Antimalarial target vulnerability of the putative Plasmodium falciparum methionine synthase

Introduction

Malaria is a devastating parasitic disease. Between 2019 and 2021, an estimated additional 13.4 million cases were attributed to malaria control disruptions, chiefly the COVID-19 pandemic (World Health Organization, 2022). Infections with Plasmodium falciparum are responsible for most malaria cases, which are treated with artemisinin combination therapy (ACT). Artemisinin-resistant P. falciparum parasites are widespread throughout Southeast Asia, and partially resistant parasites have evolved independently in parts of Africa (World Health Organization, 2022), prompting the need for antimalarials against novel targets. Given that the current standard of care for malaria is a three-day ACT regimen, any new antimalarial must be similarly fast-acting. Fast-acting antimalarials inhibit the functions of essential, vulnerable targets (Forte et al., 2021). More than half of the protein-coding genes in P. falciparum are annotated as essential based on the criterion that disruption of the protein-coding region by transposon insertion is not tolerated (Zhang et al., 2018). However, the absence of transposon insertion is not definitive annotation of essentiality because of local variation in transposon insertion efficiency. One possible example of this scenario is the PF3D7_1311700 (cyt c-2) gene, which lacks transposon insertions but was shown to be dispensable by targeted knockout (Espino-Sanchez et al., 2023).

Identifying vulnerable targets is challenging because some essential genes are non-vulnerable, including targets for which specific inhibitors with antimalarial activity are available, such as deoxyhypusine synthase (Aroonsri et al., 2019), Niemann-Pick Type C1-Related protein (Istvan et al., 2019), and plasmepsin V (Sleebs et al., 2014; Polino et al., 2020). Partial loss-of-function (LOF) mutants of these genes have latent proliferation defects, and in the case of plasmepsin V, a defect was observed after 96 h only in LOF mutants with greater than 90% knockdown of the wild-type level of expression (Polino et al., 2020). Antimalarial discovery efforts might be better focused on directly identifying vulnerable targets with alternative assays rather than proving essentiality, which may require laborious monitoring of proliferation in LOF mutants with varying degrees of knockdown and/or conditional knockout mutants over extended periods for non-vulnerable targets. We propose to define vulnerable targets for the purpose of assay development as genes for which a partial LOF mutant (with significant knockdown of about 50 to 90% reduction of the wild-type expression level) has an acute proliferation defect observable at 96 h or sooner.

In this study, we developed a target vulnerability assay for LOF mutants created with the glmS ribozyme tool (Prommana et al., 2013). To apply the tool, the gene of interest must be modified by DNA transfection. Proof of concept for the tool was previously demonstrated for the dihydrofolate reductase-thymidylate synthase (DHFR-TS) gene modified by single-crossover integration of transfected circular DNA (Prommana et al., 2013; Aroonsri et al., 2016). This transfection method has been superseded by the more efficient CRISPR-Cas9 gene editing system (Ghorbal et al., 2014). We edited the DHFR-TS gene to assess whether the glmS ribozyme was functional in the context of an edited gene, particularly one with the 3′ coding region replaced with artificial recodonized sequence as a consequence of gene editing. In addition, the LOF mutant obtained with an edited DHFR-TS gene was used to validate the target vulnerability assay. DHFR-TS is a known vulnerable target that is inhibited by antifolate drugs such as pyrimethamine and P218 (Yuthavong et al., 2012), and LOF mutants of DHFR-TS show significant proliferation defects at 72 h or earlier (Prommana et al., 2013; Aroonsri et al., 2019).

To search for new antimalarial targets, we propose testing LOF mutants of unexplored genes in target vulnerability assay. Methionine metabolic pathways contain several unexplored antimalarial targets. Methionine is an essential amino acid in P. falciparum that must be obtained from salvage because proliferation is markedly reduced in culture media lacking methionine (Divo et al., 1985; Marreiros et al., 2023). P. falciparum salvages methionine via the new permeation pathway and a neutral amino acid transporter (Cobbold, Martin & Kirk, 2011). In addition to protein synthesis, methionine is used as a cofactor to produce the essential metabolite S-adenosyl-l-methionine (SAM) by the SAM synthase (SAMS) enzyme. The methyl group from SAM is transferred to various acceptors by methyltransferases to form S-adenosyl-l-homocysteine (SAH). SAH is hydrolyzed to adenosine and l-homocysteine (Hcy) by the highly conserved enzyme S-adenosyl-l-homocysteine hydrolase (Tanaka et al., 2004). P. falciparum lacks key enzymes in the reverse transsulfuration pathway for the conversion of Hcy to cysteine. Consequently, Hcy accumulates and is effluxed from the parasite during intra-erythrocytic growth (Beri et al., 2017). However, excess Hcy is deleterious and can trigger gametocytogenesis in P. falciparum (Beri et al., 2017). In addition to efflux for the control of Hcy, Hcy can be converted to methionine by the action of the methionine synthase enzyme (MS, 5-methyl tetrahydrofolate homocysteine methyltransferase, EC.2.1.1.13). MS uses 5-methyltetrahydrofolate (5-mTHF) as a cofactor to generate tetrahydrofolate (THF) as a by-product (Banerjee & Matthews, 1990). P. falciparum can obtain 5-mTHF from salvage or by synthesis from 5,10 methylenetetrahydrofolate via a methylenetetrahydrofolate reductase enzyme (Asawamahasakda & Yuthavong, 1993).

Cobalamin-dependent MS enzymatic activity was reported previously in P. falciparum intra-erythrocytic stage protein extract. Nitrous oxide inhibits the activity of this enzyme and parasite proliferation, suggesting that parasite cobalamin-dependent MS may be a potential antimalarial target (Krungkrai, Webster & Yuthavong, 1989). However, the gene encoding the parasite MS enzyme has not yet been identified in the P. falciparum 3D7 genome (Müller & Hyde, 2013). We identified a candidate P. falciparum MS gene and created a LOF mutant for assessing the vulnerability of this target.

Materials & Methods

Results

We hypothesized that P. falciparum possesses an MS-encoding gene. We searched for P. falciparum 3D7 proteins with an Hcy-binding domain in the InterPro database (Paysan-Lafosse et al., 2023), since all MS enzymes possess an N-terminal Hcy-binding domain (Matthews, Sheppard & Goulding, 1998). The PF3D7_1233700 gene product is the only P. falciparum 3D7 protein with an InterPro-annotated Hcy-binding domain. PF3D7_1233700 is annotated in OrthoDB (Kuznetsov et al., 2023) as a member of the syntenic Plasmodium orthologous group 1324at5820 (Hcy-binding domain). PF3D7_1233700 and orthologous (single-copy) proteins from other human-infective Plasmodium spp. showed low (≈20%) identity with human and Escherichia coli cobalamin-dependent MS (Figs. S1 & S2). We searched for candidate MS using a three-dimensional structural superposition-based approach (Foldseek), which is more sensitive for identifying homologous proteins (Van Kempen et al., 2023). Full-length cobalamin-dependent MS comprises four modules (N-terminal Hcy-binding, 5-mTHF-binding, cobalamin-binding, and C-terminal adenosylmethionine-binding/reactivation; Matthews, Sheppard & Goulding, 1998). The X-ray structures of cobalamin-dependent MS protein fragments from different species have been determined, although the structure of the N-terminal fragment containing Hcy- and 5-mTHF substrate binding domains is available only for Thermotoga maritima (Evans et al., 2004). We selected the structure of the T. maritima cobalamin-dependent MS fragment co-complexed with Hcy and 5-mTHF as a query for Foldseek. The top-ranked Foldseek hit to P. falciparum 3D7 proteins was PF3D7_1233700. Lower-ranked hits had well-defined annotations unrelated to methionine metabolism, suggesting incidental structural similarity of protein folds with functions unrelated to MS (Table S2). Notwithstanding the possibility of even more diverged proteins not detectable by sequence- or structure-based homology, PF3D7_1233700 is putatively assigned as P. falciparum MS. However, definitive annotation requires direct functional data, e.g., biochemical assay of the purified PF3D7_1233700 protein for MS activity.

We edited the PF3D7_1233700 and PF3D7_0412700 (DHFR-TS) genes, placing them under the control of the glmS ribozyme (Prommana et al., 2013). The edited DHFR-TS and MS genes were confirmed by PCR genotypic assays in clonal lines of transgenic parasites (Figs. 1 & 2). One clonal line of each edited gene (designated as DHFR-TS_glmS and MS_glmS, respectively) was selected for phenotypic analysis. The expressions of target proteins in transgenic parasites were assessed by western blotting of synchronized transgenic parasites cultured for 24 h in the presence or absence of GlcN. Protein species of sizes expected for modified MS (72.7 kDa) and DHFR-TS (75.5 kDa) were detected in transgenic parasites (Fig. S3). GlcN treatment caused significant reductions in MS and DHFR-TS protein expression (DHFR-TS mean = 40%, P-adjusted = 0.004; MS mean = 51%, P-adjusted = 0.03; Fig. 3A). There was no significant difference in the mean % target protein (GlcN treated relative to untreated control) level between DHFR-TS and MS (P = 0.34).

Next, we assessed the consequences of target protein knockdown in transgenic parasites with respect to proliferation in target vulnerability assay. In previous studies of the acute effect of glmS-ribozyme mediated target knockdown on parasite proliferation, treatment was performed for up to 72 h in which GlcN has a minor inhibitory effect on wild-type strains (Prommana et al., 2013). By extending the GlcN treatment to 96 h, greater than 50% inhibition of the 3D7 wild-type strain was observed at the highest concentrations such that we could determine the EC₅₀ (8.0 mM; 95% confidence intervals 5.9 to 10.0 mM). We posited that knockdown of vulnerable target gene expression by the action of the glmS ribozyme enhances the proliferation defect caused by GlcN treatment itself over 96 h manifested as a significantly lower EC₅₀ compared with 3D7 wild-type. The EC₅₀ of the DHFR-TS_glmS parasite, but not that of the MS_glmS parasite, was significantly different from that of the 3D7 wild-type strain (Fig. 3B). Hence, reduction in DHFR-TS, but not MS expression affected parasite sensitivity to GlcN. Based on these results, DHFR-TS is defined as a vulnerable antimalarial target as expected. In contrast, MS is a non-vulnerable target.

Discussion

PF3D7_1233700 was identified as the only candidate gene encoding MS from a bioinformatic search of the P. falciparum 3D7 genome, suggesting that the parasite possesses a single MS gene. However, P. falciparum MS is not a vulnerable antimalarial target, in contrast to the vulnerability of the downstream SAMS enzyme in the P. falciparum methionine pathway (Musabyimana et al., 2022). Salvage is the major source of methionine substrate for P. falciparum SAMS because reducing exogenous methionine leads to a concomitant decrease of SAM (Harris et al., 2023). Hence, methionine synthesized by MS is of minor importance for intra-erythrocytic proliferation under standard in vitro culture conditions. However, methionine synthesis may be more important for P. falciparum proliferation in natural infections, since the methionine concentration in human serum (Barić et al., 2004) is approximately two to seven times lower than that of parasite culture medium.

The other roles of P. falciparum MS besides methionine synthesis should be considered to explain the conservation of MS in Plasmodium and its tentative annotation of essentiality in P. falciparum based on the absence of transposon insertions in the encoding gene (Zhang et al., 2018). It should be noted that definitive proof of essentiality requires demonstration of a proliferation defect from more complete knockdown using a different tool (e.g., the TetR-DOZI system with 5′ and 3′ aptamers installed at the target gene (Polino et al., 2020)) or from conditional gene knockout mutagenesis. Essential genes are expressed through the life cycle, and P. falciparum MS is detectably expressed by data-independent proteomics throughout the intraerythrocytic stages at a level approximately 40–1000-fold lower than that of DHFR-TS (Siddiqui et al., 2022). Moreover, single-cell RNA sequencing data indicate that P. falciparum MS is expressed during mosquito stages (Real et al., 2021), and the P. vivax MS ortholog PVP01_1451800 is expressed in liver stages (Mancio-Silva et al., 2022).

Although malaria parasites possess a mechanism for the efflux of Hcy (Beri et al., 2017), lack of MS function may lead to increased levels of Hcy and redox stress, which may be important for sporogonic development in the mosquito vector when Plasmodium is more dependent on glutathione (Vega-Rodríguez et al., 2009) and α-lipoic acid (Biddau et al., 2021) to mitigate oxidative stress. The conversion of 5-mTHF to THF by MS may be important for recycling folate required for other enzymatic reactions, particularly during the developmental stages with the greatest folate demand. Plasmodium can salvage folates and the folate precursor para-aminobenzoic acid (p ABA), but the levels of folates are too low, or are not in a form capable of being efficiently transported to support intra-erythrocytic development in the absence of de novo synthesis (Salcedo-Sora & Ward, 2013).

P. berghei lacking the p ABA synthetic enzyme aminodeoxychorismate synthase cannot develop in p ABA-deficient medium during intraerythrocytic stages; however, the growth of these mutant parasites is unaffected in p ABA-deficient medium during liver stages (Matz et al., 2019). Despite an inefficient folate transport system for the uptake of 5-mTHF (Salcedo-Sora & Ward, 2013), elevated levels of 5-mTHF in the liver may drive its accumulation and conversion to THF by Plasmodium MS, such that the parasite is less reliant on de novo folate synthesis during this stage of the life cycle. A knockout mutant of the P. berghei orthologous MS gene (PBANKA_1448300) shows no growth defect during intraerythrocytic stages (Bushell et al., 2017), but growth of the mutant is reduced during the transition from the sporozoite (through the liver) to the blood stage (Stanway et al., 2019). Although these data suggest a non-essential role of Plasmodium MS, dispensability in P. falciparum cannot be extrapolated from knockout data in P. berghei because of species-specific differences in Plasmodium host cell tropism. P. berghei preferentially invades reticulocytes, whereas P. falciparum invades mature erythrocytes. The reticulocyte milieu has a greater metabolic complexity than that of erythrocytes, which can support the growth of P. berghei parasites with knockouts of genes functioning in the intermediary carbon metabolic pathway, pyrimidine metabolism, and glutathione biosynthesis that are essential in P. falciparum (Srivastava et al., 2015).

Conclusions

The finding that MS is a non-vulnerable antimalarial target raises the question of what other enzymes in the Plasmodium parasite methionine pathway and other pathways related to folate metabolism (Müller & Hyde, 2013) are also non-vulnerable targets. This could be tested by target vulnerability assay of LOF mutants for other genes annotated as essential. Although the role of MS as an antimalarial target is deprioritized, it would be interesting to test whether the roles of P. falciparum MS in Hcy metabolism and folate recycling are more important in mosquito and liver stages.

Supplemental Information