Inhibition of Protein Synthesis and Malaria Parasite Development by Drug Targeting of Methionyl-tRNA Synthetases

Abstract

Aminoacyl-tRNA synthetases (aaRSs) are housekeeping enzymes that couple cognate tRNAs with amino acids to transmit genomic information for protein translation. The Plasmodium falciparum nuclear genome encodes two P. falciparum methionyl-tRNA synthetases (PfMRS), termed PfMRS^cyt and PfMRS^api. Phylogenetic analyses revealed that the two proteins are of primitive origin and are related to heterokonts (PfMRS^cyt) or proteobacteria/primitive bacteria (PfMRS^api). We show that PfMRS^cyt localizes in parasite cytoplasm, while PfMRS^api localizes to apicoplasts in asexual stages of malaria parasites. Two known bacterial MRS inhibitors, REP3123 and REP8839, hampered Plasmodium growth very effectively in the early and late stages of parasite development. Small-molecule drug-like libraries were screened against modeled PfMRS structures, and several “hit” compounds showed significant effects on parasite growth. We then tested the effects of the hit compounds on protein translation by labeling nascent proteins with ³⁵S-labeled cysteine and methionine. Three of the tested compounds reduced protein synthesis and also blocked parasite growth progression from the ring stage to the trophozoite stage. Drug docking studies suggested distinct modes of binding for the three compounds, compared with the enzyme product methionyl adenylate. Therefore, this study provides new targets (PfMRSs) and hit compounds that can be explored for development as antimalarial drugs.

INTRODUCTION

Plasmodium falciparum is the most virulent form of Plasmodium and a causative agent of malaria. The World Health Organization (WHO) estimates that there are ∼0.62 million deaths due to malaria per year (1). The P. falciparum genome is AT-rich (81%) and codes for ∼5,300 proteins, with unusual distributions of several residues (2). Almost 60% of encoded proteins appear to be unique to the parasite, reflecting great evolutionary distance between the parasite and the genomes of known eukaryotes (3). The malaria parasite (and the related apicomplexan Toxoplasma gondii) has three translationally active compartments, i.e., cytoplasm, apicoplasts, and mitochondria (4–8). All malaria parasite proteins involved in the protein synthesis machinery are encoded by the nuclear genome. Either these proteins are transported to target organelles or their modified/activated substrates are transported across the organelles to perform required functions (4–9). The primary enzymes responsible for translating genetic code into polypeptide chains are aminoacyl-tRNA synthetases (aaRSs). The canonical function of aaRSs is to ligate a specific amino acid to its cognate tRNA, which is then employed in protein synthesis. The aaRSs are an ancient class of essential enzymes, and their catalytic domains are conserved in all kingdoms (bacteria, archaebacteria, and eukaryotes). On the basis of catalytic domain architecture, aaRSs are classified into two categories, with each class containing ∼10 aaRSs (10). Although aaRSs perform the basic function of charging tRNA molecules for protein synthesis, aaRSs have acquired a wide range of sequence, structural, and functional diversity during the course of evolution (11). The aaRSs are known to incorporate additional domains that deploy the aaRSs for different noncanonical functions (12). Plasmodium falciparum possesses 36 aaRSs, which show asymmetric distributions among parasite organelles (7, 8, 13, 14). The presence of appended domains imparts characteristic functions to parasite aaRSs (13–15). For example, recent studies have revealed cytokine-like functions for malaria tyrosyl-tRNA synthetase (tyrosyl-RS) (15).

In eukaryotes, methionyl-tRNA synthetases (MRSs) possess glutathione-S-transferase (GST)-like domains at their N termini, and these are involved in protein-protein interactions (16). The appended C-terminal RNA binding domains in MRSs confer higher tRNA binding affinity, thereby contributing to catalytic function (17). The crystal structure of the GST domain of yeast MRS in complex with the GST domain of Arc1P (a cofactor) revealed a classic GST homodimer responsible for multisynthetase complex formation in higher eukaryotes (18). In some organisms (e.g., Thermos thermophilus), MRS is a homodimer driven by the presence of a C-terminal domain that also displays nonspecific tRNA binding properties. Its three-dimensional structure resembles that of Trbp111, a dimeric tRNA binding protein found in bacteria and archaeans (19). The MRS C-terminal domain in plants possesses an endothelial-monocyte-activating polypeptide II (EMAPII) domain-like structure and is monomeric at physiological concentrations (20). The C-terminal domains of MRSs, despite variations in structural folding across organisms, are functionally related and are an example of convergent evolution owing to their tRNA binding properties (21). In this light, the appended domains in P. falciparum methionyl-tRNA synthetases (PfMRSs) bear highly divergent sequences in comparison with other organisms.

The emergence of resistant strains of P. falciparum continues to fuel an urgent need for the development of new antimalarials. Malaria parasite aaRSs are currently being explored as new targets for drug development (22, 23). Within aaRSs, MRSs can serve as valuable drug targets because of their sequence and domain heterogeneity. Inhibitors that target MRSs are already under development against bacterial infections (24). Derivatives of diarylamines, quinolones, urea, and various other lead compounds with potent activities against MRSs have been tested (25–27). Therefore, we decided to explore various attributes of malarial MRSs with the aim of probing their potential for drug targeting. Here we report the localization and phylogenetic analysis of both copies of PfMRSs. We also provide parasite growth inhibition data using drug-like compounds to address the feasibility of targeting PfMRSs. Some of the “hit” compounds are able to abrogate protein translation in malaria parasites, suggesting that they likely target the active sites of PfMRSs. In summary, our data add to the growing family of parasite aaRSs that can be targeted for inhibitor development against malaria parasites.

MATERIALS AND METHODS

Cloning, expression, and purification of subdomains of PfMRSs and antibody generation.

Clones of the N- and C-terminal domains of MRS^cyt and the anticodon binding domain of MRS^api were synthesized from full-length genes using the following primer pairs: (i) forward, GCTCCATGGAATTCATGATG; reverse, GTGGTACCTTATTAATTAATGGCGGTGGTGATATAAA; (ii) forward, GCTCCATGGGCGCGAAAATTAAACTGCAG; reverse, GTGGTACCTTATTAAAAAAAGGTCAGGCTACC; (iii) forward, GTCCATGGCAAAAGAGCAGAACATCGAAAGCTTCGAACTG; reverse, GTGGTACCTTATTAAAACATCAGAATGCTGAAGTATTTCAT. The vector PetM11 was used for protein expression in BL21(DE3) cells. Culture medium for growing transformed cells was inoculated with 1% culture grown overnight at 37°C until the optical density (OD) at 600 nm reached 0.8. Protein expression was induced with 0.2 mM isopropyl-β-d-thiogalactopyranoside (IPTG) at 18°C, and cells were allowed to grow for 10 to 12 h. The cells were harvested at 5,000 × g for 30 min and sonicated, and proteins were purified using immobilized metal affinity chromatography. A further purification step of gel permeation chromatography and ion-exchange chromatography was carried out to purify target proteins. Antibodies against PfMRSs were generated in rabbits, and previously characterized antibodies against parasite proteins were used as controls where appropriate (28, 29).

Culture of Plasmodium falciparum 3D7 and D10-ACP leader-GFP-transfected cells.

Plasmodium falciparum 3D7 cells were cultured with O⁺ red blood cells (RBCs) in RPMI 1640 medium (Invitrogen) supplemented with 4.5 mg ml⁻¹glucose (Sigma), 0.1 mM hypoxanthine (Invitrogen), 25 mg ml⁻¹ gentamicin (Invitrogen), and 0.5% AlbuMax I (Invitrogen), according to standard methods. Parasites were treated with sorbitol in the ring stage to maintain synchronized cultures, as described previously (30). The Plasmodium falciparum D10-acyl carrier protein (ACP) leader-green fluorescent protein (GFP) transfectant line, in which GFP is targeted to the apicoplast by the leader peptide of ACP, was cultured similarly and supplemented with the addition of pyrimethamine (10 nM).

Confocal microscopic examination of blood-stage parasites.

Cells were washed with phosphate-buffered saline (PBS) and fixed in solution with 4% paraformaldehyde and 0.0075% glutaraldehyde in PBS for 30 min. After one wash with PBS, cells were permeabilized with 0.1% Triton X-100 in PBS for 10 min. After another PBS wash, cells were treated with 0.1 mg ml⁻¹ sodium borohydride in PBS for 10 min. Cells were washed once with PBS, blocked with 3% bovine serum albumin (BSA) in PBS for 1 h, and incubated overnight at 4°C with either rabbit antiprotein antiserum (1:200 dilution) or mouse antiserum against d-aminoacyl-tRNA deacylase (DTD), a cytoplasmic marker (1:200 dilution). Cells were washed three times (10 min each time) with PBS and incubated for 2 h at room temperature with Alexa Fluor 488-tagged goat anti-mouse IgG secondary antibody or Alexa Fluor 594-tagged goat anti-rabbit IgG secondary antibody. The cells were allowed to settle onto coverslips coated with poly-d-lysine (100 μg ml^−l), and then the coverslips were washed three times with PBS, mounted in antifade medium with 4′,6-diamidino-2-phenylindole (DAPI), and sealed. Confocal microscopy was performed with immunofluorescently labeled parasites by using a Nikon A1R microscope with diode (405 nm), argon (488 nm), and helium-neon green (543 nm) lasers, with a 100× oil immersion lens. Images were viewed and analyzed using NIS-Elements software (version 3.2).

Modeling, in silico screening, and phylogenetic analyses.

For modeling of the synthetase domains of PfMRSs, the sequences were compared with homologs of known structure through BLAST analysis with the Protein Data Bank (PDB). The MRS from Thermos thermophilus (PDB accession number 1A8H) was found to be the nearest homolog and was used as the template. Prior to modeling, low-complexity regions (LCRs) in the PfMRSs were removed. Atomic models were obtained using Modeller (31) and were selected based on discrete optimized protein energy (DOPE) and modular objective function (MOF) scores. Subsequently, models were validated using the structural analysis verification server (SAVES) (http://nihserver.mbi.ucla.edu/SAVES). PfMRS structural models were subjected to energy minimization using the Prime module in the Schrödinger suite (Prime version 3.0; Schrödinger, LLC, New York, NY), Ramachandran plot outliers were fixed, and models were revalidated using SAVES. The active sites of the PfMRS three-dimensional models were analyzed using Chimera (32), in the context of sequence alignments and conserved motif information. The small-molecule drug-like library (∼50,000 compounds) was downloaded from Specs (http://www.specs.net). The library was prepared using the LigPrep and QikProp modules in the Schrödinger suite and was passed through 10 lead-like filters and analysis with Lipinski's rule of five. The PfMRS^cyt model was prepared using the PrepWizard module and was used to build an energy grid. For Glide docking, the box was centered on active site residues identified from methionyl adenylate (MOD)-bound MRS complex structures. The Glide extra precision (XP) algorithm was used to perform virtual screening. The top-ranking ligands were selected based on docking scores and H-bond interactions. Interactions of potential inhibitors were analyzed using Chimera (32) and LigPlot (33). The average logP and logS values were calculated using the Virtual Computational Chemistry Laboratory server (http://www.vcclab.org). For phylogenetic analysis, 61 MRS sequences were aligned by using ClustalW2 and a phylogenetic tree was constructed with the maximum likelihood method by using the Mega 5.0 tool (34, 35).

In vitro parasite inhibition assays and ³⁵S labeling of Plasmodium falciparum.

The Plasmodium falciparum 3D7 strain was cultured in 96-well plates and synchronized at the ring stage using sorbitol. At ∼1% parasitemia and 2% hematocrit levels, the compounds were incubated for 48 h with parasite cultures, and the mixtures were assayed by the SYBR green I DNA staining method, as described previously (36, 37). After 48 h, 100 μl SYBR green dye at 1× concentration in lysis buffer supplemented with 0.1% saponin was added to each well. After 45 min of incubation at 37°C, fluorescence was estimated using a multiwell plate reader (Victor3; PerkinElmer), with excitation and emission wavelength bands centered at 485 and 530 nm, respectively. Anisomycin and pyrimethamine were used as control compounds, and all experiments were performed in triplicate. The 50% effective concentration (EC₅₀) values were obtained by plotting fluorescence values, expressed in terms of percent inhibition of parasite growth, at each inhibitor concentration. For protein synthesis inhibition assays, asynchronous cultures of the Plasmodium falciparum 3D7 strain were harvested at parasitemia levels of ∼7 to 10% and hematocrit levels of 4%. The infected RBCs (iRBCs) were washed three times with methionine- and cysteine-free medium (Sigma) and were aliquoted in triplicate into 96-well plates. The selected compounds were added to culture plates in serially diluted concentrations from 0.1 nM to 100 μM. Anisomycin (Sigma), a known inhibitor of protein translation, was used as a positive control. The remaining steps were performed as reported previously (38). Counts were measured in a scintillation counter (Beckman Coulter) and compared with those of precipitates of parasites not treated with inhibitor or compounds. Plots were prepared with GraphPad Prism 6 software.

Growth inhibition by REP3123 and REP8839.

The maturation of schizonts was evaluated by incubating iRBCs, at 0.5 to 1% parasitemia, in the presence of REP3123 or REP8839. Parasites in the later stages of development (∼40 h) were treated with inhibitors at 200 nM or 500 nM. The final dimethyl sulfoxide (DMSO) concentration was maintained below 1%. The maturation of schizonts was counted on thick-smear glass slides after 24 h of treatment. The percent inhibition of schizont maturation was estimated as the number of unreleased schizonts versus the number in control parasite cultures without inhibitor treatment. The experiments were repeated three times, and the final results were averages from the three experiments.

In vitro transcription of tRNA^Met and preparation of total protein extracts from parasites.

The sequence for tRNA^Met was taken from the PlasmoDB database and was synthesized commercially with the addition of a T7 promoter site in the 5′ region. The tRNA was prepared from a double-stranded DNA template. The last two bases in the reverse primer were modified with 2-O-methyl substitution at the hydroxyl group of the ribose sugar. The tRNA was then synthesized using an in vitro transcription kit (New England BioLabs), by incubating the reaction mixture for 10 h at 37°C. The tRNA was isolated by ethanol precipitation in the presence of 0.3 M sodium acetate and was resuspended in RNase-free water.

To extract total protein from Plasmodium falciparum 3D7 parasites, RBCs were lysed with 0.15% saponin and centrifuged at 14,000 rpm for 10 min at 4°C to separate the parasites. These cells were then lysed on ice for 10 min with 0.65% NP-40 (HiMedia) in buffer A, containing 30 mM HEPES (pH 7.5), 140 mM NaCl, 30 mM KCl, 40 mM MgCl₂, and 1 mM dithiothreitol (DTT), supplemented with protease inhibitors. Protein extracts were stored at −80°C.

PfMRS^cyt activity assays in the presence of inhibitors.

To determine the effects of inhibitors on the activity of the PfMRS form termed PfMRS^cyt, 100 μg of total protein extract from Plasmodium falciparum 3D7 cells was assayed for activity in the presence of tRNA prepared in vitro, using the malachite green assay described by Cestari and Stuart (39), with some modifications. Fold increases in P_i release were assessed by performing the aminoacylation reaction in a time-dependent manner in the absence and presence of tRNA^Met.

To determine the effects of inhibitors on aminoacylation, reactions were performed in 50-μl volumes in 96-well flat-bottom plates containing 100 μg total protein extract. The inhibitors were added at serially diluted concentrations (0.1 nM to 100 μM) and incubated with protein extract for 15 min at 37°C in buffer A. The reaction mixture was then supplemented with 1 μM tRNA^Met, 100 μM ATP, 50 μM methionine, and 2 U/ml pyrophosphatase (PPase) in buffer A and incubated for 10 min at 37°C. The reaction was stopped with the addition of one-fourth of the total reaction volume of malachite green reagent. Reactions performed in the absence of inhibitors were used as controls. Pyrophosphate release was quantified by the formation of complexes between malachite green molybdate and free orthophosphate. The absorbance was read at 620 nm using a 96-well enzyme-linked immunosorbent assay (ELISA) plate reader (Versamax; Molecular Biosystems). Absorbance values were compared with a standard curve, and results were plotted as percent decrease in PfMRS^cyt activity versus inhibitor concentration.

RESULTS

Domain organization of PfMRS^cyt and PfMRS^api and comparison with other MRSs.

The Plasmodium species possess two copies of MRS enzymes encoded within their nuclear genomes. In the case of P. falciparum, the genome contains the genes PF10_0340 (PfMRS^cyt) and PF10_0053 (PfMRS^api) in PlasmoDB, with each PfMRS having a distinct domain architecture (Fig. 1). Both proteins expectedly show conserved motifs such as HIGH and KMSKS, which are characteristic of class I aaRSs. However, the sequences of the synthetase subdomains of MRS^cyt and MRS^api are only ∼26% identical to each other. In addition to the synthetase domain of 497 residues in MRS^cyt, domains of 225 residues at the N terminus and 167 residues at the C terminus are appended to the core catalytic domain. In yeast and humans, the structural folding of the N-terminal MRS^cyt domain is similar to nonenzymatic glutathione S-transferase (GST) folding. The presence of the GST domain may be meant to mediate interactions of MRS with other proteins containing GST domains (40). Although the GST domain in PfMRS^cyt is conserved among Plasmodium species, it does not show any significant similarity to orthologous MRS proteins. On the other hand, MRS^api has a unique low-complexity insertion region of 35 amino acids in the connective peptide 1 (CP1) region and a highly divergent sequence of 140 residues at the N terminus (of 609 residues). Furthermore, and rather surprisingly, the N-terminal extension in MRS^api is absent in Plasmodium vivax, suggesting unusual intraspecies variation in this set of housekeeping enzymes (Fig. 1).

FIG 1.

Overall domain architecture of MRSs from prokaryotes and eukaryotes. The PfMRS proteins bear additional domains at the N and C termini, with no significant homology to other MRSs. HsMRS, Homo sapiens MRS; EcMRS, E. coli MRS; ScMRS, Saccharomyces cerevisiae MRS.

Localization of MRS^cyt and MRS^api in malaria parasites.

Despite extensive efforts, recombinant production of both PfMRSs was not successful in our laboratory. Therefore, for localization studies, polyclonal antibodies were raised against the recombinantly expressed N- and C-terminal domains of PfMRS^cyt and tRNA binding domain of PfMRS^api. These antibodies were used to study, using confocal microscopy, the localization of the respective proteins in all asexual stages of malaria parasites. The expression of both PfMRSs in parasites was confirmed using these antibodies (Fig. 2A and B; also see Fig. S1B in the supplemental material). The microscopy data revealed the presence of PfMRS^cyt solely in the cytoplasm of asexual stages, including merozoites (Fig. 2C; also see Fig. S1C, D, and F in the supplemental material). In contrast to human methionyl-tRNA synthetase, which shows nuclear localization in the presence of growth factors (41), translocation of PfMRS^cyt was not observed in any of the asexual stages even in the presence of growth factors (data not shown). In all asexual blood stages of parasite development, PfMRS^api showed distinct localization in the apicoplasts of the parasites (Fig. 2D; also see Fig. S1E in the supplemental material). From these localization studies, it is evident that PfMRSs are compartmentalized within the parasite cytoplasm and apicoplasts, to assist in protein translation in these compartments.

FIG 2.

(A) SDS-PAGE analysis of the purified recombinant N-terminal domain of PfMRS^cyt (residues 1 to 232, corresponding to a 28-kDa band) (left) and expression of PfMRS^cyt in parasite lysates assessed using antibodies against this N-terminal domain (right). (B) Purified maltose binding protein (MBP)-tagged anticodon binding domain (residues 584 to 704) of PfMRS^api. Left, SDS-PAGE gel, showing bands at 57 kDa, corresponding to the MBP-fused PfMRS^api anticodon binding domain, and at 43 kDa, corresponding to MBP. Right, Western blot of PfMRS^api in parasite lysates using antibodies against the anticodon binding domain. Lanes 1 in panels A and B show protein molecular weight markers, and lanes 2 show purified recombinant protein. (C) Localization of PfMRS^cyt using antibodies directed against the N-terminal domain. The protein localizes in the cytoplasm for all asexual parasite stages. (D) Localization of MRS^api, studied using antibodies against the anticodon binding domain. Localization of MRS^api in apicoplasts in three asexual stages (ring, trophozoite, and schizont) was verified.

Evolutionary relationships of PfMRS^cyt and PfMRS^api.

The differential localization of PfMRSs prompted us to analyze the phylogenetic relationships of MRSs in various taxa. This analysis covered organisms from primitive bacteria, fungi, and algae to complex eukaryotes. To avoid any bias in the data, we trimmed the N- and C-terminal-appended domains, as they do not share any homology with sequences of known MRSs. Similarly, a low-complexity region (LCR) of 35 amino acids in PfMRS^api was removed. The phylogenetic data revealed a lower-plant-like origin of PfMRS^cyt, whereas PfMRS^api showed close similarity to fungus-like organisms. The synthetase domains of the two PfMRSs showed a polyphyletic relationship with each other. The apicomplexan MRS^cyt included in our analysis fell close to heterokonts such as Phytopthora and Thalassiosira, while MRS^api appeared in a fungal clade of the phylogenetic tree. PfMRS^api and Leishmania major MRS are also closely related to primitive thermophilic and nitrogen-fixing bacteria, sharing a common ancestor (see Fig. S2 in the supplemental material). It was unexpected to find fungal and primitive bacterial associations of MRS^api, as this enzyme resides in parasite apicoplasts and was expected to be of plant-like origin, owing to the plastid-like nature of apicoplasts. The distinct evolutionary history of MRS^cyt and MRS^api suggests that the two proteins have not originated as a result of a gene duplication event and are not paralogs.

Modeling of PfMRS synthetase domains and in silico inhibitor screening against PfMRS^cyt.

In an attempt to target parasite MRS^cyt, we modeled the synthetase domains of both PfMRSs. The levels of sequence identity with Thermos thermophilus MRS (PDB accession number 1A8H) were 37% for PfMRS^cyt and 28% for PfMRS^api. Therefore, T. thermophilus MRS was used as the template (see Fig. S3 in the supplemental material). A structure-based sequence alignment revealed several differences in the active site residues involved in binding the l-methionyl adenylate (MOD) complex. Furthermore, some of the MOD binding residues in PfMRS^cyt were dissimilar with respect to the human cytoplasmic counterpart. Therefore, we decided to screen the PfMRS^cyt active site for small-molecule compounds that could serve as inhibitors of this enzyme. We screened the whole active site pocket, including the MOD binding pocket and the auxiliary pocket in the l-Met binding site. The presence of an auxiliary pocket and an extended l-Met binding pocket was reported previously for Trypanosoma brucei MRS (42).

A small-molecule library (∼50,000 compounds; Specs) was screened and top-ranking ligands were selected based on docking scores and bonding interactions. A total of 40 compounds were procured for in vitro parasite growth inhibition assays, and several compounds showed submicromolar parasite growth inhibition (see Fig. 5; also see Table S1 in the supplemental material). Analogs of these compounds were purchased and further screened in parasite growth inhibition assays.

FIG 5.

Chemical structures of the compounds tested in growth inhibition assays with the 3D7 strain, with their EC₅₀s and predicted average logP (AlogP) and average logS (AlogS) values.

Parasite growth inhibition by REP3123 and REP8839.

REP3123 and REP8839 are known inhibitors of bacterial MRSs and show potent and selective activity, with minimal cytotoxicity for mammalian cells (43). REP3123 is active against various Gram-positive and Gram-negative bacteria, including resistant Clostridium stains and Staphylococcus strains. REP3123 kills Clostridium difficile with a K_i of 0.02 nM, compared with values of 0.017 nM and 0.08 nM against Staphylococcus aureus and Staphylococcus pneumoniae, respectively, in vitro. REP8839 inhibits MRS activity with a K_i of 10 pM against Staphylococcus aureus (23, 24). Therefore, we tested the effects of these MRS inhibitors on malaria parasite growth in vitro. The experiments were performed with the Plasmodium falciparum 3D7 strain at ∼1% parasitemia and 2% hematocrit levels. Both of the compounds were found to inhibit parasite growth and developmental progression. REP3123 inhibited parasite growth with an EC₅₀ of 144 nM, while REP8839 showed the same effect at 155 nM (Fig. 3A and B). The two compounds also seemed to affect parasite maturation in early and late developmental stages. The compounds inhibited parasite development in the early growth stages at approximately the IC₇₀ (500 nM) and prevented the formation of mature trophozoites and schizonts (Fig. 3C). Both compounds were also found to inhibit maturation in the late stages of parasite development. The inhibitors were used to treat late trophozoites or early schizonts (∼40 h postinvasion). Synchronized iRBCs at 0.5 to 1% parasitemia were incubated with 200 nM or 500 nM inhibitors for 24 h. REP3123 and REP8839 prevented schizont maturation and subsequent rupture by ∼65% and 58%, respectively, at 200 nM. The inhibition of maturation was improved at approximately the IC₇₀ (500 nM), at which both MRS inhibitors reduced parasite development by ∼75% (Fig. 4). Interestingly, both of these MRS inhibitors are in clinical trials for the treatment of bacterial infections (24), and their derivatives may be valuable starting points for targeting of PfMRSs.

FIG 3.

(A) Inhibition of parasite growth in the presence of REP3123 and REP8839. The assay was performed in 96-well plates, and growth was estimated with the SYBR green staining method. Percent reductions in parasite growth were plotted against inhibitor concentrations from 0.1 nM to 100 μM. REP3123 and REP8839 reduced parasite growth to one-half at 144 nM and 155 nM, respectively. (B) Chemical structures of REP3123 and REP8839. (C) Effects of REP3123 and REP8839 on parasite growth progression, studied by observing Giemsa-stained parasites under a microscope at ×100 magnification. The experiment was conducted with 4- to 8-h ring-stage parasites treated with inhibitors at 300 nM and monitored up to completion of one growth cycle. The two compounds repressed parasite development rapidly, as parasites were highly stressed and no mature schizonts were observed.

FIG 4.

Effects of REP3123 and REP8839 on schizont maturation after incubation with P. falciparum 3D7 at ∼40 h in the parasite life cycle. (A) Schizont maturation was inhibited by 65% and 58% using REP3123 and REP8839, respectively, at 200 nM each. (B) Both compounds inhibited schizont maturation by ∼75% at 500 nM. The schizonts were counted at 1% parasitemia and 2% hematocrit levels, on thick-smear glass slides. (C and D) Microscopic examinations of Giemsa-stained parasites treated with 200 nM (C) or 500 nM (D) showed growth arrest of parasites in the schizont stages.

Parasite growth inhibition using compounds screened in silico.

The compounds were dissolved in DMSO (not exceeding a final concentration of 1% in parasite cultures) and tested for activity against parasite cultures with the 3D7 strain. The experiments were performed in a 96-well format with a concentration range of 0.1 nM to 100 μM, in triplicate. Estimations of parasite growth inhibition were performed with SYBR green assays. Several compounds were found to affect parasite multiplication in vitro (Fig. 5; also see Table S1 in the supplemental material), and a few of those compounds showed EC₅₀s in the nanomolar or submicromolar range. The compounds C1, C2, and C3 and their analogs showed EC₅₀ values of <500 nM (Fig. 6A and D). The other compounds listed in Fig. 5 with inhibitory effects on parasite growth are analogs of these three compounds. The effects of these inhibitors were observed by labeling parasite DNA with SYBR green, after 48 h of incubation. All compounds killed parasites in the first cycle of erythrocytic development, suggesting that their likely target was PfMRS^cyt and not the apicoplastic PfMRS. We also tested the effects of the three selected inhibitors on parasite growth progression at IC₇₀ levels in a 48-h cycle. The three compounds were found to effectively inhibit parasite progression from the ring stage to the trophozoite stage, and this effect was observed in nearly 70 to 80% of iRBCs (Fig. 6B).

FIG 6.

Testing of hit compounds against malaria parasites. (A) In vitro susceptibility of Plasmodium falciparum 3D7 to compounds C1, C2, and C3. (B) Effects of these three compounds on parasite development after treatment at IC₇₀. The parasites were observed to shrink and could not form mature trophozoites and schizonts in the 48-h cycle. (C) Graphical representation of protein synthesis inhibition by C1, C2, and C3. The effects on protein synthesis were assessed by quantifying percent inhibition of incorporation of ³⁵S-labeled cysteine and methionine into newly synthesized proteins after starving the parasites by growing them in cysteine- and methionine-free medium. The x and y axes represent the concentrations of inhibitors and percent reduction in protein synthesis, respectively. Anisomycin and pyrimethamine were used as positive and negative controls, respectively. (D) Chemical structures of C1, C2, and C3.

Protein labeling with ³⁵S-labeled methionine and cysteine.

To investigate whether the selected compounds C1, C2, and C3 were specific for PfMRS^cyt and would thereby abrogate protein synthesis, we added ³⁵S-labeled methionine and cysteine to the culture medium, to label newly synthesized proteins. The incorporation of these labeled amino acids was analyzed after starving of the cells for 30 min in methionine- and cysteine-free RPMI 1640 medium, in the presence of inhibitors at concentrations of 0.1 nM to 100 μM. We used anisomycin, a widely used inhibitor of protein synthesis in eukaryotes (44), as a positive control and pyrimethamine, a known parasite growth inhibitor that specifically targets folic acid metabolism (45), as a negative control. Cells without inhibitors or drugs were considered to demonstrate 100% incorporation of ³⁵S-labeled amino acids. Treatment with the selected compounds quickly decreased protein synthesis in the parasites. The reductions in protein synthesis were ∼50% in 1 h with EC₅₀ levels of the three inhibitors (Fig. 6C). A comparison of EC₅₀s against parasite growth and deficits in protein synthesis with C1, C2, and C3 is presented in Table 1. These experiments suggest the likelihood of the selected compounds specifically targeting the protein translational machinery via inhibition of PfMRS^cyt.

TABLE 1.

EC₅₀ (growth inhibition) and IC₅₀ (protein synthesis inhibition) values for C1, C2, and C3 against Plasmodium falciparum 3D7 cultures

Inhibitor	EC50 (nM)	IC50 (nM)
C1	153	303
C2	111	177
C3	263	293
Anisomycin	98	74

PfMRS^cyt enzyme activity assays with inhibitors.

Parasites were lysed for 10 min with 0.65% NP-40 on ice and were centrifuged to separate the cytosolic fraction from other membrane-bound organelles (46). Inhibition of PfMRS^cyt aminoacylation activity was performed in the presence of identified inhibitors, using protein extracts as the source of PfMRS. DNA strands (see Fig. S4A in the supplemental material) encoding tRNA^Met (PlasmoDB accession number PF3D7_1339100) were synthesized for this experiment. The last two bases in the reverse primer were modified by incorporation of a methyl group at the 2′-OH group of the ribose sugar, and tRNA^Met was prepared using an in vitro transcription kit, as given in the reaction scheme (see Fig. S4B in the supplemental material). The malachite green assay (which quantifies the release of P_i as a result of the aminoacylation reaction) was performed in the presence or absence of tRNA^Met. A 4-fold increase in activity was observed with the addition of external tRNA (see Fig. S4C in the supplemental material). Inhibitors were incubated with parasite protein extracts for 15 min at 37°C in buffer A (30 mM HEPES [pH 7.5], 140 mM NaCl, 30 mM KCl, 40 mM MgCl₂, and 1 mM DTT). The reaction mixture was then supplemented with 50 μM methionine, 100 μM ATP, and 1 μM tRNA^Met for 10 min at 37°C. The quantified P_i release was plotted as a percentage with respect to values for the reaction performed in the absence of inhibitors. Significant decreases in activity were observed with increasing concentrations of inhibitors. REP3123 and REP8839 blocked PfMRS^cyt activity with EC₅₀ values of ∼146 and 281 nM, respectively (Fig. 7A). Nonlinear regression with the variable slope method, following the equation y = 100/[1 + 10^{(logIC₅₀ − x)×Hill slope}] (in which IC₅₀ is the 50% inhibitory concentration), was used to fit the curves. All graphs were plotted using GraphPad Prism 6 software.

FIG 7.

Growth inhibition by bacterial anti-MRS compounds. (A) The inhibition of MRS^cyt aminoacylation activity by REP3123 and REP8839 was assessed using parasite lysates as the source of MRS. The two compounds inhibited MRS^cyt activity at 146 nM and 281 nM, as determined with the malachite green assay. (B) The hit compounds C1, C2, and C3 inhibited MRS^cyt activity at 571, 154, and 614 nM, respectively.

The crystal structure of REP3123 in complex with Clostridium difficile MRS (47) reveals the molecular basis of enzyme inhibition, as this drug binds in the methionine binding pocket and into an adjacent nonsubstrate binding pocket. Of the hit compounds we studied, C2 inhibited the enzyme activity with an EC₅₀ of 154 nM (Fig. 7B), while C1 and C3 inhibited the enzyme activity with values of ∼571 and 614 nM, respectively, and are predicted to bind to the MRS active site in a manner similar to that of REP3123. The activity assays with parasite lysates showed no inhibition if the assays were carried out in the presence of 1 mM methionine, indicating possible competition with the methionine binding site (data not shown).

Binding modes of hit compounds in the MRS active site.

The three effective compounds (C1, C2, and C3) were redocked into the active site of MRS^cyt using Glide XP in the Schrödinger suite, to study their binding modes. Visual analysis suggested that these compounds are stabilized in the active site pocket by extensive hydrophobic interactions and they show distinct modes of binding. The binding pocket for C1 is located between the extended l-Met binding site and the auxiliary pocket instead of the ATP binding site (Fig. 8A). The naphthalene ring of compound C1 is stabilized by Phe482, His483, Tyr454, Ile231, Leu451, and Trp447, while its piperidine is stabilized by hydrogen bonding interactions with Asp270 (Fig. 8B). The binding mode of compound C2 is similar to methionyl adenylate (MOD) binding, and its benzimidazole ring mimics the adenine moiety of MOD, which is stabilized by hydrophobic interactions with His505, His42, Leu509, and Phe507 (Fig. 8C). The middle aromatic ring in compound C2 is stabilized by hydrophobic interactions with His483, Ile479, and Ile231. The halogen-substituted ring occupies a hydrophobic pocket formed by extended l-Met binding pocket residues Tyr454, Phe482, Trp447, Leu451, and Val446. The halogen (Cl and F) atoms in compound C2 are stabilized by hydrogen bonding interactions with Glu351 and Asp270, respectively (Fig. 8D). The halogen-substituted aromatic ring of compound C3 occupies an extended l-Met binding pocket similar to that of the halogen-substituted ring in compound C2 (Fig. 8E). The amino group-containing chain is stabilized by bifurcated hydrogen bonding interactions with Asp270 and Glu351 (Fig. 8F).

FIG 8.

(A, C, and E) Structural basis of PfMRS-inhibitor complexes. The binding modes of C1 (violet) (A), C2 (pink) (C), and C3 (yellow) (E) in the PfMRS^cyt active site are based on in silico docking routines. The selected docking poses of compounds were superimposed on the substrate l-methionyl adenylate (MOD) (green). The three compounds showed distinct modes of binding in the hydrophobic pocket, compared to the substrate MOD. C1 and C3 partially occupy the l-Met binding site as well as the auxiliary pocket. (B and F) LigPlot analysis, showing interactions of C1 (B), C2 (D), and C3 (F) within the active site of PfMRS^cyt.

DISCUSSION

To fulfill the requirements of protein synthesis, malaria parasite protein translation enzymes and their reaction substrates and products need to be distributed between apicoplasts, mitochondria, and cytoplasm (7, 8, 13, 14, 22). In P. falciparum, dual localization of single-copy aaRSs has been observed in the cases of Ala-RS, Gly-RS, Thr-RS, and Cys-RS (7, 8, 13, 14, 23). However, several P. falciparum aaRSs remain unannotated experimentally, in terms of their cellular localization. PfMRSs typify this problem, as the presence of two copies in the parasite genome alerted us to the possibility of dual compartmentalization (7, 8). Our results demonstrate that one copy, PfMRS^api, is targeted to the parasite apicoplasts, while the other, PfMRS^cyt, resides in the parasite cytoplasm in asexual blood-stage parasites. The homolog of PfMRS^cyt in humans is known to colocalize to the nucleolus and is involved in the synthesis of rRNA under proliferative conditions (in the presence of growth factors) (40). Such nuclear translocation of PfMRS^cyt was not observed in P. falciparum. Human MRS bears a nuclear targeting signal (KKKK) at the end of the C terminus that is lacking in PfMRS^cyt. Comparison of the C-terminal domains of Homo sapiens MRS and PfMRS^cyt shows an evolutionary divergence in structural folding. Mammalian MRSs possess a C-terminal WHEP domain that does not display any sequence similarity to PfMRS^cyt. We modeled the structure of the C-terminal domain of PfMRS^cyt, which resembles an oligonucleotide/oligosaccharide-binding (OB)-fold RNA binding domain similar to those found in prokaryotic MRSs (data not shown). The association of mammalian or higher eukaryotic MRSs with noncanonical functions in ribosome biogenesis and rRNA synthesis therefore seems lacking in Plasmodium falciparum. On the other hand, PfMRS^api bears an unusually long low-complexity region (LCR) in its synthetase domain (Fig. 1). The function of this asparagine-rich LCR in the CP1 region/Rossman fold remains unclear. Our failure to produce full-length recombinant PfMRSs precluded investigation of their enzymatic attributes. Interestingly, Plasmodium has four copies of tRNAs specific for MRSs. Two each of these tRNAs are predicted to be compartmentalized to parasite cytoplasm and apicoplasts, presumably serving as initiator and elongator tRNAs.

Comparison of the synthetase domains of MRSs across organisms of diverse taxa reveals an interesting evolutionary history. From our phylogenetic tree, it is apparent that PfMRS^cyt and PfMRS^api likely originated separately and are not the result of a gene duplication event. PfMRS^cyt and PfMRS^api show lower-plant-like and fungus-like origins, respectively, closely resembling enzymes from thermophilic and nitrogen-fixing bacteria. Since both of these proteins seem only distantly related to higher eukaryotes and are closer to primitive bacteria or a fungal clade, it is again clear that parasite evolution predated the origins of hominids. The sequences of and structural divergence between P. falciparum and human MRSs provide an opportunity for selective targeting of the former.

Several critical malaria parasite processes, including nucleosome assembly and lipid import in the malaria parasites, have been studied using structure-function techniques (29, 48–51). However, in these and similar cases, a lack of experimentally verified inhibitors has hampered drug development to date. This situation stands in contrast to the more comforting position for the aaRS family, for which many natural and synthetic inhibitors have been discovered and studied in the past several decades. Of all aaRSs, MRSs are the only family members that are involved in both initiation and elongation of polypeptide chains during protein synthesis. Previous studies have progressed toward the design of potent inhibitors against staphylococcal and Escherichia coli MRSs (52, 53). T. brucei MRS has also been explored as a drug target (54). In this study, we have shown potent inhibition of malaria parasite growth with the bacterial MRS inhibitors REP3123 and REP8839. Both compounds affect pathogen survival at multiple stages of parasite development. Inhibitors of protein synthesis are expected to show immediate effects on parasite survival, and REP3123 and REP8839 showed effects by impeding the transitions of developmental stages from the ring stage to the trophozoite stage and subsequent schizont maturation. In addition to REP3123 and REP8839, several new scaffolds identified here using in silico approaches seem promising starting points for derivatization and optimization. Of the identified compounds, C1, C2, and C3 targeted parasite development and maturation effectively. In summary, our current study provides small-molecule scaffolds of utility in specifically abrogating protein translation in malaria parasites. Our results serve to validate PfMRS as feasible targets for antimalarial drug development.