The effects of dyslipidaemia and cholesterol modulation on erythrocyte susceptibility to malaria parasite infection
Malaria disease commences when blood-stage parasites, called merozoites, invade human erythrocytes. Whilst the process of invasion is traditionally seen as being entirely merozoite-driven, emerging data suggests erythrocyte biophysical properties markedly influence invasion. Cholesterol is a major determinant of cell membrane biophysical properties demanding its interrogation as a potential mediator of resistance to merozoite invasion of the erythrocyte.
Biophysical measurements of erythrocyte deformability by flicker spectroscopy were used to assess changes in erythrocyte bending modulus on forced integration of cholesterol and how these artificial changes affect invasion by human Plasmodium falciparum merozoites. To validate these observations in a natural context, either murine Plasmodium berghei or human Plasmodium falciparum merozoites were tested for their ability to invade erythrocytes from a hypercholesterolaemic mouse model or human clinical erythrocyte samples deriving from patients with a range of serum cholesterol concentrations, respectively.
Erythrocyte bending modulus (a measure of deformability) was shown to be markedly affected by artificial modulation of cholesterol content and negatively correlated with merozoite invasion efficiency. In an in vitro infection context, however, erythrocytes taken from hypercholesterolaemic mice or from human clinical samples with varying serum cholesterol levels showed little difference in their susceptibility to merozoite invasion. Explaining this, membrane cholesterol levels in both mouse and human hypercholesterolaemia erythrocytes were subsequently found to be no different from matched normal serum controls.
Based on these observations, serum cholesterol does not appear to impact on erythrocyte susceptibility to merozoite entry. Indeed, no relationship between serum cholesterol and cholesterol content of the erythrocyte is apparent. This work, nonetheless, suggests that native polymorphisms which do affect membrane lipid composition would be expected to affect parasite entry. This supports investigation of erythrocyte biophysical properties in endemic settings, which may yet identify naturally protective lipid-related polymorphisms.
KeywordsPlasmodium falciparum Red blood cell Host–parasite interactions Flicker microscopy Membrane biophysics Merozoite
green fluorescent protein
low density lipoprotein (LDL) receptor
high fat diet
Cholesterol is a key constituent of human cells and plays a key role in modulating membrane properties, influencing both membrane fluidity  and stiffness [2, 3]. The effect on these cellular properties is mediated by cholesterol’s flat rigid structure which is defined by the planar tetracyclic ring shape of the molecule. While an elevation in the cholesterol content contained within cell membranes is expected to lead to a reduction in cell elasticity, how these levels are regulated and how dynamic they are in human erythrocytes remains largely unknown. A significant change in plasma lipid levels is medically described as dyslipidaemia, a term used to categorize a number of conditions, including hypercholesterolaemia, which is associated with an increase in plasma cholesterol levels. Importantly, whether and how such an increase in plasma cholesterol would affect cellular membrane composition is not clear. For example, a number of studies have found differences in how treatment of elevated plasma cholesterol or clinical conditions with dyslipidaemia is associated with differences in erythrocyte membrane cholesterol levels [4, 5, 6].
Erythrocyte infection lies at the heart of all symptoms of malaria disease. It is established when the blood-stage merozoite form of the Plasmodium parasite attaches to and penetrates the erythrocyte with concomitant formation of a parasitophorous vacuole inside . There is a detailed appreciation of the stepwise molecular events that characterize merozoite invasion , however, the role the erythrocyte plays in the process was, up until recently, largely overlooked . There has been a growing appreciation in the past few years that parasite binding to the erythrocyte, stimulates biophysical changes in the red cell that likely facilitate entry making it energetically more favourable [10, 11]. Further, several key polymorphisms that protect against malaria infection may do so directly by modulating erythrocyte biophysical properties [12, 13]. These polymorphisms are generally associated with changes in either the erythrocyte cytoskeleton  or membrane surface proteins, such as components of the glycophorin family, a well-studied group of erythrocyte surface receptors that are known to be under natural selection, likely from malaria [13, 14, 15]. To date, however, there is little study of the effects of lipid changes in mediating susceptibility to invasion and whether changes in erythrocyte membrane lipid composition might be associated with changes in efficiency of malaria parasite entry. Several studies have explored the complex relationship between obesity, nutrition and malaria. Obesity has been implicated in being protective against cerebral malaria in a mouse model of malaria infection, although no significant difference in parasitaemia was recorded and the mechanism linking the two is still unclear . Other studies, also in mice, have noted a correlation between malaria infection and outcome in hypoglycaemia and hyperinsulinemia models , as well as with mice under calorie restriction . The latter study was found to be due to nutrient sensing by the parasites and subsequent adjustment of multiplication rates through changes in gene expression levels according to nutrient availability. Finally, in humans, assessment of clinical malaria cases in Nigeria noted a negative correlation between malaria infection and serum cholesterol levels  though the power of the study was relatively low.
Given the implied linkages between diet, cholesterol levels and malaria and the clear role cholesterol plays in defining membrane properties of cells, the effect of elevated cholesterol levels on erythrocyte biophysical properties and susceptibility to malaria parasite infection was investigated using both an in vitro human Plasmodium falciparum and murine Plasmodium berghei model.
Human blood samples and serum cholesterol measurements
Human erythrocytes (O+ , male) for parasite invasion work were obtained from the NHS Blood and Transplant. Approval for collection of clinical human blood samples was granted via the Imperial College Healthcare Tissue Bank, National Research Ethics approval number 17/WA/0161, project ID R18015 and all methods were performed in accordance with the stipulated guidelines and regulations. Blood samples were collected with informed consent from hypercholesterolaemic patients attending the Imperial College London NHS lipid clinic, and from severely hypercholesterolaemic patients undergoing lipoprotein apheraesis as a treatment for homozygous hypercholesterolaemia. EDTA whole blood was collected for parasite assays and parallel measurement of serum cholesterol was performed (Abbott Architect assay, North West London Pathology Blood Sciences Laboratories). Anonymized, normocholesterolaemic EDTA whole blood samples, originating from a primary care setting, were obtained from the hospital blood sciences laboratory. No samples from patients with haemoglobinopathy were included.
Hypercholesterolaemia mouse model and serum cholesterol measurement
C57/B6 Ldlr−/− mice were obtain directly from the Jackson Laboratory (https://www.jax.org/). The mice were fed either on normal chow (SAFE diet 105) or Western High Fat (Dietex, FAT 21%, Cholesterol 0.15%) diets for 8 weeks. Total cholesterol and HDL cholesterol were measured using an enzymatic method in a Siemens Dimensions RxL analyser, following manufacturer’s instructions.
Cholesterol (Sigma-Aldrich) was made up in ethanol at a concentration of 15 mg/ml (38.8 mM). A 5% (38.8 mM) methyl-beta-cyclodextrin (MβCD) stock was made up in MQ-water and heated up on a hotplate stirrer set to 80 °C. 4 × 10 μl aliquots of the 15 mg/ml cholesterol stock was added to 400 μl of the MβCD stock on the stirrer, leaving 10 min between each aliquot. The resulting mixture contains MβCD-Cholesterol at a ratio of 10:1. The MβCD-cholesterol mixture was left stirring on the hotplate for 1 h (to evaporate the ethanol and allow complex formation) and either used immediately or stored at − 20 °C. Erythrocytes were incubated with MβCD-cholesterol complexes for 30 min at room temperature while shaking, with concentrations of up to 3.88 mM MβCD − 388 μM cholesterol. Following incubation, erythrocytes were spun down at 800×g and resuspended into fresh RPMI media before use for further experiments.
Erythrocyte membrane cholesterol measurements
Membrane cholesterol was quantified using a fluorometric Cholesterol Quantitation Kit (Sigma-Aldrich, UK) according to manufacturer instructions. In short, membranes were extracted with a chloroform-isopropanol-IGEPAL CA-630 solution and spun at 13,000×g for 10 min, before the organic phase was transferred to a new tube and dried under nitrogen. Samples were put under vacuum for 30 min to remove any residual organic solvents before the lipid films were dissolved in Cholesterol Assay Buffer and vortexed until the mixture was homogenous. Fluorescence intensity was measured using a Tecan MPro 200 fluorescent plate reader (excitation: 535, emission: 587 nm).
Parasite in vitro culture
Plasmodium falciparum parasites (D10-PHG ) were maintained in standard culture conditions . Parasites were grown in human O+ erythrocytes at 4% haematocrit in complete RPMI1640-HEPES media (Sigma-Aldrich) supplemented with 0.3% l-glutamine, 0.05% hypoxanthine, 0.025% gentamicin and 0.5% Albumax II (Life Technologies). Parasites were grown at 37 °C in 1% O2, 5% CO2 in N2 and D10-PHG cultures were supplemented with 25 ng/ml pyrimethamine and 5 μg/ml blasticidin (Sigma-Aldrich).
Plasmdoium falciparum merozoite invasion assays
The P. falciparum merozoite assay was carried out according to published protocols [22, 23]. For a standard merozoite invasion assay, approximately 90 ml of 5% synchronized late stage D10-PHG (constitutively expressing green fluorescent protein (GFP)) parasites were magnetically isolated as described in , incubated with 10 μM of the cysteine protease inhibitor l-transepoxysuccinyl-leucylamido-(4-guanidino)butane (E64, Sigma-Aldrich) for up to 6 h under standard culture conditions until at least 50% of schizonts had developed into parasitophorous vacuole enclosed membrane structures (PEMS) . To obtain free merozoites, PEMS were centrifuged at 800×g for 5 min (no brake) and resuspended into a minimum of 750 μl of RPMI. The concentrated PEMS-solution is then passed through a 1.2 μm Ministart syringe filter (Sartorius Stedim Biotech) to release individual merozoites.
For quantifying erythrocyte invasion rates, 25 μl of the filtered merozoite solution was added to a prepared erythrocyte suspension (10 μl of 1.5% haematocrit) in a 96 well plate and incubated at 37 C on a shaker (300–500 RPM) for 20 min. An additional sample containing an invasion inhibitor (typically 100 nM cytochalasin D) was included in each experiment and used for accurate gating of the invaded population. Parasites were stained with 100 μl of 5 μg/ml Ethidium Bromide (EtBr) for 10 min at room temperature, washed twice in 100 μl phosphate buffered saline (PBS), and resuspended into 60 μl. Invasion was quantified by flow cytometry, acquiring a total of 100,000 events per well using a BD Fortessa flow cytometer equipped with a high-throughput plate reader. Cells were excited at 488 nm and emission read between 515 and 545 nm for GFP positivity and between 600 and 620 for Ethidium Bromide positivity. New ring-stage parasites were quantified based on their GFPpositive, EtBr-low staining profile . Data analysis was carried out in FlowJo v10 by gating on cells based on the side-scatter to forward-scatter profile (i.e. gating out debris), selecting single cells based on the forward-scatter width to area ratio and finally GFP and EtBr profile. Data was visualized in Prism v7 (GraphPad).
Plasmodium berghei merozoite assays
Female Theiler’s Original (TO) mice, 8–10 weeks of age (Envigo, UK) were injected with 200 μl glycerol stocks of blood infected with 3% P. berghei (strain ANKA 507) expressing GFP. When parasitaemia reached approximately 3%, parasites were harvested by cardiac puncture and incubated overnight at 37 °C in 30 ml complete RPMI-1640 media under low oxygen conditions. The following day schizonts were purified using a magnetic separation column (MACS, Miltenyi Biotec), Purified schizonts were ruptured with a 1.2 μm Ministart syringe filter (Sartorius Stedim Biotech) to release individual merozoites and 25 μl merozoite solution was added to blood obtained from mice of interest (10 μl at 100,000 cells per μl, counted as described below) in a 96 well plate. The culture was incubated for 20 min on a shaker at 37 °C. Cells were stained and quantified as per the P. falciparum merozoite assay’quantification of invasion by flow cytometry’ described above.
Flow cytometry and bead counting
The counting bead (CountBright Absolute Cell Counting Beads, Thermo-Fisher) solution was made up of 400 μl PBS, 50 μl counting bead solution and 50 μl of 1% haematocrit blood. To ensure equal number of beads across the samples, a mastermix of PBS-Counting beads was made up first, mixed well and separated into the number of samples required before the cell solution was added to the bead-mix. Analysis was carried out on a BD Fortessa flow cytometer. Cells were excited at 488 nm and emission read between 515 and 545 nm. 2000 beads were acquired per sample. Using ratiometric analyses based on the number of beads and cells acquired, the number of cells per μl of sample can be calculated. Data analysis was carried out in FlowJo v10 by gating on cells based on based on the forward-scatter and GFP negativity and beads based on their forward scatter profile and GFP positivity.
Artificial incorporation of membrane cholesterol inhibits Plasmodium falciparum merozoite invasion
Establishing an in vitro merozoite invasion assay for Plasmodium berghei
Plasmodium berghei merozoite invasion efficiency is not affected by increased serum cholesterol in a mouse model of hypercholesterolaemia
Erythrocytes from mice with high serum cholesterol do not show altered biophysical properties
Plasma cholesterol quantification of LDL-R−/− mice
Plasmodium falciparum merozoite invasion efficiency is not affected by increased serum cholesterol in patients with hypercholesterolaemia
Erythrocytes from patients with high serum cholesterol do not show altered biophysical properties or cholesterol content
This work set out to explore the role of erythrocyte lipid content and susceptibility to malaria. A key finding from the work is that artificial manipulation of erythrocyte cholesterol content clearly affects both red cell biophysics and merozoite invasion. By altering the cholesterol content in vitro, the erythrocyte bending modulus (a measure of membrane deformability) can be shown to markedly increase and this negatively correlates with merozoite invasion efficiency. Extending this observation to medical conditions associated with increased cholesterol levels, the question was then asked whether in vivo conditions also lead to stiffer erythrocytes and an associated reduction in merozoite invasion efficiency. Specifically, the biophysical properties of cells obtained from a murine hypercholesterolaemia model and from human clinical samples with reported hypercholesterolaemia were quantified and subsequently used in quantitative merozoite invasion assays. Strikingly, in both the murine model of hypercholesterolaemia and clinical samples of patients with elevated serum cholesterol, the elevated levels of cholesterol did not correlate with changes in erythrocyte cholesterol content. In line with this, hypercholesterolaemic blood samples showed no comparable changes in malaria parasite infection rates compared to matched controls. Thus, whilst changes in cholesterol do directly impact on red cell susceptibility to malaria parasite invasion, the balance of lipid homeostasis in the body likely buffers the red cell from altering its cholesterol content, even in the dramatic situation of hypercholesterolaemia.
Due to the difficulty in obtaining large numbers of fresh blood samples from hypercholesterolaemic patients not on statins, it is possible that very subtle effects on red cell cholesterol content may not be found with the given sample size. Investigating potential effects on erythrocyte membrane cholesterol content within a clinical setting matched the situation found in a mouse hypercholesterolaemia model, suggesting that in vivo buffering of red cell cholesterol does occur. Of note, blood samples of patients with homozygous familial hypercholesterolaemia were included, a rare condition which leads to extremely elevated levels of serum cholesterol. Even in these rare cases, red cell membrane cholesterol did not differ significantly from samples obtained from healthy patients. This therefore implies cholesterol content in the erythrocyte is tightly regulated and not significantly affected by serum cholesterol content.
The lack of correlation between plasma and erythrocyte membrane cholesterol was surprising considering that a similar study carried out in two animal models of hypercholesterolaemia found both plasma and membrane cholesterol levels responding to diet , suggesting that erythrocyte membrane cholesterol is dynamic and modifiable under certain conditions. In humans, the picture is not as clear, while both diet  and lipid lowering drugs  have been shown to affect erythrocyte cholesterol content, other studies report no direct correlation between plasma and cell membrane cholesterol content [5, 32]. Importantly, how erythrocyte cholesterol content is regulated is not well understood, since erythrocytes neither produce cholesterol de novo nor contain receptors, such as LDL-R, which allow uptake of lipoprotein molecules . The most likely mechanism for the induced changes in erythrocyte cholesterol content is a dynamic exchange between plasma and membrane cholesterol [34, 35], with the rate of exchange likely being influenced by the structure and composition of the cell .
Despite the lack of changes seen in the samples tested, the question of lipid modulation, its concomitant change in erythrocyte biophysical properties and invasion susceptibility is still pertinent. Manipulating the erythrocyte modulus using different methods (using 7KC in a previous study and cholesterol enrichment here) both affected Plasmodium parasite invasion efficiency. A consistent working hypothesis is that this results in a direct change in the cell’s bending modulus that significantly alters parasite invasion susceptibility. Secondary effects or interactions triggered by the induced enrichment of membrane cholesterol cannot be excluded with full certainty. Cholesterol content has been reported to influence numerous cellular processes for instance changing ligand binding activity through triggering of conformational changes in other lipid groups  and altering the membrane dipole potential, thereby affecting protein activity .
The work presented here demonstrates that erythrocyte biophysical properties are markedly affected by direct modulation of lipid, here cholesterol, content. Modulating the biophysical properties in this way negatively impacts merozoite invasion efficiency. Counterintuitively, however, an increase in serum cholesterol in vivo does not significantly affect erythrocyte membrane cholesterol. Thus, classical conditions in which serum cholesterol is markedly elevated, such as hypercholesterolaemia, are not protective directly against the process of merozoite invasion. The impact of lipid changes to the erythrocyte is nonetheless suggestive that other metabolic causes of variation in erythrocyte biophysical properties caused by lipid content would still be possible. Future work focusing on sampling and identifying individuals from malaria endemic regions with erythrocytes that have abnormal biophysical properties would be predicted to enable the identification of new protective polymorphisms against malaria disease.
We thank Jane Srivastava for assistance with flow cytometry (Imperial College Flow Cytometry Facility), Kate Wright and Fernando Sanchez-Roman Teran for assistance with P. falciparum merozoites and Jason L Johnson for generously providing additional mouse models (eventually not used in this study).
All authors participated in the design of experiments and analysis of data. MK, FA, AMB, ZM and YL performed different components of the P. berghei infection experiments and analysed data. MK JC and BJ collected clinical samples and processed cells for infections. MK performed all in vitro P. falciparum work. All authors contributed to the editing of the manuscript. MK and JB wrote the manuscript. All authors read and approved the final manuscript.
Tissue sample collection was supported by the National Institute for Health Research (NIHR) Biomedical Research Centre based at Imperial College Healthcare NHS Trust and Imperial College London. M.K. is supported by a PhD scholarship from the UK Medical Research Council (MR/K501281/1). B. J. is supported by an NIHR Clinical Lectureship. Research was directly supported by an Investigator Award from Wellcome (100993/Z/13/Z J.B.). The views expressed are those of the authors and not necessarily those of the NHS, the NIHR or the Department of Health.
Ethics approval and consent to participate
Human erythrocytes (O+ , male) for parasite invasion work were obtained from the NHS Blood Transfusion Service. Approval for collection of clinical human blood samples was granted via the Imperial College Healthcare Tissue Bank, National Research Ethics approval number 17/WA/0161, project ID R18015 and all methods were performed in accordance with the stipulated guidelines and regulations. Blood samples were collected with informed consent from hypercholesterolaemic patients attending the Imperial College London NHS lipid clinic, and from severely hypercholesterolaemic patients undergoing lipoprotein apheraesis as a treatment for homozygous hypercholesterolaemia.
Consent for publication
All authors give their consent for publication.
The authors declare that they have no competing interests.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.