Acetonitrile (ACN), water (LC-MS grade), formic acid (FA), n-hexane, methanol, acetic acid, and isopropanol were from Carlo Erba Reagenti, Milan, Italy. Standards of TXB₂, PGE₂, HETEs, and 15-epi-LXA₄ and their deuterated forms were from Cayman Chemical (Ann Arbor, Michigan, United States). ECL Western Blotting Detection Reagents were from GE Healthcare (Milan, Italy). Acetylsalicylic acid (ASA, aspirin), dimethyl sulfoxide (DMSO), ethanol (ETOH), bovine serum albumin (BSA), NaCl, KCl, MgCl₂, CaCl₂, glucose, citric acid, citrate-dextrose solution (ACD), Triton X-100, phenylmethylsulfonyl Fuoride (PMSF), the β-actin monoclonal antibody (cat# A5316), dextran from Leuconostoc spp, Hystopaque 1,077, Dulbecco′s Modified Eagle′s Medium (DMEM), Penicillin-Streptomycin, Fetal Bovine Serum (FBS), endoproteinase Glu-C Sequencing Grade (Glu-C), arachidonic acid (AA), ammonium bicarbonate (NH₄HCO₃), dithiothreitol (DTT) were from Sigma Aldrich, Milan, Italy. The columns (Aeris 1.7 μm PEPTIDE XB-C18 100 Å, 100 × 2.1 mm and Lux^® 3 μmAmylose-1, 150 mm × 3.0 mm) were from Phenomenex, Torrance, CA, United States. The Bradford protein assay, the Laemmli Buffer, β-Mercaptoethanol, the PVDF membrane, and the non-fat milk for immunoblot were from Bio-Rad, Milan, Italy. The anti-platelet type 12-LOX polyclonal antibody (item#Ab211506) and the anti-15-LOX-1 polyclonal antibody (item #Ab80221) were from Abcam, Cambridge, UK; the anti-β-actin monoclonal antibody was from Santa Cruz Biotechnology (item#sc8432); the anti-5-LOX polyclonal antibody was a gift from Dr. Dieter Steinhilber (Goethe University, Frankfurt, Germany). Colon cancer cell line HCA7 colony 29 (HCA7) was from the European Collection of Cell Cultures (ECC, Salisbury, UK). Trypsin Gold (Mass Spectrometry Grade) and ProteaseMAX™ Surfactant, Trypsin Enhancer were from Promega (Wisconsin, United States). The light standard peptides (unacetylated and acetylated) of COX-1 and COX-2 [for COX-1: IGAPFSLK and IGAPFS(Ace)LK; for COX-2: VGAPFSLK and VGAPFS(Ace)LK] and the corresponding AQUA peptides, which were incorporated with stable isotopes ¹³C and ¹⁵N (Patrignani et al., 2014; Tacconelli et al., 2020b) were synthesized by Thermo Fisher Scientific (Waltham, MA, United States) (>97% purity). NuPAGE^® Novex 7% Tris-Acetate Gel, RIPA buffer, GelCode Blue Stain Reagent, NuPAGE LDS sample buffer, EDTA-free proteases inhibitors were from Thermo Fisher Scientific (Waltham). L-ASA (IP1867B) and triacetin/saccharin (called diluent) were a gift from Innovate Pharmaceuticals, Manchester, United Kingdom.

Peripheral venous blood samples were drawn from healthy volunteers (n = 10, 23–50 years) when they had not taken any non-steroidal antiinflammatory drug (NSAID) during the 2 weeks preceding the study. This study was carried out following the recommendations of the Declaration of Helsinki after approval by the local Ethics Committee of “G. d' Annunzio” University of Chieti-Pescara (#254), and informed consent was obtained from each subject.

ASA (0.005–150 mM) was dissolved in DMSO; IP1867B (L-ASA 0.005–150 mM) was dissolved in triacetin (glycerin triacetate) and saccharin (named diluent) (kindly provided by Innovate Pharmaceuticals, United Kingdom); then 2-µL aliquots of the vehicles or the different solutions of ASA or L-ASA were pipetted directly into glass test tubes to give final concentrations of 0.01–300 µM (for ASA) and 0.01–300 µM (for L-ASA) in 1 ml of blood. One-ml aliquots of whole blood were immediately transferred into glass tubes and allowed to clot at 37 °C for 1 h. After incubation, serum was immediately separated by centrifugation (1560 g for 10 min at 4°C) and stored at –80°C until assayed for TXB₂, which reflects platelet COX-1 activity (Patrono et al., 1980) by immunoassay (Cayman Chemical, item#501020) and LC-MS/MS (Patrignani et al., 2014; Tacconelli et al., 2020a). In serum, 12R-HETE, 12S-HETE, 15R-HETE, 15S-HETE, 5R-HETE, 5S-HETE, 8R-HETE, 8S-HETE, 15-epi-LXA_4, and PGE₂ were also assessed by LC-MS/MS as previously reported (Tacconelli et al., 2020a) and briefly described below. The impact of DMSO and triacetin-saccharin on the lipidomics of serum eicosanoids at baseline (in the absence of vehicles) was also evaluated.

Human platelets were freshly isolated from leukocyte concentrates obtained from Centro Trasfusionale, at Renzetti Hospital, Lanciano, Italy, as previously described (Dovizio et al., 2012; Dovizio et al., 2013a). Washed platelets presented only 0.38 ± 0.29% (mean ± SEM) of leukocyte contamination as assessed by flow cytometry (Grande et al., 2019). Platelets (1 × 10⁸) suspended in a final volume of 0.5 ml of HEPES buffer [containing 5 mM HEPES, 137 mM NaCl, 2 mM KCl, 1 mM MgCl₂, 12 mM NaHCO₃, 0.3 mM NaH₂PO₄, and 5.5 mM glucose, pH 7.4, 3.5 mg/ml BSA] were treated with different concentrations of ASA (0.01–300 µM or DMSO) or L-ASA (0.01–300 µM or diluent) for 30 min at room temperature. Then, CaCl₂ 2 mM and MgSO₄ 1 mM were added for 2 min at room temperature, and then platelets were stimulated with AA (10 μM) for 30 min at 37°C. The reaction was stopped by keeping the tubes at 4°C before the centrifugation at 2200 g for 5 min (at 4°C) to separate the platelet pellet from the supernatant, which was collected and centrifuged at 10,000 g for 5 min at 4°C. The supernatant was stored at −80°C. TXB₂ levels were assessed in the supernatants by a specific immunoassay (Patrignani et al., 2014). In the same experimental conditions, the % acetylation of COX-1 was assessed in 2.0 × 10⁷ platelets treated with ASA, L-ASA, DMSO, or diluent by LC-MS/MS as previously described (Patrignani et al., 2014). The cells were lysed and subjected to in-gel digestion using trypsin and GluC (Drapeau et al., 1972; Brown and Wold, 1973). The samples containing the peptides obtained by the proteolysis were dried by vacuum centrifugation and stored at -80 °C until analyzed for the concentrations of endogenously formed IGAPFSLK, its acetylated form, and the AQUA peptides by LC-MS/MS (Patrignani et al., 2014).

HCA7 cells (3.0 × 10⁶) were seeded and cultured in DMEM supplemented with 10% FBS and 1% penicillin-streptomycin for 48 h, then the medium was changed to DMEM supplemented with 0.5% FBS, and 1% penicillin-streptomycin. HCA-7 cells were assessed for COX-1 and COX-2 levels by using LC-MS/MS, as previously described (Patrignani et al., 2014; Tacconelli et al., 2020b). 5-LOX, 15-LOX-1, and platelet type 12-LOX were assessed by Western blot as reported below.

The cells were treated with increasing concentrations of ASA (1–1000 µM), L-ASA (1–1000 µM), or with DMSO or diluent for 60 min at 37 °C. The cells were lysed and subjected to in-gel digestion using trypsin and GluC (Drapeau et al., 1972; Brown and Wold, 1973; Tacconelli et al., 2020b). The samples containing the peptides obtained by the proteolysis were dried by vacuum centrifugation and stored at -80 °C until analyzed for the concentrations of endogenously formed VGAPFSLK, its acetylated form, and the AQUA peptides by LC-MS/MS (Tacconelli et al., 2020b). The effect of ASA or L-ASA on the activity of COX-2 was evaluated by preincubating the cells (1 × 10⁶), cultured as reported above, with increasing concentrations of ASA (1–1,000 µM), L-ASA (1–1,000 µM), DMSO or diluent for 30 min at 37°C; finally, cells were incubated with AA (0.5, 10 and 100 µM) for 30 min at 37°C. At the end of the incubation, conditioned media were harvested and centrifuged at 10,000 g for 5 min to discard cell debris, and the supernatant was stored at -80 °C until assayed for the levels of eicosanoids by using LC-MS/MS (Tacconelli et al., 2020a). The generation of eicosanoids (12R-HETE, 12S-HETE, 15R-HETE, 15S-HETE, 5R-HETE, 5S-HETE, 8R-HETE, 8S-HETE, 15-epi-LXA₄, TXB₂ and PGE₂) was assessed by using LC-MS/MS. Samples were extracted by using a liquid-liquid extraction (Maskrey et al., 2007; Tacconelli et al., 2020a). Briefly, to 1 ml of the sample, 2.5 ml of a mixture of acetic acid/isopropanol/hexane (2:20:30, v/v/v) and internal standards (d₈-12S-HETE, d₈-15S-HETE, d₈-5S-HETE, d₄TXB₂, and d₄-PGE₂; at the final concentration of 5 ng/ml) were added. The extraction was performed by adding 5 ml of n-hexane. Then, the samples were centrifuged at 1500 g at 4°C for 5 min. The dried hexane phases were stored at −80°C until LC-MS/MS analysis. Before analysis, dried lipids were resuspended in 200 μL of methanol and analyzed by LC-MS/MS as previously described (Tacconelli et al., 2020a). The LC-MS/MS system consisted of ACQUITY UPLC I-Class/Xevo TQ-S micro IVD System (Waters) equipped with a Z-Spray ESI source under negative ionization conditions. Deuterated and non-deuterated standards (from Cayman Chemical) were analyzed in MS/MS mode to examine the collision-induced fragmentation spectrum. In Supplementary Table, the list of the molecular ions and fragments monitored for each eicosanoid is shown. Separation of 12R-HETE, 12S-HETE, 15R-HETE, 15S-HETE, 5R-HETE, 5S-HETE, 8R-HETE, 8S-HETE, 15-epi-LXA₄, TXB₂ and PGE₂ was performed using a chiral chromatographic column (Lux 3  μm Amylose-1, 150 mm × 3.0 mm; Phenomenex, Torrance, CA, United States) eluting a 20-min gradient of 50–100% solvent B (60% methanol, 40% ACN, 0.1% glacial acetic acid) and solvent A (75% water, 25% ACN, 0.1% glacial acetic acid): 50% solvent B for 5 min; 50–60% solvent B for 4 min; 60–80% solvent B for 2 min; 80–90% solvent B for 2 min; 90–100% solvent B for 1 min, 100% solvent B for 2 min and 50% of solvent B from 17 to 20 min with a flow rate of 0.2 ml/min). The linear standard curves were obtained by adding constant amounts of internal standards to eight different concentrations of each analyte (0.01–500 ng/ml), then the calibration curves were constructed by linear regression of the ratio of the peak areas of the analytes to the areas of the corresponding internal standards. For 8R- and 8S-HETE and 15-epi-LXA₄, we used d₈-12S-HETE as the internal standard because the deuterated forms are not commercially available. The eicosanoid concentrations were calculated by interpolation from the calculated regression lines. The peptide peak areas were extracted and analyzed by using MassLynx software (Waters, United Kingdom). The data were normalized to sample volume and expressed as ng/ml. The detection limit of quantification of each eicosanoid was 10 pg/ml.

The protein levels of 5-LOX, 15-LOX-1, and platelet-type-12-LOX in HCA7 lysates from HCA7 were determined by Western blot analysis, as previously described (Dovizio et al., 2013b). Briefly, an aliquot of the sample was loaded onto 9% Sodium Dodecyl Sulphate PolyAcrylamide Gel Electrophoresis (SDS-PAGE), transferred to polyvinylidene difluoride (PVDF) membrane (GE Healthcare, Milan, Italy) and blocked with a solution of 5% non-fat milk in tris-buffered saline-0.1% Tween-20 (TBS-Tween-20). The membranes were incubated overnight with primary antibodies diluted in TBS-Tween-20, platelet-type12-LOX (dilution 1:1,000), 5-LOX (dilution 1:50) or 15-LOX-1 (dilution 1:1,000), and β-actin (dilution 1:1,000). Then, the membranes were washed in TBS-Tween-20 and incubated with the secondary antibodies. Quantifying the optical density (OD) of different specific bands was calculated using Alliance 1 D software (UVITEC, Cambridge) and normalized to the OD of β-actin.

The data have been reported as mean ± SD or SEM. The statistical analysis was performed using GraphPad Prism software (version 9.00 for Windows; GraphPad, San Diego, CA). The values of p < 0.05 were considered statistically significant. The concentration-response curves were obtained using PRISM (GraphPad, version 9.00 for Windows). The IC₅₀ (compound concentration to inhibit by 50% eicosanoid biosynthesis) or EC₅₀ (compound concentration to cause 50% of maximal effect) and 95% confidence interval (CI) values of the sigmoidal concentration-response data were obtained by GraphPad Prism software.

We have previously reported the targeted chiral LC-MS/MS analysis of human serum obtained by peripheral whole blood allowed to clot for 1 h at 37°C (Tacconelli et al., 2020a). Under these conditions, thrombin is generated endogenously, and activating the PARs (protease-activated receptors) induces the release of AA from membrane phospholipids which can be transformed enzymatically and non-enzymatically into prostanoids and HETEs. The influence of DMSO (used as a vehicle in ASA experiments) and triacetin/saccharin (used in L-ASA experiments, called diluent) on serum eicosanoid generation at baseline (i.e., in the absence of vehicles) was assessed (Figures 1A–C and Supplementary Figure S1). In serum, 12S-HETE and TXB₂ were the major eicosanoids (67.72% and 23.5%, respectively). They are mainly of platelet origin (Tacconelli et al., 2020a). PGE₂ and other HETEs were minor products. Prostanoids and low levels of 15S- and 15R-HETE originate from COX-1 activity since COX-2 is not expressed in clotting whole blood (Patrignani et al., 1999). 5S-HETE, the product of the 5-LOX, was only 0.09% of all eicosanoids. Moreover, non-enzymatic oxidation of AA led to tiny levels of 12R-HETE, 8R-HETE, 8S-HETE, and 5R-HETE (i.e., 0.99%), as reported before (Mazaleuskaya et al., 2018; Tacconelli et al., 2020a).

DMSO did not significantly affect the generation of these eicosanoids (Supplementary Figure S1). In the presence of the diluent, a slight, significant increase of 12S-HETE and a decrease of 8R-HETE (a marginal product) was detected versus DMSO (Supplementary Figure S1). However, the concentration of the other eicosanoids in serum, including TXB₂, was not significantly influenced by DMSO or the diluent (Supplementary Figure one).

The heatmap representation of the effects of ASA and L-ASA on serum eicosanoids is reported in Figures 1D,E, respectively. ASA and L-ASA caused a concentration-dependent reduction of TXB₂, while 12S-HETE was substantially unaffected (Supplementary Figures S2A,B).

The heatmap of the inhibitory effect of the COX-1 products, TXA₂, PGE₂, 15S- and 15R-HETE by ASA and L-ASA is also shown in Figures 2A,B, respectively. ASA and L-ASA caused a concentration-dependent reduction of the four eicosanoids. These results are also represented as concentration-response sigmoidal curves for the inhibition of eicosanoid production by ASA and L-ASA (Supplementary Figures S2A,B). ASA and L-ASA showed concentration-dependent inhibitory effects on TXB₂, PGE₂, and 15R-HETEs. Variable and less pronounced effects on the generation of the other HETEs was found (Supplementary Figures S2A,B). L-ASA was associated with higher inhibitory potency on whole blood COX-1-derived eicosanoids than ASA (Figures 2A,B, Supplementary Figures S2A,B, Supplementary Table S2).

Due to the relevant role of TXA₂ in platelet function, we compared the concentration-response curves of ASA and L-ASA for inhibition of serum TXB₂ in seven different separate experiments; we show that L-ASA inhibited serum TXB₂ with IC₅₀ values which were 5.8-fold lower than those found with ASA, in a statistically significant fashion. However, at 100–300 μM, ASA and L-ASA caused a comparable and almost virtually complete inhibition of serum TXB₂ generation (Figure 2C).

As reported above, the diluent of L-ASA alone did not affect serum TXB₂ levels versus DMSO and baseline (without any vehicle) (Supplementary Figure S1), suggesting that it did not contribute to the higher potency of L-ASA versus aspirin.

We also assessed 15-epi-LXA₄ in serum at baseline and after ASA or L-ASA, but it was undetectable (<10 pg/ml) (not shown).

We characterized the potency and maximal inhibitory effects of ASA and L-ASA on TXB₂ production in washed human platelets stimulated with 10 μM of AA (Figure 3A). Under these experimental conditions, the IC₅₀ values for ASA and L-ASA were not significantly different (Figure 3A). The two compounds caused comparable concentration-response curves for the acetylation of platelet COX-1 (Figure 3B). The maximal % acetylation of COX-1 by ASA and L-ASA obtained at 100 μM (60.8 ± 6.7 and 61.4 ± 15%, respectively) was slightly enhanced at the higher concentration of 300 μM (69.0 ± 13.0 and 74.4 ± 13.0%, respectively), but the differences were not statistically significant. These data show that in isolated platelets, ASA and L-ASA affect platelet COX-1 activity and TXA₂ biosynthesis with comparable potency and maximal efficacy. The baseline values of TXB₂ in the presence of DMSO or diluent were similar (1,008 ± 68 and 999 ± 143 ng/ml, respectively).

Altogether these data show that in isolated platelets, the two compounds have similar pharmacodynamics, i.e., inhibiting COX-1-dependent TXA₂ generation via the acetylation of COX-1.

The human colonic adenocarcinoma cell line HCA7 was studied because it expresses COX-2, while COX-1 is undetectable. This finding was previously reported using Western blot (Tacconelli et al., 2020b); here, we confirm it using LC-MS/MS (Supplementary Figure S3A). The cells also expressed 15-LOX-1 and 5-LOX, while platelet-type12-LOX was undetectable (Western blot analysis) (Supplementary Figure S3B).

In HCA7 cells, we assessed the generation of prostanoids, HETEs, and 15-epi-LXA₄ using LC-MS/MS in response to physiological concentrations of AA, 0.5 and 10 μM, or the high AA concentration of 100 μM. The biosynthesis of eicosanoids was assessed in the presence of DMSO (Figures 4A,C,E) or the diluent (Figures 4B,D,F). In response to AA 0.5 μM (Figures 4A,B) or 10 μM (Figures 4C,D), PGE₂ represented 76%, and TXB₂ was ∼20–30% of all eicosanoids generated. At 100 μM of AA (Figures 4E,F), 15S-HETE, 15R-HETE, and PGE₂ were the primary products. However, TXB₂, 5S-HETE, 12S-HETE, and non-enzymatically formed HETEs (5R-HETE, 8R-HETE, 8S-HETE, and 12R-HETE) were detected. 15-epi-LXA₄ was undetectable. (<10 pg/ml) (not shown). Comparable results were obtained in the presence of DMSO (vehicle of ASA) or triacetin/saccharin (diluent; vehicle of L-ASA) (Figures 4E,F).

These results suggest that PGE₂ is the main eicosanoid generated in HCA7 cells in response to 0.5 and 10 μM of AA, i.e., at concentrations closer to those released endogenously from cellular membrane phospholipids.

As shown in Figure 5, ASA and L-ASA caused concentration-dependent acetylation of cancer cell COX-2 with comparable potency. However, the maximal % acetylation of COX-2 by ASA and L-ASA was obtained at 1,000 μM (74 ± 5 and 73 ± 3%, respectively). These values resulted significantly (p < 0.01) higher than those found at 100 μM (56 ± 3 and 60 ± 4%, respectively). The two compounds’ % acetylation of COX-2 at 100 μM was not significantly different from the % acetylation of platelet COX-1 (Figure 3B).

In Figure 6, heat maps display the concentration-dependent effects of ASA (panels A, C, E) and L-ASA (panels B, D, F) on eicosanoid generation at 0.5, 10, and 100 μM of AA, respectively.

ASA and L-ASA caused a concentration-dependent reduction of PGE₂ and TXB₂ production at all AA concentrations (Figures 6A,C,E, and Figures 6B,D,F, respectively). AA 100 μM was associated with the production of HETEs, and ASA and L-ASA caused marginal and variable effects (Figures 6E,F, respectively). 15-epi-LXA₄ was undetectable in all experimental conditions (not shown).

The inhibitory effect on PGE₂ biosynthesis in HCA7 cells stimulated with the different concentrations of AA by the increasing concentrations of ASA or L-ASA is reported in Figures 7A–C. The IC₅₀ values of PGE₂ inhibition were not significantly different for ASA and L-ASA at each AA concentration (Figures 7A–C and Supplementary Table S3). Comparable results were obtained for TXB₂ (Supplementary Figures S4A-C and Supplementary Table S3). These findings show the irreversible nature of COX-2 inhibition by the two compounds. If the interaction of the compounds with COX-2 was reversible, we would have seen a decrease in potency as a function of increased AA concentrations.

Figure 8 shows the effect of ASA (panel A) and L-ASA (panel B) on 15R-HETE and 15S-HETE production by HCA7 cells stimulated with AA 100 μM. The two compounds caused variable and not significant effects. 15R-HETE was enhanced 2-fold, not significantly, only by ASA 1 mM, a supratherapeutic concentration (Figure 8A). 15-epi-LXA₄ was undetectable in all conditions, although the cells presented acetylated COX-2 and active 5-LOX; in fact, 5S-HETE was detected in cell supernatant.

As shown in Figures 8C,D, ASA and L-ASA caused variable and marginal effects on the production of other HETEs.