PTHrP prevents chondrocyte premature hypertrophy by inducing cyclin-D1-dependent Runx2 and Runx3 phosphorylation, ubiquitylation and proteasomal degradation

Abstract

In chondrocytes, PTHrP maintains them in a proliferative state and prevents premature hypertrophy. The mechanism by which PTHrP does this is not fully understood. Both Runx2 and Runx3 are required for chondrocyte maturation. We recently demonstrated that cyclin D1 induces Runx2 protein phosphorylation and degradation. In the present studies, we tested the hypothesis that PTHrP regulates both Runx2 and Runx3 protein stability through cyclin D1. We analyzed the effects of cyclin D1 on Runx3 protein stability and function using COS cells, osteoprogenitor C3H10T1/2 cells and chondrogenic RCJ3.1C5.18 cells. We found that cyclin D1 induced Runx3 degradation in a dose-dependent manner and that both Myc-tagged Runx3 and endogenous Runx3 interact directly with CDK4 in COS and RCJ3.1C5.18 cells. A conserved CDK recognition site was identified in the C-terminal region of Runx3 by sequence analysis (residues 356-359). Pulse-chase experiments showed that the mutation of Runx3 at Ser356 to alanine (SA-Runx3) increased the half-life of Runx3. By contrast, the mutation at the same serine residue to glutamic acid (SE-Runx3) accelerated Runx3 degradation. In addition, SA-Runx3 was resistant to cyclin D1-induced degradation. GST-Runx3 was strongly phosphorylated by CDK4 in vitro. By contrast, CDK4 had no effect on the phosphorylation of SA-Runx3. Although both wild-type and SE-Runx3 were ubiquitylated, this was not the case for SA-Runx3. Runx3 degradation by cyclin D1 was completely blocked by the proteasome inhibitor PS1. In C3H10T1/2 cells, SA-Runx3 had a greater effect on reporter activity than SE-Runx3. The same was true for ALP activity in these cells. To investigate the role of cyclin D1 in chondrocyte proliferation and hypertrophy, we analyzed the growth plate morphology and expression of chondrocyte differentiation marker genes in Ccnd1-knockout mice. The proliferating and hypertrophic zones were significantly reduced and expression of chondrocyte differentiation marker genes and ALP activity were enhanced in 2-week-old Ccnd1-knockout mice. PTHrP significantly suppressed protein levels of both Runx2 and Runx3 in primary chondrocytes derived from wild-type mice. By contrast, the suppressive effect of PTHrP on Runx2 and Runx3 protein levels was completely abolished in primary chondrocytes derived from Ccnd1-knockout mice. Our findings demonstrate that the cell cycle proteins cyclin D1 and CDK4 induce Runx2 and Runx3 phosphorylation, ubiquitylation and proteasomal degradation. PTHrP suppresses Runx2 and Runx3 protein levels in chondrocytes through cyclin D1. These results suggest that PTHrP might prevent premature hypertrophy in chondrocytes, at least in part by inducing degradation of Runx2 and Runx3 in a cyclin-D1-dependent manner.

Fig. 1.

Cyclin D1 induces Runx3 degradation. (A) A Runx3 expression plasmid was cotransfected with different amounts of cyclin A and cyclin D1 expression plasmids (0.2, 0.6 and 1.8 μg/6-cm culture dish) into COS cells. Western blotting was performed 48 hours after transfection. Cyclin D1 induced Runx3 degradation. By contrast, cyclin A had no effect on Runx3 degradation. (B) Myc-Runx3 and cyclin D1 expression plasmids were cotransfected into COS cells. Immunoprecipitation (IP) was performed using the anti-CDK4 antibody followed by western blotting (WB) with an anti-Myc or anti-cyclin D1 antibody. CDK4 interacts with Runx3 in COS cells. The interaction between cyclin D1 and CDK4 was also detected as a positive control. (C,D) Cell lysates were collected from chondrogenic RCJ3.1C5.18 cells and immunoprecipitation was performed using either an anti-CDK4 antibody or an anti-Runx3 antibody followed by western blotting using the anti-Runx3 or anti-CDK4 antibody. Interaction was detected between endogenous Runx3 and CDK4.

Fig. 2.

Ser356 of Runx3 is critical for Runx3 stability. (A) Protein decay assay. Equal amounts of Myc-Runx3, FLAG-S356A-Runx3 or FLAG-S356E-Runx3 expression plasmid were transfected into COS cells. Cell lysates were extracted after treatment with 80 μg/ml cycloheximide for 0, 20, 60, 120 or 300 minutes and western blotting was performed. WT Runx3 protein levels were considerably reduced at 120 minutes after protein synthesis was inhibited by cycloheximide. FLAG-S356A-Runx3 remained stable during the entire 300 minute period after cycloheximide treatment. By contrast, FLAG-S356E-Runx3 expression began to be considerably reduced at 60 minutes after cells were treated with cycloheximide. (B) Myc-Runx3 and FLAG-S356A-Runx3 expression plasmids were cotransfected with different amounts of cyclin D1 plasmid (0.2, 0.6, and 1.8 μg/dish) into COS cells. Cell lysates were extracted 24 hours after transfection and western blotting was performed. WT Runx3 protein levels were considerably reduced by cyclin D1 in a dose-dependent manner. By contrast, S356A-Runx3 levels were not affected by the overexpression of cyclin D1.

Fig. 3.

Cyclin D1 induces Runx3 phosphorylation. (A) COS cells were transfected with equal amounts of Myc-Runx3 and FLAG-S356A-Runx3 plasmids and treated with PS1 (10 μM, 4 hours incubation). Cell lysates were extracted 24 hours after transfection. Phosphorylation of Runx3 was detected by western blotting (WB) using the anti-phosphoserine antibody after WT and mutant Runx3 protein was immunoprecipitated (IP) using either the anti-Myc or anti-FLAG antibody. Strong phosphorylation was detected in WT Runx3-transfected cells. However, only trace amount of phosphorylated mutant Runx3 (P-Runx3) was detected. (B) In vitro phosphorylation assay. CDK4 was immunoprecipitated from 300 μg C2C12 cell lysate using 1.5 μg anti-CDK4 antibody and was used in the Runx3 kinase assay (1 μM substrate). GST-Rb (379-928) was used as a positive control and GST-Smad4 was used as a negative control. GST-Runx3 exhibits a strong phosphorylation, whereas GST-S356A-Runx3 exhibits a weak phosphorylation. Left: Runx3 phosphorylation by CDK4. Right: protein amounts, detected by Coomassie blue staining.

Fig. 4.

Cyclin D1 induces ubiquitylation and proteasome degradation of Runx3. (A) Myc-Runx3 and HA-ubiquitin expression plasmids were cotransfected with different amounts of cyclin D1 expression plasmid (0.2 and 0.6 μg/6-cm culture dish) into COS cells in the presence or absence of PS1 (10 μM for 4 hour incubation). Cyclin D1 induced Runx3 ubiquitylation in a dose-dependent manner and this effect was further enhanced by the addition of PS1. (B) Myc-Runx3, FLAG-S356A-Runx3 or FLAG-S356E-Runx3 was cotransfected with HA-ubiquitin plasmid into COS cells in the presence of PS1 (10 μM, 4 hours incubation). Runx3 ubiquitylation was detected in the WT Runx3 and S356E-Runx3 groups but not in the S356A-Runx3 group. (C) To further determine the ubiquitylation of endogenous Runx3, chondrogenic RCJ3.1C5.18 cells were treated without or with MG132 (10 μM, 4 hour incubation) before cell lysates were collected. Immunoprecipitation was performed using the anti-Runx3 antibody followed by western blotting using the anti-ubiquitin antibody. Ubiquitylation of endogenous Runx3 was detected in the presence of MG132 in RCJ3.1C5.18 cells. (D) Myc-Runx3 expression plasmid was cotransfected with different amounts of cyclin D1 expression plasmid (0.2, 0.6 and 1.8 μg/dish) into COS cells. Cells were treated with PS1 (10 μM) for 4 hours after transfection. Cyclin D1 induced a dose-dependent degradation of Runx3 and treatment with PS1 completely reversed cyclin-D1-induced Runx3 degradation. (E) WT and mutant Runx3 (S356A) expression plasmids were transfected into COS cells. Cells were treated with PS1 (10 μM, 4 hour incubation) after transfection. The addition of PS1 increased the level of WT Runx3 protein but had not effect on mutant Runx3 (S356A) protein.

Fig. 5.

Ser356 is critical for Runx3 function. (A) Wild-type (WT) and mutant Runx3 (S356E and S356A) expression plasmids were cotransfected with 6xOSE2-Luc reporter into C3H10T1/2 cells. Cell lysates were collected 24 hours after transfection and a luciferase assay was performed. The S356A-Runx3 increased activity and S356E-Runx3 decreased activity of the 6xOSE2-Luc reporter compared with the WT Runx3. ^*P<0.05, compared with the WT Runx3; unpaired Student's t-test. (B) WT and mutant Runx3 (S356E and S356A) expression plasmids were transfected into C3H10T1/2 cells. ALP activity was measured using an ALP assay kit. S356A-Runx3 increased and S356E-Runx3 decreased ALP activity compared with the WT Runx3. ^*P<0.05, compared with the WT Runx3; unpaired Student's t-test.

Fig. 6.

Chondrocyte proliferation and differentiation are altered in Ccdn1-knockout mice. (A-F) Histological analyses, including Safranin O/Fast green (A,C) and Alcian blue/Hemotoxylin and Orange G (B) staining, showed that growth plate length (GPL; D) and the lengths of the proliferating zone (PZ; C,E) and the hypertrophic zone (HZ; C,F) were reduced (n=6) and the formation of the secondary ossification center was delayed (black arrows in A and B) in Ccdn1-knockout mice compared to their WT littermates. (G-M) Primary chondrocytes were isolated from 3-day-old Ccdn1^-/- mice and WT littermates and expression of chondrocyte marker genes and alkaline phosphatase (ALP) staining were examined. The results demonstrated that the expression of chondrocyte marker genes, such as Alp (G), collagen type X (ColX; H), Mmp9 (I), Mmp13 (J), Vegf (K) and osteocalcin (Oc; L), and ALP activity (M) were significantly increased in Ccdn1- knockout chondrocytes (n=3). ^*P<0.05, compared with the WT littermate controls; unpaired Student's t-test.

Fig. 7.

PTHrP downregulates steady-state protein levels of Runx2 and Runx3 in a cyclin-D1-dependent manner. (A,B) The time-course effects of PTHrP on the expression of cyclin D1 mRNA and CDK protein were examined by real-time PCR and western blotting in chondrogenic RCJ3.1C5.18 cells. PTHrP (10^-8 M) significantly increased the expression of cyclin D1 mRNA and CDK4 protein (4-12 hour treatment) in RCJ3.1C5.18 cells. (C,D) Effects of PTHrP on Runx2 and Runx3 protein levels were examined by western blotting in primary chondrocytes. PTHrP significantly downregulated Runx2 (C) and Runx3 (D) protein levels in WT chondrocytes. However, these effects were completely inhibited in Ccdn1-knockout chondrocytes. ^*P<0.05, one-way analysis of variance followed by Dunnett's test, compared with the control group without PTHrP treatment.