Genetics, Vol. 162, 689-703, October 2002, Copyright © 2002
The sal3+ Gene Encodes an Importin-ß Implicated in the Nuclear Import of Cdc25 in Schizosaccharomyces pombe
Gordon Chuaa,
Carol Lingnera,
Corey Frazera, and
Paul G. Younga
a Department of Biology, Queen's University, Kingston, Ontario K7L 3N6, Canada
Corresponding author:
Paul G. Young, Rm. 2443, Biosciences Complex, Queen's University, Kingston, Ontario K7L 3N6, Canada., youngpg{at}biology.queensu.ca (E-mail)
Communicating editor: P. RUSSELL
 | ABSTRACT |
|---|
In Schizosaccharomyces pombe, the nuclear accumulation of Cdc25 peaks in G2 and is necessary for the proper timing of mitotic entry. Here, we identify the sal3+ gene product as an importin-ß homolog that participates in the nuclear import of Cdc25. Loss of sal3+ results in a cell cycle delay, failure to undergo G1 arrest under nitrogen-starvation conditions, and mislocalization of Cdc25 to the cytosol. Fusion of an exogenous classical nuclear localization sequence (cNLS) to Cdc25 restores its nuclear accumulation in a sal3 disruptant and suppresses the sal3 mutant phenotypes. In addition, we show that enhanced nuclear localization of Cdc25 at endogenous levels of expression advances the onset of mitosis. These results demonstrate that the nuclear translocation of Cdc25 is important for the timing of mitotic entry and that Sal3 plays an important role in this process.
PROPER localization of gene products is necessary for their normal function to ensure coordinated access to regulators and substrates. Such is the case for elements involved in cell cycle control where subcellular localization can influence the progression of the cell cycle (reviewed in TAKIZAWA and MORGAN 2000 
; YONEDA 2000 ). The mechanism of their regulated localization is therefore important in understanding cell cycle control. The fission yeast, Schizosaccharomyces pombe, is ideal for studying this process since many of the cell cycle regulatory genes are well characterized and the consequences of perturbing their subcellular localization are easily examined through molecular genetic and imaging techniques.
The crucial event for mitotic initiation is the activation of a protein kinase complex consisting of catalytic and regulatory cyclin B subunits encoded by the cdc2+ and cdc13+ genes, respectively (SIMANIS and NURSE 1986 ; BOOHER et al. 1989 ). The activity of the kinase is regulated by the phosphorylation state of tyrosine-15 (Y-15) on Cdc2 (GOULD and NURSE 1989 ). Inhibitory phosphorylation of Cdc2 Y-15 is carried out by the Wee1 and Mik1 kinases (LUNDGREN et al. 1991 ; MCGOWAN and RUSSELL 1993 ) while activation is dependent upon Y-15 dephosphorylation by the Cdc25 and Pyp3 phosphatases (GAUTIER et al. 1991 ; MILLAR et al. 1992 ). Timing of mitotic entry and cell size at division is therefore highly sensitive to the gene dosage of the primary regulators, Cdc25 and Wee1 (RUSSELL and NURSE 1986 , RUSSELL and NURSE 1987A ). Below the cell size threshold for mitotic entry Cdc2 is held inactive by Y-15 phosphorylation, predominantly through the action of Wee1. A G2 cell size homeostasis checkpoint acting through Cdc25 monitors the size threshold and triggers mitotic entry upon attainment of a critical cell size (RUPES et al. 2001 ). Thus cells that have fulfilled the size requirement for mitosis show a rapid decrease in Y-15 phosphorylation as a consequence of Cdc25 activation (RUPES et al. 2001 ).
The localization of Cdc2, Cdc13, and Cdc25 fluctuates in the cell cycle, colocalizing in the nucleus at maximal levels from late G2 to mitotic metaphase and at minimal levels from mitotic anaphase to S phase (BOOHER et al. 1989 ; LOPEZ-GIRONA et al. 1999 ; DECOTTIGNIES et al. 2001 ). The coordinated nuclear accumulation of Cdc2-Cdc13 complexes and of Cdc25 results in Cdc2 activation in the nucleus and consequent mitotic entry. Consistent with this model is the observation that nuclear exclusion of Cdc25 by removal of a nuclear localization sequence (NLS) results in a mitotic delay (LOPEZ-GIRONA et al. 2001 ).
Since Cdc25 is too large to diffuse through the nuclear pore complex (NPC), the periodic accumulation of Cdc25 in the nucleus must be an active process. Nuclear transport is mediated by a family of importin-ß (karyopherin-ß1)-related molecules consisting of importins and exportins that transport proteins in and out of the nucleus, respectively (reviewed in MATTAJ and ENGLMEIER 1998 ; WEIS 1998 ). Three major steps involving importin-ß interactions are required for the nuclear import of proteins. First, proteins destined for the nucleus contain a NLS that is recognized by importin-ß. NLSs are broadly divided into classical (cNLS) and nonclassical types. The former requires an adapter molecule (importin-
) to bridge the interaction between the cargo protein and importin-ß, and the latter involves direct binding of importin-ß to its cargo protein. Importin-ß then associates with various nucleoporins, components of the NPC. This results in the docking of the importin-ß (importin-)-cargo complex at the NPC's cytoplasmic face and subsequent translocation to its nuclear face. Third, the binding of RanGTP to importin-ß in the nucleus causes the release of the cargo protein and importin-ß is recycled back to the cytoplasm.
Nuclear export of proteins occurs in a highly analogous fashion to the nuclear import process except that the targeting signals, importin-ß homologs and their interactions with specific nucleoporins and RanGTP, are distinct (reviewed in JANS et al. 2000 ). In this case, RanGTP-bound exportin associates with the cargo protein containing a nuclear export signal (NES) in the nucleus and then dissociates in the cytoplasm following translocation through the NPC and RanGTP hydrolysis.
In budding yeast, 14 importin-ß's and one importin- have been found to function in the nuclear transport of several hundred proteins (WOZNIAK et al. 1998 ). This indicates that each importin-ß must be capable of transporting multiple cargo proteins. Studies have shown that this is the case and that certain proteins are transported by several importin-ß's while others rely specifically on one importin-ß for their transport (reviewed in JANS et al. 2000 ).
Change in the subcellular localization of S. pombe Cdc25 during the cell cycle is dependent upon the relative rates of nuclear import and export. The concentration of Cdc25 in the nucleus at the onset of mitosis may be due to an increase in import, a decrease in export, or a combination of the two. Inhibition of the export factor exportin 1 (Crm1) by mutation or the drug leptomycin B causes a nuclear accumulation of Cdc25, indicating that Cdc25 is actively transported in and out of the nucleus and that the latter occurs via Crm1 (NISHI et al. 1994 ; LOPEZ-GIRONA et al. 1999 ; ZENG and PIWNICA-WORMS 1999 ). However, information on the import process remains limited. Although an importin- homolog has been associated with Cdc25C transport in Xenopus (KUMAGAI and DUNPHY 1999 ), the nuclear import receptors responsible for the translocation of Cdc25 into the nucleus have not been identified in S. pombe.
Here we report the isolation and the functional characterization of a fission yeast importin-ß homolog encoded by the sal3+ gene. The sal3 mutant was originally discovered as a sup3-5 allosuppressor displaying a cold-sensitive cell cycle defect (NURSE and THURIAUX 1984 ). We have reisolated sal3 in an insertional mutagenesis screen for changed division response (Cdr)- mutants that are defective in reducing the size threshold for cell division in response to nitrogen deprivation. Loss of sal3+ results in the mislocalization of Cdc25, cell elongation, and failure to undergo G1 arrest under nitrogen-starvation conditions. Addition of an exogenous classical NLS to Cdc25 restores its nuclear accumulation in a sal3 disruptant and suppresses the sal3 mutant phenotypes. Altogether, these data show that Sal3 plays an important role in the nuclear import of Cdc25.
| MATERIALS AND METHODS |
|---|
Yeast strains, media, and general methods:
All strains were derived from 972 h-, 975 h+, and 968 h90 (MITCHISON 1970 ) and are listed in Table 1. Cells were grown on YEA (MITCHISON 1970 ) or Edinburgh minimal medium (EMM; MITCHISON 1970 as modified by NURSE 1975 ). Matings were performed on SPA plates (GUTZ et al. 1974 ) and strains tested for the Cdr- phenotype on EMM minus NH4Cl (EMM - N) plates at 25° after 2–3 days (YOUNG and FANTES 1984 , YOUNG and FANTES 1987 ). Standard techniques for genetic analysis were followed as described previously (MORENO et al. 1991 ). Yeast cells were transformed by the lithium acetate method (MORENO et al. 1991 ) or by electroporation (Bio-Rad, Richmond, CA, gene pulser) as indicated (PRENTICE 1991 ). Cell length (micrometers) was determined in septated log-phase cells by the ruler function in Slidebook (Intelligent Imaging Innovations, Denver) on images of cells captured with a high-performance cooled CCD camera (Cooke SensiCam, Auburn Hills, MI).
Integration mapping of sal3+:
A PCR product containing the full-length sal3+ open reading frame (ORF) flanked by NdeI and SalI sites at the 5' and 3' ends, respectively, was amplified with high-fidelity Taq polymerase (Roche Molecular Biochemicals, Indianapolis), using the primers SAL3GC1 (5'-GGAATTCCATATGTCTAGTGGATTTCCTCCTGAATAT-3') and SAL3GC3 (5'-ACGCGTCGACTTAAAAATGTGCAGACAAAGCTCTCTGA-3'), and genomic DNA was isolated (MORENO et al. 1991 ) as template. Using standard molecular biology techniques (SAMBROOK et al. 1989 ), the PCR product was digested with NdeI and SalI and cloned into the pREP1 and pREP41 expression vectors under the control of the nmt1 and nmt41 promoters, respectively (BASI et al. 1993 ). The resulting plasmid was electroporated into sal3-33 leu1-32 to test for functionality by rescue of the Cdr- phenotype, and then stable leu+ integrants were selected as detailed in MORENO et al. 1991 . Three of the integrants were crossed to a leu1-32 strain and their tetrads analyzed. A total of 60 tetrads segregating 2:2 Leu+ to Leu- yielded only wild-type progeny, indicating that sal3+ integrated at or near the site of the mutation.
Disruption of sal3+:
A 4.9-kb HindIII-SacI fragment containing the sal3+ ORF with 950 and 612 bp of 5' and 3' flanking sequences, respectively, was PCR amplified (MBI Fermentas) using the primers KOGC3 (5'-GGGGGAAGCTTAGCGAACAATAACTTAGCTTG-3') and KOGC4 (5'-GGGGGGAGCTCAAACTTATATGACCAACATTC-3'). This PCR product was cloned into pGEM-T (Promega, Madison, WI) and subsequently digested with KpnI and BclI to remove the majority of the sal3+ ORF except for 63 bp at the C terminus. A 1.8-kb fragment of the ura4+ cassette (GRIMM et al. 1988 ) containing terminal KpnI and BclI sites was generated by PCR with the primers KOGC1 (5'-GGGGGGGTACCAGCTTAGCTCACCCTCCCACTGGC-3') and KOGC2 (5'-GGGGGTGATCATGTGATATTGACGAAACTTTTTGACAT-3') and inserted in place of the missing sal3+ ORF. The resulting plasmid was digested with HindIII and SacI to liberate the deletion construct, which was separated from the vector by Gene Clean (Quantum Biotechnologies). The sal3::ura4+ construct was transformed into an h-/h+ ura4-D18/ura4-D18 leu1-32/leu1-32 ade6-210/ade6-216 diploid (Q474) by the lithium acetate-DMSO method (BAHLER et al. 1998 ) and plated on minimal medium lacking uracil. The ura+ diploid transformants were isolated and allowed to sporulate. Tetrad dissection showed a 2:2 segregation of ura+ and ura- progeny, indicating that the sal3 null was viable. The disruption of the sal3+ ORF was confirmed by PCR analysis (data not shown). Genomic DNA isolated from either a ura+ diploid or its sporulated products was used as template for PCR. Utilizing previously mentioned primers, KOGC3 and SAL3GC3, two PCR products of sizes 4.2 and 2.6 kb were amplified from the diploid strain. These fragments corresponded to the wild-type and disrupted versions of the sal3+ gene, respectively. All ura+ and ura- progeny showed PCR amplification of the disrupted and wild-type sal3+, respectively (data not shown).
Nitrogen downshift experiments:
Log-phase cells grown in EMM at 30° were harvested and washed twice with 50 ml of EMM - N on microfiber glass filters (Whatman) before release into EMM - N (25°) at a cell density of 2–4 x 106 cells/ml. Samples were collected at various time intervals for determination of cell number and DNA content. For the growth-kinetic experiments, 50-µl cell samples were transferred to 14 ml of Coulter counter fluid (Fisher Scientific, Pittsburgh), sonicated briefly, and cell number was measured with a Coulter counter (Coulter Electronics, Hialeah, FL). For flow cytometry, samples of 
1 x 107 cells were fixed in 70% ethanol, washed in 50 mM sodium citrate (pH 7.0), incubated with 10 mM RNase (Sigma, St. Louis) for 1.5 hr at 37°, and stained with 1 µM propidium iodide (Sigma; ALFA et al. 1993 ). Cells were sonicated briefly prior to fluorescence-activated cell sorter (FACS) analysis.
Construction of C-terminal GFP fusions:
The ORFs of sal3+, cdc2+, cdc13+, cdc25+, wee1+, mik1+, pyp3+, cdr1+/nim1+, cdr2+, spc1+/sty1+, cdc10+, res1+, res2+, and rep2+ were PCR amplified using the primers listed in Table 2 and similarly cloned into pREP1/41/81-GFP (S165T) expression vectors (TARICANI et al. 2002 ) as mentioned above. The functionality of the green fluorescent protein (GFP) fusions was assessed by their ability to rescue the mutant phenotypes and the presence of GFP fluorescence in the cell.
Fluorescence microscopy:
Cells containing nmt-driven and chromosomally integrated GFP-fusion constructs regulated by the native promoter were grown in minimal media lacking thiamine and in YEA for 18–24 hr, respectively, at 25° and 35°, and subjected to methanol fixation (ALFA et al. 1993 ). Approximately 1 x 107 cells were harvested by centrifugation (3000 rpm, 5 min), resuspended in 1 ml of -20° methanol, and incubated on a rotary inverter (Barnstead/Thermolyne) for 7 min. The fixed cells were then centrifuged (3000 rpm, 5 min), washed once with 1 ml of 50 mM sodium citrate (pH 7.0), and resuspended in 10 µl of 50 mM sodium citrate (pH 7.0). The cell suspension (1.0 µl) was mixed with 1 µl of 4',6-diamidino-2-phenylindole (0.5 µg/ml) and placed on a microscope slide. Cells were examined using a Leitz DMRB fluorescence microscope with a 100x objective (Leica Microsystems) and images were captured with a high performance CCD camera (Cooke SensiCam) and Slidebook image analysis software (Intelligent Imaging Innovations).
Chromosomal integration of GFP-tagged genes:
To facilitate the chromosomal integration of cdc25-GFP at its own locus, a 1.5-kb PstI-SalI fragment carrying the nmt1 promoter was liberated from pREP1-GFP and replaced with a PCR product containing the entire cdc25+ ORF and 1550 bp of upstream sequence. The primers CDC25CF1 (5'-ACGCCTGCAGTCCGAGTTTAACAAGACAACTGGC-3') and CDC25GC3 (see above) were used for amplification. The resultant plasmid was integrated into a cdc25::ura4+ cdc2-3w ura4-D18 leu1-32 strain and when the putative integrants were outcrossed there were no cdc- progeny. For the cdc25NLS-GFP integrant, an SV40 T-antigen nuclear localization signal (PKKKRKV; underlined) was attached to the carboxyl terminus of the cdc25+ ORF by using the primer CDC25GC6 (5'-ACGCGTCGACGAGACCTTACGCTTCTTCTTAGGAAATCTTCTAAGTGTAGAGAGGGAATGCA-3') and constructed in a similar manner as above.
The cdc13-GFP integrant was isolated from a cdc13-117 background following transformation with pREP81cdc13+-GFP. Stable transformants were tested for suppression of the cdc- phenotype of cdc13-117 at 36° in the presence of 25 µM thiamine to select for insertions producing wild-type Cdc13-GFP controlled by its native promoter. Outcrossing of such integrants generated no cdc- progeny. Insertion of the plasmid into the nmt1 locus or the cdc13 locus upstream of the cdc13-117 point mutation site results in an endogenously regulated mutant cdc13-GFP and was screened against. A similar approach was employed to obtain the sal3-GFP integrant.
Protein extracts and Western blots:
Cells were harvested by centrifugation and washed once in 1 ml of ice-cold stop buffer (150 mM NaCl, 50 mM NaF, 10 mM EDTA, and 1 mM NaN3, pH 8.0). They were then resuspended in 200 µl of lysis buffer (150 mM NaCl, 50 mM Tris-HCl, pH 8.0, 50 mM NaF, 5 mM EDTA, 1 mM Na3VO4, 1 mM phenylmethylsulfonyl fluoride, 10% glycerol, 1% Nonidet P-40, and 1 mM dithiothreitol) supplemented with protease inhibitor cocktail tablets (Roche Molecular Biochemicals). A total of 200 µl of 425- to 600-µm acid-washed glass beads (Sigma) was added and the cells were broken by vortexing the contents in a 50-ml Falcon tube. The cell lysate was then transferred to microcentrifuge tubes and cleared by centrifugation at 14,000 rpm for 5 min. Protein concentration was determined by the Bio-Rad protein assay and 5 µg of each sample was resolved on 10% SDS-PAGE followed by transfer onto a polyvinylidene fluoride membrane (New England Nuclear Life Science, Boston). Using standard Western blotting procedures, 0.2 µg/ml of anti-GFP (Roche Molecular Biochemicals) and the TAT-1 monoclonal antibody (WOODS et al. 1989 ) were added and immunodetected with the horseradish peroxidase-conjugated anti-mouse secondary antibody (Roche Molecular Biochemicals) and the ECL chemiluminescence method (New England Nuclear Life Science).
| RESULTS |
|---|
The sal3-33 mutant displays a cdr- phenotype:
The sal3-33 mutation was originally identified in a complex screen for allosuppressors. The mutation was able to reconstitute the tRNA nonsense suppressing activity of an inactive sup3 allele (NURSE and THURIAUX 1984 ). Since allosuppressors are likely to influence tRNA or protein synthesis, and sal3 mutants exhibited a cold-sensitive cell-elongated phenotype, it was suggested that the sal3+ gene product may function in the transduction of nutritional or growth cues to the mitotic control (NURSE and THURIAUX 1984 ). This hypothesis was further supported by our observations that sal3-33 cells displayed a cdr- phenotype. Upon nitrogen deprivation, cell numbers per colony were fewer and the size threshold for cell division was not reduced as seen for wild type (Fig 1). The cdr- phenotype was more severe at lower temperatures (20°; data not shown). This indicated that the sal3+ gene product is necessary for proper cell division response under starvation conditions.
View larger version (73K):
In this window
In a new window
Download PPT slide
|
Figure 1.
Alleles of sal3 display a cdr- phenotype. Colony morphologies of wild type, a point mutant, and an insertional mutant of sal3 (sal3-33 and sal3-i2, respectively) on EMM and EMM - N plates are shown. Strains were grown at 25° for 2 days.
|
|
Isolation of the sal3+ gene:
Our attempts to clone sal3+ by plasmid complementation were unsuccessful. We took advantage of an insertional mutagenesis system available in S. pombe (CHUA et al. 2000 ) to isolate novel cdr genes and as a possible alternative for retrieving the sal3+ gene. Seven cdr mutant strains were recovered and by linkage analysis determined to represent three alleles of cdr2, two alleles of sal3, and a single allele each of mcs1/res2 and ssp1. All four genes have been shown to have some role in G2/M control (NURSE and THURIAUX 1984 ; YOUNG and FANTES 1984 , YOUNG and FANTES 1987 ; MOLZ et al. 1989 ; MATSUSAKA et al. 1995 ; BREEDING et al. 1998 ; KANOH and RUSSELL 1998 ; TOURNIER and MILLAR 2000 ), but the molecular identity of sal3+ and its mode of mitotic regulation were unknown. To identify the sal3+ gene, genomic sequences flanking the ura4+ gene insertion were determined from the sal3-i2 insertional mutant and the site of integration in the genome was localized by comparison with the S. pombe sequence database. Insertion occurred 179 bp upstream from a gene (GenBank accession no. CAA20126) within the cosmid c1840 on chromosome III. The upstream insertion suggests that sal3-i2 is likely a partial loss-of-function allele. This is supported by a less severe cdr- phenotype displayed in sal3-i2 compared to the sal3-33 point mutant (Fig 1).
The full-length sal3+ ORF under the control of the nmt41 promoter rescued the cdr- phenotype of sal3-33 as well as that of a sal3 disruptant (see below) in nitrogen-deprivation conditions (Fig 2A). In addition, the cold-sensitive cell-elongation phenotype of sal3 mutants to various stresses such as high osmolarity (1.2 KCl) and low pH (3.5) was also suppressed by this construct (data not shown). When the plasmid was integrated in a sal3-33 background, no Cdr- spores were found upon outcrossing, indicating integration at the sal3+ locus.
View larger version (69K):
In this window
In a new window
Download PPT slide
|
Figure 2.
Genetic and phenotypic analysis of sal3 alleles under nitrogen-deprivation conditions. (A) The cdr- phenotype of both sal3 alleles is suppressed by the full-length sal3+ ORF expressed from the nmt41 promoter. The strains were grown on EMM - N plates at 25° for 2 days. (B) sal3-33 and the sal3 disruptant fail to complement. The diploid strains were grown on EMM - N plates at 25° for 2 days. (C) Growth kinetics of wild-type (solid diamonds) and the sal3 disruptant (open squares) upon nitrogen downshift (t = 0). Error bars represent standard error of the mean with a sample size of three for each strain. (D) FACS analysis of wild type (top) and the sal3 disruptant (bottom) upon nitrogen downshift (t = 0).
|
|
The sal3+ gene encodes a fission yeast homolog of importin-ß:
The sal3+ gene product consists of 1095 amino acids with significant homology to importin-ß-3 (IB3). It shows 38% identity and 54% similarity to the Saccharomyces cerevisiae Pse1p and 33% identity and 53% similarity to human RanBP5 over the length of the protein (NCBI Blast). The regions of homology appear to be scattered throughout the protein. The gene products of this family have been demonstrated to be involved in the nucleocytoplasmic transport of proteins and RNA (reviewed in MATTAJ and ENGLMEIER 1998 ; WEIS 1998 ; STROM and WEIS 2001 ). Sal3 contains two conserved domains common to importin-ß molecules. The WPEL motif and a glutamate and an aspartate-rich region are located at approximately amino acids 130 and 330, respectively. These domains have been shown to interact with the importin- adapter protein and RanGTP, influencing the binding of importin-ß to its substrate (REXACH and BLOBEL 1995 ; ENENKEL et al. 1996 ; MOROIANU et al. 1996 ).
The sal3-33 mutation is likely a null allele:
A sal3 disruptant missing 98% of the ORF (see MATERIALS AND METHODS) was viable and indistinguishable in phenotype from the sal3-33 allele (data not shown). A sal3-33/sal3::ura4+ diploid exhibited a severe cdr- phenotype (Fig 2B), indicating that the mutations were allelic. Furthermore, no expression was detected in a C-terminal GFP fusion to sal3-33 (data not shown), suggesting that the mutant protein is unstable or possibly carries a nonsense mutation in the sal3+ gene. Together, these observations demonstrate that sal3-33 is probably a null allele. All further work was done with the deletion strain.
The sal3 disruptant fails to undergo G1 arrest in response to nitrogen starvation:
Unlike wild-type cells that arrest in G1 when starved of nitrogen, cdr mutants such as cdr2 and spc1 arrest in G2 during nitrogen deprivation, indicating a defect in the G2/M size control (SHIOZAKI and RUSSELL 1995 ; BREEDING et al. 1998 ; KANOH and RUSSELL 1998 ). To examine the nitrogen-starvation response of the sal3 disruptant, growth kinetics and flow cytometry were performed. When asynchronous wild-type cells are rapidly deprived of nitrogen, they divide approximately twice, increasing the initial cell numbers by 3.6 times before arresting in G1 (YOUNG and FANTES 1987 ). In contrast, the cell number of the sal3 disruptant increased by only about one-half of the wild-type level during nitrogen deprivation (Fig 2C), indicating the occurrence of only a single cell division. Furthermore, FACS analysis revealed that the sal3 disruptant accumulated a G2 peak during an extended period of nitrogen deprivation while wild-type cells displayed a G1 peak under the same conditions (Fig 2D). Microscopic examination of the sal3 disruptant cells after prolonged nitrogen starvation revealed that they were uninucleate but the majority displayed a stretched nuclear morphology that appeared to be the result of plasmolysis (data not shown). Altogether, these results demonstrate that loss of sal3+ causes a failure to undergo the second cell division and G1 arrest in response to nitrogen starvation.
Alleles of sal3 are synthetically lethal with cdc25 mutant strains:
Because loss of sal3+ causes a cell cycle delay and therefore abnormally long cells, the role of sal3+ in the cell cycle was addressed by investigating its genetic relationship to known elements of the mitotic size control. Inactivation of both sal3+ and cdc25+ results in a strong synthetic-lethal interaction. The partial sal3-i2 allele in a cdc25-22 background undergoes a single cell cycle arrest at the permissive temperature (25°; Fig 3). A similar interaction was observed with sal3-33 in combination with cdc25-22r1, a partial revertant of cdc25-22 (BREEDING et al. 1998 ; data not shown). In addition, sal3-33 also exhibited similar negative interactions with alleles of other cdc genes such as cdc2-M35 and cdc13-117, each of which is also synthetically lethal with cdc25-22 (BUENO and RUSSELL 1993 ; data not shown). These synthetic-lethal interactions were suppressed in a stf1-1/cut12 background, a dominant gain-of-function allele that rescues the partial and complete loss of cdc25+ (HUDSON et al. 1990 ; BRIDGE et al. 1998 ; data not shown). In summary, these results imply that the sal3+ gene product may control mitotic entry by affecting the Y-15 phosphorylation state of Cdc2.
View larger version (49K):
In this window
In a new window
Download PPT slide
|
Figure 3.
The sal3 insertional allele forms a synthetic-lethal interaction with the cdc25-22 strain. Meiotic products of a sal3-i2/cdc25-22 mating were sporulated and grown on YEA at the semipermissive temperature (25°) for 3 days.
|
|
Genetic analysis indicates that Sal3 influences the Y-15 phosphorylation state through Cdc25:
We next used genetic analysis to determine how Sal3 affected Y-15 phosphorylation/dephosphorylation. If Sal3 were acting solely on Wee1, then loss of Wee1 activity would completely suppress the cell-elongated phenotype of the sal3 disruptant. However, wee1-50 was not fully epistatic to the sal3 disruptant at 35° because sal3::ura4+ wee1-50 cells were significantly larger than wee1-50 alone (Table 2). These results are consistent with Sal3 being sensitive to Y-15 phosphorylation through a Wee1-independent pathway. If Sal3 affects Y-15 phosphorylation, then it potentially acts through Mik1. In this case, the cell elongation of the sal3 disruptant would be caused by Mik1 hyperactivity. The deletion of mik1+ would result in suppressing the cell size phenotype of sal3::ura4+. A sal3:: ura4+ mik1::ura4+ double mutant displayed the same size as sal3::ura4+ alone (data not shown), indicating that Sal3 function does not exclusively involve Mik1.
We next determined whether Sal3 regulates Y-15 phosphorylation through Cdc25 by examining the effect of the sal3 disruptant in the absence of cdc25+ gene activity. To circumvent the lethality associated with the loss of Cdc25 function, the cdc2-3w mutation that rescues strains lacking cdc25+ was used in this background (RUSSELL and NURSE 1987A ). It is expected that a cell cycle delay caused by the sal3 disruptant in a cdc2-3w cdc25-22 background is indicative of a Sal3 function independent of Cdc25, while a Cdc25-mediated process is manifested as a cdc2-3w cdc25-22 epistasis of sal3::ura4+. We discovered that the sal3::ura4+ cdc2-3w cdc25-22 triple mutant did not show a significantly different cell size from cdc2-3w cdc25-22 at either temperature (25° and 35°; Table 2), suggesting strongly that Sal3 regulates Cdc25 activity. In contrast, Sal3 appeared not to function through the secondary tyrosine phosphatase Pyp3 because the sal3-33 pyp3::ura4+ double mutant displayed a similar cell size to sal3-33 alone (data not shown).
In addition, the sal3 disruptant reverses the wee1-50 epistasis of cdc25-22 (Table 2), indicating that the cell cycle delay of sal3 mutants also occurs through a pathway independent of Wee1 and Cdc25. This result suggests that Sal3 is involved in regulating the activity of more than one cell cycle control gene.
The sal3 disruptant shows negative interactions with other cdr mutants:
Mutations in several cdr genes including cdr1+/nim1+, cdr2+, spc1+/sty1+, and mcs1+/res2+ have been reported to display a synthetic-lethal interaction with cdc25-22 (MOLZ et al. 1989 ; FEILOTTER et al. 1991 ; SHIOZAKI and RUSSELL 1995 ; KANOH and RUSSELL 1998 ; TOURNIER and MILLAR 2000 ). Because sal3 alleles also share these phenotypes, double mutants in various combinations were constructed to determine whether sal3+ belonged in a common or separate pathway from these cdr genes. We found that the sal3 disruptant showed an additive effect in combination with alleles of cdr1/nim1, cdr2, spc1/sty1, and mcs1/res2 (Fig 4), indicating that the nutritional modulation of mitotic control by Sal3 is distinct from that of these gene products. This was not surprising since both Cdr1/Nim1 and Cdr2 kinases have been shown to function through Wee1 (RUSSELL and NURSE 1987B ; YOUNG and FANTES 1987 ; FEILOTTER et al. 1991 ; COLEMAN et al. 1993 ; PARKER et al. 1993 ; WU and RUSSELL 1993 ; BREEDING et al. 1998 ; KANOH and RUSSELL 1998 ), and Spc1/Sty1 functions independently of Cdc25 (SHIOZAKI and RUSSELL 1995 ), while our results indicated otherwise for Sal3 (see above).
View larger version (63K):
In this window
In a new window
Download PPT slide
|
Figure 4.
The sal3 disruptant displays additive interactions with other cdr mutants. All strains were grown on EMM - N plates at 25° for 2 days. Top, single cdr mutants; bottom, the corresponding cdr mutants in a sal3 mutant background.
|
|
The cdr- phenotype of the sal3 disruptant is suppressed by a net increase in Y-15 dephosphorylation:
On the basis of the identification of Sal3 as an importin-ß homolog, our next approach was to construct C-terminal GFP fusions to candidate genes and examine their intracellular localization in wild type and the sal3 mutant. The GFP-tagged proteins were placed under control of various nmt promoters (BASI et al. 1993 ) and included elements of the mitotic control including Cdc2, Cdc13, Cdc25, Wee1, Mik1, and Pyp3; cdr gene products such as Cdr1/Nim1, Cdr2, and Spc1/Sty1; as well as the proteins of the Mlu1-binding factor complex (Cdc10, Res1, Res2, and Rep2; LOWNDES et al. 1992 ; TANAKA et al. 1992 ; CALIGIURI and BEACH 1993 ; MIYAMOTO et al. 1994 ; ZHU et al. 1994 ; NAKASHIMA et al. 1995 ). The functionality of the GFP fusions was established by the ability to rescue the mutant phenotypes. However, in the case of Mik1 and Pyp3, which do not exhibit a mutant phenotype by themselves, functionality was assessed by a delay and an acceleration of the cell cycle, respectively, when overexpressed. The cdr- phenotype of the sal3 disruptant was rescued by a net increase in Y-15 dephosphorylation since only overexpression of the Cdc25, Pyp3, and Cdr1/Nim1-GFP fusions under nmt41 regulation was capable of this rescue (Fig 5A). This argues strongly that the cell-elongated phenotype of the sal3 disruptant is due to a net deficiency in Y-15 dephosphorylation as a result of a failure to properly localize one of these regulators in the nucleus.
View larger version (72K):
In this window
In a new window
Download PPT slide
|
Figure 5.
(A) The cdr- phenotype of the sal3 disruptant is suppressed by an increase in Y-15 dephosphorylation. The sal3 disruptant transformed with various GFP-fusion constructs under control of the nmt41 promoter was grown to log phase in liquid EMM (top) and EMM - N media (bottom) at 25° for 24 and 48 hr, respectively. (B) Cellular localization of Sal3-GFP. The sal3-GFP integrant was grown in liquid YEA medium at 25° to log phase, fixed in methanol, and subjected to fluorescence microscopy. (C) Overexpression of Cdc25 suppresses the meiotic defect of the sal3-33h90 strain. The sal3-33h90 strain was transformed with either pREP41-GFP (top) or pREP41cdc25+-GFP (bottom), sporulated in liquid SPA medium, and tetrads were examined by light microscopy. (D) The proportions of zero- to four-spore tetrads from the sal3-33h90 strain containing pREP41-GFP (solid bars), pREP 41sal3+-GFP (lightly shaded bars), and pREP41cdc25+-GFP (dark shaded bars) are shown in the bar graph.
|
|
Intracellular localization of Sal3:
The cdr- phenotype of the sal3 disruptant was also suppressed by pREP41 sal3+-GFP and a chromosomally integrated version of sal3+-GFP (Fig 5A; data not shown). The subcellular localization of Sal3 was both nuclear and cytoplasmic with a perinuclear pattern consistent with its function as an importin-ß (Fig 5B). No changes in the subcellular localization of Sal3 were observed during the cell cycle, following nitrogen deprivation or under low pH (3.5) conditions (data not shown).
The meiotic defect of the sal3-33 homothallic strain is suppressed by overexpression of Cdc25:
A meiotic defect was observed in a homothallic sal3-33 strain where the majority of asci contained two spores (68%) and only 5% were four-spored asci (Fig 5C and Fig D). A similar defect has been reported in the sporulation of cdc25-22/cdc25-22 h-/h+ diploids and also following the meiotic induction of cdc25-22 haploids by ectopic expression of mei3+ (IINO et al. 1995 ). To determine whether the inability of sal3 mutants to complete meiosis is caused by a deficiency in Y-15 dephosphorylation, we expressed Cdc25-GFP under control of the nmt41 promoter in a homothallic sal3-33 strain. We found that overexpression of Cdc25 suppressed the meiotic defect of sal3 (Fig 5C and Fig D), reinforcing the regulatory role of Sal3 in Cdc25 activity. In agreement with this observation, the meiotic defect of sal3 was also suppressed by overexpression of Pyp3 and by Nim1 (data not shown).
Loss of sal3+ results in the nuclear mislocalization of Cdc25:
The subcellular localization of the various pREP41-based GFP fusions was either in agreement with published data or consistent with the molecular identity and function of the proteins (Table 3). Among the GFP-tagged proteins expressed from the pREP41-GFP vector, only Mcs1/Res2 showed a slight difference in localization between wild type and the sal3 disruptant (data not shown). The nuclear expression of Mcs1/Res2-GFP appeared more predominant in wild type than in the sal3 disruptant. However, this phenomenon was manifested only under overexpression conditions (data not shown) and is not the focus of this article.
Because our previous results supported a potential role for Sal3 in Cdc25 regulation, we were surprised that Cdc25 accumulated in the nucleus of the sal3 disruptant (Fig 6A). A possible explanation is that the constitutive overexpression from the pREP41-GFP plasmid may saturate the nuclear import machinery, allowing Cdc25 to enter the nucleus through other importin-ß's. To resolve this issue, the chromosomal cdc25+ was replaced with a GFP-tagged cdc25+, resulting in the latter being under control of the native cdc25 promoter (see MATERIALS AND METHODS). The cdc25-GFP integrant displayed the same size as wild type at 25° and 35° (Table 2 and Table 4), indicating normal activity of the tagged and functional protein. Nuclear accumulation of Cdc25 in wild-type cells was maximal in late G2 and early mitosis, decreased in anaphase, and had minimal levels in S phase (Fig 6A), consistent with the observations of LOPEZ-GIRONA et al. 1999 . In contrast, the sal3 disruptant did not show the nuclear accumulation of Cdc25 seen in wild-type cells, and the level of expression in the nucleus was notably lower than that in the cytoplasm (Fig 6A). The sal3::ura4+ cdc25-GFP integrant also displayed the same elongated phenotype as its untagged version (Table 2 and Table 4). Together, these results demonstrate that Sal3 is involved in the nuclear import of Cdc25 and that the mitotic defect in sal3 mutants is a consequence of Cdc25 mislocalization.
View larger version (50K):
In this window
In a new window
Download PPT slide
|
Figure 6.
Nuclear accumulation of Cdc25 is absent in the sal3 disruptant. Strains containing an integrated or a plasmid copy of the GFP-fusion construct were cultured to log phase in YEA and EMM media, respectively, at 25° for 20 hr. (A) Wild-type and sal3 disruptant strains expressing Cdc25-GFP from the pREP41-GFP vector (left) or an integrated copy (right). (B) Wild-type and sal3 cdc25-GFP integrants in a rad24 disruptant background. The septated wild-type and rad24 disruptant cells display a difference in Cdc25 nuclear localization as indicated by arrowheads. (C) Cellular localization of an integrated cdc13-GFP in cdc13-117 alone and in combination with a sal3 mutant background. The strains were grown in YEA + 25 µM thiamine at 35° to inhibit expression of the Cdc13 mutant protein.
|
|
We were next interested in whether the absence of Sal3 could affect the enhanced nuclear accumulation of Cdc25-GFP in rad24 mutants (LOPEZ-GIRONA et al. 1999 ). The rad24+ gene encodes a 14-3-3 protein that binds to several phosphorylated serines on Cdc25, resulting in a reduction or increase in nuclear import or export, respectively (FORD et al. 1994 ; ZENG and PIWNICA-WORMS 1999 ). In rad24 mutants, the nuclear localization of Cdc25 remains constitutively high throughout the cell cycle, including in septated cells (LOPEZ-GIRONA et al. 1999 ; Fig 6B, arrowheads). This leads to a semi-wee phenotype with a tapered morphology (FORD et al. 1994 ). We found that the enhanced nuclear localization of Cdc25-GFP in the rad24::ura4+ disruptant was abolished in the sal3 mutant background (Fig 6B). This indicated that the mode of Cdc25 nuclear concentration in rad24 cells is antagonized by the loss of sal3+. However, some Cdc25-GFP was present in the nucleus of the rad24::ura4+ sal3::ura4+ double disruptant since its level of nuclear Cdc25-GFP fluorescence appeared slightly higher than that of the sal3::ura4+ disruptant alone (Fig 6B). In addition, the tapered morphology of rad24 was exacerbated in a sal3 mutant background (Fig 6B).
To determine if the aberrant nuclear localization of Cdc25-GFP was specific to the sal3 disruptant, the subcellular localization of another mitotic regulator at endogenous levels of expression was examined. Cdc13 was chosen on the basis of its exclusive nuclear localization (BOOHER et al. 1989 ; DECOTTIGNIES et al. 2001 ). A pREP81 cdc13+-GFP plasmid was integrated into a cdc13-117 strain (see MATERIALS AND METHODS). The cdc13-GFP integrant was able to grow at the restrictive temperature (35°) but was slightly longer than wild type (15.0 µm compared to 13.3 µm for wild type). This was also true in a sal3 mutant background at 25° (27.9 µm compared to 25.3 µm for sal3::ura4+ alone). This indicated that the integrated cdc13-GFP had a somewhat reduced function. In both wild type and the sal3 disruptant, Cdc13 was localized in the nucleus and the protein was destroyed in late mitotic cells as expected (Fig 6C). Sal3 is not involved in the nuclear import of Cdc13.
Addition of an exogenous classical NLS to Cdc25-GFP advances mitotic entry and suppresses the cell cycle defect of the sal3 disruptant:
Previously, LOPEZ-GIRONA et al. 2001 identified the Cdc25 NLS, which did not display any sequence resemblance to cNLSs. This suggests that Cdc25 is imported into the nucleus via the nonclassical pathway by direct association with importin-ß and not by association with the importin-/importin-ß dimer. Therefore, the addition of an exogenous cNLS to Cdc25 should promote the translocation of Cdc25 into the nucleus of the sal3 disruptant by an alternate nuclear import route. We performed a C-terminal fusion of our GFP-tagged Cdc25 with a cNLS (Cdc25NLS-GFP) to determine whether this protein could be imported into the nucleus of the sal3 disruptant and suppress its cell cycle phenotype. The cdc25NLS-GFP integrant was constructed in the same manner as the cdc25-GFP integrant to ensure that expression of the tagged protein was solely under control of its native promoter (see MATERIALS AND METHODS). We observed that the cdc25NLS-GFP integrant displays an enhanced nuclear accumulation of Cdc25NLS-GFP and this is also seen in the sal3 disruptant (Fig 7A). This result indicates that the Sal3-mediated nuclear import of Cdc25 does not occur through the classical nuclear import pathway but instead via a nonclassical one. In addition, integration of cdc25NLS-GFP in the sal3 disruptant leads to a suppression of the cdr- phenotype (Fig 7B) as well as the cold-sensitive cell-elongation phenotypes on rich media and under high osmolarity and low pH stress (Table 4; data not shown). This further reinforces the conclusion that failure to localize Cdc25 appropriately is the primary reason for the cell cycle defect caused by loss of sal3+.
View larger version (45K):
In this window
In a new window
Download PPT slide
|
Figure 7.
Addition of an exogenous classical NLS on Cdc25-GFP suppresses the cell cycle defect of the sal3 disruptant. (A) Wild-type and the sal3 disruptant expressing an integrated GFP-tagged Cdc25 fused to an SV40 T-antigen NLS at the C terminus (Cdc25NLSGFP). The strains were grown to midlog phase in YEA at 30° and examined by fluorescent microscopy. (B) Wild-type and the sal3 disruptant expressing either an integrated normal GFP-tagged Cdc25 (top) or a GFP-tagged Cdc25 with an SV40 T-antigen NLS (bottom). The strains were grown in EMM - N for 2 days at 25° and examined by Nomarski. (C) Western blots of cdc25-GFP integrant strains grown to midlog phase in YEA at 30° and probed with anti-GFP (top) and anti-tubulin (bottom) antibodies. 1, cdc25GFPint; 2, cdc25NLSGFPint; 3, sal3::ura4+ cdc25GFPint; 4, sal3::ura4+ cdc25NLSGFPint.
|
|
Interestingly, we discovered that the enhanced nuclear accumulation of Cdc25NLS-GFP accelerates entry into mitosis since the cdc25NLS-GFP integrant is significantly shorter than wild type (Table 4). The ability of the cdc25NLS-GFP integrant to advance mitosis is not due to increased protein levels as a result of hyperstability of the tagged protein because Western blotting showed similar levels of protein compared to wild type (Fig 7C).
| DISCUSSION |
|---|
Identification of Sal3:
In this article, we have determined the molecular identity of the sal3+ gene and its function in the regulation of mitotic entry. The sal3+ gene product encodes an importin-ß that is involved in the nuclear import of Cdc25. An extensive attempt to clone sal3+ by plasmid complementation of two synthetic-lethal strains (sal3-33 cdr2-96 wee1-50 and sal3-33 cdc25-22r1 at 20° and 35°, respectively) was unsuccessful, yielding only multiple copies of cdc25. Cdc25-GFP can accumulate in the nucleus of a sal3 disruptant when overexpressed and suppresses its cell-elongated phenotype (Fig 5A and Fig 6A). This indicated that the constitutive overexpression of Cdc25 may allow it to enter the nucleus through other importin-ß's. We have discovered that overexpression of sal3+ causes the inhibition of cell growth (our unpublished data). This is likely the main reason why sal3+ could not be cloned by plasmid complementation since all the genomic libraries used were constructed in multicopy plasmids. In addition, the fact that these cloning strains are already compromised in overall Y-15 dephosphorylation activity without sal3-33 in their background also probably hindered cloning by this method. In this study, we demonstrate the utility of the insertional mutagenesis system in S. pombe as an alternative for gene retrieval. This is especially useful for complications that arise from cloning by plasmid complementation as in this case and for the isolation of loss-of-function suppressor genes.
Sal3 regulates Cdc25:
Several lines of evidence support the role of Sal3 as a nuclear import factor for Cdc25. First, genetic analysis revealed that sal3 mutants exhibit synthetic lethality with cdc25 alleles and various mutant backgrounds sensitive to partial cdc25+ gene activity. These synthetic-lethal interactions were suppressed by the stf1-1/cut12 mutation, which is able to compensate for the loss of cdc25+. Consistent with sal3+ functioning through cdc25+, the cell cycle defect of the sal3 disruptant was not observed in a cdc2-3w cdc25::ura4+ background. Second, multicopy plasmid suppressor studies determined that the cell-elongated phenotype of the sal3 disruptant is a manifestation of a net deficiency in Y-15 dephosphorylation, since overexpression of cdc25+, pyp3+, or nim1+ suppressed this phenotype. The sal3 meiotic defect that resembles that of cdc25 mutants was also suppressed by overexpression of these genes. Finally, fluorescence microscopy revealed that the nuclear accumulation of Cdc25 during the cell cycle is absent or very much reduced in sal3 mutants. Furthermore, the attachment of an exogenous cNLS to Cdc25 restored its nuclear accumulation in the sal3 disruptant and suppressed its cell cycle defect.
Examination of the nuclear localization of Cdc25-GFP determined that it oscillates during the cell cycle, accumulating in the nucleus throughout interphase and peaking at G2/M (Fig 6A). It then serves to activate nuclear Cdc2-Cdc13 complexes. Similar observations were reported by LOPEZ-GIRONA et al. 1999 . This pattern of nuclear Cdc25 accumulation during the cell cycle parallels the periodic fluctuations of Cdc25 protein levels (DUCOMMUN et al. 1990 ; MORENO et al. 1990 ). In anaphase, Cdc25-GFP displays a perinuclear localization (our unpublished data) similar to that observed for Cdc13 (DECOTTIGNIES et al. 2001 ), suggesting that Cdc25 is destroyed by the proteasome at this stage of the cell cycle. The rapid drop in Cdc25 protein levels in late M phase is consistent with this hypothesis (DUCOMMUN et al. 1990 ; MORENO et al. 1990 ).
Mechanism of Cdc25 nuclear transport by Sal3:
The cell elongation and nuclear exclusion of Cdc25-GFP displayed in the sal3 disruptant (Fig 6A) indicate that the nuclear accumulation of Cdc25 is necessary for the proper timing of mitotic entry in S. pombe. Previously, the nuclear import of Cdc25 was demonstrated to play an important role in the timing of mitosis and to be mediated by the nonclassical pathway (LOPEZ-GIRONA et al. 2001 ). The removal of three consecutive lysines (amino acids 212–214) with no sequence resemblance to cNLSs in Cdc25 resulted in nuclear exclusion and mitotic delay, which is suppressed by the expression of Cdc25 with an exogenous cNLS (LOPEZ-GIRONA et al. 2001 ). These observations imply that nuclear import of Cdc25 involves direct binding to importin-ß and is not through association with importin- as seen for cargo proteins containing cNLSs.
Our results reveal that Sal3 is the primary importin-ß involved in the nuclear import of Cdc25. Consistent with the observations of LOPEZ-GIRONA et al. 2001 , the sal3 disruptant displayed nuclear exclusion of Cdc25 and a mitotic delay with a comparable cell length to the cdc25 allele devoid of the nonclassical NLS (Table 4). We also demonstrated that the addition of an exogenous cNLS (identical to that used in the previous study) to wild-type Cdc25 restored the nuclear accumulation in the sal3 disruptant and suppressed its cell cycle defect. Although we cannot rule out the possibility that importin- plays a role in Cdc25 nuclear import, the evidence here points to Cdc25 nuclear import to be mediated primarily through a nonclassical route. In contrast, vertebrate Cdc25C, which exhibits a similar subcellular localization as S. pombe Cdc25, has been shown to be imported into the nucleus through the classical (importin--mediated) pathway (KUMAGAI and DUNPHY 1999 ). However, we have been unable to detect a two-hybrid interaction between full-length Sal3 and Cdc25 (our unpublished data). This may be the result of a Sal3 interaction with Cdc25 via an adapter protein such as importin-. Alternatively the phosphorylation state of the NLS, which has been shown to regulate binding to importins (HUBNER et al. 1997 ; BRIGGS et al. 1998 ), is unfavorable in a heterologous system.
We have not been able to biochemically detect a direct interaction between Sal3 and Cdc25 in vivo, using cell extracts from a strain endogenously coexpressing Sal3-GFP and Cdc25-HA. This could be due to several reasons. It is probable that only a very low proportion of the pool of Cdc25 is associated with Sal3 at steady state. These levels may be below the range of detection in our co-immunoprecipitation assays. In addition, the Sal3-Cdc25 interaction may be weak and transient, making the demonstration of an association difficult in our assays. Indeed, only a small number of nuclear proteins devoid of cNLSs have been purified in complexes containing importin-ß in budding yeast and none have been demonstrated to exhibit a direct interaction in vivo between the importin-ß and its cargo (KAFFMAN et al. 1998 ; CHAVES and BLOBEL 2001 ; MOSAMMAPARAST et al. 2001 , MOSAMMAPARAST et al. 2002 ). We cannot rule out that Sal3 is involved indirectly in the nuclear import of Cdc25, perhaps through the subcellular regulation of the cognate import factor.
The addition of an exogenous cNLS to Cdc25 in the experiments by LOPEZ-GIRONA et al. 2001 indicates that nuclear accumulation of Cdc25 per se is not sufficient for mitotic entry. Its constitutive nuclear localization at normal levels of expression fails to accelerate entry into mitosis. This is in contrast to our data, which showed that forced nuclear accumulation of Cdc25 by attaching the identical cNLS advances the onset of mitosis (Table 4). Similar to the other study, the expression level of our Cdc25NLS-GFP protein was equivalent to the wild-type GFP-tagged version, indicating that the accelerated entry into mitosis is not due to the hyperstability of this protein (Fig 7C). In addition, similar levels of Cdc25NLS-GFP expression reduced the cell elongation in the sal3 disruptant to a cell length significantly less than that of wild type at 35° (Table 4 and Fig 7C). This further supports the conclusion that nuclear accumulation of endogenous amounts of Cdc25 is sufficient for entry into mitosis. The discrepancy between these two studies may be attributed to the difference in the two Cdc25-fusion proteins containing the same cNLS. In our study, the cNLS was attached to a wild-type GFP-tagged Cdc25 that appeared to be fully functional because the cell length of this integrant was not significantly different from the untagged strain (compare Table 2 and Table 4). In contrast, the studies by LOPEZ-GIRONA et al. 2001 involved the fusion of the cNLS to a Myc-tagged Cdc25 with three internal consecutive lysine residues deleted. It is possible that the removal of these residues may have an additional effect on the catalytic activity of Cdc25, which would result in a slight cell cycle delay, thus influencing the overall timing of mitotic entry.
Effect of Rad24 on Cdc25 nuclear transport:
Our Cdc25-GFP protein displayed an enhanced nuclear localization in a rad24 mutant background (Fig 6B), consistent with the observations of LOPEZ-GIRONA et al. 1999 . The binding of Rad24 to phosphorylated serine residues on Cdc25 serves to keep the latter out of the nucleus (LOPEZ-GIRONA et al. 1999 ). Mutagenesis of several phosphorylated serines on Cdc25 results in a hindrance of Rad24 binding and nuclear exclusion of the mutant protein (ZENG and PIWNICA-WORMS 1999 ). In addition, overexpression of Rad24 excludes Cdc25 from the nucleus (ZENG and PIWNICA-WORMS 1999 ).
The nuclear exclusion of Cdc25 by Rad24 can be caused by either an increase or decrease in nuclear export and import rates, respectively, or both. Since Cdc25 reveals no obvious NES(s), it has been proposed that Rad24 provides the NES when complexed to Cdc25, thus enhancing translocation out of the nucleus (LOPEZ-GIRONA et al. 1999 ). This is supported by the observation that mutagenesis of the Rad24 NES impairs the nuclear exclusion of Cdc25 in response to irradiation (LOPEZ-GIRONA et al. 1999 ). However, similar studies demonstrate that disruption of the Rad24 NES cripples binding to Cdc25 in vivo and in vitro, suggesting rather that Rad24 may play a role in nuclear import (ZENG and PIWNICA-WORMS 1999 ). Nuclear import of Cdc25 may be mediated through direct binding to either importin- in a trimeric complex containing an importin-ß molecule or importin-ß alone. In vertebrates, the former appears to be the case since Cdc25 possesses an intrinsic NES(s) and Rad24 binding has been shown to inhibit the nuclear import of Cdc25 by disrupting its association with importin- (KUMAGAI and DUNPHY 1999 ; YANG et al. 1999 ; GRAVES et al. 2001 ).
We observed that the enhanced nuclear localization of Cdc25-GFP in the rad24 disruptant is abrogated in a sal3 mutant background (Fig 6B). However, the extent of nuclear exclusion in the sal3 rad24 double disruptant was not as severe as seen in the sal3 disruptant alone (Fig 6B). Similar to these observations, LOPEZ-GIRONA et al. 2001 found that the nuclear exclusion of Cdc25 by the removal of the putative NLS at amino acids 212–214 is abolished in a rad24 mutant background but not restored to the same degree as in wild-type Cdc25. This commonality leads us to speculate that the Sal3 import of Cdc25 into the nucleus may be mediated by its binding to this putative NLS. These results suggest that a low level of Cdc25-GFP is imported into the nucleus in the sal3 disruptant and a defect in Rad24-mediated nuclear export results in its accumulation in the nucleus. Alternatively, Rad24 may inhibit nuclear import through a weak NLS on Cdc25.
Cdc25 regulation of mitotic entry in response to nitrogen deprivation:
The cdr- phenotype and allosuppressor activity displayed by loss of sal3+ indicate that the sal3+ gene product is involved in linking nutrient status to mitotic control (Fig 1; NURSE and THURIAUX 1984 ). The sal3 disruptant's failure to properly downregulate its cell size in response to nitrogen deprivation is due to a mislocalization of Cdc25 to the cytosol since this defect can be suppressed by restoring Cdc25 accumulation in the nucleus (Fig 7B). Consistent with this, we have observed that the nuclear exclusion of Cdc25 by removal of the putative NLS also produces a cdr- phenotype (our unpublished data). Indeed, the isolation of an allele of cdc25 (sal2) in the same screen as sal3 for allosuppressors is unlikely to be coincidental (NURSE and THURIAUX 1984 ). Support for the role of nutritional modulation of the mitotic size control by Cdc25 has been demonstrated recently from the finding that Cdc25 is an effector of the cell size checkpoint that monitors the cell size threshold for mitotic entry (RUPES et al. 2001 ). The loss of regulatable Y-15 dephosphorylation in a cdc25 delete strain by expression of the T-cell PTPase delays mitotic entry in response to a nitrogen downshift (RUPES et al. 2001 ). This indicates that the reduction in the cell size threshold and acceleration of mitotic entry during nitrogen deprivation is caused in part by a net increase in Cdc25 activation and Y-15 dephosphorylation. The mechanism of Cdc25 activation in response to a nitrogen downshift remains unknown but it is unlikely to be mediated by changes in protein levels since the resetting of the size threshold is very rapid (FANTES and NURSE 1977 ; YOUNG and FANTES 1987 ). We show here that the nuclear translocation of Cdc25 participates to some extent in this process. The accumulation of Cdc25-GFP in the nucleus is never seen in the sal3 disruptant during nitrogen starvation (our unpublished data). Therefore, in this strain the absence of Cdc25 nuclear accumulation during nitrogen starvation likely results in a deficiency in Cdc25 activation in the nucleus, causing the failure to undergo the second cell division and G1 arrest. In contrast, the subcellular distribution of Cdc25-GFP in wild-type cells does not change significantly after a nitrogen downshift (our unpublished data). These results indicate that the nuclear accumulation of Cdc25 is necessary for the second cell division in response to a rapid nitrogen downshift.
Sal3 is involved in the nuclear import of a second mitotic regulator:
The observation that loss of sal3+ reverses the wee1 epistasis of cdc25 (Table 2) indicates that the cell cycle delay of sal3 mutants also occurs through a pathway independent of Wee1 and Cdc25. The acceleration of mitotic entry in the sal3 disruptant by inducing the nuclear accumulation of Cdc25 is significantly less than that in wild type at 25° but not at 35° (Table 4). This suggests that this second mitotic regulator is likely responsible for the cold-sensitive cell-elongated phenotype of sal3 mutants. However, the cell cycle delay of the sal3 disruptant is predominantly attributed to the mislocalization of Cdc25 because the cdc2-3w cdc25-22 epistasis of sal3::ura4+ is also seen at the lower temperature. One possible candidate for this secondary mitotic regulator is Mcs1/Res2, whose disruption also causes cold sensitivity, a cdr- phenotype, and the reversal of the wee1 epistasis of cdc25 (MIYAMOTO et al. 1994 ; our unpublished data). Furthermore, we observed that the nuclear localization of a GFP-tagged Mcs1/Res2 is impaired in the sal3 disruptant when overexpressed (our unpublished data).
In summary, we have demonstrated that the Sal3 importin-ß homolog in fission yeast is involved in cell cycle control by affecting the nuclear import of Cdc25. In vertebrates, importin-ß has been shown recently to inhibit spindle assembly by sequestering essential components of the spindle apparatus (reviewed in WALCZAK 2001 ). A further understanding of the regulation of these processes will provide new insights into importin-ß-mediated cell cycle control.
| ACKNOWLEDGMENTS |
|---|
We thank Drs. David Beach and Paul Russell for providing strains used in this study. This work was supported by grants from the Natural Science and Engineering Research Council of Canada to P.G.Y.
Manuscript received June 14, 2002; Accepted for publication July 22, 2002.
| LITERATURE CITED |
|---|
TOP
ABSTRACT
MATERIALS AND METHODS