Voltar
Artigo Acesso aberto Revisado por pares
BIBTEX RIS

The Role of Cyclin D3-dependent Kinase in the Phosphorylation of p130 in Mouse BALB/c 3T3 Fibroblasts

1998; Elsevier BV; Volume: 273; Issue: 11 Linguagem: Inglês

10.1074/jbc.273.11.6190

ISSN

1083-351X

Autores

Feng Dong, W. Douglas Cress, Deepak Agrawal, W. J. Pledger,

Tópico(s)

Epigenetics and DNA Methylation

Resumo

We have observed that cyclin D3-dependent kinase activity is increased in the late G1 phase in BALB/c 3T3 fibroblasts. The profile of cyclin D3-associated activity closely parallels that of cyclin D1, which is also induced after mitogenic stimulation of quiescent cells. These activities correlate with the appearance of hyperphosphorylated p130, an Rb family member important in regulating E2F-4 and E2F-5 activity in fibroblastic cells. We demonstrated, however, that only the cyclin D3 activity efficiently phosphorylated p130 in an in vitro kinase assay. This apparent specificity was further demonstrated by experiments which demonstrated that cyclin D3 was physically associated with p130 at the times when D3-dependent kinase activity and p130 hyperphosphorylation were observed. Examination of E2F by electrophoretic mobility shift assay revealed that E2F-4 DNA binding activity existed in a p130·E2F complex at times before D3-dependent kinase activity was apparent and in a free E2F-4 complex after D3 activity developed. Thus, our data suggest that cyclin D3 preferentially phosphorylates p130 and is thereby specifically targeted to overcoming growth-suppressive control mediated through p130 pathways. We have observed that cyclin D3-dependent kinase activity is increased in the late G1 phase in BALB/c 3T3 fibroblasts. The profile of cyclin D3-associated activity closely parallels that of cyclin D1, which is also induced after mitogenic stimulation of quiescent cells. These activities correlate with the appearance of hyperphosphorylated p130, an Rb family member important in regulating E2F-4 and E2F-5 activity in fibroblastic cells. We demonstrated, however, that only the cyclin D3 activity efficiently phosphorylated p130 in an in vitro kinase assay. This apparent specificity was further demonstrated by experiments which demonstrated that cyclin D3 was physically associated with p130 at the times when D3-dependent kinase activity and p130 hyperphosphorylation were observed. Examination of E2F by electrophoretic mobility shift assay revealed that E2F-4 DNA binding activity existed in a p130·E2F complex at times before D3-dependent kinase activity was apparent and in a free E2F-4 complex after D3 activity developed. Thus, our data suggest that cyclin D3 preferentially phosphorylates p130 and is thereby specifically targeted to overcoming growth-suppressive control mediated through p130 pathways. Three D-type cyclins (D1, D2, and D3) identified in the context of the delayed early response to growth factor stimulation are differentially and combinationally expressed in mammalian cells (1Sherr C.J. Matsushime H. Roussel M.F. CIBA Found. Symp. 1992; 170: 209-226PubMed Google Scholar, 2Sherr C.J. Trends Biochem. Sci. 1995; 20: 187-190Abstract Full Text PDF PubMed Scopus (883) Google Scholar, 3Matsushime H. Roussel M.F. Sherr C.J. Cold Spring Harbor Symp. Quant. Biol. 1991; 56: 69-74Crossref PubMed Scopus (66) Google Scholar). Although Cdk4 appears to be their most prominent catalytic partner in macrophages and fibroblasts (4Matsushime H. Ewen M.E. Strom D.K. Kato J.Y. Hanks S.K. Roussel M.F. Sherr C.J. Cell. 1992; 71: 323-334Abstract Full Text PDF PubMed Scopus (780) Google Scholar, 5Ewen M.E. Sluss H.K. Sherr C.J. Matsushime H. Kato J. Livingston D.M. Cell. 1993; 73: 487-497Abstract Full Text PDF PubMed Scopus (916) Google Scholar), D-cyclins can also form complexes with Cdk2, -4, -5, and -6 (6Matsushime H. Quelle D.E. Shurtleff S.A. Shibuya M. Sherr C.J. Kato J.Y. Mol. Cell. Biol. 1994; 14: 2066-2076Crossref PubMed Scopus (1025) Google Scholar, 7Xiong Y. Zhang H. Beach D. Cell. 1992; 71: 505-514Abstract Full Text PDF PubMed Scopus (899) Google Scholar, 8Matsuoka S. Edwards M.C. Bai C. Parker S. Zhang P. Baldini A. Harper J.W. Elledge S.J. Genes Dev. 1995; 9: 650-662Crossref PubMed Scopus (909) Google Scholar, 9Meyerson M. Harlow E. Mol. Cell. Biol. 1994; 14: 2077-2086Crossref PubMed Scopus (734) Google Scholar). The importance of cyclin D1 in controlling cellular proliferation has been documented both in cultured cells and in whole animals and has been linked to relief of growth suppression mediated through Rb protein (2Sherr C.J. Trends Biochem. Sci. 1995; 20: 187-190Abstract Full Text PDF PubMed Scopus (883) Google Scholar, 4Matsushime H. Ewen M.E. Strom D.K. Kato J.Y. Hanks S.K. Roussel M.F. Sherr C.J. Cell. 1992; 71: 323-334Abstract Full Text PDF PubMed Scopus (780) Google Scholar, 5Ewen M.E. Sluss H.K. Sherr C.J. Matsushime H. Kato J. Livingston D.M. Cell. 1993; 73: 487-497Abstract Full Text PDF PubMed Scopus (916) Google Scholar, 6Matsushime H. Quelle D.E. Shurtleff S.A. Shibuya M. Sherr C.J. Kato J.Y. Mol. Cell. Biol. 1994; 14: 2066-2076Crossref PubMed Scopus (1025) Google Scholar). Studies of cyclin D2- and cyclin D3-mediated control have been reported primarily with hematopoietic cells in which these cyclins are induced upon mitogenic stimulation and are the predominant partners for Cdk4 and Cdk6 (1Sherr C.J. Matsushime H. Roussel M.F. CIBA Found. Symp. 1992; 170: 209-226PubMed Google Scholar,9Meyerson M. Harlow E. Mol. Cell. Biol. 1994; 14: 2077-2086Crossref PubMed Scopus (734) Google Scholar, 10Nagasawa M. Melamed I. Kupfer A. Gelfand E.W. Lucas J.J. J. Immunol. 1997; 158: 5146-5154PubMed Google Scholar, 11Ando K. Ajchenbaum-Cymbalista F. Griffin J.D. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 9571-9575Crossref PubMed Scopus (170) Google Scholar). In Daudi Burkitts lymphoma cells, treatment with α-interferon results in a G0-like arrest accompanied by a loss of cyclin D3 and Cdc25A (12Tiefenbrun N. Melamed D. Levy N. Resnitzky D. Hoffman I. Reed S.I. Kimchi A. Mol. Cell. Biol. 1996; 16: 3934-3944Crossref PubMed Scopus (128) Google Scholar). Similarly, cyclin D3 has also been shown to play a key role in controlling proliferation in 32D myeloid precursor cells, which self-renew when cultured in IL-3 and undergo growth arrest and terminally differentiate to granulocytes in granulocyte colony-stimulating factor. Growth of 32D precursor cells in IL-3 is accompanied by expression of cyclin D2 and cyclin D3, and enforced expression of either cyclin shortened G1 and blocked the ability to differentiate in colony-stimulating factor (11Ando K. Ajchenbaum-Cymbalista F. Griffin J.D. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 9571-9575Crossref PubMed Scopus (170) Google Scholar, 13Kato J.Y. Sherr C.J. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 11513-11517Crossref PubMed Scopus (211) Google Scholar). Studies on the role of cyclin D3 in other cell types has implicated this protein in a variety of decisions concerning cell fate, including proliferation, apoptosis, and differentiation (14Hauser P.J. Agrawal D. Flanagan M. Pledger W.J. Cell Growth Differ. 1997; 8: 203-211PubMed Google Scholar, 15Janicke R.U. Lin X.Y. Lee F.H. Porter A.G. Mol. Cell. Biol. 1996; 16: 5245-5253Crossref PubMed Scopus (55) Google Scholar, 16Wang J. Walsh K. Cell Growth Differ. 1996; 7: 1471-1478PubMed Google Scholar, 17Rogatsky I. Trowbridge J.M. Garabedian M.J. Mol. Cell. Biol. 1997; 17: 3181-3193Crossref PubMed Scopus (224) Google Scholar, 18Mann D.J. Higgins T. Jones N.C. Rozengurt E. Oncogene. 1997; 14: 1759-1766Crossref PubMed Scopus (51) Google Scholar). It is clear that cyclin D3 plays an important role in proliferation in multiple cell types, but specific roles have not been clearly established. The E2F transcription factor plays an important role in cell growth control by linking the activities of the cell cycle machinery with the transcriptional regulation of genes whose products are required for entry into the S phase of the cell cycle (19Nevins J.R. Science. 1992; 258: 424-429Crossref PubMed Scopus (1365) Google Scholar). It has been well documented that this link is achieved through interactions of E2F with members of the retinoblastoma tumor suppressor protein (Rb) family (20Chellapan S. Hiebert S. Mudryj M. Horowitz J. Nevins J.R. Cell. 1991; 65: 1052-1061Abstract Full Text PDF Scopus (1095) Google Scholar, 21Cao L. Faha B. Dembski M. Tsai L.H. Harlow E. Dyson N. Nature. 1992; 355: 176-179Crossref PubMed Scopus (272) Google Scholar, 22Cobrinik D. Whyte P. Peeper D.S. Jacks T. Weinberg R.A. Genes Dev. 1993; 7: 2392-2404Crossref PubMed Scopus (320) Google Scholar). The retinoblastoma susceptibility gene product (pRb) is a nuclear phosphoprotein that has been shown to serve as a negative regulator of cell cycle progression through G1-S transition (23Reed S.I. Wittenberg C. Lew D.J. Dulic V. Henze M. Cold Spring Harbor Symp. Quant. Biol. 1991; 56: 61-67Crossref PubMed Google Scholar, 24Nevins J.R. Chellapan S.P. Mudryj M. Hiebert S. Devoto S. Horowitz J. Hunter T. Pines J. Cold Spring Harbor Symp. Quant. Biol. 1991; 56: 157-162Crossref PubMed Scopus (43) Google Scholar, 25Cobrinik D. Dowdy S.F. Hinds P.W. Mittnacht S. Weinberg R.A. Trends Biochem. Sci. 1992; 17: 312-315Abstract Full Text PDF PubMed Scopus (163) Google Scholar). In quiescent cells, pRb forms complexes with E2F, thus inhibiting the E2F capacity to activate transcription and potentially converting E2F into a dominant transcription inhibitor (26Bremner R. Cohen B.L. Sopta M. Hamel P.A. Ingles C.J. Gallie B.L. Phillips R.A. Mol. Cell. Biol. 1995; 15: 3256-3265Crossref PubMed Scopus (170) Google Scholar). G1 growth arrest by pRb is dependent on the sequences necessary for its physical interaction with the E2F family of transcription factors (20Chellapan S. Hiebert S. Mudryj M. Horowitz J. Nevins J.R. Cell. 1991; 65: 1052-1061Abstract Full Text PDF Scopus (1095) Google Scholar, 27Hiebert S. Chellapan S.P. Horowitz J.M. Nevins J.R. Genes Dev. 1992; 6: 177-185Crossref PubMed Scopus (468) Google Scholar). Binding to these proteins requires the “pocket region” of pRb, which is shared by the two other members of the Rb family, p107 and p130. p130 has been shown to bind cyclin A, E, D1, D2, and D3 in vitro and to inhibit cell growth when overexpressed (28Ewen M.E. Xing Y.G. Lawrence J.B. Livingston D.M. Cell. 1991; 66: 1155-1164Abstract Full Text PDF PubMed Scopus (346) Google Scholar, 29Li Y. Graham C. Lacy S.M. Duncan A. Whyte P. Genes Dev. 1993; 7: 2366-2377Crossref PubMed Scopus (302) Google Scholar, 30Hannon G.J. Demetrick D. Beach D. Genes Dev. 1993; 7: 2378-2391Crossref PubMed Scopus (405) Google Scholar, 31Mayol X. Grana X. Baldi A. Sang N. Hu Q. Giordano A. Oncogene. 1993; 8: 2561-2566PubMed Google Scholar). pRb, p107, and p130 each bind to members of the E2F family at defined but different stages of the cell cycle. pRb binds E2F 1–3, whereas p107 and p130 have preferential affinity for both E2F-4 and E2F-5 (32Sardet C. Videal M. Cobrinik D. Geng Y. Onufryk C. Chen A. Weinberg R.A. Proc. Natl. Acad. Sci. U. S. A. 1995; 93: 2403-2407Crossref Scopus (302) Google Scholar). E2F-4 is the most abundant E2F family member and is clearly present at high levels even in quiescent and differentiated cells (33Moberg K. Starz M.A. Lees J.A. Mol. Cell. Biol. 1996; 16: 1436-1449Crossref PubMed Scopus (304) Google Scholar), 1R. F. Kassatly and W. D. Cress, manuscript in preparation. suggesting it may be the E2F family member most relevant in growth arrest. In arrested cells, the predominant E2F species, E2F-4, appears to be complexed with hypophosphorylated p130 (22Cobrinik D. Whyte P. Peeper D.S. Jacks T. Weinberg R.A. Genes Dev. 1993; 7: 2392-2404Crossref PubMed Scopus (320) Google Scholar, 35Chittenden T. Livingston D.M. DeCaprio J.A. Mol. Cell. Biol. 1993; 13: 3975-3983Crossref PubMed Scopus (83) Google Scholar). E2F-4 associates with other Rb family members; however, pRb and p107 do not associate with E2F-4 until the cells reach the G1-to-S transition (36Schwartz J.K. Devoto S.H. Smith E.J. Chellapan S.P. Jakoi L. Nevins J.R. EMBO J. 1993; 12: 1013-1020Crossref PubMed Scopus (181) Google Scholar). The decay of the E2F-4·p130 complex requires the action of G1 cyclin-dependent kinase activity (37Smith E.J. Leone G. Degregori J. Jakoi L. Nevins J.R. Mol. Cell. Biol. 1996; 16: 6965-6977Crossref PubMed Scopus (207) Google Scholar). The presence of E2F-4·p130 complex correlates with E2F-1 gene repression, and the overexpression of p130 inhibits transcription from the E2F-1 promoter. A D-type cyclin-dependent kinase activity specifically activates the E2F-1 promoter by relieving E2F-mediated repression, suggesting that D-type cyclin-associated kinases may control cell growth through phosphorylating p130 (38Johnson D.G. Oncogene. 1995; 11: 1685-1692PubMed Google Scholar). The existence of several Rb family members, multiple E2F transcription factors, and the preferential binding patterns that have been observed among these proteins suggests that each Rb family member protein may have a specific function(s) in each E2F pathway. Similarly, it can be noted that several D-type cyclins also exist, each capable of forming active complexes with Cdk4 and Cdk6, although specific phosphorylation of specific Rb family members has not been reported. The analysis of recombinant mice deficient for cyclin D1 or cyclin D2 revealed that the loss of either genes did not compromise viability, although cyclin D1−/− offspring were significantly smaller than littermates containing a wild type allele (39Sicinski P. Donaher J.L. Parker S.B. Li T. Fazeli A. Gardner H. Haslam S.Z. Bronson R.T. Elledge S.J. Weinberg R.A. Cell. 1995; 82: 621-630Abstract Full Text PDF PubMed Scopus (892) Google Scholar), and cyclin D2−/− exhibited reduced proliferation in ovarian and testicular tissue (40Sicinski P. Donaher J.L. Geng Y. Parker S.B. Gardner H. Park M.Y. Robker R.L. Richards J.S. McGinnis L.K. Biggers J.D. Eppig J.J. Bronson R.T. Elledge S.J. Weinberg R.A. Nature. 1996; 384: 470-474Crossref PubMed Scopus (594) Google Scholar), indicating that the functions of these cyclins may be redundant. In an analogous fashion, loss of up to two of the three G1 cyclins (CLN) 2The abbreviations used are: CLN, G1cyclins; PDGF, platelet-derived growth factor; GST, glutathioneS-transferase; DMEM, Dulbecco's modified Eagle's medium; CS, calf serum. in yeast does not produce a loss of viability but does severely reduce growth rate. Loss of all three CLN genes, however, results in a loss of viability due to an inability to pass START (41Richardson H.E. Wittenberg C. Cross F. Reed S.I. Cell. 1989; 59: 1127-1133Abstract Full Text PDF PubMed Scopus (379) Google Scholar). A recent study, which has analyzed a number of CLN alleles in more detail, revealed that there are specific functions associated with individual CLN genes, and the growth defects are the result of overlapping functions (42Levine K. Huang K. Cross F.R. Mol. Cell. Biol. 1996; 16: 6794-6803Crossref PubMed Scopus (66) Google Scholar). We have examined whether there are specific functions associated with cyclin D3 in BALB/c 3T3 cells during G1 traverse. We have addressed the possibility that cyclin D3 may function preferentially to relieve growth-suppressive function of a specific Rb family member. Our data reveal that cyclin D3-associated kinase activity is observed during mid-late G1 at a time when p130 is hyperphosphorylated. We have compared cyclin D1- and cyclin D3-mediated p130 phosphorylation in vitro and found that p130 is utilized as a substrate more effectively by cyclin D3 complexes. In addition, our data reveals that cyclin D3 associates with p130 in vivo when p130 becomes hyperphosphorylated. These results suggest that cyclin D3/Cdk4 is responsible for the phosphorylation of p130in vivo, leading to the hypothesis that cyclin D3 may regulate cell cycle progression by antagonizing growth suppressive functions of p130. Biological buffers, detergents, and inorganic molecules were from Sigma or Fisher. Cell culture medium, antibiotics, and protein A-agarose beads were from Life Technologies, Inc., and platelet-derived growth factor (PDGF-BB) was from BioSource Inc. (Camarillo, CA). Nitrocellulose paper and reagents for SDS-polyacrylamide gel electrophoresis were from Bio-Rad. Anti-rabbit horseradish peroxidase, ECL reagents, and radioisotopes were from Amersham Life Science, Inc. Autoradiographic film was from Eastman Kodak Co. GST-Rb (Rb amino acids 379–928) and GST-p107 (p107 amino acids 252–936) has been previously described (43Cress W.D. Johnson D.J. Nevins J.R. Mol. Cell. Biol. 1993; 13: 6314-6325Crossref PubMed Scopus (109) Google Scholar). GST-p130 (p130 amino acids 322–1139) was constructed by cleaving pBS-(s)F-130 (a kind gift from David Johnson, University of Texas) with XbaI, filling in with Klenow, purifying the resulting 2516-base pair fragment, and ligating the fragment into SmaI-digested, alkaline phosphatase-treated pGEX3X (Amrad Corp.). The cyclin D1 (72-13G), D3 (C-16), p107 (CD)X, Rb (IF8), and p130 (sc-317) antibodies were from Santa Cruz Biotechnology (Santa Cruz, CA). The Cdk4 antibody was raised against C-terminal polypeptide of mouse Cdk4. The p27 kip1 antibody was raised against His6-kip fusion protein (44Agrawal D. Hauser P. McPherson F. Dong F. Garcia A. Pledger W.J. Mol. Cell. Biol. 1996; 16: 4327-4336Crossref PubMed Scopus (161) Google Scholar). Mouse BALB/c 3T3 fibroblasts (clone A31) were maintained in an incubator with a humidified atmosphere (5% CO2, 95% air) at 37°C. Experimental cultures were grown to confluence in 100-mm Petri dishes using Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% calf serum (CS), 4 mml-glutamine, 50 units/ml penicillin, and 50 μg/ml streptomycin. Cells were stimulated 2–3 days after density arrest with DMEM containing 10 ng/ml PDGF-BB and 10% CS. For analysis of cell cycle distribution, monolayer cells were trypsinized, fixed in ethanol, and suspended in a 1-ml solution containing 0.1% sodium citrate, 0.1% Triton X-100, and 50 μg/ml propidium iodide and treated for 30 min at 37 °C with 10 μg of RNase A. The stained cells were analyzed by a FACScan flow cytometer (Becton Dickinson, San Jose, CA), and the distribution of cells at each stage of the cycle was determined. Cell extracts were prepared as described previously (45Agrawal D.K. Dong F. Wang Y.Z. Kayda D. Pledger W.J. Cell Growth Differ. 1995; 6: 1199-1205PubMed Google Scholar). Cultures were rinsed twice in ice-cold phosphate-buffered saline, harvested by scraping, and collected by centrifugation. The pellets were resuspended in a lysis buffer (50 mm HEPES, pH 7.5, 100 mm NaCl, 2 mmEDTA, 0.5% Nonidet P-40, 10% glycerol, 0.1 mm sodium orthovanadate, 0.1 mm phenylmethylsulfonyl fluoride, 2.5 μg/ml leupeptin, and 1 mm dithiothreitol), vortexed vigorously, and incubated on ice for 30 min. After centrifugation, the supernatants were transferred into fresh tubes and then stored at −70 °C. Specific antisera were added to 40 μg of total protein from cell extracts in a volume of 250 μl of lysis buffer, and the mixture was rocked for 2–10 h at 4 °C followed by the addition of 30 μl of 25% protein A-agarose beads and an additional 30 min of mixing. Immune complexes were collected by centrifugation and washed twice with lysis buffer, once with 2× kinase reaction buffer (50 mm HEPES, pH 7.5, 10 mmMgCl2, 5 mm MnCl2, 10 mm dithiothreitol), resuspended in 8 μl of 1× kinase reaction buffer containing 10 μCi of [γ-32P]ATP, 10 μm cold ATP, and 1 μg of GST-Rb, GST-p130, or a mixture of GST-Rb, GST-p107, and GST-p130. The mixture was incubated at 30 °C for 15 min, and 10 μl of 2× loading buffer was added. The mixture was then heated to 95 °C for 5 min and separated on an 11% SDS-polyacrylamide gel. Phosphorylated protein was visualized by autoradiography and quantitated by exposure on a PhosphorImager (Molecular Dynamics). Immunoprecipitation was performed using 40 μg of total proteins from cell extracts. Specific antisera were incubated with protein A-agarose beads in 250 μl of the lysis buffer described above at 4 °C for 60 min while shaking. The antibody-linked beads were washed once with lysis buffer, followed by the addition of cell extract. Immune complexes were allowed to form by incubating this mixture for 60 min at 4 °C. For immunoblot analysis, complexes bound to protein A-agarose were washed twice with lysis buffer and separated by SDS-polyacrylamide gel electrophoresis. Immunoblotting was performed by using immunoprecipitated proteins or 40 μg of proteins from cell extracts. Proteins were separated on 6 or 11% SDS-polyacrylamide gels, and the separated proteins were transferred to nitrocellulose membrane. The blots were incubated with the primary antibody for 1 h followed by a 1-h incubation with a secondary antibody. The bound antibodies were visualized using the ECL detection system (Amersham) according to the manufacturer's instructions. E2F DNA binding assays were done as described (46Ikeda M.-A. Jakoi L. Nevins J.R. Proc. Natl. Acad. Sci. U. S. A. 1996; 93: 3215-3220Crossref PubMed Scopus (225) Google Scholar) with minor modifications. The assays used a 32P-end-labeled HaeIII-HindIII fragment, which spanned the region from −103 to −23 relative to the dihydrofolate reductase promoter region, isolated from the dihydrofolate reductase-chloramphenicol acetyltransferase plasmid (47Blake M.C. Azizkhan J.C. Mol. Cell. Biol. 1989; 9: 4994-5002Crossref PubMed Scopus (247) Google Scholar). Cell extracts were incubated for 20 min at room temperature in 20 mm HEPES, pH 7.9, 40 mm KCl, 6 mmMgCl2, 1 mm EGTA, 1 mmdithiothreitol, 10% glycerol/bovine serum albumin at 30 μg/500 ng of sonicated salmon sperm DNA/0.1 ng of 32P-end-labeled probe, following by electrophoresis in a 5% polyacrylamide gel (acrylamide/bisacrylamide, 75:1) in TBE (50 mm Tris-borate, 1 mm EGTA) containing 5% glycerol. Supershift assays were performed in an identical manner except that the cell extracts were preincubated with antibodies against E2F-1 (sc-251), E2F-2 (sc-632), E2F-3 (sc-879), E2F-4 (2–12E8, a kind gift of J. Lees, MIT), Rb (sc-102), p107 (sc-250), or p130 (Transduction Laboratories) for 30 min before the addition of 32P-end-labeled probe. Mitogenic stimulation of density-arrested Balb/c 3T3 cultures with medium containing 10% calf serum and 10 ng/ml PDGF-BB-induced G1 traverse in approximately 90% of the population, with initiation of DNA replication beginning at 12 h (Fig. 1). The majority of cells that had undergone cell cycle traverse reentered G0/G1phase between 24 and 30 h after stimulation and did not undergo further growth, as indicated by the persistence of a G0/G1 population between 30 and 48 h after stimulation. Our results demonstrated that under these conditions most of the cells were stimulated to progress through a single round of replication and reenter a quiescent state within 30 h. Using this cell cycle model, we determined whether the level of cyclin D3-dependent kinase activity varied in a periodic manner during the growth cycle. We prepared extracts from cultures of quiescent, density-arrested BALB/c 3T3 cells stimulated for various times as described in Fig. 1. As shown in Fig. 2 A, cyclin D3-associated kinase activity, measured by the ability of immunoprecipitated cyclin D3 to phosphorylate GST-Rb fusion protein, clearly fluctuated in a cell cycle-dependent manner. The kinase activity was low in quiescent cells (hour 0) and remained low during early G1 (hours 3–6). Increased activity was first apparent by 9 h after stimulation, reached a maximum by 18 h, and declined thereafter. Cells that had completed the cell cycle and reentered quiescence (hour 27) contained activity equivalent to that seen in quiescent cells before stimulation.Figure 2The kinetics of cyclin D3- and D1-dependent kinase activities and p130 proteins during the cell cycle. Density-arrested BALB/c 3T3 fibroblasts were mitogenically stimulated with DMEM supplemented with 10 ng/ml PDGF and 10% CS. Cells were harvested at the indicated time points, cell extracts were prepared, and protein amounts were determined using the Bradford assay. Whole cell extracts containing 40 μg of total proteins were subjected to immunoprecipitation with cyclin D3 (A) or D1 antibodies (B). The kinase activities in the immunoprecipitates were measured as described under “Experimental Procedures” using GST-Rb fusion protein as a substrate. For p130 Western blot analysis, 40 μg of total proteins from the same extracts used in the above kinase assays was subjected to SDS-polyacrylamide gel electrophoresis and blotted to a nitrocellulose membrane. The blot was probed with p130 antibody, and p130 was detected as described under “Experimental Procedures” (C).View Large Image Figure ViewerDownload (PPT) The timing of the appearance of cyclin D3 activity indicated that it was activated in mid to late G1, a timing that is similar to that observed for cyclin D1 as shown in Fig. 2 B. Therefore, the kinetics of cell cycle-dependent regulation of the kinase activities of both cyclin D1- and D3-dependent kinases are similar in mouse BALB/c 3T3 fibroblasts. In addition to cyclin D1 and D3, we also examined cyclin D2-dependent kinase activity. Cyclin D2-associated kinase activity, however, was undetectable in BALB/c 3T3 fibroblasts in similar conditions (data not shown). p107 and p130 are two additional Rb family members that have been identified as substrates for cyclin-dependent kinase activity (48Gibbs E. Pan Z.Q. Niu H. Hurwitz J. J. Biol. Chem. 1996; 271: 22847-22854Abstract Full Text Full Text PDF PubMed Scopus (27) Google Scholar, 49Herzinger T. Wolf D.A. Eick D. Kind P. Oncogene. 1995; 10: 2079-2084PubMed Google Scholar). p130 binds to members of the E2F family of transcription factors, resulting in the formation of a complex that is transcriptionally inert or may function as a transcription repressor. As cells traverse G0/G1, p130 becomes phosphorylated and releases free, transcriptionally active E2F-4 (50Vairo G. Livingston D.M. Ginsberg D. Genes Dev. 1995; 9: 869-881Crossref PubMed Scopus (289) Google Scholar). We determined if phosphorylation of p130 varied after quiescent BALB/c 3T3 cells were stimulated with medium supplemented with calf serum and PDGF. Changes in the phosphorylation state of p130 can be detected by alterations in the electrophoretic mobility, and as shown in Fig. 2 C, quiescent 3T3 cells contained two distinct isoforms of p130, termed 1 and 2, that persisted through the first 6 h after stimulation. By 9 h, a distinct form, isoform 3, with a reduced electrophoretic mobility was observed. The type 3 isoform persisted until 18 h after stimulation, following which the total cellular level of p130 fell to undetectable levels. Thus, the state of p130 phosphorylation changed in a manner that correlated with the appearance of cyclin D1- and D3-associated kinase activity. To provide a direct comparison between the degree of p130 phosphorylation and alterations in E2F DNA binding activity, we performed an electrophoretic mobility shift assay with a32P-end-labeled DNA fragment derived from the dihydrofolate reductase gene that is specific for E2F binding (46Ikeda M.-A. Jakoi L. Nevins J.R. Proc. Natl. Acad. Sci. U. S. A. 1996; 93: 3215-3220Crossref PubMed Scopus (225) Google Scholar, 47Blake M.C. Azizkhan J.C. Mol. Cell. Biol. 1989; 9: 4994-5002Crossref PubMed Scopus (247) Google Scholar) using extracts made from BALB/c 3T3 cells after various times of stimulation. Our results demonstrated that DNA binding complexes containing both E2F and p130 were present in quiescent cells and in cells after mitogenic stimulation for 6 h (Fig. 3 A). By 9 h, the abundance of E2F complexes containing p130 dramatically decreased, demonstrating that p130 phosphorylation results in a release of E2F. To confirm that the major E2F partner in quiescent cells was the p130 seen in Fig. 3 A, a quiescent cell extract was analyzed by gel mobility shift assay with antibody addition. As shown in Fig. 3 B, the majority of E2F DNA binding activity consisted of a single species in a quiescent cell extract. This species could be supershifted by a p130 antibody but not by Rb or p107 antibodies. Furthermore, it could also be supershifted by antibodies against E2F-4 but not E2F-1, -2, or -3. Taken together, these data demonstrate that this species represented the activity of a p130·E2F-4 complex. The E2F DNA binding activity present in a stimulated cell extract, shown in Fig. 3 C, demonstrated that E2F-4 was present both as a free species and in a complex with p107. These data strongly suggest that p130 serves as a major partner of E2F-4 in quiescent cells and that the phosphorylation of p130 may release a substantial amount of E2F activity during the G1/S phase transition. Because in vivo phosphorylation of p130 occurred at a time during which both cyclin D1- and D3-dependent kinase activity was induced, we determined the ability of each of these cyclins from extracts of BALB/c 3T3 fibroblasts to phosphorylate p130. As shown in Fig. 4, cyclin D3 immune complexes efficiently phosphorylated exogenous GST-p130 in vitro, exhibiting induction of activity identical to that obtained with a GST-pRb substrate. In contrast, GST-p130 was a very poor substrate for cyclin D1 immune complexes, despite the fact that both types of complexes are able to phosphorylate pRb efficiently in vitro. We note that cyclin D1 complexes were able to phosphorylate p130 in vitro as determined after longer exposure of the gel shown in Fig. 4 (data not shown). The level of activity, however, was less than 5% that obtained with cyclin D3 complexes. To further determine the relative substrate preference of phosphorylation of the members of Rb family protein by D-type cyclin-dependent kinases, active cyclin D1 and cyclin D3 complexes were utilized in an in vitro kinase assay using a mixture of 1 μg each of GST-Rb, GST-p107, and GST-p130 as substrate. As shown in Fig. 4 B, the ability of cyclin D3-dependent kinase to phosphorylate GST-p107 was equal to that of cyclin D1-dependent kinase, and a comparable level of phosphorylation activity was also observed for a GST-pRb substrate. There was, however, a distinct difference in their ability to efficiently utilize GST-p130, in agreement with the results obtained in the single substrate reactions shown above. To demonstrate that the absence of p130 phosphorylation activity in cyclin D1 complexes was not a property specific for the antibody we utilized, we also measured the activity of cyclin D1 immune complexes prepared with three different cyclin D1 antibodies. Data in Fig. 4 C show that the cyclin D3 immune complex phosphorylated p107 (lane 1) and p130 (lane 5) with similar efficiency. The cyclin D1 immune complexes did not phosphorylate p130 (lane 6–8) yet transferred phosphate to p107 (lane 2–4) as effectively as the cyclin D3 immune complex (lane 1), irrespective of the antibody utilized for immunoprecipitation. It has been shown that cyclin D3 forms complexes with p130 in vitro (30Hannon G.J. Demetrick D. Beach D. Genes Dev. 1993; 7: 2378-2391Crossref PubMed Scopus (405) Google Scholar). Our data suggested that cyclin D3-dependent kinase phosphorylated p130 in vivo, implying that an association of p130 and cyclin D3 would be detectable during specific intervals of the cell cycle. We tested this hypothesis by examining p130 immune complexes derived from cells at various stages of the cell cycle for the presence of cyclin D3 protein. From the data presented in Fig. 5 it is evident that cyclin D3 was associated with p130 in a periodic fashion. It became detectable 6 h after mitogenic stimulation, reached a maximum at approximately 9–12 h, and gradually dissociated over the next 6 h, corresponding to both the induction of cyclin D3 kinase activity and the hyperphosphorylation of p130 protein observed in vivo. By 21 h after stimulation, the association was barely detectable, most likely reflecting the low levels of p130 protein present at these times. We were unable to detect the association of cyclin D1 with p130 at these same times (data not shown). Thus, we could not find evidence of association between cyclin D1 and p130 in vivo. To directly test whether the differential phosphorylation of p130·E2F4 by cyclin D1 or cyclin D3 complexes could be demonstrated to result in a functional consequence, we determined whether treatment with either kinase complex affected the ability of p130 to associate with E2F-4. Fig. 5 B clearly reveals that E2F-4 can be detected in p130 immune complexes isolated in an extract prepared from quiescent cells (lane 1). Treatment with an agarose bead linked with cyclin D1 kinase complex did not substantially alter the amount of E2F-4 from p130 immune complexes (lane 2). Treatment of this cell extract with an agarose bead-linked cyclin D3 kinase complex resulted in almost complete loss of E2F-4 from the subsequently isolated p130 immune complexes (lane 3), thus supporting the hypothesis that cyclin D3 activity is directed toward the repression of E2F-4 by p130. We have investigated the role of cyclin D3 in the regulation of cell cycle traverse in BALB/c 3T3 cells and found that this cyclin specifically phosphorylates the Rb family member, p130. Cyclin D3-associated kinase activity was undetectable in quiescent cells, however, and only became apparent after mitogenic stimulation during mid-G1. Although both cyclin D1- and cyclin D3-mediated activity exhibited a similar pattern of induction based on in vitro kinase assays utilizing a pRb substrate, only cyclin D3-associated kinase activity, derived from cyclin D3/CDK4 complex, efficiently phosphorylated p130 during the same interval. The ability of D-type cyclin-dependent kinase activity to phosphorylate Rb family members has been well documented; however, specificities between particular D-type cyclins and members of the Rb family have not been well characterized. Since we observed both cyclin D1 and cyclin D3 activity over the same interval, this system presented a unique opportunity to address this question. Our data support the hypothesis that the cyclin D3 complex is the major kinase that is specifically targeted to p130 growth-suppressive control. This idea is sustained by both in vitro and in vivo data. In vitro kinase assays clearly revealed that p130 is utilized as a substrate by cyclin D3 complexes more efficiently than cyclin D1 complexes, resulting in greater than 20-fold differences in phosphorylated product (Fig. 4 A). In contrast, kinase reactions utilizing p107 as a substrate revealed more similar levels of catalytic activity (Fig. 4 B). Data in Fig. 4 C demonstrate that the lack of p130 phosphorylation by cyclin D1 immune complexes was not due to one specific antibody interaction with the cyclin D1·Cdk4 complex. The specificity of D-cyclin complexes toward a p107 substrate was not studied further, since it complexes with E2F only after G1 phase (Fig. 3 C). Although differences in the catalytic efficiency were observed with respect to p130 phosphorylation, the apparent kinetics of the appearance of D-cyclin activity during cell cycle traverse remained comparable, with activity appearing between 9 and 12 h. In vivo, analysis of p130 immune complexes revealed that cyclin D3 protein was present (Fig. 5), whereas cyclin D1 could not be found in p130 immunoprecipitates (data not shown). Furthermore, the association of cyclin D3 with p130 displayed kinetics identical to that observed for cyclin D3 activity in vitro and hyperphosphorylation of p130 in vivo (Figs. 2 and 5). Our data suggest that cyclin D3-dependent kinase is responsible for p130 phosphorylation in BALB/c 3T3 fibroblasts. Our analysis of p130 protein in these murine fibroblasts during cell cycle traverse revealed a pattern similar to that previously reported for human glioblastoma T98G cells, with three distinct phosphorylated forms that could be distinguished by differences in electrophoretic mobility (51Mayol X. Garriga J. Grana X. Oncogene. 1995; 11: 801-808PubMed Google Scholar). Only the two faster migrating forms were found during G0 and early G1. The slowest migrating form appeared in the interval between 6 and 9 h and increased such that by 15 h it comprised nearly all the p130 that was detectable. This abrupt transition was shown to occur concomitantly with the activation of D-type cyclin-dependent kinases. The specificity of p107/p130 to bind to a subset of E2F family members that is distinct from that observed with pRb, coupled to the fact that p130·E2F complexes are observed during G0 and early G1 whereas p107·E2F complexes are observed during entry into and traverse of S phase suggest that each Rb family contributes a unique regulatory function to cell cycle control. Similarly, multiple D-type cyclins have been demonstrated to bind and activate Cdk4/6 to provide a G1 cyclin-dependent kinase activity that may phosphorylate and thereby relieve Rb- and p130-mediated growth suppression separately (38Johnson D.G. Oncogene. 1995; 11: 1685-1692PubMed Google Scholar); our data suggest that different D-type cyclins may preferentially phosphorylate specific Rb family membersin vivo. The phenotypes of cyclin D1−/− and cyclin D2−/− mice suggest that this specificity of D-cyclins for Rb family members may be partially overlapping and that the absence of a particular D-cyclin may be compensated by those remaining (39Sicinski P. Donaher J.L. Parker S.B. Li T. Fazeli A. Gardner H. Haslam S.Z. Bronson R.T. Elledge S.J. Weinberg R.A. Cell. 1995; 82: 621-630Abstract Full Text PDF PubMed Scopus (892) Google Scholar, 40Sicinski P. Donaher J.L. Geng Y. Parker S.B. Gardner H. Park M.Y. Robker R.L. Richards J.S. McGinnis L.K. Biggers J.D. Eppig J.J. Bronson R.T. Elledge S.J. Weinberg R.A. Nature. 1996; 384: 470-474Crossref PubMed Scopus (594) Google Scholar). It is interesting to note that a similar redundancy has also been found for p107- and p130-deficient mice (34Cobrinik D. Lee M.H. Hannon G. Mulligan G. Bronson R.T. Dyson N. Harlow E. Beach D. Weinberg R.A. Jacks T. Genes Dev. 1996; 10: 1633-1644Crossref PubMed Scopus (378) Google Scholar), suggesting partial redundancy may exist at multiple levels in components of the D-cyclin/Rb regulatory network. A comparison of the p130 phosphorylation pattern (Fig. 2 C) with the E2F DNA binding activity (Fig. 3 A) during the cell cycle indicated that the interval (between 6–9 h) in which the majority of the p130 is converted to a hyperphosphorylated form coincides to the loss of a detectable p130·E2F complex, suggesting that, like Rb, p130 dissociates from E2F after it is phosphorylated, a finding in accordance with previous analysis of REF52 cells (38Johnson D.G. Oncogene. 1995; 11: 1685-1692PubMed Google Scholar). The role of cyclin D3·Cdk4 complex in the regulation of p130·E2F activity is strengthened by our observation that the cyclin D3·Cdk4 activity causes the complete disruption of the p130·E2F-4 complex as determined with an in vitro assay (Fig. 5 B). The relatively short duration of the interval in which p130·E2F activity is lost suggests that p130 inactivation may occur as a result of a burst of p130-directed kinase activity at a discrete point in G1. Somewhat surprisingly, we did not observe proportional changes in p130·E2F and free E2F-4 activity. Rather, there was a loss of p130·E2F (between 6 and 9 h) followed by a delayed appearance of free E2F-4 and p107·E2F activity (between 9 and 12 h). It is unlikely that this pattern is the result of a change in E2F protein levels, since the level of E2F-4 remains constant in Balb/c 3T3 cells throughout the cell cycle. 3W. D. Cress, manuscript in preparation. Given the results that have demonstrated a repressor function for the p130·E2F complex, this observation may indicate the requirement for an additional step(s) between loss of repressor function and the appearance of transactivation function. We thank the Molecular Imaging Core and the Flow Cytometry Core of the H. Lee Moffitt Cancer Center and Research Institute. We acknowledge Dr. Jacqueline Lees (MIT) for providing a monoclonal antibody directed against E2F4.

Referência(s)