The penetration of exogenous tracers through the enameloid organ of developing teleost fish teeth

TISSUE & CELL 1989 21 (3) 419430 @ 1989 Longman Group UK Ltd

KENNETH

S. PROSTAK,

PHILIP SEIFERT and ZIEDONIS

SKOBE

THE PENETRATION OF EXOGENOUS TRACERS THROUGH THE ENAMELOID ORGAN OF DEVELOPING TELEOST FISH TEETH Keywords: Enameloid, mineralization, teeth, development, cell permeability ABSTRACT. In order to determine whether exogenous materials permeate to the forming tooth enameloid matrix, teleost species were injected intramuscularly with horseradish peroxidase (HRP) or myoglobin, or; intracardially with lanthanum nitrate or HRP, then killed at predetermined intervals post-injection. Tooth bearing bones were processed for transmission electron microscopy. At the enameloid matrix formation stage, capillaries associated with the enameloid organ were few in number and rarely fenestrated. Both organic tracers reached the matrix at cervical, but not coronal, regions of the teeth in all species examined. Lanthanum was rarely observed extravascularly and never extended to the enameloid matrix at the secretion stage. At the enameloid mineralization stage, fenestrated capillaries were closely associated with the outer dental epithelial cells (ODE). All tracers were observed in the plasma membrane imaginations of the ODE. Only intracardially injected HRP compromised the apical intercellular junctions of the inner dental epithelial cells (IDE) to reach the mineralizing enameloid. Lanthanum did not extend past the ODE-IDE cell junctions. It is concluded that the close association of mineralization stage fenestrated capillaries with the highly invaginated ODE cells result in increased tracer penetration compared to the secretory stage. The deeper penetration of the organic tracers, compared with lanthanum, between mineralization stage IDE cells may be due to longer in viva circulation of the former material. The apical junctions of mineralization stage IDE cells, however, remained impermeable to the organic tracers. The absence of mineral in secretory stage enameloid mineral could not be due to specialized cell junctions preventing access of molecules to the matrix. It is suggested that controlling factors other than cellular permeability initiate enameloid mineralization.

Introduction

Skobe, 1986b). On the other hand, during mammalian tooth development, small nascent enamel crystallites form immediately as the ectodermally derived, non-collagenous enamel matrix proteins are secreted (Frank, 1979). The subsequent mineralization process of mammalian enamel and teleost enameloid matrices proceed in a similar manner progressing from both the surface layer and dentin-enamel junction towards the middle layer to form highly mineralized, apatitic tissues (Isokawa et al., 1970; Toda et al., 1970; Inoue et al., 1973; Suga, 1983). However, it is not known whether similar cytological processes are involved in the regulation of enameloid and enamel matrix mineralization. Several levels of controlling enamel

Mammalian tooth enamel and teleost fish enameloid are both highly mineralized, apatitic tissues covering the outer surface of the dentition. During development, the enameloid matrix of some teleost fish teeth is primarily an ectodermally derived collagen (Prostak and Skobe, 1985; 1986a) which is secreted to its full thickness before any mineralization begins (Prostak and Skobe, 1986b). Unlike bone or dentin collagen, however, enameloid collagen is removed during the mineralization process (Prostak and Forsyth Dental 02115, USA.

Center,

140 Fenway,

Boston,

MA

Received 22 September 1989. 419

PROSTAK

mineralization have been proposed. The mammalian enamel organ, specifically the ameloblast cell layer, is thought to regulate enamel mineralization by providing an ion diffusion barrier to the enamel matrix (Wennberg and Bawden, 1978; Bawden et al., 1982). The intercellular junctions of mammalian secretory ameloblasts limit the access of electron dense tracers to the nascent enamel surface (Takano and Crenshaw, 1980; Kallenbach, 1980), although other studies suggest these junctions are leaky (Uchida et al., 1987). Likewise, it is not clear if intercellular junctions of mammalian maturation ameloblasts are tight (Josephsen, 1984; Kallenbach, 1980b) or leaky (Sasaki, 1984; Warshawsky, 1984). Mammalian ruffleended ameloblasts have been reported to block isotope entry to the enamel whereas the smooth-ended cells do not (Crenshaw and Takano, 1982; McKee etal., 1986; MartineauDoize et al., 1986). A pulpal route for access of ions to the mammalian enamel matrix does not exist (Reith and Coty, 1962). Little is known about the role teleost odontogenic cells have in limiting the access of materials to the enameloid matrix. Secretory stage inner dental epithelial (IDE) cells do not have apical and basal junctional complexes consistently located (Shellis and Miles, 1976; Prostak and Skobe, 1986a) as observed in mammalian secretory amelo-

blasts (Warshawsky, 1978). Despite this apparent morphological lack of diffusion barriers, the formative enameloid matrix remains unmineralized initially. Teleost maturation stage IDE cells do have numerous intercellular junctions and predominantly ruffle-ended apical membranes, (Shellis and Miles, 1976; Prostak and Skobe, 198613)similar in appearance to mammalian ruffle-ended ameloblasts (RA) (Suga 1959; Warshawsky and Smith, 1974; Josephsen and Fejerskov, 1977). Mammalian RA are thought to play a role in calcium transport to the growing enamel (Crenshaw and Takano, 1982) and in protein resorption (Reith and Coty, 1967; Takano and Ozawa, 1984). These functions are also ascribed to teleost RA morphology (Shellis and Miles, 1976; Prostak and Skobe, 1986b). A maturation stage papillary layer described by Kallenbacl! (1966), is not present in teleost tooth buds. Instead, the outer dental epithelium (ODE) plasma membrane is highly invaginated, and is suggested to have ion transport or protein resorption functions (Herold, 1974; Garant, 1970; Prostak and Skobe, 1986). The high quantity of IDEODE intercellular junctions corroborates the idea that these cells may jointly form a diffusion barrier to extradental materials from reaching the enameloid matrix (Prostak and Skobe, 1986).

Fig. 1, Vascular element (V) adjacent to the outer dental epithelium (ODE) of a secretory stage tooth bud from a lanthanum perfused cichlid. Lanthanum deposits line the endothelial lumen and clefts. Very little tracer is observed extravascularly or at the basal lamina (arrow) of the ODE. x 15,WJO. Fig. 2. Vascular element (V) adjacent to the ODE of a late secretory stage tooth bud from a lanthanum perfused tautog. Lanthanum is restricted to the endothelial lumen, even though the ODE extracellular spaces are contiguous with the surrounding connective tissue (arrow). x8.300. Fig. 3. Cichhd secretory stage ODE cells in close juxtaposition to a vascular element (V). Lanthanum perfused specimen. In this rare instance, tracer was observed between the endothelial clefts to partially permeate through the ODE intercellular space. No ODE cellular junctions are apparent at the sites of maximal penetration (arrow). x 15,OlM. Fig. 4. Cichlid late maturation stage tooth bud showing thin walled capillary (C) adjacent to the ODE cell layer. Lanthanum perfused specimen. Tracer permeated between neighboring ODE cells (arrowhead), and also within the numerous plasma membrane invaginations (arrow). Lanthanum was not observed intracellularly, nor did it permeate between the IDE celk. x 12,000.

ENAMELOlD

ORGAN

OF DEVELOPING

TELEOST

This study investigates the permeability of three electron dense tracers, horseradish peroxidase, myoglobin, and lanthanum nitrate, in several teleost fish species to determine whether the secretory and maturation stage enameloid organs act as diffusion barriers to these exogenous substances. Materials and Methods Lanthanum nitrate

The teleost species utilized in this experiment were 7 small (4 cm long) and 3 medium (9 cm long) sized cichlids (Cichlasoma sp. ; freshwater), one large (400cm long) tautog (Tautog onitis; marine), and one 300 cm long sea robin (Prionotus sp.; marine). All fish were anesthetized by the addition of Finquell (Ayerst Lab; New York) at a ratio of 15000 to the water. An incision was made between the pectoral fins to expose the ventricle. A 28 gauge needle was inserted through the ventricle into the bulbous arteriosus, and ligated to prevent leakage. The sinous venosus was cut, the vasculature was cleared with 0.9% saline containing heparin, and perfused with a 3% lanthanum nitrate solution in 0.9% saline for approximately 3-5 min. Tooth bearing bones were dissected and immersed in fixative (Karnovsky, 1965) (marine fixative had 3% NaCl added) for 4 hr at room temperature. Samples were rinsed overnight in 0.1 M cacodylate buffer, post fixed in 2% 0~0, buffered with 0.1 M collidine,

FISH TEETH

423

dehydrated in an ethanol series, and embedded in Epon 812. Horseradish Peroxidase (HRP) Intracardial Injection

The ventricle of one Scm long cichlid was exposed as previously described, and injected with 10mg HRP dissolved in O-1ml saline. Ten min. post injection, the tooth bearing bones were dissected and fixed for 4 hr. Subsequent HRP cytochemistry is described below. Intramuscular Injection. (HRP and Myoglobin)

Two cichlids were injected along the lateral line with 10 mg myoglobin/O.l ml saline, killed 2 and 4hr post injection, and tooth bearing bones fixed for 4 hr. Another 12 cichlids were injected similarly with 10 mg HRP/ 0.1 ml saline, 3 each killed at 1, 2, 5 and 6 hr post injection, and tooth bearing bones fixed in the solutions described above. Control fish (2) remained uninjected. All HRP, myoglobin and control specimens rinsed in 0.1 M cacodylate buffer, and the bony plates were partially dissected to allow penetration of substrate to the tooth buds. Sample washes were changed to 0.1 M Tris buffer, pH 7.4, and incubated in diaminobenzidine (DAB) solution (10 mg/ml Tris) containing 0.01% H,O, for 2 hr in the dark. This time was found to be adequate for the DAB to penetrate the whole tissue. After

Fig. 5. ODE-IDE interface of a maturation stage tooth bud from a sea robin. Lanthanum perfused specimen. Lanthanum deposits (arrows) present between ODE cells was not observed between IDE cells. Extensive junctions(J) connect IDE to ODE cells. x22SGll. Fig. 6. Cervically located, secretory stage IDE cells from a cichlid injected intramuscullary (IM) with myoglobin, and killed 2 hr later. Reaction product is seen between cells, in pinocytotic vesicles (arrows), and within the extracellular matrix (M). Note that reaction product occurs on either side of a cellular junction (J). ~24,000. Fig. 7. Mid coronal IDE cells from the same specimen shown in Fig. 6. Intercellular reaction product (arrow) did not penetrate to the matrix (M) despite the lack of cell junctions. More coronally located IDE cells had no intercellular product. x 32,ooO. Fig. 8. Cervically located odontoblasts from the same specimen shown in Fig. 6. Reaction product is present between odontoblasts (0), on either side of intercellular junctions (J). and within the matrix (M). No reaction product is observed between more coronally located odontoblasts. x17.000.

PROSTAK

incubation, samples were rinsed in Tris, and post fixed in 2% 0~0, in O-1 M collidine buffer for 4 hr. One control specimen which had not been incubated in DAB had 1% K,Fe(CN), added to the osmium solution. Specimens were dehydrated and embedded as described above. All specimens were observed with no additional staining in either a JEOL 1OOB or 1200EX transmission electron microscope. Results Lanthanum Nitrate At the secretory stage, lanthanum deposits were primarily confined to the lumen of capillaries near the ODE cells of cichlid (Fig. 1) and tautog (Fig. 2) tooth buds. Most capiilaries observed at this stage were not fenestrated although lanthanum deposits were sometimes observed permeating between the endothelial cell clefts. In a few instances, lanthanum did escape the capillary lumen through the endothelial clefts to penetrate past the basal lamina of cichlid ODE cells (Fig. 3). The deposits extended into the intercellular space of ODE cells approximately 5 pm, but no further, despite the lack of any cellular junctions. At the enameloid mineralization stage, capillaries were highly fenestrated and closely associated with the ODE cells in all species examined. Lanthanum was observed within the capillary lumen, adherent to the basal lamina, and within the invaginations of the ODE plasma membrane (Fig. 4).

Deposits extended up to the ODE-IDE junctions but were not present between cells (Figs 4,5) in any species examined.

cell IDE

HRP and Myoglobin Intramuscular Injection At the secretory stage, both myoglobin and HRP permeated between cichlid IDE cells near the cervical loop region to reach the enameloid and predentin matrices, 1 hr post injection (Fig. 6). Fuzzy coated vesicles near the lateral and apical plasma membranes of the IDE cells also contained reaction product. Towards the mid-coronal regions, the organic tracers penetrated between the IDE cells, but did not reach the enameloid matrix at any time interval examined, even in cases where cell junctions were not discernable (Fig. 7). Cervically located odontoblasts had tracer between neighboring cells, and within the predentin matrix (Fig. 8). Tracer was not observed between more coronally located odontoblasts or in predentin matrix. At the mineralization stage, both myoglobin and HRP permeated into the ODE membrane invaginations and was endocytosed by fuzzy coated vesicles of the ODE cells, 1 hr post injection (Figs 9, 10). Both organic tracers did permeate the basal junctional complexes between IDE cells, but myoglobin did so less frequently. On the other hand, HRP was consistently found between IDE cells (Fig. lo), sometimes forming interceliular pools. Neither tracer was observed at the mineralizing enameloid surtace during the

Fig. 9. Maturation stage ODE-IDE cell interface from a cichlid injected IM with myoglobin. and killed 4 hr later. Plasma membrane invaeinations (I) and fuzzv coated vesicles of the ODE cells contain reaction product. Myoglobin had permeated between some (arrow), but not other (S)IDE cells. Reaction product did not extend to the enamel surface in either case. x 13,500. Fig. 10. Maturation and killed 1 hr later. elongated fuzzy coated reaction product. HRP

stage ODE-IDE cell interface from a cichlid injected IM with HRP. Within the ODE cell cytoplasm, reaction product is observed within vesicles (arrows). Some invaginations of the ODE cell membrane contain is also observed between IDE cells (arrowhead). x21 ,ooO.

Fig. 11. Apical region of the IDE cells shown in Fig. 10. HRP has a spotty intercellular localization (arrows), but is not observed near or at the enameloid surface (E). x9GC10. Fig. 12. Maturation stage IDE cells from a cichlid injected intracardially (arrows) extends between cells to the enameloid surface (E). x 11 ,ooO.

with HRP. Tracer

ENAMELOID

ORGAN OF DEVELOPING

TELEOST FISH TEETH

experimental time intervals (Fig. 11). The 5 and 6 hr post injection specimens were basically similar to the earlier time intervals, and control enameloid organs had no reaction product. lntracardial HRP Injection

In this cichlid, HRP was observed between neighboring IDE cells, apical to intercellular junctions (Fig. 12a) and approached closely to the highly mineralized enameloid surface (Fig. 12b). In some mineralization stage IDE cells, apical granules, some of which appeared in continuity with the enamel surface, were also filled with reaction product (Fig. 12~). No other stages were observed in this specimen. Potasium Ferricyanide Staining

This staining method in control tissues paralleled the localization of the lanthanum in mineralization stage ODE cells. Deeply stained material was observed within the capillary lumen, the membranous invaginations and fuzzy coated vesicles of the ODE cells (Fig. 13). Osmium-ferricyanide staining did not extend between mineralization stage IDE cells, nor was it observed in specimens post-fixed in osmium without ferricyanide added. Discussion

In the mineralization stage tooth buds of the species examined (cichlid, tautog, and sea robin), the three tracers used here freely penetrated the highly fenestrated capillaries surrounding the tooth buds to become incorporated within the membranous invaginations of the ODE cells. The penetration of tracers, and also the uniform ferricyanideosmium staining of the ODE membrane inva-

427

ginations, support the contention that these structures are continuous with the ODE plasma membrane (Garant, 1970; Herold, 1974; Prostak and Skobe, 1986). Because of this continuity, it is unlikely that these structures are similar to the intracellular, acid phosphatase positive, tubular vesicles observed in mammalian papillary cells (Sasaki and Higashi , 1983). Possibly, the increase in membrane surface area of ODE cells plays a role in ion regulation or protein resorption (Garant, 1970; Herold, 1974; Prostak and Skobe, 1986). On one hand, the incorporation of myoglobin and HRP into fuzzy coated vesicles of the ODE cells is similar to that observed in mammalian papillary layer cells (Skobe and Garant, 1974; Sasaki and Higashi, 1983). However, we did not observe the subsequent translocation of any tracer within lysosomal structures at the time intervals observed. In fact, teleost mineralization stage ODE cells do not have an extensive Golgi-LysosomeEndoplasmic-Reticulum (GERL) complex (Garant, 1970; Herold, 1974; Prostak and Skobe, 1986). Therefore, it is most likely that the membrane invaginations have a primary function in ion regulation, possibly by ATPase activity. This function is currently under examination. The basal cell junctions between mineralization stage ODE and IDE cells were impermeable to lanthanum, and therefore may provide a diffusion barrier to exogenous materials from reaching the mineralizing surface. This is comparable to studies on mammalian tooth development depicting tracer impermeable junctions in ruffle- and smoothended ameloblasts (Josephsen, 1984; Takano and Ozawa, 1984). However, in this study, intracardially injected HRP did permeate the mineralization stage ODE-IDE cell junctions

Fig. 13. Maturation stage IDE cells from a cichlid injected intracardially with HRP. Tracer tills apical dilations (arrow) of the plasma membrane and extends to the enameloid surface. ~28,000. Fig. 14. Control cichlid maturation stage ODE cells post-fixed with osmium containing ferricyanide. Densely stained material is observed within the plasma membrane invaginations of the ODE cells. This staining does not continue to the IDE intercellular space. X 12,000.

PROSTAK

to reach the mineralizing enameloid surface. Due to the small size of the intracardially injected specimen, it is possible that the injection pressure (which is unknown) was sufficiently high to cause an artifactual permeability of the cellular junctions similar to the effect observed in mammalian epithelial cell junctions (Schneeberger and Karnovsky, 1971). Intramuscularly injected HRP and myoglobin, unlike lanthanum, did permeate past the basal junctions of maturation stage IDE cells 1 hr post injection. HRP may have induced cichlid ODE-IDE cell junction permeability similar to that observed in mammalian epithelial junctions (Cotran and Karnovsky, 1961). Myoglobin appeared to be less permeable through IDE basal junctions than HRP, similar to that effect reported in mammalian tissues (Anderson, 1972; Mazariegos and Hand, 1984). However, the obserthat the apical junctions of vation mineralization stage IDE cells remained impermeable to either intramuscularly injected HRP or myoglobin up to 6 hr post injection, supports the contention that these junctions are tight. This implies that mineralization stage IDE cells do limit the access of extracellular materials to the mineralizing enameloid. Due to its toxicity, lanthanum could not be used for longer time intervals to determine whether it would also penetrate the basal ODE-IDE cell junctions, as did the organic tracers. In teleost secretory stage tooth buds, lanthanum does not penetrate the capillary endothelial wall freely, possibly due to fewer fenestrations than observed in the capillaries of maturation stage tooth organs. In cases where lanthanum was observed extravascularly, it did not permeate the ODE cells very deeply, even though no intercellular junction existed. This may be due to insufficient perfusion time, pressure, or volume, or ion binding to cell surface materials. However, the difference in vascular type between secretory and maturation stages may explain the retention of lanthanum in secretion but not maturation stage tooth buds. The relative lack of 45 Ca observed in early stage tooth matrix of HopIognathus (Inoue et al., 1973) could possibly

be due to a relative lack of vasculature surrounding the ODE. The permeability of myoglobin and HRP, 2 hr post injection, through cervically located IDE cells of secretory stage tooth buds, but not through more coronally located tooth buds is similar to that observed in rodent amelgenesis (Kallenbach, 1980a). However, a morphological basis for the cessation of organic tracer flow through the IDE cells to the enameloid matrix could not be determined during the time intervals examined, but certainly does not rely on perfusion pressure as may be the case for lanthanum. In mammals, the cessation of tracer flow may be due to the formation of more complete apical and basal terminal web complexes (Kallenbach, 1980a). However, in teleosts, the IDE junctional complexes are lacking or variable in location (Shellis and Miles, 1976; Prostak and Skobe, 1986a). The apparent flow of the organic tracers to the cervical enameloid matrix through both the IDE and odontoblasts in secretory stage tooth buds suggests that ions such as calcium and phosphates would also diffuse similarly. However, mineralization of the secreted enameloid matrix does not occur until its full thickness is deposited. Other factors, such as non-collagenous proteins, Gla- proteins, and proteolipids, may be involved in the regulation of mineralization of mammalian collagen (reviewed by Boskey, 1984). Protease activity detected in mammalian enamel matrix (Suga, 1970) is also found at some stage of teleost enameloid matrix (Kawasaki et al., 1987). How these factors may be involved in the mineralization of ectodermally derived collagen matrix of teleost enameloid has yet to be determined, but from the results of this study, it is apparent that the distribution and type of vascular elements (Fenestrated or not) also play a role in the access of materials to the enameloid organ cells and forming matrix.

Acknowledgements

This study was supported by NIDR Grant+ DE07677 to Dr. K. Prostak.

ENAMELOID

ORGAN

OF DEVELOPING

TELEOST

FISH TEETH

429

References Anderson,

W. 1972. The use of exogenous

myoglobin

of protein by the renal tubule. 1. Histochem.

as an ultrastructural

Cytochem.,

tracer.

Reabsorption

and translocation

20,672-684.

Bawden, .I., Crenshaw, M., Takano, Y. and Hammarstrom, L. 1982. Ion transport through the enamel organ-an update. .I. Dent.Res., 61,1551-1554. Boskey, A. 1984. Overview of cellular elements and macromolecules implicated in the initiation of minerahzation. In The Chemistry and Biology of Mineralized Tissues (ed W. Butler), Ebsco Media, Birmingham, Alabama. Cotran, R. and Karnovsky, M. 1961. Vascular leakage induced by horseradish peroxidase in the rat, froc. Sot. Exp. Biol. Med., 126,557-561. Crenshaw, M. and Takano, Y. 1982. Mechanisms by which the enamel organ controls calcium entry into developing enamel. 1. Dent. Res., 61, 15761579. Frank, R. 1979. Tooth enamel: Current state of the art. J. Dent. Res., 58(b) 684-693. Garant, P. 1970. Observations on the ultrastructure of the ectodermal component during odontogenesis in Helostoma temmincki. Anat. Rec., 166, 167-188. Herold, R. 1974. Ultrastructure of odontogenesis in the pike (Esox lucius). Role of dental epithelium and formation of enameloid layer. J. Ultrawuct. Res., 48,435-454. Inoue, Y., Uchino, S., Asano, T, Hasegawa, S. and Fujioka, Y. 1973. The mineralization pattern of developing fish tooth enameloid using X-ray probe microanalyzer, historadiography and radioactive tracer (15Ca). J. Nihon Univ. Sch. Dent., 15,52-57. Isokawa, S., Satamura, I., Yamaguchi, K., Yoshie, T. and Morimoto, M. 1970. Historadiographic observations on the outer dentin in certain osseous fishes. 1. Nihon Sch. Dent., 12, 1-15. Josephsen, K. 1984. Lanthanum tracer study on permeability of ameloblast junctional complexes in maturation zone of rat incisor enamel organ. In Tooth Enamel IV eds R. Fearnhead and S. Suga), pp. 251-255. Elsevier Science Publishers, Holland. Josephsen, K. and Fejerskov, 0. 1977. Ameloblast modulation in the maturation zone of the rat incisor enamel organ. A light and electron microscope study. J. Anat., 124,5-70. KaIIenbach, E. 1966. Electron microscopy of the papillary layer of rat incisor enamel organ during enamel maturation. 1. Ultrastruct. Res., 14,51b533. Kallenbach, E. 1980a. Fate of horseradish peroxidase in the secretion zone of the rat incisor enamel organ, Tissue & Cell, 12,491-501. Karnovsky, M. 1965. A formaldehyde glutaraldehyde fixative of high osmolarity for use in electron microscopy. 1. Cell Biol., 27, 137A (abst.). Kawasaki, K., Shimoda, S. and Fukae, M. 1987. Histological and biochemical observations of developing enameloid of the sea bream. Adv. Dent. Ref., 1,191-195. Martineau-Doize, B., McKee, M., Warshawsky, H. and Bergeron, .I. 1986. In viva demonstration by radioautography of binding sites for insulin in liver, kidney, and calcified tissues of the rat. Anat. Rec., 214,13@140. Mazariegos, M. and Hand, A. 1984. Horseradish peroxidase (HRP) damages cell membranes and junctions in the rat parotid gland. J. Dent. Res., 63,660. McKee, M., Martineau-Do&e, B. and Warshawsky, H. 1986. Penetration of various molecular weight proteins into the enamel organ and enamel of the rat incisor. Archs. oral Biol., 31,287-296. Prostak, K. and Skobe, Z. 1985. The effects of colchicine on the ultrastructure of the dental epithelium and odontoblasts of teleost tooth buds. 1. Craniofact. Genet. and Develop. Biol., 5,7M8. Prostak, K. and Skobe, Z. 1986a. Ultrastructure of the dental epithelium and odontoblasts during enameloid matrix deposition in cichlid teeth. J. of Morphol., 187,159-172. Prostak, K. and Skobe, Z. 1986b). Ultrastructure of the dental epithelium during enameloid mineralization in a teleost fish, Cichlosoma cyanoguttatum. Arch. oral Biol., 31,7%85. Reith, E. and Coty, V. 1962. Autoradiographic studies on calcification of enamel. Arch. oral Bio[. ,7,365-372. Reith, E. and Coty, V. 1967. The absorptive activity of ameloblasts during the maturation of enamel. Anat. Rec., 157,577-588. Sasaki, T. 1984. Morphology and function of maturation ameloblastsin kitten tooth germs, J. Anat., 138,3>342. Sasaki, T. and Higashi, S. 1983. A morphological, tracer and cytochemical study of the role of the papillary layer of the rat incisor enamel organ during enamel maturation. Archs oral Biol., 28,201-210. Schneeberger, E. and Karnovsky, M. 1971. The influence of intravascular fluid volume on the permeability of newborn and adult mouse lungs to ultrastructural protein tracers. J. Cell Biol., 49,319-334. Shellis, P. and Miles, A. 1976. Observations with the electron microscope on enameloid formation in the common eel (AnguiIIa anguilla: Teleostei). Proc. R. Sot. Land., 194,253-269. Skobe, 2. and Garant, P. 1974. Electron microscopy of horseradish peroxidase uptake by papillary cells of the mouse incisor enamel organ. Archs oral Biol., 19,387-395. Suga, S. 1959. Amelogenesis. ht. Dent. J., 9,394420.

430

PROSTAK

Suga, S. 1970. Histochemical observations of proteolytic enzyme activity in the developing dental hard tissues of the rat. Arch. oral Biol., 15,55-558. Suga, S. 1983. Comparative histology of the progressive mineralization pattern uidevelopmg enamel. In Mecharrwns of Tooth Enamel Formation (ed S. Suga), pp. 167-203, Quintessence Publishmg Co. Inc.. Tokyo. Takano, Y. and Ozawa, H. 1980. Ultrastructural and cytochemical observations on the alternating morphologic changes of the ameloblasts at the stage of enamel maturation. Archs. his. Jap., 43,385-399. Takano, Y. and Ozawa, H. 1984. Autoradiographic and tracer experiments on the exit route for the resorbed organic matrix of the enamel at the stage of maturation, In ToothEnamel IV(eds R Fearnhead and S. Suga) pp. 271-27.5. Elsevier Science Publishers. Holland. Takano, Y. and Crenshaw. M. 1980. The penetration of intravascularly perfused lanthanum Into the amelohlast layer of developing rat molar teeth. Archs. orul Biol., 25,505-511. Toda, Y.. Hasegawa, T., Komoda, T., Tsuchida, S., Yamaguchx. K. and Kancko, T. 1970. Study on the developing enameloid of a fish: (Hoplognathus fasciatus) II. the effects of tetracycline injection. J. Nihon Univ. Sch. Dem. 12,50-53. Uchida, T. McKee, M. and Warshawsky, H. 1987. A radmautographic study ot the effects of vinblastine on the fate of injected 45 calcium and (12X]-insulin in the rat incisor. Arch. oral Bwl.. 32,433437. Wennberg, A. and Bawden. J. 1978. Comparison of “P with “Ca distribution in developing rat molar enamel in viva and in vitro. 1. Denr. Res., 57, I1 1-l 17. Warshawsky, H. 1978. A freeze-fracture study of the topographic relationship between inner enamel- secretory ameloblasts in the rat incisor. Amer. J. Amt.. 152, 15s208. Warshawsky, H. and Smith, C. 1974. Morphological classification of rat incisor ameloblasts. Anur. Rec., 179,42.%446. Warshawsky, H. 1984. Ultrastructural studies on amelogenesis. In The Chemistr_v and Biology of Mineralized Tksue~ (ed W. Butler). 3.%45 Ebsco Media. Inc. Birmingham, Ala.