Mice deficient for p53 showed increased bone mass and increased bone formation rates. (A) Normal growth plates in p53^−/− mice. Femurs were fixed, decalcified, and stained with hematoxylin-eosin. (B) Bone mineral density (BMD) of the mutant and wild-type mice of the same age and gender. Femur bones were taken from 3–4-mo-old p53 knockout and control mice and their bone densities measured (n = 8, P = 0.004). (C) A representative von Kossa staining of the femurs of mutant and control mice showing an overall improved trabecular bone in p53^−/− mice. (D) The number of trabecular bones was increased in p53^−/− mice (n = 8, P = 0.014). (E) p53^−/− mice had increased trabecular bones (trabecular bone vol/tissue vol; n = 8, P = 0.0037). (F) Calcein labeling of the cortical bones in mutant and control mice. 3–4-mo-old mice were injected with calcein twice, with an interval of 9 d. The mice were killed and femur bones sectioned without decalcification. Pictures were taken under a fluorescence microscope (40×). (G) Interlabel width (μm; n = 8, P = 0.025). (H) Mineral apposition rates (n = 8, P = 0.025). (I) Bone formation per bone perimeter. (n = 8, P = 0.0004) (J) Mineral apposition rates in trabecular bones (n = 8, P = 0.03). (K) Trabecular bone formation rates per bone perimeter (n = 8, P = 0.032). (L) Bone formation rate per bone area in trabecular bones (n = 8, P = 0.032). (M) p53^−/− mice showed increased osteoblast surface (osteoblast surface/bone surface; n = 5, P = 0.012). (N) p53^−/− mice showed increased osteoblast numbers (osteoblast numbers/bone surface; n = 5, P = 0.034). Bar, 500 μm. Error bars represent ± SD. Asterisks mark samples significantly different from control with P < 0.05.

Calvarial osteoblasts showed enhanced differentiation. (A) p53 deficiency did not significantly alter the differentiation potential of mesenchymal cells into osteoblasts. Colonies containing >25 cells were counted (per 2.5 × 10⁶ bone marrow cells). (B) Calvarial osteoblasts were isolated from newborn mice and cultured to passage 3. The same number of cells was then plated into 12-well plates at 2.5 × 10⁵ per well. (C) Quantitation of ALP activities of p53^−/− and control wild-type cells that were normalized to protein levels. (D) Quantitation of ALP activities of p53^−/− and control wild-type cells that were normalized to the number of cells. The ALP activity for control cells at day 1 was set at 1. (E) Northern blot analysis shows that expression of osteocalcin during differentiation of p53^−/− and control calvarial osteoblasts. The level of osteocalcin in control osteoblasts at day 0 was set at 1. (F) Osteoblasts were cultured as in B, except that the cells were stained using the von Kossa method for mineral deposition. (G and H) Reconstitution of p53 led to normal differentiation of p53^−/− osteoblasts, justified by ALP staining (G) and quantitative ALP (H). (G, bottom) Expression of p53 in mutant and control cells. (I) p53^−/−osteoblasts showed reduced doubling time. 4 × 10⁵ cells (triplicates) were plated in 60-mm plates, and the number of cells were counted every day for 5 d and plotted. Doubling time was calculated from three p53^−/− osteoblasts and their control littermates. (J) Expression of proteins involved in cell proliferation. The experiments were carried out as described in I, and one plate was harvested each day for Western blot analysis. (K) Quantitation data (day 3) for the data presented in J. (L) Retroviral expression of p21 in p53^−/− osteoblasts did not significantly repress differentiation. p53^−/− and control osteoblasts were infected with p21-expressing retrovirus (or control retrovirus), selected for 5 d, and plated for ALP assays. Error bars represent ± SD. Asterisks mark samples significantly different from control with P < 0.05.

Osteoblasts deficient for p53 showed up-regulated expression of osterix, but not Runx2. (A) Mutant and control osteoblasts were collected at different stages of differentiation. Total RNA was isolated and used for RT-PCR. (B) Expression of p53 changed during osteoblast differentiation. Cells at different stages of differentiation were harvested and the levels of p53 were determined by Western blot analysis. (C) Quantitation data from A and B (similar results were obtained from four pairs of p53^−/− and control osteoblasts). The level of osterix, Runx2, or p53 in control osteoblast was set at 1. (D and E) Elevated expression of osterix mRNA in calvaria of p53^−/− mice. Total RNA was made from calvaria of p53^−/− and control mice and used for RT-PCR for osterix, Runx2, and Dlx5 (D), or real-time PCR for osterix (E) to assess the mRNA levels of osterix. The level of osterix mRNA in control calvaria was set at 1. The internal control for real-time PCR was GADPH mRNA. (F) In situ staining of osterix mRNA. Femurs of p53^−/− and control mice were OCT embedded and cryosectioned, fixed, and then hybridized with radio-labeled osterix antisense RNA probe. Osterix is mainly expressed at sites of mature osteoblasts beneath the growth plates. Bar, 500 μm. Error bars represent ± SD. Asterisks mark samples significantly different from control with P < 0.05.

p53 inhibited the promoter activity of osterix. (A) Induction of osterix mRNA by BMP2 in p53^−/− and control MEFs. MEFs were stimulated with 100 ng/ml BMP2 for different periods of time. Total RNA was isolated, and RT-PCR was performed. Similar results were obtained from three pairs of mutant and control MEFs. The level of osterix in control cells at time 0 was set at 1. (B) DNA damage inhibited the expression of osterix. p53^−/− and control osteoblasts were treated with 5 or 10 μM of cisplatin for 24 h, and total RNA was isolated for RT-PCR assays. (bottom) The values of the Y axis were normalized to the level of actin, with the basal level of control cells set at 1. (top) Western blot analysis of p53 induction by cisplatin in control osteoblasts. *, P < 0.05, compared with the same genotype without CDDP treatment. (C) Coexpression of p53 inhibited the basal level of the promoter activity in osterix promoter-Luc. (D) Effects of p53 mutants on osterix promoter activity. Each of these p53 mutant constructs was cotransfected into C2C12 cells with the reporter gene, and luciferase activity was measured. (E) Serial deletion analysis of the osterix promoter. Different constructs were cotransfected into C2C12 cells with 1× p53 expression construct and the luciferase activities were measured as in C. The repression was compared with the 6-kb promoter (+++). (F) p53 repressed Osx promoter activity in the presence of p300. The experiments were performed as in D, except that TBP or p300 was included. 0.3 μg DNA was used for each expression construct. Error bars represent ± SD.

Elevated expression of osterix was necessary and sufficient for accelerated differentiation of p53^−/− osteoblasts. (A) Ectopic expression of osterix led to increased expression of the osteoblast marker ALP. Wild-type primary osteoblasts were infected with increasing amounts of retrovirus expressing osterix, selected for 3 d, and stained for ALP. (B) Western blot analysis showed the levels of osterix in cells infected with the retrovirus (top). RT-PCR assays showed the increase of osteoblast marker osteocalcin in response to ectopic expression of osterix (bottom). (C) ALP activities normalized to the level of proteins. Control cells infected with vector were set at 1. (D) Quantitation data for the protein levels of osterix in B. Control cells infected with vector were set at 1. (E) Quantitation data for the mRNA levels of osteocalcin in B. Control cells infected with vector were set at 1. (F) Knocking down of osterix (a1 to a5) led to a reduction in ALP expression in p53^−/− osteoblasts. p53^−/− and control osteoblasts were transfected with siRNA against osterix and control oligos for 2 d and the plates were stained for ALP. Different combinations of four commercially designed siRNA species were used to obtain different degrees of knocking down. a1, 1+2 siRNA duplexes; a2, 3+4 duplexes; a3, 1–4 duplexes; a4, 1–4 duplexes (2×); a5, 1–4 duplexes (4×). (G) Western blot analysis showed the levels of osterix in cells transfected with siRNAs (top). RT-PCR assays showed the decrease of osteoblast marker osteocalcin in response to knocking down of osterix (bottom). (H) ALP activities normalized to the level of proteins. Control cells transfected with control siRNAs was set at 1. (I) Quantitation data for the protein levels of osterix in H. Control cells transfected with control siRNAs were set at 1. (J) Quantitation data for the mRNA levels of osteocalcin in H. Control cells transfected with control siRNAs were set at 1. Error bars represent ± SD.

Effects of p53 deficiency on osteoclast differentiation and resorption. (A) p53^−/− mice showed increased osteoclast surface (osteoclast surface/bone surface). n = 8, P = 0.013. (B) p53^−/− mice showed an increase in the number of osteoclasts per bone surface. n = 8, P = 0.002. (C) Urinary deoxypyridinoline (DPD) cross-links excretion was significantly increased in p53-deficient mice compared with control mice. (D) Normal osteoclast differentiation from BMMs in p53^−/− mice in the presence of M-CSF and RANKL. (E) Quantitation data from five p53^−/− and control littermate mice. (F) Normal bone resorption rates of p53^−/− osteoclasts. Pit formation assays of p53^−/− and control osteoblasts (two representative views of different magnifications; left, stained with Toluidine blue; right, stained with Gill's Hematoxylin Solution). (G) Quantitation of the resorption area by p53^−/− and control osteoblasts (n = 4). (H) p53^−/− osteoblasts showed a much greater stimulatory effect on osteoclastogenesis. p53^−/− and control osteoblasts were first plated, and then the BMMs isolated from p53^−/− and control mice were plated. They were cocultured for 6 d in the presence of 10⁻⁸ M of vitamin D3 and stained for TRAP-positive osteoclasts. (I) Quantitation data from H (n = 4). (J) Number of TRAP-positive cells per well with ≥3 nuclei. *, P < 0.05, comparison between osteoclastogenesis (+/+ or −/− monocytes) on +/+ and −/− osteoblasts. (K) Up-regulation of M-CSF in p53^−/− osteoblasts. PT-PCR analysis of the mRNA levels of M-CSF in p53^−/− and control osteoblasts. The levels of M-CSF mRNA for control cells were set at 1. (L) Ectopic expression of osterix increased the expression of M-CSF. RT-PCR was performed for the sample used in Fig. 5 A. Error bars represent ± SD.

Osteoblasts deficient for both p53 and c-Abl showed advanced differentiation. Compound knockout and control osteoblasts were isolated from two litters, plated, and stained for ALP (A), quantitative ALP (B), von Kossa (C), and for assessment of osterix mRNA levels (D). The level of osterix of control osteoblasts at day 0 was set at 1. Error bars represent ± SD. Asterisks mark samples significantly different from control with P < 0.05.