DEVELOPMENTAL DYNAMICS 215:45–53 (1999) Msx1 Is Required for the Induction of Patched by Sonic Hedgehog in the Mammalian Tooth Germ YANDING ZHANG,1 XIANG ZHAO,1 YUEPING HU,1 TARA ST. AMAND,1 MEIFENG ZHANG,1 RAJEE RAMAMURTHY,1 MENGSHENG QIU,2 AND YIPING CHEN1* 1Department of Cell and Molecular Biology, Tulane University, New Orleans, Louisiana 2Department of Anatomical Sciences and Neurobiology, University of Louisville College of Medicine, Louisville, Kentucky ABSTRACT We have used the mouse developing tooth germ as a model system to explore the transmission of Sonic hedgehog (Shh) signal in the induction of Patched (Ptc). In the early developing molar tooth germ, Shh is expressed in the dental epithelium, and the transcripts of Shh downstream target genes Ptc and Gli1 are expressed in dental epithelium as well as adjacent mesenchymal tissue. The homeobox gene Msx1 is also expressed in the dental mesenchyme of the molar tooth germ at this time. We show here that the expression of Ptc, but not Gli1, was downregulated in the dental mesenchyme of Msx1 mutants. In wild-type E11.0 molar tooth mesenchyme SHHsoaked beads induced the expression of Ptc and Gli1. However, in Msx1 mutant dental mesenchyme SHH-soaked beads were able to induce Gli1 but failed to induce Ptc expression, indicating a requirement for Msx1 in the induction of Ptc by SHH. Moreover, we show that another signaling molecule, BMP4, was able to induce Ptc expression in wild-type dental mesenchyme, but induced a distinct expression pattern of Ptc in the Msx1 mutant molar mesenchyme. We conclude that in the context of the tooth germ Msx1 is a component of the Shh signaling pathway that leads to Ptc induction. Our results also suggest that the precise pattern of Ptc expression in the prospective tooth-forming region is controlled and coordinated by at least two inductive signaling pathways. Dev Dyn 1999;215:45–53. r 1999 Wiley-Liss, Inc. Key words: tooth development; Sonic hedgehog; Msx1; Ptc; Gli1; Bmp4; signaling pathway INTRODUCTION Vertebrate organs form through sequential and reciprocal interactions between two different tissue layers, most commonly an epithelium and a mesenchyme. These processes involve a series of inductive and permissive tissue interactions (also known as secondary induction), which govern tissue-specific gene expression and morphogenesis, and eventually lead to terminal cell differentiation and organ patterning. Similar to many other embryonic organs, murine tooth development r 1999 WILEY-LISS, INC. relies largely on such tissue interactions (reviewed in Thesleff and Nieminen, 1996). Therefore, the murine developing molar tooth has been employed as one of the classical model systems for studying early inductive interactions and their underlying molecular basis. The first sign of mouse molar tooth morphogenesis occurs at embryonic day 11.5 (E11.5) as a local thickening of dental epithelium (lamina stage), which subsequently invaginates into the subjacent mesenchyme to form budlike structures (bud stage: E12.5–E13.5). This process is accompanied by the condensation of mesenchymal cells around the epithelial bud. The tooth bud then progresses to the cap (E14.5) and bell stages (E16.5). Eventually, the epithelial cells differentiate into enamelsecreting ameloblasts and the mesenchymal cells into dentin-secreting odontoblasts, pulp, and alveolar bone (Palmer and Lumsden, 1987). Rapidly accumulating evidence has indicated that peptide growth factors, such as bone morphogenetic proteins (BMPs) and fibroblast growth factors (FGFs), act as morphogenetic signals mediating the inductive interactions during tooth development (Thesleff and Sahlberg, 1996). The function of growth factors as inductive signaling molecules in epithelial–mesenchymal interactions has been demonstrated by their ability to substitute for one tissue in the induction of gene expression and morphogenetic change in an adjacent tissue (Vainio et al., 1993; Chen et al., 1996a; Neubüser et al., 1997). For example, BMP4 can substitute for dental epithelium to induce the expression of a number of genes, including homeobox genes Msx1 and Msx2, an HMG box gene Lef1 and Bmp4 itself, in dental mesenchyme (Vainio et al., 1993; Chen et al., 1996a; Kratochwil et al., 1996). The expression of growth factors in different developmental phases and different tissue layers may form signaling loops to achieve sequential and reciprocal interactions (reviewed in Chen and Grant sponsor: National Science Foundation; Grant number: IBN9796321; Grant sponsor: Tulane University (Y.P.C.); Grant sponsor: University of Louisville/NIH Human Genetics Enhancement Program; Grant sponsor: Kentucky Spinal Cord and Head Injury Trust; Grant sponsor: NSF (IBN-9808126) (M.S.Q.). *Correspondence to: Dr. YiPing Chen, Department of Cell and Molecular Biology, Tulane University, New Orleans, LA 70118. E-mail: email@example.com Received 30 November 1998; Accepted: 5 February 1999 46 ZHANG ET AL. Fig. 1. (A) Front view of the mouse face. (B) Oral view of a mandible at E11. The region (mandible) used for this study is outlined with dashed lines in Figure A. fb, forebrain; I, incisor; M, molar; mb, midbrain; md, mandible; mx, maxillary arch; np, nasal pit; T, tongue. Fig. 2. Expression of Shh , Ptc, and Msx1 in the early mouse developing tooth germ. Whole-mount in situ hybridization showed the expression of Shh (A, D, J), Ptc (B, E, K), and Msx1 (C, F, L) in the incisor (white arrows in panels A and B) and molar (black arrows in panels D, E, F, J, K, and L) teeth from E11.0 to E12.5 (oral view). Section in situ hybridization indicated that Shh transcripts are localized to the dental epithelium (G) at E11.5 (sagittal section), while Ptc and Msx1 transcripts are found in dental mesenchyme (H, I) at E11.5 (cross section). Note the weak expression of Ptc in the dental epithelium (panel H). de, dental epithelium; dm, dental mesenchyme; m, molar primordium; ma, mandibular arch; mx, maxillary arch; T, tongue. Scale bar: A–F, J–L ⫽ 500 µm; G–H ⫽ 50 µm; I ⫽ 25 µm. Msx1 IS A COMPONENT OF Shh SIGNALING PATHWAY 47 Fig. 3. Expression of Ptc but not Gli1 is altered in the Msx1 mutant dental mesenchyme. (A–F) Oral view of mandibles. (G–J) Cross section through molar teeth. Marked reduction of Ptc expression was observed in the Msx1 mutant dental mesenchyme from E11.5 to 13.5 (panels A, D, and G). Insert in panel A shows that Ptc expression at E11.5 is preserved in epithelium but is downregulated in mesenchyme of the Msx1 mutant molar tooth. Gli1 expression was preserved in the Msx1 mutant tooth germ from E11.5 to 13.5 (panels C, F, and J), as compared to the wild-type Gli1 expression at the same stages (panels B, E, and I). Panel H shows expression of Ptc in an E13.5 wild-type molar tooth germ. e, epithelial bud. Scale bar: A–F ⫽ 500 µm; G–J ⫽ 50 µm. Maas, 1998). It has been suggested that homeobox containing genes participate in epithelial–mesenchymal interactions by regulating the expression of inductive signaling molecules (Chen and Maas, 1998). This is exemplified by the fact that Msx1 controls Bmp4 expression in the dental mesenchyme (Chen et al., 1996a). Mice lacking Msx1 exhibit an arrest of tooth development at the bud stage, which is accompanied by the downregulation of several genes, including Bmp4, Fgf3, Dlx2, Lef1 and syndecan-1 (Satokata and Maas, 1994; Chen et al., 1996a; Bei and Maas, 1998). Recent studies have revealed the importance of vertebrate homologues of the Drosophila segment polarity gene hedgehog (hh) in the control of organogenesis (reviewed in Hammerschmidt et al., 1997). Several members of the vertebrate hh gene family that encode secreted signaling molecules have been identified. Among them, Sonic hedgehog (Shh) is best characterized. Shh is expressed in tissues with inductive and polarizing activities in several vertebrate organs (Hammerschmidt et al., 1997), including Hensen’s node (Levin et al., 1995, Chen et al., 1996b), limb (Riddle et al., 1993), neural tube (Echelard et al., 1993; Roelink et al., 1994), gut (Roberts et al., 1995), hair (Bitgood and McMahon, 1995), lung (Bellusci et al., 1997) and tooth (Bitgood and McMahon, 1995; Koyama et al., 1996; Vaahtokari et al., 1996; Hardcastle et al., 1998). Similar to HH action in Drosophila, SHH exerts its short- and long-range effects by activating downstream gene expression (reviewed in Johnson and Tabin, 1995). The Ptc gene, originally identified as a Drosophila segment polarity gene (Hooper and Scott, 1989), encodes a transmembrane protein which serves as a SHH receptor and plays a key role as a negative regulator in the SHH signaling pathway (Ingham et al., 1991; Marigo et al., 1996a; Stone et al., 1996). In addition, Ptc is also one of the general downstream targets of SHH signal. Ectopic Shh expression leads to ectopic Ptc expression in several vertebrate developing organs, including neural tube (Goodrich et al., 1996; Marigo and Tabin, 1996), limb (Marigo et al., 1996b), and lung (Bellusci et al., 1997). It has also been demonstrated that Gli1, one of the three vertebrate Gli gene family members that are 48 ZHANG ET AL. orthologous to the Drosophila cubitus interruptus (ci) gene, may function as a component of the Shh signaling pathway (Orenic et al., 1990; Hui et al., 1994). In the Drosophila wing disc, ci functions downstream of hh to activate Ptc expression (Alexandre et al., 1996). The involvement of ci in the regulation of Ptc by HH is conserved in vertebrates. For example, ectopic Shh expression in the chick limb and mouse lung results in upregulation of Ptc and Gli1 transcription (Marigo et al., 1996b, 1996c; Bellusci et al., 1997; Grindley et al., 1997). This upregulation of Gli1 by SHH appears to mediate the ability of SHH to upregulate Ptc expression, since misexpression of Gli1 also induces ectopic Ptc expression (Marigo et al., 1996c). Additional evidence for this idea comes from the observation that the expression of Ptc and Ptch2, a second mouse Patched gene, is downregulated in mutant mice lacking Gli2 (Motoyama et al., 1998). Although the important role of Shh in induction and patterning processes in vertebrate organogenesis has been well established by loss-of-function and misexpression studies (Chiang et al., 1996; reviewed in Hammerschmidt et al., 1997), components of the SHH signal transduction pathway and the general target genes of SHH signaling remain partly unknown. Here we present evidence that Shh, which is normally expressed in mouse dental epithelium, can induce expression of Ptc and Gli1 in mouse dental mesenchyme. In the dental mesenchyme of Msx1 mutant mice, Ptc, but not Gli1 expression, is markedly downregulated at the stages that precede overt morphologic differences in the Msx1 mutant tooth germ. This result suggests a requirement of Msx1 for Ptc expression in dental memesenchyme. Consistent with this hypothesis, we further demonstrate that induction of Ptc, but not Gli1, in dental mesenchyme by SHH requires functional MSX1 protein, implicating MSX1 as a component of the Shh signaling pathway that leads to the induction of Ptc expression. Finally we show that BMP4 is able to induce Ptc expression even in the absence of Msx1, indicating the involvement of different signaling pathways in the regulation of Ptc expression in the developing mouse molar tooth germ. RESULTS Shh and Ptc Are Expressed in the Early Tooth Germ In order to explore the Shh signaling pathway in the mouse tooth germ, we began with an analysis of Shh and Ptc expression in the early developing tooth. Whole mandibular arches from E11.0 to E13.5 mouse embryos were used for in situ hybridization (Fig. 1). At E11.0, Shh transcripts were first detected weakly in the incisor-forming region (Fig. 2A), while Ptc expression was found in the incisor as well as molar-forming regions (Fig. 2B). At E11.5, Shh expression becomes stronger in the incisors and starts to appear in the molar tooth germ (Fig. 2D). Shh expression is restricted to the dental epithelium and remains there until E13.5 (Fig. 2G,J, and data not shown). At E14.5, Shh transcripts are localized to the enamel knot of the dental epithelium (Koyama et al., 1996; Vaahtokari et al., 1996) (data not shown). Ptc transcripts were detected relatively weakly in dental epithelium but strongly in adjacent dental mesenchyme from E11.5 to E13.5 (Figs. 2E,H,K and 3H). Similar results on the expression pattern of Shh and Ptc in mouse tooth germ have recently been reported by Dassule and McMahon (1998) and Hardcastle et al. (1998). The expression pattern of Ptc always overlaps with, but is broader than, that of Shh in the tooth-forming regions at the same stages. Thus, similar to other Shh signaling centers (Goodrich et al., 1996), Ptc is also expressed in cells close to Shh-expressing cells in the developing tooth germ. In the dental mesenchyme, Ptc expression coincides with that of Msx1 temporally and spatially (Fig. 2C,F,I,L; MacKenzie et al., 1991), suggesting a potential correlation between the expression of these two genes during early tooth development. Msx1 Is Required for Ptc But Not Gli1 Expression in Dental Mesenchyme Observations that the expression pattern of Msx1 and Ptc overlaps in the molar dental mesenchyme prompted us to test the hypothesis that Msx1 and Ptc reside within the same signaling pathway in the early tooth development. We examined Ptc expression in the Msx1 mutant tooth germs of E11.5 to E13.5 embryos by in situ hybridization. At E11.5, weak signals of Ptc expression were detected in the incisor and molar tooth germ of the Msx1 mutant mandible by whole-mount in situ hybridization (Fig. 3A; cf. Fig. 2E). Sectioning of the samples revealed that Ptc expression occurred only in the dental epithelium but was downregulated from the adjacent mesenchyme (insert in Fig. 3A; cf. Fig. 2H). At E12.5, Ptc expression was almost undetectable in the Msx1 deficient tooth germ (Fig. 3D), as compared with Ptc expression in the wild-type tooth germ at the same stage (Fig. 2K). At E13.5, Ptc expression remained undetectable in the dental mesenchyme but was weakly detected in the dental epithelium of Msx1 mutants (Fig. 3G; cf. wild-type expression in Fig. 3H). However, Ptc expression was not affected in the developing tongue where Msx1 is not expressed (Fig. 3A,D), indicating the specificity of the reduction of Ptc expression in the developing tooth germ of Msx1 mutants. Since SHH is an inducer of Ptc expression, this downregulation of Ptc expression could result from altered Shh expression in the Msx1 mutant dental epithelium. However, we have found that Shh expression was preserved in the dental epithelium of E11.5 Msx1 mutants (Zhang et al., unpublished observation). Based on these results, we conclude that Ptc is a downstream gene from Msx1. It was previously reported that Gli1 may function as a component of Shh signal transduction machinery, 49 Msx1 IS A COMPONENT OF Shh SIGNALING PATHWAY residing downstream of Shh but upstream to Ptc (Marigo et al., 1996b, 1996c; Grindley et al., 1997). In addition, Gli1 expression was detected in both dental epithelium and dental mesenchyme at E14.5, coinciding with Ptc expression pattern (Hui et al., 1994) (data not shown). We therefore asked whether Gli1 also resides downstream from Msx1 in the early mouse developing tooth germ. In situ hybridization was performed to examine Gli1 expression in wild-type and Msx1 mutant tooth germ from E11.5 to E13.5. Gli1 expression was detected in the wild-type incisor and molar tooth germs of all stages examined (Fig. 3B,E,I). The transcripts were localized to dental epithelium and mesenchyme (Fig. 3I), as recently reported (Dassule and McMahon, 1998; Hardcastle et al., 1998). The wild-type expression pattern of Gli1 in the early developing tooth germ mirrors that of Ptc. However, by contrast, Gli1 expression in the Msx1 mutant tooth germ is mainly preserved (Fig. 3C,F,J). Thus, Msx1 is clearly required for the expression of Ptc, but not Gli1, in mouse dental mesenchyme. TABLE 1. Induction of Ptc and Gli1 Expression in Wild-Type and Msx1 Mutant Mandibles Probe Ptc Gli1 SHH ⫹/⫹ ⫺/⫺ 19/21 0/8 10/10 5/5 Inducer BMP4 ⫹/⫹ ⫺/⫺ 15/17* 7/7 nd nd BMP2 ⫹/⫹ ⫺/⫺ 0/10 nd nd nd nd, not done. *Biased expression; see text for details. to Ptc expression induced by SHH beads in the wildtype molar mesenchyme (Fig. 4A). These results indicate that Msx1 is required for the induction of Ptc by Shh in the mouse tooth germ. By contrast, SHH beads were able to induce Gli1 expression in the Msx1 mutant dental mesenchyme (Fig. 4D and Table 1). Thus, SHH can induce Gli1 expression in the absence of Msx1. Msx1 Is Required for Induction of Ptc But Not Gli1 by SHH in Dental Mesenchyme BMP4 Induces a Distinct Pattern of Ptc Expression in Wild-type and Msx1 Mutant Dental Mesenchyme Since SHH protein can induce Ptc and Gli1 expression in mandibular mesenchyme (Dassule and McMahon, 1998; Hardcastle et al., 1998), and since Ptc expression is downregulated in the Msx1 mutant dental mesenchyme, we asked whether the Shh induction of Ptc expression is mediated by Msx1. We first analyzed induction of Ptc and Gli1 expression by SHH in wildtype dental mesenchyme. Bacterially expressed recombinant SHH protein was prepared and purified. Whole mandibles isolated from wild-type E11.0 and E11.5 embryos were subjected to microsurgical separation of epithelium from mesenchyme after enzyme treatment. The whole mandibular mesenchyme was applied to organ culture with prospective tongue side facing up. Removal of the epithelium resulted in complete loss of Ptc expression in E11.0 mandibular explants after 24 hours in culture, indicating a requirement of epithelial signals to maintain Ptc expression at this stage. By contrast, endogenous Ptc expression is retained in E11.5 mandibular mesenchyme without epithelium after 24 hours in culture. Therefore, E11.0 mandibles were used for the induction assay. SHH-soaked beads were placed on top of the right side molar region, while BSA-soaked beads were placed on the left side molar region as controls. After 24 hours in culture, samples were assayed for gene expression by whole-mount in situ hybridization. The results, summarized in Table 1, demonstrated that beads soaked with SHH were able to induce expression of Ptc (Fig. 4A) and Gli1 (Fig. 4C) in the cells around the implanted beads in the dental mesenchyme. Whole mandibular mesenchyme from E11.0 Msx1 mutant embryos was applied to similar bead implantation culture before analysis of Ptc and Gli1 expression. Very weak or no induction of Ptc expression by SHH was observed in the Msx1 mutant molar mesenchyme (Fig. 4B and Table 1), as compared It has been demonstrated that BMP4 is a strong inducer of Msx1 expression in dental mesenchyme (Vainio et al., 1993). Bmp4 expression is found in dental epithelium at the initiation stage and then shifts to the dental mesenchyme shortly afterward (Vainio et al., 1993). Based on the observations that Msx1 is required for Ptc expression, we asked whether BMP4 can also induce Ptc expression in dental mesenchyme; if so, whether this induction requires Msx1. To address this question, BMP4-soaked beads were implanted to the molar region of mandibular mesenchyme from E11.0 wild-type and Msx1 mutant embryos. Samples were analyzed for Ptc expression after 24 hours in culture. BMP4 did indeed induce Ptc expression (Table 1). The results demonstrated that in the wild-type, unlike Msx1, which is induced in the mesenchymal cells around the BMP4 beads (Fig. 5A), Ptc expression was induced by BMP4 with a strong bias on the lingual and mesial sites of molar mesenchyme (Fig. 5B). Of the17 samples assayed, 15 exhibited a biased induction, while the other 2 samples showed induction in cells around the implanted BMP4 beads. These results indicate the presence of a distinct signaling pathway for the induction of Ptc expression in dental mesenchyme. This idea is supported by the observations that neither SHH induced Bmp4 expression nor BMP4 induced Shh expression in the dental mesenchyme (data not shown). Similar observations were also reported recently (Dassule and McMahon, 1998). However, although BMP4 could also induce Ptc expression in Msx1 mutant dental mesenchyme, the induction is seen in cells surrounding the beads in all samples examined (Fig. 5C; Table 1). The results indicate a requirement of Msx1 for the biased induction of Ptc by BMP4 in molar mesenchyme. BMP2, a BMP4 closely related signaling molecule, which is also expressed in the dental epithelium during 50 ZHANG ET AL. Fig. 4. Induction of Ptc but not Gli1 by SHH in molar mesenchyme requires Msx1. (A) Strong induction of Ptc by SHH-soaked bead is evident in the right side molar mesenchyme of an E11.0 wild-type mandible, as compared with no induction of Ptc by BSA bead placed on the left side molar mesenchyme. (B) SHH-soaked bead (blue bead on the right side) failed to induce PtcMsx1 mutant dental mesenchyme. White bead soaked with BSA placed on the left side served as a control. The sample was overdeveloped for color reaction to detect residual expression. Endogenous Ptc expression in the tongue is seen. (C) SHH bead (blue bead on the right side) induced Gli1 expression in an E11.0 wild-type molar mesenchyme, BSA-soaked bead (white bead on the left side) served as a control. (D) SHH beads were able to induce Gli1 expression (right side) in an E11.0 Msx1 mutant molar mesenchyme. BSA-soaked bead was placed on the left side as a control. Scale bar ⫽ 500 µm. Fig. 5. Induction of Ptc by BMP4 in molar mesenchyme. Blue beads on the right side of mandible are protein-soaked beads, while white beads on the left side are BSA controls. (A) BMP4-soaked bead could induced Msx1 expression in molar mesenchyme of an E11.0 wild-type mandible. Cells around the beads showed response to BMP4 induction. No induction is seen on the left side where a BSA bead was implanted. (B) In contrast to Msx1 expression, Ptc could only be induced by BMP4 bead in the cells localized on the lingual site of molar mesenchyme of an E11.0 wildtype mandible. BSA bead failed to induce Ptc expression on the left side molar mesenchyme. (C) BMP4 could induce Ptc expression in cells around the bead in an E11.0 Msx1 mutant molar mesenchyme. (D) BMP2-soaked bead did not induce Ptc expression in an E11.0 wild-type mandibular mesenchyme. Scale bar ⫽ 500 µm. early tooth morphogenesis (Turecková et al., 1995; Dassule and McMahon, 1998; Thesleff and Pispa, 1998), failed to induce Ptc expression when BMP2-soaked beads were implanted in E11.0 wild-type molar region (Fig. 4D and Table 1), as reported recently (Dussale and McMahon, 1998). This induction of Ptc expression in the molar mesenchyme is thus unique to BMP4. DISCUSSION Msx1 Is a Component of Shh Signaling Pathway A number of downstream target genes of the Shh signal, including Ptc, have been described in vertebrates in the past few years. Msx1, which coexpresses with Ptc in dental mesenchyme, is required for expression of Ptc in the dental mesenchyme. The specific reduction of Ptc expression in the Msx1 mutant dental mesenchyme establishes that Msx1 is genetically epistatic to Ptc. The requirement of Msx1 for SHH to induce Ptc expression implicates Msx1 as a key component of SHH signaling pathway. However, whether Ptc is a direct downstream gene regulated by Msx1 remains unclear. These results further indicate that Msx1 may not only be involved in epithelial–mesenchymal interactions by controlling the expression of inductive signals in the dental mesenchyme (Chen et al., 1996a; Bei and Maas, 1998), but may also participate in pattern formation by controlling Ptc expression. Our results have also demonstrated that expression of Gli1, a known downstream target of Shh and a component of Shh signaling transduction machinery, is preserved in the Msx1 mutant dental mesenchyme where Ptc expression is eliminated. Furthermore, SHH can induce Gli1 expression in the absence of Msx1. Gli1 may be either upstream of or parallel with Msx1 in the Shh signaling pathway. However, based on the fact that SHH-soaked beads failed to induce Msx1 expression in E11.0 wild-type dental mesenchyme (data not shown), we suggest that Gli1 is more likely to act in parallel with Msx1. It has been suggested that regulation of Ptc by Gli1 may be direct, since GLI consensus binding Msx1 IS A COMPONENT OF Shh SIGNALING PATHWAY sites are found in the promoter region of the Drosophila Ptc gene (Alexandre et al., 1996). Deletion of these consensus binding sites in the Ptc promoter results in failure of promoter activity in response to HH signal. In the mouse dental mesenchyme, however, Gli1 alone apparently may not directly regulate the expression of Ptc induced by SHH signal. It is possible that the Msx1 gene product interacts with GLI1 protein to activate Ptc expression. This hypothesis warrants further experiments. Since Ptc serves as a general downstream target as well as a receptor for SHH, one question that is how SHH can induce Gli1 expression in the Msx1 mutant dental mesenchyme in the absence of Ptc. The explanation could be that there exist residual Ptc transcripts that are below the sensitivity of in situ hybridization detection. Alternatively, other receptors for SHH, such as Ptch2, which is known to be present in the mouse tooth germ (Motoyama et al., 1998), may present in the Msx1 mutant tooth germ and mediate SHH signaling. BMP4 Signaling Represents a Distinct Pathway for Ptc Induction Among the inductive factors in the early dental epithelium postulated to be involved in the initiation of tooth morphogenesis, BMP4 was the first identified signal mediating the inductive interactions between epithelium and mesenchyme (Vainio et al., 1993). The function of BMP4 has been demonstrated by its ability to mimic the effect of presumptive dental epithelium in induction of a number of genes in dental mesenchyme (Vainio et al., 1993; Chen et al., 1996a). We show here that BMP4 also induces Ptc expression in dental mesenchyme. Ptc expression induced by BMP4 is limited to the lingual and mesial sites of molar mesenchyme. The specificity of BMP4 in Ptc induction in dental mesenchyme is further supported by the fact that BMP2 failed to induce Ptc expression in the similar condition (Dussale and McMahon, 1998; this study). Although BMP2 shares with BMP4 95% amino acid sequence identity and many functions, including gene induction in tooth germ (Vainio et al., 1993; Chen et al., 1996a; Thesleff and Pispa, 1998), these results provide evidence for the distinct function of BMP2 and BMP4 in early tooth development. A genetic model for the regulation of Ptc expression in the early tooth germ is thus proposed (Fig. 6). In this model, epithelial SHH induces Gli1 expression in the mesenchyme. Gli1 may activate Ptc expression by interacting with the product of Msx1, which is induced in the mesenchyme by the epithelially derived BMP4 (Vainio et al. 1993; Tucker et al., 1998). Mesenchymal BMP4, whose expression requires Msx1 and provides a positive feedback signal for maintenance of Msx1 expression (Chen et al., 1996a), can also induce Ptc expression. Epithelial BMP4 seems unlikely to be able to induce Ptc expression in the mesenchyme, since Bmp4 expression is preserved in the dental epithelium of Msx1 mutant tooth germ where Ptc expression is downregulated (Chen et al., 1996a; this study). 51 Fig. 6. A model of genetic pathway regulating Ptc expression in the early tooth germ. Epithelial BMP4 induces in the mesenchyme the expression of Msx1 which is required for the mesenchymal Bmp4 expression. Epithelially derived SHH may induce Ptc expression through the induction of Gli1. Msx1 participates in this induction of Ptc by SHH probably by interacting with Gli1. Meanwhile, mesenchymally expressed Bmp4 can also regulate Ptc expression. Ptc expression is thus regulated by at least two distinct pathways in the early mouse tooth germ. See text for details. Interestingly, BMP4 can induce Ptc expression in the absence of Msx1, but the asymmetric induction of Ptc by BMP4 obviously requires functional MSX1. Expression of Msx1 in dental mesenchyme may precondition cells in response to BMP4. Our results show that mesenchymal cells at different sites of molar tooth germ respond differentially to different inductive factors in terms of Ptc expression. These observations suggest that the prospective dental mesenchymal cells at the initiation stage have not only been specified in general to odontogenesis (Neubüser et al., 1997), but also behave differently according to their positions in the tooth germ. The neural crest derived cells that give rise to prospective dental mesenchyme may have been imprinted with different positional codes at the beginning of tooth morphogenesis. Msx1 expression, like that of Dlx genes, could be among these positional codes (Sharpe, 1995; Weiss et al., 1995; Thomas et al., 1997). Based on the results presented above, we conclude that Msx1 is a component in the Shh-Ptc pathway, required for the induction of Ptc expression by SHH. In the Bmp4-Ptc pathway, Msx1 may play a role in restricting the induction of Ptc by BMP4 to certain mesenchymal cells. Ptc expression pattern is thus regulated and coordinated by at least two distinct inductive signaling pathways in which Msx1 plays different roles in the early mouse dental mesenchyme. EXPERIMENTAL PROCEDURES Embryo Collection and Genotyping Mouse embryos used for in situ hybridization and tissue recombination were collected from matings of CD-1 mice. Msx1 mutant embryos were collected from Msx1⫹/⫺ females crossed with an Msx1⫹/⫺ male. The day of vaginal plug discovery was designated as embryonic day 0.5 (E0.5). The genotype of Msx1 mutant 52 ZHANG ET AL. embryos was determined by PCR using genomic DNA isolated from extraembryonic membranes, as previously described (Chen et al., 1996a). In Situ Hybridization Probes. All riboprobes used in this study were labeled with digoxygenin-UTP. The 0.65-kb Shh probe was made from a HindIII linearized template using T3 RNA polymerase (Bitgood and McMahon, 1995). The 2.25-kb Patched probe was transcribed with T7 RNA polymerase from a BglII linearized template covering the whole coding region (from Dr. M. Scott). The Gli1 probe was transcribed using T3 polymerase from a 1.7-kb template linearized by NotI (Hui et al., 1994). The 1.0-kb Bmp4 and 0.8-kb Msx1 probes were generated as described previously (Hill et al., 1989; Jones et al., 1991). Probe size and yield were checked by electrophoresis on a 1.5 % agarose gel with an RNA standard. Whole-mount in situ hybridization. Samples were fixed with ‘‘M’’ buffer (3.7% formaldehyde, 100 mM MOPS, pH 7.4, 2 mM EGDA, 1 mM MgSO4) at 4°C overnight, followed by bleaching with 10% H2O2 in ‘‘M’’ buffer at room temperature. Whole-mount in situ hybridization was performed as described previously (Chen et al., 1996b). Briefly, all probes were hydrolyzed to similar size before use. Hybridization and washing temperatures were set at 60°C. Proteinase K concentrations and digestion time varied, depending on the sample size and targeting tissue. Signals were visualized with NBT/BCIP (Boehringer Mannheim, Indianapolis, IN). Samples were washed with PBS and then refixed with ‘‘M’’ buffer for 30 minutes before photography and sectioning. Section in situ hybridization. Tissues were fixed in 4% paraformaldehyde/phosphate buffered saline (PBS) overnight and dehydrated with ethanol before embedding in paraffin wax. Sections of 8 µm were cut and used for nonradioactive in situ hybridization. Briefly, about 1 µg/ml of each probe was used. Hybridization and washing temperatures were set according to the size of probes. Maleic acid buffer and blocking reagent (from Boehringer Mannheim) were included in blocking and antibody washing steps. Signals were developed with BM purple alkaline phosphatase substrate. Sections were counterstained briefly with 1% Safranin and then mounted with Permount. Preparation of Protein-Soaked Beads BMP4 protein was obtained from Genetics Institute (Cambridge, MA). Recombinant SHH protein was prepared from a Shh-expression vector containing sequences encoding six histidine residues upstream to the mouse Shh-coding sequences (amino acids 25–198) (from Dr. A. McMahon). The protein was induced and purified according to the method described previously (Marti et al., 1995). The inductive activity of SHH protein was examined and confirmed by its ability to induce digit duplication of chick wing buds after SHHsoaked beads (1 mg/ml) were implanted to the anterior margin of a host wing bud as well as its ability to induce cPix2 expression in an early chick embryo (St. Amand et al., 1998). Affi-Gel blue agarose beads (Bio-Rad) were washed with PBS before incubating with BMP4 protein (60 mg/ml), BMP2 (100 mg/ml) and SHH protein (1 mg/ml), respectively, at 37°C for 30 minutes. Control beads (white heparin beads) were soaked with similar concentrations of BSA under same conditions. All protein-soaked beads were stored at 4°C and used within 1 week. Bead Implantation and Organ Culture To separate the epithelium from the whole mandibular arch, E11.0 mandibles were incubated with 2.25% trypsin/0.75% pancreatin on ice for 10 minutes, followed by incubation in PBS/horse serum on ice for 20 minutes. Mandibular epithelium was then separated microsurgically. Freshly separated mesenchymal tissues were placed on Millipore filters (pore size, 0.1 µM) supported by metal grids. The whole mandibular mesenchyme was oriented with tongue side facing up. Beads were placed on the top the molar tooth regions of the whole mandibular mesenchyme, with SHH or BMP4-soaked beads placed on the right-side molar, while BSA beads on the left-side molar regions. All explants were cultured in Dulbecco’s minimal essential medium (DMEM) with 10% FCS at 37°C for 24 hours. Samples were fixed with ‘‘M’’ buffer and processed for whole-mount in situ hybridization. ACKNOWLEDGMENTS We thank Drs. Andrew McMahon (Harvard) and Matthew Scott (Stanford) for plasmids, Anthony Celeste (Genetics Institute, Inc., Cambridge, MA) for BMP2 and BMP4 recombinant proteins, Richard Maas (Harvard), in whose laboratory this study was originally initiated, for Msx1 mutant mice, and Ken Muneoka and Carol Burdsal and members of their laboratories for sharing laboratory resources and help during the initiation phase of the Chen laboratory. Y.P.C. thanks Dr. Ken Muneoka for his encouragement. REFERENCES Alexandre C, Jacinto A, Ingham PW. 1996. Transcriptional activation of hedgehog target genes in Drosophila is mediated directly by the cubitus interuptus protein, a member of the GLI family of zinc finger DNA-binding protein. Genes Dev 10:2003–2013. Bei M, Maas R.1998. FGFs and BMP4 induce both Msx1-independent and Msx1-dependent signaling pathways in early tooth development. Development 125:4326–4333. Bellusci S, Furuta Y, Rush MG, Henderson R, Winnier G, Hogan BLM. 1997. Involvement of Sonic hedgehog (Shh) in mouse embryonic lung growth and morphogenesis. Development 124:53–63. Bitgood MJ, McMahon AP. 1995. Hedgehog and Bmp genes are coexpressed at diverse sites of cell-cell interaction in the mouse embryo. Dev Biol 172:126–138. Chen YP, Maas R. 1998. Signaling loops in the reciprocal epithelial– mesenchymal interactions of mammalian tooth development. In: Chuong C-M (ed). ‘‘Molecular basis of epithelial appendage morphogenesis.’’ Austin, TX: RG Landes, pp 265-282. Chen YP, Bei M, Woo I, Satokata I, Maas R. 1996a. Msx1 controls inductive signaling in mammalian tooth morphogenesis. Development 122:3035–3044. Msx1 IS A COMPONENT OF Shh SIGNALING PATHWAY Chen YP, Dong D, Kostetskii I, Zile MH. 1996b. Hensen’s node from vitamin A-deficient quail embryo induced chick limb duplication and retains its normal asymmetrical expression of Sonic hedgehog (Shh). Dev Biol 173:256–264. Chiang C, Litingtung Y, Lee E, Young KE, Corden JL, Westphal H, Beachy PA. 1996. Cyclopia and defective axial patterning in mice lacking Sonic hedgehog gene function. Nature 383:407–413. Dassule HR, McMahon AP. 1998. Analysis of epithelial–mesenchymal interactions in the initial morphogenesis of the mammalian tooth. Dev Biol 202:215–227. Echelard Y, Epstein DJ, St-Jacques B, Shen L, Mohler J, McMahon JA, McMahon AP. 1993. Sonic hedgehog a member of a family putative signaling molecules, is implicated in the regulation of CNS polarity. Cell 75:1417–1430. Goodrich LV, Johnson RL, Milenkovic L, McMahon JA, Scott MP. 1996. Conservation of the hedgehog/patched signaling pathway from flies to mice: induction of a mouse patched gene by hedgehog. Genes Dev 10:301–312. Grindley JC, Bellusci S, Kerkins D, Hogan BLM. 1997. Evidence for the involvement of the Gli gene family in embryonic mouse lung development. Dev Biol 188:337–348. Hammerschmidt M, Brook A, McMahon AP. 1997. The world according to hedgehog. Trends Genet 13:14–21. Hardcastle Z, Mo R, Hui C-C, Sharpe PT. 1998. The Shh signaling pathway in tooth development: defects in Gli2 and Gli3 mutants. Development 125:2803–2811. Hill RE, Jones PF, Rees AR, Sime CM, Justice MJ, Copeland NC, Jenkins NA, Graham E, Davidson DR. 1989. A new family of mouse homeobox containing genes: molecular structure, chromosomal location and developmental expression of Hox7.1. Genes Dev 3:26–37. Hooper JE, Scott MP. 1989. The Drosophila patched gene encodes a putative membrane protein required for segmental patterning. Cell 59:751–765. Hui C-C, Slusarski D, Platt KA, Holmgren R, Joyner AL. 1994. Expression of three mouse homologs of the Drosophila segment polarity gene cubitus interuptus, Gli, Gli2 and Gli3, in ectodermand mesoderm-derived tissues suggests multiple roles during postimplantation development. Dev Biol 162:402–413. Ingham PW, Taylor AM, Nakano Y. 1991. Role of the Drosophila Patched genes in positional signaling. Nature 353:184–187. Johnson RL, Tabin C. 1995. The long and short of hedgehog signaling. Cell 81:313–316. Jones CM, Lyons KM, Hogan BLM. 1991. Involvement of bone morphogenetic protein-4 (BMP-4. and Vgr-1 in morphogenesis and neurogenesis in the mouse. Development 111:531–542. Koyama E, Yamaai T, Iseki S, Ohuchi H, Nohno T, Yoshioka H, Hayashi Y, Leatherman JL, Golden EB, Noji S, Pacific M. 1996. Polarizing activity, Sonic hedgehog, and tooth development in embryonic and postnatal mouse. Dev Dyn 206:59–72. Kratochwil K, Dull M, Fariñas Galceran J, Grosschedl R. 1996. Lef1 expression is activated by BMP4 and regulates inductive tissue interactions in tooth and hair development. Genes Dev 10:1382– 1394. Levin M, Johnson RL, Stern CD, Kuehn M, Tabin C. 1995. A molecular pathway determining left–right asymmetry in chick embryogenesis. Cell 82:803–814. MacKenzie A, Leeming GL, Jowett AK, Ferguson MWJ, Sharpe, PT. 1991. The homeobox gene Hox 7.1 has specific regional and temporal expression patterns during early murine craniofacial embryogenesis, especially tooth development in vivo and in vitro. Development 111:269–285. Marigo V, Tabin CJ. 1996. Regulation of Patched by Sonic hedgehog in the developing neural tube. Proc Natl Acad Sci USA 93:9346–9351. Marigo V, Davey RA, Zuo Y, Cunningham JM, Tabin CJ. 1996a. Biochemical evidence that Patched is the Hedgehog receptor. Nature 384:176–179. Marigo V, Scott MP, Johnson RL, Goodrich LV, Tabin CJ. 1996b. Conservation in hedgehog signaling: induction of a chicken Patched homolog by Sonic hedgehog in the developing limb. Development 122:1225–1233. 53 Marigo V, Johnson RL, Vortkamp A, Tabin CJ. 1996c. Sonic hedgehog differentially regulates expression of GLI and GLI3 during limb development. Dev Biol 180:273–283. Marti E, Bumcrot D, Takada R, McMahon AP. 1995. Requirement of 19K form of Sonic hedgehog for induction of distinct ventral cell types in CNS explants. Nature 375:322–325. Motoyama J, Takabatake T, Takeshima K and Hui C–C. 1998. Ptch2, a second mouse Patched gene is co-expressed with Sonic hedgehog. Nature Genet 18:104–106. Neubüser A, Peters H, Ballings R, Martin GR. 1997. Antagonistic interactions between FGF and BMP4 signaling pathways: a mechanism for positioning the sites of tooth formation. Cell 90:147–155. Orenic TV, Slusarski DC, Kroll KL, Holmgren RA. 1990. Cloning and characterization of the segment polarity gene cubitus interuptus Dominant of Drosophila. Genes Dev 4:1053–1067. Palmer RM, Lumsden AGS. 1987. Development of periodontal ligament and alveolar bone in homografted recombinations of enamel organ and papillary, pulpal and follicular mesenchyme in the mouse. Arch Oral Biol 32:281–289. Riddle RD, Johnson RL, Tabin C. 1993. Sonic hedgehog mediates the polarizing activity of the ZPA. Cell 75:1401–1416. Roberts DJ, Johnson RL, Burke AC, Nelson CE, Morgan BA, Tabin C. 1995. Sonic hedgehog is an endodermal signal inducing Bmp-4 and Hox genes during induction and regionalization of the chick hindgut. Development 121:3163–3174. Roelink H, Augsburger A, Heemskerk J, Korzh V, Norlin S, Ruiz i Altaba A, Tanabe Y, Placzek M, Edlund T, Jessell TM, Dodd J. 1995. Floor plate and motor neuron induction by vhh-1, a vertebrate homolog of hedgehog expressed by the notochord. Cell 76:761–775. Satokata I, Maas R. 1994. Msx-1 deficient mice exhibit cleft palate and abnormalities of craniofacial and tooth development. Nature Genet 6:348–356. Sharpe PT. 1995. Homeobox genes and orofacial development. Connect Tissue Res 32:17–25. St.Amand TR, Ra J, Zhang YD, Hu YP, Baber S, Qiu M-S, Chen YP. 1998. Cloning and expression pattern of chicken Pitx2: a new component in the SHH signaling pathway controlling embryonic heart looping. Biochem Biophys Res Commun 247:100–105. Stone DM, Hynes M, Armanini M, Swanson TA, Gu Q, Johnson RL, Scott MP, Pennica D, Goddard A, Philips H, Noll M, Hooper JE, de Sauvage F, Rosenthal A. 1996. The tumour-suppressor gene Patched encodes a candidate receptor for sonic hedgehog. Nature 384:129– 134. Thesleff I, Nieminen P. 1996. Tooth morphogenesis and cell differentiation. Curr Opin Cell Biol 8:844–850. Thesleff I, Pispa J. 1998. The teeth as models for studies on the molecular basis of the development and evolution of organs. In: Chuong C-M (ed). ‘‘Molecular basis of epithelial appendage morphogenesis.’’ Austin, TX: RG Landes, pp 157-179. Thesleff I, Sahlberg C. 1996. Growth factors as inductive signals regulating tooth morphogenesis. Semin Cell Dev Biol 7:185–193. Thomas BL, Tucker AS, Qiu M-S, Ferguson CA, Hardcastle Z, Rubenstein JLR, Sharpe PT. 1997. Role of Dlx-1 and Dlx-2 genes in patterning of the murine dentition. Development 124:4811–4818. Tucker AS, Al Khamis A, Sharpe PT. 1998. Interactions between Bmp-4 and Msx-1 act to restrict gene expression to odontogenic mesenchyme. Dev Dyn 212:533–539. Turecková J, Sahlberg C, Aberg T, Ruch JV, Thesleff I, Peterková R. 1995. Comparison of expression of the msx-1, msx-2, BMP-2 and BMP-4 genes in the mouse upper diastemal and molar primordia. Int J Dev Biol 39:459–468. Vaahtokari A, Åberg T, Jernvall J, Keranen S, Thesleff I. 1996. The enamel knot as a signaling center in the developing mouse tooth. Mech Dev 54:39–43. Vainio S, Karavanova I, Jowett A, Thesleff I. 1993. Identification of BMP-4 as a signal mediating secondary induction between epithelial and mesenchymal tissues during early tooth development. Cell 75:45–58. Weiss KM, Ruddle FH, Bollekens J. 1995. Dlx and other homeobox genes in the morphological development of the dentition. Connect Tissue Res 32:35–40.