Human platelets generate phospholipid-esterified prostaglandins via cyclooxygenase-1 that are inhibited by low dose aspirin supplementation

Abstract

Oxidized phospholipids (oxPLs) generated nonenzymatically display pleiotropic biological actions in inflammation. Their generation by cellular cyclooxygenases (COXs) is currently unknown. To determine whether platelets generate prostaglandin (PG)-containing oxPLs, then characterize their structures and mechanisms of formation, we applied precursor scanning-tandem mass spectrometry to lipid extracts of agonist-activated human platelets. Thrombin, collagen, or ionophore activation stimulated generation of families of PGs comprising PGE₂ and D₂ attached to four phosphatidylethanolamine (PE) phospholipids (16:0p/, 18:1p/, 18:0p/, and 18:0a/). They formed within 2 to 5 min of activation in a calcium, phospholipase C, p38 MAP kinases, MEK1, cPLA₂, and src tyrosine kinase-dependent manner (28.1 ± 2.3 pg/2 × 10⁸ platelets). Unlike free PGs, they remained cell associated, suggesting an autocrine mode of action. Their generation was inhibited by in vivo aspirin supplementation (75 mg/day) or in vitro COX-1 blockade. Inhibitors of fatty acyl reesterification blocked generation significantly, while purified COX-1 was unable to directly oxidize PE in vitro. This indicates that they form in platelets via rapid esterification of COX-1 derived PGE₂/D₂ into PE. In summary, COX-1 in human platelets acutely mediates membrane phospholipid oxidation via formation of PG-esterified PLs in response to pathophysiological agonists.

Keywords: Oxidized phospholipids; PGE2/D2-PEs; atherosclerosis.

Fig. 1.

Identification of esterified PGs in human platelets and analysis of PGE₂/D₂-PE using LC/MS/MS. A: Precursor scanning demonstrates lipids with m/z 351.2 eluting during LC/MS/MS. Total lipid extracts from washed human platelets activated with 0.2 U/ml of thrombin for 30 min at 37°C were separated on the Q-Trap platform using LC/MS/MS as described in Materials and Methods, with online negative precursor scanning for m/z 351.2. *, region of LC trace where ions appear that are elevated by thrombin stimulation. Control, broken line. B: Identification of ions that generate m/z 351.2 daughter ions. Shown is a negative MS scan of region marked * in A. Scan shows ions eluting between 19 and 24 min. C: Characterizing phospholipid headgroups of esterified PL. Lipid extracts from thrombin-activated platelets were separated on normal-phase HPLC, as described in Materials and Methods, with fractions collected at 30 sec intervals. Twenty microliters of each fraction was analyzed specific parent → m/z 351.2 MRM transitions. PL class elution was determined using commercial phospholipid standards. Panels D–G: LC/MS/MS of PGE₂/D₂-PEs. Platelet lipid extracts were separated using LC/MS/MS as described in Materials and Methods and detected on the Q-Trap platform by parent → m/z 271.2.

Scheme 1.

Structures of PGE₂/D₂-PE molecules identified in human platelets.

Fig. 2.

Generation of free and esterified PGs by agonist-activated platelets. Generation of PGE₂/D₂-PEs and free prostaglandins in response to pathophysiological agonists. Washed platelets were activated for varying times as shown, and lipids were extracted and analyzed using LC/MS/MS as described in Materials and Methods. Platelets were activated using 0.2 U/ml thrombin and PGE₂/D₂-PEs were determined (A). Platelets were activated using 10 µg/ml of collagen (B). Platelets were activated using 10 µg/ml of collagen and 0.2 U/ml of thrombin (C). Platelets were activated using 10 µmol/L A23187 (D). Levels of PGE₂/D₂-PE are expressed as ratio analyte to internal standard with experiments repeated at least three times on different donors (n = 3, mean ± SEM).

Fig. 3.

Esterified prostaglandins are retained by platelets while free PGE₂ and PGD₂ are primarily secreted, and generation of free and esterified PGs is sensitive to COX-1 inhibition in vitro and in vivo. A, B: Esterified PGs are retained by platelets. Washed human platelets were activated with 0.2 U/ml thrombin for 30 min before centrifugation at 970 g. The supernatant was centrifuged at 16,060 g to pellet microparticles before lipid extraction and analysis by LC/MS/MS. C, D: Esterified and free PGE₂/D₂ generation is sensitive to aspirin in vitro. Platelets were incubated for 15 min with 1 mM aspirin prior to thrombin activation (0.2 U/ml, 30 min) followed by lipid extraction and analysis of free and esterified PGE₂/D₂ using LC/MS/MS. E, F: Esterified and free PGE₂/D₂ generation is sensitive to COX inhibitors. Platelets were incubated for 10 min with 1 μM Sc-560 or 10 μM indomethacin prior to thrombin activation (0.2 U/ml for 30 min) followed by lipid extraction and analysis of free and esterified PGE₂/D₂ using LC/MS/MS. For all experiments described above, n = 3; mean ± SEM; data representative of three independent donors. *P < 0.05, **P < 0.01, and ***P < 0.001 versus control using ANOVA and Bonferroni post hoc test. G, H: Free and esterified PGE₂/D₂ formation in vivo is blocked by aspirin. Blood was obtained following a 14-day NSAID-free washout period for isolation of washed platelets and determination of free and esterified PGE₂/D₂ levels, as described in Materials and Methods, following activation using 0.2 U/ml thrombin for 30 min. Subjects then received 75 mg/day aspirin for 7 days before donation of a second blood sample and repeat determination of free and esterified PGE₂/D₂ levels. Data is representative of five independent donors (n = 5, mean ± SEM); *P < 0.05, **P < 0.01, and ***P < 0.001 versus control using ANOVA and Bonferroni post hoc test. Levels of PGE₂/D₂-PE are expressed as ratio analyte to internal standard. Levels of free PGE₂ and PGD₂ are expressed as ng/2 × 10⁸ platelets.

Fig. 4.

Generation of PGE₂/D₂-PEs requires cPLA₂ and esterification of free eicosanoids. A-D: Generation of free and esterified PGE₂/D₂ requires cPLA₂ but not other PLA₂ isoforms. Washed human platelets were incubated for 10 min with each phospholipase A₂ inhibitor prior to thrombin activation (0.2 U/ml for 30 min) followed by lipid extraction and analysis by LC/MS/MS. Inhibitors: cytosolic PLA₂ (cPLA_2α) inhibitor (cPLA_2αi, 50 nmol/L), Ca²⁺-dependent secretory PLA₂ (sPLA₂) inhibitor (OOEPC2, 2 µmol/L), Ca²⁺-independent intracellular PLA₂ (iPLA₂) inhibitor (BEL, 50 nmol/L) or vehicle (DMSO, 0.5%). E, F: Inhibition of PGE₂/D₂-PE, by thimerosal or triascin C. Washed platelets were incubated for 30 min at 37°C with 75 µmol/L thimerosal or 7 mmol/L triascin C prior to thrombin activation (0.2 U/ml for 30 min) followed by lipid extraction and analysis using LC/MS/MS. For all experiments, n = 3; mean ± SEM; data are representative of three independent donors. *P < 0.05, **P < 0.01, and ***P < 0.001 versus control using ANOVA and Bonferroni post hoc test. G, H: COX isoforms generate a 2:1 ratio of PGE₂:PGD₂ in vitro. 3.5 μg of purified ovine COX-1 and 3.5 μg recombinant murine COX-2 were incubated at 37°C for 3 min with 150 µM of AA before lipid extraction and analysis using LC/MS/MS as described in Materials and Methods. PGE₂ and PGD₂ are expressed as micrograms/3.5 μg enzyme formed over 3 min (n = 3, mean ± SEM). I: PE is oxidized to PGE₂/D₂-PE during oxidation of AA by COX-1. Purified ovine Apo-PGHS-1 was reconstituted with hematin, ratio 2:1 (hematin: Apo-PGHS-1). 3.5 μg of purified ovine COX-1 was incubated at 37°C for 3 min with the following substrates: 150 µmol/L of AA; 150 µmol/L SAPE; liposomes containing AA and SAPE, in the presence or absence of 10 μM DTPA, before lipid extraction and analysis using LC/MS/MS as described in Materials and Methods. Levels of PGE₂/D₂-PE are expressed as ratio analyte to internal standard/3.5 μg enzyme generated over 3 min. PGE₂ and PGD₂ are expressed as micrograms/3.5 μg enzyme generated over 3 min (n = 3, mean ± SEM).

Fig. 5.

Elucidation of pathways involved in esterified and free PGE₂/D₂ generation by human platelets. A, B: Esterified and free PGE₂/D₂s are generated via PAR-1 and PAR-4 receptor stimulation. Washed platelets were activated with a PAR-1 agonist, TFLLR-NH₂ (20 μmol/L), and/or a PAR-4 agonist AY-NH₂ (150 μmol/L) for 30 min before lipid extraction and LC/MS/MS analysis as described in Materials and Methods. C, D: Cytosolic Ca²⁺ is required for thrombin-stimulated generation of esterified and free PGE₂/D₂. Washed platelets were incubated for 10 min with each inhibitor prior to thrombin activation (0.2 U/ml for 30 min) before lipid extraction and analysis as described in Materials and Methods. E-H: Phospholipase C, MEK1, p38 MAP kinases and src tyrosine kinases are required for thrombin-stimulated generation of PGE₂/D₂-PEs. Washed human platelets were incubated for 10 min with inhibitors prior to thrombin (0.2 U/ml, 30 min) followed by lipid extraction and analysis for free and esterified PGE₂/D₂. Inhibitors used are as follows: PI3 kinase (wortmannin, 100 nmol/L), protein kinase C (Gö 6850, 100 nmol/L), MEK1 (PD98059, 50 µmol/L), src-family tyrosine kinase (PP2, 50 µmol/L), p38 MAP kinase inhibitor (p38i, 100 nmol/L), and phospholipase C (U-73112, 5 µmol/L), or vehicle (DMSO, 0.5%). For all experiments, n = 3; mean ± SEM; data are representative of three independent donors. *P < 0.05, **P < 0.01, and ***P < 0.001 versus control using ANOVA and Bonferroni post hoc test. Levels of PGE₂/D₂-PE are expressed as ratio analyte to internal standard (n = 3, mean ± SEM).

Scheme 2.

Proposed mechanisms for formation of PGE₂/D₂-PE by human platelet COX-1. First, AA is oxidized by COX-1, then esterified into PE. Formation of the lipids takes place in intracellular membranes where COX-1 and esterification enzymes are localized.