Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Cyclin E and c‐Myc promote cell proliferation in the presence of p16INK4a and of hypophosphorylated retinoblastoma family proteins

Cyclin E and c‐Myc promote cell proliferation in the presence of p16INK4a and of... The EMBO Journal Vol.16 No.17 pp.5322–5333, 1997 Cyclin E and c-Myc promote cell proliferation in the INK4a presence of p16 and of hypophosphorylated retinoblastoma family proteins (Guan et al., 1994; Polyak et al., 1994; Toyoshima and Konstantinos Alevizopoulos, Jaromir Vlach, Hunter, 1994; Hirai et al., 1995; Quelle et al., 1995; Vlach Silke Hennecke and Bruno Amati et al., 1996), and numerous observations link CKIs to Swiss Institute for Experimental Cancer Research, CH-1066 Epalinges, growth arrest and/or differentiation in diverse cell types Switzerland (reviewed by Sherr and Roberts, 1995; Assoian, 1997). Corresponding author The best characterized substrates of cyclin–CDK com- e-mail: [email protected] plexes are the retinoblastoma family proteins pRb, p107 and p130 (or ‘pocket proteins’), which negatively regulate Retroviral expression of the cyclin-dependent kinase cell cycle progression and are inactivated through phos- INK4a (CDK) inhibitor p16 in rodent fibroblasts induces phorylation by CDKs (Weinberg, 1995). In their active, dephosphorylation of pRb, p107 and p130 and leads hypophosphorylated forms, pocket proteins bind to and to G arrest. Prior expression of cyclin E allows S-phase negatively regulate the activities of several targets, such entry and long-term proliferation in the presence of as the heterodimeric E2F–DP transcription factors. Five p16. Cyclin E prevents neither the dephosphorylation E2F (E2F-1 to -5) and three DP proteins (DP-1 to -3) of pRb family proteins, nor their association with E2F have been identified. E2F-1, -2 and -3 bind predominantly proteins in response to p16. Thus, cyclin E can bypass pRb, E2F-4 binds p107 and p130, and E2F-5 binds the p16/pRb growth-inhibitory pathway downstream predominantly p130 (reviewed by Cobrinik, 1996). of pRb activation. Retroviruses expressing E2F-1, -2 Depending on the target promoters, E2F complexes can or -3 also prevent p16-induced growth arrest but are function as either transcriptional activators or repressors Kip1 ineffective against the cyclin E–CDK2 inhibitor p27 , (in association with pocket proteins). E2F target genes suggesting that E2F cannot substitute for cyclin E include regulators of S-phase entry (e.g. B-myb, CDC2, activity. Thus, cyclin E possesses an E2F-independent cyclins E and A) and genes required for DNA replication function required to enter S-phase. However, cyclin E (e.g. DHFR, DNA polα, thymidine kinase) (DeGregori may not simply bypass E2F function in the presence et al., 1995a; reviewed by Nevins, 1992; Zwicker and of p16, since it restores expression of E2F-regulated Mu¨ller, 1997). genes such as cyclin A or CDC2. Finally, c-Myc When ectopically expressed, E2F-1 is sufficient to bypasses the p16/pRb pathway with effects indistin- induce entry of quiescent cells into S-phase (but also guishable from those of cyclin E. We suggest that this apoptosis) (Johnson et al., 1993; Qin et al., 1994; Shan effect of Myc is mediated by its action upstream of and Lee, 1994; Lukas et al., 1996), a property shared by cyclin E–CDK2, and occurs via the neutralization other E2F proteins (Lukas et al., 1996; reviewed by Kip1 of p27 family proteins, rather than induction of Adams and Kaelin, 1996). E2F-1 and E2F-4 can override Cdc25A. Our data imply that oncogenic activation of pRb- and p130-induced G arrest, respectively (Zhu et al., c-Myc, and possibly also of cyclin E, mimics loss of 1993; Qin et al., 1995; Vairo et al., 1995). E2F-1 also the p16/pRb pathway during oncogenesis. prevents G arrest imposed either by antibodies neutraliz- Keywords: CDK/cyclin/Myc/p16/retinoblastoma ing cyclin D–CDK4 function or by p16, in spite of the p16-induced dephosphorylation of cellular pRb (DeGregori et al., 1995b; Lukas et al., 1996; Mann and Jones, 1996). Expression of p16 or neutralization of cyclin Introduction D–CDK4 also fail to induce G arrest in cells lacking functional pRb (Lukas et al., 1994, 1995a,b; Koh et al., The activities of the cyclin-dependent kinases CDK4 and 1995; Medema et al., 1995). Thus, the essential role of CDK6, which associate with D-type cyclins, and of CDK2, cyclin D–CDK4 is to inactivate pRb, thereby allowing which associates with cyclins E or A, are rate-limiting for E2F function. In summary, p16, cyclin D1–3, CDK4/6, progression through G and into S-phase of the vertebrate pRb and E2F can be ordered within a single G -regulatory cell cycle (reviewed by Sherr, 1994, 1995). Besides pathway (the p16/pRb pathway). association with cyclins, the activity of CDKs is controlled The p16/pRb pathway plays a critical role in suppressing by site-specific phosphorylation or dephosphorylation and tumorigenesis. Several of its components are genetically by association with a group of inhibitory proteins collect- altered in various human malignancies, resulting either in ively called CKIs (reviewed by Morgan, 1995; Sherr and Cip1/Waf1 oncogenic activation (cyclins D1, D2 and CDK4) or loss Roberts, 1995). CKIs of the Kip/Cip family, p21 , Kip1 Kip2 of function (pRb and p16) (reviewed by Hall and Peters, p27 and p57 , can associate with and inhibit all 1996). p16-null mice are viable, but develop malignancies known G cyclin–CDK complexes. CKIs of the INK4 INK4a INK4b INK4c INK4d family, p16 , p15 , p18 and p19 bind at increased rates (Serrano et al., 1996). Strikingly, E2F-1- CDK4 and CDK6 and can interfere with cyclin–CDK null mice also have a tumour-prone phenotype (Field interactions. Ectopic expression of CKIs induces G arrest et al., 1996; Yamasaki et al., 1996). Thus, the major role 5322 © Oxford University Press Cyclin E and Myc bypass the p16/pRb pathway of E2F-1 in vivo might be to relay the repressive action complexes (Steiner et al., 1995), and this also occurs of pRb, rather than its own activating function. It should through the suppression of p27 function (Perez-Roger be noted here that pRb has other targets in cells, which et al., 1997; J.Vlach and B.Amati, unpublished data). may also be critical for its growth- and tumour-suppressive However, no link was yet established between Myc and functions (Weinberg, 1996; Wang, 1997). p16 function. CDK2 and its partners, cyclins E and A, also phos- In this study, we addressed whether cells infected with phorylate and inactivate pocket proteins (Weinberg, 1995). retroviruses expressing G cyclins, Cdc25A, E2F proteins For example, in transfected SAOS-2 cells, the G arrest or Myc could escape growth arrest induced by p16, and induced by pRb is relieved by cyclins A or E, and this compared the effects of p16 and p27 in those cells. Our correlates with hyperphosphorylation of pRb (Hinds et al., data demonstrate that cyclin E promotes cell proliferation 1992). Moreover, rapid activation of cyclin E–CDK2 in the presence of elevated p16 levels, without inducing upon induction of Myc (see below) coincides with pRb phosphorylation of pocket proteins. We suggest that cyclin phosphorylation, whereas activation of cyclin A–CDK2 E cannot simply substitute for E2F function, but rather and cyclin D–CDK4 follow later in this particular system rescues it in the presence of p16. Reciprocally, the require- (Steiner et al., 1995). Cyclin E–CDK2 appears to have ment for cyclin E in G –S progression cannot be substituted additional function(s) in G –S progression, distinct from by E2F. Finally, we demonstrate that Myc also bypasses pRb phosphorylation and E2F activation. First, inducible the p16/pRb pathway, and argue that this effect is mediated expression of cyclin E in Rat1 fibroblasts accelerates G –S by the positive action of Myc on cyclin E. progression upon serum stimulation, without affecting the kinetics of pRb phosphorylation (Resnitzky and Reed, Results 1995). Second, G arrest induced by dominant-negative CDK2 in U2OS cells is not relieved by the SV40 large T Myc and cyclin E prevent growth arrest by INK4 antigen (LT), although LT rescues E2F function in this proteins system (Hofmann and Livingston, 1996). Third, unlike The ability of Myc and various cell cycle regulators to D-type cyclins, cyclin E is essential for cell cycle progres- suppress the growth-inhibitory function of INK4 family sion in pRb-deficient cells (Ohtsubo et al., 1995). proteins (p15, p16 or p18) was investigated in Rat1 The functional hierarchy between the pRb/E2F pathway fibroblasts as previously described for Myc and p27 (Vlach and cyclin E is complicated further by the finding that the et al., 1996). First, cells were infected with retroviruses cyclin E gene is itself an E2F target. Cyclin E is induced expressing Myc, Cdc25A, cyclins A and E, or D1–3. by ectopic expression of E2F-1 (DeGregori et al., 1995a; Serial infections using the pBabe series of retroviral vectors Soucek et al., 1997), and functional E2F-binding sites in conferring resistance to neomycin (pBNeo), hygromycin the cyclin E promoter are activated by E2F-1 and repressed (pBHygro) or phleomycin (pBBleo) (Morgenstern and by pRb (Ohtani et al., 1995; Geng et al., 1996). Further- Land, 1990) allowed co-expression of up to three exo- more, pRb-deficient fibroblasts express enhanced levels genous proteins prior to INK4 proteins, which were of cyclin E (Herrera et al., 1996). These observations expressed with a puromycin resistance vector (pBPuro). suggest the existence of a positive feedback regulatory Following infection, cells were seeded at serial dilutions loop, in which cyclin E is not only an upstream regulator (e.g. 1/20, 1/200, 1/2000) in high serum, puromycin- of E2F (via the phosphorylation of pRb), but also its selective medium. All non-infected cells had detached downstream effector. If the latter were true, deregulated within 36 h of selection. Subconfluent populations of cyclin E expression might be expected to bypass growth infected cells were harvested at 48 h after a 30 min pulse arrest by p16 or pRb, exactly as E2F. with bromodeoxyuridine (BrdU) for analysis of cell cycle Myc, the transcription factor encoded by the c-myc distribution. Biochemical studies were performed at the proto-oncogene, was also linked to the activation of same time. Dilute dishes were incubated further to assess cyclin–CDK complexes in G . Myc expression is strictly long-term proliferation and colony outgrowth. controlled by mitogens, and is required for cell cycle entry Expression of p16 induced accumulation of Rat1 cells and continued proliferation. Constitutive Myc expression in G and suppressed colony outgrowth (Figure 1A and prevents growth arrest by a variety of growth-inhibitory B). Most cells in these p16-arrested populations remained signals, and activation of a conditional Myc–estrogen isolated or formed sparse colonies, with no signs of receptor chimera (MycER) induces entry into the cell apoptosis (data not shown). Retroviruses expressing cycle (but also apoptosis) in the absence of mitogens cyclins A, D1, D2, D3 or Cdc25A did not prevent p16- (Eilers et al., 1991; Alexandrow et al., 1995; Hermeking induced cell cycle arrest, although all exogenous proteins et al., 1995; reviewed by Marcu et al., 1992; Henriksson were detectable by immunoblotting (data not shown and and Lu¨scher, 1996). Since antimitogenic stimuli are medi- Table I). In contrast, cells expressing Myc or cyclin E ated by CKIs, we investigated whether Myc could over- entered S-phase and formed colonies efficiently in the come the growth-inhibitory function of these proteins. We presence or absence of p16 (Figure 1A and B, and Table previously reported that constitutive expression of Myc I). Similar results were obtained with NIH-3T3 cells (data abrogates growth arrest by retrovirally expressed p27, by not shown). As for p16, Myc and cyclin E bypassed cell preventing its association with cyclin E–CDK2 (Vlach cycle arrest by p15 and p18 in Rat1 cells (Figure 1C). et al., 1996). Myc also induces transcription of the CDK- Immunoblot analysis revealed that Myc and cyclin E did activating phosphatase Cdc25A (Galaktionov et al., 1996), not suppress expression of p16, p15 or p18 (Figure 2A), which appears not to be involved in rescue of p27-induced and had no significant effect on the expression of the arrest (Vlach et al., 1996). Moreover, MycER activation cellular p16 targets CDK4, CDK6 and D-type cyclins, in quiescent fibroblasts rapidly activates cyclin E–CDK2 apart for a slight increase in cyclin D2 levels in Myc cells 5323 K.Alevizopoulos et al. Table I. Summary of proliferation data with doubly infected Rat1 cells INF.2 p16 p27 INF.1 vector Myc cyclin A cyclins D1–3 cyclin E Cdc25A E2F-1–3  – E2F-4, 5 DP-1 E2F-4  DP-1  n.d. Data from this work are compiled together with data from Vlach et al. (1996), for a direct comparison between p16 and p27.  and – indicate proliferating and arrested cells, respectively, according to BrdU incorporation and colony outgrowth assays. All cells proliferated after the first infection (INF. 1) as efficiently as control, vector- infected cells. n.d., not determined. hypophosphorylated (Figure 3A and B, lane 2). p107, in particular, was present as a doublet of both hyper- and hypophosphorylated forms in p27-arrested cells (Vlach et al., 1996 and Figure 3A), whereas INK4-arrested cells contained only the lower band. Strikingly, Myc and cyclin E did not reverse the effect of p16 on pRb family proteins, which were also mostly hypophosphorylated in proliferating Mycp16 or cyclin Ep16 cells (Figure 3B, lanes 4 and 6). This observation was reproduced for pRb and p107 in NIH-3T3 cells (Figure 3C) in which p16 induced no changes in p130 mobility (data not shown). Thus, pocket proteins were hypophosphorylated in grow- ing Mycp16 or cyclin Ep16 cells, unlike in control Fig. 1. Myc and cyclin E overcome growth arrest by INK4 family proteins. (A) p16 induces accumulation of control cells in G , but does growing cells. This was not merely a transient phenome- not modify the cell cycle distribution of cells expressing Myc or non. Indeed, when passaged and cultured for longer cyclin E. Serial infections were performed as described in the text. periods of time, Rat1 cells co-expressing Myc or cyclin Cells were infected first (INF.1) with retroviruses expressing Myc or E with p16 proliferated with elevated p16 levels and cyclin E or with the empty control vector (pBHygro), followed by a maintained hypophosphorylated pRb, p107 and p130 pBPuro retrovirus with or without p16, as indicated (INF.2). Cells were harvested after 48 h of puromycin selection, and cell cycle (Figure 4). At all time points, the percentage of S-phase distribution was analysed by flow cytometry. The percentage of cells cells was equivalent to those of control growing cells in each phase of the cell cycle is indicated for all cell populations. (data not shown). Altogether, these observations suggested (B) p16 prevents colony outgrowth in control, but not in Myc- or that Myc and cyclin E allowed cell proliferation by cyclin E-expressing cells. Infections were performed as above. bypassing pRb activation, and that p16 remained functional Colonies were fixed and stained on dishes after 6 days in puromycin- selective medium. (C) Percentage of S-phase cells in populations in inhibiting its targets, CDK4 and CDK6. Indeed, p16– infected with retroviruses expressing Myc or cyclin E (INF.1), and CDK6 interactions were not altered by Myc or cyclin E, p15, p16 or p18 (INF.2). S-phase cells were visualized by as detected by CDK6 immunoprecipitation followed by immunocytochemical detection of incorporated BrdU. immunoblotting of p16 (Figure 2C). Retroviral expression of CDK4 (wild-type or a kinase-defective mutant) or (Figure 2B). In conclusion, Myc and cyclin E overcome CDK6 also prevented p16-induced arrest (data not shown), growth arrest by INK4 proteins. supporting the notion that these kinases were the p16 targets in cells. However, we were unable to demonstrate Myc and cyclin E do not prevent p16-induced directly that p16 suppressed the kinase activities associated dephosphorylation of pocket proteins with CDK4, CDK6 or D-type cyclins. In summary, Myc To investigate the effects of Myc or cyclin E on INK4 and cyclin E promote cell growth in the presence of p16 function, we examined the relative phosphorylation state and hypophosphorylated pocket proteins. of pRb, p107 and p130 by immunoblot analysis. In INK4- arrested Rat1 cells, pocket proteins were found exclusively Retrovirally expressed pRb cooperates with p16 in their hypophosphorylated, faster migrating form (Figure and is bypassed by Myc or cyclin E 3A), consistent with the induction of G arrest. Although If Myc and cyclin E bypass pRb activation by p16, they pRb levels were decreased consistently in p16-arrested might rescue growth in the presence of elevated pRb levels. cells in comparison with controls, the remaining pRb was To address this question, we constructed retroviruses 5324 Cyclin E and Myc bypass the p16/pRb pathway Fig. 3. Immunoblot analysis of expression and relative phosphorylation states of pRb family proteins. Rat1 cells were infected with (A) viruses expressing p15, p16, p18 or p27 alone, as indicated, or (B) with the indicated combination of viruses. (C) The same experiment as in (B) was performed with NIH-3T3 cells. The slower migrating, hyperphosphorylated forms of pRb, p107 or p130 and their faster migrating, hypophosphorylated forms are indicated on the right (hyper, hypo) in (A). Fig. 2. Expression of exogenous INK4 proteins, of cellular cyclins and CDKs, and association of exogenous p16 with cellular CDK6. (A) Immunoblot analysis of retrovirally expressed INK4 proteins in doubly infected cells, as indicated (INF.1 and INF.2). The asterisk beside the p18 panel represents a non-specific band cross-reacting with the p18 antibody. (B) Immunoblot analysis of cellular proteins (indicated on the left of each panel) in doubly infected Rat1 cells expressing Myc, cyclin E and/or p16, as indicated at the top. The asterisks below lanes 5 and 6 of the cyclin E panel indicate that only the endogenous, rodent cyclin E is visualized in this blot. (C) CDK6 was immunoprecipitated from the same cells as above, and the co-precipitated p16 protein revealed by immunoblotting. expressing full-length, hemagglutinin (HA)-tagged murine pRb (HApRb) or a derivative mutated in several potential CDK phosphorylation sites (HAΔp34) (Hamel et al., Fig. 4. p16 levels are maintained, and pRb, p107 and p130 remain hypophosphorylated in growing cells expressing exogenous Myc or 1992). Although HApRb and HAΔp34 accumulated in cyclin E together with p16. Cells infected with the indicated cells at higher levels than endogenous pRb (Figure 5A), combinations of viruses were allowed to grow for the indicated they were insufficient to affect cell cycle progression or number of passages (Pass.) and days. All cells were proliferating proliferation on their own (Figure 5B and data not shown). continuously throughout the experiment, and were subconfluent at each harvesting time point. Immunoblotting of p16 (lower panels) shows However, HApRb slightly, albeit reproducibly, enhanced that its expression is maintained. Pocket proteins (indicated to the left) the efficiency of p16-induced arrest, as shown by the are found in their hypophosphorylated forms in all cells expressing lower percentage of S-phase cells in HApRbp16 cells p16, but not in control cells. The faster migrating pRb bands seen in compared with p16-expressing cells (Figure 5C). Further- the presence of p16 represent further dephosphorylated forms, as more, HApRbp16 cells showed a marked increase in previously seen (Chen et al., 1989). 5325 K.Alevizopoulos et al. of pRb, p107 and p130. At this stage, however, inactivation of pocket proteins by a specific cyclin E-mediated phosphorylation event could not be ruled out. To address the functional status of pRb, we sought to analyse its interaction with E2F-1, -2 or -3. Because these proteins were undetectable in Rat1 cells, we co-expressed Myc and cyclin E with an E2F variant (HA-E2F1eco) which interacts with pRb but does not bind DNA (Johnson et al., 1993). In spite of a slight residual activity of E2F1eco (see Figure 9C), growth of E2F1eco cells was suppressed by p16 and restored by Myc or cyclin E (data not shown). Following superinfection with the p16 retrovirus, the interactions between HA-E2F1eco and pRb were monitored by immunoprecipitation with anti-E2F-1 anti- bodies, followed by pRb immunoblotting (Figure 6A, panels ii). Total and immunoprecipitated HA-E2F1eco levels were constant throughout the experiment (panels iii and iv), and no pRb was detectable in anti-E2F-1 immunoprecipitates from cells lacking HA-E2F1eco (panels ii, lanes 0). The results showed that p16 increased the association of pRb with HA-E2F1eco in control cells (Figure 6A, lanes 1 and 2), as well as in cells expressing Myc (lanes 3 and 4) or cyclin E (lanes 5 and 6), even though total pRb levels were decreased by p16 in all cells (Figure 6A, panels i; Figure 3B). The same experiment was repeated in cells expressing exogenous HApRb together with HA-E2F1eco, and similar controls were provided (Figure 6B). The result was identical, showing that the association of pRb (endogenous plus exogenous) with HA-E2F1eco was induced by p16 in all cells (panels ii). Taken together, these experiments demonstrate that Myc and cyclin E do not interfere with p16-induced activation of pRb. Similar experiments were performed to assess the func- Fig. 5. Retrovirally expressed pRb enhances p16-induced arrest, but does not prevent rescue by Myc or cyclin E. (A) Immunoblot analysis tional status of p130. As HA-tagged E2F-4 is not functional of retrovirally expressed HApRb of the HAΔp34 mutant with anti-pRb in the p16 rescue assay (see below, Figure 9C), we co- and anti-HA antibodies, as indicated. Hyper, hypo: forms of pRb as expressed it together with Myc or cyclin E and monitored defined in Figure 3A. (B) Percentage of cells in S-phase 48 h its interaction with endogenous p130. As above, we post-infection with the same retroviruses as in (A). (C) Percentage of S-phase cells following serial infections with the indicated viruses controlled the amounts of total p130 (Figure 7, panel i) (INF.1, INF.2, INF.3). The data presented here represent the average of and of total and immunoprecipitated E2F-4 (panels iii and three independent experiments. (D) Immunoblot analysis of HApRb in iv). No p130 was detectable in HA immunoprecipitates the same cells. in the absence of HA-E2F-4 (lane 0). The data (panel ii) showed that p16 induced binding of p130 to E2F-4 in the flattened, senescent-like phenotype, reminiscent of that control, Myc and cyclin E cells, analogous to the results seen in pRb-arrested SAOS-2 cells (Templeton et al., with pRb. The same experiment did not yield a clear 1991; Hinds et al., 1992) (data not shown). Co-expression answer for the interaction between p107 and HA-E2F-4. of Myc or cyclin E with HApRb prior to p16 fully This interaction was detectable in growing Rat1 cells, but prevented cell cycle arrest (Figure 5C) and the appeareance was not increased further upon expression of p16 (data of the ‘flat’ phenotype (data not shown). Although the not shown). differentially phosphorylated forms of HApRb were diffi- To analyse E2F–pocket protein complexes further and cult to resolve on immunoblots, a significant fraction of to assess the amount of ‘free’ E2F activity, we performed the protein was found in its faster-migrating form in all gel retardation assays with retrovirally infected cells. As p16-expressing cells (Figure 5D). These data confirmed detection of E2F–DNA complexes was very difficult in two important points: first, that pRb levels are limiting Rat1 cells, we used NIH-3T3 cells, which showed similar for p16-induced arrest; second, that Myc and cyclin responses to p16 (Figure 3C and data not shown). Three E rescue cell growth without preventing p16-induced DNA-binding complexes specific for E2F sites were dephosphorylation of either endogenous or exogenous identified with oligonucleotide competition studies (com- pRb. plexes A, B and C; Figure 8). Complex A corresponds to ‘free’ E2F–DP dimers, as treatment of the extracts with Pocket proteins are functional in growing deoxycholate (DOC) resulted in loss of B and C and a Mycp16 or cyclin Ep16 cells simultaneous increase in the amount of complex A. The The data presented so far suggested that Myc and cyclin E2F species in complex A was most likely E2F-4, since E promote cell growth in the face of p16-induced activation it co-migrated with the major E2F–DNA complex seen in 5326 Cyclin E and Myc bypass the p16/pRb pathway Fig. 6. Interaction of pRb with the E2F-1eco protein in the absence (A) or presence (B) of exogenous pRb. Cells were serially infected with the indicated combinations of viruses. The E2F-1 mutant HA-E2F1eco (see text) was immunoprecipitated with anti-E2F-1 antibodies. Immunoblotting was performed to visualize the immunoprecipitated HA-E2Feco with anti-HA antibodies (panels iii) and the associated pRb (endogenous and exogenous) with anti-pRb antibodies (panels ii). Total levels of pRb (panels i) and HA-E2F1eco (panels iv) were also measured. nd: not determined. NIH-3T3 cells expressing recombinant E2F-4, but not E2F-1, -2, -3 or -5 (data not shown). The composition of complex B, which was not supershifted by any tested antibody, is unknown. Complex C could be supershifted by p107 or CDK2 antibodies (data not shown) and presumably also contained cyclins E or A (Lees et al., 1992; Schwarz et al., 1993; Cobrinik, 1996). No other pocket protein–E2F complexes were detected in growing NIH-3T3 cells, as previously described (Lam and Watson, 1993; Schwarz et al., 1993). Two major points can be made concerning the effects of Myc, cyclin E and/or p16 on E2F DNA-binding complexes (Figure 8). First, p16 alone reduced the amount of ‘free’ E2F-4–DP (complex A) present in control cells. This decrease was minimized in the presence of Myc or cyclin E. Second, complex C was increased in response to p16 in Myc or cyclin E cells. The lack of complex C Fig. 7. Interaction of p130 with E2F-4. Cells were serially infected induction in control cells may be due to the formation of with the indicated combinations of viruses. HA-E2F-4 was alternative complexes not resolved here and, in part, to immunoprecipitated with anti-HA antibodies. Immunoblotting was slightly lower p107 levels in p16-arrested cells (Figure performed to visualize the immunoprecipitated HA-E2F-4 with 3C). Regardless of this, our data show that p16 activates anti-HA antibodies (panel iii), or the associated p130 (panel ii). the E2F-binding activity of p107 in Myc and cyclin E Total p130 (panel i) and HA-E2F-4 (panel iv) were also visualized. nd: not determined. cells. In summary, the results from co-immunoprecipitation and gel retardation studies show that Myc and cyclin E prevent growth arrest by p16 without interfering with the results is that cyclin E and Myc restore activity of some ability of pRb, p107 and p130 to associate with their E2F complexes (see Discussion), although alternative respective E2F targets. routes to restore expression of E2F target genes cannot be ruled out. E2F proteins bypass p16-, but not p27-induced If cyclin E acted solely through E2F, E2F proteins arrest: implications for an E2F-independent would be expected to bypass G arrest not only by p16, function of cyclin E as previously reported (DeGregori et al., 1995b; Lukas The data presented so far may be interpreted in two et al., 1996; Mann and Jones, 1996), but also by p27, alternative ways. First, cyclin E or Myc might allow cell which inhibits cyclin E–CDK2 function. To address this growth in the absence of E2F function. Second, cyclin E question, we infected Rat1 cells with retroviruses or Myc might rescue the transcriptional activity of some expressing HA-tagged versions of E2F-1 to -5, as well as E2F complexes in the presence of active pRb, p107 E2F-1Δ5 and E2F-4Δ4 (two mutants which are not bound and p130. Unfortunately, we could not detect significant by pocket proteins; Krek et al., 1994; Vairo et al., 1995), regulation of stably or transiently transfected E2F-respons- the DNA-binding mutant E2F-1eco and a T7-tagged form ive reporter genes by retrovirally expressed p16. However, of DP-1. Expression of all these proteins was detected cyclin A and CDC2 protein levels were suppressed by readily by immunoblot (Figure 9A and B and data not p16 in control cells, but not in cells expresing cyclin E shown). E2F-1, -2, -3 and E2F-1Δ5 prevented growth or Myc (Figure 2B). The simplest interpretation of these arrest by p16, whereas E2F-4, E2F-5 and DP-1 were 5327 K.Alevizopoulos et al. Fig. 8. Gel retardation analysis of E2F DNA-binding activities in NIH-3T3 cells. Cells were infected with the indicated viruses and complexes binding to a radiolabeled E2F-binding oligonucleotide were resolved by native gel electrophoresis. Three specific complexes were identified (A, B and C). Where known, the proteins comprising these complexes are indicated on the right (see text). The asterisks indicate non-sequence-specific complexes, as determined by competition studies. inactive (Figure 9C and Table I), consistent with previous reports (Lukas et al., 1996; Mann and Jones, 1996). The lack of activity of E2F-4 and E2F-5 may be due to the combination of several parameters, including interaction with pocket proteins and/or inefficient dimerization. Indeed, E2F-4Δ4 had a weak rescuing activity, and co- expression with DP-1 allowed E2F-4 and E2F-4Δ4to fully rescue p16-induced arrest (Figure 9; Lukas et al., 1996). In contrast to the p16 results, none of the E2F proteins prevented growth arrest by p27 (Figure 9D and Fig. 9. Retrovirally expressed E2F-1 to -3 overcome p16-, but not data not shown, and Table I). Association of E2F-1 with p27-induced cell cycle arrest. Cells were infected with viruses expressing the indicated tagged E2F and DP proteins and mutant pocket proteins did not account for its inability to bypass derivatives (see text). (A and B) Immunoblot analysis of retrovirally p27 arrest, since cells expressing E2F-1Δ5 were also expressed E2F proteins (indicated at the top), with anti-E2F-1 or arrested by p27. Thus, E2F-1 to -3 in our system behaved anti-HA in (A), and anti-HA in (B). Given that the HA-tag is like cyclin E, which rescued growth arrest by p16 (this N-terminal, the faster migrating band in the E2F-4 and E2F-4Δ4 lanes work), but not by p27 (Vlach et al., 1996) (Table I). In is probably a degradation product. (C) Percentage of cells in S-phase in pools infected with viruses expressing HA-tagged E2F proteins and summary, although E2F-1, -2 and -3 were expressed at p16, as indicated (INF.1, INF.2). The values for p16 cells (shaded levels sufficient to bypass pRb activation in response to bars) are superimposed on those for control cells (white bars). p16, they could not bypass repression of cyclin E–CDK2 (D) Same as above with p27 instead of p16. Myc was used as a by p27. This implies that cyclin E has a distinct function, positive control to overcome p27-induced arrest. for which its targets lie outside of the pRb/E2F pathway. CDK6 and induced activation of pocket proteins and their Discussion association with E2Fs. This effect was not transient since Cyclin E bypasses activation of the p16/pRb cells expressing cyclin E and p16 sustained long-term growth-inhibitory pathway proliferation with hypophosphorylated, active pocket pro- Previous observations suggested that cyclin E has a teins. Cyclins A, D1, D2 and D3 were inactive in bypassing function distinct from pRb phosphorylation required for p16-induced arrest in this system (Table I). G –S progression (see Introduction; Ohtsubo and Roberts, In normal cells, cyclin E expression is under the control 1993; Resnitzky and Reed, 1995; Hofmann and Livingston, of E2F and pRb. The promoter of the cyclin E gene 1996). We demonstrate here that cyclin E is sufficient to contains E2F-responsive elements and is repressible by bypass growth arrest by p16, and allows cell proliferation pRb (Ohtani et al., 1995; Botz et al., 1996; Geng et al., in the presence of active pRb family proteins. Infection 1996). Consistent with this, pRb-deficient cells showed with a p16-expressing retrovirus arrested Rat1 cells in G , enhanced expression of cyclin E (Herrera et al., 1996), but had no effect on cells expressing exogenous cyclin E. and overexpression of E2F-1 induced cyclin E (DeGregori p16 remained active in these cells since it bound its target et al., 1995a; Soucek et al., 1997). Since cyclin E–CDK2 5328 Cyclin E and Myc bypass the p16/pRb pathway complexes also contribute to pRb phosphorylation and activating complexes. Compatible with, but not proving, inactivation, these observations led to the proposal that a this hypothesis, our gel retardation assays showed that positive feedback loop may operate in late G (see p16 in cyclin E-expressing cells induced formation of Introduction). According to this model, the increase in E2F-4/DP/p107/cyclin–CDK2 DNA-binding complexes. cyclin E–CDK2 activity would contribute to further activ- On the other hand, cyclin E also favored the maintenance ation of E2F, and thus to enhanced transcription of the of ‘free’ E2F-4–DP complexes. However, whether and cyclin E gene. The outcome of such a loop would be to how E2F contributes to cyclin E function in overcoming boost activation of both E2F and cyclin E. However, such the p16 block remains to be investigated. a model does not predict which of E2F or cyclin E, or The functional relationship of cyclin E and E2F is both, performs the rate-limiting downstream function in highlighted further by studies in Drosophila embryos. S-phase entry. E2F- or cyclin E-deficient embryos fail to enter S-phase Ectopic activation of an essential downstream effector of embryonic cycle 17 in the endocycling cell compartment may bypass the presence of active pRb. This is the case (Knoblich et al., 1994; Duronio et al., 1995). In this for E2F-1, which suppresses both pRb- (Zhu et al., 1993; tissue, the limiting downstream target of E2F for S-phase Qin et al., 1995) and p16-induced G arrest, in conditions entry is cyclin E. Ectopic E2F–DP expression requires in which cellular pRb remains hypophosphorylated (Lukas cyclin E function to trigger S-phase entry, although the et al., 1996; Mann and Jones, 1996; this work and transcriptional function of E2F itself does not require unpublished data). Moreover, adenovirus-mediated gene cyclin E (Duronio et al., 1995, 1996). Conversely, ectopic transfer of E2F-1 suppressed G arrest and allowed comple- expression of cyclin E can trigger S-phase entry, but tion of a mitotic cycle in the presence of either p16 or requires expression of additional E2F target genes. Cyclin p27, leading to the suggestion that E2F-1 allows S-phase E expression induces E2F-dependent transcription entry in the absence of cyclin E function (DeGregori (Duronio and O’Farrell, 1995), most likely through et al., 1995b). In those studies, however, E2F-1 was phosphorylation of the pRb homolog RBF (Du et al., strongly overexpressed, which may have led to non- 1996). In the central nervous system, cyclin E is expressed physiological effects. Indeed, we and others (Mann and by an E2F-independent mechanism, and is required Jones, 1996) found that p27 is dominant over E2F-1. In upstream of E2F function (Duronio et al., 1995). Thus, fact, E2F proteins behaved similarly to cyclin E, which the dependence of one gene upon the other can be reversed rescued p16-, but not p27-induced arrest (Table I), owing in different tissues or stages, depending on which gene to the inhibition of cyclin E by p27 (Vlach et al., 1996). product becomes rate-limiting. In summary, in Drosophila These observations imply that E2F activity cannot bypass as in mammalian cells, cyclin E is both a downstream inhibition of cyclin E–CDK2 by p27, even though it can effector and an activator of E2F. In addition to this cross- bypass activation of pRb by p16. Two conclusions can be regulation, both E2F and cyclin E have at least one drawn from these studies: first, that E2F is unable to independent downstream function required for S-phase promote growth in the absence of cyclin E activity; second, entry. that cyclin E has at least one E2F-independent function What might be the E2F-independent function of cyclin required for S-phase entry. Such a function of cyclin E? In vitro experiments suggest that it might be linked to E–CDK2 was also implied by the observation that the the initiation of DNA replication. In Xenopus extracts, SV40 LT restored E2F activity, but not cell cycle progres- which replicate in the absence of transcription (and thus sion, in the presence of transiently expressed dominant- of E2F activity), addition of p21 blocked DNA replication, negative CDK2 (Hofmann and Livingston, 1996). and cyclins A or E could overcome this block (Strausfeld If cyclin E has a function downstream of E2F, could it et al., 1994). In addition, G -phase HeLa cell nuclei be sufficient for S-phase entry in the absence of E2F initiated DNA replication when co-incubated with S-phase activity? A direct test of this question would be to block nuclear and cytoplasmic extracts. The nuclear extract E2F activity in cells expressing cyclin E. We attempted could be replaced by cyclin E–CDK2 or cyclin A–CDK2 this with a dominant-negative mutant of DP-1 (DN-DP-1) (Krude et al., 1997). A direct involvement of cyclin–CDK which blocks E2F function and arrests cells in G in activity in triggering the onset of DNA replication has also transient transfections (Wu et al., 1996). However, retro- been demonstrated in yeast (reviewed by Stillman, 1996). viral expression of DN-DP-1 (or a variant tagged with a nuclear localization sequence) had no effect on cell cycle Rescue of p16-induced arrest by Myc is most likely progression. Thus DN-DP-1 may effectively block E2F a cyclin E-mediated activity activity in vivo only if overexpressed, which may have We show here that Myc bypasses p16/pRb-induced growth non-specific side effects. A direct test of the requirement arrest. Like other biological activities of Myc (Henriksson for E2F in the presence of cyclin E awaits alternative and Lu¨scher, 1996; Vlach et al., 1996), this effect depends ways to ablate E2F activity. In spite of this, several upon the formation of transcriptionally active Myc–Max observations suggest that cyclin E does not bypass, but dimers (unpublished data). As for cyclin E, cells expressing rather rescues E2F activity by an alternative route in the Myc and p16 proliferate with hypophosphorylated, active presence of p16 and active pRb. In transient transfections, pocket proteins. Thus, Myc and cyclin E rescue p16- cyclin E relieved p16 (but not p27)-mediated repression induced arrest by indistinguishable mechanisms. The main of the cyclin A promoter (Zerfass-Thome et al., 1997). In rationale for these findings is provided by the notion that our experiments, cyclin E rescued the expression of Myc can act as an upstream activator of cyclin E–CDK2. cellular cyclin A and CDC2 in the presence of p16. Constitutive expression of Myc prevented inactivation of Zerfass-Thome et al. (1997) proposed that cyclin E–CDK2 cyclin E–CDK2 and cell cycle arrest by retrovirally binding to E2F-4/DP/p107 results in the formation of expressed p27, through the sequestration of p27 into heat- 5329 K.Alevizopoulos et al. labile complexes (Vlach et al., 1996) (Table I). The same Hall and Peters, 1996). These observations strongly sug- mechanism was involved in activation of cyclin E–CDK2 gest that other genetic lesions resulting in the functional by MycER in quiescent fibrobasts (Steiner et al., 1995; suppression of the p16/pRb pathway should also be Perez-Roger et al., 1997; J.Vlach and B.Amati, unpub- tumorigenic. lished data). In some cell lines, p27 has a role in limiting Recently, aberrant expression of cyclin E has been the rate of G –S progression during proliferation (Coats linked to tumorigenesis (see, for example, Keyomarsi et al., 1996), whereas constitutive expression of cyclin E et al., 1994; Porter et al., 1997). It was suggested that (Ohtsubo and Roberts, 1993) or Myc (Karn et al., 1989) cyclin E–CDK2 replaces CDK4 function in tumor cells can accelerate it. These observations suggest that Myc, that express high p16 levels by phosphorylating pRb by suppressing p27 function, can accelerate the activation (Gray-Bablin et al., 1996). However, our data show that of cyclin E during G . This, through the positive cyclin cyclin E bypasses p16/pRb function through a different E–E2F feedback loop discussed above, should further pathway. Although cyclin E was also shown to have a enhance cyclin E expression. Consistent with this view, mild oncogenic potential in transgenic mice (Bortner and MycER activation induced expression of E2F target genes Rosenberg, 1997), its causal involvement in human tumors (Jansen-Du¨rr et al., 1993; Pusch et al., 1997b), and in remains to be investigated. particular of cyclin E (Perez-Roger et al., 1997; Pusch The widespread oncogenic potential of myc family et al., 1997a), although there is no evidence that any of genes and their involvement in numerous malignancies in these genes is a direct transcriptional target of Myc. both humans and animals are well established (reviewed The cdc25A gene, on the other hand, is a Myc target by Marcu et al., 1992; Garte, 1993). As discussed above, (Galaktionov et al., 1996). However, overexpression of Myc antagonizes p27 function, but loss of p27 is not per Cdc25A rescued neither p16- nor p27-induced arrest se a tumorigenic event (Fero et al., 1996; Kiyokawa et al., (Vlach et al., 1996; Table I), suggesting that additional 1996; Nakayama et al., 1996). This suggests that Myc Myc/Max target genes are involved. may have other direct targets, which may include the p27- In addition to its upstream action, Myc may also be a related CKIs p21 and p57, as well as p53 (Hermeking downstream target of pocket proteins. First, the c-myc et al., 1995). Independently of this aspect, bypassing the gene was proposed to be an E2F target; however, unlike tumour suppressor function of the p16/pRb pathway may other E2F targets, it was not induced upon adenovirus be an important, if not the essential, aspect of Myc’s expression of E2F-1 (DeGregori et al., 1995a). Second, oncogenic activity. In support of this hypothesis, fibroblasts p107 can associate with the Myc protein and suppress its isolated from p16-null mouse embryos are transformed transactivation potential in transient transfections by ras alone, whereas wild-type cells undergo premature (Beijersbergen et al., 1994; Gu et al., 1994; Hoang et al., senescence (Serrano et al., 1996, 1997). On the other hand, 1995). In these assays, Myc bypassed cell cycle arrest by transformation of wild-type cells requires cooperation of p107, but not by pRb (Beijersbergen et al., 1994). In ras with immortalizing oncogenes, such as c-myc (Land contrast, microinjected GST–Myc proteins prevented cell et al., 1983). Thus, loss of p16 mimics myc activation, cycle arrest by pRb (Goodrich and Lee, 1992). Since the and vice versa. Based on this, we predict that myc mechanisms underlying these effects were not studied, activation may partly release the pressure to mutate these discrepancies remain unexplained and illustrate the components of the p16/pRb pathway during tumorigenesis. potential dangers of overexpression. Although we cannot rule out that the Myc–p107 interaction plays a role in Materials and methods overcoming p16 (or p27)-induced arrest by Myc, it is unlikely to account for our findings. First, in our system, Retroviral expression vectors we did not detect the Myc–p107 interaction. Second, cells The retroviral vectors pBabe-Puro (pBP), -Bleo (pBB), - Neo2 (pBN2) expressing Myc and p16 proliferate in the presence of not and- Hygro2 (pBH2) were described previously (Morgenstern and Land, only active p107, but also of pRb and p130. Third, Myc 1990; Vlach et al., 1996). cDNAs encoding human cyclins A, D1–3 and E, as well as Myc, HApRb, HAΔp34, HA-E2F-1 to -3 and HA-E2F1eco was expressed at low levels in our studies, making titration proteins were subcloned in pBN2, pBH2 and/or pBB, as required. HA- of p107 by Myc a very unlikely mechanism. E2F-4 and HA-E2F-4Δ4 were expressed from the pRcCMVneo retroviral In summary, we propose that the overcoming of p16- vector (a kind gift from W.Krek). cDNAs encoding the human CKIs induced growth arrest by Myc is mainly the consequence p15, p16, p18 and p27 were subcloned in pBP. of its upstream action on cyclin E–CDK2: Myc suppresses Retroviral infections, cell cycle analysis and biochemical p27 function, allowing an elevation of cyclin E activity analysis above the threshold required to bypass the p16/pRb block. High-titer retroviral supernatants (510 /ml) were generated as In situations in which p16 is not overexpressed, the same described (Vlach et al., 1996). Infected Rat1 or NIH-3T3 cells were activity of Myc accelerates hyperphosphorylation of pRb selected with the appropriate drug: G418 (Calbiochem, 1000 μg/ml), by cyclin E–CDK2 in concert with cyclin D–CDK4. hygromycin (Calbiochem, 150 μg/ml), phleomycin (Cayla, 50 μg/ml) or puromycin (Sigma, 2.5 μg/ml). Serial infections of cell pools, preparation of cell lysates and biochemical analysis (immunoblots, immunoprecipit- Relevance of bypassing p16/pRb function for ations) were as previously described (Vlach et al., 1996). Gel retardation tumorigenesis assays were performed from infected cells as previously described The p16/pRb pathway has a fundamental role in sup- (Beijersbergen et al., 1995). For cell cycle analysis, cells were labeled with 33 μM BrdU for 30 min, and analyzed by either two-dimentional pressing tumorigenesis. Both p16 and pRb are encoded flow cytometry (Vlach et al., 1996), or by immunocytochemical detection by tumor suppressor genes, as indicated by loss of function of BrdU. For the latter, cells were fixed in 3.7% formaldehyde in in different tumor types. On the other hand, oncogenic phosphate-buffered saline (PBS) (5 min), permeabilized with acetone activation in human tumors has been reported for the at –20°C (30 s), treated with 1.5 M HCl (10 min), blocked in PBS with genes encoding cyclins D1, D2 and CDK4 (reviewed by 1% bovine serum albumin (BSA) (1–2 min), incubated for 1 h with the 5330 Cyclin E and Myc bypass the p16/pRb pathway primary antibody (undiluted mouse anti-BrdUNuclease, Amersham gene that regulates E2F activity and interacts with cyclin E in No. RPN202), and for 1 h with the secondary antibody (goat anti-mouse Drosophila. Genes Dev., 10, 1206–1218. Fab–FITC, Sigma, dil. 1/200 in PBS  1% BSA) with Hoechst Duronio,R.J. and O’Farrell,P.H. (1995) Developmental control of the G1 (0.5 μg/ml, Calbiochem), and mounted with FluorSave™ Reagent to S transition in Drosophila: cyclin E is a limiting downstream target (Calbiochem). All steps were performed at room temperature and of E2F. Genes Dev., 9, 1456–1468. separated by several washes in PBS (with 1% BSA after the blocking Duronio,R.J., O’Farrell,P.H., Xie,J.E., Brook,A. and Dyson,N. (1995) step). The transcription factor E2F is required for S phase during Drosophila embryogenesis. Genes Dev., 9, 1445–1455. Duronio,R.J., Brook,A., Dyson,N. and O’Farrel,P.H. (1996) E2F-induced Antibodies S-phase entry requires cyclin E. Genes Dev., 10, 2505–2513. The following antibodies were used: against CDK2, M2 (Santa Cruz Eilers,M., Schirm,S. and Bishop,J.M. (1991) The MYC protein activates sc-163); CDK4, C22 (sc-260) or H303 (sc-749); CDK6, C21 (sc-177); transcription of the alpha-prothymosin gene. EMBO J., 10, 133–141. CDC2(17), (sc-54); cyclin A, H-432 (sc-751); rodent cyclin E, M20 Fero,M.L. et al. (1996) A syndrome of multiorgan hyperplasia with (sc-481); human cyclin E, C19 (sc198); cyclin D1, 72-13G (sc-450); features of gigantism, tumorigenesis, and female sterility in p27(kip1)- cyclin D2, DCS-5 (from J.Bartek); cyclin D3, C16 (sc-182); cyclin B1, deficient mice. Cell, 85, 733–744. GNS1 (sc-245); pRb, 14001A (Pharmingen); p107, C18 (sc-318); p130, Field,S.J., Tsai,F.Y., Kuo,F., Zubiaga,A.M., Kaelin,W.G., Livingston, C20 (sc-317); E2F-1, C20 (sc-193); p15, C20 (sc-612); HA-tag, D.M., Orkin,S.H. and Greenberg,M.E. (1996) E2F-1 functions in mice HA11(Babco); T7-tag, T7 (Novagen); p16, DCS 50.2 (from G.Peters); to promote apoptosis and suppress proliferation. Cell, 85, 549–561. p18, 11256 (from Y.Xiong). Galaktionov,K., Chen,X. and Beach,D. (1996) Cdc25 cell-cycle phosphatase as a target of c-myc. Nature, 382, 511–517. Garte,S. (1993) The c-myc oncogene in tumor progression. Crit. Rev. Acknowledgements Oncogen., 4, 435–449. Geng,Y., Eaton,E.N., Picon,M., Roberts,J.M., Lundberg,A.S., Gifford,A., We acknowledge the following colleagues for essential reagents, fruitful Sardet,C. and Weinberg,R.A. (1996) Regulation of cyclin E discussions, advice on the manuscript and/or sharing unpublished data: transcription by E2Fs and retinoblastoma protein. Oncogene, 12, Michel Aguet, Nassos Alevizopoulos, Jiri Bartek, Rene´ Bernards, James 1173–1180. DeGregori, Martin Eilers, David Franklin, Susan Gasser, Peter Greasley, Goodrich,D.W. and Lee,W.H. (1992) Abrogation by c-myc of G1 phase Paul Hamel, Markus Hengstschla¨ger, Richard Iggo, Willy Krek, Hartmut arrest induced by RB protein but not by p53. Nature, 360, 177–179. Land, Nathalie Lauper, Trevor Littlewood, Kirsten Mundt, Joseph Nevins, Gray-Bablin,J., Zalvide,J., Fox,M.P., Knickerbocker,C.J., DeCaprio,J.A. Matthias Peter, Gordon Peters, Charles Sherr, Viesturs Simanis and Yue and Keyomarsi,K. (1996) Cyclin E, a redundant cyclin in breast Xiong. Our special gratitude goes to Rene´ Bernards and Ron Kerkhoven cancer. Proc. Natl Acad. Sci. USA, 93, 15215–15220. for their hospitality and precious advice on E2F gel retardation assays. Gu,W., Bhatia,K., Magrath,I., Dang,C. and Dalla-Favera,R. (1994) K.A. and S.H. were supported by the Swiss National Science Foundation, Binding and suppression of the Myc transcriptional activation domain and J.V. by post-doctoral fellowships from the Roche Research by p107. Science, 264, 251–254. Foundation and the Swiss Cancer League. B.A. is a recipient of a START Guan,K.L., Jenkins,C.W., Li,Y., Nichols,M.A., Wu,X., O’Keefe,C.L., fellowship and of a research grant from the Swiss National Science Matera,A.G. and Xiong,Y. (1994) Growth suppression by p18, a Foundation. p16INK4/MTS1- and p14INK4B/MTS2-related CDK6 inhibitor, correlates with wild-type pRb function. Genes Dev., 8, 2939–2952. Hall,M. and Peters,G. (1996) Genetic alterations of cyclins, cyclin- References dependent kinases, and cdk inhibitors in human cancer. Adv. Cancer Adams,P.D. and Kaelin,W.G. (1996) The cellular effects of E2F Res., 68, 67–108. overexpression. Curr. Topics Microbiol. Immunol., 208, 79–93. Hamel,P.A., Gill,R.M., Phillips,R.A. and Gallie,B.L. (1992) Regions Alexandrow,M.G., Kawabata,M., Aakre,M. and Moses,H.L. (1995) controlling hyperphosphorylation and conformation of the Overexpression of the c-Myc oncoprotein blocks the growth-inhibitory retinoblastoma gene product are independent of domains required for response but is required for the mitogenic effects of transforming transcriptional repression. Oncogene, 7, 693–701. growth factor β1. Proc. Natl Acad. Sci. USA, 92, 3239–3243. Henriksson,M. and Lu¨scher,B. (1996) Proteins of the Myc network: Assoian,R.K. (1997) Anchorage-dependent cell cycle progression. J. Cell essential regulators of cell growth and differentiation. Adv. Cancer Biol., 136, 1–4. Res., 68, 109–182. Beijersbergen,R.L., Hijmans,E.M., Zhu,L. and Bernards,R. (1994) Hermeking,H., Funk,J.O., Reichert,M., Ellwart,J.W. and Eick,D. (1995) Interaction of c-Myc with the pRb-related protein p107 results in Abrogation of p53-induced cell cycle arrest by c-Myc: evidence for inhibition of c-Myc-mediated transactivation. EMBO J., 13, 4080– an inhibitor of p21WAF1/CIP1/SDI1. Oncogene, 11, 1409–1415. 4086. Herrera,R.E., Sah,V.P., Williams,B.O., Makela,T.P., Weinberg,R.A. and Beijersbergen,R.L., Carlee,L., Kerkhoven,R.M. and Bernards,R. (1995) Tyler,J. (1996) Altered cell cycle kinetics, gene expression, and G1 Regulation of the retinoblastoma protein-related p107 by G1 cyclin restriction point regulation in Rb-deficient fibroblasts. Mol. Cell. Biol., complexes. Genes Dev., 9, 1340–1353. 16, 2402–2407. Bortner,D.M. and Rosenberg,M.P. (1997) Induction of mammary gland Hinds,P.W., Mittnacht,S., Dulic,V., Arnold,A., Reed,S.I. and Weinberg, hyperplasia and carcinomas in transgenic mice expressing human R.A. (1992) Regulation of retinoblastoma protein functions by ectopic cyclin E. Mol. Cell. Biol., 17, 453–459. expression of human cyclins. Cell, 70, 993–1006. Botz,J., Zerfass-Thome,K., Spitkovsky,D., Delius,H., Vogt,B., Eilers,M., Hirai,H., Roussel,M.F., Kato,J.Y., Ashmun,R.A. and Sherr,C.J. (1995) Hatzigeorgiou,A. and Jansen-Du¨rr,P. (1996) Cell cycle regulation of Novel INK4 proteins, p19 and p18, are specific inhibitors of the the murine cyclin E gene depends on an E2F binding site in the cyclin D-dependent kinases CDK4 and CDK6. Mol. Cell. Biol., 15, promoter. Mol. Cell. Biol., 16, 3401–3409. 2672–2681. Chen,P.-L., Scully,P., Shew,J.-Y., Wang,J.Y.J. and Lee,W.-H. (1989) Hoang,A.T., Lutterbach,B., Lewis,B.C., Yano,T., Chou,T.Y., Barrett,J.F., Phosphorylation of the retinoblastoma gene product is modulated Raffeld,M., Hann,S.R. and Dang,C.V. (1995) A link between increased during the cell cycle and cellular differentiation. Cell, 58, 1193–1198. transforming activity of lymphoma-derived MYC mutant alleles, their Coats,S., Flanagan,W.M., Nourse,J. and Roberts,J.M. (1996) defective regulation by p107, and altered phosphorylation of the Requirement of p27(kip1) for restriction point control of the fibroblast c-Myc transactivation domain. Mol. Cell. Biol., 15, 4031–4042. cell cycle. Science, 272, 877–880. Hofmann,F. and Livingston,D.M. (1996) Differential effects of cdk2 and Cobrinik,D. (1996) Regulatory interactions among E2Fs and cell cycle cdk3 on the control of pRb and E2Ff function during G(1) exit. Genes control proteins. Curr. Topics Microbiol. Immunol., 208, 31–61. Dev., 10, 851–861. DeGregori,J., Kowalik,T. and Nevins,J.R. (1995a) Cellular targets for Jansen-Du¨rr,P., Meichle,A., Steiner,P., Pagano,M., Finke,K., Botz,J., activation by the E2F1 transcription factor include DNA synthesis- Wessbecher,J., Draetta,G. and Eilers,M. (1993) Differential modulation and G1/S-regulatory genes. Mol. Cell. Biol., 15, 4215–4224. of cyclin gene expression by MYC. Proc. Natl Acad. Sci. USA, 90, DeGregori,J., Leone,G., Ohtani,K., Miron,A. and Nevins,J.R. (1995b) 3685–3689. E2F-1 accumulation bypasses a G1 arrest resulting from the inhibition Johnson,D.G., Schwartz,J.K., Cress,W.D. and Nevins,J.R. (1993) of G1 cyclin-dependent kinase activity. Genes Dev., 9, 2873–2887. Expression of transcription factor E2F-1 induces quiescent cells to Du,W., Vidal,M., Xie,J.E. and Dyson,N. (1996) RBF, a novel RB-related enter S phase. Nature, 365, 349–352. 5331 K.Alevizopoulos et al. Karn,J., Watson,J.V., Lowe,A.D., Green,S.M. and Vedeckis,W. (1989) Polyak,K., Lee,M.H., Erdjument,B.H., Koff,A., Roberts,J.M., Tempst,P. Regulation of cell cycle duration by c-myc levels. Oncogene, 4, and Massague,J. (1994) Cloning of p27Kip1, a cyclin-dependent 773–787. kinase inhibitor and a potential mediator of extracellular antimitogenic Keyomarsi,K., O’Leary,N., Molnar,G., Lees,E., Fingert,H.J. and signals. Cell, 78, 59–66. Pardee,A.B. (1994) Cyclin E, a potential prognostic marker for breast Porter,P.L., Malone,K.E., Heagerty,P.J., Alexander,G.M., Gatti,L.A., cancer. Cancer Res., 54, 380–385. Firpo,E.J., Daling,J.R. and Roberts,J.M. (1997) Expression of cell- Kiyokawa,H. et al. (1996) Enhanced growth of mice lacking the cyclin- cycle regulators p27(kip1) and cyclin E, alone and in combination, dependent kinase inhibitor function of p27(kip1). Cell, 85, 721–732. correlate with survival in young breast cancer patients. Nature Med., Knoblich,J.A., Sauer,K., Jones,L., Richardson,H., Saint,R. and Lehner, 3, 222–225. C.F. (1994) Cyclin E controls S phase progression and its down- ¨ Pusch,O., Bernaschek,G., Eilers,M. and Hengstschlager,M. (1997a) regulation during Drosophila embryogenesis is required for the arrest Activation of c-Myc uncouples DNA replication from activation of of cell proliferation. Cell, 77, 107–120. G1-cyclin-dependent kinases. Oncogene, 15, 649–656. Koh,J., Enders,G.H., Dynlacht,B.D. and Harlow,E. (1995) Tumour- Pusch,O., Soucek,T., Hengstschla¨ger-Ottnad,E., Bernaschek,G. and derived p16 alleles encoding proteins defective in cell-cycle inhibition. Hengstschla¨ger,M. (1997b) Cellular targets for activation by c-Myc Nature, 375, 506–510. include the DNA metabolism enzyme thymidine kinase. DNA Cell Krek,W., Ewen,M.E., Shirodkar,S., Arany,Z., Kaelin,W.J. and Biol., 16, 737–747. Livingston,D.M. (1994) Negative regulation of the growth-promoting Qin,X.-Q., Livingston,D.M., Kaelin,W.G. and Adams,P.D. (1994) transcription factor E2F-1 by a stably bound cyclin A-dependent Deregulated transcription factor E2F-1 expression leads to S-phase protein kinase. Cell, 78, 161–172. entry and p53-mediated apoptosis. Proc. Natl Acad. Sci. USA, 91, Krude,T., Jackman,M., Pines,J. and Laskey,R.A. (1997) Cyclin/Cdk- 10918–10922. dependent initiation of DNA replication in a human cell-free system. Qin,X.-Q., Livingston,D.M., Ewen,M., Sellers,W.R., Arany,Z. and Cell, 88, 109–119. Kaelin,W.G. (1995) The transcription factor E2F-1 is a downstream Lam,E.W. and Watson,R.J. (1993) An E2F-binding site mediates cell- target of RB action. Mol. Cell. Biol., 15, 742–755. cycle regulated repression of mouse B-myb transcription. EMBO J., Quelle,D.E. et al. (1995) Cloning and characterization of murine 12, 2705–2713. p16INK4a and p15INK4b genes. Oncogene, 11, 635–645. Land,H., Parada,L.F. and Weinberg,R.A. (1983) Tumorigenic conversion Resnitzky,D. and Reed,S.I. (1995) Different roles for cyclins D1 and E of primary embryo fibroblasts requires at least two cooperating in regulation of the G1-to-S transition. Mol. Cell. Biol., 15, 3463–3469. oncogenes. Nature, 304, 596–602. Schwarz,J.K., Devoto,S.H., Smith,E.J., Chellappan,S.P., Jakoi,L. and Lees,E., Faha,B., Dulic,V., Reed,S.I. and Harlow,E. (1992) Cyclin Nevins,J.R. (1993) Interactions of the p107 and Rb proteins with E2F E/cdk2 and cyclin A/cdk2 kinases associate with p107 and E2F in a during the cell proliferation response. EMBO J., 12, 1013–1020. temporally distinct manner. Genes Dev., 6, 1874–1885. Serrano,M., Lee,H., Chin,L., Cordon,C.C., Beach,D. and DePinho,R.A. Lukas,J., Muller,H., Bartkova,J., Spitkovsky,D., Kjerulff,A.A., Jansen- (1996) Role of the INK4a locus in tumor suppression and cell Durr,P., Strauss,M. and Bartek,J. (1994) DNA tumor virus oncoproteins mortality. Cell, 85, 27–37. and retinoblastoma gene mutations share the ability to relieve the Serrano,M., Lin,A.W., McCurrach,M.E., Beach,D. and Lowe,S.W. (1997) cell’s requirement for cyclin D1 function in G1. J. Cell Biol., 125, Oncogenic ras provokes premature cell senescence associated with 625–638. accumulation of p53 and p16INK4a. Cell, 88, 593–602. Lukas,J., Bartkova,J., Rohde,M., Strauss,M. and Bartek,J. (1995a) Cyclin Shan,B. and Lee,H.-W. (1994) Deregulated expression of E2F-1 induces D1 is dispensable for G1 control in retinoblastoma gene-deficient cells S-phase entry and leads to apoptosis. Mol. Cell. Biol., 14, 8166–8173. independently of cdk4 activity. Mol. Cell. Biol., 15, 2600–2611. Sherr,C.J. (1994) G1 phase progression: cycling on cue. Cell, 79, Lukas,J., Parry,D., Aagaard,L., Mann,D.J., Bartkova,J., Strauss,M., 551–555. Peters,G. and Bartek,J. (1995b) Retinoblastoma-protein-dependent Sherr,C.J. (1995) D-type cyclins. Trends Biochem. Sci., 20, 187–190. cell-cycle inhibition by the tumour suppressor p16. Nature, 375, Sherr,C.J. and Roberts,J.M. (1995) Inhibitors of mammalian G1 cyclin- 503–506. dependent kinases. Genes Dev., 9, 1149–1163. Lukas,J., Otzen Petersen,B., Holm,K., Bartek,J. and Helin,K. (1996) Soucek,T., Pusch,O., Hengstschla¨ger-Ottnad,E., Adams,P.D. and Deregulated expression of E2F family members induces S-phase entry Hengstschla¨ger,M. (1997) Deregulated expression of E2F-1 induces and overcomes p16INK4A-mediated growth suppression. Mol. Cell. cyclin A- and E-associated kinase activities independently from cell Biol., 16, 1047–1057. cycle position. Oncogene, 14, 2251–2257. Mann,D.J. and Jones,N.C. (1996) E2F-1 but not E2F-4 can overcome Steiner,P., Philipp,A., Lukas,J., Godden-Kent,D., Pagano,M., p16-induced G1 cell-cycle arrest. Curr. Biol., 6, 474–483. Mittnacht,S., Bartek,J. and Eilers,M. (1995) Identification of a Myc- Marcu,K.B., Bossone,S.A. and Patel,A.J. (1992) myc function and dependent step during the formation of active G cyclin–cdk regulation. Annu. Rev. Biochem., 61, 809–860. complexes. EMBO J., 14, 4814–4826. Medema,R.H., Herrera,R.E., Lam,F. and Weinberg,R.A. (1995) Growth Stillman,B. (1996) Cell cycle control of DNA replication. Science, 274, suppression by p16ink4 requires functional retinoblastoma protein. 1659–1664. Proc. Natl Acad. Sci. USA, 92, 6289–6293. Strausfeld,U.P., Howell,M., Rempel,R., Maller,J.L., Hunt,T. and Blow,J.J. Morgan,D.O. (1995) Principles of CDK regulation. Nature, 374, 131–134. (1994) CIP1 blocks the initiation of DNA replication in Xenopus Morgenstern,J.P. and Land,H. (1990) Advanced mammalian gene extracts by inhibition of cyclin-dependent kinases. Curr. Biol., 4, transfer: high titre retroviral vectors with multiple drug selection 876–883. markers and a complementary helper-free packaging cell line. Nucleic Templeton,D.J., Park,S.H., Lanier,L. and Weinberg,R.A. (1991) Acids Res., 18, 3587–3596. Nonfunctional mutants of the retinoblastoma protein are characterized Nakayama,K., Ishida,N., Shirane,M., Inomata,A., Inoue,T., Shishido,N., by defects in phosphorylation, viral oncoprotein association, and Hori,I., Loh,D.Y. and Nakayama,K. (1996) Mice lacking p27(kip1) nuclear tethering. Proc. Natl Acad. Sci. USA, 88, 3033–3037. display increased body size, multiple organ hyperplasia, retinal Toyoshima,H. and Hunter,T. (1994) p27, a novel inhibitor of G1 cyclin– dysplasia, and pituitary tumors. Cell, 85, 707–720. Cdk protein kinase activity, is related to p21. Cell, 78, 67–74. Nevins,J.R. (1992) E2F: a link between the Rb tumor suppressor protein Vairo,G., Livingston,D.M. and Ginsberg,D. (1995) Functional interaction and viral oncoproteins. Science, 258, 424–429. between E2F-4 and p130: evidence for distinct mechanisms underlying Ohtani,K., DeGregori,J. and Nevins,J.R. (1995) Regulation of the cyclin growth suppression by different retinoblastoma protein family E gene by transcription factor E2F1. Proc. Natl Acad. Sci. USA, 92, members. Genes Dev., 9, 869–881. 12146–12150. Vlach,J., Hennecke,S., Alevizopoulos,K., Conti,D. and Amati,B. (1996) Ohtsubo,M. and Roberts,J.M. (1993) Cyclin-dependent regulation of G1 Kip1 Growth arrest by the cyclin-dependent kinase inhibitor p27 is in mammalian fibroblasts. Science, 259, 1908–1912. abrogated by c-Myc. EMBO J., 15, 6595–6604. Ohtsubo,M., Theodoras,A.M., Schumacher,J., Roberts,J.M. and Wang,J.Y.J. (1997) Retinoblastoma protein in growth suppression and Pagano,M. (1995) Human cyclin E, a nuclear protein essential for the death protection. Curr. Opin. Genet. Dev., 7, 39–45. G1-to-S phase transition. Mol. Cell. Biol., 15, 2612–2624. Weinberg,R.A. (1995) The retinoblastoma protein and cell cycle control. Perez-Roger,I., Solomon,D.L.C., Sewing,A. and Land,H. (1997) Myc Cell, 81, 323–330. activation of cyclin E/CDK2 kinase involves induction of cyclin E gene transcription and inhibition of p27Kip1 binding to newly formed Weinberg,R.A. (1996) E2F and cell proliferation: a world turned upside complexes. Oncogene, 14, 2373–2381. down. Cell, 85, 457–459. 5332 Cyclin E and Myc bypass the p16/pRb pathway Wu,C.L., Classon,M., Dyson,N. and Harlow,E. (1996) Expression of dominant-negative mutant DP-1 blocks cell cycle progression in G1. Mol. Cell. Biol., 16, 3698–3706. Yamasaki,L., Jacks,T., Bronson,R., Goillot,E., Harlow,E. and Dyson,N.J. (1996) Tumor induction and tissue atrophy in mice lacking E2F-1. Cell, 85, 537–548. Zerfass-Thome,K., Schulze,A., Zwerschke,W., Vogt,B., Helin,K., Bartek,J., Henglein,B. and Jansen-Du¨rr,P. (1997) P27(kip1) blocks cyclin E-dependent transactivation of cyclin a gene expression. Mol. Cell. Biol., 17, 407–415. Zhu,L., van den Heuvel,S., Helin,K., Fattaey,A., Ewen,M., Livingston, D.M., Dyson,N. and Harlow,E. (1993) Inhibition of cell proliferation by p107, a relative of the retinoblastoma protein. Genes Dev., 7, 1111–1125. Zwicker,J. and Mu¨ller,R. (1997) Cell-cycle regulation of gene expression by transcriptional repression. Trends Genet., 13, 3–6. Received on March 10, 1997; revised on May 27, 1997 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png The EMBO Journal Springer Journals

Cyclin E and c‐Myc promote cell proliferation in the presence of p16INK4a and of hypophosphorylated retinoblastoma family proteins

Download PDF
12 pages

Loading next page...
 
/lp/springer-journals/cyclin-e-and-c-myc-promote-cell-proliferation-in-the-presence-of-v0nwwuSzwi

References (111)

Publisher
Springer Journals
Copyright
Copyright © European Molecular Biology Organization 1997
ISSN
0261-4189
eISSN
1460-2075
DOI
10.1093/emboj/16.17.5322
Publisher site
See Article on Publisher Site

Abstract

The EMBO Journal Vol.16 No.17 pp.5322–5333, 1997 Cyclin E and c-Myc promote cell proliferation in the INK4a presence of p16 and of hypophosphorylated retinoblastoma family proteins (Guan et al., 1994; Polyak et al., 1994; Toyoshima and Konstantinos Alevizopoulos, Jaromir Vlach, Hunter, 1994; Hirai et al., 1995; Quelle et al., 1995; Vlach Silke Hennecke and Bruno Amati et al., 1996), and numerous observations link CKIs to Swiss Institute for Experimental Cancer Research, CH-1066 Epalinges, growth arrest and/or differentiation in diverse cell types Switzerland (reviewed by Sherr and Roberts, 1995; Assoian, 1997). Corresponding author The best characterized substrates of cyclin–CDK com- e-mail: [email protected] plexes are the retinoblastoma family proteins pRb, p107 and p130 (or ‘pocket proteins’), which negatively regulate Retroviral expression of the cyclin-dependent kinase cell cycle progression and are inactivated through phos- INK4a (CDK) inhibitor p16 in rodent fibroblasts induces phorylation by CDKs (Weinberg, 1995). In their active, dephosphorylation of pRb, p107 and p130 and leads hypophosphorylated forms, pocket proteins bind to and to G arrest. Prior expression of cyclin E allows S-phase negatively regulate the activities of several targets, such entry and long-term proliferation in the presence of as the heterodimeric E2F–DP transcription factors. Five p16. Cyclin E prevents neither the dephosphorylation E2F (E2F-1 to -5) and three DP proteins (DP-1 to -3) of pRb family proteins, nor their association with E2F have been identified. E2F-1, -2 and -3 bind predominantly proteins in response to p16. Thus, cyclin E can bypass pRb, E2F-4 binds p107 and p130, and E2F-5 binds the p16/pRb growth-inhibitory pathway downstream predominantly p130 (reviewed by Cobrinik, 1996). of pRb activation. Retroviruses expressing E2F-1, -2 Depending on the target promoters, E2F complexes can or -3 also prevent p16-induced growth arrest but are function as either transcriptional activators or repressors Kip1 ineffective against the cyclin E–CDK2 inhibitor p27 , (in association with pocket proteins). E2F target genes suggesting that E2F cannot substitute for cyclin E include regulators of S-phase entry (e.g. B-myb, CDC2, activity. Thus, cyclin E possesses an E2F-independent cyclins E and A) and genes required for DNA replication function required to enter S-phase. However, cyclin E (e.g. DHFR, DNA polα, thymidine kinase) (DeGregori may not simply bypass E2F function in the presence et al., 1995a; reviewed by Nevins, 1992; Zwicker and of p16, since it restores expression of E2F-regulated Mu¨ller, 1997). genes such as cyclin A or CDC2. Finally, c-Myc When ectopically expressed, E2F-1 is sufficient to bypasses the p16/pRb pathway with effects indistin- induce entry of quiescent cells into S-phase (but also guishable from those of cyclin E. We suggest that this apoptosis) (Johnson et al., 1993; Qin et al., 1994; Shan effect of Myc is mediated by its action upstream of and Lee, 1994; Lukas et al., 1996), a property shared by cyclin E–CDK2, and occurs via the neutralization other E2F proteins (Lukas et al., 1996; reviewed by Kip1 of p27 family proteins, rather than induction of Adams and Kaelin, 1996). E2F-1 and E2F-4 can override Cdc25A. Our data imply that oncogenic activation of pRb- and p130-induced G arrest, respectively (Zhu et al., c-Myc, and possibly also of cyclin E, mimics loss of 1993; Qin et al., 1995; Vairo et al., 1995). E2F-1 also the p16/pRb pathway during oncogenesis. prevents G arrest imposed either by antibodies neutraliz- Keywords: CDK/cyclin/Myc/p16/retinoblastoma ing cyclin D–CDK4 function or by p16, in spite of the p16-induced dephosphorylation of cellular pRb (DeGregori et al., 1995b; Lukas et al., 1996; Mann and Jones, 1996). Expression of p16 or neutralization of cyclin Introduction D–CDK4 also fail to induce G arrest in cells lacking functional pRb (Lukas et al., 1994, 1995a,b; Koh et al., The activities of the cyclin-dependent kinases CDK4 and 1995; Medema et al., 1995). Thus, the essential role of CDK6, which associate with D-type cyclins, and of CDK2, cyclin D–CDK4 is to inactivate pRb, thereby allowing which associates with cyclins E or A, are rate-limiting for E2F function. In summary, p16, cyclin D1–3, CDK4/6, progression through G and into S-phase of the vertebrate pRb and E2F can be ordered within a single G -regulatory cell cycle (reviewed by Sherr, 1994, 1995). Besides pathway (the p16/pRb pathway). association with cyclins, the activity of CDKs is controlled The p16/pRb pathway plays a critical role in suppressing by site-specific phosphorylation or dephosphorylation and tumorigenesis. Several of its components are genetically by association with a group of inhibitory proteins collect- altered in various human malignancies, resulting either in ively called CKIs (reviewed by Morgan, 1995; Sherr and Cip1/Waf1 oncogenic activation (cyclins D1, D2 and CDK4) or loss Roberts, 1995). CKIs of the Kip/Cip family, p21 , Kip1 Kip2 of function (pRb and p16) (reviewed by Hall and Peters, p27 and p57 , can associate with and inhibit all 1996). p16-null mice are viable, but develop malignancies known G cyclin–CDK complexes. CKIs of the INK4 INK4a INK4b INK4c INK4d family, p16 , p15 , p18 and p19 bind at increased rates (Serrano et al., 1996). Strikingly, E2F-1- CDK4 and CDK6 and can interfere with cyclin–CDK null mice also have a tumour-prone phenotype (Field interactions. Ectopic expression of CKIs induces G arrest et al., 1996; Yamasaki et al., 1996). Thus, the major role 5322 © Oxford University Press Cyclin E and Myc bypass the p16/pRb pathway of E2F-1 in vivo might be to relay the repressive action complexes (Steiner et al., 1995), and this also occurs of pRb, rather than its own activating function. It should through the suppression of p27 function (Perez-Roger be noted here that pRb has other targets in cells, which et al., 1997; J.Vlach and B.Amati, unpublished data). may also be critical for its growth- and tumour-suppressive However, no link was yet established between Myc and functions (Weinberg, 1996; Wang, 1997). p16 function. CDK2 and its partners, cyclins E and A, also phos- In this study, we addressed whether cells infected with phorylate and inactivate pocket proteins (Weinberg, 1995). retroviruses expressing G cyclins, Cdc25A, E2F proteins For example, in transfected SAOS-2 cells, the G arrest or Myc could escape growth arrest induced by p16, and induced by pRb is relieved by cyclins A or E, and this compared the effects of p16 and p27 in those cells. Our correlates with hyperphosphorylation of pRb (Hinds et al., data demonstrate that cyclin E promotes cell proliferation 1992). Moreover, rapid activation of cyclin E–CDK2 in the presence of elevated p16 levels, without inducing upon induction of Myc (see below) coincides with pRb phosphorylation of pocket proteins. We suggest that cyclin phosphorylation, whereas activation of cyclin A–CDK2 E cannot simply substitute for E2F function, but rather and cyclin D–CDK4 follow later in this particular system rescues it in the presence of p16. Reciprocally, the require- (Steiner et al., 1995). Cyclin E–CDK2 appears to have ment for cyclin E in G –S progression cannot be substituted additional function(s) in G –S progression, distinct from by E2F. Finally, we demonstrate that Myc also bypasses pRb phosphorylation and E2F activation. First, inducible the p16/pRb pathway, and argue that this effect is mediated expression of cyclin E in Rat1 fibroblasts accelerates G –S by the positive action of Myc on cyclin E. progression upon serum stimulation, without affecting the kinetics of pRb phosphorylation (Resnitzky and Reed, Results 1995). Second, G arrest induced by dominant-negative CDK2 in U2OS cells is not relieved by the SV40 large T Myc and cyclin E prevent growth arrest by INK4 antigen (LT), although LT rescues E2F function in this proteins system (Hofmann and Livingston, 1996). Third, unlike The ability of Myc and various cell cycle regulators to D-type cyclins, cyclin E is essential for cell cycle progres- suppress the growth-inhibitory function of INK4 family sion in pRb-deficient cells (Ohtsubo et al., 1995). proteins (p15, p16 or p18) was investigated in Rat1 The functional hierarchy between the pRb/E2F pathway fibroblasts as previously described for Myc and p27 (Vlach and cyclin E is complicated further by the finding that the et al., 1996). First, cells were infected with retroviruses cyclin E gene is itself an E2F target. Cyclin E is induced expressing Myc, Cdc25A, cyclins A and E, or D1–3. by ectopic expression of E2F-1 (DeGregori et al., 1995a; Serial infections using the pBabe series of retroviral vectors Soucek et al., 1997), and functional E2F-binding sites in conferring resistance to neomycin (pBNeo), hygromycin the cyclin E promoter are activated by E2F-1 and repressed (pBHygro) or phleomycin (pBBleo) (Morgenstern and by pRb (Ohtani et al., 1995; Geng et al., 1996). Further- Land, 1990) allowed co-expression of up to three exo- more, pRb-deficient fibroblasts express enhanced levels genous proteins prior to INK4 proteins, which were of cyclin E (Herrera et al., 1996). These observations expressed with a puromycin resistance vector (pBPuro). suggest the existence of a positive feedback regulatory Following infection, cells were seeded at serial dilutions loop, in which cyclin E is not only an upstream regulator (e.g. 1/20, 1/200, 1/2000) in high serum, puromycin- of E2F (via the phosphorylation of pRb), but also its selective medium. All non-infected cells had detached downstream effector. If the latter were true, deregulated within 36 h of selection. Subconfluent populations of cyclin E expression might be expected to bypass growth infected cells were harvested at 48 h after a 30 min pulse arrest by p16 or pRb, exactly as E2F. with bromodeoxyuridine (BrdU) for analysis of cell cycle Myc, the transcription factor encoded by the c-myc distribution. Biochemical studies were performed at the proto-oncogene, was also linked to the activation of same time. Dilute dishes were incubated further to assess cyclin–CDK complexes in G . Myc expression is strictly long-term proliferation and colony outgrowth. controlled by mitogens, and is required for cell cycle entry Expression of p16 induced accumulation of Rat1 cells and continued proliferation. Constitutive Myc expression in G and suppressed colony outgrowth (Figure 1A and prevents growth arrest by a variety of growth-inhibitory B). Most cells in these p16-arrested populations remained signals, and activation of a conditional Myc–estrogen isolated or formed sparse colonies, with no signs of receptor chimera (MycER) induces entry into the cell apoptosis (data not shown). Retroviruses expressing cycle (but also apoptosis) in the absence of mitogens cyclins A, D1, D2, D3 or Cdc25A did not prevent p16- (Eilers et al., 1991; Alexandrow et al., 1995; Hermeking induced cell cycle arrest, although all exogenous proteins et al., 1995; reviewed by Marcu et al., 1992; Henriksson were detectable by immunoblotting (data not shown and and Lu¨scher, 1996). Since antimitogenic stimuli are medi- Table I). In contrast, cells expressing Myc or cyclin E ated by CKIs, we investigated whether Myc could over- entered S-phase and formed colonies efficiently in the come the growth-inhibitory function of these proteins. We presence or absence of p16 (Figure 1A and B, and Table previously reported that constitutive expression of Myc I). Similar results were obtained with NIH-3T3 cells (data abrogates growth arrest by retrovirally expressed p27, by not shown). As for p16, Myc and cyclin E bypassed cell preventing its association with cyclin E–CDK2 (Vlach cycle arrest by p15 and p18 in Rat1 cells (Figure 1C). et al., 1996). Myc also induces transcription of the CDK- Immunoblot analysis revealed that Myc and cyclin E did activating phosphatase Cdc25A (Galaktionov et al., 1996), not suppress expression of p16, p15 or p18 (Figure 2A), which appears not to be involved in rescue of p27-induced and had no significant effect on the expression of the arrest (Vlach et al., 1996). Moreover, MycER activation cellular p16 targets CDK4, CDK6 and D-type cyclins, in quiescent fibroblasts rapidly activates cyclin E–CDK2 apart for a slight increase in cyclin D2 levels in Myc cells 5323 K.Alevizopoulos et al. Table I. Summary of proliferation data with doubly infected Rat1 cells INF.2 p16 p27 INF.1 vector Myc cyclin A cyclins D1–3 cyclin E Cdc25A E2F-1–3  – E2F-4, 5 DP-1 E2F-4  DP-1  n.d. Data from this work are compiled together with data from Vlach et al. (1996), for a direct comparison between p16 and p27.  and – indicate proliferating and arrested cells, respectively, according to BrdU incorporation and colony outgrowth assays. All cells proliferated after the first infection (INF. 1) as efficiently as control, vector- infected cells. n.d., not determined. hypophosphorylated (Figure 3A and B, lane 2). p107, in particular, was present as a doublet of both hyper- and hypophosphorylated forms in p27-arrested cells (Vlach et al., 1996 and Figure 3A), whereas INK4-arrested cells contained only the lower band. Strikingly, Myc and cyclin E did not reverse the effect of p16 on pRb family proteins, which were also mostly hypophosphorylated in proliferating Mycp16 or cyclin Ep16 cells (Figure 3B, lanes 4 and 6). This observation was reproduced for pRb and p107 in NIH-3T3 cells (Figure 3C) in which p16 induced no changes in p130 mobility (data not shown). Thus, pocket proteins were hypophosphorylated in grow- ing Mycp16 or cyclin Ep16 cells, unlike in control Fig. 1. Myc and cyclin E overcome growth arrest by INK4 family proteins. (A) p16 induces accumulation of control cells in G , but does growing cells. This was not merely a transient phenome- not modify the cell cycle distribution of cells expressing Myc or non. Indeed, when passaged and cultured for longer cyclin E. Serial infections were performed as described in the text. periods of time, Rat1 cells co-expressing Myc or cyclin Cells were infected first (INF.1) with retroviruses expressing Myc or E with p16 proliferated with elevated p16 levels and cyclin E or with the empty control vector (pBHygro), followed by a maintained hypophosphorylated pRb, p107 and p130 pBPuro retrovirus with or without p16, as indicated (INF.2). Cells were harvested after 48 h of puromycin selection, and cell cycle (Figure 4). At all time points, the percentage of S-phase distribution was analysed by flow cytometry. The percentage of cells cells was equivalent to those of control growing cells in each phase of the cell cycle is indicated for all cell populations. (data not shown). Altogether, these observations suggested (B) p16 prevents colony outgrowth in control, but not in Myc- or that Myc and cyclin E allowed cell proliferation by cyclin E-expressing cells. Infections were performed as above. bypassing pRb activation, and that p16 remained functional Colonies were fixed and stained on dishes after 6 days in puromycin- selective medium. (C) Percentage of S-phase cells in populations in inhibiting its targets, CDK4 and CDK6. Indeed, p16– infected with retroviruses expressing Myc or cyclin E (INF.1), and CDK6 interactions were not altered by Myc or cyclin E, p15, p16 or p18 (INF.2). S-phase cells were visualized by as detected by CDK6 immunoprecipitation followed by immunocytochemical detection of incorporated BrdU. immunoblotting of p16 (Figure 2C). Retroviral expression of CDK4 (wild-type or a kinase-defective mutant) or (Figure 2B). In conclusion, Myc and cyclin E overcome CDK6 also prevented p16-induced arrest (data not shown), growth arrest by INK4 proteins. supporting the notion that these kinases were the p16 targets in cells. However, we were unable to demonstrate Myc and cyclin E do not prevent p16-induced directly that p16 suppressed the kinase activities associated dephosphorylation of pocket proteins with CDK4, CDK6 or D-type cyclins. In summary, Myc To investigate the effects of Myc or cyclin E on INK4 and cyclin E promote cell growth in the presence of p16 function, we examined the relative phosphorylation state and hypophosphorylated pocket proteins. of pRb, p107 and p130 by immunoblot analysis. In INK4- arrested Rat1 cells, pocket proteins were found exclusively Retrovirally expressed pRb cooperates with p16 in their hypophosphorylated, faster migrating form (Figure and is bypassed by Myc or cyclin E 3A), consistent with the induction of G arrest. Although If Myc and cyclin E bypass pRb activation by p16, they pRb levels were decreased consistently in p16-arrested might rescue growth in the presence of elevated pRb levels. cells in comparison with controls, the remaining pRb was To address this question, we constructed retroviruses 5324 Cyclin E and Myc bypass the p16/pRb pathway Fig. 3. Immunoblot analysis of expression and relative phosphorylation states of pRb family proteins. Rat1 cells were infected with (A) viruses expressing p15, p16, p18 or p27 alone, as indicated, or (B) with the indicated combination of viruses. (C) The same experiment as in (B) was performed with NIH-3T3 cells. The slower migrating, hyperphosphorylated forms of pRb, p107 or p130 and their faster migrating, hypophosphorylated forms are indicated on the right (hyper, hypo) in (A). Fig. 2. Expression of exogenous INK4 proteins, of cellular cyclins and CDKs, and association of exogenous p16 with cellular CDK6. (A) Immunoblot analysis of retrovirally expressed INK4 proteins in doubly infected cells, as indicated (INF.1 and INF.2). The asterisk beside the p18 panel represents a non-specific band cross-reacting with the p18 antibody. (B) Immunoblot analysis of cellular proteins (indicated on the left of each panel) in doubly infected Rat1 cells expressing Myc, cyclin E and/or p16, as indicated at the top. The asterisks below lanes 5 and 6 of the cyclin E panel indicate that only the endogenous, rodent cyclin E is visualized in this blot. (C) CDK6 was immunoprecipitated from the same cells as above, and the co-precipitated p16 protein revealed by immunoblotting. expressing full-length, hemagglutinin (HA)-tagged murine pRb (HApRb) or a derivative mutated in several potential CDK phosphorylation sites (HAΔp34) (Hamel et al., Fig. 4. p16 levels are maintained, and pRb, p107 and p130 remain hypophosphorylated in growing cells expressing exogenous Myc or 1992). Although HApRb and HAΔp34 accumulated in cyclin E together with p16. Cells infected with the indicated cells at higher levels than endogenous pRb (Figure 5A), combinations of viruses were allowed to grow for the indicated they were insufficient to affect cell cycle progression or number of passages (Pass.) and days. All cells were proliferating proliferation on their own (Figure 5B and data not shown). continuously throughout the experiment, and were subconfluent at each harvesting time point. Immunoblotting of p16 (lower panels) shows However, HApRb slightly, albeit reproducibly, enhanced that its expression is maintained. Pocket proteins (indicated to the left) the efficiency of p16-induced arrest, as shown by the are found in their hypophosphorylated forms in all cells expressing lower percentage of S-phase cells in HApRbp16 cells p16, but not in control cells. The faster migrating pRb bands seen in compared with p16-expressing cells (Figure 5C). Further- the presence of p16 represent further dephosphorylated forms, as more, HApRbp16 cells showed a marked increase in previously seen (Chen et al., 1989). 5325 K.Alevizopoulos et al. of pRb, p107 and p130. At this stage, however, inactivation of pocket proteins by a specific cyclin E-mediated phosphorylation event could not be ruled out. To address the functional status of pRb, we sought to analyse its interaction with E2F-1, -2 or -3. Because these proteins were undetectable in Rat1 cells, we co-expressed Myc and cyclin E with an E2F variant (HA-E2F1eco) which interacts with pRb but does not bind DNA (Johnson et al., 1993). In spite of a slight residual activity of E2F1eco (see Figure 9C), growth of E2F1eco cells was suppressed by p16 and restored by Myc or cyclin E (data not shown). Following superinfection with the p16 retrovirus, the interactions between HA-E2F1eco and pRb were monitored by immunoprecipitation with anti-E2F-1 anti- bodies, followed by pRb immunoblotting (Figure 6A, panels ii). Total and immunoprecipitated HA-E2F1eco levels were constant throughout the experiment (panels iii and iv), and no pRb was detectable in anti-E2F-1 immunoprecipitates from cells lacking HA-E2F1eco (panels ii, lanes 0). The results showed that p16 increased the association of pRb with HA-E2F1eco in control cells (Figure 6A, lanes 1 and 2), as well as in cells expressing Myc (lanes 3 and 4) or cyclin E (lanes 5 and 6), even though total pRb levels were decreased by p16 in all cells (Figure 6A, panels i; Figure 3B). The same experiment was repeated in cells expressing exogenous HApRb together with HA-E2F1eco, and similar controls were provided (Figure 6B). The result was identical, showing that the association of pRb (endogenous plus exogenous) with HA-E2F1eco was induced by p16 in all cells (panels ii). Taken together, these experiments demonstrate that Myc and cyclin E do not interfere with p16-induced activation of pRb. Similar experiments were performed to assess the func- Fig. 5. Retrovirally expressed pRb enhances p16-induced arrest, but does not prevent rescue by Myc or cyclin E. (A) Immunoblot analysis tional status of p130. As HA-tagged E2F-4 is not functional of retrovirally expressed HApRb of the HAΔp34 mutant with anti-pRb in the p16 rescue assay (see below, Figure 9C), we co- and anti-HA antibodies, as indicated. Hyper, hypo: forms of pRb as expressed it together with Myc or cyclin E and monitored defined in Figure 3A. (B) Percentage of cells in S-phase 48 h its interaction with endogenous p130. As above, we post-infection with the same retroviruses as in (A). (C) Percentage of S-phase cells following serial infections with the indicated viruses controlled the amounts of total p130 (Figure 7, panel i) (INF.1, INF.2, INF.3). The data presented here represent the average of and of total and immunoprecipitated E2F-4 (panels iii and three independent experiments. (D) Immunoblot analysis of HApRb in iv). No p130 was detectable in HA immunoprecipitates the same cells. in the absence of HA-E2F-4 (lane 0). The data (panel ii) showed that p16 induced binding of p130 to E2F-4 in the flattened, senescent-like phenotype, reminiscent of that control, Myc and cyclin E cells, analogous to the results seen in pRb-arrested SAOS-2 cells (Templeton et al., with pRb. The same experiment did not yield a clear 1991; Hinds et al., 1992) (data not shown). Co-expression answer for the interaction between p107 and HA-E2F-4. of Myc or cyclin E with HApRb prior to p16 fully This interaction was detectable in growing Rat1 cells, but prevented cell cycle arrest (Figure 5C) and the appeareance was not increased further upon expression of p16 (data of the ‘flat’ phenotype (data not shown). Although the not shown). differentially phosphorylated forms of HApRb were diffi- To analyse E2F–pocket protein complexes further and cult to resolve on immunoblots, a significant fraction of to assess the amount of ‘free’ E2F activity, we performed the protein was found in its faster-migrating form in all gel retardation assays with retrovirally infected cells. As p16-expressing cells (Figure 5D). These data confirmed detection of E2F–DNA complexes was very difficult in two important points: first, that pRb levels are limiting Rat1 cells, we used NIH-3T3 cells, which showed similar for p16-induced arrest; second, that Myc and cyclin responses to p16 (Figure 3C and data not shown). Three E rescue cell growth without preventing p16-induced DNA-binding complexes specific for E2F sites were dephosphorylation of either endogenous or exogenous identified with oligonucleotide competition studies (com- pRb. plexes A, B and C; Figure 8). Complex A corresponds to ‘free’ E2F–DP dimers, as treatment of the extracts with Pocket proteins are functional in growing deoxycholate (DOC) resulted in loss of B and C and a Mycp16 or cyclin Ep16 cells simultaneous increase in the amount of complex A. The The data presented so far suggested that Myc and cyclin E2F species in complex A was most likely E2F-4, since E promote cell growth in the face of p16-induced activation it co-migrated with the major E2F–DNA complex seen in 5326 Cyclin E and Myc bypass the p16/pRb pathway Fig. 6. Interaction of pRb with the E2F-1eco protein in the absence (A) or presence (B) of exogenous pRb. Cells were serially infected with the indicated combinations of viruses. The E2F-1 mutant HA-E2F1eco (see text) was immunoprecipitated with anti-E2F-1 antibodies. Immunoblotting was performed to visualize the immunoprecipitated HA-E2Feco with anti-HA antibodies (panels iii) and the associated pRb (endogenous and exogenous) with anti-pRb antibodies (panels ii). Total levels of pRb (panels i) and HA-E2F1eco (panels iv) were also measured. nd: not determined. NIH-3T3 cells expressing recombinant E2F-4, but not E2F-1, -2, -3 or -5 (data not shown). The composition of complex B, which was not supershifted by any tested antibody, is unknown. Complex C could be supershifted by p107 or CDK2 antibodies (data not shown) and presumably also contained cyclins E or A (Lees et al., 1992; Schwarz et al., 1993; Cobrinik, 1996). No other pocket protein–E2F complexes were detected in growing NIH-3T3 cells, as previously described (Lam and Watson, 1993; Schwarz et al., 1993). Two major points can be made concerning the effects of Myc, cyclin E and/or p16 on E2F DNA-binding complexes (Figure 8). First, p16 alone reduced the amount of ‘free’ E2F-4–DP (complex A) present in control cells. This decrease was minimized in the presence of Myc or cyclin E. Second, complex C was increased in response to p16 in Myc or cyclin E cells. The lack of complex C Fig. 7. Interaction of p130 with E2F-4. Cells were serially infected induction in control cells may be due to the formation of with the indicated combinations of viruses. HA-E2F-4 was alternative complexes not resolved here and, in part, to immunoprecipitated with anti-HA antibodies. Immunoblotting was slightly lower p107 levels in p16-arrested cells (Figure performed to visualize the immunoprecipitated HA-E2F-4 with 3C). Regardless of this, our data show that p16 activates anti-HA antibodies (panel iii), or the associated p130 (panel ii). the E2F-binding activity of p107 in Myc and cyclin E Total p130 (panel i) and HA-E2F-4 (panel iv) were also visualized. nd: not determined. cells. In summary, the results from co-immunoprecipitation and gel retardation studies show that Myc and cyclin E prevent growth arrest by p16 without interfering with the results is that cyclin E and Myc restore activity of some ability of pRb, p107 and p130 to associate with their E2F complexes (see Discussion), although alternative respective E2F targets. routes to restore expression of E2F target genes cannot be ruled out. E2F proteins bypass p16-, but not p27-induced If cyclin E acted solely through E2F, E2F proteins arrest: implications for an E2F-independent would be expected to bypass G arrest not only by p16, function of cyclin E as previously reported (DeGregori et al., 1995b; Lukas The data presented so far may be interpreted in two et al., 1996; Mann and Jones, 1996), but also by p27, alternative ways. First, cyclin E or Myc might allow cell which inhibits cyclin E–CDK2 function. To address this growth in the absence of E2F function. Second, cyclin E question, we infected Rat1 cells with retroviruses or Myc might rescue the transcriptional activity of some expressing HA-tagged versions of E2F-1 to -5, as well as E2F complexes in the presence of active pRb, p107 E2F-1Δ5 and E2F-4Δ4 (two mutants which are not bound and p130. Unfortunately, we could not detect significant by pocket proteins; Krek et al., 1994; Vairo et al., 1995), regulation of stably or transiently transfected E2F-respons- the DNA-binding mutant E2F-1eco and a T7-tagged form ive reporter genes by retrovirally expressed p16. However, of DP-1. Expression of all these proteins was detected cyclin A and CDC2 protein levels were suppressed by readily by immunoblot (Figure 9A and B and data not p16 in control cells, but not in cells expresing cyclin E shown). E2F-1, -2, -3 and E2F-1Δ5 prevented growth or Myc (Figure 2B). The simplest interpretation of these arrest by p16, whereas E2F-4, E2F-5 and DP-1 were 5327 K.Alevizopoulos et al. Fig. 8. Gel retardation analysis of E2F DNA-binding activities in NIH-3T3 cells. Cells were infected with the indicated viruses and complexes binding to a radiolabeled E2F-binding oligonucleotide were resolved by native gel electrophoresis. Three specific complexes were identified (A, B and C). Where known, the proteins comprising these complexes are indicated on the right (see text). The asterisks indicate non-sequence-specific complexes, as determined by competition studies. inactive (Figure 9C and Table I), consistent with previous reports (Lukas et al., 1996; Mann and Jones, 1996). The lack of activity of E2F-4 and E2F-5 may be due to the combination of several parameters, including interaction with pocket proteins and/or inefficient dimerization. Indeed, E2F-4Δ4 had a weak rescuing activity, and co- expression with DP-1 allowed E2F-4 and E2F-4Δ4to fully rescue p16-induced arrest (Figure 9; Lukas et al., 1996). In contrast to the p16 results, none of the E2F proteins prevented growth arrest by p27 (Figure 9D and Fig. 9. Retrovirally expressed E2F-1 to -3 overcome p16-, but not data not shown, and Table I). Association of E2F-1 with p27-induced cell cycle arrest. Cells were infected with viruses expressing the indicated tagged E2F and DP proteins and mutant pocket proteins did not account for its inability to bypass derivatives (see text). (A and B) Immunoblot analysis of retrovirally p27 arrest, since cells expressing E2F-1Δ5 were also expressed E2F proteins (indicated at the top), with anti-E2F-1 or arrested by p27. Thus, E2F-1 to -3 in our system behaved anti-HA in (A), and anti-HA in (B). Given that the HA-tag is like cyclin E, which rescued growth arrest by p16 (this N-terminal, the faster migrating band in the E2F-4 and E2F-4Δ4 lanes work), but not by p27 (Vlach et al., 1996) (Table I). In is probably a degradation product. (C) Percentage of cells in S-phase in pools infected with viruses expressing HA-tagged E2F proteins and summary, although E2F-1, -2 and -3 were expressed at p16, as indicated (INF.1, INF.2). The values for p16 cells (shaded levels sufficient to bypass pRb activation in response to bars) are superimposed on those for control cells (white bars). p16, they could not bypass repression of cyclin E–CDK2 (D) Same as above with p27 instead of p16. Myc was used as a by p27. This implies that cyclin E has a distinct function, positive control to overcome p27-induced arrest. for which its targets lie outside of the pRb/E2F pathway. CDK6 and induced activation of pocket proteins and their Discussion association with E2Fs. This effect was not transient since Cyclin E bypasses activation of the p16/pRb cells expressing cyclin E and p16 sustained long-term growth-inhibitory pathway proliferation with hypophosphorylated, active pocket pro- Previous observations suggested that cyclin E has a teins. Cyclins A, D1, D2 and D3 were inactive in bypassing function distinct from pRb phosphorylation required for p16-induced arrest in this system (Table I). G –S progression (see Introduction; Ohtsubo and Roberts, In normal cells, cyclin E expression is under the control 1993; Resnitzky and Reed, 1995; Hofmann and Livingston, of E2F and pRb. The promoter of the cyclin E gene 1996). We demonstrate here that cyclin E is sufficient to contains E2F-responsive elements and is repressible by bypass growth arrest by p16, and allows cell proliferation pRb (Ohtani et al., 1995; Botz et al., 1996; Geng et al., in the presence of active pRb family proteins. Infection 1996). Consistent with this, pRb-deficient cells showed with a p16-expressing retrovirus arrested Rat1 cells in G , enhanced expression of cyclin E (Herrera et al., 1996), but had no effect on cells expressing exogenous cyclin E. and overexpression of E2F-1 induced cyclin E (DeGregori p16 remained active in these cells since it bound its target et al., 1995a; Soucek et al., 1997). Since cyclin E–CDK2 5328 Cyclin E and Myc bypass the p16/pRb pathway complexes also contribute to pRb phosphorylation and activating complexes. Compatible with, but not proving, inactivation, these observations led to the proposal that a this hypothesis, our gel retardation assays showed that positive feedback loop may operate in late G (see p16 in cyclin E-expressing cells induced formation of Introduction). According to this model, the increase in E2F-4/DP/p107/cyclin–CDK2 DNA-binding complexes. cyclin E–CDK2 activity would contribute to further activ- On the other hand, cyclin E also favored the maintenance ation of E2F, and thus to enhanced transcription of the of ‘free’ E2F-4–DP complexes. However, whether and cyclin E gene. The outcome of such a loop would be to how E2F contributes to cyclin E function in overcoming boost activation of both E2F and cyclin E. However, such the p16 block remains to be investigated. a model does not predict which of E2F or cyclin E, or The functional relationship of cyclin E and E2F is both, performs the rate-limiting downstream function in highlighted further by studies in Drosophila embryos. S-phase entry. E2F- or cyclin E-deficient embryos fail to enter S-phase Ectopic activation of an essential downstream effector of embryonic cycle 17 in the endocycling cell compartment may bypass the presence of active pRb. This is the case (Knoblich et al., 1994; Duronio et al., 1995). In this for E2F-1, which suppresses both pRb- (Zhu et al., 1993; tissue, the limiting downstream target of E2F for S-phase Qin et al., 1995) and p16-induced G arrest, in conditions entry is cyclin E. Ectopic E2F–DP expression requires in which cellular pRb remains hypophosphorylated (Lukas cyclin E function to trigger S-phase entry, although the et al., 1996; Mann and Jones, 1996; this work and transcriptional function of E2F itself does not require unpublished data). Moreover, adenovirus-mediated gene cyclin E (Duronio et al., 1995, 1996). Conversely, ectopic transfer of E2F-1 suppressed G arrest and allowed comple- expression of cyclin E can trigger S-phase entry, but tion of a mitotic cycle in the presence of either p16 or requires expression of additional E2F target genes. Cyclin p27, leading to the suggestion that E2F-1 allows S-phase E expression induces E2F-dependent transcription entry in the absence of cyclin E function (DeGregori (Duronio and O’Farrell, 1995), most likely through et al., 1995b). In those studies, however, E2F-1 was phosphorylation of the pRb homolog RBF (Du et al., strongly overexpressed, which may have led to non- 1996). In the central nervous system, cyclin E is expressed physiological effects. Indeed, we and others (Mann and by an E2F-independent mechanism, and is required Jones, 1996) found that p27 is dominant over E2F-1. In upstream of E2F function (Duronio et al., 1995). Thus, fact, E2F proteins behaved similarly to cyclin E, which the dependence of one gene upon the other can be reversed rescued p16-, but not p27-induced arrest (Table I), owing in different tissues or stages, depending on which gene to the inhibition of cyclin E by p27 (Vlach et al., 1996). product becomes rate-limiting. In summary, in Drosophila These observations imply that E2F activity cannot bypass as in mammalian cells, cyclin E is both a downstream inhibition of cyclin E–CDK2 by p27, even though it can effector and an activator of E2F. In addition to this cross- bypass activation of pRb by p16. Two conclusions can be regulation, both E2F and cyclin E have at least one drawn from these studies: first, that E2F is unable to independent downstream function required for S-phase promote growth in the absence of cyclin E activity; second, entry. that cyclin E has at least one E2F-independent function What might be the E2F-independent function of cyclin required for S-phase entry. Such a function of cyclin E? In vitro experiments suggest that it might be linked to E–CDK2 was also implied by the observation that the the initiation of DNA replication. In Xenopus extracts, SV40 LT restored E2F activity, but not cell cycle progres- which replicate in the absence of transcription (and thus sion, in the presence of transiently expressed dominant- of E2F activity), addition of p21 blocked DNA replication, negative CDK2 (Hofmann and Livingston, 1996). and cyclins A or E could overcome this block (Strausfeld If cyclin E has a function downstream of E2F, could it et al., 1994). In addition, G -phase HeLa cell nuclei be sufficient for S-phase entry in the absence of E2F initiated DNA replication when co-incubated with S-phase activity? A direct test of this question would be to block nuclear and cytoplasmic extracts. The nuclear extract E2F activity in cells expressing cyclin E. We attempted could be replaced by cyclin E–CDK2 or cyclin A–CDK2 this with a dominant-negative mutant of DP-1 (DN-DP-1) (Krude et al., 1997). A direct involvement of cyclin–CDK which blocks E2F function and arrests cells in G in activity in triggering the onset of DNA replication has also transient transfections (Wu et al., 1996). However, retro- been demonstrated in yeast (reviewed by Stillman, 1996). viral expression of DN-DP-1 (or a variant tagged with a nuclear localization sequence) had no effect on cell cycle Rescue of p16-induced arrest by Myc is most likely progression. Thus DN-DP-1 may effectively block E2F a cyclin E-mediated activity activity in vivo only if overexpressed, which may have We show here that Myc bypasses p16/pRb-induced growth non-specific side effects. A direct test of the requirement arrest. Like other biological activities of Myc (Henriksson for E2F in the presence of cyclin E awaits alternative and Lu¨scher, 1996; Vlach et al., 1996), this effect depends ways to ablate E2F activity. In spite of this, several upon the formation of transcriptionally active Myc–Max observations suggest that cyclin E does not bypass, but dimers (unpublished data). As for cyclin E, cells expressing rather rescues E2F activity by an alternative route in the Myc and p16 proliferate with hypophosphorylated, active presence of p16 and active pRb. In transient transfections, pocket proteins. Thus, Myc and cyclin E rescue p16- cyclin E relieved p16 (but not p27)-mediated repression induced arrest by indistinguishable mechanisms. The main of the cyclin A promoter (Zerfass-Thome et al., 1997). In rationale for these findings is provided by the notion that our experiments, cyclin E rescued the expression of Myc can act as an upstream activator of cyclin E–CDK2. cellular cyclin A and CDC2 in the presence of p16. Constitutive expression of Myc prevented inactivation of Zerfass-Thome et al. (1997) proposed that cyclin E–CDK2 cyclin E–CDK2 and cell cycle arrest by retrovirally binding to E2F-4/DP/p107 results in the formation of expressed p27, through the sequestration of p27 into heat- 5329 K.Alevizopoulos et al. labile complexes (Vlach et al., 1996) (Table I). The same Hall and Peters, 1996). These observations strongly sug- mechanism was involved in activation of cyclin E–CDK2 gest that other genetic lesions resulting in the functional by MycER in quiescent fibrobasts (Steiner et al., 1995; suppression of the p16/pRb pathway should also be Perez-Roger et al., 1997; J.Vlach and B.Amati, unpub- tumorigenic. lished data). In some cell lines, p27 has a role in limiting Recently, aberrant expression of cyclin E has been the rate of G –S progression during proliferation (Coats linked to tumorigenesis (see, for example, Keyomarsi et al., 1996), whereas constitutive expression of cyclin E et al., 1994; Porter et al., 1997). It was suggested that (Ohtsubo and Roberts, 1993) or Myc (Karn et al., 1989) cyclin E–CDK2 replaces CDK4 function in tumor cells can accelerate it. These observations suggest that Myc, that express high p16 levels by phosphorylating pRb by suppressing p27 function, can accelerate the activation (Gray-Bablin et al., 1996). However, our data show that of cyclin E during G . This, through the positive cyclin cyclin E bypasses p16/pRb function through a different E–E2F feedback loop discussed above, should further pathway. Although cyclin E was also shown to have a enhance cyclin E expression. Consistent with this view, mild oncogenic potential in transgenic mice (Bortner and MycER activation induced expression of E2F target genes Rosenberg, 1997), its causal involvement in human tumors (Jansen-Du¨rr et al., 1993; Pusch et al., 1997b), and in remains to be investigated. particular of cyclin E (Perez-Roger et al., 1997; Pusch The widespread oncogenic potential of myc family et al., 1997a), although there is no evidence that any of genes and their involvement in numerous malignancies in these genes is a direct transcriptional target of Myc. both humans and animals are well established (reviewed The cdc25A gene, on the other hand, is a Myc target by Marcu et al., 1992; Garte, 1993). As discussed above, (Galaktionov et al., 1996). However, overexpression of Myc antagonizes p27 function, but loss of p27 is not per Cdc25A rescued neither p16- nor p27-induced arrest se a tumorigenic event (Fero et al., 1996; Kiyokawa et al., (Vlach et al., 1996; Table I), suggesting that additional 1996; Nakayama et al., 1996). This suggests that Myc Myc/Max target genes are involved. may have other direct targets, which may include the p27- In addition to its upstream action, Myc may also be a related CKIs p21 and p57, as well as p53 (Hermeking downstream target of pocket proteins. First, the c-myc et al., 1995). Independently of this aspect, bypassing the gene was proposed to be an E2F target; however, unlike tumour suppressor function of the p16/pRb pathway may other E2F targets, it was not induced upon adenovirus be an important, if not the essential, aspect of Myc’s expression of E2F-1 (DeGregori et al., 1995a). Second, oncogenic activity. In support of this hypothesis, fibroblasts p107 can associate with the Myc protein and suppress its isolated from p16-null mouse embryos are transformed transactivation potential in transient transfections by ras alone, whereas wild-type cells undergo premature (Beijersbergen et al., 1994; Gu et al., 1994; Hoang et al., senescence (Serrano et al., 1996, 1997). On the other hand, 1995). In these assays, Myc bypassed cell cycle arrest by transformation of wild-type cells requires cooperation of p107, but not by pRb (Beijersbergen et al., 1994). In ras with immortalizing oncogenes, such as c-myc (Land contrast, microinjected GST–Myc proteins prevented cell et al., 1983). Thus, loss of p16 mimics myc activation, cycle arrest by pRb (Goodrich and Lee, 1992). Since the and vice versa. Based on this, we predict that myc mechanisms underlying these effects were not studied, activation may partly release the pressure to mutate these discrepancies remain unexplained and illustrate the components of the p16/pRb pathway during tumorigenesis. potential dangers of overexpression. Although we cannot rule out that the Myc–p107 interaction plays a role in Materials and methods overcoming p16 (or p27)-induced arrest by Myc, it is unlikely to account for our findings. First, in our system, Retroviral expression vectors we did not detect the Myc–p107 interaction. Second, cells The retroviral vectors pBabe-Puro (pBP), -Bleo (pBB), - Neo2 (pBN2) expressing Myc and p16 proliferate in the presence of not and- Hygro2 (pBH2) were described previously (Morgenstern and Land, only active p107, but also of pRb and p130. Third, Myc 1990; Vlach et al., 1996). cDNAs encoding human cyclins A, D1–3 and E, as well as Myc, HApRb, HAΔp34, HA-E2F-1 to -3 and HA-E2F1eco was expressed at low levels in our studies, making titration proteins were subcloned in pBN2, pBH2 and/or pBB, as required. HA- of p107 by Myc a very unlikely mechanism. E2F-4 and HA-E2F-4Δ4 were expressed from the pRcCMVneo retroviral In summary, we propose that the overcoming of p16- vector (a kind gift from W.Krek). cDNAs encoding the human CKIs induced growth arrest by Myc is mainly the consequence p15, p16, p18 and p27 were subcloned in pBP. of its upstream action on cyclin E–CDK2: Myc suppresses Retroviral infections, cell cycle analysis and biochemical p27 function, allowing an elevation of cyclin E activity analysis above the threshold required to bypass the p16/pRb block. High-titer retroviral supernatants (510 /ml) were generated as In situations in which p16 is not overexpressed, the same described (Vlach et al., 1996). Infected Rat1 or NIH-3T3 cells were activity of Myc accelerates hyperphosphorylation of pRb selected with the appropriate drug: G418 (Calbiochem, 1000 μg/ml), by cyclin E–CDK2 in concert with cyclin D–CDK4. hygromycin (Calbiochem, 150 μg/ml), phleomycin (Cayla, 50 μg/ml) or puromycin (Sigma, 2.5 μg/ml). Serial infections of cell pools, preparation of cell lysates and biochemical analysis (immunoblots, immunoprecipit- Relevance of bypassing p16/pRb function for ations) were as previously described (Vlach et al., 1996). Gel retardation tumorigenesis assays were performed from infected cells as previously described The p16/pRb pathway has a fundamental role in sup- (Beijersbergen et al., 1995). For cell cycle analysis, cells were labeled with 33 μM BrdU for 30 min, and analyzed by either two-dimentional pressing tumorigenesis. Both p16 and pRb are encoded flow cytometry (Vlach et al., 1996), or by immunocytochemical detection by tumor suppressor genes, as indicated by loss of function of BrdU. For the latter, cells were fixed in 3.7% formaldehyde in in different tumor types. On the other hand, oncogenic phosphate-buffered saline (PBS) (5 min), permeabilized with acetone activation in human tumors has been reported for the at –20°C (30 s), treated with 1.5 M HCl (10 min), blocked in PBS with genes encoding cyclins D1, D2 and CDK4 (reviewed by 1% bovine serum albumin (BSA) (1–2 min), incubated for 1 h with the 5330 Cyclin E and Myc bypass the p16/pRb pathway primary antibody (undiluted mouse anti-BrdUNuclease, Amersham gene that regulates E2F activity and interacts with cyclin E in No. RPN202), and for 1 h with the secondary antibody (goat anti-mouse Drosophila. Genes Dev., 10, 1206–1218. Fab–FITC, Sigma, dil. 1/200 in PBS  1% BSA) with Hoechst Duronio,R.J. and O’Farrell,P.H. (1995) Developmental control of the G1 (0.5 μg/ml, Calbiochem), and mounted with FluorSave™ Reagent to S transition in Drosophila: cyclin E is a limiting downstream target (Calbiochem). All steps were performed at room temperature and of E2F. Genes Dev., 9, 1456–1468. separated by several washes in PBS (with 1% BSA after the blocking Duronio,R.J., O’Farrell,P.H., Xie,J.E., Brook,A. and Dyson,N. (1995) step). The transcription factor E2F is required for S phase during Drosophila embryogenesis. Genes Dev., 9, 1445–1455. Duronio,R.J., Brook,A., Dyson,N. and O’Farrel,P.H. (1996) E2F-induced Antibodies S-phase entry requires cyclin E. Genes Dev., 10, 2505–2513. The following antibodies were used: against CDK2, M2 (Santa Cruz Eilers,M., Schirm,S. and Bishop,J.M. (1991) The MYC protein activates sc-163); CDK4, C22 (sc-260) or H303 (sc-749); CDK6, C21 (sc-177); transcription of the alpha-prothymosin gene. EMBO J., 10, 133–141. CDC2(17), (sc-54); cyclin A, H-432 (sc-751); rodent cyclin E, M20 Fero,M.L. et al. (1996) A syndrome of multiorgan hyperplasia with (sc-481); human cyclin E, C19 (sc198); cyclin D1, 72-13G (sc-450); features of gigantism, tumorigenesis, and female sterility in p27(kip1)- cyclin D2, DCS-5 (from J.Bartek); cyclin D3, C16 (sc-182); cyclin B1, deficient mice. Cell, 85, 733–744. GNS1 (sc-245); pRb, 14001A (Pharmingen); p107, C18 (sc-318); p130, Field,S.J., Tsai,F.Y., Kuo,F., Zubiaga,A.M., Kaelin,W.G., Livingston, C20 (sc-317); E2F-1, C20 (sc-193); p15, C20 (sc-612); HA-tag, D.M., Orkin,S.H. and Greenberg,M.E. (1996) E2F-1 functions in mice HA11(Babco); T7-tag, T7 (Novagen); p16, DCS 50.2 (from G.Peters); to promote apoptosis and suppress proliferation. Cell, 85, 549–561. p18, 11256 (from Y.Xiong). Galaktionov,K., Chen,X. and Beach,D. (1996) Cdc25 cell-cycle phosphatase as a target of c-myc. Nature, 382, 511–517. Garte,S. (1993) The c-myc oncogene in tumor progression. Crit. Rev. Acknowledgements Oncogen., 4, 435–449. Geng,Y., Eaton,E.N., Picon,M., Roberts,J.M., Lundberg,A.S., Gifford,A., We acknowledge the following colleagues for essential reagents, fruitful Sardet,C. and Weinberg,R.A. (1996) Regulation of cyclin E discussions, advice on the manuscript and/or sharing unpublished data: transcription by E2Fs and retinoblastoma protein. Oncogene, 12, Michel Aguet, Nassos Alevizopoulos, Jiri Bartek, Rene´ Bernards, James 1173–1180. DeGregori, Martin Eilers, David Franklin, Susan Gasser, Peter Greasley, Goodrich,D.W. and Lee,W.H. (1992) Abrogation by c-myc of G1 phase Paul Hamel, Markus Hengstschla¨ger, Richard Iggo, Willy Krek, Hartmut arrest induced by RB protein but not by p53. Nature, 360, 177–179. Land, Nathalie Lauper, Trevor Littlewood, Kirsten Mundt, Joseph Nevins, Gray-Bablin,J., Zalvide,J., Fox,M.P., Knickerbocker,C.J., DeCaprio,J.A. Matthias Peter, Gordon Peters, Charles Sherr, Viesturs Simanis and Yue and Keyomarsi,K. (1996) Cyclin E, a redundant cyclin in breast Xiong. Our special gratitude goes to Rene´ Bernards and Ron Kerkhoven cancer. Proc. Natl Acad. Sci. USA, 93, 15215–15220. for their hospitality and precious advice on E2F gel retardation assays. Gu,W., Bhatia,K., Magrath,I., Dang,C. and Dalla-Favera,R. (1994) K.A. and S.H. were supported by the Swiss National Science Foundation, Binding and suppression of the Myc transcriptional activation domain and J.V. by post-doctoral fellowships from the Roche Research by p107. Science, 264, 251–254. Foundation and the Swiss Cancer League. B.A. is a recipient of a START Guan,K.L., Jenkins,C.W., Li,Y., Nichols,M.A., Wu,X., O’Keefe,C.L., fellowship and of a research grant from the Swiss National Science Matera,A.G. and Xiong,Y. (1994) Growth suppression by p18, a Foundation. p16INK4/MTS1- and p14INK4B/MTS2-related CDK6 inhibitor, correlates with wild-type pRb function. Genes Dev., 8, 2939–2952. Hall,M. and Peters,G. (1996) Genetic alterations of cyclins, cyclin- References dependent kinases, and cdk inhibitors in human cancer. Adv. Cancer Adams,P.D. and Kaelin,W.G. (1996) The cellular effects of E2F Res., 68, 67–108. overexpression. Curr. Topics Microbiol. Immunol., 208, 79–93. Hamel,P.A., Gill,R.M., Phillips,R.A. and Gallie,B.L. (1992) Regions Alexandrow,M.G., Kawabata,M., Aakre,M. and Moses,H.L. (1995) controlling hyperphosphorylation and conformation of the Overexpression of the c-Myc oncoprotein blocks the growth-inhibitory retinoblastoma gene product are independent of domains required for response but is required for the mitogenic effects of transforming transcriptional repression. Oncogene, 7, 693–701. growth factor β1. Proc. Natl Acad. Sci. USA, 92, 3239–3243. Henriksson,M. and Lu¨scher,B. (1996) Proteins of the Myc network: Assoian,R.K. (1997) Anchorage-dependent cell cycle progression. J. Cell essential regulators of cell growth and differentiation. Adv. Cancer Biol., 136, 1–4. Res., 68, 109–182. Beijersbergen,R.L., Hijmans,E.M., Zhu,L. and Bernards,R. (1994) Hermeking,H., Funk,J.O., Reichert,M., Ellwart,J.W. and Eick,D. (1995) Interaction of c-Myc with the pRb-related protein p107 results in Abrogation of p53-induced cell cycle arrest by c-Myc: evidence for inhibition of c-Myc-mediated transactivation. EMBO J., 13, 4080– an inhibitor of p21WAF1/CIP1/SDI1. Oncogene, 11, 1409–1415. 4086. Herrera,R.E., Sah,V.P., Williams,B.O., Makela,T.P., Weinberg,R.A. and Beijersbergen,R.L., Carlee,L., Kerkhoven,R.M. and Bernards,R. (1995) Tyler,J. (1996) Altered cell cycle kinetics, gene expression, and G1 Regulation of the retinoblastoma protein-related p107 by G1 cyclin restriction point regulation in Rb-deficient fibroblasts. Mol. Cell. Biol., complexes. Genes Dev., 9, 1340–1353. 16, 2402–2407. Bortner,D.M. and Rosenberg,M.P. (1997) Induction of mammary gland Hinds,P.W., Mittnacht,S., Dulic,V., Arnold,A., Reed,S.I. and Weinberg, hyperplasia and carcinomas in transgenic mice expressing human R.A. (1992) Regulation of retinoblastoma protein functions by ectopic cyclin E. Mol. Cell. Biol., 17, 453–459. expression of human cyclins. Cell, 70, 993–1006. Botz,J., Zerfass-Thome,K., Spitkovsky,D., Delius,H., Vogt,B., Eilers,M., Hirai,H., Roussel,M.F., Kato,J.Y., Ashmun,R.A. and Sherr,C.J. (1995) Hatzigeorgiou,A. and Jansen-Du¨rr,P. (1996) Cell cycle regulation of Novel INK4 proteins, p19 and p18, are specific inhibitors of the the murine cyclin E gene depends on an E2F binding site in the cyclin D-dependent kinases CDK4 and CDK6. Mol. Cell. Biol., 15, promoter. Mol. Cell. Biol., 16, 3401–3409. 2672–2681. Chen,P.-L., Scully,P., Shew,J.-Y., Wang,J.Y.J. and Lee,W.-H. (1989) Hoang,A.T., Lutterbach,B., Lewis,B.C., Yano,T., Chou,T.Y., Barrett,J.F., Phosphorylation of the retinoblastoma gene product is modulated Raffeld,M., Hann,S.R. and Dang,C.V. (1995) A link between increased during the cell cycle and cellular differentiation. Cell, 58, 1193–1198. transforming activity of lymphoma-derived MYC mutant alleles, their Coats,S., Flanagan,W.M., Nourse,J. and Roberts,J.M. (1996) defective regulation by p107, and altered phosphorylation of the Requirement of p27(kip1) for restriction point control of the fibroblast c-Myc transactivation domain. Mol. Cell. Biol., 15, 4031–4042. cell cycle. Science, 272, 877–880. Hofmann,F. and Livingston,D.M. (1996) Differential effects of cdk2 and Cobrinik,D. (1996) Regulatory interactions among E2Fs and cell cycle cdk3 on the control of pRb and E2Ff function during G(1) exit. Genes control proteins. Curr. Topics Microbiol. Immunol., 208, 31–61. Dev., 10, 851–861. DeGregori,J., Kowalik,T. and Nevins,J.R. (1995a) Cellular targets for Jansen-Du¨rr,P., Meichle,A., Steiner,P., Pagano,M., Finke,K., Botz,J., activation by the E2F1 transcription factor include DNA synthesis- Wessbecher,J., Draetta,G. and Eilers,M. (1993) Differential modulation and G1/S-regulatory genes. Mol. Cell. Biol., 15, 4215–4224. of cyclin gene expression by MYC. Proc. Natl Acad. Sci. USA, 90, DeGregori,J., Leone,G., Ohtani,K., Miron,A. and Nevins,J.R. (1995b) 3685–3689. E2F-1 accumulation bypasses a G1 arrest resulting from the inhibition Johnson,D.G., Schwartz,J.K., Cress,W.D. and Nevins,J.R. (1993) of G1 cyclin-dependent kinase activity. Genes Dev., 9, 2873–2887. Expression of transcription factor E2F-1 induces quiescent cells to Du,W., Vidal,M., Xie,J.E. and Dyson,N. (1996) RBF, a novel RB-related enter S phase. Nature, 365, 349–352. 5331 K.Alevizopoulos et al. Karn,J., Watson,J.V., Lowe,A.D., Green,S.M. and Vedeckis,W. (1989) Polyak,K., Lee,M.H., Erdjument,B.H., Koff,A., Roberts,J.M., Tempst,P. Regulation of cell cycle duration by c-myc levels. Oncogene, 4, and Massague,J. (1994) Cloning of p27Kip1, a cyclin-dependent 773–787. kinase inhibitor and a potential mediator of extracellular antimitogenic Keyomarsi,K., O’Leary,N., Molnar,G., Lees,E., Fingert,H.J. and signals. Cell, 78, 59–66. Pardee,A.B. (1994) Cyclin E, a potential prognostic marker for breast Porter,P.L., Malone,K.E., Heagerty,P.J., Alexander,G.M., Gatti,L.A., cancer. Cancer Res., 54, 380–385. Firpo,E.J., Daling,J.R. and Roberts,J.M. (1997) Expression of cell- Kiyokawa,H. et al. (1996) Enhanced growth of mice lacking the cyclin- cycle regulators p27(kip1) and cyclin E, alone and in combination, dependent kinase inhibitor function of p27(kip1). Cell, 85, 721–732. correlate with survival in young breast cancer patients. Nature Med., Knoblich,J.A., Sauer,K., Jones,L., Richardson,H., Saint,R. and Lehner, 3, 222–225. C.F. (1994) Cyclin E controls S phase progression and its down- ¨ Pusch,O., Bernaschek,G., Eilers,M. and Hengstschlager,M. (1997a) regulation during Drosophila embryogenesis is required for the arrest Activation of c-Myc uncouples DNA replication from activation of of cell proliferation. Cell, 77, 107–120. G1-cyclin-dependent kinases. Oncogene, 15, 649–656. Koh,J., Enders,G.H., Dynlacht,B.D. and Harlow,E. (1995) Tumour- Pusch,O., Soucek,T., Hengstschla¨ger-Ottnad,E., Bernaschek,G. and derived p16 alleles encoding proteins defective in cell-cycle inhibition. Hengstschla¨ger,M. (1997b) Cellular targets for activation by c-Myc Nature, 375, 506–510. include the DNA metabolism enzyme thymidine kinase. DNA Cell Krek,W., Ewen,M.E., Shirodkar,S., Arany,Z., Kaelin,W.J. and Biol., 16, 737–747. Livingston,D.M. (1994) Negative regulation of the growth-promoting Qin,X.-Q., Livingston,D.M., Kaelin,W.G. and Adams,P.D. (1994) transcription factor E2F-1 by a stably bound cyclin A-dependent Deregulated transcription factor E2F-1 expression leads to S-phase protein kinase. Cell, 78, 161–172. entry and p53-mediated apoptosis. Proc. Natl Acad. Sci. USA, 91, Krude,T., Jackman,M., Pines,J. and Laskey,R.A. (1997) Cyclin/Cdk- 10918–10922. dependent initiation of DNA replication in a human cell-free system. Qin,X.-Q., Livingston,D.M., Ewen,M., Sellers,W.R., Arany,Z. and Cell, 88, 109–119. Kaelin,W.G. (1995) The transcription factor E2F-1 is a downstream Lam,E.W. and Watson,R.J. (1993) An E2F-binding site mediates cell- target of RB action. Mol. Cell. Biol., 15, 742–755. cycle regulated repression of mouse B-myb transcription. EMBO J., Quelle,D.E. et al. (1995) Cloning and characterization of murine 12, 2705–2713. p16INK4a and p15INK4b genes. Oncogene, 11, 635–645. Land,H., Parada,L.F. and Weinberg,R.A. (1983) Tumorigenic conversion Resnitzky,D. and Reed,S.I. (1995) Different roles for cyclins D1 and E of primary embryo fibroblasts requires at least two cooperating in regulation of the G1-to-S transition. Mol. Cell. Biol., 15, 3463–3469. oncogenes. Nature, 304, 596–602. Schwarz,J.K., Devoto,S.H., Smith,E.J., Chellappan,S.P., Jakoi,L. and Lees,E., Faha,B., Dulic,V., Reed,S.I. and Harlow,E. (1992) Cyclin Nevins,J.R. (1993) Interactions of the p107 and Rb proteins with E2F E/cdk2 and cyclin A/cdk2 kinases associate with p107 and E2F in a during the cell proliferation response. EMBO J., 12, 1013–1020. temporally distinct manner. Genes Dev., 6, 1874–1885. Serrano,M., Lee,H., Chin,L., Cordon,C.C., Beach,D. and DePinho,R.A. Lukas,J., Muller,H., Bartkova,J., Spitkovsky,D., Kjerulff,A.A., Jansen- (1996) Role of the INK4a locus in tumor suppression and cell Durr,P., Strauss,M. and Bartek,J. (1994) DNA tumor virus oncoproteins mortality. Cell, 85, 27–37. and retinoblastoma gene mutations share the ability to relieve the Serrano,M., Lin,A.W., McCurrach,M.E., Beach,D. and Lowe,S.W. (1997) cell’s requirement for cyclin D1 function in G1. J. Cell Biol., 125, Oncogenic ras provokes premature cell senescence associated with 625–638. accumulation of p53 and p16INK4a. Cell, 88, 593–602. Lukas,J., Bartkova,J., Rohde,M., Strauss,M. and Bartek,J. (1995a) Cyclin Shan,B. and Lee,H.-W. (1994) Deregulated expression of E2F-1 induces D1 is dispensable for G1 control in retinoblastoma gene-deficient cells S-phase entry and leads to apoptosis. Mol. Cell. Biol., 14, 8166–8173. independently of cdk4 activity. Mol. Cell. Biol., 15, 2600–2611. Sherr,C.J. (1994) G1 phase progression: cycling on cue. Cell, 79, Lukas,J., Parry,D., Aagaard,L., Mann,D.J., Bartkova,J., Strauss,M., 551–555. Peters,G. and Bartek,J. (1995b) Retinoblastoma-protein-dependent Sherr,C.J. (1995) D-type cyclins. Trends Biochem. Sci., 20, 187–190. cell-cycle inhibition by the tumour suppressor p16. Nature, 375, Sherr,C.J. and Roberts,J.M. (1995) Inhibitors of mammalian G1 cyclin- 503–506. dependent kinases. Genes Dev., 9, 1149–1163. Lukas,J., Otzen Petersen,B., Holm,K., Bartek,J. and Helin,K. (1996) Soucek,T., Pusch,O., Hengstschla¨ger-Ottnad,E., Adams,P.D. and Deregulated expression of E2F family members induces S-phase entry Hengstschla¨ger,M. (1997) Deregulated expression of E2F-1 induces and overcomes p16INK4A-mediated growth suppression. Mol. Cell. cyclin A- and E-associated kinase activities independently from cell Biol., 16, 1047–1057. cycle position. Oncogene, 14, 2251–2257. Mann,D.J. and Jones,N.C. (1996) E2F-1 but not E2F-4 can overcome Steiner,P., Philipp,A., Lukas,J., Godden-Kent,D., Pagano,M., p16-induced G1 cell-cycle arrest. Curr. Biol., 6, 474–483. Mittnacht,S., Bartek,J. and Eilers,M. (1995) Identification of a Myc- Marcu,K.B., Bossone,S.A. and Patel,A.J. (1992) myc function and dependent step during the formation of active G cyclin–cdk regulation. Annu. Rev. Biochem., 61, 809–860. complexes. EMBO J., 14, 4814–4826. Medema,R.H., Herrera,R.E., Lam,F. and Weinberg,R.A. (1995) Growth Stillman,B. (1996) Cell cycle control of DNA replication. Science, 274, suppression by p16ink4 requires functional retinoblastoma protein. 1659–1664. Proc. Natl Acad. Sci. USA, 92, 6289–6293. Strausfeld,U.P., Howell,M., Rempel,R., Maller,J.L., Hunt,T. and Blow,J.J. Morgan,D.O. (1995) Principles of CDK regulation. Nature, 374, 131–134. (1994) CIP1 blocks the initiation of DNA replication in Xenopus Morgenstern,J.P. and Land,H. (1990) Advanced mammalian gene extracts by inhibition of cyclin-dependent kinases. Curr. Biol., 4, transfer: high titre retroviral vectors with multiple drug selection 876–883. markers and a complementary helper-free packaging cell line. Nucleic Templeton,D.J., Park,S.H., Lanier,L. and Weinberg,R.A. (1991) Acids Res., 18, 3587–3596. Nonfunctional mutants of the retinoblastoma protein are characterized Nakayama,K., Ishida,N., Shirane,M., Inomata,A., Inoue,T., Shishido,N., by defects in phosphorylation, viral oncoprotein association, and Hori,I., Loh,D.Y. and Nakayama,K. (1996) Mice lacking p27(kip1) nuclear tethering. Proc. Natl Acad. Sci. USA, 88, 3033–3037. display increased body size, multiple organ hyperplasia, retinal Toyoshima,H. and Hunter,T. (1994) p27, a novel inhibitor of G1 cyclin– dysplasia, and pituitary tumors. Cell, 85, 707–720. Cdk protein kinase activity, is related to p21. Cell, 78, 67–74. Nevins,J.R. (1992) E2F: a link between the Rb tumor suppressor protein Vairo,G., Livingston,D.M. and Ginsberg,D. (1995) Functional interaction and viral oncoproteins. Science, 258, 424–429. between E2F-4 and p130: evidence for distinct mechanisms underlying Ohtani,K., DeGregori,J. and Nevins,J.R. (1995) Regulation of the cyclin growth suppression by different retinoblastoma protein family E gene by transcription factor E2F1. Proc. Natl Acad. Sci. USA, 92, members. Genes Dev., 9, 869–881. 12146–12150. Vlach,J., Hennecke,S., Alevizopoulos,K., Conti,D. and Amati,B. (1996) Ohtsubo,M. and Roberts,J.M. (1993) Cyclin-dependent regulation of G1 Kip1 Growth arrest by the cyclin-dependent kinase inhibitor p27 is in mammalian fibroblasts. Science, 259, 1908–1912. abrogated by c-Myc. EMBO J., 15, 6595–6604. Ohtsubo,M., Theodoras,A.M., Schumacher,J., Roberts,J.M. and Wang,J.Y.J. (1997) Retinoblastoma protein in growth suppression and Pagano,M. (1995) Human cyclin E, a nuclear protein essential for the death protection. Curr. Opin. Genet. Dev., 7, 39–45. G1-to-S phase transition. Mol. Cell. Biol., 15, 2612–2624. Weinberg,R.A. (1995) The retinoblastoma protein and cell cycle control. Perez-Roger,I., Solomon,D.L.C., Sewing,A. and Land,H. (1997) Myc Cell, 81, 323–330. activation of cyclin E/CDK2 kinase involves induction of cyclin E gene transcription and inhibition of p27Kip1 binding to newly formed Weinberg,R.A. (1996) E2F and cell proliferation: a world turned upside complexes. Oncogene, 14, 2373–2381. down. Cell, 85, 457–459. 5332 Cyclin E and Myc bypass the p16/pRb pathway Wu,C.L., Classon,M., Dyson,N. and Harlow,E. (1996) Expression of dominant-negative mutant DP-1 blocks cell cycle progression in G1. Mol. Cell. Biol., 16, 3698–3706. Yamasaki,L., Jacks,T., Bronson,R., Goillot,E., Harlow,E. and Dyson,N.J. (1996) Tumor induction and tissue atrophy in mice lacking E2F-1. Cell, 85, 537–548. Zerfass-Thome,K., Schulze,A., Zwerschke,W., Vogt,B., Helin,K., Bartek,J., Henglein,B. and Jansen-Du¨rr,P. (1997) P27(kip1) blocks cyclin E-dependent transactivation of cyclin a gene expression. Mol. Cell. Biol., 17, 407–415. Zhu,L., van den Heuvel,S., Helin,K., Fattaey,A., Ewen,M., Livingston, D.M., Dyson,N. and Harlow,E. (1993) Inhibition of cell proliferation by p107, a relative of the retinoblastoma protein. Genes Dev., 7, 1111–1125. Zwicker,J. and Mu¨ller,R. (1997) Cell-cycle regulation of gene expression by transcriptional repression. Trends Genet., 13, 3–6. Received on March 10, 1997; revised on May 27, 1997

Journal

The EMBO Journal – Springer Journals

Published: Sep 1, 1997

Keywords: CDK; cyclin; Myc; p16; retinoblastoma

There are no references for this article.