Tooth morphogenesis and pattern of odontoblast differentiation.

The terminal differentiation of odontoblasts is controlled by the inner dental epithelium (IDE) and occurs according to a tooth-specific pattern. It requires temporospatially regulated epigenetic signaling and the expression of specific competence. The patterning of cusp formation was compared with that of odontoblast differentiation in the first lower molar in mice. Histology, immunostaining, and three dimensional reconstructions were completed by experimental approaches in vitro. The mesenchyme controls the pattern of cusp formation. During the cap-bell transition in the molar, a subpopulation of nondividing IDE cells from the enamel knot (EK) undergo a tooth-specific segregation in as many subpopulations as cusps will form. Epithelial cell-basement membrane interactions seem to be involved in the segregation of EK cells. The timing and spatial pattern of the segregation of EK cells correlate with cusps formation. However, the temporal pattern of odontoblast terminal differentiation is different. This discrepancy might result from cusp-specific differences either in the timing of the initiation of odontoblast terminal differentiation and/or in cell proliferation kinetics.

An address to young research workers: Inconsequences and blindness to the facts and unpublished observations.

The critical remembrance of my research activity displays aspects of inconsequence and mental rigidity. Particular probative examples are called to mind. Unpublished observations concerning immortalized dental cells are summarized. As a consequence, if I could start again, I would improve several aspects of my behavior and attitude of mind.

Cell-cell and cell-matrix interactions during initial enamel organ histomorphogenesis in the mouse.

Relationships between cell-cell/cell-matrix interactions and enamel organ histomorphogenesis were examined by immunostaining and electron microscopy. During the cap-bell transition in the mouse molar, laminin-5 (LN5) disappeared from the basement membrane (BM) associated with the inner dental epithelium (IDE), and nondividing IDE cells from the enamel knot (EK) underwent a tooth-specific segregation in as many subpopulations as cusps develop. In the incisor, the basement membrane (BM) in contact with EK cells showed strong staining for LN5 and integrin alpha 6 beta 4. LN5 seems to provide stable adhesion, while its proteolytic processing might facilitate cell segregation. In both teeth, immunostaining for antigens associated with desmosomes or adherens junctions was similar for EK cells and neighboring IDE cells. Outside the EK, IDE cell-BM interactions changed locally during the initial molar cusp delimitation and on the labial part of the incisor cervical loop. Conversely, cell-cell junctions stabilized the anterior part of the incisor during completion of morphogenesis. Time and space regulation of cell-matrix and cell-cell interactions might thus play complementary roles in allowing plasticity during tooth morphogenesis and stabilization at later stages of epithelial histogenesis.

Epigenetic signals during odontoblast differentiation.

Odontoblast terminal differentiation occurs according to a tooth-specific pattern and implies both temporospatially regulated epigenetic signaling and the expression of specific competence. Differentiation of odontoblasts (withdrawal from the cell cycle, cytological polarization, and secretion of predentin/dentin) is controlled by the inner dental epithelium, and the basement membrane (BM) plays a major role both as a substrate and as a reservoir of paracrine molecules. Cytological differentiation implies changes in the organization of the cytoskeleton and is controlled by cytoskeleton-plasma membrane-extracellular matrix interactions. Fibronectin is re-distributed during odontoblast polarization and interacts with cell-surface molecules. A non-integrin 165-kDa fibronectin-binding protein, transiently expressed by odontoblasts, is involved in microfilament reorganization. Growth factors (TGF beta 1, 2, 3/BMP2, 4, and 6), expressed in tooth germs, signal differentiation. Systemically derived molecules (IGF1) may also intervene. IGF1 stimulates cytological but not functional differentiation of odontoblasts: The two events can thus be separated. Immobilized TGF beta 1 (combined with heparin) induced odontoblast differentiation. Only immobilized TGF beta 1 and 3 or a combination of FGF1 and TGF beta 1 stimulated the differentiation of functional odontoblasts over extended areas and allowed for maintenance of gradients of differentiation. Presentation of active molecules in vitro appeared to be of major importance; the BM should fulfill this role in vivo by immobilizing and spatially presenting TGF beta s. Attempts are being made to investigate the mechanisms which spatially control the initiation of odontoblast differentiation and those which regulate its propagation. Analysis of molar development suggested that odontoblast differentiation and crown morphogenesis are interdependent, although the possibility of co-regulation requires further investigation.

Inhibition of apoptosis in the primary enamel knot does not affect specific tooth crown morphogenesis in the mouse.

The enamel knot (EK), located in the center of cap-stage tooth germs, is a transitory cluster of non-dividing epithelial cells, eventually linked to the outer dental epithelium by the enamel septum (ES). It might act as a signaling center providing positional information for tooth morphogenesis and could regulate the growth of tooth cusps through the induction of secondary signaling EKs. The EK undergoes apoptosis, which could constitute a mechanism whereby the signaling functions of this structure are terminated. Recently, we demonstrated the segregation of 5-bromo-2'-deoxyuridine (BrdU) negative inner dental epithelial (IDE) cells of the EK into as many individual groups of cells as cusps will form and suggested a morphogenetic role for these particular IDE cells. Using Z-VAD-fmk, a specific caspase inhibitor, apoptosis in the primary EK of first mouse lower cap-staged molars and lower incisors cultured in vitro was abrogated. No obvious histological alterations were observed in the incisors, whereas a prominent EK and an ES connecting the outer dental epithelium (ODE) and the BrdU negative IDE cells capping cusp L2 were observed in the molars. EK specific transcription (Shh, Msx-2, Bmp-2, Bmp-4) was down-regulated in the body of these structures with the exception of the associated IDE cells. In these experimental conditions, segregation of non-dividing transcriptionally active IDE cells occurred and a normal cusp pattern was expressed.

Effects of hepatocyte growth factor anti-sense oligodeoxynucleotides or met D/D genotype on mouse molar crown morphogenesis.

Hepatocyte growth factor (HGF) is considered to be one of the mediators of epithelio-mesenchymal interactions during early organogenesis and to be also involved in the development of murine molars. In the developing tooth, HGF is expressed in the cells of the dental papillae, and c-Met, its receptor, in the cells of dental epithelia. In order to study the functional role played by HGF in tooth development, we tested the effects of HGF translation arrest by anti-sense phosphorothioate oligodeoxynucleotides on E-14 molars cultured in vitro. We also analyzed the histo-morphogenesis and crown cytodifferentiation of transgenic met E-14 molars cultured in vitro. 3D reconstructions revealed perturbations of the cusp pattern. However, histo-morphogenesis and crown cytodifferentiation were normal at the histological level.

In vitro synchronization of embryonic mouse incisor preodontoblasts and preameloblasts: repercussions on terminal differentiation.

Preodontoblasts divide asynchronously and their terminal differentiation occurs gradually. Experimental data suggested that the expression of competence by preodontoblasts to respond to specific epigenetic signals, triggering their overt differentiation, requires a minimal number of cell cycles. The intrinsic timing mechanism could imply division counting and preodontoblasts of juxtaposed cell generations might sequentially withdraw from the last physiological cycle. To test such an hypothesis, embryonic mouse lower incisors were cultured in vitro and treated sequentially with nocodazole in order to induce a transitory synchronization of the dividing preodontoblasts and preameloblasts. This synchronization led to a disorganization of the physiological gradual terminal differentiation of the odontoblasts, giving rise to three distinct domains comprising respectively: 1) odontoblasts with altered polarization and predentin secretion; 2) odontoblasts demonstrating equivalent polarization and predentin deposition; and 3) preodontoblasts-odontoblasts involved in gradual terminal differentiation. These results strongly suggest that the gradient of odontoblast functional differentiation results from sequential withdrawal from asynchronous cell cycles of competent cells able to overtly differentiate.

Position and growth of upper and lower tooth primordia in prenatal mouse--3D study.

The secondary palate formation in mouse has been associated with the period of fast growth of the mandible from embryonic days (ED) 13.0 to 16.0. During that time, the incisors and first molars develop from the bud to the bell stage. We investigated the position and growth of the tooth during prenatal elongation of the lower and upper jaws, and searched for the developmental stage when alignment of opposing teeth was achieved. Computer-aided 3D representations allowed us to represent the position of incisors and molars in the embryonic head from ED 13.5 to 18.0 on the basis of data obtained from histological sections. The atlas-hypophysis connection exhibited minimum change in length and orientation during the prenatal period, and thus was used as a reference line. The length of the teeth was calculated from 3D data. The upper first and second molars were longer than the lower ones. When viewed from the upper side, the upper and lower molar primordia were parallel from ED 13.5 to 15.0. During this period, the upper molars had a more lateral position than the lower ones. This situation was maintained in the anterior extremity of the first molars at later stages, while the posterior part of the upper and lower molar epithelia reached opposition in the medio-lateral direction from ED 16.0. The lower incisors exhibited an apparently backward position when compared to the upper incisors at earlier stages. However, the distance between the prospective anterior tips of the opposing incisors gradually decreased. The part of Meckel's cartilage associated with the lower dental quadrant elongated more than 3-fold from ED 13.5 to 18.0, and the lower jaw grew faster than the upper one. This difference resulted from the fast growth of the lower diastema from ED 14.0 to 18.0. The different growth speeds of the upper and lower jaws did not change the relative antero-posterior adjustment of the upper and lower molars, but contributed to achieving the opposition of the gnawing ends of the incisors.

Dissection of the odontoblast differentiation process in vitro by a combination of FGF1, FGF2, and TGFbeta1.

Dental papillae (DP) isolated from first lower molars of 17-day-old mouse embryos were cultured in the presence of combinations of the following growth factors: FGF1, FGF2, and TGFbeta1. After 6 days in culture, only the DP treated with FGF1+TGFbeta1 contained differentiated odontoblast-like cells at the periphery of the explants, and these cells secreted extracellular matrix similar to predentin. Surprisingly, treatments with FGF2+TGFbeta1 induced cell polarization at the surface of the explants but no matrix secretion was observed. Electron microscopy and histochemical analysis of odontoblast markers showed that differentiated cells induced by FGF1+TGFbeta1 exhibited cytological features of functional odontoblasts with matrix vesicle secretion and mineral formation, positive alkaline-phosphatase activity, and type-I collagen production. DP cultured in the presence of FGF2+TGFbeta1 showed cell polarization and long and thin cell processes containing matrix vesicles; however, type-I collagen secretion was not detected and alkaline-phosphatase activity was completely inhibited. Our results indicate that, in our culture system, exogenous combinations of FGF1, FGF2, and TGFbeta1 interact with preodontoblasts and induce cell polarization or differentiation, which can be studied separately in vitro. Thus, FGF1 and TGFbeta1 do have a synergic effect to promote morphological and functional features of differentiated odontoblasts whereas FGF2 seems to modulate TGFbeta1 action, causing morphological polarization of preodontoblasts but limiting the functional activity of these cells in terms of type-I collagen secretion and alkaline-phosphatase activity.

Regeneration of halved embryonic lower first mouse molars: correlation with the distribution pattern of non dividing IDE cells, the putative organizers of morphogenetic units, the cusps.

Recently we demonstrated that non-cycling, cap-stage, mouse molar inner dental epithelial (IDE) cells corresponding to the primary enamel knot (EK) area underwent a coordinated temporo-spatial patterning leading to their patchy irregular segregation at the tips of the forming cusps. These non-cycling cells were suggested to perhaps represent the organizers of the morphogenetic units (OMU), the cusps. The present study has analyzed the regenerative capacity of halved cap-stage first lower mouse molars through three dimensional (3D) reconstructions. Partial regeneration of the anterior half and possible complete regeneration of the posterior half were documented. Using BrdU (5-bromo-2'-deoxyuridine) labeling and 3D reconstructions of the IDE, we have correlated the patterns of cusp regeneration with the distribution of BrdU negative IDE cells. These data support a morphogenetic role for the non-cycling IDE cells.

Differential expression of laminin-5 subunits during incisor and molar development in the mouse.

Rodent incisors are continuously growing teeth and enamel deposition is restricted to the labial side. In the present study, the expression of laminin-5 subunits (alpha3, beta3 and gamma2) has been analyzed by in situ hybridization in developing mouse lower incisors and compared to that reported in the molar. At the bud stage (E12), mRNAs for all subunits were detected in the whole epithelial thickening. At E14, when histogenesis had started, transcripts for alpha3 and gamma2 subunits were restricted to the outer dental epithelium (ODE), whereas the beta3 subunit was intensely expressed in the inner dental epithelium (IDE). A transient expression for alpha3 subunit was seen in the enamel knot area and disappeared at E15. Subsequently, all laminin-5 subunit genes were re-expressed in differentiating ameloblasts on the labial side. Similar patterns of transcription were observed in incisor and molar, suggesting that the differential expression of laminin-5 subunits in the IDE might be involved in the histogenesis of the IDE and ameloblast differentiation. At E16.5, cells of the IDE at the anterior extremity of the incisor and in the anterior part of the lingual IDE expressed transcripts for alpha3 and beta3 but not for gamma2 subunit. Similar expression patterns were observed in the enamel-free areas of the E18 molar. This specific expression might thus be related to cells that do not differentiate as functional ameloblasts. Throughout incisor development, intense expression for all laminin-5 subunits was restricted to the labial side of the cervical loop. The asymmetrical expression of laminin-5 might be related to incisor morphogenesis and to the differences in histogenesis and cytodifferentiation of the IDE that exist in the labial versus lingual aspect of the cervical loop.

The presence of rudimentary odontogenic structures in the mouse embryonic mandible requires reinterpretation of developmental control of first lower molar histomorphogenesis.

In the mouse embryonic maxilla, rudimentary tooth primordia have been identified, which can be mistaken for the first upper molar. In order to determine whether such a situation might exist in the lower jaw as well, tooth development was investigated in the mouse mandibular cheek region during ED 12.5-15.0. A combination of histology, morphometry and computer-aided 3D reconstructions demonstrated the existence of rudimentary dental structures, whose gradual appearance and regression was associated with the segmental progress of odontogenesis along the mesio-distal axis of the jaw: 1) At ED 12.5, the mesial segment (MS) was the most prominent part of the dental epithelial invagination. It included an asymmetrically budding dental lamina. The MS, although generally mistaken for the lower first molar (M1, primordium, regressed and did not finally participate in M1 cap formation. 2) At ED 13.5, a wide dental bud (called segment R2) appeared distally to the MS. Although the R2 segment transiently represented the predominant part of the dental epithelium at ED13.5, it participated only in the formation of the mesial end of the M1 cap. 3) The top of the R2 segment at ED13.5 was not the precursor of the enamel knot (EK), contrary to what has been assumed. 4) The central segment of the M1 cap as well as the EK developed later and distally to the R2 segment. 5) Time-space specific apoptosis correlated with the retardation in growth of the R2 segment as well as with strong regressive changes in the epithelium situated mesially to it. These highlight the need to reinterpret current molecular data on early M1 development in the mouse in order to correlate the expression of signalling molecules with specific morphogenetic events in the appropriate antemolar or molar segments of the embryonic mandible.

Alterations in the incisor development in the Tabby mouse.

The X-linked tabby (Ta) syndrome in the mouse is homologous to the hypohidrotic ectodermal dysplasia (HED) in humans. As in humans with HED, Ta mice exhibit hypohidrosis, characteristic defects of hairs and tooth abnormalities. To analyze the effects of Ta mutation on lower incisor development, histology, morphometry and computer-aided 3D reconstructions were combined. We observed that Ta mutation had major consequences for incisor development leading to abnormal tooth size and shape, change in the balance between prospective crown- and root-analog tissues and retarded cytodifferentiations. The decrease in size of Ta incisor was observed at ED13.5 and mainly involved the width of the tooth bud. At ED14.5-15.5, the incisor appeared shorter and narrower in the Ta than in the wild type (WT). Growth alterations affected the diameter to a greater extent than the length of the Ta incisor. From ED14.5, changes in the shape interfered with the medio-lateral asymmetry and alterations in the posterior growth of the cervical loop led to a loss of the labio-lingual asymmetry until ED17.0. Although the enamel organ in Ta incisors was smaller than in the WT, a larger proportion of the dental papilla was covered by preameloblasts-ameloblasts. These changes apparently resulted from reduced development of the lingual part of the enamel organ and might be correlated with a possible heterogeneity in the development of the enamel organ, as demonstrated for upper incisors. Our observations suggest independent development of the labial and lingual parts of the cervical loop. Furthermore, it appeared that the consequences of Ta mutation could not be interpreted only as a delay in tooth development.

Morphogenesis of the lower incisor in the mouse from the bud to early bell stage.

The development of the lower incisor in the mouse was investigated from histological sections using computer-aided 3D reconstructions. At ED 13.0, the incisor was still at the bud stage. At ED 13.5, the initial cap was delimited by a short cervical loop, the development of which proceeded on the labial side, but was largely retarded on the medial side. This difference was maintained up to ED 15.0. From ED 16.0, the bell stage was achieved. Metaphases had a ubiquitous distribution both in the enamel organ and in the dental papilla from the bud to early bell stage. Apoptosis gradually increased in the mesenchyme posteriorly to the labial cervical loop from ED 13.5 to 14.0 and then disappeared; this apoptosis was not related to the posterior growth of the incisor. From ED 13.5, a high apoptotic activity was observed in the stalk. A focal area of apoptosis was observed at ED 13.5 in the enamel organ, approaching the epithelio-mesenchymal junction at the future tip of the incisor. There, the inner dental epithelium formed a bulbous protrusion towards dental papilla, reminiscent of the secondary enamel knot of mouse molars. This epithelial protrusion was still maintained at the bell stage. The enamel knot in the incisor demonstrated specific features, different from those characterizing the enamel knot in the molar: the concentric arrangement of epithelial cells was much less prominent and the occurrence of apoptosis was very transitory in the incisor at ED 13.5. The disappearance of the enamel knot despite a low apoptotic activity and the maintenance of the protrusion suggested a histological reorganization specific for rodent incisor.

Initial aspects of mineralization at the dentino-enamel junction in embryonic mouse incisor in vivo and in vitro: a tem comparative study.

The frontier between the enamel organ and the dental papilla, the future dentino-enamel junction, undergoes coordinated modifications. The mineralization of the extracellular matrix starts within the predentine, which is a prerequisite for the formation of the first enamel crystallites in vivo. We investigated the dentino-enamel junction using the embryonic mouse incisor as a model. Our data showed that the notion of the dentino-enamel junction should not be restricted to the thin interface classically described. A temporo-spatial survey from the epithelio-mesenchymal junction to the dentino-enamel junction delineated a clear sequence of events characterized by the early deposition of electron-dense granules, followed by the appearance of patches of stippled material at the dentino-enamel junction. The first tiny enamel crystallites appeared in the vicinity of this material which presented a well-ordered alignment. The comparison of data obtained in vivo on 17-, 18-, 19-d-old embryonic incisors with those obtained in vitro using 15-d-old embryonic incisors cultured for 7 d emphasizes the relevance of this sequence. Helicoidal growing crystals were observed in cultured tooth germs but never in vivo.

Initial features of the inner dental epithelium histo-morphogenesis in the first lower molar in mouse.

First lower molar development in the mouse was investigated from the cap to early bell stage using histology, morphometry, TEM and 3D reconstructions. This period was characterized by the histogenesis of the enamel organ (EO), folding of the epithelio-mesenchymal junction and growth of the tooth. The histogenesis of the EO and appearance of the enamel knot (EK) were initiated at the early cap stage (ED14). From ED14 to ED15, the anterior and posterior extension of the EK was very prominent whilst the length of the enamel organ did not substantially change. The EK appeared as a dynamic and transitory histological structure including dying and replacement cells. At ED16, the folding of the IDE, which extended over the anterior two thirds of the molar, was the first sign of cuspidogenesis. It was accompanied by a local remodeling of the basement membrane (BM): IDE cells involved in this folding transitorily lost contact with the BM which formed a loop in the mesenchyme. During this period, the growth of the lower M1 along the antero-posterior axis was restricted to the posterior part of the molar. Histogenesis occurred in the whole EO, whilst initial cuspidogenesis was limited to the anterior part of the tooth. Distinct cell populations were thus involved in different contemporary processes leading to changes in the cell density in the mesenchyme, in the mitotic activity, in cell-shape, and cell-matrix interactions in the IDE, and remodeling of the BM where both epithelium and mesenchyme might participate.

Mouse odontogenesis in vitro: the cap-stage mesenchyme controls individual molar crown morphogenesis.

Day 14 ICR mouse first lower (M1) and upper molars (M1) as well as heterotopic recombinations of M1 epithelium/M1 mesenchyme and M1 epithelium/M1 mesenchyme were cultured for 6, 8 and 10 days on semi-solid medium. Computer-assisted 3D reconstructions were performed to follow the in vitro development of these explants. In vitro culture of cap-stage molars allowed for the emergence of unequivocal morphological features distinctive for M1 versus M1 including the cusp pattern, cusp inclination and tooth specific chronology for odontoblast and ameloblast terminal differentiations. Both M1 epithelium/M1 mesenchyme and M1 epithelium/M1 mesenchyme recombinations developed according to the known developmental fate of the mesenchyme. Our data demonstrate that the cap-stage dental ecto-mesenchyme not only directs tooth class specific morphogenesis, but also individual molar crown features. Furthermore, the mesenchyme apparently also controls the typical mirror symmetry of right and left handed teeth.

Aspects of cell proliferation kinetics of the inner dental epithelium during mouse molar and incisor morphogenesis: a reappraisal of the role of the enamel knot area.

First lower E-14 and E-16 mouse molars and E-13 lower incisors were cultured in vitro and either sequentially or continuously labelled with BrdU (5-bromo-2'-deoxyuridine). The behaviour of the non-cycling inner dental epithelial cells emerging from the enamel knot area of the molars was analysed by 3D (three dimensional) reconstructions of serial sections. These cells, as well as slow cycling cells underwent a coordinated temporo-spatial patterning leading to their patchy segregation at the tips of the forming cusps. In incisors (in vitro and in vivo), non-cycling cells were also present in the inner dental epithelium of the enamel knot area. However, these cells were not redistributed during incisor morphogenesis. These non-dividing inner dental epithelium cells of the enamel knot area which are either redistributed or not according to the tooth type specific morphogenesis might represent the organizers of morphogenetic units (OMU), the cusps.

Effects of apatite, transforming growth factor beta-1, bone morphogenetic protein-2 and interleukin-7 on ameloblast differentiation in vitro.

Preameloblasts overtly differentiate in vitro when recombined with cell-free predentin-dentin. The predentin-dentin components able to trigger ameloblast terminal differentiation have not been identified, but some growth factors (members of the TGFbeta superfamily) have been demonstrated to be associated with this matrix, and in situ hybridization has demonstrated the presence of transcripts for TGFbetas and BMP-2 in odontoblasts facing differentiating ameloblasts. Moreover, intense expression of receptors for the cytokine interleukin-7 (IL-7) in polarizing ameloblasts has been reported. In this study, isolated E-18 and E-19 mouse molar enamel organs were cultured in vitro in presence of BMP-2, TGFbeta-1, IL-7 or synthetic apatite. TGFbeta-1 and BMP-2 combined with heparin induced cytodifferentiation of ameloblasts. IL-7 maintained the polarized state of ameloblasts. BMP-2-soaked apatite induced functional differentiation of ameloblasts (secretion of amelogenin). Integrating these data with previous work, a working hypothesis concerning the control of ameloblast terminal differentiation is presented: members of the TGFbeta superfamily secreted by odontoblasts might be trapped by predentin components first and then by dentin apatites, and these growth factors might trigger the cytological-functional sequence of ameloblast terminal differentiation.

Localization of antigens associated with adherens junctions, desmosomes, and hemidesmosomes during murine molar morphogenesis.

Epitheliomesenchymal interactions are known to play a crucial role during odontogenesis. Since epithelial cell-cell and cell-matrix interactions may also be involved in enamel organ histomorphogenesis, we investigated the localization of proteins associated with junctional complexes in mouse and rat first lower molars by indirect immunofluorescence. Adherens junctions were detected using antibodies directed against E-cadherin, beta-catenin, and plakoglobin (gamma-catenin). Desmosomes were localized with antibodies against desmoglein, and hemidesmosomes using antibodies against BP-230 and HD-1 proteins. When the inner dental epithelium differentiates, a decrease of E-cadherin, plakoglobin, and BP-230 is seen. An asymmetric distribution of plakoglobin, desmoglein, and BP-230 between the lateral and medial side of the tooth exists; desmoglein, which was first restricted to the gubernaculum dentis, progressively accumulated in the stellate reticulum, the stratum intermedium, and the basal pole of ameloblasts. The specific temporospatial distributions patterns of these antigens suggests a direct involvement of adherens junctions, desmosomes, and hemidesmosomes in the development of the murine first lower molar.