Induction of proinflammatory responses in macrophages by the glycosylphosphatidylinositols of Plasmodium falciparum: cell signaling receptors, glycosylphosphatidylinositol (GPI) structural requirement, and regulation of GPI activity.

The glycosylphosphatidylinositol (GPI) anchors of Plasmodium falciparum have been proposed to be the major factors that contribute to malaria pathogenesis through their ability to induce proinflammatory responses. In this study we identified the receptors for P. falciparum GPI-induced cell signaling that leads to proinflammatory responses and studied the GPI structure-activity relationship. The data show that GPI signaling is mediated mainly through recognition by TLR2 and to a lesser extent by TLR4. The activity of sn-2-lyso-GPIs is comparable with that of the intact GPIs, whereas the activity of Man(3)-GPIs is about 80% that of the intact GPIs. The GPIs with three (intact GPIs and Man(3)-GPIs) and two fatty acids (sn-2-lyso-GPIs) appear to differ considerably in the requirement of the auxiliary receptor, TLR1 or TLR6, for recognition by TLR2. The former are preferentially recognized by TLR2/TLR1, whereas the latter are favored by TLR2/TLR6. However, the signaling pathways initiated by all three GPI types are similar, involving the MyD88-dependent activation of extracellular signal-regulated kinase, c-Jun N-terminal kinase, and p38 and NF-kappaB-signaling pathways. The signaling molecules of these pathways differentially contribute to the production of various cytokines and nitric oxide (Zhu, J., Krishnegowda, G., and Gowda, D. C. (2004) J. Biol. Chem. 280, 8617-8627). Our data also show that GPIs are degraded by the macrophage surface phospholipases predominantly into inactive species, indicating that the host can regulate GPI activity at least in part by this mechanism. These results imply that macrophage surface phospholipases play important roles in the GPI-induced innate immune responses and malaria pathogenesis.

The process of malaria pathogenesis is very complex and, despite malaria being one of the most extensively studied infectious diseases during the past decades, the precise molecular basis for disease progression remains poorly understood. However, based on our current knowledge, several key processes can be recognized (1). These include the rapid destruction of infected and uninfected erythrocytes, dyserythropoiesis, metabolic acidosis, hypoglycemia, and adherence of infected erythrocytes to the microvascular capillary, vascular obstruction, and production of high levels of proinflammatory mediators in response to parasitic factors (1)(2)(3)(4)(5).
Studies from various laboratories have demonstrated that proinflammatory responses, including the production of TNF-␣, 1 IFN␥, IL-12, and reactive oxygen and nitrogen intermediates by the innate immune system in response to the microbes are critical for killing the pathogenic organisms (5)(6)(7)(8). The inflammatory cytokines can also control infection by promoting macrophage phagocytosis and by complement activation (7,9,10). In addition, the innate immune response is critical in determining the nature of adaptive immune responses, i.e. the specificity of the adaptive immunity against pathogens depends on the pattern of early cytokine expression (5,8).
areas, leading to endothelial damage and organ dysfunction (15). Furthermore, up-regulation of endothelial cell adhesion molecules in response to TNF-␣ potentially augments the adherence, spiraling the destruction of the endothelia and causing organ-related pathological conditions (20). Understanding the underlying mechanism that controls the expression of the proinflammatory responses to P. falciparum components is likely to provide therapeutic avenues for the prevention of malaria pathogenesis.
Although it is known that excessive production of proinflammatory mediators is a key process that contributes to the falciparum malaria pathogenesis, very little is known about the nature of various potential ligands of the parasite and cellsignaling mechanisms involved. However, about a decade ago, it was shown that the glycosylphosphatidylinositols (GPIs) of P. falciparum could induce the expression of proinflammatory cytokines and nitric oxide in macrophages (21). GPIs administered to animals have been shown to cause transient pyrexia and hypoglycemia, cachexia, and death in galactosamine-sensitized animals, the events that are reminiscent of acute malaria (21). Thus, GPIs have been proposed as the dominant parasite factors responsible for malaria pathogenesis. Later studies have shown that the GPIs of P. falciparum can induce the expression of iNOS, up-regulate the expression of intracellular adhesion molecule 1, vascular cell adhesion molecule 1, and E-selectin in endothelial cells, implicating these processes in malaria pathogenesis (20).
Previously, it was proposed that parasite GPI-induced proinflammatory responses involve the activation of protein kinase C and protein-tyrosine kinase-signaling pathways that collaboratively activate NF-B/c-Rel, which translocates to the nucleus, initiating the expression of proinflammatory molecules (22). It has also been postulated that the two signaling pathways are activated by the distinct structural elements of GPIs, namely the conserved glycan and diacylglycerol moieties, respectively (23). Several hypotheses, including the insertion of GPIs into cell membrane microdomains rich in GPI-anchored proteins and membrane perturbation, hydrolysis of GPIs, and internalization of the diacylglycerol moiety or GPI translocation across the cell membrane initiating the activation of protein kinase C-dependent second messengers, have been proposed for cell signaling by GPIs (23). However, in recent years it has become clear that the proinflammatory responses by the innate immune system are mediated through the recognition of specific molecular patterns of microorganisms by the TLR family of proteins (24 -26). Recently, cell signaling by GPIs of Trypanosoma cruzi to produce proinflammatory mediators has been shown to be through the exclusive recognition of TLR2 (27). In the preceding paper (28) we have demonstrated that P. falciparum GPI-induced proinflammatory responses in macrophages proceed through the recognition of mainly TLR2 and to a lesser extent through TLR4, and the cell signaling involves the activation of MAPK-, ERK-, p38-, JNK-, and NF-B-signaling pathways.
A number of studies have shown that ERK-, p38-, JNK-, and NF-B-signaling pathways are activated by TLR/MyD88-mediated cell signaling in response to various ligands of pathogenic organisms (29). However, the requirement of various signaling molecules of these pathways for the expression of cytokines appears to be dependent on the nature of ligands and the host cell types. In many cases the reported results are controversial (30). With regard to GPIs, a recent study reported the requirement of ERK, p38, and NF-B pathways with GPIs of T. cruzi; however, the information is very limited (31). In this study we investigated the requirement of the MAPK-and NF-B-signal-ing pathways for the P. falciparum GPI-induced expression of proinflammatory mediators using macrophages from mice deficient in the activation of ERK1/ERK2, JNK1, and JNK2 and by using specific inhibitors. Our data show that the ERK1/ ERK2-, p38-, and JNK1/JNK2-signaling molecules and NF-B family proteins are differentially involved in the GPI-induced production of TNF-␣, IL-6, IL-12, and nitric oxide by macrophages.

EXPERIMENTAL PROCEDURES
Materials-PDTC, SN50 peptide, PD98059 and UO126, SB203580, and SB202190 were from Calbiochem. Griess reagent was from Sigma. SP600125 was from Tocris Cookson, Inc. (Ellisville, MO). The chemiluminescence substrate kit was from KPL (Gaithersburg, MD). Nitrocellulose membranes were from Bio-Rad. The anti-peptide and phosphospecific antibodies that recognize ERK1/ERK2, p38, JNK, anti-IB␣, and anti-␤-tubulin antibodies and horseradish peroxidase-conjugated goat anti-mouse IgGs and goat anti-rabbit IgGs were from Cell Signaling Technology Inc. (Beverly, MA). Anti-Tpl2 antibody was from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). Kits for the estimation of TNF-␣, IL-6, and IL-12 (p40) by ELISA (Duoset ELISA Development System) were from R&D Systems, Inc. (Minneapolis, MN). FBS, DMEM, and Trizol RNA isolation reagent were from Invitrogen. L929 murine fibroblast cells were from American Type Culture Collection. The PCR primers used were custom-synthesized by Invitrogen. Human blood and serum from healthy donors were from the hospital of the Hershey Medical Center. Endotoxin-free reagents, water, and buffers were used for all the experimental procedures.
P. falciparum Culturing and Isolation of GPIs-The parasites culturing and purification of GPIs are described in the preceding paper (28).
Mice-NF-B1/p105 knock-out mice and the corresponding wild type control mice (B6.129PF2/J) and MAPK8 (JNK1) and MAPK9 (JNK2) knock-out mice and the corresponding wild type control mice (C57BL6/J) were purchased from The Jackson Laboratories. The animals were maintained in a germ-free environment, and the animal care was in accordance with the Institutional Guidelines of the Pennsylvania State University College of Medicine.
Preparation of L929 Cell Supernatant-L929 cells were cultured in DMEM, 5% FBS, 1% penicillin/streptomycin, 1% L-glutamine in roller flasks at 37°C for 4 -5 days. The supernatant were collected, centrifuged at 2500 rpm for 20 min, and used as a source of macrophage colony-stimulating factor.
Preparation of Bone Marrow-derived Macrophages-Bone marrowderived macrophages were obtained as described (32). Briefly, the murine bone marrow was flushed from the femurs of mice with DMEM, 10% FBS, and penicillin/streptomycin. The tissue debris in each case was removed by centrifugation at 300 rpm, and cells were spun down at 1200 rpm. The cell pellets were washed twice with DMEM, 10% FBS, penicillin/streptomycin and dispersed in DMEM, 20% FBS, 30% L929 cell supernatant, and 1% penicillin/streptomycin and cultured at 37°C for 2 days. The non-adherent cells were removed, and the adherent cells were cultured by changing the medium every 2 days. Seven days later the macrophages were harvested using phosphate-buffered saline, pH 7.2, containing 10 mM EDTA and 20% FBS and used for stimulation with the parasite GPIs.
Stimulation of Macrophages with P. falciparum GPIs-Freshly harvested macrophages were seeded in 96-well microtiter plates (2.5 ϫ 10 4 cells/well) and cultured overnight in DMEM, 10% FBS, and penicillin/ streptomycin. The cells were treated with varying amounts of GPIs coated onto gold particles as described in the preceding paper (28). Cells treated with uncoated gold particles were used as controls. For inhibition studies the inhibitors were added to the culture medium 1 h before the addition of GPIs. In some experiments cells were primed with IFN␥ (100 units/ml) before stimulation with GPIs. The culture supernatant were collected at the indicated time points and stored at Ϫ70°C until used for cytokine and nitric oxide measurements.
For the analysis of signaling proteins, macrophages in 24-well microtiter plates (0.5 ϫ 10 6 cells/well) were cultured overnight in DMEM, 0.5% FBS and then stimulated with 200 nM GPIs. For inhibition studies cells were treated for 1 h with appropriate inhibitors before stimulation with 200 nM GPIs. At various time points, the cells were washed with ice-cold phosphate-buffered saline and then lysed with ice-cold radioimmune precipitation assay buffer (50 mM Tris-HCl, 150 mM NaCl, pH 7.2, containing 1% sodium deoxycholate, 0.1% SDS, 1% Triton X-100). The lysates were centrifuged at 4°C, and the supernatant was used for Western blotting.
For RT-PCR analysis of gene transcription, macrophages in 24-well plates (1 ϫ 10 6 cell/well) were cultured overnight in DMEM, 0.5% FBS, then stimulated with 200 nM GPIs in DMEM, 10% FBS; at various time points culture supernatants were removed, and cells were washed with ice-cold phosphate-buffered saline, pH 7.2. Cells were dissolved in Trizol, and total RNA was isolated and used for cDNA preparation.
Cytokine and Nitrite Measurement-The TNF-␣, IL-6, and IL-12 (p40) in the culture supernatants of macrophages stimulated with GPIs were determined by sandwich ELISA with horseradish peroxidase-conjugated streptavidin and 3,3Ј,5,5Ј,-tetramethylbenzidine color reagent using the Duoset ELISA development kit (R&D Systems). After stopping the color development with 1 M sulfuric acid, the absorbance at 450 nm was measured with SpectraMax Plus384 plate reader (Molecular Devices). The cytokine concentrations were calculated with reference to the standard curves.
The nitrite in the macrophage culture supernatants was assayed by microcolorimetric assay using the Griess reagent (33). The absorbance at 540 nm was determined, and the nitrite concentrations were estimated with reference to a graph generated using a standard solution of sodium nitrite.
Western Blotting-The macrophage lysates obtained as outlined above were made to 1ϫ with respect to the SDS-PAGE sample buffer containing 2-mercaptoethanol, boiled for 5 min, and electrophoresed on 10% SDS-polyacrylamide gels. The protein bands in the gels were transferred onto polyvinylidene difluoride or nitrocellulose membranes. The membranes were blocked with 5% (w/v) nonfat dry milk in phosphate-buffered saline, pH 7.4, containing 0.1% Tween 20 at room temperature for 1 h and then incubated with either anti-peptide or phospho-specific antibodies in the above buffer. After washing the membranes were incubated with horseradish peroxidase-conjugated anti-mouse or anti-rabbit secondary antibodies, treated with chemiluminescent substrate system (LumiGLO from KPL), and exposed to x-ray films.

Production of Proinflammatory Cytokines and Nitric Oxide in IFN␥-primed Macrophages Stimulated with P. falciparum GPIs
Previous studies have reported that 0.1-1 M P. falciparum GPIs are required for the efficient production of TNF-␣ in macrophage cell lines, and an appreciable level of nitric oxide was evident only at Ͼ1 M GPIs (22,23,35,36). In the case of T. cruzi trypamastigote GPIs, however, subnanomolar concentrations are sufficient to induce high levels of proinflammatory cytokines and nitric oxide production in IFN␥-primed peritoneal macrophages, and priming was essential for the expression of IL-12, IL-6, and nitric oxide (31,37). Because in previous studies with P. falciparum GPIs the activity was studied in macrophage cell lines without IFN␥ priming (22,23,35,36), the reported comparative assessment of the potency of GPIs of these two protozoan parasites measured under different conditions does not accurately reflect the relative potency of malarial GPIs. Therefore, in this study we tested P. falciparum GPIs for the production of proinflammatory cytokines and nitric oxide in mouse bone marrow-derived and peritoneal macrophages before and after priming with IFN␥ ( Fig. 1 and not shown). Additionally, to determine the pathophysiological relevance of P. falciparum GPIs in malaria, we tested the activity of the GPIs in IFN␥-primed and -unprimed human peripheral monocytes. P. falciparum GPIs induced efficient production of TNF-␣ at 5-20 nM in IFN␥-primed and -unprimed murine macrophages and human monocytes (Fig. 1, A and E). The P. falciparum GPIs exhibited an order of magnitude higher activity in macrophages than that previously reported under the conditions used for measuring the activity of T. cruzi mucin GPIs. Therefore, the potency of the P. falciparum GPIs is about 10-fold but not 100-fold lower than that of T. cruzi GPIs as reported previously (37). Furthermore, as in the case of T. cruzi GPI (31, 37), the P. falciparum GPIs were unable to induce IL-6, IL-12, and nitric oxide expression in unprimed macrophages but could induce the production of these proinflammatory mediators in IFN␥-primed macrophages (Fig. 1, B-D). Although 5-20 nM GPIs were sufficient to produce high levels of TNF-␣ and nitric oxide in macrophages, 20 -320 nM GPIs were required for the efficient production of IL-12 and IL-6 ( Fig. 1, B and C).

The Requirement of NF-B-and MAPK-signaling Pathways for the P. falciparum GPI-induced Production of Proinflammatory Cytokines and Nitric Oxide by Macrophages
In the preceding paper (28) we have shown that stimulation of macrophages with P. falciparum GPIs results in the activation of ERK-, p38-, JNK-, and NF-B-signaling pathways through the recognition mainly by TLR2 and to some degree by TLR4 (28). In this study we investigated the requirement of these downstream signaling pathways for the production of proinflammatory cytokines and nitric oxide using specific inhibitors of signaling molecules and macrophages from gene knock-out mice. The results of these studies are described below.
The Requirement of NF-B Activation for the P. falciparum GPI-induced Proinflammatory Responses-The activation and nuclear translocation of NF-B is crucial for the production of proinflammatory molecules in response to various microbial ligands (38 -41). NF-B comprises a family of proteins, which differentially regulate the expression of proinflammatory mediators in response to activation by various microbial ligands. For example, NF-B1/p105 is not essential for the LPS-induced expression of TNF-␣ and nitric oxide (41, 42), but p50 derived from NF-B1/p105 appears to be crucial for the B-dependent expression of a subset of genes in cells activated through TLRsignaling pathways (43). To determine the role of NF-B1/p105 in the GPI-induced proinflammatory responses, we used macrophages from NF-B1/p105 Ϫ/Ϫ mice (41,42). The levels of TNF-␣ and nitric oxide produced by NF-B1/p105 Ϫ/Ϫ macrophages at various time points after stimulating with P. falciparum GPIs were comparable with those by the wild type macrophages (Fig. 2). In both cells types similar levels of TNF-␣ and nitric oxide were also produced in response to different concentrations of GPI stimulation (not shown). Consistent with these results, the real time RT-PCR analysis showed that the mRNA levels of TNF-␣ and iNOS in GPI-treated NF-B1/p105 Ϫ/Ϫ macrophages are comparable with those of the GPI-treated wild type macrophages (Fig. 3). These data demonstrate that the NF-B1/p105 is not involved in the production of TNF-␣ and nitric oxide by macrophages stimulated with GPIs. In contrast to TNF-␣ and nitric oxide production, NF-B1 Ϫ/Ϫ macrophages stimulated with the parasite GPIs produced markedly lower levels of IL-12 and IL-6 expression, respectively, about 20 and 40% that of those produced in wild type macrophages (Fig. 2). Quantitative RT-PCR analysis showed that the levels of IL-12 and IL-6 mRNA in GPI-stimulated NF-B1/p105 Ϫ/Ϫ macrophages were, respectively, 5-10-and 2-4-fold lower than that in GPI-treated wild type macrophages at various time points after stimulation (Fig. 3). These data indicate that the observed lower level of IL-12 and IL-6 expression (see Fig. 2) in NF-B1/p105 Ϫ/Ϫ macrophages compared with wild type macrophages is due to the marked low levels of mRNA transcripts in the former. The data also suggest that p50 (the matured form of NF-B1/p105) has little or no role in the production of TNF-␣ and nitric oxide but is crucial for the expression of IL-6 and IL-12, presumably by the IB-mediated gene expression as reported recently for LPS (42,43).
The involvement of NF-B members in the GPI-induced production of proinflammatory responses was studied further by using PDTC and SN50, the inhibitors of NF-B activation and translocation of the p50 component of NF-B complex to nucleus, respectively (44,45). PDTC markedly inhibited the production of TNF-␣, IL-6, IL-12, and nitric oxide in GPI-stimulated macrophages (Fig. 4A). In contrast to PDTC, SN50 only marginally inhibited the expression of GPI-induced production of TNF-␣ and nitric oxide in macrophages (Fig. 4B). These results indicate that the p50-containing NF-B complex has little or no role in the production of TNF-␣ and nitric oxide, and the data are consistent with the results from NF-B1/p105 Ϫ/Ϫ macrophages (see Fig. 2). In the case of IL-6 and IL-12, 4.5 M SN50 inhibited their production by ϳ40 and 50%, respectively (Fig. 4B). These results are consistent with the requirement of NF-B1/p50 for IB-dependent production of IL-12 and IL-6 (43). Unexpectedly, however, at higher concentrations of SN50, the inhibitory effect on IL-6 and IL-12 expression was nearly abolished, and the reason for this effect is not clear.
The Requirement of the ERK-signaling Pathway for the P. falciparum GPI-induced Proinflammatory Responses-Recently, it has been shown that NF-B1/p105 Ϫ/Ϫ mice are deficient in the activation of the ERK-signaling pathway but not those of p38 and JNK (46,47). Therefore, we used bone marrow-derived macrophages from NF-B1/p105 Ϫ/Ϫ mice to study the requirement of ERK activation for the expression of proinflammatory cytokines, TNF-␣, IL-6, and IL-12, and nitric oxide in response to P. falciparum GPIs. Macrophages from the NF- B1/p105 Ϫ/Ϫ and wild type mice were stimulated with P. falciparum GPIs, and cell lysates were analyzed for the activation of ERK, p38, JNK, and NF-B pathways by Western blotting. The macrophages from NF-B1/p105 Ϫ/Ϫ mice responded normally to the parasite GPI-induced stimulation with regard to p38, JNK activation, and IB␣ degradation (indicative of NF-B activation) but were deficient in the activation of ERK1/ ERK2 even after prolonged (2 h) stimulation with the parasite GPIs ( Fig. 5A and data not shown). NF-B1/p105 is known to be involved in the stabilization of Tpl2, a MAP3K (MAPK kinase kinase) upstream of ERK1/ERK2 (46,47). Consistent with this finding, the level of Tpl2 was also markedly low in macrophages from NF-B1/p105 Ϫ/Ϫ mice even though the Tpl2 mRNA level was normal (Fig. 5, B and C). Furthermore, macrophages from NF-B1/p105 Ϫ/Ϫ mice were defective in the transcription of Egr-1, a transcription factor that is phosphorylated by ERK (Fig. 5C and see Ref. 46).
It has been previously shown that macrophages from NF-B1/p105 Ϫ/Ϫ mice when stimulated with LPS showed very low levels of ERK2 activation but not ERK1 (46). In contrast, in this study NF-B1/p105 Ϫ/Ϫ macrophages stimulated with P. falciparum GPIs exhibited a very low level of ERK1 activation but not EKR2 (Fig. 5A). Nevertheless, because NF-B1/p105 Ϫ/Ϫ macrophages are defective in ERK1/ERK2 activation, these cells served as a useful model for studying the requirement of ERK-signaling pathway for the expression of P. falciparum GPI-induced proinflammatory cytokine and nitric oxide production.
As described above (see Figs. 2 and 3), the production of TNF-␣ and nitric oxide as well as the mRNA levels of TNF-␣ and iNOS in NF-B1/p105 Ϫ/Ϫ macrophages was normal despite the cells being defective in the activation of the ERK-signaling pathway. These data show that the ERK-signaling pathway is not involved in the expression of TNF-␣ and nitric oxide by macrophages stimulated with GPIs. In contrast, the GPI-induced NF-B1/p105 Ϫ/Ϫ macrophages produced significantly lower levels of IL-12 and IL-6 compared with wild type macrophages at both protein and mRNA levels (see Figs. 2 and 3). These data suggest at a first glance that ERK1/ERK2 activation is critically required for the production of IL-12 and IL-6 in macrophages. However, when the activation of ERK1/ERK2 in wild type macrophages was inhibited with specific inhibitors, the production of IL-12 and IL-6 was markedly enhanced (see Fig. 6A). Therefore, as shown recently (43), the observed strikingly low level expression of IL-12 and IL-6 in NF-B1/p105 Ϫ/Ϫ macrophages is likely due to the functional impairment of IB rather than the lack of ERK1/ERK2 activation in NF-B1/ p105 Ϫ/Ϫ macrophages.
The requirement of the ERK-signaling pathway for the parasite GPI-induced proinflammatory responses was also assessed using the specific inhibitors of ERK1/ERK2 activation, namely PD98059 and UO126. Pretreatment of macrophages with inhibitors followed by stimulation with P. falciparum GPIs led to a 20 -30% inhibition of TNF-␣ and nitric oxide production compared with control macrophages (Fig. 6A). However, the normal levels of TNF-␣ and nitric oxide production by NF-B1/p105 Ϫ/Ϫ macrophages deficient in ERK1/ERK2 activation in response to parasite GPIs argue that the observed marginal inhibition of these inflammatory mediators is likely due to some level of either nonspecific inhibition or cytotoxicity. In contrast to these results, treatment of macrophages with PD98059 and UO126 led to markedly elevated levels of GPI- induced expression of IL-6 and IL-12 compared with cells stimulated with GPIs in the absence of inhibitors (Fig. 6A). With both inhibitors the effect was much more pronounced for the production of IL-12 than that of IL-6. These results agree with the reported negative regulation of IL-12 by T. cruzi GPIs and LPS in macrophages (31,48). Western blot analysis of the cell lysates showed that, as expected, both PD98059 and UO126 substantially inhibited the GPI-induced activation of ERK1/ ERK2 (Fig. 6B). These data taken together with the results of NF-B1 Ϫ/Ϫ macrophages (see Figs. 2 and 3) demonstrate that ERK1/ERK2 activation is not essential for the P. falciparum GPI-induced production of TNF-␣ and nitric oxide. However, ERK1/ERK2 are involved in controlling the levels of IL-6 and IL-12, i.e. the ERK signaling pathway negatively regulates the parasite GPI-induced expression of IL-6 and IL-12.
The Requirement of the p38 Pathway for the P. falciparum GPI-induced Proinflammatory Responses-The p38 MAPK is composed of four isoforms, designated p38␣ (or simply p38), p38␤, p38␥, and p38␦. However, it has been reported that only p38␣ is important for inflammatory responses (49). Because p38 knock-out mice are not available and to determine the requirement of the p38 signaling pathway for the parasite GPI-induced inflammatory responses, we used highly specific inhibitors of p38 activity, SB203580 and SB202190 (sensitive to both ␣,␤ isoforms) (50).
Stimulation of bone marrow-derived macrophages from wild type mice with the parasite GPIs in the presence of SB203580 showed only a marginal decrease in the production of TNF-␣ and nitric oxide compared with that in GPI-treated control macrophages (Fig. 7A). However, under the same conditions, the production of IL-6 was reduced by 90% and that of IL-12 was completely abrogated (Fig. 7A). The production of IL-6 and IL-12 was markedly inhibited even at 1.25-5 M concentrations of SB203580. Similar results were obtained with another related p38-specific inhibitor, SB202190 (data not shown). Consistent with the known property of these inhibitors, which is that they interfere with the p38 kinase activity but not p38 phosphorylation, Western blot analysis of the lysates from macrophages stimulated with GPIs in the presence 10 M SB203580 showed normal p38 phosphorylation (Fig. 7B). Together the above results demonstrate that the p38-signaling pathway is differentially required for the GPI-induced proinflammatory cytokine and nitric oxide production. The Requirement of the JNK Pathway for the P. falciparum GPI-induced Proinflammatory Responses-Previous studies have shown that activation of the JNK-signaling pathway is crucial for LPS-induced TNF-␣ (51)(52)(53). JNK is comprised of 10 isoforms formed by the alternative splicing of mRNA from three genes, JNK1, JNK2, and JNK3 (53). Of these, only JNK1 and JNK2 are expressed in monocytes and macrophages, and JNK3 is selectively expressed in nervous tissues and to a lesser extent in heart and testis (54). We studied the role of JNK1 (MAPK8) and JNK2 (MAPK9) in bone marrow-derived macrophages from the JNK1 and JNK2 knock-out mice. Upon stimulation with P. falciparum GPIs, the levels of TNF-␣, nitric oxide, IL-6, and IL-12 produced by the JNK1 Ϫ/Ϫ macrophages were comparable with those of the GPI-treated wild type macrophages (Fig. 8A). Because the JNK pathway is thought to be crucial for inflammatory responses (51)(52)(53), these results suggest that JNK1 is either not required for the parasite GPIinduced inflammatory responses, or JNK2 compensates for the function of JNK1. In the case of JNK2 Ϫ/Ϫ macrophages, stimulation with the parasite GPIs could induce TNF-␣, IL-6, and nitric oxide at levels compared with those produced by the GPI-treated wild type macrophages (Fig. 8A). These results together with those of JNK1 Ϫ/Ϫ macrophages indicate that JNK1 and JNK2 are either not involved in the production of TNF-␣, IL-6, and nitric oxide by macrophages stimulated with GPIs or these signaling molecules are functionally redundant. However, in GPI-treated JNK2 Ϫ/Ϫ macrophages, the level of IL-12 was substantially (ϳ80%) decreased, suggesting that JNK2 is critical for IL-12 production, and JNK1 does not compensate for JNK2 for IL-12 expression. Consistent with these results, real-time RT-PCR analysis showed that the levels of TNF-␣, IL-6, and iNOS transcripts in JNK2 Ϫ/Ϫ macrophages stimulated with GPIs were comparable with those in wild type macrophage control (Fig. 8B). However, the level of IL-12 mRNA was markedly lower in GPI-induced JNK2 Ϫ/Ϫ macrophages compared with wild type macrophages (Fig. 8B).
The requirement of JNK for the P. falciparum GPI-induced inflammatory responses in mouse bone marrow-derived macrophages was also studied by using the inhibitor of JNK enzyme activity, SP600125 (55). SP600125 has been shown to specifically inhibit JNK enzymatic activity at Ͻ10 M concentrations (55). SP600125 inhibited the GPI-induced production of TNF-␣, IL-6, IL-12, and nitric oxide in macrophages in a dose-dependent manner (Fig. 9A). The inhibitory effect of SP600125 on IL-12 production was much more pronounced than that on the production of TNF-␣, IL-6, and nitric oxide. These results taken together with the results of JNK1 Ϫ/Ϫ and JNK2 Ϫ/Ϫ macrophages indicate that JNK1 and JNK2 compensate for each other for TNF-␣, IL-6, and nitric oxide production but not for the expression of IL-12. As expected, the SP600125-treated macrophages showed normal or somewhat decreased (at 10 M SP600125) levels of JNK phosphorylation upon stimulation with GPIs (Fig. 9B). This could be due to the inhibition of an upstream kinase and/or autophosphorylation (55). DISCUSSION It is becoming increasingly clear that GPIs of various parasitic protozoa, including Plasmodia, Trypanosome, and Leishmania, play a prominent role in the immunopathology of the parasitic infection by their ability to induce potent proinflammatory responses (56). Thus, identification of the signaling molecules involved in the expression of various proinflammatory mediators may offer therapeutic targets for severe malaria. In the preceding manuscript (28) we showed that stimulation of monocytes and macrophages with P. falciparum GPIs leads to the activation of ERK, p38, JNK, and NF-B downstream signaling pathways through the TLR/MyD88-mediated signaling mechanism. The results are consistent with the notion that the TLR/MyD88-dependent signaling of macrophages by various pathogen-associated molecular patterns, such as LPS, lipoproteins, peptidoglycans, and T. cruzi GPIs involves the activation of downstream ERK-, p38-, JNK-, and NF-Bsignaling pathways. Although, the activation of these signaling pathways in response to inflammatory stimuli is a common event, their involvement in the expression of various proinflammatory mediators is not clear and in many instances is controversial (30). In this study using macrophages from gene knock-out animals and specific inhibitors, we define the requirement of downstream signaling pathways for P. falciparum GPI-induced expression of proinflammatory cytokines and nitric oxide. Our data indicate that ERK-, p38-, JNK-, and NF-B-signaling molecules are differentially involved in the expression proinflammatory mediators in macrophages stimulated with GPIs.
The results of this study show that the activation and translocation of NF-B complex is obligatory for the GPIinduced expression of proinflammatory mediators, TNF-␣, IL-6, IL-12, and nitric oxide (see Fig. 4A). These findings are in agreement with the known information that NF-B is a key transcriptional factor that regulates the gene expression of a number of cytokines, chemokines, and growth factors by macrophages in response to stimulation with microbial agents. However, NF-B comprises a family of proteins, and much needs to be learned about the involvement of specific NF-B family members in the TLR/MyD88-mediated expression of proinflammatory mediators in response to microbial ligands, including GPIs. Our data show that the GPI-induced expression of IL-6 and IL-12 but not that of TNF-␣ and nitric oxide is specifically regulated by p50 derived from NF-B1/ p105. This is evident from 1) the production of significantly lower levels of IL-6 and IL-12 and normal levels of TNF-␣ and nitric oxide by GPI-stimulated macrophages pretreated with SN50, an inhibitor of translocation of p50-containing NF-B complexes (see Fig. 4B), and 2) the expression of normal levels of TNF-␣ and nitric oxide and markedly low levels of IL-6 and IL-12 by NF-B1/p105 Ϫ/Ϫ macrophages stimulated with GPIs (see Fig. 2). These results are also consistent with the results of a previous study that reported that the GPIinduced expression of TNF-␣ and nitric oxide is regulated by NF-B/c-Rel transcription factor (23). We observed that the expression of IL-6 and IL-12 was markedly reduced at both protein and transcription levels in NF-B/p105 Ϫ/Ϫ macrophages (see Figs. 2 and 3). These data are consistent with the recent finding that IB, a new member of the IB family proteins that is inducible in macrophages upon stimulation with TLR ligands, regulates the expression of a subset of genes, including IL-6 and IL-12, by recruitment to the DNAbinding sites of the gene promoters (43,57). The function of IB appears to require its association with the p50-derived form NF-B1/p105. Thus, our data demonstrate for the first time that NF-B1/p105 specifically regulates the GPI-induced expression of IL-6 and IL-12 but not TNF-␣ and nitric oxide by macrophages.
The results presented in this paper show that the activation of ERK1/ERK2 is differentially involved in the production of proinflammatory mediators by macrophages stimulated with FIG. 5. Analysis of MAPK and NF-B-signaling pathways in macrophages from NF-B1/p105 ؊/؊ mice. Panel A, bone marrow-derived macrophages from wild type (WT) B6.129PF2/J and NF-B1/p105 Ϫ/Ϫ mice were plated in 24-well microtiter plates (5 ϫ 10 5 cells/well). The cells were stimulated with 200 nM P. falciparum GPIs. At the indicated time points the cell lysates were prepared and analyzed by Western blotting using 10% SDS-polyacrylamide gels and anti-peptide and phospho-specific antibodies against the indicated proteins. P, phospho-specific antibodies. Panel B, cells from wild type and NF-B1/p105 Ϫ/Ϫ mice were stimulated as above. At the indicated time points the cells lysates analyzed by Western blotting using antibodies against Tpl2 and ␤-tubulin. Panel C, cells from wild type and NF-B1/p105 Ϫ/Ϫ mice were stimulated as described under "Experimental Procedure." The cells were harvested, total RNA was isolated, Tpl2 and Egr-1 genes were amplified by RT-PCR and analyzed on 1% agarose gel. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA was analyzed as an endogenous control.
P. falciparum GPIs. Previously, several laboratories have studied the role of the ERK-signaling pathway for the production of TNF-␣ and other inflammatory mediators in response to various ligands, including LPS (51,58,59). It is generally believed that ERK1/ERK2 are activated by Raf isoforms in response to mitogenic or proliferative stimuli and that they are unlikely to be involved in the proinflammatory responses (59). Tpl2/Cot has also been reported to activate ERK1/ERK2 in response to proinflammatory stimulation, and one study showed that Tpl2 Ϫ/Ϫ macrophages were unable to produce TNF-␣ in response to LPS (58). However, the requirement of Tpl2-dependent activation of ERK1/ERK2 for the expression of TNF-␣ and other proinflammatory mediators has not been further studied. In this study we assessed the role of ERK1/ERK2 activation in GPI-stimulated proinflammatory responses by macrophages using the inhibitors PD98059 and UO126 that are widely used by various laboratories. As previously reported for LPS and T. cruzi GPIs (31), our results with the inhibitors showed a 20 -30% decrease in TNF-␣ and nitric oxide expression in response to P. falciparum GPIs. However, in our studies using NF-B1/p105 Ϫ/Ϫ macrophages that are deficient in ERK activation, the cells were able to produce normal levels of TNF-␣ and nitric oxide compared with wild type macrophages. These results clearly demonstrate that ERK signaling is not involved in the GPI-induced TNF-␣ and nitric oxide. These results taken together argue that the observed marginal decrease in the TNF-␣ and nitric oxide production in GPI-treated macrophages in the presence of PD98059 and UO126 is not due to the inhibition of ERK1/ERK2 activation but likely is due to either a partial cytotoxic effect of the inhibitors or a nonspecific inhibition of other signaling molecules.
In this study the inhibition of ERK1/ERK2 activation by PD98059 and UO126 led to the marked increase in GPI-induced production of IL-6 and IL-12 (see Fig. 6A). Similar results have been observed in macrophages stimulated with T. cruzi GPIs and LPS in the presence of PD98059 (31,48). Although the reason for the negative regulatory effect of the ERK activation inhibitors on IL-6 and IL-12 expression by macrophages remains to be determined, our preliminary studies showed enhanced transcription of IL-6 and IL-12 in the presence of ERK activation inhibitors. 2 Therefore, we speculate that the observed enhanced IL-6 and IL-12 expression is due to increased level of p50 and/or efficient binding of p50/IB to the gene promoters. In contrast to the results from the inhibitors, as mentioned above, the NF-B1/p105 Ϫ/Ϫ macrophages showed marked reduction in the levels of IL-6 and IL-12 produced in response to P. falciparum GPIs compared with control macrophages. These results at first glance suggest that ERK1/ERK2 activation is critically required for the production of IL-6 and IL-12. However, based on the results of a recent study, the observed reduction in the expression of these cytokines is most 2 J. Zhu and D. C. Gowda, unpublished data. likely due to the functional impairment of IB rather than the lack of normal ERK1/ERK2 activation (43). Thus, our data from the inhibitor and NF-B1/p105 Ϫ/Ϫ macrophage studies together argue that the ERK-signaling pathway is not critical for the production of certain proinflammatory mediators, such as TNF-␣ and nitric oxide by GPI-stimulated macrophages, but negatively regulates the induction of other proinflammatory mediators, IL-6 and IL-12.
The data presented here demonstrate that the p38-signaling pathway is differentially required for the P. falciparum GPIinduced expression of TNF-␣, IL-6, IL-12, and nitric oxide. Inhibition of p38 activity by SB203585 or SB202190 led to the almost complete abrogation of GPI-induced IL-6 and IL-12 expression by macrophages, whereas the production of TNF-␣ and nitric oxide was only marginally reduced, indicating that the p38-signaling pathway is crucial for IL-6 and IL-12 and is either marginally required or not essential for TNF-␣ and nitric oxide production ( Fig. 7A and data not shown). Our conclusion that p38 is not involved in nitric oxide production agrees with the results of previous studies, which reported that nitric oxide production by macrophages stimulated with T. cruzi GPIs or LPS is not affected by the inhibition of p38 activity (31,60,61). However, with respect to the regulation of TNF-␣ and IL-12 by the p38 pathway, our results are in sharp contrast to those observed in the case of T. cruzi GPIs (31). In macrophages stimulated with T. cruzi GPIs or LPS, the inhibition of p38 activity by 10 M SB203580 affected TNF-␣ production by 60%, whereas the expression of IL-12 was only slightly (ϳ25%) reduced (31). In contrast to these results, however, Lu et al. (62) showed that macrophages deficient in p38 activity (by disruption of MKK3, a specific MAPK upstream of p38) expressed drastically reduced levels of IL-12 in response to LPS, whereas TNF-␣ was normal. Similar results were observed using SB203580 or SB202190 (30,62). Furthermore, Allen et al. (63) found that the production of IL-1 and IL-6 in p38␣ Ϫ/Ϫ mouse embryonic fibroblasts was markedly compromised. Thus, these results agree with our conclusion that p38 is crucial for GPIinduced expression by macrophages of IL-6 and IL-12 but not of TNF-␣ and nitric oxide.
The requirement of the JNK-signaling pathway for the GPIinduced expression of proinflammatory mediators was not studied before. In this study we show that the JNK pathway is critical for P. falciparum GPI-induced proinflammatory responses. As shown in Fig. 9A, pretreatment of macrophages with SP600125 resulted in a complete blockade of IL-12 and a marked decrease in the level of TNF-␣ and nitric oxide production; interestingly, the production of IL-6 was only moderately inhibited. These results are in general agreement with the notion that the JNK pathway is important for the production of proinflammatory cytokines in response to LPS and doublestranded RNA (51)(52)(53)64). Our results also demonstrate that of the two JNK isoforms, JNK1 and JNK2, present in macrophages are for the most part functionally redundant. This conclusion is evident from the observed normal level of expression of TNF-␣, IL-6, and nitric oxide in macrophages from JNK1 Ϫ/Ϫ or JNK2 Ϫ/Ϫ mice (see Fig. 8A). However, for the expression of IL-12, JNK2 but not JNK1 is critical, and this requirement is at the transcriptional level (see Fig. 8B). These FIG. 8. Production of proinflammatory cytokines and nitric oxide in P. falciparum GPI-stimulated JNK ؊/؊ macrophages. Panel A, bone marrow-derived macrophages from C57BL6/J mice (wild type), JNK1 Ϫ/Ϫ , and JNK2 Ϫ/Ϫ mice were plated into 96-well microtiter plates (ϳ2.5 ϫ 10 4 cells/well). After overnight culturing the cells were primed with IFN␥ and stimulated with 40 nM GPIs. At the indicated time points the culture supernatants were assayed for TNF-␣, IL-6, IL-12, and nitric oxide. Open bars, macrophages from wild type mice; black bars, macrophages from JNK1 Ϫ/Ϫ mice; hatched bars, macrophages from JNK2 Ϫ/Ϫ mice. The experiments were performed two times, each in duplicate, average values were plotted, and the error range is shown. Panel B, the macrophages (1 ϫ 10 6 cells) from wild type and JNK2 Ϫ/Ϫ mice were stimulated with 200 nM GPIs. At the indicated time points the cells were harvested, and total RNA was isolated and analyzed for TNF-␣, IL-6, IL-12, and iNOS transcripts by real time RT-PCR as described under "Experimental Procedure." Open bars, macrophages from wild type mice; hatched bars, macrophages from JNK2 Ϫ/Ϫ mice. The experiments were performed two times, average values were plotted, and the error range is shown.
observations agree with the previous findings that the expression of IFN-␣, IFN-␤, IL-6, and IL-12 was substantially decreased in JNK2 Ϫ/Ϫ fibroblasts in response to double-stranded RNA and that JNK1 does not compensate for JNK2 (64). Although JNK1 is unable to compensate for JNK2 in fibroblasts, as shown here for P. falciparum GPI-induced macrophages, JNK1 efficiently compensated for JNK2 for the expression of IL-6. This difference is likely due to the different cell types and/or the ligands used.
In conclusion, the results of this study show that of the downstream signaling pathways that are activated in macrophages stimulated with P. falciparum GPIs, JNK and NF-B pathways are crucial for the production of key proinflammatory mediators, TNF-␣, IL-6, IL-12, and nitric oxide. The ERK-and p38-signaling pathways and JNK1/JNK2 and NF-B family transcription factors are differentially involved in the expression of TNF-␣, IL-6, IL-12, and nitric oxide. Further studies aimed at understanding the details of these signaling pathways in various cell types of the immune system and identification of P. falciparum ligands other than GPIs and their cell signaling mechanism are likely to be of therapeutic benefit for severe malaria. FIG. 9. The effect of JNK inhibition on the production of inflammatory cytokines and nitric oxide in macrophages stimulated with P. falciparum GPIs. Panel A, bone marrow-derived macrophages from C57BL6/J mice were plated into 96-well microtiter plates (2.5 ϫ 10 4 cells/well). The cells were treated with the indicated concentrations of SP600125 for 1 h, primed with IFN␥, and then stimulated with 40 nM GPIs in the presence of the inhibitor. Cells primed and stimulated but not treated with the inhibitor were used as controls. After 48 h of stimulation, the culture supernatants were assayed for TNF-␣, IL-6, IL-12, and nitric oxide. Panel B, the macrophages (5 ϫ 10 5 cells/well) were treated with SP600125 for 1 h and then stimulated with 200 nM GPIs for 25 min. The cell lysates were prepared and analyzed by Western blotting using anti-JNK peptide and phospho (P)-specific antibodies.