DPP and DSP are Necessary for Maintaining TGF-ß1 Activity in Dentin.

Porcine dentin sialophosphoprotein (DSPP) is the most abundant non-collagenous protein in dentin. It is processed by proteases into 3 independent proteins: dentin sialoprotein (DSP), dentin glycoprotein (DGP), and dentin phosphoprotein (DPP). We fractionated DPP and DSP along with TGF-ß activity by ion exchange (IE) chromatography from developing pig molars and measured their alkaline phosphatase (ALP)-stimulating activity in human periodontal (HPDL) cells with or without TGF-ß receptor inhibitor. We then purified TGF-ß-unbound or -bound DPP and DSP by reverse-phase high-performance liquid chromatography (RP-HPLC) using the ALP-HPDL system. The TGF-ß isoform bound to DPP and DSP was identified as being TGF-ß1 by both ELISA and LC-MS/MS analysis. We incubated carrier-free human recombinant TGF-ß1 (CF-hTGF-ß1) with TGF-ß-unbound DPP or DSP and characterized the binding on IE-HPLC using the ALP-HPDL system. When only CF-hTGF-ß1 was incubated, approximately 3.6% of the ALP-stimulating activity remained. DPP and DSP rescued the loss of TGF-ß1 activity. Approximately 19% and 10% of the ALP stimulating activities were retained by the binding of TGF-ß to DPP and DSP, respectively. The type I collagen infrequently bound to CF-hTGF-ß1. We conclude that both DPP and DSP help retain TGF-ß1 activity in porcine dentin.

Cytodifferentiation activity of synthetic human enamel sheath protein peptides.

BACKGROUND AND OBJECTIVE: Enamel sheath protein (ESP) is involved in the construction of the enamel sheath during tooth development. The 17 kDa ESP is a one-step cleavage product processed by proteolysis from the N-terminal side of sheathlin (ameloblastin/amelin), one of the porcine enamel matrix proteins. Enamel sheath protein exhibits periodontal ligament and cementum regeneration activity in a buccal dehiscence model in dogs, and promotes the cytodifferentiation of cultured human periodontal ligament (HPDL) cells. The aim of this study was to determine the peptide segment on the C-terminal side sequence of the human ESP that possesses a cytodifferentiation activity on cultured HPDL cells. MATERIAL AND METHODS: The peptides synthesized on the basis of human ESP C-terminal side sequence were tested for their ability to increase the alkaline phosphatase (ALP) and mineralization activity of cultured HPDL cells. The expressions of osteocalcin, osteopontin and bone sialoprotein were measured by semi-quantitative PCR and therefore were determined to be specific indicators of mineralized tissue differentiation. RESULTS: Multiple synthetic peptides from the human ESP increased the ALP activity and stimulated matrix mineralization in long-term cultures of HPDL cells. Semi-quantitative PCR demonstrated the osteocalcin, osteopontin and bone sialoprotein expressions to increase relative to the control values. The peptide SDKPPKPELPGVDF had the strongest cytodifferentiation activity among all the synthetic peptides tested. CONCLUSION: A specific peptide sequence derived from the C-terminal side of the human ESP promotes the cytodifferentiation and mineralization activity of HPDL cells in a cell culture system.

Effects of porcine 25 kDa amelogenin and its proteolytic derivatives on bone sialoprotein expression.

BACKGROUND AND OBJECTIVE: Amelogenins are hydrophobic proteins that are the major component of developing enamel. Enamel matrix derivative has been used for periodontal regeneration. Bone sialoprotein is an early phenotypic marker of osteoblast differentiation. In this study, we examined the ability of porcine amelogenins to regulate bone sialoprotein transcription. MATERIAL AND METHODS: To determine the molecular basis of the transcriptional regulation of the bone sialoprotein gene by amelogenins, we conducted northern hybridization, transient transfection analyses and gel mobility shift assays using the osteoblast-like ROS 17/2.8 cells. RESULTS: Amelogenins (100 ng/mL) up-regulated bone sialoprotein mRNA at 3 h, with maximal mRNA expression occurring at 12 h (25 and 20 kDa) and 6 h (13 and 6 kDa). Amelogenins (100 ng/mL, 12 h) increased luciferase activities in pLUC3 (nucleotides -116 to +60), and 6 kDa amelogenin up-regulated pLUC4 (nucleotides -425 to +60) activity. The tyrosine kinase inhibitor inhibited amelogenin-induced luciferase activities, whereas the protein kinase A inhibitor abolished 25 kDa amelogenin-induced bone sialoprotein transcription. The effects of amelogenins were abrogated by 2-bp mutations in the fibroblast growth factor 2 response element (FRE). Gel-shift assays with radiolabeled FRE, homeodomain-protein binding site (HOX) and transforming growth factor-beta1 activation element (TAE) double-strand oligonucleotides revealed increased binding of nuclear proteins from amelogenin-stimulated ROS 17/2.8 cells at 3 h (25 and 13 kDa) and 6 h (20 and 6 kDa). CONCLUSION: These results demonstrate that porcine 25 kDa amelogenin and its proteolytic derivatives stimulate bone sialoprotein transcription by targeting FRE, HOX and TAE in the bone sialoprotein gene promoter, and that full-length amelogenin and amelogenin cleavage products are able to regulate bone sialoprotein transcription via different signaling pathways.

Cleavage site specificity of MMP-20 for secretory-stage ameloblastin.

Ameloblastin is processed by protease(s) during enamel formation. We tested the hypothesis that MMP-20 (enamelysin) catalyzes the cleavages that generate secretory-stage ameloblastin cleavage products. We isolated a 23-kDa ameloblastin cleavage product from developing enamel and determined its N-terminus sequence. Ameloblastin was stably expressed and secreted from HEK293-H cells, purified, and digested with MMP-20 or Klk4 (kallikrein 4). The digests were analyzed by SDS-PAGE and Western blotting, and cleavage products were characterized by N-terminal sequencing. Six fluorescent peptides were digested with MMP-20 and Klk4 and analyzed by RP-HPLC and by mass spectrometry. MMP-20 cleaved each peptide exactly at the sites corresponding to ameloblastin cleavages catalyzed in vivo. Klk4 cleaved ameloblastin and the fluorescent peptides at sites not observed in vivo, and cleaved at only a single correct site: before Leu(171). We conclude that MMP-20 is the enzyme that processes ameloblastin during the secretory stage of amelogenesis, and we present a hypothesis about the sequence of ameloblastin cleavages.

Degradation of noncollagenous components by neutrophil elastase reduces the mechanical strength of rat periodontal ligament.

BACKGROUND AND OBJECTIVE: We have previously shown that increases in neutrophil elastase in periodontal ligament with chronic periodontitis results in degradation of the noncollagenous components. The purpose of this study was to investigate whether the destruction of noncollagenous components by treatment with elastase in vitro causes changes in the mechanical properties of the periodontal ligament. MATERIAL AND METHODS: The transverse sections of mandibular first molars, prepared from male Wistar rats at 6 wk of age, were digested with 0-50 microg/mL of neutrophil elastase at 37 degrees C for 4 h. Then, their mechanical properties and morphological features were examined. RESULTS: Digestion with elastase dose-dependently decreased the maximum shear stress and failure strain energy density of the periodontal ligament (p < 0.05-0.01). The histological observations after digestion revealed marked degradation of oxytalan fibers, but no marked changes of the collagen fibers, which was confirmed by the detection of very low quantities of hydroxyproline in the digest. The light and scanning electron micrographs showed that the elastase degraded the interfibrillar substances in the periodontal ligament and exposed individual collagen fibrils. CONCLUSION: These results suggest that the increased neutrophil elastase observed in periodontal disease degrades the oxytalan fibers and interfibrillar substances in the periodontal ligament to decrease its mechanical strength.

Splicing determines the glycosylation state of ameloblastin.

In developing porcine enamel, the space between enamel rods selectively binds lectins and ameloblastin (Ambn) N-terminal antibodies. We tested the hypothesis that ameloblastin N-terminal cleavage products are glycosylated. Assorted Ambn cleavage products showed positive lectin staining by peanut agglutinin (PNA), Maclura pomifera agglutinin (MPA), and Limulus polyphemus agglutinin (LPA), suggesting the presence of an O-linked glycosylation containing galactose (Gal), N-acetylgalactosamine (GalNAc), and sialic acid. Edman sequencing of the lectin-positive bands gave the Ambn N-terminal sequence: VPAFPRQPGTXGVASLXLE. The blank cycles for Pro(11) and Ser(17) confirmed that these residues are hydroxylated and phosphorylated, respectively. The O-glycosylation site was determined by Edman sequencing of pronase-digested Ambn, which gave HPPPLPXQPS, indicating that Ser(86) is the site of the O-linked glycosylation. This modification is within the 15-amino-acid segment (73-YEYSLPVHPPPLPSQ-87) deleted by splicing in the mRNA encoding the 380-amino-acid Ambn isoform. We conclude that only the N-terminal Ambn products derived from the 395-Ambn isoform are glycosylated.

Micelle structure of amelogenin in porcine secretory enamel.

Even during the secretory stage of amelogenesis, enamel crystals thicken as amelogenins (the major protein component) decrease. To explain this phenomenon, we propose a model for amelogenin structure and function based upon the hypothesis that amelogenin forms micelles. Solubility and hydrophobicity analyses suggest that all but the hydrophilic amelogenin C-terminal regions aggregate via hydrophobic bonds to form a micelle core. Amelogenin micelles may form super-assemblies via their C-termini (KTKREEVD), which contain complementary positive (KTKR) and negative (EEVD) elements. Disassembly of the micelles through controlled proteolysis provides space for crystal growth. Initial cleavage (by enamelysin) removes the surface-accessible amelogenin C-terminus, exposing the middle portion to cleavage (by EMSP1). As a result, the 13-kDa amelogenin, a rod-shaped domain based upon ultrafiltration and transmission electron microscopy studies, is released. This model explains how amelogenin is able to 'space' and support the ribbon-like crystals and continuously yield space as the crystals thicken, until they are sufficiently mature to support themselves.

Neutrophil elastase is involved in the initial destruction of human periodontal ligament.

BACKGROUND AND OBJECTIVE: It has been reported that noncollagenous proteins may provide mechanical strength to the periodontal ligament. Several proteolytic activities, including that of neutrophil elastase, are reported to increase significantly in periodontal disease. The aim of this study was to investigate the function of neutrophil elastase in the initial destruction of periodontal ligament at early stages of periodontal disease. MATERIAL AND METHODS: The detection and identification of proteinases in chronic periodontitis and healthy periodontal ligament were examined by zymographic and zymo-Western analysis. The morphological changes of periodontal ligament, digested with or without authentic proteinases, were observed using scanning electron microscopy. RESULTS: Increases in neutrophil elastase, plasminogen, and matrix metalloproteinase-9 were detected in periodontal ligament from chronic periodontitis, compared with healthy periodontal ligament. Among these proteinases, only neutrophil elastase digested the intact noncollagenous proteins of periodontium. When human healthy periodontal ligament was directly digested by neutrophil elastase in an in vitro system, the morphological features were quite similar to that of the periodontal ligament in chronic periodontitis . In healthy periodontal ligament, the collagen fibrils are covered with noncollagenous proteins containing 110 kDa acidic glycoprotein, which was degraded initially by the neutrophil elastase. CONCLUSION: It was concluded that neutrophil elastase is involved in the degradation of noncollagenous protein-covered collagen fibrils in the early destructive stages of periodontal disease.

The 17-kDa sheath protein in enamel proteins induces cementum regeneration in experimental cavities created in a buccal dehiscence model of dogs.

BACKGROUND AND OBJECTIVE: Commercially available enamel proteins, such as Emdogain, are clinically used for periodontal regeneration. However, the real mechanisms behind the bioactivities of enamel proteins is still unclear, as enamel proteins have multicomponents. The purpose of this in vivo study was to identify the cementum regeneration-promoting factor in enamel proteins that is clinically used for periodontal regeneration to induce cementum-promotive and osteopromotive activities. MATERIAL AND METHODS: Cementum regeneration, which is an important part of periodontal regeneration, was examined in experimental cavities prepared on a buccal dehiscence model of dogs. The purification of enamel protein with cementum regeneration activity was carried out by gel filtration and ion exchange chromatographies of newly formed secretory enamel. RESULTS: Cementum regeneration activity was found in the aggregate comprising 13-17-kDa sheath proteins along with a small amount of amelogenins, found in the newly formed secretory enamel. In these proteins, cementum regeneration activity was detected upon application of the 17-kDa sheath protein, but not by other lower molecular-weight sheath proteins and amelogenins. However, the purified 17-kDa sheath protein induced cementum regeneration activity only in a small area, although the regenerated cementum was thick. The activity of the 17-kDa sheath protein was believed not to have been a result of contamination by growth factors such as transforming growth factor-beta1 (TGF-beta1) found in the enamel protein, as the application of TGF-beta1 induced weak cementum regeneration activity. CONCLUSION: It is concluded that the 17-kDa sheath protein itself exhibits cementum regeneration activity, although other factors may be needed to demonstrate its full ability.

Enamel matrix derivative gel stimulates signal transduction of BMP and TGF-{beta}.

It has been shown that Emdogain Gel (Emd-Gel) containing enamel matrix proteins promotes biomineralization, such as osteogenesis and cementogenesis, during the regeneration of periodontal tissues. However, the growth factors involved in these activities of Emd-Gel remain unclear. In this study, Emd-Gel was fractionated into 22 sub-fractions by size exclusion chromatography. The osteoinductive factors, TGF-beta and BMP, were examined by a specific luciferase reporter gene assay. In the unfractionated Emd-Gel, TGF-beta-like activity was detected, while BMP activity was not. In contrast, in the fractionated Emd-Gel samples, TGF-beta-like activity was detected from fractions 8 to 13, and BMP-like activity was detected from fractions 4 to 6. Also, it was confirmed that the BMP-like activity in Emd-Gel was inhibited by authentic TGF-beta1 and TGF-beta-like activity. These results indicate that Emd-Gel contains both TGF-beta- and BMP-like growth factors that contribute to the induction of biomineralization during periodontal regeneration.

Porcine amelogenin is expressed from the X and Y chromosomes.

Amelogenin is the major enamel matrix component in developing teeth. In eutherian mammals, amelogenin is expressed from the X chromosome only, or from both the X and Y chromosomes. Two classes of porcine amelogenin cDNA clones have been characterized, but the chromosomal localization of the gene(s) encoding them is unknown. To determine if there are sex-based differences in the expression of porcine amelogenin, we paired PCR primers for exons 1a, 1b, 7a, and 7b, and amplified enamel organ-derived cDNA separately from porcine males and females. The results show that exons 1a/2a and 7a are always together and can be amplified from both males (XY) and females (XX). Exons 1b/2b and 7b are also always paired, but can be amplified only from females. We conclude that porcine amelogenin is expressed from separate genes on the X and Y chromosomes, and not, as previously proposed, from a single gene with two promoters.

Relative levels of mRNA encoding enamel proteins in enamel organ epithelia and odontoblasts.

Amelogenin, enamelin, sheathlin (ameloblastin/ amelin), enamelysin (MMP-20), and KLK4 (EMSP-1) are the major structural proteins and proteinases in developing tooth enamel. Recently, odontoblasts were reported to express amelogenin, the most abundant enamel protein. In this study, we hypothesized that odontoblasts express all enamel proteins and proteases, and we measured their relative mRNA levels in enamel organ epithelia and odontoblasts associated with porcine secretory- and maturation-stage enamel by RT-PCR, using a LightCycler instrument. The results showed that amelogenin mRNA in secretory-stage EOE is 320-fold higher than in odontoblasts beneath secretory-stage enamel, and over 20,000-fold higher than in odontoblasts under maturation-stage enamel. Similar results were obtained for enamelin and sheathlin. Enamelysin mRNA levels were equivalent in these two tissues, while KLK4 mRNA was higher in odontoblasts than in secretory-stage EOE. These results support the conclusion that odontoblasts are involved in the formation of the enamel layer adjacent to enamel-dentin junction.

Odontoblasts enhance the maturation of enamel crystals by secreting EMSP1 at the enamel-dentin junction.

The temporal expression patterns and activity distributions of enamelysin and EMSP1, which are the major proteinases in immature enamel, were characterized. Extracellular matrix fractions from developing porcine incisors, individually comprised of predentin, dentin, and four secretory-stage enamel samples, including the highly mineralized enamel (HME) at the enamel-dentin junction (EDJ), were isolated, and their resident proteinases were identified by zymography. Soft-tissue fractions, which included cells from the extension site of enamel formation (ESEF), secretory- and maturation-stage ameloblasts, and odontoblasts, were characterized histologically and by RT-PCR for their expression of enamelysin and EMSP1. A significant finding was that EMSP1, expressed by odontoblasts, concentrates in the HME, but is not detected in predentin or dentin. We conclude that odontoblasts deposit EMSP1 via their cell processes into the deepest enamel layer, which facilitates the hardening of this layer and contributes significantly to the functional properties of the EDJ.

Noggin blocks osteoinductive activity of porcine enamel extracts.

Enamel extracts induce biomineralization such as osteogenesis and cementogenesis, but the molecular component responsible for this activity remains uncertain. We fractionated enamel extracts from developing pig teeth and isolated the osteoinductive fraction. Proteins from pig enamel scrapings were extracted under alkaline conditions (pH 10.8) and fractionated with the use of a Sephadex G-100 (size exclusion) column. The ability of each fraction to enhance alkaline phosphatase (ALP) activity was assayed in ST2 cells, a mouse bone marrow stromal cell line. The osteoinductive fraction of enamel extracts (OFE) was found in fractions 44 and 45, which induced ST2 cells to express the phenotype of bone-forming osteoblasts, and to form mineralized nodules. Furthermore, the ALP activity of ST2 cells exposed to OFE was reduced by noggin, an antagonist of BMPs, and OFE reacted with BMP-2/4 antibody in dot-blot analysis. These results indicate that OFE contains BMPs that contribute to the induction of biomineralization.

Amelogenin gene expression in porcine odontoblasts.

Amelogenin is the major organic component in the enamel matrix of developing teeth and plays an important role in enamel biomineralization. Amelogenin has been reported to be a specific secretory product of ameloblasts. In this study, we examined amelogenin gene expression in various cell layers prepared from a porcine permanent tooth germ using reverse transcription-polymerase chain-reaction (RT-PCR). Amelogenin amplification products were detected only in the secretory ameloblast layer after 20 cycles of PCR. After 30 cycles of PCR, amelogenin amplification products were detected in secretory and maturation-stage ameloblasts and in odontoblasts. The relative levels of amelogenin gene expression in secretory and maturation-stage ameloblasts and odontoblasts were determined. Secretory ameloblasts expressed over 1000 times the level of amelogenin mRNA found in odontoblasts. Amelogenin gene expression in odontoblasts was confirmed in an erupted porcine permanent first molar, which has no ameloblasts. Amelogenin PCR amplification products were identified from 4 different alternatively spliced transcripts in the ameloblast samples, and the same spliced forms were detected in the odontoblast samples.

Calcium binding of enamel proteins and their derivatives with emphasis on the calcium-binding domain of porcine sheathlin.

Dental enamel is believed to form by the transfer of ions from solution, primarily calcium, phosphate, hydroxyl and carbonate, to the surface of solid-state mineral. Such precipitation phenomena can be controlled by regulating the degree of saturation of the solution with respect to the potential solid phases that can form. The concentration of free calcium is the factor that most affects the degree of saturation for calcium hydroxyapatite, and its buffering by calcium-binding proteins has been proposed as the mechanism that determines the enamel mineral structure. In this study, Stains-all staining was used to identify and isolate calcium-binding proteins from the enamel matrix, and determine their structures and association constants for calcium. Proteolytic cleavage fragments derived from the C-terminus of sheathlin, having apparent molecular weights of 13, 15, 27 and 29 kDa, were characterized by amino-terminal protein sequencing, amino acid analysis, and sugar, phosphate and sulphate determinations. Sheathlin C-terminal cleavage products were shown to have no N-linked glycosylations or phosphorylated amino acids, but Pro(350) was hydroxylated, and there was one sulphated O-linked glycosylation at Thr(386), containing galactose and N-acetylgalactosamine. The calcium-binding association constants for enamel proteins ranged from a high of 1.2 x 10(4) M(-1) to a low of 4.4x10(1) M(-1). The relative strengths of binding in order of decreasing affinity were: 13 and 15 kDa calcium-binding domain of sheathlin >27 and 29 kDa calcium-binding proteins >32 kDa enamelin >89 kDa enamelin >6.5 kDa, 25 kDa, 23 kDa, 20 kDa, 13 kDa, 5.3 kDa amelogenins. It is concluded that if enamel proteins have similar calcium-binding properties in vivo as have been measured in vitro, they would tend to buffer the free calcium ion concentration in enamel fluid.

Immunoblot detection and expression of enamel proteins at the apical portion of the forming root in porcine permanent incisor tooth germs.

There have been many immunohistochemical studies of enamel proteins during root formation. In the present article, the detection and expression of enamel proteins in tissue samples prepared from the apical portion of the forming root (APFR) in porcine permanent incisor tooth germs were studied. Amelogenin, enamelin, and sheathlin were detected by immunoblot analysis, but only in small amounts. The detection of their derivatives indicated their degradation. It is, at present, unclear as to which proteinases are involved in these degradations, because activity of enamel matrix serine proteinase 1 and enamelysin was not detected on gelatin and casein zymograms. The expression of enamel proteins was also proved in the APFR sample by the detection of polymerase chain reaction products of their cDNAs, and this may be related to cells of fragmentized Hertwig's epithelial root sheath. Amelogenin expression was not greater than that of enamelin and sheathlin. It was different from the expression pattern of secretory ameloblasts involved in enamel matrix formation. These results suggest that the amelogenins found in the APFR do not form a three-dimensional structure of amelogenin micelles, which has been proposed for the secretory enamel matrix structure. In this case, the enamel proteins could spread out easily following degradation into the matrix of future cementum. Some of their derivatives may play a role in the formation of the cementum.

Maturation ameloblasts of the porcine tooth germ do not express amelogenin.

Amelogenins are the most abundant constituent in the enamel matrix of developing teeth. Recent investigations of rodent incisors and molar tooth germs revealed that amelogenins are expressed not only in secretory ameloblasts but also in maturation ameloblasts, although in relatively low levels. In this study, we investigated expression of amelogenin in the maturation stage of porcine tooth germs by in situ hybridization and immunocytochemistry. Amelogenin mRNA was intensely expressed in ameloblasts from the differentiation to the transition stages, but was not detected in maturation stage ameloblasts. C-terminal specific anti-amelogenin antiserum, which only reacts with nascent amelogenin molecules, stained ameloblasts from the differentiation to the transition stages. This antiserum also stained the surface layer of immature enamel at the same stages. At the maturation stage, no immunoreactivity was found within the ameloblasts or the immature enamel. These results indicate that, in porcine tooth germs, maturation ameloblasts do not express amelogenins, suggesting that newly secreted enamel matrix proteins from the maturation ameloblast are not essential to enamel maturation occurring at the maturation stage.

Enamelysin (matrix metalloproteinase-20): localization in the developing tooth and effects of pH and calcium on amelogenin hydrolysis.

The formation of dental enamel is a precisely regulated and dynamic developmental process. The forming enamel starts as a soft, protein-rich tissue and ends as a hard tissue that is over 95% mineral by weight. Intact amelogenin and its proteolytic cleavage products are the most abundant proteins present within the developing enamel. Proteinases are also present within the enamel matrix and are thought to help regulate enamel development and to expedite the removal of proteins prior to enamel maturation. Recently, a novel matrix metalloproteinase named enamelysin was cloned from the porcine enamel organ. Enamelysin transcripts have previously been observed in the enamel organ and dental papillae of the developing tooth. Here, we show that the sources of the enamelysin transcripts are the ameloblasts of the enamel organ and the odontoblasts of the dental papilla. Furthermore, we show that enamelysin is present within the forming enamel and that it is transported in secretory vesicles prior to its secretion from the ameloblasts. We also characterize the ability of recombinant enamelysin (rMMP-20) to degrade amelogenin under conditions of various pHs and calcium ion concentrations. Enamelysin displayed the greatest activity at neutral pH (7.2) and high calcium ion concentration (10 mM). During the initial stages of enamel formation, the enamel matrix maintains a neutral pH of between 7.0 and 7.4. Thus, enamelysin may play a role in enamel and dentin formation by cleaving proteins that are also present during these initial developmental stages.

Immunocytochemical and immunochemical study of enamelins, using antibodies against porcine 89-kDa enamelin and its N-terminal synthetic peptide, in porcine tooth germs.

Enamelins comprise an important family of the enamel matrix proteins. Porcine tooth germs were investigated immunochemically and immunocytochemically using two antibodies: a polyclonal antibody raised against the porcine 89-kDa enamelin (89 E) and an affinity purified anti-peptide antibody against the porcine enamelin amino-terminus (EN). Immunochemical analysis of layers of immature enamel from the matrix formation stage detected immunopositive protein bands ranging from 10 kDa to 155 kDa in the outer layer enamel sample irrespective of the antibodies used. In contrast, the middle and inner enamel layer mainly contained lower molecular weight enamelins. In immunocytochemical analyses of the differentiation stage, 89 E stained enamel matrix islands around mineralized collagen fibrils of dentin, while EN stained both enamel matrix islands and stippled material. At the matrix formation stage, both antibodies intensely stained enamel prisms located in the outer layer. In the inner layer, 89 E moderately stained enamel matrix homogeneously, while EN primarily stained the prism sheath. The intense immunoreaction over the surface layer of enamel matrix at the matrix formation stage, following staining with 89 E and EN, disappeared by the end of the transition stage and the early maturation stage, respectively. The Golgi apparatus and secretory granules in the ameloblasts from the late differentiation stage to the transition stage were immunostained by both antibodies. These results suggest that expression of enamelin continues from late differentiation to the transition stage and the cleavage of N-terminal region of enamelin occurs soon after secretion. Some enamelin degradation products, which apparently have no affinity for hydroxyapatite crystals, concentrate in the prism sheaths during enamel maturation.