socket at 3, 4, 5, 6, 7, 8, 10 and 14 days after extraction. CD-RAP was expressed by preosteoblast cells maximally at 6, 7, and 8 days after extraction. Fully dierentiated osteoblasts expressed osteocalcin, a speci®c osteoblast marker. Preosteoblasts and ®broblasts did not express osteocalcin. On double immuno¯uorescent staining, some preosteoblasts coexpressed CD-RAP (indicative of chondrogenic dierentiation), and either alkaline phosphatase or osteopontin (markers of osteogenic stem-cell maturation). There was no colocalization between osteopontin and osteonectin. CD-RAP was unique amongst the cell markers used in that it was expressed by preosteoblasts, but not by osteoblasts lining the newly formed trabeculae. CD-RAP may have an important role in osteoblast cell dierentiation during bone healing. # 1999 Elsevier Science Ltd. All rights reserved. Keywords: Osteoblast; Dierentiation; Chondrogenic
1. Introduction CD-RAP is a recently described protein whose mRNA expression was isolated from cartilaginous tis-
Abbreviations: BSA, bovine serum albumin; CD-RAP, cartilage-derived retinoic acid-sensitive protein; NIH, National Institute of Health; USA; HS PBS, high-salt phosphate-buffered saline; SDS, sodium dodecyl sulphate; PBS, phosphatebuered saline. * Corresponding author. Fax: 0161-275-7822.
sue, and which was colocalized with a probe for type IIA procollagen mRNA (Dietz and Sandell, 1996). CD-RAP expression is restricted to cartilage in the adult animal (Bosserho et al., 1997) and is not present in fully dierentiated osteoblasts. No cartilage or chondrocytes are evident in the healing tooth-extraction socket of the rat (Devlin et al., 1997), but the dierentiating progenitor cells express both chondrogenic and osteogenic cell markers. Procollagen type II mRNA, normally expressed in hyaline cartilage, is detectable in socket preosteoblasts (Devlin et al., 1995). Collagen type IX, which normally
0003-9969/99/$ - see front matter # 1999 Elsevier Science Ltd. All rights reserved. PII: S 0 0 0 3 - 9 9 6 9 ( 9 9 ) 0 0 0 6 2 - X
752
Y.C. Shyng et al. / Archives of Oral Biology 44 (1999) 751±757
coats the collagen type II ®bre, has also been detected, but only in the early bone-formation stages of extraction socket healing (Jahangiri et al., 1998). This raises the intriguing possibility that proteins associated with cartilaginous dierentiation may be expressed transiently during early intramembranous healing. Using the rat tooth-extraction socket as a model allows the unique study of the dierentiation pathway of osteoprogenitor cells, in which cells committed fully to the chondrogenic pathway are not present. In early socket healing, osteogenic stem-cells may be capable of dierentiation along either an endochondral or intramembranous pathway depending on interactions with the extracellular matrix and growth factors, and may therefore express both cartilage and osteoblast markers. Osteo-progenitor cells populating the extraction socket may be derived from the remnants of periodontal ligament cells, pericytes or marrow-derived cells from the surrounding residual ridge. Osteoblast cells can express type I collagen, osteonectin, osteopontin, proteoglycan I and II, bone sialoprotein, osteocalcin, alkaline phosphatase and other proteins (Rodan and Noda, 1991). However, the most speci®c of the osteoblast markers, such as bone sialoprotein, are not expressed by a large proportion of osteoblasts during active bone deposition (Bianco et al., 1993). Alkaline phosphatase is expressed early, and osteopontin and osteocalcin expressions appear late during osteoblast dierentiation; therefore, in our experiments, the temporal expression of these markers was compared with CD-RAP. Our aim was to determine whether ``chondrocytespeci®c'' CD-RAP protein expression was evident in the rat extraction socket and, secondly, to compare its temporal expression with other markers of osteogenic stem-cell maturation. Alkaline phosphatase, osteonectin and osteopontin are present in osteogenic stem cells during development of rat endochondral and membranous skeletal tissues. They are expressed sequentially with osteogenic dierentiation in vivo (Nakase et al., 1994) and in vitro culture systems (Strauss et al., 1990). 2. Materials and methods Sixteen 29 day-old male Sprague±Dawley rats were used. Under general anaesthesia with Hypnorm1 intramuscular (Janssen Pharmaceutical Ltd, Oxford, UK), the upper right maxillary molars were extracted. The rats were killed humanely in pairs 3, 4, 5, 6, 7, 8, 10 and 14 days after the extraction; their craniomaxillary tissues were recovered and ®xed in formalin. The tissues were demineralized in EDTA (pH 7.0), dehydrated in graded alcohol solutions and embedded in paran wax. The second maxillary molar sockets were
serially sectioned in the coronal plane and stained using immunohistochemical techniques. 2.1. Immunoperoxidase technique A polyclonal osteonectin antibody (diluted 1:200), a monoclonal osteopontin antibody (diluted 1:100), a polyclonal rabbit CD-RAP antibody (diluted 1:400) and a monoclonal alkaline phosphatase (diluted 1:100) were used to detect these antigens in paran sections of the tooth socket. In addition, a polyclonal rabbit osteocalcin antibody (diluted 1:200) was used, a generous gift of Dr D. Modrowski (Institut National de la Sante et de la Recherche MeÂdicale, Paris, France). The CD-RAP antibody (1881) was a generous gift from the Roche/Boehringer Mannheim company (Penzberg, Germany). This antibody is chondrocyte-speci®c (Bosserho et al., 1997). The osteonectin antibody (LF-23) was raised in the rabbit and was generously donated by Dr L. Fisher, NIH (see Wewer et al., 1988). The alkaline phosphatase (B4-50) and osteopontin MPIIIB10(1) mouse monoclonal antibodies were developed by Lawson et al. (1985) and Gorski et al. (1990), respectively. They were both obtained from the Developmental Studies Hybridoma Bank, maintained by the University of Iowa, Department of Biological Sciences, Iowa City, IA 52242, under contract NO1HD-7-3263 from the National Institute of Child Health and Development. The sections were dewaxed in xylene and hydrated in graded alcohol solutions. Antigen was retrieved by using trypsin (diluted 1:20) for 30 min at 378C; the trypsin (Dako Ltd, Cambridge, UK) was derived from bovine pancreas (Code S2012, immunohistochemical grade; spec. act. 148 units/mg at 258C). Endogenous peroxidase activity was eliminated by incubation with 3% hydrogen peroxide for 15 min. The tissue sections were rinsed in PBS, treated with normal goat serum for 30 min, and then incubated with the primary antibody overnight at 48C. A biotinylated secondary antibody was applied for 15 min at room temperature and a DAKO LSAB1 2 kit, peroxidase (Dako Corp., Carpinteria, CA, USA) was used to produce a red precipitate at the antigen site. The sections were counterstained with haematoxylin for 5 sec and mounted with a water-based medium. Negative controls were performed by substituting normal goat serum in PBS for the primary antibodies. Positive controls with the CD-RAP antibody were performed by staining mature rat femur. These showed strong cytoplasmic expression in articular chondrocytes, with no expression in bone cells or soft tissue. 2.2. Double immuno¯uorescent staining On day 1, the sections were dewaxed, hydrated and trypsinized. They were then washed with PBS, 1%
Y.C. Shyng et al. / Archives of Oral Biology 44 (1999) 751±757
Fig. 1. Alkaline phosphatase was expressed throughout the 7day extraction socket (i.e., positively stained osteoprogenitor (Os), preosteoblast (P), and osteoblast (O) cells). Counterstained with haematoxylin. 243.
SDS in PBS for 5 min and again with PBS, all at room temperature. Auto¯uorescence was prevented by applying 100 mM glycine in 1% BSA/PBS for 10 min, two washes (5 min each) of HS PBS and a further two washes (5 min each) of PBS; 3% normal goat serum in 1% BSA/PBS was also applied to the sections for 30 min. The sections were incubated with the ®rst primary antibody (either osteopontin (1:50) or alkaline phosphatase (1:50) antibody) overnight in 48C. Then, after two washes of HS PBS and two of PBS (5 min each), the sections were incubated with the ®rst secondary antibody (goat antimouse antibody conjugated with Texas Red, diluted 1:100; Jackson Immunoresearch Ltd, USA). The incubation lasted for 1 h at room temperature. After washes of HS PBS and PBS, the sections were again blocked with 3% normal goat serum in 1% BSA/PBS for 30 min. The second primary antibody (CD-RAP, diluted 1:200) was applied and incubated overnight at 48C. On day 3, after washes of HS PBS and PBS, the sections were incubated with the second secondary antibody (goat antirabbit antibody conjugated with ¯uorescein isothiocyanate, 1:100) at room temperature for 1 h. The sections were washed with HS PBS and PBS as before and then mounted with Vectashield (Vector Ltd, CA, USA). The sections were observed under the immuno¯uorescent microscope with dierent ®lters to view the distribution of the two ¯uorescent secondary antibodies. The image was captured with a Zeiss Axioskop microscope using a cooled CCD camera and processed using Improvision Openlab and Adobe Photoshop software. Negative controls were performed by substituting normal goat serum in PBS for either primary or secondary antibodies.
753
Fig. 2. At 7 days after extraction, CD-RAP was not expressed by osteoblasts (O) lying against the bone trabeculae (counterstained with haematoxylin). Other preosteoblast (P) and osteoprogenitor (Os) cells stained positively (red stain). Counterstained with haematoxylin. 300.
3. Results 3.1. Alkaline phosphatase In 4-day-old extraction sockets, osteocytes, osteoblasts and many adjacent cells of ¯attened and more circular shape were stained positively. At 5, 6 and 7 days, preosteoblasts and osteoblasts were stained (Fig. 1). Thick collagen ®bres coursed into the socket between positively stained osteoblasts and preosteoblasts. These collagen bundles were inserted into the primary bone trabeculae. Small, circular osteoprogenitor cells surrounding the birefringent ®bres were stained positively for alkaline phosphatase. At 10 and 14 days after tooth extraction, staining was less obvious in the sockets, with a limited staining of a layer of osteoblasts against the trabeculae. 3.2. CD-RAP In 3-day-old sockets, CD-RAP did not stain the osteoblasts, ®broblasts or osteoclasts but stained the reversal lines in the bone. At 4, 5 and 6 days after extraction, small, positively stained, spherical osteoprogenitor cells surrounded the osteoblasts in the socket. Staining was at maximum intensity at 6 and 7 days (Fig. 2) and was less intense at 10 and 14 days after extraction. Many osteoblasts surrounding the newly formed trabeculae failed to express CD-RAP. CD-RAP was expressed in chondrocytes of the nasal cartilage and palatal suture. 3.3. Osteonectin At 4 days after extraction, preosteoblasts and osteoblasts lining the trabeculae and resorption cavities
754
Y.C. Shyng et al. / Archives of Oral Biology 44 (1999) 751±757
Fig. 3. At 7 days after tooth-extraction, osteoprogenitor cells throughout the socket (Os) expressed osteonectin (red stain). Counterstained with haematoxylin. 243.
were stained positively, but most of the osteoclasts were negatively stained. Cellular staining in the socket matrix was of maximum intensity at 6 days after extractions, but many ®broblasts, osteoblasts and osteocytes failed to stain. At time points after 7 days (Fig. 3), the socket had a declining number of positively stained osteoblast cells. Osteonectin was expressed in chondrocytes of the nasal cartilage and palatal suture. 3.4. Osteopontin In 3-day-old sockets, resorption was active, with many positively stained osteoclasts present. In 4-dayold sockets, ¯attened and rounded osteoblastic cells lining the new trabeculae of woven bone were strongly positively stained. About three layers of cells against the forming trabeculae were stained positively. The staining pattern at 5, 6 and 7 days (Fig. 4) after extraction was similar to that in 4-day-old sockets, but the
Fig. 4. A layer of osteoblasts strongly expressed osteopontin at 7 days after tooth-extraction. Red-stained osteoblasts (O). Counterstained with haematoxylin. 243.
Fig. 5. Osteopontin expression in socket 7 days after extraction (Texas Red-conjugated secondary antibody). 500.
positive cells lining the trabeculae had reduced in thickness to 1±2 layers. At 8 days after extraction, there was selective staining of some osteoblast and preosteoblast cells surrounding the trabeculae. At 10 and 14 days after extraction, the number of osteopontinpositive cells surrounding the trabeculae was much reduced. 3.4.1. Osteopontin and CD-RAP double immuno¯uorescent staining Some preosteoblast cells coexpressed CD-RAP (indicative of chondrogenic dierentiation) and osteopontin (a marker of osteogenic stem-cell maturation). Heterogeneous expression of either CD-RAP or osteopontin was seen in other morphologically identical cells (Figs. 5±8). When present, CD-RAP staining was expressed in the perinucleus and cytoplasm, but osteopontin was expressed only in the cytoplasm of cells.
Fig. 6. CD-RAP staining of same section of 7-day socket as Fig. 5 (¯uorescein isothiocyanate-conjugated secondary antibody). 500.
Y.C. Shyng et al. / Archives of Oral Biology 44 (1999) 751±757
755
4. Discussion
Fig. 7. Phase contrast of same tissue as shown in Figs. 5 and 6. Arrows indicate the border of the bone surface (T, bone trabeculum). 500.
3.4.2. Alkaline phosphatase and CD-RAP double immuno¯uorescent staining Some preosteoblasts expressed alkaline phosphatase and CD-RAP, whereas osteoblasts expressed only alkaline phosphatase. All controls were negative. 3.5. Osteocalcin At 6 and 8 days after tooth-extraction, osteocalcin was expressed maximally by fully dierentiated osteoblasts lying on the woven-bone trabecular surface (Figs. 9 and 10). None of the ®broblasts or preosteoblast cells in the surrounding ®brous connective tissue was stained.
Fig. 8. A heterogeneous mixture of cells expressing either osteopontin (red), CD-RAP (green) or both (orange); 7-day extraction socket (same tissue as shown in Figs. 5, 6 and 7). 500.
In the healing tooth-extraction socket, coarse collagen ®bres form a preliminary framework on which deposition of woven bone occurs (Devlin et al., 1997). No resting chondrocytes, hypertrophic chondrocytes or cartilaginous tissues are observed. Here preosteoblastic cells expressed CD-RAP from 4 days after extraction, a protein that has only previously been described in cartilage tissue. CD-RAP expression was thought to be restricted to cartilage in the adult animal (Bosserho et al., 1997). In the present study, preosteoblasts and immature osteoblasts surrounding the collagen ®bres strongly expressed alkaline phosphatase, osteonectin and osteopontin. Fully dierentiated osteoblasts expressed osteocalcin, which agreed with the in vitro ®ndings of Nefussi et al. (1997). During healing of the rat extraction socket in our experiments, the temporal and spatial distribution of CD-RAP more closely resembled that of osteonectin than osteopontin. In the nasal cartilage, chondrocytes expressed osteonectin and CD-RAP, but not osteopontin. Osteopontin is expressed at sites that become mineralized, such as the palatal suture and hypertrophic chondrocytes of the femoral epiphysis. Owen et al. (1990) observed sequential gene expression in cultures of fetal calvarial osteoblasts. In our experiments, strong expression of alkaline phosphatase and osteopontin was evident early in the healing process at 4 and 5 days after tooth-extraction. Mark et al. (1987) have shown that the young rat jaw bone contains many preosteoblasts that are immunopositive for osteopontin. The preosteoblast markers, osteonectin and CD-RAP, were expressed maximally in the socket at 6 or 7 days post-extraction. At later time points, all antigen expression became less apparent as healing progressed. The expression of osteopontin at 4±5 days after tooth-extraction is simultaneous with the onset of woven-bone formation (Hsieh et al., 1995), and may indicate a regulatory role for osteopontin in mineralization. The induction of osteopontin activity has recently been shown to be dependent on a minimal level of alkaline phosphatase activity (Beck et al., 1998), which might explain the similar temporal expression of these proteins in our experiments. In the present experiments, marked heterogeneity in the expression of osteoblast-associated markers was observed in morphologically identical cells. For example, some preosteoblasts expressed CD-RAP and osteopontin, whereas other similarly shaped cells expressed only one of these proteins. Rat calvaria cells grown in vitro express the same heterogeneity of osteoblast-associated marker proteins (Liu et al., 1994). Alternatively, the healing extraction socket may contain separate populations of chondrogenic and osteogenic precursors rather than bipotential progenitors for
756
Y.C. Shyng et al. / Archives of Oral Biology 44 (1999) 751±757
Fig. 9. Only fully dierentiated osteoblasts and osteocytes expressed osteocalcin (dark red). The area outlined by the black rectangle is shown at higher magni®cation in Fig. 10 (W, newly formed woven bone). 50.
bone and cartilage (osteochondroprogenitors; Hall, 1970). Membrane bones are certainly capable of forming cartilage, as seen in the repair of fractured rat parietal bones (Pritchard, 1946); and secondary cartilages are a normal feature of development of the mandible. Osteonectin and CD-RAP may function as growth-
regulating factors in wound-healing tissues. Lane and Sage (1990) have shown that osteonectin is a potent inhibitor of cell spreading and may locate the preosteoblasts to the bone-forming site. CD-RAP has been shown to inhibit tumour stem-cell colony formation, but whether it has any growth-regulatory function in the healing of bone tissues is unknown (Bogdahn et al., 1989). We show that CD-RAP is expressed transiently by a heterogeneous population of preosteoblasts in the healing tooth socket. It is surprising that ``chondrocytespeci®c'' proteins are expressed in the healing socket, as this tissue is derived from intramembranous ossi®cation. The function of these proteins is uncertain, but the presence of CD-RAP in cartilage and healing bone may indicate similarities at the molecular level between endochondral and intramembranous ossi®cation.
References
Fig. 10. Osteoblasts, but not preosteoblasts, expressed osteocalcin (P, osteoblasts). 500.
Beck Jr, G.R., Sullivan, E.C., Moran, E., Zerler, B., 1998. Relationship between alkaline phosphatase levels, osteopontin expression, and mineralization in dierentiating MC3T3-E1 osteoblasts. J. Cell Biochem 68, 269±280. Bianco, P., Riminucci, M., Bonucci, E., Termine, J.D., Robey, P.G., 1993. Bone sialoprotein (BSP) secretion and osteoblast dierentiation: relationship to bromodeoxyuridine incorporation, alkaline phosphatase, and matrix deposition. J. Histochem. Cytochem 41, 183±191. Bogdahn, U., Apfel, R., Hahn, M., Gerlach, M., Behl, C.,
Y.C. Shyng et al. / Archives of Oral Biology 44 (1999) 751±757 Hoppe, J., Martin, R., 1989. Autocrine tumour cell growth-inhibiting activities from human malignant melanoma. Cancer Res 49, 5358±5363. Bosserho, A.K., Kondo, S., Moser, M., Dietz, W., Copeland, N.G., Gilbert, D.J., Jenkins, N.A., Buettner, R., 1997. Mouse CD-RAP/MIA gene: structure, chromosomal localization, and expression in cartilage and chondrosarcoma. Dev. Dyn 208, 516±525. Devlin, H., Hoyland, J., Freemont, A.J., Sloan, P., 1995. Localization of procollagen type II mRNA and collagen type II in the healing tooth socket of the rat. Archs Oral Biol 40, 181±185. Devlin, H., Hoyland, J., Newall, J.F., Ayad, S., 1997. Trabecular bone formation in the healing of the rodent molar tooth-extraction socket. J. Bone Min. Res 12, 2061± 2067. Dietz, U.H., Sandell, L.J., 1996. Cloning of a novel retinoic acid-sensitive cDNA expressed in cartilage and during chondrogenesis. J. Biol. Chem 271, 3311±3316. Gorski, J.P., Grin, D., Dudley, G., Stanford, C., Thomas, R., Huang, C., Lai, E., Karr, B., Solursh, M., 1990. Bone acidic glycoprotein-75 is a major synthetic product of osteoblastic cells and localized as 75- and/or 50-kDa forms in mineralized phases of bone and growth plate and in serum. J. Biol. Chem 265, 14956±14963. Hall, B.K., 1970. Cellular dierentiation in skeletal tissues. Biol. Rev 45, 455±484. Hsieh, Y.D., Devlin, H., McCord, F., 1995. The eect of ovariectomy on the healing tooth socket of the rat. Archs Oral Biol 40, 529±531. Jahangiri, L., Devlin, H., Ting, K., Nishimura, I., 1998. Current perspectives in residual ridge remodeling and its clinical implications: A review. J. Prosthet. Dent 80, 224± 237. Lane, T., Sage, H., 1990. Functional mapping of SPARC: peptides from two distinct Ca2+-binding sites modulate cell shape. J. Cell Biol 111, 3065±3076. Lawson, G.M., Katzmann, J.A., Kimlinger, T.K., O'Brien, J.F., 1985. Isolation and preliminary characterization of a monoclonal antibody that interacts preferentially with the
757
liver isoenzyme of human alkaline phosphatase. Clin. Chem 31, 381±385. Liu, F., Malaval, L., Gupta, A.K., Aubin, J.E., 1994. Simultaneous detection of multiple bone-related mRNAs and protein expression during osteoblast dierentiation: polymerase chain reaction and immunocytochemical studies at the single cell level. Dev. Biol 166, 220±234. Mark, M.P., Prince, C.W., Oosawa, T., Gay, S., Bronckers, A.L.J.J., Butler, W., 1987. Immunohistochemical demonstration of a 44-kD phosphoprotein in developing rat bones. J. Histochem. Cytochem 35, 707±715. Nakase, T., Takaoka, K., Hirakawa, K., Hirota, S., Takemura, T., Onoue, H., 1994. Alterations in the expression of osteonectin, osteopontin and osteocalcin mRNAs during the development of skeletal tissues in vivo. Bone Min 26, 109±122. Nefussi, J.R., Brami, G., Modrowski, D., Oboeuf, M., Forest, N., 1997. Sequential expression of bone matrix proteins during rat calvaria osteoblast dierentiation and bone nodule formation in vitro. J. Histochem. Cytochem 45, 493±503. Owen, T.A., Aronow, M., Shalhoub, V., Barone, L.M., Wilming, L., Tassinari, M.S., Kennedy, M.B., Pockwinse, S., Lian, J.B., Stein, G.S., 1990. Progressive development of the rat osteoblast phenotype in vitro: reciprocal relationships in expression of genes associated with osteoblast proliferation and dierentiation during formation of the bone extracellular matrix. J. Cell Physiol 143, 420±430. Pritchard, J.J., 1946. Repair of fracture of the parietal bone in rats. J. Anat 80, 55±60. Rodan, G.A., Noda, M., 1991. Gene expression in osteoblastic cells. Crit. Rev. Eukaryotic Gene Expression 1 (2), 85±98. Strauss, P.G., Closs, E.I., Schmidt, J., Er¯e, V., 1990. Gene expression during osteogenic dierentiation in mandibular condyles in vitro. J. Cell Biol 110, 1369±1378. Wewer, U.M., Albrechtsen, R., Fisher, L.W., Young, M.F., Termine, J.T., 1988. Dierential deposition of osteonectin in basement membranes of human tissues undergoing profound changes in tissue architecture. Am. J. Path 132, 345±355.