Isolation and Characterization of the Saccharomyces cerevisiae LPP1 Gene Encoding a Mg2+-independent Phosphatidate Phosphatase*

The DPP1-encoded diacylglycerol pyrophosphate (DGPP) phosphatase enzyme accounts for half of the Mg2+-independent phosphatidate (PA) phosphatase activity inSaccharomyces cerevisiae. The LPP1(lipid phosphate phosphatase) gene encodes a protein that contains a novel phosphatase sequence motif found in DGPP phosphatase and in the mouse Mg2+-independent PA phosphatase. A genomic copy of the S. cerevisiae LPP1gene was isolated and was used to construct lpp1Δ andlpp1Δ dpp1Δ mutants. A multicopy plasmid containing the LPP1 gene directed a 12.9-fold overexpression of Mg2+-independent PA phosphatase activity in the S. cerevisiae lpp1Δ dpp1Δ double mutant. The heterologous expression of the S. cerevisiae LPP1 gene in Sf-9 insect cells resulted in a 715-fold overexpression of Mg2+-independent PA phosphatase activity relative to control insect cells. The Mg2+-independent PA phosphatase activity encoded by the LPP1 gene was associated with the membrane fraction of the cell. The LPP1gene product also exhibited lyso-PA phosphatase and DGPP phosphatase activities. The order of substrate preference was PA > lyso-PA > DGPP. Like the dpp1Δ mutant, thelpp1Δ mutant and the lpp1Δdpp1Δ double mutant were viable and did not exhibit obvious growth defects. Biochemical analyses of lpp1Δ,dpp1Δ, and lpp1Δ dpp1Δ mutants showed that the LPP1 and DPP1 gene products encoded nearly all of the Mg2+-independent PA phosphatase and lyso-PA phosphatase activities and all of the DGPP phosphatase activity in S. cerevisiae. Moreover, the analyses of the mutants showed that the LPP1 andDPP1 gene products played a role in the regulation of phospholipid metabolism and the cellular levels of phosphatidylinositol and PA.

PA 1 phosphatase in the yeast Saccharomyces cerevisiae catalyzes the dephosphorylation of PA to yield DG and P i (1). The DG derived from PA is used for the synthesis of phosphatidylethanolamine and phosphatidylcholine via the Kennedy (CDP-ethanolamine-and CDP-choline-based) pathway and is also used for the synthesis of triacylglycerols (2)(3)(4). Two types of PA phosphatase have been identified in S. cerevisiae (5). One type of PA phosphatase is Mg 2ϩ -dependent and N-ethylmaleimide-sensitive, and the other enzyme type is Mg 2ϩ -independent and N-ethylmaleimide-insensitive (5). Two membrane-associated forms (104-and 45-kDa) of the Mg 2ϩ -dependent PA phosphatase have been purified and characterized from S. cerevisiae (5)(6)(7). These enzymes are regulated by growth phase (7,8), inositol supplementation (7,8), phosphorylation via protein kinase A (9), phospholipids (10), sphingoid bases (11), and nucleotides (12). This regulation correlates with changes in the synthesis of phospholipids and triacylglycerols (8,13,14).
Much less is known about the Mg 2ϩ -independent type of PA phosphatase and its role in phospholipid metabolism in S. cerevisiae. It was first identified as an activity of a 34-kDa DGPP phosphatase enzyme (15) encoded by the DPP1 gene (16). Pure DGPP phosphatase catalyzes the removal of the ␤-phosphate of DGPP to yield PA and then catalyzes the removal of the phosphate from PA to yield DG (15). Although the DGPP phosphatase enzyme utilizes PA as a substrate in the absence of DGPP, the specificity constant for PA is 10-fold lower than that of DGPP (15). The deletion of the DPP1 gene in S. cerevisiae results in the loss of all detectable DGPP phosphatase activity and only a 50% loss in Mg 2ϩ -independent PA phosphatase activity (16). These data indicate the existence of an additional gene in S. cerevisiae encoding Mg 2ϩ -independent PA phosphatase activity (16). The isolation of the gene encoding this enzyme is required for defined studies to examine its role in phospholipid metabolism.
In this paper we report the isolation and characterization of the LPP1 (lipid phosphate phosphatase) gene in S. cerevisiae. The expression of the LPP1 gene in S. cerevisiae cells on a multicopy plasmid and in Sf-9 insect cells by baculovirus infection resulted in the overexpression of Mg 2ϩ -independent PA phosphatase activity. The overexpressed gene product also ex-hibited LPA phosphatase and DGPP phosphatase activities. Biochemical analyses of an lpp1⌬ mutant, a dpp1⌬ mutant, and an lpp1⌬ dpp1⌬ double mutant showed that the LPP1 and DPP1 gene products encoded nearly all of the Mg 2ϩ -independent PA phosphatase and LPA phosphatase activities and all of the DGPP phosphatase activity in S. cerevisiae. Moreover, the analyses of the mutants showed that the LPP1 and DPP1 gene products played a role in the regulation of phospholipid metabolism in S. cerevisiae and in particular the cellular contents of phosphatidylinositol and PA.

Materials
Growth medium supplies were purchased from Difco. Protein assay reagents were purchased from Bio-Rad. Restriction endonucleases, modifying enzymes, and recombinant Vent DNA polymerase with 5Јand 3Ј-exonuclease activity were purchased from New England Biolabs. Polymerase chain reaction (PCR) and sequencing primers were prepared commercially by Genosys Biotechnologies, Inc. The PCRScript TM AMP SK(ϩ) cloning kit was from Stratagene, and the Yeastmaker TM yeast transformation system was obtained from CLONTECH. DNA sequencing kits were obtained from Applied Biosystems. The DNA size ladder used for agarose gel electrophoresis was purchased from Life Technologies, Inc.. The baculovirus transfer vector pVL1392 was obtained from Invitrogen. Triton X-100 and bovine serum albumin were purchased from Sigma. Lipids were purchased from Avanti Polar Lipids, Sigma, and Biomol. Radiochemicals were purchased from NEN Life Science Products. Scintillation counting supplies were from National Diagnostics. Silica Gel 60 thin layer chromatography plates were from EM Science. Escherichia coli DG kinase was obtained from Lipidex Inc.

Methods
Strains, Plasmids, and Growth Conditions-The strains and plasmids used in this work are listed in Tables I and II, respectively. Methods for yeast growth, sporulation, and tetrad analysis were performed as described previously (31,32). Yeast cultures were grown in YEPD medium (1% yeast extract, 2% peptone, 2% glucose) or in complete synthetic medium minus inositol (33) containing 2% glucose at 30°C. The appropriate amino acid of complete synthetic medium was omitted for selection purposes. E. coli strain DH5␣ was grown in LB medium (1% tryptone, 0.5% yeast extract, 1% NaCl (pH 7.4)) at 37°C. Ampicillin (100 g/ml) was added to cultures of DH5␣ carrying plasmids. Media were supplemented with either 2% (yeast) or 1.5% (E. coli) agar for growth on plates. Yeast cell numbers in liquid media were determined by microscopic examination with a hemacytometer or spectrophotometrically at an absorbance of 600 nm. The inositol excretion phenotype (34) of yeast strains was examined on complete synthetic medium (minus inositol) by using growth of the inositol auxotrophic indicator strain MC13 (ino1) (33) as described by McGee et al. (35).
DNA Manipulations, Amplification of DNA by PCR, and DNA Sequencing-Plasmid and genomic DNA preparation, restriction enzyme digestion, and DNA ligations were performed by standard methods (32). Transformation of yeast (36,37) and E. coli (32) were performed as described previously. Conditions for the amplification of DNA by PCR were optimized as described previously (38). The annealing temperature for the PCRs was 55°C, and extension times were typically between 2.0 and 2.5 min at 72°C. PCRs were routinely run for a total of 30 cycles. DNA sequencing reactions were performed with the Prism DyeDeoxy Terminator Cycle sequencing kit and analyzed with an au-tomated DNA sequencer. Plasmid maintenance and amplifications were performed in E. coli strain DH5␣.
Isolation of the LPP1 Gene-A computer search of the Saccharomyces Genome Data base indicated that the LPP1 2 gene (locus, YDR503C) (GenBank TM accession no. U33057) flanked the 3Ј-end of the PSP1 gene. We obtained plasmid pSK5ϩ (from Dr. Timothy Formosa) that contained genomic copies of the PSP1 and LPP1 genes. A 1.8-kb insert, containing the LPP1 gene, 755 bp of the promoter region, and 200 bp of the untranslated region, was released from pSK5ϩ by digestion with HpaI. This DNA fragment was ligated into the SmaI site of pRS426, a multicopy E. coli/yeast shuttle vector containing the URA3 gene (39) to form plasmid pWB1-LPP1. This construct was transformed into W303-1B and the indicated mutants for the overexpression of the LPP1 gene product.
Construction of an lpp1⌬ Mutant-The plasmid pSK5ϩ was digested with HpaI/SalI to remove the LPP1 gene. This DNA fragment was ligated into the SmaI/SalI sites of pCRScript TM AMP SK(ϩ) to form pCR-LPP1. The pCR-LPP1 construct was digested with StyI/HindIII to remove the entire LPP1 coding sequence. The sites on the digested plasmid were converted to blunt ends using the Klenow fragment of DNA polymerase I. A 2.5-kb HIS3/Kan r disruption cassette, derived from plasmid pJA50 (40) by SmaI digestion, was inserted into the blunt-ended plasmid to form plasmid pCR-lpp1⌬. A linear 2.8-kb LPP1 deletion cassette was released from pCR-lpp1⌬ by digestion with XhoI/ NotI. This DNA fragment was transformed into W303-1B to delete the chromosomal copy of the LPP1 gene by the one-step gene replacement technique (41). Transformants were selected for their ability to grow on complete synthetic medium without histidine. The deletion of the chromosomal copy of the LPP1 gene was confirmed by PCR (primers, 5Ј-GAATGTCAATGAGTTTCGCAGAAGACG-3Ј and 5Ј-GTATTTTGGCT-TCGGTTAATATCTGG-3Ј) using 30 cycles at 55°C annealing temperature with a 3.5-min extension time at 72°C. The template for the PCR used to confirm the LPP1 deletion was genomic DNA isolated from transformed colonies. The PCR template for the native LPP1 gene was genomic DNA isolated from W303-1B. One of the lpp1⌬ mutants that we isolated was designated strain WBY1.
Construction of the lpp1⌬ dpp1⌬ Double Mutant-The lpp1⌬ mutant was crossed with a dpp1⌬ mutant to form a diploid that was heterozygous for the LPP1 and DPP1 alleles. Putative lpp1⌬ dpp1⌬ double mutants were selected for their ability to grow on complete synthetic media lacking both histidine and tryptophan. The deletion of the chromosomal copies of the LPP1 gene and the DPP1 gene (16) was confirmed by PCR as described above. Strain TBY1 was one of the haploid lpp1⌬ dpp1⌬ double mutants that were isolated.
Recombinant Viral Expression of the S. cerevisiae LPP1 Gene in Insect Cells-Plasmid pWB1-LPP1 was digested with BamHI/SalI to release the entire coding sequence of the LPP1 gene. This DNA fragment was ligated into the BamHI/SalI sites of the YEp352 vector resulting in the formation of plasmid pDT1-LPP1. This construct was then digested with Bsu36I/SalI to reduce the remaining promoter sequence to 26 bp upstream of the protein coding sequence. These sites on the plasmid were converted to blunt ends using the Klenow fragment of DNA polymerase I. The blunt ends were ligated together to form plasmid pDT2-LPP1. The pDT2-LPP1 plasmid was digested with BamHI/PstI to release the LPP1 open reading frame. This DNA fragment was then ligated into the BamHI/PstI site of the baculovirus vector PVL1392 to form plasmid pWW1-LPP1. The pWW1-LPP1 plasmid was subsequently co-transfected with BaculoGold TM Autographa 2 The LPP1 gene was previously referred to as the DPP2 gene (27). californica DNA (PharMingen) into a monolayer of Sf-9 cells using the CaCl 2 method. The Sf-9 cells were routinely grown in TMNFH medium (42) containing 10% heat-inactivated fetal bovine serum. General procedures for the growth, maintenance, and infection of Sf-9 cells followed the methods described by O'Reilly et al. (42). Routine infection of Sf-9 cells for LPP1 expression used 1-2 ϫ 10 7 cells grown in 75-cm 2 tissue culture flasks. The cells were infected at a viral multiplicity of 10 and grown in TMNFH medium with 10% heat-inactivated fetal bovine serum for 48 h. The infected cells were collected by gentle trituration with medium, harvested by centrifugation, and washed twice with phosphate-buffered saline. The final cell pellet was snap-frozen over dry ice and stored at Ϫ80°C.

Preparation of Cell Extracts and Subcellular
Fractions-All steps were performed at 5°C. Yeast cells were disrupted with glass beads with a Mini-Bead-Beater (Biospec Products) in 50 mM Tris-HCl buffer (pH 7.5) containing 1 mM Na 2 EDTA, 0.3 M sucrose, and 10 mM 2-mercaptoethanol (43). Glass beads and cell debris were removed by centrifugation at 1,500 ϫ g for 5 min. The supernatant was used as the cell extract. Insect cells were disrupted by sonic oscillation in 50 mM Tris-HCl buffer (pH 7.5) containing 0.3 M sucrose, 1 mM Na 2 EDTA, 10 mM 2-mercaptoethanol, 0.5 mM phenylmethanesulfonyl fluoride, 1 mM benzamide, 5 g/ml aprotinin, 5 g/ml leupeptin, and 5 g/ml pepstatin (16). The disrupted cell suspension was centrifuged at 1,500 ϫ g for 5 min to remove unbroken cells and cell debris. The supernatant was used as the cell extract. The cell extract was centrifuged at 100,000 ϫ g for 1.5 h to obtain the cytosolic (supernatant) and total membrane fractions.
Enzyme Assays and Protein Determination-Mg 2ϩ -Independent PA phosphatase activity was measured by following the release of watersoluble 32 P i from chloroform-soluble [ 32 P]PA (10,000 cpm/nmol) (45). The reaction mixture contained 50 mM Tris maleate buffer (pH 6.5), 0.1 mM PA, 1 mM Triton X-100, 2 mM Na 2 EDTA, 10 mM 2-mercaptoethanol, and enzyme in a total volume of 0.1 ml. Mg 2ϩ -Dependent PA phosphatase activity was measured with 50 mM Tris maleate buffer (pH 7.0), 10 mM 2-mercaptoethanol, 2 mM MgCl 2 , 1 mM Triton X-100, 0.1 mM [ 32 P]PA, and enzyme protein (45). LPA phosphatase activity was measured by following the release of water-soluble 32 P i from chloroformsoluble [ 32 P]LPA (20,000 cpm/nmol) (45). The reaction mixture contained 50 mM Tris maleate buffer (pH 6.5), 0.1 mM LPA, 1 mM Triton X-100, 2 mM Na 2 EDTA, 10 mM 2-mercaptoethanol, and enzyme protein in a total volume of 0.1 ml. DGPP phosphatase activity was measured by following the release of water-soluble 32 P i from chloroform-soluble [␤-32 P]DGPP (5,000 -10,000 cpm/nmol) as described by Wu et al. (15). The reaction mixture contained 50 mM citrate buffer (pH 5.0), 0.1 mM DGPP, 2 mM Triton X-100, 10 mM 2-mercaptoethanol, and enzyme protein in a total volume of 0.1 ml. All enzyme assays were conducted at 30°C in triplicate. The average S.D. of the assays was Ϯ 5%. The enzyme reactions were linear with time and protein concentration. A unit of enzymatic activity was defined as the amount of enzyme that catalyzed the formation of 1 nmol of product/min. Specific activity was defined as units/mg of protein. Protein concentration was determined by the method of Bradford (46) using bovine serum albumin as the standard.
Labeling and Analysis of Phospholipids-Steady-state labeling of phospholipids with 32 P i was performed as described previously (47)(48)(49). Lipids were extracted from labeled cells by the method of Bligh and Dyer (50) as described by Morlock et al. (8). Phospholipids were analyzed by two-dimensional thin layer chromatography on high performance silica gel thin layer chromatography plates using chloroform/ methanol/glacial acetic acid (65:25:10, v/v) as the solvent for dimension one and chloroform/methanol, 88% formic acid (65:25:10, v/v) as the solvent for dimension two (51). The 32 P-labeled phospholipids were analyzed by autoradiography and by PhosphorImager analysis. The position of the labeled lipids on chromatography plates was compared with standard lipids after exposure to iodine vapor. The amount of each labeled lipid was determined by liquid scintillation counting of the corresponding spots on the chromatograms.

RESULTS
Isolation of the S. cerevisiae LPP1 Gene and the Deduced Primary Structure of Its Encoded Protein-The LPP1 (lipid phosphate phosphatase) gene was identified in the Saccharomyces Genome Data Base on the basis that its deduced protein product showed homology to the DPP1-encoded DGPP phosphatase (16) and to the mouse Mg 2ϩ -independent PA phosphatase (24). The homologous regions of these proteins have been shown to constitute a novel phosphatase sequence motif (52). Based on this information we hypothesized that the gene encoded a Mg 2ϩ -independent PA phosphatase. The LPP1 gene is located on the right arm of chromosome IV (53). The LPP1 gene and its flanking sequences were isolated from plasmid pSK5ϩ, a multicopy plasmid that contains LPP1 on a 7.8-kb insert of genomic DNA. The LPP1 gene was sequenced twice by automated DNA sequence analysis and was shown to match the sequence in the data base.
The LPP1 DNA sequence does not have any sequence motifs that would suggest the existence of introns in the gene. The predicted protein product is 274 amino acids in length, has a minimum subunit molecular mass of 31.6 kDa, and is predicted to be an integral membrane protein (Fig. 1). This protein is predicted to have six transmembrane spanning regions distributed over the entire polypeptide sequence (Fig. 1B). The phosphatase sequence motif is comprised of three domains (52), and these domains (Fig. 1A) are predicted to be localized to the hydrophilic surface of the membrane (Fig. 1B). The PSORT computer program 3 predicts possible endoplasmic reticulum, plasma membrane, and Golgi body localization for the deduced amino acid sequence of the LPP1 gene. The PROSITE Motif program 4 predicts that the LPP1 gene product has six protein kinase C and four casein kinase II phosphorylation target sites. Mg 2ϩ -Independent PA Phosphatase, LPA Phosphatase, and DGPP Phosphatase Activities in S. cerevisiae Cells Overexpressing the LPP1 Gene-The LPP1 gene was used to construct a multicopy plasmid for the overexpression of the LPP1 gene product in S. cerevisiae. Cells bearing the multicopy plasmid were grown to the exponential phase of growth, and cell extracts were prepared and assayed for Mg 2ϩ -independent PA phosphatase. The plasmid containing the LPP1 gene directed a 1.9-fold overexpression of Mg 2ϩ -independent PA phosphatase activity when compared with cells not bearing the plasmid ( Fig.  2A). We also examined the ability of the LPP1 gene product to utilize LPA and DGPP as substrates. These substrates were used for this analysis since they are also substrates for the DGPP phosphatase from S. cerevisiae (15,54) and the Mg 2ϩindependent PA phosphatase from rat liver (26,27). The LPA phosphatase and DGPP phosphatase activities in cells bearing the LPP1 gene on the multicopy plasmid were 1.45-and 1.4fold higher, respectively, when compared with the control cells ( Fig. 2A).
Mg 2ϩ -Independent PA Phosphatase, LPA Phosphatase, and DGPP Phosphatase Activities in Sf-9 Insect Cells Overexpressing the LPP1 Gene-To test further the hypothesis that the LPP1 gene was the structural gene encoding a Mg 2ϩ -independent PA phosphatase, we used heterologous expression of the gene in Sf-9 insect cells. The LPP1 gene was placed within the genome of baculovirus under control of the polyhedrin promoter and expressed by viral infection of Sf-9 cells. Infection of the cells with the baculovirus containing the LPP1 gene resulted in the massive overexpression of PA phosphatase, LPA phosphatase, and DGPP phosphatase activities when compared with uninfected cells (Fig. 2A). These data provided strong evidence that the LPP1 gene encoded a Mg 2ϩ -independent PA phosphatase enzyme. These studies also indicated that the highest activity was obtained when PA was used as the substrate (Fig. 2A).
The computer analysis predicted that the LPP1 gene product 4 Available on-line at the following address: http://www.genome-.ad.jp/sit/motif.html.
FIG. 2. Mg 2؉ -independent PA phosphatase, LPA phosphatase, and DGPP phosphatase activities in S. cerevisiae wild-type cells and in Sf-9 insect cells overexpressing the LPP1 gene and the Mg 2؉ -independent PA phosphatase activity in subcellular fractions of Sf-9 insect cells. A, cell extracts were prepared from the indicated S. cerevisiae and Sf-9 insect cells and assayed for Mg 2ϩindependent PA phosphatase, LPA phosphatase, and DGPP phosphatase activities as described under "Experimental Procedures." The break in the activity axis is between 14 and 25 units/mg. B, the indicated subcellular fractions were prepared from Sf-9 insect cells overexpressing the S. cerevisiae LPP1 gene and assayed for Mg 2ϩ -independent PA phosphatase activity as described under "Experimental Procedures." WT, wild-type.

FIG. 1. Deduced amino acid sequence of the LPP1 gene and TMpred plot showing presumptive membrane-spanning regions.
A, the amino acids of the deduced sequence of the LPP1 gene are numbered at the left. The amino acid residues in boxed areas indicate the three domains comprising the novel phosphatase sequence motif (52). B, the deduced amino acid sequence of the DGPP phosphatase protein was analyzed with the TMpred computer program (http:// ulrec3.unil.ch:80/software) for presumptive transmembrane regions. The TMpred algorithm is based on the statistical analysis of a data base (TMbase) of naturally occurring transmembrane proteins (80). Roman numerals indicate presumptive membrane spanning regions of the polypeptide chain. The arabic numerals indicate the regions that comprise the three domains of the novel phosphatase sequence motif.
is an integral membrane protein. We examined this hypothesis using insect cells overexpressing the LPP1 gene. The cytosolic and total membrane fractions were isolated from the cell extract and used for the assay of Mg 2ϩ -independent PA phosphatase activity. The specific activity of the enzyme in the total membrane fraction was 24-fold higher than the activity in the cytosolic fraction (Fig. 2B). These data supported the conclusion that the LPP1 gene product was indeed a membraneassociated enzyme.
We examined the effects of ceramide 1-phosphate and sphingosine 1-phosphate on the Mg 2ϩ -independent PA phosphatase activity encoded by the LPP1 gene. These compounds are substrates and inhibitors of the rat liver Mg 2ϩ -independent PA phosphatase (26). The concentrations of ceramide 1-phosphate and sphingosine 1-phosphate were varied in the assay up to 16 mol % using uniform Triton X-100/lipid mixed micelles (3). At a final concentration of 16 mol %, ceramide 1-phosphate inhibited Mg 2ϩ -independent PA phosphatase activity by only 26%. This indicated that ceramide 1-phosphate was not a good inhibitor and was likely a poor substrate for the enzyme. Sphingosine 1-phosphate had no effect on activity at concentrations up to 16 mol %.
Deletion of the LPP1 Gene-The LPP1 gene was deleted to further support the hypothesis that its gene product encoded a Mg 2ϩ -independent PA phosphatase enzyme. In addition, the availability of an lpp1⌬ mutant would allow us to examine whether the LPP1 gene was essential for cell growth and to examine possible phenotypes that would shed light on the physiological role of its gene product. A genomic construct containing the LPP1 gene was manipulated to delete all of the coding sequences and small amounts of the 5Ј-and 3Ј-noncoding sequences. The LPP1 deletion construct was introduced into the genome of haploid cells by homologous recombination as described under "Experimental Procedures." Haploid lpp1⌬ mutant cells were viable and exhibited growth properties similar to wild-type control cells when grown vegetatively in complete synthetic medium and in YEPD medium at 30°C. In addition, mating and sporulation were unaffected by the deletion of the LPP1 gene. Microscopic examination of lpp1⌬ mutant cells showed no apparent gross morphological differences when compared with wild-type cells. Overall, these results indicated that the LPP1 gene was not essential for cell growth under typical laboratory growth conditions.
We examined the growth of the lpp1⌬ mutant at different temperatures and on different carbon sources. Both lpp1⌬ mutant cells and wild-type cells grew equally well at temperatures ranging from 12 to 37°C when incubated on YEPD and on complete synthetic medium plates. There was no difference in the growth rate and maximum cell density of the lpp1⌬ mutant when compared with its parent when grown at 30°C on medium with 2% glucose, 2% galactose, or 3% glycerol as the carbon source. Mutant cells were grown on medium with 1 M sorbitol, with 1.5 M NaCl, and with 0.5 M CaCl 2 to examine for sensitivity to osmotic stress. The lpp1⌬ mutant grew comparable to its wild-type parent under these conditions indicating that the LPP1 gene product did not play a role in osmohomeostasis.
The lpp1⌬ mutant was also examined for an inositol excretion phenotype (34). This phenotype is the result of the derepression of the INO1 gene (4) and is a characteristic trait of mutants defective in the structural genes for several phospholipid biosynthetic enzymes (4,55). Growth of the ino1 mutant was used as an indicator of the phenotype, and the opi1 mutant, which excretes inositol (34) due to unregulated derepression of the INO1 gene (4, 55), was used as a positive control. The mutant did not exhibit the inositol excretion phenotype.

Deletion of Both LPP1 and DPP1
Genes-The DPP1 gene encodes a DGPP phosphatase enzyme (16) that utilizes a variety of lipid phosphate substrates including DGPP, PA, and LPA (15,54). We constructed an lpp1⌬ dpp1⌬ double mutant to examine the effects of the deletion of both LPP1 and DPP1 genes on cell viability and to examine the phosphatase activities contributed by each gene product. The double mutant was constructed by crossing the lpp1⌬ mutant with a dpp1⌬ mutant as described under "Experimental Procedures." Diploid cells were induced to sporulate and analyzed for the segregation of the HIS3 and TRP1 genes that were used to disrupt LPP1 and DPP1 (16), respectively. Analysis of 12 tetrads showed that the cross yielded two parental ditype (i.e. all four spores of parental types), 8 tetratype (i.e. one spore of each parental genotype and one spore of each recombinant type), and 2 nonparental ditype (all four spores of the recombinant genotypes) tetrads for the HIS3 and TRP1 genes. This segregation pattern indicated that the LPP1 and DPP1 genes segregated independently. Haploid lpp1⌬ dpp1⌬ double mutants were viable and exhibited growth properties similar to wild-type control cells and to lpp1⌬ and dpp1⌬ mutant cells when grown vegetatively in complete synthetic medium and in YEPD medium at 30°C.
Mg 2ϩ -Independent PA Phosphatase, LPA Phosphatase, and DGPP Phosphatase Activities in the S. cerevisiae lpp1⌬ Mutant, dpp1⌬ Mutant, and lpp1⌬ dpp1⌬ Double Mutant-The lpp1⌬ mutant was grown to exponential phase, and cell extracts were prepared and assayed for phosphatase activities using PA, LPA, and DGPP as substrates. The Mg 2ϩ -independent PA phosphatase, LPA phosphatase, and DGPP phosphatase activities in the lpp1⌬ mutant were reduced by 35,22, and 20%, respectively, when compared with the activities found in the wild-type parent (Fig. 3). Transformation of the lpp1⌬ mutant with the multicopy plasmid containing the LPP1 gene resulted in a small but reproducible overexpression of Mg 2ϩ -independent PA phosphatase (1.56-fold), LPA phosphatase (1.22-fold), and DGPP phosphatase (1.27-fold) activities when compared with the activity exhibited by the mutant (Fig. 3).
To examine the expression of the phosphatase activities encoded by the LPP1 gene product in the absence of the DGPP phosphatase enzyme, we utilized a dpp1⌬ mutant. Mg 2ϩ -Independent PA phosphatase and LPA phosphatase activities were reduced in the dpp1⌬ mutant by 47 and 77%, respectively, when compared with the wild-type parent (Fig. 3). As described previously (16), DGPP phosphatase activity was not detectable in the dpp1⌬ mutant (Fig. 3). These data were consistent with the conclusion that the DPP1 gene product was responsible for all of the detectable DGPP phosphatase activity, the majority of the LPA phosphatase activity, and about half of the Mg 2ϩindependent PA phosphatase activity in the cell. Transformation of the dpp1⌬ mutant with the LPP1 gene on the multicopy plasmid resulted in increases in Mg 2ϩ -independent PA phosphatase and LPA phosphatase activities of 3.4-and 1.4-fold, respectively, when compared with these activities in the dpp1⌬ mutant (Fig. 3). In addition, the expression of the LPP1 gene in the dpp1⌬ mutant resulted in a small but reproducible expression of DGPP phosphatase activity (Fig. 3).
We expressed the LPP1 gene in the lpp1⌬ dpp1⌬ double mutant to explore further the phosphatase activities encoded by the LPP1 gene. The deletion of both LPP1 and DPP1 genes resulted in reductions of Mg 2ϩ -independent PA phosphatase, LPA phosphatase, and DGPP phosphatase activities of 92.5, 90, and 100%, respectively, when compared with wild-type cells (Fig. 3). These data indicated that together the LPP1 and DPP1 gene products accounted for nearly all of the Mg 2ϩ -independent PA phosphatase and LPA phosphatase activities and all of the DGPP phosphatase activity in the cell. The expression of the LPP1 gene in the lpp1⌬ dpp1⌬ double mutant resulted in a 12.9-fold increase in Mg 2ϩ -independent PA phosphatase activity and a 4.5-fold increase in LPA phosphatase activity when compared with these activities in the double mutant (Fig. 3). A small amount of DGPP phosphatase activity was expressed in the lpp1⌬ dpp1⌬ double mutant carrying the LPP1 gene on the multicopy plasmid (Fig. 3).
Mg 2ϩ -Dependent PA Phosphatase Activity in the S. cerevisiae lpp1⌬ Mutant, dpp1⌬ Mutant, and lpp1⌬ dpp1⌬ Double Mutant-Two membrane-associated forms of the Mg 2ϩ -dependent type of PA phosphatase have been purified and characterized from S. cerevisiae (5)(6)(7). However, the genes encoding these enzymes have not been isolated. We examined the expression of Mg 2ϩ -dependent PA phosphatase activity in the lpp1⌬ mutant, dpp1⌬ mutant, and lpp1⌬ dpp1⌬ double mutant. The level of expression of the Mg 2ϩ -dependent PA phosphatase activity paralleled the expression of the Mg 2ϩ -independent PA phosphatase activity in wild-type cells and in the lpp1⌬ and dpp1⌬ mutants (Fig. 4). However, the Mg 2ϩ -dependent PA phosphatase activity was 4.35-fold higher than the Mg 2ϩ -independent activity in the lpp1⌬ dpp1⌬ double mutant (Fig. 4). These results confirmed that separate genes encoded the Mg 2ϩ -dependent and Mg 2ϩ -independent types of PA phosphatase. The availability of the lpp1⌬ dpp1⌬ double mutant should facilitate the isolation of genes encoding the Mg 2ϩ -dependent enzymes.
Phospholipid Composition of the lpp1⌬ Mutant, dpp1⌬ Mutant, and lpp1⌬ dpp1⌬ Double Mutant-The effects of the mutations in the LPP1 and DPP1 genes on phospholipid composition were examined. Wild-type cells and the lpp1⌬ mutant, dpp1⌬ mutant, and lpp1⌬ dpp1⌬ double mutant were labeled to steady-state with 32 P i . Phospholipids were extracted from the cells and then analyzed by two-dimensional thin layer chromatography as described under "Experimental Procedures." The major changes observed in the phospholipid composition of the lpp1⌬ mutant were a 38% decrease in the amount of phosphatidylinositol and a 47% increase in the amount of DGPP when compared with wild-type cells (Fig. 5). The deletion of the DPP1 gene resulted in a 49% decrease in phosphatidylinositol, a 38% increase in PA, and a 67% increase in DGPP when compared with the control cells (Fig. 5). The deletion of both LPP1 and DPP1 genes together resulted in a 61% decrease in phosphatidylinositol, a 45% increase in PA, and a 69% increase in DGPP when compared with wild-type cells (Fig. 5). The lpp1⌬ dpp1⌬ mutant also showed small changes in the relative amounts of phosphatidylcholine (10% decrease), phosphatidylethanolamine (18% increase), and phosphatidylserine (28% decrease).

DISCUSSION
The Mg 2ϩ -independent type of PA phosphatase is postulated to play a role in lipid signaling pathways in mammalian cells (18,26,56). Owing to its tractable molecular genetic system, we are using S. cerevisiae as a model eukaryote to study this enzyme and determine its role in phospholipid metabolism. FIG. 5. Phospholipid composition of the S. cerevisiae wildtype, lpp1⌬ mutant, dpp1⌬ mutant, and lpp1⌬ dpp1⌬ double mutant cells. The indicated S. cerevisiae strains were grown in complete synthetic medium and harvested at the exponential (1 ϫ 10 7 cells/ml) phase of growth. The steady-state phospholipid composition was determined by labeling cells for six generations with 32 P i (5 Ci/ ml). The incorporation of 32 P i into phospholipids during the steady-state labeling was approximately 10,000 -12,000 cpm/10 7 cells. The phospholipid composition of the cells was determined as described under "Experimental Procedures." The percentages shown for phospholipids were normalized to the total 32 P i -labeled chloroform-soluble fraction which included sphingolipids and other unidentified phospholipids. The values reported are the average of 4 chromatographic separations Ϯ S.D. WT, wild-type. Mg 2ϩ -Independent PA phosphatase activity has previously been shown to be an activity of the DPP1-encoded (16) DGPP phosphatase enzyme (15). Biochemical analysis of a dpp1⌬ mutant has revealed that the DPP1 gene is not responsible for all of the Mg 2ϩ -independent PA phosphatase activity in S. cerevisiae (16). We hypothesized that the LPP1 gene encoded a Mg 2ϩ -independent PA phosphatase enzyme. This was based on the fact that the predicted product of the gene contains a novel phosphatase sequence motif (52) found in the yeast DPP1encoded DGPP phosphatase (16) and in the mammalian Mg 2ϩindependent PA phosphatase (24).
The LPP1 gene was isolated and expressed in wild-type cells on a multicopy plasmid. The amount of overexpression of Mg 2ϩindependent PA phosphatase activity was relatively low when compared with other phospholipid metabolic enzymes that have been expressed in S. cerevisiae (16,(57)(58)(59)(60)(61)(62). The low level of activity may have been due to regulation of LPP1 gene expression and/or regulation of its encoded activity. In addition, the presence of the DPP1-encoded DGPP phosphatase enzyme in wild-type cells masked the expression of the LPP1 gene product. The expression of the LPP1 gene in the dpp1⌬ mutant and in the lpp1⌬ dpp1⌬ double mutant resulted in an overexpression of Mg 2ϩ -independent PA phosphatase activity of 3.36-and 12.9-fold, respectively. Moreover, the heterologous expression of the S. cerevisiae LPP1 gene in Sf-9 insect cells resulted in a 715-fold overexpression of Mg 2ϩ -independent PA phosphatase activity relative to control insect cells and a 65fold overexpression of activity relative to wild-type yeast. This activity was associated with the membrane fraction of the cell. Taken together, these data provided a conclusive level of evidence for the identification of the LPP1 gene as the structural gene encoding a membrane-associated Mg 2ϩ -independent PA phosphatase. The expressed LPP1 gene product in S. cerevisiae and in insect cells also exhibited LPA phosphatase and DGPP phosphatase activities. The order of substrate preference based on the assay conditions described here was PA Ͼ LPA Ͼ DGPP. Detailed substrate specificity studies await the purification of the LPP1-encoded Mg 2ϩ -independent PA phosphatase enzyme.
The LPP1-encoded Mg 2ϩ -independent PA phosphatase was similar to the DPP1-encoded DGPP phosphatase (15,16,54) and the mammalian Mg 2ϩ -independent PA phosphatase (24 -27) insofar as these enzymes are Mg 2ϩ -independent, utilize a variety of lipid phosphate molecules as substrates, and contain a novel phosphatase sequence motif (52). However, other than the phosphatase sequence motif, these enzymes show relatively little (ϳ23% identity) overall amino acid sequence homology. These enzymes also differ with respect to their substrate specificity. For example, the LPP1-encoded Mg 2ϩ -independent PA phosphatase had a preference for PA over DGPP by about 4-fold, whereas the DPP1-encoded DGPP phosphatase has a 10-fold higher specificity constant for DGPP when compared with PA (15). The rat liver Mg 2ϩ -independent PA phosphatase utilizes PA and DGPP with about equal specificity (27). These data may suggest that each of these enzymes play different roles in phospholipid metabolism.
The great advantage of using S. cerevisiae to study the Mg 2ϩindependent PA phosphatase is the relative ease with which null allele mutants can be constructed (41). The construction of the lpp1⌬ mutant revealed that the LPP1 gene was not essential for cell growth in S. cerevisiae. The lpp1⌬ mutant lacked an identifiable phenotype. The vegetative growth and cell morphology of the mutant were indistinguishable from its wildtype parent. The construction of the lpp1⌬ mutant, the dpp1⌬ mutant (16), and the lpp1⌬ dpp1⌬ double mutant facilitated biochemical studies to examine the contribution of the LPP1 and DPP1 gene products to the lipid phosphate phosphatase activities in S. cerevisiae. The analysis of the mutants showed that the DPP1 gene product was responsible for most of the Mg 2ϩ -independent PA phosphatase and LPA phosphatase activities, and all of the detectable DGPP phosphatase activity in the cell. Together, the LPP1 and DPP1 gene products accounted for 92% of the Mg 2ϩ -independent PA phosphatase, 90% of the LPA phosphatase activities, and all of the DGPP phosphatase activity in the cell. Thus, additional gene(s) exist in S. cerevisiae to account for the remaining Mg 2ϩ -independent PA phosphatase and LPA phosphatase activities. As described previously (16), the level of DGPP was elevated in the dpp1⌬ mutant, and as expected, DGPP was elevated in the lpp1⌬ dpp1⌬ double mutant. However, the 20% reduction in DGPP phosphatase activity and the increase in DGPP content in the lpp1⌬ mutant were surprising. All of the detectable DGPP phosphatase activity was absent in the dpp1⌬ mutant. It is unclear whether the expression of the phosphatase activities in the lpp1⌬ mutant was a reflection of the loss of the LPP1 gene alone and/or a reflection of the regulation of the DPP1 gene in the lpp1⌬ mutant background. Additional studies will be required to address this question.
The phospholipid composition analysis of the mutants revealed that the LPP1 and DPP1 gene products played a role in the regulation of phospholipid metabolism. All three mutants showed decreases in the amounts of phosphatidylinositol. The most dramatic decrease was observed in the lpp1⌬ dpp1⌬ double mutant. The dpp1⌬ mutant and lpp1⌬ dpp1⌬ double mutant also showed significant increases in PA levels. The mechanism responsible for the changes in phosphatidylinositol and PA in the mutants is unknown. It is known that DGPP, which was elevated in the mutants, potently inhibits Mg 2ϩ -independent PA phosphatase activity (15). This inhibition of activity may attribute to the increase in PA levels that were observed in the dpp1⌬ mutant and lpp1⌬ dpp1⌬ double mutant. In turn, PA is a potent activator of phosphatidylserine synthase activity (63). The activation of phosphatidylserine synthase activity could have drawn upon the cellular pools of CDP-diacylglycerol, which is also a substrate for the enzyme phosphatidylinositol synthase (2)(3)(4). Reduction in CDP-diacylglycerol results in a decrease in phosphatidylinositol synthase activity and phosphatidylinositol content in S. cerevisiae (64,65). In addition, the elevation of phosphatidylethanolamine in the double mutant was consistent with a stimulation of phosphatidylserine synthesis at the expense of phosphatidylinositol. These explanations for the changes in phospholipid composition may be an oversimplification of the regulation occurring in the mutants. The regulation of phospholipid metabolism in S. cerevisiae is complex and involves both genetic and biochemical mechanisms (2)(3)(4)55).
Phosphatidylinositol is an essential membrane phospholipid in S. cerevisiae (4) that serves as the precursor for sphingolipids and the D-3 and D-4 phosphoinositides (4, 66). These inositol-containing lipids and their hydrolysis products are prominent lipid signaling molecules in S. cerevisiae and in mammalian cells (66 -70). PA regulates the activity of several lipid-dependent enzymes in S. cerevisiae and mammalian cells (63,(71)(72)(73)(74) and exhibits mitogenic effects in mammalian cells (28, 74 -77). The fact that the levels of phosphatidylinositol and PA changed in the dpp1⌬ and lpp1⌬ dpp1⌬ mutants may suggest that the LPP1 and DPP1 gene products play a role in lipid signaling pathways. Clearly, additional studies are needed to gain insight into the role(s) the LPP1 and DPP1 gene products play in phospholipid metabolism. The availability of the genes and mutants described in this work will facilitate future studies on these important phosphatase enzymes and how they impact on cell physiology.