The Digestive Food Vacuole of the Malaria Parasite Is a Dynamic Intracellular Ca 2 (cid:1) Store* □ S

The acidic food vacuole of Plasmodium falciparum has been the subject of intense scientific investigation in the 40 years since its role in the digestion of host hemoglobin was first suggested. This proposed role has important implications for the complex host-parasite inter-relationship and also for the mode of action of several of the most effective antimalarial drugs. In addition, adaptive changes in the physiology of this organelle are implicated in drug resistance. Here we show that in addition to these functions, the digestive food vacuole of the malaria parasite is a dynamic internal store for free Ca 2 (cid:1) , a role hitherto unsuspected. With the aid of live-cell laser scanning confocal imaging, spatiotemporal studies revealed that maintenance of elevated free Ca 2 (cid:1) in the digestive food vacuole (relative to cytosolic levels) is achieved by a thapsigargin (and cyclopiazonic acid)-sensitive Ca 2 (cid:1) -pump in cooperation with a H (cid:1) -dependent Ca 2 (cid:1) transporter. Redistribution of free cytosolic and vacuolar Ca 2 (cid:1) during parasite growth also suggests that vacuolar Ca 2 (cid:1) plays an essential role in parasite morphogenesis. These data imply that the digestive food vacuole of the malaria parasite is functionally akin to the vacuole of plants (tonoplast) and the small electron-dense granules of some parasites (acido-calcisomes) whereby H (cid:1) -coupled Ca 2 (cid:1) transport is involved in ion transport, Ca 2 (cid:1) homeostasis, and signal transduction. These findings have significant implications for parasite development, antimalarial drug action, and mechanisms of drug resistance. 0, 10, 20, and 50 (cid:1) M ferriprotoporphyrin IX RF-5001PC spectrofluorophotometer. Real Time Ca 2 (cid:1) Confocal Laser Scanning Microscopy— For experi- mentation, suspensions (1%) of infected erythrocytes in HEPES-buff-ered RPMI medium (no serum) were loaded with Fluo-4/AM (3 (cid:1) M , Molecular Probes BV) for 20 min at 37 °C. For imaging, malaria para-site-infected erythrocytes were immobilized using polylysine-coated coverslips in a Bioptechs FCS2 perfusion chamber and maintained at 37 °C in growth medium (no serum). Inhibitors were added to the perfusate, and the Ca 2 (cid:1) -dependent fluorescence responses were moni-tored in real time. Fluorescence signals from Fluo-4 were collected on a Zeiss LSM510 confocal microscope through a Plan-Apochromat 63 (cid:2) 1.4 N.A. oil objective using the “tracking mode.” Excitation of Fluo-4 was performed using an argon ion laser at 488 nm. Emitted light was reflected through a 505–550-nm band pass filter from a 540-nm dichroic mirror. Photobleaching (the irreversible damage of Fluo-4/AM producing a less fluorescent species) was assessed by continuous exposure (5 min) of loaded cells to laser illumination. For each experiment, the laser illumination setting that gave a minimal reduction in cytosolic and food vacuole fluorescence was used (this varied according to dye loading and confocal laser scanning microscopy (CLSM) settings). Data capture and extraction was carried out with LSM510 version 3 software (Zeiss).

Malaria remains one of the largest global health (and economic) problems, resulting in more than a million deaths, mostly in young African children (1). With this in mind, it is astonishing how little is known about the basic physiology of the intraerythrocytic malaria parasite and its organelles. The food vacuole is a major digestive organelle of the malaria parasite and a proven chemotherapeutic target. This organelle has a role in degradation of host-derived hemoglobin, is the site of action of important classes of antimalarials, and harbors transporters associated with drug resistance (2)(3)(4). Hydrolysis of hemoglobin in the malaria parasite digestive food vacuole occurs by the integrated action of aspartic, cysteine, and metalloproteases (5), resulting in the production of hemozoin (malaria pigment), a biocrystal of the toxic precursor ferriprotoporphyrin IX (FPIX) 1 (6). The internal pH of the digestive food vacuole is lower than the parasite cytosol, although the precise pH value is under debate (7). In collaboration with Prof. Kiaran Kirk's laboratory (8), we have recently described the function of two H ϩ -pumping mechanisms in the digestive food vacuole; one is a bafilomycin-sensitive V-type H ϩ -ATPase, and the other is a NaF-sensitive H ϩ -pyrophosphatase showing a partial dependence on K ϩ . The combination of a H ϩ -ATPase and a H ϩ -pyrophosphatase acting in the digestive vacuole is analogous to the situation in acidic tonoplasts of plant cells (9) and of the acidocalcisome of some protozoa (10). It should be noted that although there may be similarities with acidocalcisomes (which are small electron-dense granules/organelles that store Ca 2ϩ as a phosphate precipitate) and the malarial parasite food vacuole, they are morphologically (and functionally, with regards to hemoglobin digestion) very distinct. Based on the known H ϩ -coupled Ca 2ϩ transport in these nonmalarial organelles, we hypothesized that an analogous coupled pathway would be present in the malaria parasite digestive food vacuole, thereby contributing to cellular Ca 2ϩ trafficking and storage.
By using live-cell confocal laser scanning microscopy and a judicious choice of Ca 2ϩ probes, we have uncovered a novel function for this unique organelle. We show that the food vacuole of the malaria parasite is involved in dynamic Ca 2ϩ storage. Evidence is presented showing that Ca 2ϩ storage by the digestive food vacuole is linked directly to parasite cell signaling and asexual development. This novel discovery raises many new questions about the functions, physiology, and regulation of this critical malarial parasite organelle. As the target for the actions of a number of drugs, the requirement for tightly regulated Ca 2ϩ movement and storage within this organelle may represent a new chemotherapeutic target.

EXPERIMENTAL PROCEDURES
Parasite Culture-Plasmodium falciparum TM6 strain was kindly provided by Dr. P. Tan with 0, 10, 20, and 50 M ferriprotoporphyrin IX was performed in a Shimadzu RF-5001PC spectrofluorophotometer.
Real Time Ca 2ϩ Confocal Laser Scanning Microscopy-For experimentation, suspensions (1%) of infected erythrocytes in HEPES-buffered RPMI medium (no serum) were loaded with Fluo-4/AM (3 M, Molecular Probes BV) for 20 min at 37°C. For imaging, malaria parasite-infected erythrocytes were immobilized using polylysine-coated coverslips in a Bioptechs FCS2 perfusion chamber and maintained at 37°C in growth medium (no serum). Inhibitors were added to the perfusate, and the Ca 2ϩ -dependent fluorescence responses were monitored in real time. Fluorescence signals from Fluo-4 were collected on a Zeiss LSM510 confocal microscope through a Plan-Apochromat 63ϫ 1.4 N.A. oil objective using the "tracking mode." Excitation of Fluo-4 was performed using an argon ion laser at 488 nm. Emitted light was reflected through a 505-550-nm band pass filter from a 540-nm dichroic mirror. Photobleaching (the irreversible damage of Fluo-4/AM producing a less fluorescent species) was assessed by continuous exposure (5 min) of loaded cells to laser illumination. For each experiment, the laser illumination setting that gave a minimal reduction in cytosolic and food vacuole fluorescence was used (this varied according to dye loading and confocal laser scanning microscopy (CLSM) settings). Data capture and extraction was carried out with LSM510 version 3 software (Zeiss). Cytosolic and vacuolar [Ca 2ϩ ] was calibrated in situ by the addition of known amounts of [Ca 2ϩ ] free in the presence of nigericin (0.8 M) and the Ca 2ϩ ionophore A23187 (1 M). Using Calcium Calibration Buffer Kit 1 solutions (Molecular Probes), the [Ca 2ϩ ] free was calculated from the K d of EGTA for Ca 2ϩ (107.9 nM at 37°C (12)) using the equation

Effect of FPIX on the Fluorescence Quantum Yield of Ca 2ϩ
Indicators-Heme (FPIX) is a major component of the digestive food vacuole of P. falciparum. As the fluorescence quantum yield of some food vacuole targeted probes can be reduced dramatically by heme (FPIX), we decided to test the effect of monomeric heme E on a number of Ca 2ϩ -specific UV-and visible light-excitable fluorescent probes. Titration of FPIX to solutions of Fura-2 or Fluo-4 resulted in a progressive reduction of fluorescence signal (Fig. 1). However, this reduction was more pronounced in the case of the UV-excitable probe (e.g. Fura-2) than for visible light-excitable probes (e.g. Fluo-4 and Fluo-3 (not shown)); Fluo-4 retained ϳ60% of its fluorescent emission signal at FPIX levels, which eliminated the Fura-2 signal. Fluo-4 was therefore used as the probe of choice to measure cytosolic and food vacuole Ca 2ϩ fluctuations.
The Digestive Food Vacuole of P. falciparum Contains Elevated Free Ca 2ϩ -With the use of CLSM and the cell-permeant AM ester of Fluo-4 to probe for Ca 2ϩ in trophozoites of infected red blood cells, we observed a strong fluorescence signal deriving from the digestive food vacuole (e.g. Fig. 3e). Some Ca 2ϩ indicators have a tendency to compartmentalize, and differences in apparent Ca 2ϩ affinities in different compartments/ organelles within a cell can sometimes lead to the formation of artifactual Ca 2ϩ gradients (13). To overcome these possible artifacts, free Ca 2ϩ was calibrated independently in situ in both the parasite cytosolic and food vacuole compartments (Fig.  2). The differing slopes (and fluorescence quantum yields in response to Ca 2ϩ ) of the parasite cytosol and food vacuole calibration curves revealed that there was a difference in the respective apparent Ca 2ϩ binding affinities, probably resulting from differences in the Ca 2ϩ buffering capacity of the different intracellular milieus and possibly from a degree of dye compartmentalization. However, taking into account the differences in Fluo-4 Ca 2ϩ affinities (by using the respective in situ calibration curves), the digestive food vacuole was measured nevertheless as having elevated free Ca 2ϩ relative to the cytosol, with levels ranging between 250 and 300 nM (n ϭ 27), some 5 to 6 times higher than the cytosol (measured at ϳ50 nM, n ϭ 27).
Intracellular Distribution of Free Ca 2ϩ during the Intraerythrocytic Development of the Malaria Parasite-The distribution of free Ca 2ϩ in the malaria parasite was observed to change during asexual development. As described, early trophozoites were observed as having a Ca 2ϩ gradient between the digestive food vacuole and cytoplasmic compartments (Fig.  3e). However, as parasites underwent schizogony, free Ca 2ϩ was observed to leave the digestive food vacuole and move to the intermembrane space surrounding each segment (Fig. 3, f  and g). On rupture and release of merozoites from the erythrocytic host, there was a sustained elevation in free Ca 2ϩ ; however the resolution of the CLSM did not allow us to determine the specific intracellular distribution of the signal (Fig.  3h). This increase was lost shortly after reinvasion of new red blood cells (i.e. the early ring stage), which may imply an important role for elevated merozoite free Ca 2ϩ in the invasion process.
Spatiotemporal Dynamics of Free Ca 2ϩ in the Malaria Parasite and Digestive Food Vacuole-Fluo-4/AM is a BAPTA (1,2bis(2-aminophenoxy)ethane-N,N,NЈ,NЈ-tetraacetic acid)-based indicator with a large dynamic range (R f Ϸ 50 -200 (14)), making it very suitable for the monitoring of both elementary and global Ca 2ϩ signals (13). Using this indicator, we were able to determine spatiotemporal Ca 2ϩ fluctuations in malaria parasite-infected erythrocytes that were immobilized using polylysine coated coverslips in a Bioptechs FCS2 perfusion chamber, in response to a variety of inhibitors and agents. As described, trophozoites were shown to posses elevated Ca 2ϩ in the digestive vacuoles (Fig. 3e). The addition of thapsigargin (a plant sesquiterpene lactone that interacts irreversibly with sarcoplasmic reticulum-like Ca 2ϩ ATPases (15)) to trophozoites  Fig. 4D, Quick-Time movie) and 5B). Unlike the case of thapsigargin, however, the reduction of food vacuolar Ca 2ϩ upon addition of cyclopiazonic acid was accompanied by a conspicuous transient increase in cytosolic Ca 2ϩ . Maneuvers that caused a perturbation of the trans-vacuolar membrane pH gradient (the food vacuole is acidic relative to the cytosol), such as the addition of the selective V-type ATPase inhibitor, bafilomycin A 1 (Fig. 5C), and of NH 4 Cl (QuickTime movie, Figs. 4E and 5C) to the perfusate, also caused a rapid reduction of vacuolar Ca 2ϩ . A transient rise in cytosolic Ca 2ϩ was also evident upon addition of NH 4 Cl to the perfusate (Fig. 5C). The membrane integrity of the digestive food vacuole was maintained on the addition of the irreversible inhibitors thapsigargin, cyclopiazonic acid, and bafilomycin A 1 (see QuickTime movies, Fig. 4, B, D, and E) and the malaria pigment crystals were observed to continue to move inside the vacuoles. Washing infected erythrocytes with HEPES-buffered RPMI (pH 7.4) or Ringer solution after treatment with NH 4 Cl restored the food vacuole Ca 2ϩ gradient of the cells (data not shown).
As the digestive food vacuole of the malaria parasite contains a P-glycoprotein pump (PGH1), which can affect the compartmentalization of some fluorescence indicators, verapamil, a known PGH1 modulator (17), was added to the perfusate of the single cell chamber system. Verapamil (10 M) was not shown to effect any change in the fluorescence signal deriving from either the food vacuole or the cytoplasm, suggesting that there was no interference from PGH1 in our system. DISCUSSION It has been suggested previously that the malaria parasite contains an acidic Ca 2ϩ storage organelle (18 -20). However, negative staining of the digestive vacuole by UV-excitable Ca 2ϩ fluorescent probes precluded its involvement (20). We have shown recently that FPIX can interact with fluorescent probes such as acridine orange, resulting in a dramatic reduction in fluorescent emission, probably occurring from the absorption by the FPIX Soret peak (21). We therefore investigated whether the fluorescent emission of previously used UV-excitable Ca 2ϩ -specific probes such as Fura2 (19,20) as well as a range of visible light-excitable Ca 2ϩ probes were also affected by FPIX. Titration of FPIX to solutions of Fura-2 or Fluo-4 resulted in a progressive reduction of fluorescence signal (Fig.  1). However, this reduction was more pronounced in the case of Fura-2 than Fluo-4, which retained ϳ60% of its original fluorescent emission signal at FPIX levels that eliminated the Fura-2 signal. Importantly, the probe remained Ca 2ϩ -sensitive within this organelle, albeit with an altered apparent Ca 2ϩ affinity (Fig. 2). FPIX is reported to achieve mM concentrations in the food vacuole (22), which would fully explain the lack of food vacuole-derived Fura-2 fluorescence reported by other investigators.
With the use of CLSM and Fluo-4/AM (which is less susceptible to FPIX fluorescence quenching and has a good dynamic range) to probe for Ca 2ϩ in mature trophozoites of infected red blood cells, we observed a free Ca 2ϩ pool in the digestive food vacuole (Fig 3e) generating a Ca 2ϩ gradient some 5-6-fold greater than that measured in the cytoplasm (at rest).
It was clearly observed that the sequestration of free Ca 2ϩ altered during the asexual intraerythrocytic development of the malaria parasite (Fig 3). For example, as parasites underwent schizogony, free Ca 2ϩ was observed to leave the digestive food vacuole and move to the intermembrane space surrounding each segment (Fig. 3, f and g) producing "wagon wheel"looking Ca 2ϩ distributions. It has been accepted for more than 20 years that Ca 2ϩ in the culture medium is essential for the complete asexual development of the malaria parasite in red blood cells in culture (23). The results presented here lend further support to a role for Ca 2ϩ -dependent processes in the asexual development of the parasite.
Our findings showing Ca 2ϩ storage by the digestive food vacuole have allowed us to shed light on a recent controversy relating to the pharmacology of the sarco/endoplasmic reticulum Ca 2ϩ -ATPases of P. falciparum. A recent study demonstrated that the malaria parasite contains an endoplasmic reticulum-like Ca 2ϩ store that is controlled by a thapsigargininsensitive but cyclopiazonic acid-sensitive Ca 2ϩ -pump (20), a finding that was at odds with previous studies indicating a thapsigargin-sensitive mechanism (19,24). It is now apparent that both of these studies adopted methods that were ineffective for the measurement of vacuolar Ca 2ϩ and therefore that their fluorescence signals originated from cytosolic Ca 2ϩ fluctuations. Using our live cell CLSM system to monitor Fluo-4loaded infected red blood cells, we observed that upon addition of thapsigargin (1 M) or cyclopiazonic acid (10 M) to the perfusate, Ca 2ϩ from the digestive food vacuole was rapidly (ϳ60 s) released (Fig. 4, A and C, and Fig. 5, A and B (Quick-Time movies, Fig. 4, B and D)). In the case of cyclopiazonic acid (but undetectable for thapsigargin), the vacuolar release of Ca 2ϩ was accompanied by a small transient increase in cytosolic Ca 2ϩ (Figs. 4C and 5B). These results help explain previous observations of transient elevations of Ca 2ϩ upon addition of cyclopiazonic acid (20). In addition these results suggest the operation of more than one P-type Ca 2ϩ ATPase, one that is sensitive to both thapsigargin and cyclopiazonic acid and one that is only sensitive to cyclopiazonic acid. Significantly, and in support of our findings, the recent functional characterization of P. falciparum transporters expressed in Xenopus laevis oocytes have revealed the presence of both a cyclopiazonic-sensi- tive Ca 2ϩ -ATPase (PfATPase4 (25)) and a thapsigargin-sensitive Ca 2ϩ -ATPase (PfATPase6). 2 At this early stage in the pharmacological analysis of intracellular Ca 2ϩ signaling in the malaria parasite, however, it should be noted that thapsigargin and cyclopiazonic acid have previously been shown to have different additional nonselective effects on Ca 2ϩ signaling pathway components, e.g. capacitative Ca 2ϩ entry (26). However, it is unlikely that capacitative Ca 2ϩ entry is responsible for the observed increase in cytoplasmic Ca 2ϩ on treatment with cyclopiazonic acid, as this effect was observed by Alleva and Kirk (20) to occur in freed P. falciparum parasites suspended in Ca 2ϩ -free solutions.
The generation of a H ϩ -gradient across the parasite vacuolar membrane by the action of V-type H ϩ ATPases and H ϩ -pumping pyrophosphatases (8) suggests that, as with plant tonoplasts and acidocalcisomes of some parasites, the digestive food vacuole of the malaria parasite may also drive Ca 2ϩ accumulation by H ϩ -coupled transport mechanisms. This hypothesis is supported by our findings that a collapse of the transmembrane H ϩ gradient by the addition of NH 4 (Fig. 5C (QuickTime movie,  Fig. 4E)) or bafilomycin A 1 (a selective V-type H ϩ -ATPase inhibitor; Fig. 5C) results in a rapid discharge of vacuolar Ca 2ϩ . Given the essential role of Ca 2ϩ as a secondary messenger molecule and the fact that physiological studies in plants have established a role for vacuolar (tonoplast) H ϩ /Ca 2ϩ exchange activity in ion transport, Ca 2ϩ homeostasis, and signal transduction (27), further elucidation of the analogous H ϩ -dependent mechanisms underlying Ca 2ϩ homeostasis in the malaria parasite will be of significant importance.
Here we add another critical transport function to the malarial parasites repertoire of vacuolar transport mechanisms (Fig. 6). It will be important to determine how the interplay of Ca 2ϩ transport and H ϩ transport via either the H ϩ -pyrophosphatase or the V-type H ϩ -ATPase contributes to organelle function. The rapid redistribution of vacuolar Ca 2ϩ at schizogony (asexual multiplication) must implicate communication between the parasite nucleus and the food vacuole. In addition 2 S. Krishna, personal communication. FIG. 6. Schematic representation of known and putative transporters/channels on the P. falciparum digestive food vacuole membrane. Although morphologically and functionally (e.g. hemoglobin digestion) quite distinct, the digestive food vacuole shares properties similar to the acidocalcisome of other apicomplexa (as described by Docampo and Moreno (10)). this organelle harbors transporters implicated in drug resistance (notably PfCRT and PGH1 (reviewed in Ref. 4)). The potential for proton-coupled Ca 2ϩ transport to influence the function of these resistance transporters could have significant implications for malaria chemotherapy in the age of the quinoline-resistant parasite.