Wnt/beta-catenin signaling directs multiple stages of tooth morphogenesis

Abstract

Wnt/beta-catenin signaling plays key roles in tooth development, but how this pathway intersects with the complex interplay of signaling factors regulating dental morphogenesis has been unclear. We demonstrate that Wnt/beta-catenin signaling is active at multiple stages of tooth development. Mutation of beta-catenin to a constitutively active form in oral epithelium causes formation of large, misshapen tooth buds and ectopic teeth, and expanded expression of signaling molecules important for tooth development. Conversely, expression of key morphogenetic regulators including Bmp4, Msx1, and Msx2 is downregulated in embryos expressing the secreted Wnt inhibitor Dkk1 which blocks signaling in epithelial and underlying mesenchymal cells. Similar phenotypes are observed in embryos lacking epithelial beta-catenin, demonstrating a requirement for Wnt signaling within the epithelium. Inducible Dkk1 expression after the bud stage causes formation of blunted molar cusps, downregulation of the enamel knot marker p21, and loss of restricted ectodin expression, revealing requirements for Wnt activity in maintaining secondary enamel knots. These data place Wnt/beta-catenin signaling upstream of key morphogenetic signaling pathways at multiple stages of tooth development and indicate that tight regulation of this pathway is essential both for patterning tooth development in the dental lamina, and for controlling the shape of individual teeth.

Fig. 1. Localization of Wnt reporter gene expression, β-catenin and Wnt10b expression in developing teeth

Tissues were cryosectioned and then X-gal stained to reveal sites of Wnt reporter expression (blue) (A-C); sectioned prior to immunofluorescence staining for β-catenin (red) (D-E); or whole mount X-gal stained prior to paraffin sectioning (F-H). (A) Transverse section of E11.5 TCF/Lef-LacZ head showing reporter gene expression in molar tooth placodes (bracketed) and adjacent oral epithelium. (B) Higher magnification view of the area bracketed on the right in (A) showing LacZ expression in epithelial cells of the first molar. (C) Transverse section of E12.5 TCF/Lef-LacZ maxilla showing LacZ expression in molar epithelial cells. (D) Transverse section of E12.5 maxilla showing nuclear and cytoplasmic localization of β-catenin in epithelial cells and immediately underlying mesenchymal cells of the first left molar. Nuclei are counterstained with DAPI and appear blue. (E) Transverse section of E14 maxilla showing nuclear localization of β-catenin in epithelial cells and immediately underlying mesenchymal cells of the first left molar. Nuclei are counterstained with DAPI and appear blue. (F) Frontal section of E14.5 TOPGAL mandible showing reporter gene expression in the enamel knot of a first lower molar. (G) Frontal section of E16.5 TOPGAL mandible showing reporter gene expression in developing molar cusps. (H) Sagittal section of E18.5 BAT-gal mandible showing reporter gene expression in molar cusp cells (arrows). (I) E12.5 mandible subjected to whole mount in situ hybridization with digoxygenin-labeled probe for Wnt10b. Note hybridization signals (purple) in incisor placodes (blue arrows) and in the developing molar placodes (black arrows). (J-L) In situ hybridization of frontally sectioned mandibles with digoxygenin-labeled probe for Wnt10b. (J) E14.5 first lower molar showing hybridization to the enamel knot. (K) E16.5 first lower molar showing hybridization to developing cusp epithelium. (L) E17.5 first lower molar showing expression in cusp epithelia. Sections in (A-C,F,G) were counterstained with eosin, section (H) with hematoxylin, and sections in (J-L) with methyl green. A dashed line indicates the epithelial-mesenchymal boundary in (B-F and J).

Fig. 2. Mutation of epithelial β-catenin to a stabilized form causes abnormal dental invaginations and ectopic tooth formation

Control littermate (A,C,C',E,G,I,K,M,O,Q,S,U,W,Y) and K14-Cre Ctnnb1^(Ex3)fl/+ (Act β-catenin) mutant(B,D,D',F,H,J,L,N,P,R,T,V,X,Z) oral cavity from embryos at E14.5 (A-F), E16.5 (G,H), E17.5 (I-R) and E18.5 (S-Z) paraffin embedded and sectioned frontally (A-F) or sagitally (G-Z) and subjected to hematoxylin and eosin staining (A,B,I,J,S,T,W,X); immunofluorescence with anti-β-catenin (red) (C,D); X-gal staining (G,H) (blue); or in situ hybridization with digoxygenin-labeled probe for Ctnnb1 (C',D'), Pitx2 (E,F,Q,R), Shh (K,L), Wnt10b (M,N), ectodin (O,P), Ambn (U,V) or Dspp (Y,Z) (purple or brown signals). Sections in (G,H) were taken from control and mutant embryos that also carried the TOPGAL Wnt reporter gene and were counterstained with eosin (pink). Size bar in (B) applies to (A,B); size bar in (D) applies to (C,D); size bar in (R) applies to (C'-R); size bar in (Z) applies to (S-Z).

Fig. 3. Ectopic expression of Dkk1 or epithelial deletion of β-catenin blocks Wnt pathway activity and tooth development

(A-D) X-gal stained mandible whole mounts from E11.5 (A,B) and E12.5 (C,D) littermate control TOPGAL (A,C) and K5-rtTA tetO-Dkk1 TOPGAL (B,D) (Dkk1-expressing) embryos treated with doxycycline from E0.5. (E-H) Transverse sectioned X-gal stained heads from E12.5 control TOPGAL (E,G) and K5-rtTA tetO-Dkk1 TOPGAL (F,H) (Dkk1-expressing) embryos. Note absence of TOPGAL staining and arrest of molar tooth development in the Dkk1-expressing embryo (F,H). Panels G,H shower higher magnification photographs of the boxed regions in (E,F). (I,J) Frontally sectioned oral cavities from E12.5 control (I) and K5-rtTA tetO-Dkk1 (Dkk1-expressing) (J) embryos subjected to immunofluorescence for β-catenin (red). Sections are DAPI counterstained (blue). Note nuclear and cytoplasmic as well as membrane localization of β-catenin in dental mesenchymal and some epithelial cells in the control (I) and predominantly membrane localization in the Dkk1-expressing molar (J). (K-P) Frontally sectioned oral cavities from control (K,M,O) and littermate K5-rtTA tetO-Dkk1 (Dkk1-expressing) (L,N,P) embryos subjected to hematoxylin and eosin staining at E12.5 (K,L), E13.5 (M,N) and E15.5 (O,P). Size bar in (J) applies to (I,J); size bar in (P) applies to (K-P). (Q-T) Frontally sectioned oral cavities from E12.5 (Q, R) and E13.5 (S, T) control (Q,S) and K14-Cre Ctnnb1^fl/fl (β-catenin loss of function (LOF)) (R, T) embryos subjected to immunofluorescence for β-catenin (red). Sections are DAPI counterstained (blue). Note mosaic depletion of β-catenin at E12.5 (R, yellow arrow and white arrowhead) and more efficient depletion at E13.5 (T). (U,V) X-gal stained mandible whole mounts from E13.5 littermate control TOPGAL (U) and K14-Cre Ctnnb1^fl/fl TOPGAL (_β-catenin null) (V) embryos. (W-Z) Frontally sectioned oral cavities from control (W,W',Y) and littermate β-catenin LOF mutant (X,X',Z) embryos subjected to hematoxylin and eosin staining at E13.5 (W-X') and E15.5 (Y,Z). Size bar in (T) applies to (Q-T); size bar in (Z) applies to (W-Z).

Fig. 4. Wnt inhibition blocks expression of multiple developmental regulators

(A) Mandibles dissected from littermate control (left panels) and K5-rtTA tetO-Dkk1 (Dkk1-expressing) (right panels) embryos doxycycline treated from E0.5, sacrificed at E12.5 or E13.5, and subjected to in situ hybridization with digoxygenin-labeled probes for Bmp4, Msx1, Msx2, Shh, Ptc2, and Pitx2. Arrows indicate positive signals (purple). (B) X-gal stained mandibles dissected from E14.5 littermate control NFκB-GAL and K5-rtTA tetO-Dkk1 NFκ-GAL (Dkk1-expressing) embryos doxycycline treated from E0.5. Note that NFκ-GAL expression in the incisor (yellow arrows) and molar (blue arrows) regions persists in Dkk1-expressing mandibles. (C) Transverse sections of oral cavities from littermate control and K5-rtTA tetO-Dkk1 (Dkk1-expressing) embryos doxycycline treated from E0.5, sacrificed at the stages indicated, and subjected to in situ hybridization with digoxygenin-labeled probes for Bmp4, Msx1, Msx2, Shh, Ptc2, Lef1, Wnt10b, Pitx2, Pax9 and Eda. Arrows indicate positive signals (brown/purple). Sections are counterstained with methyl green.

Fig. 5. Epithelial deletion of β-catenin affects expression of regulatory genes in dental epithelium and mesenchyme

(A-B) Mandibles dissected from littermate control (A) and K14-Cre; Ctnnb1^fl/fl (β-catenin LOF) (B) E13.5 embryos and subjected to in situ hybridization with digoxygenin-labeled probe for Bmp4. (C-H) Transverse sections of oral cavities from littermate control (C,E,G) and K14-Cre Ctnnb1^fl/fl (β-catenin LOF) (D,F,H) E13.5 embryos subjected to in situ hybridization with the probes indicated. (I, J). Mandibles dissected from littermate control (I) and K14-Cre Ctnnb1^fl/fl (β-catenin LOF) (J) E13.5 embryos and subjected to in situ hybridization with digoxygenin-labeled probe for Shh. (K-P) Transverse sections of oral cavities from littermate control (K,M,O) and K14-Cre Ctnnb1^fl/fl (β-catenin LOF) (L,N,P) E13.5 embryos subjected to in situ hybridization with the probes indicated. Arrows indicate positive signals (brown or purple). A dashed line indicates the epithelial-mesenchymal boundary in (M,N). Sections are counterstained with methyl green. Size bar in (H) applies to (C-H); size bar in (L) applies to (K,L); size bar in (P) applies to (M-P).

Fig.6. Ectopic Dkk1 does not block the ability of exogenous BMP4 to induce expression of Msx1 and Msx2

(A-D) Representative samples of mandibles dissected from littermate control (A,C) and K5-rtTA tetO-Dkk1 (Dkk1-expressing) (B,D) embryos doxycycline treated from E0.5, sacrificed at E11.5, implanted with BMP4-coated beads, cultured in the presence of doxycycline for 24 hours, and subjected to in situ hybridization with digoxygenin-labeled probes for Msx1 (A,B) and Msx2 (C,D). Arrows indicate positive signals (purple). (E) Summary of Msx1 and Msx2 expression data from multiple experiments.

Fig. 7. Wnt/β-catenin signaling is required for normal development of molar cusps

(A) Right mandibles dissected at P21 from control littermate and K5-rtTA tetO-Dkk1 (Dkk1 DTG) mice doxycycline treated from E16. Note blunted cusp formation in Dkk1 DTG molar teeth compared with controls. (B,C) Right mandibles from P0.5 control littermate (B) and K5-rtTA tetO-Dkk1 (Dkk1 DTG) (C) mice doxycycline treated from E16.5, serially sectioned at 5μ-m in a sagittal plane from lingual to buccal side. Sections were stained with hematoxylin and eosin. The photographs shown were taken every 8 sections; each photo in each series is of a section 40μ-m from the previous photo. The images on the right represent stacked images of the serial sections, compiled using the Stack function of ImageJ software.

Fig. 8. Wnt/β-catenin signaling is required for the maintenance of secondary enamel knots

Sagittal sections of oral cavities from control littermate (A,C,E,G,I,K,M,O,Q,S) and K5-rtTA tetO-Dkk1 (Dkk1 DTG) (B,D,F,H,J,L,N,P,R,T) mice doxycycline treated from E16 and sacrificed at E17.75 (A,B,E,F,I,J,M,N,Q,R) or P0.5 (C,D,G,H,K,L,O,P,S,T) hybridized with digoxygenin-labeled probes for Edar (A-D), Eda (E-H), Bmp4 (I-L), p21 (M-P) and ectodin (Q-T). Positive signals (purple-brown) are indicated by arrows. In (O,P) black arrows indicate epithelial cells; blue arrows indicate developing odontoblasts. Size bar in (T) applies to (A-T).