Portal de Periódicos da CAPES

Induction of cyclin E-cdk2 kinase activity, E2F-dependent transcription and cell growth by Myc are genetically separable events

2000; Springer Nature; Volume: 19; Issue: 21 Linguagem: Inglês

10.1093/emboj/19.21.5813

1460-2075

Rudolf Beier, Andrea Bürgin, Astrid Kiermaier, Matthew L. Fero, Holger Karsunky, Rainer Saffrich, Tarik Möröy, Wilhelm Ansorge, Jim Roberts, Martin Eilers,

Ubiquitin and proteasome pathways

Article1 November 2000free access Induction of cyclin E–cdk2 kinase activity, E2F-dependent transcription and cell growth by Myc are genetically separable events Rudolf Beier Rudolf Beier Institute of Molecular Biology and Tumour Research, Emil-Mannkopff-Straße 2, 35033 Marburg, Germany Search for more papers by this author Andrea Bürgin Andrea Bürgin Institute of Molecular Biology and Tumour Research, Emil-Mannkopff-Straße 2, 35033 Marburg, Germany Search for more papers by this author Astrid Kiermaier Astrid Kiermaier Institute of Molecular Biology and Tumour Research, Emil-Mannkopff-Straße 2, 35033 Marburg, Germany Search for more papers by this author Matthew Fero Matthew Fero Division of Basic Sciences, Howard Hughes Medical Institute, Fred Hutchinson Cancer Research Center, Seattle, WA, USA Search for more papers by this author Holger Karsunky Holger Karsunky Institute for Cell Biology, University of Essen, Virchowstrasse 173, 45122 Essen, Germany Search for more papers by this author Rainer Saffrich Rainer Saffrich Biological Instrumentation Programme, EMBL, Meyerhofstraße 1, 69117 Heidelberg, Germany Search for more papers by this author Tarik Möröy Tarik Möröy Institute for Cell Biology, University of Essen, Virchowstrasse 173, 45122 Essen, Germany Search for more papers by this author Wilhelm Ansorge Wilhelm Ansorge Biological Instrumentation Programme, EMBL, Meyerhofstraße 1, 69117 Heidelberg, Germany Search for more papers by this author Jim Roberts Jim Roberts Division of Basic Sciences, Howard Hughes Medical Institute, Fred Hutchinson Cancer Research Center, Seattle, WA, USA Search for more papers by this author Martin Eilers Corresponding Author Martin Eilers Institute of Molecular Biology and Tumour Research, Emil-Mannkopff-Straße 2, 35033 Marburg, Germany Search for more papers by this author Rudolf Beier Rudolf Beier Institute of Molecular Biology and Tumour Research, Emil-Mannkopff-Straße 2, 35033 Marburg, Germany Search for more papers by this author Andrea Bürgin Andrea Bürgin Institute of Molecular Biology and Tumour Research, Emil-Mannkopff-Straße 2, 35033 Marburg, Germany Search for more papers by this author Astrid Kiermaier Astrid Kiermaier Institute of Molecular Biology and Tumour Research, Emil-Mannkopff-Straße 2, 35033 Marburg, Germany Search for more papers by this author Matthew Fero Matthew Fero Division of Basic Sciences, Howard Hughes Medical Institute, Fred Hutchinson Cancer Research Center, Seattle, WA, USA Search for more papers by this author Holger Karsunky Holger Karsunky Institute for Cell Biology, University of Essen, Virchowstrasse 173, 45122 Essen, Germany Search for more papers by this author Rainer Saffrich Rainer Saffrich Biological Instrumentation Programme, EMBL, Meyerhofstraße 1, 69117 Heidelberg, Germany Search for more papers by this author Tarik Möröy Tarik Möröy Institute for Cell Biology, University of Essen, Virchowstrasse 173, 45122 Essen, Germany Search for more papers by this author Wilhelm Ansorge Wilhelm Ansorge Biological Instrumentation Programme, EMBL, Meyerhofstraße 1, 69117 Heidelberg, Germany Search for more papers by this author Jim Roberts Jim Roberts Division of Basic Sciences, Howard Hughes Medical Institute, Fred Hutchinson Cancer Research Center, Seattle, WA, USA Search for more papers by this author Martin Eilers Corresponding Author Martin Eilers Institute of Molecular Biology and Tumour Research, Emil-Mannkopff-Straße 2, 35033 Marburg, Germany Search for more papers by this author Author Information Rudolf Beier1, Andrea Bürgin1, Astrid Kiermaier1, Matthew Fero2, Holger Karsunky3, Rainer Saffrich4, Tarik Möröy3, Wilhelm Ansorge4, Jim Roberts2 and Martin Eilers 1 1Institute of Molecular Biology and Tumour Research, Emil-Mannkopff-Straße 2, 35033 Marburg, Germany 2Division of Basic Sciences, Howard Hughes Medical Institute, Fred Hutchinson Cancer Research Center, Seattle, WA, USA 3Institute for Cell Biology, University of Essen, Virchowstrasse 173, 45122 Essen, Germany 4Biological Instrumentation Programme, EMBL, Meyerhofstraße 1, 69117 Heidelberg, Germany *Corresponding author. E-mail: [email protected] The EMBO Journal (2000)19:5813-5823https://doi.org/10.1093/emboj/19.21.5813 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info The c-myc gene has been implicated in three distinct genetic programs regulating cell proliferation: control of cyclin E–cdk2 kinase activity, E2F-dependent transcription and cell growth. We have now used p27−/− fibroblasts to dissect these downstream signalling pathways. In these cells, activation of Myc stimulates transcription of E2F target genes, S-phase entry and cell growth without affecting cyclin E–cdk2 kinase activity. Both cyclin D2 and E2F2, potential direct target genes of Myc, are induced in p27−/− MycER cells. Ectopic expression of E2F2, but not of cyclin D2, induces S-phase entry, but, in contrast to Myc, does not stimulate cell growth. Our results show that stimulation of cyclin E–cdk2 kinase, of E2F-dependent transcription and of cell growth by Myc can be genetically separated from each other. Introduction The proto-oncogene c-myc encodes a transcription factor of the helix–loop–helix/leucine zipper family of proteins. Expression of c-myc is frequently enhanced in human tumours, either due to mutations in the c-myc gene itself or due to mutations in signalling pathways that control c-myc expression, such as mutations of the APC gene (He et al., 1998). One possible explanation for this frequent deregulation is the unusual ability of Myc protein to interfere with genetic programs that control cellular proliferation (Obaya et al., 1999). For example, activation of conditional alleles of Myc (Eilers et al., 1989; Littlewood et al., 1995) can be sufficient to induce growth factor-independent proliferation in established rodent cell lines (Eilers et al., 1991). c-myc has been implicated in at least three distinct genetic pathways controlling progression through the G1 phase. First, activation of Myc rapidly induces activation of cyclin E–cdk2 kinase and loss of p27kip1 from cdk2 complexes (Steiner et al., 1995; Santoni-Rugiu et al., 2000); this is required for Myc to promote proliferation (Rudolph et al., 1996). Conversely, levels of p27kip1 are elevated and cyclin E–cdk2 kinase activity is suppressed in cells expressing dominant-negative alleles of Myc and in c-myc−/− fibroblasts (Berns et al., 1997; Mateyak et al., 1999). In murine fibroblasts expressing Myc, p27kip1 is initially sequestered in cyclin D–cdk4 complexes (Bouchard et al., 1999; Perez-Roger et al., 1999); the direct transcriptional induction of the cyclin D2 and, potentially, of the cdk4 gene by Myc contributes to the formation of these complexes (Bouchard et al., 1999; Coller et al., 2000; Hermeking et al., 2000). In primary cells or in response to TGF-β signalling, suppression of the p15ink4b gene by Myc further contributes to sequestration of p27 (Warner et al., 1999). Subsequently, p27 is degraded (Müller et al., 1997) and the transcriptional stimulation of the Cul-1 gene by Myc has been implicated in this process (O'Hagan et al., 2000b). As a result, expression of Myc suppresses the growth inhibitory function of physiological levels of p27kip1 (Vlach et al., 1996). Similarly, suppression of the expression of p27kip1 by antisense oligonucleotides can induce proliferation in serum-starved fibroblasts in a manner reminiscent of induction of MycER (Coats et al., 1996). Together, these findings have led to the hypothesis that Myc controls G1 progression by suppressing the function of p27kip1 (Amati et al., 1998). Secondly, activation of Myc induces the transcriptional activity of the E2F/DP family of transcription factors (Jansen-Dürr et al., 1993; Leone et al., 1997). Transcriptional activation by E2F/DP family proteins is inhibited by association with members of the pocket protein family. Since the interaction between E2F/DP proteins and pocket proteins is controlled by cdk- dependent phosphorylation, activation of E2F-dependent transcription may be an indirect consequence of regulation of cdk activity by Myc. Alternatively, Myc has been shown to activate the promoter of the E2F2 gene, and constitutive expression of Myc upregulates expression of E2F2 protein (Sears et al., 1997, 1999). Therefore, induction of E2F transcriptional activity may be an independent and direct effector pathway of Myc. Thirdly, Myc has also been implicated in cell growth, i.e. an increase in cell mass, both in Drosophila (Johnston et al., 1999) and in mammalian cells (Rosenwald et al., 1993; Mateyak et al., 1997; Iritani and Eisenman, 1999). An involvement of Myc in cell growth had also been suspected from the identity of putative target genes of Myc, many of which appear to control cellular metabolism and protein translation rather than cell cycle progression per se (Johnston et al., 1998; Dang, 1999; Coller et al., 2000; Greasley et al., 2000; O'Hagan et al., 2000a). Together, the findings have prompted the question as to whether only one pathway is the primary effector pathway downstream of Myc, with other events occurring as secondary consequences of Myc-induced proliferation. Now, we report experiments aimed at genetically dissecting signalling pathways downstream of Myc. Our results allow the conclusion that activation of cyclin E–cdk2 kinase, induction of E2F-dependent transcription and stimulation of cell growth define genetically separable pathways by which Myc promotes G1 progression. Results Immortalization and induction of apoptosis by Myc are independent of p27 Previous work had linked the ability of Myc to stimulate G1 progression to its ability to antagonize the cell cycle arrest by physiological levels of p27kip1 (Amati et al., 1998). In order to determine whether loss of p27kip1 affects Myc function in primary cells, mouse embryo fibroblasts (MEFs) were isolated from both p27+/+ and p27−/− embryos. The genotype of the embryos was verified by Southern blotting and by western blotting of cellular extracts with antibodies directed against murine p27 (not shown). Cells were plated and infected in parallel with retroviruses expressing either a resistance gene alone (pbabe-puro; Morgenstern and Land, 1990) or together with a human c-myc cDNA. Western blots confirmed that both p27+/+ and p27−/− cells expressed human Myc protein at equal levels when infected with the corresponding virus (Figure 1A). When infected cells were plated at low density in selective medium, colonies emerged after 10–14 days. The number of colonies was greatly enhanced in plates infected with a virus expressing c-myc relative to vector control cells. In several independent experiments, expression of Myc stimulated colony formation in both p27+/+ and p27−/− cells to a similar extent (Figure 1A). Pools of infected p27−/− and p27+/+ cells proliferated with no sign of senescence upon infection with retroviruses expressing Myc, but arrested and showed a senescent phenotype (not shown) upon infection with control viruses (Figure 1B). In the absence of serum, both p27+/+ and p27−/− cells expressing c-myc underwent apoptosis within 48 h (Figure 1C). FACScan experiments confirmed that DNA fragmentation took place in ∼70% of both p27+/+ and p27−/− cells expressing c-myc (not shown). We concluded that primary p27−/− and p27+/+ cells were indistinguishable with respect to immortalization and induction of apoptosis by Myc. Figure 1.Immortalization and induction of apoptosis by Myc in primary p27+/+ and p27−/− MEFs. (A) Number of colonies growing upon infection of either p27−/− or p27+/+ cells with control ('pbabe') or Myc-expressing virus ('pbabe-Myc'), after selection with puromycin and incubation for 2 weeks. The graph shows a quantitation of a representative experiment. The western blot documents equal expression of Myc proteins in p27−/− and p27+/+ cells. (B) Growth curve of pools of p27+/+ and p27−/− cells recovered after infection with the indicated viruses; 5 × 104 drug-resistant cells were plated at the start of the experiment. (C) Induction of apoptosis by Myc. p27+/+ and p27−/− MEFs, recovered after infection with the indicated viruses and selection, were plated in the absence or presence of serum; photographs were taken 48 h after plating. Download figure Download PowerPoint Activation of cyclin E–cdk2 kinase by Myc depends on p27 The extensive fragmentation of DNA that occurred in primary cells infected with Myc upon serum starvation precluded a detailed analysis as to whether Myc stimulated cell cycle progression under these conditions. Therefore, cell lines were established from both p27−/− and p27+/+ cells using a standard 3T3 protocol. Both lines that were established maintained expression of p19ARF; however, in contrast to primary cells both p27−/− and p27+/+ 3T3 cells constitutively expressed high levels of p53 that were independent of DNA damage induced by adriamycin. These findings indicate that both lines had undergone a mutation in the p53 gene during immortalization (Figure 2A). Figure 2.Induction of cell cycle progression and cell growth by Myc. (A) Western blots documenting DNA-damage-independent expression of p53 in p27−/− and p27+/+ 3T3 cell lines and two MycER clones, and expression of p19ARF in MycER clones. Where indicated, cells were treated with 0.5 μg/ml adriamycin for the indicated periods of time (24 h in the upper panel). (B) Percentage of cells incorporating BrdU in p27−/− and p27+/+ MycER cell lines. Cells were serum starved for 48 h before addition of either 200 nM 4-OHT or 10% FCS. The percentage of cells incorporating BrdU was determined 20 h later. (C) Cell cycle distribution of the indicated cell lines. Cells were starved for 48 h before re-induction. Samples were taken at the indicated time points after addition of either 200 nM 4-OHT or 10% FCS. (D) FSC profiles indicating growth of p27−/− and p27+/+ MycER cells after activation of Myc or addition of FCS for 24 h. Download figure Download PowerPoint p27+/+ and p27−/− 3T3 lines were infected with retroviruses expressing MycER proteins (Eilers et al., 1989; Littlewood et al., 1995). Both pools and clones were analysed for expression of MycER protein by western blotting. Clones and pools with detectable expression were expanded and subsequently analysed (see below). More detailed data were obtained first from one clonal p27+/+ and one clonal p27−/− MycER cell line. BrdU (5-bromo-2-deoxyuridine) labelling was used to determine whether the ability of Myc to promote progression into S phase depends on p27. As described previously (Coats et al., 1999), both p27+/+ and p27−/− cells accumulate in G0 upon growth factor starvation (Figure 2B) and can be re-induced to enter the cell cycle upon addition of growth factors. In some experiments, we noted a higher percentage of cells incorporating BrdU in p27−/− cells relative to p27+/+ cells upon serum stimulation. Activation of Myc by addition of 4-hydroxy-tamoxifen (4-OHT) induced S-phase entry to a similar extent in both cell types (Figure 2B); time course experiments revealed no significant difference in the length of G1 and S phase (not shown). To determine whether stimulated cells progressed to mitosis, FACScan profiles were recorded at different time points after stimulation with either serum (10% fetal calf serum, FCS) or 4-OHT (Figure 2C). Both p27+/+ and p27−/− cells efficiently progressed into S and G2 phase when stimulated by addition of serum; counting of mitotic figures from 4′,6-diamidine-2-phenylindole (DAPI)-stained preparations revealed a relatively synchronous passage through mitosis 20 h after stimulation (not shown). In contrast, cells stimulated by addition of 4-OHT efficiently progressed into S phase but did not progress through mitosis (Figure 2C), and no significant increase in the number of mitotic figures was seen in either p27+/+ or p27−/− MycER cells (not shown). Analysis of forward scatter (FSC) profiles to determine cell size from the same samples showed that addition of either FCS or 4-OHT also promoted growth of both p27+/+ and p27−/− cells (Figure 2D). Activation of Myc also promoted cell growth in the presence of either hydroxyurea or aphidicolin, whereas addition of either drug effectively blocked Myc-induced DNA synthesis (see Supplementary data, available at The EMBO Journal Online). The findings confirm previous observations that activation of Myc can promote cell growth in the absence of overt cell cycle progression (Schuhmacher et al., 1999). Finally, we noted that activation of Myc causes an increase in cell size in all phases of the cell cycle (see below, Figure 6F), similar to observations made in B cells of mice harbouring an Ig-Myc transgene (Iritani and Eisenman, 1999). Figure 3.Failure of Myc to activate cyclin E-dependent kinase in p27−/− cells. (A) Autoradiogram of cyclin E-dependent kinase assays. p27+/+ and p27−/− MycER cells were serum starved for 48 h and re-stimulated by addition of either 4-OHT or FCS as above. Samples were taken at the indicated time points and cyclin E-kinase activity was determined using histone H1 as substrate. (B) Quantitation of cyclin E-kinase activity (mean values from two independent experiments). (C) Western blots documenting the amount of cyclin E and p130 proteins, and cyclin E–p130 complexes in serum-starved p27+/+ and p27−/− MycER cells after induction of Myc. Samples were taken at the indicated time points. (D) Western blots documenting expression of cdk2 and cyclin A and quantitation of cyclin E-dependent kinase (relative to non-induced cells) in pools of p27−/− and p27+/+ MycER cells and in several independent clones of p27−/− MycER cells. Download figure Download PowerPoint Figure 4.Activation of E2F-dependent genes by Myc is independent of p27. (A) RT–PCR documenting the expression of target genes of E2F. Total RNA was prepared from p27−/− and p27+/+ MycER cells after stimulation with FCS or 4-OHT, and subjected to RT–PCR analysis as detailed in Materials and methods. (B) Percentage of cyclin A-positive cells after microinjection of expression plasmids encoding the indicated proteins. Cells were serum starved for 48 h before microinjection. Four hours later, cells were re-stimulated by the addition of 4-OHT. Cyclin A expression was detected by immunofluorescence 20 h later. (C) Western blot documenting expression of cyclin A in p27−/− MycER cells at the indicated time points after the addition of 200 nM 4-OHT to serum-starved cells. Where indicated, 25 μM roscovitine was added together with 4-OHT. Download figure Download PowerPoint Cyclin E–cdk2 kinase complexes are inactive in primary serum-starved p27−/− fibroblasts and are activated in response to addition of serum (Coats et al., 1999). In agreement with these observations, little kinase activity was detected in cyclin E immunoprecipitates from serum-starved p27+/+ or p27−/− MycER cells (Figure 3A). Re-addition of serum induced kinase activity in both cell lines to a similar extent. In contrast, addition of 4-OHT activated cyclin E–cdk2 kinase only in p27+/+, but not in p27−/− cells. A quantitation of the results of two representative experiments is shown in Figure 3B. Figure 5.Activation of cyclin A–cdk2 kinase activity by Myc in p27−/− MycER cells. (A) Autoradiogram of cdk2 kinase assays. p27+/+ and p27−/− MycER cells were serum starved for 48 h and re-stimulated by the addition of either 4-OHT or FCS as before. Samples were taken at the indicated time points and cdk2-kinase activity was determined using histone H1 as substrate. Under these conditions, cyclin A regulates most of the cdk2 activity. (B) Quantitation of cdk2-kinase activity (mean values from two independent experiments). (C) Western blots documenting the amount of p130, cyclin A and cdk2 in p27−/− MycER cells after stimulation with either 4-OHT or FCS as indicated. Cellular lysates were either left untreated (−) or depleted three times with α-cdk2 antibodies (+) before loading on the gel. (D) Western blots documenting the amount of p130 and of cyclin A in p27−/− MycER cells after stimulation with either 4-OHT or FCS as indicated. Cellular lysates were either left untreated (−) or depleted three times with α-p130 antibodies (+) before loading on the gel. Download figure Download PowerPoint In serum-starved primary p27−/− cells, inhibition of cyclin E-dependent kinase activity is due to complex formation with the pocket protein p130 (Coats et al., 1999). Indeed, significant amounts of cyclin E–p130 complexes were detected in serum-starved p27−/− but not in p27+/+ MycER cells (Figure 3C); immunoprecipitations with α-p130 antibodies showed that such complexes were inactive both before and after activation of Myc (not shown). Cyclin E–p130 complexes remained stable after induction of Myc in p27−/− MycER cells (Figure 3C). Consistent with a view in which Myc specifically antagonizes association of p27kip1 with cyclin E, activation of Myc actually promoted the formation of a small amount of cyclin E–p130 complexes in p27+/+ MycER (Figure 3C) and in RAT1 MycER cells (not shown). We noted that in the first clones picked for analysis, levels of p130 appeared higher in the p27−/− than in the p27+/+ cells. However, analysis of the pools and clones, shown in Figure 3D, revealed that there was not a systematic difference in p130 protein levels between p27−/− and p27+/+ cells (data not shown). Bandshift experiments showed that a fraction of complexes containing cyclin E, cdk2 and p130 were bound to E2F proteins in serum-starved p27−/− MycER cells, whereas no cyclin E–cdk2–p130–E2F complexes were detectable in serum-starved p27+/+ MycER cells (not shown). Activation of Myc did not alter the composition of E2F complexes in p27−/− MycER cells (as determined in bandshift analysis), demonstrating that activation of Myc did not antagonize association of p130 with either cyclin E–cdk2 complexes or E2F (not shown). Consistent with this view, expression of elevated levels of p130 by microinjection inhibited cell cycle induction by Myc in p27−/− cells; analysis of mutants of p130 (Castano et al., 1998) revealed that binding to both E2F and cdk2 contributed to inhibition (not shown). The data show that Myc can promote activation of cell cycle progression in the absence of activation of cyclin E–cdk2 kinase in p27−/− cells. To exclude the possibility that this phenotype is limited to the specific clones analysed, we tested both pools of p27+/+ and p27−/− MycER cells, and several additional clones of p27−/− MycER cells that expressed varying levels of MycER proteins. In all cases, activation of Myc upregulated expression of cyclin A, a marker protein of cell cycle progression (see below). However, cyclin E kinase was upregulated only in pools of p27+/+ MycER cells, but neither in pools nor in independent clones of p27−/− MycER cells (Figure 3D), demonstrating that the failure of Myc to induce cyclin E–cdk2 kinase activity is indeed due to the lack of p27. Recently, a second cyclin E(2) has been identified (Lauper et al., 1998; Gudas et al., 1999); thus, activation of cyclin E2–cdk2 kinase might substitute for cyclin E–cdk2 kinase in p27−/− cells. However, in several experiments only low cyclin E2-dependent kinase activity was found that was induced upon addition of serum, but not significantly altered in response to activation of MycER (not shown). We concluded that Myc promotes cell cycle progression and cell growth in the absence of cyclin E–cdk2 activation in p27−/− cells. Finally, we noted that ectopic expression of constitutive Myc induced extensive apoptosis in pools of serum-starved p27−/− and p27+/+ 3T3 cells, and that expression levels of Myc in stable cell lines derived from such pools were very low. These findings precluded a clear analysis of the mitogenic effects of constitutive Myc in these cell lines. Induction of target genes of E2F by Myc is independent of p27 In order to understand how activation of Myc promotes cell cycle progression in the absence of activation of cyclin E–cdk2 kinase, we measured the expression of target genes of E2F after activation of Myc by reverse transcription PCR (RT–PCR) assays (Vigo et al., 1999). In both p27+/+ and p27−/− MycER cells, we observed induction of multiple target genes of E2F after activation of Myc and no significant difference was observed between both cell lines (Figure 4A). Induction of cyclin A expression was confirmed by western blotting (see below). The results demonstrate that activation of Myc can induce transcription of E2F target genes in the absence of induction of cyclin E–cdk2 kinase. A previous analysis of the cyclin A promoter had shown that induction of cyclin A expression by Myc is mediated via regulation of E2F activity (Rudolph et al., 1996). Figure 6.Ectopic expression of cyclin D2 and of E2F2 in p27−/− MycER cells. (A) Western blots documenting expression of cyclin D2 (left) and of E2F2 (right) in pools of p27−/− MycER cells after infection with either control retroviruses (top), or retroviruses encoding cyclin D2 (lower left) or E2F2 (lower right). Cells were either growing exponentially or serum starved for 24 or 48 h. After 48 h, 200 nM 4-OHT was added and samples were harvested after the indicated times. (B) Western blot documenting the expression of cyclin A in p27−/− MycER control infected cells (top), and in cells expressing cyclin D2 (middle) or E2F2 (bottom) constitutively. (C) BrdU incorporation. The indicated cells were serum starved for 48 h before addition of 4-OHT. BrdU incorporation was measured 20 h later. (D) Cell cycle distribution as determined by FACScan of the indicated p27−/− MycER cell lines after addition of 4-OHT. (E) Top: FSC profiles of p27−/− MycER/cyclin D2 and of p27−/− MycER/E2F2 cells, relative to control cells, after 48 h of serum deprivation. Bottom: FSC profiles of serum-deprived p27−/− MycER/cyclin D2 (left) and of p27−/− MycER/E2F2 (right) cells after addition of 4-OHT, relative to non-induced control cells. Cells were deprived of serum for 48 h before addition of 4-OHT; samples were analysed after 20 h. (F) FSC profiles of cells gated for individual cell cycle phases. p27−/− MycER/E2F2 cells (as in E) were serum starved (control) and MycER was activated by addition of 4-OHT for 20 h. Download figure Download PowerPoint In RAT1 cells, both cdk2 and cdk4 kinase activities are required for Myc to activate cyclin A expression (Rudolph et al., 1996). Since p27 is sequestered by cdk4 complexes in response to activation of Myc, and since p27 is a substrate of cdk2, loss of p27 might influence the requirement for either cdk2 or cdk4 kinase activity. To test this, we microinjected expression plasmids encoding p16ink4a, pRb or non-phosphorylatable alleles of pRb into p27+/+ and p27−/− MycER cells (Figure 4B). In both cell lines, expression of any protein inhibited induction of cyclin A expression by Myc, demonstrating that cdk4 kinase activity is required for Myc to activate cyclin A expression. Similar experiments using dominant-negative alleles of cdk2 yielded equivocal results in p27−/− cells (not shown). Therefore, we used roscovitine, a specific chemical inhibitor of cdk2 kinase (Alessi et al., 1998). Addition of roscovitine abolished induction of cyclin A expression by Myc in both p27+/+ and p27−/− MycER cells (data shown for p27−/− cells in Figure 4C) at low micromolar concentrations. We concluded that the loss of p27 does not completely abolish the requirement for either cdk2 or cdk4 kinase activity in Myc-induced G1 progression. Previous work has shown that ectopic expression of E2F1 can induce S-phase entry in the absence of detectable cdk2 activity, suggesting that this might be the mechanism by which Myc acts to promote S-phase entry in p27−/− cells (Leone et al., 1999). Surprisingly, however, the total cdk2 activity strongly increased after activation of Myc in both p27+/+ and p27−/− cells (Figure 5A; a quantitation of the results of two representative experiments is shown in Figure 5B). Precipitation with α-cyclin A antibodies showed identical results (not shown). Thus, cyclin A–cdk2 complexes that formed after induction of Myc were active in both p27+/+ and p27−/− cells. To explain the difference compared with cyclin E–cdk2 complexes, we tested whether cyclin A–p130 complexes could be detected; indeed, immunoprecipitation experiments detected the presence of cyclin A in α-p130 immunoprecipitates and vice versa (data not shown). In contrast to observations made with p27kip1 (Poon et al., 1995), depletion of cell lysates by an α-cdk2 antibody not only removed cdk2, but also p130 in serum-starved cells, demonstrating that the amount of p130 is limiting in these cells (Figure 5C). Indeed, p130 is depleted by α-cdk2 antibodies both before and after activation of Myc or addition of serum in p27−/− MycER cells. At early time points after activation, newly synthesized cyclin A is also quantitatively bound to cdk2 (see 8 and 12 h after activation of Myc); taken together, both observations suggest that there is not enough p130 in these cells to inhibit quantitatively newly formed cyclin A–cdk2 complexes. Consistent with this view, depletion with an α-p130 antibody failed to deplete cyclin A from p27−/− MycER cells (Figure 5D). The data suggest that the strong increase in cyclin A protein is sufficient to account for the preferential activation of cyclin A–cdk2 versus cyclin E–cdk2 complexes. They do not exclude, however, that differences in affinity between cyclin A–cdk