Phosphatidylinositol-4-phosphate 5-Kinase Localized on the Plasma Membrane Is Essential for Yeast Cell Morphogenesis*

Phosphatidylinositol 4,5-biphosphate (PtdIns(4,5)P 2 ), an important element in eukaryotic signal transduction, is synthesized either by phosphatidylinositol-4-phosphate 5-kinase (PtdIns(4)P 5K) from phosphatidylinositol 4-phosphate (PtdIns(4)P) or by phosphatidylinositol-5-phosphate 4-kinase (PtdIns(5)P 4K) from phosphatidylinositol 5-phosphate (PtdIns(5)P). Two Saccharomyces cerevisiae genes, MSS4 and FAB1 , are homologous to mammalian PtdIns(4)P 5Ks and PtdIns(5)P 4Ks. We show here that MSS4 is a functional homolog of mammalian PtdIns(4)P 5K but not of PtdIns(5)P 4K in vivo. We constructed a hemagglutinin epitope-tagged form of Mss4p and found that Mss4p has PtdIns(4)P 5K activity. transformants that grew in SD-Trp, 15 strains grew on FOA plates at 23 °C but not at 37 °C. Finally, one temperature-sensitive mutation ( mss4-1 ) on plasmid (pYO1970) was further analyzed. The mss4-1 mutation and the wild-type MSS4 gene were integrated into the genome by one-step plasmid integration strategy. We utilized plasmids pYO1974 and pYO1975 digested with Sac I and Avr II to transform YOC802 and selected the integrants for the LEU2 marker. After being incubated at 23 °C for 2 days, the cells were streaked on FOA plates and were incubated at 23 °C for 3 days so that the MSS4 - URA3 plasmid was eliminated. Single colonies were picked up and were tested for temperature sensitivity. A temperature-sensitive mss4-1 strain was thus obtained and designated YOC808, whereas the MSS4 gene integrated in the same locus was labeled YOC807. Immunofluorescent staining of was to early phase at 30 °C HA-tagged Mss4p was visualized by indirect immunofluorescence 16B12 anti-HA mouse monoclonal antibody as first antibody an fluo-rescein isothiocyanate-conjugated goat anti-mouse IgG (Wako ,6 -diamidino-2-phenylindole di- hydrochloride, rhodamine-phalloidin white M2R

ent protein kinases and calcineurin, a type II B phosphoprotein phosphatase (4). Diacylglycerol, on the other hand, activates the conventional isoforms of protein kinase C, which in turn play a critical role in the regulation of a number of cellular functions in mammalian cells (5). In the budding yeast Saccharomyces cerevisiae, a protein kinase C-homologous gene (PKC1) was isolated (6), whose product was shown to function in cell wall integrity and cell cycle progression (7,8). In vitro studies of Pkc1p, however, indicated that Pkc1p is strongly activated by phosphatidylserine in the presence of Rho1p, but not by diacylglycerol (9). The stimulation by phosphatidylserine alone is characteristic of the atypical isoform of protein kinase C, which is stimulated by phosphatidylserine alone. Since the biochemical property of Pkc1p is different from that of the conventional isoforms of mammalian protein kinase C, it remains unclear whether and how diacylglycerol acts as an important second messenger in S. cerevisiae.
PtdIns(4,5)P 2 is also known to function as a regulator of actin-binding proteins (10) such as profilin (11), gelsolin (12), and ␣-actinin of vertebrates (13). Recently, profilin was reported to be localized both in the plasma membrane and cytosolic fractions in S. cerevisiae, with the membrane association presumably facilitated by its interaction with phosphatidylinositol metabolites (14). Therefore, it is likely that through its regulation of actin-binding proteins, phosphatidylinositol metabolites affect the cytoskeleton in yeast.
Moreover, PtdIns(4,5)P 2 stimulates GDP to GTP exchange of ADP-ribosylation factor 1 (ARF1) (15). As the GTP-bound form of ARF1 triggers the attachment of the coat proteins (16 -18), PtdIns(4,5)P 2 may play a critical role in coat assembly. Interestingly, PtdIns(4,5)P 2 was found to work as a cofactor for brain membrane phospholipase D (PLD) (19). These findings led to the proposal that PLD and phosphatidylinositol 4-phosphate 5-kinase (PtdIns(4)P 5-kinase) with their respective products, PtdIns(4,5)P 2 and phosphatidic acid, form a positive feedback loop that causes a vesicle fusion with the acceptor membrane (19). Since PtdIns(4,5)P 2 as well as phosphatidic acid activates an ARF GTPase-activating protein (20), they further postulated that the positive feedback loop is halted by the conversion of active ARF-GTP to ARF-GDP. Thus, PtdIns(4,5)P 2 may work as a crucial factor in membrane trafficking.
We report here that Mss4p has PtdIns(4)P 5K activity in vitro and that expression of murine type I␤ PtdIns(4)P 5K functionally replaces MSS4 in vivo. Unlike Fab1p, Mss4p is located primarily on the plasma membrane. Analyses of a temperature-sensitive mss4 mutant revealed that Mss4p is involved in the establishment of cell morphology.

MATERIALS AND METHODS
Yeast Strains and Genetic Manipulations-The yeast strains used are listed in Table I. The complete and minimal yeast media as well as the sporulation medium and procedures of tetrad analysis were as described (30). YPGS medium contains 2% galactose, 0.1% sucrose, 1% Bacto-yeast extract, and 2% polypepton, whereas YPA medium for pre-sporulation consists of 1% Bacto-yeast extract, 2% polypepton, and 1% potassium acetate (Wako Pure Chemical Industries, Osaka, Japan). Yeast transformation was carried out with lithium acetate (31). Plates containing 0.2% 5-fluoroorotic acid (FOA, Sigma) were used to select yeast cells capable of losing a URA3 marked plasmid. E. coli strains, DH5␣ (Life Technologies, Inc.) and SCS1 (Stratagene), were used for gene manipulation. DNA sequencing was carried out with an automated DNA sequencer (model 373A, Applied Biosystems, Foster City, CA).
Construction of Plasmids-The plasmids used in this study are described in Table II. Plasmid pYO1953 was cloned from the YEp13 genomic library (32). The insertion of the 3.9-kb BamHI-XhoI fragment of MSS4 into the vector pBluescript SK ϩ resulted in pYO1956, which was used for the construction of other MSS4-containing plasmids and as a template for error-prone polymerase chain reaction. pYO1958, which was designed to aid MSS4 gene disruption, mss4::HIS3, was constructed by ligation of the 5.0-kb EcoRI-EcoRI fragment of pYO1956 and the 1.3-kb BamHI-XhoI fragment of pJJ215 containing the HIS3 gene. pYO1959, pYO1960, and pYO1962 were made by inserting the 3.9-kb BamHI-XhoI fragment of pYO1956 containing MSS4 into the BamHI-XhoI gap of pRS315, pRS314, and pRS316, respectively. pYO1964 was formed by replacing the 1.2-kb NdeI-KpnI fragment of pYO1960 with the NdeI-KpnI linker, which was made by annealing the oligomers TATGTGAGATCTGGTAC and CAGATCTCACA. The murine PtdIns(4)P 5K type I␤ and human PtdIns(5)P 4K genes were obtained by polymerase chain reaction using the published sequences (25,23) with the BamHI and BclI restriction sites, respectively, attached at both ends.
The PtdIns(4)P 5K and PtdIns(5)P 4K genes were then inserted to the BglII site of a single-copy plasmid with the GAL1 promoter (pYO761) to yield pYO2116 and pYO2117, respectively, and to a multicopy counterpart (pYO767) to produce pYO2118 and pYO2119, respectively. We also inserted the murine and human PIPK genes to the BglII site of a single copy plasmid with the GAP promoter (pYO2141) to get pYO2142 and pYO2143, respectively. Similar insertions of the PIPK genes to a multicopy plasmid with the GAP promoter (pYO2144) resulted in pYO2145 and pYO2146. We selected and used only those plasmids with the genes whose sequences agreed with the published ones and inserted in the right direction.
A 3HA tag was introduced to the N terminus of Mss4p as follows. An AvrII adaptor was made by annealing the two oligonucleotides, CCG-GATCCTAGG and CCGGCCTAGGAT, and was inserted to the AccIII site of pYO1959. After checking the direction of insertion by DNA sequencing, we digested the plasmid with AvrII and ligated it with the NheI-digested 3HA tag of pYO1365. The SphI-XhoI fragment of the resultant plasmid carrying 3HA-tagged MSS4 was inserted to the SphI-XhoI gap of pRS314 and pQR324 to produce plasmids pYO1965 and pYO1966, respectively.
Immunoprecipitation and PtdIns(4)P 5K Assay-Cell lysates were made in RIPA buffer (50 mM Tris-HCl, pH 8.0, 1% Nonidet P-40, 0.15 M NaCl, 5 mM EDTA, 1 mM phenylmethylsulfonyl fluoride, 1 g/ml leupeptin, 2 g/ml aprotinin, and 1 mM sodium orthovanadate) by vortexing six times for 30 s each with acid-washed glass beads (425-600 m in diameter, Sigma). After preadsorption with protein A cellulofine (Seikagaku-kogyo, Tokyo), the samples (300 g of protein, assayed by Bio-Rad protein assay kit) were subjected to immunoprecipitation with saturating amounts of 16B12 anti-HA monoclonal antibody (Berkeley Antibody, Richmond, CA) and then adsorbed to protein A cellulofine. The adsorbed immunoprecipitates were then washed four times with RIPA buffer and four times further with buffer T (50 mM Tris-HCl, pH 7.5, 1 mM EDTA, 50 M ATP, 0.25 M sucrose, and 0.15 M NaCl). To determine the PtdIns(4)P 5K activity in the immunoprecipitates, we incubated 10 l of sample in 50 mM Tris-HCl, pH 7.5, 1 mM EGTA, 10 mM MgCl 2 , 50 M ATP, 80 M PtdIns(4)P (Sigma), and 5.0 or 0.5 Ci of [␥-32 P]ATP (Amersham Pharmacia Biotech) in the presence or absence of 50 M phosphatidic acid in a total volume of 50 l. After 60 min, the reaction was terminated by the addition of 0.4 ml of chloroform/methanol/12 N HCl (100:200:1 by volume). The lipids were extracted by the method of Bligh and Dyer (33), dried, and, together with PtdIns(4,5)P 2 , which was used as standard, were spotted on Merck Silica gel 60 TLC plates impregnated with 1.2% potassium oxalate, with the exception of the experiment whose result is shown in lanes 4 and 5 of Fig. 1C, in which a similarly treated Whatman 60A plate was utilized. The samples were separated with the solvent system of chloroform/methanol/ acetone/acetic acid/water (42:30:12:12:12 by volume), and [ 32 P]Ins(4,5)P 2 was visualized by autoradiography except for the product on the Whatman plate, which was processed by BAS2000 Fuji bioImaging analyzer.
Isolation of Temperature-sensitive mss4 Mutants-We first made an mss4 strain carrying the mutant gene on a centromer plasmid; the 3.1-kb BamHI-XhoI fragment of pYO1958 carrying the mss4::HIS3 gene was used to transform the diploid strain, YPH501. His ϩ transformants were selected, and the disruption of one of the chromosomal MSS4 gene copies was confirmed by Southern hybridization. The MSS4/mss4::HIS3 diploid strain, named YOC801, was transformed with pYO1962 carrying MSS4 and URA3, and the transformants were subjected to tetrad dissection. His ϩ Ura ϩ asci were selected and designated YOC802 (mss4::HIS3 (pYO1962)).
Random mutations were introduced by error-prone polymerase chain reaction mutagenesis (34) in the PI(4)P 5-kinase-conserved region of MSS4 using the two synthetic oligonucleotides, CCTTCTCAAAAGT-CAAAGCA and TCGTACTACCGTTCCGGTA, corresponding to bases 841-860 and 2055-2025, respectively. The amplified 1.2-kb fragment was purified, digested with NdeI and KpnI, and then inserted to the NdeI-KpnI gap of pYO1964. Approximately 4,000 independent clones were made, and DNA of the plasmid pool was employed for transformation of YOC802 strain. The transformants that grew on SD-Trp medium at 23°C were streaked on FOA (ϪTrp) plates and grown at 23 or 37°C so that the MSS4-URA3 plasmid is eliminated. Out of 496

Involvement of PtdIns(4)P 5-Kinase in Yeast Morphogenesis
transformants that grew in SD-Trp, 15 strains grew on FOA plates at 23°C but not at 37°C. Finally, one temperature-sensitive mutation (mss4-1) on plasmid (pYO1970) was further analyzed. The mss4-1 mutation and the wild-type MSS4 gene were integrated into the genome by one-step plasmid integration strategy. We utilized plasmids pYO1974 and pYO1975 digested with SacI and AvrII to transform YOC802 and selected the integrants for the LEU2 marker. After being incubated at 23°C for 2 days, the cells were streaked on FOA plates and were incubated at 23°C for 3 days so that the MSS4-URA3 plasmid was eliminated. Single colonies were picked up and were tested for temperature sensitivity. A temperature-sensitive mss4-1 strain was thus obtained and designated YOC808, whereas the MSS4 gene integrated in the same locus was labeled YOC807.
Immunofluorescence Microscopy-Immunofluorescent staining of yeast cells was carried out according to Pringle et al. (35). Cells were grown to early exponential phase at 30°C in YPD medium. HA-tagged Mss4p was visualized by indirect immunofluorescence using 16B12 anti-HA mouse monoclonal antibody as the first antibody and an fluorescein isothiocyanate-conjugated goat anti-mouse IgG (Wako Pure Chemical Industries, Osaka, Japan) as the second antibody. DNA, actin, and chitin were stained with 4Ј,6Ј-diamidino-2-phenylindole dihydrochloride, rhodamine-phalloidin (Molecular Probes), and calcofluor white M2R new (Sigma), respectively. Cell morphology and fluorescent staining were observed and photographed using a BX60 microscope (Olympus, Tokyo). Cell fractionation experiments were performed using the previously described techniques (36).

Mss4p
Is a Functional Homolog of Mammalian PtdIns(4)P 5-Kinase-A BLAST search of protein sequence data bases revealed that yeast Mss4p has 36, 33, and 31% identity with murine type I␣, type I␤ PtdIns(4)P 5K, and human PtdIns(5)P 4K, respectively, in agreement with previous reports. To examine whether Mss4p is a functional homolog of any of the mammalian phosphatidylinositol phosphokinases (PIPKs) in yeast, we constructed plasmids carrying the genes encoding murine type I␤ PtdIns(4)P 5K and human PtdIns(5)P 4K hooked up to either the constitutive GAP promoter or the galactose-inducible GAL1 promoter. After these expression plasmids were introduced to YOC802 strain carrying mss4::HIS3 and a URA3-MSS4 plasmid, the growth on FOA plates was examined. We found that all the transformants expressing the type I␤ PtdIns(4)P 5K gene were capable of growing on FOA plates (Fig. 1A, panels b and c). On the other hand, expression of the PtdIns(5)P 4K gene failed to complement the MSS4 gene disruption, irrespective of copy number or the promoters used (Fig. 1A, panels b and c).
We next made a temperature-sensitive MSS4 mutant and tested the suppression of the temperature sensitivity by mammalian PIPKs. Mutations were introduced into the conserved region for PIPK within the MSS4 gene, and one of the mutants that grew at 23°C but not at 37.5°C (mss4-1) was studied (see "Materials and Methods"). The temperature-sensitive mss4-1 strain was transformed with plasmids containing the two mammalian PIPK genes under the control of the two different promoters to test if the temperature sensitivity is suppressed. When murine type I␤ PtdIns(4)P 5K was expressed under the GAL1 promoter on either the single copy or multicopy plasmid, the strain grew on a galactose-containing plate at the restrictive temperature (Fig. 1A, panels e and f). On a glucose-containing plate on which the expression of type I␤ PtdIns(4)P 5K under the GAL1 promoter is reduced, however, suppression of mss4-1 was observed only with the multicopy plasmid, indicating the failure of suppression when expression is greatly reduced. When the same gene was placed on plasmids under the GAP promoter, the plasmid-harboring strains grew at the re-  (52) strictive temperature irrespective of the copy number (Fig. 1A, panels e and f). Murine type I␤ PtdIns(4)P 5K therefore suppresses the temperature sensitivity of mss4-1. In contrast, human PtdIns(5)P 4K failed to suppress the temperature sensitivity in any of the combinations of the plasmids and the promoters tested. Therefore, the functional complementation analysis with the deletion and the temperature-sensitive mutations of mss4 suggest that MSS4 encodes PtdIns(4)P 5K.
Expression of an Epitope-tagged MSS4 in Yeast Cells-To analyze the Mss4p functions, we inserted the 3HA-epitope tag at the N-terminal of Mss4p. Introduction of the 3HA-epitope tag preserves its essential function because the tagged MSS4 with a single copy plasmid can fully complement mss4::HIS3 at all the temperatures examined (23, 30, and 37°C). These results indicate that the 3HA-tagged Mss4p is functional in vivo. Western blotting analysis of the cells expressing the tagged Mss4p has shown that the anti-HA monoclonal antibody recognized a single band with a molecular mass of 86 kDa, which matched the predicted molecular weight of Mss4p (data not shown).
The Tagged MSS4 Gene Product Has PtdIns(4)P 5-Kinase Activity-To examine PtdIns(4)P 5K activity of the MSS4 gene product, we immunoprecipitated the 3HA-tagged MSS4 protein expressed in yeast with the anti-HA monoclonal antibody and determined the kinase activity in the immunoprecipitate (Fig. 1B). The immunoprecipitate from the YOC804 cells carrying the tagged MSS4 gene on a multicopy plasmid had the highest PtdIns(4)P 5K activity, followed by that from the YOC803 cells, which harbored the same gene on a single copy plasmid, whereas that from the cells with untagged Mss4p (YOC806) exhibited little activity. Furthermore, the PtdIns(4)P 5K activity in the immunoprecipitates was found to be stimulated by the addition of 50 M phosphatidic acid (Fig. 1B), a characteristic property of PtdIns(4)P 5K but not of PtdIns(5)P 4K (37). These results demonstrate that the tagged Mss4p possesses PtdIns(4)P 5K activity and that the amount of the kinase activity is copy number-dependent.
The mss4-1 Protein Has Less PtdIns(4)P 5K Activity When Cultured at the Restrictive Temperature-We examined whether the PtdIns(4)P 5K activity of the mss4-1 mutant changes at the restrictive temperature. We first made a strain with the MSS4 gene disrupted but harboring a 3HA-tagged mss4-1 gene on a multicopy plasmid and designated it YOC823. The strain was cultured at 23°C, was transferred to 38°C at early exponential growth phase, and was further cultivated for 0, 2, 4, 6, or 8 h before being harvested. The lysates were made, immunoprecipitated with the anti-HA antibody, and the PtdIns(4)P 5K activities in the immunoprecipitates were assayed. As can be seen in Fig. 1C, the kinase activity starts decreasing immediately upon the temperature shift, and the reduction is complete by 4 h at the restrictive temperature. Western blotting of cell lysates showed that the strain had less amount of the mutant protein when cultured at the restrictive temperature than at the permissive temperature (data not shown). These results indicate a temperature-sensitive defect in the synthesis and/or heat lability of the mutant protein.
The MSS4 Gene Product Is Localized on the Plasma Membrane-To examine intracellular localization of the tagged Mss4p, we first investigated the partitioning of 3HA-tagged [␥-32 P]ATP. Lanes 4 and 5, the PtdIns(4)P 5K activity in YOC804 cells was similarly assayed in the presence or absence of 50 M phosphatidic acid (PA) with 0.5 Ci of [␥ -32 P]ATP per sample. C, Lanes 1-5, the YOC823 (⌬mss4::HIS3 (YEp3HA:mss4-1)) strain was cultured at 23°C and was then placed at 38°C for indicated times, the lysates prepared and immunoprecipitated, and the PtdIns(4)P 5K assay was carried out at 23°C with 5.0 Ci of [␥-32 ]ATP in each sample.
Phenotypes of the Temperature-sensitive mss4-1 Mutant-Growth of the temperature-sensitive mss4-1 mutant (YOC808) was compared with those of the wild-type strain (YOC807) and the mss4::HIS3 strain expressing untagged Mss4p on a single copy plasmid (YOC806) cultured on YPD plates (Fig. 3) and in YPD liquid media (Fig. 4). Judging from the colony size, the mss4-1 strain grew as well as the wild-type strain at 23°C, grew slowly at 37°C, and completely failed to grow at 37.5°C (Fig. 3). The growth defect of the mss4 mutant was not suppressed by addition of 100 mM CaCl 2 . The doubling time of mss4-1 was 4.3 h at 23°C, whereas that of the wild-type cells was 4.0 h, indicating that the mss4-1 mutant grows almost as fast as the wild type at the permissive temperature. At 38°C, however, the growth of the mutant cells stopped within 4 h after the temperature shift (Fig. 4).
When observed under the light microscope, the mss4-1 mutant cells were found to stop growing at 38°C with an enlarged size and round shape (Fig. 5). There was no indication of cell lysis under the microscope after a 4-h incubation at 38°C. Most of the mutant cells contain single nuclei or divided two nuclei (Fig. 6, panel I). Even at the permissive temperature (23°C), the mutant cells were round-shaped, whereas the wild-type cells were all oval-shaped (data not shown).
As similar morphological phenotypes were observed with mutations of actin-binding proteins, such as profilin and capping protein (38,39), we then examined the actin distribution of the mss4-1 cells by staining with rhodamine-conjugated phalloidin (Fig. 6). In the wild-type cells, the actin staining showed cables running longitudinally and cortical patches, whereas in budding cells, it occurred exclusively on the bud as previously reported (40 -42) (Fig. 6, panel F). A 1-h incubation at 38°C of the wild-type cells made actin cables fainter and altered the localization of actin patches to the bud (Fig. 6, panel  G), but further incubation restored polarized distribution of actin patches to the bud and, until 8 h at the restrictive temperature, kept the localization unchanged (Fig. 6, panel H). In contrast, phalloidin staining of mss4-1 mutant cells revealed a random distribution of cortical actin patches. In the mutant cells, actin patches became partially polarized after a 4-h incubation (Fig. 6, panel C), but the polarization was completely lost after 6 and 8 h of incubation at 38°C (Fig. 6, panels D and  E). Even at 23°C, actin cables were faint, and actin patches were frequently distributed not only at buds but also in mother cells in the temperature-sensitive strain (Fig. 6, panel A). At 38°C, the staining of actin was obviously brighter in the mutant cells than in the wild-type cells after the 8-h incubation at 38°C (Fig. 6, panel E), whereas staining of cell wall chitin revealed deposition of chitin all over the cell surface in the mutant cells (Fig. 6, panel K). These results indicate that cell polarity and actin distribution are significantly impaired in mss4-1 even at 23°C, and the loss becomes greater at 38°C.
We further examined a genetic interaction between mss4-1 and cls5-1, the latter of which possesses a missense mutation in the profilin gene. 2 We found that a haploid strain with mss4-1 and cls5-1 mutations harboring a URA3-MSS4 plasmid could not grow on FOA plate at 30°C (data not shown). This indicates that the mss4-1 cls5-1 double mutant fails to grow, although either an mss4-1 or a cls5-1 single mutant grows well at both 23 and 30°C. The synthetic lethal interaction between MSS4 and the profilin gene is consistent with our observation that actin cytoskeleton is impaired in the mss4-1 mutant.

DISCUSSION
The first piece of evidence that MSS4 encodes PtdIns(4)P 5K comes from the kinase assay; immunoprecipitates from cells expressing the epitope-tagged form of Mss4p had PtdIns(4)P 5K activity, and the tagged Mss4p is functional as demonstrated by the full complementation of the mss4 deletion by single-copy expression of 3HA-Mss4p. The substrate used, PtdIns(4)P, was purified from bovine brain, is approximately 98% pure on TLC according to the manufacturer, and may contain a small amount of PtdIns(5)P. It is formally possible that MSS4 actually encodes PtdIns(5)P 4K, and the PtdIns(5)P 4K activity produces PtdIns(4,5)P 2 from the trace amount of PtdIns(5)P contained in the substrate. However, the following lines of evidence make this alternative explanation unlikely. The kinase activity of the gene product is enhanced in the presence of phosphatidic acid, a characteristic of mammalian PtdIns(4)P 5K isoforms, but not of PtdIns(5)P 4K (37). In addition, the murine PtdIns(4)P 5K gene, but not the human PtdIns(5)P 4K gene, complemented the mss4 gene disruption and, when the expression level was reasonably high, suppressed the temperature sensitivity of the mss4-1 strain. This identification is consistent with the higher homology Mss4p has with mammalian PtdIns(4)P 5K isoforms than with Pt-dIns(5) 4K. We therefore conclude that MSS4 encodes PtdIns(4)P 5K.
Our finding that MSS4 encodes PtdIns(4)P 5K explains the previous genetic studies on MSS4 well. Since overexpression of MSS4 suppresses the cell lysis phenotype of ⌬stt4 at 23°C and the temperature-sensitive stt4-1 mutation, MSS4 was suggested to function downstream of Stt4p, a PtdIns 4-kinase (29,53). The present demonstration is in agreement with the idea that Stt4p phosphorylates PtdIns to produce PtdIns(4)P, which in turn is further phosphorylated by the action of Mss4p to yield PtdIns(4,5)P 2 . In the same paper, it was reported that overproduction of MSS4 did not affect PtdIns 4-kinase activity of wild-type yeast cells and that PtdIns 4-kinase activity of the ⌬stt4 cells carrying a multicopy MSS4 plasmid was as low as that in the ⌬stt4 cells. These results are consistent with the notion that Mss4p is a PtdIns(4)P 5K and has little if any PtdIns 4-kinase activity.
Another S. cerevisiae gene, FAB1, whose product has a significant homology to mammalian PIPKs, is not essential (27). Fab1p was suggested to have PtdIns(4)P 5K activity on the vacuolar membrane, and the product of the kinase reaction, PtdIns(4,5)P 2 , was proposed to function as a regulator of vacuole homeostasis (27). However, the possibility that FAB1 encodes PtdIns(5)P 4K cannot be excluded, especially because no kinase assay has been reported. On the other hand, MSS4 is an essential gene (29) whose product is mainly localized to the plasma membrane (Fig. 2). Thus, it seems that Mss4p functions as PtdIns(4)P 5K on the plasma membrane and the product of the reaction, PtdIns(4,5)P 2 , plays an essential function at or near the plasma membrane. The idea that the two PIPKs that produce PtdIns(4,5)P 2 play different roles at different compart-FIG. 4. mss4-1 strain stops growing at 38°C. At zero time, YOC807 (MSS4) and YOC808 (mss4-1) cultures grown in YPD at 23°C to early exponential phase were diluted with the same medium, and the diluted cultures were grown at either 23 or 38°C. The cell densities at time zero and subsequent times were determined with a hemocytometer.
FIG. 5. mss4-1 strain shows aberrant cell morphology. YOC808 (mss4-1) cells were cultured at 23°C, diluted, and further incubated for 8 h at 23 or 38°C. The cells were observed under the phase-contrast microscope. ments is supported by our recent observation that MSS4 on a multicopy plasmid suppresses the temperature sensitivity of cmd1-228, a calmodulin mutant with defect in calmodulin localization (43), whereas FAB1 does not (data not shown).
One explanation of the phenotypes of the temperature-sensitive mss4-1 mutant cells is that the phenotypes are caused by a reduced level of PtdIns(4,5)P 2 , whereas another interpretation is that they are brought about by defects in hitherto unidentified function(s) of Mss4p. We favor the former possibility because it is consistent with our finding that, when the temperature-sensitive mutant cells are shifted to the restrictive temperature, PtdIns(4)P 5K activity decreases before the mutant phenotypes become evident (Figs. 1C and 6).
Through what pathways does PtdIns(4,5)P 2 give rise to the mutant phenotypes? The mss4-1 mutant phenotypes are strikingly similar to those of mutants of two actin-binding proteins, i.e. profilin null mutants (38) and the capping protein deletion mutants, ⌬cap1 and ⌬cap2 (39). Profilin is a ubiquitous actinand PtdIns(4,5)P 2 -binding protein in eukaryotic cells (44) and is required for the proper organization of actin cytoskeleton into actin cables, which occur at regions of active growth and for proper maintenance of cell polarity (38). It was also shown that depletion of PtdIns(4,5)P 2 in the plasma membrane leads to profilin translocation to the cytosol (14). The binding of capping protein to the growing end of actin filaments was demonstrated to be prevented by micromolar concentrations of PtdIns(4,5)P 2 (39). We show here a synthetic lethal interaction between PtdIns(4)P 5K and profilin. Thus, it is plausible that a lower level of PtdIns(4,5)P 2 in mss4-1 cells hinders proper functioning of profilin and capping protein, leading to disorganization of actin cables.
Ins(4,5)P 2 is known to be hydrolyzed by phospholipase C, encoded by the PLC1 gene in S. cerevisiae, to produce IP 3 and diacylglycerol. Temperature-sensitive plc1 mutant cells were reported to be swollen with large buds and two nuclei at the restrictive temperature, and the growth defect of the mutant was suppressed by addition of 100 mM CaCl 2 (45). An increased incidence of aberrant chromosomal segregation was also observed with another temperature-sensitive mutant, plc1-1 (46). Since mss4-1 cells at the restrictive temperature do not show these phenotypes, we consider it improbable that the primary effect of Mss4p is through PtdIns(4,5)P 2 hydrolysis. Alternatively, PtdIns(4,5)P 2 may be required to stimulate GDP to GTP exchange of yeast ARF, which is important for secretion (47). The localization and the mutant phenotypes, however, do not support the idea that the major pathway related to Mss4p involves ARF. PtdIns(4,5)P 2 is also known to work as a cofactor for PLD. Yeast PLD encoded by SPO14 (48) is essential for meiosis but not for vegetative growth. The necessity of this gene only for meiosis again makes PLD an unlikely candidate for the major target of PtdIns(4,5)P 2 .
In summary, we propose that the MSS4 gene product functions in regulation of actin-binding proteins through generation of PtdIns(4,5)P 2 from PtdIns(4)P in or near the plasma membrane. We believe that novel factors involved in the PtdIns(4,5)P 2 signaling pathway can be identified by genetic approach with the conditional mutant of mss4. Further investigations of the MSS4 gene will greatly elucidate phosphatidylinositol signal transduction cascade.