Smad1 plays an essential role in bone development and postnatal bone formation

Abstract

Objectives: To determine the role of Smad1 in bone development and postnatal bone formation.

Methods: Col2a1-Cre transgenic mice were bred with Smad1(fx/fx) mice to produce chondrocyte-specific Smad1 conditional knockout (cKO) mice. Embryonic skeletal preparation and staining were performed, alkaline phosphatase activity (ALP) and relative gene expression were examined in isolated primary cells. Smad1(fx/fx) mice were also bred with Col1a1-Cre transgenic mice to produce osteoblast-specific Smad1 cKO mice. Postnatal bone formation was assessed by micro-computed tomography (μCT) and histological analyses in 2-month-old mice. Mineralized bone nodule formation assay, 5-bromo-2'-deoxy-uridine (BrdU) labeling and gene expression analysis were performed.

Results: Mice with chondrocyte- and osteoblast-specific deletion of the Smad1 gene are viable and fertile. Calvarial bone development was delayed in chondrocyte-specific Smad1 cKO mice. In osteoblast-specific Smad1 cKO mice, BMP signaling was partially inhibited and mice developed an osteopenic phenotype. Osteoblast proliferation and differentiation were impaired in osteoblast-specific Smad1 cKO mice.

Conclusions: Smad1 plays an essential role in bone development and postnatal bone formation.

Fig. 1

Generation and analysis of Smad1^Col2a1 mice. (A) Heterozygous Smad1 floxed mice (Smad1^fx/wt) were bred with each other (Table I) and the production of homozygous Smad1 floxed mice (Smad1^fx/fx) were identified by PCR using primers SIF400/LoxP2 (lower panel, Lane 1, 410-bp band). Smad1^fx/fx mice were then bred with Co21a1-Cre mice (Table I, step a and b) and Smad1^Col2a1 mice were identified by PCR using Cre-5′/Cre-3′ primers (600-bp PCR product) and SIF400/LoxP2 primers (410-bp PCR product). (B–E) Total RNAs were extracted from calvarial tissues of E16.5 Cre-negative and Smad1^Col2a1 embryos. Expression of Col2a1 mRNA was detected in both Cre-negative and Smad1^Col2a1 embryos and expression of Cre mRNA was only detected in Cre-positive Smad1^Col2a1 embryos. Expression of Smad1 mRNA was significantly reduced in E16.5 Smad1^Col2a1 embryos. (F) Chondrocyte-specific Smad1^Col2a1 embryos and littermate control embryos are indistinguishable in size at embryonic day 16.5 (E16.5). Alizarin red/Alcian blue staining was performed on E16.5 embryos. E16.5 Smad1^Col2a1 embryos had delayed mineralization and ectopic cartilage formation of frontal [F] and parietal [P] bones compared to the same aged control embryos. (G) Calvarial Alizarin red/Alcian blue staining showed delayed mineralization of supraoccipital [S] bone in E16.5 Smad1^Col2a1 embryo. (H) The formation of ossification center of proximal phalangeal bone was also delayed in E16.5 Smad1^Col2a1 embryos. (I and J) μ-CT analysis showed reduced calvarial BV and mineralization in E16.5 Smad1^Col2a1 embryos. Unpaired Student’s t-test, P = 0.045 (n = 3). (K) E18.5 Smad1^Col2a1 embryos and littermate control embryos are indistinguishable in size. No delayed mineralization and ectopic cartilage formation were observed in E18.5 Smad1^Col2a1 embryos.

Fig. 1

Generation and analysis of Smad1^Col2a1 mice. (A) Heterozygous Smad1 floxed mice (Smad1^fx/wt) were bred with each other (Table I) and the production of homozygous Smad1 floxed mice (Smad1^fx/fx) were identified by PCR using primers SIF400/LoxP2 (lower panel, Lane 1, 410-bp band). Smad1^fx/fx mice were then bred with Co21a1-Cre mice (Table I, step a and b) and Smad1^Col2a1 mice were identified by PCR using Cre-5′/Cre-3′ primers (600-bp PCR product) and SIF400/LoxP2 primers (410-bp PCR product). (B–E) Total RNAs were extracted from calvarial tissues of E16.5 Cre-negative and Smad1^Col2a1 embryos. Expression of Col2a1 mRNA was detected in both Cre-negative and Smad1^Col2a1 embryos and expression of Cre mRNA was only detected in Cre-positive Smad1^Col2a1 embryos. Expression of Smad1 mRNA was significantly reduced in E16.5 Smad1^Col2a1 embryos. (F) Chondrocyte-specific Smad1^Col2a1 embryos and littermate control embryos are indistinguishable in size at embryonic day 16.5 (E16.5). Alizarin red/Alcian blue staining was performed on E16.5 embryos. E16.5 Smad1^Col2a1 embryos had delayed mineralization and ectopic cartilage formation of frontal [F] and parietal [P] bones compared to the same aged control embryos. (G) Calvarial Alizarin red/Alcian blue staining showed delayed mineralization of supraoccipital [S] bone in E16.5 Smad1^Col2a1 embryo. (H) The formation of ossification center of proximal phalangeal bone was also delayed in E16.5 Smad1^Col2a1 embryos. (I and J) μ-CT analysis showed reduced calvarial BV and mineralization in E16.5 Smad1^Col2a1 embryos. Unpaired Student’s t-test, P = 0.045 (n = 3). (K) E18.5 Smad1^Col2a1 embryos and littermate control embryos are indistinguishable in size. No delayed mineralization and ectopic cartilage formation were observed in E18.5 Smad1^Col2a1 embryos.

Fig. 2

Reduction in marker gene expression in sternal chondrocytes and calvarial cells derived from Smad1^Col2a1 mice. (A) Primary sternal chondrocytes were isolated from 3-day-old Cre-negative and Smad1^Col2a1 mice and total RNAs were extracted from these cells. Expression of genes related to chondrocyte differentiation and was measured by real-time PCR. Significant reduction in expression of these genes was observed in Smad1-deficient cells. Unpaired Student’s t-test, Col2a1, P = 0.033; Col10a1, P = 0.005; Runx2, P = 0.012; ALP, P = 0.048; Osterix, P = 0.004; OC, P = 0.080; Mmp13, P = 0.022 (Cre-negative, n = 6; Smad1^Col2a1, n = 4). (B) Primary calvarial cells were isolated from 3-day-old Cre-negative and Smad1^Col2a1 mice and total RNAs were extracted from these cells. Expression of genes related to osteoblast differentiation and mineralization was measured by real-time PCR. Unpaired Student’s t-test, Col2a1, P = 0.005; Runx2, P = 0.001; Alp, P = 0.452; Osterix, P < 0.001; OC, P = 0.002; Mmp13, P = 0.032 (Cre-negative, n = 6; Smad1^Col2a1, n = 4). (C) Primary calvarial cells were isolated from 3-day-old Smad1^Col2a1 and their littermate control mice. Cells were treated with vehicle or BMP-2 (100 ng/ml) for 24 h, followed by ALP staining. Smad1-deficient cells had decreased basal ALP activity. BMP-2 stimulated ALP activity in control cells but had no effect on Smad1-deficient cells.

Fig. 3

Generation and analysis of Smad1^Col1a1 mice. (A and B) Smad1^del/wt mice were bred with Col1a1-Cre mice and Col1a1-Cre;Smad1^del/wt mice were identified by PCR using Cre-5′/Cre-3′ primers (lower panel, Lane 1, 4, 5, 600-bp PCR product) and Rec-Smad1/LoxP2 primers (300-bp PCR product). Heterozygous Smad1 null allele mediated by CMV-Cre was detected by Rec-Smad1/LoxP2 primers. To generate osteoblast-specific Smad1^Col1a1 mice, the Col1a1-Cre;Smad1^del/wt mice were bred with Smad1^fx/fx mice. The desired progeny, Col1a1-Cre;Smad1^del/fx mice (Smad1^Col1a1 mice), were identified by PCR using SIF400/LoxP2 primers (tail tissue: 410-bp PCR product), Cre-5′/Cre-3′ primers (600-bp PCR product) and Rec-Smad1/LoxP2 primers (300-bp PCR product). (C and D) Expression of Smad1 mRNA and protein was measured by real-time PCR and Western blotting in bone marrow cells. Expression of Smad1 mRNA was reduced about 80%. Unpaired Student’s t-test, P < 0.001 (n = 3). Expression of Smad1 protein was detected by Western blotting using an anti-Smad1 monoclonal antibody in bone marrow cells derived from control mice but significantly reduced in those derived from Smad1^Col1a1 mice. (E and F) BV was analyzed in a defined area 0.5–2.5 mm from the growth plate in proximal tibiae of 2-month-old Smad1^Col1a1 mice and littermate control mice. The BV was normalized to TV. A 30% decrease in BV was observed in Smad1^Col1a1 mice compared with littermate controls. Unpaired Student’s t-test, P = 0.006 (n = 6). (G and H) The animals were labeled with calcein/calcein. In Smad1^Col1a1 mice, the distance between the two labels was significantly reduced as indicated by white arrows. The BFR/BS were measured in the area of the marrow cavity 0.5–2.5 mm from the growth plates of proximal tibiae and expressed as mm³/mm²/day (F). Unpaired Student’s t-test, P = 0.004 (n = 6). (I) Osteoblast numbers in trabecular bone in the same area were quantitated and a 19% decrease in osteoblast numbers was found in Smad1^Col1a1 mice. Unpaired Student’s t-test, P < 0.001 (n = 6).

Fig. 4

μ-CT analysis of Smad1^Col1a1 mice. μ-CT images showed a significant reduction in bone mass in Smad1^Col1a1 mice. (A) Trabecular structural parameters in the secondary spongiosa of the proximal tibia were measured using μCT analysis. Samples were examined with an isotropic resolution of 5-μm voxel size. Tb.N (B), Tb.Th (C) and trabecular connectivity density (Conn D) (E) were significantly decreased in Smad1^Col1a1 mice compared with littermate controls. Trabecular separation (Tb.Sp) (D) and SMI (F) were significantly increased in Smad1^Col1a1 mice. Unpaired Student’s t-test; Tb.N., P = 0.047; Tb.Th., P = 0.001; Tb.Sp., P = 0.007; Conn D, P = 0.012; SMI, P < 0.001 (n = 6).

Fig. 5

Inhibition of osteoblast proliferation and differentiation in Smad1^Col1a1 mice. (A and B) Paraffin-embedded sections of calvariae were stained with BrdU. The BrdU-positive osteoblasts were indicated by red arrows and a 22% reduction of BrdU-positive osteoblasts was found in Smad1^Col1a1 mice. Unpaired Student’s t-test, P = 0.002 (n = 4). (C) Primary osteoblasts isolated from Smad1^Col1a1 mice and littermate control mice were cultured for 10 days in the absence or presence of BMP-2 (100 ng/ml). von Kossa staining was performed after the cell culture was ended and the formation of mineralized bone nodules was analyzed. In osteoblasts derived from Smad1^Col1a1 mice, the formation of mineralized bone nodules was inhibited. (D) Primary osteoblasts isolated from Smad1^Col1a1 mice and littermate control mice were transfected with BMP signaling reporter construct 12× SBE-OC-Luc and treated with 4, 20 and 100 ng/ml of BMP-2. The luciferase activity was measured and normalized to β-gal activity. In osteoblasts derived from Smad1 cKO mice, BMP signaling was significantly inhibited compared with cells isolated from littermate control mice. One-way ANOVA followed by Dunnett’s test, BMP-2 (4 ng/ml), P = 0.003; BMP-2 (20 ng/ml), P = 0.014; BMP-2 (100 ng/ml), P < 0.001 (n = 3) between control and Smad1^Col1a1 cells.

Fig. 6

Reduction in expression of osteoblast marker genes in Smad1^Col1a1 mice. (A–D) mRNA expression of osteoblast marker genes was examined by real-time PCR. The expression of Runx2, Osterix (Osx), BSP and osteocalcin (D) genes was significantly decreased in osteoblasts derived from Smad1^Col1a1 mice. Unpaired Student’s t-test; Runx2, P = 0.002; Osterix, P = 0.022; BSP, P = 0.050; OC, P = 0.002 (n = 6). (E) Primary osteoblasts were starved for 12 h and then treated with BMP-2 (100 ng/ml) for 6 h. Smad1/5/8 phosphorylation was measured by Western blotting using an anti-p-Smad1/5/8 antibody. BMP-2-induced Smad1/5/8 phosphorylation was reduced in Smad1-deficient cells. (F) Primary osteoblasts were treated with BIO (1 μM) for 48 h and ALP activity in cell lysates was measured. Significant inhibition of ALP activity was found in Smad1-deficient cells. One-way ANOVA followed by Dunnett’s test, basal ALP, P = 0.127; BIO, P < 0.001 (n = 3), between control and Smad1-deficient cells.