Chemoproteomics-based profiling elucidates the antimalarial effects of amodiaquine through disruption of glycolysis process in Plasmodium falciparum

Abstract

Background

Malaria poses a significant global health threat, and 4-aminoquinolines have played a pivotal role in the decades-long fight against malaria. Amodiaquine (AQ), a prominent member of the class, has been employed clinically for decades and, in combination with artesunate (AS), constitutes one of the most widely applied artemisinin-based combination therapies (ACTs). However, the precise molecular targets and antimalarial mechanisms of AQ remain incompletely understood.

Methods

We synthesized an AQ-derived activity probe (AQP) and systematically identified AQ-binding proteins using an activity-based protein profiling (ABPP) strategy. Integrative proteomic and transcriptomic analyses were then performed to characterize the pathways and potential targets associated with AQ action.

Results

We identified three glycolysis-associated enzymes as potential antimalarial targets of AQ. Subsequent validation experiments confirmed that AQ binds to these proteins and disrupts glycolytic processes in Plasmodium falciparum (P. falciparum). Moreover, we investigated the interactions between AS and AQ and demonstrated their complementary effects on shared molecular targets, suggesting a potential mechanism underlying the enhanced efficacy of AS-AQ combination therapy.

Conlusions

Our findings reveal that AQ exerts its antimalarial effects by binding to key glycolytic enzymes in P. falciparum and highlight the interplay between AQ and AS in targeting parasite metabolism. This work deepens the mechanistic understanding of AQ and AS-AQ, providing new insights into the mode of action of ACTs and offering potential strategies for future antimalarial drug development.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12964-026-02663-z.

Background

Malaria remains one of the most formidable infectious diseases that poses a serious threat to human health worldwide, with an estimated 282 million malaria cases and 610,000 deaths attributed to malaria in 2024 [1]. Pharmacotherapy control has held a pivotal position in the centuries-long fight against malaria, particularly with the advent of the quinolines preceding the discovery of artemisinin, which played a crucial role [2, 3]. The most successful of these was chloroquine (CQ), but unfortunately the rapid emergence of widespread resistance severely limited its use [4, 5]. Consequently, as an alternative to CQ, another important antimalarial drug is amodiaquine (AQ), which by possessing a 4-aminoquinoline scaffold, was developed and introduced into clinical use [6, 7]. Compared to CQ, AQ demonstrates superior effectiveness in inhibiting in vitro the proliferation of Plasmodium falciparum (P. falciparum), along with enhanced cost-effective, improved compliance, and tolerability in children, and crucially, sustained antimalarial activity against CQ-resistant strains [7, 8]. AQ is now commonly used in combination with artesunate (AS) as one of the most widely applied ACTs regimens, serving as a first-line treatment for the uncomplicated malaria, particularly in African countries [9–11]. Despite over seventy years of clinical use, the mechanism of action (MoA) of AQ remains limited.

AQ, which is a 4-aminoquinoline but that features a Mannich base component in its structure, shares same similarities with CQ, as a diprotic weak-base profile. However, AQ exhibits a lower pKa value (pKa1 = 7.1, pKa2 = 8.1) [12, 13]. Conventionally, AQ is thought to operate through a MoA which is in part similar to that observed for CQ: it accumulates within the acidic food vacuoles of the parasite and binds to heme, thereby inhibiting the hemozoin crystal formation and impeding the heme detoxification process of the parasite [14–16]. The elevated concentration of heme triggers the Fenton reaction, which generates a large number of oxygen free radicals, denatures proteins, may causes DNA damage and cell membrane disruption, and ultimately lead to the death of the parasite [14, 17, 18]. However, studies have shown that the antimalarial activity of AQ does not appear to be entirely explained by the weak base theory [13]. Despite being slightly less alkaline than CQ, its antimalarial activity is stronger than that of the former [13]. In addition to, CQ-resistant strains of P. falciparum remain susceptible to AQ [19, 20]. Thus, it appears that there are other antimalarial pathways that collectively contribute to the favourable antimalarial effect of AQ. Nevertheless, our understanding of the relevant mechanisms of action is scanty, and few studies have been reported.

Furthermore, studies have shown that AQ has pharmacological activity against Zika and dengue viruses, as well as potential effects in the treatment of melanoma and diabetes [21–24]. Therefore, it is of utmost importance to promptly ascertain the binding target of AQ and elucidate its MoA in order to optimize its clinical applications and expand its indications, especially in light of the increasingly prominent challenge of artemisinin resistance [25].

In this work, we performed a systematic identification of antimalarial targets of AQ using an activity-based protein profiling (ABPP) strategy [26, 27]. Followed by analysis of changes in protein and gene expression after AQ treatment using proteomics and transcriptomics approaches. Through a detailed analysis of these obtained dataset, three pivotal proteins associated with the glycolysis pathway were identified as potential antimalarial targets of AQ. Finally, a series of experimental validations were performed towards these targets, indicating that AQ conceivably exert the antimalarial effects by interfering with the glycolysis process of parasites through binding crucial proteins. This work significantly contributes to the advancement and deepening of our understanding of the antimalarial mechanism of AQ.

Methods

Parasites culture

The culture and synchronization of P. falciparum 3D7 strain was performed as our previously described [27, 28].

Antimalarial activity assay in vitro

This assay was performed based on the established 72 h fluorescent SYBR Green I method [29]. Briefly, the highly synchronized parasites during ring-stage were cultured in the 96-well plate at 2% hematocrit, 0.5% parasitemia and treated with gradient diluted compounds for 72 h. After lysed, the fluorescent intensity was measured and the results was analysised by the GraphPad Prism 8 software. Three biological independent replicates were performed and results are presented as mean ± standard error.

Fluorescence labeling assay

This assay was performed primarily following the published method [26, 30]. Unsynchronized parasites were cultured with 2% hematocrit, 5% parasitemia, and incubated with increasing concentrations of AQP or AP1. For the competition labeling, parasites were preincubated with excessive AQ or AS for 2 h and then treated with AQP for 2 h. Parasite proteins were then extracted by sonication after released from RBCs. 20 µg protein were used for the click reaction. Following that, proteins were subjected to precipitation using pre-chilled acetone, then resolubilized and separated via SDS-PAGE. Finally, fluorescence scanning and visualization were done with the Sapphire Biomolecular Imager (Azure Biosystems). Coomassie Brilliant Blue was employed as the loading control.

For the fluorescence labeling of recombinant purified target proteins, 2 µg protein were treated with AQP or AP1 for 2 h at room temperature. For the competitive labeling, the recombinant purified protein was pretreated with excess AQ, IAA or AS for 2 h and then treated with AQP, IAA-P or AP1 for another 2 h. After that, the click reaction, SDS-PAGE electrophoresis and fluorescence scanning analysis were performed as described above.

Intracellular fluorescence imaging

These assay was performed as previously reported [27, 31]. Parasites were seeded in a 24-well plate and treated with AQP or DMSO. Parasites were then dripped onto coverslips after fixed and permeabilized. Subsequently, the click reaction was performed. After that, the coverslips were moved to a slide and imaged with a Leica TCS SP8 SR confocal fluorescence microscope (Leica, Munich, Germany).

For immunofluorescence, after the click chemistry reaction, parasites were incubated with specific primary and fluorophore-linked secondary antibodies. Immunofluorescence images were captured using Dragonfly 200 spinning disk confocal microscopy, and semiquantitative analysis was done with the JACop plugin.

Pull-down experiments

The pull-down experiments were conducted to identify the antimalarial target proteins of AQ [30]. In brief, highly synchronized parasites were treated with AQP (1 µmol/L) or an equal volume of vehicle (DMSO) for 4 h. The competition experiment was conducted by pretreating with excess AQ followed by AQP. Then, the soluble parasite proteins were extracted and quantified for the click reaction to conjugate biotin-azide. Subsequently, proteins were then incubated with Neutravidin beads (Thermo Scientific, USA) for enrichment. After the proteins were reduced with dithiothreitol (DTT) and alkylated with iodoacetamide (IAA), they were subjected to enzymatic digestion at 37 °C overnight. Subsequently, the supernatant peptides were desalted, spin—dried, and labeled with Tandem Mass Tag (TMT) Labeling Reagents in preparation for liquid chromatography—tandem mass spectrometry (LC—MS/MS) analysis. Following this, the MS—data and bioinformatic analysis were carried out according to our previously described methods [27].

For the validation of targets by Western blotting, after the enrichment, the proteins were separated by SDS-PAGE gel electrophoresis, transferred to polyvinylidene fluoride (PVDF) membranes, incubated with corresponding primary and secondary antibodies and visualized using ECL.

Proteomics analysis

Firstly, parasites were cultured in a 10 cm dish and treated with DMSO or AQ for 4 h. Three independent biological replicates were carried out for each group. Parasites were then harvested and lysed as described previously [27, 30]. After quantification, equal amounts of parasite lysates were taken for the reduction and alkylation before digestion as mentioned above. Finally, the peptides were desalted and spin-dried for LC–MS/MS analysis. The proteins with absolute fold change ≥ 1.5 and P-value < 0.05 were considered to be significant differences.

RNA-sequencing (RNA-seq) analysis

The parasite RNA extraction and RNA-seq analysis were performed as described previously [27, 31, 32]. Briefly, the RNA of parasites was extracted using TRIzol RNA isolation reagent (Invitrogen, USA) and purified using poly-T oligo-attached magnetic beads. Next, the polymerase chain reaction (PCR) was then performed and the qualified libraries were sequenced on Illumina platforms (Novogene, CN). Three independent biological replicates were carried out for each group. The genes with log₂FC (fold change) > 0.5 and P < 0.05 were considered as differentially expressed genes (DEGs), and the Gene Ontology (GO) analysis was subsequently performed [31].

Expression and purification of recombinant proteins

As previously described [27], the coding gene sequence of target proteins acquired from PlasmoDB were synthesized commercially and cloned into the pET28a vector (GENEWIZ, China). The recombinant proteins were induced by isopropyl-β-d-thiogalactoside (IPTG) in Escherichia coli BL21. Then, proteins were extracted under a pressure of 1200 bar and purified using Ni-nitrilotriacetic acid (NTA) chromatography column.

Cellular thermal shift assay of target proteins

The binding of AQ to target proteins was evaluated using cellular thermal shift assays (CESTA) [33]. Parasites were lysed by sonication and equal amounts of proteins were incubated with AQ or DMSO for 1 h at room temperature. The mixtures were then aliquoted into PCR tubes and heated at 37–64 ℃ in a thermal cycler (Applied Biosystems, Thermo Scientific). After centrifugation, the supernatant was mixed with 1 × SDS loading buffer, heated at 95 °C, and subjected to Western blotting as mentioned above.

Surface plasmon resonance (SPR) assay

Recombinant proteins were immobilized on the surface of the CM5 Series S sensor chip (Cytiva Life Sciences, USA). Then different concentrations of AQ were flowed over the sensor chip surface The BIA evaluation software was used for kinetic analysis. The binding affinity (KD) was evaluated based on the concentration dependence on the steady-state response.

AQ binding-site identification

The binding site of AQ to the recombinant protein PfLDH was identified as our previously described with minor modifications [30]. Briefly, PfLDH(5 µmol/L) was incubated with AQ(100 µmol/L) for 4 h at room temperature. After removal of excess AQ, the reduction and alkylation experiments were performed before digestion as mentioned above. Finally, the peptides were desalted and spin-dried for LC–MS/MS analysis.

Molecular docking simulation assay

As previously described [30, 31], the 3D structure of AQ was downloaded from PubChem (CID: 2165). The structure of PfLDH (PDB: 1T24) was downloaded from the Protein Data Bank (PDB) database. These structures were loaded into MOE for molecular docking simulation. After preparation, the N58 of PfLDH was selected as the docking site and the docking pocket was set to 4.5 Å.

Recombinant protein activity assay

The enzyme activities of the recombinant proteins PfLDH, PfPGK and PfGAPDH were measured in vitro according to the manufacturer's instructions of the LDH activity assay kit (Beyotime, China), the PGK activity assay kit (Abcam, UK), and the Glyceraldehyde-3-phosphate dehydrogenase activity assay kit (Abcam, UK) as previously described [27, 30].

Determination of lactate level

The lactic acid level was measured using the Lactic Acid (LA) Content Assay Kit (Solarbio, China, BC2235). Parasites were seeded in a 6-well plate and incubated with different concentrations of AQ. Subsequently, supernatant medium and infected red blood cells (iRBCs) were collected for the measurement of lactate according to the manufacturer's instructions.

Extracellular flux analysis using the XFe96 analyzer

The Seahorse XF Glycolysis Stress Test Kit and Agilent Seahorse XFe96 Analyzer were used for extracellular acidification rate (ECAR) as described previously [34]. As brief, the sensor cartridge was pre-hydrated at 37 ℃ overnight. And parasites were treated with increasing concentrations of AQ, released with 0.01% pre-warmed saponin. Then, parasites were resuspended in Seahorse XF RPMI 1640 medium (Agilent Technologies) containing 1 mmol/L glutamine (Agilent Technologies) and the same amounts of parasites were seeded in Seahorse mini plates precoated with 0.01% (w/w) polylysine. Immediately after, 10 mmol/L glucose, 1 µmol/L oligomycin and 50 mmol/L 2-DG were added to the corresponding wells sequentially for the measurement.

Results

Design and synthesis of AQP

AQ has the same 7-chloroquinoline ring system as CQ, but differs by possessing a Mannich base side chain. We have functionally modified the structure of AQ by using a click chemistry strategy while preserving its indispensable antimalarial active groups (7-chloroquinoline ring, Mannich base side chain) [35]. We designed and synthesized AQP, a probe based on AQ, as shown in the synthetic scheme of Fig. 1a, by introducing a click-chemistry-capable alkyne report handle at the 4’-hydroxyl position of the p-hydroxyanilino aromatic ring of its side chain. Subsequently, we assessed the in vitro antimalarial activity of AQP against P. falciparum 3D7 strain and compared with AQ. As shown in Fig. 1b, AQP probe has retained a similar antiplasmodial activity as AQ, which is in nanomolar range.

Fig. 1.

Identification of amodiaquine (AQ) target proteins by AQ activity probe (AQP) based on ABPP strategy. a Synthesis scheme of AQP. b Determination of the antimalarial activity of AQP and AQ in P. falciparum 3D7 strain. c A general workflow of AQP-mediated ABPP strategy used to label and identify AQ targets in vivo. d In situ labeling of AQP in P. falciparum in a dose-dependent manner. e Excessive AQ can specifically diminish the labeling of AQP in P. falciparum. f Fluorescence confocal imaging showed AQP (1 µmol/L) rapidly distributed in parasites and can be eliminated by excess AQ (Scale bar = 2 µm). Abbreviations: ABPP, activity-based protein profiling; CBB, Coomassie brilliant blue; HZ, hemozoin; TAMRA, carboxytetramethylrhodamine; iRBC, infected red blood cells

Fluorescence labelling of AQP targets in parasites

After confirming the favorable antimalarial effect of the AQP, we then performed in situ fluorescence labeling experiments using the AQP to observe the binding of the probe to parasite proteins, as illustrated in Fig. 1c. As shown in Fig. 1d, the fluorescence intensity of TAMRA-labelled parasite proteins increased in a dose-dependent manner with AQP. Importantly, the fluorescence signal was significantly diminished after pre-incubation with excess AQ, indicating that AQ and AQP interact with the same target proteins in parasites (Fig. 1e). Furthermore, we also carried out a live cell imaging experiment to evaluate the distribution of AQP in the parasite. As shown in Fig. 1f, the results showed that the AQP could distribute and accumulate inside the parasites, with additional excess AQ effectively displacing it. Overall, the AQP retained the antimalarial activity and target binding specificity similar to that of AQ, opening up the possibility of further utilizing it to identify antimalarial targets of AQ.

Identifying the binding targets of AQ using ABPP strategy

Given the favourable labeling efficiency and target specificity of AQP, we further used it to identify the potential antimalarial targets of AQ. After incubation of AQP with the parasites, AQP-labeled target proteins were conjugated with biotin-azide tags via the copper catalyzed azide-alkyne cycloaddition (CuAAC). Subsequently, AQP-binding proteins were retrieved by biotin-streptavidin affinity beads for subsequent quantitative proteomics using high-resolution mass spectrometry (Fig. 1c). Following identification and a rigorous filtering, a total of 50 proteins were identified with high confidence as potential targets for the antimalarial activity of AQ (Fig. 2a, S1, Table S1). We then performed bioinformatic analysis of the identified target proteins and the Gene Ontology (GO) analysis showed that several physiological pathways were enriched. Interestingly, most of them were associated with glycolysis and energy metabolism pathways (Fig. 2b), and P. falciparum relies heavily on anaerobic glycolysis for energy for its own growth and proliferation during the intraerythrocytic stage [36].

Fig. 2.

Identification of critical targets of AQ by integrating transcriptomic and proteomic analysis. a Scatter plot of 50 target proteins identified by AQP ABPP-based experiments. The x-axis represents the mean of log₂ protein abundance difference between AQP (1 µmol/L) and control (DMSO) group, while y-axis represents the mean of log₂ difference between AQP (1 µmol/L) and AQ (5 µmol/L) + AQP (1 µmol/L) group. Three independent biological replicates were set up for each experiment. The fold change (FC) of protein abundance > 1.2 (dashed line) and P < 0.05 were used as the screening criterion of target proteins. b Gene ontology (GO) analysis of the enriched biological process (BP) for 50 targets identified of AQ by AQP-based ABPP. c-d Volcano plot representing DEGs (P < 0.05, log₂ (FC) > 0.5) of P. falciparum after treatment with AQ versus the control (DMSO) in the proteomic analysis (c) and transcriptomic analysis (d). e Venn diagrams of up-regulated differential proteins of the transcriptome, proteome and identified targets by AQP. f Gene ontology (GO) analysis of the enriched biological process (BP) for overlapped targets from (e)

Integrating transcriptomics and proteomics to identify the critical targets of AQ

To further ascertain which of the AQP-hooked proteins were the critical targets, we followed up with a combination of proteomic and transcriptomic analysis to explore the targets. Initially, we analysed the global changes in the proteome and transcriptome of P. falciparum following AQ treatment. As shown in Fig. 2c, significant changes occurred at the protein level, with 317 significantly up-regulated and 340 significantly down-regulated. Similarly, significant modulation occurred at the gene expression level after AQ treatment, with 1079 genes significantly up-regulated and 995 genes significantly down-regulated with screening criteria of P < 0.05 and log₂ (FC) > 0.5 (Fig. 2d). Further GO analysis of these differentially expressed genes highlighted their involvement in various physiological processes within the parasite (Figure S2). We then combined the above results to examine the expression changes of the identified target proteins, and found that 13 targets were significantly up-regulated at both levels (Fig. 2e), while only one was down-regulated (Figure S3). Subsequent GO analysis of these 13 targets showed that the glycolytic process was significantly enriched (Fig. 2f). Notably, among the targets captured by AQP, three proteins implicated in this physiological process stood out: L-lactate dehydrogenase (PfLDH, PF3D7_1324900), glyceraldehyde-3-phosphatedehydrogenase (PfGAPDH, PF3D7_1462800), and phosphoglycerate kinase (PfPGK, PF3D7_0922500) (Figure S1). Previous studies have underscored the importance of these three proteins in the glycolysis process of P. falciparum and therefore may be crucial antimalarial targets of AQ [37–39].

Binding verification of the AQ target proteins

First, these three glycolysis-associated enzymes were recombinantly expressed, successfully purified in vitro, and further evaluated in detail for interaction with AQ. As shown in Fig. 3a-b, the fluorescence labeling results indicated that AQP could bind to these three proteins in a dose-dependent manner and could be competitively diminished by excess preincubation of AQ. Meanwhile, the labelling and competition effects in living cells were similarly observed by pull-down western blotting in situ (Fig. 3c) and immunofluorescence experiments (Fig. 3d-e). In contrast, the results of the cellular thermal shift assay (CETSA) experiment also indicated that the thermal stability of all three target proteins was improved to some extent after AQ treatment (Fig. 3f, S4). Additionally, we also performed the surface plasmon resonance (SPR) assay to determine the binding affinity between AQ and the three targets using the Biacore T200 platform, and all events were at a low level (Fig. 3g-i). Taken together, the above experimental evidence supported the binding of AQ to all three glycolysis-related targets. Having confirmed the binding between the AQ and these targets, we are now interested in the regions where the AQ binds to the target. We first performed competitive labelling experiments with a cysteine-targeted probe IAA-alkynyl probe (IAA-P) and observed that the AQ significantly attenuated the labelling of the IAA-P, supporting the possibility that the cysteine may be the binding sites of AQ (Fig. 3j). Subsequently, we used mass spectrometry to precisely localize the binding site of AQ by incubation with the PfLDH, and identified the asparagine 58 (N58) as the binding site (Fig. 3k). This result was further validated by docking simulation as well as the diminished fluorescence labelling of single point mutants (Fig. 3l-m).

Fig. 3.

Binding validation of the 3 critical target proteins of AQ. a Fluorescence labeling of recombinant PfLDH, PfGAPDH, PfPGK with AQP in a dose-dependent manner. b Pretreatment of excess AQ can diminish the AQP binding to the 3 proteins. c Validation of the binding of AQP to the critical target proteins in situ by pull-down western blotting. d-e Quantitative analysis of Co-localization imaging of AQP with 3 target proteins. f Thermal stability of 3 proteins increases after binding to the AQ. g-i Determination of binding affinity of AQ and 3 proteins by SPR assay. j AQ competes with the binding of IAA-P to the 3 proteins. k Identification of the binding site of AQ on the recombinant PfLDH protein. AQ may bind to the asparagine 58 (N58) of recombinant PfLDH. l Docking simulation of AQ binding to recombinant PfLDH (PDB:1T24) at N59 residue with binding score of −5.1. m Fluorescence labeling of AQP with the wild type and single-site mutants (N58A) of recombinant PfLDH

AQ disrupts glycolytic processes of parasites by binding to critical targets

After validating binding of AQ to the target proteins, we further investigated the effect of AQ on the function of the target proteins as well as the glycolytic pathway involved. As shown in Fig. 4a-c, the catalytic activities of the three identified critical enzymes of glycolysis could be significantly inhibited by AQ in a dose-dependent manner. Immediately after, we evaluated the changes of glycolytic processes in P. falciparum after AQ treatment using a Seahorse XF96 extracellular flux analyzer, by measuring the changes in extracellular acidification rate (ECAR) [34, 40]. As shown in Fig. 4d, the glycolytic processes of P. falciparum were significantly diminished by AQ in a concentration-dependent manner.

Fig. 4.

Functional verification of the 3 critical target proteins of AQ. a-c AQ inhibits the enzymatic activities of recombinant PfLDH protein. d The lactate levels of both in culture medium and (e) parasites decrease significantly after AQ treatment. f The inhibition of AQ on the glycolytic process of P. falciparum in a dose-dependent manner. All data are obtained from three independent experiments and presented as mean ± SEM (*P < 0.05, **P < 0.01, ***P < 0.01)

In addition to, we measured the levels of lactate in both parasites and culture media, which is the major end product of glycolysis [41]. We then observed from Fig. 4e-f that both lactate levels significantly decreased with increasing AQ concentration, which also reflected the diminishing in the glycolytic process. Thus, the above experimental results collectively suggest that AQ may exert antimalarial effects by binding to critical target proteins associated with glycolysis, and inhibiting their activity, thereby interfering with the glycolytic processes of P. falciparum.

AQ and AS synergistically target the glycolytic pathway in P. falciparum

Although we have extended our exploration into the antimalarial targets and mechanism of AQ, AQ is used in combination with AS (AQ-AS). Therefore, investigating the interactions between AQ and AS was relevant to study here. Initially, based on previous transcriptomic analysis of ATS [31], we compared the effects of individual treatments with AQ or ATS on the transcriptional levels of P. falciparum parasites. The results revealed a significant overlap in the transcriptional alterations following both treatments, which involved interference with multiple crucial physiological and metabolic processes in the parasites, such as glycolysis and hemoglobin metabolism (Figure S5).

Next, we employed AQP as well as our previously developed artemisinin activity probe (AP1) to perform in situ fluorescence labeling of parasites. As shown in Fig. 5a, AQ significantly promoted AP1 binding to parasite proteins. Interestingly, when AS was co-incubated with AQP, the fluorescence intensity of AQP markedly decreased, suggesting that AS competes more efficiently and specifically for the AQP target proteins (Fig. 5b). This observation suggests that AS likely binds a broader set of target proteins than AQ.

Fig. 5.

The interaction between AS-AQ and parasite proteins. a AQ Significantly enhances AS binding to proteins in P. falciparum. b AS markedly reduces AQ binding to proteins in P. falciparum. c Comparison of antimalarial binding targets of AQ and AS. d AS substantially weakens AQP binding to the three target proteins. e AQ significantly reduces AP1 binding to PfLDH While enhancing its binding to PfGAPDH and PfPGK. f Molecular docking simulation showing AQ and AS binding near the same pocket on PfLDH

We then further analysed the binding targets of both AQ and AS. Comparative analysis of the AQ targets identified in this study and previously identified artemisinin targets using ABPP techniques revealed 32 common proteins, including the three validated target proteins. Bioinformatics analysis indicated that these proteins are primarily associated with the glycolytic pathway in P. falciparum, implying that disruption of glycolysis may be a crucial pathway through which the combination of AS-AQ exerts their effects (Fig. 5c, S5). As shown in Fig. 5d, the fluorescence labeling of the three target proteins by AQP was markedly diminished when co-incubated with AS. Interestingly, we found that when AQ and AP1 interacted with the target proteins simultaneously, the fluorescence labeling of PfLDH by AP1 decreased significantly, while labeling of PfGAPDH and PfPGK was significantly enhanced (Fig. 5e). Additionally, we identified the same peptide binding sites for AQ on PfLDH as previously identified for AS binding to PfLDH. Molecular docking simulations also indicated that AQ and AS could bind near the same pocket on PfLDH (Fig. 5f).

These findings suggest that the presence of both synergistic and competitive binding interactions between AQ and AS can occur on the same targets, potentially due to the differences in drug-protein affinity or conformational changes in the proteins upon drug binding. Overall, the combined use of AQ and AS results in a notably enhanced antimalarial effect.

Discussion

The MoA of 4-aminoquinolines, as one of the most successful drugs in the antimalarial therapy, has been a focal point in the field. Despite significant progress made by numerous studies, our understanding remains incomplete. Moreover, the declining prevalence of resistance alleles associated with quinolines underscores the potential for repurposing this drug class [42]. Therefore, elucidating their MoA is crucial for optimizing current clinical dosing regimens. As we know, AQ was originally introduced as an alternative to chloroquine (CQ) and is one of its phenyl-substituted analogues [6]. Both drugs belong to the 4-aminoquinoline class of antimalarials and have a similar core structure, featuring the same aminoquinoline ring. However, a key dissimilarity between AQ and CQ is AQ's Mannich base side chain, which contains a p-hydroxyaniline aromatic ring. This alters the pKa, lipophilicity and flexibility of the side chain [13, 43], thereby enhancing AQ's hydrophilicity and reactivity towards nucleophilic amino acid residues [44]. Consequently, AQ exhibits a superior binding affinity to P. falciparum proteins than CQ. This structural difference also explains why amodiaquine is a more effective drug for malarial treatment and remains effective against most CQ-resistant strains.

Although AQ’s antimalarial efficacy is superior to that of CQ, we know less about its MoA compared to CQ. The accumulation effect of quinoline antimalarial drugs in parasite acidic food vacuole is usually considered as the main reason for their antimalarial activity (Fig. 6) [45].

Fig. 6.

AQ accumulates in the acidic food vacuole of P. falciparum and inhibits the heme detoxification into hemozoin crystals. This treatment simultaneously interferes with the glycolytic processes of P. falciparum, jointly contributing to the antimalarial mechanisms

Here, we used a 4′-hydroxyl group modified active probe AQP to identify the binding targets of AQ. The biologically active conformation of AQ depends on an intramolecular hydrogen bond between the phenolic hydroxyl group and the protonated nitrogen atom in the N,N-diethylamino side chain [35]. However, the exact contribution of the hydroxyl group to the antimalarial activity of AQ remains elusive. Previous studies have demonstrated that substitution of the 4-hydroxyl group yields AQ analogues with enhanced oxidative and metabolic stability by preventing the formation of cytotoxic quinone-imine metabolites–known to be responsible for AQ-related adverse effects [46, 47]. Although these analogues exhibited a moderate reduction in potency (approximately sevenfold lower activity compared to the parent AQ, consistent with our findings), they retained antimalarial potency at nanomolar range [35, 48, 49]. Moreover, derivatives that preserve the quinoline aromatic system but lack the phenolic hydroxyl group were still able to interact with heme, indicating that the hydroxyl group is not essential for heme binding [50]. Collectively, these findings suggest that while the hydroxyl moiety may contribute to AQ’s overall bioactivity, it is not indispensable, and functional modification at this position represents a rational strategy for structure optimization to mitigate toxicity while maintaining efficacy [51]. Importantly, our in situ fluorescence labeling results further confirmed that AQP-labeled P. falciparum proteins could be almost completely competed away by excess AQ, demonstrating identical target engagement and underscoring the high specificity of the AQP probe. Then combined with transcriptomic and proteomic analyses to focus the final target on three important proteins related to parasite glycolysis, which was verified by a series of experiments, indicating that interference with the glycolytic pathway may indeed be a way for AQ to produce antimalarial effects (Fig. 6).

Previous studies consistently indicate that these enzymes are predominantly cytosolic in P. falciparum, corresponding to their role in glycolysis during the intraerythrocytic stage [36, 52]. Specifically, PfLDH and PfPGK are canonical cytosolic enzymes responsible for energy generation [53–55], while PfGAPDH, although primarily cytosolic, has been reported to exhibit “moonlighting” functions and transient association with vesicular or apical structures under certain physiological conditions [56, 57]. Therefore, it is reasonable to infer that AQ likely interacts with its target proteins within the parasite cytosol or at the digestive vacuole. This also further supports our conclusion that AQ may exert antimalarial effects not only through the classical “acidic-vacuole accumulation and heme detoxification inhibition” mechanism, but also by perturbing cytosolic glycolytic processes essential for parasite energy metabolism.

It is also interesting to note that some of the targets identified in this work were also identified in our previous work on the MoA of CQ and similar conclusions were reached, further supporting the idea that interference with the glycolytic process may be an important MoA of 4-aminoquinolines [27]. It is conceivable that significant suppression of the glycolytic pathway, coupled with the parasite's inability to compensate, triggers cascading effects affecting multiple systems [58]. Concurrently, decreased lactate concentrations induce cytoplasmic alkalinization, which affects multiple biological systems [58]. For instance, reduced lactate levels result in cytoplasmic alkalinization, further impairing hemoglobin metabolism and pyrimidine biosynthesis, ultimately leading to parasite death [59]. Despite the advances made in elucidating the antimalarial mechanism of AQ through the ABPP technology, it remains to be comprehensively investigated whether the binding of these identified targets to AQ, in addition to exerting direct inhibitory effects on enzymatic activities, also has implications for secondary or downstream cellular processes beyond the glycolytic pathway. Furthermore, while our data demonstrate a clear association between AQ binding and functional inhibition of these enzymes, the precise molecular events underlying this inhibition, including potential conformational changes or alterations in protein–protein interactions, have not yet been fully delineated. Addressing these questions will require further structural and mechanistic studies to comprehensively define how AQ binding modulates enzyme function and physiological processes.

Moreover, AQ is a cost-effective drug, which has made its combination with AS one of the most widely used ACT regimens. The MoA of AS-AQ antimalarial activity is commonly attributed to the complementary pharmacokinetic/pharmacodynamic (PK/PD) characteristics of AQ and AS [60]. However, studies indicate that their pharmacokinetic properties change when co-administered, as opposed to when each is used alone, suggesting additional potential interactions [61, 62]. Our findings further reveal that in clinical antimalarial therapy, AQ not only extends the duration of antimalarial activity in the bloodstream, widening the therapeutic window, but may also enhance AS efficacy by strengthening its binding to parasite proteins. This likely occurs because AQ inhibits the detoxification process whereby free heme aggregates into hemozoin, thus raising heme concentration and leading to enhanced activation of artemisinin.

To the best of our knowledge, this is the first report employing ABPP to explore the antimalarial targets and underlying mechanisms of action of AQ and to examine the interaction between AQ and AS in ACT therapy, thereby deepening our understanding of the mechanisms of both AQ and AS-AQ. Although the use of AQ has been limited by increased adverse events due to immunologic side effects [35], it remains a first-line therapy in combination with AS in several countries, demonstrating significant advantages [63]. Importantly, AQ offers potential for developing the new artemisinin-based triple therapies [64]. This work provides some theoretical basis for further optimization of the current clinical use of AQ, its redevelopment and expansion of indications. Additionally, it offers a new research strategy for investigating interactions within ACT regimens and understanding their combined antimalarial mechanisms.

Conclusions

This work lays a foundation for a more comprehensive understanding of the antimalarial mechanisms of AQ and AS-AQ, deepening our insight into the MoA of quinoline-based drugs, particularly those leveraging the classical acid accumulation pathway. It also provides potential strategies for elucidating the antimalarial mechanisms of ACT drugs. This is of great significance for the alleviation of drug resistance, accelerating the optimization of existing clinical drug regimens and expanding indications. Meanwhile, the identified key targets and related glycolytic pathway may serve as a valuable direction for future antimalarial drug development.

Abbreviations

ABPP

Activity-based protein profiling

ACT

Artemisinin-based combination therapy

AQ

Amodiaquine

ART

Artemisinin

AS

Artesunate

BLI

Bio-Layer Interferometry

CBB

Coomassie brilliant blue

CESTA

Cellular thermal shift assays

AQP

Amodiaquine active probe

DTT

Dithiothreitol

GO

Gene ontology

HZ

Hemozoin

IAA

Iodoacetamide

P. falciparum

Plasmodium falciparum

PfLDH

P. falciparum L-lactate dehydrogenase

PfGAPDH

P. falciparum Glyceraldehyde-3-phosphatedehydrogenase

PfPGK

P. falciparum Phosphoglycerate kinase

RBC

Red blood cells

RNA-seq

RNA-sequencing

TAMRA

Carboxy tetramethyl rhodamine

Authors’ contributions

Jigang Wang, Peng Gao and Xiao Chen conceived the project and designed the experiments. Jianyou Wang performed the major experiments. Chen Wang, Ruishen Zhuge and Huan Tang performed bioinformatic analysis and data visualization. Fei Xia, and Ying Zhang, contributed to data interpretation. Junzhe Zhang, Cui Liu and Jiao Wu performed mass spectrometry data acquisition. All authors contributed to manuscript revision and approved it for submission.

Funding

The work was supported by grants from the National Natural Science Foundation of China (82521104, 82441001, 82304577 and 82274182); Taishan Scholars Program (tstp20240848); Introduce innovative team projects of Jinan (202228029); the Scientific and Technological Innovation Project of China Academy of Chinese Medical Sciences (CI2023D003 and CI2023E005TS05); the Beijing Natural Science Foundation (7262208); the Fundamental Research Funds for the Central Public Welfare Research Institutes (ZZ18-YQ-056, ZZ18-ND-10–12, ZZ13-ZD-07, ZZ17-ND-10 and ZZ18-ND-10); the distinguished Expert Project of Sichuan Province Tianfu cholar (CW202002).

Data availability

The datasets used and/or analyzed during the present study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

No applicable.

Consent for publication

No applicable.

Competing interests

The authors declare no competing interests.