| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* USM 0504, Biologie Fonctionnelle des Protozoaires, Muséum National d'Histoire Naturelle, Paris, France
INSERM UMR 585, INSA-Lyon, Physiopathologie des Lipides et Membranes, Villeurbanne, France
Published, JLR Papers in Press, April 10, 2006.
1 To whom correspondence should be addressed. e-mail: deregnau{at}mnhn.fr
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
Supplementary key words secreted phospholipase A2 • malaria • unsaturated fatty acids • oxidation
Abbreviations: AA, arachidonic acid; BHT, butylated hydroxytoluene; bvPLA2, bee venom phospholipase A2; chyl, chylomicron; DHA, docosahexaenoic acid; FA, fatty acid; F-chyl/VLDL, freshly prepared chyl/VLDL; GC, gas chromatography; LA, linoleic acid; lysoPC, lysophosphatidylcholine; lysoPE, lysophosphatidylethanolamine; lysoPL, lysophospholipid; Ox-chyl/VLDL, oxidized chyl/VLDL; Ox-LDL, oxidized LDL; PC, phosphatidylcholine; PS, phosphatidylserine; RBC, red blood cell; sPLA2, secreted phospholipase A2; TBARS, thiobarbituric acid-reactive substances
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
Phospholipases A2 enzymes (EC 3.1.1.4) exhibit a variety of physiological activities in addition to intrinsic lipolytic action. Those enzymes catalyze the hydrolysis of the sn-2 ester bond of glycerophospholipids, leading to the production of NEFAs and lysophospholipids (lysoPLs). Secreted phospholipases A2 (sPLA2s) form a large family of low-molecular-mass (13–19 kDa), water soluble, and structurally conserved enzymes that have primarily been identified in animal venoms, but are also distributed in mammalian tissues, fluids, and secretions (3), plants (4), bacteria (5), and viruses (6, 7). Interestingly, despite common catalytic properties, venom sPLA2s differ greatly in their pharmacological effects, such as neurotoxic, myotoxic, cardiotoxic, or anticoagulant properties, and some sPLA2s have been shown to display antibacterial (8, 9), antiviral (HIV) (10), or anti-Plasmodium properties (11, 12).
We previously showed that sPLA2s from snake, scorpion, or bee venoms are potent inhibitors of the in vitro intraerythrocytic development of P. falciparum (11, 12). Inhibition at low enzyme concentration occurs only in the presence of serum phospholipids (12), suggesting that hydrolysis of exogenous phospholipids, rather than hydrolysis of the infected red blood cell (RBC) membrane phospholipids, is required for sPLA2s' toxicity.
Here, we were interested in deciphering the molecular and cellular interplays between sPLA2 enzyme, human lipoproteins, and infected erythrocytes, as an approach to a better understanding of what may occur in pathophysiological situations such as malaria, in which alterations of the host lipoproteinogram (13) and sPLA2 production (14, 15) are encountered.
To understand what governs the indirect toxicity of sPLA2s, the distribution of the bee venom phospholipase A2 (bvPLA2) between the infected erythrocytes and the triglyceride-rich fraction [chylomicrons (chyls) and VLDLs] of human lipoproteins was analyzed. Incidence of lipoprotein oxidation on bvPLA2 particle binding and anti-Plasmodium activity was also analyzed, because peroxidation has been shown to increase the susceptibility of lipoproteins to hydrolysis by sPLA2 (16) and oxidized lipoproteins have been found in malaria patients (17). Finally, we attempted to establish which of the lipolyzed particles or lipid products are responsible for parasite killing. LysoPLs and NEFAs generated by the bvPLA2 activity were identified, and their individual involvement in parasite killing was established, exemplifying the major role of PUFAs.
Our results illustrate the capacity of sPLA2s to be active on biological targets via the generation of exogenous NEFAs and raise the question of a potential role for oxidized lipoproteins and/or endogenous sPLA2s in the host defense against malaria.
|
EXPERIMENTAL PROCEDURES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
Methods
Culture and synchronization of P. falciparum
The Colombian strain FcB1 of P. falciparum was used throughout the work. Cultures were grown in complete medium consisting of RPMI 1640 (Life Technologies, Inc.) supplemented with 11 mM glucose, 27.5 mM NaHCO3, 100 UI/ml penicillin, 100 µg/ml streptomycin, adjusted to pH 7.4 before addition of heat-inactivated human serum (8% final), according to the procedure of Trager and Jensen (18). Parasites were grown at 37°C in human O+ or A+ RBCs at a 2% hematocrit and a 3–6% parasitemia, in a 3% CO2, 6% O2, and 91% N2 atmosphere. Cell cultures were synchronized by successive Plasmagel® (19) and sorbitol (20) treatments.
Purification and lipolysis of the chyl/VLDL fraction The d<1.006 g/ml lipoprotein fraction comprising chyls and VLDLs was purified by centrifugation (24 h, 160,000 g, 4°C) of nonfasted human serum diluted 1:5 in sterile NaCl, 9 g/l, according to the technique by Havel, Eder, and Bragdon (21). The protein content of the fraction was determined using the Bio-Rad DC protein assay according to the manufacturer's recommendations.
Coupling of bvPLA2 to Affi-Gel 10 beads (Bio-Rad) was achieved following the instructions of the manufacturer. Efficiency of the coupling and enzymatic activity of the immobilized bvPLA2 were measured as in (12). For enzymatic lipolysis, 1 ml chyl/VLDLs (approximately 0.15 mg/ml in PBS), either freshly prepared (F-chyl/VLDL) or oxidized (Ox-chyl/VLDL), was mixed with 130 pM active immobilized bvPLA2 in the presence of 1 mM CaCl2 and incubated at 37°C for 17 h.
Chyl/VLDL oxidation To induce air/light minimal oxidation of lipoproteins, chyl/VLDLs in PBS or in NaCl (9 g/l) were stored in a transparent flask at ambient temperature and under sterile air exchange for 18 days. Lipid peroxides were measured in terms of thiobarbituric acid-reactive substances (TBARS) according to the method of Morlière et al. (22). Results are expressed in malondialdehyde equivalents.
In experiments comparing properties of fresh and oxidized lipoproteins, chyl/VLDLs were isolated from plasma that had been aliquotted and frozen at –20°C just after blood drawing. One aliquot was thawed for chyl/VLDL purification and air/light oxidation. At the end of the 2 1/2 week oxidation period and just prior to the experiment, a second aliquot was thawed for purification of F-chyl/VLDLs.
Analysis of bvPLA2 binding to erythrocytes Binding of bvPLA2 to RBCs was quantified by sedimenting the cells and measuring the fraction of enzyme remaining in the supernatant. RBCs were sedimented by centrifugation at 90 g for 3 min. The amount of bvPLA2 in the supernatant was measured using the fluorimetric assay with ß-py-C10-HPC as substrate (23). Binding experiments were performed either in PBS (10 mM NaPO4, 150 mM NaCl, pH 7.4) or in RPMI. Binding reactions (200 µl) contained 8.0 ng (2.5 nM) bvPLA2 and 2.5 x 107 RBCs, either normal or parasitized by mature forms of P. falciparum. Parasitized erythrocytes were highly enriched (70–80%) in schizont forms (36–48 h of age) recovered by Plasmagel® treatment.
The procedure was as follows: RBCs in PBS were distributed in the BSA-coated wells of a 96-well microplate at the rate of 100 µl/ per well, then 100 µl bvPLA2 (5 nM) in PBS or in PBS + 0.15 mg/ml chyl/VLDLs was added. In the experiment with oxidized lipoproteins, bvPLA2 + chyl/VLDLs were incubated for 45 min at 37°C prior to addition. After 45 min incubation at 37°C, the plate was gently centrifuged at 90 g and supernatants were collected for the spectrofluorimetric assay.
The reaction medium was 50 mM Tris, pH 7.5, 100 mM NaCl, 1 mM EDTA, 6 mM CaCl2, 0.1% BSA, and 2 µM ß-py-C10-HPC. Two hundred microliters of substrate was added to 100 µl of supernatant. The pyrene monomer fluorescence, corresponding to the phospholipid hydrolysis upon bvPLA2 activity, was measured continuously for 10 min with a luminescence spectrometer (AMINCO; Bowman, Series 2) using 345 nm (excitation) and 398 nm (emission) wavelengths. PLA2 activity is expressed in fluorescence variation per second. BvPLA2 100% activity in PBS and in PBS + lipoproteins, and in the absence of cells, was measured at time zero of the incubation. Residual activity after 45 min incubation in the absence of cells was measured to estimate bvPLA2 binding to coated BSA. Negative controls were without bvPLA2. Detection limits were 25 pM enzyme in the absence and 250 pM enzyme in the presence of chyl/VLDLs.
Lipid sequestration by BSA Two milliliters of chyl/VLDLs (0.15 mg/ml in PBS, 250 µM CaCl2) was incubated with 130 pM active Affi-Gel-immobilized bvPLA2 for 17 h at 37°C. After a brief centrifugation to pellet Affi-Gel beads, the supernatant was mixed with FA-free BSA (20 mg/ml) for a 2 h incubation at 37°C under gentle agitation. PBS was added up to 8 ml and chyl/VLDLs were separated from the lipid-charged BSA by ultracentrifugation (160,000 g, 20 h, 4°C). The buoyant lipoproteins were recovered from the top fraction, and the pelleted BSA was recovered from the bottom fraction. Both fractions were tested in dose-response assays toward the in vitro intraerythrocytic development of P. falciparum.
NEFA and lysoPL identification Lipid-enriched BSA was supplemented with 50 µM BHT and extracted twice with chloroform-methanol-water (2.5:2.5:1.25; v/v/v) using the Bligh and Dyer protocol (24). Extracted lipids were dried in a rotary evaporator and stored at –20°C under argon for further analysis. Extracted lipids in chloroform-methanol (1:1, v/v) were separated by TLC on silica gel 60G plates. Samples were duplicated, and migration was realized in chloroform-methanol-water (65:25:4; v/v/v) with standard markers. Lipids were fractionated into lysophosphatidylcholine (lysoPC) (RF 0.20), lysophosphatidylethanolamine (lysoPE) (RF 0.40), and NEFAs (RF 0.91). They were either revealed by 2'-7'-dichlorofluorescein, 0.02% in ethanol-water (95:5; v/v) vaporization, or left unstained for recovery before gas chromatography (GC) analysis.
Lipids were collected by plate scraping. LysoPLs were resuspended in 1 ml H2SO4 (5% in MeOH), and incubated for 1 h 30 at 100°C in a dry bath. The reaction was stopped by transfer at 0°C, lipids were neutralized by the addition of 1.5 ml K2CO3 (5% in H2O) and extracted by the addition of 2 ml isooctane. The organic phase was collected upon centrifugation at 900 g for 10 min and dried, and lipids were resuspended in isooctane (2 ml) for GC analysis.
NEFAs were extracted twice with 1 ml diethyl ether-methanol (9:1; v/v). The organic phase was collected after centrifugation (10 min, 450 g), dried under argon, and derivatized by a 15 min incubation in the dark with 250 µl diazomethane. Upon derivatization, the fraction was dried under argon and resuspended in 250 µl isooctane for GC analysis.
GC separation of FA methyl esters was carried out on a Supelco SP2380 capillary column (30 m x 0.25 mm) and analyzed on an HP 6890 G1530 chromatograph. Heptadecanoic acid (C17:0) and heptadecanoyl-lysoPC were used as internal standards. FAs were quantified according to the known amount of added internal standard.
Toxicity assay (dose-response assay) The diverse lipid preparations described above, as well as commercial lipids, were tested for their capacity to inhibit the in vitro intraerythrocytic development of P. falciparum. Dry preparations of commercial lysoPLs and NEFAs, to be tested either individually or upon mixing, were solubilized in 20 µl ethanol; then 1 ml RPMI 8% serum was added, and the lipid solution was shaken for 2 h at ambient temperature before being tested for parasite growth inhibition.
Dose-response assays based upon 3H-hypoxanthine incorporation by growing parasites were performed as in (12). Radioactivity was measured with a 1450 Microbeta counter (Wallac, Perkin Elmer). Percentage of growth inhibition was calculated from the parasite-associated radioactivity compared with the control (25). Values for the IC50 and IC100 minimum (IC100min) were determined from dose-response curves.
PLA2 association with lipoproteins Five milliliters of PBS containing bvPLA2 (3 µg/ml), BSA fraction V (1 mg/ml), and CaCl2 (1 mM) were supplemented with chyl/VLDLs, either fresh or oxidized (70 µg/ml final concentration), or supplemented by the same volume of PBS. Each preparation was centrifuged at 160,000 g for 24 h. Fractions (0.6 ml) were collected from the top to the bottom of each tube. They were analyzed by SDS-PAGE on an 8% polyacrylamide gel under reducing conditions. The gel was then stained by Coomassie blue or blotted for immunodetection of bvPLA2 with anti-bvPLA2 antibodies (1:5,000).
Interaction between chylomicrons and erythrocytes
F-chyl/VLDLs or Ox-chyl/VLDLs were hydrolyzed for 45 min at 37°C by 6 nM bvPLA2 in a 300 µl final volume of PBS containing 1 mM CaCl2, or treated equally in the absence of enzyme. Three hundred microliters of a 6% hematocrit suspension of normal or Plasmodium-infected RBCs (trophozoites/schizonts at a 70–80% parasitemia) were added, then 50 µl aliquots were taken at time zero and after 0.5, 1, 2, 3, 5, 10, 15, and 30 min incubation at room temperature, and immediately centrifuged (briefly) to pellet the cells. The supernatants were boiled in SDS/ß-mercaptoethanol buffer and subjected to SDS-PAGE on an 8% polyacrylamide gel. Proteins from gels were blotted onto nitrocellulose and analyzed for the presence of apoB-48 using anti-apoB antibodies diluted 1:10,000. ApoB-48 (molecular mass 
250 kDa) is the main and specific apolipoprotein of chyls. Because of its high molecular mass (about 500 kDa), apoB-100, the main apolipoprotein of the VLDLs, was confined to the stacking gel, and could not be visualized on the Western blot under our experimental conditions.
Morphological analysis of P. falciparum in the presence of fresh or oxidized lipoproteins hydrolyzed by bvPLA2 A P. falciparum culture was synchronized on a 4 h time window. After 16 h of growth under classical culture conditions, 100 µl of culture (parasite age: 16–20 h; parasitemia: 1%; hematocrit: 4%) in RPMI + 16% serum was distributed in a 96-well plate. PLA2-hydrolyzed F- and Ox-chyl/VLDLs diluted in RPMI alone were added at their respective IC100mins (100 µl/well). As a control, the effect of nonhydrolyzed lipoproteins (fresh and oxidized) at the same concentration was analyzed. Culture was carried out in a candle jar at 37°C. Aliquots of the culture were taken every 2 h for 20 h (until parasites were 36–40 h old), then after 24 h of culture (40–44 h-old parasites) and 48 h (16–20 h-old reinvaded parasites). The parasitemia, stage distribution, and morphological development of the parasites were followed by optical examination of Giemsa-stained smears and by the counting of 4,000 cells. Images were captured by the Canon Power Shot S40 camera coupled to the Canon Utilities Remote Capture 2.1.0.10 software (magnification x1140).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
Experiments were carried out at an enzyme-to-cells ratio 100-fold higher than the IC50 toxic ratio (enzyme-to-cells ratio at the bvPLA2 IC50: 1.2 ng bvPLA2:4.5 x 108 RBCs). Noninfected or infected erythrocytes with mature forms of the parasite were incubated with bvPLA2 with or without chyl/VLDLs in BSA-coated plastic wells; then enzymatic activity in the supernatant was measured. Initial velocity at time zero of the incubation (100% enzymatic activity), as well as residual activity after incubation in the BSA-coated wells in the absence of cells, was determined. The initial velocity in the presence of chyl/VLDLs could not be determined without preincubation, because phospholipids from lipoproteins competed strongly with ß-py-C10-HPC for hydrolysis by bvPLA2.
In PBS alone, a net decrease (by approximately 56%) in enzyme activity was observed after the bvPLA2 had been incubated in the BSA-coated wells, indicating that the enzyme had adsorbed to the BSA (see Fig. 1A ). By contrast, when incubation was performed in the presence of healthy or parasitized erythrocytes, 100% activity was recovered in the supernatant, showing that under these conditions, the enzyme does not bind (or not at a detectable rate) to BSA or to erythrocytes. The same results were obtained when experiments were carried out in RPMI (not shown), indicating that in culture medium also, the enzyme does not bind to erythrocytes. When the experiment was performed in the presence of chyl/VLDLs and in the absence of cells, bvPLA2 activity in the supernatant was found to be high (Fig. 1B), suggesting that the lipoproteins had largely prevented the enzyme adsorption to BSA. The presence of erythrocytes, either infected or not, did not lower the activity, indicating that in the presence of lipoproteins, as well as in PBS or RPMI alone, the bulk of enzyme does not bind to erythrocytes.
|
|
Lipid products are essential to the anti-Plasmodium toxicity of bvPLA2-hydrolyzed lipoproteins
To assess whether lysoPLs and NEFAs are involved in the anti-Plasmodium toxicity of the lipolyzed lipoproteins, lipids were sequestered from particles by using lipid-free BSA, then both the resulting lipid-charged BSA and the delipidated lipoproteins were tested for their capacity to inhibit the parasite intraerythrocytic growth.
As can be seen in Fig. 3A , toxicity of the bvPLA2-hydrolyzed chyl/VLDLs was decreased after one round of lipid extraction and almost totally lost after two rounds, indicating that lipid products are obligatory elements in the hydrolyzed lipoproteins' toxicity. In good correlation with this, the lipid-charged BSA became toxic to Plasmodium (Fig. 3B).
|
|
GC analysis of NEFAs identified seven main species of lipids (Fig. 4
). Among them, PUFAs were the most prevalent, with linoleic acid (LA) (C18:2 n-6), arachidonic acid (AA) (C20:4 n-6), docosahexaenoic acid (DHA) (C22:6 n-3), and dihomo-
-linolenic acid (C20:3 n-6), accounting, respectively, for 39.2, 15.1, 7.1, and 5.5% of the 35.4 µg NEFAs produced per mg chyl/VLDLs. The monounsaturated oleic acid (C18:1 n-9) accounted for 13.3%, and the saturated palmitic acid (C16:0) and stearic acid (C18:0) for 6.5% and 4.6%, respectively. It thus appeared that PUFAs, especially LA, are predominant products of the chyl/VLDLs hydrolysis. This was quite expected, inasmuch as phospholipids are preferentially esterified by a saturated FA at the sn-1 position and by an unsaturated FA at the sn-2 position. Also, it must be noted that the relative distribution of NEFAs in the BSA fraction reflected that of grafted FA in PC from VLDLs (27), suggesting that the FA molecular species at the sn-2 position in phospholipids has no effect on the hydrolyzing activity of bvPLA2 toward chyl/VLDLs.
|
|
Albumin decreases the bvPLA2-hydrolyzed lipoprotein toxicity
Because serum albumin is a well-known transporter of FAs in plasma and is able to capture lysoPLs and NEFAs from bvPLA2-hydrolyzed lipoproteins, we asked whether it might be involved in toxicity by facilitating the transport of the lipid products from the lipolyzed particles to the erythrocytes. We compared the anti-Plasmodium activity of the bvPLA2-hydrolyzed chyl/VLDLs in normal culture conditions and in a culture containing additional BSA (20 mg/ml final, i.e., approximately five times the normal culture concentration). The presence of BSA led to a net decrease in lipolyzed lipoprotein toxicity (IC50: 158.5 µg/ml vs. 45.7 µg/ml in the absence of BSA) (Fig. 5
), indicating that BSA does not facilitate but rather prevents the anti-Plasmodium activity of the lipid products.
|
Oxidative modification of chyl/VLDLs enhances the toxicity of the bvPLA2-digested lipoproteins
Because it has been shown that minimal oxidation of lipoproteins increases the susceptibility to phospholipid hydrolysis by the human group IIA sPLA2 (16), we asked whether it might improve the toxicity of the bvPLA2-hydrolyzed lipoproteins. The chyl/VLDL fraction was purified from an aliquot of freshly drawn plasma and settled for 2 1/2 weeks under light and sterile air exchange so that minimal oxidation of lipids occured. TBARS content was measured upon oxidation and was found to be approximately twice that in F-chyl/VLDLs; digestion by bvPLA2 did not affect the TBARS content in either fresh or oxidized lipoproteins (not shown).
Interestingly, in contrast to F-chyl/VLDLs, Ox-chyl/VLDLs were able to induce Plasmodium growth arrest in vitro, with a 117.3 ± 9.5 µg/ml IC50 (see Fig. 6 ). Upon digestion by bvPLA2, the Ox-chyl/VLDLs exhibited a slightly greater toxicity than the F-chyl/VLDLs (12.1 ± 3.5 µg/ml IC50 and 26.5 ± 8.4 µg/ml IC50, respectively). Toxicity improvement upon oxidation was established through independent experiments (n = 4) with chyl/VLDLs from different sera, and was found to increase 2- to 4-fold.
|
Ox-chyl/VLDLs and F-chyl/VLDLs preincubated with bvPLA2 were added to a suspension of healthy RBCs. An aliquot of the suspension was taken at different times, then centrifuged, and the supernatant was analyzed for the presence of apoB-48, the specific 250 kDa apolipoprotein of chyls. As can be seen in Fig. 7 , lipoprotein oxidation induced the removal of apoB-48 from the incubation medium in a time-dependent manner. Removal was completed within approximately 1 h. ApoB-48 disappearance was not observed in the absence of cells (not shown), indicating that protein removal from the supernatant was not due to Ox-chyl adsorption on the tube wall, nor to any chyl/VLDL-dependent proteolytic degradation of the apoB-48 induced by oxidation. Upon Ox-chyl/VLDL incubation with bvPLA2, the apoB-48 removal occurred faster (noticeable within 10–15 min, completed within 30 min). Kinetic analysis of apoB-48 disappearance was similar regardless of whether incubation had been performed with noninfected or with P. falciparum-infected erythrocytes (not shown). Remarkably, no depletion in apoB-48 was observed with native lipoproteins, either enzymatically digested or not.
|
BvPLA2 associates with oxidized lipoproteins
The same approach used to visualize bvPLA2 association with F-chyl/VLDLs was carried out with Ox-chyl/VLDLs. Upon ultracentrifugation of a mixture of bvPLA2, Ox-chyl/VLDLs, and BSA, the enzyme was found to be associated with the buoyant oxidized lipoproteins to virtually the same extent as with F-chyl/VLDLs (not shown). Also, enzyme association with the erythrocytes was reconsidered in the presence of oxidized lipoproteins. The enzyme was incubated for 45 min at room temperature with Ox-chyl/VLDLs or with F-chyl/VLDLs at a similar enzyme/lipoprotein (protein) ratio, then incubation proceeded for another 45 min period in the presence of parasitized or healthy erythrocytes. Measurement of the enzymatic activity in the incubation supernatants is reported in Fig. 8
. BvPLA2 preincubated with fresh lipoproteins did not associate with erythrocytes, either infected or not (Fig. 8A), confirming the results reported above and ruling out the possibility that prehydrolysis by bvPLA2 might have changed the binding properties of fresh lipoproteins. By contrast, a dramatic depletion in the supernatant activity (by approximately 88%) was observed when the enzyme had been preincubated with Ox-chyl/VLDLs (Fig. 8B). No loss in activity was observed upon incubation of bvPLA2 with Ox-chyl/VLDLs alone, ruling out the possibility that bvPLA2 depletion in the presence of cells might be due to an oxidized lipoprotein-dependent PLA2 inactivation process. Altogether, our results thus strongly suggest that oxidation of lipoproteins leads to an interaction of yet unknown nature between the lipoprotein particle and the erythrocyte, which might be involved in the oxidation-mediated lipoprotein toxicity process toward P. falciparum.
|
When incubated with lipolyzed chyl/VLDLs, either oxidized or not, trophozoites did not develop into schizonts, and presented increasing alterations ending mainly at the formation of a condensed parasite (pyknotic form) inside the host erythrocyte. Twenty-five to thirty percent abnormal forms were observed as soon as after a 4 h–6 h incubation period (Fig. 9 ). Parasitemia remained unaffected until approximately 16 h of incubation, then decreased to zero during the following period (not shown), in correlation with the observation of extra-erythrocytic parasites in the culture. Therefore, it appeared that bvPLA2-hydrolyzed chyl/VLDLs, either native or oxidized, induced similar phenotypes of parasite death, independent of (i.e., preceding) erythrocyte membrane lysis.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
Binding of free bvPLA2 to chyl/VLDLs was observed through the formation of buoyant enzyme-lipoprotein complexes. These complexes were formed in PBS/BSA, as well as in RPMI, even in the presence of normal erythrocytes (unpublished observations), indicating that under the conditions of cell culture also, the bvPLA2 associates primarily with lipoproteins. It is interesting to note that calcium is present in RPMI, inasmuch as calcium was demonstrated to increase the interfacial affinity of bvPLA2 for mixed PS/PC/cholesterol/sphingomyelin (SM) vesicles with compositions reflecting those of mammalian cell membranes (37) and might have increased the enzyme binding to erythrocytes in RPMI. Yet, inconsistent with this, we could not detect any binding to healthy erythrocytes in the absence (PBS) or in the presence (RPMI) of calcium.
Why bvPLA2 associates with lipoproteins rather than with erythrocytes remains an open question. However, some considerations can be put forward: First, the external leaflet of the erythrocyte membrane is enriched in SM (38), whereas the lipoprotein surface is not, and SM was shown to inhibit sPLA2 activity by perturbing the enzyme anchorage to the membrane (39). Second, the erythrocyte membrane is largely composed of proteins and glycoproteins that might prevent optimal binding of the enzyme. Third, membrane curvature has been postulated to modulate the sPLA2 activity (40, 41), and surface curvatures of disk-shaped erythrocytes and spherical lipoproteins are obviously different.
Oxidation of lipoproteins
We show for the first time that air/light Ox-chyl/VLDLs are inhibitory to P. falciparum in vitro. Modifications induced by such treatment are expected to be mild, as confirmed by the low increase in TBARS concentrations, indicating a low level of lipid peroxidation (42). What is known of the biological and pathophysiological effects of such oxidized lipoproteins comes mainly from studies with Ox-LDL. Minimally or mildly oxidated LDLs exhibit many bioactive properties, including cell death (42–44). Their toxicity is associated with the various bioactive lipids generated by oxidation, which remain bound mainly to the Ox-LDL. The oxidized particle may bind molecules of the cell surface and exchange oxidized lipids with the cell membrane. Then lipids interact with cellular components (molecular targets), either at the cell surface or inside the cell, through multiple mechanisms (29). The main routes of Ox-LDL to the cell are the receptor-mediated pathways (apoB/E and scavenger receptors) (45). Such receptors have not been described in erythrocytes, but the existence of low-affinity binding sites (approximately 200) mediating attachment of native LDL and HDL or lipid vesicles composed of PC or PC/cholesterol to the erythrocyte surface has been published (30). As far as we know, studies on binding of triglyceride-rich lipoproteins or oxidized lipoproteins to Plasmodium-infected erythrocytes have not been reported. Here, we show that incubation of oxidized chyls with normal or parasitized erythrocytes leads to a rapid clearance of the apoB-48 from the supernatant, and that clearance is faster after the Ox-chyl/VLDLs have been lipolyzed by bvPLA2. This was not observed with native or bvPLA2-hydrolyzed fresh lipoproteins, suggesting that physico-chemical modifications due to oxidation, but not to lipolysis, induce the particle interaction with erythrocytes. Because erythrocytes are nonendocytic cells and because it is not clear whether infection by Plasmodium confers new properties to the host cell regarding internalization of macromolecules and, a fortiori, particles, we can speculate that oxidized lipoproteins are not internalized. Yet, the nature of the interaction is currently unknown.
As demonstrated here, the anti-Plasmodium toxicity of bvPLA2-digested fresh lipoproteins is mediated by enzymatically produced NEFAs. The toxic agent(s) in Ox-chyl/VLDLs remain(s) to be identified. Because the protein moiety of lipoprotein is not affected by mild oxidation, it can be assumed that toxicity comes from the lipid part of the particle; PUFAs are very susceptible to reactive oxygen species, and plasma membranes, as well as lipoproteins, are the main targets of peroxidation. The lipid moiety of lipoproteins can undergo major modifications, among which is the formation of peroxides, aldehydes, oxysterols, lysoPLs, and NEFAs, resulting from intrinsic PLA2 (PAF-acetyl hydrolase) activity in apoB-100-containing particles (46). Interestingly, it has been shown that Plasmodium is sensitive to oxidized FAs (28) and to peroxidation-intermediate molecules (47). Changes in the structure of the lipoprotein particle may also contribute to the toxicity by conferring new properties to the lipoprotein. As an example, it has been shown that mild oxidation induces aggregation of LDL, which, in turn, enhances the macrophage-mediated particle degradation (48). However, PLA2 hydrolysis also leads to aggregation of LDL (49), thus suggesting that different and/or additional changes are responsible for the interaction of oxidized chyls with erythrocytes. In our model, this interaction might either confer direct killing potencies to the particles or facilitate the discharge of toxic oxidized lipids to the infected erythrocyte.
Interestingly, Ox-LDL and HDL with increased TBARS content and decreased phospholipid and cholesterol content have been described in malaria patients. Ox-LDL from those patients increased endothelial expression of adhesion molecules, suggesting a detrimental role for oxidized lipoproteins in the pathogenesis of the disease by an increased cytoadherence of the parasitized erythrocytes (17). Our results indicate that oxidized lipoproteins might play a beneficial role in malaria as well, by blocking or slowing down the intraerythrocytic development of P. falciparum. Interestingly, these results are in line with the previously published demonstration that tumor necrosis factor sera containing peroxidized lipoproteins are inhibitory to the in vitro development of P. falciparum (50).
Role of NEFAs
Phospholipids from chyl/VLDLs are substrates of bvPLA2, as confirmed by TLC/GC analysis of the lipids produced. Here, we show that the generated NEFAs are obligatory actors in the anti-Plasmodium toxicity of bvPLA2-digested lipoproteins, because BSA-delipidated lipoproteins are no longer toxic, whereas the lipid-charged BSA becomes toxic. However, the observation that FA-free BSA diminishes the toxicity of the lipolyzed particles suggests that the protein avidity for the lipid molecules prevents their toxic activity, and sustains the idea that lipid delivery to the erythrocytes is a prerequisite to parasite killing (because of BSA retention, fewer lipids would be available for toxicity). The lipid fraction sequestered by BSA comprises mostly NEFAs, especially PUFAs, with LA and AA representing 39.2% and 15.1% of total NEFAs, respectively. The highest anti-Plasmodium activity was found with the long PUFAs: AA and DHA were individually the most lethal to the intraerythrocytic Plasmodium, with IC50s of 5.3 ± 2.6 µM and 7.8 ± 3.2 µM, respectively. LA exhibited a higher IC50, at 45.2 ± 20.7 µM. On the basis of the individual representation of each NEFA in the lipid fraction, it appeared that AA and LA may alone be responsible for parasite killing, and that AA is the best candidate, insofar as it is found at a 4-fold IC50 concentration in the lipid fraction.
The in vitro and in vivo antimalarial effects of NEFAs have long been known (28, 51, 52). Our results are in accordance with the observation by Kumaratilake and coworkers (28) that the antiplasmodial activity of NEFAs is dependent on the number of double bonds of the carbon chain. These authors also showed that toxicity of the PUFAs might relate to their susceptibility to peroxidation. In our case, antioxidants (
-tocopherol, BHT) did not significantly alter the toxicity of the bvPLA2-digested lipoproteins (unpublished observations), suggesting that peroxidation is not primarily responsible for their toxicity. It must be noted that intraerythrocytic degeneration (condensation) of P. falciparum upon incubation with FAs was also observed in those studies (28, 51), reinforcing the idea that NEFAs are directly responsible for the parasiticidal activity of the lipolyzed lipoproteins. Interestingly, the antimalarial activity of DHA, AA, and LA has been demonstrated in vivo: DHA and AA are capable of inhibiting the growth of Plasmodium berghei in BALB/c mice and of preventing the malaria-induced anemia (28), whereas LA provokes a substantial reduction of parasitemia in Swiss mice infected with Plasmodium vinckei petteri (51). A massive supply in AA or LA is susceptible to being toxic via multiple pathways, either by perturbing the FA metabolism of the parasite, or by more specific effects, such as alteration of heme detoxification in the case of LA (53), or perturbation of the parasite prostaglandin production in the case of AA (54). Also, because many different NEFAs, with numerous bioactive properties, are generated by the enzymatic hydrolysis of lipoproteins, we cannot exclude a toxic mechanism implying all or several of these FAs. Interestingly, Rigoni et al. (55) recently reported similar effects of lysoPLs-FA mixtures and neurotoxic sPLA2s on the neuromuscular junction. LysoPL and NEFAs generated by enzymatic hydrolysis of the cell membrane are shown to act synergistically to paralyze the neuromuscular junction. In line with this, it would be worthwhile to measure the anti-Plasmodium potencies of joined NEFAs and lysoPC, because lysoPC, despite moderate toxicity, represents a large part of the hydrolysis products.
|
CONCLUSION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
It has been shown that the intrusion and development of the malarial parasite increase the systemic rate of inflammatory sPLA2-IIA in humans (14, 15), and implication of the enzyme in the host defense against the parasite has been suggested. Nine catalytically active sPLA2s have been identified in humans, with each sPLA2 displaying a specific phospholipid affinity (56) and a specific pattern of expression (57, 58). Like venomous sPLA2s, human sPLA2s can hydrolyze lipoproteins. Groups V and X sPLA2s are capable of hydrolyzing anionic and neutral PLs and are 20-fold more effective in hydrolyzing lipoproteins, as compared with sPLA2-IIA, which hydrolyzes almost exclusively anionic PLs (59). However, as far as we know, no sPLA2 different from the sPLA2-IIA in plasma, cells, or tissues from malarial patients has been reported. Remarkably, the hydrolysis efficiency of the sPLA2-IIA is enhanced 26-fold when the lipoproteins are oxidized (16), and it has been shown recently that lipoprotein fluidity and oxidative degree are enhanced in malaria patients, in relation to the severity of the case (17). If we place our results in the malaria context, the alteration of the physiological parameters in malaria patients might be considered as a modification of the host physiology toward an environment more favorable to the anti-Plasmodium activity of sPLA2s and a direct participation of the oxidized lipoproteins in the host defense. Moreover, this innate defense mechanism deserves to be considered in a more general view, because elevation of the sPLA2-IIA concentration in plasma and oxidation of lipoproteins is a recurrent scheme in infections by pathogenic microorganisms and inflammatory diseases.
|
ACKNOWLEDGMENTS |
|---|
Manuscript received March 7, 2006 and in revised form April 3, 2006.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
