QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

Vol. 19, Issue 9, 4006-4018, September 2008

This Article Abstract Full Text (PDF) Supplemental Materials All Versions of this Article: E08-03-0286v1 E08-03-0286v2 19/9/4006    most recent Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Google Scholar Articles by Keaton, M. A. Articles by Lew, D. J. PubMed PubMed Citation Articles by Keaton, M. A. Articles by Lew, D. J.

Nucleocytoplasmic Trafficking of G2/M Regulators in Yeast

Mignon A. Keaton^*, Lee Szkotnicki, Aron R. Marquitz, Jake Harrison, Trevin R. Zyla, and Daniel J. Lew

Department of Pharmacology and Cancer Biology, Duke University Medical Center, Durham, NC 27710

Submitted March 17, 2008; Revised June 5, 2008; Accepted June 9, 2008
Monitoring Editor: Mark J. Solomon

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Nucleocytoplasmic shuttling is prevalent among many cell cycle regulators controlling the G2/M transition. Shuttling of cyclin/cyclin-dependent kinase (CDK) complexes is thought to provide access to substrates stably located in either compartment. Because cyclin/CDK shuttles between cellular compartments, an upstream regulator that is fixed in one compartment could in principle affect the entire cyclin/CDK pool. Alternatively, the regulators themselves may need to shuttle to effectively regulate their moving target. Here, we identify localization motifs in the budding yeast Swe1p (Wee1) and Mih1p (Cdc25) cell cycle regulators. Replacement of endogenous Swe1p or Mih1p with mutants impaired in nuclear import or export revealed that the nuclear pools of Swe1p and Mih1p were more effective in CDK regulation than were the cytoplasmic pools. Nevertheless, shuttling of cyclin/CDK complexes was sufficiently rapid to coordinate nuclear and cytoplasmic events even when Swe1p or Mih1p were restricted to one compartment. Additionally, we found that Swe1p nuclear export was important for its degradation. Because Swe1p degradation is regulated by cytoskeletal stress, shuttling of Swe1p between nucleus and cytoplasm serves to couple cytoplasmic stress to nuclear cyclin/CDK inhibition.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Nucleocytoplasmic shuttling of regulatory proteins has been exploited in several ways by eukaryotic cells. Shuttling can allow access of a single regulator to both nuclear and cytoplasmic compartments, as exemplified by cyclin B/cyclin-dependent kinase Cdk1 complexes that need to phosphorylate both cytoplasmic targets (to promote centrosome separation) and nuclear targets (to promote nuclear envelope breakdown) as cells enter mitosis (Pines, 1999; Takizawa and Morgan, 2000). Shuttling can also provide a means to control a regulator's activity, as exemplified by transcription factors that can only access their target DNA in the nucleus and therefore cannot function when located in the cytoplasm. Furthermore, shuttling can allow communication between the two compartments. This is exemplified by stress-responsive mitogen-activated protein kinases (MAPKs) that are activated by cytoplasmic MAPK kinases and translocate to the nucleus where they phosphorylate transcription factors (Ferrigno and Silver, 1999). In all of these instances, controlling the subcellular localization of the regulatory protein serves to either promote or prevent its interaction with upstream regulators and/or downstream targets that are themselves constrained to reside in one compartment.

In this paper, we address the role(s) of nucleocytoplasmic shuttling of G2/M regulators in the model eukaryote Saccharomyces cerevisiae. Previous work indicated that the major mitotic cyclin Clb2p shuttles continuously in and out of the nucleus, with a steady-state localization pattern that is predominantly nuclear (Hood et al., 2001). Analysis of mutations in the nuclear import and export sequences in Clb2p suggested that nuclear Clb2p promotes anaphase and that cytoplasmic Clb2p promotes actin depolarization, which leads to an "apical-isotropic switch" in bud growth (Bailly et al., 2003; Eluere et al., 2007; Hood-DeGrenier et al., 2007). The activity of Clb2p/Cdc28p (cyclin B/Cdk1) is regulated by inhibitory tyrosine phosphorylation catalyzed by Swe1p (the sole Wee1 homologue) and reversed by Mih1p (the sole Cdc25 homologue). We show here that both Swe1p and Mih1p, like their target Clb2p/Cdc28p, shuttle in and out of the nucleus, and we address how this trafficking affects G2/M regulation.

An important facet of G2/M regulation is the presence of positive feedback loops that are thought to impart a concerted all-or-none character to the G2/M transition (O'Farrell, 2001). Studies in Xenopus extracts and mammalian cells established that Wee1 is inhibited and Cdc25 is activated by cyclin B/Cdk1 (Dunphy, 1994). In cell-free (largely cytoplasmic) extracts, these positive feedback loops result in a bistable system that can abruptly transition between G2 (low cyclin B/Cdk1 and Cdc25 activity, high Wee1 activity) and M (high cyclin B/Cdk1 and Cdc25 activity, low Wee1 activity) states (Xiong and Ferrell, 2003). It is unclear to what degree similar feedback loops might act in S. cerevisiae, but there is evidence for complex phosphoregulation of both Swe1p and Mih1p that is directly or indirectly influenced by Cdc28p (Harvey et al., 2005; Pal et al., 2008). Moreover, Clb2p/Cdc28p activity promotes degradation of its inhibitor, Swe1p (Sia et al., 1998; Asano et al., 2005), yielding another potential feedback pathway that is likely conserved in vertebrate cells (Watanabe et al., 2004).

Because Swe1p and Mih1p are chasing (and being chased by) a moving target (Clb2p/Cdc28p), it is difficult to predict a priori how localization of a specific regulator might impact G2/M control. At one extreme, it could be that shuttling of the regulators effectively links nucleus and cytoplasm, yielding a single G2/M switch coordinated between the two compartments. At the other extreme, each compartment may remain insulated from the other compartment, potentially allowing independent regulation of nuclear and cytoplasmic events. Distinguishing between these and other possibilities requires an examination of the effects of trapping the regulators in one or other compartment.

By identifying and mutating motifs important for nucleocytoplasmic shuttling of Swe1p and Mih1p, we have addressed how trafficking of these regulators impacts G2/M progression, both in unperturbed cells and in cells exposed to conditions that trigger G2 checkpoints. The results suggest that even when trafficking of Swe1p or Mih1p is disabled, there is sufficient flux of Clb2p/Cdc28p between cytoplasm and nucleus to effectively link the two compartments. At the same time, we find that both Swe1p and Mih1p efficacy is significantly enhanced by nuclear localization. Finally, we show that the ability of Swe1p to respond to stress is dependent on its shuttling through the cytoplasm. Thus, shuttling of Swe1p serves to link stress sensing in the cytoplasm to effective cell cycle arrest.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Yeast Strains and Plasmids
All yeast strains used in this study are in the BF264-15DU background (ade1, his2, leu2-3112, trp1-1^a, and ura3ns) (Richardson et al., 1989) and are listed in Table 1. Standard genetic and molecular biology techniques were used for generation of all strains described here. All oligonucleotides used in this study are listed in Table 2. The swe1::LEU2 (Booher et al., 1993), hsl1::URA3 (Ma et al., 1996), and mih1::LEU2 (Russell et al., 1989) disruptions have been described previously. The hsl1::URA3 allele was achieved by homologous recombination using a URA3 PCR product (Baudin et al., 1993) and the primers OJ105 and OJ106. The SWE1-12myc:TRP1 (McMillan et al., 2002), SWE1-NES A-12myc:URA3, SWE1-NES I-12myc:URA3 (McMillan et al., 2002), HSL7-HA:KAN^r (Longtine et al., 2000), cdc24-1 (Sia et al., 1996), and cdc12-6::LEU2 (Longtine et al., 2000) alleles have been described previously. The cdc15-2 allele (Surana et al., 1993) was introduced into BF264-15DU strain background by backcrossing five times. The GAL1-MIH1:TRP1 allele places the endogenous MIH1 under control of the GAL1 promoter, as described in McMillan et al. (1998).

For localization studies, BamHI/SacI fragments containing GAL1-SWE1^NES--12myc and GAL1-SWE1^NLS--12myc were subcloned into pRS306 (Sikorski and Hieter, 1989), yielding pDLB2438 and pDLB2439. These and the wild-type GAL1-SWE1-12myc plasmid pDLB955 (McMillan et al., 1998) were digested with StuI and transformed into DLY1 to target integration by homologous recombination at URA3.

The MIH1-myc allele was generated by a three-step process. First, a 767-base pair C-terminal MIH1 fragment was amplified by PCR using OJ65 and OJ66 and cloned into pRS306 using KpnI and XhoI sites at the 3' and 5' ends. Next, a 502-base pair fragment beginning 16 bases upstream of the MIH1 stop codon and extending 483 bases downstream of the stop codon was amplified by PCR using OJ64 and OJ71, cut with XhoI/SacI, and ligated into the XhoI site of the aforementioned plasmid. This creates plasmid pJM1029, an integrating plasmid containing the C terminus of MIH1 and genomic sequence downstream of MIH1 with an XhoI site just upstream of the MIH1 STOP codon in the pRS306 backbone. Next, a plasmid containing 12 tandem myc tags (pBS-12Xmyc, a gift from Paul Russell, Scripps Research Institute) was digested with XhoI and SalI, and the resulting 12 myc fragment was inserted into the XhoI site of pJM1029. The resulting plasmid pJM1034, which contains a C-terminal MIH1 fragment fused to 12 tandem myc tags, was partially digested with NdeI to target integration at the endogenous MIH1 locus to generate the MIH1-myc:URA3 allele.

Generation of the MIH1^NLS- allele was performed by mutation of MIH1 in pGEX-KG (pDLB439) by overlap PCR using the flanking oligos H114 and H115 and the mutagenic internal oligos H125 and H126. Mutant PCR products were digested with EcoRI and SacI and ligated into pDLB439 cut with EcoRI and SacI. To generate MIH1^NLS--myc, the promoter of MIH1 was amplified by PCR using oligos M13R and H113 and plasmid pDLB457 (a library clone of MIH1 in pRS316; Sikorski and Hieter, 1989) as template. This introduces a NcoI site at the start codon of MIH1 and XhoI and XbaI sites at the 5' and 3' ends of the PCR fragment, respectively. The resulting PCR product was then introduced into pRS305 (Sikorski and Hieter, 1989) using XhoI and XbaI sites. Next, an XbaI/SacI fragment from pJM1034 encoding the C terminus of Mih1p fused to 12 myc tags was introduced into the pRS305 MIH1 expression construct creating plasmid pDLB800. Wild-type or NLS-mutant MIH1 open reading frames were then introduced into pDLB800 using NcoI and XbaI sites to create full-length MIH1-myc integrating plasmids, which were digested with EcoRV to target integration to the LEU2 locus. Single integrants were confirmed by PCR with oligos OJ176 and OJ177, and equal protein levels were demonstrated by Western blotting analysis.

To append a strong NLS to Mih1p^NLS--myc, a cassette containing two copies of the simian virus 40 (SV40) T antigen NLS (Edgington and Futcher, 2001) was amplified by PCR introducing SalI and XhoI sites at the 5' and 3' ends. The SV40 NLS was then cloned into pJM1034 (containing a C-terminal fragment of Mih1p-myc) at the XhoI site just downstream of the 12 myc tags, yielding pDLB801. This plasmid was digested with AflII and transformed into a yeast strain to target integration at MIH1^NLS--myc:LEU2, generating MIH1^NLS--SV40-myc. All plasmids derived from PCR were confirmed by DNA sequencing.

Media, Growth Conditions, and Cell Synchrony
Strains were grown in YEP (1% yeast extract 2% Bacto Peptone, 0.012% adenine, and 0.01% uracil) media containing 2% dextrose (YEPD), sucrose (YEPS), or galactose (YEPG) at 30°C. GAL1-SWE1-12myc strains were grown in YEPS to 0.5 x 10⁷ cells/ml, and Swe1p expression was induced by the addition of 2% galactose. To deplete Mih1p, GAL1-MIH1 strains were grown to log phase in YEPG and then shifted to media containing dextrose (at 2.5 x 10⁵ cells/ml) for 12 h. For pheromone arrest–release experiments, log phase cells were incubated with 40 ng/ml -factor for 3 h at 30°C and then collected and resuspended in fresh growth media at 1 x 10⁷ cells/ml. For cdc24-1 checkpoint assays, log phase cells containing cdc24-1 were incubated with 40 ng/ml -factor for 3.5 h at 24°C and then collected and resuspended in fresh media at 37°C. Cells were fixed in 70% ethanol overnight by the addition of 450 µl of cells to 1050 µl of 100% ethanol, washed in phosphate-buffered saline (PBS) containing 0.2 µg/ml 4–6-diamidino-2-phenylindole (DAPI; Sigma-Aldrich, St. Louis, MO), and resuspended in mounting media (90% glycerol, 9.2 mM p-phenylenediamine [Sigma-Aldrich] in PBS). Budding and nuclear division were scored for each time point.

Immunofluorescence and Microscopy
For immunofluorescence, cells were fixed in 4.5% formaldehyde for 1 h at room temperature, washed twice in solution B (0.1 M KPO₄ pH 7.5, and 1.2 M sorbitol), and digested with lyticase (Sigma-Aldrich). Cells were then washed twice in PBS and adhered to 0.1% polyethylenimine (Sigma-Aldrich)-coated slides. Cells were incubated with primary (1:200 mouse anti-myc (9E10) or 1:200 rat anti-tubulin (YOL 1134); Santa Cruz Biotechnology, Santa Cruz, CA) and secondary antibody (1:200 goat anti-mouse Alexa-568 or goat anti-rat Alexa-568; Invitrogen, Carlsbad, CA) diluted in Tris-buffered saline containing 5% normal goat serum (Invitrogen) and 0.1% Tween-20 (Bio-Rad) for 1 h each. Slides were washed extensively with the same buffer between incubations. Nuclei were visualized by staining with 0.5 µg/ml DAPI (Sigma-Aldrich) in the mounting media. For differential interference contrast (DIC) microscopy, cells were fixed in 70% ethanol overnight, washed in PBS containing 0.2 µg/ml DAPI, sonicated, and spotted onto a 2% agarose pad. DIC and fluorescent images were acquired with a 100x objective on an Axioimager microscope (Carl Zeiss, Thornwood, NY) with an Orca ER monochrome cooled charge-coupled device camera (Hamamatsu, Bridgewater, NJ) and processed using MetaMorph (Molecular Devices, Sunnyvale, CA) and Photoshop (Adobe Systems, Mountain View, CA) software.

For experiments measuring bud length, cdc12-6 cells were grown at the permissive temperature of 24°C to 0.5 x 10⁷ cells/ml and stained with 250 µg/ml concanavalin A (ConA) Alexa Flour-488 (green; Invitrogen) in PBS for 10 min at room temperature. The cells were then washed extensively with PBS, resuspended in fresh media prewarmed to 37°C, and incubated at the nonpermissive temperature of 37°C. After 2 h, cells were stained with ConA Alexa Flour-594 (red; Invitrogen) in the same manner. Cells were then maintained at 4°C in H₂O until DIC and fluorescent imaging was performed on a 2% agarose pad. Bud length measurements were obtained using fluorescent images and NIH ImageJ software (National Institutes of Health, Bethesda, MD).

Western Blot Analysis
For Western blot analysis, cells were grown to mid-log phase, and cell lysates were generated by trichloroacetic acid (TCA) precipitation. Approximately 1 x 10⁷ cells were collected and resuspended in 225 µl of pronase buffer (25 mM Tris-HCl, pH 7.5, 1.4 M sorbitol, 20 mM NaN₃, and 2 mM MgCl₂). TCA (100%, wt/vol; Sigma-Aldrich) was added to a final concentration of 17% vol/vol, and cell suspensions were stored at –80°C. Samples were thawed, and cells were homogenized by addition of glass beads (Sigma-Aldrich) and vortexing at 4°C for 10 min. The lysate was collected, and the beads were washed two times with 5% (wt/vol) TCA to recover the remaining lysate. Precipitated proteins were collected by centrifugation in a Biofuge microcentrifuge for 15 min at 4°C. Pellets were then resuspended in Thorner sample buffer (40 mM Tris-HCl, pH 6.8, 8 M urea, 5% SDS, 143 mM β-mercaptoethanol, 0.1 mM EDTA, and 0.4 mg/ml bromphenol blue), and any residual TCA was neutralized by the addition of 2 M Tris-HCl, pH 8.0.

For immunoblotting, TCA precipitates resuspended in Thorner sample buffer were heated at 42°C for 5 min before polyacrylamide gel electrophoresis. Proteins were then transferred to nitrocellulose (Pall. East Hills, NY), blocked in 2% milk in PBS, and incubated in primary antibodies diluted in blocking buffer containing 0.1% Tween 20 overnight. Antibodies against Cdc11p (1:20,000 anti-Cdc11; Santa Cruz Biotechnology) and myc epitope (1:2000 mouse anti-myc [9E10]; Santa Cruz Biotechnology) were used at the indicated dilutions. After several washes in PBS with 0.1% Tween 20, blots were incubated with fluorescent goat anti-mouse IRdye800 and goat anti-rabbit Alexa Flour-680 secondary antibodies (1:7500; Invitrogen) in blocking buffer with 0.1% Tween 20 and 0.01% SDS for 1 h at room temperature. After several subsequent washes, membranes were scanned using an Odyssey scanner and integrated fluorescence was quantitated using Odyssey software (Li-Cor, Lincoln, NE).

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Swe1p Shuttles between the Cytoplasm and Nucleus
Swe1p is synthesized in late G1/early S phase and accumulates in the nucleus (Sia et al., 1998; Longtine et al., 2000). Following bud formation, a pool of Swe1p can also be found in the cytoplasm, concentrated at the mother-bud neck (Longtine et al., 2000). The majority of the Swe1p is then degraded during G2/M, which results in depletion of the nuclear pool (Sia et al., 1998; Longtine et al., 2000). By early anaphase, only neck-localized Swe1p can be detected, and this cytoplasmic pool disappears during anaphase/telophase. We previously suggested that this localization pattern results from export of nuclear Swe1p, followed by neck targeting and subsequent degradation (Longtine et al., 2000). The same pattern could arise, however, if the initial pool of Swe1p that accumulates in the nucleus during G1/S is also degraded in the nucleus during G2/M, whereas a later synthesized pool of Swe1p is targeted to the neck and then degraded during anaphase. To distinguish between these possibilities, we examined the localization of Swe1p that was synthesized in a pulse during early G1. Cells containing myc-tagged Swe1p expressed from the tightly regulatable GAL1 promoter were arrested in G1 by treatment with pheromone for 2 h and induced to express Swe1p by addition of galactose for 1 h in the continued presence of pheromone. The cells were then transferred to dextrose-containing media to shut off Swe1p expression and released from the G1 arrest. As shown in Figure 1, A and B, Swe1p-myc synthesized during the G1 arrest initially accumulated in the nucleus, but a neck-localized pool became apparent after bud emergence 60 min after Swe1p shut-off. This indicates that neck-localized Swe1p is recruited from the pre-existing (largely nuclear) pool of Swe1p rather than a later synthesized pool. Subsequent disappearance of early G1-synthesized Swe1p from the nucleus followed the same pattern (disappearance from the nucleus and then the neck) as that documented for endogenous Swe1p (Figure 1, A and B). We conclude that Swe1p shuttles between the nucleus and cytoplasm and that time of synthesis does not dictate localization.

View larger version (47K): [in this window] [in a new window]   Figure 1. Identification of Swe1p localization determinants. (A) Localization of Swe1p made in a pulse during G1 arrest. GAL1-SWE1-myc cells (RSY 136) were arrested with -factor for 3 h in YEPS. Swe1p-myc expression was induced for the last hour of the arrest by the addition of 2% galactose. The cells were harvested by centrifugation and resuspended in media containing 2% dextrose without -factor to repress Swe1p-myc expression and to allow cell cycle progression. At the indicated times (minutes), cells were fixed, and Swe1p-myc localization was assessed by immunofluorescence microscopy. The bud neck is indicated by arrowheads. (B) Samples from A were scored for percentage of cells with buds (circles), nuclear Swe1p (squares), and neck Swe1p (triangles). (C) Schematic of Swe1p and alignment of amino acid sequence from three Saccharomyces species (S. cerevisiae, S. castelli, and S. kluyeri). The kinase domain (green) and the conserved Box 1 (red; NES), Box 2 (blue; NLS), and Box 3 (black; Hsl7p binding) elements in the nonconserved regulatory domain are indicated. Identical residues among the three species are highlighted in yellow. Asterisks indicate residues mutated in the NES and NLS. (D) Immunofluorescence micrographs showing localization of Swe1p-myc (DLY 6941), Swe1pNLS--myc (DLY 6933), and Swe1pNES--myc (DLY 6932) expressed from the GAL1 promoter after 1.5 h of galactose induction. The bud neck is indicated by arrowheads.

Abundance of Swe1p^NES- and Swe1p^NLS- through the Cell Cycle
To assess how altering Swe1p localization motifs affects the abundance of Swe1p, we integrated a single copy of each mutant (as well as a wild-type Swe1p-myc control), expressed under control of the SWE1 promoter, into strains lacking endogenous Swe1p. To ensure that cell cycle progression was not affected by any potential differences in activity of the Swe1p variants, we also introduced GAL1-MIH1 so that overexpression of Mih1p in galactose media would rapidly reverse any Swe1p-catalyzed Cdc28p inhibition. Single-cycle synchrony experiments were performed by pheromone arrest, release, and rearrest with pheromone, all in galactose media. Like wild-type Swe1p, Swe1p localization-mutants accumulated to a peak level at 50–60 min (Figure 2A); however, there were differences in the degree of accumulation and in the subsequent degradation profiles for the different Swe1p mutants. Swe1p^NES- accumulated to lower peak levels (when normalized to the loading control) than wild-type Swe1p and then persisted throughout G2/M before gradually declining during the subsequent G1 (Figure 2A). Swe1p^NLS-, in contrast, accumulated to similar peak levels as wild-type Swe1p, but a fraction of Swe1p^NLS- persisted throughout G2/M and the subsequent G1 (Figure 2A). These differences do not reflect altered cell cycle profiles, because all strains exhibited comparable timing of budding and nuclear division (Figure 2B).

View larger version (36K): [in this window] [in a new window]   Figure 2. Abundance of Swe1p localization variants during the cell cycle. (A) Strains expressing GAL1-MIH1 were grown in YEPG at 30°C and synchronized by pheromone arrest and release. Pheromone was added back to the culture 60 min after release to rearrest the cells in G1 after one cycle. Swe1p-myc levels were assessed by Western blot analysis at the indicated times in the following strains: SWE1-myc (DLY7351), SWE1-myc hsl1 (DLY9859), SWE1NES--myc (DLY9681), SWE1NES--myc hsl1 (DLY9858), SWE1NLS--myc (DLY9697), and SWE1NLS--myc hsl1 (DLY9855). The G2/M interval during which the majority of wild-type Swe1p gets degraded is indicated by the red bracket. Blotting for the septin Cdc11p served as a loading control. (B) Synchrony data for the experiments in A: 100 cells were scored for budding and nuclear division (HSL1, open symbols; hsl1, filled symbols) at each time point. (C) Binding of Swe1p variants to Hsl7p. Expression of the indicated SWE1 alleles (under the control of the GAL1 promoter) was induced by addition of 2% galactose for 3 h in the following strains: empty vector (DLY6461), GAL1-SWE1 (DLY6462), GAL1-SWE1NES--myc (DLY6463), GAL1-SWE1NLS--myc (DLY6465), and GAL1-SWE11-myc (JMY1681). Cell lysates were prepared and incubated with beads bound to recombinant glutathione transferase (GST) or GST-Hsl7p. The beads were then washed and bound Swe1p was detected by Western blotting. (Deletion of a lane is indicated by a vertical line.) (D) Western blot analysis of Swe1p levels from asynchronously growing cells expressing GAL1-MIH1. The ratio of Swe1p abundance in hsl1 versus HSL1 cells was quantitated for each Swe1p variant as shown in the graph, right (mean ± SEM; n = 3). SWE1-myc (DLY7351), SWE1-myc hsl1 (DLY 7353), SWE1NES--myc (DLY9681), SWE1NES--myc hsl1 (DLY9685), SWE1NLS--myc (DLY9697), and SWE1NLS--myc hsl1 (DLY9699).

Figure 2A also suggested that Swe1p^NES- accumulates to lower peak levels than the other Swe1p variants and that Swe1p^NES- levels in asynchronous cells were comparable with those of wild-type Swe1p, even though Swe1p^NES- is largely resistant to Hsl1p-mediated degradation (Figure 2D). This surprising result may indicate the existence of an unsuspected degradation pathway targeting nuclear Swe1p during G1/S. It is also possible, however, that the NES mutations decrease mRNA stability or possibly affect the probability of correct Swe1p folding. A subset of newly synthesized Swe1p^NES- may misfold and be immediately degraded, whereas the remainder folds properly and is then stable.

An additional unexpected observation from these experiments was that Swe1p^NLS- is immune to the slow degradation that targets Swe1p and Swe1p^NES- during G1 arrest (Figure 2A). Although most Swe1p is degraded by the Hsl1p/Hsl7p pathway in G2/M, residual Swe1p that survives that destruction is subsequently degraded slowly in G1 (Sia et al., 1998). This is particularly evident in hsl1 strains where the G2/M degradation does not occur (Figure 2A). It has been suggested that this phase of Swe1p degradation is mediated by the anaphase-promoting complex (APC) (Thornton and Toczyski, 2003). Given that the APC is localized to the nucleus (Sikorski et al., 1993; Lim et al., 1998), the resistance of Swe1p^NLS- to this phase of degradation may stem from its location in the cytoplasm.

Swe1p^NES- and Swe1p^NLS- Mutants Retain the Ability to Inhibit Clb2p/Cdc28p
To assess whether the Swe1p localization mutants retained the ability to inhibit cell cycle progression, hsl1 GAL1-MIH1 strains containing each mutant were grown in galactose medium and then transferred to dextrose medium (which represses the GAL1 promoter) to deplete Mih1p. In cells containing wild-type Swe1p, Mih1p depletion leads to a lethal G2 arrest with elongated buds and high-stoichiometry Cdc28p phosphorylation (McMillan et al., 1999a). In this context, both Swe1p^NLS- and Swe1p^NES- promoted a similar arrest (Figure 3A), indicating that these mutants must retain robust kinase activity. In cells containing wild-type HSL1 and SWE1, depletion of Mih1p promotes a small G2 delay that makes the cells larger, but it does not slow proliferation (Russell et al., 1989) (Figure 3B). Cells containing Swe1p^NLS- and wild-type Swe1p behaved similarly in this background; however, the presence of Swe1p^NES- severely impaired proliferation and the cells became significantly elongated (Figure 3B). This finding is consistent with the failure of Swe1p^NES- to undergo Hsl1p-mediated degradation (Figure 2), and it supports the conclusion that Swe1p nuclear export is required for its destruction.

View larger version (44K): [in this window] [in a new window]   Figure 3. Swe1p localization mutants are functional in vivo. (A and B) GAL1-MIH1 strains containing the indicated SWE1 alleles were grown in YEPG (GAL) or shifted to YEPD (DEX) for 12 h to deplete Mih1p. Left, cell morphology (DIC). Right, microcolony morphology on YEPD plates. (A) SWE1-myc hsl1 (DLY7353), SWE1NLS--myc hsl1 (DLY9699), SWE1NES--myc hsl1 (DLY9685). (B) SWE1-myc (DLY7351), SWE1NLS--myc (DLY9697), SWE1NES--myc (DLY9681). (C) The following strains containing GAL1-SWE1-myc alleles were induced to express Swe1p by addition of 2% galactose for 3 h: GAL1-SWE1-myc (DLY6462), GAL1-SWE1NES--myc (DLY6463), and GAL1-SWE1NLS--myc (DLY6465). Cells were fixed, processed, and stained for Swe1p-myc (top) and DNA (bottom).

Cytoplasmic Swe1p Is Less Effective than Nuclear Swe1p in Blocking Nuclear Division under Conditions That Block Bud Formation
The major role of Swe1p in vegetative yeast cells is to enact a G2 delay in response to stresses that disorganize either the actin or septin cytoskeletons via a pathway called the morphogenesis checkpoint (Lew, 2003). During a physiological morphogenesis checkpoint response, Swe1p degradation is inhibited and Mih1p activity is reduced, leading to a delay in nuclear division (Lew, 2003). Because the nuclear pool of Clb2p/Cdc28p is primarily responsible for initiating nuclear division (Eluere et al., 2007), trapping Swe1p in the cytoplasm might be expected to impair its ability to delay this process. To assess whether Swe1p^NLS- and Swe1p^NES- promoted a normal checkpoint response, we used a cdc24-1 mutant in which polarization (and hence bud formation) is temperature sensitive (Hartwell, 1971). We previously showed that the timing of nuclear division upon shift of cdc24-1 cells to restrictive temperature was sensitive to SWE1 gene dosage, providing a quantitative assay for Swe1p function in vivo (McMillan et al., 1999a). In this context, we found that Swe1p^NLS- displayed a somewhat reduced delay in nuclear division, whereas Swe1p^NES- promoted a delay comparable with that of wild-type Swe1p (Figure 4A).

View larger version (27K): [in this window] [in a new window]   Figure 4. Swe1p localization and checkpoint delay of nuclear division. (A–C) Delay in nuclear division in strains that are unable to form buds. The indicated cdc24-1 strains were synchronized in G1 at 24°C, released from the pheromone arrest at 37°C, and scored for nuclear division (left; n > 200 cells/ time point). Swe1p levels in asynchronous 24°C cultures of the same strains are shown (right; with Cdc11p as loading control). (B) Cells expressing SWE1SV40-myc from the GAL1 promoter were induced with galactose for 1.5 h and processed for immunofluorescence. NES-A refers to "active" NES sequences, whereas NES-I refers to "inactive" NES sequences carrying a mutation in each NES (Edgington and Futcher, 2001). The following strains were used: swe1 (DLY690), SWE1 (DLY9376), SWE1NLS- (DLY9379), SWE1NES--myc (DLY9466), SWE1-NES A-myc (DLY7909), SWE1-NES I-myc (DLY7910), and SWE1SV40-myc (DLY9380).

Cytoplasmic and Nuclear Swe1p Are Equally Effective in Blocking the Apical-Isotropic Growth Switch under Conditions That Block Septin Assembly
In addition to promoting nuclear division, Clb2p/Cdc28p also triggers the apical-isotropic switch to depolarize bud growth in G2 (Lew and Reed, 1993). Because the cytoplasmic pool of Clb2p is responsible for triggering this switch (Eluere et al., 2007; Hood-DeGrenier et al., 2007), trapping Swe1p in the nucleus might be expected to impair its ability to delay this process. To assess whether Swe1p^NLS- and Swe1p^NES- promoted a normal delay of the apical-isotropic growth switch, we took advantage of the temperature-sensitive septin mutant cdc12-6. The septins are a family of filament-forming cytoskeletal proteins that assemble a specialized hourglass-shaped structure at the mother-bud neck (Gladfelter et al., 2001; Longtine and Bi, 2003). One function of this septin collar is to recruit Hsl1p and other regulators of Swe1p to this structure. On shift of cdc12-6 mutants to restrictive temperature, the septin collar disassembles, resulting in Swe1p accumulation, inhibition of Clb2p/Cdc28p, and a delay in both nuclear division and the apical–isotropic switch (Barral et al., 1999). The latter delay leads to the formation of elongated buds, whose length reflects the duration of the delay. Thus, bud elongation after cdc12-6 shift-up serves as an indicator of the effectiveness of Swe1p in inhibiting a cytoplasmic Clb2p/Cdc28p function. Using this assay, we found that both Swe1p^NLS- and Swe1p^NES- were as effective as wild-type Swe1p in promoting bud elongation (Figure 5). This indicates that either cytoplasmic or nuclear Swe1p is sufficient to block cytoplasmic Clb2p/Cdc28p activation.

View larger version (41K): [in this window] [in a new window]   Figure 5. Swe1p localization and checkpoint delay of the apical-isotropic switch. Prolonged bud growth in strains that are unable to organize septins. The indicated cdc12-6 cells were grown at 24°C, stained with ConA Alexa Flour-488 (green) and then resuspended in fresh media at 37°C. Two hours after the temperature shift, cells were stained with ConA Alexa Flour 597 (red). (A) Representative fluorescence micrographs of cdc12-6 strains carrying swe1 (JMY1489), SWE1-myc (DLY9593), SWE1NLS--myc (DLY9594), and SWE1NES--myc (DLY9595). (B) Bud length of new buds stained with ConA Alexa Flour-597 from A was measured using NIH ImageJ software (average ± SEM).

View larger version (35K): [in this window] [in a new window]   Figure 6. Mih1p localization. (A) Asynchronous cells expressing MIH1-myc (DLY5410) were grown at 24°C and processed for immunofluorescence to detect Mih1p and the septin Cdc11p. Representative images from cells in G1/S (unbudded and small budded), in S/G2 (budded), and telophase are shown. (For comparison to untagged control strains, see Supplemental Figure 1.) (B) The percentages of cells containing nuclear Mih1p at different cell cycle stages were determined for the indicated MIH1 alleles. The oval nuclei indicate elongated DAPI signals in cells undergoing nuclear division; n = 200. (C) Cells expressing MIH1-myc (JMY1319) were grown at 30°C and synchronized by pheromone arrest and release. At the indicated times, samples were scored for budding and nuclear division (n = 100) and processed for Western blotting to assess Mih1p levels (Cdc11p, loading control). (D) Schematic of Mih1p and alignment of amino acid sequence from three Saccharomyces species (S. cerevisiae, S. castelli, and S. kluyeri). The catalytic domain (green) and putative NLS (red) are indicated. Identical residues are highlighted in yellow, and basic motifs are highlighted in green. Amino acids marked by asterisk (*) were mutated to alanine. (E) Localization analysis for MIH1NLS--myc (DLY5406) and MIH1NLS--SV40-myc (DLY5405) strains performed as described in A. Quantitation for these strains is included in B. For MIH1NLS--myc only telophase cells are shown, as this is the time at which localization differs from wild type. (F) Localization of wild-type Mih1p during anaphase arrest (left) and following release (right) from a cdc15 late-anaphase block. Cells (JMY1444) were grown at 24°C and shifted to 37°C for 3 h (left) and then shifted down to 24°C and incubated for a further 30 min (right) or 60 min before fixation and processing for immunofluorescence. Parallel samples were stained with anti-myc to visualize Mih1p-myc and with anti-tubulin to visualize spindles. Two hundred cells were scored per sample for nuclear Mih1p and anaphase spindles, shown in the graph to the right.

Cytoplasmic Mih1p Is Less Effective than Nuclear Mih1p
To assess whether blocking Mih1p nuclear import affected its function, we generated hsl1 strains (in which Swe1p is stabilized) containing a single integrated copy of Mih1p or Mih1p^NLS-. As shown in Figure 7A, the cells relying on the cytoplasmic Mih1p^NLS- were significantly larger and more elongated than those containing wild-type Mih1p. They also exhibited a delay in nuclear division, as revealed by the increased proportion of uninucleate budded cells in the population (Figure 7B). These defects were rescued by appending SV40 T antigen NLS elements at the C terminus of Mih1p^NLS- (Figure 7, A and B), implying that the defect is due to lack of Mih1p nuclear import rather than some unrelated effect of the NLS mutations. All forms of Mih1p were expressed at comparable levels (Figure 7C), and Mih1p^NLS- with the appended SV40 NLS was constitutively nuclear (Figure 6, B and F). These findings suggest that nuclear Mih1p is more effective than cytoplasmic Mih1p, even when it comes to promoting a cytoplasmic Clb2p/Cdc28p function (i.e., the apical–isotropic switch).

View larger version (31K): [in this window] [in a new window]   Figure 7. Mih1p localization and function. (A) Cytoplasmic Mih1pNLS- is less effective in promoting the apical-isotropic switch. DIC images illustrating larger size and elongated cell morphology of hsl1 cells containing MIH1NLS--myc (DLY9955) compared with MIH1-myc (DLY9956) and MIH1NLS--SV40-myc (DLY9954) in asynchronous cultures grown at 24°C. (B) Cytoplasmic Mih1pNLS- is less effective in promoting nuclear division, as judged by the increase in budded preanaphase cells. Cells from A were scored as G1 (unbudded, uninucleate), S-G2 (budded, uninucleate), or telophase (budded, binucleate). n > 250. (C) Western blot analysis of Mih1p levels (Cdc11p, loading control) in the following strains grown at 24°C: cdc24-1 MIH1-myc (DLY5410), cdc24-1 MIH1NLS--myc (DLY5406), cdc24-1 MIH1NLS--SV40-myc (DLY5405), and no tag (DLY1). (D) Delay in nuclear division in strains that are unable to form buds. The cdc24-1 strains from C as well as swe1 (DLY690) and mih1 (JMY1172) controls were synchronized in G1 at 24°C, released from the pheromone arrest at 37°C, and scored for nuclear division (n > 200 cells/time point).

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
In this study, we have identified localization determinants in the CDK regulators Swe1p (NLS and NES) and Mih1p (NLS) that mediate nucleocytoplasmic shuttling and have investigated the effects of mutational inactivation of those determinants. A surprising conclusion from these studies is that even when these regulators are largely trapped in one compartment, they remain able to regulate events triggered by Clb2p/Cdc28p in the opposite compartment. Cytoplasmic Swe1p^NLS- was able to block nuclear division, and cytoplasmic Mih1p^NLS- was able to promote nuclear division. Conversely, nuclear Swe1p^NES- was effective in blocking the cytoplasmic apical–isotropic switch, whereas forcing Mih1p into the nucleus (with an appended SV40 T antigen NLS) was able to rescue inhibition of this cytoplasmic Clb2p function. The simplest way to explain these findings is that Clb2p/Cdc28p shuttles in and out of the nucleus rapidly enough that the Cdc28p tyrosine phosphorylation state reflects the total cellular (rather than compartment-specific) Swe1p/Mih1p status. This conclusion reveals that altering the localization of a regulator may not have the expected consequences when the target of that regulator is itself moving between cellular compartments.

If nuclear localization is not needed for Swe1p and Mih1p to act on Clb2p/Cdc28p, why are these proteins imported into the nucleus? Our data suggest that for both Swe1p and Mih1p, nuclear localization enhances their effectiveness as Clb2p/Cdc28p regulators. Nuclear Swe1p and Mih1p would be expected to gain increased access to the predominantly nuclear Clb2p/Cdc28p: by mass action, phosphorylation/dephosphorylation would proceed more rapidly in the compartment with higher substrate concentration. However, it may also be that upstream regulators of Swe1p and/or Mih1p are unequally partitioned, leading to differences in the relative activities of nuclear and cytoplasmic pools.

Swe1p is regulated at the level of protein degradation, and the correlation between Swe1p localization to the mother-bud neck and its degradation has been well documented (McMillan et al., 1999a; Shulewitz et al., 1999; Longtine et al., 2000; McMillan et al., 2002). That correlation suggested that the link might be causal, with neck localization being a prerequisite for degradation (Lew, 2003). We found that the nuclear export-impaired Swe1p^NES- mutant did not visibly accumulate at the neck, suggesting that most of the neck-localized Swe1p originates from the nuclear pool. More importantly, this mutant was largely immune to Hsl1p/Hsl7p-mediated degradation, even though it was fully competent to interact with Hsl7p in vitro. These findings support the hypothesis that Swe1p neck localization is an important step in its degradation.

Interestingly, the constitutively cytoplasmic Swe1p^NLS- was effectively localized to the neck, indicating that transit through the nucleus is not mandatory for neck targeting. The simplest scenario that incorporates our findings is as follows: when Swe1p is synthesized in late G1, there is no neck structure (because budding begins at the beginning of S phase) and Swe1p accumulates in the nucleus. Upon budding, Hsl1p and Hsl7p localize to the bud neck, and cytoplasmic Swe1p can be tethered to the neck through interaction with Hsl7p. The Swe1p NLS we identified is immediately adjacent to sequences implicated in Hsl7p interaction, and in fact mutations in the NLS residues partly impaired Hsl7p binding. These observations suggest that binding of Swe1p to importins and to Hsl7p might be mutually exclusive, precluding the nuclear import of Hsl7p-bound Swe1p. Hsl7p could therefore retain Swe1p in the cytoplasm to promote Swe1p phosphorylation events by neck-localized kinases that are thought to target it for degradation (Sakchaisri et al., 2004; Asano et al., 2005).

Previous work has shown that various stresses can inhibit the ability of septin-localized Hsl1p to promote Swe1p degradation (Longtine et al., 2000; Theesfeld et al., 2003; Clotet et al., 2006). Our work suggests that inhibition of Swe1p nuclear export would be an effective way for stress-responsive pathways to block Swe1p degradation and promote Cdc28p inhibition. It will be interesting to determine whether the nuclear import/export of Swe1p is regulated in response to stresses that trigger the morphogenesis checkpoint. Addressing this issue will require detection of Swe1p localization in living cells; however, we were unable to detect Swe1p-GFP, presumably due to the low abundance of Swe1p. A version of Swe1p fused to three green fluorescent protein (GFP) moieties was reported to localize to the nucleus and mother-bud neck when overexpressed from a high-copy plasmid (Sakchaisri et al., 2004), an observation that we have reproduced. On integration of a single copy of the Swe1p-3xGFP construct in the genome, however, we found that it was ineffective in sustaining a checkpoint response (data not shown), suggesting that the GFP fusion severely compromises Swe1p function. Until a functional, detectable GFP-Swe1p becomes available, the question of whether shuttling kinetics is affected by checkpoints will remain open.

The localization of several Cdc25 family proteins has been well studied (Pines, 1999; Donzelli and Draetta, 2003; Lindqvist et al., 2004), but the localization of Mih1p has not been described previously and is surprisingly distinct from that of its homologues in other systems. Cdc25 proteins in other systems undergo dramatic relocalization at the G2/M transition when dephosphorylation of Cdk1 is necessary to promote mitosis (e.g., mammalian Cdc25C translocates to the nucleus, whereas mammalian Cdc25B exports to the cytoplasm; Karlsson et al., 1999). In contrast, Mih1p remains distributed throughout the cell during G2 and M, although it remains possible that a small degree of Mih1p accumulates in the nucleus at G2/M that is not detected due to technical issues. Strikingly, Mih1p strongly accumulates in the nucleus during telophase in a MEN-dependent manner. This unexpected finding suggests that during telophase, Mih1p may play a nuclear role. Alternatively, it may be important to sequester Mih1p away from a cytoplasmic CDK population at this time. A recent report indicates that it is important to inhibit a residual cytoplasmic pool of cyclin B/Cdc2 in mammalian cells during mitotic exit (Nakajima et al., 2008). This inhibition is carried out by the Wee1-family kinase Myt1 and is presumably enabled by degradation of Cdc25 during mitosis (Nakajima et al., 2008). In yeast, Mih1p is not degraded, but sequestration of Mih1p in the nucleus may allow the remaining Swe1p (which is largely cytoplasmic during anaphase) to inhibit a residual cytoplasmic Clb/Cdc28p pool, thereby contributing to mitotic exit or cytokinesis.

In conclusion, our findings suggest that Clb2p/Cdc28p shuttles between nucleus and cytoplasm rapidly enough to integrate both nuclear and cytoplasmic Swe1p/Mih1p status, even when Swe1p or Mih1p are trapped within a specific compartment. Nevertheless, the ability of Swe1p to shuttle between the nucleus (where it can most effectively inhibit Clb2p/Cdc28p) and the cytoplasm (where it is subject to regulation by Hsl1p/Hsl7p) is important to couple the cytoplasmic process of bud formation to nuclear division.

	ACKNOWLEDGMENTS

 
We thank Sally Kornbluth, Steve Haase, and the members of the Lew laboratory for critical comments on the manuscript and for many stimulating discussions. We thank N. Edgington and Paul Russell for plasmids used in this study. This work was supported by National Institutes of Health/National Institute of General Medical Sciences grant GM-53050 (to D.J.L.).

	Footnotes

 
This article was published online ahead of print in MBC in Press (http://www.molbiolcell.org/cgi/doi/10.1091/mbc.E08-03-0286) on July 2, 2008.

Present addresses: * Department of Biochemistry and Molecular Genetics, University of Virginia, Charlottesville, VA 22908;

Lineberger Comprehensive Cancer Center, University of North Carolina, Chapel Hill, NC 27599;

Microbia Precision Engineering, Inc., Lexington, MA 02421; and

Duke-NUS Graduate Medical School Singapore, Singapore 169612.

Address correspondence to: Daniel J. Lew (daniel.lew{at}duke.edu)

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Asano, S., Park, J. E., Sakchaisri, K., Yu, L. R., Song, S., Supavilai, P., Veenstra, T. D., and Lee, K. S. (2005). Concerted mechanism of Swe1/Wee1 regulation by multiple kinases in budding yeast. EMBO J 24, 2194–2204.[CrossRef][Medline]

Bailly, E., Cabantous, S., Sondaz, D., Bernadac, A., and Simon, M. N. (2003). Differential cellular localization among mitotic cyclins from Saccharomyces cerevisiae: a new role for the axial budding protein Bud3 in targeting Clb2 to the mother-bud neck. J. Cell Sci 116, 4119–4130.[Abstract/Free Full Text]

Barral, Y., Parra, M., Bidlingmaier, S., and Snyder, M. (1999). Nim1-related kinases coordinate cell cycle progression with the organization of the peripheral cytoskeleton in yeast. Genes Dev 13, 176–187.[Abstract/Free Full Text]

Baudin, A., Ozier-Kalogeropoulos, O., Denouel, A., Lacroute, F., and Cullin, C. (1993). A simple and efficient method for direct gene deletion in Saccharomyces cerevisiae. Nucleic Acids Res 21, 3329–3330.[Free Full Text]

Bi, E., and Pringle, J. R. (1996). ZDS1 and ZDS2, genes whose products may regulate Cdc42p in Saccharomyces cerevisiae. Mol. Cell Biol 16, 5264–5275.[Abstract]

Booher, R. N., Deshaies, R. J., and Kirschner, M. W. (1993). Properties of Saccharomyces cerevisiae wee1 and its differential regulation of p34CDC28 in response to G1 and G2 cyclins. EMBO J 12, 3417–3426.[Medline]

Clotet, J., Escote, X., Adrover, M. A., Yaakov, G., Gari, E., Aldea, M., de Nadal, E., and Posas, F. (2006). Phosphorylation of Hsl1 by Hog1 leads to a G2 arrest essential for cell survival at high osmolarity. EMBO J 25, 2338–2346.[CrossRef][Medline]

Donzelli, M., and Draetta, G. F. (2003). Regulating mammalian checkpoints through Cdc25 inactivation. EMBO Rep 4, 671–677.[CrossRef][Medline]

Dunphy, W. G. (1994). The decision to enter mitosis. Trends Cell Biol 4, 202–207.[CrossRef][Medline]

Edgington, N. P., and Futcher, B. (2001). Relationship between the function and the location of G1 cyclins in S. cerevisiae. J. Cell Sci 114, 4599–4611.[Medline]

Eluere, R., Offner, N., Varlet, I., Motteux, O., Signon, L., Picard, A., Bailly, E., and Simon, M. N. (2007). Compartmentalization of the functions and regulation of the mitotic cyclin Clb2 in S. cerevisiae. J. Cell Sci 120, 702–711.[Abstract/Free Full Text]

Ferrigno, P., and Silver, P. A. (1999). Regulated nuclear localization of stress-responsive factors: how the nuclear trafficking of protein kinases and transcription factors contributes to cell survival. Oncogene 18, 6129–6134.[CrossRef][Medline]

Gladfelter, A. S., Pringle, J. R., and Lew, D. J. (2001). The septin cortex at the yeast mother-bud neck. Curr. Opin. Microbiol 4, 681–689.[CrossRef][Medline]

Hartwell, L. H. (1971). Genetic control of the cell division cycle in yeast. IV. Genes controlling bud emergence and cytokinesis. Exp. Cell. Res 69, 265–276.[CrossRef][Medline]

Harvey, S. L., Charlet, A., Haas, W., Gygi, S. P., and Kellogg, D. R. (2005). Cdk1-dependent regulation of the mitotic inhibitor Wee1. Cell 122, 407–420.[CrossRef][Medline]

Hood-DeGrenier, J. K., Boulton, C. N., and Lyo, V. (2007). Cytoplasmic Clb2 is required for timely inactivation of the mitotic inhibitor Swe1 and normal bud morphogenesis in Saccharomyces cerevisiae. Curr. Genet 51, 1–18.[Medline]

Hood, J. K., Hwang, W. W., and Silver, P. A. (2001). The Saccharomyces cerevisiae cyclin Clb2p is targeted to multiple subcellular locations by cis- and trans-acting determinants. J. Cell Sci 114, 589–597.[Abstract]

Karlsson, C., Katich, S., Hagting, A., Hoffmann, I., and Pines, J. (1999). Cdc25B and Cdc25C differ markedly in their properties as initiators of mitosis. J. Cell Biol 146, 573–584.[Abstract/Free Full Text]

Keaton, M. A., Bardes, E. S., Marquitz, A. R., Freel, C. D., Zyla, T. R., Rudolph, J., and Lew, D. J. (2007). Differential susceptibility of yeast S and M phase CDK complexes to inhibitory tyrosine phosphorylation. Curr. Biol 17, 1181–1189.[CrossRef][Medline]

Lew, D. J. (2003). The morphogenesis checkpoint: how yeast cells watch their figures. Curr. Opin. Cell Biol 15, 648–653.[CrossRef][Medline]

Lew, D. J., and Reed, S. I. (1993). Morphogenesis in the yeast cell cycle: regulation by Cdc28 and cyclins. J. Cell Biol 120, 1305–1320.[Abstract/Free Full Text]

Lim, H. H., Goh, P. Y., and Surana, U. (1998). Cdc20 is essential for the cyclosome-mediated proteolysis of both Pds1 and Clb2 during M phase in budding yeast. Curr. Biol 8, 231–234.[CrossRef][Medline]

Lindqvist, A., Kallstrom, H., and Karlsson Rosenthal, C. (2004). Characterisation of Cdc25B localisation and nuclear export during the cell cycle and in response to stress. J. Cell Sci 117, 4979–4990.[Abstract/Free Full Text]

Longtine, M. S., and Bi, E. (2003). Regulation of septin organization and function in yeast. Trends Cell Biol 13, 403–409.[CrossRef][Medline]

Longtine, M. S., Theesfeld, C. L., McMillan, J. N., Weaver, E., Pringle, J. R., and Lew, D. J. (2000). Septin-dependent assembly of a cell cycle-regulatory module in Saccharomyces cerevisiae. Mol. Cell Biol 20, 4049–4061.[Abstract/Free Full Text]

Ma, X. J., Lu, Q., and Grunstein, M. (1996). A search for proteins that interact genetically with histone H3 and H4 amino termini uncovers novel regulators of the Swe1 kinase in Saccharomyces cerevisiae. Genes Dev 10, 1327–1340.[Abstract/Free Full Text]

McMillan, J. N., Longtine, M. S., Sia, R. A., Theesfeld, C. L., Bardes, E. S., Pringle, J. R., and Lew, D. J. (1999a). The morphogenesis checkpoint in Saccharomyces cerevisiae: cell cycle control of Swe1p degradation by Hsl1p and Hsl7p. Mol. Cell Biol 19, 6929–6939.[Abstract/Free Full Text]

McMillan, J. N., Sia, R. A., Bardes, E. S., and Lew, D. J. (1999b). Phosphorylation-independent inhibition of Cdc28p by the tyrosine kinase Swe1p in the morphogenesis checkpoint. Mol. Cell Biol 19, 5981–5990.[Abstract/Free Full Text]

McMillan, J. N., Sia, R. A., and Lew, D. J. (1998). A morphogenesis checkpoint monitors the actin cytoskeleton in yeast. J. Cell Biol 142, 1487–1499.[Abstract/Free Full Text]

McMillan, J. N., Theesfeld, C. L., Harrison, J. C., Bardes, E. S., and Lew, D. J. (2002). Determinants of Swe1p degradation in Saccharomyces cerevisiae. Mol. Biol. Cell 13, 3560–3575.[Abstract/Free Full Text]

Nakajima, H., Yonemura, S., Murata, M., Pwinica-Worms, H., and Nishida, E. (2008). Myt1 protein kinase is essential for Golgi and ER assembly during mitotic exit. J. Cell Biol 181, 89–103.[Abstract/Free Full Text]

O'Farrell, P. H. (2001). Triggering the all-or-nothing switch into mitosis. Trends Cell Biol 11, 512–519.[CrossRef][Medline]

Pal, G., Paraz, M. T., and Kellogg, D. R. (2008). Regulation of Mih1/Cdc25 by protein phosphatase 2A and casein kinase 1. J. Cell Biol 180, 931–945.[Abstract/Free Full Text]

Pines, J. (1999). Four-dimensional control of the cell cycle. Nat. Cell Biol 1, E73–E79.[CrossRef][Medline]

Richardson, H. E., Wittenberg, C., Cross, F., and Reed, S. I. (1989). An essential G1 function for cyclin-like proteins in yeast. Cell 59, 1127–1133.[CrossRef][Medline]

Russell, P., Moreno, S., and Reed, S. I. (1989). C