Phosphatidic acid regulates systemic inflammatory responses by modulating the Akt-mammalian target of rapamycin-p70 S6 kinase 1 pathway.

Macrophages are pivotal effector cells in the innate immune system. When microbial products bind to pathogen recognition receptors, macrophages are activated and release a broad array of mediators, such as cytokines, that orchestrate the inflammatory responses of the host. Phosphatidic acid (PA) has been implicated as an important metabolite of phospholipid biosynthesis and in membrane remodeling and has been further suggested to be a crucial second messenger in various cellular signaling events. Here we show that PA is an essential regulator of inflammatory response. Deleterious effects of PA are associated with the secretion of proinflammatory cytokines, such as tumor necrosis factor-alpha, interleukin-1beta, interleukin-6, and the production of nitric oxide, prostaglandin E2, which are predominantly released by macrophage Raw264.7 cells. Furthermore, the administration of PA to mice increased the serum cytokine level. Moreover, direct or lipopolysaccharide-induced PA accumulation by macrophages led to the Akt-dependent activation of the mammalian target of rapamycin-p70 S6 kinase 1, a process required for the induction of inflammatory mediators. These findings demonstrate the importance of the role of PA in systemic inflammatory responses, and provide a potential usefulness as specific targets for the development of therapies.

Mononuclear phagocytes represent a large family of cell types that includes macrophages, Kupffer cells, and microglia. Macrophages exert key functions during the innate immune response, which is vital for the recognition and elimination of invasive microbial pathogens (1,2). However, under certain circumstances macrophages have deleterious effects. This is the case in septic shock, which is a severe systemic inflammatory response triggered by interaction between some bacterial components and macrophages and other host cells (3). Many pathogenic mediators of sepsis have been described, including lipopolysaccharide (LPS). 1 As the major outer membrane com-ponent of Gram-negative bacteria, LPS serves as a potent modulator of acute inflammation (4). The most important factors of this are believed to be TNF-␣, and IL-1␤, but other factors play significant roles, including IL-6, IL-8, macrophage migration inhibitory factor (5), and high mobility group 1 protein (6). LPS also triggers the activation of multiple intracellular signaling cascades, resulting in the release of immunoregulatory molecules such as nitric oxide (NO) and prostaglandin (PG) (7,8). This process may lead not only to tissue damage, but also to hemodynamic changes, multiple organ failure, and ultimately death.
Phosphatidic acid (PA) participates and regulates numerous cellular pathways, including ligand-mediated secretion, endocytosis, and respiratory burst (9 -11). Phospholipase D (PLD), an important effector enzyme in receptor-mediated signaling pathways, catalyzes the hydrolysis of the most abundant membrane phospholipid, phosphatidylcholine, and generates PA and choline. PLD was first identified in plants but has subsequently been shown to be highly conserved across all species and to be present in large amounts in bacteria, yeast, and mammalian cells (12,13). A role for PA as a key intracellular signaling molecule has been proposed, as PA has been shown to directly activate protein kinases (14), protein-tyrosine phosphatase (15), phosphoinositide 4-kinase (16), and sphingosine kinase (17). Extensive attention has been recently given to the diverse biological effects of PA released by activated macrophages and multiple other cells (18,19). Of particular interest, PA acts as an intermediary messenger with various selective pro-inflammatory targets. PA has been suggested to protect by pharmacologic inhibition in LPS-induced septic mice (20). Zhang et al. (21) found that the stimulation of LPS leading to PLD activation results in inducible nitric-oxide synthase (iNOS) induction. However, in these studies, the action mechanism of the PLD activated by LPS was not defined, and the relationship of the activation of PLD to the various signaling enzyme cascades following LPS stimulation was unknown. Although the regulation of PA generation represents a multipotent strategy for modulating the occasionally self-destructive forces unleashed by exuberant self-defensive responses, the intracellular signaling mechanisms triggered by PA in cells have not been fully elucidated.
The regulation of translation initiation is the main control point of mRNA translation. Several steps in translation are regulated by signaling events coupled to the mammalian target of rapamycin (mTOR), a Ser/Thr protein kinase that is specifically inhibited by the immunosuppressant rapamycin (22). Its best known downstream effectors include the ribosomal sub-

FIG. 2. Effects of PA injection on the production of pro-inflammatory cytokines in vivo.
A, C57BL/6 mice were treated with control saline or 1 mg of PA for the indicated times. B, mice were treated with the indicated doses of PA and then sacrificed. Serum levels of TNF-␣, IL-1␤, and IL-6 were determined at the indicated time points. Data points correspond to the mean and the S.D. of two independent experiments (5-7 mice/ treatment group). unit S6 kinase (S6K) and the eukaryotic initiation factor 4Ebinding protein (4E-BP1) (23). Both the activation of S6K and the phosphorylation of 4E-BP1 are stimulated by mitogens, and mTOR is required to trigger this mitogen response (24). In addition, mitogenic activation of S6K and 4E-BP1 requires the phosphatidylinositol 3-kinase (PI3K) pathway (25). Recently, Fang et al. (26) reported that PA (likely to be produced by PLD) directly mediates the mitogenic stimulation of mTOR signaling to S6K1. However, a major unanswered question is how PA is sensed by cells and how such signals are transduced through mTOR to downstream effectors.
We hypothesized that the PA is involved in macrophage activation leading to production of pro-inflammatory cytokines, NO, and PGE 2 . To examine this possibility, we used Raw264.7 cells to test the production of these cytokines, NO, and PGE 2 in the presence of synthetic PA or LPS. The addition of synthetic PA to macrophages led to an increase in pro-inflammatory mediators through the Akt and the mTOR-p70 S6K1 pathways. We also analyzed PLD activity in LPS-treated macrophages and investigated whether produced PA can modulate pro-inflammatory mediators. We found that LPS-stimulated PA clearly induced pro-inflammatory mediators, an effect that was mediated by the Akt-mTOR pathway. These findings support the hypothesis that PLD generating PA is an important factor in the production of macrophage systemic inflammatory mediators.

EXPERIMENTAL PROCEDURES
Reagents and Antibodies-LPS (Escherichia coli serotype 0111:B4), dioctanoyl PA (99%), 1-butanol, and 1-propranolol were obtained from Sigma; LY294002 was from BIOMOL (Plymouth Meeting, PA); and rapamycin was obtained from Cell Signaling Technology (Beverly, MA). For Western blot analysis, we used a rabbit Ab against iNOS (Santa FIG. 3. PA activation of iNOS, COX-2 gene expression, and their protein products. A, Raw264.7 cells were treated for the indicated times with 100 M synthetic PA. Nitrites and PGE 2 secreted into the medium were quantified by Griess reagent and EIA, as described under "Experimental Procedures." B, dose-response analysis of iNOS and COX-2 protein expression and of their products after 24 h of treatment with PA. Western blot analysis of iNOS or COX-2 were performed using anti-iNOS or anti-COX-2 antibodies. C, Raw264.7 cells were transiently transfected with iNOS or COX-2 promoter-driven luciferase reporter plasmid. 24 h after transfection, cells were incubated for 12 h with the indicated concentrations of PA. Luciferase activity was measured as described under "Experimental Procedures" and normalized with respect to total protein. Cruz Biotechnology, Santa Cruz, CA), cyclooxygenase-2 (COX-2; Cayman Chemicals, Ann Arbor, MI), phospho-Akt (Thr-308), Akt, phospho-mTOR (Ser-2448), phospho-p70 S6K1 (Thr-389), and p70 S6K1 (Cell Signaling Technology). Peroxidase-conjugated anti-rabbit IgG, antigoat IgG, or anti-mouse IgG were used as secondary antibodies.
Cell Culture-Raw264.7 cells were obtained from the American Type Culture Collection (Manassas, VA) and were cultured in RPMI 1640 (Invitrogen) with 10% fetal bovine serum, 2 mM L-glutamine, 10 units/ml penicillin, and 10 g/ml streptomycin at 37°C in 5% CO 2 in a water-saturated atmosphere. The cells were treated with LPS or synthetic PA for the indicated times.
Determination of NO and PGE 2 Production-NO production was estimated by measuring nitrate/nitrite in the cell culture media. Macrophages were cultured in RPMI, and samples were stored at Ϫ80°C until assayed. Reduced samples were incubated with an equal volume of Griess reagent, and absorbance was measured at 550 nm. Total nitrate/nitrite concentration was determined versus a standard curve. PGE 2 production was measured using a commercial PGE 2 enzyme immunoassay (EIA) kit (Amersham Biosciences). Cells were cultured in 6-well or 12-well plates and stimulated with PA or LPS. Supernatant samples were obtained at the indicated times and subjected to EIA analysis.
Cytokine Measurements-The concentrations of TNF-␣, IL-1␤, and IL-6 were determined by cytokine-specific enzyme-linked immunosor-bent assays, according to the instructions from the manufacturer (R&D Systems).
Akt, mTOR, and p70 S6K1 Activation-Cells were treated with PA or LPS as described above. Cells were then lysed with lysis solution, and Western blots were performed as described above; membranes so obtained were probed with Ab against active (phosphorylated) Akt, mTOR, or p70 S6K1. Blots were stripped and reproved with Ab recognizing both active and total Akt, mTOR, or p70 S6K1, to demonstrate equal loading.
PLD Activity Assay-For the PA production analysis, cells were labeled with [ 32 P]orthophosphate (5 Ci/ml) in DMEM phosphate-free medium containing 1% fetal bovine serum for 24 h. The radioactive medium was replaced with DMEM, and challenged with LPS in DMEM containing 0.1% bovine serum albumin for specific times. Incubations were terminated by the addition of 1 ml of methanol:HCl (100:1), and lipids were extracted. PA was identified by the co-spotting of standards and quantified after separating total radioactive lipids by TLC using ethyl acetate, 2,2,4-trimethyl pentane, glacial acetic acid, and water (65:10:15:50) as the developing solvent system.

FIG. 4. Involvement of Akt in PA-induced pro-inflammatory cytokines, and nitrites and PGE 2 production.
A, cells were treated with 100 M PA for the indicated times. Cells were lysed, and the lysates were analyzed by SDS-PAGE and immunoblotted with anti-phospho-Akt Ab specifically recognizing the phosphorylated Thr-308 residue. The same blot was reprobed with anti-Akt Ab. B, cells were pretreated with the PI3K inhibitor LY294002 (25 M) for 30 min before PA stimulation for 20 min. Cell lysates were analyzed for Thr-308 phosphorylation as in A. Cells were pretreated with LY294002 for 30 min before PA stimulation for 18 h. C, the concentrations of TNF-␣, IL-1␤, and IL-6 in the culture media were determined by ELISA. D, nitrites and PGE 2 secreted into the medium were quantified as described under "Experimental Procedures." Western blot analyses of iNOS and COX-2 were performed using anti-iNOS or anti-COX-2 antibodies.
Transfection and Promoter Activity Assay-Cells were transfected with full-length iNOS or COX-2 promoter DNA using LipofectAMINE 2000. After transfection, cells were incubated in complete media for 24 h at 37°C before being stimulated with PA or LPS for 12 h at 37°C. In some experiments, cells were preincubated with 1-butanol or 1-propranolol for 10 min at 37°C before adding LPS. Cell lysates were assayed for luciferase activity according to the instructions from the manufacturer (Promega) using a luminometer.
Total RNA Isolation and RT-PCR-Total RNA was isolated from PA-stimulated Raw264.7 cells using Tri reagent according to the in-structions from the manufacturer and was then reverse transcribed using a GeneAmp RNA PCR core kit (Applied Biosystems). The PCR was performed using an annealing temperature of 55°C with the following primers: TNF-␣ forward (5Ј-TTC TGT CTA CTG AAC TTC GGG GTG ATC GGT CC-3Ј) and reverse (5Ј-GTA TGA GAT AGC AAA TCG GCT GAC GGT GTG GG-3Ј, which generates a product of 354 bp), IL-1␤ forward (5Ј-GAA GCT GTG GCA GCT ACC TAT GTC T-3Ј) and reverse (5Ј-CTC TGC TTG TGA GGT GCT GAT GTA C-3Ј, which generates a product of 523 bp), IL-6 forward (5Ј-TTC CCT ACT TCA CAA GTC-3Ј) and reverse (5Ј-ACT AGG TTT GCC GAG TAG-3Ј, which generates a product of 564 bp), and the housekeeping gene ␤-actin forward (5Ј-TCC TTC GTT GCC GGT CCA CA-3Ј) and reverse (5Ј-CGT CTC CGG AGT CCA TCA CA-3Ј, which generates a product of 509 bp).
Protein Extraction and Western Blot Analysis-Cells were washed twice in cold phosphate-buffered saline and lysed on ice with lysis solution (1% Triton X-100, 50 mM Tris, pH 8.0, 150 mM NaCl, 1 mM phenylmethylsulfonyl fluoride, 1 mM Na 3 VO 4 , and protease inhibitor mixture). Sample protein concentration was determined using the Bio-Rad protein assay. Proteins from the cell lysates (30 g) were boiled at 95°C in Laemmli SDS loading buffer, separated on 8% SDS-PAGE, and electrotransferred to nitrocellulose membranes. The membranes were blocked for at least 30 min at room temperature in Tris-buffered saline plus 0.05% Tween-20 (TTBS) containing 5% nonfat dry milk and then incubated with TTBS containing the primary Ab. For immunoblotting, incubation was performed for 4 h at room temperature. After five washes of 30 min each in TTBS, the membranes were incubated with peroxidase-conjugated secondary Ab for 1 h. After five washes of 30 min with TTBS, enhanced chemiluminescence detection (Amersham Biosciences) was performed and the membranes were exposed to x-ray films.

PA Induces the Production of Pro-inflammatory Cytokines in
Macrophages-Exogenous PA added to cell culture medium incorporates rapidly into cellular membranes and subsequently participates in cellular functions (27,28). TNF-␣ is one of the earliest cytokines released by activated macrophages, and occupies a pivotal role in the pathogenesis of inflammation, endotoxic shock, and tissue injury. We measured TNF-␣ in supernatants conditioned by Raw264.7 macrophages exposed to PA for 2 h, and observed a PA dose-dependent stimulation of TNF-␣ (Fig. 1A). PA concentrations as low as 10 M significantly increased TNF-␣ release. To address the molecular basis of PA-induced macrophage activation, we measured TNF-␣ mRNA levels in PA-stimulated cells. RT-PCR results demonstrated that stimulation with PA significantly induced TNF-␣ mRNA production in a time-dependent manner (Fig. 1B). Upregulation of TNF-␣ mRNA was observed within 2 h of PA addition, and decreased the levels of TNF-␣ mRNA. To assess the specificity of cytokine response to PA, we measured the release of other macrophage-derived cytokines in PA-stimulated macrophage cell cultures. PA time-dependently induced the release of other pro-inflammatory cytokines, i.e. IL-1␤ and IL-6 mRNA and protein (Fig. 1, B and C), but failed to induce the release of the anti-inflammatory cytokines (IL-2, transforming growth factor-␤, and interferon-␥; data not shown). The kinetic profiles of IL-1␤ and IL-6 differed significantly from that observed in the TNF-␣, in which peak IL-1␤ and IL-6 mRNA levels occurred within 4 and 6 h, respectively (Fig. 1B). This activation was PA-specific, because none of the other phospholipids tested were found to stimulate these cytokines, including lysophosphatidic acid and diacylglycerol (Fig. 1D). To exclude contamination of endotoxin effect, the endotoxin content was determined by using a kinetic quantitative chromogenic Limulus amoebocyte lysate kit (BioWhittaker, Walkersville, MD) and showed no contamination of endotoxins (Ͻ0.001 ng/ml).
Administration of PA to Mice Stimulates Increased Serum Cytokines-To determine whether PA can stimulate cytokines synthesis in vivo, we administered PA to C57BL/6 mice and measured serum cytokines by ELISA. Cytokines levels in ani- mals injected intraperitoneally with PA were monitored and compared with those of animals that had received control injections of saline. The administration of doses of PA (0.5-2 mg) to mice significantly stimulated the appearance of TNF-␣, IL-1␤, and IL-6 in serum, indicating that the systemic exposure to PA activates systemic pro-inflammatory cytokines in vivo ( Fig.  2A). The kinetic pattern of PA-activated cytokines synthesis differed from that of LPS-induced mice. The administration of LPS to mice activated TNF-␣ synthesis within 1 h, and this then rapidly declined to basal levels as judged by ELISA (data not shown). In contrast to LPS, PA administration showed TNF-␣ synthesis within 2 h and this slowly reduced over at least 30 h (Fig. 2B). Lethality was monitored over time (10 days) and compared with animals that had received control injections of saline or LPS. 100% of the mice treated with PA survived as compared with 0% of the LPS-treated mice.
PA Induces iNOS and COX-2 Expression-Cytokines, such as TNF-␣ and IL-1␤, are major activator of macrophages, a cell type in which iNOS and COX-2 have been reported to be highly up-regulated upon exposure to several pro-inflammatory signals. We characterized the effect of PA on iNOS and COX-2 expression and on their products. To analyze the effect of PA on NO and PGE 2 production, we treated PA to cells and assayed nitrites and PGE 2 using Griess reagent and an EIA kit, respectively. Nitrites were detected within 10 h of incubation with PA and dramatically increased after 12 h. In contrast to nitrite production, PGE 2 was detected within 2 h of incubation with PA, peaked at 8 h, and slowly increased after 24 h (Fig. 3A). Treatment of macrophages with different doses of PA showed a dramatic increase in proteins and their products, although detectable levels of both proteins and their products were observed after treatment with as little as 10 or 50 M PA (Fig.  3B). Next, we investigated the mechanism by which PA induces both of these proteins, by transiently transfecting the cells with iNOS and COX-2 promoter luciferase reporter. PA was found to induce iNOS and COX-2 promoter activities in a dose-dependent manner (Fig. 3C). Taken together, these results indicate that iNOS and COX-2 protein synthesis are up-regulated by PA in mouse macrophages.
PA Induction of Pro-inflammatory Mediators Requires the Akt Pathway-We observed that extracellular PA stimulated three cytokines (TNF-␣, IL-1␤, and IL-6) and iNOS and COX-2 protein expressions in Raw264.7 cells. The understanding of the PA-initiated intracellular pathways leading those responses is important. Akt could participate in these signaling reactions, as it is a well known mediator, and was recently shown to be activated in macrophages (29). To determine whether Akt is involved in the PA-stimulated production of three cytokines and two proteins, we examined the activation of Akt by detecting their phosphorylated forms by Western blotting using specific anti-phospho-Akt Ab. PA strongly stimulated a rapid (within 10 min) increase in Akt phosphorylation, which peaked at 20 min, and then declined to base-line level (Fig. 4A). To confirm the involvement of the Akt pathway, we analyzed the phosphorylation of Akt in cells treated by LY294002, an inhibitor of PI3K/Akt. Fig. 4B illustrates the observed inhibition of the phosphorylation of Akt by LY294002. To determine whether the Akt pathway is required for the PA induction of cytokines and the two proteins in Raw264.7 cells, we examined the effect of LY294002 on PA-induced mediator production and expression. LY294002 strongly inhibited the PA induction of all cytokines. In addition, LY294002 inhibited the PA induction of nitrites and PGE 2 production, and iNOS and COX-2 expression (Fig. 4, C and D). These results indicate that the Akt pathway mediates the PA induction of cytokine production, and of iNOS and COX-2 expression.

Rapamycin Specifically Abrogates the PA-mediated Production of Cytokines and the Expressions of iNOS and COX-2-
Rapamycin, a specific mTOR inhibitor, was used to examine the role of mTOR signaling in the PA-induced production of inflammatory mediators. We examined whether the mTOR-S6K pathway is activated by PA in macrophages. PA strongly induced the phosphorylation of p70 S6K1 at Thr-389 from 20 min and this was sustained for at least 4 h. In addition, PA stimulated the phosphorylation of mTOR at Ser-2448 (Fig. 5A). As expected, rapamycin did not affect Akt phosphorylation (data not shown), but it completely inhibited p70 S6K1 phosphorylation (Fig. 5B). We then examined whether the mTOR-p70 S6K1 pathway is activated by PA through Akt. In contrast to rapamycin, LY294002 completely inhibited Akt and p70 S6K1 phosphorylation by PA (Fig. 5C). To confirm the sequential activation of PI3K-Akt-mTOR pathway, cells pretreated with the LY294002 before PA stimulation and evaluated the mTOR activation. PA-stimulated mTOR phosphorylation was significantly inhibited by LY294002 (Fig. 5D). Thus, it is clear that PA activates a rapamycin-sensitive Akt-mTOR-p70 S6K1 signaling pathway in macrophages. To confirm that mTOR-p70 S6K1 is recruited by PA to produce cytokines and the expressions of iNOS and COX-2, we determined the cytokine levels and iNOS and COX-2 expression by PA in the presence or absence of rapamycin. Rapamycin caused a dose-dependent abrogation of PA-induced cytokines, nitrites, and PGE 2 production and of iNOS and COX-2 expression (Fig. 6, A and B). Thus, the PA-induced production of pro-inflammatory mediators in macrophages occurs through an Akt-sensitive mTOR-p70 S6K1 pathway.
The PLD Activity Is Required for LPS-induced Pro-inflammatory Mediators-To determine whether PLD is involved in the LPS-induced production of pro-inflammatory mediators, we investigated the expression of PLDs in Raw264.7 cells, and found that expression of PLD 2 is enriched in these cells (Fig.  7A). We tested the response of these cells to LPS by monitoring the production of PA. Stimulation of these cells with LPS led to an acute increase in the amount of cellular PA within 5 min, which returned to its basal level after 30 min (Fig. 7B). PLD 2 is predominantly expressed in the Raw264.7 cells. Therefore, effects of PLD 2 on iNOS and COX-2 promoter activities were examined with HEK293 cells stably transfected with empty vector or human PLD 2 . iNOS and COX-2 promoter activities were significantly enhanced in PLD 2 -transfected cells compared with vector-transfected cells (Fig. 7C). 1-Butanol, but not 2-butanol (data not shown), significantly blocked the LPS-stimulated release of cytokines, whereas propranolol potentiated this LPS stimulatory effect (Fig. 8A), thus confirming the specificity of involvement of PA in the PLD-mediated pathway. Furthermore, the levels of iNOS and COX-2 proteins and of luciferase reporter activity achieved by LPS stimulation were significantly inhibited by 1-butanol and were conversely increased by propranolol (Fig. 8, B and C). Thus, PA production through PLD appears to be required for the LPS stimulation of pro-inflammatory mediators.
The rapid stimulation of PLD in LPS-treated macrophages is consistent with the idea that this phospholipase mediates the activation of Akt-mTOR-p70 S6K1. To test this hypothesis directly, p70 S6K1 phosphorylation was measured in Raw264.7 cells that had been treated with LY294002 or rapamycin before LPS exposure. Both compounds were found to substantially inhibit LPS-stimulated p70 S6K1 phosphorylation (Fig. 9A). We then examined whether LY294002 or rapamycin affected pro-inflammatory mediator production in LPS-treated raw264.7 cells, and found that both inhibitors partially inhibited mediators production stimulated by LPS (Fig. 9B). These functions of LPS were also mediated, in part, by the sequential activation of the Akt-mTOR-p70 S6K1 pathway. DISCUSSION These results demonstrate that PA acts as an important mediator that specifically stimulates pro-inflammatory cytokines, and NO and PGE 2 synthesis in macrophages. Macrophages occupy a central role in coordinating the immune response to injury and infection. The observation that PA can activate additional downstream cytokine cascades has widespread implications. Our findings also indicate that PLD mediates LPS-induced systemic inflammatory responses in macrophages. Moreover, a study using LY294002 and rapamycin demonstrated the role of the Akt-mTOR-p70 S6K1 pathway in the PA or LPS induction of inflammatory mediators. Our present results support a mechanism by which PA can participate in the host response to infection by regulating the production of pro-inflammatory cytokines and mediators.
Host innate responses to bacterial infections are primarily mediated by macrophages or monocytes (30). Stimulation of these cells initiates secretion of pro-inflammatory mediators (31), which promote the elimination of infectious agents and the induction of tissue repair. Excessive of inflammation owing to bacterial infections can lead to tissue damage and septic shock (32). LPS or endotoxin is a potent initiator of the inflammatory response and serves as an indicator of bacterial infection for the mammalian host defense system. Even though CD14 and toll-like receptor have been identified as the main pathway for LPS recognition (33,34), effective treatment of sepsis and septic shock has remained elusive. Moreover, although several molecules have been identified as the main LPS signal transduction (35)(36)(37)(38), accumulating evidence has suggested the possible existence of other functional regulator(s). Previous studies have suggested an important role for PA as an intermediary messenger with selective inflammatory targets. However, no evidence of a direct effect of PA on macrophage activation and animal toxicity has been presented. The discov- FIG. 7. Time response of LPS-induced PLD protein expression and activation. A, macrophages were incubated with the different doses of LPS for 24 h. Proteins (50 g) were separated on SDS gel, blotted, and then incubated with a specific anti-PLD1/2 Ab. B, Raw264.7 cells metabolically labeled with 32 P-orthophosphate were stimulated with 100 ng/ml LPS for various times. Cellular lipids were extracted and separated by thin layer chromatography. PA was identified by co-spotting standards, scraped, and quantified using a scintillation counter. C, HEK293 cells stably transfected with an empty vector (Vec) or human PLD 2 and cells transfected with an iNOS or a COX-2 promoter-driven luciferase reporter plasmid. 24 h later, cells were harvested before luciferase assays. ery that PA is involved in pro-inflammatory cytokine and mediator (NO and PGE 2 ) production has proved new insights into the inflammatory response of macrophages. Rice et al. (20) reported that the inhibition of PA generation protects mice from endotoxin lethality, and Abraham et al. (39) also reported that the PA generation inhibitor, lisofylline, attenuates lung injury in a mouse hemorrhage model, and the accumulation of neutrophils in lung tissue. Moreover, lisofylline suppresses TNF-␣, IL-1, and IL-6 production elicited by endotoxin, zymosan, and protein A in whole blood ex vivo (40). More recent studies have shown that PA can serve as an inflammatory mediator by inducing IL-8 secretion (41,42). The production of pro-inflammatory cytokines, NO, and PGE 2 by PA was specific, because none of the other phospholipids tested were found to stimulate these mediators; these included lysophosphatidic acid, diacylglycerol, phosphatidylcholine, and phosphatidylethanolamine ( Fig. 1D and data not shown). In addition, different acyl kinds of PA, i.e. dimyristoyl PA, also stimulated the production of pro-inflammatory cytokines in the same macrophages, suggesting that the structurally conserved PA mediates this response (data not shown). These results strongly suggest a modification of signal transduction through PA generation in inflammatory response.
PA has been implicated as a mediator of the mitogenic action of various growth factors and hormones in several types of mammalian cells. Accumulating evidence supports the idea that the activation of the PA-dependent signaling cascade contributes to the progression of some inflammatory diseases such as sepsis (43). However, the signaling pathways responsible for functional alterations of macrophages in inflammation are not yet fully understood because of their complexity, although much has been learned about the effects of inflammatory mediators on macrophages. A number of previous studies have shown that PA-dependent signaling involves the several protein kinases including extracellular signal-regulated kinase 1/2 (ERK1/2) (44) and phosphatidylinositol kinase (16,45) in mammalian cells. In addition, the study have also indicated that Akt is a crucial signaling molecule, which is required for macrophage activation in inflammation (46). In the present study, PA stimulates a series of kinases, including Akt, which in turn, transmit signals that stimulate the cellular responses of the macrophage, including production of pro-inflammatory cytokines, NO, and PGE 2 . This stimulation by PA was abolished by LY294002, thus implicating the PI3K/Akt pathway. The mitogen-activated protein kinase pathway may be involved in the production of pro-inflammatory mediators of PA-induced macrophages, because the PA increases ERK1/2 and p38 phosphorylation, and inhibition of the ERK1/2 and p38 kinase pathways with each specific inhibitors partially suppresses the production of pro-inflammatory mediators by PA (data not shown). These data suggest that the PA-dependent production of pro-inflammatory mediators is mainly regulated through PI3K/Akt pathways. On the other hand, the mitogen-activated protein kinase inhibitor did not completely inhibit PA-dependent production of pro-inflammatory mediators. This may point FIG. 8. Effect of 1-butanol and 1-propranolol on the LPS-induced production of pro-inflammatory mediators. Raw264.7 cells were grown in 6-well plates and treated with the indicated concentrations of LPS in the presence of 0.5% 1-butanol (f) or 100 M 1-propranolol (Ⅺ). A, the concentration of TNF-␣, IL-1␤, and IL-6 in the culture media was determined by ELISA. B, secreted nitrites and PGE 2 were quantified using Griess reagent and EIA as described under "Experimental Procedures." Western blot analyses of iNOS or COX-2 were performed using anti-iNOS or anti-COX-2 antibodies. Results are the means Ϯ S.D. of triplicate measurements and are representative of the results of three separate experiments. Con, control; 1-BuOH, 1-butanol; Propra, 1-propranolol. C, Raw264.7 cells were grown in 6-well plates and transiently transfected with an iNOS or a COX-2 promoter-driven luciferase reporter plasmid. 24 h after transfection, cells were incubated for 12 h with the indicated concentrations of LPS in the presence of 1-butanol (f) or 1-propranolol (Ⅺ). Luciferase activity was measured as described under "Experimental Procedures" and normalized with respect to total protein.
to multiple pathways potentially involved in PA-dependent macrophage activation.
Several stimuli, such as various growth and survival factors, regulate the p70 S6K1 primarily by triggering a signaling cascade dependent on sequential activation of PI3K, Akt, and mTOR kinases (47)(48)(49). mTOR activity can be effectively blocked by rapamycin, leading to inactivation of p70 S6K1. The best known function of mTOR, in the context of cell proliferation, is the regulation of translation initiation, presumably mediated by S6K (22,50). Fang et al. (26) reported that PA mediates the mitogenic activation of mTOR signaling to S6K1 and 4E-BP1 has uncovered a previously unexpected regulatory mode for mTOR. They suggested that PA signaling to p70 S6K1 specifically goes through mTOR, and not through PI3K in HEK293 cells, which supports the conclusion that PA has no effect on the activity of PI3K. We observed that PA stimulated p70 S6K1 phosphorylation in Raw264.7 cells and this stimulation was abolished by rapamycin, implicating mTOR involvement. The production of PA-stimulated pro-inflammatory cytokines and mediators was also abolished by rapamycin, which suggests a link between the mTOR pathway and these media-tors. In addition, PA was found to have a strong effect on the Akt activity of Raw264.7 cells in our experiment, suggesting that PA signaling is likely to affect the PI3K/Akt pathway and is cell type-specific. Moreover, the activation of Akt and p70 S6K1 by PA was virtually abrogated by pretreating Raw264.7 cells with PI3K inhibitor LY294002. This sensitivity of PA-dependent mTOR-p70 S6K1 activation to LY294002 strongly suggests that this response is dependent on the activation of Akt.
PA is the lipid product of PLD. To evaluate the significance of PLD, we used an established model of LPS stimulation in Raw264.7 cells. Endotoxic shock is a potentially lethal complication of systemic infection by Gram-negative bacteria (3,4). Macrophages respond to LPS, a major component of the cell wall of the invading Gram-negative bacteria, by inducing the expression of cytokines and enzymes involved in the production of the small pro-inflammatory mediators NO and PGs (7,8,(51)(52)(53). Moreover, inflammatory cytokines produced by these cells have been proposed as the primary mediators of this event. The release of LPS into the circulation activates a series of tissue responses that in their most severe forms lead to septic shock and death. Therefore, it is important to under- FIG. 9. Effect of LY294002 and rapamycin on the LPS-stimulated phosphorylation of Akt and p70 S6K1, and on the production of cytokine, nitrites, and PGE 2 . A, Raw264.7 cells were pretreated with 25 M LY294002 or 20 nM rapamycin for 30 min, and then exposed to 100 ng/ml LPS for 30 min. Cell lysates were analyzed by immunoblotting with anti-phospho-Akt or anti-phospho-p70 S6K1 antibodies. B, cells were pretreated with 25 M LY294002 or 20 nM rapamycin for 30 min and stimulated with LPS for 18 h. Cytokine, nitrites, and PGE 2 released into the culture medium were then quantified. Rapa, rapamycin. stand the regulation of cytokines and of both iNOS and COX-2 proteins. Furthermore, the signaling pathways induced by LPS and the manner in which these signals are linked to PLD activation and inflammatory mediator production are poorly understood. Alcohols compete with water as hydroxyl donors in the hydrolysis of phospholipids by PLD, which results in the production of phosphatidylalcohol at the expense of PA. Treatment of macrophages with 1-butanol attenuated the LPS-stimulated production of PA and of inflammatory mediators, but 2-butanol had no inhibitory effect. This is consistent with observations reported in other cells (26,54). Propranolols prevent PA dephosphorylation by PA phosphohydrolase and cause the accumulation of PA. The LPS-stimulated activation of inflammatory mediators was potentiated by propranolol under identical conditions, confirming the specificity of the effect of 1-butanol and involvement of PLD in the LPS-stimulated production of inflammatory mediators. However, signaling pathways not involving PLD probably also contribute to the response, because the PLD inhibitor did not completely impair the response of LPS.
Two forms of PLD have been characterized in mammalian cells, and PLD1 exists as a number of splice variants, although the significance of these is not known. PLD1 is regulated by a number of signaling molecules, including PKC and small GTPase (55)(56)(57). PLD 2 , in contrast, does not appear to be regulated by Rho, and its activation by ARF appears to be small in comparison to the effects of ARF on PLD1 activity (58). The relative roles of PLD1 and PLD 2 in receptormediated signaling remain elusive. Thus, the question of which PLD is regulated by what extracellular signals is still of special interest. Here we show that PLD 2 isoenzyme is predominantly expressed in Raw264.7 cells and that treatment with LPS did not significantly alter the expression levels of PLD 2 . These data suggest that PLD 2 is the main target of LPS-dependent PLD activation. Furthermore, transfection of PLD 2 -overexpressing HEK293 cells with an iNOS or COX-2 promoter cDNA also resulted in the activation of both promoter activities versus those of the vector transfected cells. Consistent with the finding that Raw264.7 cells specifically expresses PLD 2 and not PLD1, promoter (iNOS or COX-2) transfection experiments suggested a specific role for PLD 2 but not PLD1 in the coupling of PLD 2 to the production of inflammatory mediators.
LPS-stimulated p70 S6K1 activity was inhibited in the presence of LY294002 or rapamycin, implicating Akt and mTOR as a regulator of p70 S6K1. Thus, Akt and mTOR may participate in signaling cascades leading to p70 S6K1, which has been previously reported in LPS-treated macrophages (59,60). Our data suggest that Akt is one of the inflammatory signaling intermediates, because LY294002 substantially reduced the synthesis of the three cytokines, and of NO and PGE 2 produced by LPS-treated macrophages. Rapamycin also significantly blocked production of these mediators. This observation implies that mTOR, in addition to Akt, contributes to LPS-stimulated mediator production. Because p70 S6K1 is a downstream effector of both Akt and mTOR, and LY294002 inhibited the activation of mTOR and p70 S6K1, it is possible that the sequential activation of Akt-mTOR-p70 S6K1 participates in pro-inflammatory mediator production following LPS stimulation.
In summary, our results demonstrate that PA induces acute inflammation both in vitro and in vivo. By regulating the activities of Akt and mTOR-p70 S6K1 in macrophages, PA enhances the production of the pro-inflammatory cytokines (TNF-␣, IL-1␤, and IL-6) and of the mediators (NO and PGE 2 ), and, by so doing, PA facilitates the initiation of the host in-flammatory response. Therefore, our findings identify a crucial role for PA in inflammation and provide a rationale for an anti-PA strategy for the treatment of people with acute inflammation.