Lipid oxidation is commonly seen in the innate immune response, in which reactive oxygen intermediates are generated to kill pathogenic microorganisms. Although oxidation products of phospholipids have generally been regarded to play a role in a number of chronic inflammatory processes, several studies have shown that oxidized phospholipids inhibit the LPS-induced acute proinflammatory response in cultured macrophages and endothelial cells. We report in this study that oxidized 1-palmitoyl-2-arachidonoyl-sn-glycero-3-phosphorylcholine (PAPC), but not nonoxidized PAPC, significantly inhibits the LPS-induced TNF-α response in intact mice. Oxidized PAPC also inhibits the 2′-deoxyribo(cytidine-phosphate-guanosine) (CpG) DNA-induced TNF-α response in cultured macrophages and intact mice. To elucidate the mechanisms of action, we show that oxidized PAPC, but not nonoxidized PAPC, inhibits the LPS- and CpG-induced activation of p38 MAPK and the NF-κB cascade. These results suggest a role for oxidized lipids as a negative regulator in controlling the magnitude of the innate immune response. Further studies on the mechanisms of action may lead to development of a new type of anti-inflammatory drug for treatment of acute inflammatory diseases such as sepsis.
- reactive oxygen intermediates
- innate immune response
innate immunity is an evolutionarily ancient system that provides multicellular organisms with immediately available defense mechanisms against a wide variety of pathogens without the requirement of prior exposure (9). Innate immunity recognition of invading pathogens is mediated by a set of germline-encoded receptors that have evolved to recognize conserved molecular patterns shared by large groups of organisms (27). Recently, the Toll-like receptor (TLR) family has been identified to comprise the major components that are responsible for pathogen recognition in the innate immune response (4, 34). These include TLR2, TLR3, TLR4, TLR5, TLR6, TLR7, and TLR9 (24, 30), among which TLR4 and TLR9 have been extensively characterized. TLR4 is responsible for the recognition of LPS, whereas TLR9 recognizes bacterial and synthetic DNA containing unmethylated 2′-deoxyribo-(cytidine-phosphate-guanosine) (CpG) dinucleotides (4, 34).
The signaling pathway via the TLR family is highly homologous to that of the IL-1 receptor (IL-1R) family (1, 21). TLR and IL-1R interact with an adaptor protein, MyD88, in their Toll/IL-1R domains. When stimulated, MyD88 recruits IL-1R-associated kinase to the receptor. IL-1R-associated kinase is activated by phosphorylation and then associates with TRAF6, leading to the activation of downstream kinases, e.g., the stress kinases c-Jun NH2-terminal kinase, p38, and the IκB kinase complex (39, 50).
Activation of the innate immune response triggers a number of events, including upregulation of adhesion molecules on endothelial cells, recruitment of leukocytes to the site of infection, induction of proinflammatory cytokines, and generation of reactive oxygen intermediates (ROIs) (4). Although ROIs play an important role in killing the pathogenic micro-organisms (18), products of the associated lipid oxidation have been shown to stimulate adhesion of monocytes to endothelial cells and induce expression of inflammatory genes in endothelial cells, suggesting a role for the oxidized lipids in chronic inflammation (32). Several studies have shown that oxidized phospholipids suppress the LPS-induced NF-κB-mediated upregulation of several early genes in macrophages (11, 14, 22, 38, 44). A recent study by Bochkov et al. (7) showed that oxidized 1-palmitoyl-2-arachidonyl-sn-glycero-3-phosphorylcholine (PAPC), but not nonoxidized PAPC, inhibits the LPS-induced upregulation of inflammatory genes in human umbilical vein endothelial cells. We report in this study that oxidized PAPC, but not nonoxidized PAPC, significantly inhibits LPS-induced TNF-α production in intact mice. We also show for the first time that oxidized PAPC inhibits CpG DNA-induced TNF-α production in cultured macrophages and intact mice. These results suggest a novel role for oxidized lipids as a negative regulator in controlling the magnitude of the innate immune response. Further studies on the mechanisms of action may lead to development of a new type of anti-inflammatory drug for treatment of acute inflammatory diseases such as sepsis.
MATERIALS AND METHODS
Lipids and LPS. PAPC and LPS from Escherichia coli serotype 055:B5 were purchased from Sigma-Aldrich (St. Louis, MO). Oxidized PAPC was obtained by air oxidation of dry PAPC, as described previously (23, 51), and stored in chloroform at -70°C. Immediately before the experiment, lipids were dried in glass tubes under a stream of N2 and solubilized in culture medium by vigorous vortexing for 30 s and then sonication for 30 s.
Cells. The mouse macrophage cell line RAW 264.7 (American Type Culture Collection) was cultured in DMEM supplemented with 10% (vol/vol) FBS, streptomycin (100 μg/ml), and penicillin (100 U/ml). All cultures were incubated at 37°C in a humidified atmosphere with 5% CO2.
Preparation of primary culture mouse macrophages. Female CD-1 mice were purchased from Harlan Sprague Dawley (Indianapolis, IN) at 4 wk of age and housed in accordance with institutional guidelines. All experiments were conducted under guidelines approved by the Animal Ethics Committee at the University of Pittsburgh. Mouse peritoneal macrophages were prepared according to a standard protocol (42). Briefly, three mice were injected intraperitoneally with 1.5 ml of sterile 3% Brewer thioglycollate medium (Sigma-Aldrich) per mouse. At 5 days after injection, peritoneal lavages were harvested by washing the peritoneal cavity with sterile PBS. The pooled cells were centrifuged at 400 g for 10 min, resuspended in DMEM with 10% FCS, and distributed into the wells of a 24-well plate. After 1 h of incubation at 37°C, nonadherent cells were washed away. The macrophages were allowed to rest overnight before LPS stimulation (see below).
TNF-α ELISA. Unless otherwise specified, RAW 264.7 cells were plated at a density of 5 × 105 cells/well in 24-well cell culture plates. Oxidized PAPC or nonoxidized PAPC in complete medium was added (0.5 ml/well), and cells were incubated for 20 min. Cells were then challenged with LPS, and the supernatants were collected 4 h later. In the CpG DNA experiment, after preincubation with lipids for 20 min, cells were treated with 1668 oligodeoxynucleotide (ODN) at 1 μg/ml for 4 h before the collection of the supernatants. The supernatants were kept frozen until use. The levels of mouse TNF-α were determined with a specific ELISA kit for mouse TNF-α (R & D, Minneapolis, MN).
Preparation of liposomes. Liposomes containing oxidized PAPC or nonoxidized PAPC in a 99:1 molar ratio with rhodamine B-1,2-dihexadecanoyl-sn-glycero-3-phosphoethanolamine were prepared as follows. Lipid mixtures in chloroform were dried as a thin layer in a 25-ml round-bottomed flask that was further dried under vacuum for 15 min. The lipid film was hydrated in medium to give a final concentration of 2.5 mg lipid/ml. Small unilamellar lipid vesicles were then prepared by extrusion through 0.2-μm polycarbonate membranes.
Cellular uptake of liposomes. RAW 264.7 cells were seeded into 24-well plates at a density of 1 × 105 cells/well and allowed to grow overnight. Liposomes were added to the cells at 50 μg lipid/ml. At different intervals, the cells were washed three times with PBS, and cellular uptake of rhodamine-labeled liposomes was examined under a Nikon Eclipse TE 300 fluorescence microscope with a green filter at ×400 magnification.
Quantification of NF-κB-dependent transcription. RAW 264.7 cells were seeded into 48-well plates at a density of 5 × 104 cells/well and allowed to grow overnight. The cells were cotransfected with 0.35 μg of pNF-κB-luciferase (Stratagene) and 0.25 μg of pCMVβ with N-[1-(2,3-dioleoyl)propyl]-N,N,N-trimethylammonium methylsulfate used as a transfection reagent (positive-to-negative charge ratio = 2:1). After 5 h, the transfection medium was replaced with complete medium. After another 18 h of incubation, the cells were stimulated with LPS (100 ng/ml) or CpG ODN (4 μg/ml) in the presence or absence of oxidized PAPC or nonoxidized PAPC, and the activity of expressed luciferase and β-galactosidase was examined 6 h later. Data are presented as luciferase activity normalized by β-galactosidase.
Western blotting. RAW cells were stimulated with LPS or CpG ODN in the presence or absence of oxidized PAPC. After 20 min of incubation with LPS or CpG, the cells were scraped and analyzed by Western blotting as described elsewhere (8). Antibodies to IκB-α, phosphorylated IκB-α, p38, and phosphorylated p38 MAPK were purchased from Cell Signaling Technology (Beverly, MA).
Animals and in vivo experiments. Female C57BL/6 mice (25 g body wt) were purchased from Harlan Sprague Dawley and housed in accordance with institutional guidelines. All experiments were conducted under guidelines approved by the Animal Ethics Committee at the University of Pittsburgh. LPS (200 μg/25 g mouse), alone or mixed with various amounts of lipids in 200 μl of saline, was injected into the mice via tail vein (5 mice/group). The dose of LPS was established according to the literature (29). After 2 h, blood samples were collected and examined for TNF-α levels using a specific ELISA kit as described above. For the CpG in vivo experiment, 1668 ODN (50 μg/25 g mouse) with or without nonoxidized PAPC (1,000 μg/25 g mouse) or oxidized PAPC (1,000 μg/25 g mouse) in 200 μl of saline was injected into the mice via the tail vein (3 mice/group). The dose of CpG was established according to our pilot experiment. After 2 h, blood samples were collected and examined for TNF-α levels as described above.
In a separate experiment, mice were anesthetized by injection of tribromoethanol (3 mg/kg ip). Groups of six mice were instilled intratracheally with 5 μg of LPS, 5 μg of LPS + 100 μg of PAPC, or 5 μg of LPS + different amounts of oxidized PAPC (25, 50, or 100 μg) in 50 μl of saline, whereas control mice received 50 μl of saline. After 2 h, mice were anesthetized and the trachea was cannulated. Bronchoalveolar lavage (BAL) was performed by flushing the airway three times with PBS (1 ml) using a polypropylene 20-gauge intravenous catheter. BAL fluid samples were centrifuged at 1,100 rpm for 5 min at 4°C, and the supernatants were examined for TNF-α levels.
Statistical analysis. Each experiment was performed at least three times, and statistical analysis was performed by one-way ANOVA with post hoc test using the Prism software program (GraphPad Software, San Diego, CA). Data were considered significant at P < 0.05 or P < 0.01 and very significant at P < 0.001. Otherwise, representative data are shown.
Oxidized PAPC inhibits LPS-induced TNF-α production in cultured murine macrophages. The cultured RAW 264.7 macrophages produced a low level of TNF-α when they were cultured in vitro within a short period of time (∼4 h), and cells treated with nonoxidized PAPC or oxidized PAPC without LPS stimulation only showed a baseline level of TNF-α secretion. As expected, cells produced significantly greater amounts of TNF-α after exposure to LPS (100 ng/ml) for 4 h (Fig. 1A). The LPS-triggered TNF-α production, however, was significantly inhibited by pretreatment with oxidized PAPC, but not nonoxidized PAPC (Fig. 1A). The inhibitory effect of oxidized PAPC was concentration dependent, with maximal inhibition at an oxidized PAPC concentration of 100 μg/ml (Fig. 1A). A similar observation was made in primary culture macrophages that were isolated from the mouse peritoneal cavity (Fig. 1B). Oxidized PAPC at 100 μg/ml effectively inhibited LPS-induced TNF-α production in RAW 264.7 macrophages over a wide rage of LPS concentrations (25–500 ng/ml; Fig. 1C). Colorimetric assay with 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide showed that viability of cells in the presence of oxidized PAPC with or without LPS was comparable to that of cells treated with PAPC or LPS alone (data not shown), suggesting that the inhibitory effect of oxidized PAPC on the TNF-α response was not due to lipid-associated cytotoxicity. Lack of toxicity was further demonstrated in a later study in which oxidized PAPC was shown to have a minimal effect on expression of transfected reporter genes in RAW 264.7 macrophages (see Fig. 5C).
To examine whether free or cell-associated oxidized PAPC was involved in inhibiting LPS-triggered TNF-α production, cells were treated with oxidized PAPC for different amounts of time, and free oxidized PAPC was removed by three washes with PBS. Cells were then exposed to LPS, and the TNF-α response was examined 4 h later. To confirm the association of oxidized PAPC with cells, rhodamine-1,2-dihexadecanoyl-sn-glycero-3-phosphoethanolamine was incorporated into oxidized PAPC liposomes, and the interaction of rhodamine-labeled liposomes with cells was examined under a fluorescence microscope. Cellular association/uptake of oxidized PAPC liposomes was observed as early as 5 min after incubation with RAW 264.7 cells (Fig. 2A). The cell-associated fluorescence intensity increased significantly with time over the 2-h period. Despite a low level of association/uptake, incubation of oxidized PAPC with cells for 5 min led to a significant inhibition of TNF-α production after a subsequent challenge with LPS (Fig. 2B). Prolonging the preincubation time from 5 to 30 min appeared to be associated with an increase in the inhibitory activity. Removal of free oxidized PAPC after ≥30 min of preincubation had a minimal effect on its inhibitory activity, suggesting that it is the cell-associated oxidized PAPC that is mainly responsible for blocking the LPS-triggered TNF-α response.
Oxidized PAPC inhibits LPS-induced TNF-α production in mice. The efficient inhibition of LPS-induced TNF-α production by oxidized PAPC in cultured macrophages prompted us to examine whether oxidized PAPC can inhibit the LPS-induced TNF-α response in intact mice. Thus groups of five C57BL/6 mice received intravenous injections of LPS alone or together with oxidized PAPC or nonoxidized PAPC, and serum levels of TNF-α were examined 2 h later. Coadministration of oxidized PAPC inhibited the LPS-induced TNF-α response in a dose-dependent manner (Fig. 3A). In contrast, nonoxidized PAPC essentially had no effect on LPS-triggered TNF-α production at the highest dose examined (1 mg/mouse; Fig. 3A). A similar observation was made in CD-1 mice (data not shown). The efficiency with which oxidized PAPC inhibited the LPS-induced TNF-α response was comparable in C57BL/6 and CD-1 mice, although C57BL/6 mice respond to LPS more dramatically than CD-1 mice with respect to serum levels of TNF-α (data not shown). Oxidized PAPC, at the highest dose used (1 mg/mouse), had little toxic effect on animals, as manifested by minimal changes in serological and hematological profiles 24 h after intravenous injection of the lipid (data not shown).
Figure 3B shows the TNF-α levels in BAL fluids 2 h after intratracheal instillation of LPS alone, LPS + nonoxidized PAPC, or LPS + oxidized PAPC. Oxidized PAPC inhibited the LPS-mediated TNF-α response in a dose-dependent manner. In contrast, nonoxidized PAPC did not inhibit the TNF-α response at the highest dose examined (100 μg/mouse).
Oxidized PAPC inhibits the CpG DNA-induced TNF-α response. After demonstrating inhibition of an LPS-induced TNF-α response in cultured macrophages and intact mice, we examined whether oxidized PAPC can inhibit the CpG DNA-mediated immune response. In contrast to LPS, which binds to TLR4, CpG DNA is recognized by TLR9, although both receptors share common downstream signaling pathways (50). 1668 CpG ODN was used in this study because it is known to trigger a strong proinflammatory cytokine response, including TNF-α induction in cultured macrophages and intact mice (26, 28, 31, 46, 48). Oxidized PAPC inhibited the CpG-induced TNF-α response in cultured macrophages in a dose-dependent manner (Fig. 4A). In contrast, nonoxidized PAPC essentially had no effect (Fig. 4A). Oxidized PAPC, but not nonoxidized PAPC, also significantly inhibited the 1668 ODN-induced TNF-α response in intact mice. Lipids alone, oxidized or nonoxidized, were essentially inert with respect to TNF-α induction (Fig. 4B).
Oxidized PAPC blocks LPS- and CpG-induced phosphorylation of IκB and activation of the p38 MAPK pathway. It has been well established that induction of the innate immune response via TLRs involves activation of the NF-κB cascade and p38 MAPK (17, 19, 25, 35, 37, 43, 45, 52); thus we examined whether oxidized PAPC or nonoxidized PAPC influenced the signaling events of NF-κB and p38 MAPK. LPS or CpG DNA challenge significantly stimulated phosphorylation of IκB-α (Fig. 5, A and B), which was effectively blocked by oxidized PAPC, but not by nonoxidized PAPC. Phosphorylation of IκB-α in LPS-treated cells was associated with a concurrent decrease in the total level of IκB-α at 20 min after treatment (Fig. 5A). Interestingly, despite a significant increase in the level of phosphorylated IκB-α, there was no concurrent decrease in the total level of IκB-α in the CpG ODN-treated cells. We also failed to observe a decrease in the total level of IκB-α at a later time (1 h after treatment; data not shown). Figure 5C shows the effect of oxidized PAPC on LPS- or CpG DNA-induced activation of the NF-κB cascade in a 5× NF-κB-luciferase reporter system. LPS or CpG DNA treatment significantly activated luciferase expression in RAW cells that were pretransfected with a 5× NF-κB-luciferase reporter construct. LPS- or CpG DNA-activated luciferase expression was significantly inhibited by oxidized PAPC, but not by nonoxidized PAPC. Oxidized PAPC itself had no effect on luciferase expression (Fig. 5B). LPS or CpG DNA treatment also significantly activated phosphorylation of p38 MAPK in RAW macrophages (Fig. 5, A and B). Again, activation of p38 MAPK by LPS or CpG DNA was significantly inhibited by oxidized PAPC, but not by nonoxidized PAPC. Taken together, oxidized PAPC inhibits the LPS- and CpG-induced activation of p38 MAPK and the NF-κB cascade, providing a molecular basis for its inhibition of the TNF-α response.
We have shown in this study that oxidized PAPC, but not nonoxidized PAPC, significantly inhibits LPS- and CpG DNA-induced TNF-α production in RAW 264.7 macrophages as well as in primary culture macrophages isolated from the mouse peritoneal cavity. Coadministration of oxidized PAPC also significantly inhibits the TNF-α response triggered by intravenous injection of LPS or CpG ODN. Oxidized PAPC is also highly efficient in suppressing the TNF-α response induced by intratracheal instillation of LPS.
Results from Fig. 5 suggest that oxidized PAPC, but not nonoxidized PAPC, inhibits LPS- or CpG DNA-induced activation of p38 MAPK and the NF-κB cascade. However, the exact mechanisms by which oxidized PAPC inhibits the LPS- and CpG DNA-induced TNF-α response are not clearly understood. The study by Bochkov et al. (7) suggests that oxidized PAPC inhibits LPS-induced upregulation of a number of adhesion molecules in endothelial cells by blocking the interaction of LPS with TLR4. This finding was based on their observations from an in vitro binding assay that oxidized PAPC, but not nonoxidized PAPC, significantly inhibits the interaction of LPS with LPS-binding protein or CD14, two cofactors that are critically involved in the recognition of LPS by TLR4 (20). The inhibition of LPS-induced TNF-α production in RAW macrophages may be through a similar mechanism. The fact that cell-associated oxidized PAPC is effective in inhibiting the TNF-α response (Fig. 2) suggests that oxidized PAPC needs to be incorporated into the cell membrane to effectively inhibit the binding of LPS to TLR4. Alternatively, incorporation of oxidized PAPC into the cell membrane may inhibit the interaction of TLRs with downstream signaling molecules. This may explain the observation that oxidized PAPC is also effective in inhibiting the CpG DNA-mediated TNF-α response in macrophages and intact mice (Fig. 4). In contrast to LPS, which is recognized by TLR4, CpG DNA is recognized by a distinctly different receptor, i.e., TLR9. Additional studies are required to better understand the mechanisms of actions by oxidized PAPC. Particularly, it needs to be examined whether incorporation of oxidized PAPC into the cell membrane may affect other signal transduction pathways that could indirectly interfere with TLR signaling.
The oxidized PAPC used in this study is a heterogeneous population containing a number of lipids with different structures, one of them being arachidonic acid (AA). A number of studies have shown that AA and its metabolites have antiproinflammatory activity, including inhibition of LPS-induced TNF-α production (2, 15, 49). Inhibition of inflammation by AA and its metabolites appears to involve the activation of peroxisome proliferator-activated-γ. Our preliminary studies show, however, that oxidized PAPC is much more potent than AA in inhibiting the LPS-induced TNF-α response when compared on an equimolar basis (data not shown), suggesting that AA is not the major component responsible for the biological activity of oxidized PAPC observed in our studies. In the study by Bochkov and colleagues (7), 1-palmitoyl-2-oxovalerolyl-sn-glycero-3-phosphorylcholine was identified to be a major component in oxidized PAPC that is responsible for protecting endothelial cells from LPS-mediated damage. It remains to be examined whether 1-palmitoyl-2-oxovalerolyl-sn-glycero-3-phosphorylcholine also plays a major role in inhibiting LPS- and CpG DNA-induced TNF-α production in macrophages.
Lipid oxidation, a phenomenon that accompanies the innate immune response, results in generation of ROIs to kill pathogenic microorganisms. Although the inflammatory response is critical to control the growth of pathogenic microorganisms, an excess immune response is harmful to the host and can even be fatal (12). The fact that the products of oxidized PAPC inhibit the LPS- and CpG DNA-induced TNF-α response supported a role for oxidized phospholipids as a negative regulator in controlling the magnitude of the innate immune response. A similar observation has been made in a recent study with oxidized low-density lipoprotein (LDL) (14, 16). Oxidized LDL is generally regarded to be highly proinflammatory and to play an important role in a number of chronic inflammatory processes such as atherosclerosis (5, 6, 40, 47, 53, 54). Oxidized LDL evokes a strong oxidative burst in endothelial cells and/or monocyte/macrophage cell lines. Preincubation of monocytes/macrophages with oxidized LDL or major lipid components of oxidized LDL, such as hydroxyoctadecadienoic acid, for 16 h significantly attenuates the oxidative burst. Such desensitization might prove to be beneficial in attenuating the development of atherosclerosis (16).
In summary, we have shown in this study that oxidized PAPC are highly efficient in inhibiting LPS- and CpG DNA-induced TNF-α production in cultured RAW 264.7 macrophages and intact mice. Further studies on the mechanisms of action may lead to the development of new types of anti-inflammatory drugs. Also, along with development of lipid vectors for targeted delivery of anti-inflammatory drugs, including gene-based drugs (13, 33), incorporation of oxidized PAPC into these vectors may have a synergistic effect in anti-inflammation.
This work was supported by National Institutes of Health Grants HL-63080 (to S. Li) and HL-32154 and GM-53789 (to B. Pitt) and American Heart Association Grant 026540U (to S. Li).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2004 the American Physiological Society