The Saccharomyces cerevisiae Actin Patch Protein App1p Is a Phosphatidate Phosphatase Enzyme*♦

Background: Phosphatidate phosphatase (PAP) plays diverse roles in lipid metabolism and cell signaling. Results: A novel yeast PAP is identified as the actin patch protein encoded by APP1. Conclusion: APP1 and other known genes (PAH1, DPP1, LPP1) are responsible for all detectable PAP activity in yeast. Significance: Identification of App1p as a PAP enzyme will facilitate the understanding of its cellular function. Phosphatidate phosphatase (PAP) catalyzes the dephosphorylation of phosphatidate to yield diacylglycerol. In the yeast Saccharomyces cerevisiae, PAP is encoded by PAH1, DPP1, and LPP1. The presence of PAP activity in the pah1Δ dpp1Δ lpp1Δ triple mutant indicated another gene(s) encoding the enzyme. We purified PAP from the pah1Δ dpp1Δ lpp1Δ triple mutant by salt extraction of mitochondria followed by chromatography with DE52, Affi-Gel Blue, phenyl-Sepharose, MonoQ, and Superdex 200. Liquid chromatography/tandem mass spectrometry analysis of a PAP-enriched sample revealed multiple putative phosphatases. By analysis of PAP activity in mutants lacking each of the proteins, we found that APP1, a gene whose molecular function has been unknown, confers ∼30% PAP activity of wild type cells. The overexpression of APP1 in the pah1Δ dpp1Δ lpp1Δ mutant exhibited a 10-fold increase in PAP activity. The PAP activity shown by App1p heterologously expressed in Escherichia coli confirmed that APP1 is the structural gene for the enzyme. Introduction of the app1Δ mutation into the pah1Δ dpp1Δ lpp1Δ triple mutant resulted in a complete loss of PAP activity, indicating that distinct PAP enzymes in S. cerevisiae are encoded by APP1, PAH1, DPP1, and LPP1. Lipid analysis of cells lacking the PAP genes, singly or in combination, showed that Pah1p is the only PAP involved in the synthesis of triacylglycerol as well as in the regulation of phospholipid synthesis. App1p, which shows interactions with endocytic proteins, may play a role in vesicular trafficking through its PAP activity.

PAP 2 catalyzes the dephosphorylation of PA to produce DAG and P i (1). The product DAG is utilized for the synthesis of the neutral storage lipid TAG and for the synthesis of the phospholipids phosphatidylcholine and phosphatidylethanolamine via the Kennedy pathway (1)(2)(3)(4)(5)(6). The substrate PA is utilized for the synthesis of phospholipids (e.g. phosphatidylinositol, phos-phatidylglycerol, and cardiolipin) via the liponucleotide intermediate CDP-DAG (7)(8)(9). In addition to serving as intermediates in lipid synthesis, PA and DAG are involved in lipid signaling. For example, PA plays a role in cell growth activation, membrane proliferation, the transcription of lipid synthesis genes, secretion, and vesicular transport, whereas DAG activates protein kinase C (10 -20). Thus, by the nature of its reaction, PAP activity may regulate the proportional synthesis of phospholipids and TAG, the pathways by which phospholipids are synthesized, and control the abundance of important signaling lipids. Indeed, biochemical and genetic studies using yeast and mammalian experimental systems have shown that PAP is an important regulator of lipid homeostasis and cell physiology (21)(22)(23)(24)(25).
Our laboratory has utilized the yeast Saccharomyces cerevisiae as a model eukaryote to study the enzymology and physiological roles of PAP enzymes (22,26). In S. cerevisiae, PAP is known to be encoded by three genes, namely PAH1 (27), DPP1 (28), and LPP1 (29). Pah1p PAP is a Mg 2ϩ -dependent enzyme whose reaction is based on the DXDX(T/V) catalytic motif within the haloacid dehalogenase-like domain of the enzyme (27,30). In contrast, the PAP activities of Dpp1p (28) and Lpp1p (29) do not have a divalent cation requirement, and their reactions are based on a three-domain lipid phosphatase motif composed of the consensus sequences KX 6 RP, PSGH, and SRX 5 HX 3 D (31,32). The Pah1p enzyme is specific for PA (27), whereas the Dpp1p and Lpp1p enzymes utilize PA and other lipid phosphate molecules (e.g. DAG pyrophosphate and lysoPA) as substrates (28,29,(33)(34)(35)(36). The PAP activities in yeast also differ with respect to their cellular locations and modes of regulation. Pah1p is found in the cytosol as a phosphoprotein that is a consequence of multiple phosphorylations (37)(38)(39)(40). For catalysis, the phosphorylated enzyme translocates to the nuclear/endoplasmic reticulum membrane through its dephosphorylation (38,(41)(42)(43). Dpp1p (28,44,45) and Lpp1p (29,46) are integral membrane enzymes with six transmembrane-spanning regions that are localized to the vacuole and Golgi compartments, respectively, of the cell.
Pah1p PAP plays an important role in lipid metabolism by controlling the relative proportions of PA and DAG (27,47). An imbalance of these intermediates due to a loss of PAP activity results in cellular abnormalities that include a drastic reduction in TAG abundance and susceptibility to fatty acid-induced lipotoxicity, the misregulation of phospholipid synthesis and an aberrant expansion of the nuclear/endoplasmic reticulum membrane, and defects in lipid droplet formation and vacuole homeostasis (27,30,43,(47)(48)(49). The Dpp1p and Lpp1p are not involved in de novo lipid synthesis that occurs at the endoplasmic reticulum (21,28,29). Instead, these enzymes are thought to have roles in lipid signaling by controlling the amounts of PA, DAG pyrophosphate, and lysoPA in the organelles where they reside (21,26).
A significant amount of Mg 2ϩ -dependent PAP activity is still present in the pah1⌬ dpp1⌬ lpp1⌬ triple mutant (27). However, unlike Pah1p PAP, this activity is sensitive to inhibition by N-ethylmaleimide (27). The enzyme activity is found in both the cytosolic and the membrane fractions of the cell, and its association with the membrane is peripheral in nature (27). Studies to gain insight into the physiological role of the PAP activity have been hampered because the gene encoding the enzyme has yet to be identified. In this work, we purified PAP from the pah1⌬ dpp1⌬ lpp1⌬ triple mutant and identified APP1 as the gene encoding the enzyme. The lack of detectable PAP activity in the app1⌬ pah1⌬ dpp1⌬ lpp1⌬ quadruple mutant indicated that all PAP activity in yeast is encoded by APP1, PAH1, DPP1, and LPP1. Moreover, this work identified the molecular function of the actin patch protein App1p as a PAP enzyme.

Materials
All chemicals were reagent grade. Growth medium supplies were purchased from Difco. Polyvinylidene difluoride membrane, the enhanced chemiluminescence Western blotting detection kit, phenyl-Sepharose CL-4B, MonoQ, and Superdex 200 columns were purchased from GE Healthcare. DE52 (DEAE-cellulose) was purchased from Whatman. Affi-Gel Blue, protein assay reagents, electrophoretic reagents, and protein standards were purchased from Bio-Rad. Radiochemicals were purchased from PerkinElmer Life Sciences. Bovine serum albumin, phenylmethylsulfonyl fluoride, benzamidine, aprotinin, leupeptin, pepstatin, isopropyl-␤-D-thiogalactoside, sodium cholate, and Triton X-100 were purchased from Sigma. Lipids were obtained from Avanti Polar Lipids. Silica gel thinlayer chromatography plates were from EM Science. Restriction endonucleases, modifying enzymes, and recombinant VentR DNA polymerase were purchased from New England Biolabs. Plasmid isolation and gel extraction kits and Ni 2ϩ -NTA-agarose resin were purchased from Qiagen. Invitrogen was the source of the DNA size standards and the yeast deletion consortium strain collection. The Yeast Maker yeast transformation kit was purchased from Clontech. Stratagene supplied the QuikChange site-directed mutagenesis kit. Nourseothricin (LEXSY NTC) was purchased from Jena Bioscience. Mouse monoclonal anti-HA, and anti-His 6 antibodies were from Roche Applied Science. Anti-App1p antibodies were prepared in rabbits against the C-terminal portion (residues 490 -502) of the protein at EZBiolab. Alkaline phosphatase-conjugated goat anti-rabbit IgG antibodies and alkaline phosphatase-conjugated goat anti-mouse IgG antibodies were from Thermo Scientific and Pierce, respectively. Scintillation counting supplies were purchased from National Diagnostics.

Strains and Growth Conditions
The strains used in this work are listed in Table 1. Yeast cells were grown in YEPD medium (1% yeast extract, 2% peptone, 2% glucose) or in synthetic complete medium containing 2% glucose at 30°C as described previously (50,51). For selection of yeast cells bearing plasmids, the appropriate amino acids were omitted from synthetic complete medium. Plasmid maintenance/amplifications (strain DH5␣) and App1p expression (strain BL21(DE3)pLysS) were performed in Escherichia coli. The bacterial cells were grown in LB medium (1% Tryptone, 0.5% yeast extract, 1% NaCl, pH 7.4) at 37°C, and ampicillin (100 g/ml) was added to select for the cells carrying plasmid. For growth on solid media, agar plates were prepared with supplementation of either 2% (yeast) or 1.5% (E. coli) agar. For App1p PAP purification in yeast, pah1⌬ dpp1⌬ lpp1⌬ mutant cells were grown to late exponential phase in YEPD medium at 30°C. For heterologous expression of His 6 -tagged App1p, E. coli BL21(DE3)pLysS cells bearing pMC101 were grown to A 600 nm ϭ 0.5 at 30°C in 1 liter of LB medium containing ampicillin (100 g/ml) and chloramphenicol (34 g/ml). The culture was then incubated for 3 h with 1 mM isopropyl-B-D-thiogalactoside to induce the expression of the PAP enzyme.

DNA Manipulations, Cloning of APP1, and Construction of Plasmids
Standard methods were used to isolate plasmid and genomic DNA and for manipulation of DNA using restriction enzymes, DNA ligase, and modifying enzymes (51). Transformations of S. cerevisiae (52) and E. coli (51) with DNA/plasmids and PCR reactions (53) followed the standard protocols. The plasmids used in this work are listed in Table 1. The S. cerevisiae APP1 gene was cloned by PCR. A 3,024-bp DNA fragment that contains the entire coding sequence (1,764 bp) of APP1, the 5Ј-untranslated region (824 bp), and the 3Ј-untranslated region (436 bp) was amplified from W303-1A genomic DNA (forward, 5Ј-GATTATAAGCTTAACTGACACCCATATCGCTTGACCC-3Ј; reverse, 5Ј-GGTCAATATACGTGCATCTAGAAGGCCT-TTCCCGAC-3Ј). The APP1 DNA fragment was digested with HindIII/XbaI and inserted into plasmid YEp351 at the same restriction enzyme sites. The multicopy plasmid containing APP1 was named pMC102. The APP1 gene was used to construct APP1 HA , in which the sequence for the HA epitope (YPY-DVPDYA) was located after the start codon. The 854-bp (forward, 5Ј-GATTATAAGCTTAACTGACACCCATATCGCT-TGACCC-3Ј; reverse, 5Ј-AGCGTAGTCTGGGACGTCGTA-TGGGTACATCTTTTTATTCCTTCTCCAAAGCAATTT-TTTCCCCC-3Ј) and 2,224-bp (forward, 5-TACCCATACGA-CGTCCCAGACTACGCTAATAGTCAAGGTTACGATGA-AAGCTCTTCCTCTACTGC-3Ј; reverse, 5Ј-GGTCAATAT-ACGTGCATCTAGAAGGCCTTTCCCGAC-3Ј) APP1 DNA fragments that contain the HA tag at the 3Ј and 5Ј ends, respectively, were amplified by PCR. These PCR products containing 27-bp overlapping ends were combined by overlap extension PCR. The combined 3,051-bp DNA was then amplified, digested with HindIII and XbaI, and inserted into YEp351 at HindIII/XbaI sites. The multicopy plasmid containing APP1 HA was named pMC103. Plasmids pMC102-D281E and pMC103-D281E were produced by PCR-mediated site-directed mutagenesis with the codon change of GAT to GAA. The nucleotide change in the APP1 and APP1 HA alleles was confirmed by DNA sequencing. For expression of APP1 in E. coli, its coding sequence (except the first nucleotide) was amplified by PCR from genomic DNA template (forward, 5Ј-TGAATA-GTCAAGGTTACGATGAAAGCTCTTCC-3Ј; reverse, 5Ј-TAATCCTCGAGTTAGTTTGAATACTTCTCCCTAATTC-TGCG-3Ј). The 1,774-bp PCR product was digested with XhoI, and the vector pET-15b was digested with NdeI, filled with Klenow, and digested with XhoI. The blunt cohesive end PCR products were inserted into pET-15b at NdeI (Klenow)/XhoI sites. The plasmid bearing His 6 -tagged APP1 was named pMC101.
Step 1: Preparation of Cell Extract-The pah1⌬ dpp1⌬ lpp1⌬ triple mutant cells were harvested in the late exponential phase, and cells were resuspended in buffer A (50 mM Tris-HCl (pH 7.5), 0.3 M sucrose, 10 mM 2-mercaptoethanol, 1 mM Na 2 EDTA, 0.5 mM phenylmethanesulfonyl fluoride, 1 mM benzamide, 5 g/ml aprotinin, 5 g/ml leupeptin, 5 g/ml pepstatin). Cells (200 g, wet weight) were disrupted with glass beads (0.5-mm diameter) using a Biospec Products BeadBeater as described previously (55). The cell lysate was centrifuged at 1,500 ϫ g for 10 min to remove unbroken cells and glass beads. Protein concentration was estimated by the method of Bradford (56) using bovine serum albumin as the standard.
Step 2: Preparation of Crude Mitochondria-Crude mitochondria were collected from the cell extract by centrifugation at 32,000 ϫ g for 10 min (57).
Step 3: Preparation of NaCl Extract-Mitochondria were suspended in buffer A containing 1 M NaCl to a final protein concentration of 10 mg/ml. The suspension was centrifuged at 32,000 ϫ g for 10 min to remove salt-unextractable proteins from mitochondrial membranes. The supernatant containing the salt-extracted proteins was then dialyzed overnight against buffer B (50 mM Tris-HCl (pH 7.5), 1 mM MgCl 2 , 10 mM 2-mercaptoethanol, 1 mM phenylmethylsulfonyl fluoride) containing 0.5% sodium cholate. Sodium cholate was added to buffer B to prevent the precipitation of PAP that occurred after the removal of salt from the enzyme preparation.
Step 4: DE52 Chromatography-A DE52 column (1.5 ϫ 7 cm) was equilibrated with buffer B containing 0.5% sodium cholate. The enzyme preparation from the previous step was applied to the column at a flow rate of 40 ml/h, and the column was washed with 5 column volumes of buffer B plus 0.5% sodium cholate followed by elution of the enzyme in 5-ml fractions with 10 column volumes of a linear NaCl gradient (0 -0.5 M) in the same buffer. The peak of PAP activity eluted at 0.2 M NaCl.
Step 5: Affi-Gel Blue Chromatography-An Affi-Gel blue column (1.5 ϫ 5 cm) was equilibrated with buffer B containing 0.5% sodium cholate and 0.2 M NaCl. The DE52-purified enzyme was applied to the column at a flow rate of 60 ml/h. The column was washed with 5 column volumes of buffer B containing 0.5% sodium cholate and 0.5 M NaCl. PAP activity was then eluted in 3-ml fractions with 12 column volumes of a linear gradient of NaCl (0.5-3.0 M) in buffer B containing 0.2% sodium cholate. The peak of activity eluted at a NaCl concentration of 1.3 M.
Step 6: Phenyl-Sepharose Chromatography-A phenyl-Sepharose (0.8 ϫ 2 cm) column was equilibrated with buffer B containing 0.2% sodium cholate and 3 M NaCl. The Affi-Gel Blue-purified enzyme was made to contain 3 M NaCl and then applied to the column under gravity flow. The column was first washed with 10 column volumes of the equilibration buffer and then washed with 10 column volumes of the buffer without NaCl. PAP activity was then eluted in 1-ml fractions with 1% Triton X-100 in buffer B. Hereafter, the PAP activity was made soluble with Triton X-100 instead of sodium cholate.
Step 7: MonoQ Chromatography-A MonoQ column (0.5 ϫ 5 cm) was equilibrated with buffer B containing 0.5% Triton X-100. The phenyl-Sepharose-purified enzyme was applied to the column; the column was then washed with 5 column volumes of buffer B containing 0.5% Triton X-100 followed by elution of PAP in 1-ml fractions with 20 column volumes of a linear NaCl gradient (0 -1 M) at a flow rate 24 ml/h. The peak of activity was eluted at NaCl concentration of 0.5 M.
Step 8: Superdex 200 Chromatography-A gel filtration Superdex 200 column (1 ϫ 30 cm) was equilibrated with buffer C containing 0.5 M NaCl and 0.5% Triton X-100. The MonoQpurified enzyme was applied to the column at a flow rate of 40 ml/h. The enzyme was then eluted in 0.5-ml fractions with 1 column volume of the same buffer.

Purification of His 6 -tagged App1p PAP from E. coli
All steps were performed at 8°C. E. coli cells expressing His 6tagged App1p were suspended in 20 ml of 20 mM Tris-HCl (pH 8.0) buffer containing 0.5 M NaCl, 5 mM imidazole, and 1 mM phenylmethylsulfonyl fluoride. Cells were disrupted by a freeze-thawing cycle and by two passes through a French press at 20,000 p.s.i. Unbroken cells and cell debris were removed by centrifugation at 12,000 g for 30 min at 4°C. The supernatant (cell lysate) was gently mixed with 2 ml of 50% slurry of Ni 2ϩ -NTA-agarose for 2 h. The Ni 2ϩ -NTA-agarose/enzyme mixture was packed in a 10-ml column and washed with 20 ml of 20 mM Tris-HCl (pH 8.0) buffer containing 0.5 M NaCl, 45 mM imidazole, 10% glycerol, and 7 mM 2-mercaptoethanol. His 6 -tagged App1p was eluted from the column in 1-ml fractions with 20 mM Tris-HCl (pH 8.0) buffer containing 0.5 M NaCl, 250 mM imidazole, 10% glycerol, and 7 mM 2-mercaptoethanol. Purified enzyme preparations were stored at Ϫ80°C.

PAP Assay
PAP activity was measured for 20 min by following the release of water-soluble [ 32 P]P i from chloroform-soluble [ 32 P]PA (5,000 cpm/nmol) at 30°C as described previously (58). The [ 32 P]PA substrate was synthesized enzymatically from DAG and [␥-32 P]ATP with E. coli DAG kinase (58). 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, 10 mM 2-mercaptoethanol, and enzyme protein in a total volume of 0.1 ml. Enzyme assays were conducted in triplicate, and the average standard deviation of the assays was Ϯ 5%. All reactions were linear with time and protein concentration. A unit of enzyme activity was defined as the amount of enzyme that catalyzed the formation of 1 nmol of product per minute.

SDS-PAGE and Western Blot Analysis
SDS-PAGE (59) using 10% slab gels and Western blotting (60, 61) using polyvinylidene difluoride membrane were performed as described previously. Proteins in polyacrylamide gels were visualized by staining with Coomassie Blue R250. The anti-HA, anti-His 6 , and anti-App1p antibodies were used at dilutions of 1:500, 1:5000, and 1:5000, respectively. Alkaline phosphatase-conjugated goat anti-mouse IgG antibodies and goat anti-rabbit IgG antibodies were used at a dilution of 1:5,000. Immune complexes were detected using the enhanced chemiluminescence Western blotting detection kit. Fluorimaging was used to acquire images from Western blots.

Identification of Proteins by Liquid Chromatography/Tandem Mass Spectrometry
An SDS-polyacrylamide gel slice containing proteins purified through the Superdex 200 chromatography step was subjected to trypsin digestion (62) followed by liquid chromatography/tandem mass spectrometry using a Thermo Fisher Scientific LTQ Orbitrap Velos instrument (63). The spectra were searched against the Swiss-Prot yeast database using the MASCOT (V.2.3) search engine (62). This work was performed at the Center for Advanced Proteomics Research facility of the University of Medicine and Dentistry of New Jersey (Newark, NJ).

Labeling and Analysis of Lipids
Steady-state labeling of phospholipids and neutral lipids with [ 32 P]P i and [2-14 C]acetate, respectively, was performed as described previously (27). Lipids were extracted from labeled cells by the method of Bligh and Dyer (64). Phospholipids were analyzed by two-dimensional thin-layer chromatography on silica gel plates using chloroform/methanol/ammonium hydroxide/water (45:25:2:3, v/v) as the solvent system for dimension one and chloroform/methanol/glacial acetic acid/ water (32:4:5:1, v/v) as the solvent system for dimension two (65). Neutral lipids were analyzed by one-dimensional thinlayer chromatography on silica gel plates using the solvent system hexane/diethyl ether/glacial acetic acid (40:10:1, v/v) (66). The identity of the labeled lipids on TLC plates was confirmed by comparison with standards after exposure to iodine vapor. Radiolabeled lipids were visualized by phosphorimaging analysis. The relative quantities of labeled lipids were analyzed using ImageQuant software. NOVEMBER 23, 2012 • VOLUME 287 • NUMBER 48

Analyses of Data
Statistical analyses were performed with SigmaPlot software.
The p values Ͻ 0.05 were taken as a significant difference.

PAP Activity in the pah1⌬ dpp1⌬ lpp1⌬ Triple Mutant Is
Encoded by the APP1 Gene-The pah1⌬ dpp1⌬ lpp1⌬ triple mutant, which does not contain PAP encoded by known genes (i.e. PAH1, DPP1, and LPP1), was used to purify the unknown PAP enzyme. The PAP activity in the triple mutant was associated with the cytosolic (0.7 Ϯ 0.3 nmol/min/mg), microsomal (1 Ϯ 0.1 nmol/min/mg), and mitochondrial (5 Ϯ 0.3 nmol/min/ mg) fractions. The PAP specific activity was enriched 5-fold in the mitochondrial fraction over the cell extract, and accordingly, this fraction was used as the source of enzyme. PAP activity was dissociated from mitochondrial membranes with 1 M NaCl followed by dialysis for desalting and chromatography with DE52, Affi-Gel Blue, phenyl-Sepharose, MonoQ, and Superdex 200. We noted the precipitation of PAP activity upon the removal of salt, and thus, sodium cholate or Triton X-100 was included in the chromatography buffers to maintain its solubility. The presence of salt also stabilized PAP activity. The Superdex 200 step afforded the greatest enrichment (4.5-fold) in specific activity ( Table 2). The elution profiles of PAP activity and protein and an SDS-PAGE analysis indicated that the enzyme preparation was not pure (Fig. 1). Attempts to further purify the enzyme were unsuccessful because the PAP activity was labile. Overall, the enzyme was purified 2,143-fold over the cell extract to a final specific activity of 1,500 nmol/min/mg with an activity yield of 5% ( Table 2).
The peak of PAP activity that emerged from the Superdex 200 column correlated with the enrichment of a minor protein band that migrated just above the size of the 50-kDa molecular mass marker (Fig. 1, fraction 27, indicated by the asterisk). An SDS-polyacrylamide gel slice containing this band was subjected to trypsin digestion followed by the analysis of peptides by liquid chromatography-mass spectrometry. This analysis yielded a few proteins of unknown function that upon further analysis (e.g. expression and purification in E. coli) showed no PAP activity. This indicated that the PAP enzyme in the preparation was very low in abundance. Accordingly, we used more sensitive liquid chromatography-tandem mass spectrometry to detect low abundance proteins. This analysis revealed that the enzyme preparation contained 112 proteins (or proteolytic fragments thereof). From the list, we focused on those proteins with unknown molecular function that possessed a phosphatase motif. We also considered proteins that are thought to be involved in phospholipid synthesis. Based on these criteria, we analyzed cell extracts from 20 mutants from the yeast deletion strain collection for a loss in PAP activity. Of the mutants, only app1⌬ exhibited a 30% decrease in PAP activity when compared with the wild type parental strain. These data indicated that the PAP activity might be directed by the APP1 gene.
The acronym APP1 stands for actin patch protein (67) because App1p is a component of cortical actin patches and interacts with endocytic proteins (67-75) (Fig. 2). Pfam analysis indicated that App1p consisting of 587 amino acids in length (66.1 kDa) contains a region with weak sequence similarity to a haloacid dehalogenase-like domain (76) (Fig. 2). Contained within this domain is a DXDX(T/V) (residues 281-285) catalytic motif that is present in the superfamily of Mg 2ϩ -dependent phosphatase enzymes (77) that include yeast Pah1p and mammalian lipin PAP enzymes (27,78). App1p also contains several PXXP motifs (Fig. 2) that are important for interactions with proteins that possess SH3 domains (79,80).
To prove that APP1 encodes PAP, we used heterologous overexpression of APP1 in E. coli. The S. cerevisiae APP1 coding sequence was amplified by PCR and inserted into plasmid pET-15b for the isopropyl-B-D-thiogalactoside-inducible expression of His 6 -tagged protein. The purified His 6 -tagged App1p migrated upon SDS-PAGE at the expected size of ϳ68 kDa (Fig. 3A). In addition, the purified protein was confirmed by immune reaction with antibodies directed against the His 6 epitope

App1p Is a Phosphatidate Phosphatase
and against a peptide sequence found at the C-terminal portion of App1p (Fig. 3A). The purified protein catalyzed the Mg 2ϩ -dependent PAP reaction, and the addition of 10 mM Na 2 EDTA abolished PAP activity. Unlike the Mg 2ϩ -dependent Pah1p PAP, App1p enzyme activity was sensitive to inhibition by N-ethylmaleimide (Fig. 3B). That App1p expressed in E. coli exhibits PAP activity also indicated that a posttranslational modification that might occur in yeast is not essential for enzyme activity. PAP Activity Is Affected by the app1⌬ pah1⌬ dpp1⌬ lpp1⌬ Mutations and by the APP1(D281E) Mutation-We constructed a variety of isogenic mutants where the app1⌬ muta-tion was combined with mutations of the other PAP genes in S. cerevisiae (Table 1). In this manner, we could determine the contribution of APP1 and the other PAP genes to the total PAP activity in yeast. In wild type cells, PAP activity is known to increase in the stationary phase of growth (81), and accordingly, the effect of the mutations was examined in both exponential and stationary phase cells (Fig. 4A). The APP1 gene accounted for 32 and 37% of the PAP activity in cell extracts of exponential and stationary phase cells, respectively. The PAH1-encoded enzyme accounted for 11 and 24% of the PAP activity in exponential and stationary phases, respectively. In the exponential phase, DPP1 and LPP1 accounted for 60% of the total PAP activity. A comparison of the activities from app1⌬ pah1⌬ and dpp1⌬ lpp1⌬ double mutants and from the app1⌬ dpp1⌬ lpp1⌬ and pah1⌬ dpp1⌬ lpp1⌬ triple mutants indicated that APP1 and PAH1 were primarily responsible for the induced expression of PAP activity in the stationary phase. The induced expression of DPP1-encoded PAP activity in stationary phase cells is governed by inositol (82), a phospholipid precursor that was not supplemented to the growth medium in this work. Finally, the lack of detectable PAP activity in the app1⌬ pah1⌬ dpp1⌬ lpp1⌬ quadruple mutant indicated that all PAP activity (i.e. Mg 2ϩ -dependent and Mg 2ϩ -independent) in S. cerevisiae was attributed to APP1, PAH1, DPP1, and LPP1 (Fig. 4A).
APP1 and HA-APP1 versions were expressed on a multicopy plasmid in the app1⌬ pah1⌬ dpp1⌬ lpp1⌬ quadruple mutant, and cell extracts were prepared for the measurement of PAP activity. The PAP-deficient mutant expressing these genes exhibited PAP activities of 6.8 Ϯ 0.7 and 6.5 Ϯ 0.6 nmol/min/ mg, respectively (Fig. 4B). This corresponded to about a 10-fold overproduction of activity when compared with that found in the pah1⌬ dpp1⌬ lpp1⌬ triple mutant (Fig. 4A). That both versions of APP1 directed the same level of activity indicated that the HA epitope did not compromise enzyme activity. The first    NOVEMBER 23, 2012 • VOLUME 287 • NUMBER 48 conserved aspartate residue in the DIDDT sequence in App1p was mutated to glutamate by site-specific mutagenesis. Glutamate was chosen to replace aspartate to conserve the charge of the amino acid at this position, and thus, minimize a structural change in the enzyme. The D281E mutation abolished the overexpressed PAP activity directed by the APP1 and HA-APP1 genes (Fig. 4B). An immunoblot analysis using anti-HA antibodies showed that the D281E mutation in the HA-tagged version of App1p did not affect the expression of the enzyme (Fig.  4B, inset). These data indicated that the DIDDT sequence in App1p is responsible for its PAP catalytic function.

App1p Is a Phosphatidate Phosphatase
Effects of the app1⌬ pah1⌬ dpp1⌬ lpp1⌬ Mutations on Lipid Composition-The PAP deletion mutants were used to determine the contribution of APP1, PAH1, DPP1, and LPP1 to lipid composition. In the first set of experiments, wild type and mutant cells were labeled to steady state with [2-14 C]acetate to analyze DAG, TAG, and total phospholipids. Cells were grown to the stationary phase (the growth phase when DAG and TAG are most affected (27)), and lipids were extracted and then analyzed by one-dimensional TLC. When compared with the pah1⌬ mutation, the app1⌬ and dpp1⌬ lpp1⌬ mutations did not have a significant effect on the relative amounts of DAG, TAG, and total phospholipids (Fig. 5). Moreover, alterations in lipid composition observed in the double, triple, and quadruple mutants were attributed to the pah1⌬ mutation (Fig. 5). These results indicate that of the genes encoding PAP in S. cerevisiae, PAH1 is responsible for the synthesis of TAG and the regulation of phospholipid synthesis. In the second set of experiments, the wild type and mutant cells were labeled to steady state with [ 32 P]P i to analyze the composition of individual phospholipids. The phospholipids were extracted from exponential phase cells (the growth phase when phospholipid composition is most affected (27)) and analyzed by two-dimensional TLC. The relative amounts of the major phospholipids phosphatidylcholine, phosphatidylethanolamine, phosphatidylinositol, phosphatidylserine, and the precursor PA were not affected by the app1⌬ and dpp1⌬ lpp1⌬ mutations (Fig. 6). The relative amounts of these phospholipids, however, were affected by the pah1⌬ mutation alone and in combination with the app1⌬ and dpp1⌬ lpp1⌬ mutations (Fig. 6). As described previously (27), the pah1⌬ mutation caused decreases in the amounts of phos-phatidylcholine (37%) and phosphatidylserine (33%) and increases in the amounts of phosphatidylethanolamine (108%), phosphatidylinositol (42%), and PA (86%). Interestingly, when combined with the pah1⌬ dpp1⌬ lpp1⌬ mutations, the app1⌬ mutation reversed the effect of the pah1⌬ mutation on the composition of phosphatidylcholine and phosphatidylethanolamine. The reason for this change is unclear.
The app1⌬ Mutant Does Not Exhibit Obvious Phenotypes-We noted that the app1⌬ mutant in the BY4741 genetic background exhibited slower growth on synthetic medium and was sensitive to elevated temperature (i.e. 37°C). It is known that the BY4741 strain has more stringent growth requirements when compared with other strains used by the yeast research community (83). Accordingly, the growth requirements of the app1⌬ mutant were also examined in the W303-1A genetic background. This analysis showed that the app1⌬ mutation did not cause obvious growth defects. In addition, no striking phenotypes (e.g. changes in cellular morphology, respiratory deficiency, salt sensitivity) were identified. In striking contrast, the pah1⌬ mutant exhibits a slow growth phenotype, is temperature-sensitive in several genetic backgrounds, and exhibits defects in cellular morphology (27,43,48,49,84). None of these phenotypes were complemented by the expression of APP1, substantiating that App1p and Pah1p PAP activities have distinct functions in lipid metabolism and cell physiology.

DISCUSSION
PAP is generally recognized as an important enzyme in eukaryotic organisms because its substrate PA and product DAG play important roles in the synthesis of TAG and phospholipids and in other aspects of cell physiology (e.g. transcription, lipid signaling, and vesicular trafficking) (1)(2)(3)(4)(5)(6)(7)(8)(9)(10)(11)(12)(13)(14)(15)(16)(17)(18)(19)(20). Two basic types of PAP activity are found in eukaryotic organisms: an activity that is dependent on Mg 2ϩ and an activity that is not dependent on Mg 2ϩ or any other divalent cation (21,22,74). Mg 2ϩ -dependent PAP activity is governed by a DXDX(T/V) catalytic motif (77), whereas Mg 2ϩ -independent activity is directed by a three-domain catalytic motif consisting of the sequences KX 6 RP, PSGH, and SRX 5 HX 3 D (31). In the yeast S. cerevisiae, the Mg 2ϩ -dependent PAP is encoded by PAH1  App1p Is a Phosphatidate Phosphatase (27), whereas the Mg 2ϩ -independent type of the enzyme is encoded by DPP1 (28) and LPP1 (29). The pah1⌬ dpp1⌬ lpp1⌬ triple mutant lacking the known Mg 2ϩ -dependent and Mg 2ϩindependent PAP enzymes still possesses the Mg 2ϩ -dependent activity that is sensitive to inhibition by N-ethylmaleimide (27), and the identity of the gene(s) encoding this activity was the focus of this work.
APP1 was identified through a reverse genetics approach using protein sequence information derived from the PAP enzyme isolated from the pah1⌬ dpp1⌬ lpp1⌬ triple mutant. Obtaining App1p sequence information was not straightforward. The PAP activity was labile in the absence of high salt, which comprised the effectiveness of chromatography steps that required low salt for enzyme binding (e.g. ion exchange chromatography). Although the eight-step purification scheme resulted in a 2,143-fold enrichment in PAP specific activity, it did not result in a homogeneous enzyme preparation that could facilitate unequivocal protein sequence determination. Although the sensitive liquid chromatography/tandem mass spectrometry technology yielded many protein candidates, it allowed us to identify App1p that was present in very low abundance. In the end, the collective data (e.g. reduction of PAP activity in the app1⌬ mutant, the APP1-directed overexpression of PAP activity in the app1⌬ pah1⌬ dpp1⌬ lpp1⌬ quadruple mutant, and the heterologous expression of App1p PAP activity in E. coli) provided a conclusive level of evidence that APP1 is the structural gene encoding a PAP enzyme in S. cerevisiae.
Efforts to identify APP1 by informatics and by genetic approaches were unsuccessful. For example, a protein BLAST search using Pah1p as the query did not identify App1p or any other homologs in the Saccharomyces genome database. A BLAST search against higher eukaryotic organism databases identified mammalian lipin proteins, but this was expected because a BLAST search with mouse lipin-1 as the query identifies S. cerevisiae Pah1p (78). Likewise, a BLAST search using App1p as the query does not identify Pah1p or the mammalian lipins. Instead, the BLAST search identifies App1p homologs only found in fungi. A synthetic genetic array screen using the pah1⌬ mutant in combination with the cho2⌬ and opi3⌬ mutations defective in the phosphatidylethanolamine methylation steps of phosphatidylcholine synthesis via the CDP-DAG pathway (85,86) has also been performed. 3 The rationale for this analysis was that the loss of a gene encoding PAP in combination with the loss of Pah1p causes lethality due to a lack of DAG production required for the synthesis of phosphatidylcholine via the Kennedy pathway (85,86). The genetic screen, however, did not lead to the identification of APP1, 3 indicating that the cellular functions of App1p and Pah1p do not overlap with each other with respect to lipid synthesis. This assertion was further supported by the fact that APP1 did not complement phenotypes (e.g. temperature sensitivity) exhibited by the pah1⌬ mutant and that the analysis of cells possessing the app1⌬ mutation alone and in combination with mutations for other known PAP genes indicated that only Pah1p PAP was involved in de novo lipid synthesis.
Although there is essentially no sequence homology between App1p and Pah1p, both enzymes possess the canonical DXDX(T/V) catalytic motif that is typical of Mg 2ϩ -dependent phosphatase enzymes (77). For Pah1p, its DIDGT catalytic sequence is contained within the haloacid dehalogenase-like domain similar to that found in the mammalian lipin PAP enzymes (21,23,78). However, the haloacid dehalogenase-like domain is not found in App1p, but instead, it contains a conserved domain found only in fungi that has overlapping regions that show weak sequence similarity to the haloacid dehalogenase-like domain found in Pah1p and lipin (76). It is within this domain that the App1p DIDDT sequence is found (Fig. 2), and indeed, the D281E mutation abolished PAP activity, confirming this to be the catalytic sequence.
PAP enzymes isolated from S. cerevisiae are known to have molecular masses of 91 kDa (Mg 2ϩ -dependent) (87), 75 kDa (Mg 2ϩ -dependent) (88), 45 kDa (Mg 2ϩ -dependent) (89), and 34 kDa (Mg 2ϩ -independent) (34). Protein sequence information has confirmed that the 91-kDa enzyme is a proteolysis product of Pah1p (27) and that the 34-kDa enzyme is Dpp1p (28). Lpp1p PAP was identified based on its sequence homology with Dpp1p (29). The identity of the genes encoding the 75-and 45-kDa forms of PAP has been an enigma. Although the basic enzymological properties of these enzymes are similar, the 75-kDa enzyme is soluble, whereas the 45-kDa enzyme is associated with mitochondria (88,89). Because of the differences in size and location, it has been assumed that the 75-and 45-kDa PAP enzymes are encoded by different genes. Based on the result that no detectable PAP activity was present in the app1⌬ pah1⌬ dpp1⌬ lpp1⌬ quadruple mutant, we hypothesize that the 75-kDa PAP was the soluble form of App1p and that the 45-kDa PAP was a proteolytic fragment of App1p bound to mitochondrial membranes. Unfortunately, this hypothesis cannot be addressed because preparations of the 75-and 45-kDa PAP enzymes are no longer available. Obviously, the protein used for sequence analysis in this study was a proteolytic fragment of App1p.
The PAP enzymes in S. cerevisiae are found in different cellular locations and play diverse roles in cell physiology (74). Pah1p is a cytosolic enzyme that associates with and functions at the nuclear/endoplasmic reticulum membrane to regulate the synthesis of TAG and membrane phospholipids (21,22,27,42,74). Dpp1p and Lpp1p, respectively, are thought to control the signaling functions of PA, DAG pyrophosphate, and lysoPA in vacuole and Golgi membranes (21, 22, 28, 29, 34, 44 -46, 74, 90). Like Pah1p, App1p is a cytosolic protein (46) that associates with membranes, but the role this PAP plays in cell physiology is not yet clear.
The formation of endocytic vesicles in S. cerevisiae involves a series of processes that include actin patch assembly, actin polymerization, and changes in membrane structure and curvature (75,92). These processes involve numerous endocytic proteins, and App1p is known to physically interact with many of them (67-75) (Fig. 2). Although these interactions have yet to be studied in detail, the presence of proline-rich regions (PXXP motifs) in App1p suggests interactions with SH3 domains of Abp1p, Ysc84p, Sla1p, Bbc1p, Bzz1p, and Rvs167p (79, 80). App1p interactions with proteins that do not possess SH3 domains (e.g. Las17p, Sla2p, Cof1p, Crn1p, and Rvs161) must occur through other mechanisms yet to be defined. Based on these protein interaction data, App1p is postulated to play a role in endocytosis (67). However, until now the molecular function of App1p has been unknown. App1p PAP located at cortical actin patches (92) may regulate the local concentrations of PA and DAG. These lipids are known to facilitate membrane fission/fusion events in model systems (93)(94)(95)(96)(97)(98), and they are also known to interact with and regulate enzymes (e.g. phosphatidylinositol 4-phosphate kinase, protein kinase C, protein kinase D) that play important roles in vesicular trafficking (99 -103). For example, in mammalian cells, the inhibition of PAP activity by propranolol attenuates protein kinase D recruitment to Golgi membranes, blocking vesicle bud formation and protein transport to the cell surface (103,104). Unfortunately, propranolol does not discriminate between Mg 2ϩ -dependent and Mg 2ϩ -independent PAP activities (90,(105)(106)(107), and thus, the identity of the type of PAP enzyme involved in this process is unknown. We speculate that in yeast, App1p PAP activity plays a role in vesicle formation through its recruitment from the cytosol to cortical actin patches via endocytic proteins. These proteins may tether App1p PAP to actin patches and/or serve to regulate the relative amounts of PA and DAG, which in turn contribute to the control of vesicle formation. Clearly, the work reported here provides an impetus to pursue these questions in more detail. Although an App1p homolog does not exist in higher eukaryotes, a functional homolog exists in the form of the lipin PAP enzymes that catalyze the same reaction (27,106,108), and as indicated above, a PAP activity is implicated in vesicular trafficking in mammalian cells.