992
Mediator of Hypoxic Coronary Vasodilatation
Nitric Oxide Is
a
Relation to Adenosine and Cyclooxygenase-Derived Metabolites Kwan Ha Park, Lisa E. Rubin, Steven S. Gross, and Roberto Levi Hypoxia is a potent coronary-vasodilating signal; its mechanisms are still controversial. We have assessed the possible role of nitric oxide (NO) in hypoxic coronary vasodilatation (HCVD) in isolated guinea pig hearts perfused at constant pressure. HCVD was elicited by a 1-minute 100o N2 exposure; coronary flow doubled within 1 minute of hypoxia (early phase) and returned to baseline within 40 seconds after reoxygenation (late phase). The early phase of HCVD was associated with a rapid approximately eightfold increase in cGMP overflow, an indication of NO release. The specific NO synthase inhibitor N'-methylL-arginine (NMA, 0.1-1 mM) antagonized HCVD and the associated increase in cGMP spillover (maximum inhibition, =65%); excess arginine (1.2 mM) prevented both effects. The late phase of HCVD was associated with an increase in adenosine overflow and was attenuated by the adenosine receptor antagonist BW A1433 (1 ,M; maximum inhibition, :45%). Indomethacin (10 ,uM) inhibited HCVD in spontaneously beating hearts by =35% but had no effect in hearts paced at faster rates. NMA and BW A1433 were more effective in combination than alone (maximum inhibition, =72%). However, irrespective of the concentrations used, there was no synergism among the anti-HCVD effects of NMA, BW A1433, and indomethacin, nor was HCVD completely inhibited by the antagonists, whether alone or in combination. Our findings indicate that NO is an important mediator of the early phase of HCVD, whereas additional mechanisms and/or factors, including adenosine and vasodilatatory prostaglandins, contribute to the late phase. (Circulation Research 1992;71:992-1001) KEY WoRDs * adenosine * cGMP * hypoxia * hypoxic coronary vasodilatation * isolated guinea pig heart * nitric oxide
It is generally agreed that hypoxia has a strong vasodilating effect on the coronary circulation.1 Yet the mechanisms by which coronary arteries dilate when oxygen tension decreases are still at issue. Much attention has been devoted in the past to the possible role of vasodilatatory metabolites, and adenosine has long been considered an important mediator of coronary vasomotion in response to lowered oxygen tension.23 Adenosine is released from endothelial cells and myocytes of hypoxic hearts and reaches vasoactive concentrations in the coronary effluent.4-6 Nevertheless, there are indications that coronary vasodilatation in response to acute hypoxia is not solely dependent on adenosine. Indeed, the time course of hypoxic coronary vasodilatation (HCVD) has been found not to coincide with that of adenosine release,7 and no reduction of HCVD has been observed upon enzymatic destruction From the Department of Pharmacology, Cornell University Medical College, New York. Preliminary data were published in abstract form (FASEB J 1992;6:A2018). Supported by National Institutes of Health grants HL-34215 and HL-46403. L.E.R. is a predoctoral fellow of the Pharmaceutical Manufacturers' Association Foundation. Address for correspondence: Roberto Levi, MD, Department of Pharmacology, Cornell University Medical College, 1300 York Avenue, New York, NY 10021. Received April 20, 1992; accepted July 9, 1992.
of adenosine within the heart.8 Thus, other endogenous vasodilators are likely to play a role in HCVD. Although the earlier literature tended to discount the role of vasodilating prostaglandins in HCVD,9-11 more recent evidence suggests that the relaxation of coronary vessels in response to hypoxic stimuli is mediated at least in part by prostaglandin release.12-14 However, neither adenosine nor prostaglandins seem to account for HCVD in its entirety, and other mechanisms have been proposed, such as increased production of the vasodilator ATP15 and hyperpolarization of coronary vascular smooth muscle via the opening of ATP-sensitive K' channels.16 Since its discovery as an endothelium-derived relaxing factor17 and its identification as nitric oxide (NO),1819 it has been demonstrated both by us20 and others21 that NO is released from the coronary endothelium and modulates coronary vascular tone. Inasmuch as hypoxia stimulates the release of endotheliumderived relaxing factor from aortic and femoral artery endothelium,22 it is plausible that a decrease in oxygen tension will also promote the release of NO from the coronary vasculature and that this autacoid will contribute to the increase in coronary flow that is associated with hypoxia. Thus, the purpose of our investigation was to establish whether NO production is increased in the hypoxic heart and to assess the possible role of this
Downloaded from http://circres.ahajournals.org/ at CONS CALIFORNIA DIG LIB on March 13, 2015
Park et al Nitric Oxide and HCVD autacoid as a mediator of HCVD. Because of the controversy surrounding the role of other putative mediators, we have also sought to investigate the relation of NO release and actions to those of adenosine and vasodilatatory prostaglandins. Materials and Methods Isolated Heart Perfusion Male Hartley guinea pigs (Hilltop, Pa.) weighing 250-300 g were killed by cervical dislocation under light anesthesia with CO2 vapor. After midline thoracotomy, the heart was rapidly excised and perfused retrogradely at a constant pressure of 45 cm H20 in a Langendorff apparatus with Ringer's solution at 37°C saturated with 100% 02.23 The composition of the Ringer's solution was (mM) NaCl 154.0, KCl 5.61, NaHCO3 5.55, CaCl2 2.16, and dextrose 5.95. Coronary flow was monitored with an ultrasonic flowmeter (Transonic Systems, Ithaca, N.Y.). The probe of the flowmeter was placed right above the orifices of the coronary arteries. In addition, coronary flow was determined by measuring the volume of coronary effluent collected during consecutive 10second periods. Collected coronary perfusate was stored at -70°C for analysis of cGMP and adenosine concentrations. Isometric ventricular contractions were continuously recorded with a force-displacement transducer (model FT03, Grass Instruments, Quincy, Mass.) connected through a thread to the apex of the left ventricle. Sinoatrial and ventricular rates were continuously monitored from a bipolar surface electrogram recorded from the right atrium and left ventricle. Hearts were allowed to beat spontaneously or were paced at a constant rate of 275 beats per minute (i.e., =65 beats per minute greater than the spontaneous control rate) with square-wave pulses (6-7 V, 10-msec duration) delivered by a stimulator (model S48, Grass) connected via a stimulus isolation unit (model SIU 5, Grass) to two electrodes placed on the dorsal surface of the right ventricle. In a separate series of experiments, hearts were perfused at a constant flow of 6.5 ml/min with a peristaltic pump (Polystaltic pump, Buchler Instruments, Ft. Lee, N.J.). The perfusion pressure was continuously monitored with a physiological pressure transducer (model P23AA, Statham). Isometric ventricular contraction and the bipolar surface electrogram were recorded as noted above.
Induction of Hypoxia Hearts were first perfused with oxygenated Ringer's solution for 30-45 minutes until sinoatrial rate, contraction, and coronary flow rate reached a steady state. Hypoxia was induced by perfusing the hearts for 1 minute with Ringer's solution equilibrated with 100% N2. The 02 concentration in the perfusion solution was measured by inserting an oxygen electrode (model MI-730, Microelectrodes, Inc., Londonderry, N.H.) into the aortic cannula. The calculated P02 values were 631+± 19 mm Hg before hypoxia and fell gradually to 123+4 mm Hg by 1 minute of hypoxia (mean+SEM, n=8). A P02 of 123 mm Hg is representative of marked hypoxia in the isolated nonhemoglobin-perfused guinea pig heart.24
993
When hypoxia was induced in the presence of pharmacological agents, hearts were continuously perfused with a given drug at the desired concentration for a 20-minute period before the induction of hypoxia. In some experiments, adenosine was administered to isolated hearts by rapid intra-aortic bolus injection in a volume of 300 ,lI.
Adenosine Assay Adenosine and inosine were assayed by reverse-phase high-performance liquid chromatography according to the method of Jenkins and Belardinelli.25 The highperformance liquid chromatography system included an autoinjector (model 2157, Pharmacia LKB Biotechnology Inc., Piscataway, N.J.), an LKB 2150 pump, and a spectrophotometer (absorbance detector, model 160, Beckman Instruments, San Ramon, Calif.). The mobile phase was 0.05 M phosphate buffer (pH 5.1) containing 10% methanol. Samples of 10-50 ,ul were injected into a reverse-phase column (ultrasphere ODS, 3 gam, 4.6 mmx7.5 cm, Beckman Instruments). The absorbance was measured at 254 nm. Adenosine overflow was calculated by combining the concentration of adenosine with its metabolic product, inosine.
cGMP Assay Coronary effluent samples of 1.5-2.0 ml were concentrated to dryness under a nitrogen stream and then reconstituted in 0.2 ml distilled water. cGMP was measured using a commercial enzyme immunoassay kit (Cayman Chemical Co. Inc., Ann Arbor, Mich.).
Drugs BW A1433, a nonselective adenosine receptor antagonist,26 was kindly donated by the Burroughs Wellcome Co., Research Triangle Park, N.C. Indomethacin, adenosine hemisulfate, and L-arginine hydrochloride were purchased from Sigma Chemical Co., St. Louis, Mo.; compound U46619 was purchased from Cayman. N6Methyl-L-arginine (NMA) was synthesized by Dr. 0. W. Griffith, Department of Biochemistry, Cornell University Medical College, New York. Pinacidil was a gift from Dr. Garrett J. Gross, Department of Pharmacology, Medical College of Wisconsin, Milwaukee.
Data and Statistics All data are expressed as mean+SEM. Statistical differences were assessed by t test when two groups were compared. When multiple groups were compared, analysis of variance combined with the Newman-Keuls test was used. Values of p
Mediator of Hypoxic Coronary Vasodilatation
Nitric Oxide Is
a
Relation to Adenosine and Cyclooxygenase-Derived Metabolites Kwan Ha Park, Lisa E. Rubin, Steven S. Gross, and Roberto Levi Hypoxia is a potent coronary-vasodilating signal; its mechanisms are still controversial. We have assessed the possible role of nitric oxide (NO) in hypoxic coronary vasodilatation (HCVD) in isolated guinea pig hearts perfused at constant pressure. HCVD was elicited by a 1-minute 100o N2 exposure; coronary flow doubled within 1 minute of hypoxia (early phase) and returned to baseline within 40 seconds after reoxygenation (late phase). The early phase of HCVD was associated with a rapid approximately eightfold increase in cGMP overflow, an indication of NO release. The specific NO synthase inhibitor N'-methylL-arginine (NMA, 0.1-1 mM) antagonized HCVD and the associated increase in cGMP spillover (maximum inhibition, =65%); excess arginine (1.2 mM) prevented both effects. The late phase of HCVD was associated with an increase in adenosine overflow and was attenuated by the adenosine receptor antagonist BW A1433 (1 ,M; maximum inhibition, :45%). Indomethacin (10 ,uM) inhibited HCVD in spontaneously beating hearts by =35% but had no effect in hearts paced at faster rates. NMA and BW A1433 were more effective in combination than alone (maximum inhibition, =72%). However, irrespective of the concentrations used, there was no synergism among the anti-HCVD effects of NMA, BW A1433, and indomethacin, nor was HCVD completely inhibited by the antagonists, whether alone or in combination. Our findings indicate that NO is an important mediator of the early phase of HCVD, whereas additional mechanisms and/or factors, including adenosine and vasodilatatory prostaglandins, contribute to the late phase. (Circulation Research 1992;71:992-1001) KEY WoRDs * adenosine * cGMP * hypoxia * hypoxic coronary vasodilatation * isolated guinea pig heart * nitric oxide
It is generally agreed that hypoxia has a strong vasodilating effect on the coronary circulation.1 Yet the mechanisms by which coronary arteries dilate when oxygen tension decreases are still at issue. Much attention has been devoted in the past to the possible role of vasodilatatory metabolites, and adenosine has long been considered an important mediator of coronary vasomotion in response to lowered oxygen tension.23 Adenosine is released from endothelial cells and myocytes of hypoxic hearts and reaches vasoactive concentrations in the coronary effluent.4-6 Nevertheless, there are indications that coronary vasodilatation in response to acute hypoxia is not solely dependent on adenosine. Indeed, the time course of hypoxic coronary vasodilatation (HCVD) has been found not to coincide with that of adenosine release,7 and no reduction of HCVD has been observed upon enzymatic destruction From the Department of Pharmacology, Cornell University Medical College, New York. Preliminary data were published in abstract form (FASEB J 1992;6:A2018). Supported by National Institutes of Health grants HL-34215 and HL-46403. L.E.R. is a predoctoral fellow of the Pharmaceutical Manufacturers' Association Foundation. Address for correspondence: Roberto Levi, MD, Department of Pharmacology, Cornell University Medical College, 1300 York Avenue, New York, NY 10021. Received April 20, 1992; accepted July 9, 1992.
of adenosine within the heart.8 Thus, other endogenous vasodilators are likely to play a role in HCVD. Although the earlier literature tended to discount the role of vasodilating prostaglandins in HCVD,9-11 more recent evidence suggests that the relaxation of coronary vessels in response to hypoxic stimuli is mediated at least in part by prostaglandin release.12-14 However, neither adenosine nor prostaglandins seem to account for HCVD in its entirety, and other mechanisms have been proposed, such as increased production of the vasodilator ATP15 and hyperpolarization of coronary vascular smooth muscle via the opening of ATP-sensitive K' channels.16 Since its discovery as an endothelium-derived relaxing factor17 and its identification as nitric oxide (NO),1819 it has been demonstrated both by us20 and others21 that NO is released from the coronary endothelium and modulates coronary vascular tone. Inasmuch as hypoxia stimulates the release of endotheliumderived relaxing factor from aortic and femoral artery endothelium,22 it is plausible that a decrease in oxygen tension will also promote the release of NO from the coronary vasculature and that this autacoid will contribute to the increase in coronary flow that is associated with hypoxia. Thus, the purpose of our investigation was to establish whether NO production is increased in the hypoxic heart and to assess the possible role of this
Downloaded from http://circres.ahajournals.org/ at CONS CALIFORNIA DIG LIB on March 13, 2015
Park et al Nitric Oxide and HCVD autacoid as a mediator of HCVD. Because of the controversy surrounding the role of other putative mediators, we have also sought to investigate the relation of NO release and actions to those of adenosine and vasodilatatory prostaglandins. Materials and Methods Isolated Heart Perfusion Male Hartley guinea pigs (Hilltop, Pa.) weighing 250-300 g were killed by cervical dislocation under light anesthesia with CO2 vapor. After midline thoracotomy, the heart was rapidly excised and perfused retrogradely at a constant pressure of 45 cm H20 in a Langendorff apparatus with Ringer's solution at 37°C saturated with 100% 02.23 The composition of the Ringer's solution was (mM) NaCl 154.0, KCl 5.61, NaHCO3 5.55, CaCl2 2.16, and dextrose 5.95. Coronary flow was monitored with an ultrasonic flowmeter (Transonic Systems, Ithaca, N.Y.). The probe of the flowmeter was placed right above the orifices of the coronary arteries. In addition, coronary flow was determined by measuring the volume of coronary effluent collected during consecutive 10second periods. Collected coronary perfusate was stored at -70°C for analysis of cGMP and adenosine concentrations. Isometric ventricular contractions were continuously recorded with a force-displacement transducer (model FT03, Grass Instruments, Quincy, Mass.) connected through a thread to the apex of the left ventricle. Sinoatrial and ventricular rates were continuously monitored from a bipolar surface electrogram recorded from the right atrium and left ventricle. Hearts were allowed to beat spontaneously or were paced at a constant rate of 275 beats per minute (i.e., =65 beats per minute greater than the spontaneous control rate) with square-wave pulses (6-7 V, 10-msec duration) delivered by a stimulator (model S48, Grass) connected via a stimulus isolation unit (model SIU 5, Grass) to two electrodes placed on the dorsal surface of the right ventricle. In a separate series of experiments, hearts were perfused at a constant flow of 6.5 ml/min with a peristaltic pump (Polystaltic pump, Buchler Instruments, Ft. Lee, N.J.). The perfusion pressure was continuously monitored with a physiological pressure transducer (model P23AA, Statham). Isometric ventricular contraction and the bipolar surface electrogram were recorded as noted above.
Induction of Hypoxia Hearts were first perfused with oxygenated Ringer's solution for 30-45 minutes until sinoatrial rate, contraction, and coronary flow rate reached a steady state. Hypoxia was induced by perfusing the hearts for 1 minute with Ringer's solution equilibrated with 100% N2. The 02 concentration in the perfusion solution was measured by inserting an oxygen electrode (model MI-730, Microelectrodes, Inc., Londonderry, N.H.) into the aortic cannula. The calculated P02 values were 631+± 19 mm Hg before hypoxia and fell gradually to 123+4 mm Hg by 1 minute of hypoxia (mean+SEM, n=8). A P02 of 123 mm Hg is representative of marked hypoxia in the isolated nonhemoglobin-perfused guinea pig heart.24
993
When hypoxia was induced in the presence of pharmacological agents, hearts were continuously perfused with a given drug at the desired concentration for a 20-minute period before the induction of hypoxia. In some experiments, adenosine was administered to isolated hearts by rapid intra-aortic bolus injection in a volume of 300 ,lI.
Adenosine Assay Adenosine and inosine were assayed by reverse-phase high-performance liquid chromatography according to the method of Jenkins and Belardinelli.25 The highperformance liquid chromatography system included an autoinjector (model 2157, Pharmacia LKB Biotechnology Inc., Piscataway, N.J.), an LKB 2150 pump, and a spectrophotometer (absorbance detector, model 160, Beckman Instruments, San Ramon, Calif.). The mobile phase was 0.05 M phosphate buffer (pH 5.1) containing 10% methanol. Samples of 10-50 ,ul were injected into a reverse-phase column (ultrasphere ODS, 3 gam, 4.6 mmx7.5 cm, Beckman Instruments). The absorbance was measured at 254 nm. Adenosine overflow was calculated by combining the concentration of adenosine with its metabolic product, inosine.
cGMP Assay Coronary effluent samples of 1.5-2.0 ml were concentrated to dryness under a nitrogen stream and then reconstituted in 0.2 ml distilled water. cGMP was measured using a commercial enzyme immunoassay kit (Cayman Chemical Co. Inc., Ann Arbor, Mich.).
Drugs BW A1433, a nonselective adenosine receptor antagonist,26 was kindly donated by the Burroughs Wellcome Co., Research Triangle Park, N.C. Indomethacin, adenosine hemisulfate, and L-arginine hydrochloride were purchased from Sigma Chemical Co., St. Louis, Mo.; compound U46619 was purchased from Cayman. N6Methyl-L-arginine (NMA) was synthesized by Dr. 0. W. Griffith, Department of Biochemistry, Cornell University Medical College, New York. Pinacidil was a gift from Dr. Garrett J. Gross, Department of Pharmacology, Medical College of Wisconsin, Milwaukee.
Data and Statistics All data are expressed as mean+SEM. Statistical differences were assessed by t test when two groups were compared. When multiple groups were compared, analysis of variance combined with the Newman-Keuls test was used. Values of p
