Ultrastructural cytochemistry of proteoglycans associated with calcification of shark cartilage.код для вставкиСкачать
THE ANATOMICAL RECORD 208:149-158 (1984) Ultrastructural Cytochernistry of Proteoglycans Associated With Calcification of Shark Cartilage MINORU TAKAGI, RICHARD T. PARMLEY, FRANCIS R. DENYS, HIROSHI YAGASAKI, AND YOSHIHISA TODA Department of Anatomy, Nihon University School of Dentistry, Tokyo 101, Japan (M. T , H. Y,Y.T), Institute of Dental Research, University of Alabama in Birmingham, Birmingham, A L 35294 fM. T,R. TP.,R R.D.), and Departments of Pediatrics and Pathology, University of Texas Health Science Center at S a n Antonio, T X 78284 (R.T P ) ABSTRACT Proteoglycans (PGs) as well as sulfated glycosaminoglycans (GAGS) are closely associated with cartilage calcification. An inner zone of endoskeletal tesserae of sharks is composed of a unique calcified hyaline cartilage. Initial calcification can be seen in the cartilage close to the inner zone.We have ultrastructurally examined shark, Triakis scyllia, noncalcifying, calcifying, and calcified cartilage using the tannic acid-ferric chloride (TA-Fe), the high iron diamine (HID), and the HID-thiocarbohydrazide-silver proteinate (HID-TCH-SP)methods for localization of sulfated complex carbohydrates. In noncalcifying cartilage, TA-Fe and HID strongly stained matrix granules which were round, ovoid, elongated, or irregularly shaped and presumably represented PG monomers. The size and staining intensity of the reactive matrix granules progressively decreased in calcifying cartilage toward the calcification front of the calcified cartilage. Similarly, a progressive decrease in the size of the HID-TCH-SP stain deposits in the matrix granules was observed in the calcifying cartilage close to the calcification front and was interpreted a s a decrease in length of sulfate containing GAG chains. In the calcified cartilage, the highly calcified areas were often localized in the calcification front and contained few or no small KID-TCH-SPstain deposits, whereas the weakly calcified regions contained more stain deposits. These results indicate that partial and complete degradation of sulfated GAGs and/or PGs may be a requisite for calcification of shark cartilage. Proteoglycans (PGs) as well as glycosaminoglycans (GAGS) are thought to influence the calcification of cartilage, bone, and dentin. PGs, whose sulfated GAG molecules possess a strong bond formation with calcium, maintain calcium phosphate in a stable colloidal state and inhibit biological calcification (Di Salvo and Schubert, 1967). The removal of these PGs or their partial degradation by lysosomal enzymes would then result in calcification (Campo, 1970; Baylink et al., 1972; Glimcher, 1976). Alternatively, some degraded PGs may constitute the nucleating site for calcification (Campo, 1970; Baylink et al., 1972). Previous biochemical (Lohmander and Hjerpe, 19751, histochemical (Van den Hooff, 1964; Quintarelli and 0 1984 ALAN R. LISS, INC. Dellovo, 1965; Silbermann and Frommer, 1974), immunohistochemical (Poole et al., 1982), ultrastructural-cytochemical (Matukas and Krikos, 1968; Bonucci, 1969; Smith, 1970; Thyberg et al., 1973; Silbermann and Frommer, 1974; Davis et al., 1982; Shepard and Mitchell, 1982; Takagi et al., 1983b1, and high-resolution electron spectroscopic imaging (Arsenault and Ottensmeyer, 1983) studies of the changes of PGs in cartilage calcification have been conducted, primarily with mammalian and avian connective tissue sources. Received June 20, 1983; accepted September 29, 1983. Address reprint requests to Minoru Takagi, D.D.S., PhD., Department of Anatomy, Nikon University School of Dentistry, 1-8-13,Kanda-Surugadai,Chiyoda-Ku, Tokyo, Japan. 150 M. TAKAGI ET AL. routinely dehydrated, and embedded in Spurn (1969) low-viscosity resin. Sharks possess a n internal skeleton of cartilage which contains sulfated GAGS including chondroitin sulfate A, C, D, and variable proportions of keratan sulfates (Mathews, 1966,1975). The oversulfated nature of chondroitin sulfates in the shark cartilage presumably prevents its calcification. However, this cartilage frequently contains variable amounts of calcium salt deposits which are called endoskeletal tesserae (Kemp et al., 1975; Kemp and Westrin, 1979). Ultrastructural examination of PGs in shark cartilage may well enlighten our understanding of the role of PGs in calcification of cartilage in several species. However, only a few ultrastructural studies of shark endoskeletal tesserae have been reported (Kemp et al., 1975; Kemp and Westrin, 1979). To our knowledge, no one has demonstrated the changes in the localization of GAG containing PGs in shark cartilage calcification. The present study of shark jaw cartilage was undertaken to identify the ultrastructural localization of PGs in calcification, using the tannic acid-uranyl acetate (TA-UA)method (Sannes et al., 1978; Takagi et al., 1983a) for anionic complex carbohydrates, the TA-ferric chloride (TAFe) method (Sannes et al., 1978; Takagi et al., 1983a1, the high iron diamine (HID)method (Spicer, 1965; Spicer et al., 19671, and the HID-thiocarbohydrazide-silver proteinate (HID-TCH-SP) method (Sannes et al., 1979)for sulfated complex carbohydrates. TA-UA Method for Anionic Complex Carbohydrates The aldehyde-fixed, rinsed specimens with and without TA were routinely dehydrated in graded alcohols and propylene oxide and embedded in Spurr (1969)low-viscosityresin. Unosmicated thin sections mounted on stainless steel grids were treated for 10 min with a filtered fresh TA solution, pH 2.6-2.8, containing 5 g tannic acid (J.T. Baker Chemical Co., Phillipsburg, NJ) in 95 ml of distilled water, rinsed three times in distilled water, and treated for 5 min with a filtered fresh UA solution (pH 4.1-4.3) containing 1g uranyl acetate in 99 ml of distilled water (Sannes et al., 1978; Takagi et al., 1983a). Subsequently, stained sections were rinsed six times in distilled water. Control specimens were also examined without the TA-UA treatment. TA-Fe Method for Sulfated Complex Carbohydrates The aldehyde-fixed, rinsed specimens with and without TA were routinely dehydrated in graded alcohols and propylene oxide and embedded in Spurr (1969)low-viscosityresin. Unosmicated thin sections mounted on stainless steel grids were treated for 10 min with a filtered fresh TA solution as described above, rinsed three times in distilled water, and treated for 1min with a filtered fresh Fe solution (pH 1.4-1.6) which was prepared by adding 5 ml of 40% ferric chloride (Fisher Scientific Co., Fair Lawn, NJ) to 95 ml of distilled water (Sannes et al., 1978; Takagi et al., 1983a). Subsequently, stained sections were rinsed six times in distilled water and examined. Control specimens were also examined without the TA-Fe treatment. MATERIALS AND METHODS Tissue Preparation Endoskeletal tissues from leopard sharks, each approximately 50 cm in length, were used in this study. Cartilage specimens including endoskeletal tesserae were taken from the shark, Triakis scyllia, jaws. Some specimens were fixed in 2.7% glutaraldehyde in 0.1 M cacodylate buffer (pH 7.35) for 2 h at 4°C or 22"C, whereas others were fixed in 2.7% glutaraldehyde in 0.1 M cacodylate buffer (pH 6.8) containing 2%tannic acid (TA) for 2 h at 22°C (Takagi et al., 1983a). Subsequently the specimens were rinsed several times in 0.1 M cacodylate buffer (pH 7.35) containing 7% sucrose and then stained or postfixed as outlined below. Thin sections were examined without counterstains with a Philips 300 or Hitachi H-500 electron microscope. Routine Morphology The aldehyde-fixed, rinsed specimens without TA were postfixed for 1 h in 1%0 ~ 0 4 , HID and HID-TCH-SP Methods for Sulfated Complex Carbohydrates The aldehyde-fixed, rinsed specimens without TA were stained for 18 h a t 22°C in a n HID solution (Spicer, 1965; Spicer et al., 1967), which was prepared by adding 1.4 ml of 40% FeCl3 (Fisher Scientific Co., Fair Lawn, NJ) to a fresh diamine solution containing 120 mg of N,N-dimethyl-m-phenylenediamine (HCI)B (Eastman Kodak Co., Rochester, NY) and 20 mg of N,N-dimethylp-phenylenediamine (HC1) (Fisher Scientific) in 50 ml of HzO. Control specimens (for evaluation of intrinsic density) were incubated SHARK CARTILAGE PROTEOGLYCANS for 18 h a t 22°C in a MgC12 solution, pH 1.4, prepared by adding 1.4 ml of 40% MgClz to 50 ml of H2O and adjusting the pH with HC1. Some specimens were then postfixed in 1% Os04, routinely dehydrated, and embedded inspurr (1969) low-viscosity resin. The postosmication step was omitted for other specimens. To enhance HID staining, some thin sections were stained with a thiocarbohydrazide (Eastman Kodak)-silver proteinate (strong silver protein, Roboz Surgical Instrument Co., Washington, D.C.) sequence (TCH-SP)as described previously (Thiery, 1967; Sannes et al., 1979).The TCH-diamine andor iron complex presumably reduce SP to form electrondense stain deposits. SP background staining was eliminated by filtering (Whatman filter No. 2) the SP solution twice before use. Acid MgClz controls were similarly processed. RESULTS Morphological Observations Endoskeletal tesserae, localized in the peripheral part of shark jaw cartilage, were calcified tissues which contained a n outer zone presumed to be bone tissue and a n inner zone composed of calcified cartilage (Kemp et al., 1975; Kemp and Westrin, 1979).Contour lines, called Liesegang rings, ran in the matrix of the inner zone or the calcified cartilage matrix (Kemp et al., 1975; Kemp and Westrin, 1979) and were often concentric around chondrocytes which appeared to be trapped by calcification of cartilage. Chondrocytes, which were spherical or ovoid, were observed in the calcified and noncalcified matrix (Fig. 1). Most of these chondrocytes appeared to be engaged in the calcification of the extracellular matrix around them and did not appear degenerated. Rare degenerative chondrocytes were observed. Calcifying globular bodies or clusters of needle-like crystals were observed in the vicinity of a junction between calcifying and calcified cartilage (Fig. 1). TA-UA Staining for Anionic Complex Carbohydrates TA-UA stained matrix granules and collagen fibrils in cartilage matrix as well as intracellular glycogen (Fig. 2). The calcified cartilage included highly or weakly calcified matrix (Fig. 2), identified by contour lines which were presumably formed by a difference in mineralization degree of the calcified matrix. The weakly calcified matrix demonstrated the presence of the sparse organic 151 material after weak demineralization with TA-UA treatment at acidic pH (Fig. 2). TA-Fe Staining for Sulfated Complex Carbohydrates TA-Fe diffusely stained matrix granules in the noncalcifying extracellular matrix which was distantly located from the calcification front. Staining was not observed on plasmalemma, glycogen, lysosomes, ribosomes, mitrochondria, nucleus, and collagen fibrils. TAFe-positive matrix granules were round, ovoid, elongated, or irregularly shaped and varied in size, having a diameter of 20-50 nm. The size and staining intensity of TA-Fepositive matrix granules were similar in the noncalcifying matrix distant from or near to the calcification front (Fig. 31, but progressively decreased in calcifying cartilage close to the calcification front (Fig. 4). Often, calcified areas did not contain TA-Fe-reactive matrix granules. HID and HID-TCH-SP Staining for Sulfated Complex Carbohydrates The acidic nature of the HID solution resulted in decalcification and staining of the specimens. HID strongly stained matrix granules which varied considerably in their size and shape. The size of HID-reactive matrix granules progressively decreased from calcifying cartilage to the calcification front (Fig. 5). The osmiophilic appearance of the calcification front prevented evaluation of HID-positive material in this area. Consequently HID-TCH-SP staining was examined in unosmicated specimens as described below. HID-TCH-SP produced stain deposits of several sizes (6-20 nm in diameter) in matrix granules (Figs. 6-9). HID-TCH-SP staining was observed neither in membrane limited vesicles presumed to be matrix vesicles nor in demineralized globular bodies (Figs. 6, 7). However, HID-TCH-SP-stained material was adherent to the outer surfaces of the limiting membrane of the vesicles. In addition, a progressive decrease in the size of HID-TCH-SP stain deposits in the matrix granules was observed in the calcifying cartilage matrix close to the calcification front, which frequently lacked stain deposits (Figs. 6, 8). In the calcified cartilage, highly calcified areas contained few or no small HID-TCH-SP stain deposits (6-8 nm in diameter), whereas more abundant HID-TCH-SP-positivematerial was localized in the weakly calcified areas (Fig. 9). HID-TCH-SP-reactive matrix granules in the weakly calcified areas were distributed 152 M. TAKAGI ET AL Fig. 1. Calcifying globular bodies (arrows) in the shark jaw cartilage can be seen in the vicinity of a junction between calcifying and calcified matrix. Chondrocytes, which are localized in these sites, appear intact. Contour lines (arrowhead) are frequently observed in calcified cartilage matrix. Osmicated specimen; not counterstained. x 1,800;bar = 5 pm. Fig. 2. TA-UA stains the extracellular matrix gran- ules (arrowheads, enlarged in inset), collagen fibrils (Co, and elnarged in inset), and intracellular glycogen (gly). TA-UA treatment at acidic pH partly demineralizes calcified matrix and renders contour lines more obvious when compared to those without TA-UA. Nucleus (N). Calcifying globular bodies (arrows). Unosmicated specimen; not counterstained. ~ 9 , 2 0 0 bar ; = 1 pm. Inset ~28,000;bar = 0.5 km. SHARK CARTILAGE PROTEOGLYCANS Fig. 3. In noncalcifying cartilage, strong TA-Fe staining of the extracellular matrix is localized in matrix granules (arrowheads). The size and staining intensity of TA-Fe reactive matrix granules were similarly observed in the cartilage matrix. Chondrocyte (C). Thin section from TA-glutaraldehyde-fixedspecimen without ; = 1fim. postosmication; not counterstained. ~ 2 0 , 2 5 0bar Fig. 4. In the calcifying cartilage, the size and staining intensity of TA-Fe-positive matrix granules (arrowheads) progressively decreased toward the calcification front of the calcified cartilage matrix (CM). TA-Fe treat- 153 ment at acidic pH results in demineralization of calcified matrix, which often lacks staining. Thin section from TA-glutaraldehyde-fixed specimen without postosmication; not counterstained. Chondrocyte (C); Nucleus (N). ~20,250;bar = 1I m . Fig. 5. The size and staining intensity of HID-reactive matrix granules (arrowheads) progressively decreased from the calcifying cartilage to the osmiophilic calcified matrix (CM) in this osmicated specimen. HID treatment at acidic pH resulted in demineralization and staining of specimens. Chondrocyte (C); Nucleus (N). Not counterstained. ~20,250;bar = 1pm. more sparsely than those in the noncalcify- ground staining was not present in unosmicated specimens. ing and calcifying cartilage (Fig. 9). The aldehyde-fixed specimens without TAUA or TA-Fe treatment lacked staining. In Cytochemical Controls addition, the TA-glutaraldehyde-fixed speciHID control specimens comparably incu- mens without TA-UA or TA-Fe treatment bated in a n acid MgC12 solution lacked the demonstrated a minimal increase in matrix stain density observed in HID specimens. granule density (Fig. 111, which was attribAcid MgClz controls of unosmicated speci- uted to better preservation of the complex mens exposed to TCH-SP lacked staining carbohydrates during fixation rather than (Fig. 101, except for some fine TCH-SP stain true TA staining (Takagi et al., 1983a). The deposits (< 6 nm in diameter) which were addition of cationic reagents to fixatives has observed on a few membrane structures of previously been shown to improve cartilage osmicated control specimens. This back- preservation (Hunziker et al., 1982). Fig. 6. HID-TCH-SP stain deposits are observed in matrix granules (arrowheads, and enlarged in lower inmatrix granules but are not discernible in membrane- set) can be seen in calcifying cartilage matrix close t o limited vesicles (thick arrows, and enlarged in upper the calcification front of the calcified matrix (CM). 0 s insets) and globular bodies (thin arrow). A progressive micated specimen; not counterstained. X17,ZOO; bar = 1 decrease in the size of HID-TCH-SP stain deposits in pm. Insets X40,OOO; bar = 0.5 pm. Fig. 7. The size of HID-TCH-SP stain deposits also decreases close to the globular body (arrow) which lacks staining. Osmicated specimen; not counterstained. ~40,000; bar = 0.5pm. Fig. 8. The size of HID-TCH-SP stain deposits in the matrix granules (arrowheads, and enlarged in inset) decreases close to the calcification front of the calcified cartilage matrix (CM) which lacks staining in this unosmicated specimen. Nucleus (N). Not counterstained. x 17,200;bar = 1pm. Inset ~40,000; bar = 0.2 pm. Fig. 9. Calcified cartilage contains variable amounts of HID-TCH-SP staining which was graded from no staining to moderate staining. The variety of the staining possibly results from differences in the degree of mineralization in the calcified matrix. Areas presumed to be highly calcified matrix (thin arrows) include or lack fine HID-TCH-SPstain deposits whereas those areas presumed to be weakly calcified matrix (thick arrows) contain more HID-TCH-SP stain deposits. Intense staining can be seen in noncalcified matrix (EM) in the vicinity of the calcification front. Osmicated specimen; not counterstained. x 17,200;bar = 1 pm. 156 M. TAKAGI ET AL. The present studies have demonstrated the changes in the matrix granule morphology and sulfated GAG content associated with calcification of shark jaw cartilage, utilizing ultrastructural cytochemical methods specific for sulfated complex carbohydrates. The decrease in the size of TA-Fe- and HIDstained matrix granules and the size of HIDTCH-SP stain deposits in calcifying cartilage and the complete loss of both stained materials in calcified cartilage indicate that PGs and sulfated GAGS are partially or completely removed during calcification. The decrease in the size of matrix granules (Matukas and Krikos, 1968; Thyberg et al., 1973; Takagi et al., 1983b) and HID-TCH-SPstain deposits (Takagi et al., 1983b) in calcifying cartilage is similar to that observed in mammalian cartilage. The complete loss of staining in calcified shark cartilage differs from rat epiphyseal cartilage, where complete or overall loss of HID-TCH-SPstaining does not occur during calcification (Takagi et al., 1983b). Thus in this respect shark cartilage appears distinct, and data obtained from this system are a requisite for our understanding of the role of PGs in calcification of cartilage in several species. The morphology, size, and staining properties (i.e., TA-Fe, HID, and HID-TCH-SP staining) of matrix granules in noncalcifying shark cartilage are very similar to those observed in the proliferative and upper hypertrophic zone of rat epiphyseal cartilage (Takagi et al., 1982b).This suggests that the matrix granule may represent the PG monomer(s) observed in previous ultrastructural studies of mammalian cartilage (Hascall, 1980;Poole et al., 1982;Takagi et al., 1982b). Previous studies (Takagi et al., 1982a,b) have correlated the variation in size of the HID-TCH-SPstain deposits in a given specimen with the length or degree of sulfation of GAG chains. In the present study we demonstrate a decrease in the size of HID-TCH-SP stain deposits in calcifying cartilage and the calcification front. This HID-TCH-SP result suggests that, concomitant with the decrease Fig. 10. No TCH-SP stain deposits are present in this unosmicated control specimen treated with acid MgClz rather than HID. Nucleus (N); calcified cartilage matrix (CM). Thin section from glutaraldehyde-fixed specimen; not counterstained. x 17,000;bar = 1pm. Fig. 11. This TA-glutaraldehyde-fixedspecimen lacks the density observed in TA-UA- and TA-Fe-stained thin sections (cf. Figs. 2, 3). Nucleus (N); calcified cartilage (CM). Unosmicated specimen without counterstains. ~17,000; bar = 1 pm. DISCUSSION SHARK CARTILAGE PROTEOGLYCANS in matrix granule size demonstrated with TA-Fe staining, the GAGs in calcifying areas are decreased in size or degree of sulfation. This observation is consistent with previous ultrastructural studies which demonstrate a n accumulation of small Ruthenium Red-positive material (Silbermann and Frommer, 1974) and small HID-TCH-SP stain deposits (Takagi et al., 1983b) in calcification of mammalian cartilage. Thus in both mammalian and shark cartilage degradation of PGs andl or sulfated GAGs appears necessary for cartilage calcification. Matrix vesicles which might be initial calcification sites have been identified in previous ultrastructural studies of shark cartilage (Kemp and Westrin, 1979).The present studies identified membrane-limited bodies presumed to be matrix vesicles a t the calcification front. These vesicles and globular bodies lacked HID-TCH-SP staining so that we could not clarify whether the vesicles were related to removal of PGs and sulfated GAGs. In contrast, small HID-TCH-SP stain deposits accumulated in disrupted matrix vesicles and globular bodies in calcification of rat epiphyseal cartilage, indicating a role for matrix vesicles in sulfated GAG degradation (Takagi et al., 198313). The fact that in some areas partial and in other areas complete loss of PGs or sulfated GAGs could be observed in shark cartilage calcification indicates the presence of different types of calcified cartilage. This interpretation is consistent with previous studies describing the presence of three principal kinds of calcified cartilage in elasmobranchs (Qrvig, 1967). To our knowledge, complete removal of PGs and/or sulfated GAGs has not been biochemically or ultrastructurally demonstrated in calcification of mammalian cartilage, bone, and dentin. The loss of all stainable sulfated GAGs clearly distinguished cartilage calcification in sharks from that previously described in mammals, and suggests unique mechanisms of calcification in this system. ACKNOWLEDGMENTS The authors thank Ms. Barbara A. Woolley and Ms. Shoko Ogura for their secretarial assistance. This work was supported in part by National Institutes of Health Grant DE02670. LITERATURE CITED Arsenault, A.L., and F.P. Ottensmeyer (1983) Quantitative spatial distributions of calcium, phosphorus, and 157 sulfur in calcifying epiphysis by high resolution electron spectroscopic imaging. Proc. Natl. Acad. Sci. USA, 80:1322-1326 Baylink, D., J. Wergedal, and E. Thompson (1972) Loss of protein-polysaccharides at sites where bone mineralization is initiated. J. Histochem. Cytochem., 20: 279-292. Bonucci, E. (1969) Further investigation on the organic/ inorganic relationships in calcifying cartilage. Calcif. Tissue Res., 3:38-54. Camoo. R.D. (1970) Protein-oolvsaccarides of cartilage and bone in health and disea”se. Clin. Orthop. Relit. Res., 68t182-209. Davis, W.L., R.G. Jones, J.P. Knight, and H.K. Hagler (1982) Cartilage calcification: An ultrastructural, histochemical, and analytical x-ray microprobe study of the zone of calcification in the normal avian epiphyseal growth plate. J. Histochem. Cytochem., 30:221-234. Di Salvo, J., and M. Schubert (1967) Specific interaction of some cartilage proteinpolysaccarides with freshly precipitating calcium phosphate. J. Biol. Chem., 242: 705-7 10. Glimcher, M.J. (1976) Composition, structure, and organization of bone and other mineralized tissues and the mechanism of calcification. In: Handbook of Physiology, Section 7: Endocrinology; Vol. 7, Parathyroid Gland, R.O. Greep and E.B. Astwood, eds. American Physiological Society, Washington, D.C., pp. 25-116. Hascall, G.K. (1980)Cartilage proteoglycan: Comparison of sectioned and spread whole molecules. J. Ultrastruct. Res., 70:369-375. Hunziker, E.B., W. Herrmann, and R.K. Schenk (1982) Improved cartilage fixation by ruthenium hexammine trichloride (RHT). A prerequisite for morphometry in growth cartilage. J. Ultrastruct. Res., 81:1-12. Kemp, N.E., S.J. Smith, and R.A. Jacobe (1975) Ultrastructure of calcified jaw tesserae in the lemon shark Neguprion breuzrostris. Am. Zool., 15: 329 (abstract). Kemp, N.E., and S.K.Westrin (1979) Ultrastructure of calcified cartilage in the endoskeletal tesserae of sharks. J. Morphol.. 160~75-102. Lohmander, S., and A. Hjerpe (1975) Proteoglycans of mineralizing rib and epiphyseal cartilage. Biochim. Biophys. Acta, 404:93-109. Mathews, M.B. (1966) The molecular evolution of cartilage. Clin. Orthop. Relat. Res., 48267-283. Mathews, M.B. (1975) Connective tissue: Macromolecular structure and evolution. Mol. Biol. Biochem. Biophys., I9:126-155. Matukas, V.J., and G.A. Krikos (1968) Evidence for changes in protein polysaccharide associated with the onset of calcification in cartilage. J. Cell Biol., 39: 43-48. Orvig, T. (1967) Phylogeny of tooth tissue: Evolution of some calcified tissues in vertebrates. In: Structural and Chemical Organization of Teeth, Vol. 1. A.E.W. Miles, ed. Academic Press, New York, pp. 45-110. Poole, A.R., I. Pidoux, and L. Rosenberg (1982) Role of proteoglycans in endochondral ossification: Immunofluorescent localization of link protein and proteoglycan monomer in bovine fetal epiphyseal growth plate. J. Cell Biol., 92:249-260. Quintarelli, G., and M.C. Dellovo (1965) The chemical and histochemical properties of alcian blue. IV. Further studies on the methods for the identification of acid glycosaminoglycans. Histochemie, 5196-209. Sannes, P.L., T. Katsuyama, and S.S.Spicer (1978)Tannic acid-metal salt sequences for light and electron microscopic localization of complex carbohydrates. J. Histochem. Cytochem., 26:55-61. Sannes, P.L., S.S. Spicer, and T. Katsuyama (1979) Ultrastructural localization of sulfated complex carbohy- 158 M. TAKAGI ET AL. drates with a modified iron diamine procedure. J. Histochem. Cytochem., 27:1108-1111. Shepard, N.L., and N.S. Mitchell (1982) Proteoglycan aggregates in the longitudinal septa of rat growth plate. Transact. 28th Annu. Mtg. Orthop. Res. SOC. 7:76 (abstract). Silbermann, M., and S. Frommer (1974) Demonstration and distribution of acidic glycosaminoglycans in mouse secondary cartilage: Light and electron microscopic study. Histochemistry, 38235-93. Smith, J.W. (1970) The disposition of protein-polysaccharide in the epiphyseal plate cartilage of the young rabbit. J. Cell Sci., 653434364. Spicer, S.S. (1965) Diamine methods for differentiating mucosubstances histochemically. J. Histochem. Cytochem., 13211-234. Spicer, S.S., R.G. Horn, and T.J. Leppi (1967) Histochemistry of connective tissue mucopolysaccharides. In: The Connective Tissue, International Academy of Pathology Monograph No. 7. B.M. Wagner and D.E. Smith, eds. Williams and Wilkins, Baltimore, pp. 251-303. Spurr, A.R. (1969)A low viscosity epoxy resin embedding medium for electron microscopy. J. Ultrastruct. Res., 26:31-43. Takagi, M., R.T. Parmley, Y. Toda, and F.R. Denys (1982a) Extracellular and intracellular digestion of complex carbohydrates by osteoclasts. Lab. Invest., 46288-297. Takagi, M., R.T. Parmley, Y. Toda, and R.L. Austin (1982b) Ultrastructural cytochemistry and immunocytochemistry of sulfated glycosaminoglycans in epiphyseal cartilage. J. Histochem. Cytochem., 30:1179-1185. Takagi, M., R.T. Parmley, F.R. Denys, andM. Kageyama (1983a) Ultrastructural visualization of complex carbohydrates in epiphyseal cartilage with the tannic acidmetal salt methods. J. Histochem. Cytochem., 31:783790. Takagi, M., R.T. Parmley, and F.R. Denys (1983b)Ultrastructural cytochemistry and immunocytochemistry of proteoglycans associated with epiphyseal cartilage calcification. J. Histochem. Cytochem., 31:1089-1100. Thiery, J.P. (1967)Mise en evidence des polysaccharides sur coupes fines en microscopie electronique. J. Microsc., 6:987-1018. Thyberg, J., S. Lohmander, and U. Friberg (1973) Electron microscopic demonstration of proteoglycans in guinea pig epiphyseal cartilage. J. Ultrastruct. Res., 45:407-427. Van den Hooff, A. (1964) Polysaccharide histochemistry of endochondral ossification. Acta Anat., 57:16-28.