Specificity Protein 7 Is Required for Proliferation and Differentiation of Ameloblasts and Odontoblasts

Abstract

The Sp7/Osterix transcription factor is essential for bone development. Mutations of the Sp7 gene in humans are associated with craniofacial anomalies and osteogenesis imperfecta. However, the role of Sp7 in embryonic tooth development remains unknown. Here we identified the functional requirement of Sp7 for dentin synthesis and tooth development. Sp7-null mice exhibit craniofacial dysmorphogenesis and are completely void of alveolar bone. Surprisingly, initial tooth morphogenesis progressed normally in Sp7-null mice. Thus the formation of alveolar bone is not a prerequisite for tooth morphogenesis. Sp7 is required for mineralization of palatal tissue but is not essential for palatal fusion. The reduced proliferative capacity of Sp7-deficient ectomesenchyme results in small and misshapen teeth with randomly arranged cuboidal preodontoblasts and preameloblasts. Sp7 promotes functional maturation and polarization of odontoblasts. Markers of mature odontoblast (Col1a, Oc, Dspp, Dmp1) and ameloblast (Enam, Amelx, Mmp20, Amtn, Klk4) are barely expressed in incisors and molar tissues of Sp7-null mice. Consequently, dentin and enamel matrix are absent in the Sp7-null littermates. Interestingly, the Sp7 expression is restricted to cells of the dental mesenchyme indicating the effect on oral epithelium–derived ameloblasts is cell-nonautonomous. Abundant expression of Fgf3 and Fgf8 ligand was noted in the developing tooth of wild-type mice. Both ligands were remarkably absent in the Sp7-null incisor and molar, suggesting cross-signaling between mesenchyme and epithelium is disrupted. Finally, promoter-reporter assays revealed that Sp7 directly controls the expression of Fgf-ligands. Together, our data demonstrate that Sp7 is obligatory for the differentiation of both ameloblasts and odontoblasts but not for the initial tooth morphogenesis.

Introduction

Specific interactions between cell lineages derived from the ectomesenchyme and oral ectoderm drive the process of tooth development and formation of the craniofacial structure in mammals. Tooth development begins in mice at embryonic day 10 (E10) and transitions through bud, cap, and bell stages by E18.5.⁽¹⁾ Differentiated odontoblast express marker genes such as collagen type I (Col1), dentin sialophosphoprotein (Dspp), osteocalcin (Oc), and dentin matrix protein 1 (Dmp1). Similarly, enamel producing ameloblasts are characterized by stage-specific expression of amelogenin (Amelx), ameloblastin (Ambn), amelotin (Amtn), enamelin (Enam), matrix metalloproteinase 20 (Mmp20), and kallikrein-related peptidase 4 (Klk4).^(2–4)

At the onset of tooth development, secreted fibroblast growth factor (Fgf) initiates crosstalk between the oral epithelium and the mesenchyme.⁽²⁾ The elaborate expression of Fgf ligands and their cognate receptors is vital during tooth development but is still not completely delineated. Signaling from 13 members of the Fgf family is involved in mesenchyme and epithelial interactions during murine tooth development.⁽⁵⁾ The Fgf4, Fgf8, and Fgf9 secreted by epithelium regulates proliferation and/or apoptosis of the mesenchyme.⁽⁶⁾ Reciprocally, Fgf3 and Fgf10, expressed by the mesenchyme, control the functional development of epithelium.⁽⁷⁾ Epithelial Fgf8 and its mesenchymal counterpart Fgf3 are among the earliest and essential signaling molecules involved in the progression of tooth development. Fgf8 promotes patterning and initiation of tooth morphogenesis.⁽¹⁾ Fgf8-null mice contain rudimentary incisors but lack molars.⁽⁸⁾ Fgf3, however, is required for the differentiation of ameloblasts and progression of tooth development from the bud to the late bell stage.⁽⁹⁾ Consistent with this role, a transient expression of Fgf3 is noted in the primary enamel knot.⁽⁷⁾ Fgf8 induces the expression of Msx1 in the mesenchyme.⁽¹⁰⁾ Both Msx1 and Runx2 are known to induce expression of Fgf3 in the developing dental mesenchyme.⁽¹¹⁾

It is generally thought, that progression of tooth morphogenesis depends upon the proper development of the maxillary and mandibular structures. In humans, absence or degradation of alveolar bone causes tooth agenesis.⁽¹²⁾ The absence of bone in the maxillofacial structure in global null mouse model of Runx2,⁽¹³⁾ Msx1,⁽¹⁴⁾ and Dlx5/6⁽¹⁵⁾ genes leads to an arrest of tooth development at the bud stage and severe abnormalities in palate development. In mice, palatogenesis begins at E12 and palatal fusion is completed by E16 with the subsequent ossification of the anterior two-thirds of the palate.^(16,17) Secondary palatal shelves fail to fuse in the global Runx2 null mice.⁽¹⁸⁾ RUNX2 insufficiency in humans has been linked to the lip and or palatal clefting.⁽¹⁹⁾ However, the role of Sp7 in palatal development and palatal osteogenesis has yet to be elucidated.

Specificity protein 7 (Sp7), also known as Osterix, belongs to the zinc-finger-containing Sp1 family of transcription factors. Sp proteins bind GC-rich sequences and can induce or suppress transcription of target genes during embryonic and postnatal development. Sp7 is essential for differentiation of osteoblasts and bone development. Sp7 global null mice lack mineralized bone tissue and die shortly after birth.⁽²⁰⁾ In mammals, the Sp7 gene consists of two exons. The first exon encodes seven amino acids, while the rest of the Sp7 protein, containing the transactivation and DNA binding domain, is encoded by the second exon. A homozygous frameshift mutation in the human SP7 gene is associated with the recessive form of osteogenesis imperfecta (OI).⁽²¹⁾ A patient born to consanguineous parents presents the classic OI phenotypes of bowing of the limbs, bone deformities, and frequent bone fractures. Craniofacial abnormalities include Wormian bones in the skull, prominent forehead, midface hypoplasia, depressed nasal bridge, microstomia, micrognathia, and high arched palate. The patient did not exhibit dentinogenesis imperfecta but teeth eruption was delayed.⁽²¹⁾

Sp7 is a downstream target of Runx2, an essential transcription factor for skeletogenesis and tooth formation.^(22,13) Like Runx2, Sp7 is expressed in odontoblasts and alveolar bone.⁽²³⁾ However, specific contributions of Sp7 in the establishment of ameloblasts and odontoblasts lineage during tooth development remain elusive. Here we report functional control of craniofacial and tooth development by Sp7.

Materials and Methods

Mouse model

The Sp7^+/− heterozygous mice were generated and maintained on a C57BL/6 background. In these mice, the second exon of the Sp7 gene that encodes all functional domains of the protein was replaced with a β-galactosidase gene. The Sp7^+/− mice were interbred to obtain WT and Sp7-null littermates. Deletion of the second exon of Sp7 was confirmed by PCR genotyping and Western blot analysis as described.⁽²⁰⁾ All animal experiments were performed with the approval from the Institutional Animal Care and Use Committee of the University of Alabama at Birmingham and conformed to relevant federal and state guidelines and regulations.

Skeletal preparation and staining

Newborn WT and Sp7-null littermates (n = 3) were euthanized, de-skinned, eviscerated, and muscles were partially removed. Pups were fixed and stained with Alizarin red and Alcian blue staining solution according to the standard protocol as described.⁽²⁴⁾ Images were captured with a Nikon Cool PIX S10 digital camera.

X-gal staining

Heads (n = 3) were de-skinned, split sagittally, and fixed in 4% paraformaldehyde (PFA) for 2 hours at 4°C. Tissues were incubation in polyvinylpyrrolidone (PVP) solution (100 mg/mL PVP, 5 mg/mL sucrose, 10 mL 100mM Tris-HCl pH7.4, and double-distilled water [ddH₂O]) in the dark for 3 days. Heads were then embedded in OCT compound (Sakura Tissue-Tek) and stored at –80°C. Cryosections containing incisor tooth organs at 8 μm thickness were obtained and selected sections were stained with X-gal according to the standard protocol as described.⁽²⁵⁾ Slides were mounted with CrystalMount aqueous mounting medium (Biomeda) and imaged using a Nikon Eclipse 80i high-resolution CCD camera.

Histological analysis and immunofluorescence

Heads from WT, Sp7^+/−, and Sp7^−/− mice (n = 3) were harvested, de-skinned, and fixed in 4% PFA in PBS overnight at 4°C. Fixed tissues were dehydrated through an ethanol gradient, cleared in xylene, and embedded in paraffin. Embedded heads were sagittally sectioned at 7 μm thickness and mounted on Superfrost Plus slides (Fisher Scientific). Serial sections of WT, heterozygous, and homozygous mutant littermates containing molar and incisor tooth organs were then cleared in xylene, rehydrated, and stained with hematoxylin and eosin (H&E) according to standard protocol. Picrosirius red staining was performed, according to standard procedure, to highlight the deposition of collagen type I in the predentin (PD) and dentin (D) matrices. Stained sections were dehydrated, cleared with xylene and mounted with Cytoseal XYL (Richard-Allan Scientific). Images were captured using a Nikon Eclipse 80i inverted microscope.

To assess mineralization of alveolar bone, WT and Sp7-null heads were subject to von Kossa staining. After rehydration, slides were incubated with 1% silver nitrate solution in a clear glass Coplin jar and placed under ultraviolet light for 20 min. Tissue was then incubated in 5% sodium thiosulfate for 5 min. Slides were then counterstained with Alcian blue for 30 min followed by dehydration, mounting and imaging as described above.

Tissue samples and sections were processed for immunostaining with respective primary antibody as described.⁽²⁴⁾ The sources and concentrations of the primary antibodies were: rabbit anti-Ki-67 (1:250; Abcam; ab66155), rabbit anti-amelogenin (1:200; Sigma; HPA-005988), rabbit anti-Gm-130 (1:250; Abcam; ab52649), mouse anti-Fgf3 (1:100; Santa Cruz; sc-135), and mouse anti-Fgf8 (1:50; R&D Systems; MAB323). Specifically bound antibodies were detected by incubation with Alexa Fluor 488 goat-anti-rabbit (1:1000; Invitrogen; A-11008) or anti-mouse (1:1000; Invitrogen; A-11001) secondary antibody for 1 hour at 25°C. Slides were washed three times with TBS (with 0.1% Tween-20) after each solution. Finally, sections were incubated with 4,6-diamidino-2-phenylindole (DAPI) (1 μg/mL) for 5 min at 25°C and mounted with ProLong Gold anti-fade reagent (Invitrogen). Slides were analyzed under fluorescent light with Tetramethyl Rhodamine Iso-Thiocyanate (TRITC) and DAPI filters with a CoolSnap EZ Photometrics fluorescence microscope with high-resolution CCD camera. The fluorescent signal was captured using NIS-Elements Advanced Research software.

Scanning electron microscopy

Heads from newborn WT and Sp7^−/− littermates were fixed, and split in half sagittally. Specimens were embedded in paraffin, sectioned, and processed as above. Tissue sections were coated with platinum using a Cressington 108 auto Sputter Coater (Cressington Scientific Instruments). Samples were imaged with a JSM 6360 scanning electron microscope at 7 kV (JEOL).

BrdU labeling assay

Pregnant females were injected at E18.5 with BrdU (Sigma) at 0.1 mg/g body weight, 3 hours prior to harvest. Mice were then harvested, dissected, sectioned, and antigen retrieval was performed as previously noted. Slides were then incubated with rat anti-BrdU antibody (1:200; Abcam; ab6326) in blocking buffer in the same manner as previously described, followed by incubation with a goat anti-rat IgG biotinylated secondary antibody (1:1000; Vector Lab; BA-9400) in blocking buffer for 60 min at 25°C. Tissue was then stained with streptavidin-cy3 (1:1000; Sigma; S6402) in TBS (with 0.1% Tween-20). Finally, nuclei were stained with DAPI, slides were mounted and imaged as described above.

RNA isolation and reverse transcriptase PCR analysis

Total RNA was isolated from whole mandibles and maxilla of newborn WT and Sp7^−/− mice using Trizol reagent (Life Technologies). The cDNA was prepared from 2 μg of total RNA using a cDNA synthesis kit (BioRad). Traditional RT-PCR was performed using GoTaq polymerase and 5× green GoTaq Flexi buffer (Promega), with specific primer pairs detailed in Table 1, according to manufacturer-recommended protocol. Amplified cDNA fragments were separated on 1% agarose gel and detected using ethidium bromide (Fisher Scientific). Gels were imaged using a UV light box and densitometric analysis was performed with Alpha-Inotech software.

Table 1.

Primers used for PCR

Genes	Primers
Genotyping
Sp7	F 5ʹ-AACCATGACGACAAGGGGATAACC-3ʹ
	R 5ʹ-GTGGAGACCTTGCTCGTAGATTTC-3ʹ
	R 5ʹ-GACAATAGCAGGCATGCTGGGGAT-3ʹ
RT-PCR
β-actin	F 5ʹ-TGTTACCAACTGGGACGACA-3ʹ
	R 5ʹ-TTTGATGTCACGCACGATTT-3ʹ
Alp	F 5ʹ-ACCTTGACTGTGGTTACTGCT-3ʹ
	R 5ʹ-GACGCCGTGAAGCAGGTGTGCC-3ʹ
Col1	F 5ʹ-GAGCTGGTGTAATGGGTCCT-3ʹ
	R 5ʹ-GAGACCCAGGAAGACCTCTG-3ʹ
Oc	F 5ʹ-CCTAGGTAGTGAACAGACTCCGGCG-3ʹ
	R 5ʹ-CTGGTCTGATAGCTCGTCAACAA-3ʹ
Dspp	F 5ʹ-TGAAAACTCTGTGGCTGTGC-3ʹ
	R 5ʹ-TTGCTGTTGCTAGTGGTGCT-3ʹ
Dmp1	F 5ʹ-GCTGGGTCAACCACCACCACC-3ʹ
	R 5ʹ-GGAGGGTCCTCCCCACTGTCC-3ʹ
Amelx	F 5ʹ-CTGGGAGCAGCTTTTGCTAT-3ʹ
	R 5ʹ-CCACTTCAAAGGGGTAAGCA-3ʹ
Enam	F 5′-AAGTGGCATTGGCTCTCATC-3′
	R 5′-CAGACCCAGGAAAACAAGGA-3ʹ
Mmp20	F 5ʹ-GCTGCTGTGGAACTGAATGG-3ʹ
	R 5ʹ-TCTGACACACGACTTGGTGCG-3ʹ
Klk4	F 5ʹ-CCGGCTGCTGTATGACCCTG-3ʹ
	R 5ʹ-TGTGGGCCTTGTAGTCAGTCC-3ʹ
Amelt	F 5ʹ-TGCAACCGCAGATGTTAC-3ʹ
	R 5ʹ-GAGTGGTTCCCTGGTTGAC-3ʹ
Sp7	F 5ʹ-GGAAGGGTGGGTAGTCATTTG-3ʹ
	R 5ʹ-TCCTCTCTGCTTGAGGAAGAAG-3ʹ

In situ hybridization

Heads of WT and Sp7^−/− littermates were prepared as described.⁽²⁴⁾ Sections were hybridized with digoxigenin-labeled single-stranded RNA probes for 16 hours at 65°C. A 319-basepair (bp) fragment of mouse Dspp cDNA was used to generate sense and antisense probes as described.⁽²⁶⁾ Digoxigenin-labeled Dspp RNA antisense and sense probe (1:100) was prepared from the plasmid using a DIG RNA labeling kit (Roche). Slides were hybridized with 1 μL of synthesized digoxigenin-labeled cRNA probe according to the standard protocol (1:250 overnight at 4°C). The probe was detected for color by BM Purple (2mM; Roche) in the dark at 22°C for 2 hours. Images were captured with a Nikon Eclipse 80i microscope color camera after two PBS washes.

Western blot analysis

Whole mandibles of WT and Sp7-null mice were extracted at birth. Tissues were flash frozen and pulverized with a dounce homogenizer in ice-cold PBS containing 1× complete protease inhibitor (Roche) and 25μM MG132 at 4°C. Pellets were lysed in ice-cold NP-40 lysis buffer (10mM Tris, 3mM MgCl₂, 10mM NaCl, 1%NP-40, protease inhibitor, 25μM MG132). Equal amounts of protein samples were separated on 10% SDS-polyacrylamide gel. Proteins were then probed with the loading control goat polyclonal Lamin-B1 antibody (1:1,000; Santa Cruz; sc-6217) and either mouse anti-Fgf3 (1:500; Santa Cruz; sc-135) or mouse anti-Fgf8 (1:200; R&D Systems; MAB323) antibodies. Finally, blots were incubated with IRDye 800CW goat anti-mouse labeled secondary antibody (1:15,000; Li-COR Biotechnology; 926-32210) for 1 hour and scanned with Odyssey imaging systems.

Cloning of promoter-reporter genes and luciferase assay

The Dspp gene promoter containing Sp binding domains was PCR-amplified from mouse C57BL/6 genomic DNA. The forward primer (5ʹ-ATTACATGTGCTGTGAAGAGACACCA-3ʹ) contained a PciI site that was bound to a sequence 2444 bp upstream of exon 1 and the reverse primer (5ʹ-TAAGCTAGCGAGTCCATCTTACCTTC-3ʹ) contained a NheI site binding 70 bp downstream of exon 1. The PCR product was cloned into an intermediate vector using TOPO TA Cloning kit according to manufacturer recommendations (Invitrogen). The 2.4-kb Dspp promoter was cut from TOPO vector by digesting with PciI and NheI and ligated into a similarly digested pEGFP-Luciferase dual reporter vector. The integrity of each promoter was confirmed by direct sequencing at UAB Core Laboratory. The –1.1-kb and –0.4-kb mouse Fgf3 and –0.2-kb rat Oc gene promoter encompassing SP motifs was described (Saroff and colleagues, unpublished data).⁽²⁷⁾

HEK293T cells were cultured in DMEM media supplemented with 10% FBS, 2mM L-glutamine, 50 U/mL penicillin G, and 50 mg/mL streptomycin. Cells were maintained in 37°C in a humid atmosphere with 5% CO₂ in the air and were passed and fed every 48 hours. Cells were seeded in 12-well dishes at a density of 1.5 × 10⁵ cells per well and cultured for 24 hours before transfection. Using Polyjet reagent (SignaGen), cells were co-transfected with 100 ng Fgf3, Oc, or Dspp EGFP-Luc plasmids and 100 ng of WT Sp7 expression plasmid. The cell lysate was obtained 24 hours posttransfection using a Luciferase Reporter Assay system (Promega) according to suggested protocol. Luciferase activity was assessed using DLReady Femtomaster FB12 and FB12 Sirius Software.

Results

Tooth morphogenesis progresses to the early bell stage in absence of alveolar bone and Sp7 gene

To assess Sp7 regulatory control of craniofacial and tooth development, we established a colony of global Sp7-null mice. Deletion of Sp7 gene and the lack of Sp7 mRNA in homozygous mice (Sp7^−/−) were confirmed by PCR and RT-PCR analysis, respectively (Supporting Fig. S1A). Homozygous mice died soon after birth. Alizarin red and Alcian blue staining revealed that mineralization was completely absent in mandible, maxilla, and craniofacial bones of Sp7^−/− mice (Fig. 1A). Heads from newborn Sp7^+/+ and Sp7^−/− littermates were processed for histological analysis to assess tooth development. Surprisingly, Sp7^−/− littermates exhibited age-appropriate uni-cusp incisors and multi-cusped first and second molars (Fig. 1B, upper panels). However, the size and morphology of incisors in the Sp7^−/− mice were significantly altered when compared to Sp7^+/+ littermates (Fig. 1B). Quantification of the average volume of the tooth organ revealed maxillary incisors were 53% and mandibular incisors were 45% smaller in Sp7^−/− mice (Fig. 1C). Similarly, maxillary first molars were significantly smaller (55%) in Sp7^−/− mice. The first mandibular molar was also decreased by 33% in Sp7^−/− littermates (Fig. 1C).

Fig. 1.

Deficiency of Sp7 gene does not impair tooth morphogenesis. (A) Three independent newborn WT and Sp7^−/− littermates (n = 3) were de-skinned and double-stained with Alizarin red and Alcian blue. A representative picture of stained heads is shown. Clear craniofacial dysmorphology is noted in Sp7^−/− mice. (B) Heads from newborn littermates were processed for histological analysis. H&E-stained sagittal section showing first and second mandibular molars (M1, M2) and maxillary incisor regions captured at magnification 4×. Lack of dentin and disorganization of odontoblast in the Sp7^−/− tooth is indicated by an asterisk. Images were captured at magnification 4× and 10×. Boxed areas are shown at magnification 40×. Scale bars = 100 μm upper panels; 250 μm middle panels. (C) Length, width, and height of Max and Mand incisors and first molars were measured using NIS Elements software. Equivalent sagittal sections were selected for height and length measurements. Nine serial frontal sections were analyzed to identify the widest point of each molar and incisor. Pooled data from nine equivalent sections of three separate mice is presented in the graph. *p < 0.005, **p < 0.00005. AM = ameloblast; OD = odontoblast; PD = pre-dentin; P = pulp; Max = maxillary; Mand = mandibular.

Both molars and incisors progressed to the bell stage in Sp7^−/− mice (Fig. 1B). Consistent with its developmental stage, Sp7^+/+ incisors contain elongated odontoblasts and ameloblasts layers arranged around a distinct pre-dentin layer (Fig. 1B, lower panel). The Sp7^−/− incisors, however, exhibited disorganized odontoblasts and ameloblasts arranged around a pre-dentin-like layer (Fig. 1B). Tooth development and eruption at postnatal day (P16) in Sp7 heterozygous mutants (Sp7^+/−) was comparable to WT littermates (data not shown). Thus the haploinsufficiency of the Sp7 gene does not affect tooth development or eruption in postnatal life.

The formation of the dento-alveolar complex involves a coordinated interaction between the dental epithelium-mesenchyme and the alveolar bone. To investigate if mineralization of the alveolar bone is necessary for the developing tooth organs, sagittal sections of newborn heads were subjected to von Kossa staining (Fig. 2A). Mineralized alveolar bone surrounds the developing incisor in Sp7^+/+ mice. We also noted mineralized dentin in the Sp7^+/+ incisor. However, the mineralized alveolar bone surrounding incisor teeth was completely absent in the Sp7^−/− mice (Fig. 2A). These data demonstrate that presence of alveolar bone is not a prerequisite for tooth morphogenesis.

Fig. 2.

Mineralized alveolar bone is not a prerequisite for tooth morphogenesis. (A) Sagittal sections of newborn Sp7^+/+ and Sp7^−/− heads (n = 3 per group) were subjected to Alcian blue/von Kossa staining. Representative images of the maxillary incisor and surrounding bone is taken at magnification 4×. Arrows correspond to the mineralized maxillary bone in Sp7^+/+ and unmineralized region in Sp7^−/− mice. Scale bar = 100 μm. (B) Coronal sections of newborn heads from Sp7^+/+ and Sp7^−/− littermates were subjected to H&E and Alcian blue/von Kossa staining. Representative images of anterior (upper panels) and posterior (lower panels) palate tissue are shown at magnification 4×. Arrowheads correspond to the descent of anterior palate in Sp7^+/+ but not in Sp7^−/− mice. Asterisk indicates the absence of palatal tissue at the midline of Sp7^−/− mice. Scale bar = 1000 μm. P = palate; M = maxillary molar; NS = nasal septum; T = tongue.

Sp7 is expressed in the anterior and posterior palate but is not essential for palatal fusion

Gross morphology of Sp7^−/− mice shows clear craniofacial abnormalities highlighted by prominent forehead, hypoplasia of the mid-face region, microstomia, and micrognathia (Supporting Fig. S1B). To investigate if Sp7 loss causes additional oral anomalies, we examined palatal development in newborn littermates. Coronal sections containing anterior and posterior hard palate tissues were stained with H&E and von Kossa (Fig. 2B). The anterior and posterior palatal shelves were completely fused at the midline and elevated above the tongue in both littermates. The uniformly fused palatal segments in Sp7^+/+ mice showed a close association with the tongue. In sharp contrast, a high arched palate with clear gap between palate and tongue was noted in both anterior and posterior palatal segments of the Sp7^−/− littermates (Fig. 2B, upper panel). The palatal development in Sp7^+/− mice was identical to WT littermates (data not shown). These results suggest that Sp7 is not essential for the development and fusion of palate.

Subsequent mineralization of the anterior and posterior hard palate was examined by von Kossa staining. Mineralization of the upper palate, maxillary, and mandibular bone surrounding molars and incisors was evident in WT mice. (Fig. 2B, lower panels). However, these orofacial structures in Sp7^−/− mice were completely devoid of mineralization. Interestingly, DAPI staining revealed the number of nuclei in the mineralized region of Sp7^+/+ palate was similar to that of the Sp7^−/− littermates (Supporting Fig. S1C). Thus failure of mineralization is not due to the absence of mesenchymal cells in the palate. We took advantage of the LacZ knock-in gene in Sp7^−/− model to evaluate the spatial expression of Sp7 in both the anterior and posterior palate (Supporting Fig. S1D). The X-gal staining was present in the mesenchymally derived portion of the palate but absent in the midline ectodermal derived portion separating the two palatal shelves (Supporting Fig. S1D). Together, these results indicate that Sp7 is required for mineralization of both anterior and posterior palatal tissue.

Sp7 regulates the proliferative capacity and functional polarization of dental mesenchyme

Teeth were misshaped and smaller in size in Sp7^−/− mice (Fig. 1). To assess if this is related to changes in the proliferative capacity of Sp7-deficient ectomesenchyme and oral epithelium, BrdU assay was performed. The developing mandibular incisor of the Sp7^−/− mice exhibited poor BrdU labeling compared to Sp7^+/+ littermates (Fig. 3A). To better quantify the difference in cell proliferation, the number of BrdU-positive cells was counted from an equal number of DAPI-positive cells in identical regions of developing mandibular incisors. The abundant presence of BrdU-positive cells in the dental pulp (P) and in the labial cervical loop (LaCL) of Sp7^+/+ incisor was decreased by fivefold and twofold, respectively, in Sp7^−/− littermates (Fig. 3B). Moreover, BrdU-positive preameloblasts (AM) and preodontoblasts (OD) were reduced by 4.7-fold and twofold, respectively, in the Sp7^−/− incisors. A similar reduction of cell proliferation in the Sp7^−/− tooth organs was noted using immunohistochemistry with the α-Ki-67 antibody (data not shown). These data suggest that the smaller size of the Sp7^−/− tooth organ may be caused by a reduced proliferation of cells. Taken together, these results show that Sp7 regulates the proliferative capacity of the dental ectomesenchyme and oral epithelium.

Fig. 3.

Sp7 promotes proliferation of progenitor cells in the developing incisor. (A) Pregnant females at E18 were injected with BrdU and embryos harvested 3 hours later. Sagittal sections of heads were stained with polyclonal anti-BrdU antibody. Images were captured at magnification 20×. The lingual loop and LaCL are outlined by the white dotted line. Boxed regions are shown at magnification 40×. (B, C) The number of BrdU-positive cells in mandibular tooth organs were counted from 100 pulp cell, 50 LaCL cell, and 20 each AM and OD cells stained with DAPI. Pooled data from identical areas of littermate (n = 3 per group) is presented in the graph. *p < 0.05, **p < 0.005. P = pulp; OD = odontoblast; AM = ameloblast; x = lingual loop; LaCL = labial cervical loop.

Interestingly, stem cells in the cervical and labial loop of Sp7^−/− tooth organs showed commitment to preodontoblasts and preameloblasts. However, these cells remain cuboidal, clustered, and lack polarity. We initially compared their polarity with DAPI staining (Fig. 4; data not shown). Sp7^−/− incisors contain DAPI-stained nuclei clustered in the basal body and away from the apical end of ameloblasts at both the presecretory and secretory stage. Interestingly, the nuclei of ameloblasts in Sp7^−/− were disorganized and aggregated when compared to equivalent cell layers of Sp7^+/+ littermates. Lack of polarity was further confirmed by the cis-Golgi marker (Gm-130) that localized specifically in the apical end of polarized ameloblasts and odontoblasts. The odontoblasts and ameloblasts in Sp7^+/+ incisors gradually progress through early, middle, and late stages of differentiation and become more elongated apically. In sharp contrast, both odontoblasts and ameloblasts of Sp7^−/− incisors fail to polarize and lack elongated apical bodies projecting toward the dentin-enamel junction (Fig. 4A). These results strongly suggest that Sp7 deletion leads to a failed polarization of both ameloblasts and odontoblasts. The lack of odontoblasts and ameloblasts polarity was further confirmed through SEM (Fig. 4B). Elongated ameloblasts and odontoblasts, aligned uniformly, with cell projections penetrating into well-defined pre-dentin and the dentin matrix, were noted in Sp7^+/+ incisors. However, pre-dentin/dentin matrix was indistinguishable in Sp7^−/− mice and both cell types lack apically oriented cell projections (Fig. 4B).

Fig. 4.

Deletion of Sp7 disrupts polarization of ameloblasts and odontoblasts. Heads of newborn Sp7^+/+ and Sp7^−/− littermates were sectioned sagittally and stained with DAPI and Gm-130 antibody. (A) Images of the DAPI/Gm-130 stained whole mandibular incisor are shown at magnification 10×. Boxed areas imaged at magnification 100× correspond to apical, middle, and coronal regions that represent successive stages of odontoblast and ameloblast differentiation. Dashed white line separates cell layers at the DEJ. Double-headed arrows correspond to the length of each cell layer. Scale bars = 250 μm. (B) Failed polarization and matrix organization in the Sp7^−/− tooth organs. Sections containing maxillary incisors of newborn Sp7^+/+ and Sp7^−/− littermates were processed for scanning electron microscopy. Left panels show entire maxillary incisor organs at magnification 30×. Boxed areas are shown at magnification 800×. Asterisk denotes pre-dentin like matrix in Sp7^−/− incisor. Scale bars = 500 μm left panels; 20 μm right panels. DEJ = dentin-enamel-junction; AM = ameloblast; OD = odontoblast; P = pulp; D = dentin; PD = pre-dentin.

The gross organization of the collagen network in the dentin matrix secreted by Sp7^+/+ and Sp7^−/− odontoblasts was assessed using picrosirius red staining (Fig. 5A). Interestingly, collagen type I (Col1) was deposited at the odontoblast basement membrane in both Sp7^+/+ and Sp7^−/− incisors. At the premature stage, the Sp7-deficient odontoblasts showed less collagen in cells and in the extracellular matrix (ECM). Unlike Sp7^+/+, the collagen matrix in Sp7^−/− incisors becomes thinner and disorganized close to the cusp tip (Fig. 5A, right panels). Interestingly, the collagen is deposited in the pulp chamber and the collagen network surrounds cells at the cusp tip. This indicates that Sp7-deficient odontoblasts are disorganized and lack polarity (Fig. 5A, right panels). The orientation of collagen fibers was investigated under polarized light (Fig. 5B). The Sp7^+/+ incisor showed mostly a radiant red color indicating aggregated and organized collagen matrix (Fig. 5B, left panels). In sharp contrast, Sp7^−/− revealed a mixture of radiant red and a light white signal indicating disorganized collagen in both dentin and bone matrices (Fig. 5B). Thus defects in assembly of the collagenous matrix, cell proliferation and polarity are contributing to smaller and misshaped teeth in the Sp7^−/− mice.

Fig. 5.

Altered organization of collagen fibers in dentin and bone of Sp7-null mice. Sagittal sections from newborn littermates were stained with picrosirius red to detect collagen fibers. (A) Images of incisor tooth organs at magnification 10× reveal Col1 deposition. Boxed areas shown at magnification 40× represent regions containing premature and mature ameloblasts and odontoblasts. Scale bar = 1000 μm. (B) Images of Picrosirius red–stained Sp7^+/+ and Sp7^−/− heads were captured under polarized light. A radiant red color denotes aggregated and organized collagenous matrix whereas a light white signal is indicative of loosely disorganized fibers. Boxed areas are shown at magnification 4× and 10× to reveal the collagen layer at the cusp tip and surrounding mandibular bone. Col1 = collagen; OD = odontoblast; AM = ameloblast.

Sp7 promotes differentiation and functional maturation of odontoblasts

To better understand the failed differentiation of odontoblast in Sp7^−/− mice, we followed expression of specific markers in RNA isolated from the mandibles and maxillas of newborn littermates (Fig. 6). RT-PCR analysis showed expression of early markers of odontoblast differentiation, alkaline phosphatase (Alp) and collagen type I (Col1), was markedly reduced in Sp7^−/− mice (Fig. 6A). However, mature odontoblasts markers, osteocalcin (Oc), dentin sialophosphoprotein (Dspp), and dentin matrix protein 1 (Dmp1), were barely detected in Sp7^−/− teeth. RT-PCR data, quantified from two independent newborn litters, confirmed a significant decrease in mRNA levels of Col1 and Alp in Sp7^−/− mice (Fig. 6B). Expression of mature odontoblasts markers Oc and Dspp were decreased by fourfold and Dmp1 by 39-fold in Sp7^−/− mice (Fig. 6B). The spatial expression of Dspp mRNA in mature odontoblast was confirmed by in situ hybridization (Fig. 6C). Sagittal sections of newborn littermates were hybridized with sense and an anti-sense Dspp probe (Fig. 6C). Abundant expression of Dspp mRNA was evident in odontoblast of Sp7^+/+ mice (Fig. 6C). In sharp contrast, the odontoblasts of the Sp7^−/− incisor showed no expression of Dspp. Specificity of Dspp signal was confirmed by sense probe used as control (Fig. 6C, upper panels). These results suggest that Sp7 protein regulates expression of marker genes of odontoblasts.

Fig. 6.

Sp7 deficiency disrupts expression of late-stage odontoblast-specific marker genes. (A) Total RNA was isolated from newborn maxilla and mandibles for RT-PCR analysis of odontoblast marker genes. Representative pictures of agarose gels with β-actin used as loading control are shown. (B) Relative values for odontoblast markers Alp, ColI, Oc, Dspp, and Dmp1 are presented in the bar graphs. Densitometric analysis was performed with Alpha-Innotech software. Pooled data from n = 3 per group was normalized with β-actin. *p < 0.05, **p < 0.005, *p < 0.0005. (C) Heads from newborn littermates were embedded in paraffin and processed for Dspp in situ hybridization. The section containing mandibular incisor were incubated for 12 hours with either Dspp antisense or sense probes. Representative images of sense probe (negative control) are shown at magnification 10× and Dspp mRNA at magnification 40×. Scale bars = 100 μm. The green box represents binding sites for zinc finger proteins in the -2.4-kb Dspp promoter (D) and the –0.2-kb Oc promoter (E). HEK 293T cells were co-transfected with promoter-reporter and Sp7 expression plasmid for 24 hours. Luciferase data pooled from three separate experiments with four replicates each (n = 12) are presented in graph. AM = ameloblast; OD = odontoblast; P = pulp.

To delineate whether Sp7 directly controls the expression of odontoblast-specific markers, we performed promoter-reporter assays. The genomic sequence of Dspp loci in mouse, rat, and human were initially compared to determine conserved transcriptional start sites. We then cloned the 2.4-kb mouse Dspp promoter upstream of a dual GFP-luciferase reporter. This fragment of the Dspp promoter contains eight potential binding sites for zinc-finger transcription factors (Fig. 6D). The effect of overexpression of Sp7 on the activity of 2.4-kb Dspp promoter-GFP-Luc reporter gene construct was tested in the 293T cells. The results indicate that Sp7 significantly enhances Dspp promoter activity by 40-fold (Fig. 6D). Promoter activity of another marker of mature odontoblast Oc was evaluated to independently establish the transcriptional function of Sp7 (Fig. 6E). The –0.2-kb proximal Oc promoter contains a binding motif for zinc-finger transcription factors. Expression of Sp7 in the 293T cells consistently stimulates Oc promoter activity by eightfold to 10-fold (Fig. 6E). Thus, Sp7 is a potent activator of Dspp and Oc gene transcription.

Sp7 is required for the functional differentiation of ameloblasts

Immunofluorescence results suggest a disrupted differentiation of ameloblasts because they fail to organize and polarize from the presecretory stage through the cusp tip in the incisors of Sp7^−/− (Fig. 4). To better understand the failed differentiation of ameloblast in Sp7^−/− teeth, we evaluated the expression profile of various stage-specific genes considered hallmarks of maturing ameloblasts (Fig. 7). Total RNA isolated from mandibles and maxillas of Sp7^+/+ and Sp7^−/− littermates was used for RT-PCR analysis. We first assessed markers of secretory ameloblasts, enamelin (Enam), amelogenin (Amelx), and matrix metalloproteinase 20 (Mmp20). All three marker genes were expressed in Sp7^+/+ mice but barely detected in Sp7^−/− littermates (Fig. 7A). Similarly, mRNA of amelotin (Amtn) and kallikrein-related peptidase 4 (Klk4) that are expressed by mature ameloblasts were barely noticed in Sp7^−/− mice. Quantification of expression data from an independent Sp7^+/+ and Sp7^−/− littermates revealed a fivefold, 11-fold, and 13-fold decrease in Enam, Amelx, and Mmp20 mRNA, respectively (Fig. 7B). The absence of differentiated ameloblasts in Sp7^−/− littermates was further evident with a 14-fold decrease in expression of Amtn and Klk4.

Fig. 7.

Sp7 deficiency disrupts expression of marker genes of late-stage ameloblasts. (A) Total RNA was isolated from newborn maxilla and mandibles for RT-PCR analysis of mature ameloblasts marker genes. Representative pictures of agarose gels show levels of Enam, Amelx, Mmp20, Klk4, and Amelt mRNA in ameloblasts. (B) Relative mRNA levels of indicated genes were quantified by densitometric analysis using Alpha-Innotech software. Data was normalized with β-actin used as an internal control. Relative mRNA values pooled from n = 3 per group are presented in the bar graphs. (C) Paraffin-embedded heads of newborn Sp7^+/+ and Sp7^−/− littermates were sectioned and probed with polyclonal amelogenin antibody. Amelogenin protein was detected along labial aspects of both maxillary and mandibular Inc of Sp7^+/+ mice. Immunofluorescence pictures captured with same exposure times are shown at magnification 4×. Insets are shown at magnification 10×. Scale bar = 500 μm. *p < 0.05; **p < 0.005. Inc = incisor.

Amelx is the most abundantly expressed protein during enamel formation and is critical for organization and mineralization of the enamel matrix. To independently confirm the failure of differentiation of the enamel matrix in Sp7^−/− mice, orofacial tissues from newborn littermates were stained with an anti-Amelx antibody. Maxillary incisors of Sp7^+/+ mice show abundant Amelx protein that was restricted to their labial surfaces (Fig. 7C). The Sp7^−/− mice showed only a faint signal of Amelx protein in the maxillary incisor. Consistent with their developmental age, Amelx expression was not noted in either maxillary or mandibular molars of Sp7^+/+ mice (data not shown). These results confirm that differentiation of ameloblasts is disrupted in the Sp7-deficient mice.

Sp7 expression is restricted to the dental mesenchyme during embryonic tooth development

The failure of both ameloblasts and odontoblast differentiation prompted an examination of Sp7 expression in the developing tooth organ. To establish unequivocal expression of the Sp7 gene in the developing incisor, we took advantage of the β-galactosidase knock-in allele in Sp7^−/− model (Supporting Fig. S1D). Heads from newborn Sp7^+/+, Sp7^+/−, and Sp7^−/− littermates were sectioned and processed for X-gal staining. The activity of β-galactosidase in the craniofacial structure revealed that expression of Sp7 gene is restricted to the ectomesenchyme derived pulp and odontoblast lineage cells as well as osteoblasts in the maxillary and mandibular bone (Fig. 8). Interestingly, dental follicular cells (DF), which also originate from ectomesenchyme, showed no expression of Sp7 (Fig. 8, Supporting Fig. S2). Moreover, β-gal activity was completely absent in epithelial origin ameloblasts. Specificity of X-gal staining was confirmed with lack of any signal in tooth organs of Sp7^+/+ mice (Fig. 8). Due to double β-gal gene dosage, X-gal staining was stronger in Sp7^−/− teeth when compared with Sp7^+/− littermates (Fig. 8).

Fig. 8.

Sp7 gene is expressed specifically in the mesenchyme of the developing tooth organ. Expression of the endogenous Sp7 gene was assessed by the activity of the β-galactosidase knock-in gene. Sagittal sections of heads from newborn Sp7^+/+, Sp7^+/−, and Sp7^−/− littermates were stained with X-gal solution for 24 hours. The left column shows equivalent portions of the whole mandibular incisors at magnification 10×. Boxed areas are shown at magnification 40× in panels on the right. Specificity of the β-gal activity was confirmed with a complete lack of X-gal staining in Sp7^+/+ tissue (bottom panel). Scale bars = 250 μm. OD = odontoblast; P = pulp; AM = ameloblast; DF = dental follicle; LCL = labial cervical loop; M = middle region; AB = alveolar bone.

To independently confirm if the expression of Sp7 is limited to ectomesenchyme derived cells, we performed immunostaining of molar and incisor tooth organs. Sections through incisor and molar of Sp7^+/+ mice were stained with an affinity-purified Sp7 antibody (Supporting Fig. S2). Immunofluorescence signal was restricted to the osteoblast and odontoblasts. Sp7 protein was noted in nuclei of both cuboidal and polarized odontoblast (Supporting Fig. S2A, C). The immature cells residing in the grooves between cusps showed weaker expression and contained the lowest frequency of Sp7 positive cells (Supporting Fig. S2C). Functionally mature odontoblasts located at the cusp tips of molar tooth organs showed a progressively increased signal of Sp7 protein (Supporting Fig. S2D). Osteoblasts in the alveolar bone also showed a strong signal of Sp7 protein (Supporting Fig. S2A, C). Together, these results show that Sp7 expression is developmentally regulated in dental mesenchyme.

Sp7 regulates Fgf signaling during the differentiation of ameloblasts and odontoblasts

During tooth development, Sp7 was not expressed in epithelial ameloblasts (Fig. 8, Supporting Fig. S2). Therefore, failed ameloblast differentiation was surprising and suggested a cell nonautonomous function of Sp7. Coordinated and reciprocal signaling by secreted Fgf proteins in oral mesenchyme and epithelium is essential for embryonic and postnatal tooth development.⁽²⁸⁾ To uncover the mechanism for disrupted differentiation of odontoblasts and ameloblasts in Sp7^−/−, we evaluated Fgf8 and Fgf3 proteins in developing teeth (Fig. 9). During tooth morphogenesis, cross-talk between epithelium and mesenchyme is in part controlled by Fgf8 and Fgf3 secreted from the dental mesenchyme and oral epithelium, respectively. Head sections from newborn Sp7^+/+ and Sp7^−/− littermates were stained with either Fgf8 or Fgf3 antibodies. A strong signal of Fgf8 was evident in the secretory and mature ameloblasts of both maxillary and mandibular incisors of Sp7^+/+mice (Fig. 9A, B). In sharp contrast, mandibular or maxillary incisors of Sp7^−/− mice showed no staining of Fgf8. The red dots in the pulp and alveolar bone represent aberrant Fgf8 staining of red blood cells.

Fig. 9.

Sp7 deficiency disrupts Fgf signaling required for interaction between dental epithelium and mesenchyme. (A–C) Heads from newborn Sp7^+/+ and Sp7^−/− mice were sectioned sagittally and stained with monoclonal Fgf8 or Fgf3 antibodies and counterstained with DAPI. All images were captured at magnification 40×. Scale bars = 100 μm. (D) Nuclear extracts were isolated from maxillary and mandibular tissues of Sp7^+/+ and Sp7^−/− littermates. Western blots were incubated with either FGF3 or FGF8 antibody, stripped, and re-probed for Lamin B antigen, used as loading control. (E) The –1.1-kb and –0.4-kb mouse Fgf3 promoter fragments contain seven and two Sp binding motifs, respectively. Luciferase assays were performed using lysates from HEK 293T cells co-transfected with Sp7 expression plasmid and indicated fragments of the Fgf3 promoter. Pooled data from three independent experiments with four replicates each (n = 12) are shown in the bar graph. P = pulp; AM = ameloblast; OD = odontoblast.

Similarly, Fgf3 expressed in Sp7^+/+mice was nearly absent in the maxillary incisor (top panels) and molars (bottom panels) of Sp7^−/− littermates (Fig. 9C). To confirm the failure of Fgf8 and Fgf3 expression, total protein was extracted from the incisor tissue of newborn Sp7^+/+ and Sp7^−/− mice for Western blot analysis. Both Fgf8 and Fgf3 protein were present in Sp7^+/+ incisors but barely detected in Sp7^−/− littermates (Fig. 9D). These data demonstrate that the Fgf pathway is disrupted in Sp7^−/− tooth organs. To better understand if Sp7 directly regulates the expression of essential Fgf signaling molecules, we cloned the 1.1-kb mouse Fgf3 promoter. This promoter fragment contains seven Sp binding motifs (Fig. 9E). Overexpression of Sp7 resulted in a threefold increase in the Fgf3 promoter activity (Fig. 9E). We also evaluated the 400-bp proximal Fgf3 promoter that contains two Sp binding motifs. Sp7 induced the activity of the 0.4-kb promoter fragment by ninefold (Fig. 9E). Activation of the Fgf3 promoter further suggests that Sp7 directly regulates Fgf3 transcription in developing teeth. These results show that disturbance in the Fgf signaling is involved in failed differentiation of odontoblast and ameloblast in Sp7^−/− mice.

Discussion

Sp7 plays an essential role in osteoblast differentiation and skeletogenesis. However, the functional requirement of Sp7 in the developing tooth germ for the synthesis of dentin and enamel matrix was yet to be elucidated. Here we used a Sp7 global null mouse model to define roles of this transcription factor during embryonic tooth development. Surprisingly, Sp7-deficient dental mesenchyme and epithelium commits to odontoblast and ameloblasts lineages despite a complete absence of alveolar bone. Sp7 gene deletion, however, disrupts subsequent differentiation of both odontoblast and ameloblast. Sp7 directly regulates the expression of marker genes of mature odontoblast Dspp and Oc. Interestingly, during embryonic development, expression of Sp7 is restricted to the dental mesenchyme, which suggests that the impaired differentiation of ameloblasts is caused by cell nonautonomous mechanisms. We also demonstrate that impaired Fgf signaling in part is responsible for failed dentin and enamel synthesis in the Sp7 null mice.

The Runx2 transcription factor is required for the commitment of mesenchyme to the osteoblast lineage and bone synthesis. Interestingly, Runx2 exerts an analogous function in the odontogenic ectomesenchyme that is involved in the development of tooth organs. Resembling failed ossification, the development of tooth germs is blocked at the late bud stage in Runx2-null mice.⁽¹³⁾ Sp7 is a downstream target of Runx2, because the expression of Sp7 is absent in the Runx2 null mice.⁽²⁰⁾ Like Runx2, Sp7 protein is essential for the formation of bone tissue. Thus Sp7 and Runx2 exhibit a similar function during osteogenesis. Therefore, we anticipated that Sp7 would exert a similar role during odontogenesis. To our surprise, initial tooth morphogenesis occurred normally in the Sp7-null mice but subsequent differentiation of dentin and enamel-secreting cells was disrupted. Teeth of Sp7 mutants were significantly smaller in size. We show that the decrease in tooth size is related to the impaired proliferation of the dental progenitor cells. Our data indicates that Sp7 regulates both proliferation and differentiation of oral ectomesenchyme. These observations in Sp7-null mice are consistent with well documented properties of several transcription factors that control both proliferation and differentiation of specific cell types. For example, Runx2, Fli-1, Pax-3, C/EBPβ, and Ebf1 transcription factor regulate proliferation and differentiation of osteochondroprogenitors, erythroid cells, Schwann cells, mammary epithelial cells, and B cells, respectively.^(25,29–32)

Previous studies have indicated that teeth cannot develop in the absence of mineralized mandibular or maxillary bones.^(14,15) Consistent with this notion, tooth development is arrested at the late bud stage in Runx2-null mice that exhibit a complete lack of craniofacial mineralization. Sp7 begins to express in craniofacial bones at E13.5.⁽²⁰⁾ Normal tooth morphogenesis occurred in Sp7-null mice despite a complete failure of mineralization of the craniofacial bones. Therefore, we report for the first time that tooth morphogenesis can progress in the absence of mineralized alveolar bone. The craniofacial region is hypoplastic in both Runx2-null and Sp7-null mice. The primary palate is formed normally in Runx2-null mice. However, the secondary palate fails to fuse resulting in a cleft palate.⁽¹⁸⁾ We show that Sp7 is expressed in the palate but Sp7 deficiency does not disrupt palate development, and subsequent fusion of the palatal shelves. Interestingly, Sp7-null mice exhibit a high arched palate and the primary and secondary palates do not mineralize. Our results are consistent with the high arched palate phenotype noted in human patients carrying a homozygous frameshift mutation in the SP7 gene.⁽²¹⁾ Thus Sp7 and Runx2 play different regulatory role in the development of orofacial structures.

Mutations in the RUNX2 gene are associated with cleidocranial dysplasia (CCD), characterized by delayed or failure of tooth eruption.⁽³³⁾ Dental follicle cells (DFCs) play a key role in tooth eruption. Runx2 expression is essential for differentiation of DFCs.^(34,35) The phenotype of CCD patients suggests that Runx2 may play an essential role in the postnatal eruption of teeth. Indeed, tooth eruption is delayed in the Runx2 heterozygous mice.⁽³⁶⁾ In sharp contrast, one copy of the Sp7 gene is sufficient to promote proper development and the postnatal eruption of teeth. Interestingly, homozygous mutation in the human SP7 gene is associated with osteogenesis imperfecta, and delayed eruption of teeth but not CCD.⁽²¹⁾ Thus Sp7 and Runx2 display different functional requirements during tooth development. Runx2 is required for the commitment of the ectomesenchyme and epithelium toward odontoblast and ameloblast lineages, whereas Sp7 is responsible for the terminal differentiation of both cell types.

The Sp7-deficient ectomesenchyme commits to ameloblast and odontoblast lineages but both cell layers remain unpolarized, indicating a failure of differentiation. Expression of collagen type 1 and Alp mRNA albeit at reduced level indicate Sp7-null odontoblasts undergo initial differentiation. However, the expression of late odontoblast marker genes Dspp, Oc, and Dmp1 were nearly absent in Sp7-null tooth organs. Interestingly, marker genes of presecretory and secretory ameloblasts such as Amelx, Enam, Amelt, Mmp20, and Klk4 were barely detected in the Sp7-null mice. These results confirm that Sp7 regulates functional differentiation of dentin and enamel synthesizing cells in vivo. Our in silico analysis revealed the presence of multiple high-affinity Sp consensus motifs in the initial ~2-kb promoter fragment of Col1, Dspp, Oc, and Dmp1 genes. The relative position of several of the Sp binding motifs in mouse, rat and human gene promoters was conserved. Promoter-reporter assays confirmed that Sp7 directly binds and activates the promoter of Oc and Dspp genes. Our findings are consistent with recent reports that show Sp7 regulate transcription of several odontoblast genes during postnatal odontogenesis.^(37,38)

Because of the perinatal lethality of Sp7-null mice, we were unable to study regulatory requirement of Sp7 during postnatal development of other structures of the tooth organs such as root dentin, cementum, and the periodontal ligament (PDL). However, several recent studies have taken advantage of conditional gene ablation to uncover the role of Sp7 in the development and maintenance of postnatal dental tissue. Deletion of Sp7 gene in preodontoblasts using –2.3-kb Col1-Cre results in impaired synthesis of root and interradicular dentin. In situ hybridization demonstrated reduced expression of Col1, Dspp, and Dmp1 in 6-week-old Sp7 conditional null mice.⁽³⁷⁾ Sp7 deficiency in preodontoblasts also resulted in a lack of Alp, Oc, Phex, and Nestin expression.⁽³⁸⁾ Selective ablation of the Sp7 gene in mature odontoblasts using Oc-Cre mice impairs dentin synthesis marked by decreased expression of Dspp, Oc, Phex, and Nestin genes. Thus Sp7 continues to regulate gene transcription in mature odontoblasts.

Sp7 deletion in collagen type I–positive cells also identified a role of Sp7 in the formation of root and cellular cementum in the postnatal tooth.⁽³⁹⁾ Sp7-null mutants presented short molar roots, extremely hypoplastic interradicular dentin, and poorly differentiated odontoblasts.⁽³⁸⁾ Moreover, the number of cementocytes, mineralization rate, and cellular cementum mass was significantly reduced by 6 weeks of age. Sp7-deficient cementocytes showed a marked reduction in expression of Bsp and Dmp1.⁽³⁹⁾ Thus, in addition to the critical role of Sp7 during embryonic tooth development, it is also required for the formation and differentiation of mesenchymally derived tissues during postnatal odontogenesis.

Expression of Sp7 is seen throughout the development of osteoblasts and alveolar bone.⁽²⁰⁾ Global deficiency of Sp7 resulted in the failed differentiation of ameloblasts and odontoblasts, suggesting functional requirements of Sp7 in both dentinogenesis and amelogenesis. We examined Sp7 expression in various cell types during tooth embryonic development by utilizing the lac-Z knock-in allele in the Sp7^−/− mouse model. The activity of β-galactosidase in the Sp7^−/− model reflects an expression of the endogenous Sp7 gene. Continuously growing incisors were selected to profile the expression of Sp7 in epithelial ameloblasts and mesenchymal odontoblasts as they progress from immature precursors to fully mature secretory cells. The Sp7 gene was not expressed in progenitor cells present near the labial and lingual cervical loop. However, the Sp7 gene is turned on in dental pulp cells and in preodontoblasts and continues to express through subsequent stages of odontoblast differentiation. Consistent with our results, other groups have reported expression of Sp7 in the crown and the root odontoblasts in 2-week-old to 4-week-old mice.^(38,39) Sp7 is also expressed in the cementoblasts, cementocytes, and PDL cells of 3-week-old to 6-week-old mice. The lineage tracing experiments using Osx-Cre mice have further confirmed expression of Sp7 in mesenchymally derived odontoblasts, dental pulp, cementoblasts, and PDL cells.⁽⁴⁰⁾ Interestingly, during embryonic development, Sp7 is not expressed in mesenchyme-derived dental follicle cells or any cells of epithelial origin including ameloblasts.

The discovery of impaired ameloblasts differentiation, despite mesenchyme and odontoblasts restricted expression of Sp7, prompted us to study functional control through cross signaling. Mesenchyme-derived and epithelium-derived paracrine and autocrine signaling by Fgf3 and Fgf8 proteins play a critical role in tooth development.^(6,7) Expression of both Fgf ligands was nearly absent in the developing incisors and molars of Sp7-null mice. We further established that Sp7 directly binds and upregulates transcription of the Fgf3 promoter. Our data is consistent with earlier studies showing Runx2 promotes Fgf3 and Fgf4 signaling during tooth development.⁽⁴¹⁾ These results strongly suggest that failed proliferation and differentiation of ameloblasts and odontoblasts in Sp7-null mice is caused by direct disruption of Fgf signaling. In this regard, Fgf3 acts as a potent mitogen in several epithelial-derived and mesenchyme-derived tissues.⁽⁴²⁾ Fgf3 regulates the size, number, position, and interrelation of cusps during murine tooth development. Homozygous mutations in the FGF3 gene leads to LAMM syndrome that is characterized by microdontia and conical teeth.^(43-44) However, future studies are warranted to define the complex interaction between Sp7 and Fgf pathway cross-signaling in odontogenic epithelium and mesenchyme during tooth development.