We have isolated the Schizosaccharomyces pombe orthologue of the Saccharomyces cerevisiae MOB1 gene in a screen designed to enrich for septation mutants. The gene is essential, and cells lacking it display a phenotype typical of septation signalling network mutants. mob1p is located on both spindle pole bodies throughout mitosis. In addition it is also co-localised with the medial ring later in mitosis, and flanks the septum as the medial ring contracts. We also demonstrate that mob1p can be precipitated from cells in a complex with the septation regulating kinase sid2p.
The yeast S. pombe provides us with a relatively simple model for the study of cytokinesis and its co-ordination with other events in mitosis. Fission yeast grows mainly by tip elongation, and then divides by binary fission after formation of a division septum. The position of the septum is defined in late G2, or as a very early event in mitosis, but septum synthesis is not initiated until the end of mitosis. Mutants that are arrested in early mitosis have defined the position of the septum, but do not initiate its synthesis (Chang et al., 1996). A large number of S. pombe mutants that are defective in positioning the septum, assembly of the medial ring, or signalling the onset of septation have been identified (for reviews, see Gould and Simanis, 1997; Le Goff et al., 1999). One group of mutants fail to make a septum, becoming elongated and multinucleate. These include the genes cdc7, cdc11, cdc14 (Nurse et al., 1976) spg1 (Schmidt et al., 1997), plo1 (Ohkura et al., 1995), and sid1, sid2, and sid4 (Balasubramanian et al., 1998). Cloning of these genes and identification of their products has revealed that the spg1 gene encodes a ras-superfamily GTPase. Four of the genes encode protein kinases: cdc7 (Fankhauser and Simanis, 1994), sid1 (cited by Balasubramanian et al., 1998), plo1 (Ohkura et al., 1995), and sid2 (Sparks et al., 1999), leading to the suggestion that they form a signal transduction cascade. They are collectively referred to hereafter as the septation signalling network. The biochemical functions of cdc14p and sid4p are presently unknown. All of these genes (except cdc11, which has not been cloned to date) have been demonstrated to be essential (reviewed by Le Goff et al., 1999).
Previous studies have suggested that the spg1p GTPase switch is an important element in signalling the onset of septum formation in fission yeast (Schmidt et al., 1997). Spg1p is regulated negatively by byr4p and cdc16p, which together form a GTPase-activating (GAP) protein (Furge et al., 1998). Both cdc16 and byr4 are essential genes (Fankhauser et al., 1993; Song et al., 1996) and loss-of-function mutants undergo multiple rounds of septum formation without cell cleavage (Minet et al., 1979; Song et al., 1996). Spg1p interacts with the protein kinase cdc7p, and increased expression of cdc7p can rescue a null allele of spg1, suggesting that the essential function of spg1p is mediated via cdc7p (Schmidt et al., 1997). The protein kinase cdc7p is distributed asymmetrically on the spindle pole bodies during anaphase (Sohrmann et al., 1998), which is likely to be mediated by recruitment of the byr4p/cdc16p GAP at one pole before the other to inactivate spg1p signalling (Cerutti and Simanis, 1999). Sid2p is located at the spindle pole body at all stages of the cell cycle and is seen at the medial ring after anaphase. It then forms a double ring flanking the developing septum (Sparks et al., 1999). Plo1p appears at the spindle pole body at the onset of mitosis, and is subsequently observed on the spindle (Mulvihill et al., 1999).
Together, these studies have emphasised the importance of the spindle pole body as an integrator of cell cycle events. A major question that remains to be answered is how the signal is transduced from the spindle pole bodies, via the septation signalling network, to the medial ring to activate septum synthesis at the end of anaphase. We have used a screen designed to identify genes implicated in septation signalling. In this paper we present the identification and analysis of the S. pombe orthologue of the S. cerevisiae MOB1 gene.
MATERIALS AND METHODS
Standard genetic techniques were used for construction and growth of strains (Moreno et al., 1991). Cells were propagated on yeast extract medium (YE) or EMM2 minimal, containing appropriate supplements.
Cloning of mob1 and construction of mob1-containing plasmids
The oligonucleotides used for tagging of proteins and deletion of genes are listed below. Tagging and deletion of mob1 were performed as described by Bahler et al. (1998b). The 5’ truncated clone of mob1 was obtained by transformation of SP2198 with a fission yeast genomic library, and selecting for leucine prototrophs directly at 25°C. Plasmids were recovered from these transformants, and the insert sequenced. Full length clones were obtained by colony screening a genomic library of fission yeast (Barbet et al., 1992). To obtain a cDNA, the gene was amplified from a wild-type fission yeast cDNA library in pREP3 (Kelly et al., 1993), using primers which insert BamHI sites 5’ and 3’ to the ORF. After amplification, the product was digested with BamHI and cloned into pDW232 and pREP vectors at the BamHI site. The amplified clone was sequenced to confirm the removal of the two predicted introns, and the absence of amplification-induced mutations.
The primers used for genomic and cDNA amplification were: Forward primer: TTGGATCCGCATGTTTGGATTTAGTAATAA-GAC. Reverse primer: CGGGATCCTTAAACCATGCTATCTAC-CAAGTCT. The BamHI sites are underlined.
Construction of mob1-null and chromosomally epitope-tagged strains
The mob1 null allele and C-terminally tagged (GFP and HA3) strains were constructed by direct chromosomal integration of PCR fragments generated using plasmid pFA6a-kanMX6 as a template (Bahler et al., 1998b): forward primer for deletion: GTTGGCACA-AACCAAGCCTACCAGAACCCAACTCATACCAACTTTTTTGC-ATTTTCTTTCCGGATCCCCGGGTTAATTAA; tagging forward: TTCAGCCTTATGGATAACAAGGAATATGCACCAATGCAAGAC-TTGGTAGATAGCATGGTTCGGATCCCCGGGTTAATTAA; tagging/deletion reverse: TAAGTTTGGGAACCGAGTTTACCTCC-CATTTCCTGTTCCTGAATAACTACAAAACTGTTCGAATTCG-AGCTCGTTTAAAC. The underlined regions are the homology with the pFA6a-kanMX6 vector.
The PCR fragment was purified and transformed into either an ade6-M210/ade6M-216 leu1-32/leu1-32 ura4-D18/ura4-D18 h−/h+ diploid (for deletion) or leu1-32 h− (for tagging) strains following the LiAc protocol described by Bahler et al. (1998b). After overnight growth in YE, transformants were selected on YE G418 plates (100 mg/l). Correct integration was verified by Southern blotting and PCR amplification. The expression of tagged proteins was confirmed by western blotting.
Cells were stained with DAPI (4’,6’-diamino-2-phenylindole, Sigma) and Calcofluor (fluorescent brightener no. 28, Sigma) after fixation in 70% ethanol, as described previously. Indirect immunofluorescence was performed as described (Balasubramanian et al., 1997; Moreno et al., 1991). Exponentially growing cells were fixed with 4% formaldehyde (37% solution; Sigma F-1268) for 7 minutes at 25°C. To examine interphase microtubules glutaraldehyde was also added to a final concentration of 0.2%. Cells were digested with zymolyase (1 mg/ml) for 20 minutes at 37°C, permeabilised with 1% Triton X-100 and incubated o/n at room temperature with primary antibody.
The following antisera were used: polyclonal rabbit anti-GFP 1:100 (a gift from Paul Nurse and Ken Sawin); anti-cdc7 (1:40; Fankhauser and Simanis, 1994); dmf1/mid1(1:40; Sohrmann et al., 1996), spg1 (1:40; Schmidt et al., 1997), cdc15 (1:50; Fankhauser et al., 1995); sad1 (1:500; Hagan and Yanagida, 1995); monoclonal anti-myc 9E10 (1:100) or anti-HA HA11 (1:500), TAT1 (1:50; Woods et al., 1989). Primary antibodies were detected with goat anti-rabbit FITC used at 1:500 (Kappel) or goat anti-mouse CY3 1:500 (Jackson) or goat anti-mouse FITC 1:500 (Sigma).
F-actin was stained with rhodamine-conjugated phalloidin (Molecular Probes; R-415). After fixation with formaldehyde, cells were permeabilised with Triton X-100, and incubated with rhodamine-conjugated phalloidin (∼1.5 units) for 30 minutes-overnight at room temperature.
Immunoprecipitations and western blot analysis
Protein extracts were prepared from 3-5×108 cells in exponential phase, that had been collected by centrifugation and frozen on dry ice. All subsequent manipulations were done on ice or in the cold room (4°C). For western blotting total protein extracts were prepared as described by Moreno et al. (1989) by vortexing vigorously with glass beads (Sigma, G-9268).
For immunoprecipitation, soluble protein extracts were prepared by vortexing with glass beads in HEN buffer (50 mM HEPES, pH 8.0, 150 mM NaCl, 5 mM EDTA, 1 mM EGTA, 50 mM β-glycerophosphate with inhibitors: 0.1 mM sodium orthovanadate, 50 μg/ml leupeptin, 1% apoprotinin; 1 mM DTT, 1 mM PMSF). Beads were washed by brief vortexing in the same buffer containing 1% NP-40. Cell extracts were clarified by two successive centrifugations. Protein concentration was measured using Bradford assay (Bio-Rad). For each immunoprecipitation, 2-3 mg of soluble protein was incubated overnight with 10 μl of either 9E10 or 12CA5 monoclonal antibodies covalently coupled to the Sepharose-Protein G beads (Sigma; P3296; ∼2 μg of Ab). Beads were washed 3 times with 1 ml of HEN-NP-40 buffer (by pelleting in a microfuge for 5 seconds), resuspended in FRB loading buffer.
For detection of HA, Myc or GFP epitope tagged proteins from total and soluble protein extracts or immunoprecipitates, cell extracts or immunoprecipitates were separated on SDS-PAGE (10%), transferred to GeneScreen Plus hybridization transfer membranes (NEN, Life Science Products) using a wet blotting apparatus (Bio-Rad). Blots were probed with 12CA5, 9E10 or polyclonal rabbit anti-GFP antibodies respectively at 1:1000 dilution and detected with anti-mouse or anti-rabbit IgG HRP conjugates (Promega) at 1:2500 dilution using Amersham Life Science ECL system.
Isolation of the S. pombe mob1 gene, and identification of mob1p
To identify new genes involved in septation signalling, we used the strain SP2198, in which a temperature-sensitive mutation in the cdc7 gene rescues the lethality of a cdc16::ura4+ null allele by attenuation of septation signalling (described by Beltraminelli et al., 1999; N. Fournier et al., unpublished). These cells are able to divide and form colonies at 32°C, but not at 25°C, where they become multiseptate. SP2198 was transformed to leucine prototrophy with a wild-type S. pombe DNA library, and cells were plated directly at 25°C. Colonies were purified by restreaking at 25°C, and plasmids were rescued from those that showed plasmid-dependence for rescue of the cold-sensitive multiseptate phenotype. Among the plasmids obtained was a clone, which was found to encode the S. pombe homologue of the budding yeast gene MOB1. However, the mob1 gene in this plasmid lacked the first 39 amino acids of the open reading frame. A full length clone was isolated by screening both genomic and cDNA libraries. Sequencing of cDNAs confirmed the presence of the two introns predicted by analysis of the sequence. The gene is located on cosmid 428 on chromosome II. Introduction of the full-length mob1 gene into SP2198 did not rescue the cold-sensitivity (Fig. 1A), suggesting that the truncated protein may attenuate septation signalling by acting as a dominant-negative mutant. An alignment of the predicted sequence of the S. pombe mob1p with S. cerevisiae Mob1p is shown in Fig. 1B. Further analysis of the S. pombe database at the Sanger center identified a protein similar to S. cerevisiae Mob2p (cosmid 970, chromosome III). This paper describes our analysis of the S. pombe mob1 gene, and its role in septation. We note that the existence of a mob1 homologue in S. pombe has been mentioned previously in the discussion of Sparks et al. (1999).
To identify the mob1 protein (mob1p), we tagged the mob1 open reading frame at its C terminus by addition of three copies of the 12CA5 epitope (HA). The HA-tagged copy was crossed into a cdc25-22 mutant and cells were synchronised by arrest-release. Western blotting of protein samples taken at intervals after release from the cdc25 block showed that the steady state level of mob1p did not change significantly (<2-fold) as cells progressed from G2 through mitosis, septum formation and cytokinesis (Fig. 1C). Though the protein migrates as a doublet on gels, no significant change in the ratio of the two bands was observed. Analysis of total RNA extracted from cells synchronised by cdc25-22 arrest-release showed that the level of mob1 RNA did not change through the cell cycle (data not shown).
Isolation of thermosensitive alleles of the mob1 gene
In addition to transformation of SP2198 with a genomic library, we also selected for chromosomal mutations which are able to rescue the cold-sensitivity of this strain. Analysis of these mutants (to be described elsewhere; N. Fournier et al, unpublished), revealed that one of the complementation groups gave wild-type progeny when crossed with mutants in sid1, sid2, sid4, spg1, cdc7, cdc11, and cdc14, or a marked allele of the plo1 gene, indicating that it did not map to any of the genes previously implicated in septation signalling. Complementation of one of the mutants in this group (R4) with a wild-type library revealed that the rescuing plasmids all encoded the S. pombe mob1 gene. Crosses to the HA-tagged copy of the mob1 gene, marked with the G418 resistance cassette (kan-MX6) did not produce any recombinants, indicating that the R4 mutant is an allele of the mob1 gene.
Crosses among the 5 mutants produced fewer than 1 in 104 wild-type progeny when >106 spores were plated directly at 36°C, indicating that they were all alleles of the mob1 gene. These five alleles, named J2, M17, N1, R4, and R17 exhibited similar phenotypes, and are typified by mob1-R4, which is shown in Fig. 2A. At 25°C, cells divide normally, and in contrast to S. cerevisiae mob1 mutants (Luca and Winey, 1998), they do not show an increased tendency to diploidise, as judged by FACS analysis and appearance of colonies on Phloxin B containing media (data not shown). In liquid culture, five hours after shift to 36°C, mob1-R4 cells display defects in septum formation (Fig. 2A). Cells are frequently binucleate, without a septum, and with the nuclei in a postmitotic configuration. In addition, tetranucleate cells consisting of two binucleate compartments separated by one or more septa were also present. No significant nuclear abnormalities were seen as judged by staining with DAPI. Unseptated cells containing more than 4 nuclei were only observed rarely; however approximately 50% of cells had lysed by 5 hours after shift to 36°C. Together, these data suggest a role for mob1p in signalling the onset of septum formation.
Reduction of mob1 activity suppresses the lethality of cdc16 and byr4 null mutants
The mob1 alleles were all obtained in a genetic background containing the cdc7-A20 mutant. Thus, these mutants could be suppressors of cdc16::ura4+ by themselves, or require reduced cdc7p activity to suppress the phenotype of cdc16::ura4+. This question was addressed by crossing the mutant mob1-R4 to SP2198, and analysing ura+ progeny which were viable at 19°C. Two classes were obtained. The first died as elongated, lysed cells when replica plated to 25°C, which was similar to the phenotype of the original triple mutant cdc16::ura4+cdc7-A20 mob1-R4 (the combination of cdc7-A20 and mob1-R4 dies at 25°C; see below). The second was viable at all temperatures tested (19°C to 36°C). Backcrossing to wild-type demonstrated that the genotype of this second class was cdc16::ura4+mob1-R4, establishing that the mob1-R4 allele functions as a bypass suppressor of the cdc16 null allele. Similar data were obtained for the other four alleles (not shown). The colonies formed by these cells were darker than wild-type on medium containing the vitality stain Phloxin B indicating the presence of an elevated number of dead cells. Examination of the cells growing in liquid culture showed an elevated septation index and errors of septation at all temperatures, though these were most abundant at 19°C and 36°C (Fig. 2B). The most frequent septation defect observed was the presence of mono-nucleate, septated cells, reminiscent of type II cells produced following loss of cdc16p function (see Minet et al., 1979). Occasional septa which had been synthesised at an aberrant angle were also observed. Thus, though these mob1 mutants can rescue loss of cdc16 function, they do not restore completely normal control of septum formation.
We also tested whether these mob1 alleles could rescue a null allele of byr4, the other subunit of the spg1p GAP. The heat-sensitive alleles of mob1 were crossed to a diploid byr4::ura4+/byr4 + ade6-M210/ade6-M210 leu1-32/leu1-32 ura4-D18/ura4-D18 h+/h+ (see Materials and Methods for derivation of this strain). Haploid, ura+ progeny were obtained, after plating spores at 29°C. The absence of the wild-type byr4+ allele was verified by Southern blotting of genomic DNA (Fig. 2C). All five alleles of mob1 were found to suppress the byr4::ura4+ allele. The phenotype of the cells was similar to those suppressing the cdc16::ura4+ allele, described above (not shown). We conclude that attenuation of mob1 activity will bypass the requirement for cdc16p and byr4p function.
Increased expression of mob1 results in defects in septum formation and mitosis
Next, we examined the effect of increased expression of the mob1 gene from the full-strength nmt1 promoter in wild-type cells. Approximately 6 generations at 25°C (26 hours) after removal of thiamine from the medium (the promoter is turned on after 3-3.5 generations), cells displayed a variety of abnormalities of both septum formation and mitosis, including failure to septate (Fig. 3A; cells marked 1), formation of one or more septa without cell cleavage (Fig. 3A; cells marked 2), condensed chromosomes (Fig. 2B, cells marked 3); many cells also formed thick, aberrant septa (Fig. 3A, cells marked 4). Deposition of a broad band of septum material in the middle of the cell was the first phenotype observed, as early as 18 hours (approximately 4.5 generations) after induction. The mitotic phenotypes were observed approximately one generation-time later, suggesting that higher doses of mob1p are required to generate the mitotic abnormalities. Some of these phenotypes resemble those produced by the mob1-R4 mutant (Fig. 2A) and indicate that an excess of mob1p interferes with cell cycle progression.
We investigated whether increased expression of mob1 from the full-strength nmt1 promoter, or the attenuated form in the REP41 vector (Basi et al., 1993), could rescue any of the other septation signalling mutants. Transformation of sid1, sid2, sid4, cdc7, cdc11, cdc14, and cdc16 mutants showed that none was rescued at 36°C, whether expression from either vector was induced or not (data not shown). Interestingly, the spg1-B8 mutant was partially rescued at 36°C by induced pREP3-mob1 (Fig. 3B).
When plasmids encoding either cdc15, cdc7, cdc14, sid1, sid2, spg1, dma1, plo1 or zfs1 were introduced into mob1-R4, either expressed from their own promoter, or from the available versions of the nmt1 promoter (Basi et al., 1993), none was able to rescue the heat-sensitivity of mob1-R4 (data not shown). In S. cerevisiae, mob1p was first identified as a protein interacting with the protein kinase Mps1p in a two-hybrid screen (Luca and Winey, 1998). Using the two-hybrid system, we have not detected any significant interactions of S. pombe mob1p with mph1p (He et al., 1998), the fission yeast homologue of MPS1 (data not shown).
mob1 is an essential gene, and is implicated in signalling the initiation of septum formation
The wide variety of phenotypes after shifting the heat-sensitive mutant to 36°C suggested that the alleles selected in this screen might be hypomorphs. To test this, and examine the phenotype of a null mutant, of one copy of the mob1 gene was replaced in a diploid by a PCR-amplified G418 resistance marker (Bahler et al., 1998b). Correct integration of the cassette was verified by Southern blotting (Fig. 4A). Dissection of 20 tetrads derived from this diploid produced only two viable progeny per tetrad, which were sensitive to G418, demonstrating that the mob1 gene is essential for colony formation. Microscopic examination of germinating mob1::kanMX6 spores showed that they became highly elongated but were unable to divide. Staining with DAPI and Calcofluor showed that the cells contained multiple nuclei, but did not form a septum (Fig. 4B). Thus, the mob1 gene product is not required for the nuclear cycle, but is essential for cell division. We also conclude that all of the mob1 alleles we isolated in this screen are hypomorphs.
Next, we investigated whether components already known to be essential for medial ring formation and/or positioning and septation signalling were correctly localised in the absence of mob1p. Examination of F-actin distribution showed that medial rings were formed during mitosis, and F-actin patches were present at the growing tips in interphase (Fig. 4C). Staining of microtubules indicated that interphase arrays and mitotic spindles were present in the population (Fig. 4D), suggesting that there were no major problems with the F-actin or microtubule cytoskeletons. Staining for mid1p showed the expected nuclear localisation in interphase (Fig. 4E, right panels), and ring formation in mitotic cells (Fig. 4E, left panels). The spindle pole body component sad1p was visible on all nuclei (Fig. 4F), and in anaphase cells, cdc7p was associated with the spindle pole bodies of half the nuclei (Fig. 4G), as described previously (Sohrmann et al., 1998). Finally, the medial ring component cdc15p, which is essential for septation, also formed a ring in the absence of mob1p (Fig. 4H). We therefore conclude that medial rings seem to be assembled correctly, and that asymmetric segregation of also cdc7p occurs normally in the absence of mob1p. Transformation of the diploid heterozygous for the mob1::kanMX6 allele with plasmids encoding either the full-length mob1 or the original 5’-truncated clone demonstrated that both could rescue the null. However, when the deletion was covered by the truncated clone, the colonies formed were slower-growing, and contained many elongated and multinucleated cells, indicating that the truncated protein was not fully functional (not shown).
In addition to the full deletion of mob1, we also constructed a disruption allele by insertion of the ura4+ gene into the single EcoRI site near the 3’-end of the ORF, leaving the gene capable of encoding the first 137 of 210 amino acids. In contrast to the full deletion, where the viability of the germinating spores was high (>90%), five independent diploids examined all showed reduced spore viability. In most tetrads, only one or zero spores could germinate. Since the diploid contains one wild-type copy of the mob1 gene, and analysis of the full deletion indicates that halving the gene dosage of mob1 does not have any serious meiotic effects, the partial deletion may have a dominant negative effect during meiosis. Future studies will address the basis for this.
Localisation of mob1p
A number of components of the septum signalling network have been localised to the spindle pole body (cdc7p and spg1p; Sohrmann et al., 1998); sid2p (Sparks et al., 1999); cdc16p and byr4p (Cerutti and Simanis, 1999). To examine its subcellular localisation mob1p was tagged by addition of GFP to the C terminus. Indirect immunofluorescence of fixed cells, or direct observation of living cells for GFP fluorescence revealed the mob1p can be observed at several discrete locations during mitosis and cytokinesis. In early mitotic cells, two dots were observed associated with the nucleus (Fig. 5A). Double staining for the mitotic spindle revealed that the dots were present at the ends, suggesting that they correspond to the spindle pole bodies (Fig. 5B). This was confirmed by costaining for sid2p (Sparks et al., 1999), which is a known spindle pole antigen (Fig. 5C). Interestingly, though it has been reported (Sparks et al., 1999) and confirmed by us (data not shown) that sid2p-GFP is located on the spindle pole body throughout interphase, the sid2p-myc13 tagged protein was absent from the spindle pole body in interphase cells (Fig. 5C). It is possible that the myc13 epitope on the protein is masked by interaction with another protein during interphase, while the GFP fluorescence remains visible. Future studies will address this issue.
Later in mitosis, in addition to the staining observed at both spindle poles, a medial ring of mob1p fluorescence was observed at the cell equator. Costaining with a tagged copy of cdc15p identified this as part of the medial ring (Fig. 6A). At the time of septum formation, when the medial ring contracts, mob1p remained visible, now forming a double ring at the cell equator (Fig. 5A). Staining for cdc15p no longer showed any overlap, with the cdc15p staining sandwiched between the two mob1p rings. As septum synthesis progressed, the double rings did not contract. Once the septum had been completed, the mob1p ring staining was no longer observed, though the spindle pole body staining was still present (Fig. 5A). Newly-separated (early G2) cells showed a strong decrease in staining at the spindle pole body (Fig. 5A), which was no longer detectable in larger interphase cells, where only a faint cytoplasmic signal could be seen (Fig. 5A). This cytoplasmic signal was seen at all stages of the cell cycle, suggesting that there is a significant cytoplasmic pool of mob1p.
No mob1p-GFP rings were observed when asynchronous cells were treated for 10 minutes with latrunculin A to depolymerise F-actin (Ayscough, 1998), or with 300 μg/ml thiabendazole (Petersen et al., 1998) to depolymerise microtubules. However, in both cases spindle pole body staining was unaffected (not shown). Thus, formation and/or maintenance of mob1p rings requires both an intact microtubule and F-actin cytoskeleton, while maintenance of mob1p at the spindle pole body does not.
The tagged mob1p-GFPc allele was crossed into different mutant backgrounds and the localisation of the protein examined in living or fixed cells. In nda3-KM311 cells, which are arrested in early mitosis, and have formed a medial ring (Chang et al., 1996), mob1p was on the spindle pole bodies, but not in the medial ring (Fig. 6B). Thus, as is the case for sid2p (Sparks et al., 1999), ring formation seems to require an anaphase event, and/or an intact microtubule cytoskeleton.
Next, we examined whether the localisation to mob1p was dependent upon any of the other septation signalling genes. Mutants were grown at 25°C and shifted to the restrictive temperature for 5 hours, and the localisation of mob1p was examined in mitotic cells. No effect upon spindle pole body localisation during mitosis was observed in sid2-250 (Fig. 6C). Similar results were obtained in cdc14-118, sid1-239, and cdc11-136 mutants (not shown). The spindle pole body signal was greatly reduced in intensity in cdc7-24, spg1-B8, and sid4-SA1 mutants at 36°C. We have been unable to assess whether a temperature shift has any effect upon the mob1p medial ring in these backgrounds, since the structure seems to be very labile at high temperatures even in a wild-type background, and we are unable to detect the ring in either fixed or living cells following shift to 32°C or 36°C. However, we noted that mob1p rings were scarcely detectable in sid2-250 even at 25°C (Fig. 6C), suggesting that formation of a stable mob1p ring requires full sid2p function. Other mutants did not show this effect at the permissive temperature (data not shown). To examine whether mob1p function was required for correct localisation of sid2p, its localisation was examined in a sid2pGFP mob1-M17 mutant (the mob1-R4 mutant could not be used for this study, see below).
At 25°C, sid2p-GFP spindle pole body staining was normal, but very few sid2p-GFP rings were detectable, and those which were seen were much fainter than in a wild-type background. We also noted that the spindle pole body signal of sid2p-GFP was absent, or greatly reduced in intensity at 36°C. Thus, mob1p function is required for stable association of sid2p with the spindle pole body and for sid2p ring formation.
Studies of sid2 has shown that formation of the sid2p ring requires cdc15p function (Sparks et al., 1999). Analysis of mob1p localisation in the mutant cdc15-140 mob1-GFP showed that at 25°C and 29°C mob1p rings could be observed. However, fewer of the double rings at the time of septum formation were seen, suggesting that the stability of these structures is decreased by partial loss of cdc15p function (not shown).
Interaction of mob1p with sid2p
In S. cerevisiae, it has been reported that Mob1p interacts physically with the protein kinase Dbf2p (Komarnitsky et al., 1998), and the protein kinase Mps1p (Luca and Winey, 1998). To date, we have not detected any interactions between mob1p and mph1p, as judged by two-hybrid analysis. The localisation of mob1p is very similar to that reported for sid2p (Sparks et al., 1999), with the exception that we do not detect mob1p on the spindle pole body in interphase cells. This prompted us to examine whether a mob1p-sid2p complex could be detected in S. pombe cells. Protein extracts were prepared from cells containing single tagged copies of both sid2p and mob1p, expressed from their own promoters. Immunoprecipitates using 12CA5 from a mob1-HAc sid2-GFPc strain contained sid2p-GFPc (Fig. 7A). In the reverse experiment, immunoprecipitates using 9E10 from a mob1-GFPc sid2-myc13 strain were found to contain mob1p-GFPc (Fig. 7B). Thus, mob1p and sid2p can be precipitated as a complex from cell extracts. The ability of the two proteins to interact in vivo was confirmed by two-hybrid analysis (data not shown).
Genetic interactions of mob1-R4 with other septation signalling network mutants
Tetrad dissection was used to analyse the genetic interactions of mob1-R4 with other genes implicated in septation signalling. To maximise the likelihood of any double mutant being viable, spores were germinated at 19°C. It was found that the double mutants of mob1-R4 with sid2-1, sid2-250, sid4-SA1, cdc11-
136 and cdc14-118 were synthetically lethal. Spores germinated, but were unable to form a colony, lysing either in the first cycle or after a few divisions. Double mutants of mob1-R4 with either cdc7-24, spg1-B8, cdc16-116, or cdc4-8 formed very small, slow-growing colonies containing many dead cells. The double mutants with either cdc7-A20, sid1-239, myo2-E1, or rng2-D5 showed a greatly reduced restrictive temperature (27°C or lower) compared with either parent strain. In contrast, when double mutants with inhibitors of septum formation were constructed, it was found that the null allele of dma1 gave weak suppression of mob1-R4, while the zfs1 null allowed mob1-R4 to form colonies even at 36°C. No significant genetic interactions were observed with the cdc15-140 mutant. Interestingly, we also found that the double mutants sid2-myc13mob1-R4 and sid2-GFP mob1-R4 formed colonies at 36°C. The same was true for the equivalent double mutants with mob1-R17. This rescue was allele-specific; the mob1 alleles M17, N1, and J2 alleles were not rescued (not shown). Thus though these tagged versions of sid2p appear to function normally in a wild-type cell (Sparks et al., 1999), their interactions with other proteins do not completely mimic those of wild-type sid2p.
We have isolated heat-sensitive mutants of the mob1 gene in a screen designed to bias for mutants defective in signalling the onset of septum formation. Like its counterpart in S. cerevisiae, the S. pombe mob1 gene is essential; spores lacking mob1p function germinate and die as elongated, multinucleated cells.
Two components of the medial ring are correctly localised during mitosis, and the asymmetric segregation of cdc7p on spindle pole bodies during mitosis occurs normally. It therefore seems likely that cells lacking mob1p are defective in signalling the onset of septum formation. In contrast to the S. cerevisiae mob1 mutants, none of the thermosensitive alleles we have isolated diploidises at an elevated rate at 25°C.
There are strong genetic interactions between mob1 and cdc7, paralleling those observed between MOB1 and CDC15 (Luca and Winey, 1998). However, in contrast to the situation in S. cerevisiae, increased expression of sid2 will not rescue heat-sensitive alleles of mob1, and increased expression of mob1 will not rescue heat sensitive alleles of sid2. Heat-sensitive alleles of mob1 show strong genetic interactions with other septation signalling genes, and are rescued by deletion of the zfs1 gene, suggesting that mob1 functions in the same signal transduction network as the other septation signalling genes (see Introduction).
We have demonstrated that mob1p forms a complex with the protein kinase sid2p. The existence of a complex of Mob1p and Dbf2p has been demonstrated in S. cerevisiae (Komarnitsky et al., 1998). Partial impairment of sid2 function seems to reduce the stability of mob1p association with the medial ring and septum, but sid2p function does not seem to be required for association of mob1p with the spindle pole body. Loss of function of mob1p affects both medial ring and spindle pole body association of sid2p. Whether mob1p acts as a targeting subunit, or a substrate of sid2p will be the subject of future studies.
Localisation of biologically functional, tagged copies of mob1p has demonstrated that the protein is found on both poles of the mitotic spindle. Mutations in sid4, spg1, and cdc7 strongly reduce, but do not completely eliminate mob1p spindle pole body staining, suggesting that the activity of these three proteins is required for efficient spindle pole body association of mob1p. The cdc7p/spg1p kinase/GTPase switch appears to be a rate-limiting step in signalling the onset of septum formation (Schmidt et al., 1997; Sohrmann et al., 1998). It is possible that these proteins must be activated for mob1p to associate in a stable manner with the spindle pole body. Whether mob1p and/or sid2p are substrates of cdc7p will be the subject of future studies. It has been suggested that sid4p functions downstream of cdc7p/spg1p, but the sequence of sid4p gives no clues to its function (Balasubramanian et al., 1998). The biological function of sid4p in regulating septation awaits further analysis.
The presence of mob1p on the spindle pole body in a sid2 mutant at the restrictive temperature is surprising, since the two proteins can be found as a complex in cell extracts, and inactivation of mob1p reduces sid2p association with the spindle pole body. Several explanations can be envisaged: one is that the sid2-250 allele is only defective in certain aspects of its function, and that it retains the ability to permit spindle pole body association of mob1p. This seems unlikely, since another allele, sid2-1, produces a similar result (not shown). Alternatively, since sid2p-GFP is found on the spindle pole body throughout interphase as well as mitosis, it may be that mob1p can be recruited to the spindle pole body independently of sid2p function, and that a complex of the two proteins is only formed once they are both at the spindle pole body, which would therefore act as an assembly site for the complex. The partial rescue of spg1-B8 by increased expression of mob1 is intriguing, particularly since it does not rescue any of the other septation signalling mutants. To date we have failed to detect any stable complexes between spg1p and mob1p in vitro or by two-hybrid analysis (not shown), so the basis for this rescue remains obscure. However, it is possible that it indicates the formation of a transient complex of the two proteins in vivo. This question will be addressed in future studies.
Mob1p also co-localises with the medial ring prior to septation, but only after all the other known components of the medial ring (except sid2p) have assembled. It is not present in cells arrested in early mitosis and does not colocalise with the contractile ring seen at the leading edge of the developing septum. This pattern of localisation during mitosis is similar to that of sid2p (Sparks et al., 1999) and pob1p (Toya et al., 1999). Like the sid2p medial ring, the mob1p ring appears to be extremely heat and fixation-sensitive, and we have been unable to assess whether its formation depends upon the other septation signalling genes at the restrictive temperatures for the available mutants. Nonetheless, partial loss of sid2p function at the permissive temperature clearly impairs mob1p ring formation, and vice-versa, suggesting that the two proteins form a complex at the medial ring, or that their appearance at the ring is interdependent. An attractive model is that the activation of septation signalling causes the appearance of mob1p/sid2p at the medial ring, resulting in the phosphorylation of key substrates to bring about medial ring contraction and septum biosynthesis. Our data do not permit us to distinguish whether mob1p/sid2p complexes are first assembled at the spindle pole body, then translocated to the medial ring, or whether appearance of these proteins at the two locations is independent. Future studies will address this issue, and also attempt to identify the substrates of the mob1p/sid2p complex.
It is becoming increasingly apparent that the spindle pole body is a localisation site for molecules implicated either in formation of the medial ring (rng2p and cam1p; Eng et al., 1998; Moser et al., 1997), or signalling the onset of septum formation (spg1p, cdc7p; Sohrmann et al., 1998); sid2p (Sparks et al., 1999); cdc16p and byr4p (Cerutti and Simanis, 1999), or both (plo1p; Bahler et al., 1998a; Mulvihill et al., 1999). Other mitotic regulators, such as cdc2p and cdc13p are found at the poles of the mitotic spindle (Alfa et al., 1990). Since these proteins execute their function at various times during mitosis, the spindle pole body may provide an ideal location for their activation to be co-ordinated with the initiation of mitosis and anaphase progression.
The S. cerevisiae counterparts of the S. pombe septation regulators cdc7, plo1, sid2, mob1, cdc16, and byr4, respectively are CDC15, CDC5, DBF2, MOB1, BUB2, and BFA1 (reviewed by Morgan, 1999; Taylor, 1999). In budding yeast, cdc15, cdc5, dbf2, and mob1 mutants all arrest with elongated spindles, separated chromosomes, elevated Cdc28p-Clb2p kinase activity, and do not undergo cytokinesis. An important role for these proteins in controlling exit from mitosis is to regulate the release of the phosphoprotein phosphatase Cdc14p from the nucleolus (Shou et al., 1999; Straight et al., 1999; Visintin et al., 1999). However, certain alleles of CDC15 are defective primarily in cytokinesis (Jimenez et al., 1998), suggesting that the S. cerevisiae proteins may also be implicated directly in regulating cytokinesis, rather than the cytokinesis defect of the mutants being a consequence of a late mitotic arrest. It is possible that in S. cerevisiae this so-called mitotic exit network is used to regulate multiple events at the end of mitosis, while in fission yeast, its primary role is to control septation.
We thank Dan McCollum for sending us the S. pombe strains expressing GFP and myc13 tagged sid2p. We are grateful to Kay Hofmann for help with sequence and database analysis, to Keith Gull for TAT-1 antibody, and to Paul Nurse for the rabbit anti-GFP serum. We thank Elena Cano for technical assistance, and the members of the cell cycle control group and our colleagues at ISREC for discussions. Work in this laboratory is supported by the Swiss National Science Foundation, the Swiss Cancer League, the Fondation Forez, the Roche Research Foundation, and ISREC.