A Nondiscriminating Glutamyl-tRNA Synthetase in the Plasmodium Apicoplast

Background: Plasmodium apicoplast protein synthesis is essential, but few apicoplast tRNA synthetases have been characterized. Results: Apicoplast glutamyl-tRNA synthetase aminoacylates tRNAGlu and tRNAGln, is sensitive to a bacterial inhibitor, and is essential in blood stages. Conclusion: Formation of apicoplast Gln-tRNAGln is via indirect aminoacylation. Significance: We demonstrate that the apicoplast glutamyl-tRNA synthetase is a potential drug target. The malaria parasite Plasmodium falciparum and related organisms possess a relict plastid known as the apicoplast. Apicoplast protein synthesis is a validated drug target in malaria because antibiotics that inhibit translation in prokaryotes also inhibit apicoplast protein synthesis and are sometimes used for malaria prophylaxis or treatment. We identified components of an indirect aminoacylation pathway for Gln-tRNAGln biosynthesis in Plasmodium that we hypothesized would be essential for apicoplast protein synthesis. Here, we report our characterization of the first enzyme in this pathway, the apicoplast glutamyl-tRNA synthetase (GluRS). We expressed the recombinant P. falciparum enzyme in Escherichia coli, showed that it is nondiscriminating because it glutamylates both apicoplast tRNAGlu and tRNAGln, determined its kinetic parameters, and demonstrated its inhibition by a known bacterial GluRS inhibitor. We also localized the Plasmodium berghei ortholog to the apicoplast in blood stage parasites but could not delete the PbGluRS gene. These data show that Gln-tRNAGln biosynthesis in the Plasmodium apicoplast proceeds via an essential indirect aminoacylation pathway that is reminiscent of bacteria and plastids.

The malaria parasite Plasmodium falciparum and related organisms possess a relict plastid known as the apicoplast. Apicoplast protein synthesis is a validated drug target in malaria because antibiotics that inhibit translation in prokaryotes also inhibit apicoplast protein synthesis and are sometimes used for malaria prophylaxis or treatment. We identified components of an indirect aminoacylation pathway for Gln-tRNA Gln biosynthesis in Plasmodium that we hypothesized would be essential for apicoplast protein synthesis. Here, we report our characterization of the first enzyme in this pathway, the apicoplast glutamyl-tRNA synthetase (GluRS). We expressed the recombinant P. falciparum enzyme in Escherichia coli, showed that it is nondiscriminating because it glutamylates both apicoplast tRNA Glu and tRNA Gln , determined its kinetic parameters, and demonstrated its inhibition by a known bacterial GluRS inhibitor. We also localized the Plasmodium berghei ortholog to the apicoplast in blood stage parasites but could not delete the PbGluRS gene. These data show that Gln-tRNA Gln biosynthesis in the Plasmodium apicoplast proceeds via an essential indirect aminoacylation pathway that is reminiscent of bacteria and plastids.
Malaria is a mosquito-borne disease caused by the apicomplexan parasite Plasmodium. The World Health Organization estimated that there are ϳ216 million malaria cases and 655,000 deaths annually, the latter mostly of young children in sub-Saharan Africa (1). Others recently reported that worldwide deaths are 2-fold higher (2). Resistance to most approved anti-malarials has been documented or is emerging (3), and despite encouraging progress against the disease with insecticide-impregnated bed nets and other tools, malaria control is difficult to maintain. This concern highlights the continual requirement for new anti-malarial drugs that are directed against novel targets.
The apicoplast (4) is a relict plastid and the remnant of an ancient secondary endosymbiotic event in which the eukaryotic progenitor of the malaria parasite engulfed a photosynthetic eukaryote. The circular 35-kb Plasmodium apicoplast genome (5) encodes components of the organelle's transcriptional and translational machinery (5-7) as well as the SufB protein involved in FeS cluster formation (8) and the ClpC protease. Most apicoplast proteins, however, are encoded by the nuclear genome and imported into the organelle post-translationally (9). Over 500 apicoplast-targeted proteins have been identified (10,11), revealing apicoplast biosynthetic pathways for fatty acids (9,12), isoprenoid precursors (13), and heme (14), as well as enzymes for tRNA modification (11) and lipoylation (15). Several of these pathways exhibit prokaryotic-like features and contain potential drug targets (11,13,16). Recent studies have shown that apicoplast isoprenoid precursor biosynthesis is essential in P. falciparum asexual stages (17), indicating that the pathway cannot be bypassed by salvage of lipids from the host and may be a good drug target in asexual stages. The type II fatty acid and heme biosynthetic pathways, however, are not essential in the asexual stages (16), and although not good targets for asexual stage chemotherapy, they may prove to be valuable prophylactic targets in liver stages (18).
Antibiotics that inhibit protein synthesis in prokaryotes can inhibit growth of blood stage parasites in vitro (19,20) and possess anti-malarial properties in vivo (21,22). Although slow acting against blood stage parasites and not used as first line drugs, antibiotics such as doxycycline have been used to treat patent blood stage infections that are resistant to fast acting drugs or when these drugs are unavailable (23). They are also used for prophylaxis against malaria transmitted by anopheline mosquitoes infected with Plasmodium sporozoites (24). Most, if not all, of these antibiotics inhibit apicoplast protein synthesis (25), suggesting that other processes that support this pathway might be useful drug targets.
Aminoacylated tRNAs synthesized by aminoacyl-tRNA synthetases (aaRSs) 4 are essential substrates for protein synthesis. Consequently, aaRSs have emerged as targets for new antibiotics (26). Mupirocin, for example, is a topical antibiotic already in clinical use, and other aaRS inhibitors such as the boronated antifungal compound AN2690 (27) are being developed. Aminoacylation in malaria parasites had been little studied despite its critical role in parasite biology and potential as a drug target, but studies describing the apicoplast (28) and cytoplasmic aspartyl-tRNA synthetases (29), lysyl-tRNA synthetases (30,31), and tryptophanyl-tRNA synthetases (32,33) were recently published.
The classical route for aminoacylated tRNA formation, direct aminoacylation, is catalyzed by aaRSs specific for each cognate amino acid:tRNA pair, but some aminoacylated tRNAs are made via alternative pathways. Most bacteria lack glutaminyl-tRNA synthetase (GlnRS) and produce Gln-tRNA Gln via a two-step indirect aminoacylation pathway (34) shown as Reactions 1 and 2. First, tRNA Gln is glutamylated by a nondiscriminating glutamyl-tRNA synthetase (GluRS). The misacylated Glu-tRNA Gln is subsequently converted into Gln-tRNA Gln by glutamyl-tRNA amidotransferase (Glu-AdT).
The second step (Reaction 2) catalyzed by Glu-AdT is essential because the misacylated Glu-tRNA Gln is toxic if it is not converted to Gln-tRNA Gln by Glu-AdT (35).
By analyzing conserved apicoplast-targeted proteins in the genomes of several Plasmodia (10,36,37) and the related parasite Theileria parva (38), we noticed that both organisms encoded apicoplast-targeted GluRS and Glu-AdT enzymes but lacked an apicoplast-targeted GlnRS. We hypothesized that the apicoplast utilizes the indirect pathway for Gln-tRNA Gln biosynthesis and that inhibition of this pathway might provide a new way to inhibit apicoplast protein synthesis. We began to explore these hypotheses by characterizing the first enzyme in the pathway, the apicoplast-targeted GluRS.

EXPERIMENTAL PROCEDURES
Bioinformatics-A list of putative apicoplast-targeted proteins conserved in Plasmodium falciparum, T. parva, and Toxoplasma gondii was obtained from the supplementary information in Gardner et al. (38). Nucleotide or amino acid sequences of Plasmodium genes or proteins (10,39) and those from other species were obtained from PlasmoDB (40), the Wellcome Trust Sanger Institute GeneDB website, or UniProt (41). Multiple sequence alignments were generated using tCoffee-Expresso (42) and formatted for display using ESPript (43) with SimilarityGlobalScore ϭ 0.7, SimilarityDiffScore 0.5, Similarity-Type ϭ R (Risler), and Consensus ϭ C. The phylogenetic tree was constructed using the tools provided on line (44). The "oneclick" mode was used, employing MUSCLE (45) for sequence alignment and Gblocks (46), PhyML (47), and TreeDyn (48) for curation of the multiple sequence alignment and tree construction and rendering, respectively. The final tree was constructed using 200 bootstraps.
Isolation of the PfGluRS cDNA from Blood Stage Parasites-P. falciparum 3D7 asexual stage parasites (MRA-102, MR4 ATCC (BEI Resources), Manassas, VA) were cultured as described (49). Infected erythrocytes were lysed with 0.5% saponin (Sigma) and total RNA was isolated using TRIzol (Invitrogen). A cDNA of the PfGluRS gene (Pf3D7_1357200) was amplified by RT-PCR using the GR31_F primer (5Ј-GAGAGGAAATAG-GTGTAATGT-3Ј), which maps 25 nucleotides 5Ј to the predicted start codon and the reverse primer G33_R (5Ј-AAACT-TAAATAGATTTTTCAAATGTAA-3Ј) that is complementary to the 3Ј end of the predicted coding sequence. The resulting PCR product was cloned into the pCR-XL-TOPO vector (Invitrogen) and sequenced.
Expression and Purification of Mature PfGluRS in Wheat Germ Extracts-Two versions of the PfGluRS coding sequence were cloned for expression in wheat germ extracts. The native PfGluRS coding sequence (amino acids 78 -574), minus the predicted apicoplast leader sequence (amino acids 1-77), was amplified from first strand cDNA using forward (5Ј-CCA-CATGGAGGGTAAAGTACGGTTAAG-3Ј) and reverse (5Ј-CTATTCAGTGGTGGTGGTGGTG-3Ј) primers, and a synthetic DNA segment encoding the same amino acid sequence but codon-optimized for Triticum aestivum (GeneArt AG) was amplified by PCR using the forward 5Ј-CACTATGGAGGGC-AAGGTGCGC-3Ј and reverse 5Ј-GTACTCAGTGGTGATG-GTGGTGGTG-3Ј primers. The PCR products encoding the native and codon-optimized PfGluRS enzymes were cloned into a cell-free expression vector that carries SP6 RNA polymerase promoter and ⍀ sequence for the wheat germ ribosome-binding site. In vitro transcription and translation were carried out as described (50). PfGluRS was purified from the extracts using Ni-NTA affinity chromatography. Pure fractions observed via SDS-PAGE were pooled, dialyzed against aminoacylation buffer (100 mM Hepes-KOH, pH 7.2, 30 mM KOH, 12 mM MgCl 2 ), and concentrated using Amicon ultracentrifugal filters (Millipore). The yield was 0.75 mg of protein in a total volume of 1 ml.
Expression and Purification of the Mature PfGluRS in Escherichia coli-PfGluRS was also expressed in E. coli KRX cells (Promega) containing the pRARE2 plasmid (EMD4BioSciences) and the pET-29a vector encoding the predicted mature PfGluRS coding sequence (Pf3D7_1357200, amino acids 78 -574) codon-optimized for E. coli (GeneArt, Inc.) with N-terminal His 6 tags. After Ni-NTA chromatography and gel filtration (51), pure fractions were concentrated and dialyzed overnight against aminoacylation buffer containing 0.5 mM DTT and stored at Ϫ20°C in 30% glycerol. The absence of contaminating pyrophosphatase activity was confirmed as described (52).
tRNA Substrates-Synthetic genes encoding P. falciparum apicoplast tRNA Glu and tRNA Gln (both from GenBank TM accession number X95276) were prepared by annealing overlapping oligonucleotides, ligated into the pGFIB expression vector (53), and transformed into E. coli BL21 cells. Cultures were grown, and crude total tRNA was extracted (54), deacylated in 200 mM Tris-HCl, pH 9, precipitated with isopropyl alcohol, washed with 80% ethanol, air-dried, and resuspended in 500 l of RNase-free water. Samples were stored at Ϫ80°C until use. Total E. coli tRNA (Sigma catalog no. 9014-25-9) was used for the experiment in Fig. 4. The crude P. falciparum tRNA Glu and tRNA Gln preparations were each separately 32 Plabeled on their 3Ј termini using the E. coli CCA-adding enzyme and [␣-32 P]ATP as outlined in Ref. 55.
Aminoacylation Assay and Activation by Pyrophosphatase-The aminoacylation assay using recombinant PfGluRS expressed in E. coli was performed and quantified as described in Ref. 55. In experiments testing the effect of PPase, two reactions were initiated simultaneously without PPase and allowed to proceed for 20 min. PPase (10 units/ml) was then added to one reaction and both reactions were incubated for an additional 40 min.
Determination of Kinetic Parameters-The K m value for L-Glu was determined (55)(56)(57) in reactions consisting of varying concentrations of L-Glu (5-400 M). The K m values for tRNA Glu and tRNA Gln were determined in varying concentrations of 32 P-labeled tRNA (from 0.02 to 5 M) at an L-Glu concentration of 200 M. The K m values for all substrates were determined simultaneously with enzyme aliquots from the same dilution. In all cases the concentration of PfGluRS used was chosen so as to obtain linear kinetics of glutamyl-tRNA formation in a 3-min reaction. Steady-state kinetic parameters were calculated using nonlinear regression based on the Michaelis-Menten equation (GraphPad Prism 5). The aminoacylation discrimination factor D (58) was calculated using the following formula: D ϭ (k cat /K m ) tRNA Glu /(k cat /K m ) tRNA Glu . Myc Tagging and Attempted Deletion of the Endogenous Plasmodium berghei GluRS Gene-A 4ϫ myc tag was appended to the 3Ј end of the gene encoding the putative apicoplasttargeted P. berghei GluRS (PlasmoDB ID PBANKA_113350) as described (16). A 1.6-kb fragment of the 3Ј end of the gene without the stop codon was amplified from P. berghei ANKA genomic DNA using primers PbGluRS_F1 (TACCGCGGA-AGTGCATATATATGCAAATGCA) and PbGluRS_R (ATA-CTAGTTATGTTAAATATACTCTTTATGTATTG); the SacII and SpeI restriction sites are underlined. Polymerase chain reactions (PCR) (50 liters) contained 50 ng of genomic DNA, 0.1 M of each primer, 5 l of 10ϫ buffer, and 1 l of Advantage2 polymerase (Clontech). The PCR product was digested with SacII and SpeI and cloned into the b3D myc vector (16). P. berghei ANKA parasites were transfected, and parasites with myc insertions in the GluRS gene were selected and cloned as described (16). Plasmid integration at the 5Ј and 3Ј insertion sites was confirmed by PCR using primer pairs PbGluRS_For2 (TACCGCGGAAAAGCTTATTATTGCTTTTGCAC) and PbGluRS_int5_Rev2 (GAGACAGCTCAATTCTTTATG-TCC) for the 5Ј integration test, and PbGluRS_int3_For2 (CCTC-TTCGCTATTACGCCAGCT) and PbGluRS_int3_Rev2 (GTG-AAATACGAGTATTAATATACAGG) for the 3Ј integration test.

GluRS Inhibition Assay-The
The strategy described previously (60) was used to attempt deletion of the PbGluRS gene by double crossover recombination. The primers used to amplify the genomic regions were as follows: PbGluRS_apico_KO.Pr1For (GCCCGCGGAATT-TATTGAAAATATGGTTATCAGC); PbGluRS_apico_KO.-Pr2Rev (TCATACTGTGGGCCCATAGAAGGATTACACATA-ACAAGG); PbGluRS_apico_KO.Pr3For, TCCTTCTATGGGC-CCACAGTATGAATTAAATTTGGTCAC); and PbGluRS_ apico_KO.Pr4Rev (GCGCGGCCGCTTACTTTGTGATTCA-GTTGCGC). Primers 1 and 2 were designed to amplify a 952-bp fragment containing the last 102 bp of the PbGluRS coding sequence and 850 bp of the 3Ј UTR. Primers 3 and 4 were designed to amplify an 813-nucleotide fragment containing the first 62 bp of the PbGluRS coding sequence and 751 bp of the 5Ј UTR. Primer 1 contained an added 5Ј-terminal GC dinucleotide and a SacII site. Primer 4 contained an added 5Ј-terminal GC dinucleotide and a NotI site. Primers 2 and 3 contained an ApaI site flanked by complementary sequences (underlined) for recombinatorial PCR. The two genomic fragments were first amplified in separate reactions using the cycling parameters described above, and the resulting products were combined in a second PCR to form a single product, which was digested with SacII and NotI and cloned into the B3D KO Red vector. This construct was linearized with ApaI, and P. berghei ANKA parasites were transfected as described (61). Pyrimethamine-resistant parasites were genotyped by PCR using the following primer pairs: ORF test, PbGluRS_For1 and PbGluRS_Rev; Test 1 For (ATATTCTGCGATTTTTCTTG) and Test 1 Rev (GCAAGGCGATTAAGTTGGGT); Test 2 For (GGCTACGTCCCGCACGGACGAATCCAGATGG) and Test 2 Rev (AGTATAACACTCTCTATCCTAAAA).
Subcellular Localization-Analysis of the P. berghei clones expressing Myc-tagged PbGluRS was performed as outlined in Ref. 62. Double staining was performed using a rabbit polyclonal anti-acyl carrier protein (ACP) primary antibody (diluted 1:500) (63) as an apicoplast marker and mouse monoclonal anti-Myc antibody (Santa Cruz Biotechnology, diluted 1:500) to detect PbGluRS-myc. Fluorescent staining was achieved with Alexa Fluor-conjugated secondary antibodies (Invitrogen) specific to rabbit (Alexa Fluor 594, red) and mouse (Alexa Fluor 488, green) IgG. DAPI was used to stain nucleic acids, and the mitochondrion was stained by incubating the parasites in culture media for 30 min with 20 nM MitoTracker Red (Invitrogen), and fixing the cells as outlined above. Images were acquired using an Olympus Delta Vision imaging system (Applied Precision) with a ϫ100 objective and deconvolved using the Softworx package (Applied Precision) with the default parameters.

Components of the Apicoplast Indirect Aminoacylation
Pathway-The first step in the indirect aminoacylation pathway for Gln-tRNA Gln biosynthesis is glutamylation of tRNA Gln by a nondiscriminating GluRS. The P. falciparum genome (10) encodes two putative GluRS enzymes, one of which (Pf3D7_1357200, 68.4 kDa, 574 amino acids) possesses a predicted N-terminal apicoplast-targeting sequence (64,65). The other, Pf3D7_1349200, appears to be cytoplasmic (66), suggesting that the enzyme encoded by Pf3D7_1357200 is the apicoplast GluRS and therefore should possess nondiscriminating activity. Apicoplast-targeted PfGluRS orthologs are present in all sequenced plasmodia, Theileria and Toxoplasma, but not in apicomplexans that lack apicoplasts. Organisms, or organelles, that utilize the indirect aminoacylation pathway for Gln-tRNA Gln biosynthesis often, but not always (67), lack GlnRS. Consistent with this, malaria parasites possess only one GlnRS (Pf3D7_ 1331700) that lacks an apicoplast targeting signal and is probably cytoplasmic (66,68). Nuclear genes encoding apicoplast-targeted P. falciparum orthologs of the GatA and GatB subunits of bacterial Glu-AdT were also identified, but not GatC. 5 GatC is a small, poorly conserved protein (12 kDa) that appears to perform a purely structural role at the interface between the GatA and GatB subunits in bacterial Glu-AdTs (69). A potential Plasmodium GatC ortholog may be difficult to detect by similarity searches. Furthermore, we did not find a Plasmodium ortholog of the GatF subunit of yeast mitochondrial Glu-AdT (70). Finally, putative tRNA Glu and tRNA Gln substrates of GluRS and Glu-AdT are encoded by the apicoplast genome (5). Thus, except for an ortholog of GatC, we identified all components of the indirect aminoacylation pathway in P. falciparum (Table 1) and in other apicoplast-containing apicomplexan parasites.
Isolation of the PfGluRS cDNA from Blood Stages-The predicted Pf3D7_1357200 gene encoding PfGluRS contained seven exons, but the gene structure had not yet been confirmed experimentally. We isolated a RT-PCR product from P. falciparum blood stage total RNA that confirmed the annotated gene structure. Microarray-based gene expression analyses also showed that the PfGluRS gene was transcribed in asexual parasites, with moderate induction in late trophozoites, and that its expression occurred in-phase with genes transcribed from the apicoplast genome (71). PfGluRS transcripts were identified in blood stage parasites using RNA-Seq (72), and the protein was also detected in gametocytes via proteomics (73,74).
Sequence Comparison of the PfGluRS and Bacterial Orthologs-Aminoacyl tRNA synthetases are classified into two major groups. Class I enzymes contain an N-terminal catalytic domain and generally acylate the 2Ј-hydroxyl of the terminal tRNA adenosine, whereas the catalytic domains of class II enzymes possess a seven-stranded anti-parallel ␤-sheet fold flanked by ␣-helices and acylate the 3Ј-hydroxyl of the tRNA adenosine. Glutamyl-tRNA synthetases are class I enzymes (75). The PfGluRS amino acid sequence was aligned with those of several prokaryotic orthologs ( Fig. 1) to identify conserved and divergent features. The PfGluRS possesses the "HIGH" and "KMSKF" ATP-binding motifs that are characteristic of the Rossman fold (75). Other features common to class I GluRSs that were well conserved in PfGluRS included residues that interact with the L-glutamate, ATP, and tRNA Glu substrates (76). A zinc-binding SWIM motif (CXCX 24 CX 24 H; open circles in Fig. 1) in the E. coli GluRS is essential for catalytic activity (77) but is not completely conserved in the PfGluRS (CXCX 24 CX 24 D). The His residue in the bacterial SWIM motif is required for Zn 2ϩ binding but is replaced in PfGluRS by an Asp residue, which is a potential Zn 2ϩ ligand. Other GluRSs also lack a canonical SWIM motif and do not bind or require Zn 2ϩ for activity (78). We found that PfGluRS did not require Zn 2ϩ supplementation for in vitro activity, and the ZincPred server (79) did not identify any zinc-binding sites in the protein.
Nondiscriminating GluRSs differ from discriminating enzymes in that they can glutamylate the cognate substrate tRNA Glu as well as the noncognate substrate tRNA Gln . Recognition and discrimination of tRNA substrates by GluRS (78,80) requires that the enzyme is able to distinguish between the tRNA Glu UUC and tRNA Gln UUG anticodons. The alignment (Fig. 1) revealed similarities between the PfGluRS and nondiscriminating bacterial and cyanobacterial enzymes at two residues critical for tRNA anticodon discrimination. The thermophilic eubacterium Thermus thermophilus possesses a single discriminating GluRS (81). Structural analysis of the T. thermophilus GluRS revealed that the bulky side chain of the Arg-358 residue in the anticodon binding pocket (position 442 in Fig. 1) recognizes the pyrimidine base of C36 in the anticodon of the tRNA Glu substrate. However, in the cyanobacterium Thermosynechococcus elongatus, which has a single nondiscriminating GluRS, Arg- 358 is replaced by an uncharged Gly residue. The smaller Gly side chain forms a larger pocket that allows recognition of the purine base of G36 in the noncognate tRNA Gln as well as the tRNA Glu C36 nucleotide in the cognate tRNA Glu (78). Six other amino acid substitutions of this Arg residue (Gln, Ser, Glu, Asn, Tyr, and His) have been observed in other nondiscriminating GluRSs (80). A second residue involved in anticodon discrimination is Thr-444 in the T. thermophilus GluRS (position 537 in Fig. 1), which is substituted by Gly in T. elongatus and most other nondiscriminating GluRSs (80). The PfGluRS contains Glu and Gly at these positions, respectively, which is consistent with other nondiscriminating GluRSs and our analyses of PfGluRS aminoacylation activity (see below). Phylogenetic Analysis-A maximum likelihood phylogenetic tree was constructed of the apicoplast PfGluRS, the human and P. falciparum cytoplasmic GluRSs, and GluRSs from several bacteria and an archaeon. As shown in Fig. 2, PfGluRS resolved with bacterial enzymes, consistent with an organellar origin, and not with eukaryotic cytoplasmic or archaeal enzymes.

S E I N K A L D F I A E R E N Q Q
PfGluRS Is a Nondiscriminating Enzyme-We hypothesized that the PfGluRS encoded by Pf3D7_1357200 was the first of two enzymes in an indirect aminoacylation pathway. As such, it should possess nondiscriminating activity and glutamylate apicoplast tRNA Gln as well as tRNA Glu . To test this hypothesis, aminoacylation assays were performed with mature PfGluRS in the presence of either apicoplast tRNA Glu or tRNA Gln substrates (52,82). Briefly, crude preparations of each apicoplast tRNA expressed in E. coli were 32 P-labeled at their 3Ј-terminal AMP with [␣-32 P]AMP using the nucleotide exchange activity of E. coli CCA-adding enzyme. The labeled tRNAs were then incubated with recombinant PfGluRS, and at various time points an aliquot of the reaction was withdrawn and added to an acidic solution of nuclease P1 to simultaneously quench the reaction, prevent deacylation, and digest the tRNA to mononucleotides. Next, the reaction products were subjected to TLC to sep-  arate [ 32 P]AMP from glutamyl-[ 32 P]AMP, and the extent of aminoacylation was calculated from the ratio of the radioactivity of the two spots as determined by phosphorimaging. As shown in Fig.  4, B and C, both of the apicoplast tRNA substrates were glutamylated, demonstrating that the PfGluRS is a nondiscriminating enzyme. PfGluRS did not utilize L-glutamine as a substrate 5 and did not glutamylate cytoplasmic P. falciparum tRNA Glu and tRNA Gln or E. coli tRNA (Fig. 4, B and C). Because nondiscriminating GluRSs are often found in cells or organelles that lack GlnRS and must use indirect aminoacylation for Gln-tRNA Gln biosynthesis, this finding provides strong biochemical evidence that the indirect pathway is used in the Plasmodium apicoplast.
Effect of PPase on PfGluRS Aminoacylation Activity-Aminoacylation reactions performed with varying amounts of PfGluRS clearly showed that an increase in enzyme concentration was accompanied by an increase in the reaction rate (Fig.  5A). However, the aminoacylation activity plateaued within a few minutes after enzyme addition, an effect suggestive of product inhibition. Adding more PfGluRS to the reaction after it had plateaued did not increase aminoacylation, implying that this effect was not an artifact caused by enzyme denaturation. Because inorganic pyrophosphate (PP i ) can inhibit tRNA synthetase activity (52), we added PPase to an aminoacylation assay after the reaction had plateaued. This dramatically stimulated the PfGluRS reaction rate and increased the steady-state level of the Glu-tRNA Glu product ϳ3-fold compared with reactions without PPase (Fig. 5B). Enhanced PfGluRS activity in the presence of PPase indicated that the strong inhibition observed in its absence was due to the generation of PP i during the reaction.
Kinetic Analyses-Having demonstrated PfGluRS aminoacylation activity and inhibition by PP i , we next conducted kinetic analyses to determine optimal reaction parameters and to compare the Plasmodium enzyme to bacterial orthologs. PfGluRS demonstrated higher affinities toward the cognate as compared with the noncognate tRNA substrate (Table 2). We then determined the same kinetic values for PfGluRS in reactions contain-

TABLE 2 Kinetic analyses of PfGluRS
The K m value for the tRNA substrates was determined in reactions containing 100 mM Hepes-KOH, pH 7.2, 30 mM KOH, 12 mM MgCl 2 , 2 mM DTT, 4 mM ATP, 200 M L-glutamate, 4 g/ml recombinant PfGluRS expressed in E. coli and varying concentrations of the indicated 32 P-labeled tRNA from 0.02 to 5 M. NA means not applicable.  NOVEMBER 8, 2013 • VOLUME 288 • NUMBER 45

P. falciparum Apicoplast Glutamyl-tRNA Synthetase
ing PPase. Here, the stimulatory effect of PPase was dramatic, increasing turnover by 8.4-and 5-fold and catalytic efficiency by 13-and 4-fold, for tRNA Glu and tRNA Gln substrates, respectively, in comparison with reactions performed in its absence. The discrimination factor, D, a measure that compares enzyme turnover rates for cognate and noncognate substrates at the same concentration (83) was also calculated, revealing that PfGluRS glutamylated the cognate tRNA Glu substrate more efficiently than the noncognate tRNA Gln substrate, an effect that was most pronounced in the presence of PPase. PfGluRS Inhibition by a Bacterial GluRS Inhibitor-Glu-SA, a GluRS inhibitor, is a stable analog of the glutamyl-AMP intermediate of the aminoacylation reaction (59). It belongs to a group of synthetic aminoacyl adenylates called aminoacylsulfamoyladenosines in which the labile mixed anhydride function of the aminoacyladenylate intermediate in the aminoacylation reaction is replaced by an isosteric, nonhydrolyzable sulfamoyl group (84). Glu-SA is a strong competitive inhibitor of the E. coli GluRS (K i ϭ 2.8 nM) but is a weaker competitive inhibitor of mammalian (mouse liver) GluRS (K i ϭ 70 nM) (59). Using the aminoacylation assay employed above to determine the kinetic values for PfGluRS, we found that Glu-SA is a competitive inhibitor of the mature PfGluRS (Fig. 6). However, Glu-SA is much less potent against the P. falciparum enzyme (K i ϭ 0.7 M) than it is against either the E. coli or mammalian GluRSs.
Subcellular Localization-The P. falciparum GluRS encoded by Pf3D7_1357200 contains a predicted apicoplast targeting sequence (10,65,85), but the subcellular localization of the enzyme has never been established experimentally. To determine whether the enzyme is targeted to the apicoplast in vivo, we tagged the endogenous gene encoding the P. berghei ortholog of PfGluRS (Table 1) with a quadruple Myc tag (Fig. 7). We tagged the endogenous PbGluRS coding sequence so that the native promoter controlled the timing and level of expression, minimizing the possibility that the fusion protein would  be mis-targeted. Fixed blood stage parasites were stained with anti-Myc antibody to detect PbGluRS and ACP antisera (9) to detect the apicoplast, and the samples were observed using Deltavision deconvolution fluorescence microscopy. Structures containing the Myc-tagged PbGluRS (PbGluRS-myc) exhibited a typical apicoplast appearance (Fig. 8A). PbGluRS-Myc marked a small spherical compartment in ring stage parasites (Fig. 8A), which then elongated and developed into a complex branched form at the trophozoite stage (Fig. 8A) prior to splitting into numerous individual structures in schizonts, one for each daughter merozoite. PbGluRS-Myc co-localized with ACP (Fig. 8A, ␣-Myc/␣-ACP overlay), confirming that it is targeted to the apicoplast. This result demonstrated that the PbGluRS genomic locus is susceptible to integration of plasmid DNA.
To investigate potential localization to the mitochondrion, live parasites expressing PbGluRS-myc were incubated with MitoTracker Red and then fixed, stained with anti-Myc monoclonal antibody, and examined by fluorescence microscopy. In ring stages, the anti-Myc antibody and MitoTracker Red marked distinct organelles that were closely apposed to each other (Fig. 8B). Little co-localization between the mitochondrion and PbGluRS-Myc was observed in late trophozoites or early schizont stages with the mitochondrial staining largely distinct from that of PbGluRS-Myc, apart from a few apparent points of contact between the two organelles (Fig. 8B). Close apposition of the mitochondrion and apicoplast has been previously observed in Plasmodium spp. (86). At the late schizont stage, the mitochondria and the apicoplast again appeared as closely apposed but distinct organelles adjacent to the nuclei. Together with our demonstration above that PbGluRS co-localizes with the apicoplast protein ACP (Fig. 8A), these observations show that PbGluRS-myc is located within the apicoplast but not the mitochondrion in erythrocytic parasites. Finally, a bipartite apicoplast-targeting sequence from a Babesia bovis GluRS was recently shown (87) to direct GFP into the Babesia apicoplast. This is consistent with our localization of the PfGluRS by an analogous approach, 5 and by epitope tagging of the endogenous PbGluRS enzyme (Figs. 8 and 9).
Attempted Deletion of the PbGluRS Gene-The apicoplast GluRS should be essential because it produces two substrates for apicoplast protein synthesis, Glu-tRNA Glu directly, and Gln-tRNA Gln indirectly, in concert with a Glu-AdT. Because bioinformatic analyses suggest that Plasmodium lacks an apicoplast-targeted GlnRS (66,88), indirect aminoacylation is probably the sole route for Gln-tRNA Gln formation in the apicoplast. To test whether PbGluRS was required for blood-stage growth, we transfected P. berghei parasites with a construct (Fig. 9) designed to delete the endogenous PbGluRS gene by double crossover recombination. As a control, we transfected parasites from the same batch with the same construct (Fig. 7) used to generate the PbGluRSmyc transgenic parasites. In three independent experiments, we were unable to integrate the deletion construct into the PbGluRS genomic locus via double crossover recombination. As shown earlier, however, transgenic PbGluRSmyc parasites were readily obtained (data not shown). These results indicate that the PbGluRS locus was accessible to recombination but that the gene could not be deleted, strongly suggesting that the apicoplast-targeted GluRS is essential in blood stage parasites.

DISCUSSION
Plasmodium aaRSs have not been widely studied despite their critical role in protein synthesis, a process absolutely required for parasite replication. One early study reported aaRS activity in P. berghei cell-free extracts (89), and genome sequencing later revealed the entire complement of aaRSs in Plasmodium (10,68,88). But to date, cytoplasmic and apicoplast P. falciparum aspartyl-tRNA synthetase (28,29), lysyl-tRNA synthetase (30,31), and tryptophanyl-tRNA synthetase (32,33) are the only other malarial aaRSs besides PfGluRS to be functionally or structurally characterized.
To our knowledge, the apicoplast PfGluRS is only the second apicoplast aaRS to be biochemically characterized. Recently, aminoacylation by recombinant apicoplast LysRS (PF3D7_1416800) was shown to be inhibited by a series of novel compounds that mimicked the lysyl-adenylate reaction intermediate, and two compounds exhibited IC 50 values in the 40 -85 nM range against asexual parasites in vitro (28). Istvan et al. (90) also showed that apicoplast isoleucyl-tRNA synthetase (PF3D7_1225100) is the target of mupirocin, an inhibitor of bacterial isoleucyl-tRNA synthetases used to treat skin infections. Mupirocin also inhibited parasite growth in vitro at very low concentrations (EC 50 ϳ58 nM). Together, these results suggest that inhibitors of other apicoplast aaRSs might also exhibit good anti-parasitic activity (91).
We used recombinant PfGluRS produced in E. coli (Fig. 3) to conduct biochemical analyses to determine whether the enzyme had nondiscriminating aminoacylation activity required for indirect aminoacylation and to establish optimal reaction conditions. We chose to use an aminoacylation assay (52) that was sensitive and could be adapted to measure Glu-AdT activity (92,93). PfGluRS glutamylated apicoplast tRNA Glu and tRNA Gln , but not cytoplasmic tRNA Glu , tRNA Gln , or E. coli tRNA (Fig. 4, B and C). The aminoacylation activity was therefore both nondiscriminating and specific for the apicoplast tRNAs. The kinetic analyses showed that, like other nondiscriminating GluRSs (94), PfGluRScatalyzed glutamylation of the cognate and noncognate substrates at similar rates (Table 2). We also found that PfGluRS was sensitive to oxidation because DTT was required to maintain enzyme activity. The sensitivity of PfGluRS to oxidation might provide a means to regulate apicoplast protein synthesis in response to changes in the organelle's redox status (95).
As with other aaRSs (52), PfGluRS-catalyzed glutamylation was subject to product inhibition by PP i , which was reversed by adding PPase to the reaction (Fig. 5B). Aminoacylation of tRNAs proceeds in two steps. First, the tRNA synthetase binds ATP and amino acid to form an enzyme-bound aminoacyladenylate and PP i . Second, upon binding of the tRNA substrate to the enzyme-aminoacyl-adenylate complex, the amino acid is transferred from the aminoacyl-adenylate to the 3Ј end of the tRNA with the concomitant release of PP i . Subsequent PP i hydrolysis provides a thermodynamic push for the reaction, so PP i must be removed to prevent aaRS inhibition. Inorganic pyrophosphatase performs this function in vivo. PPase is essential in E. coli (96) and yeast (97), and chloroplasts possess a plastid-specific PPase. Transient repression of plastid PPase expression in tobacco leaves increased PP i levels, reduced total soluble leaf protein by 60%, and inhibited expression of nucleusand plastid-encoded subunits of the carbon-fixing enzyme ribulose-1,5-bisphosphate carboxylase oxygenase, suggesting that repressing plastid PPase levels reduced cytoplasmic and plastid protein synthesis (98 FIGURE 9. Attempted knock-out of the PbGluRS gene by double crossover recombination (replacement). Two fragments containing 5Ј-and 3Ј-UTRs of the PbGluRS gene, with ϳ100 nucleotides of the start and end, respectively, of the PbGluRS coding sequence were cloned into the B3D KO Red vector (60). After linearization with ApaI, the construct was transfected into P. berghei ANKA parasites (108). In three independent experiments, we were unable to delete the endogenous PbGluRS gene by integration of the deletion construct via double crossover recombination.
encodes a soluble pyrophosphatase (PF3D7_0316300) that may be dual-targeted to the apicoplast and the cytoplasm via two alternatively spliced isoforms. The enzyme is expressed in asexual stages and gametocytes (71,100); an apicoplast isoform might regulate PP i levels within the organelle to prevent PP imediated inhibition of protein synthesis or PP i -sensitive enzymes in pathways such as isoprenoid biosynthesis (13). Next, we tested whether a potent bacterial GluRS inhibitor, Glu-SA, was able to inhibit the Plasmodium enzyme. Glu-SA inhibited PfGluRS activity (K i ϭ 0.7 M, Fig. 6) but with a 250and 10-fold lower efficiency than for E. coli or murine liver GluRSs, respectively (59). The wide range of susceptibility to Glu-SA inhibition exhibited by enzymes from different lineages probably reflects structural differences between their active sites. These differences between Plasmodium and host enzymes might be exploited to produce inhibitors with selective activity against the Plasmodium apicoplast GluRS.
The PfGluRS and its apicomplexan orthologs possess a bipartite apicoplast targeting sequence, but the enzyme's subcellular localization had never been determined experimentally. We found via immunofluorescence microscopy of Myc-tagged PbGluRS that the enzyme was localized in the apicoplast in erythrocytic stage parasites (Fig. 8A). Minor overlaps between the anti-Myc and MitoTracker Red signals were observed where the apicoplast and mitochondrion appeared to contact one another (Fig. 8B), a phenomenon observed with other apicoplast-targeted proteins (86,101). Localization of PbGluRS solely in the apicoplast differs from the situation in Arabidopsis, where at least 15 nucleus-encoded aaRSs, including GluRS, are targeted to the plastid and the mitochondrion (102). Plasmodial genomes encode a second GluRS in addition to the one studied here, but that enzyme, encoded in P. falciparum by the gene Pf3D7_1349200, is predicted to be cytoplasmic (66,68,88). Dual targeting of aaRSs to the plastid and mitochondrion is common in higher plants (102), but this is unlikely in Plasmodium because the mitochondrion in the related parasite Toxoplasma imports aminoacylated tRNAs from the cytoplasm (66).
In this work, we identified components of an indirect aminoacylation pathway to produce Gln-tRNA Gln for apicoplast protein synthesis. We demonstrated that the first enzyme in the pathway, GluRS, is targeted to the apicoplast in blood stage parasites and can glutamylate both tRNA Glu and tRNA Gln in vitro. The latter property is associated with GluRSs that play a role in indirect aminoacylation, providing convincing evidence that the apicoplast indeed uses this pathway. The second enzyme in the pathway, Glu-AdT, has not been studied in Plasmodium, but we have expressed recombinant P. falciparum GatA and GatB subunits 5 and are testing whether they can amidate Glu-tRNA Gln produced by the PfGluRS to form Gln-tRNA Gln . We are also investigating a potential novel feature of the Plasmodium Glu-AdT, the apparent absence of a GatC subunit that is required to stabilize the bacterial Glu-AdT (GatCAB) (69,103), and whether the GluRS, Glu-AdT, and tRNA Gln form a ribonucleoprotein complex (the transamidosome) as they do in bacteria (104 -106). Finally, apicoplast indirect aminoacylation is probably essential in malaria parasites because the parasite genome does not encode an apicoplasttargeted GlnRS (66,68,88). This is consistent with our inability to delete the gene encoding the P. berghei apicoplast GluRS in three independent experiments, despite the fact that it was accessible to myc tagging. This pathway may be a potential drug target because apicoplast protein synthesis is essential in both blood and liver stages. Further investigations to examine the requirements for apicoplast indirect aminoacylation across the Plasmodium life cycle and to identify specific inhibitors of plasmodial GluRS or Glu-AdT enzymes may lead to novel ways in which to target this pathway for chemotherapy.