Strains and culture conditions.

L. mexicana (MNYC/BZ/62/M379) promastigotes were grown in vitro at 25°C in HOMEM medium supplemented with 10% (vol/vol) heat-inactivated fetal calf serum (35). L. mexicana (MNYC/BZ/62/M379) amastigotes were grown at pH 5.5 in Schneider's Drosophila medium (Gibco) supplemented with 20% (vol/vol) heat-inactivated fetal calf serum and gentamicin sulfate at 25 μg/ml and 32°C (13). L. mexicana promastigotes resistant to pentamidine were obtained in vitro from sensitive L. mexicana cells by stepwise exposure to the drug with final concentrations of 2.5 μM, 5 μM, 10 μM, and 30 μM until they had no effect on parasite growth, as previously described for other Leishmania species (9). L. mexicana amastigotes resistant to pentamidine were obtained axenically by transformation of promastigotes resistant to 30 μM pentamidine at 32°C without drug pressure (13). Once transformed, parasites were cultivated with 30 μM pentamidine. The stability of pentamidine resistance in vitro was determined as described previously (9).

Cross-resistance to other drugs was determined by measuring the 50% inhibitory concentration (IC₅₀) and the resistance factor (IC₅₀ ratio between resistant and wild-type cells) with the colorimetric MTT dye reduction assay (48), in which the tetrazolium salt MTT is converted to a colored formazan product by enzymes active only in living cells. The IC₅₀ value corresponds to the concentration of a given compound provoking 50% inhibition of cell proliferation compared with untreated cells after 3 days. The effect on pentamidine resistance of 5 μM trifluoperazine (TFP), 5 μM prochlorperazine (PCP), 15 μM verapamil, and 1 mM buthionine sulfoximine (BSO) was studied by incubating the parasites with these compounds and different concentrations of pentamidine for 72 h, with the inhibitory effect on growth being measured as described above.

Transport studies.

Parasites were harvested during the mid-logarithmic phase of growth by centrifugation at 2,100 × g for 10 min at room temperature. Cells were then washed three times with phosphate-buffered saline supplemented with 1% d-glucose (PBSG) at pH 7.4 for promastigotes and pH 5.5 for amastigotes. Cells were resuspended in this buffer at densities appropriate for the procedure to be used, as indicated in the text and figure legends. Parasite suspensions (160 μl, containing 2 × 10⁷ cells for promastigotes and 3 × 10⁷ cells for amastigotes) were warmed to 25°C (for promastigotes) or 32°C (for amastigotes) and mixed with 20 μl of assay buffer containing labeled molecule plus or minus other test compounds at the concentrations indicated in the text. Transport was terminated by the rapid separation of parasites from the buffer components by centrifugation through a 9:1 mixture of dibutyl phthalate (specific gravity, 1.04) and mineral oil (specific gravity, 0.875 to 0.885). The sample tubes were immediately flash frozen in liquid nitrogen, and the tubes were cut to separate the pellet from the transport medium. The pellet was dissolved in 2% sodium dodecyl sulfate (250 μl) and 3 ml of scintillation fluid (EcoscintA; National Diagnostics, Hessle, United Kingdom). These were left overnight to remove the effects of chemiluminescence, and then incorporated radioactivity was counted. The protein content was determined by the dye-binding method (21) and calculated as 6.1 and 2.0 μg of protein/10⁶ cells for L. mexicana promastigotes and amastigotes, respectively. Cell volumes were previously determined as 100 μm³ for promastigotes and 27 μm³ for amastigotes (17).

The effects of various pharmacological reagents on pentamidine accumulation in wild-type and resistant cells was investigated by resuspending cells in PBSG containing inhibitors at the concentrations indicated in the text and figure legends for 10 min at 25°C. After this treatment, 0.25 μM [³H]pentamidine was added to the mixture, and uptake over 10 s was measured.

In order to determine efflux of pentamidine, cells were loaded with 0.25 μM [³H]pentamidine for 10 min either at room temperature or on ice. After exposure to the drug, an aliquot was immediately centrifuged through oil and frozen in liquid nitrogen. Retained pentamidine was determined as described above. The remaining cells were centrifuged, washed three times with PBSG, resuspended in drug-free PBSG at 10⁸ cells/ml, and incubated at 25°C or 32°C. Aliquots of cells were removed, and intracellular pentamidine was determined over a range of time points. To evaluate the effect of energy, mitochondrial membrane potential, proton motive forces, and multidrug resistance-reversing agents on pentamidine efflux, inhibitors were added simultaneously with the [³H]pentamidine. Subsequently, cells were centrifuged, washed three times with PBSG containing inhibitors, and then resuspended in PBSG supplemented with the inhibitor but no pentamidine.

Analysis of kinetic data.

The net uptake of pentamidine into leishmanias probably involves several components, including a plasma membrane uptake system, a plasma membrane efflux system, and a mitochondrial uptake system. Technical limitations prevented us from dissecting out each of these components, and hence we measured net uptake and expressed the values from plots of net uptake against velocity as apparent kinetic parameters. Apparent K_m and V_max values for pentamidine uptake were determined by measuring apparent initial uptake rates over various concentrations of pentamidine between 1 μM and 1 mM. Nonlinear regression fits to the Michaelis-Menten equation with the GraFit version 4.09 for Microsoft Windows (Erithacus Software) were used to determine apparent K_m and V_max values. The type of inhibition and K_i values for pentamidine, arginine, and other molecules were determined with the enzyme inhibition program of GraFit 4.09.

Fluorescence microscopy.

Cells were incubated with 2 μM 4′,6′-diamidino-2-phenylindole (DAPI) for the times marked in the text. Cells were then transferred to a microscope slide and viewed by fluorescence microscopy (excitation at 330 nm, emission at 400 nm). To permeabilize the cells, 1 μM digitonin was added to the cells for 1 to 2 min before DAPI was added.

Pentamidine uptake by wild-type and resistant parasites.

Uptake of 0.25 μM pentamidine was measured over 3 h in pentamidine-resistant L. mexicana and compared with that in wild-type cells. The four pentamidine-resistant promastigote strains showed a reduced accumulation of drug in comparison with sensitive cells, with all of the resistant lines accumulating very little (Fig. ​(Fig.1).1). After 3 h of incubation, the intracellular pentamidine concentration was 10 μM for Pent^r2.5, 9 μM for Pent^r5, 7 μM for Pent^r10, and 5 μM for Pent^r30, compared with 55 μM in the wild-type cell line. Similar results were obtained with amastigotes, where an intracellular concentration of 35 μM was obtained for the resistant line Pent^r30 and 230 μM for the sensitive amastigotes (data not shown).

Open in a separate window

FIG. 1.

Time course of pentamidine uptake by wild-type and pentamidine-resistant L. mexicana promastigotes. Parasites were incubated with 0.25 μM pentamidine at 25°C, and pentamidine uptake was measured between 5 s and 3 h. Wild-type (solid circles) and resistant Pent^r2.5 (solid squares), Pent^r5 (open circles), Pent^r10 (open triangles), and Pent^r30 (open square) strains were tested. Results are the mean ± standard error of the mean of at least three experiments.

Apparent kinetic constants for pentamidine uptake into both resistant promastigotes and amastigotes were determined with a range of substrate concentrations between 1 μM and 1 mM and compared with values obtained for wild-type cells (Table ​(Table2).2). The apparent K_m value was similar for both wild-type amastigotes and promastigotes. The apparent V_max value was, however, 19-fold higher in amastigotes than in promastigotes (Table ​(Table2).2). Apparent K_m values for pentamidine in resistant cells were similar to the mean values obtained for wild-type cells. Apparent V_max values, however, were substantially lower in resistant cells and inversely correlated with the degree of resistance. The apparent V_max value of Pent^r30 amastigotes was 17-fold lower than the value obtained for wild-type cells.

TABLE 2.

Kinetic constants of pentamidine uptake by wild-type and pentamidine-resistant L. mexicana strains

Cells	Mean apparent Km (μM) ± SEM (n = 4)	Mean apparent Vmax (fmol/s/μg of protein) ± SEM (n = 4)
Promastigotes
Wild type	7.4 ± 1.8	452 ± 42
Pentr2.5	6.2 ± 2.4	196 ± 23
Pentr5	5.8 ± 1.6	148 ± 12
Pentr10	5.4 ± 0.6	68 ± 3
Pentr30	5.9 ± 2.1	57 ± 5
Amastigotes
Wild type	13.9 ± 2.9	8,540 ± 726
Pentr30	13.2 ± 5.6	491 ± 85

Open in a separate window

Effects of pharmacological inhibitors of cellular homeostasis.

Pentamidine uptake was investigated in the presence and absence of different inhibitors to investigate the nature of the force driving uptake. The results indicate that classical inhibitors of mitochondrial metabolism, potassium cyanide, sodium azide, and 2,4-dinitrophenol (Table ​(Table3;3; data not shown for amastigotes), led to significant reductions in net pentamidine uptake in wild-type parasites. Ouabain, an inhibitor of the plasma membrane Na⁺/K⁺-ATPase (31), and monensin, an Na⁺ ionophore, did not significantly decrease pentamidine uptake in either promastigotes or amastigotes. These data indicate that pentamidine uptake is independent of Na⁺ ions. The K⁺ ionophore valinomycin and the protonophore carbonyl cyanide (33, 34) reduced pentamidine uptake.

The nonspecific H⁺-ATPase inhibitor N,N′-dicyclohexylcarbodiimide, which reduced the plasma membrane pH gradient and the membrane potential in Leishmania spp. (33, 34), oligomycin A, a mitochondrial H⁺-ATPase inhibitor (3), and N-ethylmaleimide all had inhibitory effects on pentamidine uptake. Orthovanadate, a plasma membrane-type H⁺-ATPase inhibitor (61), also decreased pentamidine uptake, whereas bafilomycin A₁, a specific vacuolar-type H⁺-ATPase inhibitor (20), had no effect. Verapamil, a Ca²⁺ channel blocker and a P-glycoprotein inhibitor, inhibited pentamidine uptake. Overall, while the agents used could affect the permeability of cells independently of their effect on the proton motive force, the data are consistent with inhibition of proton motive forces associated with mitochondrial metabolism having a profound effect on the net accumulation of pentamidine into L. mexicana.

Uptake in the presence of these inhibitors was also studied in Pent^r30 promastigotes. Net pentamidine uptake in the resistant line was increased in the presence of dinitrophenol, potassium cyanide, and sodium azide compared with uptake into the resistant cells themselves (Table ​(Table3).3). Valinomycin, carbonyl cyanide, N,N′-dicyclohexylcarbodiimide, N-ethylmaleimide, oligomycin A, and verapamil also increased pentamidine accumulation in resistant cells. These results indicate that the mitochondrial membrane potential is not involved in the accumulation of pentamidine in resistant parasites. Moreover, since most of these inhibitors also deenergize cells, they also indicate that the drug may be actively effluxed from energized cells.

Specificity of pentamidine transporter.

Pentamidine was shown to be a competitive inhibitor of arginine uptake for promastigotes and amastigotes, with K_i values of 12.5 ±3.8 μM and 16.3 ±2.8 μM, respectively. However, arginine had no inhibitory activity against pentamidine uptake, even at concentrations as high as 10 mM (data not shown). K_m and V_max values of arginine uptake were essentially identical in wild-type and pentamidine-resistant parasites (K_m values were 2.3 ± 0.7 μM for the wild-type and 3.3 ± 2.7 μM for the resistant line, and V_max values were 6.6 ± 0.5 nmol s⁻¹ [10⁶ cells]⁻¹ and 5.1 ± 0.6 nmol s⁻¹ [10⁶ cells]⁻¹ for the two lines, respectively). This would indicate that pentamidine does not enter cells via an arginine transporter.

The aromatic diamidines propamidine, stilbamidine, DAPI, and berenil and the phenanthridine isometamidium (Samorin) were found to be potent competitive inhibitors of pentamidine uptake, suggesting that the nature of the lateral chain is not important in recognition (Table ​(Table4).4). para-Aminobenzoic acid, benzamide, and benzylamine did not affect pentamidine uptake, whereas benzamidine, 3-aminobenzamidine, and 4-aminobenzamidine all inhibited uptake, but noncompetitively.

In order to establish the specificity of the transporter, uptake of 4 μM pentamidine was studied in the presence of a vast excess (1,000-fold or more) of potential competitive substrates. Polyamines (putrescine, spermidine, and spermine), amino acids (all 20 naturally occurring l-amino acids), nucleobases (adenine, guanine, cytosine, and uracil), nucleosides (adenosine, inosine, guanosine, thymidine, and cytidine), sugars (d-glucose, d-fructose, and d-galactose), and a variety of other metabolites (folic acid, biotin, choline, panthenoic acid, ethanolamine, thiamine, pterine, pyridoxal, taurine, linoleic acid, sphingosine, guanidine, urea, reduced glutathione, oxidized glutathione, ascorbic acid, and imidazole) all failed to inhibit pentamidine uptake. Two peptides rich in basic amino acids (Arg-Ser-Arg and Lys-Lys-Lys-Lys) did give modest inhibition (around 50% when used at 5 mM and 10 mM, respectively).

Inhibition was also seen with some inorganic divalent cations (74% inhibition with 5 mM Ca²⁺, 47% with 0.1 mM Cd²⁺, 45% with 0.1 mM Co²⁺, and 69% with 5 mM Ni²⁺). Taurocholate also decreased pentamidine uptake by 50% when used at 10 mM. Taurocholate was a noncompetitive inhibitor, with a K_i of 10.5 ± 1.9 mM (data not shown). Pentamidine uptake was sensitive to several compounds often used as P-glycoprotein inhibitors, including TFP, PCP, and verapamil (the three drugs were competitive inhibitors, with a K_i of 10.70 ± 0.04 μM for TFP, 8.01 ± 0.09 μM for PCP, and 53.2 ± 8.1 μM for verapamil [data not shown]).

Efflux of pentamidine.

Examining pentamidine uptake in the presence of metabolic inhibitors, including dinitrophenol, azide, and KCN, revealed significant decreases in accumulated drug in the wild-type parasites. However, these compounds increased net uptake in resistant cells, indicating that an energy-dependent efflux system for pentamidine could influence the resistance phenotype. This is in agreement with the observation that P-glycoprotein inhibitors reverse the resistance (Table ​(Table33).

To study this more, parasites were loaded to steady state over 10 min with radiolabeled pentamidine at room temperature, and retained drug was measured for 1 h in both wild-type and resistant cells. Accumulated pentamidine reached substantially higher levels in wild-type parasites during the loading phase. An efflux of pentamidine was then observed in both wild-type and resistant cells. This represented a significant proportion (about 70%) of total accumulated drug in resistant parasites but only a small fraction (about 14%) of the total in wild-type cells (Fig. ​(Fig.2A2A).

Open in a separate window

FIG. 2.

Efflux of pentamidine by L. mexicana promastigotes. Wild-type and pentamidine-resistant parasites were incubated with 0.25 μM pentamidine at 25°C (A) or on ice (B) for 10 min, centrifuged, washed three times with PBSG, and resuspended in drug-free PBSG. Retained pentamidine was measured between 1 and 60 min. Results are the mean ± standard error of the mean of at least three experiments.

When cells were loaded with pentamidine on ice, both wild-type and resistant lines accumulated similar levels. When these were then transferred to room temperature and the rate of efflux was measured in both cell lines, it was shown to occur at a similar rate (Fig. ​(Fig.2B),2B), indicating that increased efflux could not explain the resistance phenotype.

Pentamidine efflux was also assayed in the presence of various metabolic inhibitors and protonophores, and this also diminished efflux (Fig. ​(Fig.3).3). Typical inhibitors of the multidrug resistance phenotype in mammalian cells which have also been used to distinguish at least three efflux systems in Leishmania spp. (30), including verapamil (used at 50 μM), TFP, and PCP (both used at 20 μM), inhibited pentamidine efflux (Fig. ​(Fig.33 and ​and4).4). Treatment of the cells with 1 mM BSO for 24 h also inhibited efflux (Fig. ​(Fig.44).

Open in a separate window

FIG. 3.

Pentamidine efflux by pentamidine-resistant promastigotes under energy- and proton-depleting conditions. Retained pentamidine was measured after 30 min in the absence (control) and presence of inhibitors: 1 mM dinitrophenol (DNP), 10 μM carbonyl cyanide (CCCP), 50 μM N,N′-dicyclohexylcarbodiimide (DCCD), and 50 μM verapamil. Results are the mean ± standard error of the mean of at least three experiments.

Open in a separate window

FIG. 4.

Inhibition effects of different multidrug resistance-associated protein inhibitors on pentamidine efflux by pentamidine-resistant promastigotes. Time course of pentamidine efflux was measured in the presence and absence of 20 μM PCP and 20 μM TFP and after 24 h of treatment with 1 mM BSO. Results are the mean ± standard error of the mean of at least four experiments.

Pentamidine accumulation over a 6-h period was also investigated in wild-type and resistant promastigotes with or without PCP and TFP. In the presence of these inhibitors, resistant cells accumulated drug to level similar to those in the wild type by 6 h (Fig. ​(Fig.55).

Open in a separate window

FIG. 5.

Effect of multidrug resistance-associated protein inhibitors on time course of pentamidine accumulation by wild-type and pentamidine-resistant promastigotes. The effect of 10 μM PCP and 10 μM TFP in sensitive wild-type cells and 20 μM PCP and 20 μM TFP in resistant cells on pentamidine influx was studied. The wild type (shaded circles), wild type plus 10 μM PCP (solid circles), wild type plus 10 μM TFP (open circles), resistant line Pent^r30 (shaded squares), resistant line Pent^r30 plus 20 μM PCP (solid squares), and resistant line Pent^r30 plus 20 μM TFP (open squares) were tested. Results are means of at least three experiments.

This work was supported by a Wellcome Trust Traveling Fellowship to M. Basselin.

We are grateful to Janice Brock for technical assistance.