Pharmacological manipulation of cyclo-oxygenase-2 in the inflamed hydronephrotic kidney.

1. Bradykinin (BK, 1 microgram) caused a small (2 fold at 6 h) increase in prostaglandin E2 (PGE2) in the normal rabbit kidney, perfused ex vivo. This was exaggerated (6 fold at 6 h) in the hydronephrotic kidney (HNK). The exaggerated release of PGE2 was attenuated by cycloheximide, an inhibitor of protein synthesis or by dexamethasone, a steroid known to inhibit the induction of cyclo-oxygenase (COX-2). BK (1 microgram) when injected at 6 h of perfusion increased the release of PGE2 from 90 +/- 33 pg ml-1 min-1 to 3069 +/- 946 pg ml-1 min-1. This was reduced to 200 +/- 30 pg ml-1 min-1 in kidneys infused with cycloheximide (1 microM) and to 250 +/- 40 pg ml-1 min-1 in kidneys infused with dexamethasone (n = 8). 2. When tested on human and murine recombinant COX-1 and COX-2 enzymes, DuP-697 was at least 50 fold more selective for COX-2 than for COX-1. 3. DuP-697 reduced the exaggerated release of PGE2 elicited by BK in the HNK (e.g., at 6 h of perfusion BK-evoked PGE2 release decreased from 3069 +/- 946 pg ml-1 min-1 to 187 +/- 22 pg ml-1 min-1 after perfusion with 1 microM DUP-697, n = 8). 4. Cycloheximide, dexamethasone or DuP-697 at doses used to inhibit completely the exaggerated release of PGE2 in the hydronephrotic kidney, failed to inhibit the release of PGE2 elicited by the injection of BK (1 microgram) in the normal contralateral kidney. 5. Indomethacin (1 microM), a non-selective COX-1 and COX-2 inhibitor, completely inhibited PGE2 release in the normal contralateral as well as in the hydronephrotic kidney. 6. We suggest that renal prostaglandin production in the normal kidney is driven by the activity of constitutive COX-1 while at sites of inflammation, such as the hydronephrotic kidney, there is induction of COX-2 that can be blocked selectively by anti-inflammatory glucocorticoids or selective COX-2 inhibitors.