Isolation and Characterization of New Fission Yeast Cytokinesis Mutants

Corresponding author: Kathleen L. Gould, Howard Hughes Medical Institute and Department of Cell Biology, Vanderbilt University School of Medicine, Nashville, TN 37232, kathy.gould{at}mcmail.vanderbilt.edu (E-mail).

Communicating editor: M. D. ROSE

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Schizosaccharomyces pombe is an excellent organism in which to study cytokinesis as it divides by medial fission using an F-actin contractile ring. To enhance our understanding of the cell division process, a large genetic screen was carried out in which 17 genetic loci essential for cytokinesis were identified, 5 of which are novel. Mutants identifying three genes, rng3⁺, rng4⁺, and rng5⁺, were defective in organizing an actin contractile ring. Four mutants defective in septum deposition, septum initiation defective (sid)1, sid2, sid3, and sid4, were also identified and characterized. Genetic analyses revealed that the sid mutants display strong negative interactions with the previously described septation mutants cdc7-24, cdc11-123, and cdc14-118. The rng5⁺, sid2⁺, and sid3⁺ genes were cloned and shown to encode Myo2p (a myosin heavy chain), a protein kinase related to budding yeast Dbf2p, and Spg1p, a GTP binding protein that is a member of the ras superfamily of GTPases, respectively. The ability of Spg1p to promote septum formation from any point in the cell cycle depends on the activity of Sid4p. In addition, we have characterized a phenotype that has not been described previously in cytokinesis mutants, namely the failure to reorganize actin patches to the medial region of the cell in preparation for septum formation.

THE fission yeast Schizosaccharomyces pombe has become a powerful model organism with which to study the process of cytokinesis. Some of its key attributes and advantages in this regard include the following: (1) a well-characterized mitotic cell cycle; (2) the fact that S. pombe, like animal cells, divides by medial fission through the use of a medial actin contractile ring; (3) the availability of mutants defective specifically in cytokinesis; and (4) the ease with which genetic, molecular, and cytological manipulations can be performed in this yeast. Furthermore, it has been established that well-known actin binding proteins, such as tropomyosin and profilin, are essential for actin contractile ring function in S. pombe, and components critical for cytokinesis that have not been identified through biochemical approaches have been identified genetically in S. pombe (reviewed in CHANG and NURSE 1996 ; GOULD and SIMANIS 1997 ).

The two most prominent actin structures in S. pombe are the contractile ring and the patches. The ring forms prior to detectable chromosome separation and begins to constrict after completion of anaphase. Actin patches are concentrated at the growing end(s) of cells during interphase, and subsequent to actin ring formation, they accumulate in the medial region of the cell adjacent to the medial actin ring where it is proposed that they regulate deposition of the septum. Previous studies of cytokinesis in S. pombe have identified at least 11 mutants defective in cytokinesis (NURSE et al. 1976 ; CHANG et al. 1996 ; SOHRMANN et al. 1996 ; KITAYAMA et al. 1997 ; MAY et al. 1997 ; SCHMIDT et al. 1997 ). These mutants can be classified into three major classes: those affecting cleavage plane specification (mid1/dmf1; hereafter referred to as mid1), those controlling actin ring assembly (cdc3, cdc4, cdc8, cdc12, cdc15, myo2, and rng2), and those controlling actin ring contraction and/or septum deposition (cdc7, cdc11, spg1, and cdc14). Most of the genes corresponding to the above-mentioned mutations have been cloned and sequenced, and, as expected, their characterization is providing a more detailed understanding of how cytokinesis proceeds.

Not only are mid1 mutants defective in actin ring positioning (CHANG et al. 1996 ) but Mid1p localizes to the nucleus in interphase cells and to a medial ring at the cortex of cells undergoing mitosis and cytokinesis (SOHRMANN et al. 1996 ). Therefore, it is likely that Mid1p mediates actin ring placement. Many of the genes corresponding to the actin ring assembly mutants encode proteins known to interact with actin and/or the actin cytoskeleton. Hence, the observation that mutations in these genes give rise to defects in cytokinesis is not unexpected. The cdc3 and cdc8 genes encode profilin and tropomyosin, respectively (BALASUBRAMANIAN et al. 1992 , BALASUBRAMANIAN et al. 1994 ). Ccd4 encodes an EF-hand protein with properties of a myosin light chain (MCCOLLUM et al. 1995 ), and the myo2 gene encodes a putative myosin heavy chain (KITAYAMA et al. 1997 ; MAY et al. 1997 ). Both Cdc4p and Myo2p localize exclusively to the medial ring during cytokinesis (MCCOLLUM et al. 1995 ; KITAYAMA et al. 1997 ). The other gene products in this category (Cdc15p and Cdc12p) also localize to the medial ring during cytokinesis (FANKHAUSER et al. 1995 ; CHANG et al. 1997 ). Cdc15p contains coiled-coil regions and an SH3 domain and its ectopic expression promotes actin ring formation in G2-arrested cells, implicating it as a key element in actin ring formation (FANKHAUSER et al. 1995 ). Cdc12p is related to the products of the budding yeast BNI1 gene (EVANGELISTA et al. 1997 ; IMAMURA et al. 1997 ) and the Drosophila Diaphanous gene (CASTRILLON and WASSERMAN 1994 ), both of which function in cytokinesis (CHANG et al. 1997 ).

At the end of anaphase, cells initiate actin ring constriction and deposition of the medial septum. How these events are regulated and coupled with the rest of the cell cycle is unclear at present although several genes involved in this step have been identified. The biochemical functions of these gene products and genetic interactions among strains containing mutations in these genes indicate that they most likely form a signal transduction pathway (reviewed in CHANG and NURSE 1996 ; GOULD and SIMANIS 1997 ). Cdc7p is a protein kinase that induces septum formation in a dosage-dependent fashion (FANKHAUSER and SIMANIS 1994 ). The spg1 gene, which encodes a putative GTPase, was found as a high-copy suppressor of cdc7 mutants. Spg1p binds to Cdc7p and is likely to be a key element in initiating septation because its overexpression drives septation from any point in the cell cycle (SCHMIDT et al. 1997 ). Cdc14p is novel in sequence and its biochemical function(s) is unknown (FANKHAUSER and SIMANIS 1993 ), although genetic interactions have been observed between cdc14 and cdc7, cdc11, as well as the fission yeast homolog of the mammalian glycogen synthase kinase-3, skp1 (MARKS et al. 1992 ; PLYTE et al. 1996 ). In addition to the above-mentioned cdc genes, the fission yeast homolog of the Drosophila polo gene, plo1, also has been shown to be important for cytokinesis, as well as aspects of mitosis, and thereby may couple the two events (OHKURA et al. 1995 ). Indeed, overexpression of Plo1p has been shown to drive septum formation regardless of the cell cycle stage, suggesting that Plo1p is another key regulator of cytokinesis.

To identify additional genes whose products function in cytokinesis, we carried out a large-scale genetic screen and have identified 17 genetic loci (6 previously unidentified by mutation) essential for various aspects of cytokinesis including actin ring placement, actin ring formation, actin patch relocalization, and actin ring constriction and septation. This study includes a description of their phenotypes. In addition, we report that rng5-E1 represents the first temperature-sensitive mutation in myo2⁺, septum initiation defective (sid)3⁺ is allelic with spg1⁺ (SCHMIDT et al. 1997 ), and sid2⁺ encodes a close relative of the Saccharomyces cerevisiae protein kinase, Dbf2p (JOHNSTON et al. 1990 ). While overexpression of spg1⁺ drives septation from any point in the cell cycle, it requires the activity of Sid4p. Hence, we have most likely identified additional components in a signal transduction cascade that regulates cell division.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

S. pombe strains and culture conditions:
S. pombe strains used in this study have been listed in Table 1. Media for vegetative growth and sporulation were as described in MORENO et al. 1991 . Cell synchronizations were carried out using a Beckman J6 centrifuge equipped with an elutriator rotor. Physiological experiments were all performed in complete medium. Repression of transcription from the nmt1 promoter (MAUNDRELL 1993 ) was achieved by addition of thiamine to a concentration of 2 µM. S. pombe transformations were carried out by electroporation (PRENTICE 1992 ).

Mutagenesis and screening for cytokinesis mutants:
The strategy used by BROEK et al. 1991 , HAYLES et al. 1994 , and CHANG et al. 1996 was employed to screen for mutants defective in cytokinesis. A mam2::LEU2 ade6-M216 leu1-32 ura1⁺ h⁹⁰ (JZ758, a kind gift of Dr. C. SHIMODA) strain was used as the starting strain for mutagenesis. This strain, even though homothallic, is incapable of mating with its progeny due to the presence of a mutation in the P-factor receptor. However, a diploid strain of the genotype mam2::LEU2/mam2::LEU2 h⁹⁰/h⁹⁰ is capable of meiosis and sporulation since conjugation is not necessary under these circumstances. These features, i.e., inability to mate with its progeny but ability to sporulate once converted to a diploid, were of importance in the selection for mutants defective in cytokinesis as described below. Another feature of importance was the ability of mam2::LEU2 ade6-M216 leu1-32 ura1^- h⁹⁰ to mate with heterothallic strains of the mating type h^-, since only the P-factor receptor function is compromised while the M-factor receptor is intact in these cells. JZ758 was mutagenized with nitrosoguanidine and cells were allowed to recover for 10 hr at 25°. The culture was then split three ways and shifted to 36° for 1.5 hr, 3 hr, and 4 hr, respectively. This step was performed to allow expression of potential heat-sensitive mutant phenotypes. Mutants of interest were those that could diploidize due to impaired cytokinesis causing the accumulation and fusion of more than one interphase nucleus in a cell. Following the temperature shifts, cells were washed three times with sporulation medium and incubated in the same medium for 3 days to allow meiosis and sporulation. Spores were prepared by digesting ascus and cell walls with glusulase and spores were suspended in water. Spores were plated on YE agar plates and incubated at 25° until colonies appeared. Colonies were replica plated to YE agar plates containing the vital dye phloxin B and incubated at 36° overnight. These plates were screened microscopically for two major phenotypes exhibited by cytokinesis mutants: highly elongated and lysed cells and elongated, dumbbell-shaped, and lysed cells. From several rounds of screening about 80 mutants that fell in one of these two categories were chosen for linkage and cytological analyses. Linkage and complementation analyses were performed by free spore analysis or by mei1-102-mediated stable diploid analysis (NURSE et al. 1976 ). Strains chosen for further analysis were backcrossed at least three times.

Fluorescence microscopic methods:
Staining with rhodamine-conjugated phalloidin, Calcofluor, and antibodies was performed as described (BALASUBRAMANIAN et al. 1997 ). The following fixatives were used in cell staining experiments: formaldehyde for F-actin staining with rhodamine-conjugated phalloidin, methanol for staining with polyclonal antibodies against Arp3p (MCCOLLUM et al. 1996 ) and monoclonal antibodies against actin [Amersham (Arlington Heights, IL) N-350], and a mixture of formaldehyde and glutaraldehyde for staining with antibodies against Cdc4p (MCCOLLUM et al. 1995 ). Following staining, images were captured using a Zeiss (Thornwood, NY) ZVS-47DEC image capturing system.

Molecular cloning:
The rng5⁺, sid2⁺, and sid3⁺ genes were cloned by complementation. The sid2-250 and sid3-106 strains were transformed with a genomic library constructed in pUR19 (BARBET et al. 1992 ) and the rng5-E1 strain was transformed with a genomic library constructed in pWH5 (a generous gift of KATHY MACH and DR. CHARLIE ALBRIGHT). Transformants were selected at 25°. Two plasmids containing overlapping genomic DNA inserts capable of rescuing the sid2 temperature-sensitive growth defect were isolated and the ends of each genomic insert were sequenced. A search of Sanger Center S. pombe genome sequencing project database revealed that these genomic clones corresponded to a sequenced region of the S. pombe genome on cosmid c24B11. A single predicted open reading frame that encoded a putative protein kinase (SPAC24B11.11c) was located in the region defined by the two clones. Genetic crosses showed that the sid2 mutation mapped to the same region of the genome as the cloned gene.

Three plasmids containing overlapping genomic DNA inserts capable of rescuing the sid3 growth defect were isolated. The smallest rescuing clone, pMB600, carried approximately 3 kb of S. pombe DNA. This plasmid was linearized by digestion with SalI and introduced into a sid3-106 ade6-210 leu1-32 ura4-D18 h^- strain. Five integrants were crossed individually to a wild-type strain. No temperature-sensitive progeny were identified among the products of meiosis in each case suggesting that the cloned gene was sid3⁺ or a gene very tightly linked to it. The rescuing activity was mapped to a 1.6-kb segment of DNA and was subjected to nucleotide sequencing by the dideoxy method. Complementary DNA clones were isolated by polymerase chain reaction using oligonucleotides (MOH58 (5' GCA GAG TAA TAT CAC TGG 3') and MOH59 (5' GAA AGA GAT GGT AAA GC 3') and the sequences of the intron-exon boundaries verified. The sid3⁺ cDNA was also obtained in a screen for cDNAs that, when overexpressed, caused an increase in the percentage of septated cells (HE et al. 1998 ). To overexpress the sid3⁺ cDNA, two approaches were taken. The first utilized a plasmid in which sid3⁺ expression was under control of the thaimine repressible nmt1 promoter in pREP3X (MAUNDRELL 1993 ). Wild-type cells carrying pREP3X-sid3⁺ were grown in media containing thiamine. After reaching mid-log growth, they were washed three times in media lacking thiamine and incubated further in the absence of thiamine. Depending on the temperature, cells began to form septa between 14 and 18 hr after thiamine deprivation began. The second approach used an integrated copy of sid3⁺ under control of the nmt1 promoter. From pREP3Xsid3⁺, a PstI-BamHI fragment containing the nmt1 promoter and sid3⁺ was removed and subcloned into pJK148 (KEENEY and BOEKE 1994 ). The resultant plasmid was linearized with MluI and transformed into sid4-SA1 leu1-32 ura4-D18 ade6-M210 h-cells. Stable Leu⁺ integrants were selected and integration of pJK148 at the leu1⁺ locus was confirmed by Southern blot hybridization. The integrated nmt1 sid3⁺ was also crossed into a sid4⁺ genetic background. The nucleotide sequence of sid3⁺ has been deposited in the GenBank database under the accession number AJ001587.

One plasmid capable of rescuing rng5-E1 cells was recovered. The ends of the ~10-kb insert were sequenced and found to match the sequence surrounding the myo2⁺ gene. By the construction of subclones and deletion clones, it was established that myo2⁺ was responsible for the complementation activity. Tight genetic linkage between rng5-E1 and the ade6 loci (unpublished observations) and physical linkage between myo2⁺ and ade6⁺ (MAY et al. 1997 ) confirmed that rng5⁺ is allelic with myo2⁺.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Isolation and classification of mutants defective in cytokinesis:
To isolate additional mutants defective in cytokinesis, we utilized a genetic screen that enriched for mutants that diploidize by one of the following routes: rereplication of chromosomal DNA, improper segregation of chromosomal DNA, or fusion of more than one interphase nucleus in the same cytoplasm (BROEK et al. 1991 ; HAYLES et al. 1994 ; CHANG et al. 1996 ). Approximately eight heat-sensitive mutants defective in cytokinesis were isolated using the procedure outlined in MATERIALS AND METHODS. Visual screening of these mutants after staining with the cell wall stain Calcofluor placed them in three phenotypic categories: (1) mutants defective in septum positioning; (2) mutants with improper septum deposits; and (3) mutants that did not deposit detectable septum material (Figure 1).

View larger version (91K): In this window In a new window Download PPT slide   Figure 1. Septation patterns of cytokinesis mutants. Cells were grown at 25° to exponential growth phase in complete medium, shifted to 36° for 2 hr, fixed and stained with Calcofluor. Representative images of (a) wild-type cells; (b1 and b2) dmf1-6 cells; (c) rng3-65 cells; and (d) sid3-106 cells.

Seven mutants defective in septum positioning were isolated. They were subsequently determined to lie in a single gene, mid1 (Table 2). dmf1-6 (dmf1 is allelic with mid1), one of the mid1 alleles isolated in this screen, has been described in detail by SOHRMANN et al. 1996 and is not discussed further here.

The actin ring serves to guide and position the septum in S. pombe (MARKS and HYAMS 1985 ). Hence, mutants defective in actin ring assembly or stability can be identified on the basis of the accumulation of disorganized septal material in the medial region of the cell (BALASUBRAMANIAN et al. 1994 ; MCCOLLUM et al. 1995 ; CHANG et al. 1996 ; MARKS et al. 1987 ; NURSE et al. 1976 ). Twenty-six such mutants were isolated. Crosses with the five previously identified temperature-sensitive mutants producing this phenotype (cdc3, cdc4, cdc8, cdc12, and rng2), followed by pairwise crosses between the newly isolated mutants established that three novel mutants defective in this stage of cytokinesis had been identified. These new loci have been named rng3, rng4, and rng5, respectively (Table 2). We found that rng5-E1 was synthetically lethal with rng3-65 (K. WONG and M. K. BALASUBRAMANIAN, unpublished data) as well as with cdc4-8. To verify that these mutants had defects in actin ring assembly, logarithmically growing rng3-65, rng4-146, and rng5-E1 cells were shifted to 36° and the actin cytoskeleton was visualized by staining with rhodamine-conjugated phalloidin (Figure 2). Upon incubation at the restrictive temperature for 4 hr, rng3-65 cells accumulated up to four nuclei. Actin rings were not formed in rng3-65 cells that were in M-phase, but in many cells, prominent actin cables were visualized. Upon incubation at the restrictive temperature for 4 hr, rng5-E1 cells also accumulated up to four nuclei as well as prominent septa. In contrast to rng3-65 cells, rng5-E1 cells accumulated actin filaments in the medial region of the cell during mitosis. However, these filaments were not organized in a compact ring like that observed in wild-type cells. rng4-146 cells did not arrest the cell cycle homogeneously within 4 hr although their actin staining patterns clearly indicated a role for Rng4p in actin ring function (data not shown). Staining of representative alleles of known genes that were isolated in this screen (cdc3-5, cdc4-D3, cdc8-A15, cdc12-242, and rng2-D5) established that, as expected, these alleles were defective in actin ring formation (data not shown).

View larger version (86K): In this window In a new window Download PPT slide   Figure 2. Actin ring formation is defective in rng3-65 and rng5-E1 mutants. Cells of the indicated genotypes were grown to exponential growth phase and shifted to 36° in complete medium. Samples were taken at 2 and 4 hr after temperature shift, fixed, and stained with rhodamine-conjugated phalloidin to visualize F-actin and DAPI to visualize DNA.

Forty mutants that did not produce a septum, despite accumulating multiple nuclei, were identified. These were crossed with cdc7, cdc11, cdc14, cdc15 mutants, the previously isolated mutants displaying a similar phenotype and then analyzed by pairwise crosses among themselves. These crosses revealed that we had obtained mutations in eight genes, four of which had not been identified previously (Table 2). The new genetic loci were named sid1, sid2, sid3, and sid4, respectively. Previous studies have shown that with the exception of cdc15, cytokinesis mutants that are defective for septum deposition are capable of making medial actin rings (MARKS et al. 1987 ; FANKHAUSER et al. 1995 ). Therefore, we tested the ability of sid1-239, sid2-250, sid3-106, and sid4-SA1 mutants to form medial actin rings. Mutant cells were grown to exponential growth phase and shifted to 36°. Samples were taken 2 and 4 hr later, fixed and stained with rhodamine-conjugated phalloidin and DAPI (Figure 3). All four mutants behaved like cdc7, cdc11, and cdc14 mutants in that they formed medial actin rings during mitosis, and during the subsequent interphases, F-actin was localized to patches at the cell ends. All four mutants accumulated up to four nuclei in 4 hr.

View larger version (64K): In this window In a new window Download PPT slide   Figure 3. Actin rings form normally in sid1-239, sid2-250, sid3-106, and sid4-SA1 mutants. Cells of the indicated genotypes were grown to exponential growth phase and shifted to 36° in complete medium. Samples were taken at 2 and 4 hr after temperature shift, fixed, and stained with rhodamine-conjugated phalloidin to visualize F-actin and DAPI to visualize chromosomal DNA.

During this analysis, we also examined cells containing the newly isolated mutations in previously described genes essential for septum deposition, cdc7-99 cdc11-205, cdc14-SA2, and cdc15-A5, for their ability to form medial actin rings. As expected, cdc7-99, cdc11-205, and cdc14-SA2 cells were capable of actin ring formation. Surprisingly, however, actin rings were also detected in cdc15-A5 cells undergoing mitosis (data not shown) although Cdc15p was reported previously to be essential for normal actin ring formation (FANKHAUSER et al. 1995 ) based on an analysis of the cdc15-140 mutant.

Actin patches do not reorganize during septation in cdc15-140 mutants:
We reexamined the ability of cdc15 mutants to form medial actin rings more thoroughly using cdc15-140 mutant cells since the previous work had utilized this allele. A synchronous population of cdc15-140 cells at the beginning of G2 phase was obtained by centrifugal elutriation, inoculated into fresh medium, and shifted to the restrictive temperature of 36°. Samples were removed at 20-min intervals, fixed, and stained with rhodamine-conjugated phalloidin and DAPI. At 100 min, all but two cells (marked with short arrows) in the field were in interphase and displayed actin staining in patches at cell ends (Figure 4, top). In the cells judged to be in mitosis because they had condensed chromosomes, a medial actin ring was deleted. At 120 min, most cells were in mitosis and all of these cells contained medial actin rings (Figure 4, middle). In some cells (marked with a white arrow) actin rings appeared to wrap around the cell circumference. The actin rings were qualitatively thinner than those typically observed in wild-type cells and in most focal planes these were detected only as two dots. At 140 min, most cells had exited mitosis and entered the following interphase; because they had failed to undergo cytokinesis, they contained two nuclei and actin patches were dispersed along the cell periphery (Figure 4, bottom). This analysis raised the possibility that the predominant defect in cdc15 mutant cells might lie not in the formation of a medial actin ring but at a later step in the cell division process.

View larger version (94K): In this window In a new window Download PPT slide   Figure 4. cdc15-140 cells form actin rings. A synchronous population of early G2 cells of the cdc15-140 mutant strain was isolated by centrifugal elutriation from exponential growth phase and inoculated into fresh complete medium at 36°. Samples were taken at 20-min intervals, fixed, and stained with rhodamine-conjugated phalloidin to visualize F-actin and DAPI to visualize chromosomal DNA. Representative fields of cells from the 100-, 120-, and 140-min time points following shift to 36° are shown.

It occurred to us that one possible explanation for the inability of cdc15 or any of the sid mutants to septate was a failure to relocalize actin patches adjacent to the actin ring during mitosis. To test this possibility, we stained synchronous cell populations of several mutant strains with antibodies against Cdc4p, which is detected only in the medial ring during the mitosis (MCCOLLUM et al. 1995 ) and antibodies to Arp3p, which is detected only in actin patches (MCCOLLUM et al. 1996 ). The use of these reagents allows a clear distinction between the two actin structures that is not always achieved with the use of rhodamine-conjugated phalloidin that binds to all filamentous actin (BALASUBRAMANIAN et al. 1997 ). We chose to examine two tight sid mutants, sid2-250, and a previously characterized mutant with the sid phenotype, cdc7-24 (FANKHAUSER and SIMANIS 1994 ), as well as wild-type and cdc15-140 cells. Synchronized populations of cells in G2 were isolated by centrifugal elutriation, inoculated into fresh medium, and incubated at 36°. Samples were taken at 10- or 20-min intervals, fixed, and stained and the percentages of cells containing actin rings and actin patches in the medial region of cells were quantified (Figure 5, A–D). As expected, cells from all four strains were able to form medial rings as detected by anti-Cdc4p staining at a time fairly coincident with the appearance of binucleate cells in the populations. The percentage of cells displaying Arp3p patches in the medial region of the cell peaked at 70% for the cdc7-24 strain and 60% for sid2-250. These data demonstrate that both cdc7-24 and sid2-250 are capable of mobilizing actin patches to the medial region of the cell following actin ring formation. In contrast, Arp3p-patches were detected in the medial region of only 9% of cdc15-140 cells. Representative fields of sid2-250 and cdc15-140 cells stained with antibodies to Arp3p are shown in Figure 5E, as are typical cdc15-140 cells stained with antibodies to Cdc4p. Analyses of other mutants with a sid phenotype (sid1-239, sid3-106, sid4-SA1, cdc11-123, and cdc14-118) demonstrated that all were capable of mobilizing actin patches to the medial region of the cell following actin ring formation (data not shown). In the course of these studies we observed that the cdc15, sid2, and cdc7 mutant cells took longer to enter mitosis following temperature-shift than did wild-type cells. As determined by the appearance of at least 25% binucleate cells, wild-type cells entered mitosis at 60 min after shift-up (Figure 5A), whereas binucleate cells appeared significantly later in the mutant cell populations (Figure 5B and Figure C). The reason for this delay is unclear at present.

View larger version (152K): In this window In a new window Download PPT slide   Figure 5. Actin patch mobilization to the medial region is impaired in cdc15-140 cells. Synchronous cultures of wild-type, cdc7-24, sid2-250, cdc15-140 cells in the G2 phase of the cell cycle were prepared by centrifugal elutriation at 25° and inoculated into complete media at 36°. Samples were taken at 10- or 20-min intervals (depending on the strain) through one nuclear division cycle and fixed with methanol or a mixture of formaldehyde and glutaraldehyde. Methanol-fixed cells were stained with antibodies against Arp3p to visualize actin patches. Aldehyde-fixed cells were stained with antibodies against Cdc4p to visualize actin rings. Nuclei were visualized by staining with DAPI. The percentage of cells containing medial rings, medial actin patches and two nuclei were quantitated. (A) Wild-type cells; (B) cdc7-24 cells; (C) sid2-250 cells; and (D) cdc15-140 cells. (E) Representative images of sid2-250 and cdc15-140 cells in anaphase stained with DAPI and antibodies to Arp3p. Typical cdc15-140 cells in anaphase stained with antibodies to Cdc4p are also shown.

Genetic interactions among the sid mutants:
To look for genetic interactions among the sid mutants and between the sid mutants and the previously isolated mutants in the same functional category (cdc7, cdc11, and cdc14), we carried out two types of analyses. First, we examined whether the two cloned genes, cdc7⁺ and cdc14⁺, could complement the sid1, 2, 3, or 4 mutants when present at high copy. Elevated levels of cdc14⁺ did not rescue any of the sid mutants. However, elevated levels of cdc7⁺ did rescue one of the mutants, sid3-106.

In the second analysis, each mutant was crossed to the other mutants to reveal synthetic lethal interactions or reciprocal suppression between them. The results of these crosses are presented in Table 3. We found several synthetic lethal interactions in these crosses when the progeny of these crosses were examined at 25°: (1) sid2 with cdc7, cdc14 and sid4, and (2) sid3 with cdc7, cdc11, cdc14, and sid4. We did not detect any reciprocal-suppression events among double mutants from these crosses. These results indicate that the sid gene products are likely to function along with the previously described cdc7, 11, and 14 proteins to bring about septation and cell division.

Gene cloning:
To understand the molecular functions of the rng and sid genes we have begun to clone and sequence the wild-type alleles of these new genes. Rng5⁺, sid3⁺, and sid2⁺ were cloned by complementation of the recessive temperature-sensitive colony formation defects of the rng5-E1, sid3-106, and sid2-250 mutants, respectively, and integration mapping confirmed that the cloned DNAs represented the indicated genes rather than high-copy suppressors. The nucleotide sequence of a 1.6-kb fragment of DNA, which was sufficient for rescue of sid3-106, was determined. Analysis of this sequence revealed the presence of a 198-amino-acid coding region present in five putative exons. The DNA sequence of the corresponding cDNA confirmed the presence and position of the introns. Database searches using the predicted 198-amino-acid polypeptide sequence as a query indicated that Sid3p was a member of the ras family of small eukaryotic GTPases (BOURNE et al. 1991 ; BARBACID 1987 ). While this work was in progress, it became clear that sid3p was identical to the product of the spg1⁺ gene (SCHMIDT et al. 1997 ) and was most closely related to the budding yeast Tem1p, a protein known to be required for the termination of mitosis (SHIRAYAMA et al. 1994 ). The rng5⁺ gene encoded the recently described myo2⁺ gene (KITAYAMA et al. 1997 ; MAY et al. 1997 ). The sid2⁺ gene encoded a potential homologue of the Dbf2p protein kinase from the budding yeast S. cerevisiae (JOHNSTON et al. 1990 ) displaying 80% sequence similarity over the entire length of the protein (Figure 6).

View larger version (73K): In this window In a new window Download PPT slide   Figure 6. Sid2p is a potential homolog of DBF2p. Alignment of Sid2p with S. cerevisiae Dbf2p. Dbf2p is on the top line of each row. The sequence of Sid2p was deduced from the DNA sequence of the cosmid c24B11. Amino acid identities and similarities between the two proteins are indicated with dashed lines.

Sid4p is required for spg1⁺-induced septation:
Overexpression of spg1⁺ induces septation from any point in the cell cycle (SCHMIDT et al. 1997 ). SCHMIDT et al. 1997 demonstrated that both the Cdc7p protein kinase and Cdc14p are required for this effect. We wished to determine whether Sid4p also was required for this effect. This question was of particular interest to us because sid4⁺ encodes a pioneer protein (L. CHANG and K. L. GOULD, unpublished results) and, hence, it is not immediately obvious what function it might play in a signal transduction cascade. Wild-type and sid4-SA1 cells containing an integrated copy of spg1⁺ under control of the thiamine repressible nmt1 promoter were grown at 25° in nonrepressive conditions for 23 hr. At this time, an increased percentage of septated cells was not observed in either culture; both cultures contained <20% septated cells (Figure 7, a–d). The cultures were then shifted to 36°. After a 5-hr incubation at 36°, wild-type cells contained multiple septa (Figure 7E and Figure F). In the culture, 32% of cells contained one septum, 40% contained two septa, and 21% contained three or more septa. In contrast, septa were not detected in the sid4-SA1 cells; they had begun to elongate and assume their terminal arrest phenotype (Figure 7G and Figure H). Thus, the induction of septum formation by spg1⁺ requires Sid4p in addition to Cdc7p and Cdc14p.

View larger version (67K): In this window In a new window Download PPT slide   Figure 7. Sid4p is required for Spg1p-induced septation. Wild-type (KGY1300; a, b, e, and f) and sid4-SA1 (KGY1346; c, d, g, and h) cells carrying a single integrated copy of sid3+ under the control of the thiamine-repressible nmt1 promoter were grown to mid-log phase at 25° in media containing thiamine. To induce overexpression of sid3+, cells were then grown in media lacking thaimine for 23 hr at 25° (a–d). Then, the cells were shifted to 36° for 5 hr (e–h). Cells were fixed with formaldehyde and stained with either DAPI (a, c, e, and g) or Calcofluor (b, d, f, and h).

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

To identify additional genes required for cytokinesis in S. pombe, we carried out a previously described genetic screen on a large scale. Characterization of the mutants derived in this screen increased by five the number of genes known to be involved in cytokinesis and provided additional evidence that polarization of actin patches to the medial region of the cell for septation is a discrete step in cytokinesis that can be defined by mutation.

Actin ring positioning and formation:
We isolated a mutant defective in the choice of cleavage plane in the course of this genetic screen; this mutant was described in an earlier study (SOHRMANN et al. 1996 ). Temperature-sensitive mutant alleles in this gene, mid1, have been isolated independently by CHANG et al. 1996 . In fission yeast the position of the premitotic nucleus (or the spindle pole body, which is embedded in the nuclear membrane) has been proposed to specify the position of the actin ring and the medial septum (CHANG et al. 1996 ). Consistent with the proposed role of Mid1p in specifying cleavage plane, Mid1p has been localized to the nucleus in interphase cells and to the medial region of the cell cortex during mitosis and cytokinesis (SOHRMANN et al. 1996 ).

Several proteins are now known to be required for actin ring formation in S. pombe including Cdc4p-putative myosin light chain, Cdc8p-tropomyosin, Cdc12p, Cdc15p, and Myo2p-myosin heavy chain. These proteins have been shown to be components of the medial ring (BALASUBRAMANIAN et al. 1992 ; MCCOLLUM et al. 1995 ; CHANG et al. 1997 ; KITAYAMA et al. 1997 ; MAY et al. 1997 ; FANKHAUSER et al. 1995 ). How these and yet to be identified proteins bring about actin ring formation, however, remains to be elucidated. Interestingly, rng5-E1 represents the first temperature-sensitive allele of myo2⁺ to be described, and its phenotype is nearly indistinguishable from that reported for the myo2 null (KITAYAMA et al. 1997 ; MAY et al. 1997 ). In rng5-E1 cells, actin rings form but they are not organized properly. Given the molecular identities of cdc4⁺ and rng5⁺, it is not surprising that we found a synthetic lethal interaction between rng5-E1 and cdc4-8; it is likely that Cdc4p is a, or the, light chain for Myo2p. Further analysis of the rng5 and rng2 mutants and the two new genes identified in this study, rng3 and rng4, should contribute significantly to the understanding of this complex process.

Cdc15p and mobilization of actin patches for septation:
During anaphase in S. pombe, actin patches become concentrated in the medial region of the cell adjacent to the actin ring (MARKS and HYAMS 1985 ). These actin patches may be responsible for the deposition of the septum and cell wall. Using both antibodies to a known component of actin patches, Arp3p, and phalloidin to detect patch distribution, we found that mutants of cdc15 do not reorganize actin patches to the medial region of the cell in preparation for septation. In contrast, all other cytokinesis mutants capable of forming medial rings accumulated actin patches adjacent to those rings. This abnormal distribution of actin patches has not been described before in a cytokinesis mutant although it has been observed previously in the apr3-c1 mutant (MCCOLLUM et al. 1996 ) and it further defines actin patch reorganization as a discrete step in the process of cytokinesis. Cdc15p was reported previously to be essential for the formation of a normal actin ring (FANKHAUSER et al. 1995 ). We have established that medial rings do form in cdc15-140 mutants; they contain both F-actin, as detected by phalloidin staining, and Cdc4p. Other studies have demonstrated that Mid1p (SOHRMANN et al. 1996 ) and Myo2p (N. I. NAQVI and M. K. BALASUBRAMANIAN, unpublished observations) also form medial rings in cdc15-140 cells. Although it is certainly possible that the medial rings formed in cdc15 mutants are defective in structure and/or function, our studies indicate a key role for Cdc15p in the reorganization of actin patches to the medial region of the cell. Cdc15p contains SH3 and coiled-coil domains and has homologs in budding yeast, mice, and flatworms (FANKHAUSER et al. 1995 ). It will be of interest to determine whether Cdc15p binds directly to components of the actin patch.

Regulation of actin ring contraction and septation:
In the sid group of mutants (cdc7, cdc11, cdc14, sid1, sid2, spg1, and sid4) actin rings are formed and actin patches are mobilized to the medial region of the cell, suggesting that the machinery necessary for cytokinesis is assembled normally in these mutants. Thus, these mutants might be defective in a signal transduction pathway that initiates actin ring contraction and/or septum deposition. Consistent with this view are the molecular identities of four gene products in this group as well-known signaling molecules. Cdc7p (FANKHAUSER and SIMANIS 1994 ), Sid1p (D. MCCOLLUM, M. K. BALASUBRAMANIAN and K. L. GOULD, unpublished results), and Sid2p (this study) are protein kinase and Spg1p is a small GTPase (SCHMIDT et al. 1997 ; this study). That these genes work together to regulate septation is supported by the numerous genetic interactions between strains defective in these gene products (MARKS et al. 1992 ; this study) and the physical association of Cdc7p and Spg1p (SCHMIDT et al. 1997 ). We also have demonstrated that although the sequence of Sid4p is not informative as to its function (L. CHANG and K. L. GOULD, unpublished results), it most likely functions downstream of Spg1p in this pathway because the induction of septation by Spg1p requires Sid4p. In S. cerevisiae, a group of gene products, Cdc5p (a protein kinase), Cdc15p (a protein kinase), Cdc14p (a protein phosphatase), Tem1p (a GTPase), Lte1p, and Dbf2p (a protein kinase) control cyclin proteolysis and regulate exit from mitosis (NASMYTH 1996 ). Cdc7p is the homolog of the budding yeast Cdc15p (FANKHAUSER and SIMANIS 1994 ), Sid2p is most closely related to the budding yeast Dbf2p (this study), and Spg1p is most closely related to the budding yeast Tem1p (SCHMIDT et al. 1997 ; this study). These similarities suggest that the sid group of genes might be functionally analogous to the CDC15, CDC5, DBF2, CDC14, TEM1, and LTE1 group of genes. However, it is not yet clear what common signals, if any, regulate these conserved groups of genes that function in cytokinesis and mitosis, respectively.

	ACKNOWLEDGMENTS

Many thanks to Dr. C. SHIMODA for providing the mam2-null mutant strain, Dr. TONY CARR for a S. pombe genomic library, Dr. VIESTURS SIMANIS for the cdc14⁺-containing plasmid as well as communicating results prior to publication, Dr. GREG DEN HAESE for the cdc7⁺-containing plasmid, Dr. FRED CHANG for the rng2-346 mutant, KATHY MACH and Dr. CHARLIE ALBRIGHT for the genomic library, Ms. MANI GUJRAL and Ms. ANNA FEOKTISTOVA for technical help, Dr. JENNIFER MORRELL for helpful comments on the manuscript, and Drs. NAM-HAI CHUA and SRI RAMACHANDRAN for providing laboratory space, facilities, and experimental advice for K.C.Y.W. This work was supported by research funds from National Science and Technology Board, Singapore to M.K.B., National Institutes of Health grant GM49119 to S.S., the Howard Hughes Medical Institute to K.L.G., of which K.L.G. is an assistant investigator.

Manuscript received February 10, 1998; Accepted for publication April 16, 1998.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

BALASUBRAMANIAN, M. K., D. M. HELFMAN, and S. M. HEMMINGSEN, 1992  A new tropomyosin essential for cytokinesis in the fission yeast S. pombe.. Nature 360:84-87[Medline].

BALASUBRAMANIAN, M. K., B. R. HIRANI, J. D. BURKE, and K. L. GOULD, 1994  The Schizosaccharomyces pombe cdc3⁺ gene encodes a profilin essential for cytokinesis. J. Cell Biol. 125:1289-1301[Abstract/Free Full Text].

BALASUBRAMANIAN, M. K., D. MCCOLLUM, and K. L. GOULD, 1997  Cytokinesis in fission yeast Schizosaccharomyces pombe.. Methods. Enzymol. 283:494-506[Medline].

BARBACID, M., 1987  ras genes. Annu. Rev. Biochem. 56:779-827[Medline].

BARBET, N., W. J. MURIEL, and A. M. CARR, 1992  Versatile shuttle vectors and genomic libraries for use with Szhizosaccharomyces pombe.. Gene. 114:59-66[Medline].

BOURNE, H. R., D. A. SANDERS, and F. MCCORMICK, 1991  The GTPase superfamily: conserved structure and molecular mechanism. Nature 349:117-128[Medline].

BROEK, D., R. BARTLETT, K. CRAWFORD, and P. NURSE, 1991  Involvement of p34^cdc2 in establishing the dependency of S phase of mitosis. Nature 349:388-393[Medline].

CASTRILLON, D. H. and S. A. WASSERMAN, 1994  diaphanous is required in Drosophila and shares domains of similarity with the products of the limb deformity gene. Development 120:3367-3377[Abstract].

CHANG, F. and P. NURSE, 1996  How fission yeast fission in the middle. Cell 84:191-194[Medline].

CHANG, F., A. WOLLARD, and P. NURSE, 1996  Isolation and characterization of fission yeast mutants defective in the assembly and placement of the contractile actin rings. J. Cell Sci. 109:131-142[Abstract].

CHANG, F., D. DRUBIN, and P. NURSE, 1997  cdc12p, a protein required for cytokinesis in fission yeast, is a component of the cell division ring and interacts with profilin. J. Cell Biol. 137:169-182[Abstract/Free Full Text].

EVANGELISTA, M., K. BLUNDELL, M. S. LONGTINE, C. J. CHOW, and N. ADAMES et al., 1997  Bni1p, a yeast formin linking Cdc42p and the actin cytoskeleton during polarized morphogenesis. Science 276:118-122[Abstract/Free Full Text].

FANKHAUSER, C. and V. SIMANIS, 1993  The Schizosaccharomyces pombe cdc14 gene is required for septum formation and can also inhibit nuclear division. Mol. Biol. Cell 4:531-539[Abstract].

FANKHAUSER, C. and V. SIMANIS, 1994  The cdc7 protein kinase is a dosage dependent regulator of septum formation in fission yeast. EMBO J. 13:3011-3019[Medline].

FANKHAUSER, C., A. REYMOND, L. CERUTTI, S. UTZIG, and K. HOFFMANN et al., 1995  The S. pombe cdc15 gene is a key element in the reorganization of F-actin of mitosis. Cell 82:435-444[Medline].

GOULD, K. and V. SIMANIS, 1997  The control of septum formation in fission yeast. Genes Dev. 11:2939-2951[Free Full Text].

HAYLES, J., K. FISHER, A. WOOLLARD, and P. NURSE, 1994  Temporal order of S-phase and mitosis in fission yeast is determined by the state of the p34^cdc2/mitotic B cyclin complex. Cell 78:813-822[Medline].

HE, X., N. HAYASHI, N. G. WALCOTT, Y. AZUMA, and T. E. PATTERSON et al., 1998  The identification of sbp1, encoding a spi1-GTP binding protein, and other cDNAs that affect the mitosis-to-interphase transition in Schizosaccharomyces pombe.. Genetics 148:645-656[Abstract/Free Full Text].

IMAMURA, H., K. TANAKA, T. HIHARA, M. UMIKAWA, and T. KAMEI et al., 1997  Bni1p and Bnr1p: downstream targets of the Rho family small G-proteins which interact with profilin and regulate actin cytoskeleton in Saccharomyces cerevisiae.. EMBO J. 16:2745-2755[Medline].

JOHNSTON, L. H., S. L. EBERLY, J. W. CHAPMAN, H. ARAKI, and A. SUGINO, 1990  The product of the Saccharomyces cerevisiae cell cycle gene DBF2 has homology with protein kinases and is periodically expressed in the cell cycle. Mol. Cell Biol. 10:1358-1366[Abstract/Free Full Text].

KEENEY, J. B. and J. D. BOEKE, 1994  Efficient targeted integration at leu1-32 and ura4-294 in Schizosaccharomyces pombe. Genetics 136:849-856[Abstract].

KITAYAMA, C., A. SUGIMOTO and M. YAMAMOTO, 1997 Type II myosin heavy chain encoded by the myo2 gene composes the contractile ring during cytokinesis in Schizosaccharomyces pombe. 137: 1309–1319.

MARKS, J. and J. S. HYAMS, 1985  Localization of F-actin through the cell division cycle of Schizosaccaromyces pombe.. Eur. J. Cell Biol. 39:27-32.

MARKS, J., I. M. HAGAN, and J. S. HYAMS, 1987  Spatial association of F-actin with growth polarity and septation in the fission yeast Schizosaccharomyces pombe.. Spec. Publ. Soc. Gen. Microbiol. 23:119-135.

MARKS, J., C. FANKHAUSER, and V. SIMANIS, 1992  Genetic interactions in the control of septation in Schizosaccharomyces pombe.. J. Cell Sci. 10:801-808.

MAUNDRELL, K., 1993  Thaimine-repressible vectors pREP and pRIP for fission yeast. Gene 123:127-130[Medline].

MAY, K. M., F. Z. WATTS, F. Z. JONES, and J. S. HYAMS, 1997  A type II myosin involved in cytokinesis in the fission yeast, Schizosaccharomyces pombe. Cell Motil. Cytoskeleton 38:1-12[Medline].

MCCOLLUM, D., M. K. BALASUBRAMANIAN, L. E. PELCHER, S. HEMMINGSEN, and K. L. GOULD, 1995  The Schizosaccharomyces pombe cdc4⁺ gene encodes a novel EF-hand protein essential for cytokinesis. J. Cell Biol. 130:651-660[Abstract/Free Full Text].

MCCOLLUM, D. A., M. FOEKTISTOVA, M. MORPHEW, K. L. BALASUBRAMANIAN AND, and K. L. BALASUBRAMANIAN ANDGOULD, 1996  The Schizosaccharomyces pombe actin related protein Arp3p is a component of the cortical actin cytoskeleton and interacts with profilin. EMBO J. 15:6438-6446[Medline].

MORENO, S., A. KLAR, and P. NURSE, 1991  Molecular genetic analysis of fission yeast, Schizosaccharomyces pombe.. Methods Enzymol. 194:795-823[Medline].

NASMYTH, K., 1996  At the heart of the budding yeast cell cycle. Trends Genet. 12:405-412[Medline].

NURSE, P., P. THURIAUX, and K. NASMYTH, 1976  Genetic control of the cell division cycle in the fission yeast Schizosaccharomyces pombe.. Mol. Gen. Genet. 146:167-178[Medline].

OHKURA, H., I. M. HAGAN, and D. M. GLOVER, 1995  The conserved Schizosaccharomyces pombe kinase plo1, required to form a bipolar spindle, the actin ring, and septum, can drive septum formation in G1 and G2 cells. Genes Dev. 9:1059-1073[Abstract/Free Full Text].

PLYTE, S. E., A. FEOKTISTOVA, J. D. BURKE, J. R. WOODGETT, and K. L. GOULD, 1996  Schizosaccharomyces pombe skp1+ encodes a protein kinase related to mammalian glycogen synthase kinase 3 and complements a cdc14 cytokinesis mutant. Mol. Cell. Biol. 16:179-191[Abstract].

PRENTICE, H. L., 1992  High efficiency transformation of Schizosaccharomyces pombe by electroporation. Nucleic Acids Res. 20:621[Free Full Text].

SCHMIDT, S., M. SOHRMANN, K. HOFMANN, A. WOOLLARD, and V. SIMANIS, 1997  The Spg1p GTPase is an essential dosage-dependent inducer of septum formation in Schizosaccharomyces pombe.. Genes Dev. 11:1519-1534[Abstract/Free Full Text].

SHIRAYAMA, M., Y. MATSUI, and A. TOH-E, 1994  The yeast TEM1 gene, which encodes a GTP-binding protein, is involved in termination of M phase. Mol. Cell. Biol. 14:7476-7482[Abstract/Free Full Text].

SOHRMANN, M., C. FANKHAUSER, C. BRODBECK, and V. SIMANIS, 1996  The dmf/mid1 gene is essential for correct positioning of the division septum in fission yeast. Genes Dev. 10:2707-2719[Abstract/Free Full Text].

This article has been cited by other articles:

				O. Hachet and V. Simanis Mid1p/anillin and the septation initiation network orchestrate contractile ring assembly for cytokinesis Genes & Dev., November 15, 2008; 22(22): 3205 - 3216. [Abstract] [Full Text] [PDF]

				H. Yan, W. Ge, T. G. Chew, J. Y. Chow, D. McCollum, A. M. Neiman, and M. K. Balasubramanian The Meiosis-Specific Sid2p-related Protein Slk1p Regulates Forespore Membrane Assembly in Fission Yeast Mol. Biol. Cell, September 1, 2008; 19(9): 3676 - 3690. [Abstract] [Full Text] [PDF]

				E. M. Neidt, C. T. Skau, and D. R. Kovar The Cytokinesis Formins from the Nematode Worm and Fission Yeast Differentially Mediate Actin Filament Assembly J. Biol. Chem., August 29, 2008; 283(35): 23872 - 23883. [Abstract] [Full Text] [PDF]

				M. Lord, T. E. Sladewski, and T. D. Pollard Yeast UCS proteins promote actomyosin interactions and limit myosin turnover in cells PNAS, June 10, 2008; 105(23): 8014 - 8019. [Abstract] [Full Text] [PDF]

				N. Matmati and Y. A. Hannun Thematic Review Series: Sphingolipids. ISC1 (inositol phosphosphingolipid-phospholipase C), the yeast homologue of neutral sphingomyelinases J. Lipid Res., May 1, 2008; 49(5): 922 - 928. [Abstract] [Full Text] [PDF]

				A. Yonetani, R. J. Lustig, J. B. Moseley, T. Takeda, B. L. Goode, and F. Chang Regulation and Targeting of the Fission Yeast Formin cdc12p in Cytokinesis Mol. Biol. Cell, May 1, 2008; 19(5): 2208 - 2219. [Abstract] [Full Text] [PDF]