Thy-1-positive cells in the subodontoblastic layer possess high potential to differentiate into hard tissue-forming cells

The cells of the subodontoblastic cell-rich layer in dental pulp are speculated to contain odontoblast progenitor cells because of their positional relationship with odontoblasts as well as their high alkaline phosphatase (ALP) activity. However, it has yet to be determined whether these cells have the ability to differentiate into odontoblastic cells. In the present study, we firstly found that the majority of cells in the subodontoblastic layer expressed Thy-1, a cell-surface marker of stem and progenitor cells. Then, we evaluated the capacity of Thy-1 high- and low-expressing (Thy-1(high) and Thy-1(low)) cells separated from rat dental pulp cells by use of a fluorescence-activated cell sorter to differentiate into hard tissue-forming cells in vitro and in vivo. Following stimulation with bone morphogenetic protein-2, Thy-1(high) cells in vitro showed accelerated induction of ALP activity and formation of alizarin red-positive mineralized matrix compared with Thy-1(low) cells. Furthermore, subcutaneous implantation of Thy-1(high) cells efficiently induced the formation of bone-like matrix. These results collectively suggest that Thy-1-positive dental pulp cells localized in the subodontoblastic layer had the ability to differentiate into hard tissue-forming cells, and thus these cells may serve as a source of odontoblastic cells.     

Cell dynamics in the pulpal healing process following cavity preparation in rat molars

Odontoblast-lineage cells acquire heat-shock protein (HSP)-25-immunoreactivity (IR) after they complete their cell division, suggesting that this protein acts as a switch between cell proliferation and differentiation during tooth development. However, there are few available data concerning the relationship between cell proliferation and differentiation following cavity preparation. The present study aims to clarify the expression of HSP-25 in the odontoblast-lineage cells with their proliferative activity after cavity preparation by immunocytochemistry for HSP-25 and cell proliferation assay using 5-bromo-2'-deoxyuridine (BrdU) labeling. In untreated control teeth, intense HSP-25-IR was found in odontoblasts and some subodontoblastic mesenchymal cells. Cavity preparation caused the destruction of odontoblasts and the disappearance of HSP-25-IR was conspicuous at the affected site, although some cells retained HSP-25-IR and subsequently most of them disappeared from the pulp-dentin border by postoperative day 1. Contrary, some subodontoblastic mesenchymal cells with weak HSP-25-IR began to take the place of degenerated cells, although no proliferative activity was recognizable in the dental pulp. Interestingly, proliferative cells in the dental pulp significantly increased in number on day 2 when the newly differentiating cells already arranged along the pulp-dentin border, and continued their proliferative activity in the wide range of the pulp tissue until day 5. These findings indicate that progenitor cells equipped in the subodontoblastic layer firstly migrate and differentiate into new odontoblast-like cells to compensate for the loss of the odontoblast layer, and subsequently the reorganization of dental pulp was completed by active proliferation of the mesenchymal cells occurring in a wide range of pulp tissue.     

Interodontoblastic collagen (von Korff fibers) and circumpulpal dentin formation: an ultrathin serial section study in the cat

The collagenous fibers of von Korff pass from the dentin matrix between the odontoblasts into the dental pulp. Although collagen fibrils are known to be present between odontoblasts, the existence of von Korff fibers has remained controversial. This may be because their continuity between the dentin matrix and the pulp has not been demonstrated ultrastructurally. In this study we have examined the odontoblast layer in the middle to apical regions of perfusion-fixed permanent canine teeth of cats by using transmission electron microscopy. Ultrathin sections of demineralized specimens revealed frequent bundles of collagen fibrils 1) entering the odontoblast layer from the predentin, 2) present between odontoblast cell bodies, and 3) passing from between the odontoblasts into the pulp. The question of continuity of these bundles from the predentin, across the odontoblast layer into the pulp was examined in ultrathin serial sections. Unbroken continuity of a collagen bundle from the predentin between the odontoblasts into the pulp was established in a reconstruction of one series of 22 serial sections and was very strongly suggested by a number of other series in which the numbers of available sections restricted their full visibility. This investigation has shown, therefore, that classical von Korff fibers are present and that these fibers are present in fully erupted teeth with closed apices, i.e., at a time when secondary circumpulpal dentinogenesis is in progress. The findings call for a reexamination of the question of von Korff fibers during mantle dentinogenesis and primary circumpulpal dentinogenesis. Resolution of their existence at the earlier stages of dentinogenesis should be possible by using the ultrathin serial-sectioning technique.     

Ultrastructure of odontogenic cells during enameloid matrix synthesis in tooth buds from an elasmobranch, Raja erinacae

The ultrastructure of the inner dental epithelial cells (IDE) and odontoblasts in elasmobranch (Raja erinacae) tooth buds was investigated by transmission electron microscopy to determine what contribution each cell type makes to the forming enameloid matrix. Row II, early stage, IDE cells contained few organelles associated with protein synthesis, whereas preodontoblasts appeared competent to initiate extracellular matrix production. Row III IDE cells are also devoid of organelles related to secretory protein synthesis, although these IDE cells accumulated large pools of intracellular glycogen. The glycogen appeared to be packaged into vesicles and exocytosed into the lateral extracellular space toward the forming enameloid matrix. Row III odontoblasts had a morphology consistent with an active protein secretory cell. No procollagen granules were present within the odontoblasts, however, nor were many collagen fibers observed in the enameloid matrix. Instead, non-collagenous "giant" fibers having 17.5-nm periodic cross striations were associated with the invaginations of odontoblast cell processes. Giant fibers, which spanned a clear zone adjacent to the odontoblasts, terminated within the enameloid matrix. Smaller 25-nm-wide "unit" fibers emanated from the giant fiber tips to form the bulk of the enameloid matrix. The clear zone, which separated the odontoblasts from the enameloid matrix at early stages, diminished in size at later stages until the odontoblast processes were completely embedded in the enameloid matrix. Nascent enameloid crystallites were observed only after a layer of unmineralized predentin was deposited beneath fully formed enameloid matrix. The results suggest that the major constituent of the enameloid matrix in skates is a non-collagenous protein derived from the odontoblasts. The inner dental epithelial cells appear to contribute large quantities of carbohydrates to the forming enameloid matrix.     

Responses of immunocompetent cells in the dental pulp to replantation during the regeneration process in rat molars

Responses of immunocompetent cells to tooth replantation during the regeneration process of the dental pulp in rat molars were investigated by immunocytochemistry using antibodies to class II major histocompatibility complex (MHC) molecules (OX6 antibody), monocyte/macrophage lineage cells (ED1 antibody) and protein gene product 9.5 (PGP 9.5), as well as by histochemical reaction for periodic acid-Schiff (PAS). Tooth replantation caused an increase in both the number of OX6- and ED1-positive cells and their immunointensity in the replanted pulp, but almost all PGP 9.5-immunoreactive nerves diminished in the initial stages. By postoperative day 3, many OX6- and ED -immunopositive cells had accumulated along the pulp-dentin border to extend their cytoplasmic processes into the dentinal tubules in successful cases. Once reparative dentin formation had begun after postoperative day 7, OX6- and ED1-immmunopositive cells became scattered in the odontoblast layer, while reinnervation was found in the coronal pulp. The temporal appearance of these immunocompetent cells at the pulp-dentin border suggests their participation in odontoblast differentiation as well as in initial defense reactions during the pulpal regeneration process. On postoperative day 14, the replanted pulp showed three regeneration patterns: (1) reparative dentin, (2) bone-like tissue formation, and (3) an intermediate form between these. In all cases, PAS-reactive cells such as polymorphonuclear leukocytes (PML) and mesenchymal cells occurred in the pulp space. However, the prolonged stagnation of inflammatory cells was also discernible in the latter two cases. Thus, the findings on PAS reaction suggest that the migration of the dental follicle-derived cells into the pulp space and the subsequent total death of the proper pulpal cells are decisive factors for eliciting bone-like tissue formation in the replanted pulp.     

Microscopic and histochemical study of odontoclasts in physiologic resorption of teeth of the polyphyodont lizard, Liolaemus gravenhorsti

Using tartrate-resistant acid phosphatase (TRAP), we examined the cytodifferentiation of odontoclast cells in resorbing areas of dental tissues during the replacement of teeth in a polyphyodont lizard, Liolaemus gravenhorsti. We also report, by means of Lectin-HRP histochemistry, the distribution pattern of some specific sugar residues of TRAPase-positive cells. For detection of TRAPase activity, the azo dye-coupling technique was used. Lectin binding sites were demonstrated by means of specific HRP-lectins. The process of tooth resorption was divided into four stages: 1) preresorption-the wall of the dental pulp is covered with an odontoblast layer, and no TRAP-positive cells are in the dental pulp; 2) early resorption-TRAP-positive multinucleate odontoclasts are present on the dental wall, but the rest of the pulp surface is still covered with an odontoblast layer; 3) later resorption-the entire surface of the pulp chamber is lined with multinucleate odontoclasts; and 4) final resorption-the tooth has been totally resorbed. Odontoclasts are usually detached from the resorbed surface, and show signs of degeneration. Of the six lectins used, PNA, ECA, and UEA-1 bind to multinucleated but not mononuclear cells. All the remaining lectins, BS-1, RCA(120), and LTA showed no binding to any cells of the teeth. The significance of saccharidic moieties such as acetyl-galactosamine, acetyl-glucosamine, and fucose sugar residues is difficult to ascertain. Perhaps these oligosaccharides might be borne on molecules associated with odontoclastic resorption or associated with multinucleation of odontoclasts after attachment to the dentine surface.     

An investigation of lymphatic vessels in the feline dental pulp

The existence of lymphatic vessels in the dental pulp has been a matter of continuing controversy. We have now used light microscopy to examine semithin transverse sections of perfusion-fixed incisors and canines in cats. Lymphatics were found in all the teeth studied. In most teeth they were present in the coronal, middle, and apical regions of the pulp; but in a few they were lacking coronally and in the middle. Within individual teeth, lymphatics were found in the subodontoblastic zone or more centrally in the pulp; but none were found in the odontoblast layer or in the pulp horns. Vessels located by light microscopy were subsequently examined by transmission electron microscopy. Their ultrastructural features were typical of lymphatics and included irregular, attenuated endothelium with adjacent cells joined in different ways. Occasional gaps connected the extracellular spaces with their lumens, and abluminal endothelial projections appeared to form open end bulbs. There was very little basement membrane, but anchoring filaments were found near the abluminal surface of the endothelium and near collagen fibrils. The total cross-sectional area of lymphatic vessels was measured in semithin sections and, with pulp area, increased from the coronal region to the middle. However, both areas decreased from the middle to the apical region suggesting either that lymph flows faster as it reaches the foramens of the apical delta or that some vessels leave the tooth through lateral root canals. Using the methods of light and transmission electron microscopy, therefore, we have shown that pulp lymphatic vessels exist. Questions remain, however, about their distribution within teeth, variations between teeth, and routes of exit from teeth.     

Induction of transforming growth factor-beta 1 on dentine pulp cells in different culture patterns

Recent studies have documented that TGF-beta1 takes part in dental pulp tissue repair. Moreover, dental pulp cells have the potential to differentiate into odontoblast-like cells and produce reparative dentine in this process. However, the molecular mechanisms and potential interactions between TGF-beta1 and dental pulp cells are not clear due to the complexity of the pulp/dentine microenvironment. In this study, we investigated the induction of TGF-beta1 on the dental pulp cells in cell culture, tissue culture and three-dimensional culture patterns. These results demonstrated that TGF-beta1 significantly increased the proliferation of cells and activity of ALPase. Dental pulp cells cultured in the presence of TGF-beta1 formed mineralization nodules. In the organ culture, dental pulp cells treated with TGF-beta1 differentiated into odontoblast-like cells and formed a pulp-dentinal complex; and TGF-beta1 significantly induced synthesis of dentine relative proteins DSPP, DMP-1. The dental pulp cells share some characteristics of the odontoblast, such as a parallel arrangement with columnar form and a unilateral cell process. Together, these data indicate that TGF-beta1 can make dental pulp cells differentiated into odontoblast-like cells and form the pulp-dentinal complex. Moreover, these results suggest that TGF-beta1 is an important regulatory factor in odontoblast differentiation during tooth development and pulp repair.     

Immunohistochemical investigation of lymphatic vessel formation control in mouse tooth development: lymphatic vessel-forming factors and receptors in tooth development in mice

The presence of lymphatic vessels in dental pulp has recently been controversial, and no conclusion has been reached. In this study, we investigated the control of lymphangiogenesis with dental pulp development in the mouse mandibular molar using VEGF-C, VEGF-D, and VEGFR-3 as indices of lymphatic vessel-controlling factors. In addition, to distinguish blood and lymphatic vascular epithelial cells, we performed immunohistochemical analysis using von Willebrand factor (vWF) and statistical analysis. In dental papilla in the bell-stage non-calcified period, mesenchymal cells positive for VEGF-C, VEGF-D, and VEGFR-3 increased and lumen-forming endothelial cells were noted, but vWF was negative, suggesting that these were actively forming lymphatic vessels. Positive undifferentiated mesenchymal cells, an increase in endothelial cells in dental pulp, and lumen expansion were noted early after birth. Positivity was also detected in the odontoblast layer and sheath of Hertwig after birth, suggesting that these factors also play important roles in odontoblast differentiation and maturation and periodontal ligament and tooth root formation. We embryologically clarified lymphatic vessel formation in dental pulp and a process of lymphatic vessel formation from blood vessels, suggesting involvement of the surrounding tissue, odontoblasts, and sheath of Hertwig in vessel formation.     

S100 protein-immunoreactive trigeminal neurons innervating the rat molar tooth pulp

S100-immunoreactivity (ir) was examined in tooth pulp primary neurons of the rat. An immunofluorescence method demonstrated that the molar tooth pulp contained S100-immunoreactive (ir) nerve fibers. In the root pulp, pulp horn and roof of the pulp chamber, S100-ir smooth and varicose fibers ramified and formed subodontoblastic nerve plexuses. All the fibers became varicose at the base of the odontoblastic layer and extended to the odontoblastic layer. Some varicose endings could be traced into the dentin. The trigeminal neurons retrogradely labeled with fluorogold (FG) from the first and second maxillary molar tooth pulps exhibited S100- and parvalbumin-ir. Approximately 60% and 24% of the labeled cells were ir for S100 and parvalbumin, respectively. Virtually all parvalbumin-ir FG-labeled cells showed S100-ir, while 40% of S100-ir ones coexpressed parvalbumin-ir. An immunoelectron microscopic method revealed that all myelinated axons and half of the unmyelinated axons in the root pulp contained S100-ir. In the odontoblastic layer, predentin and dentin, S100-ir neurites lost the Schwann cell ensheathment and made close contact with cell bodies and processes of odontoblasts. The odontoblastic layer also contained parvalbumin-ir neurites. These neurites were devoid of the Schwann cell ensheathment and in close apposition to cell bodies and processes of odontoblasts. S100-ir pulpal axons seemed to be insensitive to repeated neonatal capsaicin treatment. This study suggests that S100-ir tooth pulp primary neurons are mostly myelinated and that S100-ir unmyelinated axons in the root pulp are preterminal segments of myelinated stem axons.     

Distribution and synthesis of type I and type III collagens in developing mouse molar tooth root

Summary The distribution and synthesis of type I and type III collagens in the mouse molar tooth root have been investigated by correlating light and electron immunohistochemical data. Purified rabbit antibodies were raised against mouse type I and type III collagens and indirect immunoperoxidase procedures were used. In these conditions, predentin, pre-bone, and pre-acellular cementum were intensely immunostained for type I collagen. Both optic and ultrastructural data confirmed the presence of type I collagen at the epithelio-mesenchymal junction, but Hertwig's basement membranes remained unlabelled. The odontoblasts including the short polarized ones, osteoblasts, some cells of pulp mesenchyme and the perifollicular cells possessed type I collagen immunoreactivity in the rough endoplasmic reticulum (RER), Golgi complex and the secretory vesicles.Type III collagen immunoreactivity was strong in the perifollicular mesenchyme, light in the pulp mesenchyme and absent from the epithelio-mesenchymal junction, the predentin, pre-bone and pre-acellular cementum. Intracellular immunolabelling was detected at the ultrastructural level in the perifollicular cells by a faint homogeneous peroxidase deposit in the RER cisternae.Finally, these results, compared with previous biochemical and morphological data, represent the first dynamic aspect of collagens distribution and synthesis in the mouse molar root development. In terms of cell differentiation, our data also suggest that type III collagen synthesis does not occur during the odontoblast process of differentiation.     

An immunocytochemical study of pulpal responses to cavity preparation by laser ablation in rat molars by using antibodies to heat shock protein (Hsp) 25 and class II MHC antigen

Initial responses of odontoblasts and immunocompetent cells to cavity preparation by laser ablation were investigated in rat molars. In untreated control teeth, intense heat shock protein (Hsp) 25 immunoreactivity was found in the cell bodies of odontoblasts, whereas cells immunopositive for the class II major histocompatibility complex (MHC) antigen were predominantly located beneath the odontoblast layer in the dental pulp. Cavity preparation caused the destruction of the odontoblast layer and the shift of most class-II-MHC-positive cells from the pulp-dentin border toward the pulp core at the affected site. Twelve hours after cavity preparation, numerous class-II-MHC-positive cells appeared along the pulp-dentin border and extended their processes deep into the exposed dentinal tubules, but subsequently disappeared from the pulp-dentin border together with Hsp-25-immunopositive cells by 24 h after the operation. By 3–5 days postoperation, distinct abscess formation consisting of polymorphonuclear leukocytes was found in the dental pulp. The penetration of masses of oral bacteria was recognizable in the dentinal tubules beneath the prepared cavity. These findings indicate that cavity preparation by laser ablation induces remarkable inflammation by continuous bacterial infections via dentinal tubules in this experimental model, thereby delaying pulpal regeneration.This work was supported by Grant-in-Aid for Scientific Research to promote 2001-Multidisciplinary Research Projects in 2001–2005, and KAKENHI (C) (nos. 12671765 and 14571727 to H.O.) from MEXT     

The distribution and origin of calcitonin gene-related peptide-containing nerve fibres in feline dental pulp

Summary The origin and distribution of calcitonin gene-related peptide (CGRP)-like immunoreactivity in feline dental pulp were studied using indirect immunofluorescence. Nerve fibres with varicosities exhibiting CGRP-like immunoreactivity were observed to enter the pulp with blood vessels. Many CGRP-containing nerve fibres were found to extend along blood vessels in the central pulp, and some of these fibres exhibited a network arrangement in the walls of dental pulp blood vessels. However, some of fibres were apparently not associated with blood vessels. Some thin, CGRP-containing nerve fibres formed a part of the nerve plexus in the subodontoblastic area and penetrated into the odontoblastic layer. In animals that had undergone transection of the inferior alveolar nerve, no CGRP-containing nerve fibres were observed. Application of a double-immunofluorescence staining technique also revealed that the distribution of CGRP-containing nerve fibres is very similar to that of substance P-containing nerve fibres.     

The distribution and origin of calcitonin gene-related peptide-containing nerve fibres in feline dental pulp. Relationship with substance P-containing nerve fibres

The origin and distribution of calcitonin gene-related peptide (CGRP)-like immunoreactivity in feline dental pulp were studied using indirect immunofluorescence. Nerve fibres with varicosities exhibiting CGRP-like immunoreactivity were observed to enter the pulp with blood vessels. Many CGRP-containing nerve fibres were found to extend along blood vessels in the central pulp, and some of these fibres exhibited a network arrangement in the walls of dental pulp blood vessels. However, some of fibres were apparently not associated with blood vessels. Some thin, CGRP-containing nerve fibres formed a part of the nerve plexus in the subodontoblastic area and penetrated into the odontoblastic layer. In animals that had undergone transection of the inferior alveolar nerve, no CGRP-containing nerve fibres were observed. Application of a double-immunofluorescence staining technique also revealed that the distribution of CGRP-containing nerve fibres is very similar to that of substance P-containing nerve fibres.     

Fibromodulin-deficient mice display impaired collagen fibrillogenesis in predentin as well as altered dentin mineralization and enamel formation.

To determine the functions of fibromodulin (Fmod), a small leucine-rich keratan sulfate proteoglycan in tooth formation, we investigated the distribution of Fmod in dental tissues by immunohistochemistry and characterized the dental phenotype of 1-day-old Fmod-deficient mice using light and transmission electron microscopy. Immunohistochemistry was also used to compare the relative protein expression of dentin sialoprotein (DSP), dentin matrix protein-1 (DMP 1), bone sialoprotein (BSP), and osteopontin (OPN) between Fmod-deficient mice and wild-type mice. In normal mice and rats, Fmod immunostaining was mostly detected in the distal cell bodies of odontoblasts and in the stratum intermedium and was weaker in odontoblast processes and predentin. The absence of Fmod impaired dentin mineralization, increased the diameter of the collagen fibrils throughout the whole predentin, and delayed enamel formation. Immunohistochemistry provides evidence for compensatory mechanisms in Fmod-deficient mice. Staining for DSP and OPN was decreased in molars, whereas DMP 1 and BSP were enhanced. In the incisors, labeling for DSP, DMP 1, and BSP was strongly increased in the pulp and odontoblasts, whereas OPN staining was decreased. Positive staining was also seen for DMP 1 and BSP in secretory ameloblasts. Together these studies indicate that Fmod restricts collagen fibrillogenesis in predentin while promoting dentin mineralization and the early stages of enamel formation.     

Changes in the matrix proteins, fibronectin and collagen, during differentiation of mouse tooth germ.

The distribution of the matrix protein fibronectin was studied by indirect immunofluorescence in differentiating mouse molars from bud stage to the stage of dentin and enamel secretion, and compared to that of collagenous proteins procollagen type III and collagen type I. Fibronectin was seen in mesenchymal tissue, basement membranes, and predentin. The dental mesenchyme lost fibronectin staining when differentiating into odontoblasts. Fibronectin was not detected in mineralized dentin. Epithelial tissues were negative except for the stellate reticulum within the enamel organ. Particularly intense staining was seen at the epithelio-mesenchymal interface between the dental epithelium and mesenchyme. Fibronectin may here be involved in anchorage of the mesenchymal cells during their differentiation into odontoblasts. Procollagen type III was lost from the dental mesenchyme during odontoblast differentiation but reappeared with advancing vascularization of the dental papilla. Similarly, procollagen type III present in the dental basement membrane during the bud and cap stages disappeared from the cuspal area along with odontoblast differentiation. Weak staining was seen in predentin but not in mineralized dentin. The staining with anti-collagen type I antibodies was weak in dental mesenchyme but intense in predentin as well as in mineralized dentin.     

Fine structure of differentiating ameloblasts in the kitten.

The fine structure of differentiating ameloblasts was studied in the lower second molar of 1-week-old kittens after perfusion fixation with and without subsequent decalcification. The differentiation zone was divided into three phases. In Differentiation 1, ameloblasts are about 27 mum long and face an uninterrupted basal lamina. The predentin adjacent to the basal lamina contains a few collagen fibrils oriented mainly at right angles to the ameloblast surface. The specialized predentin forms a well-defined layer, up to 1.5 mum thick, referred to as the junctional layer. In Differentiation 2, ameloblast processes extend through the basal lamina and the thickness of the junctional layer. The processes consist of cytoplasmic sheets forming a honeycomb-like network. Dentin starts to calcify after process-formation is underway. Two distinct types of odontoblast processes, having different shapes and contents, come in contact with the ameloblasts and push into the ameloblastic layer. In Differentiation 3, stippled material appears in the extracellular spaces between ameloblasts. Later, stippled material-like substances appear in the predentin close to the ameloblast apex and close to odontoblast processes within the dentin. Ameloblasts now are up to 40 mum high. Enamel secretion starts in small circumscribed areas which gradually enlarge, leading to the disappearance of the ameloblast processes. These findings are compared with results obtained in other species, including man, and their possible functional significance is discussed.     

Notch2 protein distribution in human teeth under normal and pathological conditions

Notch signaling is essential for the appropriate differentiation of many cell types during development and, furthermore, is implicated in a variety of human diseases. Previous studies have shown that although the Notch1, -2, and -3 receptors are expressed in developing and injured rodent teeth, Notch2 expression was predominant after a lesion. To pursue the role of the Notch pathway in tooth development and disease, we have analyzed the expression of the Notch2 protein in embryonic and adult wounded human teeth. During the earlier stages of tooth development, the Notch2 protein was expressed in the epithelium, but was absent from proliferating cells of the inner enamel epithelium. At more advanced stages, Notch2 was expressed in the enamel-producing ameloblasts, while it was absent in mesenchyme-derived odontoblasts that synthesize the dentin matrix. Although Notch2 was not expressed in the pulp of adult intact teeth, it was reexpressed during dentin repair processes in odontoblasts and subodontoblastic cells. Transforming growth factor beta-1, which stimulates odontoblast differentiation and hard tissue formation after dental injury, downregulated Notch2 expression in cultured human dental slices, in vitro. These observations are consistent with the notion that Notch signaling is an important element in dental physiological and pathogenic conditions.