Arachidonate-containing oxidized phospholipids are acutely generated by 12-LOX (12-lipoxygenase) in agonist-activated platelets. In the present study, formation of structurally related lipids by oxidation of DHA (docosahexaenoic acid)-containing phospholipids is demonstrated using lipidomic approaches. Precursor scanning reverse-phase LC (liquid chromatography)–MS/MS (tandem MS) identified a new family of lipids that comprise phospholipid-esterified HDOHE (hydroxydocosahexaenoic acid). Two diacyl and two plasmalogen PEs (phosphatidylethanolamines) containing predominantly the 14-HDOHE positional isomer (18:0p/14-HDOHE-PE, 18:0a/14-HDOHE-PE, 16:0a/14-HDOHE-PE and 16:0p/14-HDOHE-PE) were structurally characterized using MS/MS and by comparison with biogenic standards. An involvement of 12-LOX was indicated as purified recombinant human 12-LOX also generated the 14-HDOHE isomer from DHA. Pharmacological studies using inhibitors and recombinant platelet 12-LOX indicate that they form via esterification of newly formed non-esterified HDOHE. HDOHE-PEs formed at significant rates (2–4 ng/4×10 7 cells) within 2–180 min of thrombin stimulation, and their formation was blocked by calcium chelation. In summary, a new family of oxidized phospholipid was identified in thrombin-activated human platelets.

The oxidative modification of low-density lipoprotein (LDL) has been implicated as a pro-atherogenic process in the pathogenesis of atherosclerosis. Macrophages rapidly take up oxidized LDL via scavenger-receptor-mediated pathways and thereby develop into lipid-laden foam cells. The uptake mechanism has been studied extensively and several types of scavenger receptors have been identified. In contrast, the intracellular fate of oxidized LDL lipids is less well investigated. We studied the degradation of specifically oxidized cholesteryl esters by murine macrophages using an HPLC-based assay, and found that oxidized substrates are hydrolysed preferentially from a 1:1 molar mixture of oxidized and non-oxidized cholesteryl esters. This effect was observed at both neutral and acidic pH. Similar results were obtained with lysates of human monocytes and with pure recombinant human hormone-sensitive lipase. These data suggest that the intracellular oxidation of cholesteryl esters may facilitate intracellular cholesteryl ester hydrolysis, and thus may represent an anti-atherogenic process.

During oxygenation by 15-lipoxygenases, polyenoic fatty acids are bound at the active site in such a way that the ω-terminus of the fatty acids penetrates into the substrate binding pocket. In contrast, for arachidonic acid 5-lipoxygenation, an inverse head to tail orientation has been suggested. However, an inverse orientation may be hindered by the large energy barrier associated with burying the charged carboxylate group in the hydrophobic environment of the substrate binding cleft. We studied the oxygenation kinetics of ω-modified fatty acids by 15-lipoxygenases and found that ω-hydroxylation strongly impaired substrate affinity (higher K m ), but only moderately altered V max . In contrast, ω-carboxylation completely prevented the lipoxygenase reaction; however, methylation of the additional carboxylate group restored the activity. Arg 403 of the human 15-lipoxygenase has been implicated in fatty acid binding by forming a salt bridge with the carboxylate group, and thus mutation of this amino acid to an uncharged residue was supposed to favour an inverse substrate orientation. The prepared Arg 403 → Leu mutant of the rabbit 15-lipoxygenase was found to be a less effective catalyst of linoleic acid oxygenation. However, the oxygenation rate of ω-hydroxyarachidonic acid was similar when the wild-type and mutant enzyme were compared, and the patterns of oxygenation products were identical for both enzyme species. These data suggest that introduction of a polar, or even charged residue, at the ω-terminus of substrate fatty acids in connection with mutation of Arg 403 may not alter substrate alignment at the active site of 15-lipoxygenases.

We investigated the geometry of the iron ligand sphere of the native rabbit 15-lipoxygenase (15-LOX) by X-ray absorption spectroscopy using synchrotron radiation. The soybean LOX-1 was used as a reference compound because its iron ligand sphere is well characterized. For structural information the X-ray absorption spectra were evaluated using the Excurve Program (CCLRC Daresbury Laboratory, Warrington, U.K.). From the positions of the absorption edges and from the intensities of the 1s–3d pre-edge transition peaks a six-coordinate ferrous iron was concluded for the rabbit 15-LOX. Evaluation of the extended region of the absorption spectra suggested six nitrogen and/or oxygen atoms as direct iron ligands, and the following binding distances were determined (means±S.D.; estimated accuracy is ±0.001 nm for bond distances, on the basis of more than 22 X-ray absorption spectra): 0.213±0.001 nm, 0.213±0.001 nm, 0.236±0.001 nm, 0.293±0.001 nm, 0.189±0.001 nm and 0.242±0.001. Lyophilization of the LOX altered the binding distances but did not destroy the octahedral iron ligand sphere. For construction of a structural model of the iron ligand sphere the binding distances extracted from the X-ray spectra were assigned to specific amino acids (His-360, -365, -540, -544 and the C-terminal Ile-662) by molecular modelling using the crystal coordinates of the soybean LOX-1 and of a rabbit 15-LOX–inhibitor complex.

The purified 15-lipoxygenase from rabbit reticulocytes is capable of oxidizing NADH in the presence of linoleic acid and oxygen. This co-oxidation proceeds at a rate that amounts to approx. 7% of linoleic acid oxygenation rates. Although NADH inhibits the lipoxygenase reaction with linoleic acid as substrate (46% inhibition at 0.2 mM NADH), the reaction specificity of the enzyme was not altered since (13 S )-hydroperoxy-(9 Z ,11 E )-octadecadienoic acid was identified as the major reaction product. NADH oxidation was inhibited by NAD + (uncompetitive with respect to linoleate and mixed/competitive with respect to NADH), and NADPH or NMNH could substitute for NADH with slightly different apparent K m values. NADH oxidation was enhanced at lower oxygen tension, but was completely prevented under anaerobic conditions. Computer-assisted modelling of 15-lipoxygenase/NADH interaction and sequence alignments of mammalian lipoxygenases with NADH-dependent enzymes suggested that there is no specific binding of the coenzyme at the putative fatty acid-binding site of lipoxygenases. These results suggest that NAD(P)H might be oxidized by a radical intermediate formed during the dioxygenase cycle of the lipoxygenase reaction but that NADH oxidation might not proceed at the active site of the enzyme. The mechanism and possible biological consequences of 15-lipoxygenase-catalysed NAD(P)H oxidation are discussed.

We have studied the expression of the 15-lipoxygenase gene in various permanent mammalian cell lines in response to interleukins-4 and -13, and found that none of the cell lines tested expressed 5-, 12- or 15-lipoxygenase when cultured under standard conditions. However, when the lung carcinoma cell line A549 was maintained in the presence of either interleukin for 24 h or more, we observed a major induction of 15-lipoxygenase, as indicated by quantification of 15-lipoxygenase mRNA, by immunohistochemistry, by immunoblot analysis and by enzyme activity assays. This effect was 15-lipoxygenase-specific, since expression of 5- and 12-lipoxygenases remained undetectable. The time course of interleukin-4 treatment indicated maximal accumulation of both 15-lipoxygenase mRNA and functional protein after 48 h. Binding studies revealed that A549 cells express about 2100 high-affinity interleukin-4 binding sites per cell. The interleukin-4 mutant Y124D, which is capable of binding to the interleukin-4 receptor but is unable to trigger receptor activation, counteracted the effect of the wild-type cytokine. Other cell lines, including several epithelial cells and various monocytic cell lines expressing comparable numbers of interleukin-4 receptors, did not express 15-lipoxygenase when stimulated with interleukin-4. These data indicate that A549 cells selectively express 15-lipoxygenase when stimulated with interleukins-4 and -13. The activation of the interleukin-4/13 receptor(s) appears to be mandatory, but not sufficient, for 15-lipoxygenase gene expression.

(15 S )-Hydroxy-(5 Z,8Z,11Z,13E) -eicosatetraenoic acid (15-HETE) suppresses in ionophore-A23187-stimulated human polymorphonuclear leucocytes (PMN) the conversion of exogenous arachidonic acid into leukotriene B 4 (LTB 4 ) and (5 S )-hydroxy-(6 E,8Z,11Z,14Z )-eicosatetraenoic acid (5-HETE). However, contrary to earlier suggestions, 15-HETE is not a genuine 5-lipoxygenase inhibitor under these conditions, but rather suppresses the 5-lipoxygenation of arachidonic acid by switching-over of substrate utilization, as judged from a sizeable formation of labelled (5 S,15S )-dihydroxy-(6 E,8Z,11Z,13E )-eicosatetraenoic acid (5,15-diHETE) from 15-[1- 14 C]HETE. Identical results were obtained with human recombinant 5-lipoxygenase. In PMN the formation of 5,15-diHETE is strongly stimulated by either hydroperoxypolyenoic fatty acids or arachidonic acid, suggesting a crucial role of the hydroperoxide tone of the cell. A comparison of a selection of hydroxypolyenoic fatty acids with respect to their capability of suppressing 5-lipoxygenation of arachidonic acid revealed that 15-monohydroxyeicosanoids throughout exhibit the highest inhibitory potencies, whereas the other HETEs, 5,15-diHETE as well as octadecanoids, are modest or poor inhibitors. The R and S enantiomers of 15-HETE do not differ from each other, excluding a receptor-like binding of the 15-hydroxy group.

Oxidative modification of low-density lipoprotein (LDL) has been implicated in foam-cell formation at all stages of atherosclerosis. Since transition metals and mammalian 15-lipoxygenases are capable of oxidizing LDL to its atherogenic form, a concerted action of these two catalysts in atherogenesis has been suggested. Cu 2+ -catalysed LDL oxidation is characterized by a kinetic lag period in which the lipophilic antioxidants are decomposed and by a complex mixture of unspecific oxidation products. We investigated the kinetics of the 15-lipoxygenase-catalysed oxygenation of LDL and found that the enzyme is capable of oxidizing LDL in the presence of the endogenous lipophilic antioxidants. In contrast with the Cu 2+ -catalysed reaction, no kinetic lag phase was detected. The pattern of products formed during short-term incubations was highly specific, with cholesterol-esterified (13 S )-hydroperoxy-(9 Z,11E) -octadecadienoic acid being the major product. However, after long-term incubations the product pattern was less specific. Preincubation with 15-lipoxygenase rendered human LDL more susceptible to Cu 2+ -catalysed oxidation as indicated by a dramatic shortening of the lag period. Addition of Cu 2+ to lipoxygenase-treated LDL led to a steep decline in its antioxidant content and to a greatly reduced lag period. Interestingly, if normalized to a comparable hydroperoxide content, autoxidation and addition of exogenous hydroperoxy fatty acids both failed to overcome the lag period. The local peroxide concentrations in various LDL subcompartments will be discussed as a possible reason for this unexpected behaviour.