Molecular Cell, Vol. 5, 501–511, March, 2000, Copyright 2000 by Cell Press

Exit from Mitosis in Budding Yeast: Biphasic Inactivation of the Cdc28-Clb2 Mitotic Kinase and the Role of Cdc20 Foong May Yeong,† Hong Hwa Lim,† Padmashree C. G.,† and Uttam Surana* Institute of Molecular and Cell Biology 30 Medical Drive Singapore 117609

Summary Cdc20, an activator of the anaphase-promoting complex (APC), is also required for the exit from mitosis in Saccharomyces cerevisiae. Here we show that during mitosis, both the inactivation of Cdc28-Clb2 kinase and the degradation of mitotic cyclin Clb2 occur in two steps. The first phase of Clb2 proteolysis, which commences at the metaphase-to-anaphase transition when Clb2 abundance is high, is dependent on Cdc20. The second wave of Clb2 destruction in telophase requires activation of the Cdc20 homolog, Hct1/Cdh1. The first phase of Clb2 destruction, which lowers the Cdc28-Clb2 kinase activity, is a prerequisite for the second. Thus, Clb2 proteolysis is not solely mediated by Hct1 as generally believed; instead, it requires a sequential action of both Cdc20 and Hct1. Introduction In recent years, it has become increasingly clear that the ubiquitin-dependent proteolytic events play a critical role in driving eukaryotic cells through mitosis. Both chromosome segregation and the final exit from M phase, the two landmark events of mitosis, require a key ubiquitin ligase (or E3 enzyme) known as the anaphasepromoting complex (APC or cyclosome) (Irniger et al., 1995; King et al., 1995; Sudakin et al., 1997; Zachariae and Nasmyth, 1999). The APC promotes chromosome separation by marking, via ubiquitination, the anaphase inhibitor Pds1 for destruction by the 26S proteasome (Cohen-Fix et al., 1996; Yamamoto et al., 1996a, 1996b). The proteolytic degradation of Pds1 liberates its binding partner Esp1, which in turn causes the cleavage of the cohesin subunit Scc1 (Michaelis et al., 1997; Ciosk et al., 1998; Uhlmann et al., 1999) allowing sister chromatid separation. Later, the APC mediates cells’ exit from mitosis by targeting for proteolysis the mitotic cyclins such as Clb2 (Glotzer et al., 1991; Holloway et al., 1993; Surana et al., 1993; Irniger et al., 1995). In the budding yeast, APC is composed of at least 12 different subunits including Cdc16, Cdc23, Cdc26, and Cdc27 proteins (Lamb et al., 1994; King et al., 1995; Tugendreich et al., 1995; Zachariae et al., 1996, 1998a). To ubiquitinate various mitotic targets, the APC requires additional proteins, namely Cdc20 and its homolog Hct1/Cdh1 (Sethi et al., 1991; Schwab et al., 1997; Visintin et al., 1997; Zachariae and Nasmyth, 1999). Although * To whom correspondence should be addressed (e-mail: mcbucs@ imcb.nus.edu.sg). † These authors contributed equally to this work.

these WD40 repeat–containing proteins are not permanent members of the core APC (Zachariae and Nasmyth, 1999), their increased binding to APC during the cell cycle correlates well with the increase in APC activity (Fang et al., 1998; Kramer et al., 1998). Therefore, Cdc20 and Hct1 are called the activators of APC (hence the terms APCCdc20 and APCHct1), though the nature of this activation remains elusive. Like Cdc20 and Hct1 in budding yeast, pairs of homologous proteins have also been identified in Schizosaccharomyces pombe, Drosophila, Xenopus, and humans (Weinstein et al., 1994; Dawson et al., 1995; Sigrist and Lehner, 1997; Matsumoto, 1997; Kitamura et al., 1998). It is generally believed that Cdc20 and Hct1 determine the substrate specificity of APC such that the activation by Cdc20 leads to Pds1 destruction at the onset of anaphase while the Hct1-activated APC targets Clb2 for proteolysis, facilitating cells’ departure from mitosis (Schwab et al., 1997; Visintin et al., 1997). It is noteworthy that this substrate specificity was seen in cells arrested in S phase, but not in M phase during which both Clb2 and Pds1 are actually degraded (Visintin et al., 1997). Nevertheless, these studies provided a simple scheme for APC-driven progression through mitosis. However, contrary to this model, it has been reported that in the budding yeast, Cdc20 function is essential not only for chromosome segregation but also for the final exit from mitosis (Yamamoto et al., 1996a; Lim et al., 1998). One of the crucial aspects of progression through mitosis is that Hct1-mediated Clb2 proteolysis occurs only after APCCdc20 consigns Pds1 for destruction. What is the mechanism that ensures the correct timing of these events? Recently, it has been shown that, while phosphorylation of Hct1 by the mitotic kinase renders it inactive (Zachariae et al., 1998b; Jaspersen et al., 1999), dephosphorylation by Cdc14 restores its activity (Visintin et al., 1998). Cdc14, a dual specificity phosphatase (Wan et al., 1992; Taylor et al., 1997), itself appears to be spatially regulated in that it is sequestered to the nucleolar RENT complex during G1 to anaphase (Shou et al., 1999). During late anaphase, Cdc14 is dispersed from the RENT complex (Shou et al., 1999) in a Tem1– dependent manner (Shirayama et al., 1994) resulting in the activation of Hct1, which in turn triggers Clb2 proteolysis. Hence, it is the antagonistic action of the mitotic kinase and the Cdc14 phosphatase on Hct1 that may determine the correct timing of Clb2 proteolysis. However, this scenario requires that, by the time cells reach late telophase, the balance is decisively tipped in favor of net dephosphorylation of Hct1. This would be achievable if the rate of Cdc14-mediated dephosphorylation were inherently higher than phosphorylation by the mitotic kinase Cdc28-Clb. Alternatively, since Cdc20 is also required for mitotic exit (Yamamoto et al., 1996a; Lim et al., 1998), it is conceivable that APCCdc20 causes destruction of a protein that normally inhibits Hct1 and thereby paves the way for rapid activation of Hct1. This possibility has recently been proposed (Tinker-Kulberg and Morgan, 1999). In this report, we have investigated the role of Cdc20

Molecular Cell 502

Table 1. Strains Used in This Study Strain

Genotype

Source

US106 US1165 US1186 US1236 US1338 US1976 US2009 US2018 US2072 US2091

MATa MATa MATa MATa MATa MATa MATa MATa MATa MATa

K. Nasmyth US lab US lab US lab US lab US lab This study This study This study This study

cdc15-2 leu2 ura3 CDC28-cmyc3-TRP1 CLB2-HA3-HIS3 ade2-1 his3 leu2 cdc15-2 CDC28-cmyc3-TRP1 CLB2-HA3-HIS3 leu2 ura3 cdc20-1 pds1-1 bar1⌬:URA3 his3 leu2 trp1 cdc15-2 leu2 pGAL-HA3-CDC20/URA3/CEN cdc15-2 cdc20⌬:LEU2 GAL-CDC20::TRP1 his3 ura3 cdc23-1 pds1-1 bar1⌬:URA3 his3 leu2 trp1 cdc15-2 4⫻GAL-CLB2::TRP1 his3 leu2 ura3 cdc15-2 CDC28-myc3-TRP1 CLB2-HA3-HIS3 hct1⌬:URA3 bar1 ⌬leu2 cdc15-2 his3 4⫻GAL-SIC-cmyc3::URA3 cdc20⌬:leu2 pMET-HA3-CDC20/TRP1/CEN

in the final exit from mitosis. We show that during mitosis, Clb2 degradation, and hence the inactivation of the Cdc28-Clb2 kinase, occurs in a biphasic manner. The first phase commences during anaphase and requires Cdc20 function. This leads to an initial lowering of the mitotic kinase activity that eventually results in, perhaps by facilitating Hct1 activation, the second phase of Clb2 proteolysis in late telophase. Our results suggest that Cdc20 is not a substrate-specific activator of APC that only promotes Pds1 degradation but its function is also essential for the rapid destruction of Clb2 during mitosis. Results Cells Lacking Cdc20 Function Fail to Exit Mitosis Our previous observation that the cdc20⌬ pds1⌬ double mutant, although able to progress through anaphase, eventually arrests in late telophase with high Cdc28Clb2 kinase activity (Lim et al., 1998) raised the possibility that Cdc20 may be required for mitotic exit. To test the validity of this notion under other physiological circumstances, we asked if Cdc20 was essential for the mitotic exit in cdc15–2 mutant cells. Cdc15 cells arrest uniformly in telophase with high Clb2 levels at the restrictive temperature (37⬚C) but exit mitosis promptly and synchronously when returned to the permissive temperature (24⬚C). We constructed a cdc15–2 cdc20⌬ strain kept alive by one copy of GAL-CDC20 integrated at the TRP1 locus (US1976). The mutant cells were first arrested in telophase at 37⬚C and then depleted of Cdc20 protein (Table 1; Figure 1, flow chart). By the end of 5 hr, when approximately 70% of the cells had arrested with divided nuclei, they were released at 24⬚C (Figure 1B, left panel). Samples were withdrawn every 10 min to monitor budding index, Cdc28-Clb2 kinase activity, the levels of Clb2 and Cdc20, and SIC1 gene expression. In a parallel experiment, cdc15–2 cells (US106) were treated and analyzed in a similar manner to serve as a control. As expected, the cdc15 mutant cells promptly degraded Clb2 and inactivated the Cdc28-Clb2 kinase within 50 min after they were returned to 24⬚C (Figure 1A). A sharp rise in the budding index indicated their entry into the new cycle. The cdc15 cdc20⌬ cells, on the other hand, remained arrested in late telophase with high levels of Clb2 and Cdc28-Clb2 kinase activity even after the restoration of Cdc15 function by a shift to the permissive temperature (Figure 1B, left panel). A complete lack of nascent buds in the culture suggested their failure to enter the subsequent cycle.

However, when CDC20 expression was induced by galactose addition, both the Clb2 level and the associated kinase activity declined sharply, reaching a minimum within 60 min (Figure 1B, lower right panel), and the cells entered a new cycle indicated by the appearance of small buds. It is noteworthy that both the levels of Clb2 and Cdc28-Clb2 kinase activity increased significantly during the period when Cdc20 protein was depleted (Figure 1B, left panel [2.5 hr and 5 hr lanes]). These observations strongly argue that CDC20 function is essential for exit from mitosis. They also imply that in the absence of Cdc20 function the Cdc28-Clb2 activity rises dramatically during the progression to telophase. Clb2 Proteolysis in the cdc15 cdc20⌬ Double Mutant Can Be Induced by Inactivation of Cdc28-Clb2 Kinase The failure of cdc15 cdc20⌬ mutant cells to exit mitosis could simply be due to their inability to inactivate the Cdc28-Clb2 kinase in the absence of Cdc20. Alternatively, the role of Cdc20 in mitotic exit may be unrelated to the inactivation of the kinase. One way to distinguish between these possibilities is to determine if inactivation of the kinase by overexpression of the Cdk inhibitor Sic1 will allow these cells to exit mitosis. Therefore, we used a cdc15 cdc20⌬ strain containing four integrated copies of GAL-SIC1-cmyc3, kept alive by MET-HA3-CDC20 (US2091). The cells were arrested in telophase in the absence of Cdc20 and analyzed as described in Figure 2. As expected, the cells in which SIC1 expression was not induced remained arrested in telophase and showed high levels of both Clb2 protein and the Cdc28-Clb2 kinase activity (Figure 2, “glu” panel). However, the induction of Sic1 not only caused inactivation of the mitotic kinase but also led to rapid destruction of Clb2 (Figure 2, “gal” panel), eventually resulting in the appearance of nascent buds (Figure 2, graph, right panel). Hence, the inability to degrade Clb2 in the absence of Cdc20 function can be compensated by the expression of Sic1. These results suggest that the lowering of the mitotic kinase is crucial for the onset of Clb2 proteolysis and that the role of Cdc20 in mitotic exit may be related to initial inactivation of the kinase. Partial Proteolysis of Clb2 and Inactivation of the Mitotic Kinase during Progression to Telophase Our observation that rapid Clb2 proteolysis during telophase-to-G1 transition requires lowering of the mitotic kinase activity implies that normally, Cdc28-Clb2 kinase

Cdc20 and Mitotic Exit 503

Figure 1. Requirement of Cdc20 Function for Exit from Mitosis (A) Exit from mitosis in cdc15–2 (US106) cells released from telophase arrest (control). cdc15–2 cells were arrested in telophase at 37⬚C for a total of 5 hr (see arrest scheme in [B]). The cells were then filtered and resuspended in YPD medium at 24⬚C. Samples withdrawn at the indicated time points were analyzed for the budding index, Clb2-associated kinase activity (PhosphorImager units), Clb2 and Cdc28 protein levels, and CDC20 and SIC1 RNAs. The FACS profile shows DNA content at 2.5 and 5 hr. (B) Cells lacking Cdc20 function fail to exit mitosis. cdc15–2 cdc20⌬:LEU2 GALCDC20::TRP1 (US1976) cells were subjected to the following experimental regime: Cycling culture→2.5 hr in gal, 37⬚C→2.5 hr in glu, 37⬚C (Cdc20 depletion)→2 hr in Raff, 24⬚C (Restoration of Cdc15 function). The culture was then divided into two halves; glucose was added to one half to keep the CDC20 transcription repressed, while galactose was added to the other half to induce Cdc20 expression. Samples were withdrawn at various time points and analyzed as in (A).

should already be partially inactivated by the time cells reach telophase. We tested the validity of this notion in cdc15–2 cells in which the endogenous CDC28 and CLB2 genes had been tagged with three copies of c-myc and HA epitopes (US1186), respectively. The cells, synchronized in G1 by ␣ factor treatment, were allowed to resume cell cycle progression at 36⬚C so that they reach telophase synchronously. The Cdc28-Clb2 kinase activity and the level of Clb2 protein associated with Cdc28 were monitored in samples withdrawn at various times. The proportion of cells with clearly divided nuclei begins to rise at 70 min, reaching its maximum at 150 min, and, as expected, remains high throughout the remaining course of the experiment (Figure 3A) as cells uniformly arrest in telophase. The Cdc28-Clb2 kinase activity increases initially and attains its maximum at around 105 min after the release; thereafter, it begins to decline but finally stabilizes at approximately 40%– 50% of its peak value. The level of Cdc28-associated Clb2 protein follows a pattern similar to that of the kinase activity (finally reaching ⵑ30% of the peak value) implying that loss of the kinase activity is most probably

due to a limited Clb2 dissociation (or degradation) (Figure 3A). To determine if the partial loss of mitotic kinase activity and Clb2 protein also occurs during progression to telophase in normal cell cycle, we monitored these two parameters in wild-type (US1165) cells released from ␣ factor–induced G1 arrest. While the proportion of cells in telophase reached a peak at 100 min, Cdc28-Clb2 kinase activity and the kinase-associated Clb2 begins to decline at 90 min (Figure 3B). An almost similar pattern was seen in wild-type cells that were allowed to resume cell cycle progression at 36⬚C after pheromone treatment (data not shown). These results clearly suggest that the abundance of Clb2 and the corresponding mitotic kinase activity decline substantially from their maximum values at metaphase by the time cells reach late telophase. It is widely believed that while Cdc20 promotes the degradation of the anaphase inhibitor Pds1, Hct1 catalyzes APC-mediated proteolysis of Clb2 (Visintin et al., 1997). To test if the partial proteolysis of Clb2 seen during the progression to telophase requires Hct1,

Molecular Cell 504

Figure 2. Overexpression of Sic1 Induces Degradation of Clb2 in the cdc15–2 cdc20⌬ Mutant cdc15–2 cdc20⌬:LEU2 cells carrying four copies of GAL-SIC1-cmyc3 integrated at the URA3 locus and MET-HA3-CDC20 on TRP1 selectable CEN plasmid (US2091) were incubated for 2 hr at 36⬚C in glucose medium lacking methionine. The culture was filtered, washed, and resuspended in raffinose medium containing methionine (0.24 mg/ml) and maintained at 36⬚C for another 2.5 hr. The culture was then divided into two halves, washed, filtered, and resuspended in methionine medium containing either glucose or galactose at 24⬚C. Samples were withdrawn at the indicated times and analyzed for state of budding, DNA content, Clb2-associated kinase activity (PhosphorImage units), and the level of Clb2 and Cdc28 proteins. The photomicrographs show the morphology of cells at the end of 4.5 hr after the shift to 24⬚C.

cdc15 hct1⌬ (US2072) cells were released at 36⬚C from an early S phase arrest induced by hydroxyurea (HU) treatment. It was necessary to use HU treatment to obtain a synchronous culture in this experiment since cdc15 hct1⌬ cells do not respond well to ␣ factor. Like cdc15–2 mutant, these cells also showed a limited loss of Cdc28-Clb2 kinase activity (ⵑ40%) as they coursed through the cell cycle to eventually arrest in telophase (Figure 3C). The decline in Clb2 level was larger compared to the kinase activity; but it eventually stabilized at a level similar to that seen in HU-arrested culture. These observations imply that the decline in Clb2 abundance as cells approach telophase is not dependent on Hct1 function. Dependence of the Limited Loss of Clb2 and the Mitotic Kinase Activity on APC and Cdc20 Next, we asked if the partial loss of Clb2 and the corresponding kinase activity during progression to telophase requires APC activity. We used the cdc23 pds1 double mutant (US2009) to explore this possibility. Although the cdc23 mutant is unable to initiate anaphase, cdc23 pds1 double mutants can segregate their chromosomes and arrest in telophase. The pheromonetreated mutant cells were released at 36⬚C from G1 arrest, and the levels of total Clb2 and Cdc28-Clb2 kinase

Figure 3. Clb2 Is Partially Degraded prior to Mitotic Exit in an Hct1Independent Manner (A) Partial proteolysis of Clb2 and inactivation of Clb2-associated kinase activity during progression to telophase. cdc15–2 CDC28cmyc3-TRP1 CLB2-HA3-HIS3 (US1186) cells were synchronized in G1 by ␣ factor (0.8 ␮g/ml) treatment for 3.5 hr at 24⬚C. The cells were then filtered, washed, and then released into YPD medium at 36⬚C. Samples were withdrawn every 15 min and analyzed for the state of mitotic spindles, DNA content, Clb2-associated kinase activity (PhosphorImage units), and Cdc28-associated Clb2 and Cdc28 protein (internal control). (B) Cell cycle progression in wild-type cells released from G1 block. Cells carrying tagged version of CDC28 and CLB2 (CDC28-cmyc3TRP1 and CLB2-HA3-HIS3) (US1165) at their respective loci were synchronized in G1 using ␣ factor (0.8 ␮g/ml) for 3.5 hr at 24⬚C. The cells were then filtered, washed, and then released into YPD medium at 24⬚C. Samples withdrawn at various time points were analyzed as in (A). (C) The partial loss of Cdc28-Clb2 kinase activity during progression to telophase does not require Hct1 function. cdc15–2 hct1⌬:URA3 (US2072) cells were synchronized in S phase with hydroxyurea (HU, 30 mg/ml) for 3.5 hr at 24⬚C. The cells were filtered, washed, and released into YPD medium at 36⬚C. Samples were withdrawn every 15 min and analyzed as in (A).

Cdc20 and Mitotic Exit 505

synchronized in G1 by ␣ factor treatment, were released from G1 arrest at 36⬚C, and the levels of total Clb2 protein and the Cdc28-Clb2 kinase activity were monitored as they progressed through mitosis. As with cdc23 pds1 double mutant, no decline in Clb2 or kinase activity was seen even after approximately 70% of cells had reached late mitosis, characterized by well-separated nuclei (Figure 4B). Together, these observations imply that the partial loss of Clb2 and of the mitotic kinase activity seen during progression to telophase requires both Cdc20 function and the APC activity.

Figure 4. Partial Proteolysis of Clb2 during Progression to Telophase Is Dependent Both on APC and Cdc20 Activities (A) Clb2-associated kinase activity in cdc23–1 pds1–1 (US2009) cells remains high. cdc23–1 pds1–1 cells were synchronized in G1 by ␣ factor (1 ␮g/ml) treatment for 3.5 hr at 24⬚C after which they were filtered, washed, and released into YPD meduim at 36⬚C. Samples were analyzed every 10 min to determine the state of nuclear division, DNA content, and the levels of Cdc28-Clb2 kinase activity, Clb2 and Cdc28 proteins. (B) Clb2-associated kinase activity in cdc20–1 pds1–1 (US1236) cells remains high. cdc20–1 pds1–1 cells were treated and analyzed as in (A).

activity were monitored as they progressed to telophase. Although approximately 70% of the cells had reached telophase by end of the experiment, neither the level of Clb2 nor mitotic kinase activity showed any reduction; instead, both continued to accumulate (Figure 4A). This suggests that the partial loss of Clb2 protein and of the corresponding kinase activity during the approach to telophase requires APC activity. Note that Clb2 protein can be detected in cdc23 pds1 mutant cells arrested in G1; this is perhaps due to defective proteolysis in these cells even at 24⬚C. Since this initial decline in the kinase activity does not require Hct1 function (Figure 3C), we suspected that it might be dependent on Cdc20. To test this, we used the cdc20 pds1 double mutant (US1236), which undergoes metaphase-to-anaphase transition at the restrictive temperature and eventually arrests in telophase with separated nuclei (Lim et al., 1998). The mutant cells,

Cdc28-Clb2 Kinase Inactivation and Degradation of Clb2 Are Biphasic A limited loss of Clb2 during the progression to telophase combined with now well-documented, rapid degradation of Clb2 during entry into the subsequent G1 phase immediately imply that Clb2 proteolysis, and thus the inactivation of mitotic kinase, is a two-step process. To visualize this more directly, we allowed cdc15 cells (US1186), synchronized in G1 by ␣ factor treatment, to resume cell cycle progression at 36⬚C. The culture was maintained at 36⬚C for 3.5 hr and then shifted to 24⬚C for another 90 min. The cells accumulated in telophase during the incubation at 36⬚C but entered a new cycle synchronously at 24⬚C as indicated by the appearance of nascent buds (Figure 5, upper right-most panel). The Cdc28-Clb2 kinase activity began to rise at approximately 60 min, peaked at 105 min, then declined to approximately 60% of the maximal value and stabilized at this level as cells accumulated in telophase at 36⬚C (Figure 5, left-most panel). The second phase (sharp drop in the kinase activity) was seen at about 20 min after the cells were transferred to 24⬚C. The level of Clb2 protein (determined at selected time points) showed similar timing and pattern of decline. These observations suggest that the proteolysis of Clb2, and therefore the inactivation of Cdc28-Clb2 kinase, is a two-step process. The first phase of inactivation does not require Cdc15 or Hct1 since it occurs in the absence of either of the gene functions (Figures 3A and 3C). The fact that the second phase of relatively rapid proteolysis sets in only after restoration of Cdc15 function implies that Cdc15 and perhaps all the elements of the exit pathway (Shou et al., 1999; Visintin et al., 1999) are required for the execution of this step. It must be noted that it is very difficult to discern the biphasic nature of kinase inactivation in synchronous wild-type cultures because anaphase-telophase-G1 transition occurs quite rapidly. Cdc20 Catalyzes Clb2 Proteolysis at High Concentrations of Clb2 Our results so far suggest that Cdc20 mediates the first step in the biphasic degradation of Clb2, while the second step requires Cdc15 and most probably other components of the exit pathway. Why does Cdc20-mediated Clb2 proteolysis cease in telophase after the Clb2 level declines to approximately 50% of its peak value and not continue until all of Clb2 is degraded? One possibility is that Cdc20 itself is degraded by the time cells reach late telophase. This seems unlikely since CDC20, a member of the CLB2 cluster genes, continues to be transcribed in telophase (Prinz et al., 1998; Goh et al.,

Molecular Cell 506

Figure 5. Cdc28-Clb2 Kinase Inactivation and Degradation of Clb2 Are Biphasic cdc15–2 CDC28-cmyc3-TRP1 CLB2-HA3-HIS3 (US1186) cells were first synchronized in G1 by ␣ factor (0.8 ␮g/ml) for 3.5 hr at 24⬚C and then released into pheromone-free YPD medium at 36⬚C. Samples were withdrawn every 15 min for 3.5 hr, after which the cells were again filtered and released into YPD medium at 24⬚C. Samples withdrawn at every 10 min were analyzed for state of mitotic spindle, budding index (in 24⬚C samples), DNA content (data not shown), SIC1 RNA, and the Cdc28-Clb2 kinase activity. The levels of Cdc28-associated Clb2 and Cdc28 protein were determined at selected time points as indicated (lowest panel).

2000), and overexpression of Cdc20 does not suppress the cdc15 mutation very effectively (our unpublished data). Alternatively, it is possible that Cdc20 catalyzes Clb2 proteolysis only at higher cellular concentrations. Once the Clb2 content falls below a certain level, APCCdc20 can no longer ubiquitinate Clb2. This notion would predict that if the Clb2 level were transiently raised in telophase arrested cells, its destruction will once again set in and continue until the Clb2 content declines to the original telophase level. To test this, cdc15–2 cells carrying four copies of GALCLB2 integrated at TRP1 locus (US2018) were treated as described in Figure 6. The fates of Clb2 protein and the associated kinase activity were determined at various times prior to and after the transient induction of CLB2 expression. The level of Clb2 and the kinase activity increased substantially due to galactose induction (Figure 6A, left panel) but declined to the initial levels within 40 min of the shift to glucose medium. There was no further decrease with continued incubation in the glucose medium. As expected, the amount of Clb2 and the corresponding kinase activity remained generally unchanged in the control cells (cdc15–2) (Figure 6A, right panel). However, there is a slow and gradual increase in the kinase activity in both cultures, perhaps due to the continued transcription in telophase of the endogenous CLB2 gene. If Cdc20 mediates Clb2 proteolysis at concentrations higher than that in telophase, then the overexpression of Cdc20 might be expected to cause only limited degradation of Clb2 in metaphase-arrested cells. To determine the validity of this notion, cdc15–2 cells carrying

GAL-HA3-CDC20 on a CEN plasmid (US1338) were arrested in prenuclear division stage by nocodazole (Noc) treatment at 24⬚C. Cdc20 synthesis was induced by addition of galactose, and the level of Clb2 was determined. Clearly, the Clb2 abundance initially declines due to excess Cdc20 (Figure 6B); however, subsequently it stabilizes at a reduced level and does not decline any further. This observation is seemingly inconsistent with a previous study (Visintin at al., 1997) in which overexpression of Cdc20 in Noc-arrested cells led to a complete destruction of Clb2. We believe this might have been due to excessive Cdc20 expression from three copies of GAL-CDC20 (R. Visintin et al., personal communication). Nevertheless, our results imply that Cdc20 catalyzes Clb2 degradation at relatively higher concentrations but that it fails to do so once Clb2 levels are below a certain threshold.

Accumulation of Cdc28-Clb5 Kinase Activity during Telophase in the Absence of Cdc20 Thus far, our results suggest that the role of Cdc20 in exit may be linked to the limited proteolysis of Clb2 to perhaps allow the activation of Hct1 by Cdc14. However, recently it has been proposed that Cdc20 facilitates mitotic exit by mediating the destruction of S phase cyclin Clb5 (Shirayama et al., 1999). We have already seen the accumulation of Cdc28-Clb2 kinase activity in the absence of Cdc20 as cells progress to telophase (Figure 1). To test if the lack of Cdc20 function in telophase also leads to an accumulation of Cdc28-Clb5 kinase activity, cdc15 cdc20⌬ cells were treated and

Cdc20 and Mitotic Exit 507

in exit may be to reduce the collective activity of Cdc28Clb kinase complexes so as to facilitate the eventual activation of Hct1 by Cdc14. These results also suggest that for a successful exit from mitosis, Cdc20 function is necessary throughout the progression to telophase. Discussion

Figure 6. Cdc20 Mediates Clb2 Degradation at Higher Levels of Clb2 (A) cdc15–2 cells carrying four copies of GAL-CLB2 integrated at TRP1 locus (US2018) were arrested in late telophase at 36⬚C (raffinose). Three samples, separated by 10 min intervals, were taken prior to induction of CLB2 for 2 hr (by addition of 2% galactose). Cycloheximide (1 mg/ml) and 2% glucose were added to inhibit Clb2 induction. Samples were taken at the indicated times and analyzed for DNA content, Clb2-associated kinase activity, and the levels of Clb2 and Cdc28 proteins. As a control, cdc15–2 cells (US106) were subjected to the same treatment. The levels of kinase activity that Clb2 and Cdc28 proteins represented in the graphs are relative to their levels in the 0 min sample. (B) Overexpression of Cdc20 induces limited Clb2 degradation in nocadazole-arrested cells. cdc15–2 cells carrying GAL-HA3-CDC20 on a CEN vector (US1338) were arrested in nocodazole (15 ␮g/ml) at 24⬚C. Galactose was added to induce Cdc20 expression. The DNA content, Clb2-associated kinase activity, and the levels of HA3Cdc20, Clb2, and Cdc28 proteins were monitored in samples withdrawn at every 15 min. The FACS profile shows continued arrest with 2 N DNA content during the experiment.

analyzed in a manner identical to that described in Figure 1. Like the Cdc28-Clb2 kinase activity, Cdc28-Clb5 activity accumulates upon depletion of Cdc20 but declines rapidly once Cdc20 expression is induced again (Figure 7). This observation implies that the role of Cdc20

The present study was prompted by our earlier observation that cdc20⌬ pds1⌬ cells arrest in telophase, implying that Cdc20 function may be required for exit from mitosis (Lim et al., 1998). We have reconfirmed the requirement of Cdc20 for the mitotic exit in another cellular context by using cdc15–2 cdc20⌬ double mutant. Our finding that cdc15–2 cdc20⌬ cells cannot degrade Clb2 even after Cdc15 function is restored (Figure 1) indicates that without Cdc20 function, the exit pathway, which requires Tem1, Cdc15, Cdc14, Hct1, and APC functions, is not fully active. These results are identical to the ones described in a recent report (Tinker-Kulberg and Morgan, 1999). It has been suggested that while dephosphorylation by Cdc14 is required to activate Hct1 (Visintin et al., 1998; Jaspersen et al., 1999), phosphorylation by the mitotic kinase leads to its inactivation (Jaspersen et al., 1999). If the rapid proteolysis of Clb2 were to commence on schedule during mitotic exit, this balance should be decisively tipped in favor of net dephosphorylation of Hct1 by the time that cells are in late telophase. What factor could be responsible for tipping the balance? It has been proposed that the accumulation of Sic1 at this juncture could help lower the kinase activity (Schwab et al., 1997) such that Hct1 dephosphorylation by Cdc14 is the favored reaction. However, we observe that in the absence of Cdc20, SIC1 is transcribed at a very low level (Figure 1B, upper right panel). Even after the first phase of decline in the kinase activity, there is only a modest increase in the level of SIC1 transcription, as evident from the cdc15–2 mutant arrested at 36⬚C (Figure 5). It is only when the kinase activity sharply falls upon release from the telophase arrest that the SIC1 transcription shows significant increase (Figure 1). This implies that it may not be the Sic1 expression, but Cdc20 itself, that catalyzes the initial decrease in the Cdc28Clb2 kinase required for both the subsequent catastrophic Clb2 destruction and SIC1 transcription. Support for this role of Cdc20 comes from our findings that the inability to degrade Clb2 in the absence of Cdc20 can be overcome by lowering the mitotic kinase activity (Figure 2). Furthermore, by the time cells progress to telophase, the Cdc28-Clb2 kinase activity and the Clb2 levels are already reduced to approximately 50% of their peak value in G2/M (Figure 3A). This initial decline in the kinase activity is independent of Hct1 function (Figure 3C), though it is dependent on both Cdc20 and the APC (Figure 4). Hence, the role of Cdc20 could be to bring about the initial drop in kinase actvity that would allow the initiation of SIC1 transcription. The concerted actions of both Cdc20 and Sic1 in lowering the kinase activity would then permit the Cdc14-mediated dephosphorylation of Hct1 to override its phosphorylation by the mitotic kinase. Our results that indicate a role for Cdc20 in the partial

Molecular Cell 508

Figure 7. Cdc28-Clb5 Kinase Activity in Telophase in the Absence of Cdc20 Function Cdc15 cdc20⌬ cells kept alive by one copy of GAL-CDC20 integrated at the TRP1 locus (US1976) were arrested in telophase and then released in a manner identical to that described in Figure 1. Cells were analyzed for Cdc28-Clb2 and Cdc28-Clb5 activities using Clb2- and Clb5-specific antibodies.

Clb2 destruction, coupled with the finding that Hct1 is also involved in Clb2 proteolysis (Schwab et al., 1997; Visintin et al., 1997) imply that Clb2 proteolysis occurs in two stages. The first phase of destruction occurs, as our data suggest, during the progression to telophase and is dependent on Cdc20 function. The second phase of Clb2 proteolysis begins during the final departure from mitosis and is, as previously established (Schwab et al., 1997; Visintin et al., 1997), mediated by Hct1. This biphasic nature of Clb2 degradation, and hence inactivation of Cdc28-Clb2 kinase, is quite clearly seen in cdc15–2 cells synchronously traversing the cell cycle (Figure 5). A partial reduction in the Cdc28-Clb2 kinase activity during cdc15–2 cells’ approach to telophase from Noc-induced arrest was also observed in a previous study (Irniger et al., 1995). That APCCdc20 and APCHct1 show apparent lack of substrate specificity toward Clb2 is perhaps not surprising, because the strict substrate specificity as originally proposed (Visintin et al., 1997) is not consistent with a number of observations reported in the literature. The addition of Cdc20 to the purified preparation of APC from both Xenopus and HeLa cells leads to ubiquitination of cyclin B (Fang et al., 1998; Kramer et al., 1998). Moreover, Fizzy and Fzr, Cdc20 and Hct1 homologs in Drosophila, both show overlapping substrate specificities (Dawson et al., 1995; Sigrist and Lehner, 1997). In budding yeast, Cdc20 may also target Clb3 and Clb5 for proteolytic degradation (Zachariae and Nasmyth, 1999). In view of these findings, it seems likely that the limited destruction of Clb2 observed during cells’ approach to telophase is due to the targeting of Clb2 by APCCdc20. This raises an important question: why does APCCdc20mediated proteolysis of Clb2 stop once the Clb2 concentration reaches approximately 50% of its peak value? We propose that APCCdc20 targets Clb2 only at a higher concentration. This is supported by our observation that cdc15–2 cells arrested in telophase can still cause degradation of Clb2 if its level were raised above

its “telophase value” (Figure 6A). The proteolysis ceases once the Clb2 declines to its “normal” telophase level. The fact that overexpression of Cdc20 can cause only limited Clb2 destruction in nocodazole-arrested cells (Figure 6B) is also consistent with this proposal. These results also explain an intriguing anomaly in a previous study (Visintin et al., 1997) where the substrate specificities of Cdc20 and Hct1 toward Pds1 and Clb2, respectively, were seen upon Cdc20 overexpression only in HU-arrested cells but not in cells blocked in mitosis due to nocodazole treatment. We have noted (data not shown) consistently that the Clb2 level in HUarrested cells is generally much lower compared to that in Noc-blocked cells. In view of our observation that Cdc20 targets Clb2 for destruction at relatively higher concentration, it is not surprising that overexpression of Cdc20 leads to discernable Clb2 destruction during mitosis but not in early S phase. If the role of Cdc20 in mitotic exit is to catalyze Clb2 proteolysis during the first phase of destruction as we have proposed, then it may seem inconsistent that cdc15–2 cdc20⌬ cells arrest in late telophase in the experiment described in Figure 1. The way the experiment had been performed, Cdc20 should have already carried out its exit-relevant function by the time the mutant cells arrived in late telophase. However, it must be noted that as Cdc20 is being depleted during the incubation in glucose medium at 37⬚C, Clb2 accumulates to a high level again (Figure 1, 2.5 hr and 5 hr lanes), perhaps due to continuing CLB2 transcription. A corresponding rise in the Cdc28-Clb2 kinase activity would thus reinstate the requirement for Cdc20 during exit from mitosis. This rationale further strengthens the importance of Cdc20 function in preventing the kinase activity from reaching an inhibitory level during progression to telophase as it would severely compromise the timely activation of Hct1 by Cdc14. What is the physiological relevance of the Cdc20induced limited loss of Cdc28-Clb2 kinase activity during the approach to telophase? We propose that the

Cdc20 and Mitotic Exit 509

overlapping specificity of APCCdc20 allows it to catalyze, during metaphase-to-anaphase transition, not only the destruction of Pds1 but also limited degradation of Clb2 due to high Clb2 concentration at this stage of the cell cycle. This would lead to lowering of the kinase activity resulting in the initiation of the Hct1 activation cycle and the expression of Cdk-inhibitor Sic1 as more Swi5 enters the nucleus (Visintin et al., 1998; Jaspersen et al., 1999). The initial rise in Hct1 activity and Sic1 expression can initiate an amplification loop causing progressively higher levels of both Sic1 and Hct1 activity, eventually culminating in rapid destruction of Clb2. Therefore, the role of Cdc20 in mitotic exit is to pave the way for the Cdc14-mediated activation of Hct1 by lowering the Cdc28-Clb2 kinase activity. Recently, Shirayama et al. (1999), using a genetic screen, have isolated mutations in the CLB5 gene that can allow the cdc20 pds1 double mutant to exit mitosis implying that APCCdc20 mediates exit from mitosis by promoting degradation of S phase cyclin Clb5. Our results would predict that the mutation in Clb2 should also permit the cdc20 pds1 mutant to undergo M-G1 transition. However, the nature of the genetic screen might have precluded isolation of mutations in the CLB2 gene since the clb2 pds1 double mutant is inviable (Shirayama et al., 1999). Thus, our study together with that of Shirayama et al. (1999) suggests that Cdc20 may promote Hct1 activation by lowering the overall Cdc28/Clb kinase activity by targeting Clb2, Clb5, and perhaps other Clbs for proteolysis. Our observation that both Cdc28-Clb2 and Cdc28-Clb5 kinase activities accumulate in the absence of Cdc20 during progression to telophase but rapidly decline upon Cdc20 induction (Figures 1 and 7) supports this notion. The possibility that Cdc20 may facilitate exit from mitosis by targeting an inhibitor of Hct1 (factor X) in telophase has been recently suggested (Morgan, 1999). We suspect that Cdc28/Clb kinase complexes collectively may be the factor X. The scheme we have proposed for the destruction of Clb2 is a clear departure from the purely deterministic model according to which Cdc20 and Hct1 are the substrate-specific activators of the APC (Visintin et al., 1997). We surmise that the overlapping specificity of Cdc20 toward Pds1, Clb2, and Clb5 destruction is not merely a “slippage in the system”; instead, this flexibility plays an important physiological role in fine tuning the timing of Clb2 degradation. The kinetic intertwining of this kind between various cellular reactions may be widely used by cells to sharply define the timing of some of the key events during cell division. Experimental Procedures Yeast Media and Reagents All strains used in this study were haploid (unless otherwise stated) and were derived from the wild-type strain W303. Cells were grown in standard yeast extract peptone (YP) or selective medium supplemented with 2% glucose or raffinose-galactose. Strains and Plasmids The strains with various genotypes (shown in Table 1) were constructed by a combination of standard molecular genetic techniques such as gene transplacement, gene disruption, and tetrad dissection. The MET-HA3-CDC20 was constructed by cloning the blunt-ended

BsgI-HindIII fragment containing CDC20 tagged with triple-HA tag at the N-terminal) behind a 600 base pair MET3 promoter. The CDC20 gene was disrupted in the required strain by transformation with a cdc20⌬:LEU2 disruption cassette in which most of the ORF of CDC20 (between XbaI to BamHI) was replaced by a 1.6 kb LEU2 marker. Southern blot analysis was carried out on the resulting transformants to confirm the disruption. Synchronization by Treatment with ␣ Factor or Hydroxyurea For experiments requiring synchronous cultures, exponential phase cells were grown in medium at 24⬚C containing either ␣ factor (5 ␮g/ml for BAR1 cells and 0.8 ␮g/ml for bar1⫺ cells) or 30 mg/ml hydroxyurea. After 3–3.5 hr of treatment, cells were filtered, washed, and resuspended in fresh medium, preincubated at the appropriate temperature. Samples were then taken at specific intervals for measurement of H1 kinase activity, Western blot analysis, Northern blot analysis, flow cytometry, and immunofluorescent staining. Cell Extracts, Kinase Assays, Immunoprecipitation, and Western Blot Analysis For determining Clb2-Cdc28 kinase activity, cells were harvested by centrifugation at 4⬚C, washed with ice-cold stop mix, and used for the preparation of cell extracts (Surana et al., 1993). For the determination of kinase activities, immunoprecipitation of Cdc28Clb2 and Cdc28-Clb5 complexes were carried out using polyclonal antibodies against Clb2 and Clb5 (a gift from Etienne Schwob) at 1:30 dilution. The kinase assay was performed as described in Surana et al. (1993). The levels of kinase activity were quantitated using PhosphorImager (Molecular Dynamics [Amersham Pharmacia Biotech AB, Uppsala, Sweden]). Since the PhosphorImager assigns arbitrary units, the y axis scales in various graphs showing kinase activity in different experiments are not identical. To monitor the amount of Cdc28-associated Clb2, immunoprecipitation of Clb2-HA3/Cdc28-cmyc3 complex using 9E10 ascites (a gift from Bor Luen Tang) cross-linked to CNBr-activated Sepharose 4B beads (Pharmacia) were used. The beads were prepared according to the Pharmacia protocol provided. Total protein (80–240 ␮g) was incubated with 30 ␮l of beads were prepared according to the Pharmacia protocol 1:1 9E10 beads in lysis buffer at 4⬚C for 1–1.5 hr on a rotator. After immunoprecipitation, the beads were washed four times with RIPA buffer and protein gel loading dye subsequently added to the beads. Proteins were resolved on 10% SDS-PAGE and transferred onto nitrocellulose membrane. For the preparation of crude extracts for Western blot analysis, precipitation of proteins by TCA was carried out. Typically, cells equivalent to OD600 ⫽ 3 were spun down and resuspended in 1 ml ice-cold water. Then 150 ␮l ice-cold YEX lysis buffer (1.85 M NaOH, 7.5% ␤-mercaptoethanol) was added, and the suspension was kept on ice for 10 min; after which, 150 ␮l 50% ice-cold TCA was added, and the suspension was left on ice for another 10 min. The precipitated proteins were obtained by spinning in microfuge for 5 min in cold room. The protein pellet was resuspended in a solution containing 100 ␮l of 1⫻ gel loading buffer and 7.5 ␮l 1 M Tris (pH 8). Of this suspension, 10–20 ␮l was used for SDS-PAGE. For Western blot analyses, immunodetection of Cdc28 was carried out using anti-Cdc28 polyclonal anitbodies (1:500 dilution), Clb2 using anti-Clb2 polyclonal antibodies (1:500 dilution), cmyc fusion proteins using anti-cmyc (1:100 dilution; Santa Cruz Biotechnology, Santa Cruz, CA), and HA fusion proteins using anti-HA monoclonal (1:100 dilution; Boehringer Mannheim, Mannheim, Germany) or antiHA polyclonal (1:250 dilution, Santa Cruz Biotechnology). Enhanced chemiluminescence kit from Amersham was used for all Western blot analyses according to the manufacturer’s instructions. Northern Blot Hybridization, Flow Cytometry, and Immunofluorescent Staining RNA extraction was carried out according to the method of Cross and Tinkelenberg (1991), and Northern blot analysis was performed as described by Price et al. (1991). The method described by Lim et al. (1996) was used for flow cytometry. Immunofluorescent staining was performed according to Kilmartin and Adams (1984) using anti-tubulin antibody YOL1/34 (a gift from J. V. Kilmartin) for the

Molecular Cell 510

staining of microtubules, and DNA was visualized after staining with diamidinophenylindole (DAPI).

APC-mediated proteolysis of both Pds1 and Clb2 during M phase in budding yeast. Curr. Biol. 8, 231–234.

Acknowledgments

Matsumoto, T. (1997). A fission yeast homolog of CDC20/p55CDC/ Fizzy is required for recovery from DNA damage and genetically interacts with p34cdc2. Mol. Cell. Biol. 17, 742–750.

We thank Ya Wang for technical assistance, Bor Luen Tang for gift of 9E10 ascites, and Raphael Tham and Richard Chng for photographic assistance. We are grateful to Etienne Schwob for the polyclonal antibodies against Clb5. This work was supported by the National Science and Technology Board, Singapore. Received November 3, 1999; revised January 21, 2000. References Ciosk, R., Zachariae, W., Michaelis, C., Shevchenko, A., Mann, M., and Nasmyth, K. (1998). An ESP1/PDS1 complex regulates loss of sister chromatid cohesion at the metaphase to anaphase transition in yeast. Cell 93, 1067–1076. Cohen-Fix, O., Peters, J.-M., Kirschner, M.W., and Koshland, D. (1996). Anaphase initiation in Saccharomyces cerevisiae is controlled by the APC-dependent degradation of the anaphase inihibitor Pds1p. Genes Dev. 10, 3081–3093. Cross, F.R., and Tinkelenberg, A.H. (1991). A potential positive feedback loop controlling CLN1 and CLN2 gene expression at the start of yeast cell cycle. Cell 65, 875–883. Dawson, I.A., Roth, S., and Artavanis-Tsakonas, S. (1995). The Drosophila cell cycle gene fizzy is required for normal degradation of cyclins A and B during mitosis and has homology to the CDC20 gene of Saccharomyces cerevisiae. J. Cell Biol. 129, 725–737. Fang, G., Yu, H., and Kirschner, M.W. (1998). Direct binding of CDC20 protein family members activates the anaphase-promoting complex in mitosis and G1. Mol. Cell 2, 163–171. Glotzer, M., Murray, A.W., and Kirschner, M.W. (1991). Cyclin is degraded by the ubiquitin pathway. Nature 349, 132–138. Goh, P.-Y., Lim, H.H., and Surana, U. (2000). Cdc20 protein contains a destruction box but, unlike Clb2, its proteolysis is not acutely dependent on the activity of anaphase promoting complex. Eur. J. Biochem. 267, 434–449. Holloway, S.L., Glotzer, M., King, R.W., and Murray, A.W. (1993). Anaphase is initiated by proteolysis rather than by the inactivation of maturation-promoting factor. Cell 73, 1393–1402. Irniger, S., Piatti, S., Michaelis, C., and Nasmyth, K. (1995). Genes involved in sister chromatid separation are needed for B-type cyclin proteolysis in budding yeast. Cell 81, 269–278. Jaspersen, S.L., Charles, J.F., and Morgan, D.O. (1999). Inhibitory phosphorylation of the APC regulator Hct1 is controlled by the kinase Cdc28 and the phosphatase Cdc14. Curr. Biol. 9, 227–236. Kilmartin, J.V., and Adams, A.E.M. (1984). Structural arrangements of tubulin and actin during the cell cycle of yeast Saccharomyces. J. Cell Biol. 98, 922–933. King, R.W., Peters, J.M., Tugendreich, S., Rolfe, M., Heiter, P., and Kirschner, M.W. (1995). A 20S complex containing CDC27 and CDC16 catalyzes the mitosis-specific conjugation of ubiquitin to cyclin B. Cell 81, 279–288. Kitamura, K., Maekawa, H., and Shimoda, C. (1998). Fission yeast Ste9, a homolog of Hct1/Cdh1 and Fizzy-related, is a novel negative regulator of cell cycle progression during G1-phase. Mol. Biol. Cell 9, 1065–1080. Kramer, E.R., Gieffers, C., Holzl, G., Hengstschlager, M., and Peters, J.M. (1998). Activation of the human anaphase-promoting complex by proteins of the CDC20/Fizzy family. Curr. Biol. 8, 1207–1210. Lamb, J.R., Michaud, W.A., Sikorski, R.S., and Heiter, P.A. (1994). Cdc16p, Cdc23p and Cdc27p form a complex essential for mitosis. EMBO J. 13, 4321–4328. Lim, H.H., Goh, P.-Y., and Surana, U. (1996). Spindle pole body separation in Saccharomyces cerevisiae requires dephosphorylation of the tyrosine 19 residue of Cdc28. Mol. Cell. Biol. 16, 6385– 6397. Lim, H.H., Goh, P.-Y., and Surana, U. (1998). Cdc20 is essential for

Michaelis, C., Ciosk, R., and Nasmyth, K. (1997). Cohesins: chromosomal proteins that prevent premature separation of sister chromatids. Cell 91, 35–45. Morgan, D.O. (1999). Regulation of the APC and the exit from mitosis. Nat. Cell Biol. 1, E47–E53. Price, C., Nasmyth, K., and Schuster, T. (1991). A general approach to the isolation of cell cycle–regulated genes in the budding yeast Saccharomyces cerevisiae. J. Mol. Biol. 218, 543–556. Prinz, S., Huang, E.S., Visintin, R., and Amon, A. (1998). The regulation of Cdc20 proteolysis reveals a role for the APC components Cdc23 and Cdc27 during S phase and early mitosis. Curr. Biol. 8, 750–760. Schwab, M., Lutum, A.S., and Seufert, W. (1997). Yeast Hct1 is a regulator of Clb2 cyclin proteolysis. Cell 90, 683–693. Sethi, N., Monteagudo, C., Koshland, D., Hogan, E., and Burke, D. (1991). The CDC20 gene product of Saccharomyces cerevisiae, a ␤-transducin homolog, is required for a subset of microtubuledependent cellular processes. Mol. Cell. Biol. 11, 5592–5602. Shirayama, M., Matsui, Y., and Toh, E.A. (1994). The yeast TEM1 gene, which encodes a GTP-binding protein, is involved in termination of M phase. Mol. Cell Biol. 14, 7476–7482. Shirayama, M., Toth, A., Galova, M., and Nasmyth, K. (1999). APCCdc20 promotes exit from mitosis by destroying the anaphase inhibitor Pds1 and cyclin Clb5. Nature 402, 203–207. Shou, W., Seol, J.H., Shevchenko, A., Baskerville, C., Moazed, D., Chen, Z.W.S., Jang, J., Shevchenko, A., Charbonneau, H., and Deshaies, R.J. (1999). Exit from mitosis is triggered by Tem1-dependent release of the protein phosphatase Cdc14 from nucleolar RENT complex. Cell 97, 233–244. Sigrist, S.J., and Lehner, C.F. (1997). Drosophila fizzy-related downregulates mitotic cyclins and is required for cell proliferation arrest and entry into endocycles. Cell 90, 671–681. Sudakin, V., Ganoth, D., Dahan, A., Heller, H., Hershko, J., Luca, F.C., Ruderman, J.V., and Hershko, A. (1997). The cyclosome, a large complex containing cyclin-selective ubiquitin ligase activity, targets cyclins for destruction at the end of mitosis. Mol. Cell. Biol. 6, 185–198. Surana, U., Amon, A., Dowzer, C., Mcgrew, J., Byers, B., and Nasmyth, K. (1993). Destruction of the CDC28/CLB mitotic kinase is not required for metaphase to anaphase transition in budding yeast. EMBO J. 12, 1969–1978. Taylor, G.S., Liu, Y., Baskerville, C., and Charbonneau, H. (1997). The activity of Cdc14p, an oligomeric dual specificity protein phosphatase from Saccharomyces cerevisiae, is required for cell cycle progression. J. Biol. Chem. 272, 24054–24063. Tinker-Kulberg, R.L., and Morgan, D.O. (1999). Pds1 and Esp1 control both anaphase and mitotic exit in normal cells and after DNA damage. Genes Dev. 13, 1936–1949. Tugendreich, S., Tomkiel, J., Earnshaw, W., and Hieter, P. (1995). CDC27Hs colocalizes with CDC16Hs to the centromere and mitotic spindle and is essential for the metaphase to anaphase transition. Cell 81, 261–268. Uhlmann, F., Lottspeich, F., and Nasmyth, K. (1999). Sister-chromatid separation at anaphase onset is promoted by cleavage of the cohesin subunit Scc1. Nature 400, 37–42. Visintin, R., Prinz, S., and Amon, A. (1997). CDC20 and CDH1: a family of substrate-specific activators of APC-dependent proteolysis. Science 278, 460–463. Visintin, R., Craig, K., Hwang, E.S., Prinz, S., Tyres, M., and Amon, A. (1998). The phosphatase Cdc14 triggers mitotic exit by reversal of Cdk-dependent phosphorylation. Mol. Cell 2, 709–718. Visintin, R., Hwang, E.S., and Amon, A. (1999). Cfi1 prevents premature exit from mitosis by anchoring Cdc14 phosphatase in the nucleolus. Nature 398, 818–823.

Cdc20 and Mitotic Exit 511

Wan, J., Xu, H., and Grunstein, M. (1992). CDC14 of Saccharomyces cerevisiae. J. Biol. Chem. 267, 11274–11280. Weinstein, J., Jacobsen, F.W., Hsu-Chen, J., Wu, T., and Baum, L.G. (1994). A novel mammalian protein, p55CDC, present in dividing cells is associated with protein kinase activity and has homology to the Saccharomyces cerevisiae cell division cycle proteins Cdc20 and Cdc4. Mol. Cell. Biol. 14, 3350–3363. Yamamoto, A., Guacci, V., and Koshland, D. (1996a). Pds1p is required for faithful execution of anaphase in the yeast, Saccharomyces cerevisiae. J. Cell Biol. 133, 85–97. Yamamoto, A., Guacci, V., and Koshland, D. (1996b). Pds1p, an inhibitor of anaphase in budding yeast, plays a critical role in the APC and checkpoint pathways. J. Cell Biol. 133, 99–110. Zachariae, W., and Nasmyth, K. (1999). Whose end is destruction: cell division and the anaphase-promoting complex. Genes Dev. 13, 2039–2058. Zachariae, W., Shin, T.H., Galova, M., Obermaier, B., and Nasmyth, K. (1996). Identification of subunits of the anaphase-promoting complex of Saccharomyces cerevisiae. Science 274, 1201–1204. Zachariae, W., Shevchenko, A., Andrews, P.D., Stark, M.J.R., Mann, M., and Nasmyth, K. (1998a). Mass spectrometric analysis of the anaphase complex from budding yeast: identification of a subunit related to cullins. Science 279, 1216–1219. Zachariae, W., Schwab, M., Nasmyth, K., and Seufert, W. (1998b). Control of cyclin ubiquitination by CDK-regulated binding of Hct1 to the anaphase promoting complex. Science 282, 1721–1724.