ANALYSIS OF HYDROPEROXIDE-INDUCED TYROSYL RADICALS AND LIPOXYGENASE ACTIVITY IN ASPIRIN-TREATED HUMAN PROSTAGLANDIN-H SYNTHASE-2

A hydroperoxide-induced tyrosyl radical has been proposed as a key cyc looxygenase intermediate for the ''basal'' isoform of prostaglandin H synthase (PGHS-1). In the present study with the ''inducible'' isoform (PGHS-2), hydroperoxide was also found to generate a radical in high yield, a wide singlet at g = 2.0058 (29 G peak to trough). Reaction of PGHS-2 with a tyrosine-modifying reagent, tetranitromethane (TNM), re sulted in cyclooxygenase inactivation and a much narrower radical EPR signal (22 G peak to trough). Addition of a cyclooxygenase inhibitor, nimesulide, similarly resulted in a narrow PGHS-2 radical. In PGHS-1, cyclooxygenase inhibition by tyrosine nitration with TNM or by active site ligands leads to generation of a narrow EPR instead of a wide EPR , with both signals originating from authentic tyrosyl radicals, indic ating that the hydroperoxide-induced radicals in PGHS-2 are also tyros yl radicals. Treatment of PGHS-2 with aspirin (acetyl salicylic acid, ASA) was previously shown to result in acetylation of a specific serin e residue, cyclooxygenase inhibition, and increased lipoxygenase activ ity. Acetylation of PGHS-1 by ASA, in contrast, inhibited both lipoxyg enase and cyclooxygenase activity. We now have found the ASA-treated P GHS-2 radical to be indistinguishable from that in control PGHS-2. Add ition of nimesulide to ASA-treated PGHS-2 inhibited the lipoxygenase a nd resulted in a narrow radical EPR like that seen in PGHS-2 treated w ith TNM or nimesulide alone. Retention of PGHS-2 oxygenase activity wa s thus associated with retention of the native radical, and loss of ac tivity was associated with alteration of the radical. Both native and ASA-treated PGHS-2 produced only the R stereoisomer of 11- and 15-HETE , demonstrating that the Lipoxygenase stereochemistry was not changed by ASA. Native and ASA-treated PGHS-2 had lipoxygenase K-m values cons iderably higher than that of the control PGHS-2 cyclooxygenase. Taken together, these results suggest that the same PGHS-2 tyrosyl radical s erves as the oxidant for both cyclooxygenase and lipoxygenase catalysi s and that acetylation of PGHS-2 by ASA favors arachidonate binding in an altered conformation which results in abstraction of the pro-R hyd rogen from C13 and formation of 11(R)- and 15(R)-HETE.