Parathyroid hormone uses multiple mechanisms to arrest the cell cycle progression of osteoblastic cells from G1 to S phase.

Parathyroid hormone (PTH) plays a major role in bone remodeling and has the ability to increase bone mass if administered daily. In vitro, PTH inhibits the growth of osteoblastic cell lines, arresting them in G(1) phase. Here, we demonstrate that PTH regulates the expression of at least three genes to achieve the following: inducing expression of MAPK phosphatase 1 (MKP-1) and p21(Cip1) and decreasing expression of cyclin D1 at both mRNA and protein levels. The induction of MKP-1 causes the dephosphorylation of extracellular signal-regulated kinase and therefore the decrease in cyclin D1. Overexpression of MKP-1 arrests UMR cells in G(1) phase. The mechanisms involved in PTH regulation of these genes were studied. Most importantly, PTH administration produces similar effects on expression of these genes in rat femoral metaphyseal primary spongiosa. Analyses of p21(Cip1) expression levels in bone indicate that repeated daily PTH injections make the osteoblast more sensitive to successive PTH treatments, and this might be an important feature for the anabolic functions of PTH. In summary, our data suggest that one mechanism for PTH to exert its anabolic effect is to arrest the cell cycle progression of the osteoblast and hence increase its differentiation.

Parathyroid hormone (PTH) plays a major role in bone remodeling and has the ability to increase bone mass if administered daily. In vitro, PTH inhibits the growth of osteoblastic cell lines, arresting them in G 1 phase. Here, we demonstrate that PTH regulates the expression of at least three genes to achieve the following: inducing expression of MAPK phosphatase 1 (MKP-1) and p21 Cip1 and decreasing expression of cyclin D1 at both mRNA and protein levels. The induction of MKP-1 causes the dephosphorylation of extracellular signal-regulated kinase and therefore the decrease in cyclin D1. Overexpression of MKP-1 arrests UMR cells in G 1 phase. The mechanisms involved in PTH regulation of these genes were studied. Most importantly, PTH administration produces similar effects on expression of these genes in rat femoral metaphyseal primary spongiosa. Analyses of p21 Cip1 expression levels in bone indicate that repeated daily PTH injections make the osteoblast more sensitive to successive PTH treatments, and this might be an important feature for the anabolic functions of PTH. In summary, our data suggest that one mechanism for PTH to exert its anabolic effect is to arrest the cell cycle progression of the osteoblast and hence increase its differentiation.
The adult human skeleton is continuously resorbed and renewed by the actions of osteoclasts and osteoblasts. The maintenance of the skeleton requires the coordinated activities of these cells. Disruption of this coordination underlies many bone diseases, including osteoporosis. Parathyroid hormone (PTH) 1 is released by the parathyroid gland in response to slightly lowered serum ionized calcium concentrations and directly acts on the osteoblast but not the osteoclast. It plays a major role in the balance between bone formation and bone resorption. Although the physiological role of PTH is to mobilize calcium from bone into blood by stimulating bone resorption, current interest has focused on its ability to increase bone mass in humans when administered once daily.
Several mechanisms have been proposed to explain this anabolic effect (for a review, see Ref. 1). Among them, a prevailing hypothesis proposes that PTH promotes osteoblast differentiation. Osteoblasts originate from bone marrow stromal stem cells. These cells undergo proliferation and differentiation into preosteoblasts and then into a final stage of differentiation concomitant with extracellular matrix mineralization. These cells are considered to be terminally differentiated mature osteoblasts that must exit the cell cycle and end their differentiation as osteocytes embedded in a mineralized matrix. Therefore, arresting cell cycle progression of the osteoblast could be part of the mechanism elicited by PTH to facilitate differentiation. In vitro studies demonstrated PTH could stimulate the proliferation of osteoblastic cells under some circumstances, such as low concentration of PTH (2), high cell density (3), short exposure to PTH (4), or in a particular cell line (TE-85) (5), through a possible mechanism of increasing Cdc2 expression (6). In chondrocytes and chondrosarcoma cells, an activating mutation of the PTH receptor (PTH1R) causes an increase in cyclin D1 and cyclin A expression (7). However, a high PTH concentration (10 Ϫ8 M) consistently inhibits the growth of several well established osteoblastic cell lines (UMR 106 -01 (8), MC3T3-E1 (9), SaOS-2 (10), and rat calvarial primary cultures). 2 Mitogen-activated protein kinase (MAPK) phosphatase-1 (MKP-1), originally named Erp/3CH134 or CL100, is a protein capable of regulating cell cycle progression. MKP-1 is an immediate early gene induced by growth factors and stress (11,12). It belongs to a family of dual specificity phosphatases capable of removing both phosphotyrosine and phosphothreonine from the Thr-X-Tyr motif, whose dual phosphorylation is critical for the activation of MAPKs (reviewed in Refs. 13 and 14). Evidence strongly suggests that one of the MAPKs, extracellular signal-regulated kinases (ERK1/2, p42/p44 MAPK ), whose dual phosphorylation is connected to cell proliferation in most cell types, are substrates for MKP-1 (15). Whereas constitutive overexpression or microinjection of MKP-1 blocks S phase entry in fibroblasts (16) and suppresses normal and oncogene-driven proliferation (17,18), antisense oligonucleotides targeted to MKP-1 prolong the activation state of MAPK (19).
In mammalian cells, the cell cycle is governed by the activities of the cyclin-dependent kinases (CDKs) and their regulatory cyclin partners. Progression from G 1 to S phase requires cyclin D-Cdk4 (6) and cyclin E-Cdk2. These complexes phosphorylate retinoblastoma protein (Rb), resulting in the dissociation of Rb and E2F. Released E2F is essential for cells to enter S phase. CDK inhibitors (CDKIs) are of two types: INK4 proteins interfere with cyclin D-Cdk4(6), and CIP/KIP members (p21 Cip1 , p27 Kip1 , and p57 Kip2 ) bind to both complexes. Whereas it acts as an assembly factor for cyclin D-Cdk4(6), CIP/KIP inhibits the activity of cyclin E-Cdk2. Upon cell cycle arrest, the levels of CIP/KIP proteins increase, saturate D-type cyclins, and then bind to cyclin E-Cdk2 to block the catalytic activity of that kinase (20).
Previous studies demonstrated that 10 Ϫ8 M PTH arrested UMR 106 cells, a rat osteoblastic osteosarcoma cell line, in G 1 phase possibly by up-regulation of p27 Kip1 (21). However, the detailed mechanism of how PTH inhibits cell cycle progression and the relationship between this feature and the anabolic effects that PTH has on bone are not fully understood. Here we report that PTH employs several mechanisms to inhibit cell cycle progression: inducing expression of MKP-1 and p21 Cip1 and decreasing expression of cyclin D1. This is true not only in osteoblastic cells (UMR 106-01 and rat calvarial primary culture) but also in PTH-treated osteoblast-rich bone samples, suggesting that arresting osteoblastic cells in G 1 could be an important part of the anabolic effect of PTH.
Cell Culture-UMR 106 -01 cells were maintained in Eagle's minimal essential medium supplemented with 5% fetal bovine serum, nonessential amino acids, 25 mM HEPES (pH 7.5), 100 units/ml penicillin, and 100 g/ml streptomycin. For experiments, cells were seeded in 100-mm dishes at a density of 1.2 ϫ 10 4 cells/cm 2 overnight and then switched to serum-free minimal essential medium for 1 day before the addition of PTH. Rat primary calvarial osteoblastic cells were obtained and cultured as described previously (22). Before PTH treatment, these cells were serum starved for 1 day.
In Vivo Injection of PTH into Rats-Sprague-Dawley rats (4-week-old male, about 75 g, and 3-month old female, about 250 g) were purchased from Hilltop (Scottdale, PA). Rats were injected subcutaneously with vehicle (0.9% saline solution) or hPTH (1-38 for 4-week-old and 1-34 for 3-month-old; 8 g/100 g) and euthanized using CO 2 at the indicated time after injection. The animal protocols were approved by the Robert Wood Johnson Medical School Institutional Animal Care and Use Committee. The primary spongiosa samples from the distal femur were harvested immediately after sacrifice as described previously (23).
Analysis of mRNA Abundance by Real Time RT-PCR-Cells or tissues were harvested at the indicated time points after PTH treatment. Total RNA was isolated using Tri Reagent (Sigma) followed by purification with the RNeasy kit (Qiagen). A TaqMan® Reverse Transcription kit (Applied Biosystems) was used to reverse transcribe mRNA into cDNA. Following this, PCR was performed on Opticon (MJ Research) using a SYBR® Green PCR Core kit (Applied Biosystems). Each analysis was performed three to four times with independent sets of cells or tissues from PTH treatment to RT-PCR to obtain mean values and S.E. shown in the figures. The primers used for the RT-PCR are summarized in Table I. For UMR 106-01 cells, ␤-actin was used as an internal control. For femoral samples and primary osteoblastic cells, glyceraldehyde-3-phosphate dehydrogenase was used as an internal control.
Cell Cycle Analysis-UMR 106-01 cells were seeded into 6-well plates at 5 ϫ 10 5 cells/well and grown overnight. Cells were then transfected with pCMV-hMKP-1 (a generous gift from Dr. Laura Mauro) or pCMV control vector along with the transfection marker pEGFP-F (Clontech), a plasmid containing enhanced green fluorescent protein with a farnesylation sequence. The next day, cells were trypsinized, washed once with PBS, and fixed in 70% ethanol for at least 1 h on ice. Fixed cells were washed with PBS and incubated with propidium iodide solution containing ribonuclease A. Cell cycle analyses were performed on a Beckman Coulter XL.
Immunoblotting-Preparation of cell lysates and Western blot analyses were performed as described previously (24).

PTH Regulates MKP-1 mRNA Abundance in Osteoblastic
Cells-Recent microarray studies from our laboratory identified MKP-1 as a target gene for rat PTH-(1-34) in UMR 106-01 cells (25). We confirmed this result by using real time RT-PCR. As shown in Fig. 1A, 10 Ϫ8 M rPTH-(1-34) greatly induced MKP-1 mRNA expression in UMR 106-01 cells, with a peak of 15-fold after 1 h of treatment. This induction occurs very quickly, since there is a 4-fold increase of MKP-1 expression after just 15 min of PTH treatment. Western blot demonstrates that PTH also regulates MKP-1 expression at the protein level ( Fig. 2A). In untreated UMR cells, there is a very low detectable amount of MKP-1. Similar to the mRNA induction, MKP-1 protein peaks after 1 h of treatment, indicating that MKP-1 mRNA translates rapidly. A similar phenomenon is also observed in another osteoblastic cell preparation, rat primary calvarial osteoblastic cells (Fig. 1B). These cells mimic the development of osteoblasts in bone and go through three stages (proliferation, differentiation, and mineralization) sequentially when cultured in dishes. Cells at different stages were treated with rPTH-(1-34) (10 Ϫ8 M) for different time periods, and the level of MKP-1 mRNA was measured by real time RT-PCR. At all developmental stages, MKP-1 levels were stimulated to the greatest level at 1 h and then decreased to base line at 4 h. The highest induction (18-fold) is observed in the mineralization stage, and the lowest (2-fold) is seen in the proliferation stage.
The Induction of MKP-1 Expression Correlates with a Decrease in ERK Phosphorylation in Osteoblastic Cells-PTH-or cAMP-generating agents such as forskolin have been shown to inhibit the basal activity of ERK1/2 or their activation by epidermal growth factor in osteoblastic cells (26,27). Consistent with those results, we found that in UMR 106-01 cells rPTH-(1-34) (10 Ϫ8 M) strongly but not completely inhibits the basal phosphorylation of ERK1/2 using Western blot analysis ( Fig. 2A). The maximum inhibition occurs 1 h after the addition of PTH. Since phosphorylated ERKs are substrates for MKP-1 activity, we compared the pattern of MKP-1 protein induction and ERK1/2 dephosphorylation after PTH treatment ( Fig. 2A, top two panels). It is interesting to note that both patterns correlate well with maximal effects at 1 h. This result implies that the induction of MKP-1 may be responsible for the decrease in ERK1/2 phosphorylation.
Next we used two approaches to test this hypothesis. First, we pretreated UMR cells with cycloheximide to inhibit de novo protein synthesis. As shown in Fig. 2B, cycloheximide completely abolished the MKP-1 protein synthesis induced by PTH. Concurrently, it also eliminated the effect of PTH on ERK dephosphorylation, suggesting that this effect requires new protein synthesis that would include MKP-1. Second, we pretreated UMR cells with orthovanadate, a tyrosine phosphatase inhibitor, to inhibit MKP-1 activity. Whereas the addition of orthovanadate alone had no effect on ERK phosphorylation in UMR cells, it totally abolished the dephosphorylation of ERK caused by PTH (Fig. 2C), providing more evidence that MKP-1 probably mediates the effect of PTH on the dephosphorylation of ERK.
Overexpression of MKP-1 Arrests UMR 106-01 Cells in G 1 Phase-MKP-1 plays an important role in cell cycle regulation. To address its role in the osteoblast, we transfected a human MKP-1 plasmid into UMR cells and measured the percentage of cells in each phase of the cell cycle by flow cytometry. Since the transfection efficiency of UMR 106-01 cells is about 20% (data not shown), we used farnesylated enhanced green fluorescent protein (EGFP-F) as a cotransfection marker to do this experiment. Cell cycle analyses were performed on EGFP-F-expressing cells, which should also have the MKP-1 vector. As shown in Table II, flow cytometry revealed that about 47% of UMR cells cotransfected with EGFP-F, and various amounts of pCMV vector controls are in G 1 phase, and 29 and 23% of cells are in S phase and G 2 ϩ M phases, respectively. Interestingly, in cells cotransfected with EGFP-F and 1 g/well MKP-1, there is a definite increase in the percentage of cells in G 1 phase (56%) as well as a decrease in the percentage of cells in S phase (20%) and in G 2 ϩ M phases (20%). More importantly, as the amount of transfected MKP-1 plasmid increased to 2 g/well, there were more cells in G 1 phase (65%) and fewer cells in S phase (16%) and in G 2 ϩ M phase (17%). These results clearly show that overexpression of MKP-1 prevents UMR cells progressing from G 1 to S phase of the cell cycle.
PTH Inhibits Cyclin D1 Expression Possibly through MKP-1 Regulation-The proliferative effect of phosphorylated ERK is partly mediated by its ability to induce cyclin D1 expression (28). Since the above data suggest that the up-regulation of MKP-1 by PTH causes the dephosphorylation of ERKs and overexpression of MKP-1 inhibits cell cycle progression, we hypothesized that similar PTH treatment will decrease cyclin D1 expression. Indeed, our previous microarray analysis suggested that cyclin D1 expression decreased to about 43% in UMR cells after 4 h of rPTH-(1-34) (10 Ϫ8 M) treatment (25). We confirmed this result using real time RT-PCR. As shown in Fig.  3A, PTH treatment significantly decreased cyclin D1 expression to about 60% of base line at 4 and 8 h (p Ͻ 0.05). This effect is more evident at the protein level. Densitometric measurement of Western blots revealed that cyclin D1 protein was 51, 36, 28, and 42% of control after 2, 4, 8, and 24 h, respectively, of PTH treatment (Fig. 3B). It is interesting to note that the effect of PTH on cyclin D1 starts at 2 h and peaks around 4 -8 h, just after the maximal effect (1 h) on MKP-1 expression and dephosphorylation of ERK.
Cyclin D1 expression is also inhibited by rPTH-(1-34) (10 Ϫ8 M) in rat primary calvarial osteoblastic cells in the mineralization phase (Fig. 3C). After 12 h of treatment, the level of cyclin D1 mRNA significantly decreased to 65% in PTH-treated cells compared with that in control cells (p Ͻ 0.05). A 1.8-fold increase of cyclin D1 mRNA was observed after 1 h of PTH treatment. However, this increase is not statistically significant (p ϭ 0.2). There was no effect of PTH on cyclin D1 in primary osteoblastic cells in the differentiation phase.
If PTH regulation of cyclin D1 is through MKP-1 and ERK pathways, then cyclin D1 must be a secondary response gene for PTH, and this response will be eliminated when MKP-1 synthesis or activity is blocked. In fact, cycloheximide abolished the PTH effect on cyclin D1 expression in UMR cells (Fig.  3D), suggesting that this effect also requires new protein synthesis. As shown in Fig. 3E, orthovanadate had no effect on the basal expression of cyclin D1 protein (compare lanes 1 and 3). However, this MKP-1 activity inhibitor completely eliminated the inhibition of cyclin D1 by PTH treatment (compare lanes 2  and 4). Therefore, it is reasonable to propose that the effect of PTH on cyclin D1 is downstream of MKP-1 and ERK.
PTH Stimulates p21 Cip1 , but Not p27 Kip1 , in UMR 106-01 and Rat Calvarial Osteoblastic Cells-To further understand the mechanism of how PTH arrests cell cycle progression of osteoblastic cells, we performed real time RT-PCR to study the expression levels of other cyclins (cyclin D2, D3, E, A, and B), CDKs (Cdk2, -4, and -6), and CDKIs (p21 Cip1 , p27 Kip1 , p57 Kip2 , p15, and p16) in UMR 106-01 cells after PTH treatment. Only p21 Cip1 , and not the other aforementioned genes, was found to be regulated by PTH. The mRNA level of p21 Cip1 was quickly stimulated by PTH, with a maximal 8-fold increase at 1 h (Fig.  4A). Western blot analysis confirmed this result. Whereas there was no detectable level of p21 Cip1 in untreated cells, p21 Cip1 protein became obvious at 1 h and peaked between 2 and 4 h after treatment (Fig. 4B). However, there was no significant change observed in both mRNA and protein levels of p27 Kip1 by PTH throughout the time course (Fig. 4, A and B). This is inconsistent with a previous study indicating that PTH stimulates the protein level of p27 Kip1 (3-fold), but not p21 Cip1 , in UMR 106 after 12 and 24 h (21). The reason for the discrepancy of the PTH effect on p27 Kip1 is not known, but it could be due to the strain difference between UMR 106 used previously and UMR 106-01 used in the present report.
Furthermore, a similar effect of PTH on p21 Cip1 and p27 Kip1 was also observed in primary calvarial osteoblastic cells. In cells in the differentiation phase, PTH stimulated p21 Cip1 mRNA 3.8-fold at 4 h, and in cells in the mineralization phase,

FIG. 2. PTH inhibits ERK phosphorylation through an increase in MKP-1 expression.
A, UMR 106 -01 cells were serumstarved for 1 day and then treated with 10 Ϫ8 M rPTH-(1-34). Protein lysates were collected at the indicated time points, and the same amounts of protein (35 g/well) were analyzed by immunoblotting using various antibodies. Western blots with total ERK1/2 antibody were used as an internal control. B, the dephosphorylation of ERK1/2 by PTH requires new protein synthesis. Control (ethanol) or 30 g/ml cycloheximide (CHX) was added to the medium 30 min before the addition of 10 Ϫ8 M rPTH-(1-34). C, blockage of MKP-1 activity by orthovanadate abolishes the effect of PTH on ERK1/2 dephosphorylation. 50 M Na 3 VO 4 was added to the medium 30 min before the addition of 10 Ϫ8 M rPTH-(1-34). the induction was 2.6-and 3.5-fold at 1 and 4 h, respectively (Fig. 4C). Consistent with the above results from the UMR 106-01 cells, we did not detect a change in p27 Kip1 mRNA level after PTH treatment (data not shown).

TABLE II Overexpression of MKP-1 in UMR 106-01 cells inhibits cell cycle progression from G 1 to S phase
Expression of MKP-1, Cyclin D1, p21 Cip1 , and p27 Kip1 during Osteoblast Differentiation-During its differentiation, the osteoblast gradually ceases proliferation and expresses various bone-specific markers at different differentiation stages. For example, in primary calvarial cultures, alkaline phosphatase has its highest expression level during the differentiation phase, whereas osteocalcin is a marker of mineralization. To further investigate the role of the above PTH-regulated cell cycle genes, we studied their expression in primary calvarial cells during differentiation using real time RT-PCR (Fig. 5). MKP-1 mRNA increased 2-fold in both differentiation and mineralization phases compared with the proliferation phase. p21 Cip1 , along with another CDKI p27 Kip1 , was expressed to a greater extent in the later two phases, since the mRNA levels of p21 Cip1 in the differentiation and mineralization phases were 6-and 5-fold that in the proliferation phase. For p27 Kip1 , it is 3-and 4-fold increased. These results demonstrate why the differentiated osteoblast ceases proliferation. Surprisingly, we observed that cyclin D1 had its highest expression in the differentiation phase (5-fold compared with that in the proliferation phase). As cells mineralized, its expression decreased about 50%, probably adding another mechanism preventing mineralized osteoblastic cells from proliferating.
PTH Regulation of MKP-1 and p21 Cip1 Is PKA-dependent and Is a Primary Response-PTH signals through both PKA and PKC pathways after binding with its receptor, PTH1R, on the osteoblast membrane (29,30). To study which pathway PTH uses to regulate MKP-1 and p21 Cip1 , we took advantage of different peptide fragments of PTH that activate different pathways and inhibitors and activators of these pathways. As shown in Fig. 6A, hPTH-(1-31) (which activates PKA but not PKC) retained the ability to stimulate MKP-1 and p21 Cip1 mRNA levels (21-and 8-fold, respectively) after a 1-h treatment of UMR 106-01 cells. However, hPTH-(13-34) (which activates PKC but not PKA) had no effect on the expression of those genes. In comparison, 8-bromo-cyclic AMP, a cell-permeable cAMP analog, stimulated the expression of MKP-1 and p21 Cip1 about 13-and 8-fold, respectively. PMA, an activator of the PKC pathway, had very little effect on MKP-1 expression (3-fold increase compared with 22-fold increase by rPTH- ). In contrast, PMA stimulated p21 Cip1 expression by 5-fold, the same level as rPTH-(1-34) achieved, indicating that p21 Cip1 could be regulated by both PKA and PKC pathways. Nevertheless, in the presence of the PKA inhibitor H-89, but not in the presence of the PKC inhibitor GF109203X, rPTH-(1-34) lost its ability to induce both MKP-1 and p21 Cip1 , suggesting that this regulation is PKA-dependent but not PKC-dependent in osteoblastic cells.
To determine whether PTH induction of MKP-1 and p21 Cip1 is a primary response, UMR 106-01 cells were treated with rPTH-(1-34) (10 Ϫ8 M) in the presence or the absence of cycloheximide. Fig. 6B shows that cycloheximide had no effect on the PTH induction of MKP-1 and p21 Cip1 , suggesting that this process does not require de novo protein synthesis (i.e. this effect is a primary response).
MKP-1, Cyclin D1, and p21 Cip1 Expression Are Similarly Regulated by PTH in Bone-A classic animal model for PTH anabolic function is repeated daily injection of PTH in rats. A single injection of hPTH-(1-34) (8 g/100g) is sufficient to modify the expression of genes involved in the actions of PTH in rat osteoblast-enriched femoral metaphyseal primary spongiosa (23). We injected human PTH-(1-38) (8 g/100 g) or saline vehicle into young male rats (4-week-old) and harvested RNA from the femoral metaphyses at various time points. Real time RT-PCR was performed to study the expression levels of MKP-1, cyclin D1, and p21 Cip1 in these animals. Fig. 7 demonstrates that one injection of PTH greatly stimulated MKP-1 expression with a 7-fold stimulation over control after 30 min (A), decreased cyclin D1 expression significantly to 60% of control levels 1 h after PTH injection (B), and increased p21 Cip1   FIG. 4. CDK inhibitor p21 Cip1 , but not p27 Kip1 , is an early response gene for PTH. UMR 106-01 cells (A and B) and primary calvarial osteoblastic cells in differentiation (day 14) and mineralization (day 21) phases (C) were serum-starved for 1 day and then treated with 10 Ϫ8 M rPTH-(1-34). RNA and protein lysates were collected at the indicated time points and were analyzed by real time RT-PCR (A and C) and immunoblotting (B), respectively. Western blot with cAMP-response element-binding protein antibody (CREB) was used as an internal control.
FIG . 5. The expression of MKP-1, cyclin D1, p21 Cip1 and p27 Kip1 during osteoblast differentiation. Real time RT-PCR was used to analyze RNA extracted from rat calvarial primary osteoblastic cells at three stages of differentiation: day 6 (proliferation), day 14 (differentiation), and day 21 (mineralization) in culture. expression more than 2-fold at both 30 min and 1 h (C). These results are consistent with our results on the effects of PTH in UMR and primary osteoblastic cells, strongly suggesting that the inhibitory effect of PTH on cell cycle progression observed in several osteoblastic cell lines is not confined only to in vitro cell culture experiments but is indeed related to the anabolic function of PTH in vivo.
Similar to previous reports (31), we observed a significant anabolic effect of PTH on trabecular bone, including an increase in bone mineral density, bone formation rate, and osteoblast numbers, in 3-month-old female rats after 14 days of daily hPTH-(1-34) (8 g/100 g) administration (data not shown). After 2 weeks of injections, RNA was harvested from femoral metaphyseal primary spongiosa, and the levels of the genes of interest were compared with that of vehicle-treated samples using real-time RT-PCR. Surprisingly, we found that whereas one injection of PTH only increased the p21 Cip1 mRNA level 2-fold at 1 h, 14 days of daily injections substantially increased the same transcript 8-fold compared with control 1 h after the last injection (Fig. 7D). Note that the p21 Cip1 mRNA level returns to basal 4 h after one PTH injection (Fig. 7C). Therefore, the basal level of p21 Cip1 mRNA before PTH injection at 14 days should be similar to that of vehicle-treated animals. It appears that repeated PTH injections make the osteoblast more sensitive to the subsequent injections, giving a reason why the anabolic effect of PTH requires multiple administrations. The stimulation of MKP-1 and inhibition of cyclin D1 after 1 h of PTH injection were unchanged between one injection and 14 days of injections. DISCUSSION There are contradictory reports regarding whether PTH stimulates or inhibits osteoblast cell proliferation. It seems that the outcome depends on the cells used, the length of treatment, PTH concentration, etc. However, how this effect is related to the anabolic function of PTH in bone is largely unknown. This is the first report demonstrating that PTH up-regulates MKP-1 and p21 Cip1 and down-regulates cyclin D1 not only in osteoblastic cells, such as UMR 106-01 and primary calvarial cultures, but also in bone cells in vivo after intermittent injection, strongly suggesting that the ability of PTH to inhibit cell proliferation in vitro is relevant to its physiological role.
Regulation of all the above three genes results in preventing cell cycle transition from G 1 to S phase. In most cases, exit from the cell cycle at G 1 is required for terminal cell differentiation. Moreover, cell cycle arrest actually stimulates differentiation. The current model of cell cycle regulation indicates that Rb is inhibited by CDKs. Therefore, the stimulation of p21 Cip1 and inhibition of cyclin D1 would lead to increased activity of Rb. Rb belongs to the pocket protein family, which also includes p107 and p130. An essential role for Rb in all cell types is proliferation control. It acts as a generic corepressor of the E2F family of transcription factors. These proteins, with their partners DP1 and DP2, transactivate genes important for G 1 to S phase transition (32). In addition, Rb also plays important roles in the differentiation of mesenchymal lineages, including myogenesis (33), adipogenesis (34), and osteogenesis (35). Osteosarcoma is the second most common tumor after retinoblas- toma, occurring in individuals with inherited heterozygous loss of the Rb gene (36). In addition, up to 60% of sporadic osteosarcomas have loss of Rb (37,38). It turns out that Rb plays a key role in facilitating expression of late markers of osteoblast differentiation through physical association with Runx2, a transcriptional regulator of osteoblast differentiation. Loss of Rb in osteoblastic cells results in blockage of late osteoblast differentiation (35). In addition to Rb, Cip/Kip CDKIs support differentiation of a variety of cells (39), possibly through increasing Rb activity. A well studied system is MyoD, a skeletal muscle-specific transcription factor that induces terminal cell cycle arrest associated with skeletal muscle differentiation, which has been shown to induce p21 Cip1 (33,40). Our data also suggest that both p21 Cip1 and p27 Kip1 levels increase as osteoblastic cells differentiate. These data are consistent with a previous report on p27 Kip1 but not p21 Cip1 (41). In that report, compared with the proliferative phase, p21 Cip1 protein levels decreased in the differentiation phase and then returned to comparable levels in the mineralization phase. Accordingly, a model is proposed to explain how PTH arrests osteoblastic cells in G 1 and stimulates terminal differentiation (Fig. 8). PTH binds to its receptor PTH1R on the mature osteoblast and activates the PKA pathway. Subsequently, posttranslational modification of some transcription factors, such as cAMP-response element-binding protein, quickly induces the expression of MKP-1 and p21 Cip1 . The increase in MKP-1 protein results in dephosphorylation of its substrate, phosphorylated ERKs, and hence decreases the expression of cyclin D1 and subsequently the activity of the cyclin D1-Cdk4(6) complex. The increase in p21 Cip1 inhibits the activity of the cyclin E-Cdk2 complex. Both result in greater activity of Rb, which would suppress proliferation and promote differentiation through interacting with Runx2.
Why is a two-pronged approach needed for the growth-inhibitory function of PTH? We can speculate that the p21 Cip1 pathway is used to arrest the intrinsic growth of the osteoblast, whereas the MKP-1/cyclin D1 pathway is used to prevent the osteoblast from responding to external growth signals. Osteoblasts reside in an environment containing a variety of cells, such as bone marrow stromal cells, hematopoietic stem cells, blood cells, adipocytes, osteoblast progenitors, osteoclasts, and many more. These cells communicate and influence each other to form a balanced microenvironment. Many growth factors, produced locally from surrounding cells or systemically, can act on the osteoblast and simulate proliferation by increasing ERK phosphorylation. These factors include FGF-2 (42), plateletderived growth factor BB (43), insulin-like growth factor-I (44), epidermal growth factor (24), amphiregulin, 2 etc. The increase in MKP-1 after PTH treatment could antagonize those signals. In fact, this explains a previous report that PTH inhibited the activation of ERK by epidermal growth factor (27). It was also reported that glucocorticoids rapidly up-regulate MKP-1 in osteoblastic cell lines, resulting in dephosphorylation of PMAstimulated ERKs and impairment of osteoblast proliferation (45). Together with our data, it seems that MKP-1 is used as a common mechanism to prevent cells from responding to growth signals.
The anabolic function of PTH requires repeated daily administration. It takes a minimum of 2 weeks to observe a bone mineral density increase in rats and mice, whereas the PTH treatment for osteoporosis patients takes months to be effective. A very recent paper using ovariectomized mice comparing PTH and estradiol treatment showed only a small overlap of the genes expressed, and notably estradiol stimulated both p21 Cip1 and cyclin D1 (46). Interestingly, we found that the stimulation of p21 Cip1 by PTH in rats is higher after 14 days of injection than that after one injection (8-fold versus 2-fold). It seems that repeated injections modify the bone cells in a way that they become more sensitive to the next PTH treatment. However, we did not detect the same change in cyclin D1 and MKP-1 expression. This could be due to the fact that we did not harvest samples at the optimum time point after 14 days of treatment or that only a fraction of PTH response genes behave similarly to p21 Cip1 . Nevertheless, this feature may be important for the anabolic function of PTH, and we are currently studying whether PTH has the same effect on other PTHregulated genes.