Axin2 controls bone remodeling through the beta-catenin-BMP signaling pathway in adult mice

Abstract

To investigate the role of Wnt-beta-catenin signaling in bone remodeling, we analyzed the bone phenotype of female Axin2-lacZ knockout (KO) mice. We found that trabecular bone mass was significantly increased in 6- and 12-month-old Axin2 KO mice and that bone formation rates were also significantly increased in 6-month-old Axin2 KO mice compared with wild-type (WT) littermates. In vitro studies were performed using bone marrow stromal (BMS) cells isolated from 6-month-old WT and Axin2 KO mice. Osteoblast proliferation and differentiation were significantly increased and osteoclast formation was significantly reduced in Axin2 KO mice. Nuclear beta-catenin protein levels were significantly increased in BMS cells derived from Axin2 KO mice. In vitro deletion of the beta-catenin gene under Axin2 KO background significantly reversed the increased alkaline phosphatase activity and the expression of osteoblast marker genes observed in Axin2 KO BMS cells. We also found that mRNA expression of Bmp2 and Bmp4 and phosphorylated Smad1/5 protein levels were significantly increased in BMS cells derived from Axin2 KO mice. The chemical compound BIO, an inhibitor of glycogen synthase kinase 3beta, was utilized for in vitro signaling studies in which upregulated Bmp2 and Bmp4 expression was measured in primary calvarial osteoblasts. Primary calvarial osteoblasts were isolated from Bmp2(fx/fx);Bmp4(fx/fx) mice and infected with adenovirus-expressing Cre recombinase. BIO induced Osx, Col1, Alp and Oc mRNA expression in WT cells and these effects were significantly inhibited in Bmp2/4-deleted osteoblasts, suggesting that BIO-induced Osx and marker gene expression were Bmp2/4-dependent. We further demonstrated that BIO-induced osteoblast marker gene expression was significantly inhibited by Osx siRNA. Taken together, our findings demonstrate that Axin2 is a key negative regulator in bone remodeling in adult mice and regulates osteoblast differentiation through the beta-catenin-BMP2/4-Osx signaling pathway in osteoblasts.

Fig. 1.

Bone mass is increased in 6- and 12-month-old female Axin2 KO mice. Three dimensional bone structure was analyzed in 6- and 12-month-old female Axin2 KO (Axin2^–/–) mice and WT controls (n=6) by micro-CT (A,J). Bone volume (BV) was increased 61 and 255% in 6- and 12-month-old Axin2 KO mice (B,K). Bone mineral density (BMD) was increased 30 and 73% in 6- and 12-month-old Axin2 KO mice when bone mineral content (BMC) was normalized to tissue volume (BMC/TV) (C,L). The structural model index (SMI) was significantly decreased (23 and 42%, respectively) in 6- and 12-month-old Axin2 KO mice (D,M). The trabecular number (Tb.N.) was increased (16 and 39%, respectively) (E,N) and trabecular thickness (Tb.Th.) was increased 15 and 21%, respectively in 6- and 12-month-old Axin2 KO mice (F,O). By contrast, trabecular separation (Tb.Sp.) was decreased (14 and 24%, respectively) in 6- and 12-month-old Axin2 KO mice (G,P). ^*P<0.05, unpaired Student's t-test, n=6. Bone marrow stromal (BMS) cells were isolated from 2- and 6-month-old WT C57BL/6J mice and expression of Axin1 and Axin2 was examined by real-time PCR. Axin1 expression was reduced 47% and Axin2 expression was increased ninefold in BMS cells derived from 6-month-old mice (H,I), suggesting that Axin2 might play a predominant role in older mice. ^*P<0.05, unpaired Student's t-test, n=3. All values are means ± s.e.

Fig. 2.

Osteoblast activity is increased and osteoclast formation was decreased in Axin2 KO mice. Histological analyses showed that trabecular bone volume was significantly increased in 6- and 12-month-old female Axin2 KO mice (A,B). Histomorphometric analyses demonstrated the 64% increase in mineral appositional rates (MAR) and 86% increase in bone formation rates (BFR) in 6-month-old Axin2 KO mice (n=9) (C,D). Histomorphometric analyses also showed that osteoclast surface and osteoclast numbers were significantly reduced (37 and 42% reduction, respectively) in 6-month-old Axin2 KO mice (n=9) (E,F). In vitro osteoclast formation assay was performed using bone marrow cells, which were cultured in the presence or absence of M-CSF (30 ng/ml) for 3 days and M-CSF (30 ng/ml) and RANKL (20 ng/ml) for an additional 4 days. TRAP staining and quantification of TRAP-positive multi-nucleus osteoclast numbers were performed. Osteoclast formation was significantly reduced in bone marrow cells derived from 6-month-old Axin2 KO mice (n=3) (G,H). The cells were also cultured on the bone slices. After cell culture, cells were removed by brushing and bone slices were stained with toludine blue. The pit areas were counted. Bone resorption was significantly reduced in Axin2 KO mice (n=3) (I). We also measured Opg, Rankl and cathepsin K expression and found that the expression of Opg but not Rankl was significantly increased and expression of cathepsin K was significantly reduced in bone marrow cells derived from 6-month-old Axin2 KO mice (n=3) (J-L). Serum OPG protein levels were measured by Elisa assay. A significant increase in OPG protein level (>threefold) was found in Axin2 KO mice (n=3) (M). ^*P<0.05, unpaired Student's t-test. All values are means ± s.e.

Fig. 3.

Osteoblast proliferation and differentiation are increased in bone marrow stromal (BMS) cells of Axin2 KO mice. BMS cells were isolated from 6-month-old WT and Axin2 KO mice. Cells were cultured for 6 days and Ki-67 staining was performed. Ki-67-positive cells were increased 5.7-fold in BMS cells derived from Axin2 KO mice (A). Consistent with this finding, cyclin D1 mRNA and protein expression were also increased in BMS cells derived from Axin2 KO mice (B,C). ^*P<0.05, unpaired Student's t-test, n=4. BMS cells were cultured for 10 days for ALP staining and activity assay. The intensity of ALP staining and ALP activity (1.8-fold increase) were significantly increased in BMS cells derived from 6-month-old Axin2 KO mice (D,E). BMS cells were also cultured for 10 days and total RNA was extracted from the cells and expression of osteoblast marker genes was measured by real-time PCR assay. The expression of Runx2, osterix (Osx), type I collagen (Col-1), osteopontin (Opn), bone sialoprotein (Bsp) (G) and osteocalcin (Oc) (H) was significantly increased in BMS cells derived from Axin2 KO mice (F-K). ^*P<0.05, unpaired Student's t-test, n=3. All values are means ± s.e.

Fig. 4.

Increased β-catenin levels are responsible for the bone phenotype observed in Axin2 KO mice. Cell lysates (cytosolic and nuclear fractions) were collected from BMS cells isolated from 6-month-old WT and Axin2 KO mice. Western blot assays were performed to examine changes in total and non-phosphorylated active form of β-catenin protein. Cytoplasm β-catenin levels (active and total) were slightly increased (A). By contrast, the nuclear β-catenin levels (active and total) were significantly increased in BMS cells derived from 6-month-old Axin2 KO mice (B). Axin2 KO mice were bred with β-catenin^fx/fx mice to produce Axin2^–/–;β-catenin^fx/fx mice. BMS cells were isolated from 6-month-old WT, Axin2^–/–, β-catenin^fx/fx and Axin2^–/–;β-catenin^fx/fx mice and were infected with lentivirus expressing Cre recombinase (Lenti-Cre). ALP activity and expression of osteocalcin (Oc) and Bmp2 and Bmp4 were examined. Deletion of the β-catenin gene by Lenti-Cre completely inhibited increased ALP activity (C) and Oc and Bmp4 expression (D,E) and significantly inhibited Bmp2 expression (F) observed in Axin2 KO mice. Deletion of the β-catenin gene was determined by measuring levels of β-catenin protein using western blot analysis (G). ^*P<0.05, unpaired Student's t-test, n=3. All values are means ± s.e.

Fig. 5.

BMP signaling is activated in Axin2 KO mice. The mRNA expression of Bmp2 and Bmp4 was examined by real-time PCR using BMS cells derived from WT and Axin2 KO mice. The expression of Bmp2 and Bmp4 was significantly increased in BMS cells derived from 6-month-old Axin2 KO mice (A,B). Western blot assay was performed to analyze changes in phosphorylated Smad1/5 protein levels. The phosphorylated Smad1/5 was significantly increased in BMS cells derived from 6-month-old Axin2 KO mice (C). In nodule formation assay, the responsiveness of the BMS cells to BMP-2 was higher in Axin2 KO cells than WT cells (13-fold increase in Axin2 KO cells versus sixfold increase in WT cells) (D,E). ^*P<0.05, unpaired Student's t-test, n=3. All values are means ± s.e.

Fig. 6.

BIO activates β-catenin signaling stimulates osteoblast differentiation and bone formation. Primary calvarial osteoblasts were isolated 3-day-old WT neonatal mice and cultured with or without BIO (1 μM) for different periods of time as indicated. The β-catenin protein levels and nuclear translocation were examined by western blot and immunostaining. BIO increased protein levels of active form of β-catenin and reached its maximal effect at 4 hours (A). BIO also induced β-catenin nuclear translocation (B). Primary calvarial osteoblasts, isolated from TOPGal transgenic mice, were treated with BIO (1 μM) for 24 hours and β-Gal activity was measured using cell lysates isolated from these cells. Wnt3a (100 ng/ml) was used as a positive control in this experiment. BIO significantly increased the β-Gal activity (C). ^*P<0.05, unpaired Student's t-test, n=3. Primary calvarial osteoblasts were cultured with 0.1 and 1 μM of BIO for 2 days and changes in ALP activity were examined by ALP staining. BIO (at both concentrations) significantly increased ALP activity in these cells (D). To further determine whether BIO induces new bone formation in vivo, 25 and 50 μg/mouse of BIO was injected into 1-month-old WT mice subcutaneously over the surface of calvariae for 5 days. Calcein labeling was performed at day 5 and 15. Mice were sacrificed 2 days after the second calcein labeling and periosteal new bone was evaluated and mineral appositional rates (MAR) were measured. FGF-1 (2 μg/mouse, 5 day injection) was used as a positive control. BIO significantly increased new bone formation and MAR in this assay (E,F). ^*P<0.05, one-way ANOVA followed by Dunnett's test (BIO versus control) and unpaired Student's t-test (FGF-1 versus control), n=5. All values are means ± s.e.

Fig. 7.

BIO induces osteoblast differentiation in a Bmp2/4-Osx-dependent manner. Primary calvarial osteoblasts were isolated 3-day-old WT neonatal mice and cultured with or without BIO (1 μM) for 24 hours. BIO significantly upregulated Bmp2 (2.9-fold) and Bmp4 (1.8-fold) expression in these cells (A,B). To determine whether BIO-induced osteoblast differentiation is Bmp2/4-dependent, primary calvarial osteoblasts were isolated from Bmp2^fx/fx;Bmp4^fx/fx mice and infected with Ad-GFP or Ad-Cre. Over 80% infection efficiency and Bmp2/Bmp4 gene deletion efficiency were observed (C,D). BIO induced Runx2 expression in Bmp2/4-deleted cells as well as in control cells (∼1.5-fold increase), suggesting that BIO-induced Runx2 expression is not Bmp2/4-dependent (E). By contrast, BIO induced Osx, Col-1, Alp and Oc expression in control cells and these effects were significantly or completely inhibited in Bmp2/4-deleted osteoblasts, demonstrating that BIO stimulates Osx and subsequent osteoblast marker genes in a Bmp2/4-dependent manner (F-I). To further determine whether BIO-induced osteoblast marker genes in an Osx-dependent manner, Osx siRNA and control scramble siRNA were transfected into these cells. Transfection of Osx siRNA efficiently knocked down Osx expression (J) and significantly inhibited BIO-induced Alp expression (K) and completely inhibited BIO-induced Col-1 and Oc expression (L,M), suggesting that BIO-induced osteoblast marker gene expression is also Osx-dependent. ^*P<0.05, unpaired Student's t-test, n=3. All values are means ± s.e.