PAH1-encoded Phosphatidate Phosphatase Plays a Role in the Growth Phase- and Inositol-mediated Regulation of Lipid Synthesis in Saccharomyces cerevisiae*

Background: Yeast Pah1p phosphatidate phosphatase produces diacylglycerol for triacylglycerol synthesis and controls phosphatidate content for phospholipid synthesis. Results: PAH1 expression was induced throughout growth, stimulated by inositol supplementation, and mediated by the Ino2p/Ino4p/Opi1p regulatory circuit and transcription factors Gis1p and Rph1p. Conclusion: Growth phase- and inositol-mediated expression of PAH1 regulates lipid synthesis. Significance: Pah1p phosphatidate phosphatase is regulated by a transcriptional mechanism throughout growth. In the yeast Saccharomyces cerevisiae, the synthesis of phospholipids in the exponential phase of growth occurs at the expense of the storage lipid triacylglycerol. As exponential phase cells progress into the stationary phase, the synthesis of triacylglycerol occurs at the expense of phospholipids. Early work indicates a role of the phosphatidate phosphatase (PAP) in this metabolism; the enzyme produces the diacylglycerol needed for the synthesis of triacylglycerol and simultaneously controls the level of phosphatidate for the synthesis of phospholipids. Four genes (APP1, DPP1, LPP1, and PAH1) encode PAP activity in yeast, and it has been unclear which gene is responsible for the synthesis of triacylglycerol throughout growth. An analysis of lipid synthesis and composition, as well as PAP activity in various PAP mutant strains, showed the essential role of PAH1 in triacylglycerol synthesis throughout growth. Pah1p is a phosphorylated enzyme whose in vivo function is dependent on its dephosphorylation by the Nem1p-Spo7p protein phosphatase complex. nem1Δ mutant cells exhibited defects in triacylglycerol synthesis and lipid metabolism that mirrored those imparted by the pah1Δ mutation, substantiating the importance of Pah1p dephosphorylation throughout growth. An analysis of cells bearing PPAH1-lacZ and PPAH1-DPP1 reporter genes showed that PAH1 expression was induced throughout growth and that the induction in the stationary phase was stimulated by inositol supplementation. A mutant analysis indicated that the Ino2p/Ino4p/Opi1p regulatory circuit and transcription factors Gis1p and Rph1p mediated this regulation.

PAP, 2 the enzyme that catalyzes the dephosphorylation of PA to form DAG and P i , was discovered by Smith et al. in 1957 (1). PA and DAG are intermediates in the synthesis of TAG and membrane phospholipids; they also influence lipid signaling, vesicular trafficking, and transcription . Thus, PAP is a key enzyme for controlling lipid homeostasis and other aspects of cell physiology.
Of the four PAP enzymes in yeast, Pah1p has the greatest effect on lipid metabolism and cell physiology. This assertion is based on the deleterious phenotypes exhibited by pah1⌬ mutant cells that lack this PAP, which include the following: a drastic reduction in TAG abundance and susceptibility to fatty acid-induced lipotoxicity; an accumulation of PA, the misregulation of phospholipid synthesis gene expression, and the aberrant expansion of the nuclear/ER membrane; defects in lipid droplet formation and vacuole homeostasis; and growth sensitivity to elevated temperature and respiratory deficiency (27, 31, 67, 76 -79). The basis for many of these phenotypes is an imbalance of PA and DAG at the nuclear/ER membrane, some of which can be suppressed by deletion of the DGK1-encoded DAG kinase that catalyzes the conversion of DAG back to PA (77,80). Thus, a PA/DAG balance, as mediated by Pah1p PAP, must be achieved to maintain lipid homeostasis and normal cell physiology. In fact, the activity of Pah1p is controlled at different levels that include transcriptional and biochemical mechanisms (24,25,71).
In this study, we explored the contribution of Pah1p to lipid metabolism throughout growth. An analysis of lipid synthesis and composition in various PAP mutant strains confirmed the essential role of Pah1p in TAG synthesis. We also showed that the Nem1p-Spo7p protein phosphatase complex was crucial for TAG synthesis throughout growth. Pah1p PAP activity increased as cells progressed from the exponential to stationary phases, with a concomitant increase in TAG synthesis. A transcriptional mechanism was responsible for the regulation of PAH1 expression throughout growth, and this was mediated by the Ino2p, Ino4p, and Opi1p regulatory circuit and transcription factors Gis1p and Rph1p.

EXPERIMENTAL PROCEDURES
Materials-All chemicals were reagent grade or better. Growth medium supplies were obtained from Difco. Lipids and silica gel 60 TLC plates were from Avanti Polar Lipids and EM Science, respectively. Radiochemicals were from PerkinElmer Life Sciences, and scintillation counting supplies were from National Diagnostics. Ampicillin, aprotinin, benzamidine, leupeptin, N-ethylmaleimide, pepstatin, phenylmethylsulfonyl fluoride, SDS, Triton X-100, o-nitrophenyl ␤-D-galactopyranoside, and ␤-mercaptoethanol were purchased from Sigma. GE Healthcare supplied the polyvinylidene difluoride paper and the enhanced chemifluorescence Western blotting detection kit. Alkaline phosphatase-conjugated goat anti-rabbit IgG antibodies, mouse anti-phosphoglycerate kinase (Pgk1p) antibodies, and alkaline phosphatase-conjugated goat IgG anti-mouse antibodies were from Thermo Scientific, Invitrogen, and Pierce, respectively. Modifying enzymes, recombinant Vent R DNA polymerase, restriction endonucleases, and nucleotides were from New England Biolabs. DNA gel extraction and plas-mid DNA purification kits were obtained from Qiagen. PCR primers were prepared by Genosys Biotechnologies, and the carrier DNA for yeast transformation was from Clontech. DNA size ladders and electrophoresis reagents were from Bio-Rad.
Strains and Growth Conditions-The strains used in this work are listed in Table 1. Plasmid amplification and maintenance was performed in Escherichia coli strain DH5␣. E. coli cells were grown at 37°C in LB medium (1% tryptone, 0.5% yeast extract, 1% NaCl (pH 7)), and ampicillin (100 g/ml) was added to select for cells carrying plasmids. Yeast cells were generally grown at 30°C in YEPD medium (1% yeast extract, 2% peptone, 2% glucose) or in standard synthetic complete (SC) medium supplemented with 2% glucose (81,82). Appropriate amino acids were omitted from the medium for selection of plasmids (81,82). To evaluate the effect of inositol on regulation, cells precultured in inositol-free SC medium (83) were cultured in the medium supplemented with varying concentrations of inositol. Cell numbers in liquid cultures were determined spectrophotometrically at an absorbance of 600 nm. The growth medium was supplemented with agar (2% for yeast, 1.5% for E. coli) for growth on plates.
DNA Manipulations and Plasmid Constructions-Standard protocols were used for the isolation of genomic and plasmid DNA, and the digestion, ligation, and PCR amplification of DNA (82,84). Plasmids used in this study are listed in Table 1. Plasmid pFP1 contains 1 kb of the PAH1 promoter fused to the coding sequence of the lacZ gene of E. coli (71). Using appropriate primers, a series of P PAH1 -lacZ plasmids with promoter deletions were constructed by PCR amplification using plasmid pFP1 as the template. Each PCR product was digested with EcoRI and KpnI and substituted for the 1.0-kb EcoRI/KpnI fragment of pFP1. Plasmid constructions were confirmed by EcoRI/KpnI digestion, and the promoter deletion plasmids were introduced into W303-1A cells for analysis of ␤-galactosidase activity. The plasmid pGH339 directs the expression of the DPP1 gene driven by the PAH1 promoter. It was constructed by replacing the AatII/SpeI fragment (PAH1 coding sequence) of plasmid pGH316 with the AatII/SpeI fragment (DPP1 coding sequence and 3ЈUTR) from pGH201. Plasmid transformations of E. coli (82) and yeast (85) were performed as described previously.
Labeling and Analysis of Lipids-Cells were grown in the presence of [2-14 C]acetate to label lipids (86), which were subsequently extracted from cells by the method of Bligh and Dyer (87). One-dimensional TLC (hexane/diethyl ether/acetic acid, 40:10:1, v/v) was used to resolve individual neutral lipids and total phospholipids (88). The identity of lipids on TLC plates was confirmed by comparison with standards after exposure to iodine vapor. Radiolabeled lipids were visualized by phosphorimaging analysis and their relative quantities analyzed with ImageQuant software. Signals were in the linear range of detectability.
Preparation of Cell Extracts and Subcellular Fractionation-All steps were performed at 4°C. Cell extracts were prepared by disruption of yeast cells with glass beads (0.5-mm diameter) in a Biospec Products Mini BeadBeater-16 (89). The lysis buffer contained 50 mM Tris-HCl (pH 7.5), 1 mM EDTA, 0.3 M sucrose, 10 mM ␤-mercaptoethanol, 0.5 mM phenylmethylsulfonyl fluo-ride, 1 mM NaF, 1 mM benzamidine, 5 g/ml aprotinin, 5 g/ml leupeptin, and 5 g/ml pepstatin. Glass beads and cell debris were removed by centrifugation at 1,500 ϫ g for 5 min, and the supernatant was used as the cell extract. The cytosol (supernatant) and total membrane (pellet) fractions were separated by centrifugation at 100,000 ϫ g for 1 h (89). The membranes were resuspended in the lysis buffer to the same volume as the cytosol fraction. Protein concentration was determined by the Bradford method (90) using bovine serum albumin as the standard.
Enzyme Assays-PAP activity was measured by following the release of water-soluble 32 P i from chloroform-soluble [ 32 P]PA (10,000 -15,000 cpm/nmol) for 20 min at 30°C (89). The reaction mixture contained 50 mM Tris-HCl buffer (pH 7.5), 1 mM MgCl 2 , 0.2 mM PA, 2 mM Triton X-100, and enzyme protein in a total volume of 0.1 ml. The 32 P-labeled PA used to assay PAP activity was enzymatically synthesized from DAG and [␥-32 P]ATP using DAG kinase; the radioactive product was purified by TLC (89). ␤-Galactosidase activity was measured for 5 min at room temperature by following the release of o-nitrophenol from o-nitrophenyl ␤-D-galactopyranoside at 410 nm (91). The reaction mixture contained 100 mM sodium phosphate buffer (pH 7.0), 1 mM MgCl 2 , 100 mM ␤-mercaptoethanol, 3 mM o-nitrophenyl ␤-D-galactopyranoside, and enzyme protein in a total volume of 0.1 ml. All enzyme assays were conducted in triplicate, and the average standard deviation was Ϯ5%. The reactions were linear with time and protein concentration. The units of PAP and ␤-galactosidase activities were defined as the amount of enzymes that catalyzed the formation of 1 nmol of product/min. Specific activity was defined as units/mg of total protein.
Western Blot Analysis-Proteins were resolved by SDS-PAGE (92) and subjected to Western blot analysis (93, 94) with polyvinylidene difluoride membrane. Rabbit anti-Pah1p (64) and mouse anti-Pgk1p antibodies were used at a concentration of 2 g/ml, and rabbit anti-Dpp1p antibodies (95) were used at a dilution of 1:1,000. Alkaline phosphatase-conjugated goat anti-rabbit and anti-mouse IgG antibodies were used at a dilution of 1:5,000. Immune complexes were detected using the enhanced chemifluorescence Western blotting detection kit. Fluorimaging was used to acquire images from immunoblots, and the densities of the images were quantified with ImageQuant software. Signals were in the linear range of detectability.
Analyses of Data-The Student's t test (SigmaPlot software) was used to determine statistical significance, and p values Ͻ 0.05 were taken as a significant difference. (96) have shown that TAG is synthesized from its precursor PA (97) throughout the exponential phase of growth and that its levels increase as cells progress to the stationary phase. Han et al. (27) subsequently showed that the pah1⌬ mutant exhibits a marked decrease (ϳ90%) in TAG content in the stationary phase, pointing to a central role of Pah1p in lipid metabolism. In this work, we examined the contribution of Pah1p to lipid synthesis during growth in more detail. Wild type and pah1⌬ mutant cells were labeled to steady state with [2-14 C]acetate, and neutral lipids and total phospholipids were analyzed by TLC. In agreement with previous work (96), the TAG content of wild type cells increased as cells progressed from the exponential to stationary phase ( Fig. 1). At the stationary phase (36-h time point), the TAG content was 6.6-fold higher than that at the exponential phase (12-h time point). Also in agreement (96), the increase in TAG occurred at the expense of phospholipids, which decreased by 72% in the stationary phase ( Fig. 1). In addition, the other storage lipid ergosterol ester increased (3.5-fold) in the stationary phase, whereas the ergosterol content decreased (2.7-fold) (Fig. 1). The pah1⌬ mutation caused dramatic effects on lipid composition throughout growth ( Fig. 1). By the stationary phase, the difference in TAG content between the mutant and wild type cells was Ͼ95%. Furthermore, ergosterol esters, fatty acids, and phospholipids were elevated in the mutant when compared with the wild type control; the differences by stationary phase (36-h time point) were 115, 206, and 220%, respectively ( Fig. 1). We also examined the effect of the pah1⌬ mutation on the levels of lipid synthesis (Fig. 2). For these experiments, cells were grown to the indicated time intervals and then pulse-labeled with [2-14 C]acetate for 20 min. The percentage of label incorporated into each lipid represents the relative rates of synthesis during the pulse (98). In the wild type control, the synthesis of TAG and DAG increased, whereas the synthesis of phospholipids decreased (Fig. 2). The pah1⌬ mutant showed dramatic reductions in the synthesis of TAG and DAG and increased the synthesis of phospholipids, fatty acids, and ergosterol esters when compared with the wild type control (Fig. 2). In particular, the rate of TAG synthesis increased by 131% in stationary phase wild type cells, although it remained relatively constant in pah1⌬ mutant cells. DAG synthesis was induced by 136% in stationary phase wild type cells compared with exponential phase cells, whereas it decreased by 54% in the pah1⌬ mutant. Wild type cells showed a 54% reduction in phospholipid synthesis in stationary phase, whereas in the pah1⌬ mutant it was decreased by only 15%. Overall, the synthetic rates of lipid synthesis were largely responsible for the lipid composition of wild type and pah1⌬ mutant cells (Fig. 1).

Pah1p PAP Regulates Lipid Synthesis and Composition throughout Growth-Taylor and Parks
Pah1p is a highly phosphorylated enzyme whose in vivo function is dependent on its dephosphorylation by the Nem1p-Spo7p protein phosphatase complex (62)(63)(64)(65)69). Accordingly, we examined the effect of the nem1⌬ mutation (loss of the protein phosphatase catalytic subunit (68)) on lipid synthesis and composition. By and large, the effects of the nem1⌬ mutation mirrored the effects of the pah1⌬ mutation on lipid metabolism (Figs. 1 and 2), substantiating the importance of Nem1p-Spo7p-mediated dephosphorylation of Pah1p for its enzymatic function and role in lipid metabolism.
Pah1p PAP Activity Is Regulated throughout Growth-The work of Hosaka and Yamashita (99) has shown that PAP activity increases as cells progress from the exponential to stationary phases of growth. The increase in PAP activity correlates with the accumulation of TAG that occurs in the stationary phase (96,99). At the time of these early studies, it was unknown that four genes (i.e. APP1, DPP1, LPP1, and PAH1) encode lipid phosphate phosphatase enzymes that utilize PA as a substrate (26 -29). Thus, it is still unclear which PAP enzyme is responsible for the synthesis of TAG throughout growth. Having shown the direct involvement of PAH1 on lipid synthesis and composition, we examined whether its role was directly related to a growth phase-mediated regulation of Pah1p PAP activity. The growth and PAP activity levels of wild type, dpp1⌬ lpp1⌬, app1⌬ dpp1⌬ lpp1⌬, and pah1⌬ dpp1⌬ lpp1⌬ mutant cells were analyzed throughout growth (Fig. 3). The growth curves of dpp1⌬ lpp1⌬ and app1⌬ dpp1⌬ lpp1⌬ mutant cells were indistinguishable from that of wild type cells. However, the pah1⌬ dpp1⌬ lpp1⌬ mutant exhibited significantly slower growth (similar to the pah1⌬ mutant alone (27,67)) when compared with the other strains, demonstrating the importance of Pah1p function in normal cell development and physiology (Fig. 3A).
Wild type and mutant cells were collected throughout growth, and cell extracts were prepared and assayed for PAP  activity (Fig. 3B). There was a marked increase (6.1-fold) in PAP activity of wild type cells as they progressed from the exponential to stationary phases, peaking at 24 h and decreasing by 25% at 36 h. The PAP activity in the dpp1⌬ lpp1⌬ mutant, which lacks the nonspecific, Mg 2ϩ -independent PAP activities, was also shown to increase (3.3-fold) during growth, but to a lesser extent than that found in wild type cells. Thus, the increase in PAP activity of wild type cells can be attributed in part to the DPP1-and LPP1-encoded lipid phosphate phosphatase enzymes. In fact, the expressions of DPP1 and LPP1 are known to be induced in stationary phase (95,100). The PAP activity profile in the app1⌬ dpp1⌬ lpp1⌬ mutant, which also lacks the actin patch-associated PAP activity (26), was similar to the profile exhibited by the dpp1⌬ lpp1⌬ mutant. This indicated that App1p did not have a major effect on the PAP activity being regulated throughout growth. Moreover, the growth phasemediated increase in PAP activity was abrogated in the pah1⌬ dpp1⌬ lpp1⌬ mutant (Fig. 3B). Taken together, these results indicated that Pah1p was primarily responsible for the synthesis of TAG and its accumulation throughout growth.
Most of the cellular Pah1p is phosphorylated and found in the cytosolic fraction of the cell (25,27,(62)(63)(64)(65)69), although a relatively small amount of the enzyme associates with the nuclear/ER membranes for its dephosphorylation by the Nem1p-Spo7p protein phosphatase complex and interaction with PA for catalysis (27,(62)(63)(64)(65)69). We examined the distribution of PAP activity between the cytosol and membrane fractions of dpp1⌬ lpp1⌬ mutant cells throughout growth. The dpp1⌬ lpp1⌬ mutant was used for this experiment to eliminate interference from the membrane-associated PAP activities encoded by DPP1 and LPP1 (28,29). As described previously (27), the specific activity of the Pah1p enzyme in this double mutant was greater in the membrane fraction when compared with the cytosolic fraction (Fig. 3C). As cells progressed from the exponential to stationary phases, the PAP activity in both fractions increased, reflecting both the induction and translocation of the enzyme activity during growth (Fig. 3C).
Expression of P PAH1 -lacZ and P PAH1 -DPP1 Reporter Genes Is Regulated throughout Growth-Given the importance of TAG synthesis throughout growth and the role that Pah1p plays in this process, we examined whether PAH1 expression was regulated by a transcriptional mechanism. These studies were facilitated by use of the P PAH1 -lacZ reporter gene (71). The ␤-galactosidase activity of wild type cells expressing the reporter gene increased as cells progressed throughout growth (Fig. 4A). The ␤-galactosidase activity was 9-fold higher in the stationary phase cells when compared with the exponential phase. In a separate experiment, the cells expressing the reporter gene were grown to stationary phase, washed with fresh growth medium, and then allowed to resume growth. The ␤-galactosidase activity of the cells resuming exponential growth declined in a time-dependent manner (5-fold) (Fig. 4B). Thus, the expression of the P PAH1 -lacZ reporter gene was induced in stationary phase and repressed in exponential phase.
The P PAH1 -DPP1 reporter gene was also used to examine the regulation of PAH1 expression throughout growth. In this experiment, the PAH1 promoter activity was monitored by the expression of Dpp1p. Western blot analysis of dpp1⌬ cells bearing the P PAH1 -DPP1 reporter gene showed that the levels of Dpp1p increased in the stationary phase of growth (Fig. 5). This result coupled to the result of the P PAH1 -lacZ reporter gene expression supported the conclusion that a transcriptional mechanism was responsible for the increase in PAH1-encoded PAP activity that occurred in the stationary phase of growth. As described previously (95), the expression of DPP1 as driven by its own promoter was also induced in the stationary phase.
The extracts from the dpp1⌬ cells containing the plasmiddirected expression of Dpp1p were also probed for Pah1p using anti-Pah1p antibodies. Paradoxically, the abundance of Pah1p decreased as cells progressed from the exponential to stationary phases of growth (Fig. 5). Choi et al. (63) have shown that the loss of Pah1p abundance in cells progressing from the early to late exponential phase is attenuated in nem1⌬ mutant cells lacking the Nem1p-Spo7p protein phosphatase complex. Moreover, phosphorylation-deficient forms of Pah1p exhibit dramatic reductions in abundance (63,64). These observations have led to the hypothesis that Pah1p abundance, as mediated by phosphorylation/dephosphorylation, is a posttranslational mechanism to control PAP function in lipid synthesis (25,63). This phenomenon and the regulation of Pah1p abundance throughout growth will be addressed in a separate communication.   DECEMBER 13, 2013 • VOLUME 288 • NUMBER 50

Effect of Inositol Supplementation on the Growth Phase-mediated Regulation of P PAH1 -lacZ Reporter Gene Expression-
Inositol is a precursor of phosphatidylinositol, which is essential to the synthesis of polyphosphoinositides, sphingolipids, and glycosylphosphatidylinositol anchors (reviewed in Refs. [101][102][103][104][105]. Inositol also plays an important role in the transcriptional regulation of phospholipid synthesis genes (e.g. INO1, CDS1, CHO1, PSD1, CHO2, and OPI3) that contain a UAS INO element in their promoters (20,101,102,106,107)). Because the inositol-mediated regulation of these genes is governed by the cellular concentration of the PAP substrate PA (see below), we examined the effect of inositol on PAH1 expression using the P PAH1 -lacZ reporter gene assay (Fig. 6). Inositol supplementation did not have a major effect on the expression of ␤-galactosidase activity in exponential phase cells. However, in stationary phase cells, the addition of inositol to the growth medium resulted in a dose-dependent increase in reporter gene activity (Fig. 6). Maximum expression was observed at 25 M inositol where the ␤-galactosidase activity was 3-fold greater than the activity of stationary phase cells grown without inositol.
Expression of P PAH1 -lacZ Reporter Genes with Promoter Deletions-To identify regions in the PAH1 promoter involved in the transcriptional regulation by inositol and growth phase, we examined the ␤-galactosidase activity resulting from P PAH1 -lacZ reporter genes with deletions from the 5Ј end of the promoter (Fig. 7). Wild type cells bearing the reporter genes were grown to the exponential and stationary phases of growth in the absence and presence of 75 M inositol. An excess of inositol was used in these and subsequent experiments to ensure that it was not depleted during growth to stationary phase. Deletion of the promoter to Ϫ800 resulted in a 46% reduction in the reporter gene activity of exponential phase cells grown without and with inositol. This deletion also caused the loss of the inositol-mediated induction of the reporter gene activity of stationary phase cells. For exponential phase cells grown without inositol, the promoter deletion to Ϫ300 caused a 42% increase in reporter gene expression, and for stationary phase cells, the deletion caused a 12-fold increase in expression (Fig. 7). These results indicated the existence of an element for an inducer and an inositol-sensitive regulatory element between positions Ϫ1000 and Ϫ800 and the presence of an element for a repressor between Ϫ500 and Ϫ200. Defective in the INO2,  INO4, and OPI1 Regulatory Genes-Two putative UAS INO elements (80% homology to the consensus sequence of 5Ј-CAT-GTGAAAT-3Ј) are present between Ϫ1,000 and Ϫ800 bp of the PAH1 promoter, the region where the inositol-mediated induction of P PAH1 -lacZ reporter gene activity was lost (Fig. 7). The UAS INO element is a binding site for an Ino2p/Ino4p heterodimer that activates transcription of several phospholipid synthesis genes in exponential phase cells grown in the absence of inositol (20,101,102,106,107). When exponential phase cells are supplemented with inositol, the Opi1p repressor, which is tethered to the nuclear/ER membrane through interactions with the membrane protein Scs2p and PA, translocates into the nucleus and binds to Ino2p for repression of UAS INO element-containing genes (20,101,102). Under these growth conditions, the interaction of Opi1p with Scs2p is destabilized by a reduction in PA content that is brought about by inositol supplementation and the synthesis of phosphatidylinositol via CDP-DAG (20,101,102). We were curious to examine the regulation of the P PAH1 -lacZ reporter gene expressed in ino2⌬, ino4⌬, and opi1⌬ mutant cells, especially because the inositol-   mediated regulation of PAH1 was opposite to that of most UAS INO -containing genes. For exponential phase ino2⌬ mutant cells, there was a 1.9-fold induction in ␤-galactosidase activity by inositol supplementation that was not observed in wild type cells. In addition, the growth phase-and inositolmediated inductions of reporter gene expression were abolished in the ino2⌬ mutant in stationary phase (Fig. 8). For exponential phase ino4⌬ mutant cells, the reporter gene activity was higher (1.6-fold) than that of wild type cells, and the expression was induced (1.6-fold) by inositol supplementation. In addition, the growth phase-and inositol-mediated inductions observed for wild type stationary phase cells were lost as a result of the ino4⌬ mutation (Fig. 8). The P PAH1 -lacZ reporter gene expression in exponential phase opi1⌬ mutant cells was 5-fold lower when compared with that of wild type cells. The growth phasemediated induction (7-fold) of ␤-galactosidase activity was intact, but the inositol-mediated induction of reporter gene activity was lost in the opi1⌬ mutant (Fig. 8).

Regulation of P PAH1 -lacZ Reporter Gene Expression by Growth Phase and Inositol in Mutants
Regulation of P PAH1 -lacZ Reporter Gene Expression by Growth Phase and Inositol in Mutants Defective in the GIS1 and RPH1 Regulatory Genes-Inspection of the PAH1 promoter revealed nine putative binding sites for the transcription factor Gis1p and six binding sites for Rph1p (Fig. 7), all of which share 75% similarity with their respective consensus sequences. In particular, there are several Gis1p-and Rph1p-binding sites between Ϫ1,000 and Ϫ800 and between Ϫ500 and Ϫ300, the regions in the PAH1 promoter that showed the major effects on the regulation of P PAH1 -lacZ reporter gene expression (Fig. 7). Gis1p can function as both a transcriptional repressor and activator and is maximally expressed as cells progress into stationary phase (108,109). Rph1p shares 35% similarity with Gis1p, but the zinc finger regions of Gis1p and Rph1p that interact with DNA are 93% identical, indicating that the two transcription factors recognize identical DNA sequences (108,110). Moreover, the consensus binding site for Rph1p (5Ј-AGGGG-3Ј) is nearly identical to the core sequence (5Ј-AGGGA-3Ј) of the Gis1p-binding site (108). To analyze the possible involvement of Gis1p and Rph1p in the regulation of PAH1 expression, the growth phase-and inositol-mediated regulations of P PAH1 -lacZ reporter gene expression were examined in gis1⌬ and rph1⌬ mutant cells (Fig. 9). For exponential phase cells, the expression of ␤-galactosidase activity in gis1⌬ and rph1⌬ mutants was similar to that found in wild type cells. However, the growth phase-mediated induction of expression for cells grown without inositol was attenuated by ϳ45%, indicating that these transcription factors induced PAH1 expression. In addition, the inositol-mediated induction observed for wild type cells in the stationary phase was abolished. Instead, inositol caused the repression (ϳ1.5-fold) of the P PAH1 -lacZ reporter gene activity in the exponential and stationary phases of both the gis1⌬ and rph1⌬ mutants (Fig. 9).

DISCUSSION
Since its initial discovery in 1957 (1), PAP has been recognized as an important regulatory enzyme in lipid metabolism (24,25,(111)(112)(113)(114)(115). Like S. cerevisiae, higher eukaryotic organisms contain multiple PAP enzymes that include Mg 2ϩ -dependent and Mg 2ϩ -independent forms involved in lipid synthesis and signaling, respectively (111)(112)(113)(114)(115)(116). Genes encoding the type of PAP responsible for de novo TAG synthesis were not identified until 2006 when Han et al. (27) established that yeast Pah1p is a PA-specific Mg 2ϩ -dependent PAP. This discovery led to the identifications of orthologous genes encoding Mg 2ϩ -dependent PAPs in humans (27,117), mice (118,119), flies (120,121), worms (122), and plants (123,124). The importance of Mg 2ϩdependent PAP in mammalian physiology is emphasized by the observations that genetic defects in the lipin PAP enzymes are manifested in metabolic disorders that include obesity, lipodystrophy, peripheral neuropathy, myoglobinuria, and inflammation (113). In S. cerevisiae, Pah1p has emerged as one of the most important and highly regulated enzymes in lipid metabolism (25,101,102). In particular, its role in the synthesis of TAG, which occurs during logarithmic growth and accumulates in lipid droplets at the stationary phase (96,(125)(126)(127)(128)(129), has become a subject of intense investigation for understanding lipid-based diseases such as obesity.
Through the analysis of mutants devoid of the four genes encoding PAP activity, we confirmed that PAH1 is responsible for the synthesis of TAG throughout growth in S. cerevisiae. On  a biochemical level, Pah1p is subject to phosphorylation/dephosphorylation, and the in vivo function of Pah1p is dependent on its dephosphorylation by the Nem1p-Spo7p protein phosphatase complex (62-65, 69, 130). This work provided the first indication that the dephosphorylation of Pah1p is crucial for the synthesis of TAG throughout growth; nem1⌬ mutant cells exhibited TAG levels and synthesis rates comparable with those of the pah1⌬ mutant. As far as we know, Pah1p is the major substrate for the Nem1p-Spo7p complex, and so, it was not surprising that the nem1⌬ mutation had such a major effect on the synthesis of TAG and overall lipid metabolism. Additional phenotypes exhibited by pah1⌬ mutant cells, which include the aberrant expansion of the nuclear/ER membrane and sensitivity to elevated temperature, are also mirrored by the nem1⌬ mutation (62,67).
The expression of PAH1 (as reflected in P PAH1 -lacZ and P PAH1 -DPP1 reporter gene activities) was induced as cells progressed from the exponential to stationary phases of growth, further confirming the relationship between Pah1p PAP activity and TAG synthesis. In addition, the growth phase-mediated induction of PAH1 was stimulated by inositol supplementation in stationary phase cells. The deletion analysis performed with the P PAH1 -lacZ reporter gene revealed regions in the promoter that were responsible for the growth phase-and inositol-mediated induction of gene expression. These regions contained putative binding sites for the Ino2-Ino4p complex (i.e. UAS INO element), Gis1p, and Rph1p. Accordingly, the regulation of PAH1 expression was examined in mutants lacking these transcription factors to shed light on their roles in this regulation. The inositol-mediated induction of the reporter gene activity was essentially lost in ino2⌬, ino4⌬, and opi1⌬ mutant cells. The ino2⌬ and ino4⌬ mutations also caused a loss of the growth phase-mediated induction of expression. In addition, the reporter gene activity was acutely reduced in opi1⌬ mutant cells. These observations supported the conclusion that the Ino2p/Ino4p/Opi1p regulatory circuit is involved in the regulation of PAH1 expression in response to growth phase and inositol supplementation. The induction by inositol and the abnormal patterns of regulation, however, were opposite to that normally shown for UAS INO -containing phospholipid synthesis genes (e.g. INO1, CDS1, CHO1, PSD1, CHO2, and OPI3) whose expressions in the exponential phase are repressed by inositol or by entrance into the stationary phase (20,101,102,106,107,131). In particular, the UAS INO -containing phospholipid genes are generally expressed at constitutively low levels in the ino2⌬ and ino4⌬ mutants and constitutively overexpressed in the opi1⌬ mutant (20,101,102,106,107,131). As discussed above, the Ino2p-Ino4p complex activates transcription, whereas Opi1p represses transcription (20,101,102,131).
PAH1 is not the only example of a lipid metabolic gene that is regulated by inositol and growth phase in a manner opposite to that of most phospholipid synthesis genes. Other genes include DPP1 and INM1, which encode a lipid phosphate phosphatase that also uses PA as a substrate and inositol monophosphate phosphatase, respectively (95,108,132,133). Like PAH1, the inositolmediated induction of DPP1 is lost in the ino2⌬, ino4⌬, and opi1⌬ mutants and, like PAH1, DPP1 is repressed in opi1⌬ and induced in ino4⌬ (95). The effects of the regulatory mutants on INM1 expression have not been reported.
That PAH1 expression was also regulated by Gis1p, and Rph1p might provide an explanation for the anomalous regulation by inositol and irregular expression patterns observed in the ino2⌬, ino4⌬, and opi1⌬ mutants when compared with that observed for phospholipid synthesis genes. The growth phasemediated induction of PAH1 was largely attenuated in gis1⌬ and rph1⌬ mutant cells indicating that Gis1p and Rph1p are activators of expression. However, in these mutants, the inositol supplementation caused the repression of PAH1 like that normally shown for the UAS INO -containing phospholipid synthesis genes. These observations raised the suggestion that the regulatory circuit involving Ino2p/Ino4/Opi1p was affected by Gis1p and Rph1p. For example, the binding of Gis1p and/or Rph1p to the PAH1 promoter might alter the interactions of Ino2p and/or Ino4p to the promoter and/or alter the interaction of Opi1p and Ino2p. Likewise, interactions of Ino2p/ Ino4p/Opi1p with the promoter might alter interactions by Gis1p and Rph1p. Interestingly, the putative binding sites for these transcription factors are clustered between Ϫ800 and Ϫ1,000 in the PAH1 promoter (Fig. 8), supporting the hypothesis for multiple interactions. Another possibility is that additional binding partners of Ino2p and/or Ino4p might affect PAH1 expression. Ino2p and Ino4p are members of the basic helix-loop-helix family of transcription factors (134,135), and an important feature of these proteins is their ability to form multiple dimer combinations conferring them with different DNA binding specificities (135). Ino4p, for example, is known to bind multiple basic helix-loop-helix proteins (136), making it possible for PAH1 expression to be regulated by yet other complex mechanisms. In this regard, the elevated expression of PAH1 in ino4⌬ mutant cells may reflect the loss of negative interaction. Confirmation that Ino2p, Ino4p, Gis1p, and Rph1p interact with the PAH1 promoter and whether or not multiple interactions occur at the promoter will require additional studies.
Prior to the identification of any of the PAP genes in S. cerevisiae, Morlock et al. (86) had observed that PAP activity is induced when wild type exponential phase cells are supplemented with inositol. We have shown here that the responsible gene could be PAH1 if Gis1p and/or Rph1p were not active. Otherwise, the responsible gene was likely DPP1, which is induced by inositol in the exponential and stationary phases of growth (95). Interestingly, DPP1 expression is also controlled by the Gis1p transcription factor, but in contrast to PAH1, Gis1p functions to repress DPP1 expression (108). Thus, the mechanisms by which Gis1p regulates the expressions of PAH1 and DPP1 are distinct. Moreover, the analysis of mutant cells lacking the various types of PAP activity has confirmed that DPP1 does not contribute to the de novo synthesis of TAG and that PAH1 is the responsible gene (26,27).
PA is the key component of the Ino2p/Ino4p/Opi1p regulatory circuit that controls the expression of UAS INO -containing phospholipid synthesis genes (20,101,102,131). The connection between the regulations of PAH1 with that of the UAS INOcontaining phospholipid synthesis genes is supported by genetic evidence. For example, several phospholipid synthesis genes (e.g. INO1, INO2, CHO1, are OPI3) are derepressed in pah1⌬ mutant cells, whereas cells that express a phosphorylation-deficient form of Pah1p that exhibits elevated PAP activity show repressed levels of these genes (27,62,67,71). These states of Pah1p function correlate with differences in the proportional synthesis of TAG and membrane phospholipids; high PAP activity favors the synthesis of TAG, whereas low PAP activity favors the synthesis of phospholipids (27, 62-64, 71, 79). It is also noteworthy that opi1⌬ mutant cells, which expressed low levels of the P PAH1 -lacZ reporter gene activity, exhibit elevated PA content (137), increased levels of phospholipids (138), and synthesize reduced amounts of TAG. 3 Collectively, the data support a model indicating that when PAH1 expression is repressed and PAP activity is low, the PA concentration in the nuclear/ER membrane is high, and thus, the synthesis of phospholipids is favored over TAG due to the derepression of phospholipid synthesis genes and by increased availability of PA for phospholipids synthesis via CDP-DAG. However, when PAH1 expression is induced and PAP activity is high, phospholipid synthesis genes are repressed via Opi1p; PA is partitioned to DAG, and the synthesis of TAG occurs at the expense of phospholipids. Clearly, the coordinated regulation of PAH1 expression with the expression of the phospholipid synthesis genes is very complex and warrants further investigations.