The most severe form of human malaria is caused by the parasite Plasmodium falciparum. Despite the current need, there is no effective vaccine and parasites are becoming resistant to most of the antimalarials available. Therefore, there is an urgent need to discover new drugs from targets that have not yet suffered from drug pressure with the aim of overcoming the problem of new emerging resistance. Membrane transporters, such as P. falciparum Ca2+-ATPase 6 (PfATP6), the P. falciparum sarcoplasmic/endoplasmic reticulum Ca2+-ATPase (SERCA), have been proposed as potentially good antimalarial targets. The present investigation focuses on: (a) the large-scale purification of PfATP6 for maintenance of its enzymatic activity; (b) screening for PfATP6 inhibitors from a compound library; and (c) the selection of the best inhibitors for further tests on P. falciparum growth in vitro. We managed to heterologously express in yeast and purify an active form of PfATP6 as previously described, although in larger amounts. In addition to some classical SERCA inhibitors, a chemical library of 1680 molecules was screened. From these, we selected a pool of the 20 most potent inhibitors of PfATP6, presenting half maximal inhibitory concentration values in the range 1–9 μm. From these, eight were chosen for evaluation of their effect on P. falciparum growth in vitro, and the best compound presented a half maximal inhibitory concentration of ~ 2 μm. We verified the absence of an inhibitory effect of most of the compounds on mammalian SERCA1a, representing a potential advantage in terms of human toxicity. The present study describes a multidisciplinary approach allowing the selection of promising PfATP6-specific inhibitors with good antimalarial activity.
Malaria is a life-threatening disease caused by an infection by parasites of the genus Plasmodium transmitted by female mosquitoes of the genus Anopheles. Plasmodium falciparum is responsible for the most severe form of human malaria. The absence of a vaccine makes drug treatment the only way to control malaria. According to the World Malaria Report of 2011, the estimations for 2010 were 216 million cases of malaria and 655000 deaths . The World Health Organization (WHO) recommends the use of artemisinin-based combination therapies as the first-line treatment for uncomplicated malaria . The increase of multiple resistances to classical antimalarials (e.g. quinolines and antifolates derivates) is becoming a major issue for public health [3, 4]. Moreover, a decrease of sensitivity to artemisinin has already been reported in Western Cambodia [5, 6]. To anticipate the emergence of new resistances, there is an urgent need to discover new targets and molecules with anti-malaria effects.
The malaria parasite is a complex cell that contains (besides the plasma membrane) several membranes delimiting organelles such as the nucleus, endoplasmic reticulum, Golgi apparatus, vestigial plastid-like apicoplast, mitochondria and food/digestive vacuole . These membranes contain several transporters that are responsible for essential cellular processes such as the movement of solutes across biological membranes, the regulation of essential nutrient uptake, ion homeostasis and the disposal of toxic waste, although transporters may also be part of the regulatory pathways [3, 7]. Membrane transporters such as P-type ATPases are already utilized as specific and potent drug targets in several human diseases [7-9], irrespective of their location on the plasma membrane or in intracellular membranes . Transporters therefore hold great potential as drug targets and have been largely exploited; for example, P. falciparum chloroquine resistance transporter [11, 12], P-glycoprotein homologue 1 [7, 11, 13, 14], P. falciparum multidrug resistance-associated protein [15, 16] and ATP/ADP translocase [7, 17].
Plasmodium falciparum Ca2+-ATPase 6 (PfATP6) is the P. falciparum sarcoplasmic/endoplasmic reticulum Ca2+-ATPase (SERCA) pump, a member of the P-type ATPase family . It is a membrane transporter considered to be localized on the parasite's endoplasmic reticulum and to play a crucial role in calcium homeostasis, and thus is expected to be essential for parasite growth and survival. The PfATP6 pump could therefore constitute a promising drug target for the treatment of malaria . PfATP6 is a protein of 139 kDa with 52% sequence identity to human SERCA1a and 51% sequence identity to rabbit SERCA1a, although it has different functional properties compared to its mammalian orthologue: PfATP6 presents a lower affinity for thapsigargin and 2,5-di(tert-butyl)-1,4-benzo-hydroquinone (BHQ) and a much stronger affinity for cyclopiazonic acid (CPA), which comprise three classical SERCA inhibitors . PfATP6 has previously been proposed as a target of the antimalarial class of artemisinin on the basis of the inhibition of PfATP6 Ca2+-ATPase activity in Xenopus oocytes membranes heterologously expressing PfATP6 . However, Cardi et al.  demonstrated that PfATP6, heterologously expressed and purified from yeast membranes, was insensitive to artemisinin .
Nevertheless, a number of early studies reported that specific mutations of the PfATP6 coding sequence (L263E and S769N) might confer resistance to artemisinin and its derivatives [21, 22], although subsequent work has provided no evidence of any causal relationship . Chavchich et al.  and Cui et al.  have also demonstrated that there is no evidence supporting the involvement of PfATP6 in artemisinin responsiveness. Tanabe et al.  demonstrated a natural and random occurrence of numerous single nucleotide polymorphisms on PfATP6, showing that this gene is naturally very polymorphic . Furthermore, the PfATP6 mutations were located in poorly conserved regions of the protein, consistent with simple genetic drift . These findings indicated that the mutations found on PfATP6 of P. falciparum strains presenting a diminished sensitivity to artemisinin may not correspond to a selective pressure of this drug on PfATP6.
Finally, recent genome-wide association studies have examined large numbers of field isolates in areas of emerging artemisinin resistance; these provide very powerful data that are highly relevant for testing whether PfATP6 could be involved in artemisinin action and resistance. Cheeseman et al.  specifically examined the PfATP6 locus in the context of emerging resistance in Western Thailand, whereas Takala-Harrison et al.  examined resistance in Western Cambodia. Both studies found no evidence for selection at the PfATP6 locus.
In summary, we consider that PfATP6 is susceptible to becoming a drug target as a result of its crucial role in cell biology, and not because it is the enzyme targeted by artemisinin. To identify new molecules that inhibit PfATP6, in the present study, we report a collaborative work that combines expertise in the heterologous expression in yeast and purification of PfATP6; ATPase activity measurements on purified PfATP6; drug screening from libraries of molecules; and, finally, testing of some of the best candidate molecules on the erythrocytic stage of P. falciparum cultures.
Results and Discussion
In vitro enzymatic inhibition and antiplasmodial activity of classical SERCA inhibitors
In previous studies, we expressed PfATP6 using a heterologous expression system in yeast. The protein was coupled to a biotin acceptor domain (BAD), which allows an in vivo biotinylation in yeast. The membrane fractionation gave rise to a light membrane fraction called P3, which contains PfATP6. The fraction was solubilized using the detergent N-dodecyl-d-maltoside (DDM). The protein was purified using a streptavidin resin, which is highly specific for biotinylated proteins. PfATP6 purified by this way is functional in terms of ATPase activity . This enzymatic activity was measured spectrophotometrically using a coupled enzyme assay [30, 31]. The effects of various mammalian SERCA1a inhibitors were investigated on PfATP6 to compare their effects on both proteins. We found that PfATP6 is less sensitive to thapsigargin and BHQ than rabbit SERCA1a, whereas CPA is a very powerful inhibitor of PfATP6 activity. This suggests significant and important differences between P. falciparum and mammalian proteins. It was possible to monitor the structural changes induced by drug binding on PfATP6 using tryptophan fluorescence measurement techniques [18, 20]. These results are summarized as the half maximal inhibitory concentration (IC50) PfATP6 in Table 1.
Table 1. Effect of SERCA classical inhibitors upon P. falciparum growth in vitro. IC50 PfATP6: values of the inhibition of the ATPase activity determined on the purified PfATP6 . Indication of the IC50 values for each drug on two P. falciparum laboratory strains differently resistant to chloroquine (3D7 and FcB1). Values (μm) correspond to at least three biological replicates. ND, not determined
6.1 ± 1.05
11.5 ± 5.7
17.2 ± 11.3
30.6 ± 13.5
4.9 ± 0.9
8.8 ± 3.4
0.13 ± 0.02
0.012 ± 0.002
The first objective of the present study was to test the effect of these classical SERCA1a inhibitors on P. falciparum growth in vitro. We wanted to validate this procedure and test whether a correlation exists between the enzymatic inhibition of PfATP6 and the antiplasmodial activity results. Accordingly, we checked these compounds for their ability to inhibit the growth of two different P. falciparum strains (differentially sensitive to chloroquine: 3D7 and FcB1), which were maintained in culture as previously described by Trager and Jensen . Drugs were assayed by using the semi-automated microdilution technique of Desjardins et al. , which consists of growing the parasites in 96-well plates, in the presence of serially diluted concentrations of tested drugs, for 48 h. One such experiment is shown for CPA in Fig. 1A. All three molecules inhibited parasite growth in erythrocytes, with IC50 values in the micromolar range, and with an approximately two-fold difference between the two strains, where 3D7 is less sensitive than FcB1.
As shown in Table 1, CPA has a more pronounced antiplasmodial activity than BHQ on both strains. This is also true when we compare the inhibitory effect of CPA on purified PfATP6 with that of BHQ [18, 20]. The results demonstrate some correlation between the antiplasmodial activity of these two compounds and their enzymatic inhibition of PfATP6, whereas this is not the case for thapsigargin, which appears to be as toxic to parasites as CPA, with this being the case for both strains (Table 1). These three classical SERCA inhibitors may not comprise good antimalarial drugs because they are also powerful inhibitors of the mammalian enzyme [18, 34] and, consequently, they are expected to be toxic to mammalian cells.
Using the SERCA1a crystallographic structure as a template, Arnou et al.  analyzed the interactions of these three classical SERCA inhibitors with the homologue residues on PfATP6 based on a model of its structure. Because thapsigargin, CPA and BHQ have different binding sites in the rabbit SERCA1a [35, 36], it is expected that the binding sites of these inhibitors also involve different residues of PfATP6 . Based on this assumption, we aimed to search for new inhibitors for PfATP6 among several categories of molecules. In this way, we hoped to increase the probability of finding a PfATP6 inhibitor with good antiplasmodial activity.
Improved expression and purification of PfATP6 for large-scale inhibitor screening
We had to adapt our previously established over-expression and purification procedures to prepare larger amounts of purified, active and concentrated PfATP6. Both expression and purification have been described previously , with modifications as indicated in the Experimental procedures. Purification is followed either by immunodetection with anti-PfATP6 serum (Fig. 2A) or by Coomassie Blue staining (Fig. 2B). Using this procedure, detergent-solubilized PfATP6 is fixed on the streptavidin-Sepharose resin as a biotinylated PfATP6-BAD protein (Fig. 2A, lane 1). After cleavage by thrombin for 30 min (Fig. 2A, lane 2) and 60 min (Fig. 2A, lane 3), PfATP6 looses its BAD tag. It is eluted from the resin (Fig. 2A, lane 4) with a small amount of PfATP6-BAD and PfATP6 still retained on the resin (Fig. 2A, lane 5). The concentration of the PfATP6 pool on Centriprep was followed (Fig. 2A, compare lanes 4 and 6). The pattern is similar in both lanes (Fig. 2A, lanes 4 and 6) and there is no significant loss of protein during the concentration step (Fig. 2A, lane 7).
To quantify the concentrated purified PfATP6, it was compared with known amounts of sarcoplasmic reticulum Ca2+-ATPase on a Coomassie Blue stained gel (Fig. 2B). The protein purified by this procedure is rather pure (Fig. 2B, lane 4) and its concentration is estimated to be 1 mg·mL−1. This value must be determined to achieve a specific activity of PfATP6. In total, we obtained 1.4 mg of concentrated purified PfATP6 from 6 L of yeast culture and 1.3 g of P3 light membranes.
To verify the enzymatic activity of the purified and concentrated PfATP6 protein, we performed ATPase activity measurements using a classical coupled-enzyme reaction [30, 31] (Fig. 3). As described previously , ATPase activity of PfATP6 was stabilized by the addition of lipids. Thus, at a octaethylene glycol mono-n-dodecyl ether (C12E8):1,2-dioleoyl-sn-glycero-3-phosphocholine (DOPC) ratio of 0.2 : 0.05 (mg·mL−1), the specific activity at 25 °C of the purified PfATP6 was 0.5 μmol of hydrolyzed ATP·min−1·mg of PfATP6−1.
In vitro enzymatic inhibition and antiplasmodial activity of potential PfATP6 inhibitors
A test measuring the inorganic phosphate liberation was established to screen several molecules and choose the most promising inhibitors. The screening was undertaken in duplicates on a small library of molecules and the IC50 for the inhibition of PfATP6 ATPase activity was determined for 1680 compounds. Twenty of them exhibited a potent inhibitory effect with an IC50 of < 10 μm. From these twenty compounds, eight were chosen to test their antiplasmodial activity based on their potent inhibitory effect on PfATP6, as well as their commercial availability. We selected compounds 1, 2, 8, 10, 17, 18, 19 and 20 for this study. The IC50 values are listed in Table 2 and the chemical structures are shown in Fig. 4. Their IC50 values on purified PfATP6 range between 1 and 9 μm (Table 2) and the best values are for compounds 2, 10, 17 and 18. Thus, these eight molecules comprised a starting point for further studies and were tested for their antiplasmodial activity (Table 2). We observed first that the compounds have a similar inhibitory effect on the growth of either FcB1 or 3D7 strains of P. falciparum (in the range 2–250 μm). The antiplasmodial activity of compound 18 is shown in Fig. 1B as an example of the results obtained for the two strains.
Table 2. IC50 of the compounds identified to be potent inhibitors of PfATP6 and chosen to be tested on P. falciparum cultures. IC50 PfATP6: values of the inhibition of the ATPase activity determined on the purified PfATP6; IC50 FcB1/IC50 3D7: effect of the identified PfATP6 inhibitors on two P. falciparum strains differently resistant to chloroquine (3D7 and FcB1). Values (μm) correspond to at least three biological replicates
3.0 ± 2.2
71 ± 25
108 ± 8
1.1 ± 0.5
57 ± 15.8
53 ± 19
3.1 ± 1.0
88 ± 33
91.5 ± 16
1.2 ± 0.6
169 ± 17
241 ± 19
1.0 ± 0.6
6.2 ± 1.7
4.9 ± 2.1
1.0 ± 0.6
2.3 ± 1.2
3.1 ± 1.2
8.5 ± 1.6
14.3 ± 3.6
14.8 ± 7.7
7.0 ± 5.4
61 ± 26
62 ± 26
Molecules 17– 20 present a range of action on P. falciparum growth in vitro; from 2 μm up to ~ 60 μm, with the compound 18 (NSC 95397) presenting the highest antiplasmodial activity (2–3 μm).
Compound 17, also known as Ebselen, presents an IC50 for inhibition of purified PfATP6 ATPase activity of ~ 1 μm (Table 2). Hüther et al.  reported an antiplasmodial activity of ~ 14 μm . Our data lie in the same order of magnitude (~ 4–6 μm) and the difference may be readily explained by the utilization of different P. falciparum strains (chloroquine sensitive strain T9-96 and chloroquine resistant strain T9-102) . Ebselen has been extensively studied as an anti-inflammatory drug and has been described to act upon several mammalian enzymes .
Compound 18, also known as NSC 95397, appears to be toxic, at least for erythrocytes, because we observed some cell lysis after 48 h of exposure to this molecule at the highest doses tested (> 30 μm) (data not shown). However, it may represent a PfATP6 inhibitor worthy of further exploration, and it is possible that structural modifications will bring a good compromise between toxicity and antiplasmodial activity.
Compound 20, also known as Tyrphostin 47, was previously tested on trypanosome in vitro cultures and it was found that nontoxic doses of 1 μm significantly reduced the in vitro growth of parasites [39, 40], which is a lower concentration than that found when Tyrphostin 47 was tested on P. falciparum in the present study (Table 2).
There is no strict correlation between the inhibition of PfATP6 activity and the antiplasmodial activity of these compounds (Table 2). Compounds can inhibit in vitro PfATP6 activity and present limited antiplasmodial activity. For example, compound 10 presents a significant inhibition of PfATP6 activity but has a low antiplasmodial effect. One explanation may be that P. falciparum possesses mechanisms to extrude drugs and metabolize them, or it may be impermeable to these compounds . Inversely, compounds can inhibit in vitro PfATP6 activity and present an antiplasmodial activity for unrelated reasons, such as overall toxicity. To investigate the mammalian cytotoxicity of our compounds, we first determined whether they were inefficient inhibitors of rabbit SERCA1a, which would be a potential advantage in terms of human toxicity. Compound 17 was the only one to significantly inhibit rabbit SERCA1a ATPase activity (IC50 of 0.5 μm), whereas the other compounds had little or no effect (50% of inhibition of SERCA1a activity was not reached at a maximum concentration of 20 μm of compound; data not shown). For further development, it will be essential to also exclude cytotoxic effects in mammalian cells.
In summary, Table 2 provides a good starting point for investigating related molecules. Our future studies aim to use these compounds, as well as screen for new compounds issued from our chemical library. The identified inhibitors of PfATP6 display little apparent structural resemblance to already known inhibitors of the SERCA1a Ca2+-ATPase. The binding pocket(s) in PfATP6 for the identified inhibitors are presently unknown and further studies will be needed to identify any possible resemblance to SERCA1a with respect to hit/target interactions. As knowledge of the structure of PfATP6 could ameliorate and accelerate this process, efforts are also being undertaken in this respect.
There is no optimal solution with respect to the search for new treatments for malaria. Worldwide, researchers have focused their energies in several directions aiming to cover the largest number of possible strategies for discovering novel ways to cure/eradicate malaria. The most commonly employed strategies are: (a) the screening of molecules issued from chemical libraries either in a phenotypic way (i.e. testing compounds upon in vitro cultures of Plasmodium; the whole cell approach)  or (b) searching for inhibitory molecules acting upon the biochemical activity of a potential drug target that comprises an essential enzyme or pathway, ideally specific to the parasite (i.e. a target-based approach) .
In the present study, we aimed to establish a collaborative work in a target-based approach, where we could benefit from the knowledge and facilities available to several teams to: (a) screen for molecules that would inhibit PfATP6 activity; (b) select the best inhibitors to subsequently test on P. falciparum growth in vitro; (c) test the toxicity on mammal cells of the best molecules that issued from these tests; (d) and, finally, test the efficiency in vivo (on a murine model for malaria) of the best inhibitors. So far, the first two objectives have been performed successfully. We are now able to screen for new PfATP6 inhibitors in a large-scale manner, reducing time-consuming issues. This is highly advantageous because we can rapidly eliminate less interesting molecules and focus our efforts on more promising compounds for investigation using a well established in vitro antiplasmodial test .
One long-term goal is to obtain a very efficient inhibitor against PfATP6, which at the same time is not very efficient for inhibiting mammalian SERCA1a. The most promising compounds should not be toxic to humans, which is an overall stringent requirement. Nevertheless, if they are found to be toxic against human cell lines, it is always possible to consider the pro-drug approach, which has already successfully applied with thapsigargin and cancer cells .
All chemical products were purchased from Sigma (St Louis, MO, USA) unless specified otherwise. High-activity bovine thrombin was obtained from Calbiochem (San Diego, CA, USA), and the streptavidin-Sepharose High Performance resin was purchased from GE Healthcare (Milwaukee, WI, USA). All products for yeast and bacteria cultures were purchased from Difco (BD Biosciences, Franklin Lakes, NJ, USA). DOPC was obtained from Avanti Polar Lipids (Alabaster, AL, USA). DDM was obtained from Anatrace (Maumee, OH, USA) and C12E8 was purchased from Nikkol Chemical (Tokyo, Japan). Precision protein standards were from Bio-Rad (Hercules, CA, USA). Immobilon-P membranes were obtained from Millipore (Bedford, MA, USA). Phosphoenol pyruvate (catalogue number P3637), l-lactic dehydrogenase solution from bovine heart (catalogue number L1006) and pyruvate kinase preparation type VII from rabbit muscle (catalogue number P7768) were obtained from Sigma. Anti-PfATP6 serum were purchased from Bethyl Laboratories (Montgomery, TX, USA).
Yeast transformation and selection of individual clones
The Saccharomyces cerevisiae yeast strain W303.1b/Gal4 (a, leu2, his3, trp1::TRP1-GAL10-GAL4, ura3, ade2-1, canr, cir+) was the same as described previously . Transformation was performed according to the lithium acetate/single-stranded carrier DNA/polyethylene glycol method . Growth conditions and criteria for expression of the Ca2+-ATPase were carried out as described for the test of individual clones and for the expression on minimal medium [44, 46].
Growth of yeast cells and large-scale expression of PfATP6 using a fermentor (Techfors-S Apparatus; INFORS HT, Massy, France)
For membrane fractionation, the procedure was performed as described previously , with some modifications. The major change consisted of using a ‘pulverisette’ for the breaking of yeasts (see below). The yeast suspensions were rapidly thawed in water bath at 20 °C, and one additional volume of Tes buffer [50 mm Tris-HCl (pH 7.5), 1 mm EDTA, 0.6 m sorbitol] with 1 mm phenylmethanesulfonyl fluoride (PMSF) and complete EDTA-free antiprotease cocktail was added before the breaking of the yeast. Cells (200 mL containing 200 g of yeast) were loaded in a grinding bowl agate (500 mL) and one equivalent volume (200 mL) of glass beads (diameter 0.5 mm) was added. Cells were broken using a planetary mill ‘pulverisette 6′ (Fritsch, Idar-Oberstein, Germany): 3 min at 450 r.p.m., 30-s pause and 3 min at 450 r.p.m. in reverse. The extract was transferred in a cristallisoire and beads were washed three times with 0.5 equivalent volumes (three times in 100 mL) of Tes buffer supplemented with 1 mm PMSF and complete EDTA-free antiprotease cocktail. The resultant crude extract must have a pH in the range 7.0–7.5. The crude extract was centrifuged at 1000 g for 20 min at 10 °C. The first supernatant S1 was centrifuged at 12 000 g for 20 min at 10 °C. The second supernatant was removed with great care and was centrifuged at 125 000 g for 1 h at 4 °C. The pellet P3, containing light membranes, was resuspended in Hepes-sucrose buffer [20 mm Hepes-Tris (pH 7.5), 0.3 m sucrose, 0.1 m CaCl2 supplemented with 1 mm PMSF and complete EDTA-free antiprotease cocktail] (i.e. 0.2 mL·g−1 yeast). The membranes were stored at −70 °C until use. The amount of the protein of interest was estimated by western blotting using the appropriate antibody.
Solubilization and batch purification of PfATP6 by streptavidin-Sepharose chromatography
Several of these steps have been modified from previous protocols, such as the washing of the light membrane fraction, the composition of some buffers, the mode of binding of the solubilized solution to the resin and the mode of elution and concentration of the purified proteins. Thus, the light membrane fraction, suspended in the Hepes-sucrose buffer, was washed once to remove contaminant and soluble biotinylated proteins (acetyl-CoA carboxylase, pyruvate carboxylase and Arc1p) by diluting the membranes at 6 mg·mL−1 in a buffer containing 50 mm Mops-Tris (pH 7), 0.5 m KCl, 20% glycerol, 1 mm CaCl2, 1 mm β-mercaptoethanol, 1 mm PMSF and complete EDTA-free antiprotease cocktail. Light membranes were then pelleted at 125 000 g for 60 min at 10 °C and the supernatant containing soluble proteins was discarded. The pellet was resuspended at a protein concentration of 12 mg·mL−1 in the solubilization buffer (same buffer with 0.1 m KCl instead of 0.5 m KCl) with a Potter homogenizer. The same volume of solubilization buffer containing 36 mg·mL−1 of DDM was prepared with stirring at 4 °C. Both solutions were mixed and were left for 30 min with stirring at 4 °C, and the solubilizate was clarified by 10–20 up and down with a Potter homogenizer. The clarified solution was left for 30 min with stirring at 4 °C and the nonsolubilized material was pelleted by centrifugation at 125 000 g for 1 h at 4 °C. All subsequent steps (unless otherwise specified) were then performed at 4 °C. The supernatant after the centrifugation step was mixed with streptavidin Sepharose High Performance resin at a ratio of 4 : 1 (w/v) (previously kept for 15 min at room temperature), using typically 4 mg of PfATP6 per millilitre of resin, and stirred gently overnight at 4 °C. The suspension was then pelleted into 50-mL tubes for 5 min at 500 g and washed, with ten resin volumes of a ‘high-salt’ buffer [50 mm Mops-Tris (pH 7), 1 m KCl, 20% glycerol, 1 mm CaCl2, 0.05% DDM) [buffer : resin, 10 : 1 (v/v)], and then with ten resin volumes of a ‘low-salt’ buffer [50 mm Mops-Tris (pH 7), 100 mm KCl, 20% glycerol, 1 mm CaCl2, 0.05% DDM] [buffer : resin, 10 : 1 (v/v)]. The resin was resuspended in the ‘low-salt’ buffer [buffer : resin, 1 : 1 (v/v)] and thrombin was added (10 U of thrombin per millilitre of resin) and the mixture was placed on a wheel and gently stirred at room temperature for 30 min, followed by a second addition of thrombin and stirring for another 30 min. To inactivate thrombin, 1 mm PMSF was then added, and the solution of resin was transferred into Handee Centrifuge columns (Perbio Science France SAS, Brebieres, France). The proteolytically cleaved PfATP6 proteins were eluted. A second elution was performed and the eluted fractions containing the Ca2+-ATPase were pooled and concentrated on Centriprep YM30 (Millipore) by successive centrifugations at 1500 g for 30 min at 4 °C until a volume of 1–2 mL was obtained. The glycerol concentration was increased to 40% before freezing the samples in liquid nitrogen and storage at −70 °C.
Protein estimation, Ca2+-ATPase quantification and rabbit SERCA1a preparation
Protein concentrations were measured by the bicinchoninic acid procedure  in the presence of 2% SDS (w/v) with BSA as standard. SERCA1a from rabbit muscle (sarcoplasmic reticulum), used as a standard for protein estimation, was prepared as described previously [49, 50]. Ca2+-ATPase quantification was performed on a Coomassie Blue stained gel after SDS/PAGE.
SDS/PAGE and western blotting
For SDS/PAGE, samples were mixed with an equal volume of denaturing buffer and loaded onto Laemmli-type 8% (w/v) polyacrylamide gels . The amounts of proteins or volumes of initial samples loaded in each well are indicated where appropriate. After separation by SDS/PAGE, gels were stained with Coomassie Blue, or proteins were electroblotted onto poly(vinylidene difluoride) Immobilon P membrane . For each gel, molecular mass markers (Precision Protein Standards; Bio-Rad) were loaded. Western blotting was followed by immunodetection with polyclonal anti-PfATP6 serum (dilution 1 : 20 000) generated in goat .
ATPase activity measurement using coupled enzymes system
ATPase activity was assayed at 25 °C using a spectrophotometric method as described previously [30, 31]. In total, 10 μg of proteins was used in 2 mL of reaction buffer [50 mm TES/Tris (pH 7.5), 0.1 m KCl, 1 mm MgCl2, 0.3 mm NADH, 1 mm phosphoenolpyruvate, 0.1 mg·mL−1 lactate dehydrogenase, 0.1 mg·mL−1 pyruvate kinase containing 0.06 mm Ca2+ and 0.2 : 0.05 mg·mL−1 C12E8 : DOPC]. The reaction was started by the addition of PfATP6 to the medium and stopped by the addition (twice) of a final concentration of 600 μm EGTA. The difference between the slopes obtained before and after the addition of EGTA is considered to be a result of Ca2+-ATPase activity. Figure 3 shows that there is some residual ATPase activity after the addition of EGTA, which may be a result of PfATP6 not being completely pure and/or a part of PfATP6 possessing a non-Ca2+-dependent activity.
Protein activity was determined by measuring the amount of liberated phosphate (Pi) from the ATP hydrolysis reaction. Some 1 μg of protein per assay and various concentrations of exogenous added compounds were incubated for 10 min in a buffer consisting of 0.1 mm CaCl2, 6 mm MgCl2, 100 mm KCl, 50 mm Tris (pH 7.5), 20% glycerol, 0.2 mg·mL−1 C12E8 and 0.05 mg·mL−1 DOPC. The reaction was initiated by the addition of Na-ATP to a final concentration of 2.5 mm, followed by incubation for 30 min at 30 °C. The amount of Pi was measured after the addition of STOP-solution [mixture of A: 170.3 μm C6H8O6 in 0.5 m HCl; B: 28.3 mm (NH4)6Mo7O24·4H2O in Milli-Q H2O; Millipore] with incubation for 10 min at room temperature followed by the addition of arsenite solution (154 mm NaAsO2, 68 mm Na3C6H5O7∙2H2O, 0.3 m CH3COOH) and incubation for 30 min. D860 was measured. A small library of 1680 compounds was screened at a concentration of 16 μm. IC50 values were determined for compounds found to have more than 75% inhibition in the initial screening. Some of these compounds were selected for in vitro antiplasmodial testing.
In vitro antiplasmodial tests
The P. falciparum strains FcB1 and 3D7 were maintained continuously in culture on human erythrocytes as initially described by Trager and Jensen . Parasite culture medium contained RPMI 1640 medium (Life Technologies, Inc., Grand Island, NY, USA), 25 mm Hepes, 27.5 mm NaHCO3 and 11 mm glucose (pH 7.4), and was supplemented with 7.5% (v/v) compatible heat-inactivated human serum. Human red blood cells were added at a haematocrit of 2% and the parasite cultures were maintained at 37 °C under an atmosphere of 3% CO2, 6% O2 and 91% N2, with daily medium changes. In vitro antiplasmodial activities were determined using a modification of the semi-automated microdilution technique of Desjardins et al. . Stock solutions of chloroquine diphosphate and tested compounds were prepared in sterile distilled water and dimethylsulfoxide, respectively. Drug solutions were serially diluted with culture medium and added to asynchronous parasite cultures (1% parasitaemia and 1% final haematocrit) on 96-well plates, which were incubated at 37 °C for 24 h in a candle jar system, before the addition of 0.5 μCi of [3H]hypoxanthin (1–5 Ci·mmol−1) per well for an additional 24 h. Parasites were then harvested on filters after a freeze-thawing cycle, and dried filters were submerged in a liquid scintillation mixture (OptiScintHisafe; Perkin Elmer, Boston, MA, USA) and counted in a 1450 Microbeta counter (Wallac; Perkin Elmer). Parasite growth inhibition was determined by comparison of the radioactivity incorporated into the treated wells with that of control wells (containing parasite cultures without drug) from the same plates. The concentrations of drugs that inhibited growth by 50% (IC50) were determined graphically from drug concentration–response curves. Final IC50 values for each compound were expressed as the mean ± SD of values determined from independent experiments (biological replicates) .
Chloroquine diphosphate (Sigma-Aldrich Chimie SARL, St Quenti Fallavier, France) was used to determine the level of resistance of the FcB1 and the 3D7 strains in the culture and test conditions, and as positive controls of antiplasmodial activities.
The authors would like to thank Dr Guillaume Lenoir for very helpful discussions and critical comments on the manuscript, as well as Dr Cedric Montigny for priceless advice and help with the protein expression experiments. This work was supported by a grant from Domaine d'Interet Majeur Maladies Infectieuses region Ile de France (DIM Malinf) (to C.J., I. F. and M.L.M.), a grant from the Agence Nationale pour la Recherche (to C.J., P.M. and M.L.M.) and by a fellowship from the ‘Ministère de l'Enseignement Supérieur et de la Recherche’ (to S.D.B.). A.L.W. was supported by a postdoctoral fellowship from the Danish National Advanced Technology Foundation.