Mechanisms of Development 67 (1997) 111123

Review article

Signalling networks regulating dental development

Irma Thesleff a ,*, Paul Sharpe b
a

Institute of Biotechnology, PO Box 56, University of Helsinki, FIN-00014, Helsinki, Finland b Department of Craniofacial Development, Guys Hospital, London, UK Received 29 May 1997; accepted 22 July 1997

Abstract There has been rapid progress recently in the identication of signalling pathways regulating tooth development. It has become apparent that signalling networks involved in Drosophila development and development of mammalian organs such as the limb are also used in tooth development. Teeth are epithelial appendages formed in the oral region of vertebrates and their early developmental anatomy resembles that of other appendages, such as hairs and glands. The neural crest origin of tooth mesenchyme has been conrmed and recent evidence suggests that specic combinations of homeobox genes expressed in the neural crest cells may regulate the types of teeth and their patterning. Signalling molecules in the Shh, FGF, BMP and Wnt families appear to regulate the early steps of tooth morphogenesis and some transcription factors associated with these pathways have been shown to be necessary for tooth development. Several of the conserved signals are also transiently expressed in the enamel knots in the dental epithelium. The enamel knots are associated with the characteristic epithelial folding morphogenesis which is responsible for the development of tooth shape and it is currently believed that the enamel knots function as signalling centres regulating tooth shape development. The developing tooth has proven to be an excellent model in studies of the molecular basis of patterning and morphogenesis of organs and it can be expected that continuing studies will rapidly increase the understanding of these mechanisms. 1997 Elsevier Science Ireland Ltd. Keywords: Tooth morphogenesis; Epithelialmesenchymal interactions; Pattern formation; Homeobox genes; FGF; Hedgehog; BMP; Wnt

1. Introduction Interactions between cells and tissues constitute a central mechanism regulating the development of all multicellular organisms. At the molecular level, these interactions involve complex signalling networks composed of signals, their receptors and transcriptional control systems. During the last 10 years it has become increasingly evident that these regulatory networks are composed of similar elements in a wide variety of organisms ranging from worms and ies to mammals. Moreover, these elements are used in the regulation of multiple tissues and organs within the same animal and even repeatedly during the development of an individual organ. Proteins that comprise intercellular signalling pathways, including ligands, receptors and transcription factors belong
* Corresponding author. Tel.: +358 9 70859401; fax: +358 9 70859560; e-mail: irma.thesleff@helsinki.

to a number of families which in the lower animals are composed of one or a few members, whereas in mammals, as a result of gene duplications, the families can often constitute more than 10 molecules. In this review we discuss recent work which indicates that the developing vertebrate tooth is no exception to the general principle and that an increasing number of the evolutionarily conserved elements have been found to regulate different aspects of tooth development.

2. Developmental anatomy of tooth morphogenesis Teeth are typical examples of epithelialmesenchymal organs. They develop from stomodeal or pharyngeal epithelium and the underlying neural crest-derived mesenchymal cells in a very similar way to skin derivatives such as hairs, feathers and scales (for review see Thesleff et al., 1995). The rst morphological sign of tooth development is a

0925-4773/97/$17.00 1997 Elsevier Science Ireland Ltd. All rights reserved PII S0925-4773 (97 )0 0115-9

112

I. Thesleff, P. Sharpe / Mechanisms of Development 67 (1997) 111123

thickening of the oral epithelium, which subsequently buds into the underlying mesenchyme. The mesenchymal cells condense around the bud and during the following cap and bell stages, the epithelium undergoes folding morphogenesis resulting in the establishment of the form of the tooth crown (Fig. 1). As in the case of all developing tissues, the development of shape occurs in the context of extensive growth and exact location implying that morphogenesis and cell division must be tightly co-ordinated and linked to position. The mesenchyme which becomes surrounded by the dental epithelium during the cap stage forms the dental papilla, giving rise to tooth pulp and the odontoblasts. The more peripheral cells of the condensed dental mesenchyme extend around the epithelial dental component, the enamel organ, forming the dental follicle. When the roots of teeth develop, the follicle gives rise to the cementoblasts depositing dental cementum as well as to the brous periodontal membrane connecting the roots of the teeth to the alveolar bone. Although the initiation and early morphogenesis of teeth occurs before bone formation starts in the jaws, the development of the teeth and the surrounding bone is later tightly co-ordinated. Postnatally the teeth have a central role in regulating the extent and direction of the growth of the alveolar processes of the jaw bones (Ten Cate, 1994). The dentine-forming odontoblasts and the enamel-forming ameloblasts are unique to teeth and they differentiate terminally during the bell stage of tooth development. This takes place at the interface of the epithelium and mesenchyme and is regulated by interactions between the two tissues (Ruch et al., 1995; Thesleff and Aberg, 1997). The odontoblasts secrete a collagenous extracellular matrix which subsequently mineralizes into dentin, a bone-like hard tissue. The ameloblasts deposit the enamel matrix which directs the mineralization of the enamel into the hardest tissue in the body. After crown morphogenesis, the roots of the teeth develop and subsequently the teeth erupt into the oral cavity (Fig. 1) (for more details see Ten Cate, 1994).

cesses are of neural crest origin (Osumi-Yamashita et al., 1994; Trainor and Tam, 1995; Imai et al., 1996). These neural crest cells derive from the midbrain region and their nal position in the maxillary and mandibular processes is associated with the original position of the cells in the neural crest as well as with the time when the cells leave the crest (Imai et al., 1996; Kontges and Lumsden, 1996). Labelled cells could be demonstrated also in the dental mesenchyme, thus conrming that neural crest cells are actually involved in tooth development (Imai et al., 1996). The earliest dental-like tissues were not restricted to the oral region and formed a dermal skeleton (Smith et al., 1996). Tooth development was subsequently conned to the oral region and during evolution, a variety of modications in teeth have been seen (Butler, 1995). The basic developmental anatomy of tooth morphogenesis has been conserved to a high degree and the morphological steps of early tooth development (Fig. 1) are very similar in all modern vertebrates. There are, however, a variety of modications in the ways in which the dentitions are organized and in which they function in different animals. The numbers of teeth vary and they come in a variety of forms which serve special functions in catching, chopping and chewing food. Changes in tooth shape and position are an important force in evolution, allowing adaptation to specialized diets and lling of different food niches. Because of their masticatory functions, the teeth are necessary for the survival of most animals.

4. Molecular control of tooth patterning Mammalian teeth are sequentially arranged structures and the different tooth groups (incisors, canines, premolars and molars) show characteristic differences in morphology. Each tooth group forms from one epithelial thickening, the dental lamina, and the development starts with the most anterior tooth and proceeds posteriorly. Transplantation experiments have shown that the morphogenetic elds for tooth development are present at the time when the dental lamina is seen (E10) (Glasstone, 1963; Kollar and Baird, 1969; Lumsden, 1988). Two different theories have been put forward to explain the segmental patterning and shape differences between the teeth. The eld theory assumes that concentrations of chemical morphogens regulate different morphogenesis of initially identical primordia (Butler, 1939). Although retinoic acid has been muted as a possible morphogen in tooth patterning based on effects of exogenous retinoids on tooth morphogenesis, the widespread effects of retinoic acid on many of the genes expressed during the stages of tooth morphogenesis and differentiation make this unlikely. According to the other theory, the so-called clone model, the stem cells giving rise to different classes of teeth differ

3. Neural crest and the evolution of teeth Teeth are only found in vertebrates and their evolution is believed to be associated with the appearance of the neural crest (Smith and Hall, 1993). The majority of craniofacial cartilages and bones are formed by neural crest derived mesenchymal cells, unlike more caudal skeletal elements which are of mesodermal origin. That cranial neural crest cells also contribute to tooth development was rst demonstrated in amphibians (Sellman, 1946). Recently, labelling of neural crest cells and analysis of their migration has been successful in mouse embryos. Labelling of neural crest cells in three to eight somite mouse embryos in vivo followed by whole embryo culture indicated that specically the mesenchymal cells beneath the surface ectoderm in the maxillary and mandibular pro-

I. Thesleff, P. Sharpe / Mechanisms of Development 67 (1997) 111123

113

Fig. 1. Schematic presentation of the morphology of tooth development. Initiation of teeth is seen as thickening of the epithelium in the facial processes. The budding of the epithelium is accompanied by condensation of the neural crest-derived mesenchymal cells. The enamel knot appears in the epithelium as it undergoes folding morphogenesis and develops into the cap stage. The form of the tooth crown is established during the bell stage and the matrices of dentin and enamel are deposited by the odontoblasts and ameloblasts, respectively. After completion of crown formation roots develop and the tooth erupts into the oral cavity. (Courtesy of Thomas Aberg.) Fig. 2. Whole mount in situ hybridization of a mouse embryo head at E10.5 showing expression of Dlx-2 in the mandibular and maxillary process. Expression in the mesenchyme is restricted to the proximal regions where molar teeth will develop indicated by the white arrows. Dlx-2 expression is also evident in the epithelium in more distal regions. The red arrows mark the distinct boundaries between epithelial and mesenchymal domains of expression. Fig. 3. Tooth phenotype of Dlx-1/2 double mutants. Maxillary (upper) molars are absent in new-born mutant animals. um, upper molars; lm, lower molars; t, tongue. (From Qiu et al., 1997 with permission.)

114

I. Thesleff, P. Sharpe / Mechanisms of Development 67 (1997) 111123

from each other initially (Osborn, 1978). In this model it is likely that the neural crest cells have acquired positional identity at the time they reach their nal destination in the region of tooth development. An updated, molecular version of the clone model was proposed by Sharpe (1995) whereby tooth shape and position were specied by the combinatorial action of different homeobox genes expressed in neural crest derived facial mesenchyme. Thus, the oral cavity is considered to be patterned by different populations of neural crest cells, determined by their expression of homeobox genes. This model is based on the observation that several homeobox genes are expressed in spatially restricted domains of oral mesenchyme prior to overt tooth development. The Dlx-1 and Dlx-2 genes, for example, are coexpressed specically in mesenchyme giving rise to molars (Fig. 2). Strong evidence for this homeobox code model has recently been provided from studies of tooth development in knockouts of the Dlx-1 and Dlx-2 genes in transgenic mice (Fig. 3) (Qiu et al., 1997). Mice carrying null mutations in both genes (double knockouts) show no development of maxillary molars whereas development of mandibular molars and all incisors is normal in the mutants. Mice lacking either Dlx-1 or Dlx-2 (single knockouts) have normal development of all teeth. Since Dlx-1 and Dlx-2 are co-expressed in maxillary mesenchyme the knockout phenotype indicates that these genes have a redundant function in the development of maxillary molars. This implies that development of teeth in different regions of the jaws involves independent genetic control. Only one gene has been identied that shows tooth type specic expression following initiation of development. The homeobox gene, Barx-1, is present in mouse molar mesenchyme and absent from incisors during early tooth morphogenesis (Tissier-Seta et al., 1995). The question of whether the determination of the position of tooth initiation is regulated by neural crest derived mesenchymal cells or by the stomatodeal ectoderm has not been satisfactorily resolved. In the case of many other ectodermal appendages, the mesenchyme determines the position of structures whereas the epithelium responds according to its inherent capabilities (Sengel, 1976). There is such evidence from interspecies tissue recombination experiments also in tooth development. Combination of mouth mesoderm from frog embryos, which normally do not develop teeth, with ank ectoderm from salamanders results in tooth development (Wagner, 1959). Heterotopic tissue recombination experiments in mouse embryos have shown that the presumptive dental epithelium instructs tooth development prior to the bud stage (Mina and Kollar, 1987; Lumsden, 1988). These studies showed that in 9-day-old mouse embryos, the rostral, but not caudal epithelium of the mandibular arch is capable of instructing tooth development in trunk neural crest as well as in the mesenchyme of the second branchial arch. However, as the neural crest cells had already reached the branchial arches at the time of the experiments, they may have induced the epithe-

lium to acquire specic tooth forming capacities before the dissection of tissues. The homeobox gene Otlx2 (RIEG) was recently shown to be expressed at embryonic day 8.5 specically in the rostral stomodeal epithelium and its expression was maintained in dental epithelium throughout tooth morphogenesis (Mucchielli et al., 1997). Mutations in the human counterpart of this gene cause Rieger syndrome, which is characterized by missing teeth and defects in the umbilicus and eye (Semina et al., 1996). Otlx2 (RIEG) is expressed very early in the dental lamina in the mouse embryo, but the expression is not restricted to different dental elds. This suggests that it is not involved in the specication of dental identity but rather may be a key regulator of the epithelial response to early positional signals from the mesenchyme.

5. Morphogenesis of individual teeth signalling networks between epithelium and mesenchyme The morphogenesis of all epithelial appendages is regulated by a sequence of reciprocal interactions between the epithelial and mesenchymal tissue components. This has been demonstrated in classic tissue recombination experiments. As mentioned above, the presumptive dental epithelium governs tooth development prior to the bud stage (Mina and Kollar, 1987; Lumsden, 1988). The potential to instruct tooth morphogenesis, however, shifts to the dental mesenchyme during budding and as demonstrated by Kollar and Baird (1969, 1970), the dental papilla of cap and bell stage teeth regulates tooth shape development, i.e. a molar tooth develops when molar mesenchyme is cultured with incisor epithelium and vice versa. Furthermore, the dental mesenchyme is able to instruct the differentiation of ameloblasts and enamel secretion in non-dental epithelium (Kollar and Baird, 1970). Hence, in addition to morphogenesis, cell differentiation is also regulated by epithelialmesenchymal interactions. The teeth belong to those organs in which the molecular basis of epithelialmesenchymal signalling has been elucidated to a signicant extent during recent years (Thesleff and Nieminen, 1996; Thesleff and Sahlberg, 1996). Numerous signal molecules and growth factors belonging to several different families as well as their specic receptors have been associated with epithelialmesenchymal signalling during tooth morphogenesis. It is apparent that the localized expression of the signals and receptors is directed by the epithelialmesenchymal interactions. In situ hybridization analysis of the patterns of expression of individual signals and their receptors indicate that in some cases they are restricted to either epithelial or mesenchymal tissues and sometimes to particular developmental stages. However, in many cases the signals seem to be used repeatedly during successive stages of morphogenesis and/or they appear to transfer messages in both directions between the interacting tissues. Some transcription factors have been identied

I. Thesleff, P. Sharpe / Mechanisms of Development 67 (1997) 111123

115

which are involved in the epithelialmesenchymal signalling during tooth morphogenesis and their roles in the signalling cascades are starting to be elucidated (Fig. 6). 5.1. BMP signalling pathway Bone morphogenetic proteins (BMPs) were the rst signals to be identied in the transmission of inductive interactions between the dental epithelium and mesenchyme (Vainio et al., 1993). BMPs constitute one of the signal families which has widespread signalling functions throughout the animal kingdom (Hogan, 1996). The best characterized member is the Drosophila morphogen Dpp, which is homologous with vertebrate BMP-2 and BMP-4. The expression of these Bmps has been associated with morphogenetic tissue interactions in a variety of organs (Hogan, 1996). Bmp-2 and Bmp-4 as well as Bmp-7 are expressed early in the dental epithelium (Fig. 4) and interestingly the expression of Bmp-4 shifts to the mesenchyme at the time when the instructive capacity shifts from the epithelium (Vainio et al., 1993). The pathway of BMP signalling in tooth development has been analyzed in organ culture studies by examining the responses of dental tissues to recombinant BMP proteins. BMP-2 and BMP-4 were shown to mimick some effects of presumptive dental epithelium on mesenchyme at the time when the instructive capacity still resided in the epithelium. In particular, they stimulated the expression of the homeobox-containing transcription factors Msx-1 and Msx-2 (Fig. 4) (Vainio et al., 1993). Unlike the epithelium, BMPs did not stimulate cell proliferation and syndecan expression suggesting that other epithelial signals are involved. Targeted mutations in Bmp-2 and Bmp-4 result in embryonic lethality prior to development of tooth germs so the contributions of these signals in early tooth development cannot be directly assessed in vivo (Winnier et al., 1995). However, analysis of the effect of mutations in the Msx-1 homeobox gene have revealed some detail of the importance of BMP signalling in early tooth development. Tooth development is arrested at the late bud stage in Msx-1 mutants and analysis of mutant embryos has indicated that Msx-1 appears to act also upstream of BMP signalling since in mutant teeth, Bmp-4 expression in mesenchyme is not upregulated as it is in wild type animals (Satokata and Maas, 1994; Chen et al., 1996). This was further supported by a rescue experiment in which the bud stage tooth germs, dissected from the Msx-1 decient embryos, reached cap stage when cultured in the presence of BMP-4 protein. Interestingly, MSX-1 was recently associated with defective tooth development also in humans. In one family, a mutation in the homeodomain of the MSX-1 gene was shown to cause oligodontia, i.e. the lack of more than six teeth in affected individuals (Vastardis et al., 1996). Analysis of the expression patterns of different Bmps indicates that Bmp-2, Bmp-4 and Bmp-7 show remarkable co-distribution as well as apparent associations with epithe-

lialmesenchymal interactions (Fig. 4) (Aberg et al., 1997). Expression of these three Bmps shifts between epithelium and mesenchyme during advancing morphogenesis. They are also expressed in the enamel knot (see below) and in association with the differentiation of odontoblasts and ameloblasts (Begue-Kirn et al., 1992; Vainio et al., 1993; Aberg et al., 1997). Bmp-5 expression, on the other hand, is conned to the ameloblasts (Aberg et al., 1997). Hence, BMPs appear to be used repeatedly during tooth morphogenesis and they are apparently capable of signalling in both directions between the epithelium and mesenchyme. Furthermore, their largely overlapping expression patterns suggest that there may be functional redundancy between the co-expressed genes. This possibility was recently supported by a study where the expression of Bmp-7 was compared with other Bmps. In the tissues which were affected in the Bmp-7 knockouts, other Bmps were not expressed, whereas co-expression was evident in several tissues not affected in the knockouts (Dudley and Robertson, 1997). Several of the serine-threonine kinase receptors for BMPs are expressed during morphogenesis but the patterns do not appear to be as clearly spatiotemporally restricted s those of the ligands (Verschueren et al., 1995; Aberg et al., unpublished data). In addition to BMPs other members of the TGFb superfamily, notably activin, regulate tooth development. Targeted disruption of the activin bA gene results in abnormal tooth development (Matzuk et al., 1996). Follistatin and activin IIA receptor knockouts can also have abnormal tooth development but this is found at a much lesser frequency that in the activin bA mice, perhaps suggesting there is functional redundancy between receptors and antagonists (Matzuk et al., 1996). Tgfb-1, Tgfb-2 and Tgfb-3 are present in teeth and may be associated with morphogenesis (Vaahtokari et al., 1991; Chai et al., 1994). 5.2. FGF signalling pathway Several of the broblast growth factor (FGF) family members have been associated with tooth morphogenesis. Fgf-3 (int-2) was the rst to be localized (Wilkinson et al., 1989). Its expression is conned to dental papilla mesenchyme and it is downregulated with advancing morphogenesis (our unpublished data). Fgf-4, Fgf-8 and Fgf-9 on the other hand are expressed exclusively in dental epithelial cells (Heikinheimo et al., 1994; Jernvall et al., 1994; Kettunen and Thesleff, 1997). Fgf-8 is present only in the presumptive dental epithelium prior to budding and Fgf-4 is restricted to the enamel knots (Fig. 5) (see below). Fgf-9 expression appears to cover the domains of both Fgf-4 and Fgf-8 and in addition it is expressed widely in the dental epithelium during the bell stage, associated with the terminal differentiation of the odontoblasts and ameloblasts (Kettunen and Thesleff, 1997). Their respective tyrosine kinase receptors are present in both epithelial and mesenchymal tissues in the tooth (Peters et al., 1992; Kettunen et al., unpublished data).

116

I. Thesleff, P. Sharpe / Mechanisms of Development 67 (1997) 111123

Fig. 4. BMP signalling during early tooth morphogenesis. (A,B) Bmp-2 and Bmp-7 are expressed early in the thickened dental epithelium. Also Bmp-4 is expressed in the E11 epithelium (not shown). (C) A bead releasing BMP-2 protein has induced the expression of Msx-1 in cultured E11 dental mesenchyme. (D) Control mesenchyme with BSA bead. (Courtesy of Thomas Aberg and Carin Sahlberg.) Fig. 5. The enamel knot is a putative signalling center. (A) At the cap stage, cells in this epithelial structure have left the cell cycle (shown by lack of BrdU incorporation) and they express Shh and Fgf-4. (B,C). Three-dimensional reconstructions of the serial sections (lower panel) show the shape of the enamel

The FGFs also use cell surface heparan sulfate proteoglycans as co-receptors. The cell surface proteoglycan syndecan-1 was among the rst molecules that were shown to be associated with and regulated by epithelialmesenchymal signalling during early tooth morphogenesis (Thesleff et al., 1988; Vainio et al., 1989). It binds FGFs and moreover syndecan expression is stimulated by FGFs in dental mesenchyme (Chen et al., 1996). Hence, syndecan-1 may play a role as a modulator of growth factor signalling especially by FGFs. Syndecan was shown to be co-expressed with the matrix glycoprotein tenascin in the organ-specic mesenchyme in several organs including the tooth and since the two proteins were shown to interact, it was suggested that this interaction may regulate the condensation of mesenchymal cells (Salmivirta et al., 1991). However, tooth development occurs normally in tenascin decient mice (Saga et al., 1992) and so the exact roles of these molecules are still unknown. Msx-1 also appears to participate in the FGF signalling pathway. Several FGFs, including FGF-2, FGF-4, FGF-8 and FGF-9 upregulate Msx-1 expression in the dental mesenchyme when applied with heparin acrylic beads in vitro (Chen et al., 1996; Kettunen and Thesleff, 1997).

Unlike BMPs which induce expression of both Msx-1 and Msx-2 (Vainio et al., 1993), FGFs are specic to Msx-1 and have no stimulatory effects on mesenchymal expression of Msx-2 (Kettunen and Thesleff, 1997). Furthermore, the analysis of the tissues of Msx-1 decient mice has indicated that the stimulatory action of FGF on syndecan expression requires a functional Msx-1 gene (Chen et al., 1996). FGFs appear to be signals which regulate pattern as well as growth. In general FGFs are potent stimulators of cell proliferation and they also stimulate cell division both in dental mesenchyme and epithelium at several stages of tooth morphogenesis (Jernvall et al., 1994; Kettunen and Thesleff, 1997). FGFs also prevent apoptosis in the dental mesenchyme thus mimicking the effect of epithelium (Vaahtokari et al., 1996b). On the other hand, FGFs have potent morphogenetic effects in several organs. FGF-8 regulates midbrain development (Crossley et al., 1996a), FGF8 and FGF-4 are involved in limb development (Crossley et al., 1996b) and FGFs stimulate the development of kidney tubules and feather buds (Karavanova et al., 1996; Widelitz et al., 1996). In common with BMPs, the biological effects of various FGFs are largely similar in different in vitro

I. Thesleff, P. Sharpe / Mechanisms of Development 67 (1997) 111123

117

assays. FGF-2, FGF-4 and FGF-8 have similar effects on cell proliferation and Msx-1 gene expression in dental tissues (Kettunen and Thesleff, 1997). Hence, it is possible that there is functional redundancy between the coexpressed Fgfs. 5.3. Hedgehog signalling pathway Of the hedgehog family, sonic hedgehog (Shh) is expressed in dental epithelium at several stages starting in the early epithelial thickenings, then reappears in the enamel knot (Fig. 5) (see below) and subsequently it is expressed in the ameloblast cell lineage (Bitgood and McMahon, 1995; Vaahtokari et al., 1996a; Iseki et al., 1996). Its receptor Patched, which is a multipass membrane protein (Chen and Struhl, 1996) is expressed widely in the dental mesenchyme as well as in epithelium not expressing Shh (our unpublished data). Targeted disruption of Shh in transgenic mice results in an early craniofacial phenotype resembling holoprosencephaly (Chiang et al., 1996). Shh expression in the early dental epithelium appears to follow those of Fgf-8 and Fgf-9, suggesting that FGFs may be upstream of Shh signalling. The role of Shh signalling in the early epithelial thickenings is unclear but its highly localized expression suggests an involvement in tooth bud initiation. During lung morphogenesis Shh is expressed intensely in the epithelium of terminal buds and its role was recently analyzed by overexpression in the distal epithelium using the surfactant promoter. The phenotype of the transgenic lungs suggested that Shh normally regulates lung mesenchymal cell proliferation (Bellusci et al., 1997). 5.4. Wnt signalling pathway In Drosophila development interactions between wingless, dpp (Bmp-2/4) and hedgehog genes are essential for correct segmentation. Wingless protein binds to its cell surface receptor, fz2 (frizzled), and activates an intracellular signalling pathway involving the armadillo protein (b-catenin) which results in nuclear transport of a Lef-1 transcription factor homologue pangolin (Brunner et al., 1997). In mice, Lef-1 knockouts lack teeth and the development is arrested at the bud stage. In the Lef-1 decient mice, hair, mammary gland and tooth morphogenesis is inhibited (van Genderen et al., 1994). Overexpression of Lef-1 in epithelial cells in transgenic mice resulted in increased invagination of epithelium and formation of extra hair follicles and a toothlike structure (Zhou et al., 1995). Elegant tissue recombination experiments using tissues from Lef-1 decient and wild type mouse embryos showed that although Lef-1 is expressed throughout tooth development and its expression is not restricted to either epithelial or mesenchymal tissues, it is needed only in epithelium during early development (Kratochwil et al., 1996). Normal teeth developed in explants which consisted of exclusively mutant epithelium and mesenchyme, if the mesenchyme

had rst been exposed to wild type epithelium before the bud stage. Hence, Lef-1 appears to be involved in the regulation of an epithelial signal acting on dental mesenchyme during the bud stage of tooth morphogenesis. b-Catenin has a dual role in cells, interacting directly with Lef-1 in the cytoplasm to mediate transport to the nucleus and interacting with cadherins in cell adhesion processes (Behrens et al., 1996; Nusse, 1997). E-Cadherin protein has been localized to the early tooth germ epithelium in similar areas to Lef-1 (Lu ning et al., 1994) suggesting a role in regulating the cell adhesive properties needed for invagination to form a bud. Several Wnt genes are expressed during tooth development including the Wnt-10 genes which are expressed in early dental epithelium (A. McMahon, pers. commun.). These or other Wnts might therefore be responsible for activating the intracellular pathway involving frizzled receptors, b-catenin and nuclear transport of Lef-1 and thus the Wnt-signalling pathway may be required for tooth bud formation. 5.5. Other signalling pathways Other conserved signals that have been elucidated during tooth morphogenesis include epidermal growth factor (EGF). EGF receptor is expressed both in dental epithelium and mesenchyme and is associated and regulated by tissue interactions (Partanen and Thesleff, 1987). Surprisingly, no dental anomalies were reported in EGF receptor decient mice although they had major defects in development of several other epithelial appendages (Miettinen et al., 1995; Sibilia and Wagner, 1995). The Notch receptor as well as its ligand Serrate (Jagged) are also expressed during several stages of tooth morphogenesis and both in epithelial and mesenchymal tissues (Mitsiadis et al., 1995, 1997). Interestingly, Notch-1, Notch-2 and Notch-3 genes are downregulated in the dental epithelium already at the time of dental lamina formation. Throughout subsequent tooth morphogenesis the cells of the ameloblast cell lineage, i.e. inner enamel epithelial cells, do not express the Notch genes whereas all other dental epithelial cells express them intensely. In Drosophila, the expression of the Notch gene is associated with cell fate specication (Artavanis-Tsakonas et al., 1995) and hence it is possible that Notch genes function in the determination of the dental epithelial cells. Furthermore, Notch and Serrate genes are regulated by tissue interactions as well as by BMPs and FGFs in dental tissues, thus suggesting that signalling through this pathway is linked with other signalling networks (Mitsiadis et al., 1995, 1997). In Drosophila, signalling through the Notch receptor is associated with the Wg signalling pathway (Couso et al., 1995). In vitro experiments in which the expression of hepatocyte growth factor (HGF, scatter factor) was inhibited by antisense oligonucleotides suggest that this signal is needed for tooth morphogenesis. Hgf is expressed in dental mesenchyme and its receptor c-Met is expressed in epithe-

118

I. Thesleff, P. Sharpe / Mechanisms of Development 67 (1997) 111123

Fig. 6. Sequential and reciprocal epithelialmesenchymal signalling regulating tooth morphogenesis. The signalling tissues are framed and the suggested signal molecules are illustrated in the arrows. Transcription factors which have been associated with dental patterning and tissue interactions are depicted in the boxes with the signalling tissues. During initiation of tooth development, the tissue source and the nature of the signals are not known, hence the dashed frames around the tissues. The epithelial dental lamina signals prior to the bud stage and the dental mesenchyme during subsequent stages. Shape development is presumably regulated by signals from both the epithelial enamel knot and the mesenchymal dental papilla.

lium (Sonnenberg et al., 1993) and the prevention of Hgf expression inhibited the progression of morphogenesis from the cap to bell stage (Tabata et al., 1996). In similar antisense experiments the prevention of expression of the precursor of substance P, a tachykinin neurotransmitter, resulted in inhibition of tooth as well as mammary gland morphogenesis (Weil et al., 1995). Defective dental development was reported in transgenic mice decient for PDGF (platelet derived growth factor) receptor alpha subunit (Morrison-Graham et al., 1992). Interestingly, neurotrophins and neurotrophin receptors (both the low afnity receptor, LANR, and tyrosine kinase trk-receptors) are also expressed during tooth morphogenesis. Their expression is largely linked with the development of innervation, but some expression domains show no relation to nerve growth and appear to be associated with epithelial mesenchymal interactions (Luukko et al., 1996; Luukko et al., 1997).

6. The enamel knot as a signalling centre After the discovery by Spemann and Mangold (1924) of the organizing activity of the blastopore lip, a number of other organizing tissues have been identied in vertebrate embryos. These regulate the establishment of patterns of tissues and shapes of organs. For instance in the limbs, antero-posterior, proximo-distal and dorso-ventral patterns are regulated by an interplay of signals from three different signalling tissues, the apical ectodermal ridge (AER), the zone of polarizing activity (ZPA) and the dorsal ectoderm

(Tickle, 1995). It has been established that signals belonging to the hedgehog, Wnt, FGF and BMP families mediate the organizing effects of different signalling tissues (Niswander and Martin, 1992; Riddle et al., 1993; Francis et al., 1994; Parr and McMahon, 1995; Yang and Niswander, 1995; Crossley et al., 1996a,b). Recent observations suggest that tooth shape may be regulated by signalling or organizing centres present in the enamel knots (Jernvall et al., 1994; Vaahtokari et al., 1996a). The enamel knots are transient clusters of dental epithelial cells which were originally observed many decades ago in cap stage teeth (Ahrens, 1913; Butler, 1956). Their functions have, however, remained obscure and they have been largely neglected and sometimes even viewed as histological artefacts. The rst indication that the enamel knot may have a signalling function came from in situ hybridization studies in which Fgf-4 gene expression was localized exclusively in the enamel knot (Fig. 5) (Niswander and Martin, 1992; Jernvall et al., 1994). Subsequently, several other signals were shown to be expressed in the enamel knot. These include Bmp-2, Bmp-4, Bmp-7 and Shh (Fig. 5) (Vaahtokari et al., 1996a) as well as Fgf-9 (Kettunen and Thesleff, 1997). The appearance and shape of the enamel knot has been analyzed in detail in the rst mandibular molar tooth in mouse embryos by three-dimensional computer based reconstruction of serial sections (Fig. 5). Morphologically, the enamel knots are rst seen at the tips of the tooth buds just prior to their development into the cap stage. A few cells of the epithelium stop proliferating, as seen by the lack of BrdU incorporation. The enamel knot grows as the

I. Thesleff, P. Sharpe / Mechanisms of Development 67 (1997) 111123

119

bud transforms to a cap, at which stage it is readily visible in histological sections as a tightly packed epithelial cell aggregate. The enamel knot was shown to be a torpedoshaped structure, the development of which started in the mesial part of the tooth germ and progressed distally (Jernvall et al., 1994). The different signals show nested expression patterns and variation in the onset and termination of expression. Fgf-4 expression correlates rather well with the non-mitotic cells although the expression domain is a few cell layers more restricted and starts slightly later (Fig. 5) (Jernvall et al., 1994). Shh, Bmp-2 and Bmp-7 are already present in budding epithelium prior to the exit of the distal tip cells from the cell cycle. Bmp-4 on the other hand starts to be expressed during the advanced cap stage starting from the distal region of the knot (Jernvall et al., 1997). The assumption that the enamel knot represents a specic cell lineage with a unique fate is supported by expression of the cyclin dependent kinase inhibitor p21, which is associated with terminal differentiation of many cell types (Parker et al., 1995). In isolated dental epithelium, p21 expression is stimulated by BMP-4 protein, suggesting that BMP-4, which is intensely expressed in the dental mesenchyme during the bud stage, might be involved in the induction of enamel knot formation (Jernvall et al., 1997). It is noteworthy that tooth development is arrested in Msx-1 and Lef-1 knockout mice at the bud stage. The transition from the bud to cap stage involves the invagination of the undersurface of the epithelial bud and the formation of the dental papilla in the adjacent mesenchyme. Hence, the arrest in the development of the knockout teeth may be associated with the lack of the formation of the enamel knot (Jernvall et al., 1997). The induction of the enamel knot may be a key event in tooth morphogenesis and since both Lef-1 and Msx-1 are needed for its induction, it is possible that proper enamel knot formation and maintenance requires the co-ordination of several signalling pathways. An interesting feature of the enamel knot is that its cells undergo apoptosis (Vaahtokari et al., 1996b). The enamel knot starts to disappear during the cap stage, starting from its distal end and progressing in the opposite direction to which it was formed (Jernvall et al., 1997). Consequently, no enamel knot is seen at the late cap stage. The disappearance of enamel knot cells is produced by apoptosis and hence apoptosis may be a general mechanism whereby the function of signalling centres is terminated (Vaahtokari et al., 1996b). Cell death is apparent also in the AER (Coelho et al., 1993; Vaahtokari et al., 1996b). Also the ZPA and the notochord are largely removed with advancing development. The expression pattern of Bmp-4 in the enamel knot shows striking co-localization with apoptosis. Since BMP-4 regulates programmed cell death during embryonic development in the rhombomeres and in the interdigital mesenchyme (Graham et al., 1994; Zou and Niswander, 1996), it is possible that this signal may have a similar function in the enamel knot. Furthermore, the timing of apoptosis in the

enamel knot may provide an important mechanism in the regulation of tooth shape. Slight variations in the activity of the enamel knot could be involved in the evolution of mammalian tooth morphologies (Jernvall et al., 1997). It has been proposed that the enamel knot determines the site of the rst cusp of teeth and regulates the formation of other cusps in molar teeth. There is an enamel knot also in the incisors during the cap stage and in molars new enamel knots appear at the sites of new cusps (Jernvall et al., 1994). These secondary enamel knots also express Fgf-4 and their cells do not divide. In the limb bud, FGF-4 which is expressed in the AER stimulates the proliferation of underlying mesenchymal cells and can replace the effect of AER in the promotion of proximodistal growth (Niswander et al., 1993). Receptors for FGFs are present both in the mesenchyme and epithelium around the enamel knot (Kettunen et al., in preparation) and FGF-4 stimulates the proliferation of both cell types (Jernvall et al., 1994). FGF-4 as well as FGF-9 may regulate the growth of the cusps by stimulation of cell proliferation and the temporospatial pattern of the enamel knots may determine the sites and timing of cusp formation. Because of the way that the tooth develops, the cusp that starts to form rst will become the highest and the subsequently forming cusps will end up being progressively shallower. Hence, the timing of initiation of cusp formation determines the relative heights of cusps (Jernvall et al., 1994; Jernvall, 1995). The roles of the different signals in the enamel knot can at present only be speculated. It is possible that some molecules, perhaps BMP-4, act as autocrine signals within the enamel knot cells. Some may act as planar signals within the dental epithelium, whereas the underlying dental papilla mesenchyme is presumably the target of some signals. Patched, the Shh receptor is expressed in the dental papilla mesenchyme suggesting that the mesenchyme may be the target for the action of Shh (our unpublished data). In Drosophila, hedgehog acts as a morphogen and different concentrations specify distinct cell types (Heemskerk and DiNardo, 1994) and Dpp mediates some of its effects in the patterning of the wing and eye. Also in the vertebrate limb bud, BMP-2 mediates the action of Shh in the ZPA. In the enamel knot, the expression of Shh, Bmp-2, Bmp-7 as well as Fgf-9 appear to spread from the enamel knot along the dental epithelium during the late cap stage and they cover the coronal part of the dental epithelium during sub sequent stages (Aberg et al., 1997; Kettunen and Thesleff, 1997). Whether this spreading is associated with differentiation of the odontoblasts and/or ameloblasts or perhaps with development of cusp morphology remains to be determined.

7. Conclusions and outlook It is apparent that the morphological features of early tooth development as well as the molecular mechanisms regulating tooth morphogenesis bear more similarities

120

I. Thesleff, P. Sharpe / Mechanisms of Development 67 (1997) 111123

than differences with the development of other epithelial appendages (Thesleff et al., 1995). In fact, no developmental mechanisms or regulatory molecules have so far been shown to be unique for tooth development. Representatives of all signal families have been found in teeth (detailed description of gene expression changes during tooth development can be found in the WWW database gene expression in tooth, 1996) and in many of these cases there is likely to be conservations of signalling interactions with Drosophila developmental processes. Thus, although teeth are unique to vertebrates, much of their development involves genetic pathways found in invertebrates. Indeed it seems remarkable that the differentiation of cells unique to teeth, ameloblasts and odontoblasts involves homologues of Drosophila signalling molecules, a species that has no equivalent cells. It remains to be seen whether the specic features of tooth development reside in specic temporo-spatial actions of various sets of the generally used regulatory molecules, or whether there exist tooth-specic regulatory molecules. In Fig. 6 we have summarized some details of signalling networks which have been shown or suggested to regulate various steps of tooth morphogenesis. The rapid progress in identifying signalling pathways in tooth initiation and cusp formation means that we are starting to understand these processes. The early events that control dental patterning via specication of neural crest cells provide a breakthrough in piecing together the mechanisms determining positional specicity. What remains to be understood is the link between spatial determination, initiation and morphogenesis. The analysis of the signalling networks in the enamel knot and their downstream effects can be expected to increase the understanding of the regulation of shape. The enamel knot can be considered a signalling centre based on the accumulation of several signals specically in this cell cluster. Although there is so far no direct evidence for an organizing capacity of the enamel knot, the theory is supported by the fact that the same signals are expressed in well known organizing centres in the embryo. In the developing limb, the zone of polarizing activity (ZPA) which regulates anteroposterior patterning, expresses Shh (Riddle et al., 1993) and the apical ectodermal ridge (AER), which controls proximodistal growth in the limb and interacts with the ZPA, expresses Fgf-4 (Niswander and Martin, 1992). In addition, both ZPA and AER express the same Bmps as the enamel knot (Francis et al., 1994). The notochord, which regulates patterning of the neural tube and somites, also expresses Shh and the ectoderm overlying the neural tube, which participates in this patterning, expresses Bmp-4 and Bmp-7 (Echelard et al., 1993; Fan and Tessier-Lavigne, 1994; Liem et al., 1995). In addition, Shh expression as well as Bmps have been localized to budding epithelium during hair and feather development and in the lungs (Nohno et al., 1995; Tingberreth and Chuong, 1996; Bellusci et al., 1997) suggesting conserved actions in budding and branching morphogenesis in various organs. However,

no morphologically distinct epithelial structures resembling the enamel knot of teeth have been described in other epithelial appendages. As in all areas of developmental biology many of the key breakthroughs in the next few years will come from transgenic mice. Knockouts in transcription factor genes have been most revealing whereas those in signalling molecules themselves have been less informative largely as a result of early embryonic lethality. Progress in direct understanding of the roles of signalling molecules using loss-of-function approaches will require more subtle techniques such as conditional inactivation. Ectopic expression of signalling molecules is likely to be fruitful but will require a detailed knowledge of ligand/receptor specicities and localization of receptors to enable interpretation of results. Both these approaches will require the availability of specic promoter sequences to target expression. The localized expression of genes in mesenchyme and epithelium prior to overt tooth development needs to be understood. How are these genes regulated and what are the early interactions between mesenchyme and epithelium? How is the highly restricted expression of genes such as Shh, Fgf-8 and Bmp-4 in the epithelium controlled? Such investigations will certainly have to involve laborious analysis of gene regulatory sequences using LacZ reporters in transgenic mice. Teeth offer some particular advantages as compared to many other organs for the studies on developmental regulation. These are due to the specic patterns and shapes as well as to the unique mineral structure of the teeth. The dental patterns as well as tooth shapes vary characteristically between animals and, hence, comparative studies should shed light on the genetic basis of pattern formation, positional information and the generation of form. In some animals, e.g. in the mouse, vestigial teeth are present in which development is interrupted and these have already been used as models to examine the molecular basis of tooth morphogenesis (Tureckova et al., 1995). A particular eld where teeth can be expected to be powerful tools is evolutionary developmental biology (Butler, 1995; Jernvall, 1995). Teeth are the only organs that have been well preserved in extinct animals and the evolution of mammals has been best documented in their teeth. Hence, successful combination of fossil data with the progressive understanding of the roles of different genes in regulation of tooth development should increase the knowledge on evolution of patterns and morphologies. The possibility that patterning of individual teeth involves independent genetic processes raises some intriguing evolutionary questions. On one hand it makes evolutionary sense to have independent genetic control of tooth shape/position to allow changes in individual teeth for adaptations in specialized feeding. On the other hand, however, since molar teeth must occlude in order to function it might be expected that the same genetic pathways would specify upper and lower molars. Identication of genes involved in patterning

I. Thesleff, P. Sharpe / Mechanisms of Development 67 (1997) 111123

121

lower molars will no doubt help elucidate this apparent contradiction. The shapes of tooth crowns reect the nature of food the animal eats. In particular, the variation in the sizes and locations of the cusps and lophs in the premolars and molars during evolution has been correlated to ecological changes (Hunter and Jernvall, 1995; Jernvall et al., 1996). There is circumstantial evidence for the role of the enamel knot in the regulation of cusp patterns of teeth and it will be interesting to see whether the extent of the enamel knot and the timing of its activity are correlated with cusp patterns in different mammalian species. Comparative studies in which the regulation of tooth development is analyzed in well selected extant species should offer the possibility to shed light on the genetic basis of evolutionary changes.

Acknowledgements Our own investigations included in this review have been supported by the Human Frontier Science Program grant RG558/95.

References
Aberg, T., Wozney, J. and Thesleff, I. (1997) Expression patterns of bone morphogenetic proteins (Bmps) in the developing mouse tooth suggest roles in morphogenesis and cell differentiation. Dev. Dyn., in press. Ahrens, K. (1913) Die Entwicklung der menschlichen Zahne. Arb. Anat. Inst. Wiesbaden 48, 169266. Artavanis-Tsakonas, S., Matsuno, K. and Fortini, M.E. (1995) Notch signalling. Science 268, 225232. Behrens, J., von Kries, M., Kuhl, M., Bruhn, D., Wedlich, R., Grosschedl, R. and Birchmeier, W. (1996) Functional interaction of b-catenin with the transcription factor LEF-1. Nature 382, 638642. Begue-Kirn, C., Smith, A.J., Ruch, J.V., Wozney, J.M., Purchio, A., Hartmann, D. and Lesot, H. (1992) Effects of dentin proteins, transforming growth factor beta 1 (TGF beta 1) and bone morphogenetic protein 2 (BMP2) on the differentiation of odontoblast in vitro. Int. J. Dev. Biol. 36, 491503. Bellusci, S., Furuta, Y., Rush, M.G., Henderson, R., Winnier, G. and Hogan, B.L.M. (1997) Involvement of Sonic Hedgehog (Shh) in mouse embryonic lung growth morphogenesis. Development 124, 53 63. Bitgood, M.J. and McMahon, A.P. (1995) Hedgehog and Bmp genes are coexpressed at many diverse sites of cellcell interaction in the mouse embryo. Dev. Biol. 172, 126138. Brunner, E., Peter, O., Schweizer, L. and Basler, K. (1997) Pangolin encodes a Lef-1 homologue that acts downstream of Armadillo to transduce the Wingless signal in Drosophila. Nature 385, 829833. Butler, P.M. (1939) Studies of the mammalian dentition. Differentiation of the postcanine dentition. Proc. Zool. Soc. London 109B, 329356. Butler, P.M. (1956) The ontogeny of molar pattern. Biol. Rev. 31, 3070. Butler, P.M. (1995) Ontogenetic aspects of dental evolution. Int. J. Dev. Biol. 39, 2534. Chai, Y., Mah, A., Crohin, C., Groff, S., Bringas, P., Jr., Le, T., Santos, V. and Slavkin, H.C. (1994) Specic transforming growth factor-beta subtypes regulate embryonic mouse Meckels cartilage and tooth development. Dev. Biol. 162, 85103. Chen, Y. and Struhl, G. (1996) Dual roles for patched in sequestering and transducing Hedgehog. Cell 87, 553563.

Chen, Y., Bei, M., Woo, I., Satokata, I. and Maas, R. (1996) Msx1 controls inductive signalling in mammalian tooth morphogenesis. Development 122, 30353044. Chiang, C., Litingtung, Y., Lee, E., Young, K.E., Corden, J.L., Westphal, H. and Beachy, B.A. (1996) Cyclopia and defective axial patterning in mice lacking Sonic hedgehog gene function. Nature 383, 407413. Coelho, C.N., Upholt, W.B. and Kosher, R.A. (1993) The expression pattern of the chicken homeobox-containing gene GHox-7 in developing polydactylous limb buds suggests its involvement in apical ectodermal ridge-directed outgrowth of limb mesoderm and in programmed cell death. Differentiation 52, 129137. Couso, J.P., Knust, E. and Martinez Arias, A. (1995) Serrate and wingless cooperate to induce vestigial gene expression and wing formation in Drosophila. Curr. Biol. 5, 14371448. Crossley, P.H., Martinez, S. and Martin, G.R. (1996a) Midbrain development induced by FGF8 in the chick embryo. Nature 380, 6668. Crossley, P.H., Minowada, G., MacArthur, C.A. and Martin, G.R. (1996b) Roles for FGF8 in the induction, initiation, and maintenance of chick limb development. Cell 84, 127136. Dudley, A.T. and Robertson, E.J. (1997) Overlapping expression domains of bone morphogenetic protein family members potentially account for limited tissue defects in BMP7 decient embryos. Dev. Dyn. 208, 349 362. Echelard, Y., Epstein, D.J., St-Jacques, B., Shen, L., Mohler, J., McMahon, J.A. and McMahon, A.P. (1993) Sonic hedgehog, a member of a family of putative signalling molecules, is implicated in the regulation of CNS polarity. Cell 75, 14171430. Fan, C.M. and Tessier-Lavigne, M. (1994) Patterning of mammalian somites by surface ectoderm and notochord: evidence for sclerotome induction by a hedgehog homolog. Cell 79, 11751186. Francis, P.H., Richardson, M.K., Brickell, P.M. and Tickle, C. (1994) Bone morphogenetic proteins and a signalling pathway that controls patterning in the developing chick limb. Development 120, 209218. Gene expression in tooth (WWW database) (1996) http://honeybee.helsinki./toothexp. Developmental Biology Programme of the University of Helsinki. Glasstone, S. (1963) Regulative changes in tooth germs grown in tissue culture. J. Dent. Res. 42, 13641368. Graham, A., Francis-West, P., Brickell, P. and Lumsden, A. (1994) The signalling molecule BMP4 mediates apoptosis in the rhombencephalic neural crest. Nature 372, 684686. Heemskerk, J. and DiNardo, S. (1994) Drosophila hedgehog acts as a morphogen in cellular patterning. Cell 76, 449460. Heikinheimo, M., Lawshe, A., Shackleford, G.M., Wilson, D.B. and MacArthur, C.A. (1994) Fgf-8 expression in the post-gastrulation mouse suggests roles in the development of the face, limbs and central nervous system. Mech. Dev. 48, 129138. Hogan, B.L. (1996) Bone morphogenetic proteins in development. Curr. Opin. Genet. Dev. 6, 432438. Hunter, J.P. and Jernvall, J. (1995) The hypocone as a key innovation in mammalian evolution. Proc. Natl. Acad. Sci. USA 92, 1071810722. Imai, H., Osumi-Yamashita, N., Ninomiya, Y. and Eto, K. (1996) Contribution of early-emigrating midbrain crest cells to the dental mesenchyme of mandibular molar teeth in rat embryos. Dev. Biol. 176, 151165. Iseki, S., Araga, A., Ohuchi, H., Nohno, T., Yoshioka, H., Hayashi, F. and Noji, S. (1996) Sonic hedgehog is expressed in epithelial cells during development of whisker, hair, and tooth. Biochem. Biophys. Res. Commun. 218, 688693. Jernvall, J. (1995) Mammalian molar cusp patterns: developmental mechanisms of diversity. Acta Zool. Fennica 198, 161. Jernvall, J., Kettunen, P., Karavanova, I., Martin, L.B. and Thesleff, I. (1994) Evidence for the role of the enamel knot as a control center in mammalian tooth cusp formation: non-dividing cells express growth stimulating Fgf-4 gene. Int. J. Dev. Biol. 38, 463469. Jernvall, J., Hunter, J.P. and Fortelius, M. (1996) Molar tooth diversity, disparity, and ecology in Cenozoic ungulate radiations. Science 274, 14891492.

122

I. Thesleff, P. Sharpe / Mechanisms of Development 67 (1997) 111123 lium precedes tooth initiation and requires mesenchyme-derived signals for its maintenance. Dev. Biol., 189, in press. Niswander, L. and Martin, G.R. (1992) Fgf-4 expression during gastrulation, myogenesis, limb and tooth development in the mouse. Development 114, 755768. Niswander, L., Ickle, C., Vogel, A., Booth, I. and Martin, G.R. (1993) FGF-4 replaces the apical ectodermal ridge and directs outgrowth and patterning of the limb. Cell 75, 579587. Nohno, T., Ishikawa, T., Saito, T., Hosokawa, K., Noji, S., Wolsing, D.H. and Rosenbaum, J.S. (1995) Identication of a human type II receptor for bone morphogenetic protein-4 that forms differential heteromeric complexes with bone morphogenetic protein type I receptors. J. Biol. Chem. 270, 2252222526. Nusse, R. (1997) A versatile transcriptional effector of Wingless signaling. Cell 89, 321323. Osborn, J.W. (1978) Morphogenetic gradients: elds versus clones. In Butler, P.M. and Joysey, K.A. (eds.), Development, Function and Evolution of Teeth, Academic Press, New York, pp. 171201. Osumi-Yamashita, N., Ninomiya, Y., Doi, H. and Eto, K. (1994) The contribution of both forebrain and midbrain crest cells to the mesenchyme in the frontonasal mass of mouse embryos. Dev. Biol. 164, 409 419. Parker, S.B., Eichele, G., Zhang, P., Rawls, A., Sands, A.T., Bradley, A., Olson, E.N., Harper, J.W. and Elledge, S.J. (1995) p53-independent expression of p21Cip1 in muscle and other terminally differentiating cells. Science 267, 10241027. Parr, B.A. and McMahon, A.P. (1995) Dorsalizing signal Wnt-7a required for normal polarity of D-V and A-P axes of mouse limb. Nature 374, 350353. Partanen, A.M. and Thesleff, I. (1987) Localization and quantitation of 125 I-epidermal growth factor binding in mouse embryonic tooth and other embryonic tissues at different developmental stages. Dev. Biol. 120, 186197. Peters, K.G., Werner, S., Chen, G. and Williams, L.T. (1992) Two FGF receptor genes are differentially expressed in epithelial and mesenchymal tissues during limb formation and organogenesis in the mouse. Development 114, 233243. Qiu, M., Bulfone, A., Ghattas, I., Meneses, J.J., Christensen, L., Sharpe, P.T., Presley, R., Pedersen, R.A. and Rubenstein, J.L.R. (1997) Role of Dlx-1 and -2 in proximodistal patterning of the branchial arches: mutations alter morphogenesis of proximal skeletal elements derived from the rst and second arches. Dev. Biol., 185, 165184. Riddle, R.D., Johnson, R.L., Laufer, E. and Tabin, C. (1993) Sonic hedgehog mediates the polarizing activity of the ZPA. Cell 75, 14011416. Ruch, J.V., Lesot, H. and Begue-Kirn, C. (1995) Odontoblast differentiation. Int. J. Dev. Biol. 39, 5168. Saga, Y., Yagi, T., Ikawa, Y., Sakakura, T. and Aizawa, S. (1992) Mice develop normally without tenascin. Genes Dev. 6, 18211831. Salmivirta, M., Elenius, K., Vainio, S., Hofer, U., Chiquet-Ehrismann, R., Thesleff, I. and Jalkanen, M. (1991) Syndecan from embryonic tooth mesenchyme binds tenascin. J. Biol. Chem. 266, 77337739. Satokata, I. and Maas, R. (1994) Msx1 decient mice exhibit cleft palate and abnormalities of craniofacial and tooth development. Nat. Genet. 6, 348356. Sellman, S. (1946) Some experiments on the determination of the larval teeth in Ambystoma mexicanum. Odontol. Tidsk. 54, 1128. Semina, E.V., Reiter, R., Leysens, N.J., Alward, W.L., Small, K.W., Datson, N.A., Siegel-Bartelt, J., Bierke-Nelson, D., Bitoun, P., Zabel, B.U., Carey, J.C. and Murray, J.C. (1996) Cloning and characterization of a novel bicoid-related homeobox transcription factor gene, RIEG, involved in Rieger syndrome. Nat. Genet. 14, 392399. Sengel, P. (1976) Morphogenesis of Skin (Developmental and Cell Biology Series), Cambridge University Press, Cambridge, MA, pp. 277. Sharpe, P.T. (1995) Homeobox genes and orofacial development. Connect. Tissue Res. 32, 1725. Sibilia, M. and Wagner, E.F. (1995) Strain-dependent epithelial defects in mice lacking the EGF receptor. Science 269, 234238.

Jernvall, J., Aberg, T., Kettunen, P., Keranen, S. and Thesleff, I. (1997) The life history of an embryonic signalling center: BMP-4 induces p21 and is associated with apoptosis in the mouse tooth enamel knot. Development, in press. Karavanova, I.D., Dove, L.F., Resau, J.H. and Perantoni, A.O. (1996) Conditioned medium from a rat ureteric bud cell line in combination with bFGF induces complete differentiation of isolated metanephric mesenchyme. Development 122, 41594167. Kettunen, P. and Thesleff, I. (1997) Expression and function of FGF-4, -8 and -9 suggest participation in signalling networks regulating morphogenesis and cell differentiation in the developing tooth. Submitted. Kollar, E.J. and Baird, G.R. (1969) The inuence of the dental papilla on the development of tooth shape in embryonic mouse tooth germs. J. Embryol. Exp. Morph. 21, 131148. Kollar, E.J. and Baird, G.R. (1970) Tissue interactions in embryonic mouse tooth germs. II. The inductive role of the dental papilla. J. Embryol. Exp. Morph. 24, 173186. Kratochwil, K., Dull, M., Farinas, I., Galceran, J. and Grosschedl, R. (1996) Lef1 expression is activated by BMP-4 and regulates inductive tissue interactions in tooth and hair development. Genes Dev. 10, 1382 1394. Kontges, G. and Lumsden, A. (1996) Rhombencephalic neural crest seg mentation is preserved throughout craniofacial ontogeny. Development 122, 32293242. Liem, K.F., Jr., Tremml, G., Roelink, H. and Jessell, T.M. (1995) Dorsal differentiation of neural plate cells induced by BMP-mediated signals from epidermal ectoderm. Cell 82, 969979. Lumsden, A.G. (1988) Spatial organization of the epithelium and the role of neural crest cells in the initiation of the mammalian tooth germ. Development 103 (Suppl.), 155169. Lu ning, C., Rass, A., Rozell, B., Wroblewski, J. and Obrink, B. (1994) Expression of E-cadherin during craniofacial development. J. Craniofac. Genet. Dev. Biol. 14, 207216. Luukko, K., Moshnyakov, M., Sainio, K., Saarma, M., Sariola, H. and Thesleff, I. (1996) Expression of neurotrophin receptors during rat tooth development is developmentally regulated, independent of innervation, and suggests functions in the regulation of morphogenesis and innervation. Dev. Dyn. 206, 8799. Luukko, K., Arumae, U., Karavanov, A., Moshnyakov, M., Sainio, K., Sariola, H., Saarma, M. and Thesleff, I. (1997) Neurotrophin mRNA expression in the developing tooth suggests multiple roles in innervation and organogenesis. Dev. Dyn. 210, in press. Matzuk, M.M., Kumar, T.R., Shou, W., Coerver, K.A., Lau, A.L., Behringer, R.R. and Finegold, M.J. (1996) Transgenic models to study the roles of inhibins and activins in reproduction, oncogenesis, and development. Recent Prog. Horm. Res. 51, 123154. Miettinen, P.J., Berger, J.E., Meneses, J., Phung, Y., Pedersen, R.A., Werb, Z. and Derynck, R. (1995) Epithelial immaturity and multiorgan failure in mice lacking epidermal growth factor receptor. Nature 376, 337 341. Mina, M. and Kollar, E.J. (1987) The induction of odontogenesis in nondental mesenchyme combined with early murine mandibular arch epithelium. Arch. Oral. Biol. 32, 123127. Mitsiadis, T., Lardelli, M., Lendahl, U. and Thesleff, I. (1995) Expression of Notch 1, 2 and 3 is regulated by epithelialmesenchymal interactions and retinoic acid in the developing mouse tooth and associated with determination of ameloblast cell fate. J. Cell Biol. 130, 407418. Mitsiadis, T.A., Henrique, D., Ish-Horowicz, D., Thesleff, I. and Lendahl, U. (1997) Mouse Serrate1 (Jagged-1): expression in the developing tooth is regulated by epithelialmesenchymal interactions and broblast growth factor-4. Development 124, 14731483. Morrison-Graham, K., Schatteman, G.C., Bork, T., Bowen-Pope, D.F. and Weston, J.A. (1992) A PDGF receptor mutation in the mouse (Patch) perturbs the development of a non-neuronal subset of neural crestderived cells. Development 115, 133142. Mucchielli, M., Mitsiadis, T.A., Raffo, S., Brunet, J., Proust, J. and Goridis, C. (1997) Mouse Otlx2/RIEG expression in the odontogenic epithe-

I. Thesleff, P. Sharpe / Mechanisms of Development 67 (1997) 111123 Smith, M.M. and Hall, B.K. (1993) A developmental model for evolution of the vertebrate exoskeleton and teeth: the role of cranial and trunk neural crest. Evol. Biol. 27, 387448. Smith, M.P., Sansom, I.J. and Repetski, J.E. (1996) Histology of the rst sh. Nature 380, 702704. Sonnenberg, E., Weidner, K. and Birchmeier, C. (1993) Expression of the met-receptor and its ligand, HGF-SF during mouse embryogenesis. In Goldberg, I.D., and Rosen, E.M. (eds.), Hepatocyte Growth Factor-Scatter Factor (HGF-SF) and the C-met Receptor, Birkhauser Verlag, Basel, pp. 382394. Spemann, H. and Mangold, H. (1924) Uber Induktion von Embyonanlagen durch Implantation artfremder Organisatoren. Roux Arch f. Entw. Mech. 100, 599638. Tabata, M.J., Kim, K., Liu, J.G., Yamashita, K., Matsumura, T., Kato, J., Iwamoto, M., Wakisaka, S., Matsumoto, K., Nakamura, T., Kumegawa, M. and Kurisu, K. (1996) Hepatocyte growth factor is involved in the morphogenesis of tooth germ in murine molars. Development 122, 12431251. Ten Cate, A.R. (1994) Oral Histology. Development, Structure and Function, C.V. Mosby, St. Louis, MO. Thesleff, I. and Aberg, T. (1997) Tooth morphogenesis and the differentiation of ameloblasts. In Dental Enamel (Ciba Foundation Symposium 205), Wiley, Chichester, pp. 317. Thesleff, I. and Nieminen, P. (1996) Tooth morphogenesis and cell differentiation. Curr. Opin. Cell. Biol. 8, 844850. Thesleff, I. and Sahlberg, C. (1996) Growth factors as inductive signals regulating tooth morphogenesis. Sem. Cell Dev. Biol. 7, 185 193. Thesleff, I., Jalkanen, M., Vainio, S. and Berneld, M. (1988) Cell surface proteoglycan expression correlates with epithelialmesenchymal interaction during tooth morphogenesis. Dev. Biol. 129, 565572. Thesleff, I., Vaahtokari, A. and Partanen, A.M. (1995) Regulation of organogenesis. Common molecular mechanisms regulating the development of teeth and other organs. Int. J. Dev. Biol. 39, 3550. Tickle, C. (1995) Vertebrate limb development. Curr. Opin. Genet. Dev. 5, 478484. Tingberreth, S.A. and Chuong, C.M. (1996) Sonic hedgehog in feather morphogenesis induction of mesenchymal condensation and association with cell death. Dev. Dyn. 207, 157170. Tissier-Seta, J.P., Mucchielli, M.L., Mark, M., Mattei, M.G., Goridis, C. and Brunet, J.F. (1995) Barx1, a new mouse homeodomain transcription factor expressed in cranio-facial ectomesenchyme and the stomach. Mech. Dev. 51, 315. Trainor, P.A. and Tam, P.P. (1995) Cranial paraxial mesoderm and neural crest cells of the mouse embryo: co-distribution in the craniofacial mesenchyme but distinct segregation in branchial arches. Development 121, 25692582. Tureckova, J., Sahlberg, C., Aberg, T., Ruch, J.V., Thesleff, I. and Peterkova, R. (1995) Comparison of expression of the msx-1, msx-2, bmp-2 and bmp-4 genes in the mouse upper diastemal and molar tooth primordia. Int. J. Dev. Biol. 39, 459468. Vaahtokari, A., Vainio, S. and Thesleff, I. (1991) Associations between transforming growth factor beta 1 RNA expression and epithelial

123

mesenchymal interactions during tooth morphogenesis. Development 113, 985994. Vaahtokari, A., Aberg, T., Jernvall, J., Keranen, S. and Thesleff, I. (1996a) The enamel knot as a signalling center in the developing mouse tooth. Mech. Dev. 54, 3943. Vaahtokari, A., Aberg, T. and Thesleff, I. (1996b) Apoptosis in the developing tooth association with an embryonic signalling center and suppression by EGF and FGF-4. Development 122, 121129. Vainio, S., Jalkanen, M. and Thesleff, I. (1989) Syndecan and tenascin expression is induced by epithelialmesenchymal interactions in embryonic tooth mesenchyme. J. Cell Biol. 108, 19451953. Vainio, S., Karavanova, I., Jowett, A. and Thesleff, I. (1993) Identication of BMP-4 as a signal mediating secondary induction between epithelial and mesenchymal tissues during early tooth development. Cell 75, 45 58. van Genderen, C., Okamura, R.M., Farinas, I., Quo, R.G., Parslow, T.G., Bruhn, L. and Grosschedl, R. (1994) Development of several organs that require inductive epithelialmesenchymal interactions is impaired in LEF-1-decient mice. Genes Dev. 8, 26912703. Vastardis, H., Karimbux, N., Guthua, S.W., Seidman, J.G. and Seidman, C.E. (1996) A human MSX1 homeodomain missense mutation causes selective tooth agenesis. Nat. Genet. 13, 417421. Verschueren, K., Dewulf, N., Goumans, M.J., Lonnoy, O., Feijen, A., Grimsby, S., Vandi Spiegle, K., ten Dijke, P., Moren, A., Vanscheeuwijck, P. et al. (1995) Expression of type I and type IB receptors for activin in midgestation mouse embryos suggests distinct functions in organogenesis. Mech. Dev. 52, 109123. Wagner, G. (1959) Untersuchungen an Bombinator-Triton-Chimaeren. Das Skelett Larvaler Triton-Kophe mit Bombinator-Mesektoderm. Wil helm Roux Arch. Entw. Mech. Org. 151, 136158. Weil, M., Itin, A. and Keshet, E. (1995) A role for mesenchyme-derived tachykinins in tooth and mammary gland morphogenesis. Development 121, 24192428. Widelitz, R.B., Jiang, T.X., Noveen, A., Chen, C.W. and Chuong, C.M. (1996) FGF induces new feather buds from developing avian skin. J. Invest. Dermatol. 107, 797803. Wilkinson, D.G., Bhatt, S. and McMahon, A.P. (1989) Expression pattern of the FGF-related proto-oncogene int-2 suggests multiple roles in fetal development. Development 105, 131136. Winnier, G., Blessing, M., Labosky, P.A. and Hogan, B.L. (1995) Bone morphogenetic protein-4 is required for mesoderm formation and patterning in the mouse. Genes Dev. 9, 21052116. Yang, Y. and Niswander, L. (1995) Interaction between the signalling molecules WNT7a and SHH during vertebrate limb development: dorsal signals regulate anteroposterior patterning. Cell 80, 939947. Zhou, P., Byrne, C., Jacobs, J. and Fuchs, E. (1995) Lymphoid enhancer factor 1 directs hair follicle patterning and epithelial cell fate. Genes Dev. 9, 700713. Zou, H. and Niswander, L. (1996) Requirement for BMP signalling in interdigital apoptosis and scale formation. Science 272, 738741.L. (1, P.A. and Hogan, B.L. (1995) Bone morphogenetic protein-4 is required for mesoderm formation and patterning in the mouse. Genes Dev. 9, 21052116..