点击显示 收起
University of Pittsburgh Graduate School of Public Health, Pittsburgh, Pennsylvania
College of Charleston, Charleston, South Carolina
VA and University of Utah School of Medicine, Salt Lake City, Utah
VA and Duke University Medical Centers, Durham, North Carolina
Hubert Kairuki Memorial University, Dar es Salaam, Tanzania
Menzies School of Health Research and Charles Darwin University, Darwin, Australia
Prostaglandins (PGs) are important mediators of macrophage activity, vascular permeability, fever, erythropoiesis, and proinflammatory responses to infection. Our recent studies have shown that plasma levels of bicyclo-PGE2 (a stable end product of PGE2 metabolism) and leukocyte cyclooxygenase (COX)2 gene expression are suppressed in children with malarial anemia. Since the role of PGs as immunomodulators of human cerebral malaria (CM) has not been examined, we investigated urinary levels of bicyclo-PGE2/creatinine in children with varying clinical outcomes of CM. Among parasitemic children, those with asymptomatic parasitemia had the highest levels of bicyclo-PGE2/creatinine, whereas those with CM had significantly lower levels of bicyclo-PGE2. Systemic levels of bicyclo-PGE2/creatinine were not significantly associated with parasitemia, hemoglobin levels, or levels of the PG-regulatory cytokine tumor necrosis factor but were positively correlated with levels of interleukin-10. The results presented here show that impaired systemic production of PGE2 is associated with adverse outcomes of CM, whereas elevated levels of PGE2 in asymptomatic parasitemia suggest a potential role for PGs in protective immunity.
Malaria is estimated to cause >2 million deaths annually [1]. Although there are diverse clinical presentations of malaria, cerebral malaria (CM) is a common feature of severe malaria in nonimmune and semi-immune individuals residing in areas with low to moderate rates of Plasmodium falciparum transmission. CM is a neurological syndrome characterized by encephalopathy, coma, and seizures. Although the underlying molecular mechanisms of CM have been only partially defined, at least 4 distinct groups of children whose conditions fulfill the World Health Organization definition of CM have been described [2]. The central features in most children with CM include sequestration of parasitized red blood cells (pRBCs) within the cerebral microvasculature [3] and excessive systemic production of proinflammatory cytokines, such as tumor necrosis factor (TNF), and of anti-inflammatory cytokines, such as interleukin (IL)10 [48]. In addition, our previous studies have shown that, in Tanzanian children with malaria, systemic levels of nitric oxide (NO) and leukocyte expression of nitric oxide synthase type 2 (NOS2) protein are inversely associated with disease severity [7].
The relative balance of pro- and anti-inflammatory cytokines in the inflammatory environment regulates the expression of inducible genes, such as NOS2 and cyclooxygenase (COX)2, which, in turn, mediate the immune response through generation of high levels of NO and prostaglandins (PGs), respectively [9]. PGs are generated in response to cellular activation that triggers the release of arachidonic acid (AA) from membrane glycerophospholipids by the actions of phospholipase A2 [10]. Free AA is then converted into prostanoids (e.g., PGE2, PGF2, PGD2, prostacyclin [PGI2], and thromboxane-A2) by the enzymatic activity of the COXs (COX-1 and COX-2 [also termed PGH synthase1 and 2]) [11]. COX-1generated PGs are typically important in maintenance of physiological homeostasis, whereas COX-2generated PGs are important in inflammation and host defense [9]. The most extensively characterized prostanoid, PGE2, regulates a number of key immunological processes relevant to malaria pathophysiology, including expression of cell adhesion molecules, release of proinflammatory mediators (such as TNF-), vascular permeability, and tissue fluid dynamics [11]. In addition, inhibition of production of PGE2, by nonsteroidal anti-inflammatory drugs, can suppress certain leukocyte functions, such as chemotaxis, phagocytosis, reactive oxygen species generation, and microbial killing (for review, see [12]).
Our recent studies of Gabonese children with severe malaria (characterized by severe anemia and high-density parasitemia) have shown that plasma levels of bicyclo-PGE2 (a stable end product of PGE2 metabolism) and leukocyte COX-2 gene expression are significantly reduced during acute malaria [13]. Although measurement of PGE2 and COX-2 gene expression in those studies was performed before treatment with antimalarials and/or antipyretics, previous studies of Gabonese children within the same geographic region have shown that treatment of malaria-infected children with acetaminophen (paracetamol) is associated with adverse outcomes, such as prolonged time to clearance of parasites and decreased production of TNF- and oxygen radicals [14]. Studies using murine models of CM have shown that blockade of formation of PGs, by aspirin, is associated with increased mortality [15] and that more-selective inhibition of COX-2 by celecoxib is associated with an earlier onset of CM [16]. Taken together, the findings of these studies suggest that enhanced production of PGE2 is associated with protective effects and that reduced levels of PGE2 may promote increased pathogenesis.
Since the role of PGE2 in mediating protection against human CM is not known, we sought to determine the relationship between systemic production of PGE2 and disease severity in children with and children without CM. We hypothesized that urinary levels of bicyclo-PGE2 in Tanzanian children would be inversely associated with malaria disease severity. We also determined the relationship between production of PGE2 and levels of the COX-2regulatory cytokines TNF- and IL-10, peripheral parasitemia, and anemia status (hemoglobin levels).
PATIENTS, MATERIALS, AND METHODS
Study participants.
The present study was performed using cryopreserved samples collected from subjects enrolled at Muhimbili Medical Centre (MMC), Dar es Salaam, Tanzania, from May 1994 to January 1995, as described elsewhere [7]. The original study was approved by the College Research and Publications Committee of MMC and the Duke University Medical Center Investigational Review Board, and informed consent was obtained from the parents of participating children in Kiswahili. Five groups of children 6 months to 9 years of age were prospectively recruited from the pediatric and surgical wards: (1) healthy control subjects (CONs) were characterized by absence of fever within the past 2 weeks, absence of parasites on thick blood film examination, a normal white blood cell (WBC) count, and a lack of any other detectable acute illnesses (a fracture >1 week old was permitted); (2) patients with asymptomatic parasitemia (AP) were characterized as were CONs, except that they had the presence of P. falciparum parasitemia on thick blood film examination; (3) patients with noncerebral malaria (NCM) were characterized by a febrile illness, with P. falciparum parasitemia >10,000 trophozoites/L, no history of convulsions, and no other evident cause of fever, and were fully alert, normoglycemic, and without severe respiratory distress; (4) patients with CM with complete recovery (CMCR) were characterized by unarousable coma with a Blantyre coma score 2 that persisted for 30 min after the last convulsion, no other cause of coma evident from clinical analysis or analysis of cerebrospinal fluid, any level of P. falciparum parasitemia on thick blood film examination, and recovery without neurological sequelae; and (5) patients with CM complicated by death or neurological sequelae (CMDS) were characterized as were patients with CMCR, except that they had coma accompanied by death or neurological sequelae on discharge. Patients with malaria were managed according to standard MMC guidelines.
Collection of samples.
Venous blood was drawn into sterile EDTA-containing vacutainers. Hemoglobin levels and WBC counts were determined by use of a Coulter counter. Thick and thin blood films were prepared; thick blood films were stained with Field's stains A and B, and thin blood films were stained with a Giemsa stain. The number of parasites per 200 WBCs was determined from thick blood films, and parasitemia (per microliter of whole blood) was calculated from the automated WBC count. An experienced microscopist examined 50 oil-immersion fields before a film was classified as negative. Urine was collected on admission and immediately stored at -70°C. Plasma (13 mL) was isolated from whole blood within 30 min of collection and was immediately frozen at -70°C, for later determination of cytokines.
Measurement of bicyclo-PGE2 and cytokines.
Since PGE2 is rapidly converted to 13,14-dihydro-15-keto-PGE2 in an in vivo environment, levels of bicyclo-PGE2 were measured. Urine (500 L) was precipitated in 2 mL of ethanol for 15 min at 4°C. Samples were purified by solid-phase extraction, by use of 500 mg/3 mL DSC-18 (octadecyl bonded silica; Suppelco) tubes. PGs were eluted with ethyl acetate containing 1% methanol and were evaporated to dryness by vacuum centrifugation, followed by resuspension in commercially available EIA buffer (500 L; Cayman Chemical). To convert PGE2 and the intermediary metabolites to bicyclo-PGE2, samples were derivatized overnight at 37°C in carbonate buffer (150 L), followed by addition of PBS (200 L) and EIA buffer (150 L). Levels of bicyclo-PGE2 were determined by use of a quantitative sandwich EIA (Cayman Chemical). Urinary levels of creatinine were determined by use of an Ektachem autoanalyzer (Eastman Kodak). Plasma levels of human TNF- and IL-10 were determined by use of a quantitative sandwich EIA, as reported elsewhere [7]. The sensitivities of detection of the cytokine assays were 15.6 pg/mL (TNF-) and 4.7 pg/mL (IL-10).
Statistical analyses.
Urinary levels of bicyclo-PGE2 (picograms per milliliter) were measured in triplicate and at several different dilutions. Urinary levels of creatinine (milligrams per milliliter) were measured in triplicate, and plasma levels of cytokines (picograms per milliliter) were measured in duplicate. To account for potential differences in renal filtration rates between the various groups, urinary levels of creatinine were measured, and data were expressed as urinary levels of bicyclo-PGE2/creatinine (picograms per milligram per milliliter). To examine potential differences between groups in parasitemia and levels of bicyclo-PGE2/creatinine, TNF-, IL-10, and hemoglobin, a 1-way analysis of covariance (ANCOVA) was performed on each variable, with age serving as a covariate. After each significant F test, post hoc pairwise mean comparisons were conducted by use of Fisher's least significant difference (LSD) multiple range test. The Fisher's LSD procedure effectively controls the family-wise type I error rate and accommodates unequal sample sizes across groups. To examine the strength of linear association between pairs of variables, bivariate zero-order and partial correlation analyses were conducted. Before performance of inferential statistical analyses, the following outcome variables were logarithmically transformed toward normality: levels of bicyclo-PGE2/creatinine, TNF-, and IL-10 and parasitemia. Hemoglobin levels did not require transformation, since they did not evidence significant skewness or kurtosis. Unless otherwise noted in Results, all ANCOVA and correlational analyses used the transformed versions of the aforementioned variables. The only exception to this rule concerned hemoglobin levels, which, as noted above, did not require transformation. For all of the analyses presented, there were no missing data for any of the subjects.
RESULTS
Patient characteristics.
The untransformed variables for the patient characteristicsincluding age (years), peripheral parasitemia (trophozoites per microliter), hemoglobin level (grams per deciliter), and levels of bicyclo-PGE2 (picograms per milliliter), creatinine (milligrams per milliliter), TNF- (picograms per milliliter), and IL-10 (picograms per milliliter)are presented in table 1 for the 5 groups participating in the study: CON (n = 18), AP (n = 8), NCM (n = 31), CMCR (n = 14), and CMDS (n = 19).
Relationship between systemic levels of PGE2 and those of cytokines.
Since TNF- and IL-10 are important for regulation of production of PGs, through induction and suppression of COX-2 gene expression, respectively [17], plasma levels of TNF- and IL-10 were examined in children with malaria, to determine whether changes in these COX-2regulatory cytokines could account for the altered production of PGs between the groups. Although levels of TNF- and IL-10 increased with disease severity, ANCOVA (with adjustment for age) failed to indicate significant between-group differences in levels of either cytokine (TNF-: F[2,44] = 1.07, P = .353; IL-10: F[2,44] = 0.36, P = .700) (table 1). The log-transformed, covariate-adjusted mean levels of TNF- (i.e., adjusted for age) for the NCM, CMCR, and CMDS groups were 1.50, 1.59, and 1.82, respectively. Similarly, the log-transformed, covariate-adjusted mean levels of IL-10 for the same 3 groups were 2.10, 2.28, and 2.30, respectively. No association was found between plasma levels of bicyclo-PGE2/creatinine and those of TNF- (r = 0.11; P = .461) (table 2). There was, however, a significant association between plasma levels of bicyclo-PGE2/creatinine and those of IL-10 (r = 0.34; P = .019) (table 2).
Partial correlations were also calculated, to determine the influence of age and disease severity on the relationship between levels of bicyclo-PGE2/creatinine and those of the cytokines. The association between levels of bicyclo-PGE2/creatinine and those of TNF- remained nonsignificant (rp = 0.08; P = .607), whereas the association between levels of bicyclo-PGE2/creatinine and those of IL-10 remained significant (rp = 0.31; P = .035).
Relationship between systemic levels of PGE2 and peripheral parasitemia.
After adjustment for age, the log-transformed mean peripheral parasitemia values for the 5 groups were as follows: CON, 0; AP, 2.68; NCM, 4.64; CMCR, 4.10; and CMDS, 4.59 (figure 2). An ANCOVA (with adjustment for age) indicated a significant between-group difference in peripheral parasitemia (F[4,72] = 124.35; P < .0001) (table 1). Results of further analyses using Fisher's LSD multiple range tests indicated that the AP group had significantly lower parasitemia than did the NCM and CM groups, whereas the NCM group had significantly higher parasitemia than did the CMCR group (figure 2). Although the zero-order correlation between the systemic level of bicyclo-PGE2/creatinine and peripheral parasitemia was not significant (r = 0.11; P = .345), when age and disease severity were taken into account via partial correlation, the 2 variables became significantly correlated (rp = 0.31; P = .007) (table 2).
The association between levels of bicyclo-PGE2/creatinine and hemoglobin levels was also examined, since PGE2 is an important soluble factor for promoting efficient erythropoiesis [18, 19]. However, no significant association was found between levels of bicyclo-PGE2/creatinine and hemoglobin levels (r = -0.03; P = .86) (table 2). When age, parasitemia, and disease severity were controlled by partial correlation, the association between levels of bicyclo-PGE2/creatinine and hemoglobin levels remained nonsignificant (rp = 0.03; P = .80) (table 2).
Our previous studies of Gabonese children with severe malarial anemia have shown a significant inverse association between plasma levels of bicyclo-PGE2/creatinine and hemoglobin levels [20]. However, all of the children representing the control group in those studies had previous clinical episodes of malaria that were recorded during their participation in a longitudinal cohort study. Since we did not know the previous exposure rate of the control group in the present study, data on parasitemic children were reanalyzed with the control group excluded from the analysis. Upon reanalysis, there was still no significant association between levels of bicyclo-PGE2/creatinine and hemoglobin levels (r = 0.06; P = .634) (table 2). The partial correlation (with adjustment for age, parasitemia, and disease severity) showed a nonsignificant positive association between levels of bicyclo-PGE2/creatinine and hemoglobin levels (rp = 0.06; P = .66) (table 2).
DISCUSSION
The lack of elevation of urinary levels of bicyclo-PGE2 in Tanzanian children with CM, along with the findings of our previous studies (which showed an inverse correlation of peripheral blood mononuclear cell COX-2 gene expression and plasma levels of bicyclo-PGE2 with disease severity in Gabonese children with malarial anemia) [13], illustrates that decreased biosynthesis of PGE2 during falciparum malaria is associated with enhanced disease severity. The findings of these 2 studies are also consistent with those of our recent study that demonstrated that production of PGE2 was suppressed in circulating placental monocytes isolated from Kenyan women with malaria during pregnancy [21]. The inverse association between systemic levels of PGE2 and malaria disease severity in all of our previous studies performed in participants from regions with differing malaria endemicity who had varied genetic backgrounds and different clinical manifestations of severe disease (i.e., malarial anemia vs. CM) suggests that suppression of systemic production of PGE2 is a universal pathophysiological marker and/or mediator of malaria pathogenesis.
Previous studies of healthy adults (nonmalaria exposed) have illustrated that salicylates and acetaminophen (paracetamol) can suppress urinary production of PGE2 [22, 23]. Since some of the children in the present study received antimalarials and/or antipyretics before enrollment, systemic levels of PGs could have been altered by common treatment interventions for malaria. Although the effects of antimalarials on systemic production of PGE2 are unknown, it is unlikely that treatment interventions alone account for the impaired production of PGE2 in children with CM. For example, urinary levels of bicyclo-PGE2 were lower in children with the most severe forms of CM, even though the number of individuals who reported having received antimalarials and/or antipyretics was slightly lower in the CMDS group (86% medication usage) than in the CMCR group (92% medication usage). Although it is possible that treatment interventions may have some effect on urinary levels of PGE2, our previous results, which showed that plasma levels of bicyclo-PGE2 are significantly suppressed in medication-naive individuals with severe malarial anemia [13], illustrate that infection with P. falciparum can impair systemic production of PGE2. Our previous studies have also demonstrated that de novo levels of PBMC COX-2 mRNA and protein were significantly lower in children with severe malaria than in children with mild disease. This finding is important since suppression of production of PGE2, by salicylates and acetaminophen, likely occurs through blockade of the COX enzymatic site in the mature protein, and not through blockade of de novo COX-2 gene expression [24, 25]. Since we did not have access to cryopreserved PBMCs in the present study, we were unable to determine whether levels of COX-2 transcript and protein were suppressed in circulating blood mononuclear cells. Although the exact molecular mechanisms by which malaria suppresses COX-2 and PGs are presently unknown, we have recently found that ingestion of hemozoin (malarial pigment) and -hematin (synthetic malarial pigment) by cultured monocytes decreases COX-2 gene expression and formation of PGE2 [26]. Thus, we postulate that acquisition of parasitic products by circulating monocytes and tissue macrophages is directly responsible for the systemic reduction of PGs in children with CM reported in the present study.
In addition to the results of the present study and those of our previous studies [13], the results of several other studies suggest that decreased production of PGs during malaria infection results in enhanced pathogenesis. In experiments by Xiao et al., treatment with aspirin (a COX-1 and COX-2 inhibitor) increased mortality in a murine model of CM [15]. In the same murine model, celecoxib, a more selective inhibitor of COX-2, was associated with an earlier onset of CM [16]. In addition, studies in Gabon have demonstrated that treatment of children with acetaminophen prolongs the time to clearance of P. falciparum and decreases production of oxygen radicals in mitogen-stimulated whole blood [14]. On the basis of previous studies by others that showed that the overuse of salicylates in areas where malaria is endemic contributes to metabolic acidosis and hypoglycemia [27], along with our studies of Gabonese children with severe malarial anemia [13] and the present study of Tanzanian children with CM, we postulate that malaria-induced suppression of peripheral PGE2 may be exacerbated by pharmacological interventions attempting to control the fever commonly associated with malaria. Although this effect of antipyretics may be deleterious, it may be balanced by potentially beneficial effects of antipyretics in reduction of seizure risk and reduction of the fever-induced cytoadherence of pRBCs within the cerebral microvasculature [28]. In previous studies, only toxic doses of antipyretics due to overmedication by caregivers have been linked to severe malaria [27]. We are currently in the process of examining the advantageous versus the adverse effects associated with use of antipyretics in both human and nonhuman primate malarias.
Formation of PGE2 is determined by the relative expression of proinflammatory cytokines, which increase COX-2 gene expression (e.g., TNF-), and anti-inflammatory cytokines, which decrease COX-2 gene expression (e.g., IL-10) [17, 29, 30]. Once formed, PGE2 can act as a negative feedback signal to decrease production of TNF- in monocytes [31, 32] and promote the release of IL-10 [33]. Thus, although PGE2 is typically associated with proinflammatory events, PGE2 has anti-inflammatory properties through its ability to down-regulate production of TNF- and up-regulate production of IL-10. Analysis of the complex interactions between cytokines and PGE2 in vivo, however, is largely unreported. Investigation of regulation of production of PGE2, by cytokines, in our previous studies of children with severe malarial anemia revealed a significant inverse association between systemic levels of PGE2 and those of IL-10 [13], which is consistent with a model suggesting that IL-10 down-regulates production of PGE2. However, in the present study, there was a significant positive correlation between systemic levels of PGE2 and those of IL-10, suggesting a model consistent with the work of Shinomiya et al. in which PGE2 was found to drive increased production of IL-10 [33]. Additional studies examining the time-dependent relationship between cytokines and production of PGE2 during the acute and convalescent stages of disease will be required to appropriately understand the complex inflammatory milieu.
Although, in the present study, there was not a significant inverse correlation between systemic levels of PGE2 and of TNF-, we postulate that inadequate peripheral production of PGE2 during malaria exacerbates the overproduction of monocyte-derived TNF-. High systemic levels of TNF- (because of systemic suppression of production of PGE2) would then circulate to the preoptic area of the anterior hypothalamus (POAH) in the central nervous system (CNS) and induce local tissue-specific COX-2 gene expression and production of PGE2 in astrocytes and/or microvascular endothelial cells. This model is consistent with the fact that localized production of PGE2 in the POAH is responsible for generation of the febrile response [34], which is the hallmark clinical sign of malaria. This may also explain why previous immunohistochemical studies have found increased levels of COX-1 in macrophages/microglial cells within Dürcks granulomas and increased expression of COX-2 in endothelial cells and astrocytes within brain parenchyma, in expatriates with CM [35]. Consistent with the fact that the systemic immune response does not always mirror what is occurring within central brain regions, children with the lowest systemic levels of PGE2 have the highest levels of feveran event requiring increased COX-2derived production of PGE2 within the POAH. On the basis of the results presented here and those presented elsewhere, we postulate that down-regulation of PGE2 in the periphery may allow overexpression of inflammatory mediators, such as TNF-, which promote enhanced pathophysiological consequences in the CNS.
Although previous studies of cultured P. falciparum isolates have shown that the parasite is capable of biosynthesizing PGs [36], the results presented here and in our previous work in Gabon [13] illustrate that children with the highest parasite burdens have the lowest levels of systemic production of PGE2. Thus, any PGE2 synthesized by parasites appears to be counteracted by the down-regulation of host production of PGE2, by the malaria disease process. The effect of the parasites' ability to synthesize PGs on the pathophysiology of malaria remains unresolved. In the present study, we found a positive, significant correlation (rp = 0.31; P = .007) between systemic levels of PGE2 and parasitemia, after adjustment for age and disease severity, suggesting that parasitemia influences levels of PGE2.
PGs also have other biological activities that could mediate the pathogenesis of malaria, such as their ability to regulate erythropoiesis [18, 19]. In our previous investigations in Gabon, we found that plasma levels of PGE2 were inversely related to hemoglobin levels [20], which is consistent with the fact that PGE2 is an important molecule for promoting appropriate erythropoiesis [19]. The data presented here differ from our previous results in that levels of bicyclo-PGE2 and hemoglobin levels were not significantly associated. When age, parasitemia, and disease severity were taken into account, the association between levels of bicyclo-PGE2 and hemoglobin levels remained nonsignificant.
High levels of PGE2 in children with asymptomatic parasitemia suggest that PGs may also contribute to the maintenance of malaria tolerance, which is the ability of children to tolerate circulating parasites without fever. We have previously found that high basal levels of COX-2 gene expression and production of PGE2 in unstimulated circulating monocytes in healthy malaria-exposed children in Gabon are higher than those in nonmalaria exposed, age-matched children in the United States (D.J.P., unpublished data). Similarly, high basal production of NO has been found in children and adults with asymptomatic parasitemia [37, 38] and healthy malaria-exposed children who develop mild versus severe disease during acute malaria [39], suggesting that high basal production of NO may contribute to malaria tolerance in endemic areas. In fact, we have previously shown that healthy malaria-exposed children who develop mild disease and produce high basal levels of NO have significantly lower basal production of TNF- than do children who produce low basal levels of NO and develop severe disease [39]. We therefore hypothesize that a common mechanism for both PGE2 and NO may be their ability to inhibit monocyte production of TNF-, one of the major fever-inducing cytokines in malaria [40]. Future studies aimed at determining the mechanisms by which P. falciparum suppresses systemic levels of PGE2 and the role that PGs play in regulating malaria pathogenesis should provide important insight into our understanding of malarial immunity and may have important therapeutic implications.
Acknowledgments
We thank Mushtaq Hassanali and Stella Stanslaus, for assistance with collection of samples, and Dennis Manyenga and Juliana Mlalasi, for assistance with processing of samples.
References
1. Breman JG, Egan A, Keusch GT. The intolerable burden of malaria: a new look at the numbers. Am J Trop Med Hyg 2001; 64(Suppl 12):ivvii. First citation in article
2. Marsh K, English M, Crawley J, Peshu N. The pathogenesis of severe malaria in African children. Ann Trop Med Parasitol 1996; 90:395402. First citation in article
3. Turner G. Cerebral malaria. Brain Pathol 1997; 7:56982. First citation in article
4. Kwiatkowski D, Hill AV, Sambou I, et al. TNF concentration in fatal cerebral, non-fatal cerebral, and uncomplicated Plasmodium falciparum malaria. Lancet 1990; 336:12014. First citation in article
5. de Kossodo S, Grau GE. Profiles of cytokine production in relation with susceptibility to cerebral malaria. J Immunol 1993; 151:481120. First citation in article
6. Baptista JL, Vanham G, Wery M, Vanmarck E. Cytokine levels during mild and cerebral falciparum malaria in children living in a mesoendemic area. Trop Med Int Health 1997; 2:6739. First citation in article
7. Anstey NM, Weinberg JB, Hassanali M, et al. Nitric oxide in Tanzanian children with malaria: inverse relationship between malaria severity and nitric oxide production/nitric oxide synthase type 2 expression. J Exp Med 1996; 184:55767. First citation in article
8. Kurtzhals J, Abadayeri V, Goka B, et al. Low plasma concentrations of interleukin 10 in severe malarial anemia compared with cerebral and uncomplicated malaria. Lancet 1998; 351:176872. First citation in article
9. Seibert K, Masferrer JL. Role of inducible cyclooxygenase (COX-2) in inflammation. Receptor 1994; 4:1723. First citation in article
10. Murakami M, Nakatani Y, Atsumi G, Inoue K, Kudo I. Regulatory functions of phospholipase A2. Crit Rev Immunol 1997; 17:22583. First citation in article
11. Vane J, Bakhle Y, Botting R. Cyclooxygenase 1 and 2. Annu Rev Pharmacol Toxicol 1998; 38:97120. First citation in article
12. Stevens D. Could nonsteroidal antiinflammatory drugs (NSAIDs) enhance the progression of bacterial infections to toxic shock syndrome Clin Infect Dis 1995; 21:97780. First citation in article
13. Perkins DJ, Kremsner PG, Weinberg JB. Inverse relationship of plasma prostaglandin E2 and blood mononuclear cell cyclooxygenase-2 with disease severity in children with Plasmodium falciparum malaria. J Infect Dis 2001; 183:1138. First citation in article
14. Brandts C, Ndjave M, Graninger W, Kremsner P. Effect of paracetamol on parasite clearance time in Plasmodium falciparum malaria. Lancet 1997; 350:7049. First citation in article
15. Xiao L, Patterson P, Yang C, Lal A. Role of eicosinoids in the pathogenesis of murine cerebral malaria. Am J Trop Med Hyg 1999; 60:66873. First citation in article
16. Ball HJ, MacDougall HG, McGregor IS, Hunt NH. Cyclooxygenase-2 in the pathogenesis of murine cerebral malaria. J Infect Dis 2004; 189:7518. First citation in article
17. Harris S, Pagilla J, Koumas L, Ray D, Phipps R. Prostaglandins as modulators of immunity. Trends Immunol 2002; 23:14450. First citation in article
18. Ortega JA, Dukes PP, Ma A, Shore NA, Malekzadeh MH. A clinical trial of prostaglandin E2 to increase erythropoiesis in anemia of end stage renal disease: a preliminary report. Prostaglandins Leukot Med 1984; 14:4116. First citation in article
19. Datta MC. Prostaglandin E2 mediated effects on the synthesis of fetal and adult hemoglobin in blood erythroid bursts. Prostaglandins 1985; 29:56177. First citation in article
20. Perkins DJ, Kremsner PG, Weinberg JB. Inverse relationship of plasma prostagladin-E2 and blood mononuclear cell cyclooxygenase-2 with disease severity in children with Plasmodium falciparum malaria: implications in the pathogenesis of disease. Amer J Trop Med Hyg 2000; 62(Suppl 3):649. First citation in article
21. Perkins D, Moore J, Otieno J, et al. In vivo acquisition of hemozoin by placental blood mononuclear cells suppresses PGE2, TNF-, and IL-10. Biochem Biophys Res Comm 2003; 311:83946. First citation in article
22. Reimann I, Golbs E, Fischer C, Frolich J. Influence of intravenous acetylsalicylic acid and sodium salicylate on human renal function and lithium clearance. Eur J Clin Pharmacol 1985; 29:43541. First citation in article
23. Bippi H, Frolich J. Effects of acetylsalicylic acid and paracetamol alone and in combination on prostanoid synthesis in man. Brit J Clin Pharm 1990; 29:30510. First citation in article
24. Donnelly MT, Hawkey CJ. COX-II inhibitorsa new generation of safer NSAIDs. Aliment Pharmacol Ther 1997; 11:22736. First citation in article
25. Vane JR. NSAIDS, Cox-2 inhibitors, and the gut. Lancet 1995; 346:11056. First citation in article
26. Keller CC, Hittner JB, Nti BK, Weinberg JB, Kremsner PG, Perkins DJ. Reduced peripheral PGE2 biosynthesis in Plasmodium falciparum malaria occurs through hemozoin-induced suppression of blood mononuclear cell cyclooxygenase-2 gene expression via an interleukin-10independent mechanism. Mol Med 2004; 10:4554. First citation in article
27. English M, Marsh V, Amukoye E, Lowe B, Murphy S, Marsh K. Chronic salicylate poisoning and severe malaria. Lancet 1996; 347:17367. First citation in article
28. Udomsangpetch R, Pipitaporn B, Silamut K, et al. Febrile temperatures induce cytoadherence of ring-stage Plasmodium falciparuminfected erythrocytes. Proc Natl Acad Sci USA 2002; 99:118259. First citation in article
29. Perkins DJ, Kniss DA. Rapid and transient induction of cyclooxygenase-2 by epidermal growth factor in human amnion-derived WISH cells. Biochem J 1997; 321:67781. First citation in article
30. Perkins D, Kniss D. Tumor necrosis factor promotes sustained cyclooxygenase-2 expression: attenuation by dexamethasone and NSAIDs. Prostaglandins 1997; 54:72743. First citation in article
31. Kunkel SL, Spengler M, May MA, Spengler R, Larrick J, Remick D. Prostaglandin E2 regulates macrophage-derived tumor necrosis factor gene expression. J Biol Chem 1988; 263:53804. First citation in article
32. Choi SS, Gatanaga M, Granger GA, Gatanaga T. Prostaglandin E2 regulation of tumor necrosis factor receptor release in human monocytic THP-1 cells. Cell Immunol 1996; 170:17884. First citation in article
33. Shinomiya S, Naraba H, Ueno A, et al. Regulation of TNF- and interleukin-10 production by prostaglandins I2 and E2: studies with prostaglandin receptordeficient mice and prostaglandin Ereceptor subtype-selective synthetic agonists. Biochem Pharmacol 2001; 61:115360. First citation in article
34. Blatteis C, Sehic E. Prostaglandin E2: a putative fever mediator. In: Mackowiak P, ed. Fever: basic mechanisms and management. Philadelphia: Lippincott-Raven, 1997:11754. First citation in article
35. Deininger M, Kremsner P, Meyermann R, Schluesener H. Focal accumulation of cyclooxygenase1 (COX-1) and COX-2 expressing cells in cerebral malaria. J Neuroimmunol 2000; 106:198205. First citation in article
36. Kubata B, Eguchi N, Urade Y, et al. Plasmodium falciparum produces prostaglandins that are pyrogenic, somnogenic, and immunosupressive substances in humans. J Exp Med 1998; 188:1197202. First citation in article
37. Anstey N, Granger D, Hassanali M, Wwaikambo E, Duffy P, Weinberg J. Nitric oxide, malaria, and anemia: inverse relationship between nitric oxide production and hemoglobin concentration in asymptomatic, malaria-exposed children. Am J Trop Med Hyg 1999; 61:24952. First citation in article
38. Boutlis C, Tjitra E, Maniboey H, et al. Nitric oxide production and mononuclear cell nitric oxide synthase activity in malaria-tolerant Papuan adults. Infect Immun 2003; 71:36829. First citation in article
39. Perkins D, Kremsner P, Schmid D, Misukonis M, Kelly M, Weinberg J. Blood monocuclear cell nitric oxide production and plasma cytokine levels in healthy Gabonese children with prior mild or severe malaria. Infect Immun 1999; 67:497781. First citation in article
40. Kwiatkowski D, Bate CAW, Scragg IG, Beattie P, Udalova I, Knight JC. The malarial fever responsepathogenesis, polymorphism and prospects for intervention. Ann Trop Med Parasitol 1997; 91:53342. First citation in article