Background
Malaria is a severe threat to global health with an estimated 300–500 million cases annually and >1 million deaths. In endemic areas of Africa, malaria accounts for 25 % of paediatric deaths [
1,
2].
Plasmodium falciparum is by far the deadliest of the four human malarial species (
P. falciparum,
Plasmodium malariae,
Plasmodium ovale, and
Plasmodium vivax). The clinical symptoms of malaria occur at the blood stage when parasites rapidly replicate and lyse red blood cells (RBCs) leading to anemia and high fever. In addition, parasitized RBCs can become trapped and cause lesions in brain capillaries resulting in cerebral malaria (CM), the most lethal form of the disease [
3].
Treatment of clinical malaria relies on a handful of drugs, the most potent being artemisinin derivatives (called the artemisinins or ADs), a group of structurally related molecules that includes artesunate (ART), artemether (ARTM) and arteether (ARTE) [
4‐
6]. To avoid appearance and spread of artemisinin resistance in
P. falciparum, the World Health Organization (WHO) has strongly discouraged use of artemisinins as a single agent, and artemisinin monotherapy is only recommended for lethal cerebral malaria [
3]. Rather, and for all other forms of clinical disease, artemisinins are administered in combination (artemisinin combination therapy; ACT) with other, long acting, anti-malarial drugs that include mefloquine, lumefantrine, amodiaquine, piperaquine and sulfadoxine/pyrimethamine [
4,
5]. Because no other drugs as potent as artemisinins are available, the potential emergence of artemisinin resistance at Thai/Cambodia border and its spread to other areas has caused significant concerns [
7‐
11], including the failure of front line ACT due to secondary partner drug resistance. Although the mechanisms of artemisinin resistance in
Plasmodium are being characterized [
12‐
16], novel adapted treatment options based on this knowledge are years away. Hence, improving the efficacy of artemisinins and of ACT represents a major short-term goal in treating and preventing the spread of artemisinin resistance in the malaria parasite. Likewise, increasing effectiveness of adjunct treatment to artemisinin monotherapy may improve the outcome of cerebral malaria, the most severe and most difficult to treat complication of malaria.
Studies in human populations from areas of endemic disease have long established the critical role of genetic factors in susceptibility and protection against malaria [
17,
18]. Examples include the protective effect of heterozygosity for loss of function variants at erythrocyte-specific proteins such as haemoglobin (sickle cell anaemia, thalassaemias), glucose 6-phosphate dehydrogenase, anion exchanger 1 (SLC4A1), Duffy antigen (DARC), ABO blood group variants and several others [
17,
19]. Likewise, studies in mouse models of blood stage malaria (
Plasmodium chabaudi AS), and of cerebral malaria (
Plasmodium berghei ANKA) have also established a strong genetic control of resistance and susceptibility to malaria, and the molecular basis has been characterized in several instances [
20], providing potentially useful entry points for discovery of novel anti-malarial drugs or other treatment modalities [
21,
22].
In a mouse model of infection with
P. chabaudi, it was reported that a loss of function in the enzyme pantetheinase (
Vnn1/Vnn3) in mouse strain AcB61 causes susceptibility to blood-stage malaria [
23]. The
Vnn1/Vnn3 pantetheinases are enzymes that hydrolyze pantetheine to pantothenic acid (vitamin B5) and Cysteamine (Cys). Furthermore, Cys displays modest but significant anti-malaria activity, and Cys treatment can significantly improve the response of mice to blood stage infection with
P. chabaudi (reduced parasitaemia, increased survival), when given either as a prophylactic (naïve animals) or as a therapeutic (infected animals) regimen [
24]. Ex vivo, Cys inhibits the degradation of hemoglobin by
Plasmodium parasites in erythrocytes [
24].
Cys is a small, naturally occurring amino thiol that has very low toxicity in vivo. Importantly, different Cys formulations are approved for life-long treatment of nephropathic cystinosis (NC), a kidney disorder caused by mutations in the lysosomal cystine carrier cystinosin [
25]. It was reported that Cys dosing regimens that display pharmacokinetic profiles similar to those measured in humans taking oral Cys for the treatment of NC, reduce replication of the malaria parasite in vivo, and reduce lethality from infection [
26]. In the present study, the potential of Cys to increase the anti-plasmodial activity of different artemisinins towards the murine
Plasmodium parasites
P. chabaudi AS (blood-stage), and
P. berghei ANKA (cerebral malaria) was investigated. Cys is shown to potentiate the activity of several artemisinins currently in clinical use in ACT, including ART, ARTE, and ARTM, against the blood stage of the infection by
P. chabaudi in mice. Cys also causes dramatic enhancement of artemisinin efficacy and protective effect against cerebral malaria induced in the mouse by infection with the ANKA strain of
P. berghei.
Methods
Mice and parasites
Eight to twelve weeks old female A/J mice were purchased from the Jackson Laboratories (Bar Harbor, ME, USA) and were housed at McGill University Animal Care Center according to the guidelines of the Canadian Council on Animal Care. Mice experimentation protocol was approved by the McGill Facility Animal Care Committee (P. Gros, Principal Investigator; protocol number: 5287), and include procedures to minimize distress and improve welfare. An isolate of
P. chabaudi AS was provided by Dr. Mary M. Stevenson (McGill University Health Center Research Institute, Montreal), and was maintained by weekly passage in A/J mice for a maximum of four consecutive passages, as previously described [
27].
Plasmodium berghei ANKA parasites were initially obtained from the Malaria Reference and Research Reagent Resource Center (MR4);
P. berghei was passaged in C57BL/6J (B6) mice for a maximum of three consecutive passages. In passage mice, blood parasitaemia was monitored daily, and the percentage of parasitized red blood cells (pRBCs) was determined on thin blood smears stained with Dif-Quik (Dade Behring, Newark, DE, USA), as described [
28]. Infected blood was diluted in pyrogen-free saline for infection of experimental animals and also for preparation of cryo-preserved stocks for long-term storage.
Drugs and drug treatments
Cysteamine hydrochloride (Cys) and N-acetyl cysteine (NAC) (Sigma, Burlington ON, Canada) were prepared fresh in PBS. Artesunate and artemether were generously provided by Dafra Pharma International (Turnhout, Belgium). Artesunate was prepared fresh as a 1 mg/mL stock in 5 % sodium bicarbonate solution which was diluted in PBS immediately before use. Artemether (Artesiane™) prepared from fractionated coconut oil stock (20 mg/mL, injectable) was diluted in sesame oil to desired concentration. Beta-Arteether (Artecef™, Artemotil; 50 mg/mL injectable in sesame oil) was diluted in sesame oil to desired concentration. Beta-Arteether was generously provided by Artecef BV (Zeewolde, The Netherlands). Drugs were administered via intraperitoneal (i.p) injection and according to a four-days treatment schedule. Mice were weighed prior to treatment to determine appropriate doses and injection volumes ranged from 100–400 μL per mouse. In the case of animals treated with two drugs, Cys was administered first, followed by artemisinin derivatives 15 min later on alternate sides of the mouse abdomen. Untreated control animals were injected with PBS alone.
Infections
For in vivo infections, mice were infected i.v. (0.2 mL) with 10
7
P. chabaudi pRBCs into the tail vein, and monitored under BSL2 containment conditions. One hour post-infection, mice received either PBS or different drugs or drug combinations (i.p), and this treatment was repeated at 24, 48, and 72 h post-infection. At day 4 and 5 post-infection, blood was sampled, and blood parasitaemia was determined (between 4–10 fields of 100 erythrocytes were counted per mouse, and results are expressed as percentage of parasitized erythrocytes). For experiments involving ex vivo drug exposure followed by in vivo infections, parasitized erythrocytes were exposed to certain compounds for 1 h at 37 °C. Treated pRBCs were then rinsed free of drug, and resuspended in drug-free PBS (at 1 × 10
6 pRBC/mL), and were then used to infect naïve mice (1 × 10
5 pRBC). The ability of the treated parasites to induce a productive infection was monitored by measuring blood parasitaemia daily, and by monitoring appearance and progression of lethal and irreversible symptoms of cerebral malaria, as we described [
29].
Pharmacokinetic studies of artemisinin in vivo
Mice were injected with Cys (140 mg/kg, 0.2 mL, i.p.) in the lower left abdomen, and exactly 15 min later, ART (0.5 mg/kg, 0.2 mL, i.p.) was injected in the lower right abdomen. At 5, 10, and 20 min, mice were euthanized and bled by cardiac puncture and blood was collected in EDTA-containing tubes to isolate plasma. Plasma samples showing hemolysis were not used for ART determination as haemolysis interfered with ART measurement. Control plasma from untreated mice was also included as a placebo for spiked calibration references covering a range between 5 and 2000 ng/mL for ART as well as DHA, and QC samples (10 and 100 ng/mL). The method consisted of a sample preparation step, followed by UPLC with MS (MRM-mode) detection. Briefly, a protecting solution consisting of potassium oxalate and sodium fluoride was added to ice-cold mouse plasma, followed by the addition of the ART-d3 and DHA-d3 internal standard solutions. After mixing, the diluted plasma samples were loaded on preconditioned µ-HLB-SPE wells, which are washed with water followed by 5 % acetonitrile in water. Elution of the compounds was done with a methanol–acetonitrile mixture, and diluted with water before injection in the UPLC system, consisting of an Acquity BEH shield RP18 column with a mobile phase consisting of water-acetonitrile-formic acid (50/50/0.1) in isocratic mode. The MRM transitions used for quantification were m/z 283→265, resp 286→268, for ART and its internal standard, and m/z 323→240, resp 326→243, for DHA and its internal standard. The method was validated and found to comply for linearity, selectivity, recovery, accuracy and precision.
Plasmodium berghei parasites were subjected to ex vivo drug exposure followed by intravenous injection in B6 mice. At day 6 post-infection, mice were injected intravenously with 0.2 mL 1 % Evans blue dye prepared in sterile PBS. One hour later, mice were exsanguinated and perfused with PBS-containing 2 mM EDTA. Brains were excised, imaged, and incubated with 1 mL of dimethyl formamide for 48 h to extract Evan’s blue dye from the tissues. Optical density was measured at 610 nm, and measurements were converted into µg of dye extravasated per gram of tissue.
Leukocyte infiltration in the brain
Six days post-infection, mice were exsanguinated and perfused as above; brains were harvested and homogenized in RPMI media-containing 0.5 mg/mL collagenase (Gibco LifeTechnologies), 0.01 mg/mL DNAse I (Roche) and 2 mM EDTA. Infiltrating cells were separated on a 33.3 % Percoll solution. Cells were stained for flow cytometry analyses with the following; Zombie Aqua Dye-V500 (BioLegend), anti-CD45-APC-eFluor780 (clone 30-F11, eBioscience), anti-CD4-PerCP Cyanine5.5 (clone RM4-5, eBioscience), anti-CD8alpha-PE (clone 53-6.7, eBioscience), anti-TCRbeta-FITC (clone H57-597, eBioscience), anti-CD19-BV421 (clone 6D5, BioLegend), F4/80-eFluor450 (clone BM8, eBioscience), anti-CD11b-APC (clone M1/70, eBioscience), anti-CD11c-PerCP Cyanine5.5 (clone N418, eBioscience), anti-Ly6C-PE (clone HK1.4, eBioscience), and anti-Ly6G-FITC (clone 1A8, BioLegend). Leukocytes were gated as CD45hi cells.
Statistical tests
Groups with normally distributed data points were compared using parametric unpaired t tests, while groups with non-Gaussian distributions were compared using non-parametric Mann–Whitney tests. Survival differences were analysed using the Log-Rank test. Standard error of percent inhibition was calculated from individual mice compared to the mean parasitaemia level of the control group.
Discussion
In previous reports from our group, Cys was shown to have low but significant anti-plasmodial activity, and could also potentiate ART in an in vivo model of blood-stage malaria with
P. chabaudi [
27]. Building on these initial results, the current studies demonstrate that (a) the effect of Cys is broad, as it is found to potentiate the anti-plasmodial activity of structurally distinct artemisinin drugs in current clinical use in ACTs, (b) Cys increases the anti-plasmoidal activity of ART against both blood stage and cerebral forms of malaria, and (c) the Cys effect is direct, does not require host factors, and can be demonstrated on isolated
Plasmodium-infected erythrocytes.
Cys can strongly potentiate the anti-plasmodial activity of the artemisinin derivatives ART, ARTM and ARTE. Artemisinin is itself a poor water-soluble and unstable sesquiterpene lactone that is not generally used clinically. ART, ARTM, ARTE, and dehydroartemisinin (DHA) are structural analogs of artemisinin that are in clinical use in current ACTs, including DHA-piperaquine (Artequick), ART-mefloquine, ART-amodiaquine, ART-sulfadoxine-pyrimethamine, ART-sulfamethoxypyrazine-pyrimethamine, and ARTM-lumefantrine [
4,
5,
32]. In a standard four-day test in mice, Cys can strongly potentiate the anti-plasmodial activity of ART, ARTM, and ARTE against the blood stage parasite
P. chabaudi, as determined by strong reduction of blood parasitaemia, and increased survival from acute infection. The observed effect of Cys on ART, ARTM, and ARTE suggests a potentiation mechanism that is independent of side chains substitutions of the main sesquiterpene lactone. Importantly, these results suggest that inclusion of Cys to current artemisinin-based anti-plasmodial drug formulations could improve efficacy of several ACTs. Furthermore, Cys potentiation of the anti-plasmodial activity of ART is observed for
Plasmodium parasites that cause different pathologies, namely haemolytic anaemia associated with blood-stage replication of
P. chabaudi AS, but also cerebral malaria associated with neuroinflammation caused by trapping of
P. berghei parasitized erythrocytes at the blood brain barrier [
23,
24,
26]. Hence, it is tempting to speculate that Cys may prove beneficial against different forms of malaria in humans.
What is the mechanism by which Cys potentiates the anti-plasmodial activity of artemisinins? This question remains unanswered. However, the present study shows that Cys potentiation may affect several aspects of the pathophysiology of
Plasmodium infection. First and foremost, in vitro studies show that Cys can directly enhance the anti-plasmodial activity of artemisinins (in a 1 h assay), and without involvement of host-based mechanisms: Indeed, erythrocytes parasitized with different
Plasmodium species (
P. chabaudi,
P. berghei) and exposed to the Cys/artemisinin combinations show significantly reduced capacity to induce a productive infection in naïve hosts (reduced blood parasitaemia and clinical symptoms, and increased survival). The observation that in this 1 h exposure assay, Cys can potentiate artemisinin activity, reducing infectivity and improving clinical endpoints (lethality) of
Plasmodium parasite infections that involve different pathogenesis and causes of death (severe anaemia, encephalitis) support a direct increase in anti-parasitic property of the artemisinin/Cys combination. This proposal is supported by studies showing that Cys has modest but significant anti-plasmodial properties in vivo [
24,
26].
In addition, Cys may have additional effects in vivo that may further enhance the anti-plasmodial properties of artemisinins. Indeed, in a parallel set of experiments, Cys pre-treatment was observed to affect pharmacokinetics of plasma ART, seemingly delaying the time of peak plasma concentration, from 5 min in controls (no Cys) to 10 min in Cys-treated mice (Additional file
3A, B), although there did not seem to be an effect on bioavailability (area under the curve; AUC). Cys also appeared to have an effect on the bio-transformation of ART to its biologically active species, dehydroartemisinin (DHA) [
33,
34]. Results in Additional file
3C, D show that levels of plasma DHA were significantly higher at 10 and 20 min in the groups receiving Cys than in the group receiving ART alone, suggesting longer persistence of DHA in animals receiving Cys. These results suggest that in addition to the direct potentiation of ART anti-plasmodial activity (Figs.
2,
3,
4), Cys treatment in vivo may favorably modify the pharmacokinetics and metabolism of artemisinin to its active form DHA (enhanced transformation, and increase in plasma AUC), further increasing its anti-parasitic activity.
The mechanistic basis of the observed in vitro and in vivo effects remain to be fully elucidated but may additionally involve the anti-oxidant properties of the thiol group of Cys. Indeed, anti-oxidants have been proposed to improve malaria outcomes, but this has remained controversial. While some studies show that anti-oxidants may improve recovery from
Plasmodium-induced oxidative stress [
35,
36], other studies have not seen these effects, in fact arguing the opposite view as artemisinin plasmodial toxicity depends on in situ generation of free radicals which could be blunted by Cys [
37,
38]. Because several of these studies used N-Acetyl Cysteine as an anti-oxidant, Cys and NAC were compared for potentiation of artemisinin in our experimental conditions. These studies showed that the Cys effect is specific, and that NAC has no activity when used in the same assay conditions (Fig.
4). This is supported by the findings that Cys anti-plasmodial activity is specific and not seen for other thiols such as dimercaptosuccinic acid (DMSA) [
24].
Cys is a small aminothiol produced by the pantetheinase reaction which has very low toxicity, and various formulations of Cys have been approved for clinical use in the life-long management of nephropathic cystinosis, a rare pathology caused by mutations in the cysteine transporter [
25]. For example, starting in early childhood, cysteamine bitartrate is given a following dose escalation schedule to reach a final dosing of ~2 g per day (divided into 4 daily doses) for a ~45–50 kg child. Continuous use of Cys at these doses over several years shows low toxicity in humans [
39]. The pharmacokinetic profile of a single dose of 1.5 g of cysteamine bitartrate in humans (
C
max = 80 μM; AUC = 2845.1 min μM) is similar to single subcutaneous (sc) injection of 50 mg/kg cysteamine hydrochloride in mice [
26]. Here, Cys (cysteamine hydrochloride) can potentiate the anti-plasmodial potency of the artemisinin derivatives ART, ARTE and ARTM in vivo over a wide dose range, and at doses (40–60 mg/kg) that are well within the range of those used to treat human cystinosis. In addition, studies in vitro show that 1 h exposure to doses of Cys as low as 10–25 μM are sufficient to significantly potentiate anti-plasmodial activity of ART against both
P. chabaudi and
P. berghei. These Cys concentrations are well within the range of the
C
max values (20–150 μM) reached by a single oral dose of cysteamine bitartrate used in humans [
39‐
42]. The observations that (a) Cys is well tolerated and already in clinical use for chronic cystinosis, (b) it has anti-plasmodial activity of its own, (c) it can potentiate the effect of artemisinin derivatives currently in use, and at concentrations within the range of those currently in use for cystinosis treatment together strongly suggest that Cys has essential features of a new partner for artemisinin derivatives in current ACT.
Artemisinin included in ACT is currently the most important drug for the clinical treatment of malaria. Resistance to artemisinin in the malaria parasites would be a major threat to global health. Unfortunately, several reports have documented the recent emergence of artemisinin resistance, in particular at the Thai-Cambodia border area [
12‐
16]. This has prompted the search for new therapeutic intervention, and new drug candidates with anti-plasmodial activity [
43‐
48]. Results from the current study suggest that modification of current ACT by incorporation of Cys could represent a significant alternative for management of artemisinin-resistant
Plasmodium malaria parasites.
Authors’ contributions
PG participated in study design, supervised data collection, and ensured the quality of the laboratory results. NM participated in laboratory experiments, data collection and analysis. ST, SG, MT helped in laboratory experiments. MMS provided expertise in study design and data collection. AF, KV, JH and BDS assisted in the design and performed the pharmacokinetics experiments. All authors participated in the writing and review of the manuscript. All authors read and approved the final manuscript.