Disrupting malaria parasite AMA1-RON2 interaction with a small molecule prevents erythrocyte invasion

Abstract

Plasmodium falciparum resistance to artemisinin derivatives, the first-line antimalarial drug, drives the search for new classes of chemotherapeutic agents. Current discovery is primarily directed against the intracellular forms of the parasite. However, late schizont-infected red blood cells (RBCs) may still rupture and cause disease by sequestration; consequently targeting invasion may reduce disease severity. Merozoite invasion of RBCs requires interaction between two parasite proteins AMA1 and RON2. Here we identify the first inhibitor of this interaction that also blocks merozoite invasion in genetically distinct parasites by screening a library of over 21,000 compounds. We demonstrate that this inhibition is mediated by the small molecule binding to AMA1 and blocking the formation of AMA1-RON complex. Electron microscopy confirms that the inhibitor prevents junction formation, a critical step in invasion that results from AMA1-RON2 binding. This study uncovers a strategy that will allow for highly effective combination therapies alongside existing antimalarial drugs.

Figure 1. Quantitative high-throughput assay to identify inhibitors of the AMA1-RON2 interaction

(a) In the AlphaScreen, streptavidin-coated donor beads captures biotin-tagged RON2L peptide and the nickel-coated acceptor beads binds to His-tagged AMA1(3D7 allele). In the absence of inhibitor, excitation of the donor beads at 680nm results in production of singlet oxygen, followed by short-distance diffusion (< 200 nm) and energy transfer to the acceptor beads, in turn resulting in emission at 520–620 nm. Disruption of the interaction leads to reduced or no signal (b) R1 peptide that specifically binds 3D7 allele of AMA1 (square) and the unlabeled RON2L peptide (black circle) were used as positive control for inhibitors in the AlphaScreen assay. Error bars show ± SEM from 2 independent experiments.

Figure 2. Small molecules block AMA1-RON complex formation and inhibit merozoite invasion

(a) Purified merozoites were used to test the effect of the three compounds on invasion of RBCs at 25 μM (white bars) and 50 μM (black bars) for 4 hr. Error bars show ± SEM from five experiments for NCGC00015280, NCGC00181034 and two for NCGC00014044. (b) Immunoprecipitation assay testing the ability of the inhibitors to block parasite AMA1-RON complex formation. Each inhibitor was used at 100 μM concentration and was immunoprecipitated using anti-RON4 antibody. RON2L peptide was used as a positive control. DMSO (1%), the solvent for the inhibitors, was used as a negative control. Experiments were performed twice and a representative western blot data is shown. (c) NCGC00015280 inhibits merozoite invasion of genetically distinct parasite clones. Purified schizonts from four different parasite clones were allowed to rupture and invade new RBCs for 4 to 6 hr in the presence of varying concentrations of the inhibitor. The number of newly invaded rings was measured by flow cytometry of SYBR green labeled parasites. IC₅₀: 12 μM (FVO), 14 μM (3D7), 13 μM (DD2) and 10 μM (HB3). Error bars show ± SEM from two experiments for each parasite clone. (d) Merozoite release from schizont-infected RBCs is not affected. The effect of the inhibitors on merozoite release was tested at 30 μM, the IC₅₀ for invasion. Error bars represent ± SEM from three experiments for NCGC00015280, NCGC00181034 and two for NCGC00014044. The number of parasites in the absence of inhibitor was considered 100%.

Figure 3. Improved efficiency of analogs in blocking merozoite invasion

(a) Structure of two analogs that showed improved potency. (b) Schizont-infected RBCs were allowed to rupture and invade new red cells for 4 hr in the presence of 15 μM of the parent compound (black bar) or the two analogs (grey bars). The number of newly invaded rings was measured by flow cytometry. The number of parasites in the absence of inhibitor was considered 100%. Error bars represent ± SEM from four independent experiments for each compound. (c) Purified invasive merozoites were allowed to invade RBCs and develop for 3 to 4 hr in the presence of varying concentrations of the two analogs (green and blue lines) and the parent compound (orange line). Invasion efficiency was measured by counting the number of newly formed rings. A 3 to 5-fold lower IC₅₀ (6 and 9.8 μM respectively) is seen for the two analogs compared to the parent compound (IC₅₀: 30 μM). orange, NCGC00015280; blue, NCGC0026250 and green, NCGC00262654. Error bars represent ± SEM from three independent experiments for each compound. (d) Merozoite invasion is not inhibited by a Src Kinase Inhibitor-1, but is blocked by AMA1-RON2 inhibitors. The concentrations of the compounds used are shown in the figure. Error bars represent ± SEM from at least two experiments for each compound. The number of parasites in the absence of inhibitor was considered 100%. (e) Dihydroartemisinin (DHA) in combination with invasion inhibitors is more efficient than by itself. Inhibitors NCGC00015280 (8 μM), NCGC00262650 (8 μM) and DHA (3 nM) alone or in combination were tested for growth inhibition. Parasite growth in the absence of any inhibitor was used as a control for no inhibition. Data represents the mean ± SEM of 3D7 and FVO parasites performed in duplicates. P values were calculated using one-way ANOVA and Bonferroni’s post test was performed to compare the effect of the combination treatment over the respective individual compounds. **P<0.01.

Figure 4. Mode of action of the inhibitor NCGC00262650 is mediated through binding of AMA1

The mode of inhibition of the small molecule was studied by a depletion assay using either his-tagged recombinant AMA1 or biotin-tagged RON2L peptide. The ability of AMA1 or RON2 to bind the inhibitor was assessed by performing invasion assays using inhibitor-depleted supernatants. 500 pmols of either recombinant AMA1 (both 3D7 and FVO allele) or RON2L peptide bound to magnetic beads was used to deplete 500 pmols of the inhibitor (final concentration 10μM). Error bars represent ± SEM from two experiments. (b) Immunofluorescence assay using FITC-labeled RON2L peptide. FITC-labeled RON2 peptide binds to AMA1 in the mature schizonts in the absence of inhibitors, while pre-incubation with inhibitor NCGC00015280 prevents binding of the peptide. Similar results were obtained with the analog NCGC00262650 and the inhibitor NCGC00181034 (data not shown). Scale bars represent 3 μm.

Figure 5. AMA1-RON2 inhibitor blocks junction formation

(a) Transmission electron microscopy of showing the different stages RBC invasion in the presence of 2 μM cytochalasin D, namely, attachment (1), re-orientation (2), junction formation (3) and rhoptry bulb secretion (4). R: rhoptry, M: micronemes, V: vacuoles; White arrow: junction. Scale bars represent 250 nm. (b) The percentage of merozoites that are attached to RBCs in the presence (white bars) or absence (black bars) of the AMA1-RON2 inhibitor NCGC00015280/NCGC00262650. (c) The percentage of apically oriented merozoites in the presence (white bars) or absence (black bars) of the inhibitor that form a junction and RBCs with vacuoles (indicative of rhoptry bulb secretion). Numbers within each bar represent the number of merozoite-attached RBCs in each category. Data was pooled from two independent experiments without inhibitor and one each with inhibitor NCGC00015280 and NCGC00262650. Scale bars represent 250 nm. (d) AMA1 secretion from micronemes is not affected. Merozoites released from schizonts in the absence (control) or presence of inhibitors NCGC00015280 (60 μM) and NCGC00181034 (60 μM) were analyzed using polyclonal antibodies to AMA1. Scale bars represent 1 μm.