Combining Stage Specificity and Metabolomic Profiling to Advance Antimalarial Drug Discovery

Abstract

We report detailed susceptibility profiling of asexual blood stages of the malaria parasite Plasmodium falciparum to clinical and experimental antimalarials, combined with metabolomic fingerprinting. Results revealed a variety of stage-specific and metabolic profiles that differentiated the modes of action of clinical antimalarials including chloroquine, piperaquine, lumefantrine, and mefloquine, and identified late trophozoite-specific peak activity and stage-specific biphasic dose-responses for the mitochondrial inhibitors DSM265 and atovaquone. We also identified experimental antimalarials hitting previously unexplored druggable pathways as reflected by their unique stage specificity and/or metabolic profiles. These included several ring-active compounds, ones affecting hemoglobin catabolism through distinct pathways, and mitochondrial inhibitors with lower propensities for resistance than either DSM265 or atovaquone. This approach, also applicable to other microbes that undergo multiple differentiation steps, provides an effective tool to prioritize compounds for further development within the context of combination therapies.

Keywords: Plasmodium falciparum; asexual blood stages; drug resistance; hemoglobin catabolism; malaria; metabolomics; mitochondria; mode of action; target identification.

Figure 1

Experimental Design for Asexual Blood Stage Specificity Profiling of Antimalarials and Profiles of Reference Drugs (A) Synchronized parasites were exposed for 8 h at the stages indicated. Survival at 60 h post-invasion was assessed by flow cytometry. (B) Unique stage specificity profiles of chloroquine, dihydroartemisinin, and KAI407. Bar plots indicate the IC₅₀^8h when parasites were exposed only during the early ring, late ring, early trophozoite, late trophozoite, or schizont stage, with error bars showing the standard error of the mean based on at least three independent repeats. KAI407, a PI4K inhibitor. All data are available in Table S1.

Figure 2

Detailed Asexual Blood Stage Susceptibility Profiles for Antimalarials with Peak Activity on All Rings or All Rings and Trophozoites Data for chloroquine and dihydroartemisinin can be found in Figure 1. Bar graphs indicate mean IC₅₀^8h values, whereas survival graphs show the most representative curves from independent repeats. Error bars indicate the standard error of the mean based on >3 independent repeats. Data are summarized in Table S1.

Figure 3

Detailed Asexual Blood Stage Susceptibility Profiles for Antimalarials with Peak Activity on All Trophozoites Bar graphs indicate mean IC₅₀^8h values, whereas survival graphs show the most representative curves from independent repeats. Error bars indicate the standard error of the mean based on >3 independent repeats. Data are summarized in Table S1.

Figure 4

Detailed Asexual Blood Stage Susceptibility Profiles for Antimalarials with Peak Activity on Late Trophozoites, or on All Trophozoites and Schizonts Data for DSM265 and atovaquone, both compounds with peak activity at the late trophozoite stage, can be found in Figure 5. Bar graphs indicate mean IC₅₀^8h values, whereas survival graphs show the most representative curves from independent repeats. Error bars indicate the standard error of the mean based on >3 independent repeats. Data are summarized in Table S1.

Figure 5

Late Trophozoites Are the Most Susceptible Stage to DSM265 and Atovaquone that Inhibit Pyrimidine Biosynthesis and the Mitochondrial Electron Transport Chain, Respectively (A) Overview of the pyrimidine biosynthesis and the mitochondrial electron transport chain pathways. DSM265 inhibits DHODH, whereas atovaquone inhibits cytochrome bc1 (Goodman et al., 2017). (B) Stage specificity profiles for DSM265 and atovaquone. IC₅₀^8h values for (B) are available in Table S1.

Figure 6

Stage of Peak Activity for Clinical and Experimental Antimalarials Peak activity illustrates the period when the parasite was most susceptible to the tested compounds. MMV020746 and MMV665939 were omitted as their IC₅₀^8h values were >10 μM. All data are available in Table S1 and Figures 1, 2, 3, 4, and 5. ATQ, atovaquone; CQ, chloroquine; DHA, dihydroartemisinin; FQ, ferroquine; LMF, lumefantrine; MB, methylene blue; MQ, mefloquine; NQ, naphthoquine; PPQ, piperaquine.

Figure 7

Metabolic Profiling of Compounds Identified Cellular Processes Targeted by Compounds Compounds were clustered based on hydrophilic metabolite response to all measured metabolites (all data available in Table S4). Compounds are listed only if they showed a >2-fold change (log₂ > 1) in metabolite levels compared with untreated controls in at least one of the treated samples. Compounds are color-coded based on peak activity as shown in Figure 6. Metabolite data for chloroquine, DSM265, MMV000248, MMV006455, MMV019017, and KAE609 were sourced from (Allman et al., 2016). Data for all other 27 compounds were generated in this study. ATQ, atovaquone; Cmpd, compound; CQ, chloroquine; DHA, dihydroartemisinin; FQ, ferroquine; LMF, lumefantrine; MB, methylene blue; mETC, mitochondrial electron transport chain; MQ, mefloquine; NQ, naphthoquine; PPQ, piperaquine.