- Email: [email protected]
Translocation of cdk2 to the Nucleus during G1-Phase in PDGF-Stimulated Human Fibroblasts
EXPERIMENTAL CELL RESEARCH ARTICLE NO.
232, 72 –78 (1997)
EX973507
Translocation of cdk2 to the Nucleus during G1-Phase in PDGF-Stimulated Human Fibroblasts CORNELIA DIETRICH, KATJA WALLENFANG, FRANZ OESCH, AND RAIMUND WIESER1 Institute of Toxicology, Johannes-Gutenberg University, Obere Zahlbacherstrasse 67, 55131 Mainz, Germany
Central key regulators of the eukaryotic cell cycle are the cyclin-dependent kinases (cdks), a family of serine/ threonine kinases which are sequentially activated and deactivated during the cell cycle [for review see 1, 2]. Several mechanisms are employed to regulate cdk activity in order to ensure that the cell’s proper cycle is tightly controlled: activation of a cdk, the catalytic subunit, is dependent on the association with a cyclin, the regulatory subunit. Synthesis and degradation of the various cyclins follow an ordered sequence. In addition, activation of the cdk–cyclin complex requires phosphorylation of the catalytic subunit at a conserved Thr residue by cdk-activating kinase (CAK) and dephosphorylation at Thr 14 and Tyr 15. The activity of the complex is further modulated by the association with small inhibitory proteins, known as p15, p16, p21, and p27 [for review see 3]. D-type cyclins (D1, D2, D3) are synthesized in early G1 in response to mitogenic stimuli and form complexes with cdk4 and cdk6 [4, 5]. In late G1, cyclin E synthesis begins with a maximum at the G1/S transition. Cyclin E primarily binds to cdk2 [6, 7]. Both complexes contribute to phosphorylation of the retinoblastoma gene product (ppRB), thus initiating S-phase [8]. DNA synthesis is sustained when cyclin A associates with cdk2. In late S-phase, cyclin B binds to cdc2 and is active during G2- and M-phase [for review see 9]. CAK, which is responsible for phosphorylation at Thr 160 of cdk2, has been shown to be itself a cyclin-dependent kinase composed of a catalytic subunit, cdk7, and a positive regulatory subunit, cyclin H [10]. Surprisingly, it has been revealed that protein levels of cdk7 and CAK activity remain constant throughout the cell cycle [11]. This suggests that phosphorylation by CAK may be regulated by the availability of the cdk/cyclin substrates. Since cdk7 is localized predominantly in the nucleus, it may be assumed that the intracellular distribution of cdks and/or cyclins change according to the cell cycle stage. Indeed, this has been demonstrated by Pines and Hunter for cyclin B, which accumulates in the cytoplasm of interphase cells and enters the nucleus at the beginning of mitosis [12]. In contrast, the G1-cyclins, cyclin D1 and cyclin E as well as cyclin A, are nuclear proteins [12 –14]. In preliminary studies
We studied the subcellular distribution of cdk2 in synchronized, PDGF-stimulated human fibroblasts (FH109). After contact inhibition and serum depletion, more than 95% of FH109 cells were arrested in G0/G1phase. PDGF-AB led to a 16-fold increase in proliferation compared with untreated cells. Cell cycle progression was studied by flow cytometric analysis, [3H]thymidine incorporation, and phosphorylation of the retinoblastoma gene product, pRB. Using Western blot analysis after subcellular fractionation, we revealed that after PDGF stimulation the phosphorylated (Thr 160), i.e., activated, form of cdk2 (33 kDa) first appeared in the nucleus at late G1-phase and persisted throughout until to the end of S-phase. Since cdk2 was not synthesized de novo, and the amount of inactive cdk2 (35 kDa) remained constant in the nucleus, we suggested a translocation from the cytosol to the nucleus in late G1. Using immunofluorescence techniques, we detected a diffuse staining in quiescent cells. Starting at late G1-phase, cdk2 immunoreactivity was concentrated to the nucleus while immunoreactivity in the cytosol disappeared. We therefore draw the conclusion that cdk2 is translocated from the cytosol into the nucleus in late G1-phase. Since protein levels and activity of cdk7, which is the catalytic subunit of cdk-activating kinase (CAK) phosphorylating cdk2, remained constant throughout the cell cycle, CAK activity might therefore be regulated by the availability of its substrate cdk2. q 1997 Academic Press
INTRODUCTION
In eukaryotic cells, ordered progression through the cell cycle is the result of a complex integration of a multitude of positive and negative signals. These include signals generated by extracellular agents such as growth factors and mitogenic hormones as well as intracellular signals generated at specific checkpoints within the cell cycle. 1 To whom correspondence and reprint requests should be addressed. Fax: /49-6131-230506. E-mail: [email protected]. uni-mainz.de.
72
0014-4827/97 $25.00 Copyright q 1997 by Academic Press All rights of reproduction in any form reserved.
AID ECR 3507
/
6i1d$$$401
03-12-97 20:27:13
eca
CELL CYCLE-DEPENDENT SUBCELLULAR DISTRIBUTION OF cdk2
we have revealed that cdk2 is distributed in the cytosol and/or nucleus in asynchronously proliferating human fibroblasts. We therefore examined in detail the subcellular distribution of cdk2 in synchronized human fibroblasts. For synchronization, human fibroblasts were arrested in G0/G1 by contact inhibition and serum depletion and then stimulated to enter cell cycle with platelet-derived growth factor. Here we show for the first time that, in human fibroblasts, cdk2 is translocated from the cytosol into the nucleus in G1-phase. MATERIALS AND METHODS Cell synchronization. FH109 human embryonal lung fibroblasts [15] were seeded to confluence and cultured for 3 days in DMEM (PAA) supplemented with 10% FCS (Gibco) and for an additional 24 h in DMEM/0.5% FCS. The quiescent cells were then stimulated with PDGF-AB (50 ng/ml, Boehringer Mannheim) for 10, 14, 18, and 22 h. Progression through the cell cycle was monitored by flow cytometry and [3H]thymidine incorporation for 30 min, as previously described [16]. Cell extracts and Western blotting. A total of 8 1 105 cells were seeded in 6-well plates and treated as described above. For total cell extracts, cells were solubilized in 500 ml of boiling SDS-sample buffer [17]. Cell fractionation for pRB detection was performed as previously described [16]. For cdk2 and cdk7 detection, nuclei were separated from the cytosol according to [18]. Briefly, cells were washed three times with ice-cold phosphate-buffered saline (PBS) containing 1 mM MgCl2 and harvested with a rubber policeman in 1 ml of icecold hypotonic buffer (10 mM Hepes, pH 7.9, 1.5 mM MgCl2 , 10 mM KCl, 300 mM sucrose, 1 mM EDTA, 1 mM DTT, 1 mM Na3VO4 , 2.5 mM NaF, 0.3 mM aprotinin, 0.1 mM soybean trypsin inhibitor, 60 mM phenylmethylsulfonyl fluoride, 0.4 mM iodoacetamid, 5 mM pepstatin, 4 mM leupeptin). After swelling on ice for 10 min, cells were disrupted by repeated aspiration through a 22-gauge needle. Cell disruption was controlled by microscopic observation. Nuclei were separated from the cytosol by centrifugation at 47C for 10 min at 960g, washed once with hypotonic buffer, and solubilized in SDS– sample buffer. Proteins of the supernatant were precipitated [19] and solubilized in SDS– sample buffer. Protein determination of each fraction was performed according to [20]. After SDS–polyacrylamide gel electrophoresis (SDS –PAGE: 7.5%, 20 mg protein/lane for pRB detection, or 10–20%, 20 mg protein/lane for cdk2 detection, or 12.5%, 30 mg protein/lane for cdk7 detection), proteins were transferred onto Immobilon membrane (Millipore) and blocked for 1 h with 5% milk powder in TBS (10 mM Tris/HCl, pH 8.0, 150 mM NaCl, 0.05% Tween 20). For immunodetection, membranes were incubated for 1.5 h with anti-Rb, anti-cdk2 (each 0.1 mg/ml, Santa Cruz), or anti-cdk7 antibodies (1 mg/ml, Santa Cruz) in blocking buffer followed by alkaline phosphatase-conjugated antirabbit antibody (1:3000, Biozol) for 1.5 h. 4-Nitroblue tetrazolium chloride and 5-bromo-4-chloro-3-indolyl phosphate (both from Boehringer Mannheim) was used as a detection system. Immunoprecipitation. A total of 8 1 105 cells were seeded in 6well plates, cultured, and treated as described. Cells were lysed for 20 min at 47C in RIPA buffer (50 mM Tris/HCl, pH 7.5, 1% sodium deoxycholate, 1% Triton X-100, 0.1% SDS, phosphatase, and protease inhibitors as mentioned) and immunoprecipitated for 4 h at 47C with anti-cyclin E antibodies (10 mg, Santa Cruz) covalently coupled to protein A– Sepharose (12.5 mg, Sigma). The immunoprecipitate was washed three times with RIPA buffer and proteins eluted with 100 ml of diethylamine (100 mM, pH 11.5) for 10 min at room temperature. The supernatant was neutralized with HCl, and the proteins were precipitated [19] and solubilized in SDS–sample buffer. SDS– PAGE and western blotting was performed as described above.
AID ECR 3507
/
6i1d$$$402
03-12-97 20:27:13
73
Immunofluorescence. For this procedure, 106 cells (60-mm dishes) were cultured on glass coverslips as described above. After washing with PBS three times, cells were fixed with acetone at 0207C. The coverslips were air-dried and washed three times with PBS. Unspecific binding was suppressed by blocking with 10% goat serum in PBS for 20 min. After three short and three 5-min washing steps with PBS, cells were incubated for 1 h with anti-cdk2 (0.5 mg/ml) antibodies in PBS containing 1% BSA. The cells were washed in three short and three 5-min washing steps and then incubated for 1 h with FITC-conjugated swine anti-rabbit immunoglobulins (1:20, DAKO) in PBS/0.5% gelatin/5% BSA/10% goat serum. The cells were again washed as described and the coverslips mounted on glass slides with Histogel (Camon) containing 2.5% 1,4-diaminobicyclooctane as an anti-fading agent (Sigma). Controls were treated the same way except that the primary antibody was omitted.
RESULTS
For cell synchronization, FH109 cells were arrested in G0/G1-phase by contact inhibition for 3 days and additionally by serum reduction to 0.5% for 24 h. As assessed by flow cytometric analysis after propidium iodide staining, more than 95% of the fibroblasts showed a 2n DNA content. (Fig. 1A). The quiescent fibroblasts were then exposed to PDGF-AB (50 ng/ml) and cell cycle progression was studied by propidium iodide staining for flow cytometry and [3 H]thymidine incorporation. Figure 1B shows that 10 h after PDGF addition the majority of the cells were still in G1-phase. DNA synthesis started 14–16 h after PDGF exposure with a maximum at 18–20 h (Figs. 1A and 1B). The maximal [3H]thymidine incorporation was 16-fold higher (16.25 { 1.01, n Å 4) in PDGF-treated compared with untreated cells. Cell cycle progression was further reflected by phosphorylation of the retinoblastoma gene product (pRB). It is generally believed that pRB becomes phosphorylated in mid G1- to S-phase. In its hypophosphorylated form pRB binds to the transcription factor E2F and blocks transcription. (Hyper)phosphorylation of pRB leads to its functional inactivation and results in a decreased affinity for the nuclear compartment. We therefore fractionated the cells into a nuclear and a cytosolic compartment (see Materials and Methods) and examined pRB phosphorylation by Western blot analysis after PDGF exposure. We observed a strong increase in the slower migrating forms of pRB (Fig. 1C) in the cytosolic compartment. These 115- to 117-kDa forms represent the hyperphosphorylated species of pRB [31]. We selected the three time points 14, 18, and 22 h after PDGF addition for investigation since then cells had reached G1/S-transition, the peak of S-phase, and late S-phase, respectively. We next examined upstream events of pRB, i.e., cdk2 activation by phosphorylation at Thr 160 and association with cyclin E using Western blot analysis, as the cdk2–cyclin E complex is known to contribute to pRB phosphorylation. In whole-cell extracts, the total
eca
74
DIETRICH ET AL.
FIG. 1. Cell cycle analysis after PDGF stimulation of quiescent FH109 cells. (A) FH109 cells were stained with propidium iodide and analyzed by flow cytometry at different time points after PDGF exposure (1, quiescent cells; 2, 16 h; 3, 20 h; 4, 24 h after PDGF exposure). (B) [3H]thymidine incorporation was measured at the indicated time points (n Å 4). (C) The proteins of FH109 cells were separated into a nuclear and cytosolic fraction at the indicated time points after PDGF stimulation. Western blot analysis was performed with anti-pRB antibodies (pRBppp, hyperphosphorylated pRB). The positions of the molecular size marker proteins are shown on the left.
amount of cdk2 was only slightly enhanced by PDGF, but we observed the appearance of a faster migrating species of cdk2 (Fig. 2, top) which is known to be the phosphorylated (Thr 160), i.e., activated, form of cdk2 [21]. In quiescent cells, this faster migrating form was absent, and maximal phosphorylation (Thr 160) occurred 22 h after PDGF stimulation. Immunoprecipitation studies with anti-cyclin E antibodies (Fig. 2, middle) revealed that a significant amount of cdk2 was associated with cyclin E at the G1/S-transition (14 h) with a maximum in early S-phase (18 h). In late Sphase (22 h), cdk2 is probably associated with cyclin A
AID ECR 3507
/
6i1d$$$402
03-12-97 20:27:13
[18]. Fig. 2, bottom, shows the results of Western blotting with anti-cdk2 antibodies after nuclear extraction. In the nucleus, the amount of faster migrating cdk2 (33 kDa) was strongly enhanced in response to PDGF, already after 10 h, when the fibroblasts were still in G1-phase. Since both the amount of slower migrating, i.e., inactive cdk2 (35 kDa) in the nucleus and the total amount of cdk2 (Fig. 2, top) remained constant throughout the cell cycle, i.e., cdk2 was not synthesized de novo, we suggested a translocation from the cytosol into the nucleus. This seemed to be reasonable, since cdk7, which, in association with cyclin H, is responsible
eca
CELL CYCLE-DEPENDENT SUBCELLULAR DISTRIBUTION OF cdk2
FIG. 2. Cell cycle-dependent phosphorylation of cdk2 (Thr 160) in response to PDGF. Total cell extracts (top), anti-cyclin E immunoprecipitates (middle), and nuclear extracts (bottom) were prepared as described under Materials and Methods at the indicated time points after PDGF-addition. Western blot analysis was performed with anti-cdk2 antibodies (cdk2p, Thr 160-phosphorylated cdk2). The positions of the molecular size marker proteins are shown on the left.
for cdk2 phosphorylation, has been described to be exclusively located in the nucleus [11]. This was confirmed in own experiments using Western blot analysis with anti-cdk7 antibodies after subcellular fractionation. Figure 3 demonstrates that cdk7 is located in the nucleus and that protein levels of cdk7 are invariant throughout the cell cycle. The hypothesis of a nuclear translocation was strengthened by the results of intracellular immunostaining with anti-cdk2 antibodies. In quiescent cells, we observed a diffuse immunostaining (Fig. 4A) which changed in response to PDGF. After 10 h (Fig. 4B), immunofluorescence appeared predominantly in the nucleus with a maximum of concentration in the nucleus 18 h after PDGF stimulation (Fig. 4D). At the end of S-phase (22 h, Fig. 4E), immunofluorescence was again detected in the nucleus and cytosol. We therefore draw the conclusion that cdk2 is translocated from the cytosol in the nucleus in late G1-phase and phosphorylated by cdk7–cyclin H.
75
for orderly progression through the cell cycle. It is now believed that cdc2 is only involved in G2/M-transition being activated by cyclin B [22]. Progression from G1- to S-phase, i.e., initiation and regulation of DNA replication, is strongly dependent on cdk2 [23, 24]. In late G1- and early S-phase, cdk2 is activated by the association of cyclin E [25] and then by cyclin A when S-phase is sustained [26]. It is generally accepted that cdk2–cyclin E— together with cdk4– cyclin D— is responsible for phosphorylation of the retinoblastoma gene product. The hypophosphorylated species of pRB binds to the transcription factor E2F, hence inhibiting transcription. (Hyper)phosphorylation of pRB leads to its functional inactivation, resulting in the loss of binding to E2F and thus permitting entry into S-phase [27, 28]. When cyclin E is degraded and pRB released from E2F, cdk2–cyclin A binds to E2F, thereby regulating DNA replication [for review see 29]. Pagano and co-workers [18] have recently reported that cdk2 is absent in quiescent human fibroblasts and appears as a nuclear protein in S-phase. Activity of cdk2 corresponded largely with cyclin A activity. Since we have revealed that cdk2 is distributed in the cytosol or nucleus in asynchronously proliferating human fibroblasts (unpublished data), we suggested that cdk2 is translocated to the nucleus already in G1-phase when it is associated with cyclin E. We therefore studied subcellular distribution of cdk2 in synchronized, PDGF-stimulated human fibroblasts, FH109 cells. After contact inhibition and additional serum depletion, more than 95% of the FH109 cells were arrested in G0/G1-phase. The quiescent fibroblasts could be restimulated to enter the cell cycle by exposure to PDGF. Late G1-phase was reached 10 h after PDGF treatment, and S-phase began 14–16 h after PDGF exposure with a peak of DNA synthesis at 18 –20 h. Progres-
DISCUSSION
In yeast, a single 34-kDa cyclin-dependent kinase (called cdc2 in Schizosaccharomyces pombe and CDC28 in Saccharomyces cerevisiae) is the key regulator for cell proliferation. In mammalian cells, cell cycle regulation is more complex and different cyclin-dependent kinases in combination with different cyclins that are activated in a finely balanced sequence are required
AID ECR 3507
/
6i1d$$$402
03-12-97 20:27:13
FIG. 3. Subcellular distribution of cdk7 and protein levels after PDGF stimulation. FH109 cell lysates were separated into a nuclear and a cytosolic fraction at the indicated time points after PDGF exposure. Western blot analysis was performed with anti-cdk7 antibodies. The positions of the molecular size marker proteins are shown on the left.
eca
76
03-12-97 20:27:13
eca
CELL CYCLE-DEPENDENT SUBCELLULAR DISTRIBUTION OF cdk2
sion through the cell cycle was monitored by an increase in pRB phosphorylation in late G1- and S-phase, which resulted in a clear shift to lower electrophoretic mobility [28]. In accordance with the work of Mittnacht and Weinberg [27], the hyperphosphorylated species of pRB had a decreased affinity for the nuclear compartment. It is generally accepted that the cdk2–cyclin E complex— in collaboration with cdk4–cyclin D —is responsible for pRB phosphorylation. With respect to our immunoprecipitation studies with anti-cyclin E antibodies, we conclude that from mid G1- to early S-phase a significant amount of cdk2 was associated with and thereby regulated by cyclin E. Cyclin E itself is known to be synthesized in late G1- and early S-phase [25], hence regulating cdk2 activity. To be active, cdk2 has to be additionally phosphorylated at a conserved Thr residue, Thr 160 (and dephosphorylated at Thr 14 and Tyr 15). Concomitant with pRB phosphorylation, a faster migrating species of cdk2 occurred. This form represented the activated form of cdk2 phosphorylated at Thr 160 [16, 21]. We therefore draw the conclusion that in late G1- to early S-phase cdk2 was activated by cyclin E association and Thr phosphorylation, hence leading to phosphorylation of pRB and finally to initiation of DNA synthesis. As mentioned above, activation of cdk2 requires phosphorylation at Thr 160 by CAK. CAK itself is composed of a catalytic and a regulatory subunit: cdk7, a nuclear protein, and cyclin H. Since both the protein level of cdk7 and the activity of CAK remain constant throughout the cell cycle (Fig. 3; [11, 30]), phosphorylation by CAK might be regulated by the substrate availability of cdk2, which would require translocation from the cytosol into the nucleus. We next examined subcellular distribution of cdk2 and cdk7 using Western blot analysis after subcellular fractionation and immunofluorescence microscopy. In the nuclear extracts, we observed a similar increase in the faster migrating, i.e., activated form of cdk2 compared with total cell extracts. Since the slower migrating, i.e., inactive form remained constant in the nucleus and cdk2 was not synthesized de novo, we suggested that a significant amount of cdk2 had to be translocated from the cytosol to the nucleus to become phosphorylated. Indeed, using immunofluorescence microscopy, we observed that cdk2 was distributed diffusely in the cytosol and the nucleus in quiescent fibroblasts. Already 10 h after PDGF treatment, a time point when the majority of the fibroblasts were still in G1-phase, there was a strong increase in nuclear immunostain-
77
ing, whereas immunoreactivity in the cytosol disappeared. Maximal immunofluorescence was observed in S-phase exclusively in the nucleus, declining at the end of S-phase. Our findings are in contrast with the work of Pagano et al. [18], who failed to detect cdk2 immunoreactivity in quiescent fibroblasts and during G1phase. Since they used a different anti-cdk2 antibody, the most likely explanation is that our anti-cdk2 antibody is more sensitive in immunofluorescence studies. Conclusively, our data suggest a translocation of cdk2 from the cytosol to the nucleus already before S-phase, which, to our knowledge, has not been described so far in human fibroblasts. Only recently, Brenot-Bosc and co-workers have reported a similar translocation of cdk2 before S-phase in nocodazole-blocked MANCAcells, a B-cell lymphoma line [31]. As has been shown by Tassan and co-workers [11], cdk7 was found to be predominantly located in the nucleus, and protein levels of cdk7 were constant throughout the cell cycle. The small amount of cdk7 in the cytosol is probably an artifact due to the experimental procedure, since in control experiments we found similar amounts of the nuclear protein cyclin D in the cytosolic fraction (unpublished data). However, first studies using immunofluorescence techniques revealed a cell cycle-dependent change in the nuclear compartimentalization of cdk7. cdk7 showed a granular pattern in mid to late G1 (10 –14 h after PDGF treatment) that was absent in quiescent or S-phase cells (unpublished data) and coincided with the appearance of cdk2 phosphorylation. Conclusively, cdk7 activity might not be regulated only by the translocation of its substrate cdk2 to the nucleus, but also by an intranuclear translocation to specific, still unknown structures. This work was supported by a grant of the Deutsche Forschungsgemeinschaft, WI 727/3-1, and is part of the Ph.D. thesis of K.W.
REFERENCES 1. Sherr, C. J. (1994) Cell 79, 551– 555. 2. Lees, E. (1995) Curr. Opin. Cell Biol. 7, 773– 780. 3. Peter, M., and Herskowitz, I. (1994) Cell 79, 181– 184. 4. Matsushime, H., Quelle, D. E., Shurtleff, S. A., Shibuya, M., Sherr, C. J., and Kato, J. (1994) Mol. Cell. Biol. 14, 2066 –2076. 5. Meyerson, M., and Harlow, E. (1994) Mol. Cell. Biol. 14, 2077 – 2086. 6. Dulic, V., Lees, E., and Reed, S. (1992) Science 257, 1958 –1061. 7. Koff, A., Giordano, A., Desai, D., Yamashita, K., Harper, W., Elledge, S., Nishimoto, T., Morgan, D., Franza, R., and Roberts, J. (1992) Science 257, 1689 – 1693.
FIG. 4. Cell cycle-dependent subcellular distribution of cdk2 after PDGF stimulation. FH109 cells were stained with anti-cdk2 antibodies before (A) or 10 h (B), 14 h (C), 18 h (D), or 22 h (E) after PDGF addition. (left: original magnification, 1180; bar, 50 mm; right: original magnification, 1720; bar, 10 mm).
AID ECR 3507
/
6i1d$$$402
03-12-97 20:27:13
eca
78
DIETRICH ET AL.
8.
Hatakeyama, M., Brill, J. A., Fink, G. R., and Weinberg, R. (1994) Genes Dev. 8, 1759 –1771.
9.
King, R., Jackson, P., and Kirschner, M. (1994) Cell 79, 563– 571.
21.
10. Fisher, R. P., and Morgan, D. (1994) Cell 78, 713– 724.
22.
11. Tassan, J.-P., Schultz, S. J., Bartek, J., and Nigg, E. A. (1994) J. Cell Biol. 127, 467–478.
23.
12. Pines, J., and Hunter, T. (1991) J. Cell Biol. 115, 1– 17. 13. Baldin, V., Lukas, J., Marcote, M. J., Pagano, M., and Draetta, G. (1993) Genes Dev. 7, 812– 821.
24. 25.
14. Ohtsubo, M., Theodoras, A. M., Schumacher, J., Roberts, J., and Pagano, M. (1995) Mol. Cell. Biol. 15, 2612 –2624. 15. Wieser, R., Heck, R., and Oesch, F. (1985) Exp. Cell Res. 158, 493–499.
26.
16. Dietrich, C., Bartsch, T., Schanz, F., Oesch, F., and Wieser, R. (1996) Proc. Natl. Acad. Sci. USA 93, 10815 –10819.
27. 28.
17. Laemmli, K. K. (1970) Nature 227, 680–685. 18. Pagano, M., Pepperkok, R., Lukas, J., Baldin, V., Ansorge, W., Bartek, J., and Draetta, G. (1993) J. Cell Biol. 121, 101 –111.
29. 30.
19. Wessel, D., and Flu¨gge, U. J. (1984) Anal. Biochem. 138, 141– 143. 20. Smith, P. K., Krohn, R. I., Hermanson, G. T., Mallia, A. K., Gartner, F. H., Provenzano, M. D., Fujimoto, E. K., Golke, N. M.,
31.
Olson, B. J., and Klenk, D. C. (1985) Anal. Biochem. 150, 76– 85. Gu, Y., Rosenblatt, J., and Morgan, D. (1992) EMBO J. 11, 3995 –4005. Schwob, E., Bo¨hm, T., Mendenhall, M. D., and Nasmyth, K. (1994) Cell 79, 233–244. Koff, A., Ohtsuki, M., Polyak, K., Roberts, J. M., and Massague, J. (1993) Science 260, 536– 539. Ohtsubo, M., Theodoras, A. M., Schumacher, J., Roberts, J., and Pagano, M. (1995) Mol. Cell. Biol. 15, 2612 – 2624. Koff, A., Giordano, A., Desai, D., Yamashita, K., Harper, W., Elledge, S., Nishimoto, T., Morgan, D. O., Franza, B. R., and Roberts, J. (1992) Science 257, 1689 –1694. Pagano, M., Pepperkok, R., Verde, F., Ansorge, W., and Draetta, G. (1992) EMBO J. 11, 961–971. Mittnacht, S., and Weinberg, R. A. (1991) Cell 65, 381–393. DeCaprio, J. A., Ludlow, J. W., Lynch, D., Furukawa, Y., Griffin, J., Piwnica-Worms, H., Huang, C.-M., and Livingston, D. M. (1989) Cell 58, 1085 –1095. Moran, E. (1991) Gene Expression 1, 281–283. Labbe, J.-C., Martinez, A.-M., Fesquet, D., Capony, J.-P., Darbon, J.-M., Derancourt, J., Devault, A., Morin, N., Cavadore, J.-C., and Marcel, D. (1994) EMBO J. 13, 5155 –5164. Brenot-Bosc, F., Gupta, S., Margolis, R. L., and Fotedar, R. (1995) Chromosoma 103, 517 –527.
Received August 12, 1996 Revised version received January 13, 1997
AID ECR 3507
/
6i1d$$$403
03-12-97 20:27:13
eca
232, 72 –78 (1997)
EX973507
Translocation of cdk2 to the Nucleus during G1-Phase in PDGF-Stimulated Human Fibroblasts CORNELIA DIETRICH, KATJA WALLENFANG, FRANZ OESCH, AND RAIMUND WIESER1 Institute of Toxicology, Johannes-Gutenberg University, Obere Zahlbacherstrasse 67, 55131 Mainz, Germany
Central key regulators of the eukaryotic cell cycle are the cyclin-dependent kinases (cdks), a family of serine/ threonine kinases which are sequentially activated and deactivated during the cell cycle [for review see 1, 2]. Several mechanisms are employed to regulate cdk activity in order to ensure that the cell’s proper cycle is tightly controlled: activation of a cdk, the catalytic subunit, is dependent on the association with a cyclin, the regulatory subunit. Synthesis and degradation of the various cyclins follow an ordered sequence. In addition, activation of the cdk–cyclin complex requires phosphorylation of the catalytic subunit at a conserved Thr residue by cdk-activating kinase (CAK) and dephosphorylation at Thr 14 and Tyr 15. The activity of the complex is further modulated by the association with small inhibitory proteins, known as p15, p16, p21, and p27 [for review see 3]. D-type cyclins (D1, D2, D3) are synthesized in early G1 in response to mitogenic stimuli and form complexes with cdk4 and cdk6 [4, 5]. In late G1, cyclin E synthesis begins with a maximum at the G1/S transition. Cyclin E primarily binds to cdk2 [6, 7]. Both complexes contribute to phosphorylation of the retinoblastoma gene product (ppRB), thus initiating S-phase [8]. DNA synthesis is sustained when cyclin A associates with cdk2. In late S-phase, cyclin B binds to cdc2 and is active during G2- and M-phase [for review see 9]. CAK, which is responsible for phosphorylation at Thr 160 of cdk2, has been shown to be itself a cyclin-dependent kinase composed of a catalytic subunit, cdk7, and a positive regulatory subunit, cyclin H [10]. Surprisingly, it has been revealed that protein levels of cdk7 and CAK activity remain constant throughout the cell cycle [11]. This suggests that phosphorylation by CAK may be regulated by the availability of the cdk/cyclin substrates. Since cdk7 is localized predominantly in the nucleus, it may be assumed that the intracellular distribution of cdks and/or cyclins change according to the cell cycle stage. Indeed, this has been demonstrated by Pines and Hunter for cyclin B, which accumulates in the cytoplasm of interphase cells and enters the nucleus at the beginning of mitosis [12]. In contrast, the G1-cyclins, cyclin D1 and cyclin E as well as cyclin A, are nuclear proteins [12 –14]. In preliminary studies
We studied the subcellular distribution of cdk2 in synchronized, PDGF-stimulated human fibroblasts (FH109). After contact inhibition and serum depletion, more than 95% of FH109 cells were arrested in G0/G1phase. PDGF-AB led to a 16-fold increase in proliferation compared with untreated cells. Cell cycle progression was studied by flow cytometric analysis, [3H]thymidine incorporation, and phosphorylation of the retinoblastoma gene product, pRB. Using Western blot analysis after subcellular fractionation, we revealed that after PDGF stimulation the phosphorylated (Thr 160), i.e., activated, form of cdk2 (33 kDa) first appeared in the nucleus at late G1-phase and persisted throughout until to the end of S-phase. Since cdk2 was not synthesized de novo, and the amount of inactive cdk2 (35 kDa) remained constant in the nucleus, we suggested a translocation from the cytosol to the nucleus in late G1. Using immunofluorescence techniques, we detected a diffuse staining in quiescent cells. Starting at late G1-phase, cdk2 immunoreactivity was concentrated to the nucleus while immunoreactivity in the cytosol disappeared. We therefore draw the conclusion that cdk2 is translocated from the cytosol into the nucleus in late G1-phase. Since protein levels and activity of cdk7, which is the catalytic subunit of cdk-activating kinase (CAK) phosphorylating cdk2, remained constant throughout the cell cycle, CAK activity might therefore be regulated by the availability of its substrate cdk2. q 1997 Academic Press
INTRODUCTION
In eukaryotic cells, ordered progression through the cell cycle is the result of a complex integration of a multitude of positive and negative signals. These include signals generated by extracellular agents such as growth factors and mitogenic hormones as well as intracellular signals generated at specific checkpoints within the cell cycle. 1 To whom correspondence and reprint requests should be addressed. Fax: /49-6131-230506. E-mail: [email protected]. uni-mainz.de.
72
0014-4827/97 $25.00 Copyright q 1997 by Academic Press All rights of reproduction in any form reserved.
AID ECR 3507
/
6i1d$$$401
03-12-97 20:27:13
eca
CELL CYCLE-DEPENDENT SUBCELLULAR DISTRIBUTION OF cdk2
we have revealed that cdk2 is distributed in the cytosol and/or nucleus in asynchronously proliferating human fibroblasts. We therefore examined in detail the subcellular distribution of cdk2 in synchronized human fibroblasts. For synchronization, human fibroblasts were arrested in G0/G1 by contact inhibition and serum depletion and then stimulated to enter cell cycle with platelet-derived growth factor. Here we show for the first time that, in human fibroblasts, cdk2 is translocated from the cytosol into the nucleus in G1-phase. MATERIALS AND METHODS Cell synchronization. FH109 human embryonal lung fibroblasts [15] were seeded to confluence and cultured for 3 days in DMEM (PAA) supplemented with 10% FCS (Gibco) and for an additional 24 h in DMEM/0.5% FCS. The quiescent cells were then stimulated with PDGF-AB (50 ng/ml, Boehringer Mannheim) for 10, 14, 18, and 22 h. Progression through the cell cycle was monitored by flow cytometry and [3H]thymidine incorporation for 30 min, as previously described [16]. Cell extracts and Western blotting. A total of 8 1 105 cells were seeded in 6-well plates and treated as described above. For total cell extracts, cells were solubilized in 500 ml of boiling SDS-sample buffer [17]. Cell fractionation for pRB detection was performed as previously described [16]. For cdk2 and cdk7 detection, nuclei were separated from the cytosol according to [18]. Briefly, cells were washed three times with ice-cold phosphate-buffered saline (PBS) containing 1 mM MgCl2 and harvested with a rubber policeman in 1 ml of icecold hypotonic buffer (10 mM Hepes, pH 7.9, 1.5 mM MgCl2 , 10 mM KCl, 300 mM sucrose, 1 mM EDTA, 1 mM DTT, 1 mM Na3VO4 , 2.5 mM NaF, 0.3 mM aprotinin, 0.1 mM soybean trypsin inhibitor, 60 mM phenylmethylsulfonyl fluoride, 0.4 mM iodoacetamid, 5 mM pepstatin, 4 mM leupeptin). After swelling on ice for 10 min, cells were disrupted by repeated aspiration through a 22-gauge needle. Cell disruption was controlled by microscopic observation. Nuclei were separated from the cytosol by centrifugation at 47C for 10 min at 960g, washed once with hypotonic buffer, and solubilized in SDS– sample buffer. Proteins of the supernatant were precipitated [19] and solubilized in SDS– sample buffer. Protein determination of each fraction was performed according to [20]. After SDS–polyacrylamide gel electrophoresis (SDS –PAGE: 7.5%, 20 mg protein/lane for pRB detection, or 10–20%, 20 mg protein/lane for cdk2 detection, or 12.5%, 30 mg protein/lane for cdk7 detection), proteins were transferred onto Immobilon membrane (Millipore) and blocked for 1 h with 5% milk powder in TBS (10 mM Tris/HCl, pH 8.0, 150 mM NaCl, 0.05% Tween 20). For immunodetection, membranes were incubated for 1.5 h with anti-Rb, anti-cdk2 (each 0.1 mg/ml, Santa Cruz), or anti-cdk7 antibodies (1 mg/ml, Santa Cruz) in blocking buffer followed by alkaline phosphatase-conjugated antirabbit antibody (1:3000, Biozol) for 1.5 h. 4-Nitroblue tetrazolium chloride and 5-bromo-4-chloro-3-indolyl phosphate (both from Boehringer Mannheim) was used as a detection system. Immunoprecipitation. A total of 8 1 105 cells were seeded in 6well plates, cultured, and treated as described. Cells were lysed for 20 min at 47C in RIPA buffer (50 mM Tris/HCl, pH 7.5, 1% sodium deoxycholate, 1% Triton X-100, 0.1% SDS, phosphatase, and protease inhibitors as mentioned) and immunoprecipitated for 4 h at 47C with anti-cyclin E antibodies (10 mg, Santa Cruz) covalently coupled to protein A– Sepharose (12.5 mg, Sigma). The immunoprecipitate was washed three times with RIPA buffer and proteins eluted with 100 ml of diethylamine (100 mM, pH 11.5) for 10 min at room temperature. The supernatant was neutralized with HCl, and the proteins were precipitated [19] and solubilized in SDS–sample buffer. SDS– PAGE and western blotting was performed as described above.
AID ECR 3507
/
6i1d$$$402
03-12-97 20:27:13
73
Immunofluorescence. For this procedure, 106 cells (60-mm dishes) were cultured on glass coverslips as described above. After washing with PBS three times, cells were fixed with acetone at 0207C. The coverslips were air-dried and washed three times with PBS. Unspecific binding was suppressed by blocking with 10% goat serum in PBS for 20 min. After three short and three 5-min washing steps with PBS, cells were incubated for 1 h with anti-cdk2 (0.5 mg/ml) antibodies in PBS containing 1% BSA. The cells were washed in three short and three 5-min washing steps and then incubated for 1 h with FITC-conjugated swine anti-rabbit immunoglobulins (1:20, DAKO) in PBS/0.5% gelatin/5% BSA/10% goat serum. The cells were again washed as described and the coverslips mounted on glass slides with Histogel (Camon) containing 2.5% 1,4-diaminobicyclooctane as an anti-fading agent (Sigma). Controls were treated the same way except that the primary antibody was omitted.
RESULTS
For cell synchronization, FH109 cells were arrested in G0/G1-phase by contact inhibition for 3 days and additionally by serum reduction to 0.5% for 24 h. As assessed by flow cytometric analysis after propidium iodide staining, more than 95% of the fibroblasts showed a 2n DNA content. (Fig. 1A). The quiescent fibroblasts were then exposed to PDGF-AB (50 ng/ml) and cell cycle progression was studied by propidium iodide staining for flow cytometry and [3 H]thymidine incorporation. Figure 1B shows that 10 h after PDGF addition the majority of the cells were still in G1-phase. DNA synthesis started 14–16 h after PDGF exposure with a maximum at 18–20 h (Figs. 1A and 1B). The maximal [3H]thymidine incorporation was 16-fold higher (16.25 { 1.01, n Å 4) in PDGF-treated compared with untreated cells. Cell cycle progression was further reflected by phosphorylation of the retinoblastoma gene product (pRB). It is generally believed that pRB becomes phosphorylated in mid G1- to S-phase. In its hypophosphorylated form pRB binds to the transcription factor E2F and blocks transcription. (Hyper)phosphorylation of pRB leads to its functional inactivation and results in a decreased affinity for the nuclear compartment. We therefore fractionated the cells into a nuclear and a cytosolic compartment (see Materials and Methods) and examined pRB phosphorylation by Western blot analysis after PDGF exposure. We observed a strong increase in the slower migrating forms of pRB (Fig. 1C) in the cytosolic compartment. These 115- to 117-kDa forms represent the hyperphosphorylated species of pRB [31]. We selected the three time points 14, 18, and 22 h after PDGF addition for investigation since then cells had reached G1/S-transition, the peak of S-phase, and late S-phase, respectively. We next examined upstream events of pRB, i.e., cdk2 activation by phosphorylation at Thr 160 and association with cyclin E using Western blot analysis, as the cdk2–cyclin E complex is known to contribute to pRB phosphorylation. In whole-cell extracts, the total
eca
74
DIETRICH ET AL.
FIG. 1. Cell cycle analysis after PDGF stimulation of quiescent FH109 cells. (A) FH109 cells were stained with propidium iodide and analyzed by flow cytometry at different time points after PDGF exposure (1, quiescent cells; 2, 16 h; 3, 20 h; 4, 24 h after PDGF exposure). (B) [3H]thymidine incorporation was measured at the indicated time points (n Å 4). (C) The proteins of FH109 cells were separated into a nuclear and cytosolic fraction at the indicated time points after PDGF stimulation. Western blot analysis was performed with anti-pRB antibodies (pRBppp, hyperphosphorylated pRB). The positions of the molecular size marker proteins are shown on the left.
amount of cdk2 was only slightly enhanced by PDGF, but we observed the appearance of a faster migrating species of cdk2 (Fig. 2, top) which is known to be the phosphorylated (Thr 160), i.e., activated, form of cdk2 [21]. In quiescent cells, this faster migrating form was absent, and maximal phosphorylation (Thr 160) occurred 22 h after PDGF stimulation. Immunoprecipitation studies with anti-cyclin E antibodies (Fig. 2, middle) revealed that a significant amount of cdk2 was associated with cyclin E at the G1/S-transition (14 h) with a maximum in early S-phase (18 h). In late Sphase (22 h), cdk2 is probably associated with cyclin A
AID ECR 3507
/
6i1d$$$402
03-12-97 20:27:13
[18]. Fig. 2, bottom, shows the results of Western blotting with anti-cdk2 antibodies after nuclear extraction. In the nucleus, the amount of faster migrating cdk2 (33 kDa) was strongly enhanced in response to PDGF, already after 10 h, when the fibroblasts were still in G1-phase. Since both the amount of slower migrating, i.e., inactive cdk2 (35 kDa) in the nucleus and the total amount of cdk2 (Fig. 2, top) remained constant throughout the cell cycle, i.e., cdk2 was not synthesized de novo, we suggested a translocation from the cytosol into the nucleus. This seemed to be reasonable, since cdk7, which, in association with cyclin H, is responsible
eca
CELL CYCLE-DEPENDENT SUBCELLULAR DISTRIBUTION OF cdk2
FIG. 2. Cell cycle-dependent phosphorylation of cdk2 (Thr 160) in response to PDGF. Total cell extracts (top), anti-cyclin E immunoprecipitates (middle), and nuclear extracts (bottom) were prepared as described under Materials and Methods at the indicated time points after PDGF-addition. Western blot analysis was performed with anti-cdk2 antibodies (cdk2p, Thr 160-phosphorylated cdk2). The positions of the molecular size marker proteins are shown on the left.
for cdk2 phosphorylation, has been described to be exclusively located in the nucleus [11]. This was confirmed in own experiments using Western blot analysis with anti-cdk7 antibodies after subcellular fractionation. Figure 3 demonstrates that cdk7 is located in the nucleus and that protein levels of cdk7 are invariant throughout the cell cycle. The hypothesis of a nuclear translocation was strengthened by the results of intracellular immunostaining with anti-cdk2 antibodies. In quiescent cells, we observed a diffuse immunostaining (Fig. 4A) which changed in response to PDGF. After 10 h (Fig. 4B), immunofluorescence appeared predominantly in the nucleus with a maximum of concentration in the nucleus 18 h after PDGF stimulation (Fig. 4D). At the end of S-phase (22 h, Fig. 4E), immunofluorescence was again detected in the nucleus and cytosol. We therefore draw the conclusion that cdk2 is translocated from the cytosol in the nucleus in late G1-phase and phosphorylated by cdk7–cyclin H.
75
for orderly progression through the cell cycle. It is now believed that cdc2 is only involved in G2/M-transition being activated by cyclin B [22]. Progression from G1- to S-phase, i.e., initiation and regulation of DNA replication, is strongly dependent on cdk2 [23, 24]. In late G1- and early S-phase, cdk2 is activated by the association of cyclin E [25] and then by cyclin A when S-phase is sustained [26]. It is generally accepted that cdk2–cyclin E— together with cdk4– cyclin D— is responsible for phosphorylation of the retinoblastoma gene product. The hypophosphorylated species of pRB binds to the transcription factor E2F, hence inhibiting transcription. (Hyper)phosphorylation of pRB leads to its functional inactivation, resulting in the loss of binding to E2F and thus permitting entry into S-phase [27, 28]. When cyclin E is degraded and pRB released from E2F, cdk2–cyclin A binds to E2F, thereby regulating DNA replication [for review see 29]. Pagano and co-workers [18] have recently reported that cdk2 is absent in quiescent human fibroblasts and appears as a nuclear protein in S-phase. Activity of cdk2 corresponded largely with cyclin A activity. Since we have revealed that cdk2 is distributed in the cytosol or nucleus in asynchronously proliferating human fibroblasts (unpublished data), we suggested that cdk2 is translocated to the nucleus already in G1-phase when it is associated with cyclin E. We therefore studied subcellular distribution of cdk2 in synchronized, PDGF-stimulated human fibroblasts, FH109 cells. After contact inhibition and additional serum depletion, more than 95% of the FH109 cells were arrested in G0/G1-phase. The quiescent fibroblasts could be restimulated to enter the cell cycle by exposure to PDGF. Late G1-phase was reached 10 h after PDGF treatment, and S-phase began 14–16 h after PDGF exposure with a peak of DNA synthesis at 18 –20 h. Progres-
DISCUSSION
In yeast, a single 34-kDa cyclin-dependent kinase (called cdc2 in Schizosaccharomyces pombe and CDC28 in Saccharomyces cerevisiae) is the key regulator for cell proliferation. In mammalian cells, cell cycle regulation is more complex and different cyclin-dependent kinases in combination with different cyclins that are activated in a finely balanced sequence are required
AID ECR 3507
/
6i1d$$$402
03-12-97 20:27:13
FIG. 3. Subcellular distribution of cdk7 and protein levels after PDGF stimulation. FH109 cell lysates were separated into a nuclear and a cytosolic fraction at the indicated time points after PDGF exposure. Western blot analysis was performed with anti-cdk7 antibodies. The positions of the molecular size marker proteins are shown on the left.
eca
76
03-12-97 20:27:13
eca
CELL CYCLE-DEPENDENT SUBCELLULAR DISTRIBUTION OF cdk2
sion through the cell cycle was monitored by an increase in pRB phosphorylation in late G1- and S-phase, which resulted in a clear shift to lower electrophoretic mobility [28]. In accordance with the work of Mittnacht and Weinberg [27], the hyperphosphorylated species of pRB had a decreased affinity for the nuclear compartment. It is generally accepted that the cdk2–cyclin E complex— in collaboration with cdk4–cyclin D —is responsible for pRB phosphorylation. With respect to our immunoprecipitation studies with anti-cyclin E antibodies, we conclude that from mid G1- to early S-phase a significant amount of cdk2 was associated with and thereby regulated by cyclin E. Cyclin E itself is known to be synthesized in late G1- and early S-phase [25], hence regulating cdk2 activity. To be active, cdk2 has to be additionally phosphorylated at a conserved Thr residue, Thr 160 (and dephosphorylated at Thr 14 and Tyr 15). Concomitant with pRB phosphorylation, a faster migrating species of cdk2 occurred. This form represented the activated form of cdk2 phosphorylated at Thr 160 [16, 21]. We therefore draw the conclusion that in late G1- to early S-phase cdk2 was activated by cyclin E association and Thr phosphorylation, hence leading to phosphorylation of pRB and finally to initiation of DNA synthesis. As mentioned above, activation of cdk2 requires phosphorylation at Thr 160 by CAK. CAK itself is composed of a catalytic and a regulatory subunit: cdk7, a nuclear protein, and cyclin H. Since both the protein level of cdk7 and the activity of CAK remain constant throughout the cell cycle (Fig. 3; [11, 30]), phosphorylation by CAK might be regulated by the substrate availability of cdk2, which would require translocation from the cytosol into the nucleus. We next examined subcellular distribution of cdk2 and cdk7 using Western blot analysis after subcellular fractionation and immunofluorescence microscopy. In the nuclear extracts, we observed a similar increase in the faster migrating, i.e., activated form of cdk2 compared with total cell extracts. Since the slower migrating, i.e., inactive form remained constant in the nucleus and cdk2 was not synthesized de novo, we suggested that a significant amount of cdk2 had to be translocated from the cytosol to the nucleus to become phosphorylated. Indeed, using immunofluorescence microscopy, we observed that cdk2 was distributed diffusely in the cytosol and the nucleus in quiescent fibroblasts. Already 10 h after PDGF treatment, a time point when the majority of the fibroblasts were still in G1-phase, there was a strong increase in nuclear immunostain-
77
ing, whereas immunoreactivity in the cytosol disappeared. Maximal immunofluorescence was observed in S-phase exclusively in the nucleus, declining at the end of S-phase. Our findings are in contrast with the work of Pagano et al. [18], who failed to detect cdk2 immunoreactivity in quiescent fibroblasts and during G1phase. Since they used a different anti-cdk2 antibody, the most likely explanation is that our anti-cdk2 antibody is more sensitive in immunofluorescence studies. Conclusively, our data suggest a translocation of cdk2 from the cytosol to the nucleus already before S-phase, which, to our knowledge, has not been described so far in human fibroblasts. Only recently, Brenot-Bosc and co-workers have reported a similar translocation of cdk2 before S-phase in nocodazole-blocked MANCAcells, a B-cell lymphoma line [31]. As has been shown by Tassan and co-workers [11], cdk7 was found to be predominantly located in the nucleus, and protein levels of cdk7 were constant throughout the cell cycle. The small amount of cdk7 in the cytosol is probably an artifact due to the experimental procedure, since in control experiments we found similar amounts of the nuclear protein cyclin D in the cytosolic fraction (unpublished data). However, first studies using immunofluorescence techniques revealed a cell cycle-dependent change in the nuclear compartimentalization of cdk7. cdk7 showed a granular pattern in mid to late G1 (10 –14 h after PDGF treatment) that was absent in quiescent or S-phase cells (unpublished data) and coincided with the appearance of cdk2 phosphorylation. Conclusively, cdk7 activity might not be regulated only by the translocation of its substrate cdk2 to the nucleus, but also by an intranuclear translocation to specific, still unknown structures. This work was supported by a grant of the Deutsche Forschungsgemeinschaft, WI 727/3-1, and is part of the Ph.D. thesis of K.W.
REFERENCES 1. Sherr, C. J. (1994) Cell 79, 551– 555. 2. Lees, E. (1995) Curr. Opin. Cell Biol. 7, 773– 780. 3. Peter, M., and Herskowitz, I. (1994) Cell 79, 181– 184. 4. Matsushime, H., Quelle, D. E., Shurtleff, S. A., Shibuya, M., Sherr, C. J., and Kato, J. (1994) Mol. Cell. Biol. 14, 2066 –2076. 5. Meyerson, M., and Harlow, E. (1994) Mol. Cell. Biol. 14, 2077 – 2086. 6. Dulic, V., Lees, E., and Reed, S. (1992) Science 257, 1958 –1061. 7. Koff, A., Giordano, A., Desai, D., Yamashita, K., Harper, W., Elledge, S., Nishimoto, T., Morgan, D., Franza, R., and Roberts, J. (1992) Science 257, 1689 – 1693.
FIG. 4. Cell cycle-dependent subcellular distribution of cdk2 after PDGF stimulation. FH109 cells were stained with anti-cdk2 antibodies before (A) or 10 h (B), 14 h (C), 18 h (D), or 22 h (E) after PDGF addition. (left: original magnification, 1180; bar, 50 mm; right: original magnification, 1720; bar, 10 mm).
AID ECR 3507
/
6i1d$$$402
03-12-97 20:27:13
eca
78
DIETRICH ET AL.
8.
Hatakeyama, M., Brill, J. A., Fink, G. R., and Weinberg, R. (1994) Genes Dev. 8, 1759 –1771.
9.
King, R., Jackson, P., and Kirschner, M. (1994) Cell 79, 563– 571.
21.
10. Fisher, R. P., and Morgan, D. (1994) Cell 78, 713– 724.
22.
11. Tassan, J.-P., Schultz, S. J., Bartek, J., and Nigg, E. A. (1994) J. Cell Biol. 127, 467–478.
23.
12. Pines, J., and Hunter, T. (1991) J. Cell Biol. 115, 1– 17. 13. Baldin, V., Lukas, J., Marcote, M. J., Pagano, M., and Draetta, G. (1993) Genes Dev. 7, 812– 821.
24. 25.
14. Ohtsubo, M., Theodoras, A. M., Schumacher, J., Roberts, J., and Pagano, M. (1995) Mol. Cell. Biol. 15, 2612 –2624. 15. Wieser, R., Heck, R., and Oesch, F. (1985) Exp. Cell Res. 158, 493–499.
26.
16. Dietrich, C., Bartsch, T., Schanz, F., Oesch, F., and Wieser, R. (1996) Proc. Natl. Acad. Sci. USA 93, 10815 –10819.
27. 28.
17. Laemmli, K. K. (1970) Nature 227, 680–685. 18. Pagano, M., Pepperkok, R., Lukas, J., Baldin, V., Ansorge, W., Bartek, J., and Draetta, G. (1993) J. Cell Biol. 121, 101 –111.
29. 30.
19. Wessel, D., and Flu¨gge, U. J. (1984) Anal. Biochem. 138, 141– 143. 20. Smith, P. K., Krohn, R. I., Hermanson, G. T., Mallia, A. K., Gartner, F. H., Provenzano, M. D., Fujimoto, E. K., Golke, N. M.,
31.
Olson, B. J., and Klenk, D. C. (1985) Anal. Biochem. 150, 76– 85. Gu, Y., Rosenblatt, J., and Morgan, D. (1992) EMBO J. 11, 3995 –4005. Schwob, E., Bo¨hm, T., Mendenhall, M. D., and Nasmyth, K. (1994) Cell 79, 233–244. Koff, A., Ohtsuki, M., Polyak, K., Roberts, J. M., and Massague, J. (1993) Science 260, 536– 539. Ohtsubo, M., Theodoras, A. M., Schumacher, J., Roberts, J., and Pagano, M. (1995) Mol. Cell. Biol. 15, 2612 – 2624. Koff, A., Giordano, A., Desai, D., Yamashita, K., Harper, W., Elledge, S., Nishimoto, T., Morgan, D. O., Franza, B. R., and Roberts, J. (1992) Science 257, 1689 –1694. Pagano, M., Pepperkok, R., Verde, F., Ansorge, W., and Draetta, G. (1992) EMBO J. 11, 961–971. Mittnacht, S., and Weinberg, R. A. (1991) Cell 65, 381–393. DeCaprio, J. A., Ludlow, J. W., Lynch, D., Furukawa, Y., Griffin, J., Piwnica-Worms, H., Huang, C.-M., and Livingston, D. M. (1989) Cell 58, 1085 –1095. Moran, E. (1991) Gene Expression 1, 281–283. Labbe, J.-C., Martinez, A.-M., Fesquet, D., Capony, J.-P., Darbon, J.-M., Derancourt, J., Devault, A., Morin, N., Cavadore, J.-C., and Marcel, D. (1994) EMBO J. 13, 5155 –5164. Brenot-Bosc, F., Gupta, S., Margolis, R. L., and Fotedar, R. (1995) Chromosoma 103, 517 –527.
Received August 12, 1996 Revised version received January 13, 1997
AID ECR 3507
/
6i1d$$$403
03-12-97 20:27:13
eca