| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* Centro de Investigaciones Biológicas, Consejo Superior de Investigaciones Cientificas, Madrid, Spain; and
Department of Oral Biochemistry, Academic Centre of Dentistry, Amsterdam, The Netherlands
1Correspondence: Centro de Investigaciones Biológicas, CSIC, Ramiro de Maeztu 9, 28040-Madrid, Spain. E-mail: luis.rivas{at}cib.csic.es
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: bioenergetics • ATP synthase • Trypanosomatidae • histidine-rich peptide
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. Despite their wide range of primary and secondary structures, most of them share a strong cationic character and the adoption of amphipathic structures on their interaction with the plasma membrane of the pathogen (2)
. This promotes the initial electrostatic interaction with the anionic phospholipids exposed at the outer leaflet of the plasma membrane. This trait is exclusive of prokaryotes and lower eukaryotes, as in higher eukaryotes this class of phospholipids is confined to the cytoplasmic face of the bilayer, precluding their interaction with AMPs. For most AMPs, this initial interaction is ensued by massive accumulation and insertion into the plasma membrane, disrupting its integrity as a permeability barrier, leading to a fast and lethal disruption of internal homeostasis and eventually to the death of the pathogen (3)
. This scenario is however being increasingly challenged by other AMPs the lethal action of which is postulated to involve the recognition of intracellular targets (reviewed in ref. 4
). These peptides would translocate across the plasma membrane, distorting its structure in a transient, nonlethal manner, and once inside the cell would recognize intracellular targets such as DNA (buforin), DnaK (drosocin and phyrrhocorrycin; ref. 4
), mitochondria [histatin 5 (Hst5); ref. 5
], and others that remain unknown (6)
.
The mitochondrion has preserved through evolution its primeval functionality as a potent energy generating organelle, although the molecular machinery in charge of its role shows a broad divergence among different phyla (7)
. The protozoan Kinetoplastida are, in evolutionary terms, the earliest branch of eukaryotes with a true functional mitochondrion (8)
. Among its members are the species of the genus Leishmania, many of which act as the causative agents of leishmaniasis, a devastating human disease with a wide variety of clinical manifestations (9)
. Leishmania spp. are digenetic protozoa, which cycle between the extracellular promastigote stage in the insect vector and the intracellular amastigote stage, responsible for the pathology in vertebrates, which dwells inside a parasitophorous vacuole in the mononuclear phagocytes (10)
. Chemotherapy is the only available method to fight this disease, and it is threatened by the rise of resistant parasites (9)
.
Aside from the clinical interest in the development of new leishmanicidal drugs, there are three additional reasons to explore the susceptibility of Leishmania to Hst5, a histidine-rich AMP present in the saliva of higher primates (11)
that displays the highest candidacidal activity among human histatins (12)
. First, the fact that drugs with both antifungal and leishmanicidal activities, such as some phenalenones, affect mitochondria (13)
suggests that the two microorganisms might share this target to a certain extent. Mitochondrial targeting in Candida spp., together with low cytotoxicity for mammalian cells, would thus make histatins attractive leishmanicidal drug candidates. Second, despite the above similarities, Candida and Leishmania have striking differences in bioenergetic metabolism. In Leishmania, it is based mainly on oxidative phosphorylation (14
, 15)
, hence lacking the partial switch between fermentative and nonfermentative pathways of Candida (16)
. Finally, the Leishmania amastigote is adapted to thrive in acidic pHs (17)
where protonation of the histidine residues of Hst5 enhances its cationic properties and, presumably, its antimicrobial action.
To gain some insight on the antimicrobial activity of Hst5 against microorganisms with the most primitive functional mitochondria, as well as to explore the feasibility of Hst5 as a potential leishmanicidal lead, we have studied the activity of Hst5 and its synthetic analog Dhvar4 on the different forms of the parasite. The all-D form of Hst5 (D-Hst5) was also assayed to explore chiral requirements in the key steps of the lethal mechanism of L-Hst5. Our results provide evidence for 1) a strong leishmanicidal activity for all three peptides, ruling out stereospecific recognition for Hst5; and 2) a preferential accumulation into the parasite mitochondrion. For Hst5, a faulty oxidative phosphorylation, mainly due to an inhibition of mitochondrial F1F0-ATPase, and in a minor extent the extracellular ATP efflux, leads to parasite death.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, 18
, 19)
.
Chemicals
L-Hst5 (DSHAKRHHGYKRKFHEKHHSHRGY), its D-enantiomer, and Dhvar4 (KRLFKKLLFSLRKY) were synthesized by Fmoc-chemistry, purified by preparative reverse phase (RP) -HPLC, and reanalyzed by RP-HPLC plus capillary zone electrophoresis. The purity of the peptides was at least 95%, and the authenticity of the peptides was confirmed by ion trap mass spectrometry (20)
. Labeling of the peptides with fluorescein-5-isothiocyanate (FITC; Molecular Probes Europe BV, Leiden, NL) was performed as described previously (5)
, resulting in a molar FITC to peptide ratio of 
0.1.
Unless otherwise stated, the reagents were purchased from Sigma-España (Madrid, Spain). Bisoxonol, [bis-(1,3-diethylthiobarbituric) trimethine oxonol], SYTOX Green, rhodamine 123 [2-(6-amino-3-imino-3H-xanthen-9-yl)benzoic acid methyl ester, chloride], DMNPE-luciferin [D-luciferin, 1-(4,5-dimethoxy-2-nitrophenyl) ethyl ester], MitoTracker Red CMXRos, and H2DCFDA (2',7'-dichlorodihydrofluorescein diacetate) were obtained from Invitrogen (Barcelona, Spain).
Fluorescence and luminescence measurements were recorded in a Polarstar Galaxy microplate reader (BMG Labotechnologies, Offenburg, Germany), fitted with the corresponding optical setting, unless otherwise stated.
Cell proliferation measurements
Parasites were harvested at late exponential phase, washed twice in sorbitol buffer (280 mM D-sorbitol, 4.0 mM Na2HPO4, 1.0 mM KCl, 4.8 mM NaHCO3, and 10 mM D-glucose, pH 7.2) at 4°C, and resuspended in the same buffer at a final concentration of 2 x 107 cells/ml. Unless otherwise stated, these conditions were maintained for the rest of the experiments.
Aliquots of 100 µl of the parasite suspension were incubated with the corresponding peptides for 4 h, either at 25 or 32°C for promastigotes or amastigotes, respectively, washed with 1 ml of sorbitol buffer at 4°C, and resuspended in 1 ml of their respective growth medium. Parasites were allowed to proliferate for 48 or 96 h for promastigotes and amastigotes, respectively. Cell proliferation was subsequently measured by reduction of 3-(4,5-dimethylthiazol-2-yl) 2,5-diphenyltetrazolium bromide (MTT; 0.5 mg/ml), followed by solubilization with 5% (w/v) SDS and readout in a 680 Bio-Rad microplate reader, equipped with a 600 nm filter (Bio-Rad, Madrid, Spain) (18)
.
The induction of apoptosis by Hst5 was monitored by analysis of the sub-G1 peak in parasites stained with propidium iodide in a Coulter XL EPICS cytofluorometer (Beckman Coulter, Miami, FL, USA) (21)
, using 15 µM miltefosine as a positive control.
Changes in plasma membrane potential
Changes in plasma membrane potential were monitored by the increase of bisoxonol fluorescence, a potential-sensitive anionic dye, after its insertion into the parasite membrane, once the cell becomes depolarized. Assays were performed under the standard conditions using as incubation medium 0.2 µM bisoxonol in sorbitol buffer. Fluorescence changes were monitored using excitation and emission filters of 544 and 584 nm, respectively. Maximal depolarization was considered as that achieved with the cecropin A-melittin hybrid peptide CA(1–8)M(1–18) at 2.5 µM (22)
.
Plasma membrane permeabilization
The procedure described previously (22)
was followed, except for the use of sorbitol buffer as incubation medium. Briefly, peptides were added to the parasite suspension (2x107 promastigotes/ml) preincubated with 1 µM SYTOX Green. The increase in fluorescence due to binding of the dye to intracellular nucleic acids was monitored using 485 and 520 nm filters for excitation and emission wavelengths, respectively. Permeation achieved by 0.1% Triton X-100 (TX-100) was considered as maximal.
Electron microscopy
Promastigotes were incubated with the peptides according to the standard procedure, washed twice in PBS, and fixed in 5% (w/v) glutaraldehyde in the same medium. Afterward, parasites were included with 2.5% (w/v) OsO4 for 1 h, gradually dehydrated in ethanol [30, 50, 70, 90, and 100% (v/v); 30 min each] and propylene oxide (1 h), embedded in Epon 812 resin, and observed in a Jeol-1230 electron microscope (Jeol Ltd., Akishima, Japan) as described (22)
.
Confocal laser microscopy
The intracellular distribution of fluoresceinated peptides was ascertained by a double-labeling experiment, using MitoTracker Red, a cell-permeable mitochondrion-selective dye. Promastigotes of L. donovani and amastigotes of L. pifanoi were loaded with 0.05 µM MitoTracker Red for 10 min at 25 and 32°C, respectively, washed twice, and incubated in darkness with the peptides for 4 h at their respective growth temperatures. Finally, unbound peptide was removed by washing twice with 1.5 ml of sorbitol buffer, and parasites were prepared with Dako fluorescence mounting medium (Dako Cytomation, Carpinteria, CA, USA) for observation. Confocal fluorescence images were obtained on a Leica TCS-SP2-AOBS-UV ultraspectral confocal microscope (Leica Microsystems, Heidelberg, Germany) Excitation and emission wavelengths for MitoTracker Red were 543 and 599 nm, respectively, and 488 and 519 nm for the labeled peptides.
Intracellular accumulation of FITC-Hst5
Aliquots of 100 µl of the parasite suspension under the standard conditions were incubated with the fluoresceinated D- or L-Hst5 for 4 h at 25 or 32°C for promastigotes or amastigotes, respectively, washed twice with 1 ml of 10 mg/ml BSA in sorbitol buffer at 4°C, and resuspended in 100 µl of sorbitol buffer. Cells were transferred into a 96-well black microplate and lysed with SDS (1% final concentration), and their fluorescence was measured using excitation and emission filters of 485 and 520 nm, respectively. A standard curve was obtained with the fluoresceinated peptides in sorbitol buffer.
Changes in intracellular ATP levels
The in vivo variation of intracellular ATP levels in L. donovani promastigotes was measured in parasites expressing firefly luciferase (19)
, except for the use of sorbitol buffer as incubation medium. Changes in luminescence were recorded for 120 min using a luminescence optics setting. The measurements were averaged every 4 s. In vitro inhibition of Photinus pyralis luciferase (EC 1.13.12.7) by the peptides was tested as described previously (19)
using a commercial purified enzyme (Roche Applied Sciences, Madrid, Spain).
The release of ATP from L. donovani promastigotes into the external medium was detected by a luminescence assay. Parasites (2x107 cells/ml) were incubated with 6 and 12 µM L-Hst5 for 10 min and centrifuged, and 50 µl of the resulting supernatant was immediately mixed with 750 µl of H2O and 200 µl of a firefly lantern extract reagent (19)
. Luminescence was measured in a LKB Bio-Orbit 1250 luminometer (LKB Bio-Orbit, Vantaa, Finland) for 3 min after mixing. The data were normalized relative to a standard curve of ATP. The inhibition of ATP release by the general anion channel inhibitor 4,4'-diisothiocyanatostilbene-2,2'-disulfonic acid (DIDS) was assayed in parallel with parasites preincubated with 1 mM of this inhibitor for 1 h before peptide addition.
Effects of Hst5 on mitochondrial activities
Oxygen consumption rates were measured on digitonin-permeabilized promastigotes as described previously (23)
in a Clark-oxygen electrode (Hansatech, KingsLynn, UK) at 25°C, using 1 ml of a parasite suspension (108 cells/ml) in sorbitol buffer, permeabilized with 60 µM digitonin, and supplemented with 100 µM ADP and 5 mM succinate as substrate. Once a steady rate was reached after peptide addition, a set of inhibitors and substrates specific for the different complexes of the respiratory chain were assayed: 0.1 mM tetramethyl-p-phenylendiamine plus 1.7 mM ascorbate (TMPD-ascorbate), 1.9 µM antimycin A, 6.5 µg/ml oligomycin, and 10 µM carbonyl cyanide 4-trifluoromethoxyphenylhydrazone (FCCP).
The mitochondrial F1F0-ATPase activity was measured in digitonin permeabilized promastigotes (24)
. Parasites permeabilized with 60 µM digitonin were resuspended at 108 cells/ml in a reaction mixture containing 150 mM Tris-HCl, pH 7.5, 1 mM MgCl2, 0.5 mM EGTA, and 100 µM Na3VO4, and they were incubated with different L- or D-Hst5 concentrations for 1 h. L-melittin (GIGAVLKVLTTGLPALISWIKRKRQQ) and its diastereomer [D]-V5,8,I17,K21-melittin (GIGAvLKvLTTGLPALiSWIkRKRQQ, small caps stand for D-amino acids; ref. 25
) were used as controls. Afterward, the reaction was started by the addition of ATP (3 mM final concentration) at 26°C and stopped after 10 min by adding SDS (1% final concentration). The resulting inorganic phosphate was determined by the colorimetric method of Fiske-Subbarow, based on the reduction of the phosphomolybdate complex at 650 nm.
Variation of the mitochondrial membrane potential (
m) in intact promastigotes was monitored by rhodamine 123 accumulation as described previously (26)
after peptide incubation in sorbitol buffer and being measured in a Coulter XL EPICS cytofluorometer (excitation and emission wavelengths 488 and 525 nm, respectively). Parasites either intact or depolarized with 10 µM FCCP were taken as controls.
Assay for cytotoxic activity against murine macrophages and intracellular amastigotes
BALB/c peritoneal macrophages were resuspended in sorbitol buffer at a final concentration of 106 cells/ml, plated in a 96-well culture microplate (100 µl/well), and incubated with Hst5 for 4 h at 37°C. Cytotoxicity was assessed using the colorimetric MTT reduction assay and was expressed as the percentage of the value obtained for control cells.
For infection experiments, the macrophages were seeded at a 105 cells/well in a Lab-Tek 16-well chamber slide system in RPMI 1640 plus 10% heat inactivated fetal calf serum and incubated with L. pifanoi axenic amastigotes at a 1:2 macrophage:amastigote ratio for 4 h. Infection was allowed to progress for 96 h at 32°C, followed by treatment with 25 µM Hst5 for 4 h in sorbitol buffer. After peptide removal, the chambers were additionally incubated for 48 h in culture medium. Finally, slides were air dried, fixed in methanol, and stained with Giemsa. The parasite burden was considered as the average number of amastigotes in 100 macrophages measured by triplicate and referred to as percentage of the untreated macrophages.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
). At neutral pH, Dhvar 4 showed the highest activity on promastigotes (LC50=1.8 µM) followed by D-Hst5 and L-Hst5. The same order resulted for the activity on amastigotes. Nevertheless, whereas D-Hst5 and Dhvar4 were equally potent on both forms of the parasite, the LC50 of L-Hst5 for amastigotes was twice as high as for promastigotes. Regardless of the assay conditions and stage of the parasite, D-Hst5 was consistently more active than L-Hst5. The fact that both enantiomers were active on Leishmania ruled out a critical role for chirality in the lethal process.
|
At acidic pH, the increase in positive charge due to histidine protonation enhanced the activity of both Hst5 enantiomers, except for D-Hst5 on promastigotes. In sharp contrast, Dhvar4, devoid of histidine residues, did not undergo any significant variation.
It was found that the leishmanicidal activity of the peptides was severely inhibited when assayed on cells previously preincubated with 1 mM KCN for 1 h, a condition that causes a reversible inhibition of cytochrome c oxidase, hence of mitochondrial ATP production, without compromising the parasite viability in a time frame of several hours (14)
.
Plasma membrane permeabilization assays
The permeabilization of the plasma membrane of L. donovani promastigotes by the peptides was monitored by two complementary approaches: 1) membrane depolarization, directly linked to the dissipation of ionic gradients across the membrane; and 2) access of the vital dye SYTOX Green (molecular weight=600) to the cytoplasm, requiring large size lesions and precluded in intact cells.
As depicted in Fig. 1
A–C, promastigotes underwent a fast and dose-dependent increase in bisoxonol fluorescence after peptide addition, typical of a depolarization process. This was followed by a dose-dependent steady recovery for >1 h. At a given concentration, Dhvar4 caused a much larger depolarization than any Hst5 enantiomer.
|
The differences in plasma membrane permeabilization between L-Hst5 and Dhvar4 were emphasized when the entrance of the vital dye SYTOX Green was measured. Figure 1D-F
shows the intracellular accumulation of SYTOX Green after peptide addition. The lesions caused by Dhvar4 and both Hst5 enantiomers at equipotent concentrations differed widely. At its LC50, the fluorescence value for Dhvar4 reached 80% of the maximal permeabilization (Fig. 1F
), defined as that achieved by the detergent TX-100 at 0.1%. In contrast, the increase produced by both Hst5 enantiomers at their respective LC50 hardly reached 10% (Fig. 1D, E
), underscoring the different nature of the lesions for Dhvar4.
Electron microscopy
In a further step, the morphological damage inflicted to the promastigotes by the different peptides was visualized by transmission electron microscopy. L-Hst5 and Dhvar4 were assayed on promastigotes at an equipotent concentration causing 80% inhibition of parasite proliferation (12 and 3 µM, respectively). Promastigotes incubated with L-Hst5 showed a dramatically distorted mitochondrion, its matrix strongly swollen and, in most parasites, with a bulging-out appearance, while the plasma membrane appeared without substantial damage (Fig. 2
B). On the contrary, Dhvar4 caused extensive blebbing of the plasma membrane of which large areas were detached from the subpellicular layer of microtubules (Fig. 2C
), an effect previously reported for other typical membrane-seeking AMPs in Leishmania (22
, 26)
. These different morphological patterns are in tune with the divergences observed for these peptides in membrane permeabilization experiments.
|
Intracellular localization of the peptides
The electron microscopy experiments were complemented with the intracellular localization of fluoresceinated peptides at their respective LC50 by confocal fluorescence microscopy. Previously, it was confirmed that peptide labeling did not affect leishmanicidal activity >5% (data not shown). The intracellular distribution of both dyes, the fluoresceinated peptides and the mitochondrion-selective dye MitoTracker Red, was largely but not exclusively overlapping, as shown in Fig. 3
. This points out mitochondrion as an organelle with a privileged accumulation of the peptides in contrast with the scarce fluorescence associated to the plasma membrane. The translocation of FITC-D-Hst5 into the intracellular space excludes a key participation of a stereospecific receptor for this process in Leishmania. Again, a difference in the fluorescence pattern between FITC-L-Hst5 and its analog FITC-Dhvar4 was observed (Supplemental Fig. 1B). For FITC-Dhvar4, the distribution was more diffuse, in contrast with that for FITC-L-Hst5, which appeared mostly confined to specific spots. Preincubation of the parasite with KCN, a condition that significantly decreased both leishmanicidal and candidacidal activity of Hst5 (5)
, prevented the accumulation of both peptides into the promastigote, being especially noticeable for FITC-L-Hst5 (Supplemental Fig. 1D).
|
Intracellular accumulation of Hst5
Once the internal localization of the peptides was established by confocal microscopy, we proceeded to measure their intracellular concentration. The accumulation of FITC-D-Hst5 at 7.5 µM in promastigotes was 3-fold higher (0.127 nmol/106cells) than FITC-L-Hst5 (0.043 nmol/106 cells) in accordance with its higher leishmanicidal activity on this form of the parasite (Table 1)
. The same trend was also found for amastigotes, with an intracellular accumulation of 0.023 nmol/106cells for FITC-D-Hst5, nearly 9-fold higher than that for FITC-L-Hst5 (2.5x10–3 nmol/106cells). These data were further used to calculate the intracellular concentration of FITC-L-Hst5 in parasites. For promastigotes, the values for internal volume (4.3 µl/mg of protein; ref. 27
) and 1 mg of protein=3.6x108 promastigotes; ref. 28
) were taken, giving an intracellular concentration 3.6 mM. Similar calculations in amastigotes resulted in a Hst5 concentration 0.7 mM, with the internal volume calculated as by Alberio et al. (29)
.
Effect of the histatins on the bioenergetic metabolism of L. donovani promastigotes
Having identified the mitochondria as an organelle with a privileged accumulation for Hst5, we studied the bioenergetic metabolism of the promastigote in greater detail to obtain a deeper understanding of the intracellular targets of L-Hst5 and Dhvar4 in this organism.
To this end, promastigotes of the cytoplasmic luciferase-encoding 3-Luc strain show a luminescence that parallels the concentration of free-cytoplasmic ATP, the limiting substrate for this reaction, when incubated in the presence of the membrane-permeable luciferase substrate DMNPE-luciferin (19
, 22)
. Both L-Hst5 and Dhvar4 inhibited luminescence in a concentration-dependent manner, reaching the final levels after 10 min (Fig. 4
). Dhvar4 induced a consistently higher inhibition relative to L-Hst5, as observed for the inhibition of proliferation; in addition, the end point for this peptide was reached faster. The specificity of this effect was shown, as the luminescence associated to the glycosomal luciferase was hardly altered after addition of L-Hst5 at the highest concentration tested (Fig. 4)
.
|
The release of intracellular ATP induced by L-Hst5 and inhibited by DIDS was very instructive about its candidacidal mechanism (30
, 31)
. In Leishmania, the external ATP accumulation caused by L-Hst5 at 6 and 12 µM was of 65 and 166 pmol x 10–6 cells, respectively, accounting for 3.2 and 15.1% of the total intracellular ATP, compared with parasites fully permeabilized with 0.1% TX-100 (887.7 pmolx10–6 cells) after subtracting the spontaneous ATP release (37.2 pmolx10–6 cells; Fig. 5
B). These values were obtained 10 min after peptide addition, a time frame at which the drop of luminescence, that is of intracellular ATP, reached its respective end point with a reduction of 17.8 and 47.8% for 6 and 12 µM, respectively (Fig. 5A
). Furthermore, parasites pretreated with 1 mM DIDS before L-Hst5 addition showed an identical rate of luminescence decay (Fig. 5A
) and identical accumulation of external ATP. Altogether, these data provide evidence that the release of ATP induced by Hst5 cannot fully account for the drop in its intracellular level, hence for a major role in its leishmanicidal activity, in opposition to its candidacidal one.
|
We therefore hypothesized that the bioenergetic collapse produced by Hst5 was due to its effect on oxidative phosphorylation, the major source of ATP synthesis for L. donovani (14)
. In fact, L-Hst5 caused a significant inhibition on two parameters directly related to this process, the mitochondrial electrochemical potential (m) and the oxygen consumption rate. Both L-Hst5 and Dhvar4 caused a substantial decrease in the levels of m-driven rhodamine 123 uptake (Fig. 6
). Nevertheless, when compared at their respective LC50, L-Hst5 produced a 30% decrease in m, whereas for Dhvar4 the inhibition was hardly perceptible. On living promastigotes, respiration was inhibited by both L-Hst5 and Dhvar4 in a dose-dependent manner. L-Hst5 was the less potent, resulting in an EC50 of 23.9 ± 1.2 µM, whereas that for Dhvar4 was 4.6 ± 0.6 µM, not correlating with its mild inhibition on m and suggesting that for this peptide the inhibition of respiration was not a consequence of a specific effect on mitochondria but rather due to the leakage of nutrients and ions from the cytoplasm caused by plasma membrane permeabilization, as described for other membrane-seeking AMPs.
|
For further identification of the molecular target of Hst5, we explored its effects on mitochondrial ATP synthesis using digitonin-permeabilized promastigotes. Under these conditions, a selective permeation of the plasma membrane, but not of mitochondria, takes place, allowing free access of substrates and inhibitors to the different complexes of the respiratory chain. The integrity of the inner mitochondrial membrane in the assay was preserved, as the respiration increased by recovery of state 3 after ADP addition (Fig. 7
, left trace). In digitonized parasites, 20 µM L-Hst5 reduced the respiratory activity by 70% relative to the value in absence of the peptide (Fig. 7
, right trace). This inhibition was reverted by the addition of the mitochondrial uncoupler FCCP (10 µM), discarding any specific inhibitory effect on the carriers of the electron transport chain. Also, addition of TMPD-ascorbate, an electron donor to cytochrome c, did not revert the inhibition (Fig. 7
, right trace). These results strongly suggested that L-Hst5 did not act at the level of the mitochondrial respiratory chain, e.g., as an uncoupler or blocker of the electron transfer. Since the electron transfer is obligatorily coupled to mitochondrial ATP-synthesis, inhibition of one process will result in inhibition of the other and vice versa. We therefore examined whether the inhibitory effect of Hst5 was due to inhibition of the F1F0-ATPase, which carries out mitochondrial ATP synthesis. Both L-Hst5 and D-Hst5 reduced the oligomycin-sensitive ATPase activity by 
50 and 75% at 20 and 50 µM, respectively. This points toward the mitochondrial F1F0-ATPase as a likely target for Hst5 (Fig. 7
, inset). L-melittin and its diastereomer [D]-V5,8,I17,K21-melittin, with a partially crippled 
-helical structure, were used as controls, inhibiting F1F0-ATPase activity with an EC50 of 0.8 and 3 µM, respectively.
|
Leishmania killing does not involve an apoptotic process
The above results have shown that the functionality of the mitochondrion was severely hampered by L-Hst5. Since the mitochondrion plays a pivotal role in the onset of apoptosis, it was investigated whether the effects of L-Hst5 involved induction of apoptotic processes in Leishmania. The percentage of promastigotes with a propidium iodide fluorescence in the sub-G1 peak region typical of apoptotic cells (21)
was identical to the untreated cells after L-Hst5 incubation, even for concentrations causing a mortality rate >80%, in oppositition to the 30% observed for those treated with miltefosine (Supplemental Fig. 2
).
Cytotoxicity on mammalian cells and intracellular amastigotes
L-Hst5 was not cytotoxic (<5%) when assayed on peritoneal mouse macrophages at 25 µM. At this concentration, the reduction of the parasite burden in L. pifanoi infected macrophages was 61.4 ± 4.6%. In comparison, the Hst5 LC50 for axenic amastigotes was 14.4 µM (Table 1)
.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, and it thrives in harsh acidic pH (17)
, where the cationicity of histatins increases due to histidine protonation. Furthermore, its plasma membrane is specially adapted to provide a partial protection against AMPs, due to the restriction of the membrane traffic to the flagellar pocket and a dense and highly anionic glycocalyx, rich in proteolytic activity (10)
.
Interaction of L-Hst5 with the plasma membrane of Leishmania
The differences reported for the action of L-Hst5 and its analog Dhvar4, endowed with a much higher amphipathicity and membrane-disrupting activity, both on model membranes (20)
and on Candida (32)
, were also mirrored on Leishmania promastigotes. Dhvar4 induces a fast and irreversible plasma membrane depolarization in Leishmania promastigotes and a large entrance of the vital dye SYTOX Green into the parasite, associated to a massive plasma membrane disruption with large blebbings, as visualized by electron microscopy, similar to the effect observed with other membrane-active AMPs on Leishmania promastigotes (33)
. Altogether, these effects are detrimental enough to account for a lethal outcome. In contrast, Hst5 produced a limited and mostly reversible depolarization of the plasma membrane and a much lower rate and final extent of SYTOX Green uptake. In addition, the electron micrographs of promastigotes treated with this peptide showed a plasma membrane that was scarcely damaged. By contrast, Hst5 induced in Candida large invaginations of its plasma membrane (5)
. The high stability of the subpellicular microtubular layer beneath the plasma membrane of Leishmania may partially account for this difference. In fact, in Candida the increase in membrane rigidity or in cytoskeleton stability leads to a reduced sensitivity to Hst5 (34)
. Nevertheless, this is a limited effect, as other membrane-seeking AMPs (26)
, including Dhvar4, even at much lower concentrations than the EC50 for L-Hst5, produced a pattern of extensive membrane disruption. This strongly suggested that in the leishmanicidal action of Hst5, membranolysis plays only a minor role; thereby, the involvement of an intracellular target is essential.
Intracellular uptake of Hst5 by Leishmania
An important issue is how L-Hst5 translocates across the membrane and gains access to the intracellular space. Several mechanisms are feasible: active uptake though the involvement of a surface receptor, endocytosis, or a passive translocation across the phospholipid matrix plasma membrane. Ssa1/2p, members of Hsp70 family, have been reported to act as cell wall-expressed major receptors for some AMPs, including L-Hst5 (35
36
37)
. To the best of our knowledge, the surface expression of Leishmania heat shock proteins has not been yet reported. By a BLAST similarity search (Blastp; available at www.ncbi.nlm.nih.gov/BLAST/Blast.cgi) for L.donovani, an Hsp70 protein (Swiss-Prot accession number P17804) with a 73% amino acid sequence identity was the closest homologue of Ssa1/2p. Preliminary experiments carried out with an extensive set of monoclonal antibodies raised against Leishmania Hsp70 were negative, both for the binding of these antibodies to intact promastigotes and for the inhibition of the leishmanicidal activity of L-Hst5 (data not shown). These results, together with the activity of the all-D-Hst5, make it quite unlikely that chiral recognition by a putative plasma membrane receptor/transporter in Leishmania is a key event in Hst5 uptake.
Preliminary experiments of colocalization of rhodamine B dextran, a marker for fluid phase endocytosis, and FITC-L-Hst5, point to the plasma membrane and to a lesser extent the endosomal membrane, as the translocation site for Hst5, discarding endocytosis, a process also inhibited by KCN, as an exclusive pathway for Hst5 uptake (data not shown).
The third option is the translocation of L-Hst5 across the phospholipid matrix of the plasma membrane. L-Hst5 showed a poor translocation capacity in anionic liposomes when compared with living Candida (20)
. This might be due to the presence of a membrane potential in living cells in agreement to the worm-hole model (3)
. On the other hand, the internalization of Hst5 in Candida is strongly dependent on the physical state of the plasma membrane, highlighting its importance in this process (34)
.
Overall, two major conclusions can be drawn: 1) damage to the plasma membrane of Leishmania promastigotes induced by L-Hst5 is less crucial for its lethal action than for other AMPs, and 2) regardless of the translocation mechanism, the peptide acts through intracellular targets, similar to its action in Candida, where the L-Hst5 lethal hit and many of its associated effects caused by extracellular L-Hst5 or L-Hst3 were reproduced by the peptide synthesized endogenously (38)
.
Hst5 targets the bioenergetic metabolism of Leishmania
Once the plasma membrane was discarded as the major target for L-Hst5 in Leishmania, we proceeded to define its intracellular targets in this parasite. A number of pleiotropic effects of Hst5 on Candida have been described, but their respective crucial or secondary role in the lethal action is not completely settled yet (16)
, although the involvement of the Hog1 MAP kinase as a global regulator was recently proposed (39)
.
Apparently, the mitochondrion is a major target for L-Hst5 in Leishmania, as 1) electron microscopy showed extensive morphological alterations of this organelle, with a swollen matrix and poor cristae definition after Hst5 treatment; 2) the distribution of the fluoresceinated Hst5 largely, but not exclusively, overlapped with mitochondrial markers, and its uptake was precluded in cells preincubated with metabolic inhibitors; and 3) both the respiration rate and m, two parameters of mitochondrial functionality crucial for an adequate production of ATP, were severely decreased by L-Hst5.
According to the results obtained in intact and digitonized promastigotes, inhibition of oxidative phosphorylation remained the most feasible option for the action of L-Hst5, as the inhibition of the oxygen consumption by this peptide was not reverted even after feeding the complex IV, the terminal electron donor, with specific artificial substrates. In contrast, addition of FCCP, a typical uncoupler, to L-Hst5-treated parasites, recovered the initial rate. Therefore, it may be concluded that L-Hst5 did not impair the functionality of the respiratory chain, in contrast with Candida, as in this pathogen the inhibition by Hst5 was not reverted by the uncoupler CCCP (32)
. Furthermore, the important role of reactive oxygen species (ROS) described for the candidacidal mechanism of L-Hst5 (16)
was also excluded in Leishmania, as in our hands, neither the coincubation of Leishmania promastigotes with L-Hst5 plus the free radical scavenger TEMPO (2,2,6,6-tetramethylpiperidine-N-oxyl) nor production of ROS, assessed by H2DCFDA, was positive (data not shown). This was not totally unexpected, since Leishmania is endowed with a large set of neutralizing mechanisms for ROS (41)
to thrive permanently or temporarily on macrophage (10)
or neutrophils (42)
, respectively, two dedicated ROS producer cells.
Our results also exclude L-Hst5 as an uncoupler for oxidative phosphorylation in Leishmania, described for other AMPs with high amphipathicity and membrane disrupting characteristics as the cecropin A-melittin peptides on purified mammalian mitochondria (40)
.
L-Hst5 and its D-enantiomer caused a significant decrease in the oligomycin-sensitive F1F0-ATPase activity, pointing toward the inhibition of ATP synthesis as a crucial step in the leishmanicidal mechanism. To our knowledge, this is the first AMP hitherto reported to act on this enzymatic activity in living Leishmania parasites. Although a detailed mechanism for this activity is still missing, it is worth noting that in bovine mitochondrial F1F0-ATPase inhibition by a variety of amphipathic peptides such as melittin or synthetic analogues of the mitochondrial import presequence of yeast cytochrome oxidase subunit IV has been reported (43)
; whether a similar assumption may be valid for Leishmania remains uncertain. Our results also support an inhibition of the Leishmania F1F0-ATPase through a promiscuous interaction, since L-melittin was also inhibitory. A certain amphipathicity threshold is critical, as demonstrated by the loss of activity of a melittin diastereomer with a partially crippled secondary structure (44)
. A similar absence of chirality requirement, the presence of an amphipathic structure but not of chirality, was described for the modulation of calmodulin by amphipathic peptides (45)
.
In AMPs, the activity of an all-D-enantiomer is taken as a touchstone to rule out intracellular targets involving stereoselective interactions (6
, 46
47
48)
in contrast to the receptor-independent killing carried out through disruption of the phospholipid matrix of the plasma membrane (49)
. To the best of our knowledge, only the all-D-enantiomers of the plant peptide Ib-AMP1 (50)
and of the pentadecamer lactoferricin B17–31 (51)
act through intracellular targets not yet identified and are as active as their all-L counterparts. The stronger leishmanicidal activity of the all-D-Hst5 relative to its natural counterpart is likely due to its higher accumulation inside the parasite due to increased resistance to enzymatic degradation.
Altogether these data show that the release of ATP induced by Hst5 cannot fully account for the drop in its intracellular level, hence for a major role in its leishmanicidal activity, in opposition to its action on Candida. A similar conclusion was reported in Candida by functional and proteomic analysis of the Hst3-resistant strains (52)
. The final outcome is the bioenergetic exhaustion and death of the parasite by a nonapoptotic process, in opposition to the apoptosis induced by indolicidin, the only leishmanicidal AMP acting through intracellular targets described so far (53)
.
Prospects for the pharmacological application of Hst5
The clinical importance of Leishmania and the paucity of new leishmanicidal drugs may support the application of histatins as a new chemotherapeutical approach for Leishmania. The feasibility of this approach is supported by successful results of histatins in oral and upper airway candidiasis (54
, 55)
, together with the lack of toxicity toward mammalian cells, and by their capability to reduce the parasite burden in infected macrophages. The lower effect of Hst5 in this system might be conveniently increased by encapsulation of Hst5 into liposomes, similar to other leishmanicidal drugs (56)
, which exploit the high rate of phagocytosis of the macrophage.
In addition, the oxidative phosphorylation in L. donovani, the main target for L-Hst5, is crucial to fulfilling the bioenergetic requirement of the parasite (14)
, and consequently, the bioenergetic flexibility of Leishmania to switch from fermentative to nonfermentative metabolism is much lower than in Candida (57)
. Therefore, Leishmania will presumably be less prone to induce resistance to Hst5 than Candida, which has been already obtained in vitro for Hst3 (52)
.
Finally, an important feature of Hst5 is its behavior as a true cell-penetrating peptide (CPP), with a rather specific toxicity for lower eukaryotes. CPPs gain access to the intracellular milieu through their translocation across the endosomal or plasma membrane of the cell (review in ref. 58
and references therein). A major practical advantage is that they may ferry a wide variety of cargo molecules, such as peptides, proteins, nucleic acids, and even nanoparticles, across the membrane without the requirement of a dedicated membrane receptor (59)
. In this way, if another leishmanicidal agent is coupled to Hst5, it might gain access to the inside of the parasite, bypassing the low expression or even the absence or faulty receptor for these drugs, converting the conjugate drug-Hst5 into a single molecule with dual leishmanicidal action. Other CPPs have been used for a similar goal in Leishmania (60
, 61)
; however, their selectivity is much lower than that for Hst5. Two additional nanotechnological advantages of Hst5 are its high intracellular concentration inside the parasite, which might allow a high accumulation of the cargo molecule attached to Hst5, and its capacity to target the mitochondria of the pathogen, a scarce trait among the CPPs currently in use (59)
. According to the inhibition of Hst5 accumulation by KCN, a functional mitochondrion is required for the lethal hit of Hst5. We may envisage that the mitochondrial potential attracts the cytoplasmic monomers of Hst5, promoting their accumulation in this organelle that acts as a sink for this peptide.
Although the lethal mechanisms of Hst5 on Leishmania shared some steps with Candida, it showed significant differences in the translocation mechanism and site of action, showing the plasticity of these set of peptides to target different pathogens. Work in progress concerns the evaluation of the feasibility of Hst5 as a new template for leishmanicidal drugs and as a cell-specific CPP.
|
ACKNOWLEDGMENTS |
|---|
Received for publication August 14, 2007. Accepted for publication January 3, 2008.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  H. Hillesland, A. Read, B. Subhadra, I. Hurwitz, R. McKelvey, K. Ghosh, P. Das, and R. Durvasula Identification of Aerobic Gut Bacteria from the Kala Azar Vector, Phlebotomus argentipes: A Platform for Potential Paratransgenic Manipulation of Sand Flies Am J Trop Med Hyg, December 1, 2008; 79(6): 881 - 886. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
EXC=544 nm; 
), 3 µM (
), 6 µM (
), 12 µM (
), and 24 µM (
).
), 5 µM (•), 8.5 µM (
), 25 µM (
). Dotted line represents luminescence obtained from L. donovani promastigotes expressing glycosomal luciferase.
