Reduced Peripheral PGE2 Biosynthesis in Plasmodium falciparum Malaria Occurs through Hemozoin-Induced Suppression of Blood Mononuclear Cell Cyclooxygenase-2 Gene Expression via an Interleukin-10-Independent Mechanism
- 3 Downloads
Molecular immunologic determinants of disease severity during Plasmodium falciparum malaria are largely undetermined. Our recent investigations showed that peripheral blood mononuclear cell (PBMC) cyclooxygenase-2 (COX-2) gene expression and plasma prostaglandin E2 (PGE2) production are suppressed in children with falciparum malaria relative to healthy, malaria-exposed children with partial immunity. Furthermore, decreased COX-2/PGE2 levels were significantly associated with increased plasma interleukin-10 (IL-10), an anti-inflammatory cytokine that inhibits the expression of COX-2 gene products. To determine the mechanism(s) responsible for COX-2-derived PGE2 suppression, PBMCs were cultured from children with falciparum malaria. PGE2 production was suppressed under baseline and COX-2-promoting conditions (stimulation with lipopolysaccharide [LPS] and interferon [IFN]-γ) over prolonged periods, suggesting that an in vivo-derived product(s) was responsible for reduced PGE2 biosynthesis. Ingestion of hemozoin (malarial pigment) by PBMC was investigated as a source of COX-2/PGE2 suppression in PBMCs from healthy, malaria-naive adults. In addition, synthetically prepared hemozoin, β-hematin, was used to investigate the effects of the core iron component of hemozoin, ferriprotoporphyrin-IX (FPIX). Physiologic concentrations of hemozoin or β-hematin suppressed LPS- and IFN-γ-induced COX-2 mRNA in a time- and dose-dependent manner, resulting in decreased COX-2 protein and PGE2 production. Suppression of COX-2/PGE2 by hemozoin was not due to decreased cell viability as evidenced by examination of mitochondrial bioactivity. These data illustrate that ingestion of FPIX by blood mononuclear cells is responsible for suppression of COX-2/PGE2. Although hemozoin induced overproduction of IL-10, neutralizing IL-10 antibodies failed to restore PGE2 production. Thus, acquisition of hemozoin by blood mononuclear cells is responsible for suppression of PGE2 in malaria through inhibition of de novo COX-2 transcripts via molecular mechanisms independent of increased IL-10 production.
Although the underlying pathophysiologic determinants of severe malaria are only partially defined, cytokines and effector molecules are important immunologic determinants for regulating disease susceptibility. Our recent investigations have focused on defining the role of prostaglandins as immunomodulatory factors in malaria. Eicosanoids, such as prostaglandin E2 (PGE2), regulate macrophage function, vascular permeability, extracellular adhesion molecules, fever, and cytokine production (1). PGE2 formation is initiated by the actions of phospholipase A2, which releases arachidonic acid from the membrane phospholipid bilayer in response to inflammatory stimuli (2). Free arachidonic acid is rapidly metabolized to prostaglandin H2 (PGH2) via the enzymatic activity of the cyclooxygenase (COX) enzymes, of which there are 2 distinct isoforms encoded by separate genes: COX-1 and COX-2 (1,3). COX-1 is constitutively expressed in most tissues and generates PGs for physiologic homeostasis, whereas COX-2 is an inducible gene that generates high levels of PGs during an inflammatory event (4). Following formation of PGH2 by COX, PGE2 synthase converts PGH2 to PGE2 (5).
Our previous studies showed that plasma levels of PGE2 were significantly reduced in children with malaria and were significantly associated with increased plasma interleukin (IL)-10 levels (6). Furthermore, peripheral blood mononuclear cell (PBMC) COX-2 gene products were inversely related to disease severity (6). We have also shown that children with cerebral malaria (CM) have suppressed systemic levels of PGE2 (7). Although suppression of COX-2/PGE2 in our previous studies was associated with enhanced severity of falciparum malaria, the mechanism(s) for decreased PGE2 synthesis remain undefined.
During the erythrocytic stage of malaria, Plasmodia digest host hemoglobin, releasing heme in the process. To avoid the toxic effects of heme, enzymes within the parasitiphorus vacuole catalyze the polymerization of heme into a nontoxic insoluble compound, hemozoin (malarial pigment) (for review, see Goldie and others ). Hemozoin is composed predominately of protein and a smaller fraction of carbohydrates, lipids, and nucleic acids, which are attached to the polymerized iron core structure, ferriprotoporphorin-IX (FPIX) (9). Upon rupture of infected erythrocytes, hemozoin is released into the blood stream and phagocytosed by circulating monocytes and tissue macrophages (10). Hemozoin is also acquired by monocytes/macrophages through phagocytosis of parasitized erythrocytes (11). Experiments in cultured blood mononu-clear cells illustrate that hemozoin augments the release of both proinflammatory cytokines, such as tumor necrosis factor (TNF)-α (12), and anti-inflammatory cytokines, such as IL-10 (13). This appears important in the current context because TNF-α induces high levels of sustained COX-2 gene expression (14), whereas IL-10 decreases COX-2 gene expression through destabilization of COX-2 message (15). Additional effects of hemozoin on blood mononuclear cells include inhibition of oxidative burst and protein kinase C activity (16), reduction of microbicidal and antitumor capabilities (17), impaired expression of MHC class II and CD54 expression (10), and an inability of monocytes to undergo repeated phagocytosis (18).
In the present investigations, PGE2 production was examined in cultured PBMCs from children with falciparum malaria under baseline conditions and following treatment with COX-2/PGE2-inducing stimuli: lipopolysaccharide (LPS) and interferon (IFN)-γ. In addition, COX-2 mRNA and protein and PGE2 formation were examined in cultured PBMCs from malaria-naive healthy donors stimulated with LPS and IFN-γ in the presence of physiologically relevant doses (10, 1.0, and 0.1 µg/mL ) of hemozoin or β-hematin (synthetic malarial pigment). The effect of hemozoin on IL-10 production and the ability of IL-10 to down-regulate PGE2 biosynthesis were also investigated.
Materials and Methods
Children (n = 30, age 2 to 7 y) were recruited from a longitudinal prospective study at the Albert Schweitzer Hospital in Lambaréné, Gabon, in the province of Moyen Ogooue. In this hyperendemic area of malaria transmission, the primary clinical manifestations of severe childhood malaria is severe anemia and/or hyperparasitemia, with cerebral malaria occurring rarely (20). Classification of malaria was defined according to our previously published methods (6,21). Only those children with mild malaria (defined by parasitemias < 100000 parasites/µL and the absence of any signs or symptoms of severe malaria [6,20]) were included in the present study because of the limited amount of blood available from severe cases. Healthy, malaria-exposed subjects were defined as participants with a previous episode(s) of malaria and the absence of a positive thick blood film for malaria, or any other illnesses, within the last 4 wk. All blood samples were obtained prior to treatment with antimalarials. Informed consent was obtained from the parents of participating children.
Healthy, malaria-naive adult donors (n = 21) were recruited from the University of Pittsburgh, PA, USA. The study was approved by the ethics committee of the International Foundation of the Albert Schweitzer Hospital, Duke University Medical Center, and the University of Pittsburgh Investigational Review Boards.
Isolation and Culture of Peripheral Blood Mononuclear Cells
For in vitro experiments in Gabonese children (n = 30) and US adults (n = 21), venous blood (3 and 40 mL, respectively) was drawn into EDTA-containing vials. PBMCs were prepared using ficoll/Hypaque, as described earlier (22), and plated at 1 × 106 cells per mL in Dulbecco’s modified Eagle medium (DMEM) containing 10% pooled human serum (heat inactivated at 56°C for 30 min). Cells were incubated with media alone (controls) or concomitant stimulation with LPS (100 ng/mL; Alexis Corp., San Diego, CA, USA) and IFN-γ (200 U/mL; BD Pharmingen, San Diego, CA, USA). For experiments in healthy adults, cultures were incubated with varying concentrations of hemozoin (10, 1.0, and 0.1 µg/mL) or β-hematin (10, 1.0, and 0.1 µg/mL) in the presence and absence of LPS and IFN-γ stimulation. For neutralization of endogenous IL-10, cultures were stimulated with IL-10 neutralizing antibodies (1.0, 0.5, and 0.1 µg/mL).
Hemozoin and β-hematin Preparation
Crude hemozoin was isolated from in vitro cultures of P. falciparum-infected red blood cells (strain Pf-D6) according to our previously published methods (19). Briefly, cultured red blood cells (RBC) were treated with 2 mL saponin for 10 min, and the solution was washed in phosphate-buffered saline and spun at 14000 rpm for 15 min (4 to 7 times). The final pellet was dried, weighed, and resuspended in filter-sterilized H2O at a concentration of 1.0 mg/mL. As a control, a parasite-free RBC lysate was prepared from uninfected RBC following the protocol outlined above for the hemozoin preparation.
β-Hematin was formed in a 4.5-M acidic acetate solution at pH 4.5 by the method of Egan and others (23). Briefly, hemin chloride (Sigma, St. Louis, MO, USA) was added to a 0.1-M solution of NaOH followed by addition of HCl and acetate. The mixture was incubated for 150 min at 60 °C without stirring, spun at 14000 rpm, washed 3 times with filter-sterilized H2O, and dried at 60 °C under vacuum. The final pellet was weighed and resuspended at 1.0 mg/mL in filter-sterilized H2O. Resuspended β-hematin was sonicated extensively to disperse the preparation. Under these preparation conditions, β-hematin has an identical infrared spectroscopy pattern as detergent-purified hemozoin, and β-hematin crystals appear morphologically similar to natural hemozoin crystals as evidenced by scanning electron microscopy (23).
PGE2 and IL-10 Measurements
Concentrations of PGE2 and IL-10 were measured in the supernatants of cultured PBMCs using a quantitative indirect sandwich enzyme immunoassay (Cayman Chemical, Ann Arbor, MI, USA) and quantitative sandwich ELISA (Pharmingen, San Diego, CA, USA), respectively, according to our previously defined methods (24). Sensitivity of detection for PGE2 and IL-10 was greater than 7.8 pg/mL.
Cell viability was determined by the MTT assay, which measures active mitochondrial conversion of MTT salt into formazan crystals. Briefly, 20 µL of MTT salt (5 mg/mL in 1X PBS [Sigma]) was added to cultured cells 5 h prior to termination of cultures. At the end of culture, supernatants were aspirated and 200 µLof dimethyl sulfoxide was added to dissolve the formazan crystals. Plates were incubated for 5 min at room temperature. Absorbance was determined at 550 nm.
Cellular extracts were prepared according to our previously defined methods (25). Protein samples (30 µg) were fractionated on a 10% sodium dodecyl sulfate-polyacrylamide gel, transferred to nitrocellulose membranes, and COX-2 antigen was measured by our previously defined methods (6). Extracts of mouse macrophage cells (RAW 264.7), stimulated with LPS (Escherichia coli LPS 0127B8, 1.0 µg/mL) and IFN-γ (10 ng/mL) were used as a positive control. Densitometry readings for COX-2 protein expression were obtained using commercially available software (Scion Image, Scion Corp, http://www.scioncorp.com).
Quantitative Real Time Reverse Transcriptase-Polymerase Chain Reaction (RT-PCR)
Total RNA was isolated from PBMCs by the GITC method (26). Total RNA (1 µg) was reverse transcribed into cDNA, and COX-2 gene expression was analyzed by quantitative real time RT-PCR on an ABI Prism 7700 sequence detection system (Applied Biosystems, Foster City, CA, USA). cDNA (100 ng) was amplified in duplicate with COX-2 specific primers and probe as follows: forward, 5′-CCTTCCTCCTGTGCCTGATG-3′; reverse, 5′-ACAATCTCATTTGAATCAGGAAGCT-3′; probe (FAM labeled), 5′-TGCCCGACTCCCTTGGGTGTCA-3′ (sequence information obtained from Sales and others ). The COX-2 cDNA was amplified for 40 cycles following the parameters specified by the manufacturer (Applied Biosystems). To control for nonspecific background fluorescence, no template controls were included in triplicate. An endogenous control gene, β-actin (Ascession number NM_001101 [Applied Biosystems]), was used as a reference gene to normalize cDNA loading between samples. Data was compared using the -ΔΔCT method as previously described (19).
Supernatant concentrations of PGE2 (pg/mL) and IL-10 (pg/mL) were measured in triplicate and at several different dilutions. COX-2 mRNA expression was measured in duplicate. Pairwise comparisons between conditions, including pairwise tests of dose-dependent effects, were performed using the Mann-Whitney U test. Multiple group comparisons for dose-dependent effects were performed using the Kruskal-Wallis test (statistical significance set at P ≤ 0.05).
PGE2 Production in Cultured PBMCs from Children with Malaria
Effect of Hemozoin on PGE2 Production
Because suppression of PGE2 by hemozoin could result from hemozoin-induced cell death, the temporal profile of cell viability was examined using the MTT assay. LPS and IFN-γ stimulation decreased cell viability at 72, 96, and 120 h (P < 0.05), with the addition of hemozoin having no effect on cell viability (Figure 2B). Thus hemozoin significantly suppresses PGE2 through mechanisms that are independent of altering cell viability.
Effect of Hemozoin and β-hematin on COX-2 Gene Products
Effect of Endogenous IL-10 on Hemozoin-Induced PGE2 suppression
Because addition of hemozoin to PBMCs increased IL-10 production and suppressed COX-2-derived PGE2 formation, we determined whether neutralizing antibodies to IL-10 could restore hemozoin-induced reductions in PGE2. As shown in Figure 7B, addition of IL-10 neutralizing antibodies failed to restore PGE2 production. Measurement of IL-10 confirmed that endogenously produced IL-10 was dose-dependently reduced in culture supernatants with increasing concentrations of IL-10 neutralizing antibodies (see Figure 7C). Determination of IL-10 transcripts by real-time RT-PCR showed that the addition of IL-10-neutralizing antibodies had no effect on IL-10 transcript levels (data not shown). In addition, the addition of isotype-matched control antibodies had no effect on IL-10 levels (data not shown). Thus, suppression of COX-2-derived PGE2 occurs through an IL-10-independent mechanism.
Our previous studies demonstrated that plasma bicyclo-PGE2 and ex vivo blood mononuclear cell COX-2 levels are suppressed in children with malaria (6). In the present study, we extend those findings by showing that cultured PBMCs from children with malaria have reduced baseline and stimulated PGE2 production. Because cells were cultured for 120 h to avoid influence of host-derived in vivo regulatory factors, we postulated that suppression of PGE2 was due to a phagocytosed parasitic product(s). Because hemozoin can persist in mononuclear cells for prolonged periods (28), the effect of hemozoin on PGE2 was examined. A crude hemozoin isolate was used because this preparation mimics the hemozoin moiety acquired during a natural infection. Consistent with studies illustrating that phagocytosis of hemozoin by monocytic cells occurs within 24 h with the biologic effects occurring over prolonged periods (29,30), hemozoin significantly decreased PGE2 synthesis in PBMCs from 48 h onward.
The molecular mechanism responsible for hemozoin-induced PGE2 suppression does not occur through reduced cell viability, because the addition of hemozoin to stimulated PBMCs did not alter mitochondrial bioactivity (that is, MTT assay) at the time points selected for the mechanism-based studies (48 and 72 h). It is important to note that after 72 h in culture, hemozoin reduced cell viability, which was further reduced by LPS and IFN-γ treatment. Thus, it was difficult to interpret experimental results in which cells were pretreated with hemozoin followed by stimulation. Although, results presented here clearly illustrate that hemozoin suppresses COX-2-derived PGE2 biosythesis, it appears that the natural acquisition of hemozoin in the cytokine-activated in vivo milieu differs slightly from that offered by the in vitro environment. This rationale is supported by the fact that hemozoin reduced PGE2 biosynthesis by 15% to 53% in cells from healthy donors versus the 70% reduction observed in children with malaria. The apparent discrepancy between in vitro hemozoin-induced reductions in PGE2 and naturally acquired hemozoin in mononuclear cells from children with malaria may be due to additional host- and/or parasite-derived factors independent of hemozoin. We are currently investigating the effects of additional host immune factors and parasitic products that may alter PGE2 synthesis during acute malaria.
While we consistently observed PGE2 suppression in PBMCs stimulated with hemozoin, there was substantial variation in PGE2 production in different individuals. To address this variation, data were expressed as percent reduction in PGE2 relative to stimulated conditions in a larger number of individuals (n = 18) at time-points when the cell viability was greatest (that is, 48 and 72 h). Expression of data in this manner “normalized” individual variation and demonstrated a dose-dependent suppressive effect of hemozoin on PGE2 biosynthesis. These experiments further demonstrated that β-hematin dose-dependently reduced PGE2 production. Because β-hematin is a synthetically prepared compound of FPIX, which lacks host- and parasite-derived proteins and lipids, and is structurally identical to purified (detergent-treated) hemozoin (31,32), it appears that FPIX is responsible for suppression of PGE2.
Our previous results showed that ex vivo PBMC COX-2 protein and mRNA were decreased in children with malaria (6). This, along with our previous data showing that de novo induction of COX-2 message is required for high levels of sustained PGE2 production (25), prompted us to investigate hemozoin-induced changes in COX-2 expression. Hemozoin and β-hematin dose-dependently suppressed COX-2 mRNA and protein, indicating that PGE2 suppression is caused by reduced levels of COX-2 gene products. Results from kinetic experiments illustrated that hemozoin inhibits de novo COX-2 message at 8 and 24 h. Thus, reductions in COX-2 protein and PGE2 formation at 48 h are consistent with decreased COX-2 transcripts at earlier time points.
Previous studies in individuals with CM showed increased COX-2 protein in Durck’s granulomas within the central nervous system (33). However, our results in plasma and PBMCs from children with malaria (6), and here in hemozoin-treated PBMCs, illustrate that COX-2/PGE2 is suppressed in circulating blood mononuclear cells and plasma. Our recent results in Tanzanian children with CM also show that peripheral PGE2 levels are decreased during acute disease, with the most profound reductions in PGE2 being associated with increased disease sequeale (7). We have further shown that placental monocytes from women with malaria during pregnancy have decreased PGE2 formation; there was a dose-dependent reduction in PGE2 with increasing amounts of naturally acquired hemozoin (24). Recent studies in a murine model of CM also showed that COX-2 transcripts were significantly increased within the brain, but non-significantly decreased in PBMCs (34). Furthermore, infection of mice with P. berghei K173, a parasite strain causing anemia in the absence of CM, significantly decreased peripheral COX-2 transcripts (34). Thus, during acute malaria, COX-2/PGE2 regulation within the brain differs from COX-2/PGE2 regulation in the local tissue compartments and in circulating blood monocytes.
Our previous studies showed that decreased plasma PGE2 levels were significantly associated with increased plasma IL-10 concentrations (6). This is consistent with data illustrating that IL-10 decreases PGE2 through suppression and/or destabilization of COX-2 transcripts (15). As such, we determined if hemozoin-induced PGE2 suppression occurred through increased IL-10 production. Hemozoin augmented LPS- and IFN-γ-induced IL-10 production, however hemozoin-induced IL-10 production does not appear responsible for PGE2 suppression because IL-10 neutralizing antibodies failed to restore PGE2. Thus, the molecular mechanism(s) by which hemozoin suppresses COX-2/PGE2 remains unclear. Previous studies show that activation of the protein kinase C (PKC) signaling pathway is essential for COX-2 induction in human intestinal myofibroblasts (35) and that inhibition of PKC attenuates IL-1β induction of COX-2/PGE2 in human pulmonary epithelial cells (36). This appears important for results presented here because previous studies found that hemozoin suppresses PKC activity in cultured monocytes (16). We are currently determining if reduced levels of COX-2 in malaria occur through inhibition of PKC signal transduction pathways.
Although decreased systemic PGE2 levels are associated with increasing disease severity, it is unclear how suppressed peripheral PGE2 promotes enhanced malaria pathogenesis. Potential pathogenic effects may occur through suppression of erythropoiesis because previous studies demonstrate that PGE2 enhances burst-forming unit erythroid formation (37), suggesting that suppression of systemic PGE2 by hemozoin may reduce formation of new RBC. Additional mechanisms by which decreased systemic PGE2 could promote increased pathogenesis may occur through overexpression of the pro-inflammatory response. For example, PGE2 is important for resolution of the late-phase immune response through promotion of TH2-derived cytokine production (for review, see Phipps and others ). PGE2 also regulates the proinflammatory response by suppressing de novo TNF-α transcription (39) and by modulating the secretion of soluble TNF-α receptors (40). PGE2 may also alter malaria pathogenesis through modulation of lymphocytic activity. Previous studies in cultured human mononuclear cells from individuals with acute falciparum malaria showed that the non-selective COX inhibitor, indomethacin, reversed malarial antigen-induced suppression of lymphoproliferation (41). However, levels of COX-2/PGE2 were not determined in these studies (41). Subsequent studies showed that exogenous addition of PGE2 failed to suppress T-cell proliferation (42). The apparent discrepancy in these results may be related to the fact that indomethacin alters a number of cellular functions in a COX-independent fashion (for review, see Tegeder and others ). We are currently investigating the effects of COX-2/PGE2 suppression on malaria pathogenesis in children with severe malarial anemia.
We thank the staff members of the Albert Schweitzer Hospital in Lambaréné, Gabon, for their cooperation and technical assistance: Anita van den Biggelaar, Judith Jans, Hanna Knoop, Doris Luckner Barbara Moritz, Anselme Ndzengue, Marcel Nkeyi, Daniela Schmid, and Milena Sovric. Furthermore, we would like to thank Dr. Venkatachalam Udhayakumar for his scientific input.
This work was conducted at the University of Pittsburgh and was supported in part by the National Institutes of Health Grants AI-51305-01 (DJP) and AI-41764 (JBW), the VA Research Service (JBW), and the University of Pittsburgh Competitive Research Development Fund (DJP). Portions of this work were presented at The 51st Annual Meeting of the American Society of Tropical Medicine and Hygiene in Denver, CO, in November 2002. The title of the presentation was Cyclooxygenase-2 and Prostaglandin E2 Down-regulation by Hemozoin is Associated with Overproduction of TNF-α Implications for the Pathogenesis of Severe Malaria (abstract number 74).
The study was approved by the ethics committee of the International Foundation of the Albert Schweitzer Hospital, Duke University Medical Center Investigational Review Board, and the University of Pittsburgh Investigational Review Board, and informed consent was obtained from the participants or the parents of participating children.
There is no conflict of interest for any of the authors of the manuscript due to commercial or other affiliations.
- 7.Perkins DJ, Hittner JB, Mwaikambo ED, Granger DJ, Weinberg JB, Anstey NM. (2004) Impaired systemic prostaglandin E2 production in children with cerebral malaria. J. Infect. Dis. In press.Google Scholar
- 13.Mordmuller B, Turrini F, Long H, Kremsner P, Arese P. (1998) Neutrophils and monocytes from subjects with the Mediterranean G6PD variant: Effect of Plasmodium falciparum hemozoin on G6PD activity, oxidative burst and cytokine production. Eur. Cytokine Netw. 9:2339–246.Google Scholar