Cholesterol-dependent enrichment of understudied erythrocytic stages of human Plasmodium parasites

Abstract

For intracellular pathogens, the host cell provides needed protection and nutrients. A major challenge of intracellular parasite research is collection of high parasite numbers separated from host contamination. This situation is exemplified by the malaria parasite, which spends a substantial part of its life cycle inside erythrocytes as rings, trophozoites, and schizonts, before egress and reinvasion. Erythrocytic Plasmodium parasite forms refractory to enrichment remain understudied due to high host contamination relative to low parasite numbers. Here, we present a method for separating all stages of Plasmodium-infected erythrocytes through lysis and removal of uninfected erythrocytes. The Streptolysin O-Percoll (SLOPE) method is effective on previously inaccessible forms, including circulating rings from malaria-infected patients and artemisinin-induced quiescent parasites. SLOPE can be used on multiple parasite species, under multiple media formulations, and lacks measurable impacts on parasite viability. We demonstrate erythrocyte membrane cholesterol levels modulate the preferential lysis of uninfected host cells by SLO, and therefore modulate the effectiveness of SLOPE. Targeted metabolomics of SLOPE-enriched ring stage samples confirms parasite-derived metabolites are increased and contaminating host material is reduced compared to non-enriched samples. Due to consumption of cholesterol by other intracellular bacteria and protozoa, SLOPE holds potential for improving research on organisms beyond Plasmodium.

Figure 1

SLOPE enrichment overview. (a) The asexual replication cycle of Plasmodium occurs inside erythrocytes. Both P. falciparum and P. knowlesi, which take 48 and 24 hours, respectively, to complete the replication cycle, can be propagated in vitro indefinitely. (b1) Hemolytic activity of Streptolysin-O (SLO) was assessed on uninfected erythrocytes to define a unit (the amount of SLO necessary for 50% lysis of 50 μl of uninfected erythrocytes at 2% hematocrit in PBS for 30 min at 37 ^oC). (b2) Ring stage synchronized cultures were treated with a defined quantity of SLO units to preferentially lyse uninfected erythrocytes. (b3) SLO treated samples were layered over a 60% Percoll gradient and centrifuged to separate lysed ghosts from intact cells. (b4) The upper layer of Percoll containing lysed ghosts was discarded while the lower, intact, infected erythrocyte enriched fraction was collected. Uninfected erythrocytes, red circles; Infected erythrocytes, red circles with black dots; lysed membranous ghosts, white circle with dashed outline.

Figure 2

SLOPE enriches ring stage Plasmodium parasites irrespective of species or media formulation. (a) SLO lysis of uninfected erythrocytes (uRBCs) and infected erythrocytes (iRBCs) from (left) ring-stage synchronized P. falciparum grown in RPMI 1640 supplemented with 20% human serum (N = 6; 3 replicates each of lines Hb3 and K1), (middle) RPMI 1640 supplemented with Albumax II (N = 9; 3 replicates each of lines Hb3, K1, and MRA 1240), and asynchronous P. knowlesi grown in RPMI 1640 supplemented with Albumax II (N = 3). (b) SYBR-Green based flow cytometry measurements before and after SLOPE enrichment. Flow plots show single cells within the erythrocyte size range. The infected erythrocyte fraction (“iRBCs”) is denoted within the dashed red gate. (c) Parasitemia fold increase upon treatment with increasing SLO units relative to untreated controls. Represented samples were grown in RPMI 1640 supplemented with Albumax II (N = 9; 3 replicates each of lines Hb3, K1, and MRA 1240). All error bars represent S.E.M.

Figure 3

Validation of ghost separation from intact erythrocytes by Percoll step. (a) Intact erythrocytes are shown as CD235a (red) positive and spectrin (green) negative. Lysed erythrocytes, termed ghosts, are shown as CD235a and spectrin double positive (yellow in merge). All images show ring-stage P. falciparum line MRA 1240 parasites stained with SYBR Green (cyan). 40X Magnification; bar represents 10 µm. Saponin samples were treated with 0.15% saponin for 5 minutes. SLO samples were treated with 40U of SLO for 6 minutes but were not centrifuged through a Percoll gradient. SLOPE samples were also treated with 40U SLO but were subjected to Percoll gradient centrifugation. (b) Proportions of lysed ghosts and intact erythrocytes quantified from fluorescence microscopy imaging across different treatments using SYBR Green dye and CD235a and spectrin antibodies (N = 3; 400 erythrocytes per condition per preparation, error bars represent S.E.M.).

Figure 4

SLOPE enriched parasites remain viable. (a) Mitoprobe DiIC₁ (5) mitochondrial membrane potential (MMP) measurements obtained by flow cytometry in untreated and SLOPE enriched ring-stage P. falciparum line MRA 1240 parasites (N = 3, error bars represent S.E.M.). (b) Six days of P. falciparum line MRA 1240 parasite growth from untreated controls or SLOPE enriched samples diluted with uninfected erythrocytes (N = 4, error bars represent S.E.M.). (c) Ring and early trophozoite stage P. falciparum line Dd2 infected erythrocytes visualized by Giemsa-stain at 100X magnification; bar represents 10 µm.

Figure 5

SLOPE enrichment increases detection of ring stage parasite metabolome. (a) Pipeline outlining metabolomics sample preparation and analysis. (1) Ring-stage synchronized P. falciparum Dd2 cultures were split into two fractions: one portion was taken for the untreated group and one portion was SLOPE enriched. (2) Equal numbers of erythrocytes from untreated and SLOPE groups were saponin lysed and washed to remove cytosolic erythrocyte metabolites. (3) Metabolism of the resulting pellet containing erythrocyte ghosts and parasites was quenched and metabolites were extracted. (4) Metabolites were identified and quantified from extracts using the AbsoluteIDQ p180 kit. These data were log transformed, centered, and scaled prior to statistical analysis. (b) Principal component analysis was performed on all metabolites detected in at least 50% of samples. Significance between untreated and SLOPE groups was determined by perMANOVA: p = 0.037. Ellipses show 95% confidence intervals.

Figure 6

SLOPE is effective on clinical samples in a cholesterol dependent manner. (a) SLO lysis of erythrocytes from P. falciparum infected patients either directly isolated from the patient (0 h) or after 6 h of incubation in complete media (RPMI supplemented with 20% serum). N = 3 patients. Error bars represent S.E.M. (b) SLO lysis of laboratory P. falciparum either grown for 48 h in complete media or complete media supplemented with 4 mM cholesterol saturated mβCD. N = 6 (left; 3 replicates each of lines Hb3 and K1); N = 3 (right; line Hb3). Error bars represent S.E.M. Black circles in selected graphs show iRBC lysis at 27U SLO.

Figure 7

SLOPE is effective on DHA-induced quiescent parasites. (a) Erythrocytes are shown by CD235a staining and P. falciparum MRA 1238 parasites are shown by SYBR Green. Within the DHA-treated image, a dead parasite (left) failed to accumulate MitoTracker Deep Red, while two quiescent parasites accumulated MitoTracker Deep Red. 63X magnification; bar represents 10 µm. (b) Flow cytometry plot measuring quiescent parasites as SYBR Green and Mitoprobe double positive events. Dead parasites are SYBR Green positive, but Mitoprobe negative; uninfected erythrocytes (uRBCs) are SYBR Green and Mitoprobe double negative. iRBC = infected erythrocyte.

Figure 8

Proposed mechanism for decreased susceptibility of infected erythrocytes to SLO lysis. Upon invasion of an erythrocyte, the parasite salvages host membrane cholesterol leading to lower cholesterol levels on the infected erythrocyte surface (top of diagram). During in vitro incubation in conditions with sub-physiological cholesterol levels, cholesterol remains low via continued parasite scavenging. Upon exposure physiological levels of cholesterol, such as in vivo, the exchange of cholesterol between plasma and erythrocytes restores erythrocyte cholesterol to near pre-invasion levels.