12-LOX catalyzes the oxidation of 2-arachidonoyl-lysolipids in platelets generating eicosanoid-lysolipids that are attenuated by iPLA2γ knockout

The canonical pathway of eicosanoid production in most mammalian cells is initiated by phospholipase A2-mediated release of arachidonic acid, followed by its enzymatic oxidation resulting in a vast array of eicosanoid products. However, recent work has demonstrated that the major phospholipase in mitochondria, iPLA2γ (patatin-like phospholipase domain containing 8 (PNPLA8)), possesses sn-1 specificity, with polyunsaturated fatty acids at the sn-2 position generating polyunsaturated sn-2-acyl lysophospholipids. Through strategic chemical derivatization, chiral chromatographic separation, and multistage tandem MS, here we first demonstrate that human platelet-type 12-lipoxygenase (12-LOX) can directly catalyze the regioselective and stereospecific oxidation of 2-arachidonoyl-lysophosphatidylcholine (2-AA-LPC) and 2-arachidonoyl-lysophosphatidylethanolamine (2-AA-LPE). Next, we identified these two eicosanoid-lysophospholipids in murine myocardium and in isolated platelets. Moreover, we observed robust increases in 2-AA-LPC, 2-AA-LPE, and their downstream 12-LOX oxidation products, 12(S)-HETE-LPC and 12(S)-HETE-LPE, in calcium ionophore (A23187)-stimulated murine platelets. Mechanistically, genetic ablation of iPLA2γ markedly decreased the calcium-stimulated production of 2-AA-LPC, 2-AA-LPE, and 12-HETE-lysophospholipids in mouse platelets. Importantly, a potent and selective 12-LOX inhibitor, ML355, significantly inhibited the production of 12-HETE-LPC and 12-HETE-LPE in activated platelets. Furthermore, we found that aging is accompanied by significant changes in 12-HETE-LPC in murine serum that were also markedly attenuated by iPLA2γ genetic ablation. Collectively, these results identify previously unknown iPLA2γ-initiated signaling pathways mediated by direct 12-LOX oxidation of 2-AA-LPC and 2-AA-LPE. This oxidation generates previously unrecognized eicosanoid-lysophospholipids that may serve as biomarkers for age-related diseases and could potentially be used as targets in therapeutic interventions.

membrane bilayer molecular dynamics and has multiple effects on complex transmembrane protein assembly through serving as a molecular scaffold (19). Remarkably, 12-LOX loss of function, through either genetic ablation or by pharmacologic inhibition, can reverse islet ␤-cell dysfunction and restore normal insulin secretion (20 -26). Moreover, 12-HETE activates platelets thereby serving as an autocrine and paracrine amplification cascade (27)(28)(29)(30)(31). Despite the profound biological and clinical importance of the downstream oxidized products of 12-LOX, investigations have been limited to previously known substrates for oxidation (i.e. arachidonic acid). Thus, the roles of 12-LOX in generating previously unknown metabolites that serve novel signaling functions remain to be clarified.
Recent structural studies of 12-LOX have revealed the fatty acid pentadiene target is positioned in close spatial proximity to the nonheme iron facilitating regiospecific hydrogen radicubal abstraction and stereospecific reaction of the radical intermediate with molecular oxygen to collectively result in the production of the chiral 12(S)-HpETE product (5,24,32,33). Notably, the crystal structure of the 12-LOX catalytic domain possesses a U-shaped substrate-binding site that facilitates the precise positioning of the C10 hydrogens in arachidonic acid for the regioselective and stereospecific oxidization of AA resulting in the production of 12(S)-HpETE (32). The open U-shape substrate-binding site of 12-LOX appeared large enough to accommodate the associated polar head group of 2-AA-lysophospholipids. Accordingly, we hypothesized that the entry of the arachidonoyl moiety of 2-AA-lysophospholipids into the active site and stereospecific oxidation appeared likely. Furthermore, the presence of the zwitterionic polar head group in 2-AA-LPC or 2-AA-LPE may allow further molecular interactions to stabilize the arachidonate moiety of 2-AA-lysophospholipids in the active site.
Previously, we identified a novel calcium-independent phospholipase, phospholipase A 2 ␥ (also known as PNPLA8) that catalyzes the highly regioselective (over 90%) cleavage of phospholipids at the sn-1 position (with a polyunsaturated fatty acid at the sn-2 position) resulting in the generation of 2-acyl-lysophospholipids (34). To gain further insight into the role of iPLA 2 ␥ in cardiac function, we generated germline iPLA 2 ␥ Ϫ/Ϫ mice by eliminating the active site of the enzyme. Genetic ablation of iPLA 2 ␥ resulted in growth retardation, compromised thermal adaption to cold, reduced exercise endurance, and a surprising ability of iPLA 2 ␥ Ϫ/Ϫ mice to remain refractory to the development of obesity and insulin resistance during prolonged HF feeding. We did not observe significant changes of the longevity between WT and iPLA 2 ␥ Ϫ/Ϫ mice (35,36). Moreover, lipidomic analysis revealed that genetic ablation of iPLA 2 ␥ Ϫ/Ϫ resulted in over a 60% decrease in the production of 2-[ 14 C]arachidonoyl-LPC from 1-palmitoyl 2-[ 14 C]arachidonoyl-PC in incubations with mitochondrial homogenates prepared from iPLA 2 ␥ Ϫ/Ϫ mice (35). The regiospecificity of iPLA 2 ␥ for the sn-1 position of polyunsaturated phospholipids suggested the existence of a previously unknown signaling pathway initiated by the generation of 2-arachidonoyl-lysophosphatidylcholine, which serves as a signaling node for multiple lipid 2nd messenger cascades (37,38).
Herein, we report the ability of platelet-type 12-LOX to efficiently catalyze the regioselective and stereospecific oxidization of both 2-AA-LPC and 2-AA-LPE into 12(S)-HpETE-LPC and 12(S)-HpETE-LPE, which are readily reduced to 12(S)-HETE-LPC and 12(S)-HETE-LPE both in vitro and in vivo, respectively. In addition, we show that multiple eicosanoid-lysophospholipid molecular species are present in myocardium and platelets with dynamic alterations after cellular stimulation. Finally, we demonstrated that aging in mice is accompanied by increased levels of 12-HETE-LPC in serum that are attenuated by genetic ablation of iPLA 2 ␥. Collectively, these results identify previously unknown iPLA 2 ␥-initiated signaling pathways mediated by direct 12-LOX oxidation of 2-arachidonoyl-lysophospholipids to generate previously unrecognized eicosanoid-lysophospholipid metabolites in critical cells of the cardiovascular system that could potentially serve as biomarkers of disease and inflammation during aging.

12-LOX efficiently catalyzes the regioselective and stereospecific oxidation of 2-AA-LPC and 2-AA-LPE to 12(S)-HpETE-lysophospholipids
To test the hypothesis that 2-AA-lysophospholipids are efficient substrates for oxidation by 12-LOX, we first cloned, expressed, and purified recombinant human platelet-type 12-LOX as described under "Experimental procedures." Additionally, we synthesized and purified 2-AA-LPC and 2-AA-LPE as previously described (34,(37)(38)(39). Next, we incubated 2-AA-LPC or 2-AA-LPE with purified 12-LOX. The reaction products were separated by HPLC and identified by high resolution high mass accuracy (HRAM) MS. Mass spectrometric analysis demonstrated the generation of previously unknown metabolites whose accurate masses and fragmentation patterns corresponded to the anticipated hydroperoxyl-oxidized eicosanoidlysophospholipids (m/z 576.3291 for HpETE-LPC and m/z 534.2820 for HpETE-LPE, respectively) ( Fig. 1, A and B). Multistage tandem MS with collision-induced dissociation (CID) fragmentation identified multiple informative fragment ions with the expected cleavage at the labile hydroperoxide bond and cleavage at the C12 position. Furthermore, accurate masses of the product ions were within 5 ppm of theoretical values and correspond to the likely fragmentation pathways (Figs. S1 and S2). Collectively, these experiments established the C12 regiochemistry of the 12-LOX-oxidized 2-AA-lysophospholipid products leading to their provisional assignments as 12-HpETE-LPC and 12-HpETE-LPE ( Fig. 1, C-F).
To further substantiate the proposed assignments by an independent method, we prepared and purified human recombinant cytosolic phospholipase A 2 ␣. Next, we treated the 12-LOX-generated eicosanoid-lysophospholipids with the purified recombinant cPLA 2 ␣ to hydrolyze the sn-2 acyl chain (40 -42). The released fatty acids were derivatized with N-(4amino-methylphenyl)-pyridinium (AMPP), and the resultant AMPP-eicosanoids were separated by RP-HPLC prior to identification by HRAM MS (43). The elution times and fragmentation patterns of the AMPP-derivatized eicosanoids originating from the 12-LOX-generated eicosanoid-lysophospholipids were indis-Eicosanoid-lysophospholipids are attenuated by iPLA 2 ␥ knockout tinguishable from authentic eicosanoid fatty acid standards, thereby confirming the proposed structural assignments (Fig. 2).

Determination of the chirality of 12-LOX-generated eicosanoid-lysophospholipids
The stereochemistry of the 12-LOX-generated eicosanoidlysophospholipid products were determined using chiral HPLC to separate the (R)-and (S)-stereoisomers of the AMPP-derivatized eicosanoids by comparisons with authentic standards. Mass spectrometric analysis of the separated reaction products revealed that 12-LOX-mediated oxidation of the arachidonoyl group of 2-AA-lysophospholipids yielded almost exclusively (Ͼ99%) the S-enantiomeric configuration (Fig. 3).

Kinetics of 12-LOX-mediated oxidation of 2-AA-LPC and 2-AA-LPE
Next, we determined the apparent kinetic values of 12-LOXmediated oxidation of both 2-AA-LPC and 2-AA-LPE substrates using substrate concentrations varying from 5 to 80 M. The results (Table S1) revealed that human platelet-type 12-LOX is capable of metabolizing 2-AA-LPC more rapidly than 2-AA-LPE by a factor of 1.9. Mouse platelet 12-LOX also shows the similar trend of substrate preference but there is no statistically significant difference under the conditions examined.

12-LOX-mediated oxidation of 2-AA-LPC and 2-AA-LPE is potently inhibited by ML355 (N-2-benzothiazolyl-4-[[(2-hydroxy-3-methoxy-phenyl)methyl] amino]-benzenesulfonamide)
Previous work has demonstrated that ML355 is a potent and selective inhibitor of 12-LOX-mediated generation of 12-HpETE from arachidonic acid (44). To further establish the role of 12-LOX in the oxidation of 2-AA-lysophospholipids, recombinant human or mouse platelet-type 12-LOX was preincubated with ML355 prior to addition of 2-AA-LPC or 2-AA-LPE substrates. Analysis of the reaction extracts by LC-MS demonstrated that production of 12-HpETE-LPC and 12-HpETE-LPE by either human or mouse platelet-type 12-LOX were dramatically inhibited (Ͼ80%) by ML355 (Fig.  4). Utilization of HRAM MS in conjunction with multistage Eicosanoid-lysophospholipids are attenuated by iPLA 2 ␥ knockout tandem MS CID fragmentation analysis and pharmacologic inhibition collectively identify 12-LOX as an efficient enzymatic mediator of the production of the identified eicosanoid-lysophospholipids from 2-arachidonoyl-lysophospholipids. These novel moieties have the potential to serve as metabolic and signaling nodes for a wide range of eicosanoid phospholipids after acyl transferase activity (or transacylase activity) or alternatively through the release of nonesterified eicosanoids after hydrolysis.

Genetic ablation of iPLA 2 ␥ markedly attenuates the production of 12-HETE-lysophospholipids in murine myocardium
To determine whether the identified 12-LOX-generated eicosanoid-lysophospholipids were present in murine myocar- Eicosanoid-lysophospholipids are attenuated by iPLA 2 ␥ knockout dium and if their production depended upon iPLA 2 ␥, we compared WT versus germline iPLA 2 ␥ Ϫ/Ϫ mice generated in our laboratory (35). Lipids from freshly isolated murine heart tissue of WT and germline iPLA 2 ␥ Ϫ/Ϫ mice were extracted using a modified Bligh and Dyer procedure, enriched for eicosanoidlysophospholipids by solid-phase extraction, and analyzed by

Calcium ionophore stimulates the production of 12-HETElysophospholipids in murine platelets, which is inhibited by genetic knockout of iPLA 2 ␥
Next, to mechanistically identify the role of iPLA 2 ␥ in the release of 2-AA-lysophospholipids and subsequent production of 12-LOX eicosanoid-lysolipids in intact platelets, we stimulated murine platelets with calcium ionophore. Platelets from WT and germline iPLA 2 ␥ Ϫ/Ϫ mice were treated with either the vehicle (DMSO) alone or the calcium ionophore A23187 (1.0 M). Ionophore-mediated calcium influx induced a robust increase in the levels of both 2-AA-LPC and 2-AA-LPE in WT platelets, which was markedly decreased in iPLA 2 ␥ knockout platelets in comparisons to WT littermate controls (Fig. 6, A  and B). Notably, abundant amounts of 1-AA-LPC and 1-AA-LPE were also detected in platelets under the experimental conditions employed (Fig. S3). Previous work has indicated that the rate of ␣-hydroxy migration from the sn-2 carbon to the sn-1 carbon is influenced by multiple factors including pH and ionic strength (45). In the current work, we investigated acyl migra-  Eicosanoid-lysophospholipids are attenuated by iPLA 2 ␥ knockout tion for 2-AA-LPC and 2-AA-LPE at selected pH values (Fig.  S4). The results demonstrated that the acyl migration rate from the sn-2 position to the sn-1 position is highly dependent upon the pH of the solution. Although we cannot entirely rule out the direct synthesis of 1-AA-LPC and 1-AA-LPE in vivo (i.e. di-AA-PC and di-AA-PE, which can be hydrolyzed by PLA 2 enzymes (e.g. cPLA 2 ␣) to generate 1-AA-LPC and 1-AA-LPE), we conclude that the majority of the 1-AA-lysophospholipids detected in platelets likely result from the cellular synthesis of 2-AA-lysophospholipids followed by nonenzymatic ␣-hydroxyl acyl migration.
Importantly, the generation of 12-HETE-lysophospholipids in platelets was increased by the influx of calcium ion, which induced a robust increase in the levels of both 12-HETE-LPC and 12-HETE-LPE in WT platelets (Fig. 6, C and D). In stark contrast, genetic ablation of iPLA 2 ␥ markedly decreased the calcium-stimulated production of 12-LOX-derived eicosanoid-lysophospholipids in iPLA 2 ␥ Ϫ/Ϫ platelets relative to WT littermate controls (Fig. 6, C and D). Moreover, ML355 did not change the production of 2-AA-LPC and -LPE, but it significantly inhibited the production of 12-HETE-LPC and -LPE (Fig. 6).

Eicosanoid-lysophospholipids increase with aging in murine serum, demonstrate gender-related differences, and are markedly attenuated in iPLA 2 ␥ ؊/؊ mice
Aging is known to result in substantial changes in serum lipids (46,47) and sex differences in free eicosanoids production have been observed (48,49). However, alterations in eicosanoid-lysophospholipid levels during aging or comparisons of male versus female mice have not been previously made. To gain insight into potential age-and gender-related changes in eicosanoid-lysophospholipid levels, we analyzed the amount of 12-HETE-LPC in serum from young (3-month-old) and old (10-month-old) male and female WT and iPLA 2 ␥ Ϫ/Ϫ mice. Notably, a 2-fold increase in serum 12-HETE-LPC was observed in 10-month-old male mice compared with 3-monthold male mice (Fig. 8). Furthermore, serum 12-HETE-LPC was markedly attenuated in iPLA 2 ␥ Ϫ/Ϫ male mice (ϳ65% decrease) relative to WT controls (Fig. 8). In contrast, we identified only modestly increased amounts of 12-HETE-LPC in serum from female mice during aging. Genetic ablation of iPLA 2 ␥ in female mice resulted in similar decreases relative to their WT controls as observed in the serum of male mice (Fig. 8). Collectively, these results demonstrate that aging is accompanied by marked increases in 12-HETE-LPC in male murine serum and imply that 12-HETE-LPC may serve as a gender-specific biomarker for various age-associated diseases.

Discussion
Previously, we demonstrated that calcium ion induces the activation of iPLA 2 ␥ leading to the production of 2-AA-LPC and 2-AA-LPE (37). Remarkably 2-arachidonoyl-lysophospholipids were directly oxidized by cyclooxygenase-2 to produce multiple eicosanoid-lysophospholipids including those containing either prostaglandin (e.g. prostaglandin E 2 ) or other hydroxylated polyunsaturated aliphatic acyl chains (e.g. 11(R)-HETE and 15-HETE) that are major participants in signaling processes (38). These findings led us to investigate whether 2-arachidonoyl-lysophospholipids were substrates for the platelet-type 12-lipoxygenase, which is the predominant 12-LOX in most cellular systems (50). In the present study, we expressed and purified recombinant human platelet-type 12-LOX and demonstrated that 12-LOX regioselectively and stereospecifically oxidizes multiple classes of 2-AA-lysophospholipids to their corresponding 12(S)-HETE-lysophospholipid derivatives. Intriguingly, 2-arachidonoyl-glycerol was not detectably oxidized by platelet-type 12-LOX (51). In stark contrast, 2-AA-LPC and 2-AA-LPE are efficiently oxidized by platelet-type 12-LOX. This difference demonstrates an interesting substrate specificity of platelet-type 12-LOX suggesting that additional interaction sites for suitable orientation and efficient catalysis by platelet 12-LOX is necessary.
To mechanistically identify the enzyme(s) responsible for the production of these previously unknown natural products, a mouse model with germline ablation of exon 5 (encoding the active site) of iPLA 2 ␥ was utilized (35). Upon stimulation with calcium ionophore, the production of both 2-AA-LPC and
Oxidized lysophospholipids are anticipated to undergo rapid lateral motion in the membrane facilitating two-dimensional interactions to engage biological partners (e.g. receptors, ion channels, etc.). Furthermore, because the membrane limits 3D spatial interactions, membranes can serve as scaffolds for complex assemblies of biomolecules (19,52,53). Thus, 12(S)-HETE-lysophospholipids are predicted to have potent pleiotropic effects on cellular membrane structure, molecular dynamics, dipole interactions, membrane potentials, and scaffolds in addition to serving as potential ligands to cellular receptors.
Moreover, we note that these oxidized lysophospholipids are also excellent substrates for sn-1 acyltransferases and transacylases, which result in the production of oxidized phospholipids that have emerged as potent signaling molecules in many systems (54). We suggest that multiple discrete eicosanoid-lysophospholipids generate a rich repertoire of molecular entities that can interact with specific receptors, differently modulate membrane surface charge and protein dynamics, and potentially serve as DAMPs (damage-associated molecular patterns) or DAMP precursors to facilitate specific cellular receptors in injured tissues. The difficulty to demonstrate a bona fide interaction due to the rapidly changing products of this class of lipids should be recognized. We have begun the synthesis of a nonhydrolyzable analog as well as an analog missing the sn-1 hydroxyl to address this important point.
Previously, we demonstrated that the sn-1 vinyl ether linkage of plasmenylcholine and plasmenylethanolamine containing arachidonic acid at the sn-2 position can be oxidatively cleaved by oxidized cardiolipin-activated cytochrome c to yield 2-AA-LPC and 2-AA-LPE, respectively (i.e. plasmalogenase activity) (55). Thus, in addition to the production of 2-AA-lysophospholipids by iPLA 2 ␥, it seems likely that a portion of the 2-arachidonoyl-lysophospholipids and their corresponding 12-LOX oxidation products observed in iPLA 2 ␥ Ϫ/Ϫ platelets is due to cleavage of the vinyl ether linkage of plasmenylethanolamine, which is rich in platelet membranes. Alternatively, other PLA 1 activities, either known or as yet undiscovered, may contribute to the generation of sn-2-arachidonoyl-lysophospholipids. Additionally, these studies also demonstrate that eicosanoidlysophospholipids can be directly hydrolyzed by a variety of intracellular phospholipases (e.g. cPLA 2 ␣) and lysophospholipases resulting in the liberation of nonesterified eicosanoids that are known to possess potent effects in multiple disease states including diabetes, cancer, cardiovascular disease, and endothelial dysfunction (6 -8).
Collectively, the present study identifies previously unknown signaling metabolites emanating from the iPLA 2 ␥-mediated sn-1 hydrolysis of arachidonate-containing phosphatidylcholine and phosphatidylethanolamine. The resultant 2-arachidonoyl-lysophospholipids are subsequently oxidized directly Eicosanoid-lysophospholipids are attenuated by iPLA 2 ␥ knockout by 12-LOX to generate eicosanoid-lysophospholipids (Fig. 7). Furthermore, our results demonstrate that aging is associated with marked alterations in 12-HETE-lysophospholipid levels in murine serum that were markedly attenuated by iPLA 2 ␥ genetic ablation. Finally, our results identify high concentrations of eicosanoid-lysolipids in murine serum that are age and gender selective. Thus, these moieties or their downstream metabolites can potentially provide new biomarkers for discovery of age-related disease states such as Alzheimer's disease.

Expression and purification of human platelet-type 12-LOX
Human platelet-type 12-LOX containing an N-terminal (His) 6 tag was constructed and inserted into the BamHI-Hin-dIII sites of pFastbac1 (Invitrogen) by GenScript. The sequence and orientation verified the identity of the recombinant pFast-Bac1 containing human ALOX12, which was transfected into DH10Bac cells (InVitrogen). The Bac-to-Bac Baculovirus Expression System (Invitrogen) was subsequently utilized for bacmid preparation and CellFECTIN-mediated transfection of Spodoptera frugiperda (Sf9) cells to produce recombinant baculovirus, which was then amplified and titered by the Neutral Red agar overlay method according to the manufacturer's protocol (Invitrogen). Expression of the N-terminal His-tagged 12-LOX was initiated by infection of 100 ml of Sf9 cells (1 ϫ 10 6 cells/ml) with recombinant virus at a multiplicity of infection of 1. After incubation for 68 h at 27°C, the cells were pelleted by centrifugation at 900 ϫ g for 10 min at 4°C, resuspended in 2 ml of ice-cold hypotonic 0.1 ϫ PBS with swelling on ice for 2 min. Swelling was then stopped by addition of 2 ml of ice-cold hypertonic 2ϫ PBS containing 10 mM imidazole (pH 7.4), 40 l of 0.1 M phenylmethylsulfonyl fluoride, and 2 l of 10 mg/ml leupeptin. The Sf9 cells were then sonicated (20 ϫ 1-s bursts utilizing a Vibra-cell sonicator at 50% output). After adding 10 l of DNase I (1.0 unit/l) and 40 l of 20% Tween 20, the Sf9 cells were sonicated again (10 ϫ 1-s bursts at 50% output), incubated for 5 min on ice, and centrifuged at 15,000 ϫ g for 15 min at 4°C. The supernatant was removed and the NaCl concentration was adjusted to 300 mM.
Purification of the recombinant 12-LOX was performed using nickel-nitrilotriacetic acid (Ni-NTA) metal-affinity chromatography per the manufacturer's protocol (Thermo Scientific number 88225). Briefly, the protein extract was loaded onto a Ni-NTA column that was pre-equilibrated with 3 ϫ 10 ml of 2ϫ PBS containing 25 mM imidazole (pH 7.4). After washing the column 6 times with 7.0 ml of 2ϫ PBS containing 25 mM imidazole (pH 7.4) to remove nonspecifically bound proteins, the His-tagged 12-LOX was fractionally eluted with 9 ϫ 1.0 ml of ice-cold 2ϫ PBS containing 150 mM imidazole (pH 7.4). The presence of the recombinant 12-LOX was confirmed by ECL Western blot analysis using an anti-(His) 6 mAb.

Kinetic analysis of the 12-LOX-mediated oxidation of 2-AA-LPC or 2-AA-LPE
Time course experiments-Human or mouse platelet-type recombinant 12-LOX (mouse platelet-type 12-LOX was purchased from Cayman Chemicals Co., Ann Arbor, MI) (2 g) was incubated with either 2-AA-LPC or 2-AA-LPE (10 M) in 400 l of 100 mM Tris-HCl buffer (pH 7.2) containing 10% glycerol at 30°C for various times. The reaction was then terminated by adding 100 l of methanol, acidified to pH 4 with glacial acetic acid and applied to a Strata-X solid-phase extraction cartridge previously preconditioned with 3 ml of methanol and equilibrated with 3 ml of 20% methanol, 80% H 2 O. After washing with H 2 O, the reaction products were eluted with methanol and analyzed by LC-MS methods as described below.
Inhibition of 12-LOX by ML355-Human or mouse platelettype recombinant 12-LOX (2 g) was incubated with ML355 (ML355 was purchased from Cayman Chemicals) (1.0 M) in 400 l of 100 mM Tris-HCl buffer (pH 7.2) containing 10% glycerol at 23°C for 15 min. 2-AA-LPC or 2-AA-LPE were then added and the reactions were incubated for various times at 30°C before termination with acidified methanol. The reaction products were isolated, concentrated, and purified by solidphase extraction using Strata-X columns as described above and analyzed by LC-MS as described below.

Eicosanoid-lysophospholipids are attenuated by iPLA 2 ␥ knockout
Substrate concentration dependence experiments-Human or mouse platelet-type recombinant 12-LOX (2 g) was incubated with varying concentrations of 2-AA-LPC or 2-AA-LPE (5 to 80 M) in 400 l of 100 mM Tris-HCl buffer (pH 7.2) containing 10% glycerol in a 30°C water bath for 2 min. The reactions were then terminated, and the reaction products were analyzed by LC-MS as described below. The kinetic values of K m and V max were determined from a nonlinear least-squares best-fit of the data to the Michaelis-Menten equation using GraphPad Prism (GraphPad Software, La Jolla, CA).

Hydrolysis of eicosanoid-lysophospholipids by cPLA 2 ␣ and identification of the released eicosanoid products
The eicosanoid-lysophospholipid products generated by 12-LOX were incubated with purified cPLA 2 ␣ (5 g) in 200 l of HEPES buffer (10 mM HEPES, pH 7.4, 100 mM KCl, 0.5 mM CaCl 2 , and 10% glycerol) for 15 min at 37°C. Reactions were stopped by addition of 50 l of methanol followed by acidification of the reaction mixture to pH 4 with glacial acetic acid. Terminated reactions were immediately applied to a Strata-X solid-phase extraction cartridge (60 mg/ml) that had been previously preconditioned with 3 ml of methanol followed by sequential washing with 3 ml of 10% methanol, 90% H 2 O. After washing with 3 ml of H 2 O, the oxidized fatty acid metabolites were eluted with 2 ϫ 0.5-ml volumes of methanol containing 1% glacial acetic acid. The combined acidified methanol eluates were evaporated to dryness using a SpeedVac concentrator prior to charge-switch derivatization with AMPP. Derivatization with AMPP was performed as previously described (43). Briefly, 20 l of ice-cold acetonitrile/N,N-dimethylformamide (4:1, v/v) and 20 l of ice-cold 640 mM (3-(dimethylamino)propyl)ethylcarbodiimide hydrochloride in HPLC grade water were added to the dried residue in a sample vial. The vial was briefly vortexed and 40 l of 5 mM N-hydroxyl benzotriazole and 15 mM AMPP in acetonitrile were added, vortexed, and placed in a 60°C water bath for 30 min. Samples were then analyzed by LC-MS n using "C18 reverse phase HPLC-tandem MS" and "chiral-phase HPLC-tandem MS" (see below).

C18 reverse phase HPLC-tandem MS (LC-MS n )
LC-MS n analyses were performed using an LTQ-Orbitrap mass spectrometer (Thermo Scientific, San Jose, CA) equipped with a Surveyor HPLC system (Thermo Scientific). Briefly, solid-phase extractions of the 12-lipoxygenase reaction mixtures utilizing 2-AA-LPC or 2-AA-LPE as a substrate were separated on a C18 reversed phase column (Ascentis Express, 2.7-m particles, 150 ϫ 2.1 mm) at 23°C using a linear gradient of solvent A (0.1% glacial acetic acid in water) and solvent B (0.1% glacial acetic acid in acetonitrile) at a flow rate of 0.2 ml/min. 100% B to 25% B followed by 10 min isocratic re-equilibration at 25% B. The sample injection volume was 10 l and the autosampler tray temperature was maintained at 4°C throughout the analysis. Mass spectrometric analyses were performed using an LTQ-Orbitrap mass spectrometer. The LTQ ion source was operated in the positive ion mode at sheath, auxil-iary, and sweep gas flows (arbitrary units) of 40, 5, and 1, respectively. The capillary temperature was set to 275°C, and the electrospray voltage was 4.1 kV. Capillary voltage and tube lens voltage were set at 30 and 110 V, respectively. The mass spectrometer was calibrated using the manufacturer's recommended positive mode calibration solution containing L-methionyl-arginyl-phenylalanyl-alanine acetate, Ultramark 1621, and caffeine. Resolving powers (at m/z 400 Th) of 30,000 in full scan mode and in MS/MS mode were used. The mass accuracy was within 5 ppm at mass values from m/z 130 to 2000. For MS/MS analyses, a normalized collision energy of 30% was applied and the activation time was set at 30 ms with an activation parameter of q ϭ 0.25. Data acquisition was performed using an Xcalibur operating system, version 2.1 (Thermo Scientific).

Chiral-phase HPLC-tandem MS
Chiral-phase LC-MS n analyses employing reverse phase LC were performed using an LTQ-Orbitrap mass spectrometer (Thermo Scientific) with a Surveyor HPLC system (Thermo Scientific). Briefly, AMPP-derivatized eicosanoids were injected and separated on a Chiralpak AD-RH chiral-phase HPLC column (5.0-m particles, 150 ϫ 2.1 mm) at 23°C using a linear gradient of solvent A (0.1% glacial acetic acid in water) and solvent B (0.1% glacial acetic acid in acetonitrile) at a flow rate of 0. , 100% B prior to isocratic re-equilibration at 5% B for 10 min. The sample injection volume was 10 l and the autosampler tray temperature was maintained at 4°C throughout the analysis. Mass spectrometric analysis conditions were identical to those described above under "C18 reverse phase HPLC-tandem MS (LC-MS n )."

General animal studies
Animal protocols were conducted in strict accordance with the National Institutes of Health guidelines for humane treatment of animals and were reviewed and approved by the Animal Studies Committee of Washington University.

Murine myocardium sample preparation and analysis of eicosanoid-lysophospholipids
Murine heart tissue samples from control C57BL/6 and iPLA 2 ␥ Ϫ/Ϫ mice (35) were rapidly removed, flash frozen in liquid nitrogen, and stored at Ϫ180°C prior to analysis for eicosanoid-lysophospholipids. Myocardial tissues were extracted using a modified Bligh and Dyer (56) procedure. Briefly, ϳ40 mg of tissue was extracted by addition of 3 ml of ice-cold CH 3 OH/CHCl 3 (1:2 (v/v) containing 2% glacial acetic acid) followed directly by addition of 17:0-LPC as an internal standard. Next, the samples were homogenized using a Polytron homogenizer at 0 -4°C and 1 ml of ice-cold H 2 O was added prior to separation of the organic and aqueous phases by centrifugation at 1,500 ϫ g for 15 min. The CHCl 3 layer was transferred to a new tube, whereas the aqueous layer was re-extracted by subsequent addition of 2 ml of CHCl 3 , vortexing, and phase separation by centrifugation at 1,500 ϫ g for 15 min. The CHCl 3 Eicosanoid-lysophospholipids are attenuated by iPLA 2 ␥ knockout extracts were combined, evaporated under a nitrogen stream, and resuspended in 100 l of 80% methanol in water prior to LC-MS/MS analyses as described above.

Murine serum sample preparation and analysis of eicosanoid-lysophospholipids
Serum samples from 10-or 3-month-old control C57BL/6 mice and iPLA 2 ␥ Ϫ/Ϫ mice (35) were prepared by centrifugation of the blood collected from submandibular veins at 1,500 ϫ g for 10 min after incubation for 30 min at room temperature. The serum was quickly removed and 10 l of fresh serum were used for lipid extractions using a modified Bligh and Dyer procedure (56) and analysis by HPLC-MS/MS as described above.

Isolation and activation of platelets
Blood from control C57BL/6 and iPLA 2 ␥ Ϫ/Ϫ mice (35) was obtained by intracardiac puncture from euthanized mice. About 0.8 ml of blood was drawn into a syringe containing 0.15 ml of 3.8% sodium citrate to prevent platelet activation. The blood was then centrifuged at 150 ϫ g for 10 min at room temperature, and the platelet-rich-plasma was subsequently centrifuged at 250 ϫ g for 5 min to remove residual red blood cells. The purified platelet-rich-plasma was spun at 1,500 ϫ g for 10 min, and the pellet was resuspended in Tyrode's buffer containing 0.38% sodium citrate. The platelets were pelleted by centrifuging at 800 ϫ g for 10 min, resuspended in Tyrode's buffer containing 1.0 mM CaCl 2 and incubated at 37°C in the presence or absence of A23817 (1.0 M) for 15 min. For inhibition experiments, ML355 (10 M) was preincubated with the isolated platelets for 15 min prior to addition of A23187. Lipids were then extracted using a modified Bligh and Dyer (56) procedure and analyzed by HPLC-MS/MS as described above.

Statistical analysis
Statistical analyses were performed using the two-tailed Student's t test. p values of less than 0.05 were considered to be statistically significant. All data are reported as the mean Ϯ S.E. unless otherwise noted.

Data availability
All data presented in this paper are contained within the manuscript.