FGF2-activated ERK mitogen-activated protein kinase enhances Runx2 acetylation and stabilization

Abstract

Runx2 is a key transcription factor regulating osteoblast differentiation and skeletal morphogenesis, and FGF2 is one of the most important regulators of skeletal development. The importance of the ERK mitogen-activated protein (MAP) kinase pathway in cranial suture development was demonstrated by the findings that the inhibition of FGF/FGF receptor (FGFR) signaling by a MEK blocker prevents the premature suture closure caused by an Fgfr2 mutation in mice. We previously demonstrated that ERK activation does not affect Runx2 gene expression but that it stimulates Runx2 transcriptional activity. However, the molecular mechanism underlying Runx2 activation by FGF/FGFR or ERK was still unclear. In this study, we found that FGF2 treatment increased the protein level of exogenously overexpressed Runx2 and that this increase is reversed by ERK inhibitors. In contrast, overexpression of constitutively active MEK strongly increased the Runx2 protein level, which paralleled an increase in Runx2 acetylation. As Runx2 protein phosphorylation mediated by ERK directly correlates with Runx2 protein stabilization, acetylation, and ubiquitination, we undertook to identify the ERK-dependent phosphorylation sites in Runx2. Analysis of two C-terminal Runx2 deletion constructs showed that the middle third of the protein is responsible for ERK-induced stabilization and activation. An in silico analysis of highly conserved ERK targets indicated that there are three relevant serine residues in this domain. Site-directed mutagenesis implicated Ser-301 in for ERK-mediated Runx2 stabilization and acetylation. In conclusion, the FGF2-induced ERK MAP kinase strongly increased the Runx2 protein level through an increase in acetylation and a decrease in ubiquitination, and these processes require the phosphorylation of Runx2 Ser-301 residue.

FIGURE 1.

FGF2 increases the Runx2 protein and acetylation levels and transcription activity and MEK enhances Runx2 stability. a, MC3T3-E1 cells were pretreated for 1 h with 10 μm U0126 and then treated or not for 8 h with 10 ng/ml FGF2. Acetylated endogenous Runx2 was detected by immunoprecipitation with an anti-acetyl-lysine antibody followed by immunoblotting with an anti-Runx2 antibody. The endogenous Runx2 protein level was estimated by immunoblotting with the anti-Runx2 antibody. b, 293T cells were transfected with the 3xMyc-Runx2 and CA-MEK or empty vector. The transfected cells treated with 50 μl/ml cycloheximide were harvested. Total Runx2 level was estimated by immunoblotting with an anti-Myc antibody. c, 293T cells were transfected with the pCDNA3.1–3xMyc-Runx2 expression vector. 24 h post-transfection, cells were serum-starved for 16 h and then treated for 8 h with 10 ng/ml FGF2. Acetylated Runx2 was detected by immunoprecipitation with an anti-acetyl-lysine antibody followed by immunoblotting with an anti-Myc antibody. The Runx2 protein level was estimated by immunoblotting with the anti-Myc antibody. β-Actin was used as a loading control. d, C2C12 cells were transiently co-transfected with the 3xMyc-RUNX2 and pGL3–6XOSE-Luc vectors plus pRL-TK vectors. After 24 h, the cells were serum-starved for 16 h and then treated for 8 h with 10 ng/ml FGF2. Runx2 transcriptional activity was estimated by a luciferase assay and normalized to Renilla luciferase.

FIGURE 2.

The ERK MAP kinase pathway is required for FGF2-enhanced Runx2 stability and transcriptional activity. a, 293T cells were transfected with the pCDNA3.1–3xMyc-Runx2 expression vector. 24 h after transfection, the cells were serum-starved for 16 h, pretreated for 1 h with 10 μm U0126, 25 μm, SB203580, or 50 μm JNKII, and then treated or not for 8 h with 10 ng/ml FGF2. Acetylated Runx2 and total Runx2 were detected as in Fig. 1. b, C2C12 cells were transiently co-transfected with the 3xMyc-Runx2 and pGL3–6XOSE-Luc vectors plus pRL-TK vectors. After 24 h, the cells were serum-starved for 16 h, pretreated or not for 1 h with the indicated inhibitor, and then treated or not for 8 h with 10 ng/ml FGF2. Runx2 transcriptional activity was estimated with the luciferase assay.

FIGURE 3.

The region between amino acids 259 and 397 of Runx2 is responsible for stability and transcriptional activity stimulated by FGF2. a, diagrammatic representation of Runx2 constructs. RHD, runt homology domain; PST, proline/serine/threonine rich domain; NMTS, nuclear matrix targeting signal. b, 293T cells were transfected with wild-type Runx2, ΔPST1 Runx2, or ΔPST2 Runx2 expression vectors. FGF2 was treated as in Fig. 1. Immunoblot analysis was performed using anti-Myc and anti-β-actin antibodies. Runx2 proteins are marked with asterisks, and β-actin is marked with arrows. c, total cell lysates were immunoprecipitated with an anti-acetyl-lysine antibody followed by immunoblotting using the anti-Myc antibody. Acetylated Runx2 is marked with asterisks. d, C2C12 cells were co-transfected with the Runx2 expression vector and reporter vector plus pRL-TK vectors. After 24 h, cells were serum-starved for 16 h and then treated with 10 ng/ml FGF2 for 8 h.

FIGURE 4.

Ser-301 in Runx2 is the major FGF2-induced ERK-dependent phosphorylation site. a, structure of Runx2 with candidate phosphorylation sites. b, 293T cells were transfected with wild-type Runx2, 282A Runx2, 301A Runx2, or 319A Runx2 expression vectors. After 24 h, cells were serum-starved for 16 h and then treated or not with 10 ng/ml FGF2 for 8 h. Acetylated Runx2 and total Runx2 were detected as described in the legend to Fig. 1. c, C2C12 cells were co-transfected with wild-type Runx2 or a Runx2 mutant (282A, 301A, or 319A) and reporter vector plus pRL-TK vectors. After 24 h, the cells were serum-starved for 16 h and treated with FGF2 for 8 h. Runx2 transcriptional activity was estimated with the luciferase assay. d, 293T cells were co-transfected with CA-MEK and Runx2 constructs. After 24 h, cells were harvested, and total cell lysates were analyzed as in Fig. 4c. e, C2C12 cells were co-transfected with wild-type Runx2 or a Runx2 mutant (282A, 301A, or 319A) and a reporter vector with (solid bars) or without (blank bars) CA-MEK plus pRL-TK vectors. 24 h post-transfection, luciferase activities were determined.

FIGURE 5.

FGF2 inhibits Runx2 polyubiquitination. a, 293T cells were co-transfected with Flag-ubiquitin and Myc-Runx2 (wild type or the S301A mutant). 24 h after transfection, cells were treated with MG132 and FGF2 for 8 h. b, 293T cells were co-transfected with Flag-ubiquitin and Myc-Runx2 (wild type or the S301A) with or without CA-MEK. 24 h after transfection, cells was treated with MG132 for 8 h. Ubiquitination of RUNX2 was analyzed by immunoprecipitation of 1 mg of lysates with the anti-Myc antibody followed by immunoblotting with an anti-Flag antibody.