Blockage of Spontaneous Ca2+ Oscillation Causes Cell Death in Intraerythrocitic Plasmodium falciparum

Malaria remains one of the world’s most important infectious diseases and is responsible for enormous mortality and morbidity. Resistance to antimalarial drugs is a challenging problem in malaria control. Clinical malaria is associated with the proliferation and development of Plasmodium parasites in human erythrocytes. Especially, the development into the mature forms (trophozoite and schizont) of Plasmodium falciparum (P. falciparum) causes severe malaria symptoms due to a distinctive property, sequestration which is not shared by any other human malaria. Ca2+ is well known to be a highly versatile intracellular messenger that regulates many different cellular processes. Cytosolic Ca2+ increases evoked by extracellular stimuli are often observed in the form of oscillating Ca2+ spikes (Ca2+ oscillation) in eukaryotic cells. However, in lower eukaryotic and plant cells the physiological roles and the molecular mechanisms of Ca2+ oscillation are poorly understood. Here, we showed the observation of the inositol 1,4,5-trisphospate (IP3)-dependent spontaneous Ca2+ oscillation in P. falciparum without any exogenous extracellular stimulation by using live cell fluorescence Ca2+ imaging. Intraerythrocytic P. falciparum exhibited stage-specific Ca2+ oscillations in ring form and trophozoite stages which were blocked by IP3 receptor inhibitor, 2-aminoethyl diphenylborinate (2-APB). Analyses of parasitaemia and parasite size and electron micrograph of 2-APB-treated P. falciparum revealed that 2-APB severely obstructed the intraerythrocytic maturation, resulting in cell death of the parasites. Furthermore, we confirmed the similar lethal effect of 2-APB on the chloroquine-resistant strain of P. falciparum. To our best knowledge, we for the first time showed the existence of the spontaneous Ca2+ oscillation in Plasmodium species and clearly demonstrated that IP3-dependent spontaneous Ca2+ oscillation in P. falciparum is critical for the development of the blood stage of the parasites. Our results provide a novel concept that IP3/Ca2+ signaling pathway in the intraerythrocytic malaria parasites is a promising target for antimalarial drug development.


Introduction
Malaria continues to be a worldwide public health problem causing significant morbidity and mortality and its resistance to existing antimalarial drugs is a growing problem [1]. The life cycle of Plasmodium species is complex (Fig. 1). Infection of humans begins with a small inoculum of sporozoites from the salivary glands of a blood-feeding Anopheles mosquito. Sporozoites penetrate liver cells, transform and multiply asexually to produce thousands of free merozoites (liver stage). Each of these asexual merozoites invades an erythrocyte and enters into another phase of asexual reproduction, and then bursts the cell, releasing 8-32 more merozoites to invade more erythrocytes (blood stage). In infected erythrocytes, development of the parasites is accompanied by morphological changes such as ring form, trophozoite and schizont stages. P falciparum is responsible for the lethal form of human malaria. The mature forms of the intraerythrocytic parasite (trophozoite and schizont) remodel the cytoskeleton and plasma membrane to create cytoadherence knobs as well as nutrient permeation pathways and alter the mechanical stability of the erythrocytes, causing them to stick to blood vessels [2,3]. This leads to blockage of the microcirculation and results in dysfunction of multiple organs, typically the brain in cerebral malaria [4].
In this study, we observed the intracellular dynamics of Ca 2+ throughout the intraerythrocytic stages of the FCR-3 strain of P. falciparum and found that stage-specific spontaneous Ca 2+ oscillations which can be blocked by the inositol 1,4,5-trisphosphate (IP 3 ) receptor inhibitor 2-aminoethyl diphenylborinate (2-APB) occur in the ring form and trophozoite. Examination of the effects of 2-APB on the in vitro intraerythrocytic parasite development and electron microscopic observations revealed that blockage of Ca 2+ oscillations caused severe degeneration and breakdown of successive asexual reproduction in the intraerythrocytic parasites, resulting in death of them. Furthermore, 2-APB showed a similar effect against the chloroquine-resistant K1 strain of P. falciparum.

Results and Discussion
IP 3 -induced Ca 2+ Oscillation in P. falciparum Figure 2 and S1 shows fluorescence Ca 2+ images of each intraerythrocytic developmental stage of the FCR-3 strain: early ring forms (ERf, parasites with smaller cell size than trophozoite without malaria pigment), late ring forms (LRf, parasites with cell size between early ring form and trophozoite without malaria pigment), early trophozoites (ET, parasites with a single nucleolus, malaria pigment and immature food vacuole), late trophozoites (LT, parasites with a single nucleolus and mature food vacuole), schizonts (S, parasites with multiple nuclei) and merozoites (M). Ca 2+ imaging of parasites was performed in culture chambers at 37uC in an atmosphere of 5% O 2 and 5% CO 2 , conditions identical to those in conventional in vitro parasite culture. The Fluo-4 fluorescence in a parasite cytoplasm (F) was calculated by subtraction of the background fluorescence and normalized to the minimum fluorescence during the imaging period (F min ). In early ring forms (ERf) and early trophozoites (ET), spontaneous Ca 2+ oscillations were observed ( Fig. 2A and B, left). Dimethyl sulfoxide (DMSO) was used as a solvent control. The frequency of Ca 2+ oscillations was higher in early ring forms than that in early trophozoites. The subcellular distribution of Fluo-4 in the early trophozoites indicates that free Ca 2+ were evenly distributed in the cytoplasm (Fig. 2B), whereas in the late trophozoites with mature food vacuole, Ca 2+ gradient between the digestive food vacuole and cytoplasm, similar to that previously reported [19,20] was observed being independent of the addition of 2-APB (Fig. S2). 2-APB was a well-established inhibitor of IP 3 receptor/Ca 2+ channels developed in our previous study [21,22] and the blockage of melatonin-induced Ca 2+ release by 2-APB in P. falciparum has been demonstrated [23]. Treatment with 100 mM 2-APB almost completely blocked Ca 2+ oscillations ( Fig. 2A and B, right). On the other hand, in the late ring forms (LRf), late trophozoites (LT), schizonts (S) and merozoites (M), small periodic Ca 2+ fluctuations were observed, and notable effects of 2-APB were not detected (Fig. S1). To investigate the effects of 2-APB in detail, we performed quantitative analysis of the effect of 2-APB on the amplitude of periodic Ca 2+ fluctuations in late ring form, late trophozoite, schizont and merozoite stages. The mean amplitude was calculated by subtracting the mean minimal value of F/F min from its mean maximal value. A statistically significant effect of 2-APB was observed in merozoites (*, P = 0.0116, two-tailed unpaired t test) (Fig. S3). This result supports those of previous studies demonstrating that an increase in cytosolic Ca 2+ concentration is involved in the regulation of merozoite invasion of the erythrocytes [16][17][18]. In apicomplexan parasites, Ca 2+ transients and oscillations have been reported to be evoked by several inducers [10,13,24,25]; however, our study is the first to demonstrate spontaneous Ca 2+ oscillations, i.e. without the addition of exogenous inducers.
Blockage of spontaneous Ca 2+ oscillations by 2-APB in early ring forms and early trophozoites strongly suggests that the observed Ca 2+ oscillations are regulated by a putative IP 3 receptor/Ca 2+ channel that is activated by IP 3 binding. In Plasmodium the phospholipase C (PLC) pathway is known to be involved in the release of Ca 2+ from the intracellular Ca 2+ store, and the effectiveness of U73122, a commonly used PLC inhibitor, has been demonstrated [25]. PLC catalyses the hydrolysis of 1phosphatidyl-D-myo-inositol-3,4,5-trisphosphate to the second messenger molecules diacylglycerol and IP 3 . As shown in Fig.  S4A, after 5 min of pre-treatment with 10mM U73122, Ca 2+ oscillations in early ring forms (ERf, left) and early trophozoites (ET, right) were almost completely inhibited. In apicomplexan parasites, two intracellular Ca 2+ stores are known to be involved in Ca 2+ release: the endoplasmic reticulum (ER) and acidocalcisomes including the food vacuole [26,27]. To investigate the source of the observed Ca 2+ oscillations, we depleted Ca 2+ stores before imaging by using thapsigargin (Tg), a specific inhibitor of sarco/endoplasmic reticulum Ca 2+ -ATPase, and concanamycin A (CMA), a specific inhibitor of vacuolar-type H + -ATPase. The effects of these inhibitors have been demonstrated in Plasmodium species [28][29][30]. We also confirmed the effectiveness of these compounds in depleting Ca 2+ by Ca 2+ imaging in early ring forms and early trophozoites using perfusion experiments ( Fig. S4B and C). As shown in Fig. S4D, Ca 2+ depletion by pre-treatment with 2 mM Tg for 30 min significantly reduced Ca 2+ oscillations in early ring forms (ERf, left) and early trophozoites (ET, right). In contrast, Ca 2+ depletion by pre-treatment with 100 nM CMA had no effect on Ca 2+ oscillations in early ring forms (ERf, left) and early trophozoites (ET, right) (Fig. S4E). In these two stages the digestive food vacuole, which is known to be a Ca 2+ pool sensitive to both thapsigargin and a vacuolar-type H + -ATPase inhibitor, bafilomycin A 1 [19], is not formed. Taken together, these results indicate that spontaneous IP 3 -induced Ca 2+ release from a thapsigarginsensitive Ca 2+ store, ER occurred in early ring forms and early trophozoites during intraerythrocytic P. falciparum development. Next, we investigated the effects of 2-APB on the intraerythrocytic P. falciparum development to understand the physiological roles of these spontaneous IP 3 -induced Ca 2+ oscillations.

2-APB Blocks P. falciparum Development
The effects of 2-APB on intraerythrocytic development were investigated using synchronized parasite cultures in the ring form stage, with initial parasitaemia of approximately 1%. Figure 3A shows parasitaemia of each developmental stage at 20, 30 and 40 h of the assay. In the presence of 100 mM 2-APB intraerythrocytic development of the parasites was delayed compared to that in the presence of dimethyl sulphoxide (DMSO). Figure 3B shows the morphology of intraerythrocytic parasites. Parasites cultured with DMSO developed into early schizonts (parasites with fewer than 8 nuclei) at 20 h of the assay (Fig. 3B, panel 1). These schizonts developed into healthy late schizonts (parasites with at least 8 nuclei) and produced ring forms in the next developmental cycle at 30 and 40 h of the assay (Fig. 3B, panels 2 and 3). In contrast, parasites cultured with 2-APB remained at the trophozoite stage (parasites with a single nucleus) with abnormal morphology at 20 h of the assay (Fig. 3B, panel 4). These abnormal trophozoites could develop into early schizonts and late schizonts but exhibited abnormal morphology at 30 and 40 h of the assay (  At 70 h parasites cultured with 2-APB developed into a few late schizonts or ring forms with abnormal morphology (1 in 5000-8000 erythrocytes), although those cultured with DMSO developed normally into late schizonts or ring forms in the next developmental stage. On the other hand, parasites could develop normally in 2-APB-pretreated erythrocytes similar to that in cultures with DMSO-pre-treated erythrocytes (Fig. S5). This indicates that the effect of 2-APB is not due to the disruption of erythrocyte physiology, which is significant to intraerythrocytic parasite development and invasion. From these results we conclude that 2-APB directly inhibits intraerythrocytic parasite development by blocking its normal cell cycle, resulting in a failure to maintain the successive developmental stages of asexual blood forms.

2-APB Acts on a Chloroquine-resistant Strain
Next, we examined the effect of 2-APB on the intraerythrocytic development of the chloroquine-resistant strain K1 of P. falciparum using synchronized parasite cultures in the ring form stage with initial parasitaemia of approximately 2%. At 24 h of the assay a tendency towards decreasing trophozoite parasitaemia was reproducibly observed in parasites cultured with 100 mM 2-APB ( Fig. 3E, 24 h). The slight inhibitory effect of 2-APB at 24 h of the assay was confirmed by measuring the area, perimeter and maximum diameter of the parasites (Fig. S6). At 48 h of the assay, intraerythrocytic parasite development in the presence 2-APB was delayed compared to that in the presence of DMSO, similar to that observed in the FCR-3 strain (Fig. 3E, 48 h). A further 24 h of the assay revealed that the number of infected erythrocytes in the presence of 2-APB was much lesser than that in the presence of DMSO, in which there was a high level of parasitaemia (Fig. 3E, 72 h).

Trophozoites are the Main Target of 2-APB
To investigate the stages of intraerythrocytic development that 2-APB may block, 2-APB was removed from or added to cultures at different time intervals and parasitaemia was determined after 40 h of the assay for each developmental stage. Two independent results are shown in Fig. 4A-E. When 2-APB was removed from culture at the ring form stage (10 h of the assay) (Fig. 4A), parasites present at 40 h of the assay developed into schizonts with normal morphology, but ring form parasitaemia in the next developmental cycle was significantly decreased. On the other hand, when 2-APB was removed between trophozoite to early schizont stages (21 h of the assay) (Fig. 4B), parasites present at 40 h of the assay developed into schizonts with either normal or abnormal morphology, and the effect of 2-APB in decreasing ring form parasitaemia in the next developmental cycle was greater than that shown in Fig. 4A. When 2-APB was added at the same time (21 h of the assay) (Fig. 4C), parasites present at 40 h of the assay developed into schizonts with abnormal morphology and ring form parasitaemia in the next developmental cycle was significantly decreased. These results indicate that the time at which 2-APB is most effective in blocking intraerythrocytic parasite development is between the trophozoite to early schizont stages. This is supported by the observation that a higher percentage of trophozoites at the time of addition of 2-APB resulted in greater effects on ring form parasitaemia (Fig. 4D). Furthermore, when 2-APB was added at a later stage (late schizont stage, 28 h of the assay), the parasites were able to develop into fully mature schizonts with segmented merozoites and produce ring forms like those observed in the control culture (Fig. 3B); however, ring form parasitaemia at 40 h of the assay was significantly decreased (Fig. 4E). This result suggests that 2-APB affects late schizont maturation and erythrocyte egress and/or invasion of merozoites. Finally, to further investigate the effect of 2-APB on the ring form stage, parasite size was measured at 10 and 20 h of the assay (Fig. 4F). In parasites cultured with DMSO, increases in area, perimeter and maximum diameter at 20 h of the assay were higher than those observed at 10 h of the assay as expected from the results in Fig. 3C. A similar result was obtained when DMSO was removed at 10 h of the assay. In contrast, in parasites cultured with 2-APB at 10 and 20 h of the assay, all 3 parameters were significantly smaller than those in parasites cultured with DMSO; however, when 2-APB was removed at 10 h of the assay, all 3 parameters were slightly smaller than those in parasites cultured with DMSO but recovered to a level at which no statistically significant difference was detected. From these results we propose 2 possibilities to explain the finding that removal of 2-APB at 10 h of the assay caused a significant decrease in ring form parasitaemia at 40 h of the assay (Fig. 4A). (1) The growth of trophozoites that were exposed to 2-APB before 10 h of the assay was inhibited. and (2) a proportion of ring forms did not recover and resume normal development. Together with results of Ca 2+ imaging experiments (Fig. 2), the inhibitory effects of 2-APB on Ca 2+ oscillations and intraerythrocytic parasite development in the blood stage are summarized in Fig. S7. Ca 2+ oscillations observed in early ring forms and early trophozoites reduced on treatment with 2-APB ( Fig. 2A and B), but treatment at the ring form stage showed a significantly weaker effect on intraerythrocytic development than that between trophozoite to schizont stages. Considering the reversible effect of 2-APB on the early ring form stage (Fig. 4F), we conclude that the lethal effect of 2-APB on intraerythrocytic parasite development was caused mainly by the blockage of Ca 2+ oscillations in the early trophozoite stage. The importance of Ca 2+mediated signals for merozoites invasion of erythrocytes has been reported previously [17,18], and hence, the results in Fig. 4E suggest that IP 3 -induced periodic Ca 2+ fluctuations in the merozoites stage has an important role in parasite invasion. Moreover, the weak effect of 2-APB on strain K1 at 24 h of the assay (Fig. 3E) may be attributable to the fact that most of the K1 parasites did not reach the stage (trophozoite to early schizont stage) at which 2-APB had a lethal effect in the FCR-3 strain (Fig. 4). However, it remains possible that the effectiveness of 2-APB in the ring form stage of strain K1 differs from its effectiveness in the FCR-3 strain.

Severe Parasite Degeneration Caused by 2-APB
Ultrastructural changes induced by 2-APB were observed by transmission electron microscopy. As shown in Fig. 5A and Fig.  S8, parasites cultured with DMSO at 30 h of the assay maintained a normal structure. In contrast, highly dense chromatin masses in the nucleus and highly dense degeneration were consistently observed 30 h after the assay ( Fig. 5B and C). The formation of Maurer's cleft and malaria pigment in the food vacuoles suggests that degeneration induced by 2-APB occurred after intraerythrocytic development to some extent (Fig. 5B). In Plasmodium species, the nuclear envelope is considered the main ER compartment [31][32][33][34]. We confirmed that the ER-Tracker signals surrounded the nuclei of parasites stained blue with Hoechst 33342 and cultured with DMSO, whereas in parasites cultured with 2-APB the ER-Tracker signals became broad and extended to the cytosol (Fig. 6). Similarly, in electron micrographs parasites cultured with 2-APB characteristically showed a dilated nuclear envelope and ER ( Fig. 5B and C) and the dilated nuclear envelopes connected with the dilated ER, forming a reticular ER structure (Figs 5D and  6). In a previous report, a nucleus surrounded by rough ER in a late schizont with the nuclear envelope bearing numerous ribosomal granules was observed by electron microscopy [35] and similar electron micrographs were obtained with parasites cultured with DMSO in this study (Fig. S8B). Interestingly, parasites cultured with 2-APB frequently exhibited an increased number of ribosomal granules, distributed at even intervals in a line along the dilated nuclear envelope (Fig. 5E). Similar morphological alterations in the nuclear envelope and rough ER were also observed in some parasites exposed to artemisinin [36], which is known to inhibit the sarco/endoplasmic reticulum Ca 2+ -ATPase orthologue (PfATP6) of P. falciparum [37], suggesting that these ultrastructural changes are commonly induced by interference with Ca 2+ signalling. Dilatation of the nuclear envelope and ER is possibly among the features of cell degeneration, but it may also compensate for disrupted Ca 2+ signalling, since ER is a multifunctional and highly dynamic organelle, undergoing constant movement and reorganization depending on cellular conditions [38,39]. Although most of the parasites cultured with 2-APB showed severe degeneration ( Fig. 5B and C), schizonts in which merozoites with normal micro-organelles were formed were also present (Fig. 5F). At 40 h of the assay, the number of merozoites formed in each schizont of parasites cultured with 2-APB was significantly smaller than that formed in parasites cultured with DMSO ( Fig. S9A and B), suggesting that parasite development was inhibited by 2-APB even in parasites in which merozoites were formed normally.

Conclusive Remarks
In higher eukaryotes, disrupted intracellular Ca 2+ signalling and accumulation of aberrant proteins is known to cause ER stress, a hallmark of cell death that is associated with many neurodegenerative diseases [40][41][42][43]. However, very little is known about the biological significance and molecular composition of Ca 2+ signaling systems in unicellular organisms. In this study, we clearly showed the stage-specific spontaneous Ca 2+ oscillations in the intraerythrocytic stages of P. falciparum, a unicellular eukaryote, without any exogenous extracellular stimulation. We further demonstrated that the blockage of these Ca 2+ oscillations by 2-APB caused severe cellular degeneration resulting to the death of the parasites. The half maximal inhibitory concentration (IC 50 ) value of 2-APB for inhibition of IP 3 -induced Ca 2+ release in mammalian cells was ,40 mM [21]. Thus, a high concentration (100 mM) 2-APB was used in this study. Such a high dose 2-APB possibly exerts pleiotropic effects on both host cells and parasites. However, our findings in this study strongly support the idea that the severe inhibitory effect of 2-APB on the intraerythrocytic development of the parasites is primarily due to the specific blockage of Ca 2+ oscillations of the early trophozoites. First, the normal development of the parasites was observed in 2-APBpretreated erythrocytes (Fig. S5). Secondly, the critically effective time window of 2-APB on the intraerythrocytic development of the parasites was coincident with the early trophozoite stage in which spontaneous Ca 2+ oscillation was observed (Fig. 2, 4 and  S7). Thirdly, the delayed development of the parasites was recovered when 2-APB was removed at 10 h of the assay (Fig. 4F). Lastly, 100 mM 2-APB did not disrupt Ca 2+ gradient between the food vacuole and cytoplasm unlike a mM order of chloroquine treatment permeabilized the food vacuole membrane, resulting in the cell death [20]. Interestingly, a much higher contribution of the Ca 2+ oscillations in trophozoite stage to the intraerythrocytic development of the parasites was observed than that in the ring form stage. This result indicates that the Ca 2+ oscillation observed in ring form stage has a different physiological role from that in the trophozoite stage. In Plasmodium species, gametocytogenesis delivers the sexual-stage of the parasite known as gametocyte involved in the transmission from the mammalian host to the mosquito. Gametocyte development can be induced by host factors or drug treatment, and of which signal transduction pathways are involved in this process [44]. There is consistent evidence that phorbol ester inducing pathways and cyclic AMP (cAMP) signalling pathway are involved in triggering gametocy-

(D) Percentages of T and S in 2 independent experiments
shown in (C) just before 2-APB was added at 21 h of the assay. (E) 2-APB was added at 28 h of the assay. Parasitaemia at 40 h of the assay is shown as mean + S.D. of 3 independent counts of 3 wells. Stages with parasitaemia of less than 0.1% are not shown. The difference in Rf parasitaemia between the DMSO and 2-APB groups was analysed statistically (two-tailed unpaired t test ) and P values are given in each panel (A, B, C and E). (F) Effects of 100 mM 2-APB on the area, perimeter and maximum diameter of parasites. Three experimental groups were tested as follows. 2-APB was added at the beginning of the assay during synchronization, and cell size was analysed at 10 or 20 h of the assay. 2-APB was added at the beginning of the assay, removed at 10 h of the assay and cell size was measured at 20 h of the assay. Fifty parasites were measured in each experimental group. P values compared with DMSO controls are given in each panel (two-tailed unpaired t test with Welch's correction). doi:10.1371/journal.pone.0039499.g004 togenesis of P. falciparum [45][46][47][48][49] and interplay between cAMP and Ca 2+ as second messengers was also reported [24]. Taken together, our results suggest the possibility that Ca 2+ oscillations during ring form stage might be involved in triggering gametocytogenesis, rather than in maintaining asexual erythrocytic cycle.
During the erythrocytic cycle, Plasmodium species repeat the drastic morphological and functional changes; invasion, feeding, multiplication, secretion and structural modification of the parasite-infected erythrocytes. Our study clearly indicated that Ca 2+ signaling plays a pivotal role for cell growth and differentiation of P. falciparum, suggesting that it could be a useful experimental model organism for understanding fundamental roles and mechanisms of Ca 2+ signaling conserved from unicellular organisms to humans. In clinical aspects, the mature-stage of P. falciparum modifies the surface structure of the host erythrocytes. The parasite-infected erythrocytes adhere to endothelium and sequester the microvasculature of several organs and block the blood circulation. This pathology called ''sequestration'' causes severe symptoms including coma, acute respiratory distress, kidney failure and death in humans. Thus, consequences of malaria are closely associated with intraerythrocytic P. falciparum development.
Although functional evidence for IP 3 -induced Ca 2+ release has been reported in Plasmodium species [50,51], little attention has been paid to the molecules upstream of Ca 2+ release from intracellular Ca 2+ stores due to the lack of molecular identities for them. An IP 3 R gene, as defined in metazoans, has not been identified in the Plasmodium genome. This apparent absence could be due to the lack of homology with IP 3 Rs in metazoans as in plants [26,52]. Our results increase the possibility that the identification of molecules responsible for generating Ca 2+ oscillations including IP 3 R will provide promising targets for the development of novel antimalarial drugs.

P. falciparum Culture
The FCR-3 and K1 strains of P. falciparum were cultured using the modified method of Trager and Jensen [53] in RPMI medium (Invitrogen/Gibco) supplemented with 0.5% Alubumax I (Invitrogen), 25 mM HEPES, 24 mM sodium bicarbonate, 0.5 g/L Lglutamine, 50 mg/L hypoxanthine, 25 mg/mL gentamicin (Sigma) and human erythrocytes (from a healthy Japanese volunteer) at a haematocrit of 5% was used for culture. Growth synchronization was achieved with 5% D-sorbitol [54].

Inhibition of P. falciparum Development by 2-APB
The 2-APB concentration used was determined by preliminary experiments with a variety of concentrations ranging from 1 to 150 mM (data not shown). Effects of 2-APB on the intraerythrocytic parasite development were assayed using parasite cultures in the ring form stage with initial parasitaemia of approximately 1%-2%. Cultures (500 ml) were placed in each well of a tissue culture plate (24-well flat-bottomed; Corning). 2-APB was dissolved in DMSO (Hybri-MaxH, Sigma) at 10 mM. Stock solutions were diluted with PPMI medium and added to each well of the culture plate to give a defined concentration. DMSO diluted with medium served as control. A drop of cultured erythrocytes was smeared on a glass slide and stained with Giemsa. The number of parasiteinfected erythrocytes in 2000 erythrocytes was counted and defined as the level of parasitaemia.

Parasite Size Calculation and Transmission Electron Microscopy
Giemsa-stained smears were observed under the Nikon Eclipse 80i microscope (Nikon), photographed using the Nikon DXM 1200F camera and uploaded on a personal computer using digital photo manager software (ACT-1; Nikon). To measure parasite size, we randomly selected 50 parasites and manually delineated areas containing parasites with lines on the screen. The area, perimeter and maximum diameter of the parasites were calculated by WinROOF software package Ver.5.8.1 (Mitani, Japan).

Transmission Electron Microscopy
Transmission electron microscopy was performed as previously described [55]. Specimens for transmission electron microscopy were fixed for 2 h in 2.5% (v/v) glutaraldehyde buffered with 0.1 M phosphate buffer, pH 7.4, at 4uC. They were postfixed in 1% (w/v) osmium tetroxide for 1 h. Fixed specimens were dehydrated in ascending concentrations of ethanol followed by propylene oxide for 15 min, and embedded in Epon 812. The blocks were cut with an ultramicrotome (Porter-Blim MT-2; Ivan Sorvall) with a diamond knife (Diatome). The sections were mounted on 200-mesh copper grids and stained with uranyl acetate and lead citrate, and examined under the JEOL JEM-1011 transmission electron microscope.

Nucleus and ER Staining
The nucleus and ER of the parasites was stained with Hoechst 33342 and ER-Tracker Red (Invitrogen). Staining with Hoechst 33342 was performed as described above for fluorescence Ca 2+ imaging. ER-Tracker Red was then added at a final concentration of 0.5 mM to the erythrocyte suspension and shaken at 200 rpm for 30 min at 37uC. Erythrocytes were then washed once and resuspended in RPMI1640 medium without phenol red and observed under fluorescence microscopy (Leica DM 2500).

Ethics Statement
The human erythrocytes stock for the parasite culture was provided by the Hokkaido Kushiro Red Cross Blood Centre under the ethical guidelines for the blood products and obtained according to their acquisition guidelines. Written informed consent from the donor of the human erythrocytes was obtained. This study was done without a regular review from ethics committees of the Obihiro University of Agriculture and Veterinary medicine, where the parasite culture was made, because the erythrocytes stock was provided as blood product from the Red Cross after completing their ethical review for the experiment.  Figure S5 Pre-treatment of erythrocytes with 2-APB did not inhibit intraerythrocytic development of P. falciparum. To evaluate the effects of 2-APB on host cells, erythrocytes were pre-treated with 100 mM 2-APB for 1 h at 37uC, washed with RPMI medium and resuspended at a haematocrit of 5% in complete culture medium. Cultures of late schizonts (LS; 0.5% parasitaemia) were diluted four times with pre-treated erythrocytes, and the culture was continued. Cultures (three wells per experimental group) were terminated at 40 h of the assay, and thin smears of erythrocytes were prepared for parasite counting. Parasitaemia with ring forms did not differ significantly between dimethyl sulfoxide (DMSO)-and 2-APB-pretreated groups (P = 0.408, two-tailed unpaired t test). Parasitaemia is shown as mean + S.D. of three independent counts of three wells. Stages with parasitaemia of less than 0.1% are not shown. (TIF) Figure S6 Effect of 2-APB on the area, perimeter and maximum diameter of the chloroquine-resistant strain K1. No statistically significant difference was observed in the area, perimeter and maximum diameter of intraerythrocytic parasites between DMSO-and 100 m M 2-APB-cultured groups after 24 h of the assay, but the 3 parameters showed a tendency to decrease. Error bars represent mean + S.D. (n = 50). P values are given in each panel (two-tailed unpaired t test). (TIF) Figure S7 Timetable of the effects of 2-APB on Ca 2+ dynamics and parasite development in the blood stage of P. falciparum. Images show Giemsa-stained parasites. Blue bars above the parasite images indicate stages of the parasites in which spontaneous Ca 2+ oscillations or small periodic Ca 2+ fluctuations were observed. Red lines represent the period of 2-APB treatment. Red circles indicate the initiation of 2-APB treatment. Black, dark and light grey boxes show the extent of the effect of 2-APB during each sampling period: black, severe effect with developmental delay and abnormal morphology; dark grey, weaker effect with developmental delay and abnormal morphology than that when parasites were exposed to 2-APB for 40 h; light grey, slight effect with developmental delay or abnormal morphology compared to that when parasites were exposed to 2-APB for 40 h.