THE ANATOMICAL RECORD 249:86–95 (1997)

Immunodetection of Amelogenin-Like Proteins in the Ganoine of Experimentally Regenerating Scales of Calamoichthys calabaricus, a Primitive Actinopterygian Fish L. ZYLBERBERG,1* J.-Y. SIRE,1 AND A. NANCI2 ‘Formations Squelettiques’, Laboratoire d’Anatomie Compare´e, Universite´ Paris, Paris, France 2Laboratory for Electron Microscopy, Department of Stomatology, Faculty of Dentistry, Universite´ de Montre´al, Centre-Ville, Que´bec, Canada 1Equipe

ABSTRACT Background: The account of the present study is to test our previous hypothesis that ganoine, a highly mineralized layer found at the scale surface of primitive actinopterygian fish, could be homologous with the enamel covering the crown of vertebrate teeth. Methods: Immunocytochemical techniques have been carried out on regenerating scales of a primitive polypterid, Calamoichthys calabaricus, with three antibodies to mammalian amelogenins. Results: The present study provides the first evidence that ganoine contains molecules which cross-react with mammalian amelogenin proteins. Conclusions: This result is consistent with our previous findings that ganoine and enamel can be considered as homologous tissues. Moreover, the presence in ganoine of a primitive actinopterygian of amelogenin-like proteins, which share epitopes with amelogenins of mammalian enamel, indicates that the gene(s) coding for these proteins appeared earlier than previously suggested and supports the hypothesis that amelogenins show a highly conserved structure through vertebrate evolution. Anat. Rec. 249:86–95, 1997. r 1997 Wiley-Liss, Inc. Key words: ganoine; amelogenins; immunocytochemistry; ganoid scales; actinopterygian fish; Calamoichthys calabaricus; scale regeneration Ganoine (Williamson, 1849) is a highly mineralized, superficial tissue covering different elements of the dermal skeleton (cranial bones, fin-rays, and scales) in numerous fossil taxa (Ørvig, 1978a–c; Richter and Smith, 1995) and in a few living primitive actinopterygian fish, the Cladistia (Polypteridae) and Ginglymodi (Lepisosteidae). It is deposited as single monolayer in odontodes, dermal units homologous with teeth (Schaeffer, 1977; Reif, 1982; Smith and Hall, 1990), or as superimposed layers in more advanced stages of evolution of the dermal skeleton (e.g., odontocomplexes in ganoid scales; Ørvig, 1967, 1977). Ganoine has been considered homologous with the highly mineralized layer coating the crown of vertebrate teeth, and consequently has been thought to be either enamel proper or enameloid, the two types of highly mineralized tissues which are recognized in vertebrate teeth. These tissues differ in their developmental origin: enamel is an entirely ectodermal production whereas enameloid is formed by a cooperative secretion between cells from ectoderm and mesenchyme (Sasagawa and Akai, 1992; Sasagawa, 1995; Smith, 1995). For a long time the tissue origin of the cells producing ganoine (mesenchymal, ectodermal, or r 1997 WILEY-LISS, INC.

both) has been a subject of controversy (e.g., Goodrich, 1907; Kerr, 1952; Ørvig, 1967; Moss 1968a,b; Meinke, 1982; Thomson and McCune, 1984). Ultrastructural examination of the developmental origin of ganoine in regenerating scales of Polypteridae (Sire et al., 1987) and Lepisosteidae (Sire, 1994) strongly suggested that in these living primitive actinopterygians ganoine is entirely an ectodermal product, as is enamel in tetrapod teeth. The basal epidermal cells show a differentiation pattern similar to that described for the ameloblasts during mammalian tooth morphogenesis (Sire et al., 1987; Sire, 1994). They synthesize a non-collagenous organic matrix, the ganoine matrix, within which long crystallites of apatite twisting in bundles toward the scale surface subsequently appear. Furthermore, ganoine matrix mineralizes to a high degree like a

Contract grant sponsor: CNRS (LZ, JYS); contract grant sponsor: MRC of Canada (AN). *Correspondence to: Louise Zylberberg, Universite´ Paris 7 - Denis Diderot, Laboratoire d’Anatomie Compare´e, 2, place Jussieu, 75251 cedex 05, France. E-mail: [email protected] Received 14 November 1996; accepted 24 March 1997.

87

AMELOGENIN-LIKE PROTEINS IN GANOINE

‘‘true’’ enamel (Sire, 1995) and shows mineral contents similar to that of tetrapod tooth enamel (Ørvig, 1967). Two broadly defined groups of proteins have been described in the enamel matrix of mammalian teeth: amelogenins (Eastoe, 1964) and non-amelogenins (reviewed in Smith and Nanci, 1996). Among these, amelogenins are the predominant proteins (over 90% of the protein content) found in the developing enamel matrix (reviews by Brookes et al., 1995; Deutsch et al., 1995; Simmer and Fincham, 1995). These low molecular weight proteins (#30kDa) are hydrophobic and have been proposed to play a crucial role in controlling crystallite size, morphology, and orientation (Aoba et al., 1992; Fincham et al., 1992; Diekwisch et al., 1993). Using polyclonal antibodies against amelogenin and non-amelogenins (enamelin) proteins of various mammalian species, enamel-like proteins have been immunolocalized in the enameloid of the teeth in lower vertebrates (Herold et al., 1980; Slavkin et al., 1983a; Samuel et al., 1987), in carp scales and teeth (Krejsa et al., 1984), and even in hagfish ‘‘teeth’’ (Slavkin et al., 1983a; Slavkin and Diekwisch, 1996). Using monoclonal antibodies, Herold et al. (1989) claimed that only enamelin proteins are present in sharks, bony fish, and larval amphibians, whereas both enamelins and amelogenins are found in adult amphibians, reptiles, and mammals. They suggested that 1) the distinction between tooth enamel and enameloid coincides with a differential distribution of the enamel proteins, amelogenins being absent in enameloid, and 2) enamel evolved from enameloid and the gene for amelogenin developed in tetrapods only. This is in contradiction with the observation that mouse amelogenin cDNA hybridizes with the genomic DNA of a teleost fish at high stringency (Lyngstadaas et al., 1990). This result, and our recent findings that ganoine in living primitive actinopterygian fish shows characteristics of a true ectodermal enamel and can be homologized with tetrapod enamel (Sire et al., 1987; Sire, 1994, 1995), reopens the question of the origin of the amelogenin gene(s) during vertebrate evolution. Thus, in order to identify amelogenin-like proteins in the ganoine of the scales of a polypterid, Calamoichthys calabaricus, we have applied post-embedding colloidal gold immunocytochemistry using three antibodies developed against mammalian amelogenin from different sources. MATERIALS AND METHODS

Immunodetection of amelogenin-like proteins in polypterid scales must be carried out during the first stages of ganoine formation, since, as with enamel, they may be largely lost as the tissue matures and acquires an extremely high degree of mineralization. Moreover, it is difficult to obtain young specimens which show the first stages of ganoine formation. Ganoine deposition occurs cyclically at unforeseeable periods in growing scales of adult specimens. However, in experimentally regenerating scales, we can provoke ganoine formation in C. calabaricus approximately three months after scale removal (Sire et al., 1987), a time which shows both stages of ganoine deposition and maturation. Adult specimens of the polypterid fish C. calabaricus, used as experimental animals, were raised in a tank at

25°C and fed daily with Tenebrio and Chironomus larvae. Scale Regeneration

The fishes were anaesthetized by immersion in MS222 at a concentration of 1:5,000. Some scales were removed from the posterior left side of the flank without extensive damage to the skin. Then the fish were left to regenerate their scales for a period of three months (Sire et al., 1987). Tissue Preparation for Routine Morphology and Immunocytochemistry

The three-month regenerated scales were readily distinguished from normal ones since they differ morphologically (Sire et al., 1987). The regenerated scales were dissected and fixed by immersion for 2 hr at 4°C in 1% glutaraldehyde in 0.1 M sodium cacodylate buffer (pH 7.4), washed in cacodylate buffer overnight, and postfixed in 1% potassium ferrocyanide-reduced osmium tetroxide for 2 hr (Neiss, 1984). Some samples were decalcified for 15 days at 4°C using 0.1 M EDTA added to the fixative solution. All the samples were dehydrated in a graded series of ethanol, infiltrated, and embedded in Epon. Semithin sections (1 µm thick), stained with buffered toluidine blue solution, were examined by light microscopy to select appropriate areas for transmission electron microscopy examination (TEM) and for immunolabeling. Thin sections used for ultrastructural study were double-stained with uranyl acetate and lead citrate, and were viewed at 80 kV in a Philips 201 transmission electron microscope. Immunocytochemical Labeling

Tissue sections were processed for postembedding protein A-gold immunolabeling, as previously described (Nanci et al., 1987). Briefly, the grid-mounted tissue sections were first treated with sodium metaperiodate (Bendayan and Zolliger, 1983) and then floated for 15 min on a drop of 0.01M phosphate buffered saline (PBS) containing 1% ovalbumin (Oval), pH 7.4. They were then transferred for 1 hr onto a drop of rabbit antirecombinant mouse amelogenin [M179; courtesy of J.P. Simmer, University of Texas Health Science Center at San Antonio (Simmer et al., 1994)] or sheep antiporcine amelogenin antibodies [Bio-Gel peak F (LRAP) and C (SAP) affinity purified amelogenins; courtesy of H. Limeback, Faculty of Dentistry, University of Toronto (Limeback and Simic, 1990)]. After incubation with the primary antibodies, the sections were washed with PBS and again blocked with PBS-Oval. Sections incubated with sheep primary antibodies were then incubated with a secondary rabbit anti-sheep IgG (Cappel, Organon Teknika, Scarborough, ON) antibody for 1 hr, washed, and blocked with PBS-Oval. All grids were subsequently placed on a drop of protein A-gold for 30 min to reveal the sites of antigene–antibody binding. The protein A-gold complex was prepared as described by Bendayan (1995) using colloidal gold particles of approximately 14 nm (Frens, 1973). After incubation with the protein A-gold complex, the grids were jetwashed with PBS followed by dH2O, stained with

88

L. ZYLBERBERG ET AL.

Fig. 1. Schematic drawing of a longitudinal section through a non-regenerated ganoid scale of Calamoichthys calabaricus, showing the location of the ganoine layer on the scale surface. Abbreviations for figures 1 and 2: bp 5 osseous basal plate; cv 5 vascular canal; d 5 dermis; de 5 dentine; ep 5 epidermis; ga 5 ganoine; iel 5 inner epidermal layer; is 5 isopedine; pga 5 preganoine; sf 5 Sharpey’s fibers.

uranyl acetate and lead citrate, and examined by TEM using a JEOL JEM 1200 EX-II operated at 60 kV. As negative controls, sections were incubated with rabbit anti-sheep IgG and a rabbit anti-rat bone sialoprotein (BSP) (LF-87-; courtesy of L.W. Fisher, NIH, Bethesda, MD) followed by protein A-gold or protein A-gold alone. All incubation steps were carried out at room temperature. Positive controls for the three antibodies to enamel proteins used in this study can be found in Nanci et al. (1996a). RESULTS

Before describing the sites of detection of amelogeninlike proteins within the ganoine matrix, the organization of a ganoid scale in a polypterid and the main events occurring before and during ganoine formation in experimentally regenerating scales will be reviewed (see Sire et al., 1987; Sire, 1995, for detailed descriptions). Ganoine Formation

A non-regenerated ganoid scale of a polypterid is composed of a thick basal plate made of cellular bone covered by two layers of dental tissues, dentine and ganoine, constituting a so-called odontocomplex (Ørvig, 1968). The bony plate is anchored into the dermis by large Sharpey’s fibers arising from the deep surface of the scale. The dentine is reduced and the ganoine is stratified, hypermineralized, and covered by the epidermis (Fig.1). Within three months, regenerated scales of C. calabaricus have nearly regained their initial shape and show a structural organization similar to that of nonregenerated scales, except that dentine is absent (Fig. 2). During these three months, the bony plate is regenerated and the epidermis comes in close contact with the bone surface. The inner epidermal layer (IEL) cells differentiate into columnar ameloblast-like cells and deposit ganoine matrix. As the ganoine layer thickens, it becomes more mineralized (Fig. 3). During these events the scale surface is always lined by the IEL cells, which flatten where the ganoine layer is well-mineralized (Fig. 4). At the ultrastructural level, the first deposited ganoine matrix appears as rounded patches deposited close to the IEL cell surface, surrounded by cytoplasmic extensions. The patches are located in the superficial

Fig. 2. Schematic drawing of part of a three-month regenerated scale of Calamoichthys calabaricus, showing the newly deposited ganoine.

region of the scale, within the collagen fibrils of the osseous tissue (Fig. 5). These patches contain a thin fibrillar material and long crystallites showing a radial organization. Then adjacent patches fuse to form an uninterrupted layer of ganoine, which thickens in direct contact with the IEL cells. The long (250 nm) and thin (8 nm diam.) crystallites become oriented perpendicular to the IEL cell surface (Fig. 6) like the organic fibrillar material seen in demineralized sections (Fig. 7). At this stage of ganoine maturation, the plasmalemma of the IEL cells is smooth and cytoplasmic extensions are no longer observed. As the ganoine mineralizes, the organic matrix becomes less abundant and its spatial organization less obvious, a process similar to that described in mammalian enamel (Warshawsky and Smith, 1974). A thin membrane, the ganoine membrane, then appears between the IEL cell surface and the maturing ganoine (Fig. 8). As the ganoine becomes more and more mineralized, the ganoine membrane thickens and differentiates (Fig. 9). The ganoine membrane is attached to the IEL cells by a network of thin filaments (Fig. 9). Immunocytochemical Labeling

The detection of amelogenin-like proteins was carried out in two stages of ganoine formation, i.e., during the deposition phase (Figs. 10 and 11, corresponding to Figs 5, 6) and during the first stages of maturation (Fig. 12, corresponding to Figs 7, 8). All antibodies used (SAP, LRAP, and M179) showed an immunoreactivity against proteins located within the ganoine matrix. However, the density of the labeling differed according to the stage examined and the antibody used. Ganoine matrix deposition

During the first stages of ganoine formation, enamel proteins were immunodetected with LRAP antibodies in the small-sized patches of ganoine located close to the IEL cells and in association with membrane invaginations at the apical surface of the IEL cells (Fig. 10). The antibodies to Bio-Gel peak C affinity purified porcine amelogenins showed the most intense reactivity, and the gold particles appeared to be uniformly distributed over the ganoine matrix (Fig. 11a). The labeling obtained with the other two antibodies, porcineLRAP and murine recombinant, M179, was considerably weaker (Fig. 11b,c).

Fig. 3. One µm-thick longitudinal section of a three-month regenerated scale of Calamoichthys calabaricus, showing the darkly stained (metachromatic with toluidine blue) ganoine (pg) deposited on the surface of the osseous basal plate (bp) and lined by the cuboidal inner epidermal layer (IEL) cells. d 5 dermis. Scale bar 5 20 µm. Fig. 4. One µm-thick longitudinal section through another scale of the same specimen showing the ganoine maturation stage. The mineralized ganoine (g) appears unstained. The IEL cells are flat. d 5 dermis. Scale bar 5 20 µm. Figs. 5 to 9. TEM. Longitudinal cross-section of a three-month regenerated scale of Calamoichthys calabaricus. Details of the interface between the inner epidermal layer cells and the upper region of the scale. Figs. 7 and 8: EDTA decalcified samples. All figures at the same magnification. Scale bar 5 0.5 µm. Fig. 5. Patches of ganoine (pg) containing long crystallites organized in a radiating pattern are close to the IEL cells and are surrounded by collagen fibrils (arrows) of the osseous basal plate.

Fig. 6. The forming ganoine layer contains long and thin crystallites (arrows) and microfilaments oriented perpendicularly to the IEL cell plasma membrane. Fig. 7. The loose aspect of the matrix in the deep part of the ganoine layer denotes the beginning of the maturation process. The arrow points to microfilaments of the ganoine matrix which are perpendicular to the IEL cell surface. Fig. 8. In this advanced stage of maturation of the ganoine matrix, the microfilaments no longer show a preferential orientation and form a loose network. The arrow indicates the anlage of the ganoine membrane which forms below the IEL cells. Fig. 9. Mature ganoine (g) made of long and thin crystallites oriented perpendicularly to the surface of the IEL cells. The well-developed ganoine membrane consists of a homogenous layer (asterisk) in contact with the ganoine and of microfilaments (arrows) perpendicular to the IEL cell membrane and linking it to the cell by numerous hemidesmosomes (arrowheads). Vesicles containing a fuzzy content are observed in the IEL cells.

90

L. ZYLBERBERG ET AL.

Fig. 10. Immunocytochemical preparation illustrating the labeling obtained with the LRAP antibody to porcine amelogenin when the first patches of ganoine are formed. Note the presence of gold particles over the patches of ganoine (g). However, little or no labeling is observed over

the collagen fibrils (co) forming the basal plate. Some gold particles are also observed on the material within membrane invaginations of the apical surface of IEL cells. Scale bar 5 0.5 µm.

Ganoine maturation

Polypterus senegalus, another polypterid (Sire, 1995), and for Lepisosteus tropicus, a lepisosteid (Sire, 1994). Thus, regeneration, which largely repeats ontogeny, is a good means to induce ganoine deposition on a large surface of the scale according to a well-defined sequence at a given time. Previous structural observations (Sire et al., 1987) have shown that the synthesis and secretion of ganoine organic matrix by the IEL cells follows a similar pathway as enamel proteins in ameloblasts (Nanci et al., 1985; Nanci and Smith, 1992). The antibody to porcine amelogenin Bio-Gel peak C (SAP) resulted in a more intense immunoreaction over ganoine than the other two antibodies. This differential labeling contrasts with what is observed in rat enamel, where all three antibodies give high densities of labeling. The more intense reactivity observed with the SAP antibody could be explained by the possibility that the antigene in ganoine is recognized with more avidity and/or that it reacts with more than one protein and/or its breakdown products. During ganoine maturation, gold particles are distributed close to or within the ganoine membrane, a well-organized layer which remains unmineralized at the ganoine surface (Zylberberg et al., 1985; Sire et al., 1987; Sire, 1994). The lack of mineralization of this structure could, in part, be accounted for by the persistence of supramolecular aggregates of proteins, which, like mammalian amelo-

A relatively dense and homogeneous distribution of the gold particles was observed over both ganoine and the ganoine membrane with the SAP antibody (Fig. 12a). Some gold particles were also present over the material within small vesicular structures found in some IEL cells (Fig. 12a). With the other two antibodies, LRAP and M179, the labeling over ganoine was generally weaker, but there was a concentration of gold particles over the ganoine membrane present at the interface between the cells and the extracellular matrix (Fig. 12b,c). Control incubations resulted in only a few gold particles randomly distributed throughout the tissue sections (Fig. 13a,b). Since the antibodies used resulted in differential labelings, they also serve as internal control for non-specific sticking. DISCUSSION Immunolocalization of Amelogenin-Like Proteins in Ganoine

The present study provides the first evidence that ganoine contains molecules which cross-react immunologically with mammalian amelogenin proteins. Light and TEM observations of the regenerating scales of the polypterid C. calabaricus confirm that ganoine matrix is entirely produced by the IEL cells, as previously shown for this species (Sire et al., 1987), for

AMELOGENIN-LIKE PROTEINS IN GANOINE

Fig. 11. Immunocytochemical preparations illustrating the labeling obtained when patches of ganoine fuse to form a continuous layer. (a): SAP antibody to porcine amelogenin. Gold particles are found throughout the ganoine matrix. (b): LRAP antibody to porcine amelogenin. (c): Antibody to murine recombinant amelogenin (M179). Note the paucity of gold particles obtained with these two antibodies over ganoine. co 5 collagenous matrix of the basal plate; g 5 ganoine. Scale bar 5 0.5 µm.

91

Fig. 12. Immunocytochemical preparations at the stage of ganoine membrane formation. (a): SAP antibody to porcine amelogenin. Gold particles are present over the ganoine and over the ganoine membrane (arrowheads). Some gold particles are also seen over material within vesicles of the IEL cells (arrow). (b): LRAP antibody to porcine amelogenin. Note the weaker immunolabeling over the ganoine but not over the ganoine membrane (arrowheads). (c): Antibody to murine recombinant amelogenin (M179). Gold particles are scarce over the ganoine (g) but appear to accumulate at the interface between ganoine and the ganoine membrane (arrowheads). Scale bar 5 0.5 µm.

92

L. ZYLBERBERG ET AL.

Fig. 13. Controls. (a): Control for LRAP antibody. The sections were incubated with the secondary antibody and protein A-gold only. (b): Labeling obtained with an unrelated antibody (rabbit anti-rat bone

sialoprotein). In both sections, very few gold particles are randomly distributed throughout the tissue. co 5 collagen fibrils of the basal plate; g 5 ganoine. Scale bar 5 0.5 µm.

genins, may regulate crystal formation and must be removed prior to final mineralization (Fincham et al., 1992; Fincham and Moradian-Oldak, 1995). Alternatively, this structure may represent a basal lamina-like structure similar to the one found at the interface between maturing enamel and ameloblasts (Nanci et al., 1993). It has been hypothesized that the ganoine membrane could have adhesive properties allowing the soft tissues to keep in contact with the ganoine surface during swimming (Zylberberg et al., 1985). Thus, like the enamel-like proteins detected in the epithelial attachment of the junctional epithelium (Nanci et al., 1995, 1996b), some proteins of the ganoine membrane may have cell-matrix binding functions.The labeling over the ganoine membrane further suggests that some enamel proteins may be multifunctional in that they not only have roles in matrix formation but may participate in other formative events.

enamel (Kemp, 1984; Richter and Smith, 1995). Ganoine, which has been found to be located at the surface of the hard tissues in the dermal scales of a number of lower actinopterygians and perhaps in some acanthodians, has been recently classified as ‘‘enamel-like tissue’’ (Richter and Smith, 1995), and is distinguished from enameloid by its ‘‘pseudoprismatic’’ arrangement of the crystallites (Ørvig, 1967; Schaeffer, 1977). However, tooth enamel and enameloid are definitely believed to differ from a developmental point of view, enamel being entirely ectodermal, whereas enameloid (in sharks and osteichthyans) is both ectodermal and mesodermal in origin. Our previous studies have shown that ganoine is also an ectodermal product, a result which strongly suggests that enamel and ganoine could be homologous tissues (Sire et al., 1987; Sire, 1994, 1995). Our knowledge of the chemical composition of enamel matrix has improved considerably in recent years, primarily as regards mammalian teeth (review in Robinson et al., 1995; Simmer and Fincham, 1995; Smith and Nanci, 1996). The chemical nature of enameloid matrix among non-tetrapods has been analyzed in the scales and teeth of sharks (Moss et al., 1964; Levine et al., 1966; Kawasaki et al., 1980), and in teleost teeth (Moss et al., 1964; Shellis, 1975). Enameloid was found to contain collagen and proteins similar to those found in mammalian enamel. The chemical nature of the ganoine is, however, still unknown. The biochemical properties of the two main classes of enamel proteins, amelogenins and non-amelogenin proteins, have likewise been well defined in mammals.

Evolutionary Implications

Evolution has produced different types of hypermineralized tissues in lower vertebrates, i.e., ganoine, enamel, and enameloid. There is a considerable amount of information regarding the mineral, chemical, and biochemical composition of mammalian tooth enamel; however, less is known about the other tissues in lower vertebrates. Such data are of major importance in view of their evolutionary implications. The wide diversity of crystallite arrangement and the general aspect of mineralization among vertebrates do not allow us to strictly classify ganoine, enameloid, and

AMELOGENIN-LIKE PROTEINS IN GANOINE

Amelogenins are now characterized by amino acid sequences which have a high degree of homology between all the mammalian species investigated, and their gene(s) have been identified and partially sequenced (see the recent reviews by Brookes et al., 1995; Fincham and Moradian-Oldak, 1995; Sasaki and Shimokawa, 1995; Simmer and Fincham, 1995). In an attempt to trace the evolution of enamel proteins, immunocytochemistry using antibodies prepared against mammalian enamel proteins has been extensively used in lower vertebrates. Such studies demonstrated the presence of enamel-like proteins in the enamel or enameloid of various lower vertebrates (Herold et al., 1980; Graham et al., 1981; Slavkin et al., 1983b; Krejsa et al., 1984), including the outer covering of the Pacific hagfish ‘‘teeth’’ (Slavkin et al., 1983a; Slavkin et al., 1991). Nevertheless, most of these studies failed to distinguish amelogenin and non-amelogenin proteins, probably because of the lack of specificity of the polyclonal antibodies used and of the progressive degradation of the enamel proteins during the maturation process (review in Smith and Nanci, 1996). The improvement of biochemical techniques (Graham, 1985; Samuel et al., 1987) and the use of monoclonal antibodies with a high specificity have finally led to the conclusion that amelogenins are absent from the teeth of sharks, bony fish, and larval amphibians, while both non-amelogenin and amelogenin proteins are present in adult amphibian, reptilian, and mammalian teeth (Herold et al., 1989). The immunological studies indicate that proteins belonging to the non-amelogenin group share common antigenic determinants across a wide range of vertebrates, and are present in enamel as well as in enameloid. Thus, it has been concluded that ‘‘enamelins’’ have been conserved throughout 500 million years (mya) and that they have appeared prior to amelogenins during vertebrate evolution. The present study brings a new element into the puzzle of the evolution of the enamel proteins in showing that in the scales of a lower actinopterygian fish (polypterid), ganoine contains epitopes shared with mammalian amelogenins. Indeed, polypterid scales have conserved ancestral characters (i.e., ganoine) which were present in the first osteichthyan fishes, more than 350 mya. These data, 1) support our previous hypotheses that ganoine can be homologized with enamel, and 2) indicate that the gene(s) responsible for the production of such proteins is older than previously supposed by Herold et al. (1989). This finding is also supported by investigations showing that the genomic DNA from a teleost, Anarhichas lupus, hybridizes with a mouse amelogenin cDNA probe (Lyngstadaas et al., 1990). Given that ganoine is structurally similar in polypterids, lepisosteids, and various fossil actinopterygian fish (Richter and Smith, 1995; Sire, 1995), we can suppose that amelogenin-like proteins are also present in all these taxa. Indeed, homology of ganoine and ectodermal tooth enamel strongly suggests that the amelogenin gene(s) was shared at least by the common ancestor of sarcopterygians (including mammals) and actinopterygians. Accordingly, its occurrence is a plesiomorphy within both clades but can be tentatively suggested as an apomorphy of osteichthyans as a whole. Indeed, teeth and scales are considered to

93

originate from a single, modifiable morphogenetic system, the odontodes (i.e. ‘‘dental’’ units) which were present in the first vertebrates, 500 mya (Schaeffer, 1977; Reif, 1982; Smith and Hall, 1990, 1993). When gnathostomes appeared, 450 mya, odontodes were also formed into the buccal cavity, where they evolved into teeth, whereas odontodes persisted on the body surface. During evolution, modifications of the odontodes (changes of tissue organization and alterations of the phenotypes) led to the actual diversity of the postcranial dermal skeleton (various types of scales), whereas teeth have roughly conserved the original ‘‘odontodal’’ structure, probably because of stronger selective pressures. Even if teeth and scales have evolved separately since at least 450 mya, and even if scales have been considerably modified compared to teeth and odontodes, the present study strongly suggests that the products of the amelogenin gene(s) are nevertheless found in the ganoid scales, and thus represent common ancestral proteins. ACKNOWLEDGMENTS

We thank Prof. A. Huysseune (Gent University, Belgium) for her helpful advice and F. Allizard and M. Fortin for technical assistance. TEM and photographic works (L.Z. and J.Y.S.) were performed in the Centre Interuniversitaire de Microscopie Electronique (CIME Jussieu, Ile de France) and immunocytochemical analyses in the Laboratory for Electron Microscopy (A.N.) of the University of Montreal. LITERATURE CITED Aoba, T., S. Shimoda, H. Shimokawa, and T. Inage 1992 Common epitopes of mammalian amelogenins at the C-terminus and possible functional roles of the corresponding domain in enamel mineralization. Calcif. Tissue Int., 51:85–91. Bendayan, M. 1995 Colloidal gold post-embedding immunocytochemistry. Prog. Histochem. Cytochem., 29:1–163. Bendayan, M., and M. Zollinger. 1983 Ultrastructural localization of antigenic sites on osmium-fixed tissues applying the protein A-gold technique. J. Histochem. Cytochem., 31:101–109. Brookes, S.J., C. Robinson, J. Kirkham, and W.A. Bonass. 1995 Biochemistry and molecular biology of amelogenin proteins of developing dental enamel. Arch. Oral Biol., 40:1–14. Deutsch, D., J. Catalano-Sherman, L. Dafni, S. David, and A. Palmon. 1995 Enamel matrix proteins and ameloblast biology. Connect. Tissue Res., 32:97–107. Diekwisch, T., S. David, P. Bringas, V. Santos, and H.C. Slavkin 1993 Antisense inhibition of AMEL translation demonstrates supramolecular controls for enamel HAP crystal growth during embryonic mouse molar development. Development, 117:471–482. Eastoe, J.E. 1964 The chemical composition of bone and teeth. In: Advance in Fluorine Research Dental Caries Prevent, Vol. 3. J.L. Hardwick, H.R. Held, and K.G. Konig, eds. Pergamon Press, Oxford, pp. 5–17. Fincham, A.G., and J. Moradian-Oldak. 1995 Recent advances in amelogenin biochemistry. Connect. Tissue Res., 32:119–124. Fincham, A.G., A.B. Belcourt, D.M. Lyaruu, and J.D. Termine 1982a Comparative protein biochemistry of developing dental enamel matrix from five mammalian species. Calcif. Tissue Int., 34:182– 189. Fincham, A.G., A.B. Belcourt, and J.D. Termine 1982b Changing patterns of enamel matrix proteins in the developing bovine tooth. Caries Res., 16:64–71. Fincham, A.G., E.C. Lau, J. Simmer, and M. Zeichner-David 1992 Amelogenin biochemistry—Form and function. In: Chemistry and Biology of Mineralized Tissues. H.C. Slavkin and P. Price, eds. Elsevier, Amsterdam, pp. 187–201. Frens, G. 1973 Controlled nucleation for the regulation of particles size in monodispersed gold suspension. Nature Phys. Sci., 221:20– 22. Goodrich, E.S. 1907 On the scales of fish living and extinct, and their importance in classification. Proc. Zool. Soc. Lond., 77:751–774.

94

L. ZYLBERBERG ET AL.

Graham, E.E. 1985 Isolation of enamelin-like proteins from blue shark (Prionace glauca) enameloid. J. Exp. Zool., 234:185–191. Graham, E.E., M. Zeichner-David, M. Ferguson, J.D. Termine, D. Eskinazi, and H.C. Slavkin 1981 Phylogenetic studies of amelogenesis. Anat. Rec., 199:98A. Herold, R.C., H.T. Graver, and P. Christner 1980 Immunohistochemical localization of amelogenins in ameloid of lower vertebrate teeth. Science, 207:1357–1358. Herold, R.C., J. Rosenbloom, and M. Granovsky 1989 Phylogenetic distribution of enamel proteins: Immunohistochemical localization with monoclonal antibodies indicates the evolutionary appearance of enamelins prior to amelogenins. Calcif. Tissue Int., 45:88–94. Kawasaki, H., T. Kawaguchi, T. Yano, S. Fujimura, and M. Yago 1980 Chemical nature of proteins in the placoid scale of the blue shark, Prionace glauca L. Arch. Oral Biol., 25:313–320. Kemp, N.E. 1984 Organic matrix and mineral crystallites in vertebrate scales, teeth and skeletons. Am. Zool., 24:965–976. Kerr, T. 1952 The scales of primitive living actinopterygians. Proc. Zool. Soc. Lond., 121:58–78. Krejsa, R.J., N. Samuel, C. Bessem, and H.C. Slavkin 1984 Immunogenetic and phylogenetic comparisons between teleost scale and dental enameloid with mammalian enamel antigens. In: Tooth Morphogenesis and Differentiation, Vol. 125. J.V. Ruch, ed. Colloques INSERM, pp. 369–376 Levine, P.T., M.J. Glimcher, J.M. Seyer, J.I. Huddleston, and J.W. Hein 1966 Noncollagenous nature of the proteins of shark enamel. Science, 154:1192–1194. Limebach, H., and A. Simic 1990 Biochemical characterization of stable high molecular-weight aggregates of amelogenins formed during porcine enamel development. Arch. Oral Biol., 35:459–468. Lyngstadaas, S.P., S. Risnes, H. Nordbø, and A.G. Flønes 1990 Amelogenin gene similarity in vertebrates: DNA sequences encoding amelogenin seem to be conserved during evolution. J. Comp. Physiol., 160:469–472. Meinke, D.K. 1982 A light and scanning electron microscope study of microstructure, growth and development of the dermal skeleton of Polypterus (Pisces, Actinopterygii). J. Zool. Lond., 197:355–382. Moss, M.L. 1968a The origin of vertebrate calcified tissues. In: Current problems of lower vertebrate phylogeny. Proc. of the 4th Nobel Symposium. T. Ørvig, ed. Almqvist and Wiskell, Stockholm, pp. 359–371. Moss, M.L. 1968b Comparative anatomy of vertebrate dermal bone and teeth. I. The epidermal coparticipation hypothesis. Acta Anat., 71:178–208. Moss, M.L., S.J. Jones, and K.A. Piez 1964 Calcified ectodermal collagens of shark tooth enamel and teleost scale. Science, 145:940– 942. Nanci, A., and C.E. Smith 1992 Development and calcification of enamel. In: Calcifications in Biological Systems. E. Bonucci, ed. CRC Press, Boca Raton, pp.313–343. Nanci, A., M. Bendayan, and H.C. Slavkin 1985 Enamel protein biosynthesis and secretion in mouse incisor secretory ameloblasts as revealed by high-resolution immunocytochemistry. J. Histochem. Cytochem., 33:1153–1160. Nanci, A., H.C. Slavkin, and C.E. Smith 1987 Application of high resolution immunocyto-chemistry to the study of the secretory resorptive and degragative functions of ameloblasts. Adv. Dent. Res., 1:148–161. Nanci, A., S. Zalzal, and Y. Kogaya 1993 Cytochemical characterization of basement membranes in the enamel organ in the rat incisor. Histochemistry, 99:321–331. Nanci, A., H. Kawaguchi, T. Ogawa, S. Zalzal, and M.D. McKee 1995 Immunocytochemical cross-reactivity between enamel proteins and rat molar epithelial attachment. J. Dent. Res., 74:417. Nanci, A., J. Hashimoto, S. Zalzal, and C.E. Smith 1996a Transient accumulation of phosphorylated and short-lived glycosylated enamel proteins at interrod and rod growth sites. Adv. Dent. Res., 10:135–149. Nanci, A., T. Ogawa, H. Kawaguchi, S. Zalzal, and M.D. McKee 1996b Interfacing of enamel-related proteins with noncollagenous proteins in rat molars. Connect. Tissue Res. 35:415 (abst). Neiss, W.F. 1984 Electron staining of the cell surface coat by osmiumlow ferrocyanide. Histochemistry, 80:231–242. Ørvig, T. 1967 Phylogeny of tooth tissues: Evolution of some calcified tissues in early vertebrates. In: Structural and Chemical Organization of Teeth, Vol. I. A.E.W. Miles, ed. Academic Press, New York, pp. 45–110. Ørvig, T. 1968 The dermal skeleton; general considerations. In: Current Problems of Lower Vertebrate Phylogeny. Proc. of the 4th

Nobel Symposium. T. Ørvig, ed. Almqvist and Wiskell, Stockholm, pp. 373–397. Ørvig, T. 1977 A survey of odontodes (‘dermal teeth’) from developmental, structural, functional, and phyletic points of views. In: Problems in Vertebrate Evolution. Linnean Society Symposium 4. S.M. Andrews, R.S. Miles, and A.D. Walker, eds. Academic Press, London & New York, pp. 53–75. Ørvig, T. 1978a Microstructure and growth of the dermal skeleton in fossil actinopterygian fish: Birgeria and Scanilepis. Zool. Scripta, 7:33–56. Ørvig, T. 1978b Microstructure and growth of the dermal skeleton in fossil actinopterygian fish: Boreosomus, Plegmolepis and Gyrolepis. Zool. Scripta, 7:125–144. Ørvig, T. 1978c Microstructure and growth of the dermal skeleton in fossil actinopterygian fish: Nephrotus and Colobobus, with remarks on the dentition in other forms. Zool. Scripta, 7:297–326. Reif, W.E. 1982 Evolution of dermal skeleton and dentition in vertebrates. The odontode regulation theory. In: Evolution Biology, Vol. 15. M.K. Hecht, B. Wallace, and G.T. Prance, eds. Plenum Press, New York, pp. 287–368. Richter, M., and M.M. Smith 1995 A microstructural study of the ganoine tissue of selected lower vertebrates. Zool. J. Linn. Soc. Lond., 114:173–212. Robinson, C., J. Kirkham, S.J. Brookes, W.A. Bonass, and R. Shore 1995 The chemistry of enamel development. Int. J. Dev. Biol., 39:145–152. Samuel, N., C. Bessem, P. Bringas, and H.C. Slavkin. 1987 Immunochemical homology between elasmobranch scale and tooth extracellular matrix proteins in Cephaloscylium ventriosum. J. Craniofac. Genet. Dev. Biol., 7:371–388. Sasagawa, I. 1995 Fine structure of tooth germs during the formation of enameloid matrix in Tilapia nilotica, a teleost fish. Arch. Oral Biol., 40:801–814. Sasagawa, I., and J. Akai. 1992 The fine structure of the enameloid matrix and initial mineralization during tooth development in the sting rays, Dasyatis akajei and Urolophus aurantiacus. J. Electron Microsc., 41:242–252. Sasaki, S., and H. Shimokawa 1995 The amelogenin gene. Int. J. Dev. Biol., 39:127–133. Schaeffer, B. 1977 The dermal skeleton in fishes. In: Problems in Vertebrate Evolution. Linnean Society Symposium 4. S.M. Andrews, R.S. Miles, and A.D. Walker, eds. Academic Press, London & New York, pp. 25–52. Shellis, R.P. 1975 A histological and histochemical study of the matrices of enameloid and dentine in teleost fishes. Arch. Oral Biol., 20:183–187. Simmer, J.P., and A.G. Fincham 1995 Molecular mechanisms of dental enamel formation. Crit. Rev. Oral Biol. Med., 6:84–108. Simmer, J.P., E.C. Lau, C.C. Hu, T. Aoba, M. Lacey, D. Nelson, M. Zeichner-David, M.L. Snead, H.C. Slavkin, and A.G. Fincham 1994 Isolation and characterization of a mouse amelogenin expressed in Escherichia coli. Calcif. Tissue Int., 54:312–319. Sire, J.-Y. 1994 Light and TEM study of nonregenerated and experimentally regenerated scales of Lepisosteus aculatus (Holostei) with peculiar attention to ganoine formation. Anat. Rec., 240:189–207. Sire, J.-Y. 1995 Ganoine formation in the scales of primitive actinopterygian fishes, lepisosteids and polypterids. Connect. Tissue Res., 33:213–222. Sire, J.-Y., J. Ge´raudie, F.J. Meunier, and L. Zylberberg 1987 On the origin of ganoine: Histological and ultrastructural data on the experimental regeneration of the scales of Calamoichthys calabaricus (Osteichthyes, Brachyopterygii, Polypteridae). Am. J. Anat., 180:391–402. Slavkin, H.C., and T. Diekwisch 1996 Evolution in tooth developmental biology. On morphology and molecules. Anat. Rec., 245:131– 150. Slavkin, H.C., E.E. Graham, M. Zeichner-David, and W. Hildemann 1983a Enamel-like antigenes in hagfish; possible evolutionary significance. Evolution, 37:404–412. Slavkin, H.C., N. Samuel, P. Bringas, A. Nanci, and V. Santos 1983b Selachian tooth development. II. Immunolocalization of amelogenin polypeptides in epithelium during secretory amelogenesis in Squalus acanthias. J. Craniofac. Genet. Dev. Biol., 3:43–42. Slavkin, H.C., R.J. Krejsa, A.G. Fincham, P. Bringas, V. Santos, Y. Sasano, M.L. Snead, and M. Zeichner-David 1991 Evolution of enamel proteins: A paradigm for mechanisms of biomineralisation. In: Mechanisms and Phylogeny of Mineralization in Biological Systems. S. Suga, ed. Springer-Verlag, Tokyo, pp. 283–289. Smith, C.E., and A. Nanci 1996 The protein dynamics of amelogenesis. Anat. Rec., 245:219–234.

AMELOGENIN-LIKE PROTEINS IN GANOINE Smith, M.M. 1995 Heterochrony in the evolution of enamel in vertebrates. In: Evolutionary Change and Heterochrony. K.J. McNamara, ed. Wiley & Sons, New York, pp. 125–150. Smith, M.M., and B.K. Hall 1990 Development and evolutionary origins of vertebrate skeletogenic and odontogenic tissues. Biol. Rev., 65:277–373. Smith, M.M., and B.K. Hall 1993 A developmental model for evolution of the vertebrate exoskeleton and teeth: The role of cranial and trunk neural crests. Evol. Biol., 27:387–448. Termine, J.D., A.B. Belcourt, P.V. Christner, K.M. Conn, and M.U. Nylen 1980 Properties of dissociatively extracted fetal tooth matrix proteins. I. Principlal molecular species in developing bovine enamel. J. Biol. Chem., 255:9760–9768.

95

Thomson, K.S., and A.R. McCune 1984 Development of the scales in Lepisosteus as a model for scale formation in fossil fishes. Zool. J. Linn. Soc., 82:73–86. Warshawsky, H., and C.E. Smith 1974 Morphological classification of rat incisor ameloblasts. Anat. Rec., 179:423–446. Williamson, W.C. 1849 On the microscopic structure of the scales and dermal teeth of some ganoid and placoid fishes. Philos. Trans. R. Soc. Lond., 139:435–475. Zylberberg, L., J. Ge´raudie, J.-Y. Sire, and F.J. Meunier 1985 Mise en e´vidence ultrastructurale d’une couche organique entre l’e´piderme et la ganoı¨ne du dermosquelette des Polypteridae. C.R. Acad. Sci., 301:517–522.