Voltar
Artigo Acesso aberto Revisado por pares
BIBTEX RIS

Fibroblast Growth Factor Inhibits Chondrocytic Growth through Induction of p21 and Subsequent Inactivation of Cyclin E-Cdk2

2001; Elsevier BV; Volume: 276; Issue: 31 Linguagem: Inglês

10.1074/jbc.m101859200

ISSN

1083-351X

Autores

Tomonao Aikawa, Gino V. Segre, Kaechoong Lee,

Tópico(s)

Ubiquitin and proteasome pathways

Resumo

Fibroblast growth factor (FGF) and its receptor (FGFR) are thought to be negative regulators of chondrocytic growth, as exemplified by achondroplasia and related chondrodysplasias, which are caused by constitutively active mutations in FGFR3. To understand the growth-inhibitory mechanisms of FGF, we analyzed the effects of FGF2 on cell cycle-regulating molecules in chondrocytes. FGF2 dramatically inhibited proliferation of rat chondrosarcoma (RCS) cells and arrested their cell cycle at the G1 phase. FGF2 increased p21 expression in RCS cells, which assembled with the cyclin E-Cdk2 complexes, although the expression of neither cyclin E nor Cdk2 increased. In addition, the kinase activity of immunoprecipitated cyclin E or Cdk2, assessed with retinoblastoma protein (pRb) as substrate, was dramatically reduced by FGF-2. Moreover, FGF2 shifted pRb to its underphosphorylated, active form in RCS cells. FGF2 not only induced p21 protein expression in proliferating chondrocytes in mouse fetal limbs cultured in vitro but also decreased their proliferation as assessed by the expression of histone H4 mRNA, a marker for cells in S phase. Furthermore, inhibitory effects of FGF2 on chondrocytic proliferation were partially reduced in p21-null limbs, compared with those in wild-type limbs in vitro. Taken together, FGF's growth inhibitory effects of chondrocytes appear to be mediated at least partially through p21 induction and the subsequent inactivation of cyclin E-Cdk2 and activation of pRb. Fibroblast growth factor (FGF) and its receptor (FGFR) are thought to be negative regulators of chondrocytic growth, as exemplified by achondroplasia and related chondrodysplasias, which are caused by constitutively active mutations in FGFR3. To understand the growth-inhibitory mechanisms of FGF, we analyzed the effects of FGF2 on cell cycle-regulating molecules in chondrocytes. FGF2 dramatically inhibited proliferation of rat chondrosarcoma (RCS) cells and arrested their cell cycle at the G1 phase. FGF2 increased p21 expression in RCS cells, which assembled with the cyclin E-Cdk2 complexes, although the expression of neither cyclin E nor Cdk2 increased. In addition, the kinase activity of immunoprecipitated cyclin E or Cdk2, assessed with retinoblastoma protein (pRb) as substrate, was dramatically reduced by FGF-2. Moreover, FGF2 shifted pRb to its underphosphorylated, active form in RCS cells. FGF2 not only induced p21 protein expression in proliferating chondrocytes in mouse fetal limbs cultured in vitro but also decreased their proliferation as assessed by the expression of histone H4 mRNA, a marker for cells in S phase. Furthermore, inhibitory effects of FGF2 on chondrocytic proliferation were partially reduced in p21-null limbs, compared with those in wild-type limbs in vitro. Taken together, FGF's growth inhibitory effects of chondrocytes appear to be mediated at least partially through p21 induction and the subsequent inactivation of cyclin E-Cdk2 and activation of pRb. fibroblast growth factor fibroblast growth factor receptor signal transducers and activators of transcription cyclin-dependent kinase CDK inhibitor rat chondrosarcoma wild type FGFs1 are a large family of at least 23 related polypeptides that bind to and activate a family of four tyrosine kinase receptors, FGFRs. They play important roles in regulating proliferation and differentiation of various types of cells, including those involved in limb development and long bone formation (1Szebenyi G. Fallon J.F. Int. Rev. Cytol. 1999; 185: 45-106Crossref PubMed Google Scholar, 2Xu X. Weinstein M. Li C. Deng C.X. Cell Tissue Res. 1999; 296: 33-43Crossref PubMed Scopus (127) Google Scholar). Long bones form by endochondral ossification, which is characterized by mesenchymal condensation, chondrogenic differentiation, chondrocytic proliferation, synthesis of cartilage matrix, hypertrophic differentiation, and replacement by bone. These sequential growth and differentiation processes are regulated by numerous growth factors and their receptors, such as parathyroid hormone-related protein, Indian hedgehog, and insulin-like growth factor-I (3Karp S.J. Schipani E. St-Jacques B. Hunzelman J. Kronenberg H. McMahon A.P. Development. 2000; 127: 543-548Crossref PubMed Google Scholar, 4Liu J.P. Baker J. Perkins A.S. Robertson E.J. Efstratiadis A. Cell. 1993; 75: 59-72Abstract Full Text PDF PubMed Scopus (2533) Google Scholar). Recently, achondroplasia, thanatophoric dysplasia, and hypochondroplasia have been shown to be caused by constitutively active mutations in the FGFR3 gene (5Webster M.K. Donoghue D.J. Trends Genet. 1997; 13: 178-182Abstract Full Text PDF PubMed Scopus (269) Google Scholar). Also, FGFR3-deficient mice displayed overgrowth of long bones (6Deng C. Wynshaw-Boris A. Zhou F. Kuo A. Leder P. Cell. 1996; 84: 911-921Abstract Full Text Full Text PDF PubMed Scopus (902) Google Scholar, 7Colvin J.S. Bohne B.A. Harding G.W. McEwen D.G. Ornitz D.M. Nat. Genet. 1996; 12: 390-397Crossref PubMed Scopus (742) Google Scholar), and mice carrying dominant active FGFR3 genes exhibited dwarfism similar to that in patients with achondroplasia and thanatophoric dysplasia (8Wang Y. Spatz M.K. Kannan K. Hayk H. Avivi A. Gorivodsky M. Pines M. Yayon A. Lonai P. Givol D. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 4455-4460Crossref PubMed Scopus (191) Google Scholar, 9Li C. Chen L. Iwata T. Kitagawa M. Fu X.Y. Deng C.X. Hum. Mol. Genet. 1999; 8: 35-44Crossref PubMed Scopus (192) Google Scholar, 10Iwata T. Chen L. Li C. Ovchinnikov D.A. Behringer R.R. Francomano C.A. Deng C.X. Hum. Mol. Genet. 2000; 9: 1603-1613Crossref PubMed Scopus (175) Google Scholar, 11Chen L. Adar R. Yang X. Monsonego E.O. Li C. Hauschka P.V. Yayon A. Deng C.X. J. Clin. Invest. 1999; 104: 1517-1525Crossref PubMed Scopus (210) Google Scholar, 12Naski M.C. Colvin J.S. Coffin J.D. Ornitz D.M. Development. 1998; 125: 4977-4988Crossref PubMed Google Scholar, 13Segev O. Chumakov I. Nevo Z. Givol D. Madar-Shapiro L. Sheinin Y. Weinreb M. Yayon A. Hum. Mol. Genet. 2000; 9: 249-258Crossref PubMed Scopus (89) Google Scholar). Moreover, targeted overexpression of FGF9 in cartilage results in dwarfism in mice similar to the dwarfism in achondroplasia (14Garofalo S. Kliger-Spatz M. Cooke J.L. Wolstin O. Lunstrum G.P. Moshkovitz S.M. Horton W.A. Yayon A. J. Bone Miner. Res. 1999; 14: 1909-1915Crossref PubMed Scopus (80) Google Scholar). Although these findings support the hypothesis that FGF and its receptors are negative regulators of endochondral bone development, little is known about the mechanisms by which FGF inhibits chondrocytic growth. Sahni et al. recently reported that FGF inhibited chondrocytic growth through activating the signal transducer and activator of transcription 1 (STAT1) pathway and suggested that p21 might be a factor responsible for inhibiting chondrocytic proliferation (15Sahni M. Ambrosetti D.C. Mansukhani A. Gertner R. Levy D. Basilico C. Genes Dev. 1999; 13: 1361-1366Crossref PubMed Scopus (317) Google Scholar). Cell cycle progression is known to be governed by the assembly of cyclins, cyclin-dependent kinases (CDKs), and CDK inhibitors (CKIs) and subsequent regulation of retinoblastoma protein (pRb) activity. In the early G1 phase of the cell cycle, for example, cyclin D-Cdk4 phosphorylates pRb, and in the late G1 phase cyclin E-Cdk2/cyclin D-Cdk4 phosphorylates pRb. Hyperphosphorylated pRb then releases E2F family transcription factors, which activate genes that are necessary for S phase entry (16Dyson N. Genes Dev. 1998; 12: 2245-2262Crossref PubMed Scopus (1954) Google Scholar). The CKI p21cip1, waf-1 belongs to a Cip/Kip family with p27kip1 and p57kip2 and broadly inhibits the activity of cyclin D-, E-, and A-dependent kinases (17Sherr C.J. Roberts J.M. Genes Dev. 1999; 13: 1501-1512Crossref PubMed Scopus (5071) Google Scholar, 18Vidal A. Koff A. Gene (Amst.). 2000; 247: 1-15Crossref PubMed Scopus (371) Google Scholar). In contrast, the second class of CKIs (i.e. Ink4 (inhibitors for Cdk4) proteins (p16INK4a, p15INK4b, p18INK4c, and p19INK4d), bind to and inactivate only Cdk4 and Cdk6 but not other CDKs or D-type cyclins (17Sherr C.J. Roberts J.M. Genes Dev. 1999; 13: 1501-1512Crossref PubMed Scopus (5071) Google Scholar, 18Vidal A. Koff A. Gene (Amst.). 2000; 247: 1-15Crossref PubMed Scopus (371) Google Scholar). p21 has been shown to be expressed in postmitotic, hypertrophic chondrocytes but not in chondrocytes in the proliferating zone (19Parker S.B. Eichele G. Zhang P. Rawls A. Sands A.T. Bradley A. Olson E.N. Harper J.W. Elledge S.J. Science. 1995; 267: 1024-1027Crossref PubMed Scopus (1016) Google Scholar), and activation of STAT1 and increased expression of p21 in cartilage have been demonstrated in thanatophoric dysplasia human embryos (20Su W.C. Kitagawa M. Xue N. Xie B. Garofalo S. Cho J. Deng C. Horton W.A. Fu X.Y. Nature. 1997; 386: 288-292Crossref PubMed Scopus (268) Google Scholar). Based on these data, p21 has been suggested to be involved in FGF's growth inhibition of cartilage. Interestingly, in some breast cancer cells, FGF has been shown to inhibit their proliferation through activation of STAT1 and induction of p21 expression (21Wang H. Rubin M. Fenig E. DeBlasio A. Mendelsohn J. Yahalom J. Wieder R. Cancer Res. 1997; 57: 1750-1757PubMed Google Scholar, 22Johnson M.R. Valentine C. Basilico C. Mansukhani A. Oncogene. 1998; 16: 2647-2656Crossref PubMed Scopus (53) Google Scholar). In this study, we first examine the effects of FGF2 on cell cycle progression and p21 expression in rat chondrosarcoma (RCS) cells, whose proliferation is potently inhibited by FGF2. Second, we analyze how p21 induction by FGF2 leads to cell cycle arrest in RCS cells by examining the binding of FGF2-induced p21 to cyclins and CDKs as well as by assessing the changes in CDK- and cyclin-dependent kinase activities after FGF2 treatment. Finally, we examine the functional involvement of p21 in FGF's growth inhibition by comparing the effects of FGF2 on cartilaginous growth in vitro in limbs isolated from p21-deficient and wild-type mice. RCS cells were grown in high glucose Dulbecco's modified Eagle's medium (Life Technologies, Inc.) supplemented with 10% fetal bovine serum (HyClone), 100 units/ml penicillin, and 100 µg/ml streptomycin (23Mukhopadhyay K. Lefebvre V. Zhou G. Garofalo S. Kimura J.H. de Crombrugghe B. J. Biol. Chem. 1995; 270: 27711-27719Abstract Full Text Full Text PDF PubMed Scopus (129) Google Scholar). Cells were passaged using trypsin-EDTA (Life Technologies) and were split at a 1:15 ratio on a 100-mm dish (Falcon; Becton and Dickinson). Cell proliferation was assessed by [3H]thymidine incorporation assay as well as by measuring DNA contents. RCS cells were plated at a density of 3 × 104 cells/24-well plate (Falcon), and growing cells (70% confluent) were treated with vehicle (0.5 m NaCl, 0.1% bovine serum albumin/phosphate-buffered saline) or the indicated dose of FGF2 (R&D Systems) for 24 h. Cells were labeled with 0.5 µCi/ml [3H]thymidine for the last 2 h of culture and dissolved with 0.1 n NaOH, 0.01% bovine serum albumin, 1% Triton X-100 (Sigma). Samples were mixed with the same volume of ice-cold 10% trichloroacetic acid for 10 min at 4 °C and then centrifuged at 14,000 rpm at 4 °C for 10 min. After washing with 5% trichloroacetic acid, pellets were dissolved in 0.1 n NaOH, 1% Triton X-100, and their radioactivity was measured by a scintillation counter (Beckman). DNA content was measured by a fluorometric method. Briefly, RCS cells were plated at a density of 1 × 104 cells/24-well plate and treated with vehicle or indicated doses of FGF2. Cells were digested with 20 µg/ml of papain (Sigma) in 0.1 m sodium acetate, 50 mm EDTA, 10 mm cysteine-HCl (pH 5.53) at 55 °C for 16 h. The DNA content in digested lysates were measured using PicoGreen DNA quantitation kit (Molecular Probes, Inc., Eugene, OR). RCS cells were plated at a density of 8 × 105 cells/60-mm culture dishes, and subconfluent cells were treated with 3 nm of FGF2 for 6, 12, and 24 h. Cells were then trypsinized, washed twice with phosphate-buffered saline, fixed in 70% ethanol, and stained with propidium iodide (50 µg/ml; Sigma), RNase A (200 µg/ml; Sigma), 0.1% glucose in phosphate-buffered saline for 4 h at room temperature. Samples were analyzed on an Epics Alta flow cytometer (Beckman Coulter). For Western blotting, cells were lysed in 2× SDS buffer containing 125 mm Tris-HCl (pH 6.8), 4% SDS containing phosphatase inhibitor mixtures I and II (Sigma), 2 mmphenylmethylsulfonyl fluoride (Sigma), and 132 µg/ml aprotinin (Sigma). Protein concentration of the lysates was determined using protein assay kit DC (Bio-Rad). Samples were then diluted by adding the same volume of 2× dye containing 20% glycerol, 4% 2-mercaptoethanol (Sigma), and equal amounts of protein were resolved on 6.5% SDS gel for pRb or 12.5% gel for all other proteins. Proteins were then transferred onto ImmobilonTM-P membranes (Millipore Corp.). After being washed twice with 0.1% Tween 20, 10 mmTris-HCl, 150 mm NaCl (pH 8.0) (T-TBS), blots were incubated with blocking buffer (5% nonfat dry milk in T-TBS) for 1 h at room temperature and then incubated with primary antibodies diluted in blocking buffer for 1 h at room temperature or overnight at 4 °C. Antibodies used for Western blotting were diluted as follows: anti-p21 (F-5, 1:750), anti-p27 (F-8, 1:200), anti-cyclin D2 (M-20, 1:2000), anti-cyclin A (C-19, 1:2000), anti-cyclin E (M-20, 1:2000), anti-Cdk2 (M2, 1:2000), anti-Cdk4 (C-22, 1:2000) (all from Santa Cruz Biotechnology, Inc., Santa Cruz, CA), and anti-pRb (clone G3–245, 1:750, Pharmingen). Secondary antibodies used were horseradish peroxidase-conjugated goat anti-rabbit IgG and goat anti-mouse IgG (Santa Cruz Biotechnology) diluted in blocking buffer at 1:10,000. Western blots were developed with chemiluminescence reagent (PerkinElmer Life Sciences). In some experiments, blots were incubated with 62.5 mm Tris-HCl (pH 6.8), 2% SDS, 100 mm 2-mercaptoethanol at 50 °C for 30 min and reprobed with other primary antibodies. RCS cells growing on 100-mm dishes were treated with vehicle or 3 nm FGF2 for 12 h. Cells were then lysed in 50 mm Tris-HCl (pH 7.4), 150 mm NaCl, 0.1% Triton X-100 containing 10 µm β-glycerophosphate, phosphatase inhibitor mixtures I and II, 1 mm phenylmethylsulfonyl fluoride, and 66 µg/ml of aprotinin for 30 min on ice. Lysates were precleared with 0.5 µg/ml nonimmune rabbit IgG (Santa Cruz Biotechnology) for 30 min at 4 °C and then with 40 µl/ml of protein A-agarose (Life Technologies) for 30 min at 4 °C with agitation. Lysates were then cleared by centrifugation, and 500 µg of protein of the supernatant was incubated with 2.0 µg of rabbit polyclonal antibodies to p21 (C-19), cyclin D2 (M-20), Cdk2 (M2), cyclin E (M-20 and Upstate Biotechnology), or mouse monoclonal antibody to Cdk4 (clone DCS-35; Pharmingen) for 1 h at 4 °C and then with 40 µl of protein G-agarose (Santa Cruz Biotechnology) for Cdk4 or protein A-agarose (Life Technologies) for all other antibodies overnight at 4 °C with agitation. Samples were then washed four times with lysis buffer, dissolved in sample buffer, and resolved on 12.5% SDS gel, followed by Western blotting. For CDK kinase assay, immunoprecipitates were washed four times with lysis buffer and then twice with kinase buffer (50 mmHepes, 10 mm MgCl2, 1 mmdithiothreitol). The kinase reaction was carried out in 30 µl of kinase buffer containing 2 µg of GST-fusioned carboxyl-terminal pRb (New England Biolabs), 30 µm ATP, and 10 µCi of [γ-32P]ATP (6000 Ci/mmol; PerkinElmer Life Sciences) for 30 min at 30 °C. Reactions were terminated by boiling after the addition of the same volume of 2× SDS sample buffer. Samples were resolved on 10% SDS gel and analyzed using a CycloneTMphosphor imager (Packard). Animals were maintained in accordance with the NIH Guide for the Care and Use of Laboratory Animals. Mutant mice deficient for p21 (B6;129-Cdkn1atm1tyj) (24Brugarolas J. Chandrasekaran C. Gordon J.I. Beach D. Jacks T. Hannon G.J. Nature. 1995; 377: 552-557Crossref PubMed Scopus (1136) Google Scholar) and wild-type (WT) mice (B6129SF2/J) were purchased from Jackson Laboratory. Methods for mouse fetal limb culture were described previously (25Vortkamp A. Lee K. Lanske B. Segre G.V. Kronenberg H.M. Tabin C.J. Science. 1996; 273: 613-622Crossref PubMed Scopus (1615) Google Scholar). In brief, limbs of embryonic day 16.5 fetuses were stripped of skin and soft tissue and placed on a filter paper (type AA, 0.8-µm pore size; Millipore) supported by stainless steel grids (Wire Mesh Corp.). Limbs were cultured with daily replacement of BGJb medium (Life Technologies, Inc.) containing 0.1% bovine serum albumin, 100 units/ml penicillin, 100 µg/ml streptomycin, and 250 ng/ml amphotericin B. After the termination of cultures, samples were fixed in 10% formalin/phosphate-buffered saline. 35S-Labeled riboprobes for histone H4 were made using genomic DNA (1.2-kilobaseEcoRI/XbaI fragment of the plasmid pF0108A) (26Sierra F. Stein G. Stein J. Nucleic Acids Res. 1983; 11: 7069-7086Crossref PubMed Scopus (77) Google Scholar) obtained from Dr. Gary Stein (University of Massachusetts School of Medicine, Worcester, MA). Antisense [35S]cRNAs were synthesized from linearized plasmids using the Gemini Transcription Kit (Promega, Madison, WI) and [35S]UTP (1289 Ci/mmol; PerkinElmer Life Sciences). In situ hybridization was performed as described previously (27Lee K. Lanske B. Karaplis A.C. Deeds J.D. Kohno H. Nissenson R.A. Kronenberg H.M. Segre G.V. Endocrinology. 1996; 137: 5109-5118Crossref PubMed Scopus (235) Google Scholar). Slides were dipped into NTB-2 (Eastman Kodak Co.) and stored at 4 °C for 7 days. After development, sections were counterstained with hematoxylin and eosin and mounted. After sections (6 µm) were deparaffinized and hydrated, antigen was unmasked by boiling twice in 10 mm citrate buffer (pH 6.0) for 5 min in a microwave oven. Immunohistochemistry was performed using anti-p21 mouse IgG (F-5, 1:50) and mouse ABC staining kit (Santa Cruz Biotechnology). All of the experiments were repeated at least three times. RCS cells display numerous chondrocytic characteristics, including expression of type II and type IX collagens and production of cartilaginous matrix (23Mukhopadhyay K. Lefebvre V. Zhou G. Garofalo S. Kimura J.H. de Crombrugghe B. J. Biol. Chem. 1995; 270: 27711-27719Abstract Full Text Full Text PDF PubMed Scopus (129) Google Scholar). In preliminary studies, we confirmed that RCS cells expressed FGFR2 and R3, which were phosphorylated at tyrosine residues upon treatment with FGF2 (data not shown). First, we assessed the effects of FGF2 on cellular proliferation by measuring [3H]thymidine incorporation as well as DNA content. As shown in Fig.1 A, FGF2 suppressed thymidine incorporation in a dose-dependent manner from 10 to 1000 pm. Growth inhibition by FGF2 was confirmed by dose-dependent decreases of DNA content in RCS cells continuously treated with 3–100 pm FGF2 (Fig.1 B). Then we analyzed the effects of FGF2 on cell cycle distribution of RCS cells. As shown in Fig. 2, vehicle-treated cells showed an almost identical distribution of cells across the cycle throughout the time course studied (G1, 56.9–63.4%; S, 16.3–17.8%; G2/M, 16.9–19.7%) except for a slight increase in the G1 population at 24 h. FGF2-treated cells first accumulated in G2/M phase (43.17%), along with a reduction in cells in S phase (5.79%) at 6 h. Then FGF2-treated cells accumulated in G1 phase at 12 and 24 h (73.90 and 77.94%, respectively), which was associated with a further reduction of cells in S phase (1.17% at 24 h). The G2/M population of FGF2-treated cells decreased dramatically between 6 and 12 h and remained at this level at 24 h. The distribution pattern of FGF2-treated cells at 24 h remained unchanged at least up to 36 h (data not shown). To understand the mechanisms of FGF-induced G1 arrest of RCS cells, we examined the effects of FGF2 on expression of G1/S cell cycle proteins. As shown in Fig.3, FGF2 induced p21 protein expression by 6 h, and its expression peaked at 12 h. In contrast, p57 was undetectable either in the presence or absence of FGF2 (data not shown). Also, whereas p27 expression was detectable at 12 and 24 h, FGF2 did not affect the level of its expression. We then examined the phosphorylation state of pRb by Western blotting, using an antibody to pRb that recognizes both hyperphosphorylated (ppRb in Fig. 3) and underphosphorylated pRb (pRb in Fig. 3). FGF2 shifted pRb from hyperphosphorylated, inactive forms to underphosphorylated, active forms by 6 h. By 12 h and thereafter, the majority of pRb was in the underphosphorylated, active form in FGF2-treated cells. This paralleled the increased p21 protein expression as well as the increase of cells in G1. A partial shift of pRb to underphosphorylated forms was observed in vehicle-treated cells also at 12 and 24 h, which might be related to the slight increase in G1 cells in vehicle-treated RCS at 24 h. Parallel to the shift of pRb to its active form, expression of cyclin A, whose expression is regulated by E2F, was dramatically reduced at 12 and 24 h. In contrast, whereas cyclin D2 expression was dramatically increased by FGF2, cyclin E expression was unaffected during the time studied. Furthermore, FGF2 did not change the expression levels of Cdk2 or Cdk4. Next, we examined which cyclin-CDK complexes were associated with FGF-induced p21. We treated RCS cells with vehicle or FGF2 for 12 h and then immunoprecipitated with antibodies to p21, cyclin D2, cyclin E, Cdk2, and Cdk4, followed by Western blotting with antibodies to p21, p27, cyclin D2, Cdk2, and Cdk4. As shown in Fig.4 A, FGF2-induced p21 coimmunoprecipitated with cyclin D2, cyclin E, Cdk2, and Cdk4, whereas p21 expressed in vehicle-treated cells was associated with them at much lower levels. Although cyclin D2 was associated with Cdk4 in vehicle-treated cells, it became associated with both Cdk4 and Cdk2 after FGF2 treatment. Also, whereas the association between cyclin E and Cdk2 was unaffected by FGF2 treatment, the cyclin E-Cdk2-p21 complex assembled only after FGF2 treatment. In summary, these data suggest that FGF2 treatment generates a cyclin E-Cdk2-p21 complex and quantitatively increases cyclin D2-Cdk4-p21 and cyclin D2-Cdk2-p21 complexes. FGF2 did not change the expression of p27 or its association with cyclin D2, cyclin E, or Cdk2. We then examined which cyclin-dependent kinase activity is regulated by FGF2. To assess the kinase activity of CDKs, cell lysates were first immunoprecipitated with antibodies to cyclin D2, cyclin E, Cdk2, or Cdk4, and precipitated proteins were then incubated with [32P]ATP and GST-Rb protein. As shown in Fig.4 B, 12 h of FGF treatment did not change the activity of Cdk4 or cyclin D2-dependent kinase but dramatically reduced the activity of Cdk2 and cyclin E-dependent kinases, as assessed by the phosphorylation of exogenous pRb protein. Reduction in the activity of Cdk2 and cyclin E-dependent kinases was also observed using histone H1 as substrate (data not shown). Taken together, these data suggest that FGF2 inhibited G1/S transition of cell cycle through the induction of p21 and subsequent inactivation of cyclin E-Cdk2 in RCS cells. We then examined whether FGF2 induces p21 in chondrocytes in the growth plate in limbs isolated from mouse fetuses. As shown in Fig.5, p21 protein expression was limited to hypertrophic chondrocytes and some proliferating chondrocytes residing in the peripheral region of the lower half of the proliferating zone in vehicle-treated limbs (Fig. 5 a). In contrast, after 8-h treatment with FGF2, p21 immunoreactivity was seen in almost all of the chondrocytes (Fig. 5, b andc). This widespread pattern of p21 immunoreactivity in growth plate chondrocytes was observed also at 16 and 24 h after FGF2 treatment (data not shown) Next, we examined whether inhibitory effects of FGF2 on chondrocytic proliferation differ in the growth plate of p21 −/− mice, compared with WT mice, by comparing mRNA expression for H4 histone (H4), a marker for cells in S phase (27Lee K. Lanske B. Karaplis A.C. Deeds J.D. Kohno H. Nissenson R.A. Kronenberg H.M. Segre G.V. Endocrinology. 1996; 137: 5109-5118Crossref PubMed Scopus (235) Google Scholar). To avoid the interference caused by developmental variability among littermates, we treated one hind limb with vehicle alone and the contralateral hind limb with the indicated amounts of FGF2. As shown in Fig.6, H4 expression was detected in cells in the proliferating zone but not in the hypertrophic zone in vehicle-treated tibias from WT and p21 −/− mice. In both WT and p21 −/− tibias, FGF2 dose-dependently decreased the number of H4-positive cells in the proliferating zone. However, whereas 1.5 nm FGF2 dramatically decreased the number of H4-positive chondrocytes in the proliferating zone in WT limbs, the number of H4-positive chondrocytes in the proliferating zone in p21 −/− limbs was unaffected. Moreover, whereas 6 and 24 nm FGF2 completely suppressed the chondrocytic proliferation in WT tibias, as evidenced by the absence of H4 mRNA expression, chondrocytic proliferation was only partially inhibited in p21 −/− tibias at these doses of FGF2. As shown in Fig. 7, when tibias were treated with 1.5 nm FGF2 or vehicle alone for 2 days, FGF2 dramatically reduced both the overall size of the growth plate and the thickness of the hypertrophic zone in WT tibia, compared with the vehicle-treated control. In contrast, neither the growth plate size nor the thickness of the hypertrophic zone in p21 −/− tibia was affected as dramatically as WT tibia. 2 days of 24 nm FGF2 treatment decreased both of these in p21 −/− tibias as it did in WT (data not shown), consistent with the effects on H4 expression. Although FGF and its receptor are known to negatively regulate cartilage proliferation, little is known about the mechanisms involved. In this study, we examined the effects of FGF2 on the cell cycle inhibitor, p21, in both cell and fetal limb cultures to understand the growth-inhibitory pathway of FGF signaling in chondrocytes. Unlike the effects of FGF2 on primary chondrocytes and ATDC5 cells, FGF2 inhibits growth of RCS cells, thus mimicking its effects in vivo. In this study, we took advantage of the antiproliferative response of RCS cells to analyze the growth-inhibitory mechanisms of FGF in chondrocytes. FGF2 arrests the cell cycle of RCS cells mainly at G1, a stage known to be controlled by pRb activity (28Weinberg R.A. Cell. 1995; 81: 323-330Abstract Full Text PDF PubMed Scopus (4275) Google Scholar). pRb is underphosphorylated and exerts its antiproliferative function in G1, and hyperphosphorylation of pRb by cyclin D-Cdk4 and cyclin E-Cdk2 complexes inactivates it at the G1/S transition, allowing the cell to proceed into S phase (29Hinds P.W. Mittnacht S. Dulic V. Arnold A. Reed S.I. Weinberg R.A. Cell. 1992; 70: 993-1006Abstract Full Text PDF PubMed Scopus (867) Google Scholar, 30Akiyama T. Ohuchi T. Sumida S. Matsumoto K. Toyoshima K. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 7900-7904Crossref PubMed Scopus (193) Google Scholar, 31Kato J. Matsushime H. Hiebert S.W. Ewen M.E. Sherr C.J. Genes Dev. 1993; 7: 331-342Crossref PubMed Scopus (1079) Google Scholar, 32Hatakeyama M. Brill J.A. Fink G.R. Weinberg R.A. Genes Dev. 1994; 8: 1759-1771Crossref PubMed Scopus (221) Google Scholar). Several lines of evidence indicate that the cell cycle arrest at G1is linked to the decreased kinase activity of cyclin D-Cdk4 or cyclin E-Cdk2 complexes as well as to the underphosphorylation of pRb (32Hatakeyama M. Brill J.A. Fink G.R. Weinberg R.A. Genes Dev. 1994; 8: 1759-1771Crossref PubMed Scopus (221) Google Scholar, 33Polyak K. Kato J.Y. Solomon M.J. Sherr C.J. Massague J. Roberts J.M. Koff A. Genes Dev. 1994; 8: 9-22Crossref PubMed Scopus (1813) Google Scholar, 34Gius D.R. Ezhevsky S.A. Becker-Hapak M. Nagahara H. Wei M.C. Dowdy S.F. Cancer Res. 1999; 59: 2577-2580PubMed Google Scholar, 35Cover C.M. Hsieh S.J. Tran S.H. Hallden G. Kim G.S. Bjeldanes L.F. Firestone G.L. J. Biol. Chem. 1998; 273: 3838-3847Abstract Full Text Full Text PDF PubMed Scopus (244) Google Scholar, 36Toyoshima H. Hunter T. Cell. 1994; 78: 67-74Abstract Full Text PDF PubMed Scopus (1925) Google Scholar). Consistent with these, the time course of G1accumulation and exclusion from S phase in RCS cells after FGF2 treatment was synchronized to the underphosphorylation of pRb and decrease in cyclin A, whose expression is regulated by E2F family members through inactivation (hyperphosphorylation) of pRb (16Dyson N. Genes Dev. 1998; 12: 2245-2262Crossref PubMed Scopus (1954) Google Scholar). Moreover, all of these events (i.e. G1accumulation, underphosphorylation of pRb, and decrease in cyclin A) were associated with p21 expression but not with p27, cyclin D2, cyclin E, Cdk2, and Cdk4 (Figs. 2 and 3). On the other hand, a slight increase in G1 population and partial underphosphorylation of pRb also was observed in vehicle-treated RCS cells. These events were associated with the expression of p21 and p27 at 12 and 24 h (Fig. 3). Since RCS cells grow very rapidly, with the doubling time shorter than 16 h, and their growth is contact-inhibited, and because we started the experiment when cells were 70% confluent, it is highly likely that the apparent G1 arrest and the associated increase in p21 and p27 expression at later time points were caused by contact inhibition. The involvement of p21 in G1 cell cycle arrest of RCS cells was further examined by coimmunoprecipitation and cyclin-dependent kinase assay. We found that FGF2 induced the assembly of cyclin E-Cdk2-p21, cyclin D2-Cdk4-p21, and cyclin D2-Cdk2-p21 complexes. Also, we demonstrated that FGF2 dramatically reduced the activity of Cdk2 and cyclin E-dependent kinases. Although FGF2 increased the expression of cyclin D2 in RCS cells, cyclin D2-dependent kinase activity was not affected by FGF2. Our findings with limb explants show that the increase in p21 expression induced by FGF2 in vitro also is evident in whole growth plates. Moreover, higher doses of FGF2 are required to inhibit cartilaginous growth in limbs from p21 −/− mice compared with those from WT mice. This further highlights the importance of p21 in the mechanisms by which FGF2 mediates inhibition of cartilage growth. In contrast to our data, Li et al. (9Li C. Chen L. Iwata T. Kitagawa M. Fu X.Y. Deng C.X. Hum. Mol. Genet. 1999; 8: 35-44Crossref PubMed Scopus (192) Google Scholar) have reported that Ink4 family proteins but not p21 were activated in postnatal growth plates of mice carrying the mutant FGFR3 gene (K644E). We did not examine Ink4 protein expression in this study, because FGF2 did not change the activity of Cdk4 or cyclin D2-dependent kinase in RCS cells. The apparent discrepancy between our in vitrodata and their in vivo data may be explained by a possible qualitative difference in signaling pathways in ligand-induced activation of WT FGFR and ligand-independent activation of mutant FGFR3 (K644E), or FGF's cell cycle regulation may involve different cell cycle inhibitors in fetal and postnatal growth plates. In summary, our data indicate that FGF's inhibitory effects on chondrocytic proliferation appear to be mediated at least partially through p21 induction and the subsequent inactivation of cyclin E-Cdk2 and activation of pRb. We thank Dr. Henry M. Kronenberg for critical reading of the manuscript and Dr. Toshihisa Kawai and Dr. Jean W. Eastcott for technical assistance with flow cytometry analysis and helpful discussion.

Referência(s)