APC ste9/srw1 promotes degradation of mitotic cyclins in G 1 and is inhibited by cdc2 phosphorylation
2000; Springer Nature; Volume: 19; Issue: 15 Linguagem: Inglês
10.1093/emboj/19.15.3945
ISSN1460-2075
AutoresM. A. Blanco, Alberto Sánchez‐Díaz, José M. de Prada, Sergio Moreno,
Tópico(s)Plant Molecular Biology Research
ResumoArticle1 August 2000free access APCste9/srw1 promotes degradation of mitotic cyclins in G1 and is inhibited by cdc2 phosphorylation Miguel A. Blanco Miguel A. Blanco Instituto de Microbiología Bioquímica, Departamento de Microbiología y Genética, CSIC/Universidad de Salamanca, Edificio Departamental, Campus Miguel de Unamuno, 37007 Salamanca, Spain Search for more papers by this author Alberto Sánchez-Díaz Alberto Sánchez-Díaz Instituto de Microbiología Bioquímica, Departamento de Microbiología y Genética, CSIC/Universidad de Salamanca, Edificio Departamental, Campus Miguel de Unamuno, 37007 Salamanca, Spain Search for more papers by this author José M. de Prada José M. de Prada Instituto de Microbiología Bioquímica, Departamento de Microbiología y Genética, CSIC/Universidad de Salamanca, Edificio Departamental, Campus Miguel de Unamuno, 37007 Salamanca, Spain Search for more papers by this author Sergio Moreno Corresponding Author Sergio Moreno Instituto de Microbiología Bioquímica, Departamento de Microbiología y Genética, CSIC/Universidad de Salamanca, Edificio Departamental, Campus Miguel de Unamuno, 37007 Salamanca, Spain Search for more papers by this author Miguel A. Blanco Miguel A. Blanco Instituto de Microbiología Bioquímica, Departamento de Microbiología y Genética, CSIC/Universidad de Salamanca, Edificio Departamental, Campus Miguel de Unamuno, 37007 Salamanca, Spain Search for more papers by this author Alberto Sánchez-Díaz Alberto Sánchez-Díaz Instituto de Microbiología Bioquímica, Departamento de Microbiología y Genética, CSIC/Universidad de Salamanca, Edificio Departamental, Campus Miguel de Unamuno, 37007 Salamanca, Spain Search for more papers by this author José M. de Prada José M. de Prada Instituto de Microbiología Bioquímica, Departamento de Microbiología y Genética, CSIC/Universidad de Salamanca, Edificio Departamental, Campus Miguel de Unamuno, 37007 Salamanca, Spain Search for more papers by this author Sergio Moreno Corresponding Author Sergio Moreno Instituto de Microbiología Bioquímica, Departamento de Microbiología y Genética, CSIC/Universidad de Salamanca, Edificio Departamental, Campus Miguel de Unamuno, 37007 Salamanca, Spain Search for more papers by this author Author Information Miguel A. Blanco1, Alberto Sánchez-Díaz1, José M. de Prada1 and Sergio Moreno 1 1Instituto de Microbiología Bioquímica, Departamento de Microbiología y Genética, CSIC/Universidad de Salamanca, Edificio Departamental, Campus Miguel de Unamuno, 37007 Salamanca, Spain ‡M.A.Blanco and A.Sánchez-Díaz contributed equally to this work *Corresponding author. E-mail: [email protected] The EMBO Journal (2000)19:3945-3955https://doi.org/10.1093/emboj/19.15.3945 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions Figures & Info Fission yeast ste9/srw1 is a WD-repeat protein highly homologous to budding yeast Hct1/Cdh1 and Drosophila Fizzy-related that are involved in activating APC/C (anaphase-promoting complex/cyclosome). We show that APCste9/srw1 specifically promotes the degradation of mitotic cyclins cdc13 and cig1 but not the S-phase cyclin cig2. APCste9/srw1 is not necessary for the proteolysis of cdc13 and cig1 that occurs at the metaphase–anaphase transition but it is absolutely required for their degradation in G1. Therefore, we propose that the main role of APCste9/srw1 is to promote degradation of mitotic cyclins when cells need to delay or arrest the cell cycle in G1. We also show that ste9/srw1 is negatively regulated by cdc2-dependent protein phosphorylation. In G1, when cdc2–cyclin kinase activity is low, unphosphorylated ste9/srw1 interacts with APC/C. In the rest of the cell cycle, phosphorylation of ste9/srw1 by cdc2–cyclin complexes both triggers proteolysis of ste9/srw1 and causes its dissociation from the APC/C. This mechanism provides a molecular switch to prevent inactivation of cdc2 in G2 and early mitosis and to allow its inactivation in G1. Introduction Ubiquitin-mediated proteolysis plays an important role in the control of cell cycle progression. Targeted protein degradation is necessary for multiple processes in mitosis and at the G1–S transition (reviewed in Krek, 1998; Peters, 1998; Morgan, 1999; Zachariae and Nasmyth, 1999). Proteolysis of Cut2/Pds1, which triggers sister chromatid separation, is required for the metaphase–anaphase transition, and degradation of mitotic cyclins is essential to exit from mitosis (Glotzer et al., 1991; Holloway et al., 1993; Irniger et al., 1995; Cohen-Fix et al., 1996; Funabiki et al., 1996). Ubiquitin is transferred to target substrates through several enzymatic reactions involving the E1, E2 and E3 enzymes. The ubiquitin ligase or E3 interacts with both the substrate and the E2 and determines the substrate specificity and the timing of degradation. Anaphase-promoting complex/cyclosome (APC/C) is the cell cycle-regulated ubiquitin ligase or E3 that mediates the degradation of Cut2/Pds1 and mitotic cyclins. APC/C is a multiprotein complex that it is activated at the metaphase–anaphase transition and remains active until late G1 (Amon et al., 1994; Brandeis and Hunt, 1996). APC/C activity and substrate specificity are regulated by its association with highly conserved regulatory activators that are part of the subfamilies of the WD40 repeat proteins, called Cdc20 and Hct1/Cdh1 in budding yeast (Schwab et al., 1997; Visintin et al., 1997), slp1 and ste9/srw1 in fission yeast (Yamaguchi et al., 1997; Kim et al., 1998; Kitamura et al., 1998), Fizzy and Fizzy-related in Drosophila and Xenopus (Dawson et al., 1995; Sigrist et al., 1995; Sigrist and Lehner, 1997; Lorca et al., 1998) and p55CDC/CDC20 and HCT1/CDH1 in humans (Weinstein et al., 1994; Fang et al., 1998; Kramer et al., 1998). The APCCdc20 (APCslp1 in fission yeast) complex promotes sister chromatid separation by ubiquitylating the anaphase inhibitor (or securin) Pds1 (cut2 in Schizosaccharomyces pombe) and by liberating the separin Esp1 (cut1 in S.pombe), which in turn causes either cleavage or modification of the cohesin subunit Scc1 (rad21 in S.pombe) (Michaelis et al., 1997; Ciosk et al., 1998; Uhlmann et al., 1999; see Yanagida, 2000 for a review). APCHct1/Cdh1 (APCste9/srw1 in S.pombe) triggers mitotic exit by targeting mitotic cyclins for destruction (Schwab et al., 1997; Sigrist and Lehner, 1997; Visintin et al., 1997; Yamaguchi et al, 1997; Kitamura et al., 1998; Kramer et al., 1998). Hct1 interaction with APC/C is negatively regulated by Cdk1–cyclin-dependent phosphorylation (Zacchariae et al., 1998; Jaspersen et al., 1999; Kramer et al., 2000) and activated by Cdc14 protein phosphatase (Visintin et al., 1998, 1999; Jaspersen et al., 1999; Shou et al., 1999). APCCdc20 activation controls not only Pds1 degradation but also that of Clb5 and Clb2 (Shirayama et al., 1999; Baümer et al., 2000; Yeong et al., 2000). Recently it has been proposed that degradation of the mitotic cyclin Clb2 occurs in two steps. First, a fraction of Clb2 is degraded by APCCdc20 at the metaphase–anaphase transition and later, in telophase, APCHct1/Cdh1 degrades the rest of Clb2 (Baümer et al., 2000; Yeong et al., 2000). Fission yeast ste9/srw1 (ste9 from here on) has been proposed to be involved in the degradation of cdc13 B-type cyclin (Yamaguchi et al., 1997; Kitamura et al., 1998). Here we show that the main role of APCste9 is to target the mitotic cyclins cdc13 and cig1 for degradation in G1. ste9 interacts with APC/C only in G1. Cdk-dependent phosphorylation of ste9 in S-phase and G2 has a dual effect: it triggers the proteolysis of ste9 and also causes its dissociation from APC/C. Results cdc13 and cig1 cyclins are targets for APCste9 Previous studies have shown that overexpression of ste9 promotes cdc13 degradation and induces multiple rounds of S-phase in the absence of mitosis (Yamaguchi et al., 1997; Kitamura et al., 1998), equivalent to deletion of the cdc13+ gene (Hayles et al., 1994). To test whether ste9 induces degradation of other B-type cyclins in addition to cdc13, we overexpressed the ste9+ gene from the nmt1 promoter and measured the levels of cig1, cig2 and cdc13 cyclins in S.pombe. As shown in Figure 1A, ste9 overexpression promotes proteolysis of cdc13 and cig1 but not of cig2. In fact, cig2 levels increased in these cells in agreement with previous observations showing that down-regulation of cdc2/cdc13 allows accumulation of cig2 (Jallepalli and Kelly, 1996; C.Martín-Castellanos and S.Moreno, unpublished data). This result provides an explanation for the endoreplication phenotype induced by overproduction of the ste9+ gene (Yamaguchi et al., 1997; Kitamura et al., 1998; Figure 1B) as inactivation of cdc2/cdc13 will prevent mitosis while activation of cdc2/cig2 will trigger multiple rounds of S-phase. Figure 1.Overproduction of ste9+ promotes degradation of cdc13 and cig1 cyclins but not of cig2. An S.pombe leu1-32 strain was transformed with the plasmid pREP3X-ste9+. Transformants were selected on plates containing minimal medium with thiamine. Cells were grown in the presence (+T, repressed conditions) or absence of thiamine (−T, derepressed conditions) and samples were taken at the indicated times. (A) Extracts were prepared from these samples and the amounts of cdc13, cig1, cig2 and α-tubulin were determined. (B) FACS analysis of the cells. Download figure Download PowerPoint This experiment also suggests that cig1 cyclin is another target of APCste9. In contrast to cdc13 or cig2, analysis of the cig1 amino acid sequence reveals no obvious destruction box (Bueno et al., 1991) or KEN box (Pfleger and Kirschner, 2000). We examined cig1 protein levels during the mitotic cell cycle in cells synchronized by centrifugal elutriation and found that cig1 protein is destroyed in mitosis with similar kinetics to cdc13 (Figure 2A). For this experiment, we used the temperature-sensitive wee1-50 mutant where the G1 phase of the cell cycle is extended when incubated at the restrictive temperature of 36°C (Nurse, 1975). To confirm this result, we performed an additional experiment using the cold-sensitive nda3-KM311 β-tubulin mutants that arrest the cell cycle in metaphase. nda3-KM311 cells were pre-synchronized in early G2 by centrifugal elutriation. The resulting culture was incubated at the restrictive temperature of 20°C for 4 h and then released to 32°C. Samples were taken at 0, 2 and 4 h during the block and at 5, 15, 30 and 45 min during the release to measure cdc13, cig1 and rum1 protein levels. As shown in Figure 2C, cig1 was completely degraded during mitosis. Degradation of cig1 occurred slightly earlier than that of cdc13. Interestingly, in metaphase cells, cig1 levels started to decrease while cdc13 levels were still high (Figure 2C, t = 4 h); at this time point, the rum1 cdk inhibitor (a target for cdc2/cig1 phosphorylation) is already present at high levels (Figure 2C). This is consistent with previous observations indicating that cdc2/cig1 promotes phosphorylation and degradation of rum1 (Correa-Bordes et al., 1997; Benito et al., 1998). Taken together, these results indicate that cig1 cyclin is destroyed in mitosis and that cig1 may be an additional target of APCste9. Figure 2.Cig1 oscillates during the mitotic cell cycle. (A) Cig1 protein levels were measured in a synchronous culture of the temperature-sensitive wee1-50 strain. A homogeneous population of cells in early G2 was selected by centrifugal elutriation of the wee1-50 strain at 25°C. This culture was incubated for 20 min at 25°C and then shifted up to 36°C. Samples were taken every 20 min to determine cig1, cdc13 and cdc2 protein levels. Both cig1 and cdc13 protein levels decreased in anaphase and increased at the end of G1. A cig1 deletion (Δ) cell extract was used as a negative control. (B) Percentage of G1 cells and mitotic index of the synchronous culture. (C) Cig1 is degraded during mitosis. Early G2 cells of nda3-KM311 were isolated by centrifugal elutriation at 32°C and blocked in metaphase for 4 h at 20°C. The culture was then released at 32°C. Samples were taken at the indicated times to determine cig1, cdc13, rum1 and α-tubulin protein levels. (D) Percentage of cells in interphase, metaphase and anaphase determined by DAPI staining. Download figure Download PowerPoint APCste9 is required for the degradation of mitotic cyclins in G1 Next, we wanted to test if ste9 was required in every cell cycle to degrade cdc13 and cig1 in mitosis, in G1 or both. In yeast, inactivation of cdc2–cyclin complexes in mitosis and G1 is thought to depend on two mechanisms: cyclin proteolysis by APC/C and cdk inhibition. Budding yeast cells lacking Hct1 and Sic1 are not viable, presumably because Cdc28/Clb2 cannot be down-regulated at the end of mitosis (Schwab et al., 1997). In contrast, fission yeast cells deleted for ste9 and rum1 are viable. We have found that the double mutant ste9Δ rum1Δ is wild-type in size and, like the single mutants ste9Δ and rum1Δ, is unable to arrest the cell cycle in G1 in response to nitrogen starvation and sterile (data not shown). We compared wild-type, ste9 Δ, rum1 Δ and ste9 Δ rum1 Δ cells synchronized in G2 using a cdc25-22 mutant and determined levels of cdc13 and cig1 protein as these cells proceeded through mitosis. As shown in Figure 3, degradation of cdc13 and cig1 took place with similar kinetics in wild-type, in the single mutants ste9 Δ and rum1 Δ and in the double mutant ste9 Δ rum1 Δ, suggesting that degradation of cdc13 and cig1 at the metaphase–anaphase transition can occur through a ste9-independent mechanism. However, ste9 is absolutely required for the degradation of cdc13 and cig1 in cells arrested in G1 using the cdc10-129 mutant (Kitamura et al., 1998; see Figure 7A, lanes 2–3 and 5–6). These experiments indicate that ste9 is not the only APC/C activator necessary for the proteolysis of mitotic cyclins as cells exit mitosis but it plays a fundamental role in G1. Perhaps APCslp1 contributes to the degradation of fission yeast mitotic cyclins during anaphase like in Saccharomyces cerevisiae, where APCCdc20 can also target Clb2 for degradation (Baümer et al., 2000; Yeong et al., 2000). We propose that the main role of APCste9 in fission yeast is to target mitotic cyclins for degradation in G1. Figure 3.Degradation of cdc13 and cig1 in mitosis does not require ste9 or rum1. Wild-type, ste9 Δ, rum1 Δ and ste9 Δ rum1 Δ cells were synchronized in G2 using the cdc25-22 mutant. After 4 h at 36°C, the cultures were release to 25°C and samples were taken to measure cig1, cdc13 and α-tubulin levels. Download figure Download PowerPoint Figure 4.ste9 is phosphorylated in S-phase and G2 but not in G1. (A) ste9, cdc13 and cig1 protein levels in cells arrested in G1 with the cdc10-129 mutant, in S-phase with hydroxyurea and in G2 with the cdc25-22 mutant. As a negative control, we used an extract prepared from the ste9 Δ mutant and α-tubulin as loading control. (B) A His6-ste9 allele introduced by gene replacement into the ste9 locus was purified on an Ni2+-NTA column from cdc10-129 cells growing at 25°C or after 4 h at 36°C. The purified His6-ste9 was treated with calf intestine alkaline phosphatase (CIAP) in the absence (−) or presence (+) of phosphatase inhibitors (Inh). (C) ste9 phosphorylation occurs at the G1–S transition. A cdc10-129 culture grown at 25°C to mid-exponential phase in minimal medium was shifted to 36°C for 4 h and then released at 25°C. Samples for western blot and flow cytometry were taken before the shift to 36°C (Asn), 4 h after the shift to 36°C (t = 0 min) and every 15 min after the release to 25°C. The levels of ste9, cdc13 and α-tubulin were determined. ste9 was unphosphorylated at the block point (t = 0 min) and became phosphorylated at the onset of S-phase (t = 60 min). Cdc13 protein levels were low in G1 cells while ste9 was unphosphorylated and started to accumulate when ste9 became phosphorylated and the cells initiated S-phase. Δ is a negative control from ste9 Δ and Asn is an extract from the asynchronous culture of cdc10-129 at 25°C. (D) Percentage of cells in G1 during the course of the experiment. Download figure Download PowerPoint ste9 is regulated by protein phosphorylation We raised specific antibodies to ste9 and measured ste9 protein levels in extracts of cells blocked at different points during the mitotic cell cycle. Extracts were made from cells blocked in G1 using the cdc10-129 mutant, in S-phase with the DNA synthesis inhibitor hydroxyurea (HU) and in G2 using the cdc25-22 mutant. We observed that ste9 migrated more rapidly in protein extracts from cells blocked in G1 (cdc10-129 at 36°C) than in S-phase (wt + HU) or G2 (cdc25-22 at 36°C) (Figure 4A). Treatment with alkaline phosphatase converted the upper bands into the lower band, indicating that ste9 is phosphorylated in vivo (Figure 4B). Phosphorylation of ste9 was then analysed in cells synchronized in G1 by blocking the cdc10-129 mutant during 4 h at 36°C and then releasing these cells to 25°C. ste9 was completely dephosphorylated in G1 (Figure 4C, t = 0) and became phosphorylated ∼60 min after the release as cells were undergoing S-phase (Figure 4C and D). Cdc13 cyclin, one of the targets of APCste9, began to accumulate in these cells at the time when ste9 became phosphorylated. These experiments suggest that, similarly to Hct1/Cdh1 in S.cerevisiae, ste9 is negatively regulated during S-phase and G2 by phosphorylation. Figure 5.ste9 phosphorylation depends on cdc2 function. Wild-type and cdc2-33 cells were nitrogen starved for 12 h at 25°C. NH4Cl was added to the culture and half of the cells were incubated at 25°C and the rest at 36°C. Samples were taken for flow cytometry (A) and for western blots (B) at 0, 2, 4 and 6 h after the addition of NH4Cl to determine ste9 and α-tubulin protein levels. Asn corresponds to a sample taken from the asynchronous culture before nitrogen starvation. (C) ste9 mobility in cdc25-22 and cdc2-33 extracts prepared from cells growing exponentially at 25°C (t = 0) and then shifted to 36°C for 4 h (t = 4). The levels of cdc13, cdc2 and α-tubulin in these extracts are also shown. As a control, we used an extract from cdc10-129 cells incubated at 36°C for 4 h. Download figure Download PowerPoint Cdc2–cyclin phosphorylates ste9 at multiple sites ste9 phosphorylation occurs in S-phase and G2 when the cdc2–cyclin kinase activity is high. In order to study whether ste9 phosphorylation is dependent on cdc2–cyclin activity, we synchronized the wild-type and the cdc2-33 mutant in G1 by nitrogen starvation at 25°C for 12 h. Nitrogen was then added back and the cultures were divided into two. Half of the cells were incubated at the permissive temperature (25°C) and the other half at the restrictive temperature (36°C) for the cdc2ts mutant. Wild-type and cdc2-33 cells incubated at 25°C underwent S-phase ∼4 h after the addition of nitrogen (Figure 5A). At 36°C, wild-type cells completed S-phase after 4 h whereas the cdc2-33 mutant remained in G1 for up to 6 h (Figure 5A). As shown in Figure 5B, ste9 phosphorylation in the cdc2-33 mutant was detected after 4 h at 25°C but not at 36°C, suggesting that phosphorylation of ste9 is dependent on the activation of cdc2–cyclin kinase at G1/S. We also observed that ste9 protein levels decreased at least 4-fold as it became phosphorylated (Figure 5B); expression of the ste9+ gene is constant under this experimental condition (data not shown), suggesting that phosphorylation of ste9 reduces its stability. We consistently found a reduction in ste9 protein levels in G2 cells compared with cells in G1 (see also Figure 4A, compare levels in cdc25ts with levels in cdc10ts). Figure 6.Expression of ste9 phosphorylation mutants induces diploidization. Three mutants, ste9-4A, ste9-10A and ste9-13A, containing four, 10 and the 13 putative cdk phosphorylation sites were mutated to alanine by site-directed in vitro mutagenesis. These mutant alleles were introduced into the fission yeast genome by gene replacement. (A) Schematic representation of the ste9 protein with the seven WD repeats and the position of the 13 putative cdk phosphorylation sites. (B) FACS profile of cells replaced with the different ste9 mutant alleles. (C) Electrophoretic mobility of the different ste9 alleles. The asterisk corresponds to a non-specific band recognized by the anti-ste9 antibody that it is also detected in ste9 Δ. (D) Half-lives of ste9, ste9-4A, ste9-10A and ste9-13A in exponentially growing cultures after the addition of 100 μg/ml of cycloheximide. The asterisk corresponds to a non-specific band recognized by the anti-ste9 antibody that it is also detected in ste9 Δ and serves as loading control. Download figure Download PowerPoint To confirm that phosphorylation of ste9 depends on cdc2 kinase activity, we performed an additional experiment. cdc2-33 and cdc25-22 mutant cells growing at 25°C were shifted to 36°C and samples were taken at 0, 2 and 4 h after the shift. ste9 became dephosphorylated in cdc2-33 cells incubated at the restrictive temperature but not in cdc25-22 cells (Figure 5C), indicating that active cdc2 kinase is needed to maintain ste9 in its phosphorylated form. Examination of the ste9 amino acid sequence revealed the presence of four putative cdc2 phosphorylation sites (S62, T98, T177 and S214) with the consensus S/T-P-X-K/R (where X represents any amino acid). There are nine additional sites (S130, T134, T143, T159, T174, S187, S425, S513 and S547) with the sequence S/T-P of which six are located at the N-terminus and three at the C-terminus of the protein within the seven WD repeats (Figure 6A). We generated three mutant alleles of ste9+ by site-directed in vitro mutagenesis where these putative phosphorylation sites were mutated to alanine. ste9-4A contained the four putative phosphorylation sites with the strict consensus sequence, ste9-10A contained in addition the six S/T-P sites at the N-terminus and ste9-13A has all the sites mutated to alanine. The three mutant alleles were introduced into the ste9+ locus by gene replacement. Expression of these mutant forms of ste9 were able to rescue fully the sterility defect of ste9-deleted cells (data not shown). Cells expressing ste9-10A and ste9-13A were elongated compared with wild-type cells and, when they were streaked onto YES plates containing phloxin B, many dark red colonies were observed, suggesting that these cells were undergoing diploidization at high frequency. To confirm this observation, we measured the DNA content of these cells by flow cytometry and found that 8% of the cells were diploids in ste9-4A, 18% in ste9-10A and 38% in ste9-13A (Figure 6B). There was a direct correlation between the number of phosphorylatable residues mutated to alanine and the ability of the ste9 mutants to induce diploidization. This result indicates that cdk phosphorylation of ste9 is important to down-regulate ste9 in S-phase and G2. If ste9 is not phosphorylated in G2, it can promote cdc13 degradation and, as a consequence, the cells endoreduplicate their DNA. We have also found that the electrophoretic mobility of ste9-10A and ste9-13A was similar to that of unphosphorylated ste9 from cells arrested in G1 with cdc10-129 at 36°C (Figure 6C). Thus, ste9 is phosphorylated at multiple sites in vivo and this phosphorylation results in its inactivation. Figure 7.ste9 associates with APC/C only in G1. (A) Extracts from cdc10-129, cdc10-129 cut9·HA ste9 Δ and cdc10-129 cut9·HA mutants grown at 25 or 36°C were immunoprecipitated with anti-HA antibodies (IP α-HA) and then western blotted with anti-HA or anti-ste9 antibodies. Total cell extracts were separated on an SDS–polyacrylamide gel, transferred to a nitrocellulose membrane and probed with anti-HA, anti-ste9, anti-cdc13, anti-cig1 and anti-α-tubulin antibodies. (B) ste9 phosphorylation mutants associate with APC/C in G2. Extracts from cdc25-22, cdc25-22 cut9·HA and cdc10-129 cut9·HA mutants grown at 36°C for 4 h were immunoprecipitated with anti-HA antibodies (IP α-HA) and then western blotted with anti-HA and anti-ste9 antibodies. Total cell extracts were separated on an SDS–polyacrylamide gel, transferred to a nitrocellulose membrane and probed with anti-HA, anti-ste9 and anti-α-tubulin antibodies. Download figure Download PowerPoint As shown in Figure 5B, phosphorylation of ste9 by cdc2–cyclin complexes correlates with a significant decrease in ste9 protein levels. To investigate the possibility that cdc2 phosphorylation regulates ste9 stability, we compared the half-life of wild-type ste9 with that of ste9-4A, ste9-10A and ste9-13A mutant proteins. Cultures of wild-type and the three mutant strains generated by gene replacement with the ste9 mutant alleles were grown to mid-exponential phase and the half-lives of ste9, ste9-4A, ste9-10A and ste9-13A were measured after the addition of the protein synthesis inhibitor cycloheximide. Figure 6D shows that ste9 is short-lived (half-life 120 min) (Figure 6D). This result confirms that ste9 protein levels are down-regulated in vivo by cdc2-dependent phosphorylation. ste9 phosphorylation prevents its interaction with APC/C Previous reports have shown that cdk phosphorylation of Hct1/Cdh1 in budding yeast and animal cells prevents its association with APC/C (Zacchariae et al., 1998; Jaspersen et al., 1999; Lukas et al., 1999; Kramer et al., 2000). To test whether ste9 associates with APC/C, we used a strain where one of the APC/C subunits cut9 was epitope tagged with three copies of haemagglutinin (HA) (Berry et al., 1999). ste9 co-precipited with cut9-HA in extracts prepared from cells arrested in G1 with the cdc10-129 mutant (Figure 7A and B, lanes 6). In contrast, ste9 was not associated with APC/C in extracts from cells arrested in G2 with the cdc25-22 mutant (Figure 7B, lane 3). Thus ste9 interacts with APC/C only in G1 when it is not phosphorylated and it does not interact with APC/C in G2 when it becomes phosphorylated. We then analysed whether the mutant proteins ste9-10A and ste9-13A were able to interact with APC/C in G2 as this may be the reason for the diploidization phenotype observed in these mutants. To test this, we prepared extracts from cells arrested in G2 with the cdc25-22 mutant and then immunoprecipitated APC/C using anti-HA antibodies. ste9 co-precipitated with APC/C in extracts expressing ste9-10A and ste9-13A but not in extracts expressing wild-type ste9+ (Figure 7B, lanes 3–5), suggesting that phosphorylation of ste9 causes its dissociation from APC/C in G2. Discussion In this study, we provide biochemical evidence for a role for ste9 as a negative regulator of cell cycle progression in G1. ste9 is a member of a highly conserved family of proteins containing seven WD repeat domains of which the prototypes are Hct1/Cdh1 of budding yeast and Fizzy-related of higher eukaryotes (Schwab et al., 1997; Sigrist and Lehner, 1997; Visintin et al., 1997; Kramer et al., 1998). These proteins function as activators of APC/C to promote polyubiquitylation and degradation of mitotic cyclins in mitosis and G1. Here we show that in fission yeast: (i) APCste9 promotes degradation of the mitotic cyclins cdc13 and cig1 but not of the S-phase cyclin cig2. The fact that cig2 levels are high in cells overexpressing ste9 provides an explanation for the re-replication phenotype associated with these cells. (ii) APCste9 is not necessary for the proteolysis of mitotic cyclins at the end of mitosis because in cells lacking ste9, degradation of cdc13 and cig1 still occurs. However, ste9 is absolutely required to degrade mitotic cyclins completely when cells need to delay or to stop the cell cycle in G1. This is important for small cells that had to lengthen the G1-phase until they reach the minimum cell size required to initiate DNA replication or to prevent entry into mitosis from G1. (iii) ste9 is phosphorylated by the cdc2 kinase in vivo at multiple sites. Cdk phosphorylation of ste9 in G2 has two effects. First, to promote the degradation of ste9 and secondly, to prevent ste9 association with APC/C. APC/C–ste9 interaction occurs only in G1 when ste9 is in its dephosphorylated form. In S-phase and G2, ste9 becomes phosphorylated and it does not interact with APC/C. Two APC complexes are involved in the degradation of mitotic cyclins We have found that cdc13 and cig1 are targets for APCste9 in G1. Two observations support this idea: first cdc13 and cig1 protein levels decrease when ste9 is overproduced. In addition, cig1 and cdc13 are not degraded when a cdc10-129 ste9 Δ mutant is incubated at the restrictive temperature. The fact that both cyclins are destroyed during mitosis in cells lacking ste9 suggests that another APC complex is responsible for the degradation of cig1 and cdc13 at the metaphase–anaphase transition. Perhaps slp1, the fission yeast homologue of Cdc20, in association with APC/C may trigger the degradation of cig1 and cdc13 as cells exit mitosis. This is reminiscent of the situation in budding yeast where proteolysis of Clb2 by the APC/C occurs in two stages. First, a fraction of Clb2 is destroyed during anaphase by APCCdc20 and the rest is degraded at the end of mitosis by APCHct1 (Baümer et al., 2000; Yeong et al., 2000). It is interesting to note that cig1 does not have a clear destruction box sequence as is the case with cdc13, nor does it have a KEN box (Pfleger and Kirschner,
Referência(s)