Activation of Artemisinin and Heme Degradation in Leishmania tarentolae Promastigotes: A Possible Link

Abstract

Endoperoxides (EPs) appear to be promising drug candidates against protozoal diseases, including malaria and leishmaniasis. Previous studies have shown that these drugs need an intracellular activation to exert their pharmacological potential. The efficiency of these drugs is linked to the extensive iron demand of these intracellular protozoal parasites. An essential step of the activation mechanism of these drugs is the formation of radicals in Leishmania. Iron is a known trigger for intracellular radical formation. However, the activation of EPs by low molecular iron or by heme iron may strongly depend on the structure of the EPs themselves. In this study, we focused on the activation of artemisinin (Art) in Leishmania tarentolae promastigotes (LtP) in comparison to reference compounds. Viability assays in different media in the presence of different iron sources (hemin/fetal calf serum) showed that IC₅₀ values of Art in LtP were modulated by assay conditions, but overall were within the low micromolar range. Low temperature electron paramagnetic resonance (EPR) spectroscopy of LtP showed that Art shifted the redox state of the labile iron pool less than the EP ascaridole questioning its role as a major activator of Art in LtP. Based on the high reactivity of Art with hemin in previous biomimetic experiments, we focused on putative heme-metabolizing enzymes in Leishmania, which were so far not well described. Inhibitors of mammalian heme oxygenase (HO; tin and chromium mesoporphyrin) acted antagonistically to Art in LtP and boosted its IC₅₀ value for several magnitudes. By inductively coupled plasma methods (ICP-OES, ICP-MS) we showed that these inhibitors do not block iron (heme) accumulation, but are taken up and act within LtP. These inhibitors blocked the conversion of hemin to bilirubin in LtP homogenates, suggesting that an HO-like enzyme activity in LtP exists. NADPH-dependent degradation of Art and hemin was highest in the small granule and microsomal fractions of LtP. Photometric measurements in the model Art/hemin demonstrated that hemin requires reduction to heme and that subsequently an Art/heme complex (λ_max 474 nm) is formed. EPR spin-trapping in the system Art/hemin revealed that NADPH, ascorbate and cysteine are suitable reductants and finally activate Art to acyl-carbon centered radicals. These findings suggest that heme is a major activator of Art in LtP either via HO-like enzyme activities and/or chemical interaction of heme with Art.

1. Introduction

Artemisinin (Art) is a natural endoperoxide (EP) usually isolated from Artemisia annua L. plants (1). The content of Art in those plants, which can be as much as 2% of their dry weight, strongly varies depending on breed, growth conditions, time of collection, storage conditions, part of the plants extracted, and many other factors. The pharmacological benefits of Art and related compounds were discovered during a Chinese antimalarial drug research initiative starting in 1967. As a result, since the 1980s in many countries Art is used to treat malaria arising from infection with Plasmodium falciparum and other plasmodia species (2). Attempts were made to repurpose this drug for other diseases. For example, the efficiency of Art against cancer cells was shown in vitro, in vivo, and in clinical settings (3–7). Furthermore, its application as antimycobacterial and antileishmanial agent was suggested (8,9). Despite of decades of research on Art in plasmodia there are still some ongoing debates about its mechanism and its specific targets (10). For the application of Art against Leishmania even less is known about its specific mechanism of action. We have previously shown that in Leishmania Art influences mitochondrial functions in the course of cell death. However, it does not directly target mitochondrial proteins themselves at concentrations relevant for its action against Leishmania (11). In a previous study about the antileishmanial mechanism of ascaridole (Asc) we observed a strong role of the labile iron pool (LIP) in the activation of this EP (12). However, biomimetic experiments with Art suggested that the activation may be different from Asc for this more complex EP.

Leishmania are protozoal organisms, which infect mammalian hosts and parasitize inside their macrophages. Leishmaniasis is a vector-borne disease with several hosts and reservoirs, including Phlebotominae sandflies, humans and dogs. This makes management of this zoonotic disease a challenge on a global scale (13). Depending on the Leishmania species and immunological response of the host, these intracellular parasites cause different clinical forms of leishmaniasis. Worldwide millions of people specifically in tropical and subtropical regions are infected by these parasites (14). According to the specific clinical manifestation of leishmaniasis, this can lead to severe disfigurings of skin, mucocutaneous tissues or in the case of visceral leishmaniasis even to death. Due to global warming the risk of infection with these parasites is also increased in previously non-endemic regions, including parts of Europe (15). Current treatment strategies in endemic regions are still partly based on antimony-containing drugs, which, however, cause increasing parasite resistance. Although there are effective alternatives, such as liposomal amphotericin B (16), the use of these drugs is on the one hand limited by costs and on the other hand by toxic adverse effects, which require tight clinical monitoring of their application. In clinical trials or as alternative treatment options compounds, such as miltefosine, paromomycin, pentamidine or sitamaquine have been used against leishmaniasis (17). None of these compounds is an EP or directly targets the iron metabolism in Leishmania.

As Leishmania parasites' survival in their intracellular amastigote form essentially depends on the supply of iron/heme from the host cell, the acquisition of this metal and its inherent redox activity provide interesting targets for potential drugs against Leishmania. Due to the variable environment of Leishmania amastigotes in the phagolysosome of macrophages they are adapted to acquire iron in different forms, such as low molecular iron (18–20), hemin/heme and endocytosed hemoglobin (21,22). Besides the acquisition of heme from macrophages, Leishmania possess the ability to synthesize heme compounds de novo (23,24). Little is known about the interconversion of heme compounds to low molecular iron in Leishmania (25), but it is assumed that Leishmania can break down heme as a source of iron (24,26). In mammalian cells this function is exerted by heme oxygenase (HO) enzymes, however, little is known about these activities in Leishmania (24).

Art and Asc are only two examples of the diversity of naturally occurring EPs (27). Although the therapeutic potential of most EPs is unknown, it is expected that the EP group contributes to a possible mode of action of those molecules. However, also the structural environment of the EP group has certainly tremendous influence on EP reactivity, protein-binding, target, metabolism, and pharmacokinetic/pharmacodynamic properties. In the application of the EP Art against plasmodia inside mammalian erythrocytes, heme continuously is available as reaction partner of Art from hemoglobin degradation. In contrast, Leishmania have to deal with the variable supply of different forms of heme and non-heme iron as possible EP reaction partners. This makes iron acquisition in Leishmania more difficult than in plasmodia. In the phagolysosome low molecular iron is liberated from internalized transferrin due to the low pH value and, on the other hand, also heme compounds are likely to occur in this mammalian cell compartment (20). In a previous report we have shown that the EP Asc is effective against Leishmania and that its action essentially depends on the available iron in Leishmania (12). However, so far it remained unclear how mechanisms in Leishmania, which are involved in iron uptake and metabolism, influence the efficiency of Art against Leishmania. In our previous work we have shown that the EP Asc, possessing a less complex structure than Art, interacted with the LIP in Leishmania (12). In addition, we observed in biomimetic experiments that Art has a much higher reactivity with heme than Asc (12), suggesting that heme could also play an important role in Art activation in Leishmania. In the physiological context the nutrition supply of intramacrophagal amastigotes may strongly depend on the location of the macrophage and its preferred phagocytized material. Preferably degradation of senescent erythrocytes by macrophages may result in high intracellular heme availability (28).

In the current study we elucidate how heme influences the efficiency of Art against Leishmania. Our results show that heme degradation in Leishmania and activation of Art to radical intermediates are closely linked to each other. A major role in these events play enzymatic processes of heme degradation (HO-like activities) in Leishmania, mainly associated with the small granule fraction and the microsomal fraction of these parasites. Consequently, inhibitors of mammalian HO exert toxic effects and antagonize the efficiency of Art against Leishmania. These findings may help to clarify the pharmacological mode of action of Art against Leishmania.

2. Materials and Methods

2.1. Chemicals

Acetonitrile (ACN), ascorbic acid, benzene, bovine serum albumin (BSA), caffeine, CuSO₄, ethylenediaminetetraacetic acid (EDTA), chloroform, FeCl₃, FeSO₄, formaldehyde, glucose, glycerol, HCl, HCOOH, HNO₃, K₂HPO₄, KCl, KCN, KH₂PO₄, KI, KOH, MgSO₄, Na₂HPO₄, NaCl, NaHCO₃, NaOH, n-hexane, potassium sodium tartrate, sodium acetate, sodium benzoate, sodium dithionite, sucrose, trichloroacetic acid, trifluoroacetic acid, and Tris were obtained from Merck (Darmstadt, Germany). Hemin, penicillin-streptomycin solution, resazurin, pentamidine isethionate (Pen), artemisinin (Art), L-cysteine, brain heart infusion (BHI) medium, M199 medium, alkane standard solution C₈ - C₂₀, heneicosane (C₂₁), 5,5-dimethyl-1-pyrroline N-oxide (DMPO), catalase (Cat), biliverdin (BV), and bilirubin (BR) were purchased from Sigma-Aldrich (St. Louis, Missouri, USA). Dulbecco's modified eagle medium (DMEM) was from Thermo Fisher Scientific (Waltham, MA, USA) and low-endotoxin fetal calf serum (FCS, FCS.LE.0500) from Bio&Sell (Nürnberg, Germany). Dimethyl sulfoxide (DMSO) was from Roth (Karlsruhe, Germany). Ascaridole (Asc) was synthesized according to a previously published procedure (29). Yeast extract powder was supplied by Amresco (Solon, Ohio, USA). Yeast extract medium (YEM) consisted of 20.7 g/L yeast extract powder, 1.2 g/L K₂HPO₄, 0.2 g/L KH₂PO₄, 2.9 g/L glucose, pH 7.4. Phosphate-buffered saline (PBS) contained 137 mM NaCl, 1.4 mM KH₂PO₄, 4.3 mM Na₂HPO₄, 2.7 mM KCl, pH 7.4. Pyridoxal isonicotinoyl hydrazone (PIH) was bought from Abcam Biochemicals (Cambridge, UK). Tin mesoporphyrin (SnMP), chromium mesoporphyrin (CrMP), zinc protoporphyrin (ZnPP), dihydroartemisinin (DHA), artesunate (Arsu), and deoxyartemisinin (DeoxyArt) were from Cayman Chemical (Ann Arbor, MI, USA). N-acetylcysteine (NAC) was from Alfa Aesar (Karlsruhe, Germany). Schneider's medium was from Biowest (Nuaille, France). Desferrioxamine (DFO) was from Novartis (Basel, Switzerland). Reduced nicotinamide adenine dinucleotide phosphate (NADPH) was purchased from Applichem (Darmstadt, Germany).

2.2. Cell culture

Leishmania tarentolae promastigotes (LtP) strain P10 were obtained from Jena Bioscience (Jena, Germany). Cells were used according to manufacturer's instructions up to passage 50. LtP were cultivated at 26 °C in brain heart infusion (BHI) medium (37 g/L) supplemented with 5 mg/L hemin, 25,000 U/L penicillin, and 25 mg/L streptomycin in 50 mL TubeSpin® bioreactors with gas-permeable caps under constant agitation in a tube shaker (0.05 s^-1). LtP were passaged three times a week. LtP were always used one day after passaging in the logarithmic growth phase. The number of LtP per mL was determined by measuring the optical density (OD) with a Hitachi U1100 photometer (Tokyo, Japan) at a wavelength of 600 nm in disposable polystyrene cuvettes. Before the measurements, the cell suspension was diluted 1:10 to achieve OD_{600 nm} values in the linear range (0 - 1 OD_{600 nm}). The cell-free BHI medium completed with hemin and penicillin/streptomycin was used as a reference. The cell density was calculated by the following formula of Fritsche (30): Cells (10⁶/mL) = OD_{600 nm} × 0.969 × 124 (0.969: conversion factor [g/L] dry weight, 124: 1 g dry weight/L corresponds to 124 × 10⁶ cells/mL).

The murine macrophage cell line J774A.1 (ATCC®, TIB-67™, Wesel, Germany) was cultivated in DMEM (high glucose) with 1.5 g/L NaHCO₃, 25,000 U/L penicillin, 25 mg/L streptomycin, and 10% FCS (heat-inactivated) in 50 mL TubeSpin® bioreactors. J774 cells were incubated at 37 °C and 5% CO₂ on a roller culture apparatus (5 rpm). J774 cells were passaged twice a week in a ratio of 1:10 up to passage 30. To determine the cell number, the cell suspension was diluted 1:2 with PBS. After fixation of the cells by mixing the cell suspension 1:1 with a formaldehyde solution (2%, v/v), aliquots of the resulting suspension were transferred into a standard Neubauer chamber (VWR, Austria) and counted manually using a light microscope (Motic BA310, Austria) at a magnification of 400 times.

2.3. Viability assays

A suspension of LtP in YEM/PBS (1:1 v/v, including antibiotics and 6 µM hemin or 10% FCS) or Schneider's medium (including antibiotics and 6 µM hemin or 10% FCS) with a final concentration of 2 × 10⁶ LtP/mL was prepared. Aliquots of 200 µL of this suspension were then distributed in 96-well plates (non-treated plates). Respective amounts of compound stocks (vehicle DMSO, max. 1% final concentration) were added and five 1:5 serial dilutions were performed. Control rows with medium only (negative control) and with untreated LtP (100% activity) were also loaded. The plate was incubated at 26 °C for 48 h. Then to each well 50 µL of resazurin stock solution (in PBS) was added giving a final concentration of 20 µM. After another 4 h of incubation, the fluorescence of resazurin was measured at 560 nm (excitation wavelength) and 590 nm (emission wavelength) using a plate reader (Perkin Elmer Enspire, Germany). The viability of cells in the absence and presence of drugs was calculated based on these data. Each concentration was tested in triplicates.

To assess the viability of J774 cells (1 × 10⁵ J774/mL) DMEM containing 1.5 g/L NaHCO₃, 25,000 U/L penicillin, 25 mg/L streptomycin, and 10% FCS was prepared. Two hundred µL of this mixture were pipetted in each well of a cell culture treated 96-well plate. The first row served as negative control and contained only DMEM without J774 cells. In the remaining rows the cells were incubated for 24 h at 37 °C and 5% CO₂ to allow attachment. Afterwards the medium was discarded, non-adherent cells were removed by washing each well with 200 µL PBS, and DMEM was added again to the wells. With the exception of one row that acted as positive control without the addition of any test substances, compounds were added with a starting concentration of 25 µM or 200 µM and then 5 serial dilutions of 1:5 were carried out. Each concentration was assayed in triplicates. After incubating the plates for another 24 h at 37 °C and 5% CO₂, 50 µL of resazurin solution (final concentration in the wells 20 µM) was added to each well and after 4 h of incubation, the fluorescence was measured at 560 nm excitation and 590 nm emission using a plate reader (Perkin Elmer Enspire, Germany).

For the combination experiments a suspension of LtP in YEM/PBS (1:1 v/v, including antibiotics and 6 µM hemin) was prepared (2 x 10⁶ LtP/mL). A part of the cell suspension was mixed either with 50 µM SnMP, 50 µM CrMP, 100 µM pih or 2 mM NAC.

Aliquots of 200 µL of the respective suspension were distributed in non-treated 96-well plates. Control rows with medium only and with untreated cell suspension were also loaded. Half of the remaining wells were loaded with LtP in YEM/PBS, whereas the other half was loaded with LtP in YEM/PBS containing the respective HO inhibitor, NAC or PIH.

After addition of Asc or Art five 1:5 serial dilution steps were performed. The plates were incubated at 26 °C for 48 h. Fifty µL of resazurin solution (in PBS) was added to each well leading to a final concentration of 20 µM. The fluorescence of resazurin was measured after another 4 h of incubation at 560 nm (excitation) and 590 nm (emission) using a plate reader (Perkin Elmer Enspire, Germany). Based on these data the viability of the cells was calculated regarding the presence or absence of an HO inhibitor, NAC or PIH. Each concentration was tested in triplicates.

2.4. Low temperature EPR spectroscopy of LtP

A culture of LtP (3.6 × 10⁸ parasites/mL) was split into 50 mL aliquots and centrifuged at 2000 × g, for 10 min at room temperature. The cell pellets were resuspended in 10 mL PBS (pH 7.4)/glucose (15 mM) solution. Samples were then incubated for 1 h at 26 °C under constant agitation with the respective amounts of desferrioxamine (DFO, 20 mM), Art (1 mM) or Asc (1 mM) and PBS as control. Then, cell pellets were obtained by centrifugation at 2000 × g for 10 min at room temperature and were resuspended in 300 µL PBS (pH 7.4, with addition of 10% glycerol). Aliquots of 500 µL were then transferred to syringes (Braun Omnifix F 1 mL, with luer connector removed) and immediately shock frozen in liquid N2. For measurement the frozen samples were transferred to a quartz finger dewar and then measured by EPR spectroscopy at 77 K with following instrument settings: microwave frequency 9.439 GHz; microwave power 50 mW; modulation frequency 100 kHz; modulation amplitude 8 G; time constant 0.655 s; center field 1600 G; sweep width 100 G; scan time 671 s; attenuation 2 × 10⁵; scans 1. The concentration calibration with Fe³⁺/DFO complex was performed as described previously (12). The LtP samples had protein concentrations of 50 - 70 mg/mL. For comparison, data were normalized to 60 mg protein/mL.

2.5. Metal measurements in LtP and media

Metal measurements in cells were designed to verify the uptake and the influence of HO inhibitors in LtP. LtP in growth medium (3.6 × 10⁷ LtP/mL at day 0, OD_{600 nm} = 0.3) were incubated with no inhibitor (control), 10 µM SnMP or 1 µM ZnPP for 0 - 3 days at 27 °C in an incubator. At every time point (0, 1, 2, 3 d) the volume of LtP suspension corresponding to 1 × 10⁹ cells was centrifuged for 10 min at 2000 × g and 25 °C. The respective pellet was then collected in an 1.5 mL reaction tube and frozen to 77 K for analysis. Per incubation type triplicates were prepared and analyzed. According to Price et al. (31) the frozen cell pellets were thawed, and subsequently 500 µL of concentrated nitric acid (trace metal grade) were pipetted into the tubes. The cell pellets were resuspended by repeated pipetting. Afterwards, the homogenized suspension was pipetted into digestion flasks and incubated for 16 h. Then, the samples were heated to 75 °C for 2 h to complete the digestion. Afterwards, the solution was diluted to 5 mL with ultra-pure water. Aliquots of the solution were analyzed for their metal content by inductive coupled plasma - optical emission spectrometry (ICP-OES) measurements. ICP-OES measurements were made using an Optima 5300 DV instrument (Perkin Elmer) applying routine operating conditions for multielement analysis. The metal content of samples was calculated from diluted standard (1.4 mol/L HNO₃) calibration samples with the following multielement concentrations: 0.01 ppm, 0.05 ppm, 0.1 ppm, 0.5 ppm, 1 ppm, 5 ppm, 10 ppm. Since the Sn concentrations in most samples were too low for detection by ICP-OES, aliquots of the digestion solution were further diluted by a factor of 3 and then analyzed by inductive coupled plasma - mass spectrometry (ICP-MS) (Agilent 7700x). The content of ¹²⁰Sn in the digestion solutions was calculated from diluted standard (0.5 mol/L HNO₃) Sn calibration samples (0.1 ppb, 0.5 ppb, 1 ppb, 5 ppb, 10 ppb, 50 ppb and 100 ppb) and corrected with the isotope abundancy of Sn. Results were calculated in amol/cell for the individual metals. Iron contents of media used for LtP cultures were measured by ICP-OES after 10-fold dilution and acidification targeting a HNO₃ concentration of 0.14 mol/L by using diluted standard (0.14 mol/L HNO₃) calibration samples. However, BHI and FCS were acidified with concentrated HNO₃ (volume ratio 1:1), kept overnight and then after 5-fold dilution analyzed at a targeted HNO₃ concentration of 1.14 mol/L by ICP-OES using diluted standard (1.14 mol/L HNO₃) calibration samples.

2.6. Subfractionation of LtP

The method used for subfractionation of LtP was adapted from Concepcion et al. (32). For LtP subfractionation 1450 mL of cell suspension with approximately 8 × 10⁸ cells/mL were used. The LtP suspension was centrifuged for 10 min at 1900 × g and 20 °C in a Sorvall RC5 (Dupont, Wilmington, DE, USA). The supernatant was discarded and cell pellets were resuspended in 720 mL buffer A containing 250 mM sucrose, 1 mM EDTA and 25 mM Tris, pH 7.4. LtP were again centrifuged for 10 min at 1900 × g and 20 °C. The supernatant was again discarded and cells were resuspended in 360 mL hypoosmotic lysis buffer B containing 25 mM Tris and 1 mM EDTA, pH 7.4 and incubated for 10 min at 20 °C. The suspension was distributed to 35 mL Sorvall centrifuge tubes and three freeze-thaw cycles using liquid N2 as coolant were performed. Then, the lysate was homogenized in a dounce homogenizer (30 mL) with four strokes. Afterwards 17.1 mL of a 5.26 M sucrose solution was added to the total amount of 360 mL of LtP homogenate to ensure a final isosmotic sucrose concentration of 250 mM. The homogenate was centrifuged in twelve Sorvall centrifuge tubes (35 mL) at 1000 × g for 10 min at 4 °C resulting in a nuclear and cell debris fraction (NCF). The supernatants were pooled and then centrifuged at 5000 × g for 10 min at 4 °C giving the pellets of the large granule fraction (LGF). The resulting supernatants were centrifuged at 33,000 × g for 15 min at 4 °C sedimenting the pellets of the small granule fraction (SGF). To obtain the pellets of the microsomal fraction (MCF), supernatants were centrifuged at 105,000 × g for 65 min at 4 °C in a Sorvall Ultra Pro 80 (Dupont, Wilmington, DE, USA). Aliquots of the subfractions were taken for protein determination. NCF, LGF, SGF and MCF were shock frozen in liquid N2 and then stored at -20 °C until use.

2.7. Heme oxygenase-like activity in LtP

Five mL of LtP suspension (1965 × 10⁶ LtP) per sample were centrifuged at 3000 × g for 10 min at 20 °C. The supernatant was then removed and the cell pellet was resuspended in 50 µL PBS. The cell suspension was then disintegrated by five times freezing and thawing using liquid N2 and a heat block set to 25 °C giving a 50 µL LtP homogenate.

The assay was performed by adding the following to a 2.5 mL reaction tube: 55 µL KH₂PO₄/K₂HPO₄ buffer (125 mM, pH 7.4), 10 µL rat liver supernatant (RLS, supernatant resulting from a 100,000 × g centrifugation for 1 h at 4 °C of rat liver homogenized in a buffer containing 300 mM sucrose, 20 mM Tris, 2 mM EDTA, pH 7.4, as a source of biliverdin reductase, as described in (33)), 50 µL LtP homogenate, 6.67 µM hemin, 1 mM NADPH in a total volume of 150 μL. Either tin (SnMP) or chromium (CrMP) mesoporphyrin (38 µM, 77 µM) or 2520 U/mL catalase in the absence and in the presence of 500 µM Art were added. The samples were then incubated in the dark for 30 min at 27 °C. Afterwards the formed bilirubin was extracted by adding 50 µL of caffeine solution (250 mM caffeine, 520 mM sodium benzoate, 900 mM sodium acetate in H2O), 50 µL saturated KCl solution and 1 mL benzene. The samples were vortexed for 5 min and centrifuged at 3000 × g for 10 min. The supernatants were then transferred into new reaction tubes and stored at -20 °C. For the spectrophotometric measurement the Hitachi U-3300 spectrophotometer (Tokyo, Japan) was used with following instrument settings: scan range 600 nm - 400 nm; slit: 5 nm; scan rate: 120 nm/min; photomultiplier: autogain, high resolution=off, repeats: 2; software: UV Solutions 2.2. The bilirubin (BR) content in the extracts was quantified using the differential OD450 nm - 520 nm. BR calibration curves were created by extracting 9 bilirubin standards (between 0.391 µM and 100 µM in KH₂PO₄/K₂HPO₄ buffer) in 1 mL benzene.

2.8. HPLC detection of hemin and its degradation products

For chromatographic determination of hemin degradation products by UV detection (Fig. 4) standard solutions containing 13 µM biliverdin (BV), 13 µM hemin, and 13 µM bilirubin (BR) in DMSO were analyzed by HPLC. LtP homogenates obtained from 1965 × 10⁶ LtP in 1335 µL were incubated with 7.5 µM hemin and 1.1 mM NADPH for 4 h. In addition, 8 µM BV was incubated with 1 mM NADPH in the absence and in the presence of 65 μL/mL RLS (rat liver supernatant). Furthermore, LtP subfractions (SGF/MCF, 5 mg protein) in 1335 µL were incubated with 7.5 µM hemin and 1.1 mM NADPH for 4 h in the absence and in the presence of 75 μL/mL RLS. For probing the influence of substrate and inhibitor on heme degradation (Fig. 6) Leishmania tarentolae homogenates (37 mg protein/mL) or combined subfractions (SGF/MCF) (3.7 mg protein/mL) were distributed to three 15 mL tubes each. Assay buffer (31.3 mM KH₂PO₄, 122.8 mM K₂HPO₄ in H2O, pH 7.4) was added to dilute the samples. The suspensions were supplemented with hemin only or hemin in combination with NADPH or ZnPP giving final concentrations of 7 µM hemin, 1.1 mM NADPH and 40 µM ZnPP. The three mixtures were resuspended and split to three new 15 mL tubes each (final volume of the samples: 4 mL). The tubes were mounted on a shaker (agitation: 0.2 s^-1) and incubated at 26 °C for 0, 2 or 4 h. Following the incubation the respective samples were supplemented with 200 µL of saturated KCl solution, 200 µL caffeine solution and 200 µL HCl (12 N). For extraction 1100 µL CHCl₃ was added to each sample and the mixture was vortexed for 5 min. After centrifugation at 2000 × g for 10 min at 20 °C in a Sorvall RC26 Plus centrifuge the supernatant was discarded and the lower organic phase was harvested. The solvent was evaporated using N2 gas. The residue was dissolved in 200 µL DMSO and transferred to HPLC vials with inserts (VWR, Vienna, Austria). HPLC measurements were carried out by injection of 20 µL of the respective sample onto a LiChrospher® 100 column (RP-18, 250 x 4 mm, 5 µm, Merck, Germany). Mobile phase A was ultrapure water with 5% acetonitrile (ACN) and 0.05% trifluoroacetic acid. Mobile phase C was ACN with 0.05% trifluoroacetic acid. The elution of the injected sample started with 100% A for the first 10 min. Then the amount of A was decreased to 25% whereas phase C was increased to 75% over 40 min. Phase C was held at 75% for 20 min and then further increased to 100% over 10 min. Finally, phase C was kept at 100% for 70 min. For the whole elution the flow rate was 0.4 mL/min. HPLC measurements were carried out using a Shimadzu LC20 HPLC system and data were acquired using the LCsolution software version 1.22 (Shimadzu Corporation, Kyoto, Japan). Calculations were carried out using Microsoft Excel. The measurements were performed in triplicates.

Fig. 4.

Degradation of hemin in homogenates and subfractions of Leishmania tarentolae promastigotes (LtP) followed by HPLC analysis. (A) For comparison purposes a standard solution containing 13 µM biliverdin (BV), 13 µM hemin, and 13 µM bilirubin (BR) in DMSO was measured. (B) Homogenates obtained from 1965 × 10⁶ LtP were incubated with 7.5 µM hemin and 1.1 mM NADPH for 4 h. (C) 8 µM BV were incubated with 1 mM NADPH over 30 min. (D) 8 µM BV were incubated with 65 µL/mL RLS (rat liver supernatant) and 1 mM NADPH over 30 min. Additional peaks at 46 and 108 min originate from RLS. (E) An LtP subfraction (SGF/MCF, 5 mg protein) was incubated with 7.5 µM hemin and 1.1 mM NADPH for 4 h. (F) Analog incubation to (E) but in the presence of 75 µL/mL RLS.

Fig. 6.

Degradation of hemin in homogenates and cellular subfractions (SGF/MCF) of Leishmania tarentolae promastigotes (LtP) measured by HPLC. LtP homogenates (37 mg protein/mL) (A) or LtP subfractions (3.7 mg protein/mL) (B) were incubated with 7 µM hemin and either 1.1 mM NADPH or 40 µM ZnPP in assay buffer for 0, 2 or 4 h at 26 °C. Data were normalized against the respective samples of 0 h incubation. Data represent mean ± SD (n = 3). × indicates significant differences to the time-matched LtP control at the level of p < 0.05.

2.9. HPLC/MS of hemin degradation products

LtP homogenates (50 mg protein) were supplemented with hemin (6.67 µM final concentration), NADPH (1 mM final concentration) and filled up with assay buffer (125 mM KH₂PO₄/K₂HPO₄; pH 7.4) to a final volume of 1335 µL per sample and incubated for 1 h at 25 °C under constant agitation (0.05 s^-1). Every sample was supplemented with 200 µL of saturated KCl and 200 µL 12 N HCl. The samples were mixed vigorously after addition of every compound. Each sample was extracted using 1100 µL CHCl₃ and mixing at high agitation for 5 min. The mixture was centrifuged for 10 min at 3300 × g and 20 °C in a Sorvall RC26 Plus centrifuge to support the separation of the phases. One mL of the lower organic phase was harvested and dried completely using N2 gas. The residues of the extracts were dissolved in 200 µL DMSO and transferred to HPLC vials with inserts. HPLC/MS measurements were carried out on a Waters Alliance 2695 separations module equipped with a Waters 996 PDA detector and a Waters Quattro Micro Mass Spectrometer (Waters Corporation, Milford, MA, USA) and the MassLynx (V.4.1) software was used for data acquisition. For HPLC measurements the following conditions were used: 20 µL of the respective sample was injected onto a Phenomenex Synergy MAX column (150 x 4.6 mm, 4 µm, 80 Ä, 30 °C, 0.5 mL/min) and separated using the following gradient: Solvent A was ultrapure water containing 0.1% HCOOH, and solvent B was ACN containing 0.1% HCOOH. The applied gradient was as follows: 0-10 min: 5% B; 10-50 min: 575% B; 50-70 min: 75% B; 70-130 min: 75-100% B; 130-145 min: 100% B; 145.1-170 min: 5% B. MS conditions were: electrospray ionization (ESI positive mode): m/z 100-700; sampling rate: 1019 ms; capillary: 3.5 kV; cone: 30 V; extractor 2 V; source temperature: 150 °C; desolvation temperature: 350 °C; gas flow: 600 L/h; UV detector: 190-400 nm (resolution: 1.2 nm); sampling rate: 1000 ms.

2.10. Measurement of artemisinin by GC/FID

LtP homogenates (4 mg protein) and LtP subfractions (4 mg protein) were resuspended in PBS and then supplemented with Art (1 mM) and different hemin concentrations (10 µM or 100 µM) and NADPH (2 mM) if required giving a final volume of 150 µL (LtP homogenates) or 300 µL (LtP subfractions). Samples were prepared and incubated at 26 °C for 0.5 or 24 h. For extraction of Art 400 µL (LtP homogenates) or 500 µL (LtP subfractions) of n-hexane per tube was added. Tubes were vortexed for 5 min and then centrifuged for 5 min at 20,160 × g for phase separation. Then, 350 µL (LtP homogenates) or 450 µL (LtP subfractions) were transferred to new tubes, small amounts of anhydrous MgSO4 were added and tubes were again centrifuged at 20,160 × g for 5 min at 20 °C. Finally, 300 µL (LtP homogenates) or 400 µL (LtP subfractions) were transferred to new tubes and stored at -20 °C until use. Prior to chromatographic analysis, hexane from sample tubes was evaporated under a stream of nitrogen gas. The dry residues were redissolved in 50 µL hexane and transferred to mini vial inserts (Perkin Elmer). For gas chromatographic (GC) analysis, the Perkin Elmer GC PE Autosystem XL, equipped with a flame ionization detector (FID), was used. The column was an Elite-5 MS (5% diphenyl-, 95% dimethyl-polysiloxane; low bleed) non-polar column (ID: 0.25 mm, film thickness: 0.25 µm) with a length of 15 m. Helium was the carrier gas with a flow rate of 20 cm/s (split 5:1). The injection volume was 1 uL. For analysis of Art the method established by Liu et al. (34) was adapted resulting in the following settings: Injection temperature was set to 240 °C and FID temperature was 250 °C. Initial oven temperature was set to 180 °C for 1 min. Then a temperature gradient with 2.7 °C per minute to 198 °C and thereafter another gradient with 20 °C/min to 280 °C were used. This temperature was maintained for 3 min. The total run time was 14.77 min. For each sample group the GC measurement was performed in triplicates. The software Turbochrom was used for GC control and signal processing as well as integration of chromatographic peak areas. To compare retention time data from GC/FID with GC/MS a mixture of alkane standards (C8-C21) was injected in both instruments and their retention times were used to calculate Kovats indices (Ki) of Art and its thermolysis products according to the following formula:

$$K_i={100}^{*}\left[\text{n}+(\text{N}-\text{n})\frac{{\log }_{10}\left({\text{RT}}_{\text{unknown }}\right)-{\log }_{10}\left({\text{RT}}_{\text{n}}\right)}{{\log }_{10}\left({\text{RT}}_{\text{N}}\right)-{\log }_{10}\left({\text{RT}}_{\text{n}}\right)}\right]$$

• Ki … Kovats index

• n … number of carbon atoms in the smaller n-alkane

• N … number of carbon atoms in the larger n-alkane

• RT … retention time

2.11. Measurement of artemisinin by GC/MS

In analogy to GC/FID measurements, GC/MS (mass spectrometry) measurements using an Agilent Technologies 5975 C VL MSD chromatograph (with Triple Axis detector) with a HP-5MS column (length 30 m, ID 0.25 mm, film thickness 0.25 µm) were performed. Helium was the carrier gas with a flow rate of 29.51 cm/s. The injection volume was 1 μL, split ratio 5:1. Injection temperature was set to 240 °C and MS detector data were: resulting EM voltage 1224, mass range 40 - 400, MS source 230 °C, MS quad 150 °C. Initial oven temperature was set to 180 °C for 1 min. Then a temperature gradient with 2.7 °C per minute to 198 °C and thereafter another gradient with 20 °C/min to 280 °C were used. This temperature was maintained for 3 min. MS data in the scan mode were recorded using Agilent MSD Chemstation software. Chromatographic traces and MS fragmentations were analyzed by the OpenChrome software.

2.12. Photometry of heme adducts

Hemin (8 µM) was dissolved in 1 mL mixture of PBS/DMSO (v/v 8:2), which was flushed with N2 for 20 min and placed in a cuvette (1 cm path length) with stopper for anaerobic experiments. The cuvette was placed in a diode-array photometer (Shimadzu MultiSpec 1501) and spectra from 350 to 650 nm in 30 s intervals were recorded. After 90 s hemin was reduced by adding an excess of sodium dithionite (5 mM). After 120 s either Art (16 µM) or Asc (16 µM) were added and spectral changes were measured until 1200 s. 3-D spectra were exported and for selected wavelength regions density plots were constructed by Python/Matplotlib.

2.13. EPR spin-trapping

For EPR spin-trapping 200 µL mixture (ACN/H2O 2:8 v/v) with 20 mM DMPO, 4 mM Art was prepared. Depending on the experiment then different reduction systems were added: 1 mM FeSO4, or 160 µM hemin and 336 µM cysteine (Cys), or 160 µM hemin and 400 µM ascorbate (Asco), or 160 µM hemin and 420 µM NADPH. After incubation for 10 min at 45 °C, 50 µL aliquots were aspirated into two 50 µL glass capillary micropipettes (BLAUBRAND intraMARK, BRAND) each and placed in the MD5 resonator. Then, measurements using the EMX Digital Upgrade spectrometer from Bruker (Rheinstetten, Germany) were started with following parameters: microwave frequency 9.676 GHz; microwave power 20 mW, modulation frequency 100 kHz; modulation amplitude 1 G; time constant 0.01 s; center field 3449 G; sweep width 100 G; scan time 60 s; attenuation 70 dB; scans 10. The obtained spectra of the EPR spin-trapping experiments were simulated in the software WINSIM (35) to reveal the coupling parameters by generating another curve with the parameters of the radicals and then using the “optimize” feature and fine-tuning until a sufficient fit of the generated curve with the recorded curve was reached.

2.14. Protein determination

An appropriate volume of Leishmania suspension or subfractions were resuspended in 1000 µL distilled water. Two hundred µL of ice cold 3 M trichloroacetic acid were added and the solution was incubated for 10 min at room temperature. The samples were then centrifuged (10 min, 2500 × g), the supernatant was discarded and the pellet was dissolved in 1000 µL Biuret solution (12.02 mM CuSO₄ 31.89 mM potassium sodium tartrate tetrahydrate, 30.12 mM KI, 0.2 M NaOH). The absorbance of this solution at 546 nm was determined, then a few grains of KCN were added, and after the blue color of the solution disappeared a second measurement was performed. The difference between these two absorbance values was used for quantification using a bovine serum albumin calibration curve.

2.15. Data analysis

Calculations were carried out using Microsoft Excel (Microsoft). The graphs were created using Origin 6.1 (OriginLab Corporation) or Python/Matplotlib. The half maximal inhibitory concentrations (IC₅₀) were determined from non-linear concentration-response curves using a four-parameter logistic model (36) and expressed as the mean ± standard deviation (SD). Statistical significant differences between groups of sample data were identified by Student's t-test.

3. Results

Art derivatives (Art: artemisinin, DHA: dihydroartemisinin, Arsu: artesunate, DeoxyArt: deoxyartemisinin) were studied for their influence on the viability of LtP and J774 cells (Tab. 1). Reference compounds pentamidine (Pen) and ascaridole (Asc) showed IC₅₀ values for both cell types in the range obtained already in previous studies. All artemisinin derivatives showed IC₅₀ values for the viability of LtP in the low micromolar range in contrast to the non-peroxide analog DeoxyArt. The toxicity of Art derivatives in J774 cells was low for Art itself; but higher for DHA and Arsu. Since Art showed in this system the highest ratio of IC₅₀ values in J774 and LtP cells, the study focused on the actions of Art in LtP.

Tab. 1.

Influence of reference compounds (Pen and Asc) and Art derivatives (Art, DHA, Arsu, DeoxyArt) on the viability of LtP (2 × 10⁶ LtP/mL) in PBS/YEM (1:1 v/v, including antibiotics and 6 µM hemin) and J774 macrophages (0.1 × 10⁶ J774/mL) in DMEM with 10% FCS and antibiotics. The selectivity index was calculated from IC50, J774/IC50, LtP. Data represent mean ± SD, n = 3.

Compound	IC50 LtP (µM)	IC50 J774 (µM)	IC50, J774/IC50, LtP
Pen	0.81 ± 0.17	12.27 ± 3.96	15
Asc	28.3 ± 6.7	202.13 ± 38.23	7
Art	2.96 ± 0.71	192.30 ± 41.07	65
DHA	0.71 ± 0.10	14.78 ± 6.49	20
Arsu	1.87 ± 0.27	16.9 ± 5.2	9
DeoxyArt	137 ± 66	>200	1

To explore the influence of the cellular environment on the drug action in LtP we studied the viability of LtP in the presence of EPs in different assay media (Tab. 2).

Tab. 2.

Influence of Pen as reference compound and EPs (Asc and Art) and on the viability of LtP (2 × 10⁶ LtP/mL) in different assay media containing different iron sources (hemin, FCS). LtP were incubated in these media supplemented with the test compounds for 48 h at 26 °C. Then resazurin was added and after 4 h the fluorescence intensity was determined. Data represent mean ± SD, n = 3 - 5.

Compound	Assay medium	Iron source	IC50 (µM)
Pen	YEM/PBS	6 µM hemin	0.81 ± 0.17
	YEM/PBS	10% FCS	0.36 ± 0.34
	Schneider’s	6 µM hemin	2.05 ± 1.44
	Schneider’s	10% FCS	0.78 ± 0.37
Asc	YEM/PBS	6 µM hemin	28.3 ± 6.7
	YEM/PBS	10% FCS	25.2± 5.4
	Schneider’s	6 µM hemin	40.9 ± 10.0
	Schneider’s	10% FCS	175.0 ± 83.4
Art	YEM/PBS	6 µM hemin	2.96 ± 0.71
	YEM/PBS	10% FCS	3.23 ± 1.92
	Schneider’s	6 µM hemin	1.49 ± 0.94
	Schneider’s	10% FCS	23.2 ± 15.6

Pen showed the least variation in IC₅₀ values, while EPs especially in Schneider's medium exhibited a strong variance. In contrast to Pen, EPs are for mechanistic reasons prone to premature activation by iron in media limiting the effective drug concentration acting on cells. This was strongly visible in Schneider's medium with different iron sources. In the presence of FCS the IC₅₀ values of Art was more than tenfold higher than in the presence of hemin.

Major iron sources for the LtP culture are media and supplements. Therefore, we studied the total iron concentration by ICP-OES in different media, which were used in LtP culture and related purposes (Tab. 3).

Tab. 3.

Total iron concentrations of media and supplements used for LtP culture determined by ICP-OES measurements. Data represent mean ± SD, n = 3.

Medium/Supplement	Fe [µM]
FCS (Bio&Sell, low endotoxin)	39.46 ± 0.27
Schneider’s	1.52 ± 0.25
M199	0.19 ± 0.07
DMEM	0.03 ± 0.04
PBS	0.16 ± 0.16
YEM/PBS (1:1 v/v)	12.10 ± 0.09
BHI	6.02 ± 0.09

Based on these data for YEM/PBS with 6 µM hemin and YEM/PBS with 10% FCS total iron contents around 18 µM and 16 µM are expected. Calculated values for Schneider's medium with 6 µM hemin or 10% FCS are lower with about 8 µM and 6 µM, respectively.

To identify possible forms of iron required for Art activation we studied the influence of Art on the LIP in LtP by low temperature EPR spectroscopy. At g = 4.26 EPR spectroscopy detects the Fe³⁺ portion of the chelatable iron in the cell (LIP). In the presence of DFO all Fe²⁺ of the LIP was converted to Fe³⁺ as well and the resulting signal reflects total Fe in the LIP.

Data in Fig. 1 demonstrate that Art influences the redox state of the LIP in LtP to a much lesser extent than Asc under these experimental conditions. In context with findings from our previous heme-reactivity studies (12) this suggests that Asc in LtP is more effectively cleaved than Art by low molecular iron in the LIP. However, since IC₅₀ values for Art in LtP are lower than for Asc this suggests other additional (but not exclusive) activation mechanisms for Art, such as the interaction with heme compounds.

Fig. 1.

The influence of EPs on the labile iron pool (LIP) in LtP. (A) EPR spectrum at g = 4.26 corresponding to the LIP obtained at 77K from cell pellets of LtP after incubation of LtP with 20 mM DFO for 1 h. (B) Apparent Fe³⁺ concentrations obtained from the double integral of EPR signals after incubation of LtP in buffer (LtP), 20 mM DFO (LtP + DFO), 1 mM ascaridole (LtP + Asc), and 1 mM artemisinin (LtP + Art). The apparent Fe³⁺ concentration is obtained from 35 × 10⁹ LtP/mL. Data represent mean ± SD, n = 4. * indicates significant differences to the LtP samples at the level of p < 0.05.

This observation motivated our studies on Art activation in LtP by heme compounds. At the cross road of low molecular iron and heme metabolism in mammalian cells the heme oxygenase (HO) enzymes are important. Porphyrin analogs containing metals other than iron were reported to competitively inhibit HO enzymes in mammalian cells (37). In Leishmania donovani promastigotes an HO-like activity was described in the past (25). Therefore, we tested the influence of HO inhibitors, the low molecular iron chelator PIH, and the radical scavenger NAC on the activation of EPs. HO inhibitors themselves exhibited a limited toxicity in LtP (SnMP: IC₅₀ = 64.1 ± 21.3 µM, n = 4, CrMP: IC₅₀ = 58.6 ± 16.4 µM, n = 3, ZnPP: IC₅₀ = 5.21 ± 0.57, n = 4). Therefore, we used SnMP and CrMP in subtoxic concentrations for maintaining LtP alive and included only assays in the evaluation in which the viability was compromised less than 15%. In contrast, PIH and NAC were less toxic in LtP (IC₅₀ > 600 µM, n = 3).

As clearly visible from these data (Fig. 2), the chelator of low molecular iron PIH non-selectively inhibited the action of both Asc and Art in a similar fashion. In contrast, the HO inhibitors SnMP and CrMP inhibited the antileishmanial activity of Art stronger than that of Asc. In addition, the thiol-containing radical scavenger NAC inhibited the action of Asc and even stronger that of Art. This points to the possibility that HO-like proteins or heme-importing enzymes are a prerequisite for the activation of Art in LtP. It is known that Leishmania possess a Leishmania heme response 1 (LHR1) transporter which is involved in heme/hemin import (21) and can be inhibited (38). Since the mechanism of action of mammalian HO inhibitors in Leishmania is completely unknown, the first question was to verify whether these compounds are imported into and can access their target in Leishmania at all. Since we have seen strong effects of SnMP and CrMP on the IC₅₀ value of Art we studied the enrichment of Sn arising from SnMP and the influence of SnMP and ZnPP on total Fe in LtP by ICP methods.

Fig. 2.

Relative changes in IC₅₀ values of Asc and Art in viability assays for LtP in the presence of the iron chelator PIH (100 µM), the HO inhibitors SnMP (50 µM) and CrMP (50 µM) as well as the radical scavenger NAC (2 mM). Individual bars were normalized to the IC₅₀ values of Asc and Art in the absence of these additional inhibitors (Asc IC₅₀ = 28.3 ± 6.7 µM, Art IC₅₀ = 2.96 ± 0.71 µM). Data represent means ± SD, n = 3-4.

In a first step we verified whether SnMP and ZnPP inhibit the iron import into Leishmania (Fig. 3A). Since most iron under our conditions for LtP was supplied by hemin in the cell culture medium we measured the total cellular Fe amount in LtP by ICP-OES. Data in Fig. 3A show that the iron content in LtP in the presence and absence of SnMP was almost identical over several days. ZnPP application resulted in a small decline but no total inhibition of iron import into LtP. Measurements of the Fe content in LtP indicate that SnMP and ZnPP are not preferably acting via inhibition of heme import into Leishmania. As can be seen from Fig. 3B, Sn was virtually absent in Leishmania (LtP). However, in the presence of 10 µM SnMP this metal was taken up into LtP. Since complexation of Sn in SnMP is rather strong, this suggests that SnMP and other porphyrins were taken up into LtP as well.

Fig. 3.

Total Fe and Sn contents in LtP. (A) Content of Fe in Leishmania upon incubation of LtP (36 × 10⁶ LtP/mL at day 0) in the absence or the presence of 10 µM SnMP or 1 µM ZnPP. (B) Accumulation of Sn in Leishmania upon incubation of LtP (36 × 10⁶ LtP/mL at day 0) in the presence of 10 µM SnMP. Fe content was determined by ICP-OES and Sn content by ICP-MS over three days. Data represent mean ± SD, n = 3.

This observation points to the possibility that in fact SnMP and similar HO inhibitors in Leishmania are acting on hemin/heme-degrading HO-like enzymes and that effects of SnMP on Art action could be related to this heme degradation.

In order to identify any HO-like activity and the influence of HO inhibitors in LtP, we measured the degradation of hemin and possible product formation in LtP. To apply a coupled photometric assay, commonly used to determine HO activity in mammalian cells (39) (Fig. 4, scheme), in LtP we validated this method using HPLC (Fig. 4).

In the coupled photometric assay (Fig. 4, scheme) cellular fractions are supplemented with hemin/NADPH and after incubation the reaction product bilirubin (BR) was extracted by organic solvents and analyzed photometrically at 450 nm. Our HPLC method was able to distinguish the major components hemin (RT 53 min), biliverdin (BV, RT 47 min), and BR (RT 88 min) (Fig. 4A). However, incubation of LtP homogenates with hemin and NADPH did not lead to a detectable formation of BR in our assay (Fig. 4B). Also, only tiny amounts of BV were detected, but this could be due to the high instability of this product. Addition of biliverdin reductase (BVR) for this coupled assay was proposed to convert as much as possible BV to BR (40). Further HPLC/MS (ESI+) analysis of LtP homogenates suggested the presence of more reduced tetrapyrroles in the samples. For example, a compound with m/z (M + H)+ of 593 was detected which could correspond to mesobilirubinogen (data not shown). This suggests that Leishmania can degrade heme cleavage products to highly reduced tetrapyrroles (beyond the reduction state of BR) as known from bacteria and other microorganisms (41). Therefore, we validated rat liver supernatant (RLS, post-microsomal fraction obtained by ultracentrifugation), which has no inherent HO activity, as possible source of BVR (see (33)). BV incubated with NADPH for 30 min did not result in BR formation (Fig. 4C), however, corresponding samples supplemented with RLS gave detectable amounts of BR (Fig. 4D). Therefore, we used RLS as a source of BVR in the LtP experiments. As shown in Fig. 4E, without RLS no BR formation from LtP takes place (although hemin degradation takes place also without RLS, see results in Fig. 6). However, in the presence of RLS, LtP subfractions supplemented with NADPH and hemin generated detectable amounts of BR (Fig. 4F). Therefore, in subsequent experiments the coupled photometric assay was applied to LtP in the presence of RLS or, alternatively, the NADPH-dependent hemin degradation in LtP was assessed directly. Using the coupled assay typical BR formation (marker for HO-like enzymatic activity) in LtP homogenates in the presence of RLS (66 μL/mL) yielded photometric absorptions equivalent to 69.7 ± 0.22 pmol BR (n = 2) in 30 min, while in identical experiments without LtP RLS showed only values of 2.22 ± 0.45 pmol BR (n = 2) equivalents in 30 min. This unequivocally demonstrates the absence of HO activity in the RLS itself.

As indicated in Fig. 5A, an HO-like activity in LtP was only obtained in the presence of external biliverdin reductase (from RLS). Without RLS addition to LtP (LtP - RLS) no formation of BR was detected, which fits to our findings by HPLC (Fig. 4E). Data in Fig. 5 show that LtP possess an enzymatic formation of BR at 27 °C which was strongly diminished by lowering the temperature to 0 °C. Both SnMP and CrMP inhibit this HO-like activity as one would expect from their action on mammalian HO enzymes.

Fig. 5.

Formation of bilirubin (BR, HO-like activity) in LtP homogenates obtained from 1965 × 10⁶cells (5 mg protein) supplemented with hemin (6.67 µM), NADPH (1 mM), and rat liver supernatant (RLS, 66 µL/mL, source of BVR). (A) The influence of HO inhibitors SnMP (38 µM, 77 µM) and CrMP (38 µM, 77 µM) on the HO-like activity. The activity was expressed as percentage of BR formation of the respective control sample (LtP). The absolute activity of LtP homogenates at 27 °C was 0.83 ± 0.05 pmol BR/min/mg protein and the negative control without LtP but with RLS, hemin, and NADPH was -0.03 ± 0.00 pmol BR/min/mg protein. (B) The acceleration of the HO-like activity by catalase (2520 U/mL). The absolute activity of the control (LtP) was 1.44 ± 0.17 pmol BR/min/mg protein. (C) Inhibition of the HO-like activity by Art (500 µM) in the presence of catalase (2520 U/mL). The absolute activity of the control (LtP + Cat) was 1.35 ± 0.37 pmol BR/min/mg protein. Data represent mean ± SD, n = 3. * indicates significant differences to the respective control samples at the level of p < 0.05.

HO-like activity in LtP was further increased by the presence of catalase (Fig. 5B) suggesting an inhibition of heme degradation by excessive H2O2 as observed also by others in mammalian cells (40). The EP Art, which is not degraded by catalase (42), inhibited the HO-like activity in LtP homogenates (Fig. 5C).

The presence of this HO-like activity in LtP and its sensitivity to the HO inhibitors CrMP and SnMP suggest that also the effects of the inhibitors on Art activation are linked to this HO-like enzyme activity and, therefore, that this HO-like enzyme activity is a major activator for the EP Art in LtP. To further track the subcellular location of HO-like activity and, hence, Art activation in LtP we isolated subcellular fractions of LtP: nuclear/cell debris fraction (NCF), large granule fraction (LGF), small granule fraction (SGF), and microsomal fraction (MCF). In a first step we analyzed the hemin degradation in LtP homogenates and the pooled SGF/MCF fractions (Fig. 6).

As can be seen NADPH-dependent hemin degradation takes place both in homogenates (Fig. 6A) and SGF/MCF fractions (Fig. 6B) of LtP on top of hemin degradation without addition of external NADPH, possibly due to remaining endosubstrates. In both sample types the hemin degradation was strongly inhibited by the HO inhibitor ZnPP. As known from mammalian cells Art metabolism (degradation) often involves breakage of the peroxide bond and activation to radical intermediates (43). To analyze the influence of HO-like activities on Art degradation in LtP we validated a GC method (Fig. 7) to quantify Art in LtP homogenates and subcellular fractions.

Fig. 7.

Analysis of artemisinin (Art) by GC/FID. (A) Chromatogram obtained upon injection of a commercial Art sample. Art is partially decomposed upon evaporation in the GC injector giving peaks at 4.4 min (P1), 5.9 min (P2), 7.8 min (P3, Ki 2178), and 8.3 min (P4, Ki 2195). (B) In analogy this Art sample was studied under the same temperature regime in GC/MS and the two most abundant peaks P3 (11.1 min, Ki 2129) and P4 (12.0 min, Ki 2217) were identified as a thermal rearrangement product of Art (Art-TP1) and Art itself, respectively. (C) A calibration curve using the sum of GC/FID peak areas of P3 and P4 vs. the Art concentration in the sample (1 µL injection volume) gave a linear correlation (r = 0.994).

The great advantage of the GC/FID method over HPLC methods was that it allows analysis of Art in a very complex matrix of LtP. However, the down side was that Art partially decomposes upon evaporation during analysis, giving 4 different peaks (Fig. 7A). GC/MS analysis revealed that P4 was Art and P3 was a thermal rearrangement product of Art (Fig. 7B). As suggested by other publications (34,44,45) the sum of the GC/FID peak areas of P3 and P4 resulted in a linear correlation with the amounts of Art injected, which was, therefore, used for further analysis.

The degradation of Art in LtP homogenates was studied (Fig. 8) using 10 µM (A, B) and 100 µM (C, D) hemin as substrate in the absence (A, C) and the presence of LtP homogenates (B, D). While a partial NADPH-dependent Art degradation was also observed in the absence of LtP (Fig. 8C), it was most pronounced in LtP homogenates supplemented with NADPH and 100 µM hemin (Fig. 8D), suggesting an enzymatic hemin- and NADPH-dependent degradation of this drug.

Fig. 8.

Degradation of Art by LtP homogenates. Art (1 mM) was incubated in PBS (A, C) or in PBS with LtP homogenates (4 mg protein; B, D), with co-substrates NADPH (2 mM) and 10 µM (A, B) or 100 µM (C, D) hemin. Extracts were analyzed for the amount of remaining Art by GC/FID after 0.5 and 24 h incubation time. All peak areas were normalized to the amount of Art detected in PBS after 0.5 h, which was set to 100%. Data represent mean ± SD of duplicate incubations. × indicates significant differences to the respective PBS/Art sample at the level of p < 0.05.

Likewise, we analyzed the distribution of HO-like activities and Art decomposition in LtP subfractions (Fig. 9). HO-like activities appeared to be higher in the MCF and SGF fractions (Fig. 9A), suggesting preferred location of this enzyme activity within the SGF/MCF fractions, similarly to what is known from mammalian HO enzyme. Following the Art decomposition in these subfractions we observed highest rates in SGF and MCF. Again, Art degradation was increased at 100 µM vs. 10 µM hemin. The observation of highest HO-like activities and highest Art degradation in MCF/SGF suggested a causal relationship with the presence of an HO-like activity.

Fig. 9.

HO-like activities (A) and degradation of Art (B) in LtP subfractions (nuclear/cell debris fraction (NCF), large granule fraction (LGF), small granule fraction (SGF), and microsomal fraction (MCF)). HO-like activities in LtP subfractions (4 mg protein) were analyzed by photometry using 1 mM NADPH, 33 µM hemin, 20 µL RLS and 800 U/mL catalase (total volume 150 µL) after 30 min incubation. Data represent mean ± SD, n = 3. In analogy to Fig. 8 the degradation of Art (1 mM) in subfractions (4 mg protein) was assessed by GC/FID after 0.5 h incubation in the presence of 2 mM NADPH and 10 or 100 µM hemin (total volume 150 µL). Data represent mean ± SD, n = 2. × indicates significant differences to the respective PBS/Art sample at the level of p < 0.05.

In addition to these enzyme-linked activities, in most experiments also a portion of non-LtP dependent decomposition of Art was observed, raising the question how Art is activated in conjunction with hemin. To approach this question we studied the reaction of Art with heme photometrically (Fig. 10).

Fig. 10.

Reaction of endoperoxides (EPs) with heme. The graphs show photometric time scans in the range of 380 - 435 nm (upper part) and 435 - 500 nm (lower part). Hemin (8 µM in PBS with 20% DMSO, λ_max = 400 nm) was reduced after 90 s to heme (λ_max 418 nm) by adding an excess of sodium dithionite (5 mM). After 120 s the EPs (A: 16 µM Art, B: 16 µM Asc) were added. Exclusively Art, but not Asc, gave rise to a new absorption peak at λ_max 474 nm, which decays slowly over time. Time scans represent examples of triplicate measurements.

As known from previous findings Art activation required reduction. Therefore, we reduced hemin (λ_max = 400 nm) by sodium dithionite causing an immediate spectral shift to λ_max = 418 nm due to the formation of heme (Fig. 10A, B, upper graphs). Then, we added anaerobically either Art or Asc. For Art this resulted in the appearance of a new peak at λ_max = 474 nm, which indicates adduct formation, in contrast to Asc which caused no spectral change. The adduct of Art with heme is obviously the starting point of Art degradation. To study this degradation we used the EPR spin-trapping technique (Fig. 11).

Fig. 11.

Activation of Art by hemin and relevant biological reductants assessed by EPR spin-trapping. The spin trap DMPO (20 mM) and Art (4 mM) dissolved in ACN/H2O (2:8 (A) or 1:1 (BD)) were incubated with (A) 1 mM FeSO4, (B) 160 µM hemin and 336 µM cysteine (Cys), (C) 160 µM hemin and 400 µM ascorbate (Asco), (D) 160 µM hemin and 420 µM NADPH for 10 min at 45 °C. Then, 50 µL aliquots were aspirated into two 50 µL glass capillaries and measured in the MD5 EPR resonator. Resulting EPR spectra in the presence and in the absence of Art are shown in the right and left column in black, respectively. Spectra represent examples of triplicate measurements. For all spectra in which a spin adduct was detected, spectra were simulated using WinSIM and are shown below the respective experimental spectra in red.

Unfortunately, we could not measure the radical formation under conditions used in Fig. 10 because dithionite-derived SO₂ ^•- radicals (a_N = 14.46 G, a_H = 15.74 G; DMPO/SO2'^-, spectra not shown) made detection of other spin adducts impossible. Therefore, we studied physiologically more relevant reductants. During EPR spin-trapping experiments in the presence of FeSO4 Art generated a DMPO spin adduct with a_N = 14.98 G / a_H = 18.41 G (66%, DMPO/•CO acyl radical) and a second adduct with a_N = 16.24 G / a_H = 22.21 G (33%, DMPO/•C). In the absence of Art FeSO4 did not generate detectable amounts of spin adducts. Experiments with hemin in the presence of Cys gave overlapping signals from spin adducts of a_N = 14.88 G / a_H = 17.69 G (84%, DMPO/•CO acyl radical) and a_N = 13.44 G / a_H = 15.29 G (16%, probably DMPO/•SH). The corresponding experiment without Art but in the presence of hemin and Cys showed no spin adducts. Using ascorbate (Asco) as reductant of hemin in the presence of Art, spin adducts with a_N = 14.84 G / a_H = 17.86 G (47%, DMPO/•CO acyl radical) and a_N = 14.46 G / a_H = 13.53 G (53%) were observed. However, in the absence of Art spin adducts with a_N = 15.74 G / a_H = 22.24 G (95%, DMPO/•C) and an ascorbyl radical (a_H = 1.70 G, 5%) were recorded. NADPH as reductant for hemin generated from Art spin adducts with a_N = 14.98 G / a_H = 17.94 G (29%, DMPO/•CO acyl radical) and a_N = 13.78 G / a_H = 15.10 G (71%). In the absence of Art in this experiment an adduct with a_N = 13.88 G / a_H = 14.86 G (100%) was observed. Although unequivocal assignments of all different spin adducts was difficult, it is obvious that DMPO/^'CO acyl radical adducts were only detected in the presence of Art but never in its absence. This suggests that these radical intermediates are formed from an interaction of Art with heme irrespective of the reductant for hemin leading to heme, although with different quantitative efficiency. This demonstrates that hemin/heme plays a major role for enzymatic and chemical activation of Art in LtP.

4. Discussion

Art is a natural EP which is extracted from the leaves of the plant Artemisia annua (46). Subsequent studies demonstrated its efficiency against malaria caused by Plasmodium falciparum and related species. Nowadays, Art and related compounds, such as artesunate (Arsu) and artemether, are approved as antimalarial drugs in several countries. In addition to the first generation of Art derivatives, second-generation synthetic Art-related compounds, such as trioxolanes, have been developed, which, however, are not yet in widespread use.

Several factors contribute to the action of Art against plasmodia. Art itself is lipophilic and is enriched in erythrocytes infected with plasmodia by a carrier-mediated transport in addition to its passive diffusion (47,48). Mechanistic studies about Art suggest that it exerts its actions against plasmodia by generating free radicals, which change the cellular redox balance causing parasites' death. It was shown that in plasmodia the conversion of Art to free radicals is essentially caused by cellular heme and iron (46). This notion suggests that Art-derived radicals rather randomly attack biomolecules possibly with a slight preference for heme-containing proteins. However, it has also been shown that Art in plasmodia specifically binds to a sarco-/endoplasmatic reticulum calcium ATPase (SERCA) ortholog (PfATP6) (49). While the action of Art at PfATP6 is iron-dependent, the non-peroxide analog thapsigargin inhibits this enzyme in a non-iron-dependent fashion in plasmodia. For Art dimers it has been shown that they likewise exhibit an iron-dependent inhibition of SERCA in cancer cells triggering thereby endoplasmatic reticulum (ER) stress (50). These observations suggest that activation of Art in plasmodia by heme and iron is superimposed by site-specific effects, which are mediated by lipophilicity and molecular structure of Art, which cause its binding to specific proteins.

In addition to antimalarial and antischistosomal properties of Art (10), at least initial evidence exists for Art activities against certain cancer cell lines (3), viruses (51), and other protozoal organisms. Occasionally, also antileishmanial activities of Art in vitro and in vivo were reported (52). While IC₅₀ values for Art against Plasmodium falciparum are much lower in comparison to those against Leishmania donovani, in both species the IC₅₀ values of the EP Art are lower than IC₅₀ values of the non-endoperoxide analog DeoxyArt (53).

A significant number of studies have demonstrated in vitro and in vivo in rodent models that Art is able to control Leishmania infection and that it has a rather high safety margin compared to other drugs in the mammalian organism (52).

A systematic study of trioxolanes, including Art, regarding their antileishmanial activity was performed by Cortes et al. (54). They observed that Art had a limited selectivity index for Leishmania infantum intracellular amastigotes versus THP-1 macrophages of 2.43, while some synthetic trioxolanes studied in their system achieved selectivity indices higher than 50. This suggests that respective synthetic trioxolane EPs should be considered as an alternative treatment option to approved antileishmanial drugs. In vivo experiments performed in a BALB/c mouse model of visceral leishmaniasis by Sen et al. (9) demonstrated that Art orally administered at 25 mg/kg body weight can reduce the splenic parasite burden slightly more than sodium antimony gluconate (SAG) administered at 20 mg/kg body weight. Therefore, the potential benefit of Art could be its alternative use to SAG to prevent further resistance development to SAG. Numerous other studies were performed with Art and related compounds in vitro using Leishmania amastigotes and promastigotes as well as in vivo with experimental animal models (for an overview see (52)). Rather surprising is the fact that even in standardized promastigote systems IC₅₀ values for Art range from low micromolar to tens or even hundred micromolar concentrations in different studies, suggesting rather variable results in these systems.

Our results (Tab. 1) indicate that Art is effective against LtP with IC₅₀ values in the low micromolar range, while in J774 macrophages higher concentrations around 200 µM were required to achieve IC₅₀ conditions. Other Art derivatives are similarly effective in LtP but show some toxicity in J774 macrophages. As expected, DeoxyArt without EP group has only little antileishmanial effect in LtP. Viability assays using different assay media and different iron sources (Tab. 2) demonstrate the influence of assay conditions on IC₅₀ values. While IC₅₀ values for the less lipophilic and non-redox active Pen are not strongly dependent on assay media and iron source, the IC₅₀ values for EPs are increased in Schneider's medium vs. YEM/PBS and specifically in the presence of FCS for EPs. The background of this observation could be that FCS binds lipophilic Art stronger than the less lipophilic Pen thereby reducing the effective drug concentration. In addition, Schneider's medium contains less total iron than YEM/PBS (Tab. 3), which is an important trigger for the degradation/activation of EPs in Leishmania. These findings suggest that viability assays for Art and related compounds may not always be strictly comparable between different studies except if they are performed under absolute identical conditions. From biomimetic studies of Art's antimalarial actions it was concluded that in principle both low molecular Fe²⁺ and heme can trigger Art activation (55). Our current experiments (Fig. 1) suggest that low molecular iron in the LIP of Leishmania is more influenced by Asc than by Art at identical EP concentrations and, therefore, alternative triggers for Art activation in Leishmania seem to be likely.

If one considers the influence of additional inhibitors (PIH, SnMP, CrMP, NAC) on the antileishmanial effect of Asc and Art (Fig. 2) some differences become visible. While the lipophilic iron chelator PIH boosts IC₅₀ values for both compounds to a similar extent, heme oxygenase (HO) inhibitors (metalloporphyrins) preferably increase IC₅₀ values for Art. Likewise, the IC₅₀-increasing effect of NAC was stronger for Art than for Asc (Fig. 2). This could be attributed to the fact that Asc can exert its antileishmanial effect not only via radical formation but also by formation of Michael adducts (12,59).

Metalloporphyrins were tested in photodynamic therapy against cutaneous leishmaniasis (56). For plasmodia it was shown that some metalloporphyrins have a moderate activity against these parasites in the absence of light irradiation (57). In our applications against LtP in the absence of light irradiation only ZnPP showed a stronger toxicity against Leishmania. That these metalloporphyrins do not inhibit heme import we concluded from the observation that they did not decrease total amount of iron per cell (Fig. 3A). In addition, we have shown that Sn from SnMP was taken up by Leishmania (Fig. 3B), suggesting that these metalloporphyrins (similarly to mammalian cells) can inhibit intracellular heme-binding enzymes. Fang et al. (58) have observed that in cancer cells the chemotherapeutic response to some organic peroxides and other anticancer drugs is enhanced by HO inhibitors, such as polyethylenglycol-functionalized ZnPP. They suggest an inhibition of the antioxidant function of HO in these cells. The publication of Stockwin et al. described that the artemisinin-dimer toxicity for cancer cells was enhanced by the HO inhibitor Sn protoporphyrin IX (SnPPIX), but inhibited by the HO inducer Co protoporphyrin IX (CoPPIX) (50).

However, the action of these metalloporphyrins in Leishmania is not well understood so far. The effects of the HO inhibitors on the Art action in Leishmania prompted us to study the heme degradation in conjunction with Art degradation in Leishmania in more detail. Recent publications about the iron acquisition machinery in Leishmania suggest that there are independent mechanisms for the acquisition of low molecular iron and iron bound in heme molecules. In Leishmania low molecular Fe²⁺ is taken up by the Leishmania iron transporter 1 (LIT1) (18,60) from the external medium (phagolysosome). Simultaneously, Fe²⁺ is generated from low molecular Fe³⁺ by Leishmania ferric reductase 1 (LFR1) (19,20) at the outer face of the Leishmania membrane. As an alternative source of iron, Leishmania are able to take up porphyrin/iron complexes (hemin/heme) via the LHR1 transporter (21). Depending on the type of iron supply, one can expect that intramacrophagal Leishmania amastigotes can use alternatively low molecular iron complexes and/or hemin/heme as iron source. So far only one publication exists describing iron liberation from heme via an HO activity (25). In trypanosomatids the formation of verdoheme and biliverdin was identified in epimastigotes (61), suggesting a heme degradation pathway via an HO-like protein. Hemin was shown to be an essential growth factor in LtP cultures (62). It has even been demonstrated that Leishmania can acquire iron by internalization of hemoglobin and its subsequent breakdown (26). In contrast to other protozoal parasites, Leishmania can use later precursors of heme synthesis because their DNA codes for enzymes that catalyze the last steps of heme synthesis as well (24). Therefore, heme imported into Leishmania may serve both for biosynthetic purposes and acquisition of iron (26). The ability to degrade heme was demonstrated for several parasites in order to benefit from this alternative iron source (24).

Our results indicate that degradation of hemin in Leishmania does not stop at BV or BR state (Fig. 4E), but possibly proceeds to higher reduced tetrapyrrole products, as it is known from bacteria (41). Only in the presence of an external source of BVR, accumulation of BR as in mammalian cells was observed (Fig. 4). This indicates the existence of a heme degradation product, which can serve as a substrate for BVR. The BVR may possibly interfere with the endogenous reduction of tetrapyrroles at the point of a verdoheme intermediate in Leishmania. The HO-like activity, which we measured in LtP is sensitive to known HO inhibitors (Fig. 5A), is enhanced by removal of H2O2 (Fig. 5B) and partially inhibited by Art (Fig. 5C). We observed that hemin degradation in Leishmania is dependent on NADPH and is partially inhibited by the metalloporphyrin ZnPP in LtP homogenates (Fig. 6A) and LtP SGF/MCF subfractions (Fig. 6B). This supports our idea about an HO-like enzyme system in Leishmania. The degradation of Art as studied by GC (Fig. 7) was preferably observed in the same SGF/MCF subfractions of LtP. The results so far suggested that there is a causal link between Art activation and heme degradation in Leishmania. In biomimetic models of Art actions in plasmodia Fe²⁺ was identified as a major trigger of Art activation (53,63).

Alternatively, heme as a trigger of Art decomposition was proposed (64). Quantum chemical calculations for Art/heme complexes according to density functional theory (DFT) support the concept that electron transfer from heme iron to the EP bridge of Art is important for its activation (65). Metabolism of Art in mammalian cells is partially mediated by cytochrome P450 enzymes of phase 1, and glucuronidation takes place in phase 2 metabolism. These metabolic steps, including hydroxylation in phase 1 reactions, may proceed with or without cleavage of the EP bridge (66,67). Therefore, at least metabolism of Art in the mammalian organism does not always include homolytic EP cleavage and subsequent radical formation. In the non-enzymatic reaction of Art with heme, it was shown that alkylated heme derivatives of Art are formed, which in part have redox properties different from the original heme molecule (68). Corresponding Art/heme adducts were also observed in vivo in a mouse model experimentally infected with plasmodia under Art treatment (69).

In spite of these detailed findings, the role of heme in the activation of Art in plasmodia is still partially controversial: whether it is a target or/and activator for Art (70). In addition to heme and hemoglobin, a role of hemin in the activation of Art has been proposed (71). We and others (72,73) observed an adduct of Art with heme in photometric measurements with maximal absorption around 474 nm, which in our experiments was only obtained by reaction of heme with Art but not in the reaction of heme with Asc (Fig. 10). Other reports suggest that this intermediate might be an alkylation product of the porphyrin ring from the reaction with Art-derived radicals (72,73).

For Leishmania so far no such detailed knowledge about the activation mechanism of Art in these protozoal organisms exists. Hemin degradation in LtP homogenates and LtP subfractions is accelerated by NADPH and retarded by the HO inhibitor ZnPP (Fig. 6). In parallel, Art degradation in LtP homogenates is highest in the presence of NADPH and hemin (Fig. 8B, D). In addition, Art degradation in Leishmania increases at higher hemin concentrations (Fig. 8B, D). In subfractions of Leishmania the HO-like activity was highest in SGF and MCF (Fig. 9A). This observation is compatible with the fact that HO enzymes in mammalian cells are also located in the microsomal fractions (37). The attempt to find homologs of mammalian HOX1 and HOX2 genes in the Leishmania genome failed so far, which is, however, not extremely surprising taking into account the large difference in the phylogenetic development between mammals and Leishmania. Identification of the genes in Leishmania involved in heme degradation is still ongoing. In our pharmacological study it is remarkable that Art degradation in Leishmania also peaks in the SGF and MCF subfractions of Leishmania (Fig. 9D, C).

Besides the enzymatic degradation of Art in Leishmania possibly linked to an HO-like activity, also the direct interaction of hemin with Art in the presence of reductants was studied by EPR spin-trapping (Fig. 11). In addition to various reductant-specific spin adducts in this system, in all samples with Art DMPO/^•OC acyl adducts with aN around 15 G and aH around 18 G were observed (Fig. 11B-D), which were also present when Art was activated by low molecular Fe²⁺(Fig. 11A). This suggests that in spite of very different systems reducing hemin to heme, primary radicals generated from Art reduction are at least similar. Furthermore, this demonstrates that hemin in conjunction with various cellular reductants can trigger Art degradation in Leishmania. In a previous study we demonstrated that Art targets mitochondrial electron transfer only at rather high concentrations, while, in contrary, cellular ATP levels in Leishmania were already decreased at therapeutic Art concentrations (11). For Art dimers it was shown that they cause ER stress in cancer cells (50). Likewise, disturbance of heme degradation in Leishmania by Art (as we observed in the current study) could be a trigger for ER stress in Leishmania and an increased ATP demand, which was observed in a previous study on mitochondrial Art effects (11).

In conclusion (Fig. 12), our results demonstrate that Art degradation in Leishmania could be linked to the presence of hemin/heme compounds and their interaction with cellular reductants including enzymatic pathways of heme degradation in the small granule fraction/microsomal fraction of Leishmania. The strong preference of Art for heme-triggered activation suggests a possible effectivity of this drug in Leishmania-infected macrophages involved in erythrophagocytosis, occurring in the spleen, which is one of the most affected organs in patients suffering from visceral leishmaniasis. Of course, further in vivo experiments with Art and clinical studies are required to prove the effectiveness of this drug in the different forms of leishmaniasis.

Fig. 12.

The different activation pathways for artemisinin (Art) in LtP. Major iron uptake pathways involve LFR1 (Leishmania ferric reductase 1) and LIT1 (Leishmania iron transporter 1) for low molecular iron in the labile iron pool (LIP) and LHR1 (Leishmania heme response 1 - transporter) for hemin. While Art decomposition is only partially influenced by the LIP, it is strongly dependent on hemin-derived products containing Fe²⁺, such as heme or verdoheme, which are generated by chemical reduction or enzymatic reduction by an HO-like enzyme.

Abbreviations

ACN

acetonitrile

Arsu

artesunate

Art

artemisinin

Asc

ascaridole

Asco

ascorbate

BHI

brain heart infusion

BR

bilirubin

BV

biliverdin

BVR

biliverdin reductase

Cat

catalase

CrMP

chromium mesoporphyrin

Cys

cysteine

DeoxyArt

deoxyartemisinin

DFO

desferrioxamine

DHA

dihydroartemisinin

DMEM

Dulbecco’s modified eagle medium

DMPO

5,5-dimethyl-pyrroline-N-oxide

DMSO

dimethyl sulfoxide

EDTA

ethylenediaminetetraacetic acid

EP

endoperoxide

EPR

electron paramagnetic resonance

ER

endoplasmatic reticulum

ESI

electrospray ionization

FCS

fetal calf serum

FID

flame ionization detection

GC

gas chromatography

HO

heme oxygenase

HPLC

high performance liquid chromatography

IC₅₀

half maximal inhibitory concentration

ICP-OES

inductively coupled plasma-optical emission spectrometry

J774

J774A.1 murine macrophage cell line

LFR1

Leishmania ferric reductase 1

LGF

large granule fraction

LHR1

Leishmania heme response transporter 1

LIP

labile iron pool

LIT1

Leishmania iron transporter 1

LtP

Leishmania tarentolae promastigotes

MCF

microsomal fraction

MS

mass spectrometry

NAC

N-acetyl cysteine

NADPH

reduced nicotinamide adenine dinucleotide phosphate

NCF

nuclear and cell debris fraction

OD

optical density

PBS

phosphate-buffered saline

PIH

pyridoxal isonicotinoyl hydrazone

RLS

rat liver supernatant

RT

retention time

SAG

sodium antimony gluconate

SD

standard deviation

SERCA

sarco-/endoplasmatic reticulum calcium ATPase

SGF

small granule fraction

SnMP

tin mesoporphyrin

YEM

yeast extract medium

ZnPP

zinc protoporphyrin