Coronary autoregulation is an important physiological

Transcription

1 Hydrogen Peroxide, an Endogenous Endothelium-Derived Hyperpolarizing Factor, Plays an Important Role in Coronary Autoregulation In Vivo Toyotaka Yada, MD; Hiroaki Shimokawa, MD; Osamu Hiramatsu, PhD; Tatsuya Kajita, MD; Fumiyuki Shigeto, MD; Masami Goto, MD; Yasuo Ogasawara, PhD; Fumihiko Kajiya, MD, PhD Background Recent studies in vitro have demonstrated that endothelium-derived hydrogen peroxide (H 2 O 2 ) is an endothelium-derived hyperpolarizing factor (EDHF) in animals and humans. The aim of this study was to evaluate our hypothesis that endothelium-derived H 2 O 2 is an EDHF in vivo and plays an important role in coronary autoregulation. Methods and Results To test this hypothesis, we evaluated vasodilator responses of canine (n 41) subepicardial small coronary arteries ( 100 m) and arterioles ( 100 m) with an intravital microscope in response to acetylcholine and to a stepwise reduction in coronary perfusion pressure (from 100 to 30 mm Hg) before and after inhibition of NO synthesis with N G -monomethyl-l-arginine (L-NMMA). After L-NMMA, the coronary vasodilator responses were attenuated primarily in small arteries, whereas combined infusion of L-NMMA plus catalase (an enzyme that selectively dismutates H 2 O 2 into water and oxygen) or tetraethylammonium (TEA, an inhibitor of large-conductance K Ca channels) attenuated the vasodilator responses of coronary arteries of both sizes. Residual arteriolar dilation after L-NMMA plus catalase or TEA was largely attenuated by 8-sulfophenyltheophylline, an adenosine receptor inhibitor. Conclusions These results suggest that H 2 O 2 is an endogenous EDHF in vivo and plays an important role in coronary autoregulation in cooperation with NO and adenosine. (Circulation. 2003;107: ) Key Words: endothelium-derived factors microcirculation ischemia nitric oxide adenosine Coronary autoregulation is an important physiological compensatory mechanism that permits the myocardium to relax when coronary perfusion pressure is decreased. 1 Nitric oxide (NO) and adenosine are known to be involved in ischemic autoregulation during low perfusion pressure. 1,2 Vascular endothelial cells play an important role in modulating vascular tone by releasing at least 3 vasodilator factors, including NO, prostacyclin (PGI 2 ), and endotheliumderived hyperpolarizing factor (EDHF). 3,4 NO synthase inhibitors largely suppress acetylcholine (ACh)-induced dilation of human 5 and canine 6 epicardial coronary arterioles in vivo, indicating that EDHF may not be involved in the ACh-induced vasodilation of large coronary arteries in vivo. Conversely, in the coronary microcirculation, inhibition of NO synthase does not abolish ACh-induced vasodilation in dogs. 7 Residual dilation to ACh after inhibition of NO synthase and cyclooxygenase has been attributed to EDHF. However, the nature of EDHF has been controversial since the first report on its existence. 3,4 The natures of EDHF that have been proposed include cytochrome P-450 metabolites, 8,9 endothelium-derived K, 10 and electrical communications through gap junctions between endothelial cells and vascular smooth muscle cells. 11 Recently, Matoba et al 12,13 established that endotheliumderived hydrogen peroxide (H 2 O 2 ) is a primary EDHF in mice and humans. Indeed, H 2 O 2 is known to activate calcium-activated K channels (K Ca ) 14 and to cause hyperpolarizations of vascular smooth muscle cells. Blockade of K Ca channels inhibits the adenosine-induced coronary vasodilation. 15 However, it remains to be examined whether H 2 O 2 is an EDHF in the coronary circulation in vivo, and if so, whether H 2 O 2 contributes to autoregulation as a compensatory mechanism for NO and adenosine. The aim of the present study was to elucidate those important issues in dogs. The results demonstrated that H 2 O 2 is a primary EDHF in vivo and plays an important role in coronary autoregulation. Methods Animal Preparation This study conformed to the Guideline on Animal Experiments of Kawasaki Medical School and the Guide for the Care and Use of Received September 10, 2002; revision received November 5, 2002; accepted November 6, From the Departments of Medical Engineering and Systems Cardiology, Kawasaki Medical School, Kurashiki (T.Y., O.H., T.K., F.S., M.G., Y.O.); Cardiovascular Medicine, Kyushu University Graduate School of Medical Sciences, Fukuoka (H.S.); and Cardiovascular Physiology, Okayama University Graduate School of Medicine and Dentistry, Okayama (F.K.), Japan. Correspondence to Toyotaka Yada, MD, PhD, Kawasaki Medical School, 577 Matsushima, Kurashiki, Okayama, , Japan ( yada@me.kawasaki-m.ac.jp); reprint requests to Fumihiko Kajiya, MD, PhD, Okayama University Graduate School of Medicine and Dentistry, Shikatacho, Okayama, , Japan American Heart Association, Inc. Circulation is available at DOI: /01.CIR

2 Yada et al H 2 O 2 and Coronary Autoregulation 1041 Laboratory Animals published by the US National Institutes of Health. Mongrel dogs (n 41, 10 to 25 kg) of either sex were anesthetized with morphine (3 mg/kg IM) and sodium pentobarbital (25 mg/kg IV). After intubation, each animal was ventilated with a highfrequency jet ventilator (model VS600, IDC) with room air supplemented by 100% oxygen. Aortic pressure and left ventricular pressure were measured with an 8F pigtail double manometer catheter (SPC-784A, Millar). Heart rate was kept constant at 100 bpm by right ventricular pacing after atrioventricular node blocking by 40% formaldehyde. 16 Needle-Probe Intravital Microscope Briefly, the needle probe (4.5 mm in diameter) contains a gradient index lens surrounded by light guide fibers and a double-lumen sheath. A doughnut-shaped balloon on the tip avoids direct compression of the vessels by the needle tip. 16 Measurements of Arteriolar Diameters We placed the needle probe gently on the subepicardial microvessel. When a clear vascular image was obtained, the end-diastolic vascular images were taken at a rate of 30 pictures per second. 16 Coronary Sinus Cannulation A Sones catheter was inserted into the right external jugular vein and advanced into the coronary sinus. Blood samples drawn from the coronary sinus catheter were analyzed for plasma adenosine concentration. 17 Lactate Measurements Arterial and coronary venous lactate samples were drawn into syringes. Lactate concentration was measured with a YSI 2300 Stat Plus model lactate analyzer. 17 Myocardial percent lactate extraction was calculated as (arterial venous)/arterial values 100 (%). Plasma Adenosine Measurements The samples were then concentrated by evaporation and resuspended in 50 L of high-performance liquid chromatography buffer. 18 The adenosine in each sample was separated on a Shimadzu LC10 high-performance liquid chromatograph with a C-18 column using an ion-pairing buffer solution of tetrabutylammonium hydrogen sulfate and potassium phosphate with an acetonitrile gradient. Application of System for Controlled Perfusion of Coronary Arteries To manipulate coronary arterial pressure, the heart was perfused with blood from the left femoral artery. 19 An in-line flow probe (Transonic 4-Fr, connected to a T206 Transonic flowmeter) just proximal to the Gregg cannula in the left main coronary artery was used to measure phasic coronary blood flow. Coronary perfusion pressure was measured in the first diagonal branch of the left anterior descending coronary artery. Experimental Protocols After the surgical procedure and instrumentation, at least 30 minutes were allowed for stabilization while hemodynamic variables were monitored. The following protocols were examined. (1) ACh (1.0 g/kg IC for 2 minutes)-induced, EDHF-mediated coronary vasodilation was evaluated before and after inhibition of NO synthase (N G -monomethyl-l-arginine [L-NMMA], 2 mol/min for 20 minutes) 20 with cyclooxygenase blockade (ibuprofen, 12.5 mg/kg IV, an inhibitor of the synthesis of vasodilator prostaglandins to evaluate the role of EDHF and NO alone without PGI 2 ) and catalase ( U/kg IV and U kg 1 min 1 IC for 10 minutes, an enzyme that selectively dismutates H 2 O 2 into water and oxygen) 21 or tetraethylammonium (TEA, 10 g kg 1 min 1 IC for 10 minutes, an inhibitor of large-conductance K Ca channels). 22 (2) Coronary perfusion pressure was changed in a stepwise manner from 100 to 30 mm Hg before and after L-NMMA plus Figure 1. Role of H 2 O 2, an endogenous EDHF, in endotheliumdependent vasodilation of canine coronary microvessels in vivo. L-NMMA (in presence of ibuprofen) attenuated vasodilator responses to ACh primarily in small arteries (left), whereas L-NMMA followed by catalase attenuated vasodilation of both small arteries and arterioles (right). Number of vessels per animal used was 10/4 (small arteries) and 9/5 (arterioles). *P 0.05, **P 0.01 vs control; #P 0.05 vs L-NMMA. ibuprofen. Coronary perfusion pressure was slowly reduced ( 1 minute) to the next lower level. (3) To determine the nature of EDHF and the mechanism of EDHF-mediated vasodilation, additional experiments were performed in coronary subepicardial arterioles with a CCD intravital microscope. The inhibitors TEA and catalase were used. In additional experiments, feedback vasodilator responses during coronary autoregulation were examined before and after catalase, and after catalase followed by L-NMMA compared with L-NMMA followed by catalase. (4) To evaluate the compensatory effects of adenosine, coronary venous blood samples were drawn. To evaluate the interaction among EDHF, NO, and adenosine, we also evaluated vasodilator responses after TEA or catalase with L-NMMA followed by adenosine receptor blockade (8-sulfophenyltheophylline, 8-SPT, 25 g kg 1 min 1 IC for 5 minutes). 22 All drugs were obtained from Sigma Chemical Co. Statistical Analysis Results are expressed as mean SEM. Vascular and coronary blood flow responses were analyzed by 2-way ANOVA followed by Scheffé s post hoc test. Student s t test was used for both paired and unpaired comparisons. The criterion for statistical significance was a value of P Results Role of EDHF in Endothelium-Dependent Vasodilation of Coronary Microvessels In Vivo The vasodilator responses of small coronary arteries ( 100 m) to ACh were significantly attenuated by L-NMMA (in the presence of ibuprofen) and were further attenuated by catalase (Figure 1, left) or TEA. By contrast, the vasodilator responses of arterioles ( 100 m) were relatively resistant to L-NMMA (in the presence of ibuprofen) but were markedly attenuated by catalase (Figure 1, right) or TEA. We confirmed that the inhibitory effects of catalase were comparable to those of TEA (data not shown). Hemodynamics and Blood Gases During Decreasing Coronary Perfusion Pressure In each experimental condition, mean aortic pressure at baseline was constant and comparable (control, 72 7 mm Hg; L-NMMA, 74 8 mm Hg; and L-NMMA

3 1042 Circulation February 25, 2003 Figure 2. Transmural coronary blood flow during decreasing perfusion pressure. Coronary blood flow was reduced significantly at 50 and 30 mm Hg of perfusion pressure in all 3 conditions. Number of animals used was 7. **P 0.01 vs 100 mm Hg. followed by catalase, 72 6 mm Hg). PO 2,PCO 2, and ph were maintained within the physiological range (ph, 7.35 to 7.45; PCO 2,25to40mmHg;PO 2, 70 mm Hg) throughout the experiments. Transmural coronary blood flow at 70 mm Hg of perfusion pressure was comparable among the 3 conditions (Figure 2). However, the flow was significantly reduced to a comparable extent at 50 and 30 mm Hg of perfusion pressure in all 3 conditions (Figure 2). Role of EDHF in Coronary Autoregulation Under control conditions, coronary autoregulatory vasodilator responses to decreasing perfusion pressure were noted in arterioles but not in small arteries (Figure 3, left). In coronary arterioles, the autoregulatory vasodilator responses were significantly attenuated by L-NMMA (in the presence of ibuprofen) and were further attenuated by catalase (Figure 3, right). Compared with the responses after L-NMMA alone (Figure 4A), L-NMMA followed by TEA (Figure 4B) or catalase (Figure 4C) caused a comparable extent of inhibition on the responses of arterioles. The arteriolar responses were also comparable after L-NMMA followed by catalase (Figure Figure 3. Coronary microvascular responses during coronary autoregulation. Under control conditions (n 7), autoregulatory vasodilator responses to decreasing coronary perfusion pressure were noted in arterioles (right, 12 vessels) but not in small arteries (left, 9 vessels). Arteriolar vasodilation was attenuated significantly by L-NMMA (in presence of ibuprofen) and was further attenuated by catalase. **P 0.01 vs control; ##P 0.01 vs L-NMMA. Figure 4. Inhibitory effect of catalase or TEA during coronary autoregulation. Both inhibitors caused a comparable extent of inhibition of arteriolar vasodilation (B and C) vs L-NMMA alone (A). Arteriolar vasodilation was comparable after L-NMMA followed by catalase (C) and after catalase followed by L-NMMA (D). Number of vessels per animal used was 12/6 for L-NMMA followed by catalase, 11/8 for L-NMMA followed by TEA, and 8/3 for catalase followed by L-NMMA. *P 0.05, **P 0.01 vs 70 mm Hg of coronary perfusion pressure. 4C) and after catalase followed by L-NMMA (Figure 4D) irrespective of the order of drug administration. The arteriolar responses at 30 mm Hg of coronary perfusion pressure were significantly reduced after L-NMMA, catalase, or TEA alone and were further decreased after L-NMMA followed by catalase, catalase followed by L-NMMA, or TEA followed by L-NMMA with decreased myocardial lactate extraction (Figure 5 and the Table). Compensatory Effects of Adenosine in Coronary Autoregulation Coronary venous adenosine concentrations were increased in response to decreasing coronary perfusion pressure after L-NMMA alone (Figure 6A), L-NMMA followed by catalase (Figure 6A), catalase followed by L-NMMA (Figure 6B), and TEA followed by L-NMMA (Figure 6C) but were not further increased after catalase or TEA alone (Figure 6, B and C). The blockade of adenosine receptor with 8-SPT suppressed the residual EDHF-mediated arteriolar dilation after L-NMMA followed by catalase at 30 mm Hg of coronary perfusion pressure with decreased myocardial lactate extraction (Figure 7A, Table). Coronary venous adenosine concentrations were increased in response to decreasing perfusion pressure after L-NMMA but were not further increased after L-NMMA followed by catalase (Figure 7B) or TEA. This was also the case with L-NMMA followed by TEA and 8-SPT (data not shown). Discussion The major findings of the present study are that (1) H 2 O 2 is a primary EDHF in the canine coronary circulation in vivo and (2) H 2 O 2 plays an important role in coronary autoregulation as a compensatory mechanism for NO and adenosine. To the best of our knowledge, this is the first report that demonstrates the importance of H 2 O 2 as an endogenous EDHF that plays an important role in coronary autoregulation in vivo.

4 Yada et al H 2 O 2 and Coronary Autoregulation 1043 Figure 5. Coronary arteriolar responses during low perfusion pressure as a function of vessel size. Arteriolar responses at 30 mm Hg of coronary perfusion pressure were significantly reduced after L-NMMA (A), catalase (B), or TEA (C) alone and were further decreased after L-NMMA followed by catalase (A), catalase followed by L-NMMA (B), or TEA followed by L-NMMA (C). Number of vessels per animal used was 12/6 (A), 8/3 (B), and 7/4 (C). **P 0.01 vs control; ##P 0.01 vs L-NMMA, catalase, or TEA. Critique of Experimental Model and Methodology On the basis of the previous reports, we chose adequate doses of L-NMMA, catalase, TEA, and 8-SPT to inhibit NO, H 2 O 2,K Ca, and adenosine, respectively. The methodological validity of the present study was confirmed previously. 16 H 2 O 2 as an Endogenous EDHF in the Coronary Circulation In Vivo Recently, Matoba et al 12,13 showed that H 2 O 2 is a primary EDHF in mesenteric arteries of mice and humans. Indeed, H 2 O 2 is a reasonable candidate for endogenous EDHF, because it is produced by endothelial cells and causes vascular smooth muscle relaxation through activation of K Ca channels. 12 It is conceivable that H 2 O 2 is produced from superoxide anions that are derived from several sources in endothelial cells, including endothelial nitric oxide synthase, cyclooxygenase, lipoxygenase, cytochrome P-450 enzymes, and NAD(P)H oxidases. 12 In the present study, inhibition of NO synthesis did not affect ACh-induced vasodilation of coronary arterioles, whereas catalase markedly attenuated the vasodilation in vivo (Figure 1). Furthermore, catalase inhibited residual arteriolar dilation after inhibition of NO synthase during coronary autoregulation, indicating that H 2 O 2 functions as a primary EDHF in canine coronary microvessels in vivo. Furthermore, in the present study, EDHF/H 2 O 2 - mediated vasodilation was greater in arterioles than in small arteries, supporting the notion that the importance of EDHF increases as vessel size decreases A previous study using miconazole suggested a possible role of cytochrome P-450 metabolites in EDHF-mediated Myocardial Lactate Extraction Rate (%) at 30 mm Hg of Coronary Perfusion Pressure Protocol n vasodilation in dogs in vivo. 25 However, miconazole inhibits intermediate-conductance K Ca channels and thus directly inhibits hyperpolarization of vascular smooth muscle. 26 Furthermore, inhibitors of cytochrome P-450 metabolites had no inhibitory effect on the EDHF-mediated vasodilation in mice or humans in vitro. 12,13 Indeed, in the present study, EDHFmediated vasodilation of coronary arterioles was markedly inhibited by catalase, indicating that H 2 O 2 is a primary EDHF in the canine coronary circulation in vivo. Compensatory Feedback Mechanism Among EDHF, NO, and Adenosine It is well known that coronary artery tone is regulated by the interactions among several vasodilators, including NO, adenosine, and EDHF. 27,28 These relaxing factors may play a crucial role in causing vasodilation of coronary microvessels in a cooperative manner. In the present study, the arteriolar vasodilation during coronary autoregulation was not completely inhibited by L-NMMA plus ibuprofen or by catalase, and the residual arteriolar dilation was further inhibited after administration of all 3 inhibitors. These results indicate the negative feedback interaction between NO and EDHF during coronary autoregulation in vivo. The compensatory effect of adenosine may also be important. Stepp et al 1 suggested that adenosine plays an important role in the transition to ischemia at a coronary perfusion pressure of 50 mm Hg after K ATP channel blockade with glibenclamide. In the present study, coronary venous adenosine concentrations were increased at a perfusion pressure of 50 mm Hg within the vasoactive range for endogenous adenosine, as previously reported. 29 Control Response First Inhibitor Second Inhibitor Third Inhibitor L-NMMA 23 4* L-NMMA TEA 14 4* TEA 30 1* TEA L-NMMA 20 4* L-NMMA 21 5* L-NMMA catalase 12 3* Catalase 22 1* Catalase L-NMMA 15 1* L-NMMA 23 1* L-NMMA TEA 13 1* L-NMMA TEA 8SPT 2 7* L-NMMA 17 1* L-NMMA catalase 12 1* L-NMMA catalase 8SPT 2 2* Results are expressed as mean SEM. *P 0.05 vs control; P 0.05 vs L-NMMA; P 0.05 vs TEA; P 0.05 vs catalase; P 0.05 vs L-NMMA plus TEA; P 0.05 vs L-NMMA plus catalase.

5 1044 Circulation February 25, 2003 Figure 6. Coronary venous adenosine concentrations and coronary autoregulation. Coronary venous adenosine concentrations were increased in response to decreasing coronary perfusion pressure after L-NMMA alone (A), L-NMMA followed by catalase (A), catalase followed by L-NMMA (B), and TEA followed by L-NMMA (C) but were not further increased after catalase or TEA alone (B and C) compared with control. Number of animals used was 6 (A), 3 (B), and 4 (C). **P 0.01 vs control; ##P 0.01 vs catalase, TEA, or L-NMMA followed by catalase. The residual arteriolar dilation ( 5%) after combined administration of ibuprofen, L-NMMA, and TEA or catalase was completely blocked by 8-SPT, indicating that an increased concentration of adenosine compensated for the loss of action of NO and EDHF. It remains to be examined why the adenosine concentrations did not increase further after administration of L-NMMA followed by TEA compared with L-NMMA alone. Cabell et al 15 reported that TEA inhibited adenosine-induced vasodilation of canine subepicardial coronary arteries in vitro. Furthermore, Gao and Vanhoutte 30 reported that H 2 O 2 relaxed canine bronchial smooth muscle and elevated camp concentrations. Chaytor et al 31 reported that EDHF-mediated relaxation to ACh was associated with an increase in smooth muscle camp concentrations. These findings suggest that a camp-mediated pathway is involved, at least in part, in the autoregulatory coronary vasodilation through large-conductance K Ca channels. Figure 7. Comparison of coronary arteriolar responses and venous adenosine concentrations during coronary autoregulation after L-NMMA and catalase with 8-SPT. Residual EDHFmediated arteriolar dilation after L-NMMA plus catalase was abolished by 8-SPT at 30 mm Hg of perfusion pressure (A). Coronary venous adenosine concentrations were not increased further in response to decreasing perfusion pressure after L-NMMA followed by catalase (B) with 8-SPT vs control. Number of vessels per animal used was 10/4 (A and B). *P 0.05, **P 0.01 vs control; #P 0.05, ##P 0.01 vs L-NMMA; P 0.01 vs L-NMMA followed by catalase. Role of EDHF in Myocardial Ischemia During Coronary Autoregulation K Ca channels contribute substantially to the coronary vasodilation in myocardial ischemia. 22 However, it remains to be examined whether EDHF contributes to autoregulatory coronary vasodilation. The present results demonstrate that EDHF/H 2 O 2 contributes substantially to the vasodilation during coronary autoregulation in vivo. Regarding the mechanism involved in the K Ca channel opening during coronary autoregulation, cellular acidosis 32 and increase in intracellular Ca 2 concentration after ischemia 33 have been postulated to open K Ca channels as a compensatory mechanism after inhibition of NO synthesis. We have previously demonstrated that subendocardial arteriolar dilation during reactive hyperemia is more sensitive to L-NMMA than subepicardial arteriolar dilation. 27 These findings indicate that the perfusion of the subendocardium is more dependent on NO than that of the subepicardium. EDHF/H 2 O 2 may compensate coronary blood flow, especially in the endocardium, in myocardial ischemia during coronary autoregulation. Clinical Implications and Conclusions The synthesis and action of endothelium-derived NO are impaired under various pathological conditions, such as hypertension and hyperlipidemia. 34 Previous studies suggested that hypertension causes a compensatory increase in the activity of potassium channels. 35 Cosentino et al 36 reported that in a transgenic mouse model of hyperphenylalaninemia, reduction in arterial tetrahydrobiopterin, a cofactor of endothelial NO synthase, was accompanied by a decrease in endothelial NO production, whereas H 2 O 2 production was increased. In the present study, NO and EDHF/H 2 O 2 apparently compensated each other to cause coronary autoregulatory vasodilation in vivo. These results may represent a negative feedback interaction between the 2 relaxing factors. In conclusion, we were able to demonstrate that H 2 O 2 is a primary EDHF in the canine coronary circulation in vivo and plays an important role in coronary autoregulation as a compensatory feedback mechanism for NO and adenosine. These findings may have important clinical implications, because hyperpolarizing mechanisms contribute substantially to endothelium-dependent vasodilation in myocardial ischemia. Acknowledgments This work was supported in part by grants and from the Japanese Ministry of Education, Culture, Sports, Science, and Technology, Tokyo, Japan.

6 Yada et al H 2 O 2 and Coronary Autoregulation 1045 References 1. Stepp DW, Kroll K, Feigl EO. K ATP channels and adenosine are not necessary for coronary autoregulation. Am J Physiol. 1997;273: H1299 H Smith TP, Canty JM Jr. Modulation of coronary autoregulatory responses by nitric oxide: evidence for flow-dependent resistance adjustments in conscious dogs. Circ Res. 1993;73: Feletou M, Vanhoutte PM. Endothelium-dependent hyperpolarization of canine smooth muscle. Br J Pharmacol. 1988;93: Chen G, Suzuki H, Weston AH. Acetylcholine releases endotheliumderived hyperpolarizing factor and EDRF from blood vessels. Br J Pharmacol. 1988;95: Lefroy DC, Crake T, Uren NG, et al. Effects of inhibition of nitric oxide synthesis on epicardial coronary artery caliber and coronary blood flow in humans. Circulation. 1993;88: Ming Z, Parent R, Lavallee M. Nitric oxide independent dilation of conductance coronary arteries to acetylcholine in conscious dogs. Circ Res. 1997;81: Komaru T, Lamping KG, Eastham CL, et al. Effects of an arginine analogue on acetylcholine-induced coronary microvascular dilation in dogs. Am J Physiol. 1991;261:H2001 H Bauersachs J, Hecker M, Busse R. Display of the characteristics of endothelium-derived hyperpolarizing factor by a cytochrome P450- derived arachidonic acid metabolite in the coronary microcirculation. Br J Pharmacol. 1994;113: Campbell WB, Gebremedhin D, Pratt PF, et al. Identification of epoxyeicosatrienoic acids as an endothelium-derived hyperpolarizing factors. Circ Res. 1996;78: Edwards G, Dora KA, Gardener MJ, et al. K is an endothelium-derived hyperpolarizing factor in rat arteries. Nature. 1998;396: Taylor HJ, Chaytor AT, Evance WH, et al. Inhibition of the gap junctional component of endothelium-dependent relaxations in rabbit iliac artery by 18-alpha glycyrrhetinic acid. Br J Pharmacol. 1998;125: Matoba T, Shimokawa H, Nakashima M, et al. Hydrogen peroxide is an endothelium-derived hyperpolarizing factor in mice. J Clin Invest. 2000; 106: Matoba T, Shimokawa H, Kubota H, et al. Hydrogen peroxide is an endothelium-derived hyperpolarizing factor in human mesenteric arteries. Biochem Biophys Res Commun. 2002;290: Barlow RS, White RE. Hydrogen peroxide relaxes porcine coronary arteries by stimulating BK Ca channel activity. Am J Physiol. 1998;275: H1283 H Cabell F, Weiss DS, Price JM. Inhibition of adenosine-induced coronary vasodilation by block of large-conductance Ca 2 activated K channels. Am J Physiol. 1994;267:H1455 H Yada T, Hiramatsu O, Kimura A, et al. In vivo observation of subendocardial microvessels of the beating porcine heart using a needle-probe videomicroscope with a CCD camera. Circ Res. 1993;72: Yada T, Richmond KN, Van Bibber R, et al. Role of adenosine in local metabolic coronary vasodilation. Am J Physiol. 1999;276:H1425 H Hermann SC, Feigl EO. Subtraction method for the high-performance liquid chromatographic measurement of plasma adenosine. J Chromatogr. 1992;574: Merkus D, Vergroesen I, Hiramatsu O, et al. Stenosis differentially affects subendocardial and subepicardial arterioles in vivo. Am J Physiol. 2001;280:H1674 H Yamabe H, Okumura K, Ishizaka H, et al. Role of endothelium-derived nitric oxide in myocardial reactive hyperemia. Am J Physiol. 1992;263: H8 H Jeroudi MO, Triana FJ, Patel BS, et al. Effect of superoxide dismutase and catalase, given separately, on myocardial stunning Am J Physiol. 1990; 259:H889 H Node K, Kitakaze M, Kosaka H, et al. Bradykinin mediation of Ca 2 -activated K channels regulates coronary blood flow in ischemic myocardium. Circulation. 1997;95: Nagao T, Vanhoutte PM. Hyperpolarization as a mechanism for endothelium-dependent relaxations in the porcine coronary artery. J Physiol. 1992;445: Shimokawa H, Yasutake H, Fujii K, et al. The importance of the hyperpolarizing mechanism increases as the vessel size decreases in endothelium-dependent relaxations in rat mesenteric circulations. J Cardiovasc Pharmacol. 1996;28: Nishikawa Y, Stepp DW, Chilian WM. In vivo location and mechanism of EDHF-mediated vasodilation in canine coronary microcirculation. Am J Physiol. 1999;277:H1252 H Jensen BS, Stobaek D, Christophersen P, et al. Characterization of the cloned human intermediate-conductance Ca 2 -activated K channel. Am J Physiol. 1998;275:C848 C Yada T, Hiramatsu O, Kimura A, et al. Direct in vivo observation of subendocardial arteriolar responses during reactive hyperemia. Circ Res. 1995;77: Yada T, Hiramatsu O, Tachibana H, et al. Role of NO and K ATP channels in adenosine-induced vasodilation on in vivo canine subendocardial arterioles. Am J Physiol. 1999;277:H1931 H Stepp DW, Van Bibber R, Kroll K, et al. Quantitative relation between interstitial adenosine concentration and coronary blood flow. Circ Res. 1996;79: Gao Y, Vanhoutte PM. Effects of hydrogen peroxide on the responsiveness of isolated canine bronchi: role of prostaglandin E 2 and I 2. Am J Physiol. 1992;263:L402 L Chaytor AT, Taylor HJ, Griffith TM. Gap junction dependent and independent EDHF-type relations may involve smooth muscle camp accumulation. Am J Physiol. 2002;282:H1548 H Siegel G, Emden J, Wenzel K, et al. Potassium channel activation in vascular smooth muscle. Adv Exp Med Biol. 1992;311: Steenbergen C, Murphy E, Levy L, et al. Elevation in cytosolic free calcium concentration early in myocardial ischemia in perfused rat heart. Circ Res. 1987;60: Shimokawa H. Primary endothelial dysfunction: atherosclerosis. J Mol Cell Cardiol. 1999;31: Liu P, Hudetz G, Knaus HG, et al. Increased expression of Ca 2 -sensitive K channels in the cerebral microcirculation of genetically hypertensive rats: evidence for their protection against cerebral vasospasm. Circ Res. 1998;82: Cosentino F, Barker JE, Brand MP, et al. Reactive oxygen species mediate endothelium-dependent relaxations in tetrahydrobiopterindeficient mice. Arterioscler Thromb Vasc Biol. 2001;21: