Roles of calcium-activated and voltage-gated delayed rectifier potassium channels in endothelium-dependent vasorelaxation of the rabbit middle cerebral artery.

1. The cellular mechanism(s) of action of endothelium-derived vasodilator substances in the rabbit middle cerebral artery (RMCA) were investigated. Specifically, the subtypes of potassium channels involved in the effects of endothelium-derived relaxing factors (EDRFs) in acetylcholine (ACh)-induced endothelium-dependent vasorelaxation in this vessel were systematically compared. 2. In the endothelium-intact RMCA precontracted with histamine (3 microM), ACh induced a concentration-dependent vasorelaxation, which was sensitive to indomethacin (10 microM) or N(G)-nitro-L-arginine (L-NOARG; 100 microM); pD2 values 8.36 vs 7.40 and 6.38, P < 0.01 for both, n = 6 and abolished by a combination of both agents. ACh caused relaxation in the presence of high K+ PSS (40 mM KCl), which was not affected by indomethacin, but abolished by L-NOARG and a combination of indomethacin and L-NOARG. 3. In the presence of indomethacin, relaxation to ACh in the endothelium-intact RMCA precontracted with histamine was unaffected by either glibenclamide (10 microM), an ATP-sensitive K+ channel (K[ATP]) blocker, 4-aminopyridine (4-AP, 1 mM) or dendrotoxin (DTX, 0.1 microM), delayed rectifier K channel (Kv) blockers. However, relaxation responses to ACh were significantly inhibited by either LY83583 (10 microM) and 1H-[1,2,4]oxadiazolo[4,3,-a]quinoxalin-1-one (ODQ, 10 microM), guanylyl cyclase inhibitors, or charybdotoxin (CTX; 0.1 microM), iberiotoxin (ITX, 0.1 microM) and apamin (APA, 0.1 microM), large conductance Ca2+-activated K+ channels (BK[Ca]) blocker and small conductance Ca2+-activated K+ channel (SK[Ca]) blocker, respectively. 4. In the presence of L-NOARG, relaxation to ACh was unaffected by glibenclamide or the cytochrome P450 mono-oxygenase inhibitor, clotrimazole (1 microM), but was significantly inhibited by either 9-(tetrahydro-2-furanyl)-9H-purin-6-amine (SQ 22,536, 10 microM) and 2',3'-dideoxyadenosine (2',3'-DDA, 30 microM), adenylyl cyclase inhibitors, or 4-AP, DTX, CTX, ITX and APA. 5. In the endothelium-denuded RMCA precontracted with histamine, authentic NO-induced relaxation was unaffected by glibenclamide, 4-AP and DTX, but significantly reduced by ODQ, ITX and APA. Authentic prostaglandin I2 (PGI2)-induced relaxation was unaffected by glibenclamide, but significantly reduced by 2',3'-DDA, 4-AP, DTX, ITX and APA. Forskolin-induced relaxation was significantly inhibited by high K+, CTX and 4-AP. 6. These results indicate that: (1) in the RMCA the EDRFs released by ACh are NO and a prostanoid (presumably PGI2), and there is no evidence for the release of a non-NO/PGI2 endothelium-derived hyperpolarizing factor (EDHF), (2) K(Ca) channels are involved in NO-mediated relaxation of the RMCA but both K(Ca) and Kv channels are involved in PGI2-mediated relaxation.