In Vivo Evidence for the Specificity of Plasmodium falciparum Phosphoethanolamine Methyltransferase and Its Coupling to the Kennedy Pathway*

Unlike humans and yeast, Plasmodium falciparum, the agent of the most severe form of human malaria, utilizes host serine as a precursor for the synthesis of phosphatidylcholine via a plant-like pathway involving phosphoethanolamine methylation. The monopartite phosphoethanolamine methyltransferase, Pfpmt, plays an important role in the biosynthetic pathway of this major phospholipid by providing the precursor phosphocholine via a three-step S-adenosyl-l-methionine-dependent methylation of phosphoethanolamine. In vitro studies showed that Pfpmt has strong specificity for phosphoethanolamine. However, the in vivo substrate (phosphoethanolamine or phosphatidylethanolamine) is not yet known. We used yeast as a surrogate system to express Pfpmt and provide genetic and biochemical evidence demonstrating the specificity of Pfpmt for phosphoethanolamine in vivo. Wild-type yeast cells, which inherently lack phosphoethanolamine methylation, acquire this activity as a result of expression of Pfpmt. The Pfpmt restores the ability of a yeast mutant pem1Δpem2Δ lacking the phosphatidylethanolamine methyltransferase genes to grow in the absence of choline. Lipid analysis of the Pfpmt-complemented pem1Δpem2Δ strain demonstrates the synthesis of phosphatidylcholine but not the intermediates of phosphatidylethanolamine transmethylation. Complementation of the pem1Δpem2Δ mutant relies on specific methylation of phosphoethanolamine but not phosphatidylethanolamine. Interestingly, a mutation in the yeast choline-phosphate cytidylyltransferase gene abrogates the complementation by Pfpmt thus demonstrating that Pfpmt activity is directly coupled to the Kennedy pathway for the de novo synthesis of phosphatidylcholine.

With more than 1 million deaths and 300 million clinical cases annually, malaria is a major worldwide health concern and a major economic burden to the developing world (1). The illness is caused by intraerythrocytic protozoan parasites of the genus Plasmodium. Plasmodium falciparum, which is responsible for the most severe form of the disease, has developed resistance to almost all the available drugs in the antimalarial armamentarium (2). New chemotherapeutic strategies are therefore urgently needed to combat this disease. One strategy is to target the ability of the parasite to synthesize new membranes and inhibit its replication within host erythrocytes. Lipid inhibitors such as choline and phosphocholine analogs have been shown to inhibit P. falciparum membrane biogenesis and to block parasite proliferation (3)(4)(5)(6)(7). Some of these compounds are currently being evaluated for treatment of drugresistant malaria (5).
Biochemical studies have indicated that the synthesis of phosphatidylcholine (PtdCho), 1 the major phospholipid in P. falciparum membranes can occur via two metabolic pathways, the de novo choline 3 phosphocholine 3 CDP-choline pathway, also named the Kennedy pathway, and a newly identified route that we named the serine decarboxylation-phosphoethanolamine methylation (SDPM) pathway (6,8). The Kennedy pathway initiates with the transport of choline from host serum into the parasite cytoplasm by unidentified erythrocyte and P. falciparum choline transporters. Choline is then converted into PtdCho by the sequential action of choline kinase, choline-phosphate cytidylyltransferase, and CDP-diacylglycerol-cholinephosphotransferase (9 -12). The SDPM pathway initiates from serine that is either transported from the host or obtained from active degradation of host proteins (13,14). Serine is first decarboxylated to form ethanolamine by a serine decarboxylase and then phosphorylated to produce phosphoethanolamine (P-Etn) by an ethanolamine kinase (8). P-Etn is subsequently methylated to generate phosphocholine (P-Cho) by a phosphoethanolamine methyltransferase (Pfpmt) (6). The P-Cho enters the Kennedy pathway to form PtdCho. The SDPM pathway has thus far been characterized only in P. falciparum and plants (6,(15)(16)(17), and available genomic data suggest that it might also take place in Caenorhabditis elegans, Caenorhabditis briggsae, and Anopheles gambiae (6).
Pfpmt is expressed throughout the intraerythrocytic cycle as well as during the gametocyte and sporozoite stages of the parasite (6,18,19). Pfpmt does not share homology with phosphatidylethanolamine (PtdEtn) methyltransferases from lower and higher eukaryotes, and no other homologs of this protein could be found in human or other mammalian data bases (6). Unlike the bipartite structure of plant phosphoethanolamine methyltransferases with two AdoMet-dependent catalytic domains, the malarial Pfpmt is only half the size of plant phosphoethanolamine methyltransferases and possesses a single catalytic domain solely responsible for the three-step AdoMetdependent methylation of P-Etn into P-Cho (6). Although biochemical studies in vitro showed that Pfpmt and plant phosphoethanolamine methyltransferases were specific for P-Etn and did not catalyze the methylation of PtdEtn (6), one cannot exclude that these phosphoethanolamine methyltransferases (PMT) also catalyze the methylation of PtdEtn in vivo. The difficulty in genetically manipulating P. falciparum, and the fact that yeast cells lack the ability to convert P-Etn into P-Cho make yeast an attractive system to assess the specificity of Pfpmt in vivo. In yeast grown in the absence of choline, the synthesis of PtdCho occurs primarily via the transmethylation of PtdEtn by two methyltransferases encoded by the PEM1 (CHO2) and PEM2 (OPI3) genes (20 -22). Individual deletions of PEM1 or PEM2 cause no discernable phenotypes, whereas disruption of both genes is lethal unless choline is provided exogenously and transported via the choline transporter Hnm1 (20,21). Here we have taken advantage of the extensive biochemical and genetic knowledge of phospholipid metabolism in yeast to examine the in vivo substrate specificity of Pfpmt. Our data provide biochemical and genetic evidence demonstrating that in vivo Pfpmt does not catalyze the transmethylation of PtdEtn to form PtdCho but instead catalyzes the methylation of P-Etn into P-Cho. Furthermore, we show that Pfpmt plays an important role in PtdCho biosynthesis by coupling its activity to the Kennedy pathway.

MATERIALS AND METHODS
Strain Construction, Growth Conditions, and Media-The Saccharomyces cerevisiae strains used in this study are described in Table I. Standard methods for yeast culture and manipulation were used (23). Yeast was cultivated at 30°C in YPD (yeast extract/peptone/dextrose) (24) or in synthetic minimal media containing 2% glucose (SD medium) or 2% galactose (SG medium) supplemented with histidine (30 g/ml), uracil (30 g/ml), leucine (100 g/ml), methionine (100 g/ml), or choline (140 g/ml) as required to maintain cell growth.
Plasmid Construction-The codon-optimized PfPMT CO was synthesized using the forward and reverse primers shown in Fig. 2. PfPMT CO was first assembled and amplified as four 200-bp long fragments that were subsequently used as templates to amplify the whole gene. Assembly reactions were performed as described by Santiago et al. (25). 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 a 3-min final elongation at 68°C. The resulting products were purified and used as templates for amplification using the following PCR program: 2 min at 94°C, 20 cycles of 30 s at 94°C, 30 s at 63°C, and 45 s at 68°C and a 2-min final elongation at 68°C. The full-length PfPMT CO was TAcloned into the vector PCR2.1 (Invitrogen). Using the primers 5Ј-PfP-MT CO 5Ј-ATGACTTTGATCGAGAACTTGAACTCTGATAAGACCTTC-CTGGAGAA-3Ј and 3Ј-PfPMT CO 5Ј-GTTCTTGGTGGCCTTGAAGTAA-C-3Ј, PfPMT CO was amplified and cloned into the vector pYES2.1-/V5-His-TOPO (Invitrogen), obtaining pYES2.1 GAL1::PfPMT CO . The sequence of PfPMT CO was verified by DNA sequencing.
Phospholipid Analysis-Wild-type-pYES2.1 and pem1⌬pem2⌬-PfP-MT CO cells were grown overnight in SG medium supplemented with 2 mM ethanolamine and in the absence of choline. The cells were diluted to A 600 ϭ 0.005 in 75 ml of fresh SG medium and grown to A 600 ϳ1.5. Cells were harvested and washed twice with water. Lipids were extracted by the Bligh and Dyer method (26) and separated on a twodimensional thin layer chromatography (TLC). For the first dimension a solvent containing chloroform/methanol/ammonium hydroxide (84.5: 45.5:6:5) was used. The second solvent system was composed of chloroform/glacial acetic acid/methanol/water (90:30:6:2.6). Lipids were stained with iodine vapor. Each lipid spot was excised from the plate and quantified by measuring phosphorus (27). The results are shown as the percentage of total lipid phosphorus in each phospholipid fraction.
Purification of Pfpmt in Yeast and Protein Immunoblotting-S. cerevisiae wild-type and pem1⌬pem2⌬ cells harboring pYES2.1 and pYES2.1 GAL1::PfPMT CO were grown in 50 ml of SD or SG media to A 600 ϭ 1 at 30°C. Extract preparation and purification of the His 6 -V5tagged Pfpmt by affinity chromatography were performed as described previously (25). Western blot analysis was performed using affinitypurified proteins from cells grown in minimal medium glucose or galactose. Proteins were separated by SDS-PAGE, transferred to a nitrocellulose membrane, and analyzed by immunoblot using anti-V5 primary (Invitrogen, 1:5000 dilution) and anti-mouse secondary (1:10 000 dilution) antibodies.
Enzyme Assays-Yeast cells were grown in 50 ml of SD or SG media to A 600 ϭ 1 at 30°C and lysed in 50 mM Tris-HCl buffer, pH 8.0, 10% glycerol, 1 mM dithiothreitol, and protease inhibitor mix by agitation with acid-washed glass beads. The 10,000 ϫ g supernatant was dialyzed against 5 mM Hepes, pH 7.8, and 0.5 mM dithiothreitol. Pfpmt was then affinity-purified as recommended by the manufacturer. PMT activity was determined by measuring the incorporation of radioactivity from [methyl-14 C]AdoMet into P-Etn as described previously (6). The incubation mixture in a final volume of 100 l, contained 100 mM Hepes, pH 8.6, 2 mM EDTA, 10% glycerol, 100 M P-Etn, 100 M AdoMet (100 nCi of [methyl-14 C]AdoMet), and 50 g of crude extract or affinity-purified recombinant enzyme. The reaction was incubated for 30 min at 30°C and terminated by the addition of 1 ml of ice-cold H 2 O. The product was purified through a AG (Hϩ) resin, and the identity of the reaction product was confirmed by TLC using [ 14 C]P-Cho (55 mCi/ mmol; ARC) as a standard. Pfpmt activity was also measured using 100 M ethanolamine and 100 m PtdEtn as described previously (6). For determination of PtdEtn methyltransferase activity, a PtdEtn emulsion was prepared by sonication in the presence of 0.02% Triton X-100 as described previously (16,28), and the reaction was performed in Tris buffer, pH 8.5, using 50 g of crude extract.

Wild-type Yeast Cells
Expressing the Codon-optimized PfP-MT CO Acquire Phosphoethanolamine Methyltransferase Activity-The inability of the yeast S. cerevisiae to catalyze the three-step methylation of P-Etn for synthesis of P-Cho ( Fig.  1) made it possible to assess biochemically and genetically the substrate specificity of P. falciparum Pfpmt in vivo. Because of the high AϩT content of PfPMT (72.5% AϩT-rich) and to allow its successful expression in yeast, the full codon sequence of the cDNA was reconstructed in vitro to augment its GϩC content. Thirty-two 50-nucleotide overlapping primers were first assembled, and the resulting assembly product was used as a template for PCR amplification to create a codon-optimized gene, PfPMT CO , with 51.3% AϩT content (Fig. 2). PfPMT CO was then cloned downstream of the GAL1inducible promoter in the yeast expression vector pYES2.1, which allowed addition of a C-terminal V5-His 6 epitope tag to monitor protein expression. As shown in Fig. 3A, expression of Pfpmt could be detected in wild-type yeast cells grown in the presence of galactose, which induces GAL1 promoterregulated genes but not in the presence of glucose, which represses GAL1 promoter-regulated expression.
To determine whether the expression of Pfpmt can provide yeast with the ability to catalyze the three-step methylation of P-Etn into P-Cho, protein extracts were prepared from wildtype cells expressing either GAL1-PfPMT CO or, as a control, the empty vector under galactose and glucose conditions and incubated with the substrate P-Etn and the methyl donor [ 14 C]AdoMet. The products of the reactions were purified by ion exchange chromatography, analyzed by thin layer chromatography, and quantified. Wild-type cells expressing Pfpmt grown in the presence of galactose were able to catalyze the three-step AdoMet-dependent methylation of P-Etn to form P-Cho, whereas the same cells grown in presence of glucose or wild-type cells expressing the empty vector failed to catalyze this reaction (Fig. 3B). Pfpmt was further purified from wildtype cells grown in the presence of galactose by nickel affinity chromatography, and its activity and substrate specificity were analyzed. Similar to the PMT activity measured in P. falciparum extracts or that of the Escherichia coli recombinant-purified enzyme (6), Pfpmt purified from yeast catalyzed the threestep AdoMet-dependent methylation of P-Etn to form P-Cho (Fig. 4, A and B). The enzyme preparation showed only background activity when ethanolamine or PtdEtn were used as substrates (Fig. 4C).
PfPMT CO Suppresses the Choline Auxotrophy of a Yeast pem1⌬pem2⌬ Mutant-Yeast cells lack PMT activity. However, they express two genes PEM1 and PEM2, which encode two enzymes required for the three-step transmethylation of PtdEtn into PtdCho. Deletion of either PEM1 or PEM2 produces viable strains, whereas deletion of both genes is lethal unless external choline is provided (20,21). To determine whether expression of Pfpmt can complement the loss of PtdEtn transmethylation, a yeast strain, pem1⌬pem2⌬, lacking PEM1 and PEM2 was constructed and transformed with the GAL1-PfPMT CO plasmid or the empty vector. The resulting transformants were confirmed to express Pfpmt by immunoblotting using anti-V5 monoclonal antibodies (Fig. 5A) and were further tested for their ability to grow in the absence of choline under inducing (in presence of galactose) and repressing (in presence of glucose) conditions (Fig. 5B). Because of residual choline in agar containing media, the assays were performed in both solid and liquid media. As expected pem1⌬pem2⌬ cells expressing the empty vector required choline for growth under both glucose and galactose conditions. In contrast, pem1⌬pem2⌬ cells expressing GAL1-PfPMT CO were able to grow in the absence of choline when galactose was the sole carbon source but required choline when grown in the presence of glucose (Fig. 5B). These data demonstrate that Pfpmt suppresses the choline auxotrophy of the pem1⌬pem2⌬ mutant.
PtdCho Biosynthesis by Pfpmt Is Independent of PtdEtn Transmethylation-To confirm that Pfpmt suppression of the choline auxotrophy of pem1⌬pem2⌬ is via restoration of the synthesis of PtdCho, phospholipids were prepared from wildtype and pem1⌬pem2⌬-GAL1-PfPMT CO cells, grown in absence of choline, analyzed by TLC, and quantified. PtdCho could be detected in both strains (Fig. 6A), although the level of this phospholipid in the pem1⌬pem2⌬-GAL1-PfPMT CO strain constituted only 32% of what is synthesized by wild-type cells (Fig.  6B). In addition, the PtdEtn level in the pem1⌬pem2⌬-GAL1-PfPMT CO strain was almost double that of the wild-type cells (Fig. 6), suggesting continuous accumulation of this phospholipid as a result of synthesis via phosphatidylserine decarboxylation and the CDP-ethanolamine pathway. Interestingly, significant amounts of phosphatidyldimethylethanolamine (PtdEtn(Me) 2 ) could be detected in wild-type cells but no intermediates of PtdEtn transmethylation could be detected in the pem1⌬pem2⌬-GAL1-PfPMT CO cells (Fig. 6). Together these results suggest that Pfpmt-mediated synthesis of PtdCho is not via methylation of PtdEtn but rather via the methylation of P-Etn. To further validate this interpretation we assessed the specificity of protein extracts prepared from wild-type, pem1⌬pem2⌬, and pem1⌬pem2⌬-GAL1-PfPMT CO cells for P-Etn versus PtdEtn substrates. Protein extracts from wild-type cells catalyzed the transmethylation of PtdEtn into PtdCho, but showed no specificity toward P-Etn substrate (Fig. 7A). On the other hand, protein extracts from pem1⌬pem2⌬-GAL1-Pf-PMT CO cells catalyzed the three-step methylation of P-Etn into P-Cho but showed no specificity for PtdEtn substrate (Fig. 7). As expected protein extracts from pem1⌬pem2⌬ grown in the presence of choline lacked both P-Etn and PtdEtn methylation activities (Fig. 7).
PfPMT CO 3. Pfpmt expression in S. cerevisiae. A, Western blot analysis was performed using crude extracts from BY4741-pYES2.1 (wild-type, WT) and BY4741-GAL1-PfPMT CO (WTϩPfpmt) grown at 30°C in minimal medium supplemented with 1 mM choline and containing glucose (Glu) or galactose (Gal) as described under "Materials and Methods." B, S. cerevisiae was grown in SD (Glu) or SG (Gal) media, and the PMT activity was measured in absence (WT) or presence (WTϩPfpmt) of the malarial Pfpmt. PMT activity in S. cerevisiae extracts was determined using phosphoethanolamine (P-Etn) and [ 14 C]AdoMet as substrates. The activity was measured at 30°C in the presence of 50 g of S. cerevisiae protein extract, and the product phosphocholine (P-Cho) was purified using AG-50(H ϩ ) ion exchange resin. Each value is the mean of duplicate experiments.

FIG. 4. Substrate specificity of Pfpmt in S. cerevisiae.
A, PMT activity was determined after affinity purification of Pfpmt from an extract of pem1⌬ pem2⌬-GAL1-PfPMT CO cells grown at 30°C in SD (Glu) or SG (Gal) media. B, TLC analysis of the reaction product P-Cho after purification using AG-50(H ϩ ) ion exchange resin. Lane S1, standard [ 14 C]P-Cho. C, substrate specificity of affinity purified Pfpmt, in the presence of 100 M P-Etn, Etn, and PtdEtn as described under "Materials and Methods." Yeast cells were grown at 30°C in the presence of glucose or galactose. Each datum represents an average of a duplicate.

FIG. 5. Functional complementation of Pfpmt in S. cerevisiae.
A, Western blot analysis was performed using crude extracts from pem1⌬ pem2⌬-pYES2.1 (pem⌬) and pem1⌬ pem2⌬-GAL1-PfPMT CO (pem⌬ Pfpmt) grown at 30°C in minimal medium supplemented with 1 mM choline and containing glucose or galactose as described under "Materials and Methods." B and C, the strains BY4741-pYES2.1 (wild type, 1), pem1⌬ pem2⌬-pYES2.1 (2), and pem1⌬ pem2⌬-GAL1-PfP-MT CO (3) were grown in solid (B) or liquid (C) minimal medium galactose (Gal) or glucose (Glu) with (ϩCho) or without (ϪCho) 1 mM choline for 24 h. The growth measurements were performed in both agar medium (left panels) and liquid medium (right panels). tion activity and altered in the de novo synthesis of PtdCho via the Kennedy pathway (29). This mutant harbors, in addition to a genetic deletion of the PEM1 gene, a spontaneous thermosensitive mutation in the PCT1 gene encoding choline-phosphate cytidylyltransferase, the second step in the CDP-choline pathway, thus resulting in a growth defect at 37°C. Expression of Pfpmt did not complement the thermosensitivity of the pem1⌬pct1 ts (Fig. 8). As a control the pem1⌬pem2⌬ strain harboring Pfpmt grew normally in the absence of choline at 30 (Fig. 5B) and 37°C (not shown), whereas the pem1⌬pem2⌬ strain harboring an empty vector required choline for growth at both temperatures (not shown). Together, these results suggest that Pfpmt function is directly coupled to the Kennedy pathway, further demonstrating the specificity of this enzyme for P-Etn in vivo.

DISCUSSION
Membrane biogenesis is essential for P. falciparum development and multiplication with human erythrocytes. Understanding this process is an important step toward designing better therapeutic strategies to block parasite replication and eliminate the pathological symptoms associated with the intraerythrocytic cycle of the parasite. Upon infection of human erythrocytes, the phospholipid content of P. falciparum increases by at least 5-6-fold (30). Available data suggest that the synthesis of PtdCho, the major phospholipid in P. falciparum membranes, can occur via two metabolic pathways: the Kennedy pathway that proceeds from choline and a newly identified route that we named the SDPM pathway that proceeds from serine. The Pfpmt enzyme of P. falciparum plays an important role in the SDPM pathway by providing P-Cho via methylation of serine-derived P-Etn (6). The resultant P-Cho then enters the Kennedy pathway for the synthesis of PtdCho. Pfpmt is expressed throughout the intraerythrocytic cycle as well as during the gametocyte and sporozoite stages of the parasite, and possesses a single catalytic domain solely responsible for the three-step AdoMet-dependent methylation of P-Etn to form P-Cho (6).
Because no homologs of PtdEtn methyltransferases could be found in the finished genome sequence of P. falciparum, it remained unclear whether in vivo Pfpmt might also catalyze the three-step methylation of PtdEtn. In the present study we provide genetic and biochemical evidence demonstrating that the primary function of Pfpmt in vivo is to catalyze the methylation of P-Etn to form P-Cho. Six lines of evidence support this important conclusion. First, expression of Pfpmt in wild-type S. cerevisiae cells, which inherently lack P-Etn methyltransferase activity, resulted in their acquiring this activity. Second, PMT activity of wild-type yeast expressing Pfpmt was specific for P-Etn, and no methylation of ethanolamine or PtdEtn could be detected. Third, expression of Pfpmt in a yeast mutant, pem1⌬pem2⌬, lacking the two genes PEM1 and PEM2, which encode the enzymes Pem1 and Pem2 essential for the synthesis of PtdCho from PtdEtn, resulted in a complete suppression of their requirement for exogenous choline. Fourth, lipid analysis of the pem1⌬pem2⌬ cells expressing Pfpmt demonstrated the synthesis of PtdCho in the absence of choline, but unlike the wild-type strain, no intermediates of the transmethylation of PtdEtn could be detected in the complemented strain. Fifth, the PtdEtn level in the pem1⌬pem2⌬ strain expressing Pfpmt was almost double that of the wild-type cells, suggesting continuous accumulation of this phospholipid as a result of synthesis via phosphatidylserine decarboxylation and the CDP-ethanolamine pathway in conjunction with the elimination of PtdEtn methylation. Finally, complementation of pem1⌬pem2⌬ choline auxotrophy by Pfpmt required a functional Kennedy pathway, because alteration of this pathway by mutation in the PCT1 gene abolishes this complementation.
The ability of Pfpmt to complement pem1⌬pem2⌬ mutant depends on the availability of the P-Etn substrate. In fact, the growth of the complemented strain was much better when the medium was supplemented with ethanolamine (not shown). In the absence of exogenous ethanolamine, this precursor could be obtained either via degradation of sphingosine-1-phosphate by the lyase encoded by the DPL1 gene, or via hydrolysis of Pt-dEtn by the Ca 2ϩ -dependent phospholipase D activity. Interestingly, although the complemented and wild-type strains grew equally well in the absence of choline, the level of PtdCho (as percent total phospholipid) in the complemented strain was only ϳ4% in the absence (not shown) and ϳ12% in the presence (Fig. 6) of ethanolamine. In contrast PtdCho was ϳ40% total phospholipid present in wild-type cells. These data suggest that although PtdCho is the major phospholipid in yeast membranes, its level in wild-type cells far exceeds what is needed for growth and survival, at least under laboratory conditions.
To enhance expression of Pfpmt in yeast, we completely redesigned its cDNA by reducing its AϩT content from 72.5 to 51.3% to prevent early transcriptional termination. Successful expression of Pfpmt was demonstrated by inserting a V5-His 6 tag in the C-terminal region of the encoded protein and monitoring expression levels using anti-V5 monoclonal antibodies. FIG. 7. Methyltransferase activity in S. cerevisiae. A, PtdEtn methyltransferase (MT) activity was determined at 30°C in 50 g of protein present in crude extracts from BY4741-pYES2.1 (wild type, 1), pem1⌬ pem2⌬-pYES2.1 (2), and pem1⌬ pem2⌬-GAL1-PfPMT CO (3) using PtdEtn and AdoMet as substrates as described under "Materials and Methods." Each datum represents an average of a duplicate. B, PMT activity in 50 g of protein present in crude extracts from BY4741-pYES2.1 (wild-type, 1), pem1⌬ pem2⌬-pYES2.1 (2), and pem1⌬ pem2⌬-GAL1-PfPMT CO (3) using P-Etn and AdoMet as substrates. Each datum represents an average of a duplicate. Pfpmt was further purified from yeast cells by affinity chromatography and used in enzymatic assays. Similar success in complementing yeast after codon optimization was achieved with PfGAT gene of P. falciparum, which encodes a glycerol-3phosphate acyltransferase of the endoplasmic reticulum important for the initial step of malarial glycerolipid metabolism (25). Pfpmt purified from yeast exhibited the same biochemical properties as the native PMT activity obtained from P. falciparum or recombinant Pfpmt purified from E. coli (6).
Our present results are concordant with previous biochemical analysis using recombinant Pfpmt purified from E. coli and provide strong evidence that Pfpmt exhibits high specificity toward the P-Etn substrate and does not catalyze the methylation of PtdEtn in vivo. Thus it seems that the most critical precursor for the synthesis of PtdCho in P. falciparum is P-Cho, which can be obtained either via Pfpmt methylation of serinederived P-Etn or by phosphorylation of choline transported by the parasite choline transporter from the host. The relative contribution of the two pathways to the total cellular pool of PtdCho is not yet clear. Recent studies have provided new information about the biochemical properties of the parasite choline transporter (31,32). However, the genes encoding choline transport activity have not yet been identified in Plasmodium or any other protozoan parasite. More detailed genetic studies in P. falciparum are now needed to evaluate the importance of choline transport and Pfpmt in phospholipid synthesis as it relates to parasite development and survival.
In summary, we have shown that Pfpmt is a highly specific methyltransferase enzyme acting exclusively on P-Etn substrate in vivo. The essential dependence of the pem1⌬pem2⌬ mutant on Pfpmt for growth and survival makes this strain an ideal system to screen for inhibitors that can specifically inhibit Pfpmt activity.