Viability Screen of LOPAC¹²⁸⁰ Reveals Tyrosine Kinase Inhibitor Tyrphostin A9 as a Novel Partner Drug for Artesunate Combinations To Target the Plasmodium falciparum Ring Stage

The emergence of artemisinin-resistant Plasmodium falciparum poses a major threat to current frontline artemisinin combination therapies. Artemisinin resistance is widely associated with mutations in the P. falciparum Kelch13 (PfKelch13) propeller region, leading to delayed parasite clearance and increased survival of early-ring-stage parasites.

ABSTRACT

The emergence of artemisinin-resistant Plasmodium falciparum poses a major threat to current frontline artemisinin combination therapies. Artemisinin resistance is widely associated with mutations in the P. falciparum Kelch13 (PfKelch13) propeller region, leading to delayed parasite clearance and increased survival of early-ring-stage parasites. There is therefore a need to discover novel drugs that are effective against artemisinin-resistant P. falciparum. In view of this, our study aimed to identify compounds from the Library of Pharmacologically Active Compounds¹²⁸⁰ (LOPAC¹²⁸⁰) that could increase the efficacy of artesunate and be used as a potential partner drug for treatment against artemisinin-resistant falciparum malaria. By using a modified ring-stage survival assay, we performed a high-throughput screening of the activities of the 1,280 compounds from the LOPAC library in combination with artesunate against the P. falciparum IPC 5202 field isolate harboring the R539T mutation in the PfKelch13 propeller region. The potencies of the hits against both the IPC 5202 and CamWT_C580Y field isolates were determined through dose-dependent isobologram analyses; CamWT_C580Y has the more prevalent C580Y mutation characteristic of strains with artemisinin resistance. We identified tyrphostin A9 to have synergistic and additive activity against both parasite strains when dosed in combination with artesunate. These findings provide promising novel artesunate combinations that can target the P. falciparum artemisinin-resistant ring stage and insights that may aid in obtaining a better understanding of the mechanism involved in artemisinin resistance.

INTRODUCTION

Malaria is one of the world’s deadliest infectious diseases that has been around for many years. In 2016, it is estimated that there were roughly 216 million cases and 445,000 malaria deaths (1). Human malaria caused by Plasmodium falciparum accounts for 99% of the estimated cases and is mostly widespread in sub-Saharan Africa (1). The current first-line treatment for falciparum malaria is the artemisinin (ART)-based combination therapies (ACTs) (2). Artemisinins are a class of antimalarial compounds from the sweet wormwood plant (Artemisia annua). Artemisinin is a sesquiterpene lactone which contains an endoperoxide bridge that is essential for its antimalarial activity. Although the mechanism of action of artemisinin is not fully understood, it is generally accepted that the activation of artemisinin into toxic free radicals via its interaction with heme damages specific intracellular targets, possibly through alkylation (3, 4). In support of this, a recent study has also shown that artemisinin activation is primarily dependent on heme (5). As heme is present almost throughout the different stages of the parasite life cycle, including the early ring stage, from de novo heme synthesis, and also at later stages, from proteolysis of the host cell hemoglobin, artemisinin is capable of targeting multiple stages of the life cycle (5). Hence, artemisinin and its derivatives are regarded as one of the most effective classes of antimalarial drugs that are currently available, as they can target almost all asexual and sexual parasite stages of P. falciparum (6–9).

More importantly, artemisinins clear parasitemia more rapidly than any other antimalarial drugs, resulting in a rapid clinical and parasitological response to treatment (10). The immediate and rapid parasite clearance by artemisinins is thought to be caused by the killing of young intraerythrocytic ring-stage parasites, which further prevents the parasite from developing into the more pathological mature stages (10–12). Nevertheless, due to the short half-life of artemisinins (6), their use as monotherapy is discouraged. Hence, they are usually coupled with a longer-lasting partner drug, more commonly known as ACTs, to ensure improved antimalarial efficiency and to prevent or delay the onset of resistance (13, 14).

Despite this, the emergence and spread of artemisinin resistance in Plasmodium falciparum strains in Southeast Asia pose a great threat to the control and elimination of malaria (15–17). Artemisinin resistance is characterized by slow parasite clearance (18, 19) and is associated with the mutation in the propeller domain of the P. falciparum Kelch13 (PfKelch13) gene (20–22). Several studies have since demonstrated that PfKelch13 mutations lead to a prolonged or arrested ring stage with an increased stress response (20, 23, 24). The increased tolerance to artemisinin at the ring stage could be attributed to the lower drug activation and drug pressure at this stage compared to the later stages, where higher drug activation causes more extensive damage to the parasite, leading to parasite death (5). It is therefore crucial to discover artemisinin partner drugs that can sustain drug pressure for a longer duration to kill these dormant rings.

In view of this, we set out to discover new compounds from the Library of Pharmacologically Active Compounds¹²⁸⁰ (LOPAC¹²⁸⁰) that could potentially act in synergy with artesunate (AS) using a ring-stage survival assay (RSA) targeted against young rings at 0 to 3 h postinvasion (p.i.), the stage when artemisinin resistance has been widely reported. Of a total of 1,280 compounds, 3 compounds, consisting of tyrphostin A9, (2′Z,3′E)-6-bromoindirubin-3′-oxime (BIO), and cyclosporine (CsA), were found to act synergistically with artesunate against strain IPC 5202, while tyrphostin A9 exerted an additive effect against strain CamWT_C580Y. Of the three compounds, the one that was the most promising in combination with artesunate was tyrphostin A9, as this pairing presented the best synergistic profile. Our study also demonstrated that the combination of tyrphostin A9 and artesunate significantly lowered the 50% inhibitory concentration (IC₅₀) value of tyrphostin A9 against artemisinin-resistant strain IPC 5202 by half compared to that achieved when tyrphostin A9 was administered alone. Taken together, we identified novel compounds from LOPAC¹²⁸⁰ that could synergize with artesunate and potentially be used as an artesunate partner drug against artemisinin-resistant P. falciparum.

RESULTS

Artemisinin-resistant IPC 5202 displays elevated early-ring-stage survival.

Previous studies have demonstrated that ART-resistant parasites are able to remain dormant in the ring stage after exposure to ART and that the recovery of these ART-resistant parasites is always observed (25, 26). We therefore conducted a ring-stage survival assay at 0 to 3 h postinvasion (RSA^{0–3 h}) (27) to compare the percentage of parasite survival between ART-sensitive strains (3D7, K1, IPC 5188) and an ART-resistant strain (IPC 5202) after 6 h of pulse treatment with artesunate (AS). As expected, the RSA^{0–3 h} result showed that the ART-resistant strain (IPC 5202) had a significantly higher percentage of surviving parasites, which was about 29.5% (an ∼6-fold increase), than the ART-sensitive strains (3D7, K1, and IPC 5188), which showed less than 5% surviving parasites posttreatment (Fig. 1). This result therefore verifies that the ART-resistant strain (IPC 5202) indeed displays elevated early-ring-stage survival.

FIG 1.

Artemisinin-resistant IPC 5202 displays elevated early-ring-stage survival. Graphical representation of the percent survival of laboratory strains 3D7 and K1 and laboratory-adapted field isolates IPC 5188 and IPC 5202 after treatment with AS for 6 h after normalization to the value for the drug-free DMSO control. Error bars indicate SEM (n ≥ 3 experiments). ***, P < 0.001, evaluated by unpaired Student's t test. ART-S, ART sensitive; ART-R, ART resistant.

Identification of true AS-potentiating LOPAC¹²⁸⁰ compound hits with activity against P. falciparum IPC 5202.

Next, we used the IPC 5202 isolate to screen LOPAC¹²⁸⁰ to identify potential compounds that could potentiate the effect of AS against ART-resistant P. falciparum. We decided to group the LOPAC¹²⁸⁰ compounds according to the percentage of ring-stage survival (RSS) 6 h after treatment with each of the compounds in combination with AS (Fig. 2). Groups A (Fig. 3a), B (see Table S4 in the supplemental material), and C (Fig. 3b) represent compounds potentiating the effect of AS, compounds with no effect on the activity of AS, and compounds with antagonistic effects toward AS, respectively. To determine if the group A compounds truly potentiate the effects of AS on parasite survival, we decided to screen group A compounds alone without their combination with AS. The baseline effect of the group A compounds on parasite survival demonstrated that only 5 compounds from group A truly potentiated the effect of AS (Fig. 3c). Of the 19 hits, tyrphostin A9, BIO, CsA, U-74389G maleate, and WIN 62,577 led to more than 60% parasite survival when used alone without AS, which was more than 2-fold higher than the survival for the control group treated with AS only (Fig. 3c). When used in combination with AS, these 5 group A compounds therefore resulted in a decrease of more than 6-fold in the percentage of parasite survival compared to their effect on IPC 5202 when used alone. In contrast, 14 of the other compounds in group A, excluding the 5 hits mentioned above, caused less than 10% to 20% parasite survival posttreatment, suggesting that they do not potentiate the effect of AS, as there was no significant change in the percentage of parasite survival when used alone or in combination with AS. For strain IPC 5202, treatment with group C compounds alone further increased the percentage of surviving parasites posttreatment compared to that after combination treatment with AS, thus confirming their antagonistic effects on AS (Fig. 3d).

FIG 2.

Summary of compound groupings and compound hits based on HTS of the compounds in LOPAC¹²⁸⁰.

FIG 3.

Identification of activities of true AS-potentiating LOPAC¹²⁸⁰ compound hits against P. falciparum IPC 5202. (a) Graphical representation of the percent RSS of IPC 5202 posttreatment with group A compounds (ring-stage survival ≤ 10%) with AS for 6 h after normalization of the results to those for the drug-free DMSO-treated control. Error bars indicate SEM (n ≥ 3 experiments). (b) Graphical representation of the percent RSS of IPC 5202 posttreatment with group C compounds (ring stage survival ≥ 40%) with AS for 6 h after normalization of the results to those for the drug-free DMSO-treated control. Error bars indicate SEM (n ≥ 3 experiments). (c) Graphical representation of the percent RSS of IPC 5202 posttreatment with group A AS-potentiating compounds (ring stage survival ≤ 10%) without AS for 6 h after normalization of the results to those for the drug-free DMSO control. CsA, BIO, WIN 62,577, U-74389G maleate, and tyrphostin A9 led to a more than 2-fold increase in parasite survival compared to that for the AS-only-treated control. Error bars indicate SEM (n ≥ 3 experiments). (d) Graphical representation of the percent RSS of IPC 5202 posttreatment with group C antagonistic compounds (ring stage survival ≥ 40%) without AS for 6 h after normalization of the results to those for the drug-free DMSO-treated control. All of the group C compounds led to parasite survival 60% greater than that achieved with the AS-only control. ODQ and 5HP-33 refer to 1H-[1,2,4]oxadiazolo[4,3-a]quinoxalin-1-one and thalidomide derivative,respectively. Error bars indicate SEM (n ≥ 3 experiments). Compound names with ellipses refer to hexahydro-sila-difenidol hydrochloride, p-fluoro analog, (R,R)-cis-dethyl tetrahydro-2,8-chrysenediol, quinacrine dihydrochloride, dequalinium chloride hydrate, emetine dihydrochloride hydrate, amsacrine hydrochloride, naltriben methanesulfonate, nicardipine hydrochloride, and GR 55562 dihydrobromide.

Interaction of AS-potentiating LOPAC¹²⁸⁰ compound hits.

Having validated the 5 compound hits to be potentiating only in combination with AS through our modified RSA, we carried out a 48-h in vitro reinvasion assay against IPC 5202 (Table S1) and CamWT_C580Y (Table S2) parasites for IC₅₀ determination. To expand upon our findings regarding the types of interactions between the 5 compound hits and AS, we performed a fixed-ratio isobologram assay, adapted from that of Fivelman et al. (28), with a total of 6 different combinations consisting of fixed ratios ranging from 5:0 to 0:5 (Table S3) and calculated the combination index (CI) (29). Based on the results obtained with IPC 5202 (Table 1), the interactions between AS and the 3 compound hits exhibited various extents of synergism. On the other hand, only tyrphostin A9 displayed an additive effect against CamWT_C580Y parasites when it was dosed in combination with AS (Table 2). These data therefore show that while tyrphostin A9, CsA, and BIO could be potential drug partners, tyrphostin A9 emerged as the overall most potent drug from the LOPAC¹²⁸⁰ library for use in combination with AS.

TABLE 1.

Interaction of AS with AS-potentiating LOPAC¹²⁸⁰ compound hits BIO, tyrphostin A9, and CsA against IPC 5202, based on sum FIC₅₀ and overall drug interaction

Drug combination	Sum FIC50 at the following AS/drug ratio:				Overall drug interaction
	4:1	3:2	2:3	1:4	Sum FIC50	Effect
AS + BIO	0.801 ± 0.011	0.840 ± 0.029	0.744 ± 0.008	0.787 ± 0.019	0.793 ± 0.017	Moderate synergism
AS + CsA	0.727 ± 0.030	0.754 ± 0.007	0.707 ± 0.016	0.527 ± 0.014	0.679 ± 0.017	Synergism
AS + U-74389G maleate	0.727 ± 0.030	0.880 ± 0.011	1.075 ± 0.032	1.262 ± 0.025	0.986 ± 0.025	Additive
AS + tyrphostin A9	0.308 ± 0.032	0.188 ± 0.005	0.241 ± 0.011	0.170 ± 0.006	0.227 ± 0.014	Strong synergism

TABLE 2.

Interaction of AS with AS-potentiating LOPAC¹²⁸⁰ compound hits against CamWT_C580Y, based on sum FIC₅₀ and overall drug interaction

Drug combination	Sum FIC50 at the following AS/drug ratio:				Overall drug interaction
	4:1	3:2	2:3	1:4	Sum FIC50	Effect
AS + BIO	1.627 ± 0.051	1.785 ± 0.053	1.578 ± 0.059	1.319 ± 0.096	1.577 ± 0.065	Antagonism
AS + CsA	1.299 ± 0.062	1.500 ± 0.046	1.592 ± 0.020	1.140 ± 0.033	1.383 ± 0.040	Moderate antagonism
AS + U-74389G maleate	1.162 ± 0.004	1.152 ± 0.041	1.300 ± 0.054	1.390 ± 0.058	1.250 ± 0.049	Moderate antagonism
AS + tyrphostin A9	1.221 ± 0.129	0.990 ± 0.152	1.083 ± 0.086	0.702 ± 0.085	0.999 ± 0.113	Additive

DISCUSSION

The increasing prevalence of P. falciparum strains that are resistant to current frontline ART combination therapies highlights the urgent need to discover new antimalarial drugs, in particular, drugs for use in combination therapy. Combination therapy is a promising approach to control the spread of resistance and delay the selection of multidrug resistance (14). Nevertheless, drug combinations can also lead to a plethora of outcomes due to the different effects of drug interactions. In view of this, drug synergy is preferred for its maximum therapeutic effect and specificity (30, 31). In this study, we carried out a single-dose screen of LOPAC¹²⁸⁰ compounds in combination with AS against field isolate IPC 5202 to identify possible drug combinations with synergistic activity against an ART-resistant P. falciparum strain. AS is a semisynthetic derivative of artemisinin with improved bioavailability that is used for the treatment of mild to severe malaria parasite infection (32). The use of ART and its derivatives in drug combinations has several advantages over the use of other antimalarial drugs, as they are more effective than any other antimalarial in lowering parasite numbers, reducing gametocyte carriage, and decreasing the rate of evolution of drug resistance, as they are very rapidly eliminated from the body (14). Among the ART derivatives, such as artemether and dihydroartemisinin, AS is the most therapeutically versatile drug, as it can be administered via the oral, rectal, intravenous, and intramuscular routes (32). Besides, intravenous AS is also the preferred therapy for severe malaria and is the only ART derivative that can be administered intravenously (33). Considering the therapeutic relevance of AS and the emergence of ART resistance, the discovery of readily available new partner drugs for AS that synergize its antimalarial effect will be a useful approach in an effort to provide more combination options to control malaria effectively.

Previously, studies have shown that mutations in the propeller domain of the PfKelch13 gene are central to ART resistance (20). ART resistance has also been associated with delayed parasite clearance and the increased survival of early-ring-stage parasites (34). We demonstrated that the selected IPC 5202 field strain harboring the R539T mutation at the PfKelch13 propeller region indeed displayed elevated early-ring-stage survival in our RSA^{0–3 h} assay. RSA is commonly used instead of the IC₅₀ assay to study ART resistance since drug-induced dormancy is particularly specific at the early ring stage (26, 27). Exposing the parasites to all different stages of its erythrocytic cycle in the IC₅₀ assay will therefore give an inaccurate result for parasite survivability, since trophozoites and schizonts are known to be highly susceptible to ART (23, 35). While the original RSA was carried out on different parasitic stages (0 to 3 h p.i. for young rings, 9 to 12 h p.i. for rings, and 18 to 21 h p.i. for trophozoites), only parasites in the young ring stage (0 to 3 h p.i.) were used in this study since it had already been shown previously that the median survival rate of slow-clearing parasites was 47 times greater than that of fast-clearing parasites at the 0- to 3-h-p.i. stage (27). In contrast, there were no significant differences in the survival rates between fast-clearing and slow-clearing parasites of the 9- to 12-h-p.i. rings and 18- to 21-h-p.i. trophozoites (27). Therefore, it was more relevant to use tightly synchronized 0- to 3-h-p.i. rings as opposed to the later stages for our study. We also decided to use a 500 nM AS drug pulse for 6 h since it has previously been shown to be a clinically achievable dosage (5, 36), although the standard RSA protocol established by the WorldWide Antimalarial Resistance Network (WWARN) administered a 700 nM DHA drug pulse for 6 h (27, 37). In addition, the 6-h AS drug pulse was used in the RSA to mimic in vivo clinical drug exposure and the pharmacokinetics of DHA or artesunate (27, 32). We also used 5 μM as the starting concentration for the screening of LOPAC¹²⁸⁰ compounds since it fell within the range of the 1 μM to 10 μM deemed suitable for first-pass high-throughput screening (HTS) (38, 39). When antimalarial quinolines, such as quinidine and chloroquine (CQ), were administered together with AS in a 6-h exposure, these compounds demonstrated an overall lack of a parasiticidal effect (Fig. 3d). These corroborative results highlight the robustness of the modified RSA, as quinoline compounds are generally thought to interfere with the heme detoxification pathway of the mature parasite trophozoite stage (40). Among the compounds screened, only CsA, BIO, tyrphostin A9, WIN 62,577, and U-74389G maleate had true potentiating effects on the efficacy of AS against ART-resistant strain IPC 5202 parasites. When administered alone, these five hits did not result in appreciable parasite cell death. Unfortunately, due to a lack of drug availability, the interaction between WIN 62,577 and AS could not be further investigated in the present study.

Tyrphostin A9 demonstrated the greatest efficacy because it had the strongest synergistic effect with AS compared to that of the other compound hits. Tyrphostin A9 is an inhibitor of the platelet-derived growth factor receptor pathway and has roles in the inhibition of receptor tyrosine kinase (RTK), though its precise inhibitory mechanism remains unknown (41). Since this is also the first study that has demonstrated the potential role of tyrphostin A9 as a possible antimalarial agent for ART combination therapy, a more thorough study is needed to understand its mechanism of action and its pharmacokinetics. It seems possible that the synergistic and additive effect of tyrphostin A9 and AS, respectively, on the IPC 5202 and CamWT_C580Y strains may partly be attributed to the inhibition action of tyrphostin A9 on RTKs, leading to the further reduction of P. falciparum phosphatidylinositol 3-kinase (PfPI3K) and phosphatidylinositol 3-phosphate (PI3P). The PI3P level is known to be elevated basally in ART-resistant parasites carrying the R539T or C580Y mutation (42, 43). Therefore, lowering of the PI3P level in resistant strain may prove to be an important aspect in the synergism of the drug partner with AS.

Knowing that BIO is a competitive ATP inhibitor (44), it is possible that it competes with ATP for the binding of ATP to the parasite protein kinase PfPI3K, hence lowering parasite PI3P levels and compromising overall parasite proliferation. Regarding drug targets, it is also worth noting that protein kinases are attractive antimalarial targets (45) due to their crucial roles in the life cycle of the malaria parasite (46). In this case, imatinib, a well-known tyrosine kinase inhibitor used in the treatment of cancer, has also previously been shown to have antimalarial activity and synergism with AS, albeit against chloroquine-resistant P. falciparum strains (47).

Another important mechanism that could be targeted to overcome ART resistance in P. falciparum is the parasite’s cell stress response. Studies have shown that an increased unfolded protein response (UPR) in Kelch13 mutant parasites is the foundation for ART resistance in P. falciparum (24, 48). It is thought that ART-resistant parasites are less susceptible to the oxidative damage exerted by ART, as mutated PfKelch13 is unable to bind to Nrf2, a transcription factor for cytoprotective enzymes, which therefore leads to constant activation of the cytoprotective enzyme and increased parasite survival (22, 49). Based on the mechanism described above, it is possible that synergism between CsA and AS could be due to the function of CsA in disrupting the UPR of the parasite, as CsA is known to inhibit the chymotrypsin-like activity of the proteasome (50, 51). Hence, the induction of oxidative stress by AS (5), coupled with the disruption of UPR by CsA, may lead to the persistence of misfolded proteins and, subsequently, parasite death.

Overall, our findings demonstrate that the combinations of tyrphostin A9, BIO, and CsA with AS are more effective against the IPC 5202 field isolate than against CamWT_C580Y. The poorer efficacy of the artesunate combinations against the CamWT_C580Y (C580Y) field strain than against IPC 5202 (R539T) could be explained by the lack of a fitness cost in strains with the more prevalent mutation Kelch13 C580Y, which has been shown to be the fittest Kelch13 allele (52). This is consistent with the findings of previous studies which demonstrated the possible role of different Kelch13 single nucleotide polymorphisms (SNPs) and their effects on in vitro resistance and fitness (52, 53). Parasites with the R539T mutation were shown to exhibit the greatest fitness cost, and the mutation affected parasite growth in vitro, thus possibly rendering IPC 5202 parasites inherently more vulnerable to the parasiticidal effects and CamWT_C580Y fitter and less susceptible to the drug combinations used in the present study.

In conclusion, we showed that the modified RSA protocol can be used for high-throughput screening to identify compounds that have potential synergistic effects against ART-resistant strains. Out of 5 compound hits, 3 compounds, including tyrphostin A9, BIO, and CsA, showed synergistic effects with AS against strain IPC 5202. Among these 3 compound hits, only the tyrphostin A9 and AS combination had an additive effect against CamWT_C580Y, while the other combinations showed antagonistic effects. The phenotypes seen from the drug interaction against different strains suggest an important association between Kelch13 SNPs, resistance, fitness, and drug potency. The discovery of the compound hits is therefore useful since they could be used as potential drug partners for combination therapies and also provide new insights into the mechanism of ART resistance.

MATERIALS AND METHODS

Selection of P. falciparum strains.

P. falciparum field strains which originated from the Thailand-Cambodia border were ordered from the Malaria Research and Reference Reagent Resource Center (MR4). Based on the SNPs of each strain tested, P. falciparum ART-sensitive strain IPC 5188 (MRA-1239; MR4; ATCC, Manassas, VA, USA) and ART-resistant strain IPC 5202 (MRA-1240; MR4; ATCC) were selected for use in the experiments. DNA extraction (QIAamp DNA blood minikit; Qiagen) was performed first on P. falciparum field isolates (MRA-1239 and MRA-1240) to select relevant strains, while laboratory CQ-sensitive strain 3D7 (MRA-102; MR4; ATCC, Manassas, VA, USA) and CQ-resistant strain K1 (MRA-159; MR4; ATCC) were selected to serve as reference strains.

PCR and sequencing of the PfKelch13 gene.

Three sets of P. falciparum Kelch13-specific forward and reverse primers were designed to sequence the PfKelch13 gene spanning the predicted open reading frame: (i) set A, forward primer 5′-GGG AAT CTG GTG GTA ACA GC-3′ and reverse primer 5′-CGC CAG CAT TGT TGA CTA AT-3′; (ii) set B, forward primer 5′-GCC TTG TTG AAA GAA GCA GA-3′ and reverse primer 5′-GGA GTG ACC AAA TCT GGG A-3′; and (iii) set C, forward primer 5′-GGT GGT AAT AAC TAT GAT TAT AAG GC-3′ and reverse primer 5′-AGC TGC TCC TGA ACT TCT AGC-3′. DNA extracted from each strain was subjected to PCR (Phusion High Fidelity DNA polymerase kit; New England Biolabs Incorporated) to obtain a sufficient amount of the PfKelch13 segment of each strain’s DNA for sequencing. The final concentrations of reagents for each reaction were as follows: (i) 1× high-fidelity buffer, (ii) 0.2 mM deoxyribonucleotide triphosphates (dNTPs), (iii) 0.5 μM PfKelch13-specific forward and reverse primers, (iv) 30 ng of DNA template, and (v) 0.02 U/μl of Phusion polymerase. The denaturation temperature was set at 98°C for 10 s, the annealing temperature was set at 62.5°C for 30 s, and the extension temperature was set at 72°C for 30 s for a total of 25 cycles. After PCR, the amplified PfKelch13 DNA segments for each parasite strain were then sent to First Base Laboratories for Sanger sequencing. Nucleotide BLAST analysis was conducted to identify SNPs of the PfKelch13 gene in each strain, and the sequenced DNA of each strain was aligned to the PfKelch13 gene sequence of P. falciparum 3D7 obtained from PlasmoDB (gene identifier PF3D7_1343700; The Plasmodium Genomics Resource).

In vitro culturing of P. falciparum parasites.

Laboratory CQ-sensitive strain 3D7 (MRA-102; MR4; ATCC, Manassas, VA, USA) and CQ-resistant strain K1 (MRA-159; MR4; ATCC) and field isolates consisting of ART-sensitive strain IPC 5188 (MRA-1239; MR4; ATCC) and ART-resistant strain IPC 5202 and CamWT_C580Y (MRA-1240 and MRA-1251, respectively; MR4; ATCC) were cultured and maintained in 25-cm² or 75-cm² nonvented cell culture flasks at 1.5% hematocrit with complete malaria culture medium (MCM), comprising a filter-sterilized homogeneous mixture of RPMI 1640 (Life Technologies, USA), 2.2 g/liter sodium bicarbonate, 0.5% (wt/vol) AlbuMAX II (Gibco, Thermo Fisher Scientific, MA, USA), 0.005% (wt/vol) hypoxanthine (Sigma-Aldrich, MO, USA), 0.03% (wt/vol) l-glutamine (Sigma-Aldrich, St. Louis, MO, USA), and 25 μg/liter gentamicin (Gibco, Thermo Fisher Scientific, MA, USA). The pH was adjusted to 7.4 using 1 N NaOH. Type O-positive human erythrocytes in citrate-phosphate-dextrose with adenine (CPDA-1) anticoagulant were obtained from the Interstate Blood Bank (Memphis, TN, USA) and tested negative for infectious agents according to FDA guidelines. The flasks were incubated at 37°C without light and gassed with 5% CO₂, 3% O₂, and 92% N₂. Routine PCR-based mycoplasma testing was conducted on a monthly basis to ensure the lack of contamination in cultures.

Tight synchronization of P. falciparum parasites.

Parasites were synchronized with a modified 5% (wt/vol) sorbitol (Sigma-Aldrich) lysis method (54) at 37°C for 10 min for the selection of ring-stage parasites, following which the cells were washed twice and resuspended to 1.5% hematocrit in MCM. Cultures were checked on every alternate day when they were at the ring stage. Once most parasites were at the schizont stage, the culture was passed through a 65% Percoll gradient (55). The schizont fraction was aspirated and placed in a fresh nonvented 25-cm² airtight flask with naive erythrocytes to a 1.5% hematocrit. The cultures were then placed in the 37°C incubator for 3 h to allow reinvasion. After 3 h, the cultures were synchronized with 5% sorbitol again to kill off any remaining schizonts. The remaining rings (0 to 3 h p.i.) were then maintained at 1.5% hematocrit with MCM or frozen.

Drug preparations and LOPAC¹²⁸⁰ storage.

AS (catalog number A3731-100MG; Sigma-Aldrich) was dissolved in sterile dimethyl sulfoxide (DMSO) to stock aliquots of 1 mM, frozen at −80°C, and protected from light. For the LOPAC¹²⁸⁰ compounds (Sigma-Aldrich), each compound was reconstituted to stock aliquots of 10 mM in DMSO. A 20× master drug plate was prepared by transferring 2 μl of each compound in the library into the corresponding wells of a sterile 96-well U-bottom plate (Greiner). For the AS-only control wells and drug-free DMSO control wells in the 20× master drug plate, 2 μl of 100% DMSO was added. The master drug plate was then kept at −20°C before the experiment. On the day of the experiment, the master drug plate was thawed and diluted with chilled MCM to a final concentration of 100 μM.

RSA, parasite drug treatment, and staining.

Parasites were synchronized 2 cycles prior to the start of the assay. On the assay day, the parasites were quantified, checked for 0- to 3-h-p.i. young rings, resuspended to a 1% parasitemia and 1.25% hematocrit, and dispensed into a 96-well flat-bottom plate. The parasites were treated with 500 nM AS in a final volume of 200 μl in each AS-containing well, while the drug-free control wells contained 0.1% DMSO. For the ring-stage survival assay (RSA), IPC 5202 parasites were treated with a final concentration of 5 μM the LOPAC¹²⁸⁰ compounds independently or in combination with 500 nM AS. The assay plate was then incubated at 37°C with a mixture of gases in a humidified hypoxic chamber for 6 h. After 6 h, the cells in all wells except the 72-h control wells were washed thrice with prewarmed MCM, resuspended with 200 μl MCM, and reincubated in the hypoxic chamber for another 66 h (Fig. 4). After incubation of the assay plates for 66 h, the cells were transferred to a 96-well V-bottom plate to concentrate the pellet and stained with 2 μg/ml Hoechst 33342 dye (Sigma-Aldrich, St. Louis, MO, USA) for 30 min. The cells were washed with phosphate-buffered saline (PBS) and resuspended in 200 μl of PBS for data acquisition.

FIG 4.

Overview of the modified RSA for HTS of LOPAC¹²⁸⁰ compounds to identify novel artesunate combinations that target the Plasmodium falciparum ring stage.

RSS screen with a flow cytometer.

Samples were acquired using a CyAn ADP analyzer (Dako/Beckman Coulter, Brea, CA, USA). Fluorescence detection of the Hoechst 33342 dye was carried out using an excitation wavelength of 405 nm and an emission wavelength of 450 ± 25nm. At least 20,000 events were collected for each sample, and the events were displayed on side scatter (log)-versus-Hoechst dye fluorescence (log) dot plots. The results were analyzed using FlowJo software, version 9.3.2, by gating the parasite-positive population in the side scatter (log)-versus-Hoechst dye fluorescence (log) plots. The final quantification of ring-stage survival (RSS) for each sample was calculated by the following equation: RSS = (raw parasitemia − average for AS-treated samples at 72 h)/average for DMSO controls, where raw parasitemia is the level of parasitemia for infected erythrocytes which contained DNA that was stained with the Hoechst dye, average for AS-treated samples at 72 h is the average of the raw levels of parasitemia from the 2 AS-containing wells at 72 h, and average for DMSO controls is the average of the raw levels of parasitemia from the 2 drug-free DMSO-treated control wells.

To ensure the robustness of each separate experiment, the results for any particular assay were accepted only if the sample wells with AS gave approximately 20% to 30% survival at 6 h. Based on the RSS for IPC 5202, samples treated with AS in combination with a LOPAC¹²⁸⁰ compound were subsequently grouped into three groups: (i) compounds with a potentiating effect (RSS ≤ 10%) (group A), (ii) compounds with no effect (10% < RSS < 40%) (group B), and (iii) compounds with an antagonistic effect (RSS ≥ 40%) (group C). Results were obtained from at least 3 separate experiments.

A 48-h reinvasion assay for IC₅₀ determination.

The 0- to 3-h-p.i. IPC 5202 and CamWT_C580Y young rings at 1.25% hematocrit and 1% parasitemia were incubated for 48 to 52 h with serially diluted AS and the compounds of interest. For parasite labeling, parasites were stained with 2 μg/ml Hoechst 33342 dye for flow cytometry, as elaborated above. IC₅₀s were then determined by the use of GraphPad Prism (version 7) software using a variable-slope logistic curve. Values were obtained from at least 3 separate triplicate experiments (see Tables S1 and S2 in the supplemental material).

Fixed-ratio isobologram assay.

The protocol for the fixed-ratio isobologram method was adapted from Fivelman et al. (28). For the IC₅₀ of individual compounds to fall at about the fourth 2-fold serial dilution for the combination assay, final fixed ratios were prepared at 8 times the IC₅₀ of each individual IC₅₀. The dilutions of AS and the compound of interest were prepared in fixed ratios, as illustrated in Table S3. For each combination ratio, 200 μl of diluted AS or compound (8 times the IC₅₀) at a 7-point serial dilution was used, in addition to a no-drug 0.5% DMSO vehicle control in designated wells, and tests with each fixed ratio were conducted in triplicate in every independent experiment. The 0- to 3-h-p.i. IPC 5202 and CamWT_C580Y rings at 1.25% hematocrit and 1% parasitemia were used for this assay. The assay plate was incubated at 37°C with a mixture of gases in a humidified hypoxic chamber for 48 to 52 h, before the parasites were stained with Hoechst 33342 dye for flow cytometry, as elaborated above. Raw parasitemia values were used for analysis of the activity of AS and the compound of interest in combination. Figure S1 shows the representative parasitemia levels from each independent isobologram experiment. The average raw parasitemia values of each fixed ratio were obtained and normalized, and six IC₅₀ curves (with four curves being derived from drug-drug interactions and two curves being derived from single drugs only) were determined by the use of GraphPad Prism (version 7) software. Each drug’s 50% fractional inhibitory concentration (FIC₅₀) at each ratio of its combination with the partner drug was calculated by the following equation (56): IC₅₀ of AS or the drug of interest in the combination/IC₅₀ of AS or the drug of interest alone.

The sum of the FIC₅₀ (sum FIC₅₀) for each ratio was then calculated using the following formula: (IC₅₀ of AS in the combination/IC₅₀ of AS alone) + (IC₅₀ of the drug of interest in the combination/IC₅₀ of the drug of interest alone).

To assess the overall drug interaction involving the different combination ratios, the following formula was used: sum FIC₅₀ of all ratios (excluding the drugs alone)/number of ratios.

Such drug-drug assessments were similarly conducted by Gorka et al., who utilized fixed-ratio combinations and normalized sigmoidal curve fits to study drug interactions in P. falciparum (56). The types of drug interaction were interpreted according to the range of combination indexes adapted from Chou (29).

Statistical tests and analyses.

Graphs were prepared and statistical tests were performed with GraphPad Prism (version 7) software, which was also used to obtain P values by two-sample Student's t tests. Error bars represent the standard errors of the means (SEM) from at least 3 separate independent experiments.