Mechanism of Action of Miltefosine on Leishmania donovani Involves the Impairment of Acidocalcisome Function and the Activation of the Sphingosine-Dependent Plasma Membrane Ca²⁺ Channel

ABSTRACT

Leishmania donovani is the causing agent of visceral leishmaniasis, a common infection that affects millions of people from the most underdeveloped countries. Miltefosine is the only oral drug to treat infections caused by L. donovani. Nevertheless, its mechanism of action is not well understood. While miltefosine inhibits the synthesis of phosphatidylcholine and also affects the parasite mitochondrion, inhibiting the cytochrome c oxidase, it is to be expected that this potent drug also produces its effect through other targets. In this context, it has been reported that the disruption of the intracellular Ca²⁺ homeostasis represents an important object for the action of drugs in trypanosomatids. Recently, we have described a plasma membrane Ca²⁺ channel in Leishmania mexicana, which is similar to the L-type voltage-gated Ca²⁺ channel (VGCC) present in humans. Remarkably, the parasite Ca²⁺ channel is activated by sphingosine, while the L-type VGCC is not affected by this sphingolipid. In the present work we demonstrated that, similarly to sphingosine, miltefosine is able to activate the plasma membrane Ca²⁺ channel from L. donovani. Interestingly, nifedipine, the classical antagonist of the human channel, was not able to fully block the parasite plasma membrane Ca²⁺ channel, indicating that the mechanism of interaction is not identical to that of sphingosine. In this work we also show that miltefosine is able to strongly affect the acidocalcisomes from L. donovani, inducing the rapid alkalinization of these important organelles. In conclusion, we demonstrate two new mechanisms of action of miltefosine in L. donovani, both related to disruption of parasite Ca²⁺ homeostasis.

INTRODUCTION

Leishmaniasis is a parasitic neglected tropical disease affecting millions of people all over the world. There are three main forms of this disease: visceral leishmaniasis (VL), cutaneous leishmaniasis (CL), and mucocutaneous leishmaniasis, all of which are caused by 20 different Leishmania species, which are transmitted by phlebotomine sand flies. Current estimates indicate that from 0.2 to 0.4 million people are affected by visceral leishmaniasis, which is the most severe clinical form of the disease and usually leads the patient to death if untreated. Its etiologic agents are the trypanosomatid parasites Leishmania infantum (in the Americas) and Leishmania donovani (Asia, Middle East and Africa) (1). The classical treatments against leishmaniasis include pentavalent antimonials (glucantime and pentostan), which present serious disadvantages, such as variable efficacy, parenteral, and marked side effects. More recently, amphotericin B administered in liposomal complex has been shown to be very efficient (2). Another class of compounds, alkyl phosphorylcholines and related derivatives, have shown efficacy against L. donovani (3). A similar compound derived from phosphocholine, miltefosine, was first used as an anti-neoplastic drug (4) and has shown large efficacy against L. donovani and other trypanosomatids like Trypanosoma cruzi and Trypanosoma brucei (5). Miltefosine also showed antiparasitic action in vivo on VL-infected patients in India (6). Accordingly, in the last few years miltefosine efficacy against different Leishmania species has been reported (7, 8). Furthermore, miltefosine has shown a synergistic effect with several drugs, among others, with nanotized curcumin against L. donovani (9), with amiodarone against L. mexicana (10), with allopurinol against canine VL produced by L. infantum (11), and with pentamidine against L. infantum-HIV coinfection (12). Despite its reported side effects as inducer of resistance and teratogenic action, evidence proving miltefosine antileishmanial action in vitro and in vivo led to its use as the first oral treatment for VL (13, 14).

In concerns to the mechanism of action of miltefosine, several compounds have been shown to act as inhibitors of lipid biosynthesis in kinetoplastid parasites. Among them, lysophospholipids produced a marked effect on the phospholipid composition of trypanosomatids, in which the biosynthesis of phosphatidylcholine (PC) is inhibited at the level of phosphatidylethanolamine N-methyltransferase (15). Miltefosine, as an alkyl-lysophospholipid, showed a reduction of the concentration of phosphatidylcholine in T. cruzi. Remarkably, it has been claimed that miltefosine inhibits the biosynthesis of PC in T. cruzi (16) with 10 to 20 times more potency compared to that of mammalian cells (17), thus explaining its high selectivity as antiparasitic drug. The same mechanism has also been reported in L. donovani, in which phosphatidylcholine concentration is decreased and phosphatidylethanolamine concentration is enhanced (18).

Previous reports demonstrate that miltefosine causes decreases in the oxygen consumption rate and ATP levels of L. donovani through inhibition of the mitochondrial cytochrome c oxidase (19). Furthermore, miltefosine also produces an apoptosis-like death in L. donovani promastigotes (20).

With regard to Ca²⁺ signaling, it is known that the mechanisms involved in Ca²⁺ regulation in trypanosomatids constitute a target for chemotherapeutic agents like amiodarone and dronedarone, which disrupt Ca²⁺ homeostasis in T. cruzi and L. mexicana (21–24) through their action on two organelles acting as Ca²⁺ compartments, the mitochondrion and the acidocalcisomes. Moreover, the antituberculosis compound SQ109, which also possesses a very potent trypanocidal effect, was recently found to act on T. cruzi (25) and L. mexicana (26) through the same mechanism of Ca²⁺ and mitochondrial disruption. Also in concerns to disruption of Ca²⁺ regulation, it has been reported that many Ca²⁺ channel antagonists produce a marked effect in several trypanosomatids (27), including L. donovani (28). In fact, a plasma membrane Ca²⁺ channel homolog to the human L-type voltage-gated Ca²⁺ channel (VGCC) has been described in L. mexicana (29). This channel shares many characteristics with its human homolog, such as antagonism by classical human channel blockers (nifedipine and verapamil). However, remarkably, the parasite channel is selectively stimulated by the sphingolipid sphingosine, while the VGCC is not (29). In the present work we show new mechanisms of action of miltefosine, demonstrating that this drug is able to activate a Ca²⁺ channel in the plasma membrane of L. donovani similar to the sphingosine-activated channel mentioned above for L. mexicana. Although miltefosine simulated the effect of sphingosine, the activation of the parasite channel by this drug was not completely blocked by dihydropyridines such as nifedipine, the classical human L-type VGCC antagonist. Furthermore, in the present work we also demonstrate that miltefosine has a direct effect on L. donovani acidocalcisomes.

RESULTS

Effect of miltefosine on the intracellular Ca²⁺ concentration of L. donovani promastigotes.

Several mechanisms have been proposed for the mode of action of miltefosine on Leishmania spp. These include disturbances of the lipid-dependent signaling pathways (16), inhibition of cytochrome C oxidase (19), and an apoptosis-like cell death (30). However, there is increasing evidence that Ca²⁺ homeostasis could be a target for the action of drugs against trypanosomatids (21–24), and the role of Ca²⁺ on different cellular processes, including cell death by apoptosis and necrosis, is well known. In order to determine the effect of miltefosine on the [Ca²⁺]_i (intracellular Ca²⁺ concentration) in L. donovani promastigotes, the parasites were loaded with the fluorescent Ca²⁺ indicator Fura-2. It can be observed (Fig. 1) that the addition of miltefosine (4 μM) induced a large increase in the [Ca²⁺]_i. We used this concentration because it has been previously shown, based on a dose-response curve, that at 4 μM miltefosine exerts its maximal effect on the magnitude of the [Ca²⁺]_i increase in L. mexicana (10). Addition of sphingosine (10 μM), which at this concentration is known to optimally activate the plasma membrane Ca²⁺ channel in L. mexicana (29), showed no further effect. Accordingly, when miltefosine was added after the rapid increase in the [Ca²⁺]_i induced by sphingosine, the drug did not produce any further increase in fluorescence. These results suggest that miltefosine and sphingosine share the same mechanism of action, namely, the opening of a Ca²⁺ channel at the plasma membrane.

FIG 1.

Effect of miltefosine and sphingosine on the intracellular Ca²⁺ concentration of L. donovani promastigotes. Promastigotes were loaded with Fura-2 and the indicated compounds were added directly to the cuvette, as described in Materials and Methods. (A) Miltefosine (4 μM) was added (arrow) in the presence of 2 mM extracellular Ca²⁺, followed by the addition of sphingosine (10 μM). (B) Sphingosine (10 μM) was added as indicated (arrow), followed by miltefosine (4 μM). Traces are representative of at least three independent experiments.

We then sought to verify if the observed Ca²⁺ channel activated by miltefosine corresponds to the same entity of the sphingosine-sensitive plasma membrane Ca²⁺ channel already described in L. mexicana (29). For this, we used Bay K 8644, a very specific agonist of the human L-type VGCC, widely used for its functional characterization, and which has been demonstrated to indeed also activate the Ca²⁺ channel reported in L. mexicana (29). At 4 μM the agonist is known to induce the maximal opening of the human L-type VGCC and also the similar channel in L. mexicana (29). It was observed that upon addition of Bay K 8644 (Fig. 2) this agonist totally substituted the effect of miltefosine. That is, Bay K 8644 (4 μM) did not produce any further effect after addition of miltefosine (Fig. 2A) and, accordingly, addition of miltefosine after Bay K 8644 did not induce any further Ca²⁺ release (Fig. 2B). These results support the notion that both miltefosine and sphingosine act on the same channel.

FIG 2.

Effect of miltefosine and the Ca²⁺ channel agonist Bay K 8644 on the intracellular Ca²⁺ concentration of L. donovani promastigotes. (A) Miltefosine (4 μM) and then Bay K 8644 (4 μM) were added (arrows) directly to the cuvette in the presence of 2 mM extracellular Ca²⁺. (B) Bay K 8644 (4 μM) was added (arrow), followed by miltefosine (4 μM) when indicated, in the presence of 2 mM extracellular Ca²⁺. Traces are representative of at least three independent experiments. (See Materials and Methods section for details).

We performed experiments to determine whether the effect of a dihydropyridine (nifedipine), a classical human L-type VGCC inhibitor, was able to also block the effect of miltefosine. It was observed (Fig. 3A) that nifedipine (4 μM) partially blocks the effect generated by miltefosine, while this channel blocker produces total blockade of the sphingosine action, as previously reported in L. mexicana (29), and shown here in Fig. 3B. In these experiments, we used nifedipine at 4 μM, since this concentration is twice the amount of this antagonist known to totally block the sphingosine-activating effect on the L. mexicana channel (29). Addition of the mild detergent digitonin (40 μM), known to disrupt the permeability barrier of the plasma membrane without affecting intracellular organelles, induced a further increase in the Ca²⁺ signal reaching the maximal fluorescence level, as expected. Further addition of EGTA (10 mM) to chelate all extracellular Ca²⁺ caused the fluorescence signal to reach the lowest level. These results suggest that the mechanism of action of miltefosine is similar but not identical to that of sphingosine.

FIG 3.

Effect of the L-type VGCC channel blocker nifedipine on the action of miltefosine on the intracellular Ca²⁺ concentration of L. donovani promastigotes. (A) Black line indicates effects of miltefosine (4 μM) in the presence of extracellular CaCl₂ (2 mM), followed by digitonin (40 μM) and EGTA (arrows), respectively. Gray line indicates effects of nifedipine (4 μM), followed by miltefosine (4 μM) in the presence of 2 mM extracellular Ca²⁺, followed by digitonin (40 μM) and EGTA (arrows), respectively. (B) Effect of nifedipine (4 μM) followed by sphingosine (10 μM), miltefosine (4 μM), digitonin (40 μM) and EGTA, respectively (arrows) in the presence of 2 mM extracellular Ca²⁺. (C) Effect of miltefosine in the absence of extracellular Ca²⁺. EGTA was added to chelate any contaminating extracellular Ca²⁺ (arrow), followed by miltefosine (4 μM), CaCl₂ (2 mM), digitonin (40 μM), and EGTA, respectively. (D) Effect of miltefosine after addition of nifedipine in the absence of extracellular Ca²⁺. EGTA was added to chelate any contaminating extracellular Ca²⁺ (arrow), followed by nifedipine (4 μM), miltefosine (4 μM), CaCl₂ (2 mM), digitonin (40 μM), and EGTA, respectively. Traces are representative of at least three independent experiments.

We then determined the effect of miltefosine in the absence of extracellular Ca²⁺. Figure 3C shows that addition of the drug, instead of inducing an increase in the [Ca²⁺]_i, reduced it to well below the basal level. This is due to the presence of EGTA, which chelates all extracellular Ca²⁺ and forces the intracellular basal Ca²⁺ to leave the cytoplasm toward the outside medium. When Ca²⁺ is restored at the extracellular milieu, a large increase was now observed, indicating that the channel had indeed been opened by miltefosine.

The effect of miltefosine after the addition of nifedipine and in the absence of extracellular Ca²⁺ was then tested (Fig. 3D). According with the results obtained in Fig. 3A and B, the release of intracellular Ca²⁺ obtained after miltefosine addition in the presence of the blocker was less than that with miltefosine alone (Fig. 3C), indicating a partial blockage of the channel and confirming that nifedipine does not completely block the activating effect exerted by miltefosine on this Ca²⁺ channel.

Effect of miltefosine on intracellular organelles of L. donovani.

We also studied the possible effect of miltefosine on the intracellular organelles known to be involved in Ca²⁺ homeostasis, such as the acidocalcisomes and the unique giant mitochondrion present in these parasites. Concerning this last organelle, it was previously shown that miltefosine has a mitochondrial depolarizing effect, reported as an impairment of the ability of the parasites to accumulate rhodamine 123 after 14 h of treatment with miltefosine in L. donovani promastigotes (19). This effect was predictable since miltefosine inhibits the cytochrome c oxidase, which in turn would affect the mitochondrial membrane potential. We now show that miltefosine produces a very short-term collapse of the mitochondrial electrochemical membrane potential, since a fast, large increase in rhodamine 123 fluorescence was observed upon addition of miltefosine (Fig. 4A). In principle, this effect was also predictable, since miltefosine induces the entrance of Ca²⁺ and therefore its accumulation in the mitochondrion. This is known to occur via a mitochondrial Ca²⁺ uniporter (31) also present in the Leishmania parasite (32), whose driving force is the mitochondrial electrochemical membrane potential. Thus, any Ca²⁺ entry would dissipate this potential, which will be translated into the release of rhodamine 123. However, the depolarizing effect of miltefosine on the parasite's mitochondrion was also observed, albeit at a lesser extent, in the absence of extracellular Ca²⁺ (Fig. 4B), indicating that this effect is partially independent of the entrance of the cation to the cell, but is a direct effect of miltefosine on this organelle. In both extracellular Ca²⁺ condition experiments we added the mitochondrial electron chain uncoupler fluorocarbonyl cyanide-p-(trifluoromethoxy)phenylhydrazone (FCCP) (2 μM), which is expected to completely deenergize the mitochondria at this concentration (32, 33). Only a small response was obtained for this effector after miltefosine was added in the absence of extracellular Ca²⁺ (Fig. 4B), confirming again the large effect of miltefosine on this organelle and its partial dependence on the entrance of extracellular Ca²⁺.

FIG 4.

Effect of miltefosine on the mitochondrial electrochemical potential of L. donovani promastigotes. Parasites were incubated in the presence of rhodamine123 (10 mg/ml) for 30 min at room temperature, as indicated in Materials and Methods. (A) In the presence of 2 mM extracellular Ca²⁺, miltefosine (4 μM) was added (arrow), followed by FCCP (2 μM). (B) Miltefosine (4 μM) was added (arrow), followed by addition of FCCP (2 μM). This was performed in the absence of extracellular Ca²⁺. Traces are representative of at least three independent experiments.

We then studied the effect of miltefosine in another very relevant compartment, associated as well with intracellular Ca²⁺ regulation and also involved in the L. donovani parasite's bioenergetics, the acidocalcisomes (34). We determined whether miltefosine had an effect on this organelle by the use of acridine orange, which is known to be accumulated in acidic reservoirs. These experiments were performed in the absence of extracellular Ca²⁺, to exclude the possible effect associated with the entrance of Ca²⁺ through the plasma membrane channel to the cytoplasm that could interfere with the basal Ca²⁺ content in the acidocalcisomes, and therefore with the degree of acidity. Figure 5A shows that the addition of miltefosine (4 μM) to promastigotes loaded with the fluorescent indicator produced a large increase in fluorescence due to the release of the fluorophore from the acidocalcisomes after its alkalinization by action of the drug. The consecutive addition of nigericin (2 μM), a known K⁺/H⁺ exchanger and therefore an inducer of the alkalinization of these organelles, produced a further increase in fluorescence. This might mean that miltefosine was not able to completely alkalinize the acidocalcisomes, or alternatively it would imply the existence in the parasite of other acidic compartments affected by this drug. In Fig. 5B we performed the same experiment but inverted the order of addition of the effectors. It can be noticed that nigericin induced alkalinization of the acidocalcisomes. Addition of miltefosine after nigericin induced a further effect, which could be attributed to the action of this compound on different acidic organelles in which acridine orange accumulates.

FIG 5.

Effect of miltefosine on acidocalcisomes in L. donovani promastigotes. Parasites were loaded with acridine orange (2 mM) as described in Materials and Methods. The excitation wavelength was 488 nm, and emission was at 530 nm. (A) Miltefosine (4 μM) was added (arrow) directly to the stirring cuvette with promastigotes loaded with acridine orange, followed by the addition of nigericin (2 μM). This was performed in the absence of extracellular Ca²⁺. (B) In the absence of extracellular Ca²⁺, nigericin was added at 2 μM (arrow), followed by miltefosine (4 μM). Traces are representative of at least three independent experiments.

DISCUSSION

Miltefosine is the first oral drug prescribed against leishmaniasis and its effects against L. donovani are well known, the Old World visceral-leishmaniasis-causing agent. Nevertheless, until the last decade little was known about the mechanism of action of this drug. One of its most remarkable effects is the inhibition of phosphatidylcholine synthesis, being 10 to 20 times more selective for the phosphatidylethanolamine N-methyl-transferase from the parasite compared to its human counterpart, thus explaining its large selectivity against trypanosomatids (15). Other relevant well-described effects of miltefosine are associated with its action on the cytochrome c oxidase, explaining the disruption of overall mitochondrial function (19). In this work we report new mechanisms of action for miltefosine. First, we demonstrated that this compound activates a plasma membrane Ca²⁺ channel in L. donovani, similar to the human L-type VGCC homolog previously described in L. mexicana. Thus, similar to its human counterpart, the channel is activated by the specific L-type VGCC agonist Bay K 8644 and is blocked by dihydropyridines (like nifedipine), which are classical L-type VGCC antagonists. Also similar to the Ca²⁺channel from L. mexicana, this channel is opened by the sphingolipid sphingosine, which is a distinctive feature of the trypanosomatid channel (29). In this context, disruption of intracellular Ca²⁺ homeostasis has been recognized as a putative target for drug action on trypanosomatids (35). The action of many other drugs used against these parasites, such as pentamidine (36), amiodarone (10, 21, 22, 22), dronedarone (23, 24, 37), and SQ109 (25, 26) is mainly through disruption of Ca²⁺ regulation. Accordingly, a large Ca²⁺ entrance to the cell induced by miltefosine should produce a massive impairment of Ca²⁺ function, causing the death of the parasite.

We also demonstrated in this work that miltefosine produced a dramatic, fast, and direct effect on the acidocalcisomes of L. donovani. This would also cause an increase in cytoplasmic Ca²⁺, since alkalinization of these organelles would lead to the release of this cation, thus adding its effect to the action produced by Ca²⁺ entrance through the plasma membrane Ca²⁺ channel. Besides, acidocalcisome impairment would have consequences on the bioenergetics of the parasite, since this organelle is involved in the production and accumulation of pyrophosphates (38), which are considered an alternative energetic coin in trypanosomatids. In turn, this effect should reinforce the well-recognized action of miltefosine on the mitochondrion, since, as mentioned, this drug inhibits the cytochrome c oxidase, which produces impairment of the membrane electrochemical membrane potential, the driving force for Ca²⁺ accumulation inside this organelle (32). Related to this point, we cannot discard a possible direct effect of miltefosine on the mitochondrial function, beyond its action on the cytochrome c oxidase. Thus, the experiments performed in this work showing the total collapse of the mitochondrial electrochemical membrane potential in seconds, very different from the previously reported long-lasting effect of miltefosine on the membrane potential observed after several hours (19), would support a third effect of this compound in these parasites. This is reinforced by the fact that the results obtained when the experiments were performed in the absence of extracellular Ca²⁺ were very similar, thus discarding an effect that could be attributed to Ca²⁺ entry to the cell through the just-opened plasma membrane Ca²⁺ channel, which would induce the entry of the cation to the mitochondria, causing the collapse of its membrane potential. Although these experiments are not conclusive, this possibility remains open. The results suggest that the effect discussed on the mitochondrial membrane potential induced by miltefosine would reinforce its global effect on the increase in the intracellular Ca²⁺ concentration, with the expected overall consequences on the parasite biology. This large increase in the intracellular concentration of this cation could be also the basis for the apoptotic effects on these parasites attributed to miltefosine (20), since it have been demonstrated that an increase in cytoplasmic Ca²⁺ concentration is a condition for cells to take the decision to start the apoptotic fate (39).

In concerns to the presence of a sphingosine-activated homolog of the human L-type VGCC in L. donovani, Fig. 6 depicts the sequence alignment of the α1C subunit of the human channel with L. mexicana and L. donovani homologs on the relevant domains, including the binding sites for the specific channel blocker nifedipine. This sequence alignment shows that, although there is a 26% homology between the human and the L. donovani sequences, there is a 100% homology sequences of the two Leishmania species. Furthermore, the complete sequence of the gene for the L. mexicana channel is about 94% similar to the L. donovani sequence (40). This high homology between this two species explains the similarities observed during this work concerning the opening of the channel by sphingosine, as well as the Bay K 8644 activation and nifedipine antagonism of this parasite channel.

FIG 6.

Sequence alignments of the IIIS6 and IVS6 domains of human L-Type VGCC (NCBI accession number NP_955630.3) with L. mexicana (NCBI accession number XP_003878633.1, gene ID according to TriTrypDB LmxM.33.0480) and L. donovani (NCBI accession number CBZ37533.1, gene ID according to TriTrypDB LdBPK_340500.1) homologs. The amino acid sequences next to the selectivity filter are highlighted in gray and the amino acids associated with dihydropyridines (nifedipine) responsiveness are highlighted in gray and underlined (29).

In conclusion, the results shown demonstrate a double effect of miltefosine on L. donovani, namely, the opening of the sphingosine-activated plasma membrane Ca²⁺ channel and a direct effect on the acidocalcisomes, which in combination should produce a large intracellular Ca²⁺ accumulation. Interestingly, both mechanisms of actions are parasite-specific. Both effects are correlated with the abrupt increase in the intracellular Ca²⁺ concentration observed in L. donovani upon addition of miltefosine. Since the disruption of the parasite Ca²⁺ homeostasis has been claimed as a target for the action of several drugs against trypanosomatids, the results presented here, added to the well-recognized action of miltefosine on phospholipid synthesis and on cytochrome c oxidase inhibition, would contribute to the dramatic parasite death induced by this drug and could explain the large benefits attributed to miltefosine.

MATERIALS AND METHODS

Chemicals.

Miltefosine (hexadecyl phosphocholine), sphingosine, Bay K 8644, verapamil, EGTA, digitonin, fluorocarbonyl cyanide-p-(trifluoromethoxy)phenylhydrazone (FCCP), and nigericin were from Sigma (St. Louis, MO). Fura-2-acetoxymethyl ester (Fura-2-AM), acridine orange, and rhodamine 123 were from Molecular Probes (Eugene, OR).

Culture of L. donovani promastigotes.

L. donovani (DD8 strain) promastigotes were cultured in liver infusion tryptose (LIT) medium supplemented with 10% of fetal bovine serum at 26°C as reported previously (7).

Intracellular Ca²⁺ measurements.

L. donovani promastigotes were loaded with the Ca²⁺ radiometric indicator Fura-2 as reported previously (24). The fluorophore Fura-2 is excited by two different wavelengths, 340 nm when it is Ca²⁺-bound and 380 nm when it is free of Ca²⁺, and emission is recorded at a unique wavelength of 510 nm. Briefly, 1 × 10⁷ parasites were collected by centrifugation at 600 × g for 2 min and washed twice in a loading buffer (137 mM NaCl, 4 mM KCl, 1.5 mM KH₂PO₄, 8.5 mM Na₂HPO₄, 11 mM glucose, 1.8 mM CaCl₂, 0.8 mM MgSO₄, 20 mM HEPES-NaOH [pH 7.4]). The pellet was resuspended in 1 μM Fura-2-AM (the acetoxymethyl ester derivative of Fura-2), and probenecid (2.4 mM) and pluronic acid (0.05%) were added to the loading buffer. The parasites were incubated at 29°C in the dark with continuous agitation for 2 h. Fura-2-AM-loaded parasites were washed twice in the same buffer, in either the presence or absence of Ca²⁺. The CaCl₂ concentration used in all the experiments where the cation was present was done in the presence of 2 mM Ca²⁺, mimicking the extracellular concentration present in the growth medium. Additionally, EGTA (500 μM) was added when measurements were made in the absence of extracellular Ca²⁺. This concentration of EGTA is high enough to chelate all possible contaminant Ca²⁺ and to lower its concentration to a level which favorably competes with the fluorescent Ca²⁺ indicator, making it possible to obtain the minimal fluorescence value. Digitonin (40 μM) is known to permeabilize the parasite cell membrane, allowing Ca²⁺ entrance from the extracellular medium (22). 10 mM EGTA was added at the end of the experiments in order to obtain the maximal and minimal fluorescence values, respectively (29). Fluorescence measurements were carried out on a stirred cuvette at 29°C, using a Perkin-Elmer LS-55 fluorescence spectrometer with a double wavelength excitation beam (340 nm and 380 nm), recording the emission at 510 nm.

Determination of the mitochondrial membrane potential.

The effect of miltefosine on the mitochondrial membrane potential of L. donovani promastigotes was evaluated using the fluorescent dye rhodamine 123 as reported previously (21), taking advantage of the internationalization of the fluorophore, according to the mitochondrial electrochemical membrane potential (Δφ). Briefly, 8 × 10⁶ parasites were collected by centrifugation at 600 × g for 2.5 min and washed in phosphate-buffered saline (PBS) plus 1% glucose. The pellet was resuspended in the same buffer in the presence of rhodamine 123 (20 μM) and incubated for 45 min at 29°C in the dark with continuous stirring. Subsequently, parasites were washed twice and resuspended in the same buffer, and then transferred to a stirred cuvette. Measurements (excitation wavelength [λ_ext], 488 nm; emission wavelength [λ_em], 530 nm) were made in a Hitachi 7000 spectrofluorimeter at 29°C. The protonophore FCCP (2 μM) was used as a positive control.

Determination of acidocalcisome alkalinization.

The effect of miltefosine on acidocalcisomes was evaluated using acridine orange, which is accumulated in acidic compartments (22). Promastigotes (8 × 10⁶ cells/ml) were collected, washed, and incubated in a loading buffer (the same used in mitochondrial membrane potential measurements) with acridine orange at 2 μM for 5 min at 29°C in the dark and with constant stirring. Measurements were performed with λ_ext at 488 nm and λ_em at 530 nm at 29°C in a Hitachi 7000 spectrofluorimeter under magnetic stirring. Nigericin, a K⁺/H⁺ exchanger that is known to alkalinize the acidocalcisomes, was used at 2 μM as a positive control. This concentration of nigericin exceeds the amount required for complete release of acridine orange from acidocalcisomes (41).