Perlecan,a Basement Membrane-type Heparan Sulfate Proteoglycan,in the Enamel Organ: Its Intraepithelial Localization in the Stellate Reticulum

Abstract

The localization and biosynthesis of perlecan, a basement membrane-type heparan sulfate proteoglycan, were studied in developing tooth germs by using murine molars in neonatal and postnatal stages and primary cultured cells of the enamel organ and dental papilla to demonstrate the role of perlecan in normal odontogenesis. Perlecan was immunolocalized mainly in the intercellular spaces of the enamel organ as well as in the dental papilla/pulp or in the dental follicle. By in situ hybridization, mRNA signals for perlecan core protein were intensely demonstrated in the cytoplasm of stellate reticulum cells and in dental papilla/pulp cells, including odontoblasts and fibroblastic cells in the dental follicle. Furthermore, the in vitro biosyntheses of perlecan core protein by the enamel organ and dental papilla/pulp cells were confirmed by immunofluorescence, immunoprecipitation, and reverse transcriptase-polymerase chain reaction. The results indicate that perlecan is synthesized by the dental epithelial cells and is accumulated in their intercellular spaces to form the characteristic stellate reticulum, whose function is still unknown.

Keywords

tooth germ heparan sulfate proteoglycan intraepithelial stroma stellate reticulum enamel organ perlecan

Perlecan, a heparan sulfate proteoglycan (HSPG), was originally known as one of the constituent molecules of the basement membranes. However, it has been demonstrated in recent years that perlecan is distributed not only in the basement membranes but also in the stromal space in various pathophysiological conditions (Costell et al. 1999). The stromal space in which perlecan is accumulated has been histologically characterized by myxoid appearances, which are seen in the stromata of pleomorphic adenomas (Saku et al. 1990), adenoid cystic carcinomas (ACCs), including the characteristic pseudocystic space (Cheng et al. 1992), adenomatoid odontogenic tumors (Murata et al. 2000), odontogenic hamartomatous tissues (Yonemochi et al. 1998) or gastrointestinal neoplasms (Sabit et al. 2001), and in the myxedematous stroma of granulation tissues (Murata et al. 1997; Yamazaki et al. 2004). More recently, widening of intercellular space due to perlecan deposition has been demonstrated in squamous epithelial dysplasia of the oral mucosa (Ikarashi et al. 2004) and in ameloblastoma foci (Ida-Yonemochi et al. 2002). Because the term ameloblastoma is derived from the resemblance of its tumor cell nest to the stellate reticulum of tooth germs, the biosynthesis of perlecan by ameloblastoma cells and the deposition of perlecan in the stellate reticulum-like tumor cell nests have indicated the possibility of perlecan production/deposition within the tooth stellate reticulum. These studies have also disclosed enhanced perlecan mRNA signals in the invading front of tumor cells (Ida-Yonemochi et al. 2002) or in the germinal focus with the epithelial layer (Ikarashi et al. 2004). Hence, it has been suggested that perlecan functions in an accelerated proliferation of tumor cells. In a series of experiments using ACC2/ACC3 cell systems derived from human ACCs, the biosynthesis of perlecan by ACC cells has been related to their proliferation (Kimura et al. 1999). In the course of these investigations, we have come to pay much more attention to the myxoid appearance or the dilatation of the intercellular space due to perlecan deposition and the function of perlecan in these particular tissue architectures, which are inevitably poor in vascularity.

The enamel organ of tooth germs is histologically characterized by the stellate reticulum, which is composed of multiple layers of uniquely differentiated epithelial cells with pronounced intercellular spaces between them. Only a small number of investigators have given attention to this peculiar structure of the stellate reticulum. But the existence of extracellular matrix (ECM) molecules, such as syndecan 1 (Thesleff et al. 1988), fibronectin (Thesleff et al. 1981), and laminin 5 (Yoshiba et al. 1998), as well as functional molecules, such as transforming growth factor-β1 (Cam et al. 1990), bone morphogenetic protein (BMP)-2, BMP-4 (Vainio et al. 1993), and fibroblast growth factor (FGF)-2 (Cam et al. 1992) has been previously demonstrated in the stellate reticulum. However, these investigations had no specific point of view in terms of how ECM molecules function in this particular milieu of extended intercellular space, and any further functions of the enamel organ itself remain unknown. Many other ECM molecules have been shown to play important roles in normal (Lesot et al. 2002) and abnormal (Yonemochi et al. 1998,1999) tooth morphogenesis, although little is known about perlecan in terms of its function or even its distribution in normal tooth germ development. Only cation histochemistry (Goldberg and Septier 1987) and heparan sulfate lyase histochemistry (Kogaya et al. 1990) have suggested the presence of HSPG in the tooth germ basement membranes and in the stellate reticulum. Furthermore, little attention has been given to ECM function as a carrier for nutrient transport by diffusion, which is thought to be indispensable in tissues with poor vascularity, such the enamel organ.

The purpose of this study is to determine the mode of perlecan expression during murine molar development, paying special attention to the enamel organ, and to demonstrate the biosynthetic activity of perlecan in tooth germ cells in culture. On the basis of this primary data regarding perlecan localization in the tooth germ, we discuss newly suggested properties of perlecan in odontogenesis.

Materials and Methods

Animals

ICR mice were obtained from Nihon Doubutsu (Osaka, Japan) and maintained under standard conditions at the Laboratory Animal Center of Niigata University Graduate School of Medical and Dental Sciences. The day when a vaginal plug appeared was designated as E0 (embryonic day 0), and the day of birth was designated as P0 (postnatal day 0). Two mice each from E11.5, E13.5, E15.5, E17.5, P1, P6, P10, P12, and P28 were used in this study.

Tissue Preparation

Heads of mice were removed under ether anesthesia and fixed with 100% methanol or 4% paraformaldehyde in 0.1 M phosphate-buffered saline (PBS) (pH 7.4) overnight at 4C and processed for embedding in paraffin. Serial sections (5-μm) cut in the sagittal or frontal planes were stained with hematoxylin-eosin and immunohistochemically with the antibody described below. Another set of sections was used for in situ hybridization.

Antibodies

Mouse perlecan core protein was prepared from the murine Engelbreth-Holm-Swarm tumor, and the antibodies against mouse perlecan were raised in rabbits as described elsewhere (Saku and Furthmayr 1989).

Immunohistochemistry

The Envision+/HRP system (DAKO Japan; Kyoto, Japan) was used for immunohistochemical staining. Before incubation with the primary antibodies, sections were treated with 3 mg/ml bovine testicular hyaluronidase (type I-S, 440 U/mg; Sigma Chemical Co., St Louis, MO) in PBS for 30 min at 37C or with 0.4% pepsin (Sigma) in 0.01 N HCl for 30 min at 37C. The primary antibody was diluted to concentrations of 50 μg/ml. For visualization of reaction products, sections were treated with 3,3′-diaminobenzidine (Dohjin Laboratories; Kumamoto, Japan) in the presence of 0.005% hydrogen peroxide, and the sections were counterstained with hematoxylin. For control experiments, the primary antibodies were replaced with preimmune rabbit IgGs.

Preparation of RNA Probes

RNA probes for the mouse perlecan were prepared by using digoxigenin (DIG) RNA labeling kits (F. Hoffmann-La Roche; Basel, Switzerland) and T7/T3 RNA polymerases (Promega; Madison, WI). Template cDNA (516 bp, corresponding to domain I of mouse perlecan) was subcloned with RT-PCR, and the resultant 516-bp fragment was ligated into the plasmid vector (pBluescript II, Promega). The vector plasmid was linearized with BamHI and HindIII, and the linearized plasmids were used as templates to synthesize DIG-labeled RNA antisense probes by T7 RNA polymerase and sense probes by T3 RNA polymerase (Ikarashi et al. 2004).

In Situ Hybridization

After deparaffinization, sections were washed in three changes of 2x standard saline citrate (SSC) and treated with 5 μg/ml of proteinase K (Sigma) for 20 min at 37C. They were then washed with 0.2% glycine in PBS, fixed with 4% paraformaldehyde in 0.1 M phosphate buffer (pH 7.5) for 5 min, dehydrated with a series of ethanol (70% to 100%), and air dried. Hybridization was performed for 15 hr at 48C in a moist chamber. The hybridization solution contained 10% dextran sulfate, 1x Denhardt's solution, 100 μg/ml of salmon sperm DNA, 125 μg/ml of yeast tRNA, 3x SSC, 50% formamide, and 500 ng/ml of probes in 10 mM phosphate buffer (pH 7.4). After hybridization, the sections were rinsed in 2x SSC, and then the hybridized probes were detected with DIG detection kits (Roche). The sections were counterstained with methyl green (Ikarashi et al. 2004).

Primary Culture of Enamel Organ and Dental Papilla/Pulp Cells

First molar tooth germs of the mandible from day 1 postnatal mouse pups were dissected in Hanks solution under a stereomicroscope. They were incubated with 0.25% trypsin (Roche) in 1 mM ethylenediaminetetraacetic acid solution for 10 min at 37C. Then, the enamel organ and the dental papilla were separated manually under a stereomicroscope. For immunofluorescence and immunoprecipitation experiments, these isolated tissues were placed in 0.1% collagenase/DMEM solution for 4 hr at 37C to dissociate the enamel pulp cells and the dental papilla cells. The suspended enamel organ cells and the dental papilla/pulp cells were plated into 35-mm plastic dishes in 2 ml of DMEM containing 10% fetal calf serum (FCS; ICN Pharmaceuticals, Costa Mesa, CA), 1% glutamine, 50 μg/ml streptomycin, and 50 IU/ml penicillin. They were incubated at 37C under a humidified 5% CO₂/95% air atmosphere. When the cells became confluent, they were split again and used for RT-PCR and immunoprecipitation experiments.

Reverse Transcriptase-Polymerase Chain Reaction (RT-PCR)

Total RNA was isolated from enamel organ and dental papilla tissues using the ISOGEN system (Nippon Gene Co.; Tokyo, Japan). cDNA was synthesized from the RNA with the SuperScript First-Strand Synthesis System (Invitrogen; Carlsbad, CA). Following the RT, PCR was carried out in an Astec thermal cycler PC-800 (Astec; Fukuoka, Japan). Oligonucleotide primers flanking the exon of domain I of perlecan core protein (nucleotide number 183–550, No. M85289, GeneBank) were synthesized as follows: 5′-CCTGA GGACA TAGAG AC-3’ forward and 5′- TCGGA AGGGA ATGCG GA-3’ reverse, to generate a 503-bp product. For the competitive PCR, we also synthesized oligonucleotide primers of mouse β-actin as follows: 5′-TGGAA TCCTG TGGCA TCCAT GAAAC-3’ forward and 5′-TAAAA CGCAG CTCAG TAACA GTCCG-3’ reverse, to generate a 348-bp product. The thermocycling protocol during 30 amplification cycles was as follows: denaturation at 94C for 1 min, annealing at 55C for 1 min, extension at 72C for 1 min, and termination with a final cycle: annealing at 55C for 1 min and extension at 72C for 7 min. The amplified DNA fragments were analyzed by electrophoresis on 3% agarose gels. The electrophoresis image was analyzed quantitatively: amounts of perlecan relative to β-actin were determined with NIH Image, an image processing and analyzing program.

Immunoprecipitation

Cell labeling and immunoprecipitation experiments were performed as described elsewhere (Kimura et al. 1999). Briefly, cells were preincubated with methionine-free minimum essential medium (MEM) for 1 hr and then incubated in fresh MEM containing 50 μCi of [³⁵S]-methionine for 3 hr. After removal of the medium, the cell layer was lysed and both the cell lysate and medium were centrifuged at 15,000 × g for 10 min. The resultant supernatants were subjected to immunoprecipitation. The precleared lysates and media were incubated with the antibodies to the perlecan core protein overnight, and immune complexes were isolated with protein A-Sepharose (Amersham-Pharmacia Biotech; Uppsala, Sweden). For antibody control experiments, the antibodies to the perlecan core protein were replaced with preimmune rabbit IgGs. Immunoisolated materials were dissolved in Laemmli's sample buffer, boiled for 5 min, and centrifuged at 10,000 × g for 5 min to remove beads. The supernatants were analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). SDS-PAGE was done on a 5% polyacrylamide slab gel with 2.5% 2-β-mercaptoethanol, according to Laemmli (1970). Gels were stained with Coomassie Brilliant Blue and then air dried. The dried gels were fluorographed on X-ray film (Hyperfilm-MPTM; Amersham, Buckinghamshire, UK). Apparent molecular mass was determined by co-electrophoresis of marker proteins.

Results

Perlecan expression in murine molar tooth germs and teeth was stage specific. It was not limited to the basement membranes only; it was demonstrated in the enamel organ as well as in the dental mesenchymal tissues. In the following sections, the modes of perlecan expression in the developing tooth germs are described in time order.

Dental Lamina Stage (E11.5)

The initial morphological sign of tooth organogenesis was a site-specific thickening of the oral epithelium, which was presumed to be dental epithelial cells (Figure 1A). Perlecan was definitely immunolocalized along the basement membrane of the oral epithelium and faintly and diffusely in the underlying mesenchyme (Figure 1B). mRNA signals for perlecan core protein were detected correspondingly to its immunolocalization in both the oral epithelial cells and the underlining mesenchymal cells (Figure 1C).

Bud Stage (E13.5)

At E13.5, the dental epithelium was further invaginated to form epithelial buds in which high epithelial cells were aligned on the basement membrane (Figure 1D). Perlecan was strongly immunolocalized in the basement membrane as well as in the central area of the buds (Figure 1E). Within the bud, it was demonstrated in the intercellular space of dental epithelial cells as well as in their cytoplasm. Condensed dental mesenchymal tissues around the bud showed faint and diffuse immunopositivities for perlecan. In situ hybridization signals for perlecan mRNA were demonstrated in dental epithelial cells within the bud, particularly in cells in the suprabasal layer at the primary enamel knots. Most of the mesenchymal cells around the bud also showed mildly positive signals, with scattered cells showing strong signals (Figure 1F).

Figure 1

Embryonal tooth germ development. (A-C) E11.5, (D-F) E13.5, (G-I) E15.5, and (J-L) E17.5. (A,D,G,J) Hematoxylin-eosin (HE) stain, (B,E,H,K) immunoperoxidase stain for perlecan, counterstained with hematoxylin, and (C,F,I,L) in situ hybridization for perlecan core protein mRNA, digoxigenin (DIG)-immune alkaline phosphatase, methyl green counterstain. (A-C) x960; (D-L) x400. Perlecan started to be immunolocalized definitely in the basement membrane and slightly and diffusely in the underlying mesenchyme at E11.5 (B). With the morphological development toward the enamel organ, perlecan came to be strongly positive in the central area of the stellate reticulum as well as in the dental papilla and dental sac (E,H,K). mRNA signals for perlecan core protein were demonstrated in both enamel organ epithelial cells, particularly in the suprabasal layer at the enamel knots (arrow), and in dental papilla or surrounding mesenchymal cells (C,F,I,L).

Cap Stage (E15.5)

At the cap stage, the dental epithelial cells became differentiated into the enamel organ, in which the inner and outer enamel epithelia, stratum intermedium, and stellate reticulum were distinguished (Figure 1G). The immunolocalization mode for perlecan in the stellate reticulum was almost the same as that in the bud stage, although the positive area was enlarged more with its growth. On the other hand, the positive staining in mesenchymal cells condensed around the enamel organ was more enhanced (Figure 1H). Although perlecan mRNA signals were found in most of the epithelial cells of the enamel organ, they were more strongly positive in those of the stellate reticulum areas, especially in the secondary enamel knots. In addition, the surrounding mesenchymal cells, including those in the dental papilla area enclosed by the invaginated enamel organ, had definite hybridization signals (Figure 1I).

Bell Stage (E17.5)

During the early bell stage, the characteristic cusp pattern of the enamel organ started to be formed, and mesenchymal areas were condensed and enclosed by the enamel organ as definite forms of the dental papilla (Figure 1J). The intercellular immunolocalization for perlecan within the stellate reticulum was much more enhanced with the development of the enamel organ. At the same time, strongly positive reactions were obtained within the dental papilla and in the surrounding interstitium (Figure 1K). Perlecan mRNA signals were most intensely localized at inner ameloblasts as well as at dental papilla cells facing tall ameloblasts. In addition, signals were also seen in the stellate reticulum and outer ameloblasts as well as in mesenchymal cells in the dental papilla and its surrounding areas (Figure 1L).

Postnatal Differentiation Stage (P1)

The tooth germ began to assume the crown shape of molar teeth. Cusps were clearly recognized, with largely regressed enamel organ spaces where blood vessels entered with extravasation of erythrocytes, and the dental papilla space was instead proportionally enlarged. Alignments of tall columnar ameloblasts and odontoblasts were obvious (Figure 2A). Immunohistochemically, perlecan was localized diffusely in the narrowed stellate reticulum and in the immature dental pulp (Figure 2B). Its mRNA signals were detected in the cells with perlecan immunopositivities in both stellate reticulum and dental pulp (Figure 2C).

Postnatal Root Developmental Stage (P6)

By day 6 after birth, tooth crowns were mostly formed and tooth roots started to be elongated in the molar tooth germ. With the apical elongation of the cervical loop of the enamel organ, the stellate reticulum started to be reduced in volume, and both enamel and dentin matrices started to be calcified cuspids emerging (Figure 2D). Immunohistochemically, the staining intensity for perlecan became stronger in the atrophic stellate reticulum. In the dental pulp, the immunolocalization for perlecan was heterogeneous, with a stronger staining intensity on the crown side and a weaker intensity toward the root side (Figure 2E). Perlecan mRNAs were detected strongly in aligned odontoblasts as well as diffusely and weakly in other dental pulp cells (Figure 2F).

At 4 weeks of age, molar roots were completely formed, with the periodontal ligament in its functional state (Figure 2G). In the dental pulp, the immunolocalization of perlecan was scarcely found in the crown part but focally preserved in the root part. Within the pulp, it was strongly immunolocalized in the predentin. The periodontal ligament, especially at the cervical area, was diffusely immunopositive for perlecan (Figure 2H). Perlecan mRNA signals were demonstrated in odontoblasts, especially those in the root area, and in cementoblastic, fibroblastic, or osteoblastic cells in the periodontal ligament, which was not always synchronous to the immunolocalization (Figure 2I). In the histological experiments described above, neither immunopositivities nor mRNA signals for perlecan core protein were obtained when the antibodies were replaced with preimmune rabbit IgGs or when the antisense probes were replaced with sense probes (not shown).

PCR Confirmation of Perlecan Expression in Tooth Germ Tissues

To analyze the perlecan mRNA expression level in each dental component during the tooth development, enamel organ and dental papilla/pulp tissues at postnatal days 1, 6, 10, and 12 were separated, and total RNA was extracted from each fraction. Total RNA samples, 3 μg each, were reverse-transcribed with oligo-dT primers, and the resultant cDNAs were further amplified for perlecan core protein domain I with β-actin oligonucleotide primer pairs as an internal control. A 503-bp PCR product for perlecan was obtained from all samples with different band densities, whereas a 348-bp PCR product for β-actin was almost stable among the samples (Figure 3A). The PCR products for perlecan and β-actin were determined by densitometry, and the relative values for perlecan expression levels against those of β-actin in the time course after birth were plotted and are shown in Figure 3B. Perlecan mRNA levels in the enamel organ decreased gradually after birth, which was in accordance with regression of the enamel organ space and the immuno-histochemical results (Figure 2E). In contrast, perlecan mRNA expression levels in the dental papilla/pulp increased with aging, which was also similar to histological and immunohistochemical observations (Figure 2I).

Figure 2

Postnatal tooth development. (A-C) P1, (D-F) P6, (G-L) P28. (A,D,G) HE stain, (B,E,H) immunoperoxidase stain for perlecan, counterstained with hematoxylin, and (C,F,I) in situ hybridization for perlecan core protein mRNA, DIG-immune alkaline phosphatase, methyl green counterstain. (A-F) x200; (G-I) x160. At day 1 after birth, the staining for perlecan became weaker with the narrowing of the enamel organ space with vascular entry (B). At day 6, the perlecan staining was condensed to be a trace-like space of the stellate reticulum, and the staining became heterogeneous in the dental pulp (E). At day 28, when tooth roots were fully developed, perlecan was immunolocalized in immature dental pulp tissue in the root part and the periodontal ligament (H). mRNA signals for perlecan were demonstrated in cells in the perlecan-immunopositive areas but were most enhanced in odontoblasts in the root apex and in spindle cells along the surface of alveolar bone or cementum (I).

Tooth Germ Cells in Primary Culture

Enamel organ-derived cells in the primary culture showed large stellate shapes with many cytoprocesses, whereas dental papilla/pulp-derived cells were rather small and polygonal in shape. Immunofluorescence studies showed that both cells were immunopositive for perlecan. However, their staining patterns were completely different. In enamel organ cells at the initial stage of culture, thread-like immunofluorescence signals for perlecan were recognized over the cytoplasm or along the cytoprocesses, suggesting that the deposition of perlecan was along focal contacts (Figure 4A). Such an immunolocalization pattern became enhanced and thick at the subconfluent stage of culture (Figure 4B). In addition to this thread-like immunofluorescence, the cells had intracellular signals diffusely around their nuclei, suggesting perlecan localization in the Golgi area. On the other hand, in the dental papilla/pulp cells, immunofluorescence signals for perlecan were observed diffusely in the cytoplasm in the initial stage (Figure 4C). In the later stage, perlecan signals became deposited mainly in the intercellular space to form mesh-like structures over the cell layers (Figure 4D). These contrasting modes of deposition in the two cell types suggested that perlecan produced by enamel organ cells tended to be released more into the culture medium, but that dental papilla/pulp cells tended to trap more perlecan molecules around the cell surface. No immunofluorescence was observed when the anti-perlecan antibodies were replaced with preimmune rabbit IgGs (not shown).

Figure 3

Gene expression of perlecan core protein in enamel organ and dental papilla tissues. (A) RT-PCR for cells from enamel organ (Lanes 1–4) and dental papilla (Lanes 5–8). Postnatal day 1 (Lanes 1 and 5), day 6 (Lanes 2 and 6), day 10 (Lanes 3 and 7), and day 12 (Lanes 4 and 8). Competitive PCR products for perlecan (503 bp) and β-actin (348 bp), 3% agarose gel electrophoresis. (B) Relative ratios of competitive PCR products for perlecan against β-actin plotted from the data in (A). Open box, enamel organ; closed box, dental papilla. Perlecan mRNA levels in the enamel organ decreased gradually with time after birth in accordance with the regression of the space of the stellate reticulum. In contrast, perlecan gene expression levels increased with time in the dental papilla tissues that developed into the dental pulp.

To confirm the biosynthesis of perlecan by enamel organ and dental papilla cells biochemically, cells in the primary culture were labeled with [³⁵S]-methionine for the core protein, and the cell lysates and the culture media were separately immunoprecipitated with anti-perlecan core protein. In the cell layer of both cell types, precipitants were seen as rather broad bands, with molecular masses of ∼400 kDa, a band at the interface between the stacking gel and the resolving gel, and a band at the bottom of the well on SDS-PAGE under reducing conditions (Figure 4E, Lanes 1 and 3). The 400-kDa bands were regarded as perlecan core proteins ready for glycosylation. Those stacked in the well bottom and the gel interface were considered as intact forms of perlecan because heparan sulfate (HS) chains interfered with the entry of the molecules into the gel. From the media, only faint bands for intact forms were seen at the gel interface, but there were no 400-kDa bands, suggesting no presence of core protein or stacks in the well bottom (Figure 4E, Lanes 2 and 4). Thus, perlecan molecules were shown to be produced by both cell types, and the findings were consistent with the morphologic evidence by immunofluorescence (Figures 4A-4D). Most of the synthesized molecules were still within the cytoplasm or deposited in the cell layer, whereas only a small number were secreted out into the media in this short labeling period. No precipitates were obtained when the anti-perlecan antibodies were replaced with preimmune rabbit IgGs (not shown).

Discussion

In the present study, we have demonstrated the expression of perlecan in murine developing tooth germs at both protein and mRNA levels by using molar samples in their neonatal and postnatal stages as well as cells from the enamel organ and the dental papilla/ pulp biochemically. Our hypothesis that the stellate reticulum structure resulted from an intercellular accumulation of proteoglycan-rich extracellular matrices including perlecan has been confirmed in the present investigation.

Perlecan, a basement membrane-type HSPG, bears three major HS side chains and 12 probable N-linked carbohydrates on its large core protein of ∼400 kDa molecular mass (Noonan et al. 1991). This core protein consists of five distinct domains, and this multidomain structure suggests that it possesses multifunctional properties of the molecule, although its actual functions are still poorly understood. On the other hand, the role of HS chains has been studied extensively, especially in binding with various growth factors by their negative charges (Jiang and Couchman 2003). Their primary function as a member of glycosaminoglycan chains is thought to be retention of water molecules around the straight chains of disaccharide units (Jackson et al. 1991). In previous studies, we showed that perlecan was most enhanced in the early phase of granulation tissue formation processes, which showed myxoid appearance, in the oral cavity (Murata et al. 1997; Yamazaki et al. 2004) and gastrointestinal tracts (Ohtani et al. 1993). Similar results were obtained in the myxoid stroma of adenoid cystic carcinoma (Cheng et al. 1992), pleomorphic adenoma (Saku et al. 1990), and adenomatoid odontogenic tumor (Murata et al. 2000). Thus, the myxoid histology resulting from the retention of water via HS chains takes place irrespective of the tissue, epithelial or mesenchymal.

The histological phenotype of the stellate reticulum-like appearance is characteristic of ameloblastomas, whose term has been adopted due to the resemblance of their tumor cell nests to the enamel organ. We have already demonstrated that this characteristic stellate reticulum-like histology of ameloblastoma is caused by perlecan retention in the intercellular space of the tumor cells (Ida-Yonemochi et al. 2002). On the basis of the results, we have predicted that perlecan must be present within the authentic stellate reticulum of the enamel organ. However, there has been no direct evidence on the localization of any concrete proteoglycan species, including perlecan, in the enamel organ, although the presence of HSPG was suggested by cation histochemistry (Goldberg and Septier 1987) and HS lyase histochemistry (Kogaya et al. 1990). Thus, in the present study, the existence of perlecan has been confirmed for the first time in the intercellular spaces within the enamel organ, and the nomenclature of ameloblastoma has been confirmed functionally by this aspect of perlecan deposition in the intercellular space of epithelial cells.

Figure 4

Biosynthesis of perlecan core protein by enamel organ and dental papilla cells. Immunofluorescence for perlecan in cells in primary culture from the enamel organ (A,B) and dental papilla (C,D) in their subconfluent (day 3, A,C) and confluent (day 7, B,D) states. (A-D) x400. (E) Immunoprecipitation for perlecan in cells in primary culture from the enamel pulp and dental papilla. (Lanes 1 and 2) Enamel organ cells; (Lanes 3 and 4) dental papilla cells; (Lanes 1 and 3) cell layers; (Lanes 2 and 4) culture media. Fluorographies of SDS-PAGE, [³⁵S]-methionine labeling for 3 hr. Arrow, intact form of perlecan. Threadlike signals of immunofluorescence for perlecan were seen over the cytoplasm of enamel organ cells even at the initial stage (A), and they fused with each other with their growth (B). On the other hand, dental papilla cells contained only weak immunofluorescence for perlecan in the cytoplasm in the initial stage (C), but in the later stage, perlecan was deposited over the cell layer, showing mesh-like arrangements (D). In immunoprecipitation experiments, both perlecan core protein appearing as broad bands of ∼400 kDa and its intact form with heparan sulfate chains were obtained in enamel organ (Lane 1) and dental papilla (Lane 3) cells. A small number of the intact forms were also found in the medium (Lanes 2 and 4). Top bands in the well bottom were intact perlecan molecules that could not enter the gels.

The stellate reticulum has been believed to function as a spacer device for mechanical protection for the tooth crown formation as well as for nutritional recruitment from the outlying vascular circulation to the stellate cells (Kallenbach 1980). However, the real function of the enamel organ and the stellate reticulum has been mostly unexplored to date. The present evidence of perlecan deposition in the intercellular space of the stellate reticulum may indicate that it acts as a carrier for transport of nutrients to epithelial cells for enamel formation, because there is no entry of blood vessels into the enamel organ before birth, and because no nutrient supplies can be expected from the dental papilla side because of the presence of dental hard matrices. Because proteoglycans and/or some groups of polyanions have been functionally implicated in the transport and diffusion of calcium ions in the secretory ameloblast layer (Goldberg and Septier 1987), it is likely that the stellate reticulum also takes part in transport or condensation of calcium ions from the blood circulation to the ameloblastic layers through the plentiful HS of perlecan.

Furthermore, we have demonstrated the presence of HS chains in the cell membrane of the stratum intermedium cells and in the papillary layer of the tooth germs and have concluded that HS chains regulate the transport of minerals through their negatively charged cell membranes and play an important role in cell-cell interaction by preserving local growth factors in the tooth germ development (Nakamura et al. 1995). Basic fibroblast growth factor (bFGF), which is known to bind HS chains, has been reported to exist in the enamel organ as well as in the basement membranes (Cam et al. 1992), and the immunolocalization pattern of bFGF is almost the same as that of perlecan, as shown in the present study. On the other hand, other basement membrane molecules, such as laminins and collagen type IV, have been localized only within the basement membrane zones at the epithelial-mesenchymal interface (Thesleff et al. 1981). Thus, it is highly likely that perlecan participates not only in basement membrane assembly but also in the regulation of the proliferation and differentiation of epithelial cells during tooth morphogenesis.

In the present immunofluorescence study of cells in primary culture, it was obvious that perlecan was deposited on the cell surface, especially along plentiful and long cytoprocesses of stellate-shaped enamel organ cells. The perlecan deposition in thread-like and parallel fashions may represent a trace of focal contacts by which cells are stretched. This further suggests that cytoskeletal networks specific to such cell shapes are affected by perlecan, although there have been no reports on the relationship between perlecan and cytoskeletal fibers.

The mRNA expression levels for perlecan by RT-PCR as well as the protein level of perlecan by immunoprecipitation in enamel organ cells/tissues, supported the immunofluorescence data. Although it was difficult to evaluate quantitatively their modes of biosynthesis, a gradual decrease in perlecan mRNA expression levels by enamel organ cells in the 12-day period after birth was at least confirmed by the RT-PCR result, which should represent the regression of the enamel organ space. On the other hand, the modes of biosynthesis of perlecan in the dental papilla/pulp cells characterized by immunofluorescence, immunoprecipitation, and RT-PCR were distinct from those in enamel organ cells. Their production of perlecan was, rather, increased with time until day 12 after birth, which was a reverse tendency in comparison to that by enamel organ cells, and the mesh-like mode of perlecan deposition in the monolayer culture of dental papilla/pulp cells was not obtained in enamel organ cells. Thus, the present results indicate that perlecan functions differently between the enamel organ and the dental papilla/pulp.

In addition to the enamel organ, the immunolocalization of perlecan was also demonstrated to occur throughout the dental mesenchyme during murine odontogenesis. The results suggest that perlecan is involved in the formation of the dental papilla/pulp and the periodontal ligaments. The perlecan immunolocalization in the predentin and its transcripts in odontoblasts indicate its function in odontoblastic differentiation and dentin matrix maturation. By using immunoelectron microscopy for HSPG, laminin, type IV collagen, and fibronectin in the dental papilla of the Macaca fuscata monkey, Sawada and Inoue (1998) suggested the function of the basement membrane in supporting and positioning of odontoblasts toward their differentiation. Moreover, in the periodontal ligament, perlecan expression in osteoblasts and cementoblasts was also first confirmed in the present study. The modes of perlecan expression seemed to be dependent on the developing stage of the structure and cell types. A more extensive examination for the dental papilla/pulp and periodontal ligament is needed for a more accurate understanding of the whole function of perlecan in dentinogenesis and cementogenesis.

Footnotes

Acknowledgements

This work was supported in part by grants-in-aid for scientific research from the Japan Society for the Promotion of Science and from the Ministry of Education, Science, Sports and Culture, Japan.

References

Cam

Neumann

Oliver

Raulais

Janet

Ruch

(1992) Immunolocalization of acidic and basic fibroblast growth factors during mouse odontogenesis. Int J Dev Biol 36: 381–389

Cam

Neumann

Ruch

(1990) Immunolocalization of transforming growth factor β₁ and epidermal growth factor receptor epitopes in mouse incisors and molars with a demonstration of in vitro production of transforming activity. Arch Oral Biol 35: 813–822

Cheng

Saku

Okabe

Furthmayr

(1992) Basement membranes in adenoid cystic carcinoma. An immunohistochemical study. Cancer 69: 2631–2640

Costell

Gustafsson

Aszodi

Morgelin

Bloch

Hunziker

Addicks

(1999) Perlecan maintains the integrity of cartilage and some basement membranes. J Cell Biol 147: 1109–1122

Goldberg

Septier

(1987) Visualization of proteoglycans and membrane-associated components in rat incisor enamel organ using ruthenium hexamine trichloride. J Biol Buccale 15: 59–66

Ida-Yonemochi

Ikarashi

Nagata

Hoshina

Takagi

Saku

(2002) The basement membrane-type heparan sulfate proteoglycan (perlecan) in ameloblastomas: its intercellular localization in stellate reticulum-like foci and biosynthesis by tumor cells in culture. Virchows Arch 441: 165–173

Ikarashi

Ida-Yonemochi

Ohshiro

Cheng

Saku

(2004) Intraepithelial expression of perlecan, a basement membrane type heparan sulfate proteoglycan, reflects dysplastic changes of the oral mucosal epithelium. J Oral Pathol Med 33: 71–78

Jackson

Busch

Cardin

(1991) Glycosaminoglycans: molecular properties, protein interactions and role in physiological processes. Physiol Rev 71: 481–539

Jiang

Couchman

(2003) Perlecan and tumor angiogenesis. J Histochem Cytochem 51: 1393–1410

10.

Kallenbach

(1980) Access of horseradish peroxidase (HRP) to the extracellular spaces of the maturation zone of the rat incisor enamel organ. Tissue Cell 12: 165–174

11.

Kimura

Cheng

Toyoshima

Oda

Saku

(1999) Basement membrane heparan sulfate proteoglycan (Perlecan) synthesized by ACC3, adenoid cystic carcinoma cells of human salivary gland origin. J Biochem 25: 406–413

12.

Kogaya

Kim

Akisaka

(1990) Changes in ultrastructural distribution of dental basement membrane heparan sulfate during early mouse tooth development. J Biol Buccale 18: 109–115

13.

Laemmli

(1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227: 680–685

14.

Lesot

Combeau

Fausser

Meyer

Schmitt

Peterkova

Peterka

(2002) Cell-cell and cell-matrix interactions during initial enamel organ histomorphogenesis in the mouse. Connect Tissue Res 43: 191–200

15.

Murata

Cheng

Horino

Hara

Shimokawa

Saku

(2000) Enamel proteins and extracellular matrix molecules are co-localized in the pseudocystic stromal space of adenomatoid odontogenic tumor. J Oral Pathol Med 29: 483–490

16.

Murata

Hara

Saku

(1997) Dynamic distribution of basic fibroblast growth factor during epulis formation: an immunohistochemical study in an enhanced healing process of the gingiva. J Oral Pathol Med 26: 224–232

17.

Nakamura

Kimura

Kenmotsu

Sakai

Saku

Ozawa

(1995) Immunolocalization of CD44 and heparan sulfate chains on the stratum intermedium and papillary layer in the rat enamel organ. Arch Histol Cytol 58: 323–334

18.

Noonan

Fulle

Valente

Cai

Horigan

Sasaki

Yamada

(1991) The complete sequence of perlecan, a basement membrane heparan sulfate proteoglycan, reveals extensive similarity with laminin A chain, low density lipoprotein-receptor, and the neural cell adhesion molecules. J Biol Chem 266: 22939–22947

19.

Ohtani

Nakamura

Watanabe

Mizoi

Saku

Nagura

(1993) Immunocytochemical localization of basic fibroblast growth factor in carcinomas and inflammatory lesions of the human digestive tract. Lab Invest 68: 520–527

20.

Sabit

Tsuneyama

Shimonishi

Harada

Cheng

Ida

Saku

(2001) Enhanced expression of basement-membrane-type heparan sulfate proteoglycan in tumor fibro-myxoid stroma of intrahepatic cholangiocarcinoma. Pathol Int 51: 248–256

21.

Saku

Cheng

Okabe

Koyama

(1990) Immunolocalization of basement membrane molecules in the stroma of salivary gland pleomorphic adenoma. J Oral Pathol Med 19: 208–214

22.

Saku

Furthmayr

(1989) Characterization of the major heparan sulfate proteoglycan secreted by bovine aortic endothelial cells in culture: homology to the large molecular weight molecule of basement membranes. J Biol Chem 264: 3514–3523

23.

Sawada

Inoue

(1998) Basement membrane-like structures occurring on the surface of dental papilla mesenchymal cells during odontogenesis in the monkey Macaca fuscata. Eur J Oral Sci 106: 126–131

24.

Thesleff

Barrach

Foidart

Vaheri

Pratt

Martin

(1981) Changes in the distribution of type IV collagen, laminin, proteoglycan, and fibronectin during mouse tooth development. Dev Biol 81: 182–192

25.

Thesleff

Jalkanen

Vainio

Bernfield

(1988) Cell surface proteoglycan expression correlates with epithelial-mesenchymal interaction during tooth morphogenesis. Dev Biol 129: 565–572

26.

Vainio

Karavanova

Jowett

Thesleff

(1993) Identification of BMP-4 as a signal mediating secondary induction between epithelial and mesenchymal tissues during early tooth development. Cell 75: 45–58

27.

Yamazaki

Cheng

Hao

Takagi

Jimi

Itabe

Saku

(2004) Basement membrane-type heparan sulfate proteoglycan (perlecan) and low-density lipoprotein (LDL) are co-localized in granulation tissues: a possible pathogenesis of cholesterol granulomas in jaw cysts. J Oral Pathol Med 33: 177–184

28.

Yonemochi

Noda

Saku

(1998) Pericoronal hamartomatous lesions in the opercula of the teeth delayed in eruption: an immunohistochemical study for extracellular matrix. J Oral Pathol Med 27: 441–452

29.

Yonemochi

Noda

Shimokawa

Saku

(1999) Immunohistochemical localization of extracellular matrix molecules in complex odontoma. Dent Jpn 35: 13–19

30.

Yoshiba

Aberdam

Meneguzzi

Perrin-Schmitt

Stoetzel

Ruch

(1998) Expression and localization of laminin-5 subunits during mouse tooth development. Dev Dyn 211: 164–176