Fig 1.
FoxO6 expression during mouse embryonic development.
A) Schematic diagram of FoxO6 knockout strategy. The targeting construct carries a LacZ reporter gene and a constitutively expressed neomycin resistance gene (Neo) embedded in two loxp sites. The entire FoxO6 coding region from start codon to stop codon is replaced by LacZ-Neo (ES cells obtained from KOMP biorepository). B-F) FoxO6 expression is shown by X-gal staining in FoxO6+/- mice. B) FoxO6 expression is not detectable at E10.5. C) At E12.5, FoxO6 expression is observed in the brain (frontal lobes, FL; cerebellum primordium, CP; trigeminal ganglion, TG; somites, S. D) At E14.5, FoxO6 expression is in the brain, craniofacial regions, also in the posterior maxilla (PMx) and posterior mandible (PMn). E) At E18.5, FoxO6 has strong expression in the craniofacial region, such as mandibular bone(MB), incisor (IN), molar (M) and palate (PL).
Fig 2.
The heads of FoxO6-/- mice are larger than those of wild type mice.
A,B) Alcian Blue/Alizarin Red staining of cartilage and bone skeletal preparations of FoxO6+/- mice and FoxO6-/- littermates. A) E18.5 head skeletal preparations show a larger skull and head in FoxO6-/- embryos. Moreover, a reduced ossification (red stain) of the interparietal bone (INT), exoccipital bone (EXO) and nasal bone (NB) were observed in E18.5 FoxO6-/- embryos compared to their heterozygous littermates. B) Skeletal preparations of P1 FoxO6+/- mice (left) and FoxO6-/- littermates (right) show that FoxO6-/- mice have an expanded head compared with control mice. The frontal bone (FB), parietal bone (PB) and occipital bone (OB) are slightly larger in FoxO6-/- mice. C) Size of head in 2 month-old FoxO6-/- mice compared to WT mice (N>5). The anterior region of the craniofacial complex continues to grow in the FoxO6-/- mice after weaning, until 2–3 months old (see yellow arrow), while the body size is normal. D) The submandibular gland was increased after 6 months in FoxO6-/- mice compared to WT. E) Results of brain MRI scans of 2 month-old control (FoxO6+/-) and FoxO6-/- mice, comparing the sizes of the different compartments of the brains of FoxO6+/- and FoxO6-/- mice. The percent increase in the brain regions are shown on top and the percent decrease on the bottom.
Fig 3.
The heads of FoxO6-/- mice are expanded in the anterior direction.
A-D) Microcomputed tomography (μCT) analysis of 2 month FoxO6-/- and WT mice heads. A) μCT of wild type and FoxO6-/- heads used for morphometric measurements. B-D) Morphometric measurements of bone length and percent (%) change between FoxO6-/- and WT mice (N = 3 for each). B) Measurements of the nasal length, facial length, mandibular length and mandibular height are all increased in the FoxO6-/- mouse head, N = 3. C) The frontal length, parietal length and skull length were also increased in the FoxO6-/-mouse head, N = 3. D) Midsagittal section measurements showed a slight increase in the cranial base angle, N = 3.
Fig 4.
FoxO6-/- mouse incisors are positioned anteriorly and distal to the molars compared to WT mice.
A-G) Microcomputed tomography (μCT) analysis of 1 month-old FoxO6-/- and WT mice heads, mandibles, lower incisors and molars. A) μCT maximum projection images of FoxO6-/- and WT heads used for analyses. B, C) Mineralization of the FoxO6-/- lower incisor resulting in detectable mineral density of enamel is shifted more anteriorly and distally relative to the first molars, the commonly used anatomical landmark for stages of enamel mineralization, compared to the WT incisor (pink arrows). D, E) Hemimandible seen in parasagittal plane through the molars showing less mineralization of enamel and dentine, indicative of an anteriorly extended delay of mineralization in the FoxO6-/- lower incisor at this position (pink arrows). F, G) Image of mandibles anterior to the plane through the distal root of the first molar in coronal plane showing that the FoxO6-/- lower incisor at this posterior position has thinner dentin and decreased enamel formation compared to the WT incisor (pink arrows).
Fig 5.
FoxO6 regulates Lats1 and pYAP expression.
A) Schematic illustration of the mouse lower incisor, showing the labial cervical loop (LaCL, stem-cell niche), inner enamel epithelium (IEE), outer enamel epithelium (OEE), and ameloblast layer derived from the LaCl cells. B-G) FoxO6 expression in the E18.5 lower incisor (B, C), molar (D, E) and oral epithelium (F, G). FoxO6 expression indicated by immunostaining using an X-gal antibody (green) of WT (B, D, F) and FoxO6+/- (C, E, G) mice (only FoxO6+/- mice express lacZ). Scale bars represent 100μm. H-K) Lats expression in FoxO6-/- vs WT mice. A series of sagittal sections of lower incisors from P1 animals were examined by immunofluorescence staining for Lats1 protein, using an Alexa-488 labeled antibody. I, K) Magnified views of the boxed regions in (H) and (J), respectively. The LaCL, dental epithelium and ameloblast layer is outlined. L-O) Expression of pYap. Less Yap is phosphorylated (activated) in incisors of FoxO6-/- vs WT mice. M, O) Magnified views of the boxed regions in (L) and (N), respectively. In all sections, staining was used to identify nuclei. Scale bars represents 100μm.
Fig 6.
GSEA plot showing the down regulation of the Hippo pathway in FoxO6-/- mandibular tissue.
A) All genes are ranked based on fold change between FoxO6 null and wild type mandible and maxilla tissue, with most up regulated genes (red color) in the left and most down regulated genes (blue color) in the right. Each black line represents a gene in the Hippo pathway. As shown in this figure, Hippo pathway genes tends to located in the right (down regulated) region. B) A volcano plot showing multiple genes regulated by FoxO6, both increased and decreased expression (4 biological replicates of WT and FoxO6 mutant embryos were combined to yield the bioinformatics data).
Fig 7.
In FoxO6-/- mice, the incisors are enlarged, the LaCL is expanded and the polarity of the dental epithelium is abnormal.
Sections of incisors from E16.5, E18.5, and P0 WT and FoxO6-/- mice stained with Hematoxylin and Eosin. A-B) E16.5 mandibles, with lower incisor (LI) framed by box to highlight differences in incisor size. C-D) Higher magnification of the boxed region of the lower incisors in A and B, showing that they are enlarged and that they have larger labial cervical loops (LaCLs) in FoxO6-/- embryos. E-F) In E18.5 mandibles, a longer lower incisor is still observed in FoxO6-/- embryos. G-H) Higher magnification of E and F (LaCL is outlined in black). I-J) P0 mandible. The lower incisor and LaCL are much larger in the FoxO6-/- mice. K-L) Higher magnification of the red box of I and J, revealing that the LaCL is larger in FoxO6-/- mice. M-N) Higher magnification of the ameloblast and odontoblast cell layers (blue boxed region in I, and J) show that the cells are not well polarized in FoxO6-/- mice compared to WT mice. O,P) Amelogenin staining in the lower incisor (LI) of FoxO6+/- P1 mice compared to FoxO6-/- littermates. Q,R) Higher magnification of the boxed area in O and P showing amelogenin in the ameloblast layer. Increased enamel (En) formation is observed in the FoxO6-/- mice. Scale bar represents 100μm. Abbreviations: Epi, dental epithelia; Mes, dental mesenchyme; LaCL, labial cervical loop. LiCL, lingual cervical loop; Am, Ameloblast; Od; Odontoblast; SI; stratum intermedium.
Fig 8.
In FoxO6-/- mice, the incisors exhibit increased cell proliferation.
A-B) Cell proliferation in lower incisors from WT or FoxO6-/- E18.5 embryos, as assessed by staining with Ki67 and DAPI. C-D) Higher magnification of the LaCL region from WT or FoxO6-/- embryos in A and B. The Ki67 positive cells are located mainly in the Transient Amplifying (TA) zone of the IEE and adjacent mesenchymal tissue. E) Quantitation of the Ki67-positive cells in the sections of lower incisors. The number of Ki67-positive cells in FoxO6-/- epithelial and mesenchymal tissue is increased compared to WT embryos. F-G) Cell proliferation in E17.5 WT and FoxO6-/- embryos, as assessed by BrdU injection (2 hours prior to sacrifice). H-I) Higher magnification of posterior or proximal lower region of incisor in WT and FoxO6-/- embryos in F and G, demonstrating increased epithelial and mesenchymal cell proliferation compared to the WT. J) Quantitation of the BrdU-positive cells in sections of lower incisors. Scale bar represents 100μm. Epi, epithelium; Mes, mesenchyme.
Fig 9.
FoxO6 activates Lats1/2, Runx2, Shh and the Hippo signaling pathway.
Genes identified as being differentially expressed in FoxO6-/- and WT mandibles in microarray screen were assessed. A) Relative expression of genes in mandible and maxilla tissue of E18.5 FoxO6-/- and WT (FoxO6+/+) was assessed by real-time PCR and normalized to expression of β-actin. Experiments were repeated at least three times each, from multiple biological samples. Error bars indicate S.E. *: p-value<0.01. B) Expression and phosphorylation of gene product in CHO cells over-expressing FoxO6. CHO cells were transfected with 5 μg empty vector (control), 5 μg pcDNA-FoxO6, or 10 μg pcDNA-FoxO6. Lysates were collected after 2 days for immunoblotting. C) Expression and phosphorylation of gene product in LS-8 (oral epithelial-like) cells over-expressing FoxO6 over-expression. 5 μg empty vector (control) or 5 μg pcDNA-FoxO6 were used for electroporation. Lysates were collected after 2 days for immunoblotting. D) Expression of FoxO6 and Lats1 in ET-16 cells (oral epithelial cells) subjected to FoxO6 knockdown. 10 μg shScramble DNA or 10 μg shFoxO6 DNA were transfected and cell lysates were collected after 2 days for RNA extraction and real-time PCR.
Fig 10.
FoxO6 directly binds to and activates the Lats1 promoter.
A) Schematic of the Lats1 1.7kb promoter, with the predicted FoxO6 binding motif (-2467bp). Primers were designed to flank the predicted FoxO6 binding site (-2362 to -2519 bp; ChIP primers) and to an upstream region that lacks a FoxO6 binding site (-4735 to -4913bp; Control primers). B) PCR products from ChIP assays involving immunoprecipitation of endogenous FoxO6 in LS-8 cells. PCR products were resolved in agarose gels. Input chromatin was used as a control to show the primer product; FoxO6 Ab (antibody) immunoprecipitated (IP) the chromatin containing the FoxO6 binding site (lane2); IgG alone did not IP the chromatin and the no template PCR reaction did not produce a band. Control primers to an upstream region of the Lats1 promoter did not detect an IP product with the FoxO6 Ab or the IgG control. C) Quantitation of the enrichment of binding by endogenous FoxO6 chromatin compared to IgG immunoprecipitated DNA. D) Activation of the Lats1 1.7kb promoter by FoxO6 using a Lats1 luciferase reporter in transfected LS-8 cells compared with empty vector. E) Activation of an Lats1 enhancer luciferase reporter whose expression is driven by a duplicated 80 bp DNA fragment derived from the Lats1 promoter region containing the FoxO6 binding site. This construct was transfected into LS-8 cells with or without FoxO6 expression plasmid. In parallel, a reporter with a mutated FoxO6 binding site (Lats1 80bp Mut enhancer) was transfected as control.
Fig 11.
Pitx2 activates FoxO6 expression.
A) Heat map from gene expression microarray showing FoxO6 is increased in Krt14-PITX2 overexpression mouse incisor dental epithelium compared to WT. DNA microarray analyses of five different E18.5 mouse dental epithelial tissue samples reveals an increase in FoxO6 expression in the PITX2 over-expression incisor. B) Real-time PCR demonstrates increased endogenous FoxO6 transcripts in LS-8 cells and MDPC cells that over-express PITX2 compared with empty vector. C, D) LS-8 cells (C) and MDPC cells (D) were infected with Pitx2 or scrambled lentivirus and probed for Pitx2 and FoxO6 expression by Western Blot. β-Tubulin served as a loading control.
Fig 12.
FoxO6 expression is directly activated by Pitx2.
A) Schematic of the FoxO6 4kb promoter with the Pitx2 binding motif (TAATCC) highlighted by red letters. Primers were designed to flank the predicted Pitx2 binding site (-1366 to -1541bp; ChIP primers) and an upstream region that lacks a Pitx2 binding site (-3525 to -3698bp; Control primers). B-C) Endogenous ChIP assay was performed in LS-8 cells. ChIP assay was performed as in Fig 6. D) Quantitation of the ChIP assays show a 40-fold enrichment of endogenous Pitx2 binding to the FoxO6 promoter compared to IgG IP. E) The FoxO6 promoter (4 Kb-5’ flanking region) was cloned into a luciferase vector and co-transfected with PITX2 or empty vector in LS-8 cells. Luciferase assays from four independent experiments were analyzed (N = 4). PITX2 activates the FoxO6 promoter at over 50-fold. F) A FoxO6 Pitx2 enhancer element (~70 bp) was cloned into a TK minimal promoter luciferase construct to test for enhancer activity. The FoxO6 enhancer construct was co-transfected with either PITX2 or empty vector. A similar FoxO6 enhancer reporter with the mutated Pitx2 binding motif was transfected in parallel as a control. All luciferase activities are normalized to β-galactosidase activity and shown as mean-fold activation compared with the FoxO6 enhancer plasmid co-transfected with empty CMV expression plasmid (SEM from five independent experiments).
Fig 13.
Model for FOXO6 regulation of Hippo signaling in controlling growth of the head and craniofacial features.
Mammalian Ste20 family kinases Mst1 and Mst2 form an active complex with Salvador (Salv), to further phosphorylate large tumor suppressor homolog (Lats1 and Lats2) kinase. Lats1 and Lats2 bind to a scaffold protein, Mob, to further phosphorylate two transcriptional co-activator Yap and Taz. Yap and Taz are downstream Hippo signaling components and modulate the expression of genes involved in cell growth, proliferation and apoptosis. Phosphorylated Yap and Taz are retained in the cytoplasm. Unphosphorylated Yap and Taz can enter the nucleus to activate gene expression in concert with TEAD, P73, Pax3, Runx2 transcription factors and Smad and β-catenin. Pitx2 activates FoxO6 to maintain FoxO6 expression in craniofacial tissues. In this paper, we conclude that FoxO6 binds and activates Lats1 to further inhibit Yap and cell proliferation, thereby controlling head and craniofacial growth regionally.