A Shh-Foxf-Fgf18-Shh Molecular Circuit Regulating Palate Development

Abstract

Cleft palate is among the most common birth defects in humans. Previous studies have shown that Shh signaling plays critical roles in palate development and regulates expression of several members of the forkhead-box (Fox) family transcription factors, including Foxf1 and Foxf2, in the facial primordia. Although cleft palate has been reported in mice deficient in Foxf2, whether Foxf2 plays an intrinsic role in and how Foxf2 regulates palate development remain to be elucidated. Using Cre/loxP-mediated tissue-specific gene inactivation in mice, we show that Foxf2 is required in the neural crest-derived palatal mesenchyme for normal palatogenesis. We found that Foxf2 mutant embryos exhibit altered patterns of expression of Shh, Ptch1, and Shox2 in the developing palatal shelves. Through RNA-seq analysis, we identified over 150 genes whose expression was significantly up- or down-regulated in the palatal mesenchyme in Foxf2-/- mutant embryos in comparison with control littermates. Whole mount in situ hybridization analysis revealed that the Foxf2 mutant embryos exhibit strikingly corresponding patterns of ectopic Fgf18 expression in the palatal mesenchyme and concomitant loss of Shh expression in the palatal epithelium in specific subdomains of the palatal shelves that correlate with where Foxf2, but not Foxf1, is expressed during normal palatogenesis. Furthermore, tissue specific inactivation of both Foxf1 and Foxf2 in the early neural crest cells resulted in ectopic activation of Fgf18 expression throughout the palatal mesenchyme and dramatic loss of Shh expression throughout the palatal epithelium. Addition of exogenous Fgf18 protein to cultured palatal explants inhibited Shh expression in the palatal epithelium. Together, these data reveal a novel Shh-Foxf-Fgf18-Shh circuit in the palate development molecular network, in which Foxf1 and Foxf2 regulate palatal shelf growth downstream of Shh signaling, at least in part, by repressing Fgf18 expression in the palatal mesenchyme to ensure maintenance of Shh expression in the palatal epithelium.

Fig 1. Histological analysis of palate developmental defects in Foxf2 mutant mouse embryos.

Representative frontal sections of wildtype (A, E, I), Foxf2^-/- (B, F, J), Foxf2^c/-Wnt1-Cre (C, G, K), and Foxf2^c/-Osr2^IresCre/+ (D, H, L) embryos at E13.5 (A-D), E14.5 (E-H) and E16.5 (I-L). p, palatal shelf; t, tongue.

Fig 2. Analysis of cell proliferation in the developing palate in E13.5 Foxf2^-/- and wildtype embryos.

(A-F) Representative images of sections through the anterior (A, B), middle (C, D) and posterior (E, F) regions of the palate in wildtype (A, C, E) and Foxf2^-/- mutant (B, D, F) embryos showing distribution of immunostained BrdU-labeled nuclei (green). Sections were counterstained with DAPI (blue). White line divides the palatal shelf into oral and nasal sides for cell counts. (G) The percentage of BrdU-labeled cells in the E13.5 palatal mesenchyme (***p<0.001). AO, oral half of the anterior region; AN, nasal half of the anterior region; MO, Oral half of the middle region; MN, nasal half of the middle region; PO, oral half of the posterior region; PN, nasal half of the posterior region.

Fig 3. Comparison of patterns of Shh, Ptch1, Shox2, and Barx1 mRNA expression in Foxf2^-/- mutant and wildtype control embryos.

(A, B, E, F) Expression patterns of Shh mRNAs in the developing palatal shelves in wildtype (A, E) and Foxf2^-/- mutant (B, F) embryos at E13.5 (A, B) and E14.5 (E, F). (C, D, G, H) Expression patterns of Ptch1 mRNAs in the developing palatal shelves in wildtype (C, G) and Foxf2^-/- mutant (D, H) embryos at E13.5 (C, D) and E14.5 (G, H). In specific anterior (arrows) and posterior (arrow heads) subdomains of the palatal shelves, Shh and Ptch1 expression is dramatically downregulated in Foxf2^-/- mutant embryos. R1-R7 mark the individual ruga in the order of their formation. (I-L) Expression patterns of Shox2 (I, J) and Barx1 (K, L) mRNAs in the palatal shelves in wildtype (I, K) and Foxf2^-/- mutant (J, L) embryos at E13.5. White dashes demarcate the palatal shelf on the right side. Arrows in I and J point to the anterior region of the palatal shelves, where the expression of Shox2 mRNAs is reduced in Foxf2^-/- mutant embryos.

Fig 4. Differential gene expression analysis of palatal mesenchyme cell in control and Foxf2 mutant embryos by using RNAseq and real time RT-PCR assays.

(A, B) The patterns of RFP expression in the Osr2^RFP/+ (A) and Foxf2^-/-Osr2^RFP/+ (B) palatal shelves are comparable. White dashes demarcate the palatal shelf on the right side. (C) Real time RT-PCR analysis of the levels of expression of Foxf2, Osr2, Fgf18, Shox2, Barx1, and Ptch1 mRNAs in E13.5 palatal mesenchyme cells in control and Foxf2^-/-Osr2^RFP/+ embryos (**p<0.01, ***p<0.001).

Fig 5. Comparison of expression of Fgf18 mRNAs in the palatal shelves in wildtype and Foxf2^-/- mutant embryos.

(A, B) Whole-mount in situ hybridization detection of Fgf18 mRNAs in the developing palatal shelves in wildtype (A) and Foxf2^-/- mutant (B) embryos at E13.5. White dashes demarcate the palatal shelf on the right side. Note that Fgf18 is ectopically expressed in specific anterior (arrow) and posterior (arrow head) subdomains of Foxf2^-/- mutant palatal shelves. (C-F) Frontal sections showing expression of Fgf18 mRNAs in the anterior (C, D) and posterior (E, F) regions of the developing palatal shelves in wildtype (C, E) and Foxf2^-/- mutant (D, F) embryos at E13.5. p, palatal shelf; t, tongue.

Fig 6. The patterns of Foxf1 and Foxf2 expression during palate development.

(A-D) Whole-mount in situ hybridization detection of Foxf1 (A, C) and Foxf2 (B, D) mRNAs in the developing palate shelves in E12.5 (A, B) and E13.5 (C, D) mouse embryos. White dashes demarcate the palatal shelf on the right side. Yellow dashes mark the molar tooth germ. (E-H) Immunofluorescent staining of Foxf1 (green) and RFP (red) on sections from the anterior to posterior regions of the palatal shelves in E13.5 Osr2^RFP/+ embryos. (I-L) Immunofluorescent staining of Foxf2 (green) and RFP (red) on sections in E13.5 Osr2^RFP/+ embryos. p, palatal shelf; t, tongue; l, lower molar; u, upper molar.

Fig 7. Comparison of expression of Fgf18 and Shh mRNAs in the palatal shelves in Foxf1^c/cFoxf2^c/c control and Foxf1^c/cFoxf2^c/cWnt1-Cre mutant embryos.

(A, B) Whole-mount in situ hybridization detection of Fgf18 mRNAs in the developing palatal shelves in Foxf1^c/cFoxf2^c/c control (A, C) and Foxf1^c/cFoxf2^c/cWnt1-Cre mutant (B, D) embryos at E12.5 (A, B) and E13.5 (C, D). White dashes demarcate the palatal shelf on the right side. Arrowhead (A, C) and arrow (B, D) point to the Fgf18 expression domains in the control and mutant palatal shelves, respectively. (E, F) Whole-mount in situ hybridization detection of Shh mRNAs in the developing palatal shelves in Foxf1^c/cFoxf2^c/c control (E) and Foxf1^c/cFoxf2^c/cWnt1-Cre mutant (F) embryos at E13.5. Note that expression of Shh is lost in Foxf1^c/cFoxf2^c/cWnt1-Cre mutant palatal shelves.

Fig 8. Regulation of Shh and Fgf18 expression during palate development.

(A) Whole-mount in situ detection of Shh mRNAs in wildtype embryonic palatal explants treated with BSA beads (left side) and Fgf18 beads (right side). (B) The patterns of GFP expression in Shh^GFP/+ embryonic palatal explants treated with BSA beads (left side) and Fgf18 beads (right side). Note that Shh expression is inhibited by Fgf18 beads. (C) A model depicting the molecular regulation of Foxf1, Foxf2, Fgf18, and Shh expression in the developing palatal shelf. Regulation of Shh expression by Fgf10/Fgfr2b, Msx1/Bmp4, and Dlx5/Fgf7 pathways is summarized from References 8, 7, and 28, respectively.