Loss of cyclin-dependent kinase 1 impairs bone formation, but does not affect the bone-anabolic effects of parathyroid hormone

Bone mass is maintained by a balance between osteoblast-mediated bone formation and osteoclast-mediated bone resorption. Although recent genetic studies have uncovered various mechanisms that regulate osteoblast differentiation, the molecular basis of osteoblast proliferation remains unclear. Here, using an osteoblast-specific loss-of-function mouse model, we demonstrate that cyclin-dependent kinase 1 (Cdk1) regulates osteoblast proliferation and differentiation. Quantitative RT-PCR analyses revealed that Cdk1 is highly expressed in bone and is down-regulated upon osteoblast differentiation. We also noted that Cdk1 is dispensable for the bone-anabolic effects of parathyroid hormone (PTH). Cdk1 deletion in osteoblasts led to osteoporosis in adult mice due to low bone formation, but did not affect osteoclast formation in vivo. Cdk1 overexpression in osteoblasts promoted proliferation, and conversely, Cdk1 knockdown inhibited osteoblast proliferation and promoted differentiation. Of note, we provide direct evidence that PTH's bone-anabolic effects occur without enhancing osteoblast proliferation in vivo. Furthermore, we found that Cdk1 expression in osteoblasts is essential for bone fracture repair. These findings may help reduce the risk of nonunion after bone fracture and identify patients at higher risk for nonresponse to PTH treatment. Collectively, our results indicate that Cdk1 is essential for osteoblast proliferation and that it functions as a molecular switch that shifts osteoblast proliferation to maturation. We therefore conclude that Cdk1 plays an important role in bone formation.

Bone mass is maintained by a balance between bone formation by osteoblasts and bone resorption by osteoclasts (1). Disruption of this balance in bone turnover leads to bone-related diseases such as osteoporosis. As the global population ages, the proportion of patients with osteoporosis has increased dramatically, a trend that is associated with enormous health care costs.
Treatment strategies for osteoporosis include inhibiting bone resorption (2), promoting bone formation (3), or both (4). Although many effective antiresorptive drugs are available, including bisphosphonates, estrogen, selective estrogen receptor modulators, and anti-receptor activator of NF-B ligand antibodies, the only Food and Drug Administration-approved bone-anabolic drugs are parathyroid hormone (PTH) 2 and PTH-related protein (5).
Growing clinical evidence suggests that intermittent PTH administration increases the number of osteoblasts, resulting in stimulation of new bone formation (6). However, the mechanisms by which PTH increases osteoblast numbers are not completely understood. For example, although PTH exerts an anabolic effect by reducing osteoblast apoptosis in mice (7), PTH-stimulated bone formation is associated with an increase in osteoblast apoptosis in humans (8). Moreover, the effects of PTH on osteoblast proliferation remain controversial. One study has shown that PTH inhibits osteoblast proliferation and increases differentiation (9), whereas others have found that PTH stimulates osteoblast proliferation (10,11). Furthermore, although recent genetic studies have uncovered various mechanisms regulating osteoblast differentiation, including transcription factors, growth hormones, and noncoding RNAs (1,(12)(13)(14), the mechanisms regulating osteoblast proliferation are largely unknown. To develop a strategy to promote bone formation, investigations of the regulatory mechanisms of osteoblast proliferation are overdue.
Generally, cell proliferation is promoted by the activity of cell cycle regulators (15). The cell cycle is regulated by cyclins and cyclin-dependent kinases (Cdks) (16). Cdks are a group of kinases consisting of 20 members in mammals, which play critical roles in cell cycle control, transcription, and development (15). However, despite the theoretical importance of Cdks, the physiological roles of Cdks in bone metabolism have not been defined. Moreover, no Cdks have thus far been found to regulate bone mass through in vivo cell-specific loss-of-function experiments.
In this study, we generated osteoblast-specific Cdk1-knockout mice to investigate the role of cyclin-dependent kinase 1 (Cdk1) in regulating bone formation.

Cdk1 plays an important role in osteoblast proliferation
To investigate the functions of Cdks in bone, we first examined the expression of Cdks in the femurs of 3-month-old mice using quantitative PCR (qPCR) analysis. Cdk1 was expressed at higher levels than other Cdks in the bone (Fig. 1A). Cdk1, the first Cdk to be identified, is conserved in all organisms and has been the most extensively studied cell cycle regulator to date (17). We also examined the expression of Cdk1 in various tissues. Although Cdk1 expression was high in testis tissue, as reported previously (18), Cdk1 expression was also higher in bone than in kidney, heart, lung, brain, and liver (Fig. 1B). To Figure 1. Cdk1 is essential for osteoblast proliferation. A, expression of Cdks in mouse bone was examined by qPCR. Cdk1 expression is significantly higher than other Cdks. *, p Ͻ 0.05 versus Cdk1 expression. B, Cdk1 expression in murine tissues. Note that Cdk1 is expressed well in bone tissue. C, changes in Cdk1 mRNA and protein expression during osteoblast differentiation, as determined via qPCR (upper figure) and Western blotting (lower figure), respectively. Cdk1 expression gradually decreased over the course of osteoblast differentiation. *, p Ͻ 0.05 versus day 0. D, the relative number of MC3T3-E1 cells treated with vehicle or RO-3306 was counted for 5 days. Note that RO-3306 -treated MC3T3-E1 cells showed impaired proliferation compared with proliferation of vehicletreated cells. *, p Ͻ 0.05 versus controls. E, relative number of MC3T3-E1 cells infected with pcDNA (control) or pcDNA Cdk1 (Cdk1 overexpression) was counted at days 1, 3, and 5. Cdk1 overexpression significantly increased MC3T3-E1 cell numbers. *, p Ͻ 0.05 versus controls. All data represent the mean Ϯ standard deviation (S.D.). n.s., not significant. ALP, alkaline phosphatase.

Loss of cyclin-dependent kinase 1 impairs bone formation
confirm Cdk1 expression in vitro, we used MC3T3-E1 cells, a murine preosteoblast cell line (12). Interestingly, Cdk1 expression gradually decreased during osteoblast differentiation (Fig.  1C). To study the importance of Cdk1 kinase activity, we tested whether inhibiting Cdk1 would affect osteoblast proliferation. To this end, we treated MC3T3-E1 cells with RO-3306, a specific inhibitor of Cdk1 (19). As expected, compared with the proliferation of vehicle-treated cells, the proliferation of MC3T3-E1 cells was significantly impaired by RO-3306 treatment (Fig. 1D), which agrees with the results in other types of cells (17,19,20). Conversely, overexpressing Cdk1 significantly promoted MC3T3-E1 cell proliferation (Fig. 1E). Taken together, these results indicate that Cdk1 plays an important role in osteoblast proliferation in vitro.

Osteoblast-specific Cdk1-knockout mice have low bone mass
Because Cdk1-null mice exhibit early embryonic lethality (17,20), we generated conditional osteoblast-specific Cdk1knockout mice to investigate the role of Cdk1 in bone metabolism. To achieve this, we crossed Cdk1 f/f (hereafter, "control") mice with transgenic mice expressing Cre recombinase under the control of the Osterix promoter (hereafter, "Osx-Cre mice") to generate Osx-Cre tg /Cdk1 f/f mice (hereafter, "Cdk1 osb Ϫ/Ϫ mice") (21). These mutant mice were recovered in the expected Mendelian ratio, indicating that embryonic development can proceed without Cdk1 expression in osteoblasts. Although the deletion of Cdk1 was confirmed in the calvaria of 3-month-old mice by qPCR and Western blotting, no significant differences in the expression of other Cdks were detected ( Fig. 2A and Fig.  S1). Osx-Cre expression alone has been shown to negatively affect skeletal growth in young mice, but the effect of Cre is diminished by 12 weeks of age (22,23). Thus, we decided to use 3-month-old mice for analysis. To confirm any potential effect of Osx-Cre on bone morphology in our studies, control mice were compared with Osx-Cre mice via microcomputer tomography (CT) and histological analysis. Indeed, bone histomorphometric analysis of the vertebrae and CT analysis of the distal femur in 3-month-old mice showed no significant differences in bone volume between control and Osx-Cre mice, in line with previous reports (Fig. S2). These results show that the Osx-Cre allele per se had a minimal effect on the bone histomorphometric and CT analyses in our study. A bone histomorphometric analysis of the vertebrae of 3-month-old Cdk1 osb Ϫ/Ϫ mice demonstrated a significant decrease in bone mass compared with that of control mice as measured by BV/TV (Fig. 2B). This effect was likely caused by a significant decrease in the number of osteoblasts, which may result in a decrease in the bone formation rate, whereas the osteoclast surface was not affected (Fig. 2C). CT analysis indicated that bone volume/total volume, bone mineral density, trabecular number, trabecular thickness, and cortical thickness were also decreased in the long bones of Cdk1 osb Ϫ/Ϫ mice compared with control mice (Fig. 2D). Accordingly, serum levels of procollagen type 1 N-terminal propeptide (P1NP), a biomarker correlated with bone formation, was decreased in Cdk1 osb Ϫ/Ϫ mice (Fig. 2E). However, serum C-terminal telopeptide of type 1 collagen (CTX-I), a biomarker for bone resorption, showed no difference between Cdk1 osb Ϫ/Ϫ mice and control mice (Fig. 2E). Collec-tively, these results indicate that Cdk1 is important for bone formation.

Osteoblast proliferation is decreased in osteoblast-specific Cdk1-knockout mice
We aimed to examine the molecular mechanism behind the decrease in osteoblast numbers observed in Cdk1 osb Ϫ/Ϫ mice. To test whether decreased bone formation was caused by altered osteoblast survival and/or proliferation, we performed TUNEL and BrdU assays and examined proliferating cell nuclear antigen (PCNA) and Ki67 expression. TUNEL assays performed in femoral sections from 3-month-old female mice failed to reveal any difference in apoptotic cells between Cdk1 osb Ϫ/Ϫ mice and control mice (Fig. 3A). However, BrdU assays performed in femoral sections from 3-month-old female mice suggested a 55% decrease in the number of proliferating osteoblasts in Cdk1 osb Ϫ/Ϫ mice compared with control mice (Fig. 3B). Moreover, there were fewer PCNA-and Ki67-positive osteoblasts in the Cdk1 osb Ϫ/Ϫ sections than in the control sections (Fig. 3C). To further test whether the decrease in bone formation was caused by decreased proliferative activity of osteoblastic progenitors and/or a decreased pool of osteoblast progenitors, we analyzed the expression of Osx and Runx2. Although Runx2 expression was similar in control and Cdk1 osb Ϫ/Ϫ mice, Osx expression levels were lower in Cdk1 osb Ϫ/Ϫ mice than in control mice (Fig. 3D). Thus, the decrease in bone formation is likely caused by decreased proliferative activity of osteoblastic progenitors rather than by a decreased progenitor pool. Collectively, these results indicate that Cdk1 plays an important role in osteoblast proliferation in vivo.

Cdk1 regulates osteoblast differentiation
Because Cdk1 regulates osteoblast proliferation, we wondered whether Cdk1 also affects osteoblast differentiation. To address this question in vitro, we isolated calvarial osteoblasts from WT mice, treated them with RO-3306 or vehicle, and then compared their behavior during differentiation. Cdk1-inhibited osteoblasts formed more mineralized nodules than vehicle-treated osteoblasts (Fig. 4A). To test whether osteogenic differentiation is affected by the Cdk1 status when progenitors are stimulated with osteogenic BMP2, we treated ST2 mesenchymal progenitor cells with RO-3306 followed by BMP2. As expected, inhibiting Cdk1 promoted osteoblastic differentiation of ST2 cells (Fig. 4B), as evidenced by alkaline phosphatase (Alp) and osteocalcin (Ocn) expression. Furthermore, knocking down Cdk1 with shRNA promoted osteoblast differentiation, as confirmed by the elevated expression of the osteoblast differentiation markers Col1a1 and Ocn (Fig. 4C). To address this question in vivo, we performed in situ hybridization analysis for osteoblast differentiation markers. Consistent with our observations in vitro, Cdk1 deletion resulted in accelerated osteoblast differentiation, which was confirmed by the elevated expression of the osteoblast differentiation markers Col1a1 and Ocn (Fig. 4D). In addition, qPCR analysis revealed increased expression of osteoblast differentiation marker genes (Alp, Col1a1, and Ocn) in the calvaria of 3-month-old Cdk1 osb Ϫ/Ϫ mice (Fig.  4E). These results indicate that the biological function of Cdk1 in osteoblasts correlates with their ability to differentiate. Over-

Loss of cyclin-dependent kinase 1 impairs bone formation
all, Cdk1 is important for not only the proliferation but also the proper differentiation of osteoblasts.

Cdk1 is dispensable for the bone-anabolic effect of PTH in vivo
Intermittent PTH treatment enhances bone formation and increases osteoblast number by a variety of mechanisms, including stimulation of osteoblast proliferation (24). Because Cdk1 is important for osteoblast proliferation, we hypothesized that Cdk1 may be involved in the bone-anabolic response to intermittent PTH treatment in vivo. To test this hypothesis, Cdk1 osb Ϫ/Ϫ mice and control mice were treated with either PTH or vehicle for 4 weeks (Fig. 5A). These mice were then analyzed by bone histomorphometry and CT for bone parameters. As expected, PTH increased bone mass by 30% in control mice (Fig. 5B). Surprisingly, PTH also successfully increased BV/TV by 60% in Cdk1 osb Ϫ/Ϫ mice ( Fig. 5B) in the presence of low osteoblast proliferation. Bone histomorphometric analysis revealed that although the number of osteoblasts and the bone formation rate were substantially lower in Cdk1 osb Ϫ/Ϫ mice than in control mice, they were both increased to approximately the same levels by PTH (Fig. 5C). With regard to bone resorption, although there was a trend toward higher bone resorption induced by PTH administration in control and Cdk1 osb Ϫ/Ϫ mice, there was no significant difference between control and Cdk1 osb Ϫ/Ϫ mice (Fig. 5C). We then examined Ki67 expression to test whether the bone-anabolic effect of PTH in Cdk1 osb Ϫ/Ϫ mice was caused by altered osteoblast proliferation. PTH treatment induced osteoblast proliferation, as shown by the markedly increased number of Ki67-positive cells in the control femoral sections; however, the effect of PTH on osteoblast proliferation was blunted in Cdk1 osb Ϫ/Ϫ sections (Fig. 5D). Collectively, these results indicate that Cdk1 in osteoblasts may be dispensable for the bone-anabolic effect of intermittent PTH treatment, suggesting that the bone-anabolic effect of PTH is independent of osteoblast proliferative ability.

Cdk1 is essential for bone fracture healing
The bone fracture healing process consists of intramembranous and endochondral bone formation (25). To examine the role of Cdk1 during bone fracture repair, we performed femoral fractures on Cdk1 osb Ϫ/Ϫ mice and control mice at 3 months of age. X-ray analysis revealed no significant differences between control and Cdk1 osb Ϫ/Ϫ mice in callus formation at 6 weeks after surgery (Fig. 6A), indicating that the bone repair process was activated normally. CT analysis demonstrated bony bridging across the fracture site at 6 weeks after surgery in control mice, suggesting bony union. However, no bony bridging was observed in Cdk1 osb Ϫ/Ϫ mice, suggesting failed union (Fig. 6B). These results indicate that Cdk1 in osteoblasts is not essential for callus formation in bone fracture repair but is required for bony union. Finally, we tested whether the administration of PTH prevented nonunion after bone fracture in the Cdk1 osb Ϫ/Ϫ mice. Indeed, the Cdk1 osb Ϫ/Ϫ mice treated with PTH showed a trend toward a higher bony union rate (50%) than non-PTHtreated Cdk1 osb Ϫ/Ϫ mice (0%) (Fig. 6C). In addition, the mean gap size was significantly lower in the PTH group than in the non-PTH group (41.5 versus 172.2 m in the non-PTH-treated Cdk1 osb Ϫ/Ϫ mice; p Ͻ 0.05) (Fig. 6D).

Discussion
In this study, we investigated the roles of Cdks in bone formation. First, we showed that Cdk1 is expressed in bone and that its protein levels are down-regulated upon osteoblast differentiation. Then, we demonstrated that deletion of Cdk1 in osteoblasts leads to osteoporosis due to a reduced number of osteoblasts and low bone formation. Finally, we found that the bone-anabolic effect of PTH is achieved without promoting osteoblast proliferation in vivo. Furthermore, we demonstrated that Cdk1 in osteoblasts plays an important role in bone fracture repair. To the best of our knowledge, this is the first study to demonstrate through in vivo osteoblast-specific loss-offunction experiments that Cdk1 has an important role in bone remodeling.
Ovariectomized mice (hereafter, "ovx mice") have been widely used as a mouse model of osteoporosis (13). In fact, ovx mice display low bone mass due to increased osteoclastogenesis, which reflects the characteristics of postmenopausal osteoporosis. However, this mouse model does not reflect the phenotype of age-related osteoporosis because in age-related osteoporosis, bone formation is reduced due to low osteoblast numbers (26). We demonstrated here that Cdk1 regulates osteoblast proliferation and that Cdk1 deletion results in osteoporosis that resembles the phenotype of age-related osteoporosis. Thus, the use of this mouse model may be beneficial for the establishment of new treatment strategies for age-related osteoporosis. Furthermore, deleting Cdk1 in osteoblasts impairs osteoblast proliferation without altering the number of osteoclasts or bone resorption (Fig. 2C). Thus, increasing Cdk1 activity in osteoblasts to promote their proliferation may enhance bone formation without affecting osteoclast parameters.
Interestingly, Cdk1 expression is high in undifferentiated osteoblasts but low in mature osteoblasts. Inhibition of Cdk1 expression and/or activity promoted osteoblast differentiation and inhibited proliferation. These results suggest that Cdk1 functions to maintain osteoblasts in a proliferating state and to delay terminal differentiation, analogous to the function of Cdk1 in maintaining proliferation and self-renewal in mouse embryonic stem cells (27). Thus, down-regulation of Cdk1 expression in osteoblasts would be important for terminal osteoblast differentiation. However, the specific molecular

Loss of cyclin-dependent kinase 1 impairs bone formation
mechanism responsible for down-regulating Cdk1 during osteoblast differentiation remains to be elucidated. Because Cdk1 is an E2F target gene (17), we hypothesize that the expression level and/or functions of E2F or Cdk1 itself should be reg-ulated by transcription factors involved in osteoblast differentiation, such as Runx2 and Osterix. Indeed, pRb, which is an upstream regulator of E2F, has been shown to directly interact with Runx2 (28). Moreover, many putative binding sites for

Loss of cyclin-dependent kinase 1 impairs bone formation
these factors are present in the sequences upstream of the Cdk1 locus, and Runx2 has been shown to directly interact with Cdk1 (29).
Our study showed that the bone-anabolic effect of PTH can be achieved despite impaired osteoblast proliferative ability. PTH has been shown to enhance the commitment of the progenitor proportion to an osteogenic fate in vitro (30). Collec-tively, these results suggest that the anabolic effect of intermittent PTH in vivo may result from an enhancement of the progenitor population to an osteogenic fate rather than promotion of osteoblast proliferation. Although PTH 1 receptor agonists are currently the only bone-anabolic agents indicated for patients with osteoporosis at a high risk of bone fracture, some patients do not respond to PTH treatment. To predict the effec- Ϫ/Ϫ mice 6 weeks (w) after surgery. Although all control mice showed complete cortical continuity, all Cdk1 osb Ϫ/Ϫ mice showed delayed union of bone fracture. n ϭ 3. C, CT images of fracture sites in Cdk1 osb Ϫ/Ϫ mice treated with or without PTH for 6 weeks after surgery. PTH-treated Cdk1 osb Ϫ/Ϫ mice showed a higher bony union rate than non-PTH-treated mice. D, CT gap-size analysis in Cdk1 f/f mice, Cdk1 osb Ϫ/Ϫ mice, and Cdk1 osb Ϫ/Ϫ mice treated with PTH 6 weeks after surgery. Among Cdk1 osb Ϫ/Ϫ mice, mean gap size was significantly reduced by PTH treatment.

Loss of cyclin-dependent kinase 1 impairs bone formation
tiveness of PTH in the early treatment period, several risk factors associated with nonresponse to PTH have been proposed. One study suggested that lower baseline bone turnover marker levels were associated with nonresponse (31), and another study suggested that lower early increases in bone turnover markers after starting PTH treatment were associated with nonresponses (32). Based on our research, we suggest adopting the latter strategy to predict response to PTH, because PTH can act effectively even in a lower bone formation state. In addition to osteoblast proliferation and differentiation, Cdk1 influences bone fracture healing (Fig. 6B). Clinically, older age, malnutrition, alcoholism, and smoking are considered risk factors for nonunion after surgery (33). However, the molecular mechanisms by which these factors affect the development of nonunion are not understood. A recent clinical study showed that lower bone formation and/or a higher tendency for bone resorption are plausible mechanisms for nonunion (34). We suggest that the loss of Cdk1 in osteoblasts leads to low osteoblast numbers due to decreased proliferative activity of osteoblastic progenitors, resulting in low bone formation. Thus, we provide evidence that the low bone formation state caused by Cdk1 deficiency resulted in nonunion after bone fracture, confirming the hypothesis that low bone formation is a significant risk factor for nonunion. Because individual bone formation states can be clinically assessed by bone formation markers (32,34), the use of a bone-anabolic agent to enhance bone formation in patients with a low bone formation state may lead to a reduced risk of nonunion after bone fracture. Although Cdk1 deficiency resulted in nonunion after bone fracture, no significant differences in callus formation were observed between groups. Thus, Cdk1 in osteoblasts may play a role in mineralization of cartilage into lamellar bone via endochondral ossification rather than in cartilage formation at the fracture site. In addition, the partial recovery of bone fracture healing by PTH administration in Cdk1 osb Ϫ/Ϫ mice suggests that not only Cdk1 expression but also sufficient osteoblast numbers may be important for bone fracture healing.
Concerning the regulation of bone resorption by Cdk1, osteoclast-specific Cdk1-knockout mice constructed using the cathepsin-K promoter showed no significant differences in bone mass or osteoclast parameters, 3 although in this study, we showed that bone formation was repressed in osteoblast-specific Cdk1-knockout mice. Thus, Cdk1 does not seem to be essential for osteoclasts but regulates bone metabolism through directly stimulating osteoblasts.
In conclusion, we have demonstrated that Cdk1 is required for proper osteoblast proliferation and differentiation and is important for bone formation. Because Cdk1 inhibitors are currently being tested for the treatment of various diseases, Cdk1 may be a potential therapeutic target for the treatment of diseases involving abnormal osteoblast proliferation, such as osteosarcoma. Moreover, increasing Cdk1 activity in osteoblasts to promote their proliferation may enhance bone formation without affecting osteoclast parameters. In this context, it will be important to develop new strategies for cell typespecific activation or inhibition of Cdk1.

Animals
Cdk1 f/f mice (17) and the Osx-Cre mouse line (21) have been described previously. We crossed Osx-Cre mice with Cdk1 f/f mice to obtain Cdk1 osb Ϫ/Ϫ . All mice were maintained under standard conditions with food and water available ad libitum under a 12-h light/dark cycle. All animal experiments were performed with the approval of the Animal Study Committee of Tokyo Medical and Dental University and conformed to relevant guidelines and laws.

Histomorphometric analysis
Bone histology and histomorphometry were performed at the L3 and L4 vertebrae in 3-month-old female mice, as described previously (35). In brief, vertebrae were dissected out and fixed in 4% formaldehyde for 18 h at 4°C. Undecalcified bones were embedded in methyl methacrylate, and 5-7-m sections were prepared on a rotating microtome. To determine the ratio of mineralized BV/TV, sections were stained with von Kossa reagent (3% silver nitrate) and counterstained with van Gieson solution (Wako). Bone formation rate (BFR) was analyzed by the calcein double-labeling method. Accordingly, mice were injected twice with 20 mg/kg calcein (Sigma) at 7 and 2 days, respectively, before sacrifice. BFR was calculated as the product of the mineral apposition rate (MAR) and the mineralizing surface/bone surface (MS/BS), (BFR ϭ MAR ϫ (MS/ BS)). MAR is the distance between the midpoints of the two labels divided by the time between the midpoints of the interval. MS/BS represents the percentage of bone surface exhibiting mineralizing activity. For the analysis of osteoblast parameters (osteoblast number per tissue area), bone sections were stained with 1% toluidine blue. For the analysis of osteoclast parameters (osteoclast surface per bone surface), bone sections were incubated for 30 min in TRAP staining solution at 37°C and then counterstained with hematoxylin. Osteoclasts were defined as multinucleated dark red cells along the bone surface. Histological sections were viewed under a microscope (Olympus) using a ϫ20 objective lens. Histomorphometric analysis was performed using the Osteomeasure System (OsteoMetrics). For each group, 4 -6 mice were analyzed.

Cell culture
Cells were purchased from the Riken Cell Bank (Tsukuba, Japan). MC3T3-E1 cells were maintained in ␣-minimal Eagle's medium (Sigma) containing 2 mM L-glutamine, 100 units/ml penicillin, 10 g/ml streptomycin, and 10% fetal bovine serum (FBS; Sigma) in 5% CO 2 . ST2 cells were cultured in RPMI 1640 medium (Sigma). Cells were treated with osteogenic medium (containing 10% FBS and 30 ng/ml recombinant human BMP2 (funakoshi)) with or without 10 M RO-3306 (Sigma) to inhibit Cdk1 activity. For cell counting, we counted the total number of cells in the visual field of four different regions. The results are representative of more than four independent experiments.

Transfection and infection
Cdk1 cDNA fragments were amplified by PCR. The PCR fragments were cloned into the pcDNA3.2/V5-DEST vector (Invitrogen). For Cdk1 overexpression studies, MC3T3-E1 cells were seeded and transfected using the Lipofectamine LTX reagent (Invitrogen) according to the manufacturer's instructions. To establish stable cell lines, we constructed retroviruses expressing shRNA against Cdk1 using the RNAi-Ready pSIREN Vector (Clontech) and Platinum Retroviral Expression System (Cell Biolabs, Inc), as directed by the manufacturers. The shRNA sequence was 5Ј-AATTCCTCGAGAAAAAATGCC-AGAGCGTTTGGAATATCTCTTGAATATTCCAAACGC-TCTGGCACG-3Ј (for Cdk1). Stable clones expressing shRNA against Cdk1 or the luciferase gene were selected using 3 g/ml puromycin.

Quantitative real-time PCR analysis
To acquire RNA from mouse bones, we flushed mouse bone marrow out of the femur with PBS and extracted bone RNA with TRIzol reagent (Invitrogen); RNA from other tissues and cultured cells was also extracted using TRIzol reagent. Reverse transcription was performed using the High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems) according to the manufacturer's instructions. We performed quantitative analysis of gene expression using the Mx3000p qPCR system (Agilent Technologies). Gapdh expression was used as an internal control. Primer sequences are available upon request.

Western blot analysis
For immunological detection, 20 g of cell lysate was separated via SDS-PAGE (7.5-10% Tris gel). After the proteins were blotted onto a PVDF membrane, the membrane was incubated with the PVDF blocking reagent Can Get Signal (TOYOBO). Proteins were probed with primary antibodies against Cdk1 (MBL) and glyceraldehyde-3-phosphate dehydrogenase (MBL). A horseradish peroxidase-conjugated goat anti-rabbit antibody was then added, and secondary antibodies were detected through autoradiography using enhanced chemiluminescence (ECL Plus, GE Healthcare).

Microcomputed tomography analysis
We obtained three-dimensional images of distal femurs via CT (Comscan). We examined more than six mice for each group for bone morphometric analysis.

Biochemistry
Blood samples were collected through cardiac puncture, kept at room temperature for 30 min, and centrifuged at 12,000 ϫ g for 15 min at 4°C. Serum P1NP and C-terminal telopeptide of type 1 collagen (CTX-I) levels were measured by an ELISA kit for mouse P1NP (Cloud-Clone) and CTX-I (Ids) according to the manufacturers' instructions, respectively.

BrdU labeling
Mice were injected intraperitoneally with BrdU (100 g/g body weight) 24 and 2 h before being sacrificed. Limbs were dissected and embedded in 4% CMC sodium. BrdU was detected using a BrdU immunohistochemistry kit (Abcam) according to the manufacturer's protocol. After incubation with streptavidin-horseradish peroxidase conjugate, we used a tyramide signal amplification system (PerkinElmer Life Sciences) to detect fluorescent signals. Ki67-positive cells in the osteoids of femur were counted and normalized the counts to the total cell numbers in the same area. Ki67-positive hematopoietic cells in the bone marrow were not counted.

TUNEL assay
Osteoblast apoptosis in Cdk1 f/f and Cdk1 osb Ϫ/Ϫ mice at the age of 12 weeks was examined via TUNEL assays. TUNEL assays were performed with the ApopTag system (Millipore) according to the manufacturer's instructions. After applying anti-digoxigenin conjugate, we used a tyramide signal amplification system (PerkinElmer Life Sciences) to detect fluorescent signals.

In situ hybridization
In situ hybridization was performed using a DIG-labeled riboprobe according to the standard protocol, as described previously (36). Hybridizations were performed at 55°C. Riboprobes for Ocn and type I collagen were provided by Dr. Shu Takeda (Tokyo Medical and Dental University, Tokyo, Japan). For detection, signals were developed using anti-DIG antibody conjugated with alkaline phosphatase. After antibody treatment, the sections were incubated with BM Purple (Roche Applied Science).

PTH treatment protocol
The study design is shown schematically in Fig. 5A. Control female mice and Cdk1 osb Ϫ/Ϫ female mice at 12 weeks of age were treated with either vehicle or 80 g/kg/day hPTH(1-34) (Bachem) for 4 weeks. Either vehicle or PTH was administered subcutaneously to the respective groups (5 days a week). All mice were sacrificed after 4 weeks of treatment.

Mouse femoral fracture model
Anesthetized mice were placed on their backs, and the right leg was maximally flexed at the knee. An anterior longitudinal midline incision was made centered over the knee joint using a No. 15 scalpel blade. A subsequent incision was made just medial to the patella and extensor mechanism. The extensor mechanism was elevated and displaced in a lateral fashion using forceps. After the extensor mechanism was subluxated laterally, the distal end of the femur was exposed. Using a blade, the Loss of cyclin-dependent kinase 1 impairs bone formation femur was broken transversely at approximately the distal third of the femoral bone. Then, we inserted a 25-gauge needle from the trochlear groove of the femur into the medullary canal in a retrograde manner. Through the fracture site, we made sure that the needle was completely inserted into the distal and proximal parts of the bone. The depth of insertion could be manually felt due to the resistance upon meeting the cortical bone of the greater trochanter. The needle was cut flush with the distal end of the femur with wire cutters. The extensor mechanism was pulled back to its anatomic location using forceps. The incision was then closed with 5-0 nonabsorbable sutures.
The callus index was defined as the ratio of the maximum callus diameter to the bone diameter at the same level as the callus (37). We measured the callus index at regular intervals using radiography during the healing period to observe callus growth and remodeling. For PTH treatment after bone fracture, 3-month-old Cdk1 osb Ϫ/Ϫ female mice were fractured and then treated with 80 g/kg/day hPTH(1-34) for 6 weeks.

Statistics
All data are presented as the means Ϯ S.D. (n Ն 3). We performed statistical analysis using Student's t test, and p Ͻ 0.05 was considered statistically significant.