In vitro transformation of chondroprogenitor cells into osteoblasts and the formation of new membrane bone.код для вставкиСкачать
THE ANATOMICAL RECORD 206:373-383 (1983) In Vitro Transformation of Chondroprogenitor Cells Into Osteoblasts and the Formation of New Membrane Bone MICHAEL SILBERMANN, DINA LEWINSON, HEDVA GONEN, MARIA ANTONIA LIZARBE, AND KLAUS VON DER MARK Division of Morphological Sciences, Faculty o f Medicine, Technion-Israel Institute of Technology, H a i f q Israel (M.S., D. L., H. GJ and The MaxPlanck-Institute f i r Biochemie, Munich, Federal Republic of Germany (M.A.L., K.u.d.M.) ABSTRACT Mandibular condyles of fetal mice 19 to 20 days in utero were kept in a n organ culture system for up to 10 days. After 2 days in culture the cartilage of the mandibular condyle appeared to have maintained all its inherent structural characteristics, including its various cell layers: chondroprogenitor, chondroblastic, and hypertrophic. After 5 days in culture no chondroblasts could be seen and, instead, the entire cartilage was occupied by hypertrophic chondrocytes. At the same time, the mesenchymal cells at the chondroprogenitor zone differentiated into osteoblasts which produced osteoid. Light microscopic examinations showed that the newly formed osteoid did not stain with acidic toluidine blue or with alcian blue, but stained intensively with the van Gieson stain and with Periodic acid-Schiff (PAS). The osteoid reacted with antibodies against type I collagen but not with antibodies against type I1 collagen, Electron microscopic examinations showed that the mineralization appeared to be associated with collagen fibers in bone rather than with matrix vesicles in the cartilage. The process of bone formation progressed with time and by the 10th day new bone replaced almost the entire cartilage, thus forming an expanded layer of membrane bone. This in vitro system represents a n experimental model whereby undifferentiated precursor cells transform into osteoblasts with the subsequent formation of a typical membrane bone. Recently, attempts were made to develop in vitro systems which would provide appropriate models for studies on the mechanisms involved in bone induction and bone formation. Simmons et al. (1982) described a combined in vitro-in vivo system for bone formation, in which isolated osteoblasts were placed in diffusion chambers and subsequently implanted in the peritoneal cavity of mice. If, however, isolated bone cells are grown in vitro in either primary cultures or in established cell lines, they usually do not form bone. The reason for that is as yet not fully understood, and it has been postulated that the currently used culture systems lack some fundamental provisions essential for bone morphogenesis (Wong, 1980). The present study used a n organ culture system of embryonic mandibular condylar cartilage, a tissue that in vivo undergoes endochondral ossification (Silbermann and Frommer, 1972, 1974). In vivo studies with this tissue have shown that following trans@ 1983 ALAN R. LISS, INC plantation the cartilage progenitor cells differentiate into osteoblasts (Meikle, 1973). Charlier and Petrovic (1967) and Petrovic (1972) were the first to culture isolated mandibular condyles, but they did not report on bone formation in their explants. Hall (1981) cultured quadratojugal cartilages of embryonic chicks and found chondroid bone formation if ascorbic acid was deleted from the medium. Hence, the present study was undertaken to reevaluate the potential of condylar organ culture for studies related to in vitro bone formation. We, hereby, report on a reproducible system in which de novo membranous bone is produced within 7 days in culture. MATERIALS AND METHODS ICR mice which were pregnant 19-20 days were anesthetized with ether. The uteri were removed aseptically, placed in sterile Hank’s Received November 30,1982; accepted April 25, 1983 374 M. SILBERMANN ET AL. solution, and the fetuses were delivered surgically. Using a surgical microscope, the mandibular condyles were dissected away from the mandibles, cleaned of all soft tissues and of the underlying bone (parts of the mandibular ramus) and were then placed in a fresh Hank’s solution a t 4°C. Explants consisting of clean cartilage (enveloped by a n intact perichondrium) were transformed onto Millipore filters (0.45 pm; Millipore Corp., Bedford, MA) cemented to stainless steel grids (Falcon, Oxnard, CA) that were placed in plastic disposable culture dishes (30 mm; Falcon, CAI. The culture medium consisted of Modified Eagle’s medium in Earle’s powder (Synthetic media, #410-1100, Gibco, Grand Island, NY), supplemented with 10% fetal calf serum (Gibco, Grand Island, NY) and with Sigma’s (St. Louis, MO) ascorbic acid (100 pg/ml), glycine (50 pg/ml), penicillin (50 U/ml) and streptomycin (50 pg/ml). The phosphate concentration in the medium was 3.0 mM and the pH 7.4. The medium was pipetted into the culture dishes until it just wetted the Millipore filters so that the explants were maintained at the medium-air interface, in a humidified incubator at 37°C in a n atmosphere of 5% COZ: 95% air. The medium was changed every 48 hours and specimens were obtained after 2, 5,7, and 10 days in culture. antibodies were used. The latter were prepared against lathyritic rat skin collagen. The immunization procedure was carried out as described by Timpl et al. (1972).The purification of the antibodies was achieved via immunoadsorption and their purity was determined by a n enzyme-linked immunosorbent assay (ELISA) as described by Gosslau and Barrach (1979). For the localization of type I1 collagen, guinea pig anti-chick type I1 collagen antibodies were used.’ Their preparation, purification, and specificity tests were carried out as described by von der Mark et al. (1976). We have also used rabbit antichick type I1 collagen antibodies, but they did not cross react with the mouse type I1 collagen. Following incubation with the above antibodies (undiluted, 30 minutes, room temperature in a moist chamber) the sections were rinsed three times with PBS and were subsequently reacted with fluorescein isothiocyanate (F1TC)-conjugated goat anti-rabbit yG (1:20) or with rabbit anti-guinea pig yG (1:20) for 30 minutes at room temperature. The above antibodies were purchased from Behringwerke (Marburg, Federal Republic of Germany). Sections incubated with FITCconjugated rabbit anti-guinea pig yG alone served as controls. Electron Microscopy Light Microscopy Tissues were fixed in 5%glutaraldehyde in Tissues that were designated for general 0.1 M cacodylate buffer, pH 7.4, at 4°C for morphology and histochemical examinations l?h hours; postfixed in 1%Os04 in 0.1 M were fixed in 80% ethanol for 30 minutes, cacodylate buffer, pH 7.4 for 1 hour; dehydehydrated in ethanols, and embedded in drated through graded ethanols and embedParaplast (Fisher Scientific Company, Pitts- ded flat in Epon. Thick sections (1pm) from burgh, PA). Serial sections (6 pm in thick- the same blocks used for electron microscopy ness) were stained with hematoxylin and were cut in the coronal plane and stained eosin, toluidine blue (0.1%; pH 2.5), alcian with 1%toluidine blue or with Pearson’s silblue (0.5M; pH 5.8), van Gieson, von Kossa, ver gelatin stain. Adjacent thin sections were and with the Periodic acid Schiff (PAS) stain. cut on LKB Ultratome I11 with a diamond Explants that were designated for immuno- knife, mounted on No. 300 copper grids, and fluorescence studies of collagen typing were stained for 10 minutes in saturated aqueous not fixed but were immediately frozen in liq- uranyl acetate followed by staining for 10 uid nitrogen (-80”C), cut in a cryostat minutes in 0.2% lead citrate before exami(-25°C) and the sections (8 pm in thickness) nation with a Jeol 100 B electron microscope, were mounted on microscope slides. A11 the operating at 60 kV. frozen sections were first treated with 0.3 M RESULTS ethylenediaminetetraacetic acid (EDTA), pH 7.5, for 30 minutes a t room temperature folThe normal structure of mouse condylar lowed by a 30-minute treatment with hy- cartilage, as seen in late fetal life, newborns, alase (ovine testes hyaluronidase salt-free and neonatal animals, has been previously analytical grade Serva #25118), 1 mg/ml in described (Silbermann and Frommer, 1973; phosphate buffered saline (PBS), pH 7.2 at room temperature. For the localization of ‘Kindly provided by Dr Waltraud Dessau, the Max Planck type I collagen, rabbit anti-rat type I collagen Institut fur Biochemie, Munich, Federal Republic of Germany BONE FORMATION IN VITRO 375 Silbermann and Lewinson, 1978); thus, rather than reiterating the material contained in these articles, we will refer to them whenever appropriate. At the age of 19-20 days embryonic life (the time when the explants were removed from the animals) the mandibular condyle is elongated supero-inferiorly, while a large proportion of the cartilage is comprised of chondroblasts and hypertrophic chondrocytes. At 7 days postpartum the hypertrophic zone is somewhat narrower, yet the progenitor and chondroblastic zones are still well developed (Silbermann and Lewinson, 1978). Bone or osteoid are never present within any of the above cell layers; and replacement of mineralized cartilage occurs only a t the condylar lower border along the ossification front (Silbermann and Frommer, 1972). Condyles that were cultured for 48 hours showed a distinct morphological resemblance to their age-matched tissues in vivo (1 to 2 days postpartum) (Fig. 1).The following cell layers could be noted very distinctly: progenitor, chondroblastic, and hypertrophic. The perichondrium was found to be very well organized and the cells along the chondroprogenitor zone formed a strip of closely packed cells with many intercommunicating cell processes. The intercellular matrix in the above zone contained relatively few collagen fibrils (Fig. 2). Farther away from the articular surface and below the chondroprogenitor zone the explants revealed a wide layer of chondroblasts. The latter cells contained the typical lakes of glycogen as well as a very well-developed system of rough endoplasmic reticulum with occasionally dilated cisternae. The Golgi apparatus appeared well developed and a small number of rounded mitochondria1 profiles were randomly dispersed throughout the cytoplasm. The matrix surrounding the chondroblasts contained fine fibrils of collagen and electron-dense granules which represent matrical proteoglycans (Fig. 3). Early hypertrophic chondrocytes retained the above-mentioned lakes; however, the glycogen was partly replaced by electron-lucent vacuoles resembling lipid droplets. These vacuoles varied in size and appearance and they were never bounded by a membrane (Fig. 4).In there was Fig. 1. Frontal section through condylar cartilage a n increasing number of mitochondria which after 48 hours in culture. The tissue shows all the typical were located mostlv along the DeriDheral cell layers as are normally seen in a condyle of a new* animal. A, articular surface; P, perichondrium; C, parts of the cells. yiunghypertrophic chon- born chondroprogenitor zone; B, chondroblastic zone; H, hy. . had a drocytes alike chondroblasts pertrophic zone. Stained with Pearson’s silver stain. surface while multiple electron-dense gran- x 2407 376 M. SILBERMANN ET AL. Fig. 2. Electron micrograph of cells in the chondroprogenitor zone of a condyle that was cultured for 48 hours. The cells reveal a high degree of resemblance to their homologs in viva. The matrix contains fine collagen fibrils which lack any specific orientation. x 6,000. ules appeared to be attached to their membrane. Hypertrophic chondrocytes were surrounded by a lacuna and the intercellular matrix farther away from the cells exhibited vesicular structures. The latter were bounded by a trilaminar membrane and were often associated with precipitates of hydroxyapatite (Fig. 5 ) .Fine collagen fibrils were always encountered in this portion of the matrix (Fig. 5). Immunofluorescent analysis indicated a positive reaction with anti type I antibodies along the perichondrium (Fig. 6) and a strong positive reaction with antiserum against type I1 collagen throughout the cartilaginous tissue (Fig. 7). Following 5 days in culture, the cartilage became occupied by hypertrophic chondrocytes which were surrounded by mineralized matrix. At this same time the zone of chondroprogenitor cells as well as some areas along the perichondrium revealed clear signs of new bone formation (Fig. 8). Immunofluorescent examination indicated that the matrix in the new site of bone formation contained type I collagen (Fig. 9). Electron micrographs from the above area showed that the newly formed collagen was irregularly oriented (Fig. 10)and that the fibers revealed the typical cross-banding pattern as normally seen in bone collagen. Furthermore, in the osteoid the hydroxyapatite crystals were noted to lie in close association with collagen fibers (Fig. 11)rather than with matrix vesicles, as was shown in cartilage (Fig. 5 ) . The ossification process continued to progress and by 7 days larger portions of the condylar cartilage were replaced by the new osteoid which was undergoing mineralization (Fig. 12). By this time the explants revealed the appearance of multinucleated giant cells; which were located close to the newly formed bone (Fig. 13).When 7-day-old explants were stained with either acidic toluidine blue or with alcian blue the new bone did not stain while the cartilage stained strongly. On the other hand, when sections were stained with either the van Gieson stain or with PAS the Fig. 3. A section through a more mature chondroblast and its adjacent matrix as seen in an explant cultured for 48 hours. Note the prominent nucleolus, the welldeveloped rough endoplasmic reticulum, the Golgi apparatus (G), and the intracellular lakes of glycogen (GI) so characteristic for these cells. The matrix contains few collagen fibrils and proteoglycan granules. x 8,000. Fig. 4. The appearance of an early hypertrophic chondrocyte in a condyle after 48 hours in culture. The cell’s cytoplasm shows remnants of the glycogen lakes (G1) which are largely replaced by lipidlike vacuoles (v). Multiple rounded mitochondria (m) are seen along with multiple strands of rough endoplasmic reticulum. A newly formed lacuna (L) is seen pericellularly. X 8,000. Fig. 5. A section through the intercellular matrix at the level of the mineralization front in a condyle cultured for 48 hours. Matrix vesicles (rnv) which appear to be bounded by a membrane are interspersed among the collagen fibers. Most of these vesicles which serve as the initial locus of matrix mineralization contain hydroxyapatite crystals. x 30,000. Fig. 6. Frontal section through an explant that has been cultured for 48 hours. The tissue was reacted with antibodies against type I collagen. Note the positive reaction along the perichondrium (PI, but not within the cartilage (C). Frozen section. X 96. Fig. 7. A similar section to that shown in Figure 6. but reacted with antibodies against type I1 collagen. The cartilage (C) responds strongly whereas the perichondrium (PI lacks any response. x 96. BONE FORMATION IN VITRO Fig. 8. Coronal section through a n explant that has been cultured for 5 days. It can be seen that almost the entire cartilage is occupied by hypertrophic chondrocytes. The matrix surrounding many of the above cells has undergone mineralization. Clear signs of new bone formation are seen along the perichondrium (single arrows) and especially at the upper portion of the condyle 379 (double arrows). Epon embedded and stained with Pearson’s silver stain. X 192. Fig. 9. A coronal section through the upper portion of a condyle that has been cultured for 5 days. The tissue was reacted with antibodies against type I collagen. A strong reaction can be seen in the region of new bone formation (B). x 384. Fig. 10. Electron micrograph showing portions of two osteoblasts (OBI and their adjacent matrix in the newly formed osteoid. A large number of collagen fibers (COL) occupy the matrix. x 10,000. Fig. 11. A section through collagen fibers in the newly formed osteoid. One of these fibers (arrow) is associated with hydroxyapatite crystals, a characteristic feature for osteoid mineralization. x 160,000. Fig. 12. A coronal section of an explant that was cultured for 7 days. In this specimen, the upper portion of the condyle underwent an advanced stage of bone formation (B). Note the alignment of giant cells along the lower right margin of the condyle. x 192. Fig. 13. A similar section to that shown in Figure 12, showing a group of giant clasts (CL) around an isolated area of mineralized matrix (M). Cartilage cells tC) at various stages of maturation are seen nearby. x 768. Fig. 14. The appearance of a portion of an explant that has been cultured for 10 days. A widespread area of the original explant underwent ossification (B). The newly formed bone is surrounded by a well organized periosteum (PI. Remnants of cartilage (C) are seen alongside the bony tissue. Pearson’s silver stain. x 192. BONE FORMATION IN VITRO 381 382 M. SILBERMANN ET AL. bone stained heavily while the cartilage did not. The new bone also reacted strongly with antibodies against type I collagen but not with the anti type I1 antibodies. By the tenth day most explants showed large areas of newly formed membranous bone (Fig. 14). Osteocytes could be easily identified within this tissue as well as osteoblasts that were lining the periphery of the mineralized osteoid. The latter cells were surrounded by several layers of fibroblasts thus forming a continuous periosteum. At this time interval remnants of cartilage tissue were still present adjacent to the new bone. DISCUSSION In the present study we were able to show that if fetal condylar cartilage is cultured in a medium containing serum the tissue maintains its original structure and organization for a t least 48 hours. Recent studies in our laboratory have indicated that during this time period the progenitor cells continue to synthesize DNA and proliferate (Maor and Silbermann, 19811, while the chondroblasts and chondrocytes participate actively in the mineralization of the matrix (Silbermann et al., 1981).This study also revealed that after 5 days in culture almost all the chondrocytes in the tissue became hypertrophic. This could suggest that 1) many of the hypertrophic chondrocytes present at the time of explant could be maintained in culture system; and 2) that the young chondroblasts became hypertrophic during the period of the culture. It is interesting to note that very similar findings have been described by Ronning and Koski (19691, who transplanted rat mandibular condyles into the brains of inbred littermates, and found that after 5 days the thickness of the proliferative zone had been reduced and that hypertrophic cells were closer to the upper surface of the cartilage. The present findings also indicate that in vitro the progeny of the chondroprogenitor cells differentiate into osteoblasts instead of into chondroblasts. This shift in differentiation and the subsequent formation of bone are apparent in areas that in vivo never show bone or osteoid tissue (Silbermann and Frommer, 1972, 1973, 1974; Silbermann and Lewinson, 1978; Lewinson and Silbermann, 1982).The results of the various methodological approaches used (histological, histochemical, immunofluorescent and ultrastructural) pointed toward a genuine formation of osseous tissue which eventually re- placed most of the original cartilage. The premise for the de novo elaboration of bone is supported by the following findings: 1)the newly formed tissue did not stain with alcian blue nor did it respond with metachromasia when stained with acidic toluidine blue; 2) it stained strongly with PAS and with the van Gieson stain; 3) it reacted positively with antibodies against type I collagen but not with antibodies against type I1 collagen; 4) the collagen fibers were banded and wider (two to five times) in comparison to those of cartilage; and 5) the mineralization process appeared to be associated with collagen fibers rather than with matrix vesicles as is normally the pattern in cartilage. However, in order to further substantiate the specificity of the newly formed tissue, one ought to check the presence of bone specific proteins, i.e., osteonectin via the application of the appropriate antibodies. The apparent change in the level of oxygen tension (in vitro versus in vivo) could be one of the factors responsible for the shift in cell differentiation as was observed in the present system. It became also apparent that our organ culture supported the mineralization of the newly formed osteoid. Recent, and as yet unpublished, data from our laboratory indicated that 45Ca is taken up by the mineralizing osteoid as well as oxytetracycline. Of particular interest was the finding that related to the appearance of multinucleated giant cells in the region of the newly formed bone. The concurrent presence of osteoblasts and osteoclasts in the cultured tissue could possibly indicate the existence of a coordinate control mechanism in this system of bone formation. The apparent “coupling” of bone formation to bone resorption has been attributed to a compensatory mechanism involved in bone homeostasis (Ivey and Baylink, 1981). A possible precursor of these newly formed osteoclasts are uninucleated cells that were transferred to the culture system in the small capillaries of the perichondrium. The cells could be mononuclear phagocytes (monocytes) or derivatives from the same progenitor cells which give rise to mononuclear phagocytes (Malone et al., 1982). In the present study osteoblasts appeared to have preceded the osteoclasts (although the exact sequence of events has not been as yet carefully checked); however, once mineralization started osteoclasts were always present. The present observations also tend to indicate that there seems to exist a BONE FORMATION IN VITRO spatial and temporal relationship between hypertrophying chondrocytes and osteodifferentiation. Scott-Savage and Hall (1980) have indeed argued that in order for progenitor cells to differentiate into osteoblasts they require a suitable microenvironment which could be provided by hypertrophic chondrocytes and their mineralized matrix. In conclusion, this study presented a n in vitro system which could serve as a useful tool for studies on bone induction and bone formation. The tissue utilized comprised embryonic cartilage which contains a population of committed but undifferentiated progenitor cells. In a n in vitro environment these cells differentiate into bone-forming cells and synthesize type I collagen and osteoid which eventually mineralizes and ossifies. ACKNOWLEDGMENTS The authors thank Mrs. A. Yahalomi and M. Brenman for their excellent technical assistance. This study was supported in part by the Israel National Council for Research and Development, Grant No. 2099. LITERATURE CITED Charlier, J.P., and A. Petrovic (1967) Recherches sur la maiidibule de r a t en culture d’organes: Le cartilage condylian a-t-il un potential de croissance independant? L’Orthodontie Franqaise, 38: 165-175. Gosslau, B., and H.J. Barrach (1979) Enzyme-linked immunosorbent microassay for quantification of specific antibodies to collagen type I, 11,111. J. Immunol. Methods, 29:71-77. Hall, B.K. (1981) Modulation of chondrocyte activity in vitro in response to ascorbic acid. Acta Anat., 109: 51-63. Ivey, J.L., and D.J. Baylink (1981)Postmenopausal osteoporosis: proposed roles of defective coupling and estrogen deficiency Metab. Bone Dis. Relat. Res., 3:3-7. Lewinson, D., and M. Silbermann (1982) Landmarks in chondrocyte differentiation and maturation a s envisaged by changes in the distribution of calcium complexes: an ultrastructural histochemical study. Metab. Bone Dis. Relat. Res., 4:143-150. 383 Malone, J.D., S.T. Teitelbaum, G.L. Griffin, R.M. Senior, and A.J. Kahn 11982) Recruitment of osteodast precursors by purified bone matrix constituents. J. Cell Biol., 92:227-230. Maor, G., and M. Silbermann (1981) In vitro effects of glucocorticoid hormones on the synthesis of DNA in cartilage of neonatal mice. FEBS Lett., 129:256-260. Meikle, M.C. (1973) In vivo transplantation of the mandibular joint of the rat; a n autoradiographic investigation into cellular changes a t the condyle. Arch. Oral Biol., 18r1011-1020. Petrovic, A.G. (1972) Mechanisms and regulation of mandibular condylar growth. Acta Morphol. Neer1.Scand., 10:25-34. Ronning, O., and K. Koski (1969) The effect of the articular disc on the growth of condylar cartilage transplants. Trans. Eur. Orthodont. SOC., 4599-108. Scott-Savage, P., and B.K. Hall (1980) Differentiative ability of the tibia1 periosteum from the embryonic chick. Acta Anat., 206;129-140. Silbermann, M., and J. Frommer (1972) The specific nature of endochondral ossification in the mandibular condyle of the mouse. Anat. Rec., 172659-668. Silbermann, M., and J . Frommer (1973) Heterogeneity among chondrocytes of the mandibular condyle in foetal and postnatal mice. Arch. Oral Biol., 18t1549-1554. Silbermann, M., and J. Frommer (1974) Electron microscopy of the secondary cartilage in mandibular condyle in the mouse. Acta Anat., Y0:330-346. Silbermann, M., and D. Lewinson (1978) An electron microscopic study of the premineralizing zone of the condylar cartilage in the mouse mandible. J. Anat., 12555-70. Silbermann, M., D. Lewinson, Z. Toister, H. Gonen, and H. Gitai (1981) An in vitro study on the effects of various steroid hormones upon the ultrastructure of young cartilage with special reference to the induction of matrix vesicles. In: Matrix Vesicles. A. Ascenzi, E. Bonucci, and B. de Bernard, eds. Wichtig Editore, Milano, pp. 129-134. Simmons, D.J., G.N. Kent, R.L. Jilka, D.M. Scott, M. Falcon, and D.V. Cohn (1982) Formation of bone by isolated, cultured osteoblasts in millipore diffusion chambers. Calcif. Tissue Int., 34:291-294. Timpl, R., H. Furthmayr, and W. Beil(1972) Maturation of the immune response to soluble rat collagen: Late appearance of antibodies directed to N-terminal sites of the &chain. J. Immunol., 108:119-125. von der Mark, H., K. von der Mark, and S. Gay (1976) Immunofluorescence studies on differential collagen synthesis during development of the chick embryo. Dev. Biol., 48:237-249. Wong, G.L. (1980) Bone cell cultures a s a n experimental model. Arthritis Rheum., 23:1081-1086.