A Dual Role for Ikk in Tooth Development

Transcription

1 Developmental Cell, Vol. 6, , February, 2004, Copyright 2004 by Cell Press A Dual Role for Ikk in Tooth Development Atsushi Ohazama, 1 Yinling Hu, 2,3 buds are initiated by invagination of epithelium at de- Ruth Schmidt-Ullrich, 4 Yixue Cao, 2 fined positions. Initiation of tooth germs has been shown Claus Scheidereit, 4 Michael Karin, 2 to require epithelial signals and involve molecules beand Paul T. Sharpe 1, * longing to the Bmp, Wnt, Fgf, and Hedgehog families. 1 Department of Craniofacial Development In addition, transcription factors such as Msx1, Lef1, GKT Dental Institute Pax9, Barx1, Lhx6/7, and Dlx family members act down- King s College stream of these epithelial signals to regulate tooth initia- Guy s Hospital tion and morphogenesis (for reviews, see Thesleff and London Bridge Sharpe, 1997; Cobourne and Sharpe, 2003). London SE1 9RT Tooth type is determined early in development by United Kingdom spatially restricted expression of homeobox genes in 2 Laboratory of Gene Regulation and the ectomesenchyme. Mice lacking functional Dlx1 and Signal Transduction Dlx2 genes have a tooth patterning phenotype where Department of Pharmacology development of maxillary molar teeth is inhibited but University of California, San Diego development of all other teeth is normal (Thomas et al., La Jolla, California ). Misexpression of Barx1 in presumptive incisor 3 Science Park, Department of Carcinogenesis mesenchyme results in a transformation of incisors into The University of Texas MD Anderson molars (Tucker et al., 1998a; unpublished data). The Cancer Center spatial domains of these homeobox genes in jaw primor- Smithville, Texas dia ectomesenchyme are generated by FGF8, expressed 4 Max-Delbrück-Center of Molecular Medicine proximally, and BMP4, expressed distally, in oral epithe- Robert-Rössle Straße 10 lium, acting to restrict expression of homeobox genes Berlin in ectomesenchyme (Tucker et al., 1998a). A crude proxi- Germany mal-distal (molar-incisor) prepattern of signaling molecules is thus first present in the oral epithelium. The molecules that control the mechanisms of cusp Summary formation to generate teeth of different shapes are poorly understood. Mice with mutations in genes that cause hypohidrotic (anhidrotic) ectodermal dysplasias IKK is a component of the I B kinase (IKK) complex in humans such as EctodysplasinA1 (EdaA1) (tumor nethat plays a key role in the activation of NF- B. In Ikk crosis factor [TNF] ligand; Tabby), Edar (TNF receptor; mutant mice and mice expressing a transdominant Downless), and Edaradd (death domain adaptor; Crinnegative mutant of I B (c I B N ), molars have abnor- kled), have molars with abnormal cusp formation (Kere mal cusps, indicating that Ikk is involved in cusp et al., 1996; Ferguson et al., 1997; Srivastava et al., 1997; formation through the NF- B pathway. However, Ikk Monreal et al., 1999; Pispa et al., 1999; Tucker et al., mutant incisors also have an earlier phenotype where 2000; Headon et al., 2001; Laurikkala et al., 2001). These epithelium evaginates outward into the developing genes are expressed at the cap stage of tooth developoral cavity rather than invaginating into the underlying ment where they affect the formation of the enamel knot mesenchyme. A similar evagination of epithelium was signaling centers, postulated to have a role in cuspal also observed in whisker development, suggesting that morphogenesis. EdaA1, Edar, and Edaradd form a cell Ikk contributes to the direction of epithelial growth signaling pathway whose intracellular effects are beduring the early stages of development in many ectoder- lieved to be mediated by nuclear factor kappa B (NF- B) mal appendages. Since c I B N mice have normal incisor (Yan et al., 2000; Döffinger et al., 2001; Kumar et al., epithelial invagination, Ikk s role appears to be NF- B 2001; Schmidt-Ullrich et al., 2001). independent. Changes in Notch1, Notch2, Wnt7b, and The NF- B pathway plays a major role in many physio- Shh expression found in incisor epithelium of Ikk mu- logical and pathological processes (for reviews, see tants suggest that this NF- B-independent function is Karin and Delhase, 2000; Israël, 2000). NF- B activity is mediated by Notch/Wnt/Shh signaling pathways. regulated by interaction with an inhibitor, Inhibitor of kappa B (I B), which in nonstimulated cells acts to retain Introduction NF- B in the cytoplasm by masking the nuclear localization sequence. Exposure to proinflammatory stimuli (e.g., interleukin-1 or TNF ) results in rapid phosphorylation Mammalian teeth develop from oral epithelium and of I B that leads to site-specific ubiquitination and deg- (ecto-)mesenchymal cells through a series of reciprocal radation by 26S proteasomes. The resulting free NF- B signaling interactions (for review, see Thesleff et al., dimers translocate to the nucleus and regulate target 1995). The first morphological signs of the beginning of gene transcription. The protein kinase that phosphorytooth development are localized thickenings of the oral lates I B in response to stimuli is a multiprotein complex, epithelium along the developing jaw primordia. Tooth I B kinase (IKK), composed of two catalytic subunits, IKK and IKK, and a regulatory subunit, IKK (also *Correspondence: paul.sharpe@kcl.ac.uk called NEMO). Mice with targeted mutations in Ikk are

2 Developmental Cell 220 Figure 1. The Expression Pattern of Ikk during Early Tooth and Whisker Development Radioactive in situ hybridization on frontal sections showing Ikk expression in embryo heads at E11.5 (A), E13.5 (B D and F), and E14.5 (E). Upper incisor (A and B), lower incisor (C), molar (D and E), and whisker follicle (F). Weak expression could be detected in presumptive incisor epithelium from E11.5 (arrowheads in [A]). Expression was restricted to the outer portion of the invaginating epithelium at bud and early cap stages (B F). Tooth germ and whisker follicle epithelium is outlined in red. Molar Tooth Development Dissected teeth and sections of Ikk mutant mice em- bryos were compared with those of heterozygous and wild-type littermates. Ikk heterozygous mice were fer- tile and revealed phenotypically normal tooth development. Analysis of molar teeth in newborn mutants re- vealed rounded and shallow cusps that resembled those found in EdaA1, EdaR, and Edaradd mutants (Figures 2A 2D). Since molecular changes in primary and sec- ondary enamel knots have been observed in EdaA1, EdaR, and Edaradd mutant molar tooth germs, the ex- pression of enamel knot marker genes Bmp4, Shh, and Wnt10b was examined in Ikk mutant molar tooth germs. At E , when the primary enamel knot is visible histologically, the size of the mutant enamel knot was slightly smaller in mutants (Figures 2E and 2F). Shh and Bmp4 expression domains were also slightly smaller and expression of Wnt10b was weaker in mutant enamel knots (Figures 2G 2L). These changes in the expression of enamel knot marker genes in the Ikk mutant molar tooth germs were similar to those that have previously been described in EdaA1 mutants (Pispa et al., 1999; Laurikkala et al., 2001). perinatally lethal and limb outgrowth is severely impaired despite unaffected skeletal development (Hu et al., 1999; Li et al., 1999; Takeda et al., 1999). Newborn mutant mice have shiny, taut, and sticky skin without whiskers, and histological analysis shows a thicker epidermis that is unable to differentiate. Limbs are wrapped inside the skin and do not extend properly out of the body trunk. Ikk mutants have also been reported to have craniofacial and tooth abnormalities. We report here that Ikk is expressed in tooth epithelium and has an essential role in tooth development, being required for development of both incisor and molar teeth. Molar teeth in Ikk mutants resemble those from EdaA1, Edar, and Edaradd mutants, having an ab- normal cusp morphology. The abnormal cusp formation is found to be caused by the reduction of the NF- B pathway. Incisors, however, have a more severe and earlier phenotype where tooth buds fail to invaginate into the underlying mesenchyme but instead evaginate into the developing oral cavity. A similar evagination of epithelium was also observed in whisker follicles. Ikk thus regulates the direction of epithelial invagination of ectodermal appendages. Unlike in cusp morphogenesis, the NF- B pathway is not involved in the invagination of incisor tooth epithelium, consistent with previous results on the skin development (Hu et al., 2001). Changes in the expression of Notch, Shh, and Wnt7b in Ikk mutant incisor epithelium indicate a possible role for these molecules in the Ikk -dependent invagination process. Results Expression of Ikk in Early Tooth Development Expression of Ikk was analyzed in the developing heads of mouse embryos between days 10.5 and 15.5 of gestation (E10.5 E15.5) using radioactive in situ hybridization. Weak expression could be detected at E11.5 (Figure 1A) and E12.5 (data not shown) in presumptive incisor epithelium. Ikk expression was clearly visible by E13.5 in both incisor and molar germs where it was restricted to the outer edges of invaginating tooth epithelium (Figures 1B 1D). The expression pattern was retained in molar epithelium at E14.5 (Figure 1E). The Ikk expression pattern in whisker follicles was similar to that in tooth primodia at E13.5 (Figure 1F). The NF- B Pathway in Tooth Development In order to examine the role of the NF- B pathway in tooth development, we first analyzed expression of an- other molecule in the pathway. Localized expression of I B was observed in tooth germ epithelium of both molars and incisors. Since there is feedback of I B by active NF- B, transcription of I B should be increased as NF- B is activated (Hoffmann et al., 2002). Weak expression of I B was detected at E in tooth germs, but from E13.5, expression was evident at the tips of tooth buds (Figure 3A). At E14.5 expression was clearly visible in the enamel knots of both molar and incisor tooth buds (Figures 3C and 3E). In order to confirm NF- B activity, we also used transgenic mice with a B-dependent nuclear localization sequence- -galactosidase (lacz) reporter gene, (Ig ) 3 conalacz (Schmidt-Ullrich et al., 1996). In (Ig ) 3 conalacz mice, LacZ expression was similar to the I B expression in tooth buds and the enamel knots (Figures 3B, 3D, and 3F). Since both primary and secondary enamel knots are believed to have a role in cusp morphogenesis and expression of Edar and Edaradd is localized to enamel

3 A Dual Role for Ikk in Tooth Development 221 Figure 2. The Molar Tooth Phenotypes of Ikk Mutant Mice First lower molar at E14 (G L), E15.5 (E and F), and E18 (A and D) of wild-type (A, C, E, G, I, and K) and mutant (B, D, F, H, J, and L). Dissected molar (A and B) and sagittal sections (C and D) show that mutant newborn molars (B and D) have rounded and shallow cusps. (E and F) E15.5 frontal sections showing the histology of late cap stage molar tooth germs. (G L) E14, frontal sections showing molar tooth germs at the cap stage of development. Mutant enamel knots (arrow in [F]) were slightly smaller than wild-types (arrow in [E]). Shh (H) and Bmp4 (arrow in [J]) expression domains were slightly smaller in mutant enamel knots. Expression of Wnt10b (L) was weaker in mutant enamel knots. Tooth germ epithelium is outlined in red. Figure 3. Activation of the NF- B Pathway in Tooth Development Incisors at E14.5 (E), molars at E13.5 (A, B, and G), E14.5 (C, D, and H), and newborn (F). In situ hybridization on frontal sections showing I B (A, C, and E), Ikk (G), and Ikk (H) expression in embryos. LacZ expression in (Ig ) 3 conalacz mice (B, D, and F). LacZ expression is observed in primary enamel knots (D) and secondary enamel knots (arrowheads in [F]). Ikk expression was restricted to the outer portion of invaginated epithelium at bud stage (G). Ikk expression was found in molar internal enamel epithelium at cap stage (H). Tooth germ epithelium is outlined in red. NF- B Pathway in Molar Development To determine whether NF- B was required for cusp formation of molar tooth germs, we examined the molar tooth phenotype in transgenic embryos expressing a super-repressor form of I B (c I B N ) (Schmidt-Ullrich et al., 2001). Among I B family members, I B is thought to knots, these results support the role of the NF- B pathway be the most important and is involved in all extracellular in EdaA1 signaling in tooth development. signal pathways that lead to rapid NF- B activation. In order to investigate whether other IKK subunits are Deletion of the N-terminal destruction box of I B creinvolved in tooth development, expression of Ikk and ates a super-repressor, I B N, which inhibits NF- B Ikk was analyzed in the developing heads of embryos release (Krappmann et al., 1996). Transgenic embryos between E10.5 and E14.5 using radioactive in situ hy- expressing this super-repressor were examined and bridization. Weak expression of Ikk could be first detected were found to have an abnormal cusp phenotype that at E12.5 in epithelial cells of whisker follicles (data was more severe than observed in Ikk mutants. The not shown). Ikk was expressed in the outer regions of molar cusps of c I B N mice were reduced in size and invaginated molar epithelium at E13.5 similar to Ikk number (Figures 4A 4D) and third molars were absent (Figure 3G). Ikk expression in developing tooth primodia (Figure 4D). The expression of primary enamel knot was clearly visible at E13.5 in both molars and marker genes Shh, Wnt10b, and Bmp4 was examined incisors (Figure 3H). In molars, Ikk was only expressed in c I B N mice, and these domains were found to be in internal tooth epithelium that seemed to be a different smaller at E15.5 (Figures 4E 4L). A form of Ikk mutant expression pattern from Ikk, whereas the expression mice (Ikk AA/AA ) have been produced where alanines re- pattern in incisors was similar to expression of Ikk (data placed serines in the activation loop (Cao et al., 2001). not shown). These mice have been shown to retain some NF- B

4 Developmental Cell 222 the oral cavity at birth, whereas the incisors of wildtypes were covered with a thin layer of developing alveolar bone and a thick layer of connective tissue and did not start to erupt until 1 2 weeks after birth (Figures 5C 5F). Molar teeth did not show these phenotypes in newborn mutants (Figure 2D). To establish the timing of the first appearance of the incisor tooth anomalies, frontal sections of mutant embryos from E10 to 14 were examined. The first distinguishable anomalies were observed with an abnormal protrusion visible in mutant presumptive incisor regions at E13 (Figures 5G 5J). Invaginating epithelium that is normally observed in wildtype teeth as they form buds was not evident in mutant incisors. Instead, localized areas of epithelium were visible bulging outward into the developing oral cavity. Mutant incisor epithelium thus does not invaginate into the underlying mesenchyme but rather evaginates outward into the developing oral cavity (Figures 5H and 5J). Primary enamel knots in incisor tooth germs express the same marker genes as molars. Shh and Wnt10b were clearly expressed in the enamel knots of the mutant tooth germs but the position of the positive cells was obviously different. Shh and Wnt10b positive enamel knot cells were localized at the outermost tip of the mutant evaginated incisor buds, whereas they were localized in the center of invaginated incisor epithelium in wild-type littermates (Figures 5K 5N). In addition to incisor tooth buds, evagination of epithelium was also observed in whisker development at E14 (Figures 5O and 5P). Shh positive cells were also localized at the outermost tip of the mutant evaginated epithelium of hair follicles, whereas they were localized at the tip of invaginated epithelium in wild-type littermates (Figures 5Q and 5R). Ikk is thus an essential gene that directs the direction of epithelial ingrowth during development of skin appendages. To determine whether NF- B was required for epithelial invagination of incisor tooth germs, we examined tooth bud development in c I B N mice. Normal tooth germ invagination were found in incisors of c I B N mice (Figures 5S and 5T), indicating that the evagination of incisor epithelium in Ikk mice was not caused by lack of NF- B activity. Normally developed incisors were observed in 6-week-old c I B N mice (Fig- Figure 4. The Molar Tooth Phenotype of c I B N Mice and Ikk AA/ AA Mice ures 5U and 5V), confirming that incisor development is First lower molar at 6 weeks old (C, D, M, and N), newborn (A and not affected by loss of NF- B activity. Normal incisors B), and E15.5 (E L) of wild-type (A, C, E, G, I, K, and M), c I B N (B, D, F, H, J, and L), and Ikk AA/AA mice (N). Radioactive in situ hybridization were also found in Ikk AA/AA mice (Figures 5W and 5X). (G L). Frontal section showing enamel knot is absent in c I B N mice (F). In c I B N mice, Shh (H) and Wnt10b (J) domains are slightly Notch Expression in Incisor Invagination smaller in enamel knots. Expression of Wnt10b (J) and Bmp4 (L) is In order to begin to elucidate the molecular basis of the weaker in c I B N mice. Tooth germ epithelium is outlined in red. incisor phenotype in Ikk mutant embryos, the expression of well-characterized tooth epithelial and mesenchymal marker genes was examined using in situ hybridactivity in certain tissues. Significantly, we observed ization. The expression patterns of Activin A, amelogenin, normal molar tooth development in these mice (Figures Barx1, Bmp4, Cbfa1, dentin sialophosphoprotein (DSPP), 4M and 4N), consistent with observed normal NF- B Dlx2, Lef1, Lunatic Fringe, Mfz6, Msx1, Notch1, Notch2, activity in mammary gland epithelial cells following treat- Pax9, Pitx2, Ptc1, Serrate1, Shh, and Wnt7b were examment with TNF (Cao et al., 2001). ined in Ikk mutant embryos from E10 to newborn. Expression of all these genes except Notch1, Notch2, Shh, and Wnt7b were found to be normal in Ikk mutant Incisor Tooth Development incisor tooth germs (Figure 6). Expression of Shh is first Analysis of the teeth in newborn Ikk mutants showed detectable in presumptive dental epithelium from E10.5, severely abnormal incisors (both maxillary and mandib- and subsequently, at the bud-cap stages, it is localized ular) that were shorter and had enlarged crowns (Figures in the epithelial enamel knot cells. Expression of Shh 5A 5F). The mutant incisors were already exposed in in Ikk mutant incisor epithelium at E10.5 was slightly

5 A Dual Role for Ikk in Tooth Development 223 Figure 5. The Incisor Tooth and Whisker Phenotypes of Wild-Type, Ikk /, c I B N, and Ikk AA/AA Mice Incisors at E13 (G J), E13.5 (S and T), E14 (K N), E18 (A F), and 6 weeks old (U X). Whisker follicles at E14 (O R). Incisors (A, C, E, G, I, K, M, U, and W) and whiskers (O and Q) of wild-types. Incisors (B, D, F, H, J, L, and N) and whiskers (P and R) of Ikk mutant. Incisors of c I B N (S, T, V) and Ikk AA/AA (X). Dissected incisors (A and B) and sagittal sections (C F) show that both upper and lower mutant incisors are short, and their anterior extents are enlarged and exposed to the oral cavity (arrowheads in [D] and [F]). At E13, frontal sections showing evagination of tooth epithelium in the mutants (arrows in [H] and [J]). (K N, Q, and R) In situ hybridization on frontal sections shows expression of marker genes in the developing jaws. At E14, expression of Shh was localized at the outermost tip of evaginating epithelium of incisors in mutants (L), whereas Shh was localized at the bottom of invaginating epithelium of incisor in wild-types (K). (M and N) Expression of Wnt10b showed a similar pattern to Shh. Evagination of epithelium was seen in mutant whisker follicles (arrowhead in [P]) at E14. Expression of Shh in mutant whisker follicles was also localized at the outermost tip of evaginating epithelium (arrowheads in [R]). Normal invaginating epithelium of both upper incisors (S) and lower incisors (T) was observed in frontal sections of c I B N mice ( N) at E13.5. Normal lower incisors at 6 weeks old of c I B N (V) and Ikk AA/AA mice (X). Tooth germ and whisker follicle epithelium is outlined in red. expanded (data not shown) but was significantly ex- and immune responses, but also for bone morphogenesis panded at E12 compared to controls (Figures 6G and (Franzoso et al., 1997; Iotsova et al., 1997). c I B N 6G ). Wnt7b acts to repress Shh expression in oral ecto- mice expressing a super-repressor form of I B revealed derm, thus maintaining the boundaries between oral and defective early morphogenesis of hair follicles dental ectodermal cells (Sarkar et al., 2000). The expres- and exocrine glands, identical to those found Eda, Edar, sion domain of Wnt7b in Ikk mutant incisor epithelium and Edaradd mutants that are linked to hypohidrotic was smaller than in wild-types, consistent with the (anhidrotic) ectodermal dysplasia in humans (Kere et expansion of Shh expression (Figures 6H and 6H ). al., 1996; Ferguson et al., 1997; Srivastava et al., 1997; Notch1/2 are both expressed in oral/dental epithelium Monreal et al., 1999; Headon et al., 2001). The observation from E10. At E11, Notch1 was found to be downregulated of abnormal cusps of molar teeth in c I B N mice in Ikk mutant presumptive incisor epithelium (Fig- provided confirmation that the NF- B pathway is required ures 7A and 7B). In E13 and E14 Ikk mutant embryos, for molar cusp formation. Although the molar Notch1 expression was downregulated in both incisor cusp phenotype observed in Ikk mutant mice further and molar epithelium but was still present in the neural supports the role of NF- B, the fact that the teeth were tube (Figures 7C 7E and 7I 7K, and data not shown). normal in Ikk AA/AA mutants suggest that sufficient kinase Notch2 expression was also downregulated in mutant activity via Ikk or some yet-unknown kinase is present incisor epithelium but was preserved in molar epithelium in tooth epithelium in these mutants. Abnormal molar (Figures 7F 7H and 7L 7N). In order to investigate cusp morphogenesis in Ikk mutants and c I B N mice whether Notch signaling is downstream of the NF- B is similar to that in Eda, Edar, and Edaradd mutants pathway, we examined Notch1/2 expression in c I B N (Pispa et al., 1999; Tucker et al., 2000; Headon et al., mice. Both Notch1 and Notch2 were retained in incisor 2001; Laurikkala et al., 2001). The expression of I B epithelium of c I B N mice (Figures 7O 7R), suggesting and LacZ expression in (Ig ) 3 conalacz mice were found that Notch signaling is independent of the NF- B in enamel knot precursor cells similar to Edar and pathway. Edaradd in developing tooth germs. Moreover, similar changes in the expression of enamel knot marker genes Discussion were observed in the Ikk mutant and in c I B N mice molar tooth germs as have been previously described NF- B Pathway in Molar Tooth Development in Eda, Edar, and Edaradd mutants (Pispa et al., 1999; TNFs can activate the NF- B pathway, and NF- B tran- Tucker et al., 2000; Laurikkala et al., 2001). Taken together scription factors are essential not only for inflammation with previous reports indicating that Ikk (NEMO),

6 Developmental Cell 224 Figure 6. Expression of Epithelial and Mesenchymal Marker Gene Expression Early (E10.5 E13.5; A P ) to late stage (Newborn; Q R ) tooth development in Ikk mutants. In situ hybridization on frontal sections shows expression of marker genes in oral epithelium (A J, Q, and Q ) and mesenchyme (K P, R, and R ) of mutant and wild-type or heterozygous littermates. Arrowheads show gene expression in (Q) (R ). The probes used are indicated on the panels. All these genes except Shh and Wnt7b show a normal pattern of expression in Ikk mutants. The expansion of Shh expression was shown in mutant presumptive incisor epithelium (arrowheads in [G] and [G ]). The domain of mutant non- Wnt7b expressing epithelium (between arrows in [H ]) was expanded compared to wild-type (between arrows in [H]). Tooth germ epithelium is outlined in red. encoded by a X-linked gene in mice and humans, shows Ikk. The Ikk expression pattern is similar to Ikk, and a similar phenotype to ectodermal dysplasia including Ikk expression is also observed in tooth germs. It is thus abnormal teeth, these results show that NF- B activa- conceivable that NF- B activity was partially rescued in tion lies downstream of Eda signaling to mediate molar tooth cusp morphogenesis in Ikk mutant mice. cusp morphogenesis (Makris et al., 2000; Schmidt-Sup- Activation of Ikk requires phosphorylation of serines prian et al., 2000; Zonana et al., 2000; Döffinger et al., in its activation loop (Delhase et al., 1999; Mercurio et 2001). al., 1997). The same serines that are present in Ikk, in The molar phenotype of c I B N mice was found to be principle, should be required for its activation. However, more severe than the Ikk mutant. It has been shown defects in Ikk or NF- B activation were not observed that rescued NF- B activity is retained in embryonic in embryonic fibroblasts from Ikk AA/AA mice, whereas fibroblasts of Ikk mutants (Hu et al., 1999; Li et al., abnormal mammary gland development was found to 1999; Takeda et al., 1999), while it is inhibited in c I B N be due to defects in the NF- B pathway (Cao et al., mice (Schmidt-Ullrich et al., 2001). In Ikk mutants, 2001). Significantly, however, the Ikk AA/AA mutant does NF- B activity is believed to be compensated for by not interfere with NF- B activation by TNF in mammary

7 A Dual Role for Ikk in Tooth Development 225 Figure 7. Notch 1/2 Signaling in Tooth Development of Ikk Mutant Mice In situ hybridization on frontal sections shows expression of genes in the developing jaws of embryos at E11 (A and B), E13 (C, D, F, G, I N), and E14 (E, H, O R) of wild-type (A, C, F, I, L, O, and Q), c I B N mice ( N) (P and R), and Ikk mutant (B, D, E, G, H, J, K, M, and N). At E11, Notch1 expression was not detected in mutant presumptive incisor epithelium (arrowheads in [B]). Expression of Notch1 and Notch2 was downregulated in mutant incisors at E13 (D and G). Both Notch1 and Notch2 were not expressed in mutant incisors at E14 (E and H) that had a similar thickness of epithelium as wild-types at E13. Notch1 expression was also downregulated in mutant molars (J), whereas Notch2 expression was maintained in mutant molars (M). Notch1 and Notch2 were expressed in incisor of c I B N mice (P and R). Tooth germ epithelium is outlined in red. (S) A scheme for the putative dual Ikk pathway in tooth development. 1999). This was supported by the observation that incisor development was normal in Ikk AA/AA mice which lack the skin phenotype of Ikk mutant mice, and thus the kinase activity of Ikk is not needed for incisor devel- opment. Following an in situ hybridization screen of genes ex- pressed during tooth invagination, we identified abnormal expression of Notch1 and Notch2 in the incisor tooth germs of Ikk mutant embryos. Notch1 and Notch2 expression was downregulated in Ikk mutant incisors. Downregulation of Notch1/2 coincided with the epithelial thickening stage of incisor development. The maintenance of Notch2 expression during molar invagination in Ikk mutant embryos may suggest a reason for the normal invagination of molar epithelium in these mutants. Moreover, Notch1/2 were also downregulated in whisker follicles in Ikk mutant embryos (data not shown). These results reveal that Notch signaling may be involved in an NF- B-independent Ikk pathway. During the initial stages of incisor development, the developing incisors rotate antero-posteriorly following the forma- tion of the epithelial bud. The Notch pathway has been suggested to be involved in this incisor rotation (Muc- chielli and Mitsiadis, 2000; Pouyet and Mitsiadis, 2000). These results may also support the hypothesis that Notch is involved in determining of direction of ectodermal epithelium growth (Figure 7). Expansion of Shh expression was observed in developing incisors of Ikk mutants. It has been reported that there is crosstalk between the Notch and Shh signaling cascades in neuronal cells. Since Wnt7b has been shown to repress Shh in embryonic oral epithelium, it is conceivable that the observed expansion of Shh ex- pression in incisor epithelium in Ikk mutants results epithelial cells. These suggested that activation of NF- B and Ikk is cell-type-specific during ectodermal appendage development. Consistent with this, normal molar cusp formation was observed in Ikk AA/AA mice. Since initiation of tooth development comes from the epithelium (in contrast to mammary gland development where initiation signals originate in the mesenchyme), this suggests that in common with mammary gland development in these mice, NF- B activity is preserved in epithelial cells. NF- B Independence of Ikk in Early Incisor Development In Ikk mutant embryos, the incisor epithelium evaginates into the developing oral cavity rather than invagi- nating into underlying mesenchyme, and a similar phe- notype was also observed in hair follicles. This suggests that Ikk is an essential gene that directs the direction of epithelial ingrowth during development of skin ap- pendages. Normal invagination of incisor tooth germ epithelium was observed following downregulation of NF-kB activity in c I B N mice. These results suggest that epithelial invagination of incisors requires Ikk to function via an NF- B-independent pathway. One such role for Ikk has been proposed in skin development involving unknown soluble factors (Hu et al., 2001). Epi- thelial evagination was also observed in whisker follicles, suggesting this same pathway may also be in- volved in incisor tooth development. Interestingly, c I B N mice showed that whereas NF- B is necessary for the development of ectodermal appendages via the EdaA1/ EdaR pathway (Schmidt-Ullrich et al., 2001), skin proliferation and differentiation are independent of the NF- B pathway (Hu et al., 1999; Li et al., 1999; Takeda et al.,

8 Developmental Cell 226 from loss of Wnt7b (Sarkar et al., 2000). Shh signaling Received: October 10, 2003 has been implicated in establishing the polarity of the Revised: November 26, 2003 Accepted: December 1, 2003 floorplate, neural tube, somites, and limbs (reviewed by Published: February 9, 2004 Hammerschmidt et al., 1997; and references therein). The precise position of Notch in relation to Shh and Wnt7b cannot be fully determined since although References Notch1/2 expression is clearly reduced as early as Angerer, L.M., and Angerer, R.C. (1992). In situ hybridisation to cellu- E10.5, there is also a small, albeit weak, expansion of lar RNA with radiolabelled RNA probes. In In Situ Hybridisation: A Shh expression at this time, which becomes much more Practical Approach, D.G. Wilkinson, ed. (Oxford: Oxford University obvious by E13. However, the specific changes in ex- Press), pp pression of these genes in mutant incisor epithelium Cao, Y., Bonizzi, G., Seagroves, T.N., Greten, F.R., Johnson, R., indicate that a Notch-Wnt-Shh pathway interaction is Schmidt, E.V., and Karin, M. (2001). IKK provides an essential link involved in determining the direction of tooth (and between RANK signaling and cyclin D1 expression during mammary whisker follicle) epithelial growth during invagination. gland development. Cell 107, Cobourne, M.T., and Sharpe, P.T. (2003). Tooth and jaw: molecular Experimental Procedures mechanisms of patterning in the first branchial arch. Arch. Oral Biol. 48, Production and Analysis of Transgenic Mice Delhase, M., Hayakawa, M., Chen, Y., and Karin, M. (1999). Positive Ikk mutant mice were produced as described by Hu et al. (1999). and negative regulation of I B kinase activity through IKK subunit (Ig ) 3 conalacz mice and c I B N mice were produced as described phosphorylation. Science 284, by Schmidt-Ullrich et al. (1996, 2001, respectively). Ikk AA/AA mice were produced as described by Cao et al. (2001). Day E0 was taken Döffinger, R., Smahi, A., Bessia, C., Geissmann, F., Feinberg, J., to be midnight prior to finding a vaginal plug. To accurately assess Durandy, A., Bodemer, C., Kenwrick, S., Dupuis-Girod, S., Blanche, the age of embryos, somite pairs were counted and the stage connodeficiency S., et al. (2001). X-linked anhidrotic ectodermal dysplasia with immu- firmed using morphological criteria, e.g., relative sizes of maxillary is caused by impaired NF- B signaling. Nat. Genet. and mandibular primodia, extent of nasal placode invagination, and 27, the size of limb buds. Embryos were harvested at the appropriate Ferguson, B.M., Brockdorff, N., Formstone, E., Ngyuen, T., Krontime and genotyped using PCR and Southern blot analysis of genomiller, J.E., and Zonana, J. (1997). Cloning of Tabby, the murine mic DNA extracted from unused embryonic or extraembryonic tishomolog of the human EDA gene: evidence for a membrane-associsue. PCR assays and Southern blot hybridization were carried out. ated protein with a short collagenous domain. Hum. Mol. Genet. c I B N mice, Ikk mutant, and wild-type mice heads from E10 to 6, newborn were fixed in 4% paraformaldehyde (PFA), wax embedded, and serially sectioned at 7 m. Sections were split over five to ten Franzoso, G., Carlson, L., Xing, L., Poljak, L., Shores, E.W., Brown, slides and prepared for histology or radioactive in situ hybridization. K.D., Leonardi, A., Tran, T., Boyce, B.F., and Siebenlist, U. (1997). Decalcification using 0.5 M EDTA (ph 7.6) was performed after fixa- Requirement for NF- B in osteoclast and B-cell development. Genes tion of newborn mice. Dev. 11, Hammerschmidt, M., Brook, A., and McMahon, A.P. (1997). The -Galactosidase Staining world according to hedgehog. Trends Genet. 13, Embryo heads of (Ig ) 3 conalacz mice were fixed for min at Headon, D.J., Emmal, S.A., Ferguson, B.M., Tucker, A.S., Justice, 4 C in 1% formaldehyde and 0.2% glutaraldehyde in PBS. X-gal M.J., Sharpe, P.T., Zonana, J., and Overbeek, P.A. (2001). Gene staining was performed at 30 C in 4 mm potassium ferrocyanide, defect in ectodermal dysplasia implicates a death domain adapter 4 mm potassium ferricyanide, 2 mm MgCl 2, and 400 g/ml X-gal in development. Nature 414, (4-chloro-5-bromo-3-indolyl- -galactoside, Sigma) in PBS for 16 hr. Afterwards, embryo heads were washed several times in PBS and Hoffmann, A., Levchenko, A., Scott, M.L., and Baltimore, D. (2002). postfixed in the fixing solution mentioned above. The embryo heads The I B-NF- B signaling module: temporal control and selective were frozen and embedded in Tissue-Tek OCT compound (BHD). gene activation. Science 298, m sections were cut using a cryostat. Hu, Y., Baud, V., Delhase, M., Zhang, P., Deerinck, T., Ellisman, M., Johnson, R., and Karin, M. (1999). Abnormal morphogenesis but In Situ Hybridization intact IKK activation in mice lacking the IKK subunit of I B kinase. Radioactive section in situ hybridization using 35 S-UTP radiolabeled Science 284, riboprobes was carried out according to Angerer and Angerer (1992) with modifications as described by Tucker et al. (1998b). The radio- Hu, Y., Baud, V., Oga, T., Kim, K.I., Yoshida, K., and Karin, M. (2001). active antisense probes were generated from mouse cdna clones IKK controls formation of the epidermis independently of NF- B. that were gifts from several laboratories: Activin A (A.J.M. van den Nature 410, Eijnden-van Raaij), Amelogenin (M.L. Snead), Barx1 (J.-F. Brunet), Iotsova, V., Caamano, J., Loy, J., Yang, Y., Lewin, A., and Bravo, R. Bmp4 (B. Hogan), Cbfa1 (G. Karsenty), DSPP (M. MacDougall), Dlx2 (1997). Osteopetrosis in mice lacking NF- B1 and NF- B2. Nat. Med. (J. Rubenstein), I B (P. Schreider), IKK (S. Yamaoka), Lef1 (R. 3, Grosschedl), Lunatic Fringe (I. Thesleff), Mfz6 (Y. Wang), Notch1 Israël, A. (2000). The IKK complex: an integrator of all signals that and Notch2 (T.A. Mitsiadis), Pax9 (R. Balling), Pitx2 (J.C. Murray), activate NF- B? Trends Cell Biol. 10, Ptc1 (M. Scott), Serrate1 (D. Horowicz), Shh (A. McMahon), Wnt7b (T. Dale), and Wnt10b (G.M. Shackleford). For making a Ikk probe, Karin, M., and Delhase, M. (2000). The I B kinase (IKK) and NF- B: key RT-PCR of total RNA from E12.5 embryo bodies with primers a and elements of proinflammatory signalling. Semin. Immunol. 12, b, a 5 -ATGTCCTCAGCGGGTGTCG, and the reverse primer b 5 - Kere, J., Srivastava, A.K., Montonen, O., Zonana, J., Thomas, N., CAACGGTCACGGTGTACTTC was carried out. RT-PCR products Ferguson, B., Munoz, F., Morgan, D., Clarke, A., Baybayan, P., et were cloned into pcr II-TOPO vector (Invitrogen) and subjected to al. (1996). X-linked anhidrotic (hypohidrotic) ectodermal dysplasia automated DNA sequence analysis. is caused by mutation in a novel transmembrane protein. Nat. Genet. 13, Acknowledgments Krappmann, D., Wulczyn, F.G., and Scheidereit, C. (1996). Different We thank Abigail S. Tucker for critically reading the manuscript. mechanisms control signal-induced degradation and basal turnover A.O. is supported by Showa University, Tokyo, and Odontis Ltd. of the NF- B inhibitor I B alpha in vivo. EMBO J. 15, This work was supported by the Wellcome Trust and MRC. Kumar, A., Eby, M.T., Sinha, S., Jasmin, A., and Chaudhary, P.M.

9 A Dual Role for Ikk in Tooth Development 227 (2001). The ectodermal dysplasia receptor activates the nuclear fac- Tucker, A.S., Matthews, K.L., and Sharpe, P.T. (1998a). Transformation of tooth type induced by inhibition of BMP signaling. Science tor- B, JNK, and cell death pathways and binds to ectodysplasin A. J. Biol. Chem. 276, , Laurikkala, J., Mikkola, M., Mustonen, T., Aberg, T., Koppinen, P., Tucker, A.S., Al Khamis, A., and Sharpe, P.T. (1998b). Interactions Pispa, J., Nieminen, P., Galceran, J., Grosschedl, R., and Thesleff, between Bmp-4 and Msx-1 act to restrict gene expression to odon- I. (2001). TNF signaling via the ligand-receptor pair ectodysplasin togenic mesenchyme. Dev. Dyn. 212, and edar controls the function of epithelial signaling centers and is Tucker, A.S., Headon, D.J., Schneider, P., Ferguson, B.M., Overregulated by Wnt and activin during tooth organogenesis. Dev. Biol. beek, P., Tschopp, J., and Sharpe, P.T. (2000). Edar/Eda interactions 229, regulate enamel knot formation in tooth morphogenesis. Development 127, Li, Q., Lu, Q., Hwang, J.Y., Buscher, D., Lee, K.F., Izpisua-Belmonte, J.C., and Verma, I.M. (1999). IKK1-deficient mice exhibit abnormal Yan, M., Wang, L.C., Hymowitz, S.G., Schilbach, S., Lee, J., Goddard, development of skin and skeleton. Genes Dev. 13, A., de Vos, A.M., Gao, W.Q., and Dixit, V.M. (2000). Two-amino acid molecular switch in an epithelial morphogen that regulates binding Makris, C., Godfrey, V.L., Krahn-Senftleben, G., Takahashi, T., Robto two distinct receptors. Science 290, erts, J.L., Schwarz, T., Feng, L., Johnson, R.S., and Karin, M. (2000). Female mice heterozygous for IKK /NEMO deficiencies develop a Zonana, J., Elder, M.E., Schneider, L.C., Orlow, S.J., Moss, C., Godermatopathy similar to the human X-linked disorder incontinentia labi, M., Shapira, S.K., Farndon, P.A., Wara, D.W., Emmal, S.A., pigmenti. Mol. Cell 5, and Ferguson, B.M. (2000). A novel X-linked disorder of immune deficiency and hypohidrotic ectodermal dysplasia is allelic to incon- Mercurio, F., Zhu, H., Murray, B.W., Shevchenko, A., Bennett, B.L., tinentia pigmenti and due to mutations in IKK- (NEMO). Am. J. Li, J., Young, D.B., Barbosa, M., Mann, M., Manning, A., and Rao, Hum. Genet. 67, A. (1997). IKK-1 and IKK-2: cytokine-activated I B kinases essential for NF- B activation. Science 278, Monreal, A.W., Ferguson, B.M., Headon, D.J., Street, S.L., Overbeek, P.A., and Zonana, J. (1999). Mutations in the human homologue of mouse dl cause autosomal recessive and dominant hypohidrotic ectodermal dysplasia. Nat. Genet. 22, Mucchielli, M.L., and Mitsiadis, T.A. (2000). Correlation of asymmetric Notch2 expression and mouse incisor rotation. Mech. Dev. 91, Pispa, J., Jung, H.S., Jernvall, J., Kettunen, P., Mustonen, T., Tabata, M.J., Kere, J., and Thesleff, I. (1999). Cusp patterning defect in Tabby mouse teeth and its partial rescue by FGF. Dev. Biol. 216, Pouyet, L., and Mitsiadis, T.A. (2000). Dynamic Lunatic fringe expression is correlated with boundaries formation in developing mouse teeth. Mech. Dev. 91, Sarkar, L., Cobourne, M., Naylor, S., Smalley, M., Dale, T., and Sharpe, P.T. (2000). Wnt/Shh interactions regulate ectodermal boundary formation during mammalian tooth development. Proc. Natl. Acad. Sci. USA 97, Schmidt-Supprian, M., Bloch, W., Courtois, G., Addicks, K., Israël, A., Rajewsky, K., and Pasparakis, M. (2000). NEMO/IKK -deficient mice model incontinentia pigmenti. Mol. Cell 5, Schmidt-Ullrich, R., Memet, S., Lilienbaum, A., Feuillard, J., Raphael, M., and Israël, A. (1996). NF- B activity in transgenic mice: developmental regulation and tissue specificity. Development 122, Schmidt-Ullrich, R., Aebischer, T., Hulsken, J., Birchmeier, W., Klemm, U., and Scheidereit, C. (2001). Requirement of NF- B/Rel for the development of hair follicles and other epidermal appendices. Development 128, Srivastava, A.K., Pispa, J., Hartung, A.J., Du, Y., Ezer, S., Jenks, T., Shimada, T., Pekkanen, M., Mikkola, M.L., Ko, M.S., et al. (1997). The Tabby phenotype is caused by mutation in a mouse homologue of the EDA gene that reveals novel mouse and human exons and encodes a protein (ectodysplasin-a) with collagenous domains. Proc. Natl. Acad. Sci. USA 94, Takeda, K., Takeuchi, O., Tsujimura, T., Itami, S., Adachi, O., Kawai, T., Sanjo, H., Yoshikawa, K., Terada, N., and Akira, S. (1999). Limb and skin abnormalities in mice lacking IKK. Science 284, Thesleff, I., and Sharpe, P. (1997). Signalling networks regulating dental development. Mech. Dev. 67, 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, Thomas, B.L., Tucker, A.S., Qui, M., Ferguson, C.A., Hardcastle, Z., Rubenstein, J.L., and Sharpe, P.T. (1997). Role of Dlx-1 and Dlx-2 genes in patterning of the murine dentition. Development 124,