Intrapericardial Delivery of l-Arginine Reduces the Increased Severity of Ventricular Arrhythmias During Sympathetic Stimulation in Dogs With Acute Coronary Occlusion
Nitric Oxide Modulates Sympathetic Effects on Ventricular Electrophysiological Properties
Background Nitric oxide (NO) modulates autonomic effects on myocardial contractility and sinus and atrioventricular nodal function of the heart. Whether NO influences autonomic actions on ventricular electrophysiological properties and arrhythmogenesis is not known.
Methods and Results Four groups consisting of 43 autonomically denervated dogs were studied. To “superfuse” sympathetic nerves innervating the ventricles, test drugs were introduced into the pericardial sac for 30 minutes, and their effects on ventricular effective refractory period (ERP) and arrhythmia development were assessed before and during sympathetic stimulation (SS). In group 1 (n=12), ventricular ERPs showed no significant difference between control and superfusion with l-arginine, a NO precursor (222±20 versus 222±19 ms, P=.485). However, l-arginine significantly reduced SS-induced ERP shortening compared with control (9±7 versus 13±7 ms, P<.001). Simultaneous administration of NG-monomethyl-l-arginine (2 mg/mL) abolished the inhibitory effects of l-arginine (13±7 versus 13±7 ms, P=.885). In group 2 (n=15), the severity of ventricular arrhythmias significantly increased during SS. l-Arginine reduced this increase caused by SS. In group 3 (n=8), plasma norepinephrine spillover measured from the coronary sinus significantly increased during SS and was reduced by pericardial superfusion with l-arginine compared with control (6005.2±1525.6 versus 8503.4±2044.5 pg/min, P=.012). In group 4 (n=8), l-arginine pericardial superfusion significantly increased NO overflow measured from the coronary sinus during SS (93.25±59.20 versus 114.82±74.92 nmol/min, P=.043).
Conclusions Pericardial l-arginine reduces ERP shortening and increased severity of ischemic ventricular arrhythmias during SS in dogs. NO-induced reduction of norepinephrine release in the heart may be one of the underlying mechanisms.
Substantial evidence indicates that heightened sympathetic activity can increase the prevalence and severity of life-threatening ventricular arrhythmias in both animals and humans. Because NO inhibits the positive inotropic response of the heart to β-adrenergic stimulation,1 perhaps by increasing production of cGMP via guanylyl cyclase,1,2 and modulates the autonomic responses of the sinus and atrioventricular nodes,3 we hypothesized that NO might influence the autonomic actions that modulate electrophysiological properties of ventricular myocardium and influence the pathogenesis of some ventricular arrhythmias produced by increased sympathetic activity. The purpose of this study was to test this hypothesis and to determine a potential mechanism by investigating the impact of NO on norepinephrine release from cardiac sympathetic nerves.
Forty-three adult mongrel dogs, weighing 20 to 30 kg, were allocated into four groups in which we measured ventricular ERP (group 1), ventricular arrhythmia (group 2), norepinephrine release (group 3), and NO production (group 4). The ERP group was further divided into two subgroups to assess the effects of l-arginine on autonomic modulation of ventricular ERP in the absence (group 1a) and presence of l-NMMA (group 1b), respectively. The experimental protocol was approved by the Indiana University Animal Experimentation Committee.
All dogs were anesthetized with 5% pentobarbital sodium (30 mg/kg IV), and additional doses of 1 mL were given as needed to maintain anesthesia during experiments. Dogs were intubated and ventilated with room air using a constant volume-cycled respirator (Harvard Apparatus). The left femoral artery and vein were cannulated to monitor blood pressure via a transducer (model p-23Db, Statham) and to administer fluids and pentobarbital sodium, respectively. The heart was exposed through a median sternotomy. The left and right stellate ganglia were isolated and decentralized. The cervical vagi were also isolated, doubly ligated, and transsected. A small hole (2×2 cm) was made in the anterior surface of the pericardium as we previously reported.4–6 A thermistor (model 400, Yellow Springs Instruments) was put into the pericardial sac and used to monitor epicardial temperature. With a heating pad underneath the animal and an operating lamp above the opened chest, body temperature was maintained between 36° and 37°C. The sinus node was crushed if the sinus rate remained >130 bpm after autonomic denervation.
Bipolar electrodes (0.5 to 1 cm apart) hooked in the right atrial appendage were used to record and stimulate the right atrium, and two bipolar electrodes (≈1 cm apart) were inserted into the right and left ventricular anterior walls for ventricular recording. The hook electrodes were made from Teflon-coated wires, insulated except for their tips. Standard lead II ECG, bipolar right atrial and right and left ventricular epicardial electrograms, and arterial blood pressure were recorded simultaneously using a multiple-channel physiological recorder (Honeywell VR16). In animals of group 1, eight hook electrodes were inserted into basal and apical right and left ventricular walls, respectively, to serve as the cathode for unipolar stimulation to determine ERPs (Fig 1⇓). An anodal electrode was placed in the abdominal wall. In animals of group 2, the left anterior descending coronary artery was dissected between the first and second diagonal branches, with care taken to not damage the perivascular nerves during this procedure. A silk suture, passed through a plastic tube, was placed around the isolated coronary artery for repeated occlusions.
In groups 3 and 4, a balloon catheter (Arrow) was inserted into the right internal jugular vein, and its tip was passed deep into the coronary sinus. The balloon catheter was fixed onto the lateral wall of the superior vena cava with a suture to prevent its slipping out of the coronary sinus. The position of the catheter was confirmed visually during and after completion of experiments. During collection of blood, the balloon was inflated, and coronary sinus blood flow was determined by the volume of blood withdrawn within a minute during very gentle pulling on the plunger of a syringe. The catheter was flushed with normal saline containing heparin after each collection. The heart rate was held constant (150 bpm) by pacing the right atrium during coronary sinus blood sampling. Heparin (a loading dose of 3000 U followed by 500 U/h IV) was administered to prevent blood clotting. In group 3, desipramine hydrochloride (1 mg/kg) was administered intravenously before coronary sinus sampling to block the neuronal reuptake of norepinephrine released from sympathetic nerve endings. Each blood sample was placed in a tube containing EDTA and heparin. The plasma, which was separated by centrifugation (6000 rpm) at a temperature of −10°C for 10 minutes, was stored in a freezer at a temperature of −77°C for ≥ 24 hours until norepinephrine or NO concentrations was measured. Norepinephrine was determined by radioenzymatic assay using phenylethanolamine N-methyltransferase.7
Analysis for NOx was performed using an established method8 by running samples through a NO analysis system (NOA 280, Sievers). Briefly, plasma samples were frozen at −77°C for ≥24 hours before analysis.9 On the day of the NOx assay, all samples were deproteinized by the addition of 2 μL of 10 N NaOH to 200 μL of plasma followed by the addition of 120 μL of 0.15 mol/L ZnSO4. Tubes were vortexed and placed on ice for 15 minutes and then centrifuged at 12,000 rpm for 10 minutes. The supernatant was used for analysis. A standard calibration curve for NOx measurement was generated at each assay with a standard solution of nitrate. Four microliters of supernatant/standard solution were injected quickly with a microliter syringe (Hamilton) through a silicone septum into a purge vessel containing vanadium chloride in hydrochloride and kept at 95°C. In the purge vessel, NOx was reduced to NO, which was transferred to the NO analysis system by a vacuum pump (model E2M5, Edwards). The amount of NO produced was determined by the luminescence generated in the presence of ozone in the NO analysis system. The luminescence measured is directly proportional to the amount of NO produced and, in turn, to the NOx content of the samples. Each sample and standard solution were measured in duplicate during all analytical runs.
Shielded bipolar electrodes were placed on the left and right stellate ganglia near the exit of the ansae subclaviae to stimulate the efferent cardiac sympathetic nerves through separate constant voltage isolators from a programmable two-channel stimulator (S88, Grass). Stimuli were 4 ms in duration at a frequency of 4 Hz. The output to the right stellate ganglia was adjusted to increase the heart rate to ≈140 bpm. The level of output to the left stellate ganglia was initially the same as that for the right one and adjusted to shorten left ventricular ERP by ≈10 to 20 ms.
Measurement of Ventricular ERP
Ventricular ERP was measured by a single extrastimulus (S2) delivered in decrements of 1 ms after every eight basic ventricular drives (S1) at a basic cycle length (S1S1) of 400 ms. S1 and S2 (2-ms duration and twice-diastolic pacing threshold) were delivered from a Medtronic stimulator (model 5325). The output of stimuli was kept constant throughout the entire study. The longest S1S2 coupling interval at which S2 did not evoke a ventricular depolarization and confirmed twice was defined as the ventricular ERP.
Cardiac sympathetic nerves were superfused4–6 by the introduction of normal Tyrode’s solution (≈40 to 80 mL) containing either l-arginine (the physiological precursor of NO; 100 mmol/L, Sigma Chemical) or l-arginine and l-NMMA (a NO synthase inhibitor; 2 mg/mL, Sigma). The millimolar composition of normal Tyrode’s solution was MgCl2 0.5, NaH2PO4 0.9, CaCl2 2.0, NaCl 137.0, NaHCO3 12.0, KCl 4.0, and glucose 5.0. This resulted in pH 7.35.6 The test solutions were prepared by the addition of substances into normal Tyrode’s solution before it was put into the pericardial sec. All the solutions were prewarmed to 37°C and gassed with 95% O2 and 5% CO2.
Baseline data were determined 30 minutes after the Tyrode’s solution was put into the pericardial sac. The Tyrode’s solution was replaced by the Tyrode’s solution containing l-arginine or l-NMMA. Each parameter was measured again 30 minutes later. Then the Tyrode’s solution containing drugs was carefully washed out at least three times using the Tyrode’s solution without any drugs. Data collection was repeated 30 minutes later.
Group 1: Ventricular ERP
ERPs were measured in the presence of pericardial superfusion with l-arginine (n=7) or l-arginine and l-NMMA (n=5) before and during SS. The mean value of ERPs in the presence of pericardial superfusion with Tyrode’s solution before and after superfusion with l-arginine was taken as control.
Group 2: Ventricular Arrhythmia
Ventricular arrhythmias were induced by repeated 7-minute occlusions of the left anterior descending coronary artery between the first and second diagonal branches during constant atrial pacing (cycle length, 400 ms) (n=15). The latter was begun 90 sec before coronary occlusion and continued for 90 sec after release of occlusion. SS, when applied, was started 60 sec before coronary artery occlusion and was continued for 60 sec after reperfusion. The first coronary artery occlusion was performed 30 minutes after normal Tyrode’s solution was put into the pericardial sac. Because intramyocardial conduction delay and ventricular arrhythmias are usually more exaggerated during the first occlusion than subsequent ones,5 possibly due to the effects of preconditioning,10 results from this first occlusion were discarded and the second occlusion was used as control. The second coronary occlusion was performed 15 minutes after the first occlusion (Table 1⇓). The third occlusion was performed 15 minutes later in the presence of SS. Then, the Tyrode’s solution containing l-arginine was put into the pericardial sac to replace normal Tyrode’s solution, and the fourth coronary artery occlusion was performed 30 minutes later. The fifth coronary artery occlusion was performed after superfusion with normal Tyrode’s solution for 30 minutes (washout). Both the fourth and the fifth coronary occlusions were performed in the presence of SS.
Groups 3 and 4: Norepinephrine Release and NO Overflow
Five-milliliter samples from coronary sinus blood were obtained during pericardial superfusion with (1) Tyrode’s solution, (2) Tyrode’s solution and l-arginine, and (3) washout with Tyrode’s solution after atrial pacing at 150 bpm for 3 minutes. These were repeated during SS for 3 minutes. The concentration of norepinephrine in coronary sinus blood reaches steady state within a minute of SS in normal dog hearts.11 Norepinephrine spillover (group 3; n=8) and NO overflow (group 4; n=8) was defined as the product of norepinephrine or NO concentration and blood flow from the coronary sinus.
All data are expressed as mean±SD. Paired t test, one-way ANOVA, Friedman’s test, and Wilcoxon’s signed rank test were used where appropriate. A two-tailed p value of <.05 was considered statistically significant.
Effects of NO on Ventricular ERPs (Group 1)
In the baseline state without SS in 7 dogs (group 1a), there was no significant difference in ventricular ERP during pericardial superfusion with Tyrode’s solution (control) and during pericardial superfusion with l-arginine (222±20 versus 222±19 ms, P=.485). SS significantly shortened ventricular ERPs by 13±7 ms during pericardial superfusion with Tyrode’s solution (P<.001), whereas during superfusion with Tyrode’s solution containing l-arginine, SS shortened ventricular ERP by 9±7 ms (P<.001) (Fig 2⇓, top). Thus, l-arginine reduced the SS-induced shortening of ventricular ERPs by 31% (P<.001).
In an additional 5 dogs (group 1b), during simultaneous pericardial superfusion with l-NMMA and l-arginine, there was no significant difference in SS-induced change in ERPs compared with controls (13±7 versus 13±7 ms, P=.885) (Fig 2⇑, bottom).
Effects of NO on Ischemic Ventricular Arrhythmias (Group 2)
Ventricular arrhythmias were classified into five categories of increasing severity5: (1) no ventricular arrhythmias, (2) isolated ventricular ectopic beats, (3) couplets, (4) nonsustained ventricular tachycardia (≥3 consecutive beats but <15 sec), and (5) ventricular fibrillation. The highest category of ventricular arrhythmias was scored for each occlusion. Data from 7 dogs were eliminated because they had no ventricular arrhythmias before and during SS. The severity of ventricular arrhythmia significantly increased during SS in the remaining 8 dogs (Fig 3⇓). l-Arginine significantly reduced this increase in ventricular arrhythmias caused by SS (P=.018). There was no significant difference in the severity of ventricular arrhythmia after washout of l-arginine using normal Tyrode’s solution compared with that before pericardial superfusion with l-arginine (P=.600).
Effects of NO on Release of Norepinephrine From Sympathetic Nerves of the Heart (Group 3)
The coronary sinus blood flow was increased during superfusion with l-arginine compared with control before (control, 12.2±2.8; l-arginine, 13.7±3.0 mL/min, P=.012) and during SS (control, 14.3±2.9; l-arginine, 15.8±2.9 mL/min, P=.068). Before SS, there were no significant changes in both norepinephrine concentration (control, 113.1±26.2; l-arginine, 114.8±25.8 pg/mL, P=.674) and spillover (control, 1347.0±456.7; l-arginine, 1470.4±407.8 pg/min, P=.161) during superfusion with l-arginine compared with controls (Fig 4⇓). During SS, both norepinephrine concentration (control, 682.9±204.9; l-arginine, 436.1±155.3 pg/mL, P=.012) and spillover (control, 8503.4±2044.5; l-arginine, 6005.2±1525.6 pg/min, P=.012) were significantly decreased during superfusion with l-arginine compared with controls. Thus, l-arginine reduced both the norepinephrine concentration and spillover by 36.1% and 29.4%, respectively, during SS, despite an increase in coronary blood flow.
Coronary Sinus NO Overflow During Superfusion of the Heart With l-Arginine (Group 4)
In group 4, the coronary sinus blood flow was significantly increased during l-arginine superfusion before (from 10.50±1.76 to 12.75±1.78 mL/min, P=.020) and during SS (from 11.67±1.99 to 14.03±2.53 mL/min, P=.042). NO overflow was significantly increased during l-arginine superfusion compared with control during SS (from 93.25±26.47 to 114.82±33.51 nmol/min, P=.043). NO overflow was also greater during superfusion with l-arginine without SS (from 91.69±27.80 to 100.26±28.77 nmol/min), but the difference did not reach statistical significance (P=.225). The difference in NO concentrations between superfusion with l-arginine and control was not statistically significant (Fig 5⇓).
Main Findings of the Present Study
In this study, we tested the hypothesis that NO is involved in the autonomic modulation of ventricular electrophysiological properties in canine hearts. We found that pericardial superfusion with l-arginine reduced the extent of sympathetically induced ERP shortening and the increased severity of ventricular arrhythmias produced during coronary occlusions. The fact that the specific NO synthase inhibitor l-NMMA prevented the action of l-arginine on ventricular ERPs during SS suggests that the underlying mechanism is through a NO pathway. In fact, we demonstrated that superfusion of the heart with l-arginine increased NO production in the heart in this experimental preparation. We found that these results can be explained by decreased norepinephrine spillover in the coronary sinus, presumably by reducing norepinephrine release from sympathetic nerves in the heart, and are consistent with the observations from our previous study.3
Potential Mechanisms Underlying NO-Mediated Modulation of Autonomic Effects on Ventricular Electrophysiological Properties
NO, produced by either constitutive or inducible isoforms of NO synthase, influences myocardial inotropic and chronotropic responses. The NO pathway has been studied using NO donors or NO synthase inhibitors.12 Basal NO production seems to exert little tonic inhibition of β-adrenergic responses.13 Because cGMP is associated with the negative effects of muscarinic cholinergic stimulation of the heart,14 NO-mediated intracellular production of cGMP via guanylyl cyclase2 may account for many of the observed physiological actions of NO. NO has been shown to be an important modulator of β-adrenergic effects on slow inward calcium currents through cGMP-dependent protein kinase.15 Mohan et al16 recently reported that cGMP is involved in the NO-induced biphasic contractile response in a concentration-dependent manner. This effect of NO is modulated by the status of endocardial endothelium and by concomitant autonomic stimulation.13
The potential sites of NO-mediated changes in ventricular electrophysiological properties during SS may be ventricular myocytes, cardiac neurons, or both. Horackova et al17 recently reported that NO did not directly affect cultured adult guinea pig myocytes. The beating rate of such myocytes could be indirectly affected when they were cultured with intrinsic or stellate ganglia neurons. In consideration that there was a significant l-arginine–mediated effect on ventricular ERPs during SS in our study but no significant change without SS, a likely location of the effects of NO is the signaling between cardiac neurons and ventricular myocardium. This is consistent with the finding in the study of Horackova et al17 that the effect of NO on the beating rates of ventricular myocytes depended on neuronal activation. Several investigators18–21 reported that nonadrenergic and noncholinergic actions may be involved in certain physiological responses to NO. Our finding that norepinephrine release in the heart was decreased in response to SS makes nonadrenergic and noncholinergic actions of NO less important in the NO-mediated autonomic modulation of ventricular electrophysiology.
Because NO is involved in the modulation of coronary blood flow,22,23 the decrease in the incidence of ventricular arrhythmias in the presence of l-arginine may be caused by improvement of myocardial perfusion in addition to changes in ventricular electrophysiological properties. Several studies24–27 have demonstrated that NO donors and l-arginine ameliorate ischemia/reperfusion injury and improve postischemic mechanical function, whereas NO synthase inhibitors are associated with a significantly poorer recovery. The fact that protection of NO donors and l-arginine has been demonstrated in subvasodilatory doses25,28 suggests that dilation of coronary vessels and increase of coronary blood flow are not the only mechanisms responsible for the protective action of NO. In addition, although changes in coronary blood flow can account for the influence of NO on ventricular arrhythmias, they cannot account for the modulating effects of NO on ventricular ERPs or norepinephrine spillover in this study. NO-induced alteration of norepinephrine release from the heart and of autonomic modulation of the electrophysiological properties of the ventricles may play a more important role in the reduction of the severity of ventricular arrhythmias during coronary artery occlusion and SS.
Potential Study Limitations
Because a higher concentration of l-arginine may cause nonspecific negative inotropic effect on papillary muscles,13 the use of a relatively high concentration of l-arginine in this study may also result in nonspecific changes in electrophysiological properties of ventricular myocardium. However, the finding that the electrophysiological effects of l-arginine could be abolished by l-NMMA suggests the NO pathway is the underlying mechanism. On the other hand, the concentration of l-arginine reached at the myocardium might be much less than that in the pericardial sac. Although every effort was made to maintain the experimental conditions stable, it was difficult to eliminate minor changes with time in cardiac function and electrophysiology during the entire experiment. These changes with time should be compensated by taking the average value between the baseline and washout. However, we could not eliminate a potential residual effect of l-arginine after washout.
To maintain the pericardial sac intact, we were unable to secure the cannula around the coronary sinus close to its ostium to collect blood samples; instead, we used a balloon catheter advanced deep in the coronary sinus. Thus, coronary sinus blood flow measured in this manner may be originating mostly from left ventricle and various parts of the right ventricle. This may explain the fact that coronary sinus blood flow was lower than that reported in the literature.11 It should be pointed out the method used to measure coronary sinus blood flow is not as accurate as other methods, such as thermodilution. Furthermore, the vasodilation effects of NO cannot be excluded in this study.
Selected Abbreviations and Acronyms
|cGMP||=||3′,5′-cyclic guanosine monophosphate|
|ERP||=||effective refractory period|
This study is supported in part by the Herman C. Krannert Fund and by grants HL-52322 and DK-42469 from the National Heart, Lung, and Blood Institute of the National Institutes of Health. We thank Dzung D. Nguyen, MSc, and Claude Arnett, BSc, for surgical preparation; Judith A Eudaly, BSc, for measurement of plasma norepinephrine; Yonggang Wu, MD, for measurement of plasma NO; and Naomi Fineberg, PhD, for statistical help.
- Received June 19, 1997.
- Revision received July 29, 1997.
- Accepted August 27, 1997.
- Copyright © 1997 by American Heart Association
Chung MK, Gulick TS, Rotondo RE, Schreiner GF, Lange LG. Mechanism of cytokine inhibition of β-adrenergic agonist stimulation of cyclic AMP in rat cardiac myocytes: impairment of signal transduction. Circ Res. 1990;67:753–763.
Elvan A, Rubart M, Zipes DP. Nitric oxide modulates autonomic effects on spontaneous sinus discharge rate and atrioventricular nodal conduction in open chest anesthetized dogs. Am J Physiol. 1997;272:H263–H271.
Miyazaki T, Pride HP, Zipes DP. Modulation of cardiac autonomic neurotransmission by epicardial superfusion: effects of hexamethonium and tetrodotoxin. Circ Res. 1989;65:1212–1219.
Miyazaki T, Zipes DP. Pericardial prostaglandin biosynthesis prevents the increased incidence of reperfusion-induced ventricular fibrillation produced by efferent sympathetic stimulation in dogs. Circulation. 1990;82:1008–1019.
Miyazaki T, Pride HP, Zipes DP. Prostaglandins in the pericardial fluid modulate neural regulation of cardiac electrophysiological properties. Circ Res. 1990;66:163–175.
Archer S. Measurement of nitric oxide in biological models. FASEB Jl. 1993;7:349–360.
Moshage H, Kok B, Huizenga JR, Jansen PL. Nitrite and nitrate determinations in plasma: a critical evaluation. Clin Chem. 1995;41:892–896.
Miyazaki T, Zipes DP. Protection against autonomic denervation following acute myocardial infarction by preconditioning ischemia. Circ Res. 1989;64:437–448.
Masuda Y, Levy MN. Heart rate modulates the disposition of neurally released norepinephrine in cardiac tissues. Circ Res. 1985;57:19–27.
Weyrich AS, Ma XL, Buerke M, Murohara T, Armstead VE, Lefer AM, Nicolas JM, Thomas AP, Lefer DJ, Vinten-Johansen J. Physiological concentrations of nitric oxide do not elicit an acute negative inotropic effect in unstimulated cardiac muscle. Circ Res. 1994;75:692–700.
Mery PF, Pavoine C, Belhassen L, Pecker F, Fischmeister R. Nitric oxide regulates cardiac Ca2+ current: involvement of cGMP-inhibited and cGMP-stimulated phosphodiesterases through guanylyl cyclase activation. J Biol Chem. 1993;268:26286–26295.
Wahler GM, Dollinger SJ. Nitric oxide donor SIN-1 inhibits mammalian cardiac calcium current through cGMP-dependent protein kinase. Am J Physiol. 1995;268:C45–C54.
Mohan P, Brutsaert DL, Paulus WJ, Sys SU. Myocardial contractile response to nitric oxide and cGMP. Circulation. 1996;93:1223–1229.
Horackova M, Armour JA, Hopkins DA, Huang MH. Nitric oxide modulates signaling between cultured adult peripheral cardiac neurons and cardiomyocytes. Am J Physiol. 1995;269:C504–C510.
Quyyumi AA, Dakak N, Andrews NP, Gilligan DM, Panza JA, Cannon RO. Contribution of nitric oxide to metabolic coronary vasodilation in the human heart. Circulation. 1995;92:320–326.
Parent R, al-Obaidi M, Lavallee M. Nitric oxide formation contributes to β-adrenergic dilation of resistance coronary vessels in conscious dogs. Circ Res. 1993;73:241–251.
Nakanishi K, Vinten-Johansen J, Lefer DJ, Zhao Z, Fowler WCd, McGee DS, Johnston WE. Intracoronary l-arginine during reperfusion improves endothelial function and reduces infarct size. Am J Physiol. 1992;263:H1650–H1658.
Weyrich AS, Ma XL, Lefer AM. The role of l-arginine in ameliorating reperfusion injury after myocardial ischemia in the cat. Circulation. 1992;86:279–288.
Siegfried MR, Erhardt J, Rider T, Ma XL, Lefer AM. Cardioprotection and attenuation of endothelial dysfunction by organic nitric oxide donors in myocardial ischemia-reperfusion. J Pharmacol Exp Ther. 1992;260:668–675.