Asymmetric acetylation of the cyclooxygenase-2 homodimer by aspirin and its effects on the oxygenation of arachidonic, eicosapentaenoic, and docosahexaenoic acids

Abstract

Prostaglandin endoperoxide H synthases (PGHS)-1 and -2, also called cyclooxygenases, convert arachidonic acid (AA) to prostaglandin H(2) (PGH(2)) in the committed step of prostaglandin biosynthesis. Both enzymes are homodimers, but the monomers often behave asymmetrically as conformational heterodimers during catalysis and inhibition. Here we report that aspirin maximally acetylates one monomer of human (hu) PGHS-2. The acetylated monomer of aspirin-treated huPGHS-2 forms 15-hydroperoxyeicosatetraenoic acid from AA, whereas the nonacetylated partner monomer forms mainly PGH(2) but only at 15 to 20% of the rate of native huPGHS-2. These latter conclusions are based on the findings that the nonsteroidal anti-inflammatory drug diclofenac binds a single monomer of native huPGHS-2, having an unmodified Ser530 to inhibit the enzyme, and that diclofenac inhibits PGH(2) but not 15-hydroperoxyeicosatraenoic acid formation by acetylated huPGHS-2. The 18R- and 17R-resolvins putatively involved in resolution of inflammation are reportedly formed via aspirin-acetylated PGHS-2 from eicosapentaenoic acid and docosahexaenoic acid, respectively, so we also characterized the oxygenation of these omega-3 fatty acids by aspirin-treated huPGHS-2. Our in vitro studies suggest that 18R- and 17R-resolvins could be formed only at low rates corresponding to less than 1 and 5%, respectively, of the rates of formation of PGH(2) by native PGHS-2.

Fig. 1.

Time-dependent inhibition of native huPGHS-2 by diclofenac and indomethacin. Purified huPGHS-2 (2 μM) was pretreated with no inhibitor (a, b, and c) for 10 min at 37°C or 12 μM diclofenac (d) for 10 min at 37°C or 100 μM indomethacin (e) for 5 min at 37°C. No inhibitor was added to the O₂ electrode assay chamber when the sample (a) was assayed. Diclofenac (0.16 μM) was included in the assay chamber when the sample (b) was assayed; sample b served as the negative control for sample d. Indomethacin (1.3 μM) was included in the assay chamber when sample c was assayed; sample c served as the negative control for sample e. Dilution of the aliquot of sample d yielded a final concentration of 0.16 μM diclofenac in the assay chamber, and no additional diclofenac was added to this assay chamber. Dilution of the aliquot of sample e yielded a final concentration of 1.3 μM indomethacin in the assay chamber, and no additional indomethacin was added to this assay chamber. Assays were performed at 37°C in 3 ml of 0.1 M Tris-HCl, pH 8.0, containing 100 μM arachidonic acid, 1 mM phenol, and 1 μM hematin. B, purified huPGHS-2 (2.1 μM) was pretreated with the indicated concentrations of diclofenac at 37°C for 10 min and then assayed for COX activity with 100 μM AA essentially as described above. Error bars show standard deviations from multiple kinetic trials.

Fig. 2.

Inhibition of huPGHS-2 by aspirin. A, purified native huPGHS-2 homodimer was incubated at 24°C with or without freshly prepared aspirin (500 μM) for the indicated times and then assayed for COX activity using an O₂ electrode. Values are derived from the average of triplicate determinations ± S.E. Similar experiments were performed at least three times with different preparations of enzyme and yielded quantitatively similar results. B, purified native huPGHS-2 was pretreated with 500 μM aspirin for 1 h at 37°C, which causes maximal inhibition. The sample was then incubated without (E + ASA) or with (E + ASA + Diclo) 12.5 μM diclofenac (Diclo) for 10 min and assayed for COX activity using an O₂ electrode. For the samples pretreated with diclofenac for 10 min, the assay mixture also contained 12.5 μM diclofenac. A control with huPGHS-2 without inhibitor (E) was treated for 1 h at 37°C with vehicle (in place of aspirin) and then incubated for an additional 10 min without any inhibitor. C, purified native huPGHS-2 was incubated with or without ASA (500 μM) at 37°C for 1 h, and the oxygenation of [1-¹⁴C]AA was assayed by radio thin-layer chromatography in the presence and absence of 12.5 μM Diclo as described under Materials and Methods. Thin-layer chromatography was used to separate the endoperoxide (PGH₂) and monohydroxy acid products, mainly 15-HETE. To perform the assays, [1-¹⁴C]AA (100 μM) was mixed with ∼9 μg of native huPGHS-2 in an assay volume of 0.10 ml, and the reactions were allowed to proceed for 40 s. The reactions were stopped, and the products were extracted, separated, and visualized by autoradiography as shown. The thin-layer plates were subsequently scraped, and the amount of radioactivity cochromatographing with AA, PGH₂, and 15-HETE standards was determined by scintillation counting. Numbers obtained from scintillation counting were used to calculate the percentage of total radioactivity found in each product. The experiment was performed three times with consistent results.

Fig. 3.

Products formed from EPA by aspirin-acetylated huPGHS-2. EPA (100 μM) was mixed with 9 μg of enzyme, and the reactions continued for 40 s. The products were extracted as described in Fig. 2C and reconstituted with 50% methanol for LC-MS/MS analysis.

Fig. 4.

Products formed from DHA by aspirin-acetylated huPGHS-2. DHA (100 μM) was mixed with 9 μg of enzyme, and the reactions continued for 40 s. The products were extracted as described in Fig. 2C and reconstituted with 50% methanol for LC-MS/MS analysis.