Characterization of ypa1 and ypa2, the Schizosaccharomyces pombe Orthologs of the Peptidyl Proyl Isomerases That Activate PP2A, Reveals a Role for Ypa2p in the Regulation of Cytokinesis

Abstract

The Schizosaccharomyces pombe septation initiation network (SIN) regulates cytokinesis. Cdc7p is the first kinase in the core SIN; we have screened genetically for SIN regulators by isolating cold-sensitive suppressors of cdc7-24. Our screen yielded a mutant in SPAC1782.05, one of the two fission yeast orthologs of mammalian phosphotyrosyl phosphatase activator. We have characterized this gene and its ortholog SPAC4F10.04, which we have named ypa2 and ypa1, respectively. We find that Ypa2p is the major form of protein phosphatase type 2A activator in S. pombe. A double ypa1–Δ ypa2–Δ null mutant is inviable, indicating that the two gene products have at least one essential overlapping function. Individually, the ypa1 and ypa2 genes are essential for survival only at low temperatures. The ypa2–Δ mutant divides at a reduced cell size and displays aberrant cell morphology and cytokinesis. Genetic analysis implicates Ypa2p as an inhibitor of the septation initiation network. We also isolated a cold-sensitive allele of ppa2, the major protein phosphatase type 2A catalytic subunit, implicating this enzyme as a regulator of the septation initiation network.

SCHIZOSACCHAROMYCES pombe cells take the form of a cylinder capped by hemispherical ends. During interphase, cells grow mainly at their tips. Cell elongation ceases after mitotic commitment, and the cytoskeleton is reorganized in preparation for nuclear and cell division (reviewed by McCully and Robinow 1971; Mitchison and Nurse 1985; Marks et al. 1986; Hagan and Hyams 1988; Hagan 1998). The cell divides by fission after forming a septum in the middle of the cell. The site of division is defined at the onset of mitosis (Daga and Chang 2005), by the formation of an actomyosin contractile ring (CAR), whose assembly continues throughout mitosis (Wu et al. 2003; Pollard 2008; Pollard and Wu 2010). At the end of anaphase, the CAR is thought to guide synthesis of the multilayered division septum (reviewed by (Ishiguro 1998; Sipiczki 2007).

Septation Initiation Network

The S. pombe septation initiation network (SIN) is a network of protein kinases that is essential for cytokinesis in the mitotic cycle; reviewed by (Gould and Simanis 1997; Simanis 2003; Wolfe and Gould 2005; Goyal et al. 2011). It is tightly regulated through the mitotic cell cycle to assure proper coordination of mitosis and cytokinesis, and during meiosis, where it is essential for spore formation (Krapp et al. 2006). Signaling requires the activity of three protein kinases, each of which has a regulatory subunit (kinase regulator); Cdc7p-Spg1p (Fankhauser and Simanis 1994; Schmidt et al. 1997; Mehta and Gould 2006), Sid1p-Cdc14p (Fankhauser and Simanis 1993; Guertin et al. 2000; Guertin and McCollum 2001), and Sid2p-Mob1p (Sparks et al. 1999; Hou et al. 2000; Salimova et al. 2000). SIN signaling is modulated by the nucleotide status of the GTPase Spg1p (Schmidt et al. 1997; Sohrmann et al. 1998), which is determined by the balance of spontaneous nucleotide exchange, a putative GEF, Etd1p (Daga et al. 2005; Garcia-Cortes and McCollum 2009), and a GAP, Cdc16p (Minet et al. 1979; Fankhauser et al. 1993), with which Spg1p interacts through a scaffold, Byr4p (Song et al. 1996; Furge et al. 1998; Furge et al. 1999). The SIN is also activated by the mitotic regulator Plo1p (Tanaka et al. 2001). Loss of SIN signaling produces multinucleated cells, while constitutive activation of the SIN results in multiseptated cells (Minet et al. 1979; Fankhauser et al. 1993; Song et al. 1996). Ectopic activation of the SIN promotes CAR and septum formation from any stage of the cell cycle (Fankhauser and Simanis 1994; Ohkura et al. 1995; Schmidt et al. 1997; Guertin et al. 2002). Analysis of the localization of SIN proteins and their activity in mutant backgrounds has led to the proposition that the order of action of the SIN proteins in the mitotic cell cycle is Cdc7p-Spg1p, then Sid1p-Cdc14p, and finally Sid2p-Mob1p (Sparks et al. 1999; Guertin et al. 2000). Active Cdc2p inhibits the SIN early in mitosis, and CDK inactivation is required for septum formation (Yamano et al. 1996; He et al. 1997; Guertin et al. 2000; Chang et al. 2001; Dischinger et al. 2008).

Components of the SIN have been conserved through evolution; in Saccharomyces cerevisiae, the equivalent signaling network is known as the mitotic exit network (MEN), which regulates mitotic exit and is also important for cytokinesis (reviewed by Bedhomme et al. 2008; Meitinger et al. 2012). The SIN/MEN protein kinases belong to the nuclear DBF-2-related (NDR) and Ste20 families, the mammalian counterparts of which are important in cell proliferation and growth control (Zhang et al. 2009; Zhao et al. 2011).

SIN signaling requires association of SIN proteins with the spindle pole body

Association of SIN proteins with the spindle pole body (SPB) at various points of the cell cycle plays an important role in SIN regulation, and therefore in the coordination of mitosis and cytokinesis (reviewed by Simanis 2003; Wolfe and Gould 2005; Krapp and Simanis 2008; Lattmann et al. 2009; Goyal et al. 2011). SIN proteins associate with the SPB via a tripartite scaffold composed of Ppc89p, Sid4p, and Cdc11p (Chang and Gould 2000; Krapp et al. 2001; Tomlin et al. 2002; Morrell et al. 2004; Rosenberg et al. 2006). To date, all SIN functions require Sid4p and Cdc11p (Balasubramanian et al. 1998; Hachet and Simanis 2008). Laser ablation of SPBs suggests that at least one SPB must be intact during anaphase B for cytokinesis to occur (Magidson et al. 2006). Duplication of the S. pombe SPB appears to be conservative (Ding et al. 1997; Grallert et al. 2004), generating distinguishable old and new SPBs (Grallert et al. 2004). Some of the SIN proteins (Cdc7p, Cdc16p, Byr4p, Sid1p, and Cdc14p) show asymmetric localization on one of the two SPBs during mitosis (Sohrmann et al. 1998; Cerutti and Simanis 1999; Guertin et al. 2000; Li et al. 2000), which is thought to be important for regulating SIN activity (Garcia-Cortes and McCollum 2009). A summary of the localization behavior of the SIN proteins in mitosis and meiosis can be found in Simanis (2003); Krapp and Simanis (2008); Lattmann et al. (2009); Goyal et al. (2011).

PP2A in S. pombe

PP2A is a heterotrimer, composed of a scaffold subunit (A) that assembles a catalytic subunit (C) and regulatory/targeting subunit (B, B′, B′′, or B′′′). The activity of protein phosphatase 2A activity is regulated by its subunit composition, phosphorylation, and PTPA proteins (reviewed by Shi 2009b). The S. pombe genome contains one PP2A scaffold protein encoded by paa1, which is essential (Kinoshita et al. 1996); germinating spores fail to proliferate and form a rounded cell. Three catalytic subunits have been characterized, which are encoded by the ppa1 and ppa2 genes (Kinoshita et al. 1990), and the ppa3 gene (Singh et al. 2011). Neither ppa1 nor ppa2 is essential, but the ppa1–Δ ppa2–Δ mutant is inviable (Kinoshita et al. 1990). The ppa3 gene is not essential, and the double null mutants ppa2–Δ ppa3–Δ and ppa1–Δ ppa3–Δ are also viable (Singh et al. 2011). The major subunit is encoded by ppa2; deletion of ppa1 has little effect upon the cell, while ppa2–Δ cells are reduced in size (Kinoshita et al. 1990). Ppa3p and Paa1p are part of the SIN inhibitory phosphatase (SIP) complex, which establishes the asymmetry of SIN protein partitioning on the anaphase SPBs (Singh et al. 2011). S. pombe has a single B-family subunit, which is encoded by the pab1 gene (Kinoshita et al. 1996); deletion of pab1 affects cell morphology and renders cells cold sensitive (Kinoshita et al. 1996). The par1 and par2 genes encode the two B′-family subunits. Par1p is the major form (Jiang and Hallberg 2000; Le Goff et al. 2001; Tanabe et al. 2001); deletion of par1 affects septum positioning and also renders cells cold sensitive and hypersensitive to loss of PP2B/calcineurin (Jiang and Hallberg 2000; Le Goff et al. 2001; Tanabe et al. 2001); in contrast, deletion of par2 has little effect on the cell per se (Jiang and Hallberg 2000; Tanabe et al. 2001).

Regulation of the SIN by phosphoprotein phosphatases

The B and B′ regulatory subunits of PP2A (Jiang and Hallberg 2001; Le Goff et al. 2001; Lahoz et al. 2010), the CDC14-family phosphoprotein phosphatase Flp1p/Clp1p (Cueille et al. 2001; Trautmann et al. 2001), and calcineurin/PP2B (Lu et al. 2002) have all been identified as regulators of the SIN through genetic analysis, although their targets are mostly unknown. Mutation of par1, rescues mutants of spg1, cdc7, and cdc11 (Jiang and Hallberg 2001; Le Goff et al. 2001) and increases the number of cells in which Cdc7p segregates symmetrically during anaphase. Mutation of pab1 can rescue the loss of etd1 and mutations in cdc11, mob1, and sid2. It also restores localization of Cdc7p to one SPB during anaphase in etd1–Δ cells (Lahoz et al. 2010). Mutants of the SIP complex show symmetrical segregation of Cdc7p during anaphase and increased sensitivity to SIN signaling (Singh et al. 2011), suggesting that the SIP complex is an inhibitor of SIN signaling. Ablation of either SIP or PP2A-Par1p activity delays the dephosphorylation of Cdc11p at the end of mitosis (Krapp et al. 2003; Singh et al. 2011). Together, these data suggest that PP2A complexes regulate localization of the SIN proteins during anaphase.

Materials and Methods

S. pombe manipulation

Standard techniques were used for the growth and manipulation of fission yeast (Moreno et al. 1991). Unless otherwise indicated, cells were grown in yeast extract (YE) medium. Phloxin B was added to solid media contained to aid identification of colonies containing dead cells as required. Crosses were performed on EMM2 without a nitrogen source; tetrads were dissected using a Singer micromanipulator. FACS analysis was performed according to the protocol of Sazer and Sherwood (1990). To define the position of cells that have G1 DNA content for each strain, hydroxurea (HU) was added to exponentially growing cells to a final concentration of 12 mM, and samples were taken 1 and 2 hr thereafter.

Genetic screen

To isolate spontaneous revertants of cdc7-24, 1 × 10⁸ cdc7-24 cells of exponentially growing culture at 25° were incubated at 32° for 5 days. Approximately 700 colonies were obtained. These were replica plated onto YE plates containing phloxin B at 19°; 6 colonies were cold sensitive. All mutants were backcrossed to wild type at least twice before detailed analysis.

Molecular biology

Standard DNA manipulation techniques were used (Sambrook et al. 1989). A list of oligonucleotides used in this study is given in Supporting Information, Table S3.

Sequencing of ppa2 and ypa2

The ppa2 gene was amplified from wild type and csr7-6 strains using primers VS 1125 (5′ SacI restriction site) and VS 1126 (5′ EcoRI restriction site) and cloned into the pBSKS(−) vector. The ypa2 gene was amplified, cloned, and sequenced from wild type and csr7-7 strains using primers VS 1177 (5′ EcoRI restriction site) and VS 1178 (5′ ClaI restriction site). Sequencing was undertaken by Microsynth.

Deletion and tagging of strains

The ypa2–Δ and ypa1–Δ strains were constructed by oligonucleotide-mediated targeting (Bahler et al. 1998). The ypa2 gene was deleted with oligonuleotide pair VS 1186 and VS 1187, and ypa1 was deleted using pair VS 1201 and VS 1202, selecting ura4⁺ and G418 resistant colonies, respectively. ypa2-HA was generated using VS 1207 and VS 1208 and ypa1-HA was made using VS 1203 and VS 1204, selecting G418-resistant colonies. Correct integration was confirmed by PCR. The mutants were crossed to wild type twice before use.

The integrated N-terminal GFP-tagged copy of ypa2 gene expressed under atb2 promoter at leu1 locus was created using VS 1355 (5′ NheI restriction site) and VS 1365 (5′ SacI restriction site) to amplify the ypa2 gene, which was subsequently cloned into pINT5 (Fankhauser and Simanis 1994). The ypa1 gene was GFP-tagged using VS 1351 (5′ NheI restriction site) and VS 1352 (5′ SmaI restriction site) to amplify the gene.

Cell synchronization and protein extracts

To analyze the expression of Ypa2p-HA and Ypa1p-HA through the cell cycle, cells were synchronized by the addition of hydroxyurea (Sigma) to logarithmically growing cultures to a concentration of 12 mM for 5 hr at 29°; cells were washed and resuspended in YE at 29°. Samples were taken every 20 min and proteins were extracted from 2 × 10⁸ cells using the trichloroacetic acid (TCA) protocol (Foiani et al. 1994). Protein extracts were analyzed on 12% SDS–PAGE gels and transferred to nitrocellulose membranes (Protran, Whatman). The membranes were probed with primary antibodies against HA (12CA5) and α-tubulin (TAT-1; Keith Gull, Oxford, UK). Bound antibodies were detected using secondary antibodies conjugated to horseradish peroxidase and ECL reagents (Amersham).

Microscopy

Exponentially growing cells were fixed with 70% ethanol and stained with DAPI and calcofluor as described previously (Balasubramanian et al. 1997). The samples were visualized using an Axiophot microscope (Zeiss) with a 100× N.A. 1.4 lens and images were captured with a Nikon Coolpix 990 camera. The transmission images of exponential growing cultures were taken with a Perkin-Elmer spinning disk confocal microscope using a Plan-S-Apo 100×/1.40 lens. Dividing cells were measured and quantified using ImageJ software. To analyze F-actin patches, the crn1-GFP allele (Pelham and Chang 2001) was crossed into the strains of interest; exponentially growing cultures were fixed by addition of paraformaldehyde to 3.7% (w/v from 37% w/v solution). GFP-atb2, expressed from the atb2 promoter and integrated at leu1 (Krapp et al. 2006), was crossed into the strains and living cells were imaged using a Plan-S-Apo 60× N.A. 1.42 objective lens mounted on a Perkin-Elmer spinning disk confocal microscope. To analyze the localization of Sid2p-GFP, cells were fixed for 1 min by addition of paraformaldehyde to 3.7% (w/v from 37% w/v solution). Cells were then washed and examined. Cdc7p-GFP was detected in living cells.

Results

Identification of cold-sensitive suppressors of the SIN mutant cdc7-24

We screened for regulators of the SIN by isolating suppressors of cdc7-24, applying the classical genetic screening logic of Moir and Botstein (Moir et al. 1982) to identify regulators of a pathway. Survivors were selected using a low restrictive temperature, followed by analysis of mutants that displayed a cold-sensitive phenotype; ∼700 colonies were formed at 32°, of which 6 were cold sensitive. These mutants were named cold-sensitive revertant of cdc7 (csr7-X); tetrad dissection of crosses to wild type revealed that the cold sensitivity and ability to rescue cdc7-24 were tightly linked and segregated 2:2. Allelism tests indicated that there were two complementation groups (A and B); group A was defined by a single mutant (csr7-6), suggesting that this screen is probably not saturated.

Complementation group A mutant maps to the PP2A catalytic subunit ppa2

Rescue of the cold sensitivity of the csr7-6 mutant by transformation with a genomic library revealed that the mutant was rescued by a plasmid that carried genes SPBC14H5.09c, the ppa2 gene, and SPBC14H5.07c. Subcloning indicated that only the ppa2 gene would rescue the csr7-6 mutant (not shown). A cross of ppa2–Δ with csr7-6 did not produce any recombinants in >300 progeny examined. Sequencing of the ppa2 gene in the csr7-6 mutant revealed a point mutation that changes D70 to N. This mutant is referred to hereafter as ppa2-6. Analysis of nonsporulating diploids indicated that ppa2-6 was recessive to ppa2⁺ with regard to both its cold sensitivity and ability to rescue cdc7-24 at 32°. In contrast to ppa2–Δ, which continues to divide at 19° (Figure 1, A and B) (Kinoshita et al. 1990), ppa2-6 is cold sensitive and undergoes only a few divisions after a shift to 19° (Table S1, Figure 1, A and B).

Figure 1 .

Characterization of suppressors of cdc7-24 that are cold sensitive. (A) Cells of the indicated genotype were grown to late exponential phase at 25° in YE medium. Serial 10-fold dilutions were made and spotted on YE agar plates at the indicated temperatures. The plates were incubated until the wild-type cells had formed colonies at the indicated temperature (2–5 days). (B) Exponentially growing cells of the indicated genotype were spread onto YE agar plates containing phloxin B. The plates were incubated at 19° for 3 days, and typical colonies were photographed. The green channel of the captured image is shown; dead cells are dark, as they are unable to exclude phloxin B. Bar, 10 μm. (C) The indicated mutants were grown to exponential phase at 25° in YE and then shifted to 29° for 6 hr or 36° for 5 hr. Cells were fixed, and stained with DAPI and calcofluor. Bar, 10 μm. (D) Mutants were grown to exponential phase at 25° in YE and then shifted to 32° for 6 hr. Cells were fixed and stained with DAPI and calcofluor. Bar, 10 μm.

Previous studies have demonstrated that the deletion of ppa2, but not ppa1, reduced the amount of PP2A activity in cell extracts (Kinoshita et al. 1990). The mutation found in ppa2-6 affects an amino acid involved in coordinating the metal ion required for catalysis and the equivalent mutation reduces the activity (k_cat) of λ phosphatase >10,000-fold (Zhuo et al. 1994). The double mutant ppa2-6 ppa1–Δ was synthetically lethal (Table 4), even when tetrads were dissected at 36° and 32°, the permissive temperature for ppa2-6. Since the ppa1–Δ ppa2–Δ double mutant is inviable (Kinoshita et al. 1990), this suggests that Ppa2-6p has little residual catalytic activity even at its permissive temperature.

Table 4. Genetic interactions between ppa2-6, ppa2–Δ, ppa1–Δ, ypa1–Δ, ypa2-7, and ypa2–Δ.

Mutant	ppa1−Δ	ppa2−Δ	ppa2-6	ypa1−Δ	ypa2−Δ	ypa2-7
ppa1−Δ	ND	Synthetically lethal: single, small, rounded cell	Synthetically lethal: 1 or 2 small rounded cells at 32° and 36°	Cold sensitive at 19°, similar to ypa1−Δ	Cold sensitive at 19°, similar to ypa2−Δ	Cold sensitive at 19°, similar to ypa2-7
ppa2−Δ		ND	ND (allelic)	Cells are wee-like ppa2−Δ at >29° and cold sensitive-like ypa1−Δ at 19°	Additive negative effect: cells are cold sensitive at 19°, very slow growing wee at >29°	Cells are wee at >25° and cold sensitive at 19°
ppa2-6			ND	Slow growing colony of rounded, lysed cells	Synthetically lethal: 1–2 rounded cells after germination	Synthetically lethal: 1–2 rounded cells after germination
ypa1−Δ				ND	Synthetically lethal: cells undergo 1–2 divisions dying as rounded cells	Synthetically lethal: cells undergo 1–2 divisions dying as rounded cells
ypa2−Δ					ND	ND (allelic)
ypa2-7						ND

Mutants were crossed at 25°. Tetrads were dissected and spores were germinated on complete medium at 32°. In cases were lethal interactions were observed, other temperatures were tested, but no temperature in the range of 19°–36° altered the result obtained. At least 10 tetrads were analyzed for all double mutant combinations. ND, not done.

Complementation group B mutants map to the PP2A activator (PTPA) homolog SPAC1782.05

We chose one representative mutant from complementation group B, csr7-7, for further study. Analysis of the diploids indicated that the csr7-7 was recessive to csr7⁺ both with regard to its cold sensitivity and rescue of cdc7-24 at 32°. During genetic analysis of csr7-7, we discovered a close linkage (40PD in 40 tetrads) of the mutation to flp1 (Cueille et al. 2001); genes in the vicinity of flp1 were therefore amplified and sequenced. A mutation was found in the gene SPAC1782.05, which changes E142 to K. This gene encodes the S. pombe ortholog of the mammalian PTPA gene (RRD2/YPA2 in S. cerevisiae), a peptidyl proline isomerase that activates PP2A catalytic subunits (reviewed by Shi 2009a). Analysis of the S. pombe genome has shown, as in S. cerevisiae, there is a second gene, SPAC4F10.04, whose product is closely related to SPAC1782.05 (Leulliot et al. 2006). We have named these genes ypa1 (SPAC4F10.04) and ypa2 (SPAC1782.05) (ypa: yeast phosphatase activator). Hereafter, the csr7-7 mutant will be referred to as ypa2-7. Oligonucleotide-mediated targeting was used to generate deletion alleles of the ypa1 and ypa2 genes (referred to hereafter as ypa1–Δ and ypa2–Δ). Analysis revealed that neither ypa1 nor ypa2 is essential for spore germination and colony formation at temperatures above 25°; however, neither of the null mutants was able to form colonies at 19° (Table S1, Figure 1, A and B). Overall, the phenotype and genetic interactions of ypa2-7 resembles that of ypa2–Δ (see below) in all the assays presented in this study, suggesting that the mutant allele retains little or no activity. This is supported by the observation that the ypa2⁺ gene rescues the cold-sensitive ypa2-7 mutation, while the mutant form of the gene does not (data not shown).

Analysis of the genetic interactions of ppa2-6, ppa2–Δ, ypa1–Δ, and ypa2–Δ with SIN mutants

Since the csr7-X mutants were obtained in the background of cdc7-24, we examined whether they could suppress other SIN mutants. Double mutants were made by tetrad dissection, and their colony formation ability was examined at different temperatures. The data are presented in Table 1, Table S1, and Figure 1A, and the salient points are summarized below. The ypa2-7 and ypa2–Δ mutants rescued heat-sensitive alleles of spg1, cdc14, cdc7 (Figure 1A), cdc11, plo1, sid2 (Figure 1D, Figure S1), and mob1. The cold sensitivity of etd1–Δ was rescued by ypa2–Δ, permitting colony formation down to 25°. A strong negative genetic interaction was observed between cdc16-116 and both ypa2-7 and ypa2–Δ, reducing the restrictive temperature to <29° (Table 1, Table S1). Thus, loss of Ypa2p function rescues mutants that have reduced SIN signaling, while being hypersensitizing cells to increased SIN signaling. The failure to rescue cdc7-24 and other SIN alleles at high temperatures suggests that ypa2–Δ does not bypass the requirement for the SIN.

Table 1. Summary of genetic interactions between SIN mutants and ppa2-6, ppa2-D, ppa1-D, ypa1-D, ypa2-7, and ypa2-D.

SIN mutant	ypa1–Δ	ypa2–Δ ypa2-7	ppa2–Δ ppa2-6	ppa1–Δ
cdc7	None	Good	Good	Very weak
spg1	None	Good	None	None
sid1	None	None	None	None
cdc14	None	Good	None	None
sid2	None	Good	Weak	None
mob1	Weak	Good	Good	Weak
etd1	ND	Good	Good	ND
cdc16	None	Negative	Negative	None
sid4	None	None	None	None
cdc11	None	Good	Good	None
plo1	None	Good	Good	ND

This table is an overview of the data presented in Table S1. A summary for each SIN mutant is given. Good, good to excellent colony formation; weak, poor rescue, or only at low temperatures; none, no significant difference with the single SIN mutant. For details of the temperature range and alleles used, please consult Table S1.

The genetic interactions of ppa2–Δ and ppa2-6 with SIN mutants was similar to those of ypa2-7 and ypa2–Δ, the main difference being a failure to rescue mutants in spg1 and cdc14 (Table 1, Table S1); like ypa2–Δ, these mutants do not bypass the requirement for the SIN. A strong negative genetic interaction was observed between cdc16-116 and both ppa2–Δ and ppa2-6, with the restrictive temperature decreased to <29° (Table 1, Table S1, Figure 1C). Examination of the mutant cdc16-116 ppa2-6 revealed the presence of swollen septated cells, as well as septated mononucleated cells, characteristic of the inactivation of cdc16 (Minet et al. 1979), while cdc16-116 cells retained a normal morphology at 29° (Figure 1C). Since Ppa2-6p has little, if any, activity, these data are consistent with PP2A being an inhibitor of the SIN, as suggested by the analysis of regulatory subunit mutants (see Introduction).

Analysis of SIN–ypa1–Δ double mutants revealed weak rescue of all the tested alleles of mob1 and one allele of sid2 (Table 1, Table S1), but no other components of the SIN. In contrast to ypa2–Δ, no negative interaction was observed between ypa1-D and cdc16-116. Deletion of ppa1, the minor catalytic subunit of PP2A, produced only a weak rescue of mob1 and cdc7 (Table 1, Table S1). Mutants in sid1 and sid4 were not rescued by either ppa1–Δ, ppa2-6, ppa2–Δ, ypa1–Δ, ypa2–Δ, or ypa2-7. None of the SIN mutants could rescue the cold sensitivity of ypa1–Δ, ypa2–Δ, ypa2-7, or ppa2-6 at 19°.

Analysis of Ypa1p and Ypa2p protein levels and localization

To study the levels of Ypa1p and Ypa2p through the cell cycle, the proteins were tagged C terminally with a HA epitope (Bahler et al. 1998); in both cases, cells expressing the tagged allele appeared phenotypically wild type and were not cold sensitive. Protein samples from synchronous cultures were analyzed by Western blotting; the steady state levels of Ypa1p-HA (Figure 2A) and Ypa2p-HA (Figure 2B) did not vary significantly through the cell cycle. Analysis of proteins extracted from a ypa1-HA ypa2-HA strain revealed that Ypa2p-HA was >30-fold more abundant than that of Ypa1p-HA (Figure 2C, Figure S2A), indicating that Ypa2p is the major PTPA homolog in fission yeast.

Figure 2 .

Characterization of Ypa1p and Ypa2p. (A and B) Strains expressing Ypa1p-HA or Ypa2p-HA were synchronized as described in Materials and Methods. Protein samples were prepared at the indicated times (minutes) after release from the block. Western blots were probed with either 12CA5 (HA), or TAT-1 (α-tubulin) as indicated. Cell cycle progression data are represented by the percentage of septated cells, indicated at each time point. (C) Proteins were prepared from strains expressing Ypa1-HA (lane 2), Ypa2-HA (lane 3), or both alleles (lane 1). Serial dilution of the sample indicates that the Ypa2p-HA signal in lane 1 is at least 30 times stronger than the Ypa1p-HA signal (Figure S2A). (D) Cells of the indicated genotype were grown to exponential phase and GFP was visualized as described in Materials and Methods. Bar, 10 μm.

To localize Ypa1p and Ypa2p, we tagged the proteins with GFP. Though cells bearing the C terminally GFP-tagged ypa1 and ypa2 alleles appeared normal, we did not detect any discrete signal (not shown). We therefore tagged the proteins with GFP at their N termini, expressing them using the S. pombe atb2 promoter; for technical reasons, these constructions were integrated at the leu1 locus (Fankhauser and Simanis 1994), followed by crossing of the tagged alleles into their respective deletion backgrounds. Both the GFP-Ypa1p and GFP-Ypa2p proteins showed a uniform cytoplasmic localization and a faint nuclear signal. There was no apparent change in the localization of the tagged proteins at any cell cycle stage (Figure 2D). Strongly increased expression of either ypa1 or ypa2 from the nmt1 promoter (Maundrell 1990) did not affect the colony forming ability or phenotype of wild-type cells (Figure S2B).

Analysis of the phenotype of ypa2–Δ cells

Measurement of ypa2–Δ cells revealed that they divide at a reduced cell length at both the permissive and restrictive temperatures, similar to ppa2–Δ (Table 2). Since a reduced cell length can indicate that a cell is advanced into mitosis (Nurse 1975), we crossed the ypa2–Δ mutant with a mutant in the mitotic inhibitor wee1 (Nurse 1975; Russell and Nurse 1987). The double mutant wee1-50 ypa2–Δ was synthetically lethal, with the germinating spore giving rise to one or two small, rounded cells at 29° (data not shown). The double mutant ppa2–Δ wee1-50 is also synthetically lethal at high temperatures (Kinoshita et al. 1993), indicating that, like Ppa2p, Ypa2p shares an essential function for survival with Wee1p.

Table 2. Cell length at division of S. pombe ypa and ppa mutants.

	Cell length (μm)
Strain	32°	19°
Wild type	15.8 ± 0.8	14.1 ± 1.1
ppa1−Δ	13.4 ± 0.7	13.1 ± 1.0
ppa2−Δ	10.1 ± 1.4	11.8 ± 1.0
ppa2-6	11.7 ± 1.0	9.6 ± 0.9
ypa1−Δ	14.1 ± 1.4	12.6 ± 1.4
ypa2−Δ	9.6 ± 0.9	9.6 ± 0.7
ypa2-7	10.4 ± 1.0	10.1 ± 0.9

Cells were grown to exponential phase in YE medium at 32°; a fraction of the culture was shifted to 19° for 20 hr. Samples were taken before and after shift and cell length was determined as described in Materials and Methods. Length of septating cells is indicated, with standard deviation: n = 50 for each condition.

Cells such as wee1 mutants, which are advanced into mitosis at a reduced cell size, expand the G1 phase of the cell cycle to permit growth before commitment to the next cell cycle (Nurse 1975). To determine whether the reduced cell size at division of ypa2–Δ cells also resulted in the appearance of G1 cells, we performed a FACS analysis on exponentially growing cells (Figure S3). Since the FACS profile of S. pombe is affected by cell size and other factors (Sazer and Sherwood 1990; Carlson et al. 1997), we established the position of G1 cells for each population by treatment with HU to inhibit DNA synthesis. In the case of wild-type cells, the single peak seen in exponential growth first developed a “shoulder” to the left and then moved to the left completely as all the cells arrested (Figure S3). Examination of ypa2–Δ cells also revealed a single peak, with a shoulder to the right-hand side, probably reflecting the presence of cells that had not completed separation (see below). Following addition of HU, these peaks shifted to the left, before coalescing into a single peak. Due to the smaller cell size, these peaks are shifted to the left compared to wild type. These data suggest that the ypa2–Δ population consists mainly of G2 cells and is consistent with data from the Yanagida laboratory concerning the ppa2–Δ strain (Kinoshita et al. 1993), which also divides at a reduced size, but shows no G1 population in FACS analysis, confirmed in our analysis (Figure S3). These data were supported by monitoring of the cell number increase after addition of HU. In the wild-type culture, the number of cells increased 2.01-fold before reaching a plateau (not shown), giving an “execution point” of ∼0, consistent with previous data (Nasmyth and Nurse 1981). The cell number increase in the ypa2–Δ culture was 2.07, which also gives a similar value. If there had been a significant proportion of G1 cells, a value significantly <2.0 would have been observed. Together, these data indicate that, despite dividing at a reduced cell size, the ypa2–Δ mutant does not expand the G1 phase significantly. The reason for the reduced cell size at division and lethality with wee1 will be the subject of future studies.

Approximately one-quarter of the dividing cells misplaced the division septum at the permissive temperature, rising to almost half at the nonpermissive temperature (Table 3). Staining with DAPI and calcofluor revealed a delay in completing cell separation at the restrictive temperature (at 19°, 3.7%, n = 381 and at 32°, 0.3%, n = 313; Figure 3A). This phenotype was not observed in wild type (n = 220). We did not detect any significant abnormalities in nuclear structure or mitosis (Figure 3A).

Table 3. Analysis of septation in ypa and ppa mutants.

Strain	Temperature	Total	Septation (%)	Eccentrically placed septum (%)
Wild type	32°	519	11.9	0.0
	19°	521	11.5	1.7
ppa1−Δ	32°	515	10.1	0.0
	19°	516	10.3	1.9
ppa2−Δ	32°	531	14.3	6.6
	19°	503	9.9	10.0
ppa2-6	32°	511	12.7	23.1
	19°	524	12.4	26.2
ypa1−Δ	32°	536	11.8	0.0
	19°	535	9.7	3.8
ypa2−Δ	32°	506	17.2	21.8
	19°	526	16.9	46.1
ypa2-7	32°	544	16.2	20.5
	19°	513	15.4	48.1

Percentage of septated cells was determined by examining transmitted light images of the cultures analyzed above. Percentage of septated cells was determined. Number of cells analyzed is indicated. Percentage of septated cells with an eccentrically placed septum was calculated.

Figure 3 .

Analysis of the phenotypes of ypa1–Δ and ypa2–Δ. Cells of the indicated genotype were grown at 32° and then shifted to 19° for 18 hr. Cell were fixed and stained with DAPI and calcofluor. Bar, 10 μm throughout. (A) Analysis of ypa1–Δ and ypa2–Δ. The arrows in ypa2–Δ indicate cells that have delayed complete separation. (B) Analysis of ypa2–Δ ppa2–Δ. Note the presence of asymmetrically placed septa, cells with elongated “stretched” nuclei, indicative of problems with chromosome separation, and (inset) two multinucleate cells that have failed to separate. (C) Circled 1 indicates a representative cell with an eccentrically placed septum. Note that in these cells, the nuclei are also flattened and positioned at the cell periphery. The arrowed 2 indicates representative cells showing lagging chromosomes. The circled 3 (inset) shows a pair of cells that have failed to separate, one of which has initiated the subsequent mitosis. The circled 4 indicates a cell with aberrant morphology. (D) ppa2-6 cells at 32°. (E) Tetrads from a cross ypa1–Δ ppa2-6 were dissected at 32°. Colonies were photographed prior to replica plating. The genotype of the microcolony was inferred from the phenotypes after replica plating.

Localization of F-actin patches using crn1-GFP revealed some cells in which they were delocalized all over the cortex, consistent with the rounded morphology of the cells (Figure 4B). We used GFP-Atb2p to visualize microtubules; the expected microtubule configurations (Hagan and Hyams 1988) were seen in wild-type cells (Figure 4F). Examination of ypa2–Δ revealed the presence of numerous short microtubules in interphase cells at both permissive (Figure 4G) and restrictive temperatures (not shown). The mitotic spindles rarely had visible astral arrays and were often bent along one side of the cell. A strong medial signal was often observed in these cases, which may correspond to the equatorial microtubule organizing center (EMTOC) (Heitz et al. 2001) or the onset of reformation of the interphase microtubule array.

Figure 4 .

Analysis of the cytoskeleton of ypa1–Δ, ypa2–Δ, and ppa2-6. Mutants expressing either crn1-GFP (A–E) or GFP-atb2 (F–I) were grown to exponential phase at 32° and GFP was visualized. For the 19° images, cells were shifted to 19° for 18 hr prior to analysis. Bar, 10 μm throughout. The circled I is an interphase cell, the circled M is a mitotic cell, and the circled P is a cell with a postanaphase array.

Together, these data lead us to conclude that Ypa2p is involved in determining the timing of mitotic commitment, establishing cell morphology, positioning of the division site, regulation of the SIN, and in completion of cytokinesis.

Analysis of the genetic interactions of ypa2–Δ with PP2A subunits

Tetrad dissection was used to create the double mutant ypa1–Δ ypa2–Δ, which was found to be inviable (Table 4). The spores germinated and underwent a maximum of two divisions before arresting as small, rounded cells. Similar results were obtained from making the mutant ypa2-7 ypa1–Δ (not shown). Thus, at the permissive temperatures for the individual null mutants, Ypa1p and Ypa2p have a shared essential function.

Next, we examined the genetic interactions of ypa2–Δ with mutants of PP2A regulatory and catalytic subunits. The double mutant ypa2–Δ ppa2-6 was synthetically lethal (Table 4), while the ypa2–Δ ppa2–Δ double mutant was viable, but slow growing and incapable of colony formation at 25° (Table 4, Figure S4). At 32° the septation index was very high (>50%), while at 19°, cells divided at a very small size (<7 μm), often asymmetrically. We also saw evidence of mitotic aberrations and loss of coordination between mitosis and cytokinesis (Figure 3B). The ypa2–Δ mutant was also strongly synthetically sick with par1–Δ and pab1–Δ, forming only small, slow-growing colonies (Table 5).

Table 5. Genetic interactions of ppa and ypa mutants with PP2A regulatory subunit mutations.

	par1−Δ	par2−Δ	pab1−Δ
Wild type	Cold sensitive at 19°	No phenotype	Weak cold sensitive
ypa1−Δ	Phenotype of par1−Δ: reduced restrictive temperature	Phenotype of ypa1−Δ: no additive effect	Phenotype of pab1−Δ and cold sensitivity of ypa1−Δ at 19°
ypa2−Δ	Strong negative interaction	Phenotype of ypa2−Δ: no additive effect	Strong negative interaction
ppa1−Δ	par1−Δ phenotype: no negative interaction	Wild-type cells	Phenotype of pab1−Δ: no negative interaction
ppa2−Δ	Phenotype of par1−Δ: strong negative interaction	ppa2−Δ phenotype	Phenotype of pab1−Δ: strong negative interaction

Double mutants of the PP2A regulator subunit null alleles indicated at the top of the table were made with the ppa and ypa mutants indicated. Crosses were analyzed by tetrad dissection and replica plating: spores were initially allowed to form colonies at 32°.

The cold sensitivity of ypa2–Δ was not rescued by increased expression of ypa1, indicating that it cannot provide the essential function of Ypa2p at low temperature. Likewise, overexpression of either ppa2, which encodes the major catalytic subunit of PP2A, or ppa1, or the S. pombe SIT4 ortholog ppe1, could not rescue ypa2–Δ cold sensitivity (data not shown).

Analysis of the phenotype and genetic interactions of ypa1–Δ

Measurement of ypa1–Δ cells at 32° showed that, in contrast to ypa2–Δ, they divided at wild-type length, with normal positioning of the division plane (Figure 3A, Table 2, Table 3) and no apparent morphological abnormalities. At 19°, the cell length at division decreased, but cells also displayed a more rounded morphology, which may account for this (Figure 3A); ∼4% of dividing cells now showed errors in septum positioning (Table 3). The presence of anucleate cell compartments (Figure 3A) also indicates a loss of coordination between mitosis and cytokinesis. Examination of microtubules at 32° did not reveal any major abnormalities, but at 19°, we observed that interphase cells contained many short microtubules and the few anaphase spindles we observed lacked astral arrays (Figure 4H). Localization of Crn1-GFP did not reveal any striking anomalies (Figure 4C).

Crosses between ypa1–Δ and either ppa2–Δ, ppa1–Δ, or par2–Δ, did not show any significant synthetic negative effect, and the double mutants always showed the cold sensitivity of the ypa1–Δ mutant (Table 4, Figure S4). The double mutant ypa1–Δ ppa2-6 displayed a negative genetic interaction; colonies contained many misshapen and lysed cells and were small and slow growing compared to wild type (Figure 3E). A double mutant between ypa1–Δ and par1–Δ, which encodes the major B′ regulatory subunit of PP2A, showed a reduced restrictive temperature (Table 5, Figure S4). The double mutant ypa1–Δ pab1–Δ failed to form colonies at 19°, similar to ypa1–Δ, but showed the “wee” phenotype of pab1–Δ (Table 5, Figure S4).

The cold sensitivity of ypa1–Δ was not rescued by increased expression of ypa2, indicating it cannot provide the essential function of Ypa1p at low temperature (data not shown). We also found that overexpression of either ppa2, ppa1, or the S. pombe SIT4 ortholog ppe1 could not rescue the cold sensitivity of ypa1–Δ (data not shown). In contrast to ypa2–Δ, a cross of ypa1–Δ with wee1-50 was viable at 25° and above, but grew more slowly than either single mutant (not shown). The weaker genetic interactions of ypa1–Δ with the SIN, and with mutants in components of PP2A and its lower abundance relative to Ypa2p, suggests that it may play a subordinate role to Ypa2p in activating PP2A. Nonetheless, the fact that it is cold sensitive indicates that it has an essential role for cell viability under some conditions.

Analysis of the phenotype and genetic interactions of ppa2-6

Measurement of cell length at division showed that the ppa2-6 mutant is semi-wee at its permissive temperature (Table 2). Moreover, we observed that at both permissive temperature and restrictive temperature, approximately one-quarter of the dividing cells had misplaced the division septum (Table 3, Figure 3C). Analysis of ppa2–Δ also revealed defects in positioning the division plane, although the frequency was lower than ppa2-6 (Table 3). Staining of ppa2-6 cells with DAPI at permissive temperature did not reveal any significant abnormalities in nuclear morphology (Figure 3C). In contrast, at the restrictive temperature, we observed mitotic abnormalities such as lagging chromosomes (Figure 3D), and incomplete separation of cells (Figure 3D), similar to the phenotype of ypa2–Δ (see above). The cell separation defect was more pronounced at the restrictive temperature (3.7% of the population, n = 267) than the permissive temperature (0.5% of the population, n = 372). This phenotype was not observed in either wild-type (n = 220) or ppa2–Δ (n = 374) cells. At both the permissive and restrictive temperatures, we also observed cells with pear- or lemon-shaped morphology, indicating defects in polar growth in both ppa2–Δ and ppa2-6 (Figure 3D; data not shown for ppa2–Δ). Consistent with this, examination of Crn1p-GFP (Pelham and Chang 2001) revealed the presence of delocalized F-actin patches in both ppa2–Δ and ppa2-6 (Figure 4, D and E). This phenotype was not observed in wild type (Figure 4A) or ppa1–Δ cells (data not shown).

We used a GFP-atb2 allele to examine microtubules in ppa2-6. In ppa2-6 at permissive temperature, we observed apparently normal mitotic spindles, with astral microtubule arrays in anaphase (Figure 4I). At the nonpermissive temperature, we did not observe any astral microtubules during anaphase; occasional bent spindles were also seen (Figure 4I). In interphase cells, the microtubule bundles appeared shorter and more numerous than in wild type; this was exacerbated at the nonpermissive temperature (Figure 4I). Similar results were obtained for ppa2–Δ (data not shown). We conclude that PP2A–Ppa2p activity is required to maintain normal microtubule arrays and for positioning the division plane, cell morphology, and mitotic progression.

The mutant ppa2–Δ divides at a reduced cell size and is synthetically lethal with a wee1 mutant at elevated temperatures (Kinoshita et al. 1990). The double mutant of ppa2-6 wee1-50 was also synthetically lethal, with the germinating spore giving rise to one or two small, rounded cells at 32° (data not shown). Crosses of ppa2–Δ with par1–Δ and pab1–Δ revealed a strong synthetic negative interaction (Table 5, Figure S4). The cold sensitivity of ppa2-6 was not rescued by increased expression of either ypa1 or ypa2 (not shown). The similarity of the phenotypes and genetic interactions of ppa2-6 and ypa2–Δ suggest that many of the effects of ypa2–Δ are due to reduced PP2A activity.

Since the ppa2–Δ allele is not cold sensitive, we were surprised to obtain a recessive, cold-sensitive mutant of ppa2. We propose that the mutant protein Ppa2-6p acts as a “poison subunit” at low temperatures, reducing PP2A activity. Alternatively, it may titrate the A and B/B′ subunits away from Ppa1p. Our finding that the negative genetic interactions of ppa2-6 with ypa1–Δ and ypa2–Δ (synthetically negative and lethal, respectively) are stronger than those of ppa2–Δ (little effect and synthetic negative, respectively) are consistent with this view. In the presence of the wild-type Ppa2p in a heterozygous diploid, the amount of active PP2A–Ppa2p complexes produced would be sufficient for viability. Precedents for this type of mutant exist from previous analysis of S. pombe phosphoprotein phosphatases; the PP1 catalytic subunit dis2 is not essential, but a recessive cold-sensitive mutant was isolated in a genetic screen (Kinoshita et al. 1990).

Which PP2A activators and subunits are required for rescue of SIN mutants?

The data presented above show that inactivation of either ypa2 or ppa2 rescues many SIN mutants; however, the double mutant ypa2–Δ ppa2–Δ showed a synergistic negative effect, suggesting that their products do not function in a linear pathway (Table 4). This prompted us to examine the effects of inactivating combinations of PP2A activators and PP2A core subunits upon the ability to rescue cdc7-24 and mob1-R4; the data are presented in Table S2, and the key points are outlined below. Deletion of either the ppa1 or ppa2 catalytic subunit in a ypa2–Δ background did not compromise the ability to rescue either cdc7-24 or mob1-R4 (Table S2, rows 3, 7, and 9). Inactivation of ypa1 in ppa2–Δ did not affect the rescue of either cdc7-24 or mob1-R4 (Table S2, rows 1, 4, and 8). The double mutant ppa1–Δ ypa1–Δ gave the same result as either single mutant (Table S2, rows 2, 4, and 10).

Previous studies indicated that cdc7-24 could not be rescued by par1–Δ at 36° (Jiang and Hallberg 2001); we have confirmed this (not shown), but have also found that, like ppa2–Δ, par1–Δ will permit growth of cdc7-24 at 32°. Deletion of ypa1, ppa1, or ppa2 did not affect the rescue of cdc7-24 by par1–Δ (Table S2, rows 5, 12, 15, and 16). We also found that par1–Δ rescues mob1-R4 at both 32° (Table S2, row 5) and 36° (not shown). Examination of triple mutants showed that deletion of ypa1, ppa1, or ppa2 does not compromise the rescue of mob1-R4 by par1–Δ (Table S2, rows 5, 12, 15, and 16). Due to the strong negative genetic interaction between ypa2–Δ and par1–Δ (Table 5), we were unable to isolate the triple mutants with either cdc7-24 or mob1-R4.

Par2p is the minor B′ subunit in S. pombe; we found that the par2–Δ will rescue weakly mob1-R4, but will not rescue cdc7-24 (Table S2, row 6). Examination of triple mutants showed that deletion of par2 does not affect the rescue profile of ypa1–Δ, ypa2–Δ, ppa1–Δ, or ppa2–Δ (Table S2, rows 13, 14, 17, and 18). At present, due to the low spore viability in crosses involving pab1–Δ, we have not been able to construct SIN–ypa/ppa–pab1–Δ triple mutants.

Reduction of PP2A activity affects Cdc7p anaphase asymmetry and the localization of Sid2p to the CAR

Previous studies have implicated the B (Lahoz et al. 2010) and B′ (Jiang and Hallberg 2001) regulatory subunits of PP2A and the SIP complex (Singh et al. 2011) in regulating the localization of Cdc7p during mitosis. Therefore, we investigated whether Ppa2p, Ypa1p, and Ypa2p also affect the asymmetric segregation of Cdc7p to the new SPB during anaphase. In wild-type cells, Cdc7p is associated with both SPBs in early mitosis in all cells, becoming asymmetric in late anaphase as it remains associated only with the new SPB (Sohrmann et al. 1998; Grallert et al. 2004). In late anaphase Cdc7p-GFP was associated with only one SPB in the great majority of cells (59 of 60 cells; Figure 5A), as described previously (Sohrmann et al. 1998; Grallert et al. 2004). Examination of late anaphase ypa2–Δ cells revealed an increase in the number of cells that retained Cdc7p-GFP on both SPBs (8 of 60, compared to 1 of 60 in wild type; Figure 5A). Late anaphase ppa2–Δ (14 of 85) and ppa2-6 cells (10 of 70), also showed a failure to establish Cdc7p asymmetry (Figure 5A). These data therefore implicate Ypa2p and Ppa2p in establishing SIN protein asymmetry during anaphase. Our results are at variance with the study of Singh et al. (2011) with regard to the effect of the ppa2–Δ mutant on SIN asymmetry; although the reason for this is not clear, it may reflect the difference in the tagged reagent used; we have used a GFP tag, while Singh et al. used a m-Cherry tag.

Figure 5 .

Localization of Cdc7p-GFP and Sid2p-GFP in ypa2 and ppa2 mutants. Cells were grown to exponential phase at 25° and then shifted to 32° for 3 hr. They were fixed and analyzed as described in Materials and Methods. (A) Localization of Cdc7p-GFP in anaphase cells. The indicated strains were grown to exponential phase. Note the presence of Cdc7p on both SPBs in the ppa2–Δ and ypa2–Δ cells. See text for details. (B) Localization of Sid2p-GFP to the CAR is rescued in cdc7-24 ppa2–Δ and cdc7-24 ypa2–Δ mutants: Cells of the indicated genotypes were grown at the temperatures shown. Note the absence of Sid2p between the two anaphase nuclei in cdc7-24 at 32° (the nuclei are seen as dark patches, and the SPB signals are at the tip-proximal edge of the nucleus, indicating that the cell is in anaphase).

Association of Sid2p with the CAR is thought to be important for cytokinesis, and in a cdc7-24 mutant, Sid2p-GFP association with the SPB and the CAR is compromised at 36° (Sparks et al. 1999). In cdc7-24 cells at 25°, we observed Sid2p-GFP associated with the CAR and SPB as expected (Sparks et al. 1999). However, at 32°, when cdc7-24 dies, we found that Sid2p-GFP associated with the SPB but not the CAR in mitotic cells (Figure 5B). Analysis of the localization of Sid2p-GFP in the mutants cdc7-24 ypa2–Δ and cdc7-24 ppa2–Δ revealed that it was associated with the CAR, consistent with the fact that these double mutants can form colonies at 32°. Together, these data implicate PP2A as an important regulator of the establishment of SIN asymmetry and association of Sid2p with the CAR.

Discussion

We have screened for regulators of the SIN by isolating suppressors of cdc7-24, which encodes the presumed first kinase in the core SIN pathway. Our screen yielded a cold-sensitive allele of the main PP2A catalytic subunit, ppa2, and a mutant in one of the two S. pombe orthologs of PTPA.

S. pombe PTPA genes

Ypa1p and Ypa2p are two clear orthologs of the mammalian PTPA proteins in S. pombe. One of the roles of PTPA is thought to be to facilitate removal of the methylesterase that associates with the C subunit and reverses the ubiquitous methylation of the carboxyl group of the C-terminal residue of PP2A (reviewed by Shi 2009a). The Glu-to-Lys mutation found in ypa2-7 is located in a conserved domain that is essential for function in human PTPA (Van Hoof et al. 1998) and lies close to the region required for PTPA dimer formation (Leulliot et al. 2006; Magnusdottir et al. 2006). S. pombe has orthologs of both the PP2A methylase SPBP8B7.08c and esterase (SPBP4H10.17c), although these have not been extensively characterized to date.

The steady-state level of Ypa2p is significantly higher than that of Ypa1p. This suggests that the major activator of PP2A in S. pombe is likely to be Ypa2p. Both the ypa1–Δ and ypa2–Δ alleles are cold sensitive per se, and in neither case can this be rescued by expression of the other gene, or PP2A/SIT4 catalytic subunits. This indicates that at low temperatures, both Ypa1p and Ypa2p have distinct, essential functions. This is consistent with the phenotypic differences between the two null alleles; ypa2–Δ cells show strong morphological and polarity defects and divide at a reduced cell length, while ypa1–Δ are more wild type in appearance at permissive temperature.

However, since a ypa1–Δ ypa2–Δ null mutant is inviable at all temperatures at their permissive temperatures, Ypa1p and Ypa2p clearly share one (or more) overlapping essential function(s); given the similarity between the phenotypes of ypa1–Δ ypa2–Δ and ppa1–Δ ppa2–Δ, it seems likely that PP2A activation will be one of these essential functions.

Since ypa2–Δ is cold sensitive, while ppa2–Δ is not, and the ypa2–Δ mutant shows a very strong negative genetic interaction with both ppa2–Δ and ppa2-6, it seems unlikely that Ypa2p functions solely via activation of Ppa2p-containing PP2A complexes. S. cerevisiae Rrd1p and Rrd2p are required for the biogenesis of active PP2A that is specific for phosphorylated serine and threonine residues (Fellner et al. 2003; Hombauer et al. 2007). Rrd2p binds directly to the PP2A catalytic subunits Pph21p and Pph22p, and is more potent in the activation of PP2A catalytic subunits than Rrd1p. It has also been suggested that Rrd1p and Rrd2p may cooperate in the activation of PP2A (Van Hoof et al. 2005). Rrd1p also activates the PP2A-related phosphatase Sit4p (Douville et al. 2004; Mitchell and Sprague 2001). It is possible that Ypa2p or Ypa1p will also activate Ppa1p–PP2A complexes, the SIP complex, or another related phosphatase, such as the SIT4-family phosphoprotein phosphatase Ppe1p (Shimanuki et al. 1993), which is known to associate with RRD proteins in S. cerevisiae (Douville et al. 2004). However, increased expression of ppa2 does not rescue ypa2–Δ or ypa1–Δ, whereas ppa2 overexpression will rescue the cold-sensitive phenotype of ppe1–Δ (Shimanuki et al. 1993). Our preliminary analysis indicates that a disruptant of ppe1 (Shimanuki et al. 1993) does not rescue cdc7-24 (data not shown), suggesting that reduction of Ppe1p activity does not make a major contribution to the rescue of cdc7-24 by ypa2–Δ.

Neither ypa2–Δ, ypa1–Δ, nor ppa2-6 arrest in the first cell cycle after shift to the restrictive temperature; it is likely that the heterogeneous cell death reflects the diversity of cellular processes that depend upon PP2A activity, including cell morphogenesis, control over the entry into mitosis, and proper organization of the cytoskeleton. In this context it is noteworthy that previous studies have implicated Ypa1p in G0 phase survival (Su et al. 1996), while the ypa2–Δ mutant has been shown to be sensitive to DNA damaging agents in a genome-wide screen (Deshpande et al. 2009).

The phenotype of the pab1–Δ mutant suggests that this form of PP2A is important for cell morphogenesis (Kinoshita et al. 1996; Tanabe et al. 2001; Lahoz et al. 2010). The similarity of this to the ppa2-6 and ypa2–Δ indicates that PP2A plays an important role in cell morphogenesis and also implicates Ypa2p in this process. It is possible that Ypa2p might be important for activating PP2A–Pab1p. However, this cannot be its sole role, since the double ypa2–Δ pab1–Δ mutant shows a synthetic-negative effect.

In S. cerevisiae, a double rrd1rrd2 mutant is inviable (Rempola et al. 2000), similar to the results for the ypa1–Δ ypa2–Δ mutant obtained in this study. The rrd1 mutant also displays abnormal morphology and an aberrant actin cytoskeleton (Van Hoof et al. 2000, 2001). Rrd1p has also been implicated in regulating the G2–M transition, where it cooperates with the protein kinase Cla4p (Mitchell and Sprague 2001) and in TOR signaling, where one of its targets may be RNA polymerase II (Douville et al. 2006; Jouvet et al. 2010). Whether S. pombe Ypa1p and Ypa2p have targets other than phosphoprotein phosphatases will be the subject of future studies. Biochemical analysis of the interaction of S. pombe PTPA proteins with PP2A will be the subject of future studies.

PP2A and SIN signaling

Ablation, or loss of function of Ypa2p rescues many SIN mutants and renders cells sensitive to increased SIN signaling. The strong similarity in the allele specificity of the rescue by ypa2–Δ with that of ppa2–Δ leads us to propose that this rescue results from reduced PP2A activity. Loss of Ypa2p function rescues loss of SIN function to a greater extent than loss of Ypa1p. Given the relative levels of the two proteins, this may simply reflect the degree of reduction of PP2A activity. However, overexpression of ypa1 cannot rescue ypa2–Δ, so it remains possible that Ypa1p and Ypa2p activate different forms of PP2A, which act at different points in the SIN pathway. It is also possible that Ypa1p may regulate other pathway(s) that impact upon SIN signaling.

Many of the proteins that make up the SIN are known to be phosphorylated (Fankhauser and Simanis 1994; Krapp et al. 2001, 2003, 2008; Tomlin et al. 2002; Hou et al. 2004; Daga et al. 2005). The presumptive effector kinase of the SIN, Sid2p–Mob1p, belongs to the NDR protein kinase family, which is known to be inhibited by PP2A in other organisms (Millward et al. 1999). The activating phosphorylation sites are conserved in Sid2p, and are important for its function (Hou et al. 2004). It will be of interest to determine whether the steady-state level of phosphorylation of these sites in Sid2p is affected in PP2A mutants. In S. pombe, the NDR-family kinase Mob2p–Orb6p (Hou et al. 2003) is a component of the morphology network (MOR), which regulates polar growth in S. pombe. It has been suggested that increased SIN activity is antagonistic to the MOR (Ray et al. 2010), and that SIN activity in mitosis turns off polar growth in preparation for cytokinesis; reviewed by Gupta and McCollum (2011). It is therefore possible that the morphology problems observed in the ypa2–Δ mutant may result from increased SIN activity, a direct effect upon the MOR, or a combination of these. Future studies will investigate the interaction of PP2A mutants with the MOR.

Previous studies have shown that deletion of the PP2A B′ regulatory subunit Par1p and loss of SIP function both delay dephosphorylation of Cdc11p at the end of mitosis (Krapp et al. 2003; Singh et al. 2011); our preliminary analysis did not reveal any major differences in the migration of Cdc11p in ppa2–Δ mutants compared to wild type in synchronous cultures (data not shown). It is possible that dephosphorylation of Cdc11p can be accomplished by a combination of the SIP and Ppa1p-containing PP2A complexes. Whether a subset of phosphorylated sites on Cdc11p is affected by ppa2–Δ will await detailed mapping of these sites. This, and identification of the target(s) of PP2A in the SIN will be the focus of our future studies.

The data presented here indicate that PP2A is an inhibitor of SIN signaling, consistent with previous studies. We have demonstrated that reduced PP2A activity permits reassociation of Sid2p with the CAR; the isoform of PP2A responsible for this is unclear, but it is noteworthy that the only regulatory subunits associated with the SPB and CAR are the B′ subunits Par1p and Par2p (Jiang and Hallberg 2000; Le Goff et al. 2001). In this study, we have shown that reduced PP2A activity also delays establishment of the asymmetric state of the SIN; on the basis of the phenotypic penetrance, the major regulatory complex in this regard is most likely to be the SIP (Singh et al. 2011), although PP2A–Par1p is also likely to contribute to this process, as previous studies have noted more cells with Cdc7p signals on both SPBs in par1–Δ (Jiang and Hallberg 2001). Both par1 and pab1 loss-of-function mutants rescue mob1 and cdc11 mutants (Lahoz et al. 2010), and loss of SIP function also rescues a cdc11 mutant (Singh et al. 2011). However, beyond this, the mutants have different spectra of rescue of SIN mutant alleles (Jiang and Hallberg 2001; Le Goff et al. 2001; Lahoz et al. 2010). It is possible that the common rescue of mutants reflects the effects of an overall reduction of PP2A activity, while the differences may indicate that different forms of PP2A act upon particular SIN proteins.

In budding yeast, PP2A has also been implicated in regulating mitotic exit and the spindle orientation checkpoint (Queralt et al. 2006; Wang and Ng 2006; Chan and Amon 2009). Studies in other systems have reported that some forms of PP2A are inactivated to permit entry into mitosis and that these are important subsequently for mitotic exit (Barr et al. 2011). In contrast, the data presented here and in previous studies clearly indicate that PP2A (directly or indirectly) inhibits the SIN, which is required for cytokinesis, the last step in the cell cycle. The reason for this difference in organization is unclear; however, we note that the SIN is inhibited by mitotic CDK (Dischinger et al. 2008), so it is possible that the regulatory elements coordinating mitosis and cytokinesis are organized differently in S. pombe. For example, other phosphoprotein phosphatases, such as Flp1p/Clp1p and PP1 complexes may be more important for mitotic exit than PP2A. Alternatively, a subset of PP2A may remain active during mitosis and block SIN activity until the end of anaphase.