Wnt Signaling Regulates Pulp Volume and Dentin Thickness


  • Won Hee Lim,

    1. Division of Plastic and Reconstructive Surgery, Department of Surgery, Stanford School of Medicine, Stanford, CA, USA
    2. Department of Orthodontics, School of Dentistry & Dental Research Institute, Seoul National University, Seoul, Korea
    Search for more papers by this author
  • Bo Liu,

    1. Division of Plastic and Reconstructive Surgery, Department of Surgery, Stanford School of Medicine, Stanford, CA, USA
    Search for more papers by this author
  • Du Cheng,

    1. Division of Plastic and Reconstructive Surgery, Department of Surgery, Stanford School of Medicine, Stanford, CA, USA
    Search for more papers by this author
  • Daniel J Hunter,

    1. Division of Plastic and Reconstructive Surgery, Department of Surgery, Stanford School of Medicine, Stanford, CA, USA
    Search for more papers by this author
  • Zhendong Zhong,

    1. Center for Musculoskeletal Health, Internal Medicine Department, University of California at Davis, Medical Center, Sacramento, CA, USA
    2. Center for Skeletal Disease Research and Laboratory of Cell Signaling and Carcinogenesis, Van Andel Research Institute, Grand Rapids, MI, USA
    Search for more papers by this author
  • Daniel M Ramos,

    1. Department of Orofacial Sciences, University of California at San Francisco, San Francisco, CA, USA
    Search for more papers by this author
  • Bart O Williams,

    1. Center for Skeletal Disease Research and Laboratory of Cell Signaling and Carcinogenesis, Van Andel Research Institute, Grand Rapids, MI, USA
    Search for more papers by this author
  • Paul T Sharpe,

    1. Department of Craniofacial Development, GKT Dental Institute, London Bridge, London, UK
    Search for more papers by this author
  • Claire Bardet,

    1. Pôles de Recherche et D'enseignement Supérieur (PRES), Université Paris Descartes, Montrouge, France
    Search for more papers by this author
  • Su-jung Mah,

    1. Department of Orthodontics, Kyung Hee University Hospital at Gangdong, Seoul, Korea
    Search for more papers by this author
  • Jill A Helms

    Corresponding author
    1. Division of Plastic and Reconstructive Surgery, Department of Surgery, Stanford School of Medicine, Stanford, CA, USA
    • Address correspondence to: Jill A Helms, DDS, PhD, Division of Plastic and Reconstructive Surgery, Department of Surgery, Stanford School of Medicine, Stanford, CA 94305, USA. E-mail: jhelms@stanford.edu

    Search for more papers by this author


Odontoblasts, cementoblasts, ameloblasts, and osteoblasts all form mineralized tissues in the craniofacial complex, and all these cell types exhibit active Wnt signaling during postnatal life. We set out to understand the functions of this Wnt signaling, by evaluating the phenotypes of mice in which the essential Wnt chaperone protein, Wntless was eliminated. The deletion of Wls was restricted to cells expressing Osteocalcin (OCN), which in addition to osteoblasts includes odontoblasts, cementoblasts, and ameloblasts. Dentin, cementum, enamel, and bone all formed in OCN-Cre;Wlsfl/fl mice but their homeostasis was dramatically affected. The most notable feature was a significant increase in dentin volume and density. We attribute this gain in dentin volume to a Wnt-mediated misregulation of Runx2. Normally, Wnt signaling stimulates Runx2, which in turn inhibits dentin sialoprotein (DSP); this inhibition must be relieved for odontoblasts to differentiate. In OCN-Cre;Wlsfl/fl mice, Wnt pathway activation is reduced and Runx2 levels decline. The Runx2-mediated repression of DSP is relieved and odontoblast differentiation is accordingly enhanced. This study demonstrates the importance of Wnt signaling in the homeostasis of mineralized tissues of the craniofacial complex. © 2014 American Society for Bone and Mineral Research.


Wnt signals play decisive roles in the development and homeostasis of mineralized tissues, and a key strategy for demonstrating these functions has been to study how gain-of-function and loss-of-function mutations in components of the Wnt signal transduction pathway affect osteogenesis. For example, reductions in Wnt signaling, either by overexpression of Wnt antagonists, a deficiency of Wnt ligands, or mutations in genes encoding Wnt receptors, cause bone loss in humans and in mice.[1-3] Conversely, enhancement of Wnt signaling results in increased bone volume, abnormal bone density (ie, hyperostosis), and pathological thickening of bone, a condition known as sclerosing bone dysplasia.[4-7] Wnt signaling also influences the program of odontogenesis, but much less is understood about its role in this complex process. For example, Wnt signaling controls morphological aspects of tooth development including the number,[8] size,[9] position,[10] and shapes[11-15] of teeth (reviewed in Thesleff[16]). Loss-of-function experiments suggest that Wnt signals regulate the maturation of dental mesenchyme into dentin-secreting odontoblasts[12] and bone-secreting cementoblasts,[17] but there are also conflicting reports.[18]

Much less is known about the postnatal functions of Wnt signaling in maintaining mineralized tissue homeostasis. Although Wnt signaling remains a critical regulator of bone homeostasis throughout life,[19, 20] the function of postnatal Wnt signaling in the maintenance of dental mineralized tissues is unclear. A standard knockout approach to abrogate Wnt signaling is complicated by the fact that there are 19 closely related mammalian Wnt genes which are largely redundant.[21-23] To overcome this difficulty we made use of a Cre-LoxP system to conditionally delete Wntless (Wls), a chaperone protein that escorts lipid-modified Wnt from the Golgi to the cell surface.[24] When Wls is inactivated, the secretion of all mammalian Wnt proteins is blocked[24, 25] and Wnt signaling is abrogated.[26] To block Wnt signaling specifically in mineralized tissues, Wlsfl/fl mice were crossed with OCN-Cre mice to generate OCN-Cre;Wlsfl/fl offspring.[27] Osteocalcin is a major noncollagenous protein synthesized by osteoblasts, odontoblasts, ameloblasts, and cementoblasts.[28] Consequently, these mice allowed us to focus on how loss of Wnt protein secretion from cells affects mineralized tissues in the head, regardless of whether they originated from the somatic mesoderm, paraxial cephalic mesoderm, or cranial neural crest.[29, 30]

Subjects and Methods

Generation of OCN-Cre;Wlsfl/fl mice

The generation of OCN-Cre;Wlsfl/fl mice was performed after review and approval by the Institutional Animal Care and Use Committee (IACUC) of the Van Andel Research Institute (protocol #13-03-015). Both OCN-Cre and Wlsfl/fl mice are available at Jackson Laboratories (Bar Harbor, ME, USA) (129S-Wlstm1.1Lan/J, 012888 for Wlsfl/fl and B6N PVB-Tg1Clem/J, 019509 for OCN-Cre). Genomic DNA was prepared from tail biopsies using an AutoGenprep 960 automated DNA isolation system (AutoGen). PCR-based strategies were used to genotype the mice (details are available upon request). Thirty mice were analyzed; 10 were 1 month old and 20 were 3 months old.

Generation of Axin2LacZ/+ mice; detection of beta galactosidase activity

The generation of Axin2LacZ/+ mice (Jackson Laboratories 129P2-Axin2tm1Wbm/J) was performed in accordance with animal welfare based on an approved IACUC protocol #13146 from Stanford University. Ten 2-month-old mice were used in this study. Cells responsive to Wnt signaling express the LacZ gene product, beta-galactosidase. Beta galactosidase is then detected by X-gal staining.[31] To perform X-gal staining, tissues were fixed in 0.4% paraformaldehyde overnight before being decalcified with 19% EDTA and infused with 30% sucrose for 24 hours. Samples were embedded in optimum cutting temperature (OCT) medium and cryosectioned at a thickness of 8 µm. Tissues were then fixed in 0.2% gluteraldehyde for 15 minutes and stained with X-gal overnight at 37°C.

Micro–computed tomography analyses

Micro–computed tomography (µCT) was performed on 10 mice at the age of 3 months (5 wild-type, 5 OCN-Cre;Wlsfl/fl littermates) using Imtek/Siemens MicroCAT II/SPECT system (Siemens, Knoxville, TN, USA) at 80 kV, 400 µA, and a resolution of 40 µm. Scans of the skulls were acquired using an isotropic voxel size of 6400 µm3, with 650 CT slices evaluated per skull. To determine the mineralized bone fraction, the threshold was standardized with the bone/water/air phantom for each sample. Individual CT slices were reconstructed with COBRA reconstruction software, and data were analyzed with GE Microview (GE Healthcare, Waukesha, WI, USA). The frontal, parietal, and occipital bones were selected automatically with the threshold within the region of interest including each of the areas.

µCT analyses of the teeth in 10 mice at the age of 3 months (5 wild-type, 5 OCN-Cre;Wlsfl/fl littermates) were taken using MicroXCT-200 (SkyScan, Belgium) at 60 kV, 7.98 W, and a resolution of 2 µm. Scans were acquired using an isotropic voxel size of 8 µm3, with 800 CT slices evaluated in the incisor area. For analyses, individual CT slices were reconstructed with MicroXCT7.0 reconstruction software (SkyScan, Belgium), and data were analyzed with Inveon Research Workplace (IRW) (Erlangen, Germany).

Sample preparation, processing, and histology

Maxillae from 1-month-old mice (5 wild-type, 5 OCN-Cre;Wlsfl/fl littermates) and 3-month-old mice (6 wild-type, 6 OCN-Cre;Wlsfl/fl littermates) were harvested and fixed in 4% paraformaldehyde overnight at 4°C. Samples were decalcified in a heat-controlled microwave in 19% EDTA for 2 weeks. After demineralization, specimens were dehydrated through an ascending ethanol series prior to paraffin embedding. Then 8-µm-thick longitudinal sections were cut and collected on Superfrost-plus slides for histology.

In situ hybridization

Tissue sections were deparaffinized following standard procedures. Relevant digoxigenin-labeled mRNA antisense probes were prepared from complementary DNA templates for Osteocalcin, Axin2, and Osterix. Sections were dewaxed, treated with proteinase K, and incubated in hybridization buffer containing the relevant RNA probe. The probe was added at an approximate concentration of 1 µg/mL. Stringency washes of saline sodium citrate solution were done at 65°C and further washing was done in maleic acid buffer with 1% Tween 20. Slides were treated with an antibody to anti-digoxigenin-AP (Roche). For color detection, slides were incubated in nitro blue tetrazolium chloride (Roche) and 5-bromo-4-chloro-3-indolyl phosphate (Roche). After developing, the slides were coverslipped with permount mounting medium.


Movat's pentachrome staining was performed.[32] Nuclei stain blue to black, cytoplasm stains red, collagen stains yellow to greenish yellow, and fibrous tissue stains an intense red. Tissues were also stained with the acidic dye, Picrosirius red,[33] to discriminate tightly packed and aligned collagen molecules. Under polarized light, well-aligned fibrillary collagen molecules present polarization colors of longer wavelengths (red) as compared to less organized collagen fibrils that show colors of shorter wavelengths (green-yellow).


Tissue sections were deparaffinized following standard procedures. Endogenous peroxidase activity was quenched by 3% hydrogen peroxide for 5 minutes, and then washed in PBS. Slides were blocked with 5% goat serum (Vector S-1000) for 1 hour at room temperature. The appropriate primary antibody was added and incubated overnight at 4°C, then the slides were washed in PBS. Samples were incubated with appropriate biotinylated secondary antibodies (Vector BA-x) for 30 minutes, and washed in PBS. An avidin/biotinylated enzyme complex (Kit ABC Peroxidase Standard Vectastain PK-4000) was added and incubated for 30 minutes and a DAB substrate kit (Kit Vector Peroxidase substrate DAB SK-4100) was used to develop the color reaction. Antibodies used include Ki67 (Thermo Scientific; dilution 1: 100), Runx2 (Origene; dilution 1: 200), Osteopontin (NIH LF 175; dilution 1: 4000), dentin sialoprotein (DSP; Millipore; dilution 1: 2000), dentin phosphoprotein (DPP; generated by the Department of Dental Science for Health Promotion, Division of Cervico Gnathostomatology, Hiroshima University, Hiroshima, Japan; dilution 1: 2000). Counterstain was performed with hematoxylin after development with a DAB substrate in order to count cells.

Quantitative real-time PCR

Dental pulp tissues were carefully removed from formalin-fixed paraffin-embedded (FFPE) tissue and transferred into the processing plate. RNA isolation was carried out using FFPE RNA Isolation kit (Ambion, Austin, TX, USA). Digestion buffer, protease and RNA digestion additives were added into the processing plate and incubated at 60°C for 45 minutes and at 80°C for 30 minutes. Nucleic acid binding beads were added to the samples, then the samples were placed on a titer shaker for 3 minutes; afterward, the supernatant was removed. Beads were washed again and the supernatant was again removed. Samples were treated with Dnase at 37°C for 20 minutes, then the beads were washed and the supernatant was removed. RNA was quantified using a Nanodrop spectrophotometer (Thermo Fischer Scientific, DE, USA). cDNA was synthesized using Superscript first-strand synthesis supermix following the manufacturers instructions (Life Technologies, Austin, TX, USA). Quantitative RT-PCR was performed as described.[34] Expression levels were calculated using the 2−ΔΔCt comparative cycle threshold method, normalized to GAPDH[35] and converted to fold expression. The following primer sets were used: GAPDH, forward primer (ACCCAGAAGACTGTGGATGG) and reverse primer (GGATGCAGGGATGATGTTCT). Runx2, forward primer (ATCGCCTCAGTGATTTAGGG) and reverse primer (TGCCTGGGATCTGTAATCTG). DSP, forward primer (TGGCTGTGCCTCTTCTAACA) and reverse primer (GCTGTTGCTAGTGGTGCTGT).

Statistical analyses

Results are presented as the mean ± SD. Student's t test was used to quantify differences described in this article. One asterisk (*) denotes a p value <0.05, and two asterisks (**) denotes a p value of <0.01.


Skeletal tissues have a universal requirement for Wnt signaling

In a first study we demonstrated that Cre-mediated recombination in OCN-Cre mice occurred in bone.[36] Here, examination of the craniofacial skeleton of OCN-Cre;Wlsfl/fl mice revealed that similar to the appendicular and axial skeleton, deletion of Wls caused a dramatic reduction in bone volume and bone mineral density of both cranial neural crest-derived skeletal elements (bone volume: p < 0.01; bone density: p < 0.05; Fig. 1A, B) and mesoderm-derived skeletal elements (bone volume and bone density: p < 0.01; Fig. 1C, D; quantified in Fig. 1E). Deletion of Wls, however, did not affect the size of the skeletal elements (Supplemental Fig. 1).

Figure 1.

Downregulation of Wnt signaling results in reduced bone density in skeleton but intact dentition of OCN-Cre;Wlsfl/fl mice. (A, B) The deletion of Wls causes significant reduction in bone volume and bone density of OCN-Cre;Wlsfl/fl frontal bone. (C, D) Substantially reduced bone volume and bone density are observed in OCN-Cre;Wlsfl/fl parietal and occipital bones. (E) Quantification reveals a statistically significant reduction in bone volume and bone density in OCN-Cre;Wlsfl/fl craniofacial skeleton. (F, G) Overall volume of the teeth is equivalent between the wild-type and OCN-Cre;Wlsfl/fl mice. (H, I) Gross morphology and appearance of molars are equivalent between the wild-type and OCN-Cre;Wlsfl/fl mice. (J, K) Wild-type and OCN-Cre;Wlsfl/fl incisors appear to have equivalent size and appearance. *p < 0.05 and **p < 0.01.

Unlike the dramatic skeletal phenotype, the dentition of OCN-Cre;Wlsfl/fl mice was largely intact. For example, the overall size, shape, and position of the teeth was equivalent between wild-type and OCN-Cre;Wlsfl/fl mice (Fig. 1F, G; Supplemental Fig. 2A,B). The gross morphology of the molars (Fig. 1H, I) and incisors (Fig. 1J, K) was similar in wild-type and mutant mice. We confirmed that in addition to its expression in osteoblasts, Osteocalcin is also expressed by ameloblasts that form enamel (Fig. 2B),[37] by cementoblasts that line the root surface (Fig. 2C),[38, 39] by odontoblasts that produce dentin,[40-42] and by pulp cells that maintain the stroma of the pulp cavity (Fig. 2D).[43] In addition, the expression of Osteocalcin in the inner enamel epithelium was first clearly observed at E18.5.

Figure 2.

A disruption of Wnt signaling causes anomalies in the OCN-Cre;Wlsfl/fl mineralized dental tissues. (A) Osteoblasts, (B) ameloblasts, (C) cementoblasts, and (D) odontoblasts and pulp cells of 3-month-old wild-type mice show osteocalcin expression (purple color). The brown color is produced by red blood cells due to abundant iron porphyrin complexes. (E, F) In 3-month-old OCN-Cre;Wlsfl/fl mice, enamel covers more root surface including near the incisor apex than wild-type mice. (G, H) Wild-type mice lack enamel in region closer to the apex, whereas OCN-Cre;Wlsfl/fl mice have a well-mineralized enamel but thin alveolar bone surrounding the incisor. (I, J) The total volume of the tooth remains constant between wild-type and OCN-Cre;Wlsfl/fl mice. (L, M) The dentin volume is significantly increased in OCN-Cre;Wlsfl/fl mice near the incisive edge. (K) Quantification reveals equivalent tooth volume but a statistically significant increase in dentin volume in OCN-Cre;Wlsfl/fl mice. Scale bar: (AD) 50 μm, (IM) 50 μm.

Postnatal ameloblasts, odontoblasts, and pulp cells are Wnt-responsive

Given the Osteocalcin expression patterns and the fact that Wnt signaling is implicated in the development of each of these dental tissues, we more closely examined the dentition in OCN-Cre;Wlsfl/fl mice for anomalies resulting from a disruption of Wnt signaling. We focused on the incisors of OCN-Cre;Wlsfl/fl mice and in a separate, ongoing study we evaluated the molar phenotypes.

Using 2-μm-resolution µCT, reconstructions of the incisors revealed that removal of Wls in Osteocalcin-expressing ameloblasts resulted in increased enamel production (arrows, Fig. 2E, F) without disrupting the overall size or shape of the incisor (Supplemental Fig. 2). Normally, enamel covers about 75% of the buccal incisor surface; in OCN-Cre;Wlsfl/fl mice enamel covered about 90% of the surface. Transverse µCT sections near the incisor apex (dotted lines, Fig. 2E, F) illustrated this point: mineralized enamel is normally detected ∼1.43 μm from the root apex; regions closer to the apex therefore lack enamel in wild-type mice (Fig. 2G). In OCN-Cre;Wlsfl/fl mice, however, more apical sections showed a well-mineralized enamel matrix (arrow, Fig. 2H). This gain in enamel mineralization stood in sharp contrast to the thin alveolar bone surrounding the OCN-Cre;Wlsfl/fl incisor (asterisks, Fig. 2G, H).

Removal of Wls from odontoblasts also resulted in a significant gain in mineralized tissue formation. The total volume of the tooth remained constant between wild-type and OCN-Cre;Wlsfl/fl specimens (p = 0.08 at root apex and p = 0.06 at incisive edge; Fig. 2I, J; quantified in Fig. 2K; see also Supplemental Fig. 2C,D) but serial sections through the incisor clearly showed that in OCN-Cre;Wlsfl/fl mice, the percentage of tooth structure occupied by dentin was significantly increased (p < 0.01; Fig. 2L, M; quantified in Fig. 2K). This increase in dentin volume was compensated by a reduction in pulp volume in OCN-Cre;Wlsfl/fl mice, most obviously at the incisive edge (dotted circle, p < 0.01; Fig. 2L, M; quantified in Fig. 2K; see also Supplemental Fig. 2B). The molecular mechanisms underlying this gain in dentin volume became the focus of our next experiments.

Adult odontoblasts and pulp cells maintain their dependency on endogenous Wnt signaling

The gain in dentin volume we observed in OCN-Cre;Wlsfl/fl mice is predicated on odontoblasts maintaining a dependency on Wnt signaling into adulthood. Previous reports, however, suggest that after postnatal day 15, molar odontoblasts and odontoblasts at the incisor tip lose their Wnt responsiveness.[44] We revisited this issue of adult odontoblast dependency on a Wnt signal using the same strain of Wnt reporter (eg, Axin2LacZ/+) mice and cryosectioned tissues from 2-month-old mice. We found that both incisor tip odontoblasts and pulp fibroblasts were X-gal positive (Fig. 3A, B). We examined the other mineralizing dental tissues and confirmed the Wnt-responsive status of cementoblasts (Fig. 3C, D), ameloblasts (Fig. 3E, F), and osteoblasts (Fig. 3G, H) in skeletally mature Axin2LacZ/+ mice. Thus, all mineralizing tissues in the craniofacial complex maintain their Wnt-responsive status into adulthood.

Figure 3.

Wnt responsiveness is maintained until adulthood. (A, B) Odontoblasts and pulp fibroblasts in 2-month-old Axin2LacZ/+ incisors are X-gal staining positive. (C, D) Cementoblasts, (E, F) ameloblasts, and (G, H) osteoblasts are Wnt-responsive in 2-month-old Axin2LacZ/+ mice. Scale bar: (AD) 50 μm.

Differential regulation of Runx2 mediates OCN-Cre;Wlsfl/fl bone loss and dentin gain

We sought to reconcile two apparently disparate findings, that reduced Wnt signaling in OCN-Cre;Wlsfl/fl mice simultaneously produces a loss in bone volume (Fig. 1) and a gain in dentin volume (Fig. 2). We first verified that Wnt signaling was reduced in the pulp of OCN-Cre;Wlsfl/fl mice, using expression of the Wnt target gene Axin2 as a readout of pathway activity.[45, 46] Compared to its expression in wild-type osteoblasts, Axin2 expression was reduced to nearly undetectable levels in OCN-Cre;Wlsfl/fl osteoblasts (Fig. 4A, B). A similar reduction in Axin2 expression was observed in OCN-Cre;Wlsfl/fl odontoblasts (compare wild-type, Fig. 4C with Fig. 4D).

Figure 4.

Downregulation of Wnt signaling results in reduction of Axin2 expression. (A, B) Expression of Axin2 is reduced in OCN-Cre;Wlsfl/fl osteoblasts compared to wild-type osteoblasts. (C, D) Odontoblasts in OCN-Cre;Wlsfl/fl mice shows reduction of Axin2 expression compared to wild-type mice. Scale bar: (A, B) 50 μm; (C, D) 50 μm.

Despite significant increases in dentin volume and density, the organization of the dentin matrix in OCN-Cre;Wlsfl/fl appeared to be the same in wild-type mice (Fig. 5A–C). Ki67 immunostaining was nonexistent in the pulp cavities of both wild-type and mutant mice (Fig. 5D, E), so it was unlikely that the increased dentin volume was attributable to differences in proliferation of OCN-Cre;Wlsfl/fl odontoblasts. We also examined the protein expression of DPP, DSP, Osteopontin, Osterix, and Runx2 (Supplemental Fig. 3), and of these dentin markers we only noticed a difference in Runx2 expression levels.

Figure 5.

Bone loss and dentin gain in OCN-Cre;Wlsfl/fl mice result from Runx2 alteration. (A, B) Picrosirius red staining shows similar dentin matrix organization between wild-type and OCN-Cre;Wlsfl/fl mice. (C) OCN-Cre;Wlsfl/fl mice has significantly denser dentin compared to wild-type mice. (D, E) Ki67 expression shows no difference in cell proliferation of odontoblasts and pulp between wild-type and OCN-Cre;Wlsfl/fl mice. The small figure in D shows Ki67 expression in rugae area. (F, G) Significantly reduced Runx2 expression is observed in OCN-Cre;Wlsfl/fl odontoblasts and pulp. (H) Using immunostaining and cell nuclei counting, significantly higher level of Runx2 was observed in wild-type mice compared to OCN-Cre;Wlsfl/fl mice (unit, fold difference relative to WT). (I, J) DSP expression is considerably increased in odontoblasts and pulp of OCN-Cre;Wlsfl/fl mice. (K) Using qRT-PCR, DSP expression in OCN-Cre;Wlsfl/fl mice is significantly higher than wild-type mice (unit, fold difference relative to WT). **p < 0.01. Scale bar: (A, B, D, E, I, J) 50 μm; (F, G) 50 μm. DSP = dentin sialoprotein; WT = wild type.

Compared to wild-type odontoblasts, the protein was noticeably reduced in OCN-Cre;Wlsfl/fl odontoblasts (Fig. 5F, J). Using immunostaining and cell nuclei counting, we observed significantly higher level of Runx2 expression in wild-type mice compared to OCN-Cre;Wlsfl/fl mice (Fig. 5H), although we observed statistically insignificant change in qRT-PCR. Concomitant with Runx2 downregulation, DSP expression was markedly increased in OCN-Cre;Wlsfl/fl odontoblasts (Fig. 5I, J). Using qRT-PCR, we showed a statistically significant, 1.7-fold increase in DSP expression in the OCN-Cre;Wlsfl/fl mice (Fig. 5K). This is in keeping with reports demonstrating that Runx2 negatively regulates the differentiation of cells into odontoblasts[47] via inhibition of DSP.[47] Taken together, the loss in bone volume and the gain in dentin volume caused by the reduction of Wnt signaling in OCN-Cre;Wlsfl/fl mice can be jointly explained by reduced Runx2 expression. In osteoblasts, a reduction in Runx2 results in an arrest in differentiation; conversely, reduced Runx2 expression in odontoblasts results in an acceleration in differentiation.


We used a conditional loss-of-function strategy to test the requirement for Wnt signaling in mineralized tissue homeostasis. In examining the skeleton of OCN-Cre;Wlsfl/fl mice, we found that regardless of embryonic origin, function, or role in weight bearing, all bones appeared to be equally affected by the loss of Wls (Fig. 1; also see Zhong and colleagues[36]). Thus, there appears to be a universal requirement for Wnt signaling for the maintenance of bone mineral density and bone volume, a conclusion strongly supporting by the existing literature.[48] There is not, however, a corresponding, well-defined role for Wnt signaling in odontogenesis. This became the subject of our investigation.

A reduction in Wnt signaling is associated with a gain in dentin formation

Wnt signaling controls tooth shape, size, position, and number; consequently, we anticipated that OCN-Cre;Wlsfl/fl mice would exhibit a tooth-related phenotype. We were surprised to discover that there was no overt disruption in odontogenesis (Figs. 1, Fig. 22; Supplemental Fig. 2). µCT reconstructions revealed that instead of a loss, there was actually a gain in dentin volume (Fig. 2). How could a reduction in Wnt signaling simultaneously lead to bone loss, and dentin accumulation?

We initially suspected that the negative impact of Wls removal did not extend to odontoblasts and pulp cells because of the conditional nature of the deletion. We verified, however, that adult odontoblasts and pulp cells, as well as ameloblasts and osteoblasts, expressed Osteocalcin (Fig. 2; see Papagerakis and colleagues[43]). Therefore, these cell types would be subjected to Wls deletion in OCN-Cre;Wlsfl/fl mice.

We then wondered if dental tissues escaped the negative effects of Wls removal, because Wnt responsiveness is thought to be a feature of fetal and early postnatal odontoblasts but not of adult cells.[44] We used the same Axin2LacZ/+ reporter mice as Lohi and colleagues,[44] but instead of paraffin embedding, we used cryoembedding. In these adult tissues X-gal staining clearly revealed that odontoblasts, pulp fibroblasts, ameloblasts, and cementoblasts, along with periodontal ligament cells all maintained their Wnt responsive status into adulthood (Fig. 3). Therefore, odontoblasts and pulp cells require Wnt signaling and in OCN-Cre;Wlsfl/fl mice, the requisite Wnt signal would not be secreted and both cell types would be devoid of this stimulus.

Wls-deficient cells maintain their ability to respond to a Wnt stimulus (authors AZ and BOW, unpublished observations); therefore, it is formally possible that OCN-Cre;Wlsfl/fl odontoblasts and pulp cells still receive—and respond to—a Wnt signal and thus the increased mineral density is not related to decreased Wnt signaling. The reduced Axin2 expression in the mutant pulp cavity, however, argues against this possibility (Fig. 4). Thus, we conclude that whatever its source, Wnt signaling is abrogated in the pulp cavity of OCN-Cre;Wlsfl/fl mice, and this reduction in Wnt signaling is directly responsible for a gain in dentin formation.

Wnt regulates Runx2 in osteoblasts and odontoblasts

The function of Runx2 in the differentiation of odontoblasts and osteoblasts is still a point of much debate (reviewed in Komori[49]). Initially, Runx2/Cbfa1 was thought to specifically regulate osteoblast maturation,[50] but this was eventually disproven. Investigators have since demonstrated key roles for Runx2 in chondrocyte maturation[51] as well as many other cell types.[52] James and colleagues[53] were the first to identify a function for Runx2 in odontoblasts, where its action appeared to inhibit the maturation process. New insights into the function of Runx2 came with the discovery that Wnt signaling promotes osteogenesis by stimulating Runx2 activity,[54] but whether it also serves as positive stimulus in odontoblasts has not been resolved. We showed that Wnt-deficient OCN-Cre;Wlsfl/fl osteoblasts have reduced Runx2 expression (Supplemental Fig. 3G, H) and that reduced Runx2 expression is accompanied by reduced Osteocalcin expression and dramatic bone loss (Fig. 1, Supplemental Fig. 3I, J). Wnt-deficient OCN-Cre;Wlsfl/fl odontoblasts also exhibit reduced Runx2 expression (Fig. 5), but then the differentiation process diverges: reduced Runx2 expression in OCN-Cre;Wlsfl/fl odontoblasts is accompanied by an increase in DSP expression (Fig. 5) with no significant change in Osteocalcin expression (Supplemental Fig. 3K, L). When these data are considered together, they support a model whereby Wnt signaling acts a positive stimulus for Runx2 expression in both odontoblasts and osteoblasts, but its transcriptional activity in these two cell types differs. In osteoblasts, Runx2 drives maturation, whereas in odontoblasts Runx2 appears to inhibit differentiation. This interpretation is in keeping with other reports examining the function of Runx2 in mineralizing tissues.[53]

Does Wnt signaling regulate odontoblast differentiation?

OCN-Cre;Wlsfl/fl mice exhibit stronger DSP expression and they have denser dentin (Fig. 5); presently the mechanisms behind this increase in dentin density are unknown. There are two plausible explanations: first, each odontoblast in OCN-Cre;Wlsfl/fl mice could secrete more dentin than a wild-type odontoblast; second, and alternatively, pulp cells in OCN-Cre;Wlsfl/fl mice might differentiate into odontoblasts more readily than wild-type pulp cells. The latter hypothesis is particularly intriguing because it provides a means by which to explain an increase in dentin volume without a commensurate increase in cell proliferation. If true, then pulp cells located in the subodontoblastic layer[55] of OCN-Cre;Wlsfl/fl mice may be responsible for generating more dentin, even without an injury stimulus.[56] A simple histologic examination of the subodontoblastic layer in wild-type and OCN-Cre;Wlsfl/fl mice, however, did not reveal any notable alterations, so this hypothesis remains speculative at the current time. Nonetheless, a role for Wnt signaling in the activation of progenitor cells that contribute to the repair of a mineralized tissue has indirect support from data on bone healing,[57] and may well also be involved in odontogenic repair.

Wnt signaling and cellular aging of the dental pulp

Throughout life odontoblasts continue to secrete an extracellular matrix that becomes mineralized[58-60] and as a consequence the pulp chamber narrows with age.[61] The molecular regulation of this process is unknown. We observe an exaggerated version of this same process in OCN-Cre;Wlsfl/fl mice: the pulp chamber reduces while the dentin gradually thickens (Fig. 2). Simultaneous with this gain in mineralized dentin we observe a profound loss in bone mass in OCN-Cre;Wlsfl/fl mice,[36] which is also a hallmark of human aging.[62, 63] In other work, we demonstrated that aging is associated with a decline in Wnt signaling, specifically in the bone marrow cavity.[64] Coupled with new data demonstrating that Wnt signaling regulates telomerase activity,[65, 66] which is tightly correlated with a decline in osteogenesis and the onset of cellular senescence,[67, 68] it is tempting to speculate that some of the changes we observe in the OCN-Cre;Wlsfl/fl dentition are an exaggerated version of an aging phenotype brought about by a gradual loss of Wnt signaling.[69] If this hypothesis holds true, some age-related effects on mineralized tissues of the body may be treatable via a Wnt-based approach.


All authors state that they have no conflicts of interest.


This research project was supported by a grant from the California Institute of Regenerative Medicine (CIRM) TR1-0219 and NIH ROI AR053293. Micro XCT imaging work was performed at the Division of Biomaterials and Bioengineering Micro-CT Imaging Facility, UCSF, supported by NIH Grant S10RR026645, and the Departments of Preventive and Restorative Dental Sciences and Orofacial Sciences, School of Dentistry, UCSF.

Authors' roles: Study design: WHL, BL, and JAH. Data collection: WHL, BL, DC, DJH, and DR. Data analysis: WHL, BL, DC, CB, and JAH. Drafting manuscript content: WHL, BL, DC, and JAH. Revising manuscript content: WHL, AZ, BW, DR, PS, SJM, and JAH. Approving final version of manuscript: WHL, BL, DC, DJH, AZ, DR, BW, PS, CB, SJM, and JAH. JAH takes responsibility for the integrity of the data analysis.