Mixed Alkyl/Aryl Phosphonates Identify Metabolic Serine Hydrolases as Antimalarial Targets

Summary

Malaria, caused by Plasmodium falciparum, remains a significant health burden. One major barrier for developing anti-malarial drugs is the ability of the parasite to rapidly generate resistance. We previously demonstrated that Salinipostin A (SalA), a natural product, potently kills parasites by inhibiting multiple lipid metabolizing serine hydrolases, a mechanism which results in a low propensity for resistance. Given the difficulty of employing natural products as therapeutic agents, we synthesized a small library of lipidic mixed alkyl/aryl phosphonates as bioisosteres of SalA. Two constitutional isomers exhibited divergent anti-parasitic potencies which enabled the identification of therapeutically relevant targets. The active compound kills parasites through a mechanism that is distinct from both SalA and the pan-lipase inhibitor, Orlistat and shows synergistic killing with Olistat. Our compound induces only weak resistance, attributable to mutations in a single protein involved in multidrug resistance. These data suggest that mixed alkyl/aryl phosphonates are promising, synthetically tractable anti-malarials.

Graphical Abstract

eTOC blurb

The human parasite and cause of malaria, Plasmodium falciparum, requires lipids for survival inside the host. Bennett et al. identify mixed alky/aryl phosphonates that block parasite growth by covalent targeting multiple serine hydrolases involved in lipid metabolism. These compounds have a low propensity to induce resistance mechanisms.

Introduction

Plasmodium falciparum (P. falciparum) is an intracellular parasite responsible for the most severe cases of malaria. While significant progress has been made in reducing malaria cases and fatalities through public health efforts and new treatment strategies over the past two decades, the global malaria burden currently remains on the rise¹. In 2021 alone, there were over 247 million reported cases and more than 600,000 deaths.² Moreover, there has been a continued spread of partial resistance to frontline treatments such as artemisinin (ART) and resistance to other drugs used in artemisinin-based combination therapies (ACTs), underscoring the need to develop agents with novel mechanisms of action that do not induce rapid resistance in the parasite.^3–7

Natural products have long served as valuable sources for the discovery of potential drugs in P. falciparum.⁸ Salinipostin A (SalA) belongs to a group of bioactive natural compounds known as cyclophostins, characterized by the presence of a bicyclic enolphosphate structure.⁸ Notably, these natural products potently inhibit the human serine hydrolase acetylcholinesterase, have antibacterial activity against mycobacterium, and have antimalarial properties.^9–11 We posited that cyclophostins, including SalA, may act via the enolphosphate electrophile to inhibit enzymes. Using an activity-based probe analog of SalA we identified a series of α/β serine hydrolases whose inhibition is responsible for the antimalarial effects of the natural product.¹² Furthermore, we found that SalA exhibits a low propensity to induce resistance, a trait shared with artemisinin (ART) and its derivatives. The ability of compounds such as SalA and ART to target multiple proteins likely contributes to the low level of resistance generated.^13–15 This underscores the potential of targeting multiple serine hydrolases within P. falciparum as a potentially valuable therapeutic strategy. However, due to the challenges of isolation and synthesis of SalA, it remains a non-viable lead molecule.

Serine hydrolases are a large superfamily of enzymes that use a catalytic serine to hydrolyze various macromolecules, such as carbohydrates, proteins, and lipids.¹⁶ In humans, serine hydrolases have been successfully targeted for diabetes and dementia with many candidates for analgesia also in the clinic.¹⁷ Nevertheless, serine hydrolases remain underexplored in P. falciparum. Serine hydrolases commonly play important roles in lipid metabolism and membrane homeostasis. Throughout the erythrocytic lifecycle of the parasite, the plasma membrane is continuously remodeled to enable the growth of the parasite.¹⁸ Moreover, P. falciparum is unable to synthesize all of its necessary lipids and must scavenge various lipids from the host.^19,20 Therefore, the parasite likely dynamically regulates the expression and activity of metabolically important enzymes at various points in the lifecycle to ensure proper membrane composition and lipid homeostasis. The activity of serine hydrolases in parasites was first investigated with a broad spectrum fluorophosphonate-desthiobiotin probe which enabled the identification of over 20 active serine hydrolases in the lysate of late-stage schizont parasites.²¹ More recently, a fluorphosphonate probe was used to identify serine hydrolases in live parasites at different stages within the life cycle, clearly demonstrating the dynamic regulation of lipolytic serine hydrolases in P. falciparum.²²

A wide range of strategies and warheads have been employed for the development of covalent inhibitors targeting serine hydrolases.²³ Phosphonates are a valuable bioisostere because they can mimic phosphate groups, while maintaining reactivity with catalytic serine residues.²⁴ In fact a range of diphenyl phosphonates have been developed to inhibit serine proteases.^25–27 Therefore, we postulated that phosphonates could be used as a replacement for the enolphosphate group of SalA (Figure 1A). However, only the broadly reactive fluorophosphonate has been used for activity-based protein profiling (ABPP) and the characterization of serine hydrolases within Plasmodium.²⁸ The mixed alkyl/aryl phosphonates are a relatively recently reported class of phosphonate-based serine hydrolase inhibitors.²⁹ This class of serine-specific electrophiles has great potential for use in serine hydrolase inhibitors because reactivity can be tuned through the substitution of various phenol leaving groups appended to the phosphonate.²⁹ Similar to the widely used carbamates and ureas, the leaving group of the electrophile can be optimized to tune both the selectivity and potency of inhibitors toward specific serine hydrolases.^30,31

Figure 1: Structure of Salinipostin A and Synthesized Phosphonates.

(A) Structure of Salinipostin A and General Mixed Alkyl/Aryl Phosphonate. Salinipostin A (left) is drawn with electrophilic enolphosphate colored red and the mixed alkyl/aryl phosphonate (right) with electrophilic phosphonate group in red.

(B) Synthesis of mixed alkyl/aryl phosphonates and structures of molecules. PPh₃, triphenyl phosphine; I₂, iodine; DCM, dichloromethane; P(OEt)₃, triethyl phosphite; Tf₂O, triflate anhydride; pyr., pyridine; ROH, phenol.

(C) Structures of Short Chain Hexyne-based Phosphonates

(D)Structures of Pentadecane-based Phosphonates

We thus set out to develop mixed alkyl/aryl phosphonates that could recapitulate the antimalarial properties of SalA but that would be more synthetically tractable. We used various phenol leaving groups and different alkyl chain lengths to investigate the structure-activity relationship of our molecules on purified enzymes and their phenotypic effects on P. falciparum parasites. Interestingly, we identified a pair of mixed aryl/alkyl phosphonates that are constitutional isomers with similar biochemical potency against purified serine hydrolases but dramatically different potencies when used on intact parasites. Using these compounds, we performed chemoproteomics to identify a serine hydrolase, abH112, as the primary anti-malarial target of the compound with potent anti-malarial activity. Interestingly, conditional knockdown of abH112 did not result in growth defects suggesting that the killing effect of our compound is likely mediated by inhibition of multiple targets which must include abH112. Phenotypic analysis and metabolic profiling studies demonstrated that our mixed phosphonate compound kills parasites through a mechanism that is distinct from that of SalA or the pan-lipase inhibitor Orlistat. Drug pressure selections demonstrated that the antimalarial mixed phosphonate generated resistant parasites with only minor shifts in sensitivity to the compound that were mediated by the same drug resistance protein identified in SalA resistant parasites. This result is consistent with a mechanism of killing involving multiple targets and provides further support for targeting serine hydrolases for anti-malarial therapeutics.

Results

Short Chain Mixed Alkyl/Aryl Phosphonates are potent inhibitors of PfMAGL

While synthesis routes have been reported for salinipostins and other similar natural products, synthesis of the bicyclic enolphosphate group remains the most significant challenge of the total synthesis.³² Additionally, limited product yield makes salinipostins unlikely to be used as therapeutics. We therefore developed a synthetic strategy for the rapid production of mixed alkyl/aryl phosphonates that mimic the main core functional elements of SalA (Figure 1B). This route utilizes an Appel reaction to convert the terminal hydroxyl of an alkyl chain, into an iodine which can then be converted to the corresponding diethyl phosphonate using an Arbuzov reaction. Subsequent aryloxylation using triflate anhydride and pyridine enable us to develop a library of racemic mixed alkyl/aryl phosphonates (Figure 1C).³³ Our initial library comprised 20 hexyne-based mixed phosphonates with phenolic leaving groups varying in their stereoelectronics.

To characterize the compounds, we conducted a biochemical screen against Plasmodium falciparum monoacyl glycerol lipase (PfMAGL), a serine hydrolase identified as a primary target of SalA. PfMAGL is annotated as an essential protein based on a piggyBac transposon saturation screen of essential genes in P. falciparum, and we previously showed that the biochemical inhibition of PfMAGL correlates with potency against intraerythrocytic parasites.^12,34 We screened our library at high (25 µM) and medium (2.5 µM) concentrations and observed that phosphonates with electron-withdrawing substitutions on the phenols effectively inhibited the enzyme (with the exception of compound 13) whereas those with either electron donating or sterically bulky groups did not (Figure S1A). Compound 2, containing a p-sulfonyl fluoride substitution, had an IC₅₀ of 20 nM and was the most potent inhibitor of PfMAGL in our initial set of compounds (Figure S1B). We obtained similar results screening against the human homolog, hMAGL, with only the phosphonates bearing an electron-withdrawing group exhibiting inhibitory effects and compound 2 showing the highest potency against the human enzyme (Figure S1C; S1D).

Following biochemical characterization of our compounds, we assessed their effects against blood stage parasites in a 72-hour parasite growth assay (Figure S1E). Only compound 2 had significant antiparasitic effects, killing over 50% of the parasites at the 10 µM concentration. In contrast, most compounds, including those with nanomolar inhibition of PfMAGL, showed limited to no killing at both 10 µM and 1 µM, suggesting that inhibition of PfMAGL may not be sufficient for parasite killing. Compound 2 exhibited moderate potency as an inhibitor of parasite growth, with an EC₅₀ of 1.6 µM (Figure S1F). However, this compound is problematic due to the fact that it contains an electrophilic p-sulfonyl fluoride that can react with nucleophilic amino acid residues other than the catalytic serine of the intended target.^35,36 Thus, some of the potent anti-parasitic activity may be the result of off target inhibition or general toxicity from the reactive p-sulfonyl fluoride group.

Long chain lipids containing the mixed phosphonate electrophile are potent inhibitors of PfMAGL and parasite growth

Since our initial set of molecules did not exhibit potent antiparasitic properties, we synthesized a second set of mixed alkyl/aryl phosphonates with a longer alkyl chain on the phosphorus to better mimic the structure of SalA. We chose a C15 lipid to match the length of the western chain on SalA. The synthesis of long chain phosphonates proved to be more challenging, but we were able to synthesize three C15 mixed phosphonates (Figure 1D; Figure 2A). We first tested these compounds for direct inhibition of recombinant PfMAGL. Two constitutional isomers, 21, with a 2,6-difluorophenol, and 22, with a 3,5-difluorophenol, were picomolar inhibitors of the enzyme (Figure 2B). Compound 23, with a phenol leaving group, had only low micromolar potency, and its short-chain counterpart 11 showed no inhibition at concentrations up to 25 µM (Figure S2A), suggesting that the combination of warhead and lipid chain length contribute to the potency of the compounds. We tested the three compounds for inhibition of hMAGL and found similar trends in potency (Figure S2B).

Figure 2: 3,5 difluorophenol Substituted Long Chain Phosphonate is a potent inhibitor of parasite growth.

(A) Strucutre of Salinipostin A and constitutional isomers of long chain mixed alkyl/aryl phosphonates.

(B) IC₅₀ curves of short chain mixed alkyl/aryl phosphonates for inhibition of 4-methylumbelliferyl caprylate processing by recombinant PfMAGL. (mean ± SD, N,n = 2,3).

(C) EC₅₀ values of Compound 21 and 22 in a 72 h treatment of synchronized ring stage W2 parasites with error bars, SEM. (mean ± SD, n = 4 from 3 independent experiments).

(D) Rate of kill of Compound 22 measured against compared against known fast (chloroquine), medium (pyrimethamine), and slow (atovaquone) acting inhibitors of parasite growth in which compounds were dosed at their EC₅₀ and parasite viability was monitored over time and ± SEM. N,n = 6,2.

(E) HFF Toxicity for alkyl cyclic peptide inhibitor and linear counterparts. Points are plotted as mean ± SEM values of each inhibitor for HFF growth. Data were generated with 72 h assays (N, n = 2,3).

After characterizing the biochemical activities of our compounds, we tested the long chain phosphonates in the 72-hour parasite growth assay. Surprisingly, we observed that compound 22 was an extremely potent inhibitor of parasite growth (EC₅₀ = 188 nM), while the constitutional isomer 21, showed no killing at concentrations up to 20 µM (Figure 2C). These results further suggest that even though PfMAGL may be an essential protein, inhibition of this enzyme alone does not seem to be sufficient to kill parasites. In addition, we reasoned that these two isomers would be ideal tools for use in comparative chemoproteomic experiments to identify the protein target(s) responsible for the observed anti-parasitic activity. In fact, this strategy of using a paired set of isomers of an electrophilic small molecule has recently been demonstrated to be an effective strategy to find phenotypically relevant targets when compounds react with multiple proteins.³⁷

To further investigate the mechanism of the antiparasitic activity of 22, we used a rate of kill assay. These results confirmed that 22 kills parasites at a rate that is faster than pyrimethamine, but not as fast as chloroquine, suggesting that our inhibitor has the potential to quickly clear an infection (Figure 2D).³⁸ Moreover, when parasites were treated with compound 22 for 1 hour followed by a washout and then continued growth for 72 hours, we observed only a ten-fold drop in potency (EC₅₀ = 1.33 µM) compared to continuous treatment, suggesting rapid and prolonged binding to targets consistent with a covalent mechanism of action (Figure S2C). Lastly, we found that, at concentrations up to 50 µM, we saw no impact on Human Foreskin Fibroblast (HFF) viability for either compound 21 or 22 (Figure 2E). This result is consistent with studies of SalA and other human serine hydrolase inhibitors, which have no cytotoxic effects on mammalian cells.

Activity-Based Probes Identify Differentially Targeted Serine Hydrolases

To perform a chemoproteomic identification of the targets of our mixed phosphonates, we synthesized terminal alkyne bearing derivatives that would enable click chemistry to isolate labeled targets. Using a synthetic route based on the previous synthesis of light activated probes of ceramides (Supplemental Scheme 1), we synthesized activity-based probe versions of compounds 21 and 22 (21-alk and 22-alk; Figure 3A).³⁹ Both compounds had similar potencies against the purified PfMAGL and hMAGL compared to the original non-alkynlyated inhibitors suggesting that addition of the tag did not impact target binding (Figure S3A). We also confirmed that probe 21-alk remained inactive in the parasite killing assay, and compound 22-alk remained active with only a modest loss in potency (EC₅₀ = 1.04 µM) compared to the parent compound (Figure S3B). This change in potency upon addition of the alkyne resulted in a decrease in potency of a magnitude that was similar to what was observed when an alkyne tag was added to the Salinipostins.¹¹ Regardless of the drop in potency, compound 22-alk remained a potent inhibitor of P. falicparum growth and retained the fast rate of kill observed for the parent compound (Figure S3C).

Figure 3: Chemoproteomics identifies distinct serine hydrolases between active and inactive difluorophenol isomers.

(A) Structures of Activity-Based Probes used for chemoproteomics.

(B) Workflow for Competitive ABPP experiment. Infected red bloods are either pretreated with DMSO or an inhibitor (pink) and followed by treatment with an activity-based probe (yellow), before being submitted through typical proteomics workflows.

(C) Volcano plots of Activity-based probe 21-alk targets in P. falciparum. The x-axis shows the logarithm values of between 21-alk-treated and DMSO vehicle-treated parasites. The statistical significance was determined without correcting for multiple comparisons in triplicates and the y-value in the volcano plot is the negative logarithm of the p value. Significantly enriched serine hydrolases are denoted in red for parasite enzymes and blue for human enzymes.

(D) Volcano plots of Activity-based probe 22-alk targets in P. falciparum. The x-axis shows the logarithm values of between 22-alk-treated and DMSO vehicle-treated parasites.

(E) Venn Diagram showing overlap of significantly enriched serine hydrolases between the inactive probe 21-alk red, the active probe 22-alk, and the natural product Salinipostin A.

To identify the target(s) responsible for the anti-parasitic effects of 22, we performed a competitive ABPP proteomic experiment in infected red blood cells by pretreating with either vehicle or 21 or 22, before labelling with 21-alk or 22-alk (Figure 3B). When labeling with the activity-based probe 22-alk, we identified seven putative serine hydrolases and one uncharacterized protein (PF3D7_0619300) as being significantly enriched (Figure 3C). The biologically inactive probe 21-alk, significantly enriched four of those serine hydrolases (Figure 3D). Both probes enriched two human serine hydrolases, Neutral cholesterol ester hydrolase 1 and acetylcholine esterase both of which were previously identified in ABPP experiments and whose inhibition has little effect on parasite growth.²² Moreover, both probes labeled PfMAGL (PF3D7_1038900) and the lysophospholipase PfPARE (PF3D7_0709700). PfPARE was previously identified as an esterase that activates the antimalarial drug pepstatin.⁴⁰ However, its esterase activity is dispensable for parasite growth. This finding coupled with the fact that it was targeted by both the active and inactive probes suggest that it is likely not a relevant target for the observed anti-parasitic activity of our compounds (Figure 3E). PfMAGL, on the other hand, has been previously identified as an essential protein in a PiggyBac transposon saturation screen, and we demonstrated that the biochemical inhibition of PfMAGL is associated with the inhibition of parasite growth when other serine hydrolases are also inhibited.^12,34 Interestingly, labeling of PfMAGL was competed by the inactive compound 21 while active inhibitor 22 did not compete, suggesting 22 has stronger affinity for other serine hydrolases (Figure S3 E,F). This suggests that inhibition of PfMAGL alone is not sufficient for parasite killing. Two serine hydrolases (PF3D7_0805000 and PF3D7_1116100) enriched by 22-alk were also identified as targets of SalA but neither were competed by 22 nor were they annotated as essential in the PiggyBac screen (Figure 3E; Table 1).¹² Of the proteins competed by 22, only PF3D7_1120400, previously identified as AbH112, was annotated as essential, was competed only by the active inhibitor 22, and was not identified as a target of SalA (Table 1; Figure S3G). Therefore, our proteomic data suggests that abH112 is likely the primary target of 22 that, together with other essential serine hydrolase targets such as PfMAGL, mediates the antiparasitic activity of this mixed akyl/aryl phosphonate compound.

Table 1.

Targets of Activity-Based Probes and Competition Experiments

Protein ID	Annotation	Essential	Competed by 21	Competed by 22	Salinipostin A
PF3D7_1038900	Esterase (PfMAGL)	Yes	Yes	No	Yes
PF3D7_0805000	α/β hydrolase	No	No	No	Yes
PF3D7_0709700	lysophospholipase (PfPARE)	No	Yes	Yes	Yes
PF3D7_1120400	α/β hydrolase (abH12)	Yes	No	Yes	No
PF3D7_0619300	unknown function	No	No	Yes	No
PF3D7_1116100	serine esterase	No	No	No	Yes
NCEH 1	Neutral cholesterol ester hydrolase 1	N/A	Yes	Yes	N/A
AChE	Acetylcholinesterase	N/A	No	No	N/A

AbH112 is predicted to be an essential gene but knock down of the protein does not impact parasite growth

To gain insight into the possible function of abH112, we performed bioinformatic analysis of the enzyme. We compiled a list of all the proteins in human, P. falciparum, and the related Toxoplasma gondii proteomes that share the Protein family (PFam) domain of Hydrolase 4 (PF12146) with abH112.⁴¹ Using the Clustal Omega multiple sequence alignment algorithm, we performed phylogenetic analysis of the putative serine hydrolase protein sequences. (Figure 4A; Figure S4A).⁴² The two proteins with the most homology to abH112 are the uncharacterized Toxoplasma gondii protein TGVAND_223510 and the human serine hydrolase ABHD17A. The human ABHD17A is of particular interest because it was recently identified as a depalmitoylase for nRAS and we have previously shown that depalmitoylases are viable drug targets in Toxoplasma gondii.^43–45 However, no depalmitoylases have been identified in P. falciparum. We overlayed the AlphaFold predicted structures for both ABHD17A and abH112 and found that the catalytic triad aspartate, histidine, and serine are aligned (Figure 4B).⁴⁶ Furthermore, both proteins show good alignment of lipid binding motifs, the catalytic triad, as well as the GxSxG serine hydrolase motif (Figure 4C). It should be noted that abH112 contains a GHSxG serine hydrolase motif which has been found in enzymes with dual serine hydrolase and acetyl transferase activity, suggesting a possible function beyond hydrolysis.⁴⁷ To assess whether abH112 has a function in regulating global palmitoylation, we incubated infected red blood cells for 16 hours with 17-ODYA to incorporate the clickable analog of palmitate into the proteome. We then treated the parasites with either vehicle, compound 21, 22, or the palmitoyltransferase inhibitor 2-bromopalmitate (2-BP) before lysing and clicking on TAMRA-azide for a fluorescent SDS-PAGE readout. While we observed changes in patterns of palmitoylation upon treatment with 2-BP, we did not observe any changes in global palmitoylation between the vehicle and the two inhibitor treated samples, suggesting that the inhibition of abH112 by compound 22 likely does not impact global palmitoylation (Figure S4C).

Figure 4: Conditional knockdown of abH112 does not lead to a growth defect.

(A) Partial Dendrogram of P. falciparum, Toxoplasma gondii, and human serine hydrolases that share a protein family domain generated using the Clustal Omega algorithm with abH112 and hAB17A colored in red.

(B) Overlay of predicted AlphaFold structure of abH112 (orange) and hAB17A (blue) with active site catalytic triad in cyan.

(C) Sequence alignment of abH112 and hABHD17A. Lipid binding motif is colored magenta, GXSXG serine hydrolase motif is shown in gold, and catalytic triad is shown in cyan.

(D) Western blotting of abH112 cKD lines grown in +/− aTC. Data were measured by ChemiDoc^™ MP Imaging System (Biorad), with bip used as a loading control.

(E) Plot of abH112 expression levels from (D) normalized to Bip.

(F) Plot of growth levels of parent WT cells (NF54) and abH112 cKD parasites in the presence or absence of aTC. Means are plotted ± SEM (n = 3). Tests for significance used a two-way-ANOVA.

G) Plot of EC₅₀ values of parasites for compound 22 in the WT parent cells (NF54) and in the cKD lines in the presence and absence of aTc. Parasites were maintained in 30 nM aTc for 72 hours before treatment with compound 22. Knockdowns were then induced at time of treatment with compound 22 by varying the concentration of aTc. EC₅₀s are plotted as means ± SEMs (N, n = 4–5, 2) and statistical significance determined using Mann-Whitney U tests comparing the parental data vs the abH112 cKD line where *p < 0.05.

To further characterize abH112, we attempted to express and purify the recombinant enzyme, but we were unable to isolate active forms of the protein. A recent study suggested that the catalytic activity of abH112 may be dispensable to the survival of the parasite.²² This study made use of a conditional mutant line in which a loxP DiCre system was used to conditionally replace the catalytic serine-227 with an alanine to catalytically inactivate abH112.^48,49 This report, however, did not confirm that the abH112 catalytic activity was disabled after recombination nor whether the wild-type gene had been fully excised from the parasite population. To explore this further, we propagated two parasite loxP-tagged recombinant clones (kindly provided by Dr. M. Blackman, the Francis Crick Institute, London) in the presence of rapamycin treatment, which causes recombination between the loxP sites (inserted in artificial introns) and leads to introduction of the mutated coding sequence fragment containing the mutated alanine (Figure S4D). PCR analyses of these parasites demonstrated the correct length for the excised product in rapamycin-treated parasites (Figure S4E). However, we also observed PCR products corresponding to the wild-type sequence in both pre and post-excision samples, which suggests an incomplete recombination or duplication of the gene (Figure S4F).

To further investigate the role of abH112 in asexual blood stage parasites, we used the TetR-DOZI aptamer system to generate cell lines in which the level of expression of the gene is conditionally knocked down (cKD) (Figure S5A).⁵⁰ We confirmed site-specific integration by diagnostic PCR (Figure S5B). The TetR-DOZI aptamer system regulates the expression of the target gene through anhydrotetracycline (aTC). Withdrawal of aTC from the culture drastically reduced the level of abH112 protein expression as determined by western blot. However, this reduction in expression of abH112 did not result in any developmental delay or growth defect in asexual blood stage parasites compared to control (Figure 4D–F). This lack of a growth phenotype in abH112 knockdown parasites may be the result of residual abH112 activity or functional redundancy in serine hydrolases during the erythrocytic cycle. To test whether the conditional knockdown impacted global serine hydrolase activity, we performed fluorescent ABPP on lysates of the knockdown parasites under both induced and uninduced expression of abH112. Using the broad-spectrum fluorophosphonate probe, we observed that there were no differences in global serine hydrolase activity between the induced and uninduced expression of abH112, suggesting that knockdown of abH112 does not lead to compensatory increase in the activity of other serine hydrolases. We also tested the susceptibility of abH112 cKD and control NF54 parasites to compound 22 in the presence and absence of aTC. We found that reduction in the levels of abH112 protein at the time of compound treatment did not affect parasite susceptibility to compound 22 (Figure 4G). In addition, parasites in which abH112 levels were reduced 72 hours prior to treatment with compound 22 showed a trend towards increased sensitivity to compound 22 compared to parasites in which abH112 levels were maintained until the time of drug treatment (Figure S5D). However, this difference was not statistically significant making the result inconclusive, further supporting the conclusion that inhibition of abH112 alone is not sufficient to kill parasites. Given the results from our proteomic studies, it is likely that inhibition of abH112 combined with inhibition of one or more of the other three serine hydrolase targets of compound 22 (PfMAGL, PF3D7_0805000 or PF3D7_1116100) is responsible for parasite killing by compound 22.

The Mechanism of Parasites killing by compound 22 is Distinct from the Pan-Lipase Inhibitor Orlistat.

Because abH112 is a hydrolase that is target by compound 22 but was not identified as a target of SalA, we reasoned that its mode of action is likely to be distinct from SalA. Therefore, we performed a detailed phenotypic analysis of infected red blood cells treated with either inactive compound 21 or active inhibitor 22. We visualized Giemsa-stained blood smears every twelve hours to determine the blood stage where growth was arrested. Ring stage parasites treated with the inactive 21 progressed through the blood stage lifecycle, egressed, and were able to successfully reinvade new red blood cells (Figure 5A). In contrast, parasites treated with 22 were arrested in growth at 24 hours, corresponding to a block in the transition from the late ring to early trophozoite stage. Previous studies have shown that depletion of palmitate and oleate lead to similar phenotypes.⁵¹ However, supplementation of both fatty acids did not rescue the killing phenotype of 22 (Figure S6A). A more recent report showed that the loss of the P. falciparum lysophospholipase, PfLPL1, caused parasite growth arrest in late ring stage via inhibition of food vacuole formation.⁵² Therefore, we evaluated the food vacuoles of parasites treated with either the vehicle, our inactive compound, or our active compound. We observed a small reduction in food vacuole size only in parasites treated with 22. However, this reduction was not significant and is likely not the primary mechanism associated with parasite killing (Figure S6B–C). Our results suggest a distinct mode of action of killing by 22 compared to both SalA and the pan-lipase inhibitor, Orlistat, both of which block progression at the late schizont stage.¹²

Figure 5: Compound 22 has a mechanism of action different from the pan lipase inhibitor Orlistat.

(A) Giemsa-stained parasites. Synchronized ring-stage parasites were treated at 0–6 h post-invasion with DMSO (vehicle control), 10 µM compound 21 or 22, and imaged at 24 h, 36 h, and 48 h.

(B) Lipidomic analysis of compound 22 treated parasites. Synchronized ring stage parasites were treated with either 900 nM compound 22 or DMSO for 24 h before lysis and collection of lipids. Significantly changed lipid classes are denoted on volcano plot. N= 3–4

(C) Isobolograms depicting P. falciparum in vitro culture-based fractional EC₅₀ (FIC50) values in the presence of compound 22 and Orlistat, each point represents a different biological replicate done in technical duplicate.

We next performed metabolomic studies of compound 22 treated parasites to determine if specific metabolic changes occur upon compound treatment. In contrast to Orlistat, which induces accumulation of triglycerides and arrest of parasites in the late schizont stage⁵³, we did not observe any significant accumulation of individual lipid classes in cells treated with 22 compared to vehicle-treated cells. However, there was an apparent downregulation of both ceramides and diacylglycerides in cells treated with 22 (Figure 5B). However, our efforts to rescue parasites with ceramides and cell-permeable diacylglyceride supplementation were unsuccessful (Figure S6D), suggesting that 22 may cause early mis-regulation of lipid metabolism pathways essential for the synthesis of neutral lipids such as ceramides and diacylglycerides. Nevertheless, our results suggest that 22 targets a pathway distinct from Orlistat. To further confirm this, we conducted drug synergy assays by co-treatment with 22 (EC₅₀ = 200 nM) and Orlistat (EC₅₀ = 250 nM). We observed that the resulting EC₅₀ values using four different combinations of the two molecules were lower (FIC₅₀<1) than what would be expected for additive inhibition (FIC₅₀=1) along a single metabolic pathway (Figure 5C). These results demonstrate a synergistic effect of combining the two compounds and suggest that Orlistat and 22 likely target distinct metabolic pathways. However, these results do not preclude the possibility of some overlap in the targets inhibited by each molecule.

Resistance mediated by compound 22 is mild and similar in mechanism to SalA resistance

Given the similarities in structures between SalA and 22, we attempted to study potential cross-resistance between SalA and 22 and assessed the relative susceptibility of this compound to induce resistance. Cross-resistance screening of SalA selected parasite lines that we previously generated in Dd2 DNA polymerase δ mutant (Dd2-Polδ) parasites^12,54 revealed that both P102L and 145I mutations in the PRELI domain-containing protein conferred a similar reduction in susceptibility to 22 compared to that reported for SalA (Figure 6A; 6C). Compared to the Dd2-Polδ parental line, PRELI mutant parasites also showed a 7-10-fold increase in EC₅₀ for 22, suggesting that there may be a common genetic mechanism of resistance.

Figure 6: Cross resistance and in vitro selection of parasites.

(A) Plot of EC₅₀ values for two SalA resistant lines (harboring mutations in PRELI domain-containing protein at positions P102L and T145I) for compound 22. Values are plotted as means ± SEMs and statistical significance determined using Mann-Whitney U tests where *p < 0.05.

(B) EC₅₀ shifts of mutant and parental lines against 22. EC₅₀s for each line are presented as means ± SEMs (N, n = 4, 2)

(C) Summary table of EC₅₀s for each line are presented as means ± SEMs (N, n = 4, 2).

To further study the mechanism of resistance to 22 we performed in vitro single step selection on the hypermutator Dd2-Polδ parasite line. We pressured 10⁹ asexual blood stage parasites per flask, with duplicate flasks, at a continuous concentration of 5×EC₅₀ of compound 22. We observed recrudescence microscopically in both flasks at day 17 and were able to recover specific clones by limiting dilution. Phenotypic analysis of cloned parasites in 72 hour dose-response assays showed an 11 to15-fold shift in the EC₅₀ values of 22 compared with the parental line (Figure 6B).

To identify the mechanism of resistance, we performed whole genome sequencing on eight resistant clones and their parental Dd2-Polδ line. Thia analysis yielded good quality reads corresponding to a high depth of coverage for the resistant clones (Supplemental Table 1). No copy number variations (CNVs) were found in any of the recrudescent parasite genomes relative to 3D7-A in either sequencing experiment except for FL2_A5 which had CNV gain of 1.5 fold in the amplified region of 82kb containing MDR1 (Supplemental Table 1). Analysis revealed that compared to the parental line, resistant clones from both flasks possessed either a P102L, N144I or a T145I mutation in the PRELI domain-containing protein. The P102L and T145I mutations are the same mutations observed in SalA selections¹², with N144I being new to selections with 22.

To determine the minimum inoculum of resistance (MIR), which is an indirect measure of the propensity to generate resistance in the field, we exposed Dd2-B2 parasites to 22 at 5×EC₅₀. The selection had a starting inoculum of 3.3 × 10⁸ parasites per flask in three separate flasks (covering the range from 3.3 × 10⁸ to a total of 1 × 10⁹). Parasites recrudesced on day 23 in one out of three flasks, resulting in an MIR value of 1× 10⁹ (Supplementary Table 2). Based on prior observations, drugs with a MIR ≥ 10⁹ have a much lower chances of treatment failure.⁵⁵ Based on these results, compound 22 is less prone to induce resistance compared to other commonly used antimalarials.

Discussion

Previously, we used the natural product SalA to identify several α/β serine hydrolases that, when inhibited, caused parasite death in P. falciparum. In this study, we describe the synthesis of mixed alkyl/aryl phosphonates as bioisosteres of SalA and further characterize their inhibition of serine hydrolases and antiparasitic effects. We identified a pair of constitutional isomers that enabled a comparative proteomic approach in which targets responsible for the parasite killing effects of the active analog can be identified. Using these compounds as probes, we identified abH112, a previously annotated essential enzyme, that when inhibited in combination with other serine hydrolases, leads to strong anti-parasitic activity of the active compound. Our compound exhibited different stage-specific killing compared to SalA and the pan-lipase inhibitor Orlistat, and the metabolic changes under treatment with our compound differed significantly from previously explored mechanisms of action. This phenotype is likely due to the inhibition of abH112, which was not identified as a target of SalA. By continuing to improve our understanding of how serine hydrolases regulate lipid metabolic pathways in Plasmodium parasites it is possible to further validate novel drug targets and mechanisms of action that disrupt these pathways.

Previous studies have attempted to assess the importance of essential serine hydrolases using either complete or catalytic inactivation of the enzymes of interest. However, in these cases, either the knockout could not be achieved, or there was insufficient evidence to determine whether the catalytic activity of the enzyme was eliminated.^12,22 These findings all point to an essential role for metabolic serine hydrolases in the parasite lifecycle. To further validate the essentiality of abh112, we further analyzed a recently reported parasite line in which tetracycline was used to induce a genetic rearrangement in which only the catalytic Ser is mutated to Ala, thus maintaining expression of the protein while inactivating the enzyme.²² We found that the loxP diCre system does in fact enable integration and proper excision of the wild type gene, but we also found evidence that the wild type gene was present in all samples. This suggests that the gene, after being excised, was likely reintegrated into the genome leading to expression of the wild-type protein. This would account for the lack of phenotypic changes associated with the conditional inactivation mutants. Moreover, this provides further support beyond the transposon mutation screens, that abH112 is essential.

Given that conditional knock down of abH112 did not result in parasite growth inhibition, and that knockdown generally did not sensitize parasites to inhibition by compound 22, it is likely that our inhibitor acts via a pleiotropic mechanism, similar to the parent compound, SalA, on which its design was based. Both compounds 21 and 22 targeted the essential hydrolase PfMAGL with pM inhibitory activity towards the recombinant enzyme. Compound 22 also binds two additional serine hydrolases, PF3D7_0805000 and PF3D7_1116100. However, because only compound 22 kills parasites, the combination of inhibition of abH112 plus other hydrolases, such as PfMAGL, and/or PF3D7_0805000 and/or PF3D7_1116100 is likely necessary for parasite killing. For SalA, which also potently inhibits PfMAGL, parasite killing is likely mediated by inhibition of PfMAGL plus other essential serine hydrolase targets such as the BEM46-like protein (PBLP) or the exported lipase 2 (PfXL2). To clarify which combination of hydrolases need to be inhibited to effectively kill parasites would require the conditional knockdown or knockout of each of the possible combinations of serine hydrolases in tandem, something that would be challenging to do even with current genetic tools in P. falciparum.

Nevertheless, the fact that Sal A and compound 22 have different targets likely results in different mechanisms of action for each compound. Specifically, we determined that 22 kills parasites at a much earlier stage than previously studied serine hydrolase inhibitors. This phenotype is likely the result of blockages of distinct points in essential metabolic pathways.^12,53 Furthermore, we found that 22 synergizes with the pan-lipase inhibitor Orlistat, suggesting that additional serine hydrolase and lipid metabolic pathways could be co-inhibited to increase compound potency and perhaps reduce chances for resistance even further. These avenues for the development of serine hydrolase inhibitors are especially promising, considering that mammalian serine hydrolases can typically be targeted with minimal cytotoxic effects.¹⁷ The ability to tune the reactivity of the mixed alkyl/aryl phosphonates also makes it possible to screen for molecules that target parasite serine hydrolases with limited targeting of host-derived enzymes that could lead to toxicity. Given the substantial work on the development of specific serine hydrolase inhibitors for human targets with favorable pharmacokinetics and bioavailability, combinations of serine hydrolase inhibitors initially developed for human serine hydrolases could potentially be repurposed and serve as a valuable reservoir for the development of antimalarials.

We also sought to characterize the serine hydrolases targeted by compound 22 using resistance selections, which yielded parasites with only moderate resistance to our compound. We isolated six clones with decreases of potency for compound 22 ranging from 11x-15x compared to the parent parasite line and performed whole-genome sequencing on these clones. Only the PRELI domain-containing protein was mutated in every clone sequenced. The PRELI domain-containing protein is not a serine hydrolase but was also found to be mutated in all clones of SalA-resistant parasites.¹² This protein is found in a wide variety of eukaryotic organisms and localizes to mitochondria where it regulates lipid accumulation by shuttling phospholipids across the intermembrane space.⁵⁶ Previously, the PRELI domain-containing protein was characterized as regulating a mechanism of multidrug resistance in the related Toxoplasma gondii parasite.⁵⁷ Our results indicate that it likely functions similarly in P. falciparum to mediate a mechanism of multidrug resistance, as we demonstrated that compound 22 and SalA likely have different mechanisms of actions and therefore do not likely act directly on the PRELI domain-containing protein as a primary target. The MIR value for 22 is high and therefore it has high barrier to resistance. Mechanism of action is not always tied to the mechanism of resistance. For example, ART is well established as a drug with a pleiotropic mechanism of action, but resistance to ART is governed by mutations in the k13 gene⁷. Moreover, in Plasmodium and other organisms, ABC and other efflux pumps are often mutated or upregulated to enable cellular export of drugs with various mechanisms of action.⁵⁸ Ultimately, we did not observe any mutations in serine hydrolases in our resistance selections, which suggests the parasite has difficulty in evolving resistance through mutations in the direct targets of the compounds. Furthermore, the fact that all the identified serine hydrolase inhibitors with potent anti-parasite activity seem to function by inhibition of multiple essential enzymes, may explain why these compounds all have a low propensity to generate resistance in the parasite. These findings further support targeting multiple serine hydrolases to develop new classes of antimalarial drugs or using existing serine hydrolase inhibitors with non-overlapping targets in the parasite to generate therapeutics that prevent resistance. This strategy is also not likely limited to Plasmodium parasites, as recent studies in Staphyloccus auerus showed that the pleiotropic targeting of serine hydrolases along with other metabolic enzymes leads to potent killing of the bacteria.⁵⁹

Here, we have described the use of mixed alkyl/aryl phosphonates as a nascent class of serine hydrolase inhibitors with potent antimalarial properties. These compounds have underscored the value of target-agnostic screening to identify new drug susceptible pathway and have the potential for application in other diseases and pathogens where serine hydrolases also play essential roles in maintaining homeostasis.^60,61 Mixed alkyl/aryl phosphonates are synthetically tractable and provide a starting point for the development of molecules that enable the validation of hydrolases that are optimal drug targets that can be targeted simultaneously. The continued development of chemical tools to study lipid metabolizing pathways and proteins governing such pathways will be essential for the development of next-generation therapeutics for P. falciparum. Furthermore, future studies will likely benefit from combining phenotypic screening, target identification, and characterizations of changes in metabolism to prioritize the critically important serine hydrolases as candidate drug targets.

Limitations of the study

The work presented here serves to further demonstrate that targeting multiple serine hydrolases in Plasmodium falciparum is an effective way to kill parasites without inducing rapid resistance. Our work makes use of a pair of active and inactive analogs of a mixed arylalkyl phosphonate to identify a set of potentially relevant targets within the family of serine hydrolases in P. falciparum. While the work demonstrates the potential value of targeting multiple members of this family, significant additional work will be required to define the roles of individual or groups of serine hydrolases in carrying out essential processes that the parasite needs to survive. Further studies using mutant lines of parasites in which multiple serine hydrolases are disrupted may provide a better understanding of which enzymes can be effectively targeted by small molecules to create next-generation anti-malarial agents. Significant efforts will also likely be required to express active recombinant forms of multiple serine hydrolases in order to begin to develop potent and selective inhibitors that are optimized to prevent unwanted cross reactivity with host-derived serine hydrolases. Our results presented here, combined with the genetic studies using transposon mutants, suggest that abH112 has important cellular functions for the parasite. However, our results also show that protein levels of this target can be dramatically reduced without significant effect on blood-stage growth of the parasite. Therefore, our current study does not allow us to definitively assign abH112 as an important new therapeutic target without a better understanding of which other serine hydrolases must be simultaneously inhibited to achieve effective parasite killing.

STAR Methods

RESOURCE AVAILABILITY:

Lead Contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact, Matthew Bogyo (mbogyo@stanford.edu).

Materials availability:

We may require a payment and/or a completed Materials Transfer Agreement if there is potential for commercial application. Materials availability may be limited by synthesis of stocks.

Data and code availability

• All data are available in the main text or the supplemental information. Proteomics data are publically accessible via the PRIDE repository with the dataset identifier PXD051499. Lipidomics data are publically accessible via the DRYAD repository and can be accessed at https://doi.org/10.5061/dryad.b5mkkwhmj.

• This paper does not report any original code.

• Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.

EXPERIMENTAL MODEL AND STUDY PARTICIPANT DETAILS:

Human Foreskin Fibroblasts were kindly provided by Boothroyd Lab (Stanford University). HFFs were maintained in Dulbecco’s modified Eagle’s medium (DMEM) with 10% FetalPlex animal serum complex (Gemini Biotech, catalog no. 100602).

P. falciparum W2, 3D7, and Dd2_Polδ cultures were were grown in human erythrocytes (no blood type discrimination) purchased from the Stanford Blood Center (approved by Stanford University) or the Interstate Blood Bank (for Columbia University). at 3% hematocrit in RPMI-1640 media, supplemented with 25 mM HEPES, 50 mg L-hypoxanthine, 2 mM L-glutamine, 0.225% sodium bicarbonate, 0.5% (wt/vol) AlbuMAXII (Invitrogen) and 10 μg/mL gentamycin. Cultures were maintained at 37°C in modular incubator chambers (Billups-Rothenberg) at 5% O₂, 5% CO₂ and 90% N₂.

METHOD DETAILS

P. falciparum Parasite Culture: P. falciparum W2, 3D7, and Dd2_Polδ cultures were maintained, synchronized, and lysed as previously described.⁶³ Parasite cultures were grown in human erythrocytes (no blood type discrimination) purchased from the Stanford Blood Center (approved by Stanford University) or the Interstate Blood Bank (for Columbia University). at 3% hematocrit in RPMI-1640 media, supplemented with 25 mM HEPES, 50 mg L-hypoxanthine, 2 mM L-glutamine, 0.225% sodium bicarbonate, 0.5% (wt/vol) AlbuMAXII (Invitrogen) and 10 μg/mL gentamycin. Cultures were maintained at 37°C in modular incubator chambers (Billups-Rothenberg) at 5% O₂, 5% CO₂ and 90% N₂. Parasites were synchronized at ring stage by treating parasites with 5% sorbitol and incubating at 37 °C for 10 minutes. The synchronized parasite were then washed and new culture media was added. Parasite pellets were collected by pelleting cultured parasites in red blood cells through centrifugation (484x g) and the red blood cell pellet was treated with 0.1% saponin for 5 min on ice, then the cultures are pelleted and washed 3x with PBS and the purified parasite pellets can be used for downstream analyses.

PfMAGL and hMAGL Inhibition Assays:

5 µL of 20 nM PfMAGL or hMAGL (final concentration 5 nM) in PBS containing 0.1% Triton was preincubated with 5 µL of inhibitor diluted in PBS for 30 minutes at 37°C in a 384-flat black well plate. 10 µL of 80 µM 4-methylumbelliferyl caprylate (final concentration 40 µM) in 1% DMSO in PBS was then added and fluorescence (lex = 365 nm and lem = 455 nm) was read at 37°C in 1 min intervals on a Cytation 3 imaging reader (BioTek, Winooski, VT, USA) for 60 min. Turnover rates were calculated over the linear phase of the reaction in RFU/min.

P. falciparum Replication Assay:

Synchronized ring stage parasites culture at 1% parasitemia and 0.5% hematocrit were plated in a 96-well plate and dilutions of compounds in 1xPBS were added to parasites. The culture was incubated for 72 h before the samples were fixed in a final concentration of 1% paraformaldehyde in PBS for 30 minutes at room temperature. The fixed samples were then stained with a 50 nM final concentration of the YOYO-1 nuclear stain. Parasite replication was monitored by observing either YOYO-1 positive red blood cells (infected) or YOYO-1 negative cell (uninfected) using a BD Accuri C6 automated flow cytometer. For one hour pulse assay, cultures were treated with drug for one hour before being washed and resuspended in fresh media, and then grown for an additional 71 h. Rescues were performed with concentration of lipid added at beginning of assay.

P. falciparum Rate of Kill Assay:

Parasite killing kinetics was adapted from the method previously outlined in previously.⁶⁴ In brief, 3D7 ring-stage parasites at 0.5% parasitemia and 2% hematocrit and were treated with 10 x EC₅₀ of drug. At either 24 h or 48 h, drug free parasites were then cultured in fresh erythrocytes and culture media for the remaining time until readout at 72 h post treatment. Cultures were then fixed and stained as above.

Human Cell Cytotoxicity Assay:

Human Foreskin Fibroblasts in Dulbecco’s modified Eagle’s medium (DMEM) with 10% FetalPlex animal serum complex (Gemini Biotech, catalog no. 100602) were seeded at 2500 cells (nonconfluent) per well <24 h before addition of the compound. Compounds were diluted for dose–response concentrations in media and added to the cells for 72 h. Cell viability was measured using the CellTiter-Blue Assay (Promega) as per the manufacturer’s instructions.

Preparation of P. falciparum Lysates for Proteomics:

75 mL cultures of W2 parasites at 12–15% parasitemia in 4% hemotocrit were either pretreated with 0.2% DMSO in PBS, 5µM Compound 21, or 5µM Compound 22 for 2 h at 37°C. Then either DMSO, 5µM 21-alk, or 22-alk was added to the cultures and incubated for another 2 h at 37°C. Labelled parasites pellets were harvested via saponin lysis and pellets were lysed in 0.1% SDS in PBS using probe sonication on ice.

Competitive LC-MS/MS target identification

Protein concentration was measured using the Bradford assay and 250 µg of samples (0.5 mg/mL, 500 µL) were used. Click chemistry was performed for one hour at r.t. with the addition of biotin azide (50 µM, 50x stock in DMSO), tris(2-carboxyethyl)phosphine hydrochloride (TCEP) (1 mM, 50x fresh stock in water), tris[(1-benzyl-1H-1,2,3-triazol-4-yl)methyl]amine (TBTA) (100 µM, 16x stock in DMSO:tButanol 1:4), and copper(II) sulfate (1 mM, 50x stock in water). Protein was precipitated by adding 4 vol. −20 °C acetone and samples were incubated overnight at −20 °C. Samples were centrifuged for 10 min at 20’000 x g at 4 °C yielding a protein pellet which was air dried for 20 min. Pellets were solubilized in 2% SDS in PBS via sonication. Tubes were centrifuged at 4’700 x g for five min and the soluble fraction was transferred to a new tube. PBS was added to give a final SDS concentration of 0.2%. 60 µL of streptavidin agarose beads (ProteoChem) were added and the mixture was rotated for four hours at r.t. Beads were washed with 1% SDS in PBS (1× 10 mL), PBS (3× 10 mL), and water (3× 10 mL). Beads were resuspended in 6 M urea in PBS (500 μL), reduced with 10 mM neutralized TCEP (20x fresh stock in water) for 30 min. at r.t., and alkylated with 25 mM iodoacetamide (400 mM fresh stock in water) for 30 min. at r.t. in the dark. Beads were pelleted by centrifugation (1’400 x g, two min.) and resuspended in 150 μL of 2 M Urea, 1 mM CaCl₂ (100x stock in water), and trypsin (Thermo Scientific, 0.5 μL of 0.5 μg/μL) in 50 mM NH₄HCO₃. The digestion was performed overnight at 37 °C. Samples were acidified with 1 vol. of isopropanol 1% TFA and desalted using styrenedivinylbenzene reverse-phase sulfonate (SDB-RPS) StageTips as described previously.⁶⁵ Briefly, samples were loaded on a 200 µL StageTip containing two SDB-RPS disks and centrifuged at 1500 x g for 8 min. This was repeated until all the sample was loaded on the StageTip. StageTips were washed three times with 200 µL of isopropanol 1% TFA at 1500 x g for 8 min, then eluted with 100 µL of 80% MeCN, 19% water, and 1% ammonia and dried. Samples were then analyzed by LC-MS/MS, MaxQuant, and Perseus.

LC-MS/MS analysis

Peptides were resuspended in water with 0.1 % FA and analyzed using an EASY-nLC 1200 nano-UHPLC coupled to a Q Exactive HF-X Quadrupole-Orbitrap mass spectrometer (Thermo Scientific). The chromatography column consisted of a 50 cm long, 75 μm i.d. microcapillary capped by a 5 μm tip and packed with ReproSil-Pur 120 C18-AQ 2.4 μm beads (Dr. Maisch GmbH). LC solvents were 0.1 % FA in H₂O (Buffer A) and 0.1 % FA in 90 % MeCN: 10 % H₂O (Buffer B). Peptides were eluted into the mass spectrometer at a flow rate of 300 nL/min. over a 90 min linear gradient (5–35 % Buffer B) at 65 °C. Data was acquired in data-dependent mode (top-20, NCE 28, R = 15,000) after full MS scan (R = 60,000, m/z 400 – 1,300). Dynamic exclusion was set to 10 s, peptide match to prefer and isotope exclusion was enabled.

MaxQuant analysis

The MS data were analyzed with MaxQuant and searched against the Plasmodium falciparum proteome (isolate 3D7, Uniprot) or human proteome (Uniprot) and a common list of contaminants (included in MaxQuant).⁶⁶ The first peptide search tolerance was set at 20 ppm, 10 ppm was used for the main peptide search and fragment mass tolerance was set to 0.02 Da. The false discovery rate for peptides, proteins, and site identification was set to 1 %. The minimum peptide length was set to 6 amino acids and peptide re-quantification and “match between runs” were enabled. MaxLFQ without normalization was used. Methionine oxidation and N-terminal acetylation were searched as variable modifications and carbamidomethylation of cysteines as fixed modification.

Phylogenetic and Structural Model

The human, Toxoplasma gondii, and Plasmodium falciparum proteomes were filtered for proteins annotated to be in the Hydrolase 4 (PF12146) protein family domain. All sequences for these proteins were input into the Clustal Omega algorithm for multiple sequence alignments and a dendrogram was constructed based on homology. For structural models, the AlphaFold structures of both ABHD17a and abH112 were placed into PyMOL and the alignment feature was used.

Parasite culture for knockdown parasites

abH112-cKD and Cas9/T7 RNA polymerase-expressing NF54attB parental control (NF54) asexual blood stage (ABS) parasites were cultured at 2% hematocrit in human O+ red blood cells (RBCs) in 500 nM aTC-containing RPMI-1640 media, supplemented with 25 mM HEPES (Fisher), 50 mg/L hypoxanthine (Sigma Aldrich), 2 mM L-glutamine (Cambridge Isotope Laboratories, Inc.), 0.21% sodium bicarbonate (Sigma Aldrich), 0.5% (wt/vol) AlbuMAXII (Invitrogen), and 10 µg/mL gentamycin. Cultures were propagated in tissue culture flasks gassed with a mixture of 5% O₂, 5% CO₂ and 90% N₂ and maintained at 37°C.

Drug susceptibility assays for abH112 cKD parasites

To define the 50% (IC50) and 90% (IC90) growth-inhibitory drug concentrations for abH112-cKD and NF54 control parasites, asynchronous cultures maintained in 500 nM aTC-containing media were washed 3 times in aTC-free culture media 72 hours before each dose-response assay and divided into media containing 30 nM aTC or 0 nM aTC. On the day of the experiment, cultures were washed to remove aTC, and exposed for 72 h to a range of ten concentrations of compound 22 (or compound 21) that had been 3-fold serially diluted in duplicates along with drug-free controls at 0∙3% parasitaemia and 1% haematocrit in media containing 30 nM, 3 nM, or 0 nM aTC. Serial dilutions of the drug were performed using a Tecan D300e digital dispenser and parasite dilutions were added manually. Parasite survival was assessed by flow cytometry on an Intellicyt iQue Screener PLUS (Sartorius) or a BD FACSCelesta (BD Biosciences) using 1µM SYBR Green (Invitrogen) and 200 nM MitoTracker Deep Red FM (Life Technologies) as nuclear stain and vital dyes, respectively. IC50 and IC90 values were derived from growth inhibition data using linear interpolation as means ± SEM from 5 independent experiments with two technical replicates each.

Stage Specific Killing Assay

Tightly synchronized W2 ring stage cultures at 5% parasitemia in 2% hematocrit were treated with either 10 µM compound 21 or 22. Every 12–24 h, a blood smear was taken from the cultures and visualized with 5% giemsa stain to monitor the growth of the parasite.

Preparation of Lipidomic Samples

For lipid analysis, synchronized W2 ring stage of 50 mL cutlures of 10% parasitemia in 4 % hematocrit were treated with 920nM (5x EC₅₀) of compound 22 or DMSO vehicle for 16 h. Red blood cells were saponinin lysed and the parasite pellet was collected. Parasite pellets were subsequently collected and resuspended in 1 ml of 1xPBS buffer and transferred into glass vials pre-loaded with 2:1 chloroform/methanol to a final ratio of 2:1:1 chloroform/methanol/1xPBS. Samples were vigorously shaken spun down at 1000g and the organic layer was collected. Three vials of lipids extracted from each treatment were dried down under a stream of argon.

Lipid analysis using high-performance liquid chromatography-mass spectroscopy Lipidomics was performed on an Agilent 6545 Q-TOF LC/MS as previously described and data was anaylzed using the online XCMS platform.⁴⁴ Metabolites shown to be significantly changed between the two treatment groups were then searched by their mass and classified using the online LipidMaps tool.⁶⁷

Food Vacuole Visualization

Parasites were plated at 5% parasitemia, 4% hematocrit in a 6 well plate and treated with either DMSO, 21, or 22 at 5 µM for 16 hrs. Nile Red was then added to a final concentration of 1 μg/ml into the parasite culture and the cultures were incubated on ice for 30 min. The culture subsequently washed with 1xPBS before analysis by confocal microscopy (excitation/emissions maxima is 552/636 nm). The resulting images were then analyzed using ImageJ and the area of food vacuole was measured by circling the Nile Red positive areas.

Drug Synergy Assay

Synchronized ring stage W2 parasites culture at 1% parasitemia and 0.5% hematocrit were plated in a 96-well plate. Drug mixtures in four different ratios diluted in PBS were added to the cultures and the parasites were grown for 72 h. EC₅₀ values were derived for each compound tested alone, and fractional EC₅₀ (FIC50) values were determined for each compound tested in combination (FIC₅₀ =EC₅₀ of the drug alone/EC₅₀ of the drug in combination) and plotted for each drug combination.

In vitro drug selection studies:

P. falciparum hypermutator (Dd2-Polδ) was used for in vitro drug selection. To select for compound 22 resistant parasites, duplicate flasks of 2×10⁹ parasites were exposed to continuous drug pressure at concentrations of 5×EC₅₀ for 60 days until recrudescent parasites were observed. Drug media was changed daily for first 6 days. Cultures were monitored daily by microscopic examination of Giemsa-stained smears. Recrudescent cultures were assayed to analyze degree of resistance and the resistant parasites were obtained by limiting dilution. Dd2-Polδ line was maintained in culture throughout the in vitro selection.

Minimum inoculum of resistance studies:

To determine MIR, 3.3×10⁸ Dd2-B2 parasites were cultured per flask in triplicates continuous under 5×EC₅₀ drug pressure. Media was changed daily for first 6 days and culture was monitored daily by Giemsa-stained smears for complete killing and recrudescence. Cultures were passaged every week by giving fresh blood and were maintained for 60 days. The MIR is defined as the minimum parasite inoculum used to obtain resistance and calculated as follows: total number of asexual blood stage (ABS) parasites cultured ÷ total number of recrudescent cultures.

In vitro drug susceptibility assays:

To define the baseline EC₅₀ of Dd2-Polδ to compound 22 and shifts in susceptibility in recrudescent lines, parasite cultures at 0.2% parasitemia and 1% hematocrit (HCT) were incubated for 72 h with a range of ten drug concentrations that were 2-fold serially diluted in duplicates along with drug-free controls in complete medium. Parasite survival was assessed by flow cytometry on IntelliCyt iQue3 (Sartorius) using SYBR Green and MitoTracker Deep Red FM (Life Technologies) as nuclear and mitochondria stains, respectively. Growth inhibition was plotted using nonlinear regression function to calculate EC₅₀ values (using Graphpad Prism 7).

DNA Extraction and Whole-Genome Sequencing:

DNA libraries for each sample were prepared using the Illumina Nextera DNA Flex library kit with dual indices. The samples were multiplexed and sequenced on an Illumina MiSeq to obtain 300 bp paired end reads and Nextseq to obtain 150 bp paired end reads at 19–44x depth of coverage across the samples. The sequence data generated were aligned to the P. falciparum 3D7 genome (PlasmoDB version 48.0) and variants filtered based on quality scores and read depth to obtain high quality SNPs. The list of variants from the resistant clones were compared against the Dd2-Polδ parent to obtain homozygous SNPs present exclusively in the resistant clones.

Generation of abH112 cKD parasites for translation regulation:

To generate abH112 cKD parasites, CRISR-Cas9 based TetR-DOZI aptamer system was used to regulate its expression.⁶⁸ The left homology region fused with recodonized 3’ end of abH112 and right homology region as well as guide sequence were cloned into pSN054 donor plasmid containing 3XHA epitope tag at C-terminus as described previously.⁶⁹ The final construct was transfected into Cas9 expressing NF54 parasites and maintained in 500nM aTC and 2.5 mg/mL of Blasticidin S. Edited parasites were confirmed by site specific integration PCR. All the primers, guide sequence and synthetic fragment sequence are given the key resource table.

Key resource table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
Mouse Anti-HA Tag Antibody	Biolegend	Cat# 901516
Rabbit Anti-Bip	BEI Resources	Cat# MRA-1246
HRP Secondary Antibody Rabbit	Abcam	Cat# ab205718
HRP Secondary Antibody Mouse	Abcam	Cat# ab205719
Bacterial and virus strains
E. coli BL21DE3	New England Biolabs	Cat# C2527
Chemicals, peptides, and recombinant proteins
Mixed Alkyl/Aryl Phosphonate Library	This Paper	N/A
Recombinant human monoacylglycerol lipase (MAGL)	Cayman Chemical	Cat# 10007812
Recombinant P. falciparum PfMAGLLP	Yoo et al.12	N/A
4-Methylumbelliferyl caprylate	Santa Cruz Biotech.	Cat# sc-281420
Fluorophosphonate-rhodamine (FP-Rho)	Lentz et al.62	N/A
Orlistat	Cayman Chemical	Cat#10005426
YOYO-1 iodide	Thermo Fisher	Cat# Y3601
anhydrotetracycline	Cayman Chemical	Cat#10009542
Critical commercial assays
Cell Titer Blue Cell Viability Assay Kit	Promega	Cat# G8080
BCA protein assay kit	Pierce	Cat# 23225
Bradford assay kit	ThermoFisher	Cat# 23200
Deposited data
Mass spectrometry proteomics (PRIDE)	This Paper	https://www.ebi.ac.uk/pride/ ; Accession number: PXD051499
Mass spectrometry lipidomics (DRYAD)	This Paper	https://datadryad.org/stash doi:10.5061/dryad.b5mkkwhmj
Experimental models: Cell lines
Human foreskin fibroblast (HFF)	John Boothroyd Lab	N/A
Experimental models: Organisms/strains
P. falciparum strain W2	MR4	Cat # MRA-157
P. falciparum strain 3D7	MR4	Cat # MRA-102
P. falciparum abH112-cMUT	Davison et al.22	N/A
P. falciparum strain NF54	MR4	Cat # MRA-1000
P. falciparum strain NF54 abH112 cKD	This Paper	N/A
P. falciparum strain Dd2 PfPRELI T145I	Yoo et al.12	N/A
P. falciparum strain Dd2 PfPRELI P102L	Yoo et al.12	N/A
P. falciparum strain Dd2_Polδ	Yoo et al.12	N/A
Oligonucleotides
AGAGGTTCGACGGAAGTATCTGTTATAATG (P1)	Eurofins	N/A
ACAACTCCAGTGAAAAGTTCTTCTCCT (P2)	Eurofins	N/A
ATCATTCTTTCCATGCATAATAAATAAAGG (P3)	Eurofins	N/A
GTGAGTACATAAATATATTATATAAACTAGACTAGGTTAACTGGCCAAGATCTCCCGGG (P8408)	Eurofins	N/A
TCCTCCTCAGAAATTAGCTTCTGCTC (P8470)	Eurofins	N/A
GTTGAGTTGGGAAAATACTGGTGAACTAG (P8510)	Eurofins	N/A
CGGGTTTAGGGGACGCATGG (P8977)	Eurofins	N/A
GCTTTTATCACTCAAGCGAAAAGCATGAA (P8978)	Eurofins	N/A
AGTTAAAATAAGGCTAGTCCGTTATCAACTTG (P9141)	Eurofins	N/A
taatacgactcactataggGAGCTATTTTATCATAAATTgttttagagctagaaatag (Guide)	Eurofins	N/A
Software and algorithms
MaxQuant	Max Planck Institute of Biochemistry	https://maxquant.net/perseus/
Perseus	European Molecular Biology Laboratory	https://www.ebi.ac.uk/jdispatcher/msa/clustalo
Clustal Omega	Seivers et al.42	https://www.ebi.ac.uk/jdispatcher/msa/clustalo
Alphafold	Jumper et al.46	https://alphafold.ebi.ac.uk/
PyMol	The PyMOL Molecular Graphics System, Schrodinger, LLC	https://pymol.org/
GraphPad Prism	GraphPad Software Inc	https://www.graphpad.com/scientific-software/prism/

Western blot analysis:

Synchronized abH112 cKD ring stage parasites were cultured in presence and absence of aTC for 72 h. Cultures were harvested and RBCs were lysed using 0.05% saponin in 1×PBS at 4°C. Parasites were washed two times with cold1×PBS containing cOmplete proteases inhibitors (4693159001, Sigma) and pellets were resuspended in SDS buffer (8 M Urea, 5% SDS, 50 mM Bis-Tris, 2 mM EDTA, 25 mM HCl, pH 6.5) and Laemmli buffer. Protein samples were run on 10% Mini-PROTEAN^® TGX^™ Precast Gels (Biorad) and transferred onto nitrocellulose membranes (Biorad). The membrane was blocked in 3% BSA-TBS and probed with either mouse anti-HA (1: 1000) or rabbit anti-Bip followed by HRP conjugated secondary antibodies. Protein blots were developed using ECL Chemiluminescent Substrate and imaged and analyzed by ChemiDocTM MP Imaging System (Biorad).

Asexual growth rate analysis:

To assess asexual parasite growth rate, synchronous ring-stage NF54 and abH112 cKD parasites were cultured at different aTC concentrations (30 nM, 3 nM and no aTC) over two intra-erythrocytic developmental cycles. Parasite expansion rate was assessed by flow cytometry on IntelliCyt iQue3 (Sartorius) using SYBR Green and MitoTracker Deep Red FM (Life Technologies) as nuclear and mitochondria stains, respectively.

General Synthetic Procedures:

Appel Reaction Alcohol to Iodo:

Alcohol (1 equivalent) was dissolved in dichloromethane (final concentration of 0.2 M). To the solution, iodine (1.2 equivalents) and triphenylphosphine (2 equivalents) were added and the reaction was protected from light and stirred for 3 hrs at room temperature. The reaction was quenched with sodium thiosulfate and the aqueous phase was extracted with dichloromethane (2x) and the combined organic phase was washed with brine and then dried over sodium sulfate and concentrated under reduced pressure. The crude reaction mixture was then purified via silica flash chromatography and concentrated under reduced pressure to yield iodo products.

Arbuzov Reaction:

Iodo containing compound (1 equivalent) is dissolved in triethyl phosphite (4 equivalents) and the solution is stirred at 130°C for 16 h. The product is concentrated via constant airflow over the solution and the crude product was carried forward.

Direct Aryloxylation:

Diethyl phosphonate (1 equivalent) was dissolved in dichloromethane (final concentration 0.2 M) and to the solution triflate anhydride (1.5 equivalents) and pyridine (2 equivalents) were added sequentially (the reaction will bubble release fumes and turn orange) and stirred for 10 minutes at room temperature. Various phenols (2.5 equivalents) were then added and the reaction and stirred for 30 more minutes at room temperature. The reaction was then concentrated under reduced pressure and the reaction mixture was purified via silica flash chromatography and concentrated under reduced pressure to yield mixed alkyl/aryl phosphonates.

Compound 26:

Alcohol (1 g, 1 eq), DMAP (367 mg, 1 eq), and triethylamine (2.5 mL, 6 eq) were stirred at room temperature for 10 min in 30 mL of DCM. Tosyl chloride (1.15 g, 2 eq) was added and the reaction was stirred overnight. The reaction was quenched with cold water, extracted with ethyl acetate 3x, washed with 1M HCl, concentrated NaHCO₃, brine, dried over sodium sulfate. The residual material was then purified via silica flash chromatography (0–20% ethyl acetate in hexanes) (1.548 g, 96% yield).

Compound 27:

Lithium Aluminum Hydride (450 mg, 6 eq) was dissolved in dry THF (35 mL). Ester (1.54 g, 1 eq) was dissolved in 3 mL of THF and added dropwise to solution and stirred for 4 hours at 55°C. Reaction was quenched with water, filtered through celite, redissolved in ethyl acetate and dried over sodium sulfate. The residual material was then The residual material was then purified via silica flash chromatography (0–40% ethyl acetate in hexanes) (50% yield).

Deprotection of Compound 28:

Silyl protected compound (450 mg, 1 eq) and potassium hydroxide (112mg, 4eq) were stirred in methanol at 50°C. The reaction was quenched with water and the pH was adjusted to 1 using 1M HCl and extracted with ethyl acetate 3x, and the combined organic layer was washed with brine and dried over sodium sulfate. The residual material was then purified via silica flash chromatography (0–40% ethyl acetate in hexanes) to yield (400 mg, 97 % yield).

NMR spectra were recorded on a Varian 400 MHz (400/100), Varian 500 MHz (500/125) or a Varian Inova 600 MHz (600/150 MHz) equipped with a pulsed field gradient accessory. Chemical shifts are given in ppm (δ) relative to tetramethylsilane as an internal standard. Coupling constants are given in Hz. Purity of compounds tested in assays were analyzed via ¹H NMR.

3,5-difluorophenyl ethyl hex-5-yn-1-ylphosphonate (1):

¹H NMR (400 MHz, cdcl₃) δ 6.83 – 6.76 (m, 2H), 6.62 (dd, J = 10.1, 7.8 Hz, 1H), 4.28 – 4.08 (m, 2H), 2.22 (td, J = 6.9, 2.6 Hz, 2H), 1.97 – 1.72 (m, 6H), 1.63 (p, J = 7.0 Hz, 2H), 1.31 (t, J = 7.1 Hz, 3H). ¹³C NMR (101 MHz, cdcl₃) δ 100.65, 83.39, 68.95, 62.86, 62.79, 28.94, 28.77, 26.04, 24.64, 21.33, 21.27, 17.88, 16.30. ³¹P NMR (162 MHz, cdcl₃) δ 29.47. Purity: 98%.

ethyl (4-(fluorosulfonyl)phenyl) hex-5-yn-1-ylphosphonate (2):

¹H NMR (400 MHz, cdcl₃) δ 8.03 – 7.94 (m, 2H), 7.52 – 7.41 (m, 2H), 4.31 – 4.07 (m, 2H), 2.22 (td, J = 6.9, 2.7 Hz, 2H), 2.06 – 1.91 (m, 2H), 1.96 – 1.74 (m, 2H), 1.70 – 1.58 (m, 2H), 1.31 (t, J = 7.1 Hz, 3H). ¹³C NMR (101 MHz, cdcl₃) δ 130.77, 121.59, 121.54, 116.35, 69.03, 63.14, 63.07, 28.88, 26.26, 24.86, 21.28, 17.88, 16.36, 16.31. ³¹P NMR (162 MHz, cdcl₃) δ 29.91. ¹⁹F NMR (376 MHz, cdcl₃) δ −63.03, −65.87. Purity: 96%.

2,6-difluorophenyl ethyl hex-5-yn-1-ylphosphonate (3):

¹H NMR (400 MHz, cdcl₃) δ 7.08 – 7.01 (m, 1H), 6.97 – 6.89 (m, 2H), 4.37 – 4.18 (m, 2H), 2.22 (td, J = 7.0, 2.7 Hz, 2H), 2.09 – 1.96 (m, 2H), 1.94 (t, J = 2.7 Hz, 1H), 1.90 – 1.79 (m, 2H), 1.71 – 1.61 (m, 2H), 1.35 (t, J = 7.1 Hz, 3H). ¹³C NMR (101 MHz, cdcl₃) δ 124.84, 112.33, 112.11, 83.62, 68.77, 62.34, 29.15, 28.97, 26.40, 24.99, 21.59, 21.54, 17.98, 17.96, 16.25. ³¹P NMR (162 MHz, cdcl₃) δ 31.07. Purity: 98%.

2,6-dimethylphenyl ethyl hex-5-yn-1-ylphosphonate (4):

¹H NMR (400 MHz, cdcl₃) δ 7.02 – 6.89 (m, 4H), 4.13 – 4.02 (m, 1H), 3.99 – 3.88 (m, 1H), 2.23 (td, J = 7.0, 2.5 Hz, 2H), 1.99 (d, J = 7.4 Hz, 1H), 1.98 – 1.82 (m, 3H), 1.66 (p, J = 7.2 Hz, 2H), 1.15 (td, J = 7.1, 1.6 Hz, 4H). ¹³C NMR (101 MHz, cdcl₃) δ 130.44, 128.94, 128.92, 124.80, 83.62, 68.82, 62.74, 53.44, 29.23, 26.77, 25.34, 21.73, 17.99, 17.46, 16.33. ³¹P NMR (162 MHz, cdcl₃) δ 28.60. Purity: 95%.

ethyl (4-(methylthio)phenyl) hex-5-yn-1-ylphosphonate (5):

¹H NMR (400 MHz, cdcl₃) δ 7.21 – 7.17 (m, 1H), 7.12 – 7.08 (m, 2H), 4.22 – 4.05 (m, 2H), 2.42 (d, J = 1.8 Hz, 3H), 2.22 – 2.14 (m, 2H), 1.95 – 1.80 (m, 2H), 1.61 (q, J = 7.2 Hz, 2H), 1.26 (td, J = 7.0, 1.8 Hz, 3H). ¹³C NMR (101 MHz, cdcl₃) δ 148.38, 134.43, 128.47, 121.05, 83.56, 68.86, 62.51, 29.06, 28.89, 25.89, 24.49, 21.46, 17.93, 16.63, 16.40. ³¹P NMR (162 MHz, cdcl₃) δ 29.18. Purity: 97%.

4-(tert-butyl)phenyl ethyl hex-5-yn-1-ylphosphonate (6):

¹H NMR (400 MHz, cdcl₃) δ 7.34 – 7.26 (m, 2H), 7.12 – 7.05 (m, 2H), 4.25 – 4.05 (m, 2H), 2.19 (td, J = 7.0, 2.7 Hz, 2H), 1.94 – 1.71 (m, 6H), 1.67 – 1.57 (m, 2H), 1.27 (s, 9H). ¹³C NMR (101 MHz, cdcl₃) δ 147.65, 126.55, 119.83, 83.62, 68.80, 62.31, 34.33, 31.38, 29.11, 28.94, 25.87, 24.47, 21.52, 21.47, 17.95, 16.39, 16.33. 31P NMR (162 MHz, cdcl₃) δ 28.91. Purity 98%.

ethyl naphthalen-2-yl hex-5-yn-1-ylphosphonate (7):

¹H NMR (400 MHz, cdcl₃) δ 7.37 (t, J = 8.8 Hz, 3H), 7.08 – 6.88 (m, 2H), 3.88 – 3.68 (m, 2H), 1.79 (td, J = 7.0, 4.0 Hz, 2H), 1.59 – 1.37 (m, 5H), 1.26 – 1.20 (m, 1H), 0.91 – 0.86 (m, 3H). ¹³C NMR (101 MHz, cdcl₃) δ 148.24, 148.16, 133.92, 130.81, 129.84, 127.68, 127.46, 126.67, 125.37, 120.59, 116.83, 83.59, 68.88, 62.58, 29.09, 28.92, 25.99, 24.58, 21.48, 17.96, 16.44 ³¹P NMR (162 MHz, cdcl₃) δ 29.18. Purity: 97%.

4-bromo-2,6-dimethylphenyl ethyl hex-5-yn-1-ylphosphonate (8):

¹H NMR (400 MHz, cdcl₃) δ 7.16 – 7.11 (m, 2H), 4.14 – 4.01 (m, 1H), 4.01 – 3.87 (m, 1H), 2.31 (q, J = 0.8 Hz, 9H), 2.23 (td, J = 6.9, 2.6 Hz, 2H), 2.02 – 1.78 (m, 4H), 1.72 – 1.60 (m, 2H), 1.17 (t, J = 7.1 Hz, 3H). ¹³C NMR (101 MHz, cdcl₃) δ 146.85, , 132.75, 131.53, 117.57, 83.55, 68.89, 62.98, 29.18, 26.79, 25.36, 21.67, 17.99, 17.35, 16.39. ³¹P NMR (162 MHz, cdcl₃) δ 29.04. Purity: 98%.

ethyl (perfluorophenyl) hex-5-yn-1-ylphosphonate (9):

¹H NMR (400 MHz, cdcl₃) δ 4.36 – 4.17 (m, 2H), 2.23 (td, J = 7.0, 2.8 Hz, 2H), 2.12 – 1.77 (m, 6H), 1.65 (p, J = 7.0 Hz, 2H), 1.36 (t, J = 7.0 Hz, 3H). ¹³C NMR (101 MHz, cdcl₃) δ 81.06, 66.42, 66.15, 58.63, 58.56, 26.54, 26.37, 26.19, 23.46, 22.35, 22.03, 18.77, 15.40, 13.73. ³¹P NMR (162 MHz, cdcl₃) δ 32.67. Purity 97%.

[1,1’-biphenyl]-4-yl ethyl hex-5-yn-1-ylphosphonate (10):

¹H NMR (400 MHz, cdcl₃) δ 7.57 – 7.48 (m, 4H), 7.46 – 7.36 (m, 2H), 7.36 – 7.30 (m, 1H), 7.30 – 7.22 (m, 2H), 4.30 – 4.05 (m, 2H), 2.22 (tdd, J = 7.0, 2.7, 0.8 Hz, 2H), 1.99 – 1.77 (m, 4H), 1.71 – 1.59 (m, 2H), 1.31 (t, J = 6.9 Hz, 3H). ¹³C NMR (101 MHz, cdcl₃) δ 150.17, 140.33, 138.08, 128.90, 128.52, 127.39, 127.09, 120.89, 120.85, 83.71, 68.99, 62.63, 29.21, 26.11, 24.70, 21.64, 18.09, 16.49. ³¹P NMR (162 MHz, cdcl₃) δ 26.55. Purity: 98%.

ethyl phenyl hex-5-yn-1-ylphosphonate (11):

¹H NMR (400 MHz, cdcl₃) δ 7.34 – 7.27 (m, 2H), 7.22 – 7.10 (m, 3H), 4.25 – 4.05 (m, 2H), 2.20 (tdd, J = 6.9, 2.7, 0.8 Hz, 2H), 1.95 – 1.74 (m, 5H), 1.67 – 1.58 (m, 2H), 1.28 (td, J = 7.1, 0.8 Hz, 3H). 13C NMR (101 MHz, cdcl₃) δ 127.15, 122.26, 117.93, 81.02, 66.25, 59.87, 26.52, 23.39, 21.98, 18.9315.38, 13.82. ³¹P NMR (162 MHz, cdcl₃) δ 26.31. Purity: 97%.

4-cyclopentylphenyl ethyl hex-5-yn-1-ylphosphonate (12):

¹H NMR (400 MHz, cdcl₃) δ 7.19 – 7.13 (m, 2H), 7.07 (dt, J = 8.7, 1.2 Hz, 2H), 4.27 – 4.03 (m, 2H), 2.93 (ddd, J = 17.2, 9.7, 7.4 Hz, 1H), 2.20 (tdd, J = 6.9, 2.7, 0.9 Hz, 2H), 2.06 – 1.96 (m, 2H), 1.96 – 1.82 (m, 3H), 1.64 (ddt, J = 14.3, 9.4, 5.2 Hz, 4H), 1.54 – 1.44 (m, 1H), 1.28 (td, J = 7.1, 0.9 Hz, 3H). ¹³C NMR (101 MHz, cdcl₃) δ 148.40, 143.03, 128.22, 120.13, 120.08, 83.64, 68.81, 62.33, 62.26, 45.24, 34.62, 29.12, 28.95, 25.86, 25.39, 24.46, 21.53, 21.48, 17.97, 16.41, 16.35. ³¹P NMR (162 MHz, cdcl₃) δ 28.92. Purity: 98%.

ethyl (4-nitrophenyl) hex-5-yn-1-ylphosphonate (13):

¹H NMR (400 MHz, cdcl₃) δ 8.27 – 8.20 (m, 2H), 7.41 (m, 2H), 4.32 – 4.09 (m, 2H), 2.24 (td, J = 6.9, 2.6 Hz, 2H), 2.03 – 1.97 (m, 2H), 1.97 – 1.76 (m, 4H), 1.67 (q, J = 7.0 Hz, 4H), 1.33 (t, J = 7.1 Hz, 3H). ³¹P NMR (162 MHz, cdcl₃) δ 27.14. Purity 98%.

4-cyanophenyl ethyl hex-5-yn-1-ylphosphonate (14):

¹H NMR (400 MHz, cdcl₃) δ 7.65 – 7.57 (m, 3H), 7.35 – 7.27 (m, 3H), 4.29 – 4.02 (m, 4H), 2.20 (td, J = 6.9, 2.6 Hz, 3H), 1.96 – 1.92 (m, 1H), 1.92 – 1.71 (m, 7H), 1.67 – 1.53 (m, 4H), 1.26 (s, 3H). ¹³C NMR (101 MHz, cdcl₃) δ 154.04, 134.08, 121.42, 121.38, 118.27, 108.61, 83.39, 68.99, 62.93, 62.86, 28.93, 28.76, 26.18, 24.78, 21.34, 21.29, 17.91, 17.89, 16.39, 16.33. ³¹P NMR (162 MHz, cdcl₃) δ 29.60. Purity: 97%.

ethyl (4-methoxyphenyl) hex-5-yn-1-ylphosphonate (15):

1H NMR (400 MHz, cdcl₃) δ 7.12 – 7.04 (m, 3H), 6.84 – 6.75 (m, 3H), 4.18 – 4.07 (m, 2H), 3.74 (s, 3H), 2.18 (td, J = 7.0, 2.7 Hz, 3H), 1.88 – 1.75 (m, 6H), 1.61 (q, J = 7.5 Hz, 4H), 1.26 (d, J = 7.1 Hz, 3H). ¹³C NMR (101 MHz, cdcl₃) δ 156.54, 121.37, 121.33, 114.63, 114.62, 83.62, 68.81, 62.41, 62.34, 55.58, 29.11, 28.94, 25.78, 24.38, 21.51, 21.46, 17.95, 17.94, 16.42, 16.36. 31P NMR (162 MHz, cdcl₃) δ 29.20. Purity: 96%.

ethyl (4-isopropoxyphenyl) hex-5-yn-1-ylphosphonate (16):

¹H NMR (400 MHz, cdcl₃) δ 7.17 – 7.10 (m, 2H), 7.10 – 7.03 (m, 2H), 4.25 – 4.03 (m, 2H), 2.19 (td, J = 7.0, 2.8 Hz, 2H), 1.80 (s, 1H), 1.61 (m, J = 7.3 Hz, 2H), 1.30 – 1.17 (m, 9H), 0.88 – 0.77 (m, 2H). ¹³C NMR (101 MHz, cdcl₃) δ 127.56, 120.20, 120.16, 83.63, 68.80, 62.33, 62.26, 33.47, 29.11, 28.94, 25.87, 24.46, 24.04, 21.52, 21.47, 17.96, 17.94, 16.39, 16.33. ³¹P NMR (162 MHz, cdcl₃) δ 28.92. Purity: 98%.

ethyl (4-(pentafluoro-sulfaneyl)phenyl) hex-5-yn-1-ylphosphonate (17):

¹H NMR (400 MHz, cdcl₃) δ 7.76 – 7.65 (m, 2H), 7.33 – 7.26 (m, 2H), 4.30 – 4.06 (m, 2H), 2.22 (td, J = 6.9, 2.6 Hz, 2H), 1.99 – 1.73 (m, 5H), 1.70 – 1.58 (m, 2H), 1.31 (t, J = 7.1 Hz, 3H). ¹³C NMR (101 MHz, cdcl₃) δ 127.86, 120.49, 120.44, 83.40, 68.96, 62.85, 62.78, 28.96, 28.78, 26.12, 24.72, 21.36, 21.31, 17.91, 16.39, 16.33. ³¹P NMR (162 MHz, cdcl₃) δ 29.55. Purity: 96%.

4-acetylphenyl ethyl hex-5-yn-1-ylphosphonate (18):

¹H NMR (400 MHz, cdcl₃) δ 7.96 – 7.89 (m, 2H), 7.31 – 7.23 (m, 2H), 4.27 – 4.04 (m, 2H), 2.55 (d, J = 0.4 Hz, 3H), 2.20 (td, J = 7.0, 2.7 Hz, 2H), 1.98 – 1.89 (m, 2H), 1.80 (s, 2H), 1.69 – 1.56 (m, 2H), 1.28 (t, J = 7.1 Hz, 3H). 13C NMR (101 MHz, cdcl₃) δ 196.80, 154.42, 133.77, 130.39, 120.45, 120.41, 83.47, 68.93, 62.76, 62.68, 29.01, 28.83, 26.57, 26.14, 24.74, 21.41, 21.36, 17.93, 17.92, 16.40, 16.34. 31P NMR (162 MHz, cdcl₃) δ 29.20. Purity: 98%.

ethyl mesityl hex-5-yn-1-ylphosphonate (19):

¹H NMR (400 MHz, cdcl₃) δ 6.86 – 6.74 (m, 2H), 4.14 – 4.03 (m, 1H), 3.93 (dddd, J = 14.9, 10.2, 7.5, 5.1 Hz, 1H), 2.29 (s, 9H), 2.25 – 2.18 (m, 7H), 2.02 – 1.77 (m, 8H), 1.76 – 1.59 (m, 3H), 1.17 (ddd, J = 7.2, 6.3, 1.0 Hz, 3H). ¹³C NMR (101 MHz, cdcl₃) δ 129.97, 129.54, 129.52, 83.68, 68.80, 62.67, 29.27, 29.09, 26.75, 25.32, 21.76, 21.71, 20.59, 18.02, 18.01, 17.40, 16.39, 16.33. ³¹P NMR (162 MHz, cdcl₃) δ 28.67. Purity: 97%.

ethyl (2,3,5,6-tetrafluorophenyl) hex-5-yn-1-ylphosphonate (20):

¹H NMR (400 MHz, cdcl₃) δ 6.90 (ttd, J = 10.0, 7.2, 1.2 Hz, 1H), 4.41 – 4.19 (m, 2H), 2.24 (td, J = 6.9, 2.7 Hz, 2H), 2.12 – 1.99 (m, 2H), 1.99 – 1.79 (m, 3H), 1.73 – 1.61 (m, 2H), 1.38 (t, J = 7.1 Hz, 3H). Purity: 97%.

2,6-difluorophenyl ethyl pentadecylphosphonate (21):

¹H NMR (400 MHz, cdcl₃) δ 7.10 – 7.00 (m, 1H), 6.93 (t, J = 8.0 Hz, 2H), 4.40 – 4.18 (m, 2H), 2.06 – 1.94 (m, 2H), 1.69 (dd, J = 16.6, 7.7 Hz, 4H), 1.29 (d, J = 43.2 Hz, 26H), 0.86 (t, J = 6.7 Hz, 3H). ¹³C NMR (101 MHz, cdcl₃) δ 117.82, 111.41, 61.54, 61.47, 31.85, 30.58, 30.41, 29.61, 29.58, 29.55, 29.51, 29.29, 28.99, 26.18, 24.78, 22.60, 22.24, 22.18, 16.34, 16.28, 14.00. ³¹P NMR (162 MHz, cdcl₃) δ 31.97. Purity: 98%.

3,5-difluorophenyl ethyl pentadecylphosphonate (22):

¹H NMR (400 MHz, cdcl₃) δ 6.79 (ddd, J = 7.8, 2.4, 1.2 Hz, 2H), 6.61 (t, J = 8.9 Hz, 1H), 4.28 – 4.07 (m, 2H), 1.87 (dt, J = 18.0, 8.3 Hz, 2H), 1.64 (s, 3H), 1.36 – 1.22 (m, 26H), 0.86 (t, J = 6.3 Hz, 3H). ¹³C NMR (101 MHz, cdcl3) δ 104.75, 104.39, 100.43, 62.73, 31.89, 30.41, 29.94, 29.43, 29.31, 28.98, 25.11, 22.66, 22.16, 16.33, 14.08. ³¹P NMR (162 MHz, cdcl₃) δ 30.32. ¹⁹F NMR (376 MHz, cdcl₃) δ −107.84 (t, J = 8.0 Hz), −110.00 (t, J = 7.9 Hz ). Purity: 98%.

ethyl phenyl pentadecylphosphonate (23):

¹H NMR (400 MHz, cdcl₃) δ 7.40 – 7.24 (m, 3H), 7.18 (s, 2H), 7.23 – 7.09 (m, 3H), 4.26 – 4.07 (m, 2H), 1.94 – 1.81 (m, 2H), 1.73 – 1.62 (m, 2H), 1.27 (d, J = 16.7 Hz, 28H), 0.88 (t, J = 6.6 Hz, 3H). ³¹P NMR (162 MHz, cdcl₃) δ 27.11. Purity: 98%.

1-iodopentadecane (24):

¹H NMR (400 MHz, cdcl₃) δ 3.17 (t, J = 7.1 Hz, 2H), 1.80 (p, J = 7.1 Hz, 2H), 1.40 – 1.34 (m, 2H), 1.25 (d, J = 3.9 Hz, 22H), 0.86 (t, J = 6.7 Hz, 3H). ¹³C NMR (101 MHz, cdcl₃) δ 33.56, 31.90, 30.49, 29.66, 29.65, 29.63, 29.59, 29.52, 29.40, 29.33, 28.52, 22.67, 14.09, 7.33.

diethyl pentadecylphosphonate (25):

¹H NMR (400 MHz, cdcl₃) δ 4.34 – 3.85 (m, 1H), 1.79 – 1.63 (m, 1H), 1.54 – 0.97 (m, 12H), 0.86 (t, J = 6.6 Hz, 1H).

10-(tosyloxy)-15-(trimethylsilyl)pentadec-14-yn-1-yl propionate (26):

¹H NMR (400 MHz, cdcl₃) δ 7.78 (dd, J = 8.3, 3.3 Hz, 2H), 7.32 (t, J = 8.1 Hz, 2H), 4.08 (dt, J = 29.3, 6.4 Hz, 2H), 2.43 (d, J = 2.1 Hz, 3H), 2.35 – 2.21 (m, 2H), 2.13 (t, J = 6.9 Hz, 1H), 1.83 (p, J = 6.6 Hz, 1H), 1.72 – 1.37 (m, 4H), 1.37 – 1.08 (m, 6H), 0.14 – 0.04 (m, 9H). ¹³C NMR (101 MHz, cdcl₃) δ 129.83, 129.65, 127.88, 127.67, 68.88, 27.86, 25.87, 21.61, 16.06, 0.03.

15-(trimethylsilyl)pentadec-14-yn-1-ol (27):

¹H NMR (400 MHz, cdcl₃) δ 3.62 (t, J = 6.6 Hz, 2H), 2.19 (t, J = 7.2 Hz, 2H), 1.58 – 1.45 (m, 4H), 1.25 (s, 7H), 0.13 (s, 9H). ¹³C NMR (101 MHz, cdcl₃) δ 63.07, 32.79, 29.59, 29.58, 29.56, 29.55, 29.45, 29.40, 29.05, 28.77, 28.61, 25.71, 19.83.

pentadec-14-yn-1-ol (28):

¹H NMR (400 MHz, cdcl₃) δ 3.62 (t, J = 6.6 Hz, 2H), 2.16 (td, J = 7.1, 2.7 Hz, 2H), 1.60 – 1.44 (m, 4H), 1.37 (s, 3H), 1.25 (d, J = 3.9 Hz, 17H). ¹³C NMR (101 MHz, cdcl₃) δ 67.99, 63.07, 32.78, 29.57, 29.46, 29.39, 29.08, 28.73, 28.47, 25.70, 18.37.

15-iodopentadec-1-yne (29):

¹H NMR (400 MHz, cdcl₃) δ 4.11 – 4.02 (m, 4H), 2.16 (td, J = 7.1, 2.7 Hz, 2H), 1.77 – 1.64 (m, 2H), 1.64 – 1.44 (m, 4H), 1.40 – 1.24 (m, 19H). ¹³C NMR (101 MHz, cdcl₃) δ 67.99, 30.67, 29.55, 29.46, 29.35, 29.07, 28.73, 28.46, 24.96, 18.37, 16.41. ³¹P NMR (162 MHz, cdcl₃) δ 32.69.

diethyl pentadec-14-yn-1-ylphosphonate (30):

¹H NMR (400 MHz, cdcl₃) δ 4.11 – 4.02 (m, 4H), 2.16 (td, J = 7.1, 2.7 Hz, 2H), 1.77 – 1.64 (m, 2H), 1.64 – 1.44 (m, 4H), 1.40 – 1.24 (m,19). 13C NMR (101 MHz, cdcl3) δ 67.99, 30.67, 29.55, 29.46, 29.35, 29.07, 28.73, 28.46, 24.96, 18.37, 16.41. 32.69. 31P NMR (162 MHz, cdcl₃) δ 32.69.

2,6-difluorophenyl ethyl pentadec-14-yn-1-ylphosphonate (21-probe):

¹H NMR (400 MHz, cdcl₃) ¹H NMR (400 MHz, cdcl₃) δ 7.11 – 6.99 (m, 1H), 6.98 – 6.88 (m, 2H), 4.37 – 4.17 (m, 2H), 2.16 (td, J = 7.1, 2.6 Hz, 2H), 2.06 – 1.89 (m, 3H), 1.71 (q, J = 13.7 Hz, 2H), 1.58 (s, 1H), 1.49 (q, J = 7.2 Hz, 2H), 1.36 (dt, J = 14.1, 7.4 Hz, 9H), 1.28 – 1.22 (m, 15H). ¹³1C NMR (101 MHz, cdcl3) δ 124.72, 112.28, 112.06, 84.78, 67.99, 30.58, 30.40, 29.55, 29.52, 29.46, 29.31, 29.07, 29.01, 28.73, 28.46, 26.87, 25.47, 22.36, 22.30, 18.37, 16.25.³¹P NMR (162 MHz, cdcl₃) δ 31.86 (t, J = 1.9 Hz). Purity>99%.

3,5-difluorophenyl ethyl pentadec-14-yn-1-ylphosphonate (22-probe):

¹H NMR (400 MHz, cdcl₃) δ 6.81 – 6.75 (m, 2H), 6.61 (tt, J = 8.9, 2.3 Hz, 1H), 4.27 – 4.06 (m, 2H), 2.15 (td, J = 7.1, 2.6 Hz, 2H), 1.93 – 1.80 (m, 3H), 1.64 (ddt, J = 15.6, 11.9, 7.5 Hz, 2H), 1.55 – 1.44 (m, 2H), 1.37 (dq, J = 12.5, 7.1 Hz, 4H), 1.33 – 1.19 (m, 22H). ¹³C NMR (101 MHz, cdcl₃) δ 104.72, 104.44, 100.79, 100.53, 100.27, 84.76, 67.99, 62.69, 31.55, 30.48, 30.31, 29.53, 29.44, 29.28, 29.06, 28.98, 28.72, 28.45, 26.50, 25.11, 22.61, 22.18, 18.36, 16.36, 14.07. ³¹P NMR (162 MHz, cdcl₃) δ 30.29. Purity: >99%

Quantification and statistical analysis

Statistical details can be found in the corresponding figure legends. p < 0.05 was considered significant. For biochemical assays, the N numbers denoted in the figure legends refer to the number of times the experiment was run and the n describes the number of technical replicates for each experiment. The total number of replicates was used to caculate the standard deviation (SD) is used for biochemical assays For cell-based assays, n describes the biological replicates of cell treatments, performed at different times with biologically distinct samples. The number n was used to calculate standard error of mean (SEM) is used for cell-based assays. The number of technical replicates is described in the legends for each cell-based assay No statistical methods were used to predetermine sample size and investigators were not blinded to outcome assessment. GraphPad Prism 9 was used for statistical analyses using default settings.

Supplementary Material

2

Data S1: NMR Spectra for compound characterization, related to STAR methods

Significance.

Lipid metabolism and the dynamic regulation of lipid membranes are essential pathways within P. falciparum. Previously we used the potent antimalarial natural product Salinipostin A to identify a series of serine hydrolases which, when inhibited, results in parasite death. Here, we developed mixed alkyl/aryl phosphonates as tool compounds to aid in dissecting and identifying important regulators of lipid metabolism. We described the use of a pair of constitutional isomers to identify a serine hydrolase, abH112, as a possible drug target. Moreover, we showed how a previous attempt of a conditional knockout of abH112 led to gene duplication and that further methodologies are needed to investigate essential lipid metabolizing enzymes within the parasite. We also described a mechanism of killing not previously identified with SalA or the pan-lipase inhibitor Orlistat, opening up the possibility of multiple lipid metabolizing pathways that can be targeted in drug development. Lastly, we demonstrated that PRELI domain-containing protein is likely a mechanism of multidrug resistance with mutations found in it rather than targeted serine hydrolases, similar to SalA. Ultimately, we have described how mixed alkyl/aryl phosphonates can be used to study lipid metabolizing pathways and further bolstered serine hydrolases as an important class for antimalarial drug development.

Highlights.

• Mixed alkyl/aryl phosphonates target multiple serine hydrolases in Plasmodium falciparum.

• Inhibition of serine hydrolases blocks growth of P. falciparum parasites

• Serine hydrolase inhibitors with multiple targets induce only mild resistance in parasites

• Alkyl/Aryl phosphonate inhibitors synergize with orlistat to kill parasites