Background
Malaria remains a major public health problem which affected about 207 million people and caused an estimated 627,000 deaths in 2012 [
1]. In the context of the widespread and increasing occurrence of
Plasmodium falciparum resistance against current anti-malarial therapy, new anti-malarial compounds are urgently needed to treat this major endemic disease. In this perspective, it is interesting to note that for many synthetic anti-malarial drugs,
P. falciparum-resistant isolates were observed one to 12 years after the first use, whereas it was longer for the natural compounds [
2]. Indeed, the first reported resistance towards quinine appeared 278 years after its introduction [
3]. The use of artemisinin combination therapy (ACT) as first-line treatment of uncomplicated malaria caused by
P. falciparum was officially recommended by the WHO in 2006 [
4]. Unfortunately, 2,000 years after the use of
Artemisia annua in the Chinese Pharmacopoeia to treat fever, the emergence of resistance to artemisinin derivatives was recently reported from Southeast Asia [
5,
6]. Molecules, structurally different from the available anti-malarial drugs and targeting innovative and independent metabolism pathways, are particularly needed to prevent the apparition of resistance and to improve care. Drawing from the rich plant biodiversity, new chemical structures may be helpful in the fight against malaria [
7].
The ethnopharmacology, based on traditional medicine, offers interesting possibilities in the discovery of new bioactive compounds isolated from the nature. A collaboration between the Cambodian and French (UMR-MD3) faculties allowed inquiries on 28 Cambodian plants used in traditional medicine [
8]. This work allowed the selection of
Stephania rotunda,
Brucea javanica,
Phyllanthus urinaria and
Eurycoma longifolia, among which,
S. rotunda (Menispermaceae), a creeping plant growing on calcareous cliffs of Cambodian mountain areas [
9], exhibited the most interesting antiplasmodial activity
in vitro. Concentrations inhibiting 50% of parasitic growth (IC
50) of the dichloromethane and water extracts of
S. rotunda tuber were below 5 μg/ml on the
Plasmodium strain W2 [
8]. The fractionation of dichloromethane extracts allowed the isolation of nine alkaloids. The main compound is a bisbenzylisoquinoline, named cepharanthine. This alkaloid has recently been extracted by green chemistry using ultrasound and microwave technologies [
10].
Possessing an interesting IC
50 measured by flow cytometry (0.61 μM on W2 strain), the antiplasmodial activity of cepharanthine was tested in mice infected by
Plasmodium berghei at a dose of 10 mg/kg [
8]. By intraperitoneal injection and oral administration, this alkaloid decreased the parasitaemia by 47 and 50%, respectively. Despite the absence of mice sterilization, this molecule is interesting in combination with other anti-malarial drugs. Indeed, cepharanthine possesses a synergistic activity with chloroquine [
8,
11] but the mechanism of this potentiation is not known currently. Two hypotheses have been proposed to explain this phenomenon: an alteration of the parasite membrane potential [
12] or a modulation of P-glycoprotein [
13] by cepharanthine.
Previous work showed that cepharanthine seemed to possess a putative mechanism of action different from those of anti-malarial drugs commonly used. Indeed, cepharanthine did not affect the crystallization of haem, unlike chloroquine. The measurement of mitochondrial membrane depolarization after labelling with DiOC6 did not show any effect on the mitochondrial membrane potential by cepharanthine, contrary to atovaquone. The use of ascorbic acid as a potential inhibitor of free radical production did not reveal any activity of free radicals production for cepharanthine, contrary to artemisinin and its derivatives [
14].
The work presented here highlights potential plasmodial targets of cepharanthine using both phenotypic and transcriptional approaches.
Discussion
A previous study had suggested that cepharanthine could have an antiplasmodial activity differing from those of available anti-malarial drugs [
14]. This assumption was not contradicted by the evaluation of IC
50 levels on strains that have various sensitivities to anti-malarial drugs. Indeed, assays performed on 3D7 strain (CQ-sensitive/MQ-resistant) and on the Asian W2, FCM29 and K1 strains (CQ-resistant/MQ-sensitive) showed that the IC
50 ratios of cepharanthine on CQ and MQ moved depending on the stains. For example, the susceptibility of W2 to cepharanthine was higher than K1 whereas the opposite was observed with both CQ and MQ. Furthermore, morphologic characterization of cepharanthine impact on the life cycle of
Plasmodium showed a particular effect depending on the parasite stage. Indeed, at any time of the parasite life cycle, when cepharanthine was incubated with a constant drug pressure,
Plasmodium was blocked at the ring stages and parasitaemia followed down to 0% after a 96- to 144-hr period. When the drug was removed, the life cycle of the parasite returned to normal. This stage effect of cepharanthine is highlighted for the first time. This specificity has enabled the design of a microarray assay at the ring stage and on closely synchronized cultures.
At the transcriptional level, a tight synchronization of parasites was performed within a four-hour time-window, allowing good accuracy of microarray results and reflecting the quality of the study. The transcriptional assay was performed during a short time of eight hours and on a relatively stable parasitic stage. Indeed, the ring stage corresponded to the beginning of the Plasmodium life cycle in which gene expression is slightly modified, few metabolic pathways being established. Despite this, high variations of gene expression were observed between the three conditions studied. This transcriptional study highlights the set of genes whose expression is directly and indirectly disrupted by cepharanthine. Transcriptional analysis performed on ring stage confirmed that the parasitic blockage, microscopically observed, might be related to the metabolism pathways.
The limitations of this study are mainly due to Plasmodium model. The quickness of Plasmodium cycle and the constancy of the parasitic evolution explain the choice of working on short time. The design of the transcriptional experiment performed on ring stages was not easy, involving a large amount of synchronized ring stages and a sufficient incubation time of cepharanthine. So, time points have been chosen to be far from the invasion by merozoites and the transformation into trophozoite stages.
Asahi
et al. studied factors controlling intra-erythrocyte development of
P. falciparum[
27]. They used various chemically defined mediums and after transcriptome profiling, they found 26 transcripts predicted to be associated with the schizogony stunting. Two of their over-expressed transcripts were also significantly upregulated in the study presented here (CSP-TRAP PFC0640w and MYND finger protein PFF0350w). The upregulation of these two genes could be responsible of the blockage of parasites into ring stage by cepharanthine.
A parasitostatic effect on ring stage was also observed after a treatment with a natural triterpene, named limonene [
28]. The freezing of parasite progression from the ring to the trophozoite stage by the limonene could involve a decrease of isoprenylation of proteins as well as rhoptry-associated proteins (RAP). In the study described here, the isoprenylation metabolic pathway was significantly decreased by cepharanthine. This biochemical pathway localized in the apicoplast of the parasite is an interesting target because of its absence in the human host. Moreover, the isoprenoid precursor synthesis is essential for the parasite survival [
28,
29].
At the genomic level, four enzymes of the glycolysis (glucose-6-phosphate isomerase PF14_0341, triose phosphate isomerase PF14_0378, phosphoglycerate mutase PF11_0208 and enolase PF10_0155) and one enzyme of the gluconeogenesis (phosphoenolpyruvate carboxykinase PF13_0234) were down-regulated by cepharanthine and could be responsible for this parasitic blockage. For example, the glucose-6-phosphate isomerase interferes with the second step of the glycolysis corresponding to the conversion of glucose 6-phosphate (G6P) in fructose 6-phosphate (F6P) [
30]. In the GO analysis, the gluconeogenesis was enriched and in the metabolic pathway, the glycolysis was significantly impacted by cepharanthine. According to Bozdech
et al., genes of the glycolysis are induced during the ring and early trophozoite stages [
31]. A downregulation of genes of this metabolic pathway could be responsible for inhibiting the passage of the ring stage to trophozoite stage.
Among the gene families whose expression is downregulated by cepharanthine, some were quoted as essential for parasite survival. The carbon catabolite repressor protein 4-associated factor 1 (CAF-1) exerts a regulation mainly on the red blood cell invasion by merozoite [
32]. The cyclin-dependent kinase cdc2 possesses a regulatory function on the cell cycle evolution in
Plasmodium[
33,
34]. The caseinolytic proteases (ClpB) are chaperones located in the apicoplast and involved in the cellular homeostasis [
35]. The heat shock proteins (HSP) 40, 70, 90 are among chaperones playing an important role in the cellular processes of the parasite survival and pathogenicity [
26,
36,
37]. The SURFIN corresponds to an antigen transported to the red blood cell surface by Maurer’s clefts and located at the merozoite apex. This antigen seems fundamental for the merozoite invasion and parasite survival [
38‐
41]. Others were proposed as potential anti-malarial targets. The acyl-CoA binding proteins (ACBP1 and ACBP2) are involved in the
de novo apicoplast fatty acid biosynthesis [
42]. The aquaglyceroporins are responsible of the urea and glycerol transport [
43]. The 4-(cytidine 5’-diphospho)-2-C-methyl-D-erythritol kinases (CMK) play a catalytic role in the biosynthesis of isopentenyl pyrophosphate [
44]. Two antigens were proposed as malaria vaccine candidates. The glycosylphosphatidylinositol (GPI)-anchored proteins are responsible of the membranous protein binding and merozoite invasion [
45]. The rhoptry-associated proteins (RAP) are immunogenic and involved in the merozoite invasion too [
46,
47]. As both protein targets are well characterized, a Western blot experiment could be performed to confirm the results obtained with the transcriptional assay.
The pathways analysis showed that few pathways are upregulated by cepharanthine. Thus, for a better knowledge of the cepharanthine mechanism and its possible targets, the study focused on genes involved in parasite survival and virulence, downregulated by cepharanthine. The genes’ coding for proteins exported to the host cell by Maurer’s clefts, named the exportome [
26], are generally downregulated by anti-malarials. These clefts are membranous structures involved in the export of parasitic proteins to the erythrocyte membrane [
38] and are widely affected by cepharanthine treatment (GO over-represented with a p-value of 10
-25). This expression modulation is observed for structural proteins of Maurer clefts (
Pf mc-2TM [
26,
48‐
50] and antigen 332) and proteins exported by Maurer’s clefts (ring-infected surface antigen (RESA)-like with PHIST and Dna J domain [
26],
Plasmodium helical interspersed subtelomeric (PHIST) a, b and c [
26], kinase named FIKK [
26,
51‐
54]). Currently, there are no available anti-malarial drugs acting directly on the transport mediated by Maurer’s clefts. It has been shown that the Maurer’s clefts decreased in number but not in function, when treated with artesunate, quinine and piperaquine [
55]. However, this observation does not represent the main mechanism of action of these three drugs. It will be necessary to study the functional modifications of these structures before and after treatment by cepharanthine, with additional methods as imaging techniques, to explain the relationship between cepharanthine and Maurer’s clefts.
The GOs of host cell plasma membrane and antigenic variations, containing mainly genes coding for
Pf EMP1,
RIFIN and
STEVOR, were significantly under-represented in this study (10
-30 < p < 10
-4).
Pf EMP1 is addressed to the host erythrocytes by Maurer’s clefts and is responsible of cyto-adherence inducing cerebral malaria [
56]. Rifin and stevor contribute to the antigenic variation of
Plasmodium conferring its adaptability towards all the antiplasmodial treatments. The downregulation of these three kinds of genes by cepharanthine seems interesting for the inhibition of
Plasmodium virulence. Furthermore, cepharanthine seems to inhibit genes related to pathways involving mitochondrion (p = 1.5 × 10
-4) and apicoplast (p = 8.4 × 10
-5). Proteomic studies showed that these organelles seem to be targeted by doxycycline [
57]. Moreover, targeting mitochondrion electron transport, atovaquone induced a static state on the ring stages [
58]. The parasitostatic effect observed with cepharanthine treatment could be due to its activity on mitochondrion but also on apicoplast. Indeed, a “delay death” has been observed with drugs inhibiting apicoplast as tetracycline and fosmidomycin [
59]. The mechanism involved in this phenomenon is not yet elucidated but it is traduced by a blockage of parasitic growth after the reinvasion of erythrocytes [
29]. This property has also been observed with cepharanthine that induced a blockage of ring stages during the second parasitic cycle. Moreover, being absent in humans, the apicoplast is a specific target. The compounds acting on this organelle could induce good safety in humans. So, it would be interesting to study the effect of cepharanthine on this organelle with further complementary and specific experiments to a better understanding and characterization of this inhibition of isoprenoid precursor biosynthesis.
In the goal to identify the real targets of cepharanthine, the potential targets underlined at the transcriptomic level in this work have to be confirmed at the proteomic level. Moreover the use of imagery technics could be interesting to check the activity of cepharanthine on some targets as Maurer clefts and mitochondrion. Elsewhere the activity of cepharanthine on cytoadherence could be evaluated
in vitro and
in vivo using the Palo-Alto (FUP)1
P. falciparum strain [
60].
Previous pharmacokinetic studies performed in mouse [
61], beagle dog and human [
62] showed a quite long elimination half-life for cepharanthine that could be a potential candidate in combination with faster-acting anti-malarials in the treatment of multidrug-resistant
falciparum malaria in seriously ill patients (ACT combination for example) [
11].
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
EO, DP, CC, and NT designed and coordinated the study. CD and CT performed the transcriptomic assay. CD, CC, JD, and NT analyzed the transcriptomic data and wrote the manuscript. AP and CD performed the cultivation of Plasmodium strains and the qRT-PCR experiment. BB provided cepharanthine for the study. All the authors read and approved the final manuscript.