Disruption of c-Kit Signaling in KitW-sh/W-sh Growing Mice Increases Bone Turnover

c-Kit tyrosine kinase receptor has been identified as a regulator of bone homeostasis. The c-Kit loss-of-function mutations in WBB6F1/J-KitW/W-v mice result in low bone mass. However, these mice are sterile and it is unclear whether the observed skeletal phenotype is secondary to a sex hormone deficiency. In contrast, C57BL/6J-KitW-sh/W-sh (Wsh/Wsh) mice, which carry an inversion mutation affecting the transcriptional regulatory elements of the c-Kit gene, are fertile. Here, we showed that Wsh/Wsh mice exhibited osteopenia with elevated bone resorption and bone formation at 6- and 9-week-old. The c-Kit Wsh mutation increased osteoclast differentiation, the number of committed osteoprogenitors, alkaline phosphatase activity and mineralization. c-Kit was expressed in both osteoclasts and osteoblasts, and c-Kit expression was decreased in Wsh/Wshosteoclasts, but not osteoblasts, suggesting an indirect effect of c-Kit on bone formation. Furthermore, the osteoclast-derived coupling factor Wnt10b mRNA was increased in Wsh/Wsh osteoclasts. Conditioned medium from Wsh/Wsh osteoclasts had elevated Wnt10b protein levels and induced increased alkaline phosphatase activity and mineralization in osteoblast cultures. Antagonizing Wnt10b signaling with DKK1 or Wnt10b antibody inhibited these effects. Our data suggest that c-Kit negatively regulates bone turnover, and disrupted c-Kit signaling couples increased bone resorption with bone formation through osteoclast-derived Wnt 10 b.

were infertile due to a lack of germ cells in the ovary and had reduced estrogen and progesterone levels, leading to increased FSH level 19 . It is unclear whether the observed skeletal phenotype in W/W v mice resulted from cell-autonomous effects in osteoclasts or was a consequence of changes in sex hormone levels. In the present study, we focused on the skeletal phenotype of C57BL/6J-Kit W-sh/W-sh (W sh /W sh ) mice that were fully fertile and determined the mechanism by which altered c-Kit signaling affected bone turnover. Our data indicated that the c-Kit W sh mutation resulted in decreased cancellous bone volume with an increase in bone resorption and bone formation in growing mice. Calvarial osteoblasts derived from W sh /W sh mice showed an increase in osteoblast precursors, alkaline phosphatase (ALP) activity and mineralization in vitro. Moreover, the RANKL/OPG ratio was increased in osteoblasts derived from these mice, leading to increased number of osteoclasts in c-Kit mutants. Quantitative real-time PCR (qPCR) indicated that c-Kit expression was decreased in W sh /W sh bone marrow macrophage (BMM) and osteoclasts but not osteoblasts, suggesting that increased bone formation in W sh /W sh mice was not an osteoblasts intrinsic effect. Conditioned medium derived from W sh /W sh osteoclasts contained increased levels of the osteoclast-derived coupling factor Wnt10b, and enhanced ALP activity and mineralization by osteoblasts. Blocking Wnt10b activity inhibited these effects. These findings demonstrate that c-Kit regulates bone turnover by suppressing osteoblast and osteoclast differentiation. Thus, c-Kit mutation increased bone formation by increasing the generation of osteoclast-derived coupling factor Wnt10b.

W/W v mice exhibited osteopenic phenotype. Previous studies indicated that both male and female
W/W v mice are infertile 19,20 . To determine whether a sex hormone deficiency contributed to changes in the skeletal phenotype of growing W/W v mice, we first analyzed the skeletal phenotype of 6-week-old W/W v mice. Theses mutants were smaller compared with wild type (WT) littermates and their body weight was 20% less (24.50 ± 0.88 g in WT vs 19.49 ± 0.42 g in W/W v mice, p < 0.05). μ CT analysis showed a decrease in cancellous bone volume, trabecular thickness, trabecular number and connectivity density with a concomitant increase in trabecular separation ( Fig. 1A and Supplementary Table S1). Cross-sectional volume, cortical volume, and cortical thickness were also decreased in W/W v mice compared with controls. Histomorphometric analysis confirmed a significant decrease in cancellous bone volume (Fig. 1B). The cancellous bone was less dense and thinner in W/W v mice with decreased trabecular number and thickness and increased trabecular separation (data not shown). Bone formation was reduced due to a slight decrease in mineralizing surface per bone surface (p = 0.052) and a significant decrease in mineral apposition rate. Although osteoblast surface per bone surface was not changed in the mutants, osteoclast surface per bone surface was significantly increased. Therefore, the reduction in bone volume in the mutants was the result of decreased bone formation and increased bone resorption. As expected, W/W v mice had decreased serum P1NP and increased serum CTX, confirming uncoupled bone turnover (Fig. 1C). Seminal vesicle weight, an index of androgen deficiency, was lower in W/W v mice (0.122 ± 0.009 g in WT vs 0.071 ± 0.005 g in W/W v mice, p < 0.05). Serum testosterone was significantly decreased in W/W v mice (2.21 ± 0.30 ng/ml) compared with WT controls (5.02 ± 1.19 ng/ml).
Growing W sh /W sh mice are osteopenic. To eliminate the possible effect of sex hormones on the skeletal phenotype in c-Kit mutants, we analyzed the skeletal phenotype of male W sh /W sh mice. The body weight of the mutants and WT were similar (data not shown). μ CT analysis of the cortical bone indicated that c-Kit mutation resulted in a significant decrease in total cross sectional volume, cortical volume, and marrow volume at 6, but not 9 and 13 weeks of age ( Fig. 2A and Supplementary Table S2). A significant decrease in cancellous bone volume, trabecular number and connectivity density with a concomitant increase in trabecular separation was observed in 6-and 9-week-old W sh /W sh mice. Unlike the W/W v mice, seminal vesicle weight was similar in W sh /W sh mice and WT controls (data not shown). The serum testosterone levels in 6-week-old mice (6.05 ± 1.08 ng/ml in WT vs 5.84 ± 1.44 ng/ml in W sh /W sh mice, NS) confirmed that male W sh /W sh mice are not androgen deficient. W sh mutation increases bone formation and bone resorption in growing mice. Histomorphometric analysis of the tibiae confirmed a decrease in cancellous bone volume at 6 weeks of age in W sh /W sh mice (Supplementary Table S3). Although mineralizing surface was not affected, mineral apposition rate was higher in 6-and 9-week-old W sh /W sh mice, leading to increased bone formation rate ( Fig. 2B and Supplementary Table S3). Indices of bone formation; osteoblast surface per bone surface, osteoid surface per bone surface, osteoid volume per tissue volume, and osteoid thickness, were markedly increased at 6 weeks of age. A trend toward an increase in serum P1NP was also observed (Fig. 2C). However, the skeletal phenotype was milder in older mice. There was no statistical significant difference between control and mutant mice in all indices of bone formation at 13 weeks of age. In contrast, osteoclast surface per bone surface was dramatically increased compared with age-matched controls in all age groups. As shown in Fig. 2C, increased bone resorption in the mutants was confirmed by increased serum CTX levels.
We then examined the skeletal phenotype of 6 weeks old female mice. Similar to male W sh /W sh mice, female mice had increased bone turnover. As shown in Supplementary Table S4, mineral apposition rate was higher in female W sh /W sh mice compared with WT, leading to an increase in bone formation rate expressed per bone surface and bone volume. Osteoblast surface per bone surface, osteoblast number per tissue area and osteoblast number per bone perimeter were significantly increased in 6-week-old W sh /W sh mice. Osteoclast surface per bone surface, osteoclast number per tissue area and osteoclast number per bone perimeter were also increased. The magnitude of change in bone formation rate was higher in female (40-57%) compared with male mice (30-37%). Therefore, there was no net change in bone volume in female mice. Male W sh /W sh and their controls were selected for further investigation.
Osteoblast and osteoclast marker gene expression was examined in 6-weeks-old male W sh /W sh mice and their controls. As shown in Fig. 3A, qPCR indicated that c-Kit mutation increased the expression of several osteoblast marker genes in femora including osteocalcin, Osterix, ALP, type I collagen and Runx2. The mRNA levels of both RANKL and OPG were increased therefore the RANKL/OPG ratio was not significantly changed. Expression profiling of osteoclast target genes showed increased expression of M-CSF, c-Fms, NFATC1 and TRAP in 6-week-old W sh /W sh mice (Fig. 3B). These data suggest that the increased bone turnover observed in 6-week-old W sh /W sh mice is likely to be due to increased bone formation and bone resorption in vivo.

c-Kit mutation increases osteoclast differentiation and precursors.
We examined the expression of c-Kit in BMM, osteoclasts, and osteoblasts. The mRNA level of c-Kit was much lower in osteoblasts compared with BMM and osteoclasts in W sh /W sh mice (Fig. 4A). c-Kit mutation reduced the c-Kit mRNA levels in BMM and osteoclasts by 43 and 35%, respectively, whereas the c-Kit mRNA level in osteoblasts was not altered.
Mutation of c-Kit increased osteoclast number in all age groups. We examined whether the increased number of osteoclasts in W sh /W sh mice was a cell-autonomous effect. Consistent with the increased in vivo bone resorption, TRAP staining showed increased osteoclast number in cultured BMM derived from W sh /W sh mice To determine the effect of c-Kit on osteoclast resorbing ability, equal numbers of WT and mutant osteoclasts were cultured on dentin slices and the resorbed areas were quantified. No differences were observed in resorption between the mutant and WT cells (Fig. 4B).
W sh /W sh mice had increased osteoclast number in vivo and in vitro. To examine whether the increased osteoclast number was secondary to a greater number of osteoclast precursors, FACS analysis was performed on spleen and bone marrow cells. The results revealed that the percentages of CD11b + , c-Fms + and CD11b + c-Fms + cells in W sh /W sh spleen cells were all increased compared with WT cells (Fig. 4D). There was an increase in the CD11b + bone marrow cells (23.95 ± 1.25 in WT vs 31.10 ± 1.44 in W sh /W sh mice, p < 0.05), whereas c-Fms + and CD11b + c-Fms + cells were not altered in W sh /W sh mice.

Mutation of c-Kit increases osteoblast progenitors and differentiation.
To determine whether the increased bone formation observed in growing W sh /W sh mice was due to enhanced osteoblast proliferation and/ or differentiation, we analyzed the proliferation and differentiation of calvarial osteoblasts in vitro. One-day-old pups were genotyped (Fig. 5A). Osteoblasts derived from W sh /W sh mice had higher ALP activity and formed more mineralized bone nodules than those from WT controls (Fig. 5B), whereas osteoblast proliferation, determined by BrdU labeling, was unchanged (data not shown). Because c-Kit expression was similar in W sh /W sh and WT osteoblasts, we hypothesized that an increase in the osteoblast precursor population contributed to increased osteoblast differentiation in W sh /W sh mice. To test this hypothesis, we analyzed the capability of calvarial osteoblasts derived from W sh /W sh mice and their WT littermates to form ALP-positive colony forming units (CFU-ALP) and CFU-osteoblasts (CFU-OB). We found that although CFU-ALP and CFU-OB were increased, the number of CFU-fibroblasts (CFU-F) was not changed in W sh /W sh osteoblasts compared with WT controls (Fig. 5C). qPCR analysis of calvarial osteoblasts derived from W sh /W sh mice showed increased expression of the osteoblast marker genes osteocalcin, Bsp, and Dmp1 (Fig. 6A). The mRNA level for RANKL increased, whereas OPG mRNA level decreased, leading to an increase in the RANKL/OPG mRNA ratio in W sh /W sh osteoblasts. To further investigate whether the increased RANKL/OPG ratio in these cells enhanced osteoclast differentiation, co-culture of W sh /W sh or WT osteoblasts with W sh /W sh or WT BMMs was performed (Fig. 6B). Compared with WT osteoblasts, W sh /W sh osteoblasts induced greater osteoclast differentiation regardless of the BMM genotype, suggesting that the increased RANKL/OPG ratio in W sh /W sh osteoblasts led to increased osteoclast differentiation.
Osteoclast-coupling factor Wnt 10b is increased in W sh /W sh osteoclasts. We examined whether altered local regulation between osteoblasts and osteoclasts contributed to increased osteoanabolic factors and subsequent bone formation in W sh /W sh mice. WT calvarial osteoblasts were cultured in conditioned medium from either W sh /W sh or control osteoclasts. Calvarial osteoblasts treated with conditioned medium derived from W sh /W sh osteoclast cultures demonstrated increased ALP staining, indicating that the conditioned medium contained Osteoclasts generated by co-culture with osteoblasts on collagen gel in media containing vitamin D 3 and PGE 2 were replated onto dentin slices and cultured for 48 h (lower). Resorption pits were stained with toluidine blue. (C) TRAP-positive osteoclast number per total area (/mm 2 ) was quantified by OsteoMeasure software. (D) FACS analysis of CD11b + , c-Fms + and CD11b + cFms + cells from spleen derived from WT and W sh /W sh mice. Results are mean ± SEM. * p < 0.05 versus WT. more anabolic coupling factors (Fig. 7A). We then determined the expression of the known osteoclast-derived coupling factor genes, Efnb2, Sphk1, Sphk2, BMP6, Sema4D, Cthrc1 and Wnt10b 21-26 from W sh /W sh and WT osteoclasts using qPCR. As shown in Fig. 7B, c-Kit mutants increased their expression of Wnt10b but not that of Efnb2, Sphk1, Sphk2, BMP6, Sema4D, or Cthrc1. Confocal immunofluorescence analysis of osteoclasts revealed increased Wnt10b staining in W sh /W sh osteoclasts cultured on dentin slices (Fig. 7C). Western blot analysis of osteoclast-conditioned media confirmed an increase in Wnt10b protein level in mutants (Fig. 7D). Before blotting, the membranes were stained with Ponceau S for protein detection. We found that the protein levels were comparable between the W sh /W sh and WT conditioned medium. Blocking Wnt10b signaling decreased W sh /W sh osteoclast conditioned medium-induced ALP activity and mineralization. To evaluate whether osteoclast-derived Wnt10b contributed to the coupling-mediated enhancement of osteoblast formation, 0.2 μ g/ml DKK1, a Wnt antagonist, was added to the WT calvarial osteoblast cultures with conditioned medium from either W sh /W sh or control osteoclasts. DKK1 markedly inhibited the increase in ALP activity and mineralization induced by W sh /W sh osteoclast conditioned medium (Fig. 8A). To further confirm the effects of the Wnt antagonist, we used antibody to neutralize the influence of Wnt10b on ALP activity and mineralization. Neutralizing Wnt10b inhibited the W sh /W sh osteoclast conditioned medium-induced increase in ALP activity and mineralization. Therefore, Wnt10b is the major coupling factor responsible for W sh /W sh osteoclast-mediated increase in bone formation.

Discussion
Our data suggest that c-Kit plays a crucial role in bone remodeling process. In the present study, we first examined the skeletal phenotype of W/W v mice that carry a c-Kit point mutation and are sterile. W/W v mutants had decreased cortical and cancellous bone volume. The reduction in cancellous bone volume was the result of a marked decrease in osteoblast surface and increase in osteoclast surface, indicating uncoupled bone turnover. To gain further insight into the precise role of c-Kit in bone metabolism, we used W sh /W sh mice that possess an inversion mutation upstream of the c-Kit region and are fertile. This c-Kit mutation, which reduced c-Kit expression in BMMs and osteoclasts but did not affect its expression in osteoblasts, resulted in osteopenia associated with increased bone formation and increased bone resorption in growing W sh /W sh mice. The skeletal phenotype was milder when animals were mature. The increased osteoclast number was a consequence of an increased RANKL/OPG ratio in osteoblasts. It appears that the alteration in the osteoclast-osteoblast coupling mechanism contributes to increased bone formation in W sh /W sh mice. Mutation of c-Kit stimulates Wnt10b secretion from osteoclasts that promotes osteoblast mineralization and subsequently bone formation. Blocking Wnt10b markedly inhibited the increased ALP activity and mineralization that were induced by W sh /W sh osteoclast conditioned medium.
Physiologically, c-Kit expression is tightly regulated and the reduction or loss of c-Kit activity by known mutations is associated with bone loss. Osteoclasts express c-Kit receptor on their cell membrane and respond to its ligand directly through cell-to-cell contact 7 or indirectly through paracrine factors. The skeletal phenotype of W/W v mice is subtle and these mice are infertile due to germ cell depletion. It has been reported that male W/W v mice have normal plasma testosterone levels but elevated FSH levels 27 . However, our data indicated that seminal vesicle weight, an index of androgen deficiency, decreased by 42% in W/W v mice. The serum testosterone was also decreased in these mice. Male hypogonadism increases the production of osteoclasts and osteoblasts, leading to an increase in cancellous bone turnover [28][29][30] . These changes were quite distinct from those found in W/W v mice. Thus, the low bone mass observed in growing W/W v mice is unlikely to be solely the consequence of androgen deficiency. W sh /W sh mice that are fully fertile and have normal testosterone level also exhibited osteopenia. However, the cellular mechanism of bone loss in these mice was different from that of W/W v mice. Increased Calvarial osteoblasts were treated with either WT or W sh /W sh osteoclast-conditioned medium with or without DKK1. ALP activity (n = 5 per group) and mineralization (n = 5 per group) were quantified. (B) Osteoblasts were cultured with either WT or W sh /W sh osteoclast-conditioned medium pretreated with either isotype control (Ctrl) or Wnt10b antibody (Wnt10b). ALP activity (n = 5 per group) and mineralization (n = 5 per group) were quantified. Results are mean ± SEM. * p < 0.05 versus WT and + p < 0.05 versus corresponding controls.
Scientific RepoRts | 6:31515 | DOI: 10.1038/srep31515 osteoclast number and increased osteoblast function with a net increase in bone resorption contributed to the skeletal phenotype observed in W sh /W sh mice.
Bone undergoes renewal and repair through bone remodeling process, with no net change in bone volume when the amount of bone removed is precisely replaced by that of bone formed. Failure to elicit a corresponding increase in bone formation following a dramatic increase in osteoclasts causes net bone loss in growing male W sh / W sh mice. c-Kit mutation decreased the mRNA level of c-Kit in BMMs and osteoclasts leading to increased osteoclast differentiation in vitro. These results suggest a cell-autonomous effect in W sh /W sh mice. Osteoclast formation is driven by the key effector RANKL derived from osteoblasts or other cell lineages within the bone microenvironment. RANKL activity is moderated by a decoy receptor OPG. c-Kit mutation induced bone resorption by increasing RANKL expression in both in vivo and in vitro. W sh /W sh osteoblasts had an increased RANKL/OPG mRNA ratio, which has been shown to promote osteoclastogenesis 31,32 . Our co-culture experiments using osteoblasts and osteoclasts also confirmed that the increased RANKL/OPG ratio in W sh /W sh osteoblasts was responsible for the increased osteoclast differentiation observed in vitro.
According to the osteoclast commitment and differentiation pathway, CD11b is expressed during the differentiation of mononuclear early progenitor cells to mature multinucleated osteoclasts. The expression is higher in mononuclear cells and low in mature osteoclasts. It has been reported that c-Fms is a major determinant in osteoclast differentiation 33,34 . FACS analysis of spleen cells derived from W sh /W sh mice revealed an increase in the percentage of CD11b + , c-Fms + , and CD11b + c-Fms + cells. Although the number of c-Fms + , and CD11b + c-Fms + cells in W sh /W sh bone marrow was not altered, CD11b + cell number was increased. These data suggest that reduced c-Kit signaling acts to expand the pool of osteoclast precursors, leading to increased osteoclast differentiation and bone resorption in W sh /W sh mice. Although histomorphometric analysis indicated no changes in bone architecture at 9 weeks of age, three-dimensional μ CT showed decreases in cancellous bone volume, trabecular thickness and trabecular number with concomitant increase in trabecular separation. The CTX level in W sh /W sh mice was increased by 68, 44 and 41% at 6, 9 and 13 weeks old, respectively, indicating a reduction in osteoclast activity in the mutants as the animals grew.
The mechanism by which loss-of-function mutation of c-Kit led to osteopenia in W sh /W sh mice remains unclear. It has been reported that c-Kit mediates cell-to-cell interactions between osteoclasts and osteoblasts/ stromal cells through membrane bound KL 7 . Soluble KL, in concert with other factors, stimulates osteoclast formation and activity 35 . Gleevec, which inhibits c-Kit as well as c-Fms, c-Abl, and PDGF receptor 36,37 , decreases osteoclast number in rodents 38 and inhibit osteoclast differentiation in vitro 39 . When used therapeutically to treat chronic myeloid leukemia, it sometimes induces secondary hyperparathyroidism with inconsistent reports of changes in bone formation and bone resorption 40,41 . The extent to which the effects of Gleevec on bone are mediated by inhibition of c-Kit has not been determined.
W/W v and W sh /W sh mice have reduced numbers of mast cells in various soft tissues [42][43][44] . Mast cell deficiency is associated with low bone turnover, whereas excessive mast cell number induces bone loss 45 . Although the relationship between mast cells and osteoclasts is not completely understood, mast cells can modulate osteoclast activity through their release of granule-associated cytokines. However, it has been reported that mast cells are rarely found in mouse bone marrow 18,46 . Therefore, it is unlikely that mast cells contribute to the skeletal changes observed in our study.
The role of c-Kit in bone formation remains undefined. Although it has been reported that c-Kit is expressed in primary rat osteoblasts and SaOS-2 cells but not MC3T3-E1, ST2 or RAW 264.7 cells 47 , we found that c-Kit expression was much lower in mouse calvarial osteoblasts compared with osteoclasts. The W sh mutation affects the tissue-specific expression of c-Kit during embryonic development and adulthood 16 . Although c-Kit expression in W sh /W sh osteoclasts was decreased by 35%, its expression in osteoblasts was not altered, indicating that the increased bone formation in W sh /W sh mice was not due to an intrinsic effect in osteoblasts. The fact that calvarial osteoblasts isolated from W sh /W sh mice had increased ALP activity and formed more bone nodules together with an increase in the number of CFU-ALP and CFU-OB without any change in total CFU-F suggests that the increased osteoblast number in bone in vivo resulted from an increase in the number of committed osteoblast precursors. Our finding indicates that c-Kit mutation leads to increased bone formation in W sh /W sh mice through indirect osteoclast-mediated effects. The evidence that bone resorption triggers bone formation suggests that certain coupling factors derived from osteoclasts are responsible for recruiting osteoblasts progenitors to the remodeling site and stimulating bone formation. Our finding that conditioned medium derived from W sh /W sh osteoclast culture increased ALP activity and mineralization in osteoblasts confirmed an increase in osteoclast-secreted osteoanabolic factors. We demonstrated that c-Kit mutation stimulated Wnt10b, a known coupling factor, production, and secretion from osteoclasts. The increase in Wnt10b production in c-Kit mutant osteoclasts, together with the increased osteoblast differentiation induced by W sh /W sh osteoclast-conditioned medium and the increased bone formation in vivo, strengthen the evidence that osteoclasts can enhance bone formation through secreted coupling factors. Binding of Wnt10b to Wnt receptors, LRP-5 and LRP-6, on osteoblasts stimulates new bone formation 48 . Antagonizing Wnt10b blunted the anabolic effects of the osteoclast-conditioned medium in vitro. Therefore, it is likely that Wnt10b is an osteoclast-derived molecule responsible for the enhancement of bone formation in W sh /W sh mice. However, the mechanism by which c-Kit mutation regulates Wnt10b production by osteoclasts remains to be determined. Our findings do not exclude a contribution of matrix-derived growth factors, such as TGF-β 1, released from the bone matrix during bone resorption. Other investigators have shown that TGF-β 1 stimulates Wnt10b production in osteoclasts that enhances the coupling of bone resorption with formation 26 . Further studies are required to address this question.
In conclusion, this study is the first to report the importance of c-Kit as a negative regulator of bone turnover and that Wnt10b is a physiologically important osteoclast-secreted molecule that promotes bone formation in c-Kit mutants. Targeting c-Kit may provide a new insight to develop therapeutic intervention for skeletal disorders.

Materials and Methods
Animals. W sh /W sh , W/W v and WBB6F1/J-Kit +/+ wildtype (WT) mice were purchased from Jackson Laboratory (Bar Harbor, ME, USA). W sh /W sh mice were crossed to C57BL6/J (Jackson Laboratory) to produce heterozygotes. W sh / + mice were then crossed to generate W sh /W sh mice and littermate controls. W/W v and W sh /W sh mice are white, and black-eyed, whereas their controls are black. Male and female mice were fed standard mouse chow ad libitum and maintained under a 12:12 h light/dark cycle. Animals were maintained in accordance with the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. The experimental protocols were approved by the Institutional Animal Care and Use Committee at the Harvard Medical School.
Mice were subcutaneously injected with 20 mg/kg calcein (Sigma, St Louis, MO, USA) and 40 mg/kg demeclocycline (Sigma) and the interlabeling periods were 4, 5, and 6 days for 6-, 9-, and 13-week-old mice, respectively. At the end of the experiment, the mice were weighed and anesthetized with isoflurane. Blood samples were collected and centrifuged and the serum was kept at − 80 °C for determination of P1NP and CTX. The seminal vesicles, tibiae, and femora were removed. The right femora and tibiae of W/W v mice were fixed in 70% alcohol for μ CT analysis and bone histomorphometry, respectively. For W sh /W sh mice, the left tibiae were used for μ CT analysis, whereas the right tibiae were analyzed for bone histomorphometry. The left femora of W sh /W sh mice were frozen in liquid nitrogen and stored at − 80 °C until processed for RNA isolation and qPCR analysis. μCT. A three-dimensional reconstruction of bone microarchitecture was performed using a desktop μ CT35 (Scanco Medical) following the recommended guidelines 50 . Cortical bone at the tibial or femoral midshaft and cancellous bone at the proximal tibial metaphysis or femoral distal metaphysis were scanned using a 7-μ m isotropic voxel size, 50 kVp, and 144 μ A. For cortical bone, 86 transverse μ CT slices were evaluated using a threshold at 35% of the maximal gray scale value to assess the total cross-sectional volume (mm 3  Osteoblast differentiation and proliferation and CFU assays. One-day old pups were genotyped by PCR using primer pair P1 (TTTGCACGTGCTAGTTACAC) and P2 (TTAAGATGGCACCCTGCTG) for WT template and primer pair P3 (AGGCTTGCAGCGCATTAT) and P4 (GAGGATTCATAGTTGTTCAATGTCC) for W sh /W sh template 51 . Primary calvarial osteoblasts derived from W sh /W sh mice and their control littermates were plated at 1 × 10 4 cells/well in 12-well plates and cultured in differentiation medium containing α -MEM supplemented with 10% FBS, 100 unit/ml penicillin and 100 μ g/ml streptomycin, 5 mM β -glycerophosphate, 10 μ M dexamethasone, and 50 μ g/ml ascorbic acid. The cells were fixed with 3.7% formaldehyde, and stained for ALP and bone nodules with Fast Blue Alkaline phosphatase (Sigma) and 2% alizarin red (Sigma) on days 7 and 21, respectively. ALP activity and mineralized bone nodules were quantified as previously described 52 . For the DKK1 experiments, WT calvarial osteoblasts were plated at 0.5 × 10 4 cells/well in 24-well plates and cultured with osteoclast-conditioned medium. Recombinant DKK1 (R&D systems) at 0.2 μ g/ml was added to the medium and ALP activity and mineralization bone nodules were determined. For the Wnt10b neutralization experiments, osteoclast-conditioned medium was pretreated with either isotype control or Wnt10b antibody (R&D) at 4 μ g/ml before addition to the osteoblast cultures. The cells were stained with ALP and alizarin red.
Calvarial osteoblasts derived from W sh /W sh mice and controls (7 × 10 3 cells/well) were cultured in 96-well plates for 24 h and exposed to BrdU labeling reagent for 18 h. Osteoblast proliferation was determined by a BrdU incorporation assay per the manufacturer's instructions (Cell Signaling). For CFU assays, calvarial osteoblasts derived from W sh /W sh and WT littermates were plated at 0.25 × 10 4 cells/well in 12-well plates and cultured in differentiation medium. CFU-ALP was determined as ALP-positive colonies on day 7. Mineralized nodules were stained with alizarin red on day 23 for the determination of CFU-OB. The cells were then stained with toluidine blue to determine CFU-F on days 7 and 23 in the same tissue culture plate.

Preparation of osteoclasts.
Bone marrow cells were cultured in α -MEM containing 10%FBS, 100 unit/ml penicillin and 100 μ g/ml streptomycin for 24 h to generate BMMs. The BMMs were then cultured on tissue culture plastic or coverslips in α -MEM for 2 days with 20 ng/ml M-CSF and for an additional 6 days in the same medium with 20 ng/ml M-CSF and 3.3 ng/ml RANKL. Osteoclast-conditioned medium was collected. Multinucleated cells were identified by TRAP staining. For the resorption assay, osteoclasts were generated by culturing BMMs with primary CD1 calvarial osteoblasts in media containing 10 nM 1,25-dihydroxyvitamin D 3 and 1 μ M prostaglandin E 2 on a collagen gel. The collagen was digested with 0.1% collagenase and the osteoclasts were replated onto dentin slices for an additional 48 h to determine their bone-resorbing activity using toluidine blue staining.
For the co-culture studies, calvarial osteoblasts derived from WT and mutants were cultured in α -MEM containing 10% FBS, 100 unit/ml penicillin and 100 μ g/ml streptomycin for 24 h. BMMs were added and cultured in media containing 1,25-dihydroxyvitamin D 3 and prostaglandin E 2 for 5 additional days. TRAP staining was performed. qPCR. Total RNA was isolated from femora using Trizol reagent according to the manufacturer's instructions (Invitrogen) and purified using an RNeasy Mini kit (Qiagen). The RNA yields were determined spectrophotometrically at 260 nm. One μ g of total RNA was used to synthesize cDNA using SuperScript VILO (Invitrogen). The qPCR was performed at 57 °C for 40 cycles using an iCycler (Biorad) and the results were normalized to GAPDH expression. The primer sequences used are shown in Supplementary Table S5. FACS analysis. Bone marrow and spleen cells were removed and red blood cells were lysed with lysis buffer (eBioscience). The cells were incubated with Alexa Fluor ® 488-conjugated anti-mouse CD11b (eBioscience) for 30 minutes at 4 °C and washed twice with washing buffer. The cells were incubated with PE-conjugated anti-mouse c-Fms (eBioscience) for 30 minutes and washed twice. The stained cells were suspended in PBS and flow cytometry was performed using BD LSRFortessa (Becton Dickinson).

Confocal microscopy.
Osteoclasts plated on dentin slices were fixed with 3.7% formaldehyde in phosphate-buffered saline (PBS) for 10 min. Cells were permeabilized in PBS containing 0.05% saponin, 0.1% BSA and 5% normal serum for 30 min, incubated with anti-mouse/rat/human Wnt10b antibody (Santa Cruz) for 1 h, washed, incubated with fluorescent secondary antibody (Alexa Fluor 488), washed again, and mounted in FluorSave (Calbiochem). For actin labeling, the cells were incubated in a 1:40 dilution of rhodamine phalloidin stock solution (Invitrogen) for 1 h and washed with PBS. The nuclei were labeled with TO-PRO-3 (1:1000) in the rhodamine phalloidin solution. Osteoclasts were visualized using a 510 Meta laser scanning confocal microscope (Carl Zeiss) and images were recorded.

Western blot analysis.
Osteoclasts were generated in a 6-well-plate and conditioned medium was collected and concentrated 50-fold using Amicon ultra centrifuge filter units (Millipore). Protein concentration was determined using a BCA protein assay kit (Thermo Scientific). Samples (200 μ g) were resolved using 4-12% Mini-PROTEAN TGX precast gel (Biorad) and transferred to nitrocellulose membranes. The membranes were stained with Ponceau S dye for protein detection then washed in TBST buffer (50 mM Tris-HCl, pH 7.5, 150 mM NaCl, and 0.1% Tween 20) and incubated in TBST buffer containing anti-Wnt10b antibody overnight at 4 °C. The membranes were washed in TBST and incubated with anti-rabbit IgG horseradish peroxidase-conjugated secondary antibody (Promega). The membranes were then washed and developed with an enhanced chemiluminescence detection kit (PerkinElmer, Inc.). Serum analysis. Serum P1NP and CTX were determined using Rat/Mouse P1NP and RatLaps TM EIA kit, respectively per the manufacturer's protocol (Immunodiagnostic systems). Serum testosterone concentration was measured by enzyme-linked immunosorbent assay (ELISA) per the manufacturer's directions (R&D Systems).

Statistical analysis.
All data are expressed as the mean ± SE. Unpaired Student's t-test was used to compare between 2 groups. Multiple comparisons were determined using one-way ANOVA followed by Fisher's protected least significant difference test. The in vitro experiments were repeated three times. Statistical significance was considered at p < 0.05.