The Plasmodium falciparum PfGatp is an Endoplasmic Reticulum Membrane Protein Important for the Initial Step of Malarial Glycerolipid Synthesis*

During its 48-h asexual life cycle within human erythrocytes, Plasmodium falciparum grows to many times its own size and divides to produce 16–32 new parasites. This rapid multiplication requires active synthesis of new membranes and is fueled by phospholipid precursors and fatty acids that are scavenged from the human host. Plasmodium membrane biogenesis relies heavily on the expression of parasite enzymes that incorporate these precursors into phospholipids. However, little is known about the genes involved in membrane biogenesis or where this process takes place within the parasite. Here, we describe the analysis in P. falciparum of the first step of phospholipid biosynthesis that controls acylation of glycerol 3-phosphate (GPAT) at the sn-1 position. We show that this activity is of parasite origin and is specific for glycerol 3-phosphate substrate. We have identified the gene, PfGAT, encoding this activity in P. falciparum and reconstituted its codon composition for optimal expression in the yeast Saccharomyces cerevisiae. PfGAT complements the lethality of a yeast double mutant gat1Δgat2Δ, lacking GPAT activity. Biochemical analysis revealed that PfGatp is a low affinity GPAT enzyme with a high specificity for C16:0 and C16:1 substrates. PfGatp is an integral membrane protein of the endoplasmic reticulum expressed throughout the intraerythrocytic life cycle of the parasite but induced mainly at the trophozoite stage. This study, which describes the first protozoan GPAT gene, reveals an important role for the endoplasmic reticulum in the initial step of Plasmodium membrane biogenesis.

Plasmodium species are obligate intraerythrocytic protozoan parasites that undergo a number of developmental stages in the vertebrate host. In humans, they annually cause clinical illness in 300 -500 million people with 1.5-2.7 million deaths, mainly caused by Plasmodium falciparum (1). Drug resistance is widespread, and the need for more efficacious and less toxic agents that exploit pathways and targets unique to the parasite is acute.
In the 48 h after invasion of human red blood cells, P. falciparum grows to many times its original size and then divides to produce 16 -32 daughter parasites. This high rate of growth and multiplication requires synthesis of new membranes. Accordingly, the phospholipid content of malaria-infected erythrocytes increases by up to 5-fold during parasite maturation, with 85% of the newly synthesized phospholipids being either phosphatidylcholine or phosphatidylethanolamine (2). Parasite infection is also accompanied by a marked increase in neutral lipid species like fatty acids, diacylglycerol (DAG) 1 and triacylglycerol (TAG) (3). This synthesis of parasite phospholipids and neutral lipids relies upon transport of choline, inositol, and fatty acids from host plasma (2, 4 -10) and de novo synthesis of fatty acids by type II fatty acid-synthesizing enzymes (11). The finished malaria genome sequence revealed the presence in P. falciparum of type II fatty acid and phospholipid genes (12,13). Whereas all of the known and predicted enzymes of the type II fatty acid-synthesizing enzyme pathway contain a signal peptide and a targeting sequence for apicoplast (14,15), a plastid-like organelle in Apicomplexa parasites, most known and predicted enzymes for the synthesis of phosphatidylcholine, phosphatidylethanolamine, phosphatidylserine, and phosphatidylinositol lack these signals, suggesting that the synthesis of the malarial phospholipids does not occur in the apicoplast and that this process takes place in other cellular organelles, the identity of which is not yet known.
Because of their importance for parasite development, the pathways of synthesis of phospholipids have long been considered attractive targets for chemotherapy. Accordingly, quaternary ammonium compounds, analogs of choline, have been shown to interfere with phospholipid metabolism, to inhibit parasite growth in vitro, and to clear malaria infection in mice and monkeys (16). In eukaryotes, the initial step of phospholipid synthesis involves acylation of glycerol 3-phosphate at the sn-1 position by glycerol-3-phosphate acyltransferases (GPAT) to form lysophosphatidic acid (17)(18)(19)(20)(21). Lysophosphatidic acid acyltransferases then catalyze the acylation of lysophosphatidic acid at the sn-2 position to generate phosphatidic acid (19,20). Phosphatidate phosphatase and CDP-DAG synthase en-zymes convert the phosphatidic acid formed into DAG and CDP-DAG, respectively. DAG subsequently enters the de novo CDP-choline and CDP-ethanolamine Kennedy pathways for synthesis of phosphatidylcholine and phosphatidylethanolamine, respectively. CDP-DAG enters the CDP-DAG pathway for synthesis of phosphatidylserine and phosphatidylinositol (22)(23)(24). DAG also serves as a substrate to DAG acyltransferases for the synthesis of TAGs (25,26).
Here we describe the identification and characterization in P. falciparum of the gene, PfGAT, encoding GPAT activity. We show that the encoded protein, PfGatp, is a yeast-like GPAT enzyme localized in the endoplasmic reticulum membrane and exists as a large multimeric protein complex. PfGatp activity is required for survival of yeast cells lacking GPAT activity because of the loss of the two GPAT-encoding genes GAT1 and GAT2 (27,28). The identification of PfGAT will set the stage for a better understanding of glycerolipid biosynthesis in P. falciparum and could lead to better therapeutic strategies to inhibit this process and block parasite proliferation.

EXPERIMENTAL PROCEDURES
Parasite Culture-All reagents were from Sigma unless otherwise specified. P. falciparum clones were grown using the method developed by Trager and Jensen (29). Serum was replaced with 0.5% Albumax (Invitrogen) in the culture medium.
Plasmid Construction-Codon-optimized PfGAT CO was synthesized using the forward and reverse primers shown in Fig. 5 and those described below. PfGAT CO was first assembled and amplified as four small fragments that were subsequently used as templates to amplify the whole gene. Assembly reactions were performed using Platinum Taq High Fidelity enzyme (Invitrogen), with 4 M final concentration for each primer. The program for assembly is the following: 2 min at 94°C, 25 cycles of 30 s at 94°C, 30 s at 55°C, and 1 min at 68°C, and terminated by a final elongation at 68°C for 3 min. The resulting products were purified and used as templates for amplification using Platinum Taq High Fidelity enzyme and the following PCR program: 2 min at 94°C, 20 -25 cycles of 30 s at 94°C, 30 s at 63°C, and 45 s at 68°C, and terminated by a final elongation at 68°C for 2 min. The full-length PfGAT CO was cloned into XmaI and KpnI sites of the pBEVY-L ADH1 LEU2 plasmid thus yielding the vector pBEVY-L ADH1::PfGAT CO LEU2. For construction of the plasmid pYES2.1 GAL1::PfGAT URA3, PfGAT was PCR-amplified using genomic DNA from the P. falciparum 3D7 clone and the primers OCHO101 (CGCG-GATCCATGCCAGATTTTTACTTTTTAATAAGATGG) and ScPfGat-FЈ (TAAGATCTCTTCCTTATATTCTAATTG) and subsequently cloned into the pYES2.1/V5-His-TOPO vector (Invitrogen). Similarly the pYE-S2.1 GAL1::PfGAT CO URA3 was generated by PCR amplification of PfGAT CO using pBEVY-L ADH1::PfGAT CO LEU2 as a template and the primers PfScGat1 primer 1 (CCCCCCGGGATGCCAGATTTCTACTTC-CTAATCAGATGGCTGTGTAAGGTTATCGTTAA) and ScPfGat-FЈ (T-AAGATCTCTTCCTTATATTCTAATTG) followed by cloning into the pYES2.1/V5-His-TOPO vector. The sequence contiguity of PfGAT and PfGAT CO was confirmed by DNA sequencing.
Purification of PfGatp, Antiserum Production and Purification, and Protein Immunoblotting-The 3Ј-end of PfGAT open reading frame was PCR amplified using the oligonucleotides Ct-X-PfGAT-5Ј (CCGCTCG-AGATGGGAAAGGAAAAAACACAT) and Ct-B-PfGAT-3Ј (CGCGGAT-CCTCAGTAAAGGTTTCGACAACC). The resulting PCR product was digested with XhoI and BamHI restriction enzymes and cloned into the XhoI and BamHI sites of the expression vector pET-15b (Novagen), thus creating pET-15b-PfGAT-His 6 -78 plasmid. This plasmid was expressed in BL21-CodonPlus-RIL Escherichia coli strain (Stratagene). A 200-ml culture of E. coli was grown to an A 600 of 0.6 in Luria broth at 37°C. The cells were induced with 0.4 mM isopropyl ␤-D-thiogalactoside and incubated for an additional 6 h at 37°C. Cells were collected by centrifugation at 2,000 ϫ g for 15 min at 4°C and lysed. Histidine-tagged protein was purified by Ni 2ϩ chromatography at 4°C (Qiagen) under denaturing conditions. Purified recombinant Ct-PfGatp was used to immunize rabbits, which was performed by Cocalico Biologicals, Inc.
To generate an antigen for screening and purification of PfGatpspecific antibodies, the pET-15b-PfGAT-His 6 -78 plasmid was digested with XhoI and HindIII restriction enzymes, and the 0.6-kb fragment containing the 3Ј-end of PfGAT was cloned into the SalI and HindIII sites of the expression vector pMalC2-X (New England Biolabs), thus creating pMalC-PfGat-MBP-78 plasmid. This plasmid synthesizes the PfGatp C-terminal fragment as a fusion protein to the N-terminal portion of the E. coli maltose-binding protein. pET-15b-PfGat-MBP-78 plasmid was expressed in BL21-CodonPlus-RIL E. coli strain. The purification of the PfGatp-MBP-78 fusion protein was performed using maltose affinity chromatography according to the manufacturer's instructions (New England Biolabs). The crude serum obtained from Cocalico Biologicals was affinity-purified over an Affi-15 gel (Bio-Rad) matrix to which PfGatp-MBP-78 fusion protein was bound covalently. For immunoblots, parasite lysates were resuspended in SDS-PAGE sample buffer, and the proteins were separated by electrophoresis on 12% SDS-polyacrylamide gels and transferred to nitrocellulose membranes. Preincubation, antibody incubations, and washes were conducted in 10 mM Tris-Cl, pH 8, 150 mM sodium chloride, and 0.05% Tween 20 with 5% skim milk. Preimmune and purified antibodies were used at 1:100 dilution. A chemiluminescence kit (ECL, Amersham Biosciences) was used to detect the immunological reaction.

GPAT and Dihydroxyacetone
Phosphate Acyltransferase (DHAPAT) Assays-Parasites from 3D7 asynchronous and synchronous cultures (2% hematocrit, 10% parasitemia) were isolated from infected erythrocytes by treatment with 0.07-0.1% saponin for 15 min at 0°C followed by centrifugation at 2,061 ϫ g for 15 min. The pellet was washed in PBS and resuspended in 500 l of 50 mM Tris-HCl, pH 7.5, 10% glycerol, 1 mM dithiothreitol, 1 mM EDTA. After sonication followed by centrifugation at 1,500 ϫ g, the supernatant was recovered and used for GPAT and DHAPAT assays. Yeast extracts were obtained as described previously (28). For the GPAT assay, 200 g of protein extracts was added to 200 l of GPAT buffer (75 mM Tris-HCl, pH 7.5, 1 mM dithiothreitol, 2 mM MgCl 2 , 45 M fatty acyl-CoA, 1 mg/ml bovine serum albumin, and 0.4 mM [ 14 C]glycerol 3-phosphate (2.5 Ci/mol) and incubated at 37°C for 1 h except as otherwise mentioned. The reaction was stopped by the addition of 600 l of 1% HClO 4 . The DHAPAT assay was performed as described by Athenstaedt et al. (30). For both activities, lipid extraction was performed by adding 3 ml of chloroform:methanol (1:2, v/v) to the mixture followed by adding 1 ml of chloroform and 1 ml of 1% HClO 4 . After centrifugation at 1,250 ϫ g for 5 min, the organic phase was recovered and washed with 2 ml of 1% HClO 4 followed by centrifugation at 1250 ϫ g for 5 min. The chloroform phase was transferred to a scintillation vial, dried, and counted. Aqueous and organic phases were also analyzed by thin layer chromatography (TLC) using silica gel plates (Whatman). The hydrophilic phase of the first extraction was dried in a SpeedVac and resuspended in choroform for loading on TLC. A solvent made of chloroform:methanol:water:acetic acid (70:30:4:2) was used to separate radiolabeled products and to confirm their identity. The main product detected in the organic phase of the GPAT assay was phosphatidic acid. No phosphatidic acid or lysophosphatidic acid could be detected in the aqueous phase. For the DHAPAT assay, low counts could be measured from the organic phase, and no radiolabeled acyldihydroxyacetone phosphate could be measured in the water phase after TLC separation. Standards used were 1-oleoyl-sn-glycerol 3-phosphate and 1,2-dioleoyl-sn-glycerol 3-phosphate (Sigma).
Gel Filtration Assay-Infected red blood cells from 84 ml of P. falciparum asynchronous culture (2% hematocrit, 10% parasitemia) were treated with 0.15% saponin for 15 min at 0°C. After centrifugation at 1,875 ϫ g for 10 min the pellet was washed in PBS, resuspended in PBS containing a mixture of protease inhibitors, sonicated, incubated in 1% Triton X-100 for 30 min at 0°C, and centrifuged at 16,300 ϫ g for 15 min. The supernatant was concentrated and separated on a Superose 12 HR 10/30 column at a flow rate of 0.2 ml/min. Fractions of 1 ml were collected and tricholoroacetic acid precipitated. The precipitates were resuspended in 20 l of SDS-PAGE loading buffer, neutralized and separated by electrophoresis on 10% SDS-polyacrylamide gels, and analyzed by Western blotting, using affinity-purified anti-PfGatp antibodies.
Analysis of PfGatp Membrane Association-Infected red blood cells from a 36-ml asynchronous culture (2% hematocrit, 10% parasitemia) were treated with 0.07% saponin for 15 min at 0°C. After centrifugation at 1,875 ϫ g for 10 min, the pellet was washed in PBS and resuspended in PBS. After sonication, the extract was subjected to various treatments followed by a 10-min centrifugation at 100,000 ϫ g. Treatments included a 30-min incubation at 0°C with buffer alone, 1% Triton X-100, 1% Triton X-114, 1% n-decyl-␤-D-maltoside (Anatrace), 0.5 M potassium acetate or 0.1 M Na 2 CO 3 at pH 11. Supernatant and pellet fractions were separated by SDS-PAGE and immunoblotted with affinity-purified anti-PfGatp and anti-PfNT1 antibodies (31). Bound antibodies were visualized by ECL.
Immunofluorescence Microscopy-Synchronous cultures of P. falciparum-infected erythrocytes were washed twice in PBS, placed onto coverslips, and dried at room temperature. Fixation, washes, and mounting were performed as described by Rager et al. (31). Coverslips were incubated simultaneously with affinity-purified anti-PfGatp antibodies (diluted 1:10) and either mouse monoclonal antibodies (Sigma) to the red blood cell Band 3 protein (diluted 1:500) or rat polyclonal antibodies (MR4) to the P. falciparum endoplasmic reticulum marker BiP (32) at 37°C with gentle shaking for 1 h. The coverslips were washed and then incubated with anti-rabbit fluorescein isothiocyanate (FITC) conjugate and anti-mouse conjugated to Texas Red (Molecular Probes) or anti-rabbit rhodamine and anti-rat FITC-conjugated secondary antibodies for 1 h at 37°C. Nuclei were stained by incubating the coverslips in PBS containing 3 g/ml Hoechst stain (Molecular Probes) for 5 min at room temperature. Mitochondrial staining was performed by incubating infected red blood cells with 250 nM MitoTracker Red CMXRos (Molecular Probes) for 5 min prior to fixation and incubation with affinity-purified PfGatp antibodies (diluted 1:10). The coverslips were washed and then incubated with goat anti-rabbit conjugated to FITC (Molecular Probes) secondary antibodies for 1 h at 37°C. Images were analyzed by high resolution fluorescence and confocal microscopy.

RESULTS
GPAT and DHAPAT Activities in P. falciparum-To examine the presence of GPAT and/or DHAPAT activities in P. falcipa- rum-infected erythrocytes and to determine their origin (i.e. red blood cell or parasite), hemolysate and parasite fractions were prepared from red blood cells infected with an asynchronous culture of the 3D7 clone of P. falciparum and analyzed for their GPAT and DHAPAT activities using glycerol 3-phosphate and dihydroxyacetone phosphate substrates, respectively. Parasite extracts were able to catalyze the acylation of glycerol 3-phosphate substrate and dihydroxyacetone phosphate substrates (Fig. 1, A and B). However, the P. falciparum DHAPAT represented less than 1% of the GPAT activity (Fig. 1B). No GPAT or DHAPAT activities could be detected in red blood cell hemolysates from infected as well as control uninfected red blood cells (Fig. 1, A and B). To determine the GPAT activity during P. falciparum intraerythrocytic development, extracts were prepared from a highly synchronous culture of 3D7-infected erythrocytes and analyzed for GPAT activity. Equal amounts of proteins from each developmental stage resulted in relatively similar acylation activities (not shown). However, determination of the total activity per developmental stage indicated a 1.  1C). Consistent with data from asynchronous parasites, only residual DHAPAT activity (ϳ0.15% of the cellular GPAT activity) could be detected from ring, trophozoite, and schizont extracts (data not shown). These results indicate that in P. falciparum the first step of glycerolipid biosynthesis is directed mostly toward the acylation of glycerol 3-phosphate substrate. As a control, yeast extracts were used, and both activities were detected (Fig. 1, A and B) in agreement with data published previously (20,28,33).
PfGatp Is a Yeast-like GPAT Protein-To identify the enzyme(s) responsible for the malarial GPAT activity, we searched for GPAT-like proteins in the P. falciparum genome data bases using known GPAT proteins as query. A gene that we named PfGAT was identified based on its homology with yeast GPAT enzymes. PfGAT was cloned from the P. falciparum 3D7 clone (The Netherlands) and was found to encode a polypeptide of 583 amino acids which exhibits ϳ28% identity and 49% similarity to S. cerevisiae Gat1p and Gat2p ( Fig. 2A) and other putative GPAT enzymes from the fission yeast Schizosaccharomyces pombe and the filamentous yeast Candida albicans (not shown). In contrast, PfGatp shares little or no homology with bacterial, mammalian, and plant known and putative GPAT proteins. The four motifs known to play a crucial role in GPAT activity are present in PfGatp and are highly homologous to those found in yeast GPAT proteins (34). Interestingly, these motifs are different from the previously characterized or predicted human, mouse, bacterial, and plant GPAT enzymes (Table II). Furthermore, whereas PfGatp and its yeast homologs Gat1p and Gat2p have a long stretch of 102, 120, and 109 amino acid residues between motifs II and III, respectively, this linker is much shorter in the human, mouse, bacterial, and plant known and putative GPAT enzymes with, respectively, 35, 35, 29, and 57 amino acid residues only (Table II). The derived amino acid sequence of PfGatp was analyzed to determine the hydrophobic character of the protein, using the TMHHM program (35). PfGatp is predicted to possess three hydrophobic membrane spanning domains with a long Nterminal stretch (1-382) exposed outside the membrane (Fig.  2B). This topology is similar to that predicted for the yeast Gat1p and Gat2p proteins (28).
PfGatp Is an Endoplasmic Reticulum Membrane Protein Expressed throughout the Intraerythrocytic Life Cycle-We have expressed and purified the C-terminal region of PfGatp and used it to immunize rabbits and produce polyclonal antibodies. These antibodies were affinity purified over a PfGatp maltosebinding protein affinity matrix and used in Western blot assays to monitor PfGatp temporal and spatial expression during the P. falciparum intraerythrocytic life cycle. Although no immunoreaction could be detected in the culture supernatant or the hemolysate fractions of uninfected or P. falciparum-infected erythrocytes, a single band with a molecular mass of 67 kDa was detected in the parasite fraction (Fig. 3A). This size is consistent with that predicted from the PfGatp translation product. Analysis of PfGatp expression during the intraerythrocytic life cycle of the parasite showed that it is expressed in all the stages (ring, trophozoite, and schizonts), but its level increases during the trophozoite stage (24 -36 h) (Fig. 3B). As a positive control, expression of the P. falciparum elongation factor 1␤ was regulated, with higher expression observed during the later stages of the parasite development (data not FIG. 3. PfGatp expression during P. falciparum intraerythrocytic life cycle. A, Western blot analysis was performed using protein extracts from supernatant (S), hemolysate (H), and parasite (P) fractions from an asynchronous culture of P. falciparum 3D7 clone. Proteins were separated by SDS-PAGE, transferred to a nitrocellulose membrane, and analyzed by immunoblot using affinity-purified PfGatp-antibodies as described under "Experimental Procedures." B, Western blot analysis was performed using protein extracts prepared from a highly synchronous culture of P. falciparum 3D7 clone at different times after P. falciparum invasion of red blood cells. C, Western blot analysis was performed using soluble (S) and pellet fractions (P) of parasite extracts treated or not with 1% Triton X-100, 1% Triton X-114, 1% n-decyl-␤-D-maltopyranoside (DM), 0.5 M potassium acetate or 0.1 M carbonate, pH 11, followed by 100,000 ϫ g ultracentrifugation. Immunoblot analysis was performed using anti-PfGatp and anti-PfNT1 antibodies.
shown), as we have reported previously (31,36). To determine whether PfGatp is an integral or peripheral membrane protein, parasite lysates prepared from P. falciparum-infected erythrocytes were treated with various detergents and salts. Membrane and soluble fractions were then separated by ultracentrifugation. In the absence of salts and detergents, PfGatp was associated with the membrane fraction (Fig. 3C). Treatment with salts or detergents had little or no effect on PfGatp cellular distribution (Fig. 3C). As a control, the nucleoside transporter, PfNT1, an integral membrane protein of the plasma membrane of the parasite (31,37), partitioned in the soluble fraction in the presence of detergents but remained associated with the membrane fraction in the presence of salts (Fig. 3C). Together, these data suggest that PfGatp is an integral membrane protein that is associated with a detergent-resistant fraction.
GPAT activities characterized in different organisms have been shown to exist in the endoplasmic reticulum, lipid particles, peroxisomes, and mitochondria (38). To examine PfGatp localization during P. falciparum intraerythrocytic development, immunofluorescence analyses were performed using affinity-purified PfGatp antibodies and specific markers of red cell membrane (Band 3) and parasite organelles (BiP and Mi-toTracker). Fluorescence signals specific for PfGatp were detected in all three intraerythrocytic developmental stages (rings, trophozoites, and schizonts) and were limited to the parasite with no PfGatp staining detected in the erythrocyte membrane or cytoplasm (Fig. 4). To localize PfGatp within the parasite further, we performed colocalization studies with the mitochondrial and endoplasmic reticulum markers, Mito-Tracker and BiP, respectively (39) (Fig. 4, B-E). The PfGatp signal was proximal and only partly overlapping with Mito-Tracker (Fig. 4B), whereas the signals of PfGatp and BiP completely overlapped in all the three stages (Fig. 4, C-E). Collectively, these data suggest that PfGatp is a component of the endoplasmic reticulum. This staining was proximal to the nucleus as revealed by Hoechst labeling (Fig. 4, A-E). Similar results were obtained using confocal microscopy (data not shown).
Yeast Complementation and PfGatp-mediated GPAT Activity-For functional analysis of PfGatp, we have used yeast as a model system to characterize the protein at the biochemical and genetic levels. In S. cerevisiae, two genes, GAT1 and GAT2, encode GPAT activities (27,28). Single disruption of GAT1 or GAT2 causes no discernible growth defects; however, disruption of both genes is lethal (27,28). To overcome expression problems caused by the high AϩT content of PfGAT, we used a PCR-based approach to synthesize a codon-optimized version of PfGAT, PfGAT CO , thus changing its AϩT composition from 73.4% to 65.5% (Fig. 5). Although immunoblot analysis revealed no expression of PfGatp from the original PfGAT gene in yeast, a major induction of PfGatp expression from PfGAT CO could be detected (Fig. 6A). To determine whether expression of PfGatp in the gat1⌬gat2⌬ mutant could replace the yeast FIG. 4. Immunofluorescence microscopy of P. falciparum-infected red blood cells using PfGatp antibodies. A, double-labeling immunofluorescence of erythrocytes infected with P. falciparum at the schizont stage of the parasite intraerythrocytic development with PfGatp-and Band 3-specific antibodies. In green, PfGatp conjugated to the FITC-conjugated goat anti-rabbit secondary antibody. In red, Band 3 conjugated to the Texas Red-conjugated anti-mouse secondary antibody. B, double-labeling immunofluorescence of erythrocytes infected with P. falciparum at the schizont stage of the parasite intraerythrocytic development with PfGatp-specific antibodies and MitoTracker. In green, PfGatp conjugated to the FITC-conjugated anti-rabbit secondary antibody. In red, MitoTracker. DNA was counterstained with Hoechst. C-E, double-labeling immunofluorescence of erythrocytes infected with P. falciparum at the ring (C), trophozoite (D), and schizont (E) stages of the intraerythrocytic development with PfGatp-and BiP-specific antibodies. In red, PfGatp conjugated to the rhodamine-conjugated anti-rabbit secondary antibody. In green, BiP conjugated to the FITC-conjugated anti-rat secondary antibody. DNA was counterstained with Hoechst (blue). Yellow represents regions of overlap beteween red and green.
Gat1p GPAT activity, PfGAT CO was expressed under the regulatory control of the ADH1 constitutive promoter (pBEVY-L ADH1::PfGAT CO LEU2 plasmid) in the yeast strain CMY228, which is deleted for both GAT1 and GAT2 and contains the plasmid pGAL1::GAT1 URA3, which harbors the yeast GAT1 gene under the regulatory control of the inducible GAL1 promoter (27). The CMY228 strain is not viable on medium containing glucose and grows only on galactose (27). CMY228 cells expressing PfGAT CO were able to grow on glucose (Fig. 6B), whereas CMY228 control cells expressing the empty vector pBEVY-L ADH1 LEU2 (40) resulted in clones that were unable to grow on glucose (Fig. 6B). Furthermore, because the endogenous pGAL1::GAT1 URA3 plasmid contains the URA3 posi-tive/negative marker, we applied a negative selection using 5-fluorotic acid to eliminate this plasmid. The strain Sc-CHO104, which harbors the plasmid pBEVY-L ADH1::Pf-GAT CO LEU2 and therefore relies solely on PfGatp expression for survival, was obtained and confirmed further for the loss of the endogenous pGAL1::GAT1 URA3 plasmid (Fig. 6C). These studies thus provide genetic evidence that PfGatp plays the same cellular function as the yeast Gat1p and Gat2p. Protein extracts from ScCHO104 strain were prepared and used to characterize further the PfGatp-mediated GPAT activity in the presence of glycerol 3-phosphate and dihydroxyacetone phosphate substrates. Similar to our results in P. falciparum, Pf-Gatp activity was specific for glycerol 3-phosphate substrate. The PfGatp GPAT activity measured at 37°C was linear during the first 6 min, after which it reached a plateau (Fig. 7A). No significant activity could be detected at 0°C (Fig. 7A). We have measured the kinetic parameters of the GPAT activity using increasing concentrations of glycerol 3-phosphate substrate. PfGatp displayed an apparent affinity (K m ) for glycerol 3-phosphate of 2.55 Ϯ 0.58 mM and a maximum velocity (V max ) of 55.1 Ϯ 7.2 nmol ϫ mg Ϫ1 ϫ min Ϫ1 (Fig. 7B). The substrate specificity of PfGatp was measured using unsaturated and saturated fatty acyl-CoA substrates with different chain lengths (C12:0, C14:0, C16:0, C16:1, C18:0, C18:1, and C20:0). PfGatp displayed a major preference for palmitoyl-CoA (C16:0) and palmitoleoyl-CoA (C16:1), low preference for C14:0, C18:0, C18:1, and C12:0 substrates, and no specificity for C20:0 (Fig.  7C). Unlike its GPAT activity, PfGatp-mediated DHAPAT was found to be very low and represented only 0.5-2.5% of its GPAT activity (Fig. 7D).
PfGatp Exists as a Large Multimeric Protein Complex in the Endoplasmic Reticulum Membrane-To examine whether the native PfGatp exists as a monomer or is part of protein complex, native proteins were separated under native conditions and analyzed by Western blot using anti-PfGatp-specific antibodies. Native PfGatp migrated as a high molecular mass polypeptide of an estimated Ͼ450 kDa (Fig. 8A). Cross-linking studies using increasing concentrations of the alkylating agent ethylene glycol bis(succinimidylsuccinate) followed by SDS-PAGE analysis showed a shift of PfGatp from a monomeric form to higher molecular masses, some of which could not enter the stacking gel (data not shown). To gain further insight into the approximate size of the native PfGatp, a gel filtration analysis of the native enzyme was performed and revealed the presence of PfGatp at the peak of migration of thyroglobulin (650 kDa) (Fig. 8B). As a control, the fractions collected by gel filtration were analyzed by Western blot using antibodies against P. falciparum elongation factor 1␣ and showed, as expected, the presence of this protein both as a monomer (free PfEF-1␣) and as a large protein complex (EF-1 complex) as described previously (Fig. 8, C and D) (36). DISCUSSION During its asexual 48-h development and multiplication cycle within human erythrocytes P. falciparum produces between 16 and 32 new merozoites that subsequently invade new red blood cells. This rapid multiplication of the parasite within human erythrocytes is accompanied by a marked increase in phosphatidylcholine, phosphatidylethanolamine, phosphatidylserine, phosphatidylinositol, fatty acids, DAG, and TAG content (3). This increased metabolic need of P. falciparum to generate new membranes has stimulated efforts to identify compounds that can interfere with parasite membrane biogenesis and block malaria proliferation. Analysis of the available P. falciparum genomic sequences points to the presence of the genes of the major pathways for synthesis of glycerolipids (12). The few of those genes that have been characterized thus far show major structural and catalytic differences from their human counterparts, thus opening future avenues for lipid-based therapeutic strategies to fight malaria.
Here, we have characterized the initial step of glycerolipid synthesis in Plasmodium-infected erythrocytes. We found that P. falciparum catalyzes the acylation of glycerol 3-phosphate into 1-acylglycerol 3-phosphate, which is the main precursor for phosphatidic acid and subsequently for the phospholipid precursors DAG and CDP-DAG. P. falciparum also catalyzes the acylation of dihydroxyacetone phosphate, although less efficiently compared with the GPAT substrate. This low DHAPAT activity suggests that malaria parasites may not require specialized DHAPAT enzymes and may not synthesize ether lipids. This idea is supported further by the lack in the finished genomic sequence of P. falciparum of homologs of . Cell extract preparation followed by Ni 2ϩ affinity chromatography was performed as described under "Experimental Procedures." Immunoblot analyses were performed using anti-PfGatp affinity-purified antibodies (1:100) and anti-V5 antibodies (1:5,000). B, PfGAT CO complementation of the conditional lethality of S. cerevisiae CMY228 grown on glucose. CMY228 cells harboring pBEVY-L ADH1 LEU2 or pBEVY-L ADH1::PfGAT CO LEU2 vector were grown to mid-log phase in galactose-containing minimal medium, washed twice in ice-cold water, and plated on YPD and YPG plates. Identical numbers of cells were serial 1:10 diluted and applied (starting with 3 ϫ 10 5 cells). C, PCR analysis using primers specific for S. cerevisiae GAT1 and P. falciparum PfGAT CO  Our studies revealed that the P. falciparum PfGAT gene encodes a GPAT enzyme. To our knowledge, this is the first GPAT gene to be identified in protozoa. Affinity-purified polyclonal antibodies against PfGatp indicated that this protein is expressed throughout the asexual life cycle of the parasite but induced mainly during the trophozite stage during which an active synthesis of phospholipids takes place, likely to provide membranes for the newly formed parasites. A similar regulation pattern was observed for the PfGAT transcript using large scale microarray analyses (41,42). Our characterization of the native PfGatp demonstrated that it is an integral membrane protein of the endoplasmic reticulum and suggests that this organelle plays an important role in phospholipid biosynthesis in P. falciparum. Furthermore, we found that native PfGatp exists as a high molecular mass protein. We do not know at this stage whether this high molecular complex is composed solely of PfGatp or whether this enzyme associates with other parasite proteins.
Analysis of the sequence of PfGatp protein suggests that it is a yeast-like GPAT enzyme. The four motifs known to be important for GPAT catalysis are present in PfGatp and are highly similar in residue composition as well as in their spatial distribution to those of the yeast GPAT proteins, Gat1p and Gat2p. Interestingly, these motifs are highly divergent from those of mammalian and bacterial GPAT enzymes. Motifs II and III in PfGatp are separated by 102 amino acid residues, whereas the human and mouse GPATs have only 35 residues between these two motifs. The fact that yeast possesses two genes GAT1 and GAT2 that catalyze the GPAT activity and that disruption of both genes is lethal has made it possible for us to functionally characterize PfGAT at the genetic and the biochemical levels using yeast as a surrogate system. The finished sequence of the P. falciparum nuclear genome indicated that its overall AϩT composition is 80.6% and rises to ϳ90% in FIG. 7. Biochemical characterization of PfGatp activity. A, PfGatp time-dependent GPAT activity. ScCHO104 (gat1⌬gat2⌬ϩ(pBEVY-L ADH1::PfGAT CO LEU2)) strain was grown on glucose-based rich medium (YPD) and harvested at A 600 ϭ 1.5. Protein extracts were prepared and assayed for GPAT activity as described under "Experimental Procedures." The assay was performed using 200 g of proteins, 0.4 mM glycerol 3-phosphate, C16:1-CoA, and incubated for the indicated times at 37°C (black circles) or 0°C (white circles). Each point represents an average of duplicate experiments Ϯ S.D. B, PfGatp-mediated GPAT kinetics. PfGatp activity was measured at 37°C for 2.5 min in the presence of C16:1-CoA and the indicated concentrations of glycerol 3-phosphate. The curve was fitted to the Michaelis-Menten equation V ϭ V max ϫ S/(K m ϩ S). The Lineweaver-Burk representation of the saturation curve is shown as an inset. C, PfGatp substrate specificity. The GPAT activity mediated by PfGatp was assayed as described under "Experimental Procedures" using fatty acid substrates with different chain lengths and incubated for 10 min at 37°C. D, PfGatp DHAPAT activity. The DHAPAT assay was performed as described under "Experimental Procedures" for 10 min at 37°C using extracts derived from wild-type (WT; ScCHO1) and ScCHO104 (gat1⌬gat2⌬ PfGatp) strains and C16:1-CoA as an acyl donor.
introns and intergenic regions, making it the most AϩT-rich genome sequenced to date (12). This unusual property of P. falciparum genes has hampered efforts to perform straightforward expression and complementation analyses in heterologous systems. Because of low expression levels in yeast, initial attempts to express PfGAT gene in gat1⌬ or gat2⌬ single knock-outs to measure activity or in the double knock-out gat1⌬gat2⌬ to complement its lethal phenotype were not successful. To overcome this problem, we synthesized a codonoptimized version, PfGAT CO . This resulted in an increase in the GϩC composition of PfGAT from 26.6 to 34.5% and a dramatic increase in the expression of PfGatp in yeast. Furthermore, expression of PfGAT CO in the double knock-out gat1⌬gat2⌬ mutant could complement its lethal phenotype. In accordance with our results in P. falciparum, PfGatp was found to be specific for glycerol 3-phosphate and showed a very low activity toward dihydroxyacetone phosphate substrate. Kinetic studies revealed that PfGatp is a low affinity GPAT enzyme that displays high substrate specificity. PfGatp activity was higher when the acyl-CoA substrates, C16:0 and C16:1, were used. Interestingly, C16:0 has been shown to be transported by the parasite from host plasma and is essential for P. falciparum growth and survival (43).
In summary, the work reported here supports the conclusion that PfGAT encodes an unusual yeast-like GPAT enzyme of P. falciparum expressed throughout the asexual life cycle (ring, trophozoite, and schizont stages) of the parasite within the host red blood cells. The identification of PfGatp is a critical step toward understanding membrane biogenesis in this parasite. The finished genome sequence of P. falciparum has also revealed a second gene, PfPLSB, encoding a polypeptide that shares homology with plant GPAT enzymes. Future studies are needed to determine whether the encoded protein catalyzes the acylation of glycerol 3-phosphate and/or dihydroxyacetone phosphate. Our attempts to target PfGAT gene for disruption have resulted in integration events to different loci in the genome but not to PfGAT locus, suggesting that PfGAT might be essential for parasite survival and that PfGatp and PfPlsB functions might not be redundant. Future complementation studies in P. falciparum to confirm the essential role of PfGAT are warranted and could provide useful information for the rational design of compounds that could specifically inhibit PfGatp activity and block parasite membrane biogenesis and multiplication.