Naturally Acquired Kelch13 Mutations in Plasmodium falciparum Strains Modulate In Vitro Ring-Stage Artemisinin-Based Drug Tolerance and Parasite Survival in Response to Hyperoxia

Abstract

The ring-stage survival assay was utilized to assess the impact of physiological hyperoxic stress on dihydroartemisinin (DHA) tolerance for a panel of Plasmodium falciparum strains with and without Kelch13 mutations. Strains without naturally acquired Kelch13 mutations or the postulated genetic background associated with delayed parasite clearance time demonstrated reduced proliferation under hyperoxic conditions in the subsequent proliferation cycle. Dihydroartemisinin tolerance in three isolates with naturally acquired Kelch13 mutations but not two genetically manipulated laboratory strains was modulated by in vitro hyperoxic stress exposure of early-ring-stage parasites in the cycle before drug exposure. Reduced parasite tolerance to additional derivatives, including artemisinin, artesunate, and OZ277, was observed within the second proliferation cycle. OZ439 and epoxomicin completely prevented parasite survival under both hyperoxia and normoxic in vitro culture conditions, highlighting the unique relationship between DHA tolerance and Kelch13 mutation-associated genetic background. IMPORTANCE Artemisinin-based combination therapy (ACT) for treating malaria is under intense scrutiny following treatment failures in the Greater Mekong subregion of Asia. This is further compounded by the potential for extensive loss of life if treatment failures extend to the African continent. Although Plasmodium falciparum has become resistant to all antimalarial drugs, artemisinin "resistance" does not present in the same way as resistance to other antimalarial drugs. Instead, a partial resistance or tolerance is demonstrated, associated with the parasite's genetic profile and linked to a molecular marker referred to as K13. It is suggested that parasites may have adapted to drug treatment, as well as the presence of underlying population health issues such as hemoglobinopathies, and/or environmental pressures, resulting in parasite tolerance to ACT. Understanding parasite evolution and control of artemisinin tolerance will provide innovative approaches to mitigate the development of artemisinin tolerance and thereby artemisinin-based drug treatment failure and loss of life globally to malaria infections.

Keywords: K13 mutation; Plasmodium falciparum; artemisinin tolerance; genetic background; hyperoxia.

FIG 1

Comparison of first versus second cycle in vitro parasite proliferation reduction in response to hyperoxia, expressed as percent reduction in parasitemia in comparison to normoxia controls. (A) Flow chart of hyperoxia exposure of cultures. Ring-stage parasites were obtained from the constant normoxia control (CN) at schizont rupture (dashed blue line). Half the culture volume was exposed to hyperoxia (H; green squares), and the remainder was maintained in normoxia (CN) for completion of the parasite development cycle. After the next schizont rupture, ring-stage parasites were isolated and once again divided in half, with half being placed under hyperoxic conditions and half under normoxic conditions, resulting in a total of 4 cultures. CN and H1 represent a first hyperoxia cycle comparison culture set, and H2 and RN (return to normoxia) the second hyperoxia exposure cycle comparison set. (B) First-cycle reduction in parasitemia for group A (black and gray) versus group B (dark and light blue columns) Pf strains. Blue brackets define groups A and B and the K13 status of the 12 strains. (C) Second-cycle reduction in parasitemia for group A and group B Pf strains. (D) Comparison of group B first-cycle reduction in parasitemia between clinical isolates versus laboratory strains. (E) Statistical significance between group A and group B reduction in parasitemia within the first or second cycle of proliferation. Statistical significance between group A and group B (panels D and E) was determined using the two-tailed nonparametric Mann-Whitney U test with 95% confidence intervals. **, P = 0.005; ***, P = 0.0005.

FIG 2

Effect of 1-cycle versus 2-cycle in vitro hyperoxia exposure on DHA tolerance of the K13 mutant parasite strain Cam5 I543T. (A) Experimental flow diagram. At the schizont stage (dashed blue lines), 0- to 3-h ring-stage CN parasites were isolated, and half were exposed to hyperoxia (H; green squares) while the other half remained in normoxia (CN; red squares) for the remainder of the parasite development cycle. At the following schizont stage, 0- to 3-h ring-stage parasites were isolated, and each culture was divided into two volumes for testing in the RSA, in the presence (H and H2) or absence (CN and RN) of hyperoxia. The percent reduction in survival after DHA exposure, in comparison to the continuous normoxic control (CN), was calculated from the %S values determined for each culture in the RSA. (B) Effect of hyperoxia on DHA tolerance. The first-cycle hyperoxia RSA data (H1) represents an acute exposure to both hyperoxia and DHA simultaneously with no prior hyperoxia exposure. The second-cycle data represent the effect of hyperoxia during the cycle prior to performing the RSA in normoxia (RN) or in conjunction with hyperoxia (H2). Statistical significance was calculated (Student’s t test, two tailed, 95% confidence intervals) was calculated for the comparison between test conditions and constant normoxia (CN). *, P < 0.05; **, P < 0.005; ***, P < 0.0005. Values are means (n = 3) and standard deviations (SD).

FIG 3

Influence of hyperoxia exposure time and parasite stage of development on subsequent DHA tolerance. (A) Experimental flow diagram. CN 0- to 3-h ring-stage parasites were divided into eight 150-mm petri dishes. At time zero (after 0- to 3-h ring-stage parasite isolation), six dishes were exposed to hyperoxia (green) and two remained in normoxia (CN and red 24 to 48 h trophozoite evaluation). After 5 h of hyperoxic exposure, a single dish was returned to normoxia, as was one dish at 5-h intervals thereafter (shown by the changing line color from red [normoxia] to green [hyperoxia]), resulting in 5-, 10-, 15-, 20-, and 30-h hyperoxia-exposed cultures in the first parasite development cycle. A single petri dish of culture was maintained under normoxic conditions until 24 h after RBC invasion and then exposed to hyperoxia for the remainder of the first cycle of parasite development. The CN control remained under normoxic conditions for the complete first development cycle to mature schizonts, and a single dish remained in hyperoxia for the complete cycle of development (H). At the schizont stage of the first cycle (dashed blue line), 0- to 3-h parasites were isolated from all 8 individual cultures. Each of the 8 cultures in the second parasite developmental cycle was evaluated in the RSA under both hyperoxic and normoxic conditions, apart from the CN control, which was tested only under normoxic conditions. The percent survival in the RSA was then used to determine the reduction in parasite survival due to the hyperoxia exposure time/condition as a percentage of the constant normoxia control: 100 − (%S test culture condition/%S CN × 100) = % reduction in survival after DHA exposure. DMSO, dimethyl sulfoxide. (B) Percentage reduction in survival after DHA exposure in comparison to the continuous normoxic culture (CN) plotted against time after RBC invasion. n = 3 for control sets (3 under constant hyperoxia and 3 under constant normoxia), and n = 3 individual biological replicates for timed cultures for each time point. Statistical significance for the reduction in parasite survival for each hyperoxia exposure duration in the first cycle of development, in either normoxia or hyperoxia in the second cycle of parasite development, was calculated in comparison to the survival of the CN control (Student’s t test, two tailed, 95% confidence intervals). *, P < 0.05; **, P < 0.005; ***, P < 0.0005. Values are means (n = 3) and SD. H2 and RN indicate the hyperoxia controls maintained under hyperoxia for a full cycle within the first cycle of parasite development, prior to performing the RSA in normoxia (RN) or hyperoxia (H2), i.e., equivalent to second-cycle data in Fig. 2B. Green and red “24-48hr” columns reflect the effect of hyperoxia on trophozoite stages and the subsequent influence on DHA tolerance.

FIG 4

Effect of hyperoxia exposure of 0- to 20-h ring-stage parasites on parasite proliferation. (A) Experimental flow diagram. Early-ring-stage 0- to 3-h normoxia control (CN) parasite cultures were isolated and exposed to hyperoxia for the first 20 h of parasite development. The CN in this evaluation was not performed. Early-ring-stage parasites (0 to 3 h) of the 20-h-hyperoxia-exposed parasite cultures were isolated, and each culture was divided in half. One half was incubated in hyperoxia (return to hyperoxia [RH]) and one half in normoxia (20 h) for a further 65 h. (B) Influence of 0- to 20-h hyperoxic exposure on parasite proliferation for three group A strains (Cam3.1 R539T, Cam3.1WT^rev, and Cam5 I543T) and three group B strains (Cam WT^C580Y, IPC3663, and Dd2^R539T), in conjunction with hyperoxia or normoxia during the second cycle of proliferation, expressed as percent reduction in parasitemia. Statistical significance (Mann-Whitney U P < 0.001) between groups A and B is shown in the inset (Mann-Whitney U test, two tailed, 95% confidence intervals). ***, P = 0.0005.

FIG 5

Effect of 20 h hyperoxia on DHA tolerance of K13 mutant containing parasite strains with and without associated delayed-PCT genetic background. (A) Experimental flow diagram. Control normoxia (CN) early ring-stage parasites (0 to 3 h) were isolated and exposed to hyperoxia (H) for the first 20 h of ring-stage parasite development before returning to normoxic conditions for the remainder of the first development cycle (20 h; red cross-hatched squares). The normoxic control (CN) was processed only in normoxia throughout the evaluation. At the schizont stage for the first development cycle (dashed blue line), 0- to 3-h early ring-stage parasites were isolated for both the 20-h hyperoxic (20 h)-exposed and constant normoxia control (CN) cultures. The ring-stage parasites were then tested for their DHA tolerance (RSA) in the second development cycle, all under normoxic conditions (CN and 20 h). (B) Percent survival after DHA exposure for all strains and conditions evaluated. The dashed line represents the average survival for all five strains after 20 h hyperoxia exposure in the first cycle of proliferation (average, 14.02 ± 5.1). The line, therefore, indicates a residual level of parasite survival proposed to be directly influenced by K13 mutations and not impacted by hyperoxia. P values were calculated using unpaired Student’s t test, one tailed (one directional), with 95% confidence limits using GraphPad Prism. **, P < 0.005; ns, not significant. Values are means (n = 3) and SD. (Inset) Comparison between survival control data for group A and group B (solid red) and 20-h-hyperoxic-exposure data (red hatched). Statistical significance between group A and B was calculated using the Mann-Whitney U two-tailed test, with 95% confidence intervals. ***, P = 0.0005.

FIG 6

Effect of hyperoxic exposure on parasite tolerance to ART, ATS, OZ277, OZ439, and the proteasome inhibitor epoxomicin. (A) Experimental flow diagram. Continuous normoxia control (CN) 0- to 3-h ring-stage parasites were isolated and maintained in normoxia or exposed to hyperoxia for the first 20 h of ring stage development before returning to normoxia. In the second cycle of development, 0- to 3-h ring-stage parasites were used to evaluate the activity of DHA and other drugs in the RSA for both continuous normoxia (CN) and 20-h-hyperoxia-exposed ring-stage parasites. (B) Comparison of percent survival of Cam5 I543T to endoperoxides for continuous normoxia or 20 h hyperoxia. P values were calculated using one-tailed nonpaired Student’s t test (one directional) with 95% confidence limits using GraphPad Prism for reduction in %S in response to hyperoxia for each compound set. *, P < 0.05; **, P < 0.005; ***, P < 0.0005; ns, not significant. Values are means (n = 3) and SD.