Macrophage activation by lipopolysaccharide, interferon-γ and interleukin-4: effect of fatty acid metabolism

The aim of this study was to investigate the effects of interferon-γ and -β (IFN-γ, -β), interleukin-4 and -10 (IL-4, -10) and Hpopolysaccharide (LPS) on the metabolism and composition of phospholipid fatty acids in macrophages. Murine J774.2 macrophages were incubated with radiolabelled fatty acids and the appropriate stimulus and the incorporation and composition of the phospholipid classes was determined. IFN-γ and IL-4 specifically stimulated enhanced incorporation of [14C]-linoleic acid into the phosphatidytethanolamine fraction. IL-4 (in contrast to IFN-γ and LPS) reduced incorporation of [14C]- arachidonic acid into phosphatidylinositol. Incubation of J774.2 cells with linoleic acid significantly increased TNFα and nitric oxide production; arachidonic acid enhanced TNFα production but reduced nitric oxide production. It is concluded that IFN-γ, IL-4 and IL-10 may differentially regulate macrophage activation via effects on the metabolism of polyunsaturated fatty acids.


Introduction
Bacterial endotoxin (lipopolysaccharide, LPS) is an important inducer of the sepsis syndrome, a rapidly fatal illness which remains a major cause of morbidity and mortality in the modern medical centre. Some patients become hypersusceptible to the lethal effects of endotoxin, although the mechanism for this sensitivity remains poorly understood. Animal models of endotoxic shock have revealed that concomitant bacterial or mycobacterial infections greatly increase susceptibility to the lethal effects of endotoxin and that antibody to tumour necrosis factor-0t (TNF00 or interferonq, (IFN-T) are beneficial. 2, It is becoming increasingly clear that IFN-T is an important mediator of hypersensitivity to endotoxin. IFN-T has been shown to induce expression of LPS binding sites on the cell surface of lung macrophages which lack binding sites for LPS. Furthermore, IFN-T has been identified as the mediator of Propionibacterium acnes-induced LPS hypersensitivity in mice. Macrophages are primary target cells for endotoxin and by production of endogenous mediators such as superoxide anion, nitric oxide, cytokines (interleukin-1, -6, -8 (IL-1, -6, -8) and TNF) and lipid mediators (for example, prostaglandins and leukotrienes), contribute to the pathophysiology of endotoxic shock.
Infection of mice with BeG results in increases in the ratio of polyunsaturated:saturated fatty acids in vivo in peritoneal macrophages and it has recently been shown that IFN-T is able to increase macrophage phospholipid polyunsaturated in ( 1995 Rapid Communications of Oxford Ltd vitro. 6,7 It is believed that this could be a possible mechanism of endotoxin sensitivity and we have demonstrated that IFN-T and exogenous polyunsaturated fatty acids increase binding of LPS to mouse macrophages. Metabolic  btg/ml Escherichia coli Olll:B4 LPS, and murine recombinant IL-4 and IL-10 (50 U/ml) was carried out for 18 h at 37C, in a humidified incubator containing 5% CO,. At the end of the incubation period the cells were dislodged by gentle agitation, washed three times with DMEM and collected by centrifugation at 1 000 x g for 5 min and finally resuspended in DMEM Preparation of ammonium salts of the radiolabelled fatty acids: To aid solubility of the radiolabelled fatty acids, they were converted into their ammonium salts as follows. The fatty acids were supplied in either ethanol or toluene and as a first stage the solvent was evaporated under a stream of nitrogen. The ammonium salt of the fatty acid was then prepared by incubating the fatty acids in 0.2 ml of 2 M ammonia solution at 60-70C under nitrogen for 30 min. The solution was evaporated to dryness under a stream of nitrogen and the resulting ammonium salts were resuspended in a known volume of growth medium. Incorporation ofl4C-fatty acids: Cells were incubated in the presence of 0.2 t.tCi of the radiolabelled fatty acids (ammonium salt) with or without IL-4 and IL-l0 for 18 h at 37C in a humidified CO,. incubator. After the appropriate incubation time, the cells were centrifuged at 1 000 x g for 3 min, washed in 10 ml phosphate buffered saline and collected by centrifugation at I 000 x g for 3 min and finally resuspended in 1 ml deionized water and sonicated in an ultrasonic water bath at maximum power. Once complete lysis of the cells had been achieved (verified by light microscopy) the cells were processed for phospholipid extraction.
Extraction ofphospholipids: The method of Garbus et al. 16

Results
The effects of interferons or LPS on the uptake of radiolabelled exogenous fatty acids by macrophages have recently been reported, and in this investigation we examined the effects of IL-4 and IL-10, whose biological actions are reported to contrast with some of the activating properties of IFN-T on J774.2 cells.
Figure la shows that IL-4 induces a statistically significant increase in the incorporation of 4Clinoleic acid into the phosphatidylethanolamine (PE) (p < 0.01 compared with control cells) fraction of the macrophages. The phosphatidylcholine (PC) and phosphatidylinositol (PI) fractions showed reduced incorporation (p < 0.05 vs. control cells) of this fatty acid. Figure lb shows that IL-4 induces a small but statistically significant increase in the incorporation of 14C-arachidonic .acid into the PC and a reduced incorporation into the PI fraction of the J774.2 cell membrane phospholipids.
In contrast to IL-4, IL-10 did not affect the incorporation of linoleic acid into the PE fraction. However, there was a statistically significant reduction in the incorporation of this fatty acid into the PC fraction. Furthermore, IL-10 induced an increase in the uptake of 4C-arachidonic acid into the PE fraction and a decrease in the PI (data not shown).   IFN-Y 173% (p < 0.01 vs. control cells) and 212% (p < 0.01 vs. control cells) respectively. IFN-, induced an increase of 101% (p < 0.01 vs. control cells). In contrast, IFN-3 did not increase the content of linoleic acid.
As it is believed that the polyunsaturated components of the plasma membrane play an important role in predisposing to endotoxin sensitivity, 17 stearic acid composition was used as a control. IFN-T, IL-4, LPS and IFN-]3-pretreatment of the cells all result in a reduction in the percentage composition of stearic acid, in comparison with control cells (Fig. 2c). The greatest reduction in the content of this saturated fatty acid is seen in IFNq,, LPS and IL-4-pretreated cells, which showed decreases of 68%, 48% and 38% respectively (p < 0.001 vs. control cells for all three). IFN-[3 caused a 24% reduction in the composition of stearic acid in the PE fraction of these cells (p < 0.05). Thus, these effects are in marked contrast to the differential effects of mediators on linoleate levels in J774.2 cells. Figure 3 shows the results of the ELISA assay for TNF0t release in control cells and in cells preincubated with arachidonic and linoleic acids for 18 h and subsequently incubated with LPS for 4 h. Control cells produced very little TNF0t but when they were incubated with LPS for 4 h there was a great increase in the release of this cytokine (p < 0.00001 vs. control cells). Cells which had been preincubated with linoleic acid and then with LPS for 4 h showed increased TNF0t release (p < 0.05 vs. control cells which were subsequently incubated with LPS). Although arachidonic acid pretreatment resulted in a slight increase in TNF0t release, the increase was not statistically significant. Cells preincubated with these fatty acids without subse- quent incubation with LPS did not show increased production of TNF. Figure 4 shows the release of nitric oxide (measured as nitrite) from control cells and from cells pretreated with IFN-y, IFN-[ and IL-4, linoleic and arachidonic acids with or without LPS. IFN-y and LPS on their own induced increases in the production of nitric oxide (p < 0.00001 vs. control cells for both) but this increase was much more marked in cells preincubated with IFN-y in the presence of LPS. IFN[ did not have any effect on the production of nitric oxide. IL-4, on the other hand, induced a statistically significant inhibition in the LPS-mediated production of nitric oxide (p < 0.05 vs. LPS treated cells).
When cells were incubated with arachidonic acid and LPS together there was a slight but statistically significant decrease in nitric oxide production (p < 0.05 vs. LPS treated cells). Linoleic acid with LPS induced a statistically significant increase in nitric oxide production (p < 0..01 vs. LPS treated cells).
When the cells were preincubated with linoleic acid or arachidonic acid alone there was no change in nitric oxide levels (data not shown).

Discussion
It has previously been reported that IFN-y may exert at least some of its effects on macrophage cells by increasing the polyunsaturation of the fatty acyl side chains of membrane phospholipids. These may be important in the subsequent interaction of macrophages with endotoxin. This study has shown that LPS increases the content of the polyunsaturated fatty acids, arachidonate and linoleate and decreases the content of the saturated fatty acid, stearate, in the membrane PE fraction. Interestingly, LPS-pretreated cells showed greater increases in the content of unsaturated fatty acids and greater decreases in the content of the saturated fatty acid, in comparison with IFN-y pretreated cells.
In this study it was found that IL-4 acts similarly to IFN-y and LPS in increasing the linoleic acid and decreasing the stearic acid content of J774.2 cells.
IFN-[ caused a decrease in the content of arachidonic acid (Fig. 2a) and a slight increase in the linoleic acid (Fig. 2b)  It has recently been reported that IFN-[ induces a significant decrease in membrane lipid bilayer fluidity of J774.2 cells and the decreased fluidity is in keeping with the slightly decreased incorporation of 14C-linoleic acid in these macrophage membrane phospholipids. 18 IFN-[ has also been reported to increase the saturated fatty acid content of mouse sarcoma S-180 cells. 19 These are in agreement with the effects of IFN-[ that were found for murine macrophages.
It has been found that IL-4 acted similarly with LPS and IFN-y in increasing the incorporation of 14Clinoleic acid into the PE fraction of J774.2 cell membranes. However, IFN-y induced increases in incorporation of this fatty acid into all fractions of membrane phospholipids, whereas IL-4 induced decreases in the PC and PI fractions. The effects of IL-l0 were not examined since this cytokine did not stimulate uptake of 14C-linoleic acid.
The fact that IFN-y and LPS both increase the content of linoleic and arachidonic acids and IFNdoes not, and that IFN-y priming can lead to enhanced LPS binding which can be mimicked by linoleic and arachidonic acids but not by interferon-[ also supports our hypothesis that a possible mechanism for IFN-y-induced LPS hypersensitivity involves fatty acid changes in the membrane phospholipids. The consequence of these lipid changes are that IFN-y increases membrane fluidity and in so doing can render greater molecular mobility of membrane receptors. This results in increased LPS binding to IFN-y-pretreated macrophages and also greater expression of CD-14 (unpublished results) recently reported to be a receptor for complexes of LPS and LPS-binding protein.
Increasing evidence suggests that macrophage activation and cytokine production can be regulated by polyunsaturated fatty acids (PUFA). Thus, n-3 PUFA have been shown to suppress the ability of macrophages to produce IL-1, IL-6 and TNF, but the n-6 PUFA, typified by linoleic acid, can enhance their production. ,. The oxidative metabolites of PUFA (eicosanoids) have been shown to regulate macrophage inflammatory reactions including cytokine synthesis. Among the arachidonic acid metabolites, prostaglandin E2 (PGE2) has been found to have important feedback actions on macrophages. In many cases PGE provides a negative feedback signal for the production of cytokines such as TNF(X 21 and IL-1. 22 Interestingly, it has been suggested that IFN-T can down-regulate PGE production, perhaps via a decrease in cyclooxygenase activity, and this enhances the production of the pro-inflammatory cytokines. 23 Several recent studies provide evidence that lipoxygenases are involved in the activation of mononuclear phagocytes. 24 It is well recognized that TNF mediates many of the lethal effects of endotoxin and recently 13-hydroxylinoleic acid has been reported to be of functional importance in TNF formation by macrophages treated with LPS. 25 Pretreatment of macrophages with IFN-T and IL-4 results in LPSinduced TNFot release and it is believed that this could be as a result of increases in the linoleic acid content of these macrophages.
From our results it can be seen that IL-4 induces increases in the linoleic acid but not the arachidonic acid content in J774.2 macrophage membranes. IL-4 decreases LPS-induced nitric oxide production but increases LPS-induced TNFot production, in agreement with recent findings. 26 Therefore, induction of TNFz in this case is not due to increases in endogenous arachidonic acid and suggests that linoleic acid may play a significant role in TNFz production by macrophages. Induction of nitric oxide release, on the other hand, may require the presence of both endogenous arachidonic and linoleic acids.
Linoleic acid, therefore, may be an important mediator in LPS-induced responses in macrophages, and IFN-T and IL-4, by increasing the linoleate content in phospholipids, could prime these cells for a heightened response to LPS.