Enhancement of myocardial reactive hyperemia with manganese-superoxide dismutase: role of endothelium-derived nitric oxide

Transcription

1 Gzrdiozmscular Research ELSEVER Cardiovascular Research 31 (19%) Enhancement of myocardial reactive hyperemia with manganese-superoxide dismutase: role of endothelium-derived nitric oxide Ryusuke Tsunoda, Ken Okumura *, Hiroshi shizaka, Toshiro Matsunaga, Toshifumi Tabuchi, Shinji Tayama, Hirofumi Yasue Division of Cardiology, Kumamoto University School of Medicine, Honjo l-1-1, Kumamoto City, Kumamoto 860, Japan Received 13 June 1995; accepted 9 October 1995 Abstract Objective: To test the hypothesis that superoxide radicals generated during myocardial ischemia and reperfusion influence reactive hyperemia (RH) by reacting with endothelium-derived nitric oxide (EDNO), we examined the effect of manganese (Mn)-superoxide dismutase (SOD) on RH in anesthetized dogs. Methods: Twelve dogs were pretreated with 8-phenyltheophylline (8PT) to block adenosine s effect. Five dogs were pretreated with 8PT and NC-nitro-L-arginine methyl ester (L-NAME) to block adenosine s and EDNO s effects. Following occlusion of the left circumflex artery (LCX) for 10 and 60 s, RH was observed before and after Mn-SOD. n another group of 6 dogs pretreated with 8PT, RH following 60-s LCX occlusion was observed before and after Mn-SOD and catalase. For comparison with the effect of Mn-SOD, that of copper, zinc (Cu,Zn)-SOD was also examined in another group of 5 dogs. Results: n the dogs pretreated with 8PT, Mn-SOD significantly increased excess flow and repayment of flow debt during RH after 60-s LCX occlusion but did not affect RH after 10-s LCX occlusion. Mn-SOD-induced augmentation of RH following 60-s LCX occlusion was not affected by catalase, while it was completely abolished by L-NAME. n contrast to Mn-SOD, Cu,Zn-SOD showed no effect on RH following 60-s LCX occlusion in the dogs pretreated with 8PT. Conclusions: Superoxide radicals generated during ischemia for 60 s and reperfusion attenuates myocardial RH through inactivation of EDNO. Mn-SOD shows more beneficial effects on myocardial RH than Cu,Zn-SOD. Keywords: Reactive hyperemia; Nitric oxide; Adenosine; Dogs; Myocardial &hernia; L-NAME; Reperfusion; Free radicals; Free radical scavengers; Dog, anesthetized 1. ntroduction Nitric oxide (NO) from the vascular endothelium has been shown to play important roles in the regulation of coronary circulation [l-5]. Previous studies showed that NO is synthesized from L-arginine [6] and this synthesis is inhibited by L-arginine analogues in both in vivo and in vitro experiments [7,8]. We and other investigators recently reported that endothelium-derived NO (EDNO) is involved in the mechanism of myocardial reactive hyperemia following transient interruption of coronary blood flow, which is one of the important pathophysiological aspects of coronary circulation [ Reactive hyperemia has been shown to be also mediated by other mechanisms including the effect of reduced oxygen tension on the vascular smooth muscle of the coronary resistance vessels [ 12,131, myogenic relaxation of coronary vascular smooth muscle caused by decrease in coronary transmural pressure during coronary occlusion [ 14,151, activation of ATP-sensitive potassium channels during coronary occlusion [ 16,171 and hypoxia-induced release of vasodilator metabolites such as adenosine [18,19]. Recently, superoxide dismutase (SOD) was reported to augment reactive hyperemia following 1-min coronary occlusion by enhancing adenosine release, indicating that superoxide radicals generated during -min ischemia and * Tel.: ( + Sl-96) ; fax: (+ 81-%) Time for primary review 28 days. OCO8-6363/%/$ Elsevier Science B.V. All rights reserved SD/ (95)00213-S

2 538 R. Tsunoda et al./cardiovascular Research 31 (1996) subsequent reperfusion attenuate reactive hyperemia by decreasing adenosine release [20]. Superoxide radical is also known to inactivate EDNO [21,22], but it is still unclear whether superoxide radicals generated during myocardial ischemia and reperfusion influence reactive hyperemia by reacting with EDNO. n the present study, we examined the effect of manganese (M&SOD on reactive hyperemia in anesthetized dogs in which adenosine s effect was blocked by pretreatment with 8-phenyltheophylline (8PT) or both adenosine s and EDNO s effects were blocked by pretreatments with 8PT and NC-nitro-Larginine methyl ester (L-NAME). n the dogs pretreated with 8PT, we also examined the effect of simultaneous administration of Mn-SOD and catalase on reactive hyperemia. Furthermore, the effect of copper, zinc (Cu,Zn)-SOD on reactive hyperemia was also examined in the dogs pretreated with 8PT for comparison with that of Mn-SOD. 2. Methods 2.1. Surgical preparation Forty-eight adult mongrel dogs of either sex weighing 11 to 18 kg were anesthetized by intravenous administration of sodium pentobarbital (30 mg/kg body weight), intubated, and mechanically ventilated with room air mixed with oxygen. Arterial blood gases and blood ph were maintained within physiological ranges by adjusting respiratory rate and tidal volume. A left thoracotomy was performed at the fifth intercostal space, and the heart was suspended in a pericardial cradle. Approximately 1 cm of the proximal portion of the left circumflex coronary artery (LCX) was gently dissected free before its first marginal branch. After heparin (200 U/kg bolus injection, followed by 1000 U every 30 min) and aspirin (17 mg/kg bolus injection) were administered intravenously, the LCX was cannulated and perfused with arterial blood from the left common carotid artery through an extracorporeal bypass tube. An electromagnetic flow probe (Nihon-Kohden, FF- 045T, Tokyo, Japan) was placed in the middle of the bypass tube. Tbe LCX blood flow was continuously measured with an electromagnetic flowmeter (MFV 3200, Nihon-Kohden) and recorded using a polygraph (MC- 9800, Fukuda Denshi, Tokyo, Japan) together with the electrocardiogram (lead ), arterial blood pressure and left ventricular pressure which was measured via the catheter inserted into the left ventricle at the apical dimple. A Goodale-Lubin catheter (USC11 inserted through the left cervical vein was placed in the coronary sinus for sampling coronary venous blood and measuring oxygen content Preparation of drugs Mn-SOD (10200 U/mg, Bacillus sp.). Cu,Zn-SOD (3 800 U/mg, bovine erythrocyte), catalase (6 700 U/mg, bovine liver), adenosine, MnCl,, CuCl, and ZnCl, were obtained from Wako Pure Chemical ndustries (Osaka, Japan), L-NAME and 8PT from Sigma Chemical (St. Louis, MO) and acetylcholine from Daiichi Pharmaceutical (Tokyo, Japan). Mn-SOD and C&Q-SOD were dissolved with normal saline to get a concentration of 10 pg/kg/min and 25 pg/kg/min, respectively, when infused into the LCX al. a rate of 0.5 ml/min. These doses of SOD were almost equal to 100 U/kg/min. Catalase was dissolved with normal saline to get the concentration of 30 pg/kg/min when infused into the LCX at a rate or 0.5 ml/mm. That dose of catalase was almost equal to 200 U/kg,/min. L-NAME was dissolved with normal saline to get a concentration of 300 PM in the LCX blood when infused into the LCX at a rate of 0.5 ml/min. 8PT was dissolved with 1 ml of 80% ethanol containing 0.2 N NaOH and diluted with normal saline to get a concentration of 30 pg/kg/min when infused into the LCX at a rate of 0.5 ml/min. Acetylcholine and adenosine were dissolved with normal saline to get a concentration of 100 rig/kg and 1 pg/kg body weight per 0.1 ml, respectively. CuCl, and ZnCl, were dissolved with normal saline so that the concentrations of Cu and Zn would be equal to their concentrations when Cu,Zn-SOD was prepared by dissolving with normal saline. MnCl, was dissolved with normal saline so that the concentration of Mn would be equal to its concentration when Mn-SOD was prepared by dissolving with normal saline. These drugs were injected into the LCX from a middle portion of the bypass tube Eqerimental protocol After a 30-min period was allowed for stabilization, the baseline LCX blood flow, heart rate, arterial blood pressure and left ventricular pressure were measured. The dogs were assigned randomly to one of the following seven study protocols. The protocols included the experiments without any pretreatments, with pretreatment with 8PT, and wilh pretreatments with 8PT and L-NAME (Fig. 1). Protocol A: The effects of Mn-SOD and Cu,Zn-SOD on a reactive hyperemic response following 60-s coronary occlusion were examined in 8 dogs. After baseline observation of reactive hyperemia following 60-s coronary occlusion, Mn-SOD (10 pg/kg/min; n = 4) or Cu,Zn-SOD (25 pg/kg/min; n = 4) was infused into the LCX. Ten minutes after the continuous infusion of either SOD, the LCX was occluded for 60 s and reactive hyperemia was again observed. Protocol B: The effect of Mn-SOD on reactive hyperemia was examined in 12 dogs pretreated with 8PT. After a baseline measurement of the LCX flow and hemodynamic parameters, 8PT (30 pg/kg/min) was continuously infused into the LCX. Five minutes after the initiation of 8PT, reactive hyperemia following 10-s (n = 6) or 60-s (n = 6) LCX occlusion was observed. After the LCX flow and hemodynamic parameters returned to the baseline

3 R. Tsunoda et al./cardiouascular Research 31 (1996) Protocol A (n-8) Protocol B (n=l2) RH locf60 RH loor Ado1.opJdcg. smin 2 v 8PT RH 60 Protocol C (n=5). Mn-SOD 10 min Adol.Opghg l 5min. T 8PT cu~soo 10 mhl RH 80 * Protocol D (n=6) RH 6~ RH ea Ado1.Opgkg AChlWn& kg Protocol E (n=6) l l smbl 1.f Ado1.opgkg l 5nM a* RH 80 l - loti CatBtllaSO f RH 60 l 1 Fig. 1. Experimental protocols. See text for details. RH = reactive hyperemia; Mn-SOD = manganese-superoxide dismutase; Cu,Zn-SOD = copper, zinc-superoxide dismutase; 8FT = &phenyltheophylline; Ado = adenosine; LNAME == NC-nitro-L-arginine methyl ester; ACh = acetylcholine. values, Mn-SOD (10 pg/kg/min) was continuously infused into the LCX in addition to 8PT. Ten minutes after Mn-SOD infusion, the LCX was occluded for 10 or 60 s and reactive hyperemia was observed. The effect of 8PT was tested by the intracoronary infusion of adenosine (1 pg/kg) before and 5 min after administration of 8PT. 8PT and Mn-SOD infusion was continued throughout the study. n 4 of the 12 dogs, blood samples were obtained from the aorta and coronary sinus before and 10 min after Mn-SOD infusion for the measurements of oxygen content and myocardial oxygen consumption rate. Protocol C: For comparison with the effect of Mn-SOD on reactive hyperemia, the effect of Cu,Zn-SOD infused into the LCX at 25 pg/kg/min on reactive hyperemia following 60-s coronary occlusion was examined in 5 dogs pretreated with 8PT. The experimental protocol was the same as in Protocol B except for the type of SOD used. Protocol D: The influence of L-NAME on the effect of Mn-SOD was examined in 5 dogs pretreated with 8PT. L-NAME (300,uM in the LCX blood) was administered into the LCX for 6 min. This dose of L-NAME was found to abolish a vasodilator response to acetylcholine without markedly altering systemic hemodynamic parameters One minute after the initiation of L-NAME infusion, 8PT (30 pg/kg/min) was also administered into the LCX. The effects of 8PT and L-NAME were tested by the intracoronary infusion of adenosine (1 pg/kg) and acetylcholine (100 rig/kg), respectively, before and after both 8PT and L-NAME infusion. After a baseline measurement of reactive hyperemia following 60-s LCX occlusion, Mn- SOD (10 pg/kg min) was continuously infused into the LCX. Ten minutes after the infusion of Mn-SOD, reactive hyperemia after 60-s LCX occlusion was again observed. The infusion of 8PT and Mn-SOD was continued throughout this study protocol. n 4 of the 5 dogs, myocardial oxygen consumption rate was measured before and 10 min after Mn-SOD infusion as described above. Protocol E: To examine the role of hydrogen peroxide, other oxygen-derived free radicals, in the effect of Mn-SOD on reactive hyperemia, we administered Mn-SOD and catalase simultaneously in 6 dogs pretreated with 8PT. Five minutes after the initiation of 8PT infusion into the LCX (30 pg/kg/ min), reactive hyperemia following 60-s coronary artery occlusion was observed. After catalase (30 pg/kg/min) and Mn-SOD (10 pg/kg/min) were continuously administered into the LCX in addition to 8PT for 10 min, reactive hyperemic response after 60-s coronary artery occlusion was again observed. ProtocoE F: The reproducibility of reactive hyperemia was examined in 4 dogs pretreated with 8PT. Reactive hyperemia following 60-s coronary occlusion was observed twice after 8PT administration. The time interval between these two observations was equal to that between the first observation of reactive hyperemia after 8PT and the second one after Mn-SOD in Protocol B. Protocol G: To examine whether Mn or Cu and Zn themselves affect reactive hyperemic response, we observed the effect of Mn or Cu and Zn on a reactive hyperemic response following 60-s coronary occlusion in 8 dogs in which adenosine s effect was blocked. Five minutes after continuous infusion 8PT (30 pg/kg/min) reactive hyperemia following 60-s coronary occlusion was observed. Mn (n = 4) or both Cu and Zn (n = 4) was infused into the LCX for 10 min and reactive hyperemia was again observed. The concentrations of Mn and both Cu and Zn in the LCX blood were the same as those in the Protocols B and C. At the conclusion of each protocol, Evans blue was injected into the LCX to determine the perfusion area. The heart was excised and the weight of the perfusion area was measured. All experiments were in accordance with the guidelines on experimental animals issued by Kumamoto University School of Medicine, and approved by the Center of Laboratory Animals Dgfinitions and calculations The blood flow debt, excess flow during reactive hyperemia and repayment of flow debt were calculated as described by Coffman and Gregg [24]: Blood flow debt

4 540 R. Tsunoda et al./cardiovascular Research.31 (1996) (ml) = control flow rate (ml/s) X occlusion time (s); excess flow during reactive hyperemia (ml) = total flow during reactive hyperemia - [control flow rate (ml/s) X duration of reactive hyperemia (s)]; repayment of flow debt (%) = [excess flow during reactive hyperemia (ml)/blood flow debt (ml)] x 100. The duration of reactive hyperemia was the time interval from the release of LCX occlusion to the point at which LCX flow returned to the baseline value. Myocardial oxygen consumption rate (ml/min/ 100 g) was calculated as LCX flow (ml/min/loo g) x [oxygen content difference between aorta and coronary sinus (ml/dl)] Data analysis All data are expressed as mean + s.e.m. The effects of acetylcholine and adenosine on LCX flow were expressed as an absolute value of the increased flow induced by the drugs (i.e., the area under the flow curve). Student s paired t-test was used for statistical analysis of the differences in LCX flow, hemodynamic parameters, myocardial oxygen consumption rate and reactive hyperemia before and after administration of each of 8PT, L-NAME, Mn-SOD, Cu,Zn-SOD, Mn-SOD and catalase, MnCl,, CuCl, and ZnCl,. The differences between the LCX flow response to acetylcholine and adenosine before and after L-NAME and/or 8PT were also analyzed by Student s paired t-test. The level of significance was P < Results 3.1. Effect of Mn-SOD and Cu,Zn-SOD on reactive hyperemia in the dogs without pretreatment with 8PT (Protocol A) shown)., while each of them significantly augmented excess flow and repayment of flow debt during reactive hyperemia after 60-s coronary occlusion (Table 1). n addition, Mn-SOD increased the duration of reactive hyperemia Efect of Mn-SOD on reactive hyperemia (Protocol B) ntracoronary infusion of 8PT (30 pg/kg/min) had no appreciable effects on LCX flow and hemodynamic parameters (Table 2), although it significantly decreased adenosine-induced LCX flow increase from 61.8 f 19.1 to 3.5 f 1.0 ml/100 g (P < 0.01). Additional infusion of Mn-SOD did not affect either the LCX flow or any of the hemodynamic parameters. Myocardial oxygen consumption rates before and after Mn-SOD were 13.2 f 1.5 ml/min/loo g and ml/min/loo g, respectively (P = NS). Fig. 2 demonstrates the representative recording of the change in reactive hyperemia following 60-s LCX occlusion after Mn-SOD administration in a 8PT-pretreated dog. Mn-SOD augmented the excess flow and repayment of flow debt during reactive hyperemia, although there was no change in the peak flow rate. The augmentation of reactive hyperemia by Mn-SOD was mainly seen in the later portion of reactive hyperemia. The upper panel of Table 3 summarizes the effects of Mn-SOD on reactive hyperemia. Mn-SOD significantly increased excess flow and repayment of flow debt following 60-s LCX occlusion (Fig. 3) although it did not affect peak flow rate. The duration of reactive hyperemia did not significantly change after h/m-sod. n contrast, Mn-SOD did not affect reactive hyperemia following 10-s LCX occlusion Eflect of Cu,Zn-SOD on reactive hyperemia (Protocol C) Each of Mn-SOD or Cu,Zn-SOD did not change any of ntracoronary infusion of 8PT (30 pg/kg/min) did not the LCX flow and hemodynamic parameters (data not affect the LCX flow and the systemic hemodynamic pa- (muminute+wo g) baseline (ml/minuwloo g) Mn-SOD ; :: 100 J 0 3 1,+ ~~ :- ~~ 1 minute Fig. 2. Representative recordings of reactive hyperemia following 60-s coronary occlusion at baseline and after administration of manganese-superoxide dismutase (Mn-SOD) in a dog pretreated with 8-phenyltheophylline. Mn-SOD significantly augmented excess flow and repayment of flow debt during reactive hyperemia although it showed little effect on peak flow rate.

5 R. Tsunoda et al./cardiovascular Research 31 (1996) Table 1 The effect of Mn- and Cu,Zn-SOD on reactive hyperemic response to 60-s coronary occlusion Mn-SOD Cu,Zn-SOD Before After Before Basal flow rate (ml/mitt/100 g) 98.0 f f f 7.1 Peak flow rate (ml/mm/ 100 g) 361.3f f f 40.8 Excess flow (ml/100 g) f f 57.9 b f 59.7 Repayment of flow debt (o/o) k55 b 277 f 64 Duration of reactive hypetemia 6) f f 26.0 a f 46.8 After 93.1* f f 64.3 a 321 k70 b f 22.3 Mn-SOD = manganese-superoxide dismutase; Cu,Zn-SOD = copper, zinc-superoxide dismutase. Values are mean f s.e.m. a P < b P < 0.01 vs before. Table 2 LCX flow and hemodynamic parameters before and after each treatment 8PT 8PT + Mn-SOD 8PT + L-NAME 8PT + L-NAME + Mn-SOD Before After Before After Before After Before After LCX flow (ml/min/ 100 g) 91.7* f f k rt & f lt21.8 Heart rate (beats/min) 159f6 159k6 160&-6 16Of6 157*9 150*7 145*9 147k8 Mean BP hmhg) 126k8 127f8 134k6 135*7 118f6 123&6 13Oic8 127f8 LVSP hmhd 145*7 147f7 155k6 155k6 135k6 139k6 146*9 145f9 LVdp/dt (mmhg/s) 2513f k f f k160a 2122f f 170 LVEDP (mmhg) 5.1 f f f rt ~b & *2.2 LCX = left circumflex coronary artery; 8PT = 8-phenyltbeophylline; L-NAME = NC-nitro-L-arginine methyl ester; BP = blood pressure; LVSP = left ventricular systolic pressure; LVdp/dt = left ventricular dp/dt; LVEDP = left ventricular end-diastolic pressure. Other abbreviations are as in Table 1. Values are meanfs.e.m. a P < 0.05 vs. before. Table 3 Effects of Mn-SOD on reactive hypemmia following lo- and 60-s coronary occlusion in PT-pretreated and 8pT- and L-NAME-pretreated dogs 10-s reactive hypemmia 60-s reactive hyperemia 8PT-pretreated dogs Basal flow rate (ml/mm/ 100 g) 101.3* f 11.7 Peak flow rate (ml/mm/100 g) 237.4* f 27.0 Excess Plow (ml/ 100 g) 34.2 f f 7.8 Repayment of Plow Debt (%) 196k35 193f34 Duration of reactive hyperemia 6) 88.1 f 18.5 loo.6* PT- and L-NAME-pretreated dogs Basal flow rate (ml/min/loo g) Peak flow rate (ml/mm/100 g> Excess flow (ml/100 g) Repayment of flow debt (%) Duration of reactive hyperemia (9 Baseline Mn-SOD Baseline Mn-SOD 97.3 f f * f k f22.1 b 147* &18a f f f rt f k *31.3 %.O f f * f f 52.4 Abbreviations are as in Tables 1 and 2. Values are mean f s.e.m. P < 0.05, b P < 0.01 vs. baseline. Table 4 The effect of Cu. Zn-SOD on reactive hyperemic response to 60-s coronary occlusion in PT-pretreated dogs Baseline Basal flow rate (ml/min/loo g) f 12.9 Peak flow rate (ml/min/loo g) f 53.6 Excess flow (ml/100 g 185.6k21.7 Repayment of flow debt (% 165f 18 Duration of reactive hyperemia ( f 29.3 Cu,Zn-SOD f f f f 49.2 Abbreviations are as in Tables 1 and 2. Values are mean f s.e.m.

6 542 R. Tsunoda et al./ Cardiovascular Research 31 (1996) lmw1w gl PQ.01 i P&O5-1omc wsec aosec WT-pretreated Dogs WT. and LNAME. Retmated Cegs Fig. 3. Excess flow and repayment of flow debt during reactive hyperemia in S-phenyltheophylline (@ T&pretreated and both WT- and NC-r&o-L-arginine methyl ester (LNAME)-preheated dogs. Manganese-superoxide dismutase (Mn-SOD) significantly increased excess flow and repayment of flow debt following 60-s but not 10-s coronary occlusion in 8PT-pretreated dogs. Mn-SOD-induced augmentation of reactive hyperemia was completely abolished in SPT- and LNAME-pretreated dogs. rameters (data not shown), while it significantly decreased adenosine-induced LCX flow increase from to ml/100 g (P < 0.01). n contrast to the effect of Mn-SOD on reactive hyperemia following 60-s coronary occlusion, Cu,Zn-SOD did not affect any of the parameters of reactive hyperemia in the dogs pretreated with W (Table 4) nfluence of L-NAME on the efect of Mn-SOD on reactive hyperemia (Protocol D) Simultaneous administration of L-NAME and 8PT did not alter the LCX flow and hemodynamic parameters except for left ventricular dp/dt which were slightly but significantly decreased after 8PT and L-NAME (Table 2). The flow-increasing effects of adenosine and acetylcholine were both significantly attenuated after infusion of 8PT and L-NAME. Adenosine increased LCX flow by f 6.4 ml/100 g before 8PT, but by 1.9 f 0.8 ml/100 g after 8PT (P < 0.01). Acetylcholine increased LCX flow by ml/100 g before L-NAME, but by ml/ 100 g after L-NAME (P < 0.05). The administration of Mn-SOD in addition to 8PT and L-NAME had no appreciable effects on LCX flow and hemodynamic parameters. As shown in the lower panel of Table 3 and Fig. 3, any of the parameters of reactive hyperemia following 60-s LCX occlusion did not change after Mn-SOD. Myocardial oxygen consumption rates before and after Mn-SOD were 8.1 f 2.2 and 8.0 rtr 2.1 ml/min/loo g, respectively (P = NS) The effect of simultaneous administration of Mn-SOD and catalase on reactive hyperemia (Protocol E) Simultaneous administration of catalase and Mn-SOD showed no appreciable effects on LCX flow, hemodynamic parameters and myocardial oxygen consumption rate (data not shown). Additional catalase administration of Mn-SOD did not influence the enhancing effect of Mn-SOD on reactive hyperemia after 60-s coronary occlusion (Table 5) Reproducibility of reactive hyperemic response (Protocol F) There was no significant difference between the first and second reactive hyperemic responses to 60-s coronary occlusion in 8PT-pretreated dogs. Repayments of flow debt after first and second 60-s coronary occlusions were 188 f 35 and 181 f 34%, respectively (P = NS) Effect of Mn and both Cu and Zn themselves on reactive hyperemia in the dogs pretreated with 8PT (Protocol rs) ntracoronary infusion of Mn or both Cu and Zn did not affect LCX flow and systemic hemodynamic parameters Table 5 The effect of simultaneous administration of Mn-SOD and catalase on reactive hypetemic response to 60-s coronary occlusion in PT-pretreated dogs Baseline Mn-SOD + catalase Basal flow rate (ml/mm/ 100 g) 93.0& zk 13.6 Peak flow rate (ml/min/ 100 g) 313.4k * 33.9 Excess flow (ml/100 g) 166.6& f 20.0 b Repayment of flow debt (a) 18!)f f19a Duration of reactive hyperemia ( f rt 14.5 a Abbreviations are as in Tables 1 and 2. Values are mean f s.e.m. a P < 0.05, b P < 0.01 vs. baseline.

7 R. Tsunoda et al./ Cardiovascular Research 31 (1996) (data not shown) and did not enhance the reactive hyperemit response after 60-s coronary occlusion. Repayments of flow debt after 60-s coronary occlusion before and after Mn were 184 f 11 and 178 &- 17%, respectively (P = NS). Those before and after Cu and Zn were 193 k 11 and %, respectively ( P = NS). 4. Discussion Nitric oxide (NO) from the vascular endothelium has been shown to play important roles in the regulation of coronary circulation [l-5]. We have shown that intracoronary infusion of L-NAME significantly decreased basal LCX flow without markedly changing systemic hemodynamic parameters and indicated the presence of the basal release of nitric oxide from the endothelium [23,25]. f superoxide radical is generated in coronary circulation in the basal condition and react with basally released nitric oxide as reported previously 121,221, SOD is expected to increase basal LCX blood flow. However, either Mn-SOD or Cu,Zn-SOD showed no effect on basal LCX flow in the present study. None of the hemodynamic parameters and myocardial oxygen consumption rate changed after SOD treatment. Thus, in the basal condition, superoxide radical is not produced in an amount enough to affect a vasodilator action of nitric oxide and thereby modify coronary flow Mechanism of Mn-SOD-induced augmentation of reactive hyperemia The present study showed that in both groups of dogs with and without pretreatment with 8FT, Mn-SOD significantly augmented a reactive hyperemic response to 60-s coronary artery occlusion but not to 10-s coronary occlusion. Furthermore, this study showed that this Mn-SOD-induced augmentation of reactive hyperemia was completely abolished by pretreatment with L-NAME. All dogs were pretreated with 8pT in order to eliminate the vasodilator mechanism of adenosine in reactive hyperemia which has been shown to be enhanced by SOD treatment [20]. Mn- SOD administered simultaneously with 8PT augmented the reactive hyperemic response to 60-s coronary occlusion without any significant changes in systemic hemodynamic parameters and myocardial oxygen consumption rate, a major determinant of the coronary blood flow [26]. This indicates that Mn-SOD-induced augmentation of reactive hyperemia was not secondary to the effects on systemic hemodynamics but was due to its primary scavenging effect on superoxide radicals. Furthermore, the possibility that Mn which might have leached from SOD influenced nitric oxide was excluded, since the equivalent dose of Mn showed no effect on reactive hyperemia. Oxygen-derived free radicals are known to affect the coronary arteries directly [27,28]. f Mn-SOD-induced aug- mentation of reactive hyperemia is due to protection from damage of the vascular wall induced by free radicals, it should be also seen in the dogs pretreated with 8PT and L-NAME. However, pretreatment with L-NAME completely abolished Mn-SOD-induced augmentation of reactive hyperemia. Furthermore, if hydrogen peroxide which is generated through the dismutation of superoxide radicals by Mn-SOD might have augmented the effect of nitric oxide on reactive hyperemia, Mn-SOD-induced augmentation of reactive hyperemia would have been eliminated in the dogs in which catalase was administered simultaneously. However, catalase did not influence the effect of Mn-SOD. Thus, Mn-SOD was likely to augment a reactive hyperemic response through enhancement of the effect of nitric oxide which has been shown to be inactivated by superoxide radical [21,22]. Mn-SOD-induced augmentation of reactive hyperemia was not observed after 10-s coronary occlusion but prominent after 60-s coronary occlusion. This suggests that a sufficient amount of superoxide radicals enough to attenuate reactive hyperemia is generated after a relatively long period of ischemia. A previous study by Kitakaze et al. also showed that SOD augmented a reactive hyperemic response to 60-s coronary artery occlusion but not those to 15, 30-, or 45-s occlusion [20] Role of EDNO in reactive hyperemia This study showed that Mn-SOD significantly increased excess flow and repayment of flow debt but did not affect peak flow rate of reactive hyperemia. There was no significant difference in the duration of reactive hyperemia before and after Mn-SOD, although that after Mn-SOD following 60-s coronary occlusion tended to be prolonged as compared with that before SOD. Thus, it is suggested that nitric oxide derived from the endothelium mediates mainly the later portion of reactive hyperemia, especially after the peak. The former portion of reactive hyperemia and peak flow are probably mediated by other mechanisms such as myogenic tone [ 14,151. t has been shown that the peak flow rate of reactive hyperemia occurred approximately 6 s after reperfusion of a 15-s coronary occlusion, while flowmediated vasodilation of a large coronary artery, which has been shown to be mediated by EDNO 129,301, was delayed, not reaching a maximum until 60 s after reperfusion [31]. This previous observation also suggests that a nitric oxide mechanism in vasodilation is delayed after reperfusion of coronary artery occlusion. Our previous report [ 111 showed that 8PT attenuated repayment of flow debt after 20-s coronary occlusion by 30%, while N G-monomethyl-L-arginine (L-MMA) by 34%. Simultaneous administration of L-MMA and 8PT further decreased repayment of flow debt by 57%, indicating that each effect of ST and L-MMA was additive to each other. These results suggest that myocardial reactive hyperemia was mediated by nitric oxide as well as adeno-

8 544 R. Tsunoda et al./cardiovascuiar Research 31 (1996) sine. Our previous report also showed that L-MMA-induced change in the repayment of flow debt following 60-s coronary occlusion was significantly smaller than those following lo- and 20-s coronary occlusion [ 111. This may be explained by inactivation of nitric oxide by superoxide radicals which were generated after 60-s coronary occlusion but not after 10-s and 20-s coronary occlusion. Thus, it is scsgested that the role of endothelium-derived nitric oxide is important in reactive hyperemia following a short duration of coronary occlusion, while it is less important or may be of no importance in reactive hyperemia following a long duration of coronary occlusion because of concomitant generation of superoxide radicals Comparison of the effects of Mn-SOD and Cu,Zn-SOD on reactive hyperemia Both Mn-SOD and Cu,Zn-SOD significantly augmented a reactive hyperemic response to 60-s coronary artery occlusion in dogs not pretreated with 8PT. n dogs pretreated with 8PT, however, Mn-SOD significantly augmented reactive hyperemia following 60-s coronary occlusion, while Cu,Zn-SOD did not. n the dogs not pretreated with 8PT, however, Cu,Zn-SOD significantly augmented the hyperemic response to 60-s coronary occlusion as well as Mn-SOD did, indicating that Cu,Zn-SOD also has superoxide radical scavenging activity. Furthermore, simultaneous intracoronary infusion of Cu and Zn had no appreciable effect on reactive hyperemia following 60-s coronary occlusion. This result suggests that Cu and Zn themselves did not affect nitric oxide. Kitakaze et al. reported that SOD augmented myocardial reactive hyperemia after 60-s coronary occlusion by protecting 5 -nucleotidase activity from superoxide radicals and thus increasing adenosine release, and that this SODinduced augmentation of reactive hyperemia was completely abolished by pretreatment with 8PT [20]. The present findings that each of Mn-SOD and Cu,Zn-SOD significantly augmented a reactive hyperemic response to 60-s coronary artery occlusion and that Cu,Zn-SOD failed to augment reactive hyperemia in dogs pretreated with 8PT are consistent with this previous report. n contrast to Cu,Zn-SOD, Mn-SOD did augment a reactive hyperemic response to 60-s coronary occlusion even in dogs pretreated with 8PT. This may be explained by the difference in the type of SOD used. The present study did not clarify the mechanism responsible for the difference in effect of the two types of SOD. Previous in vitro study showed that Cu,Zn-SOD produces hydroxyl radical, which is a more potent toxic radical and was reported to damage endothelium-dependent relaxation selectively [32], by reacting with hydrogen peroxide which is generated through the dismutation of superoxide radicals, while Mn-SOD does not [33]. n addition, Mn-SOD was reported to equilibrate much faster than other types of SOD in the interstitial fluid and the concentration of SOD was correlated with its protec- tive effect on ischemia and reperfusion injury in the Langendorff rabbit heart [34]. These may explain the difference in effect of Mn-SOD on reactive hyperemia in 8PTpretreated dogs from that of Cu,Zn-SOD. Further studies are required to clarify the precise mechanism responsible for the difference. 5. Conclusion We demonstrated that Mn-SOD augmented reactive hyperemia by protecting EDNO from superoxide radicals generated during 60-s coronary ischemia and reperfusion. Mn-SOD seems to have more beneficial effects on coronary circulation during ischemia and reperfusion than CuZn-SOD. Acknowledgements This study was partly supported by Grant-in-Aid for Scientific Research from the Ministry of Education, Science and Culture (CO and by Research Grant from the Smoking Research Foundation, Tokyo, Japan. References 111 Furchgott RF, Zawadzki JV. The obligatory role of endothelial cells in the relaxation of arterial smooth muscle by acetylcholine. Nature 1980;288: l21 Furchgott RF. Role of endothelium in responses of vascular smooth muscle. Circ Res 1983;53: [31 Kelm M, Schlader J. Control of vascular tone by nitric oxide. Circ Res 1990;66: [41 Lhscher TP, Vanhoutte PM eds. The endothelium: modulator of cardiovascular function. Boston: CRC Press, 1990: l-6. [51 Palmer RMJ, Ferrige AG, Moncada S. Nitric oxide release accounts for the biological activity of endothelium-derived relaxing factor. Nature 1987;327: l61 Palmer RMJ, Ashton DS, Moncada S. Vascular endothelial cells synthesize nitric oxide from L-arginme. Nature 1988;333: [71 Palmer RMJ, Rees DD, Ashton DS, Moncada S. L-Arginine is the physiological precursor for tbe formation of nitric oxide in endothelium-derived relaxation. Biochem Biophys Res Commun 1988;153: la1 Rees DD, Palmer RMJ, Schulz R, Hodson HF. Moncada S. Characterization of three inhibitors of endothelial nitric oxide synthase in vitro and in vivo. Br J Pharmacol 1990;101: [91 Ahman JD, Kimr J, Duncker DJ, Bathe RJ. Effect of inhibition of nitric oxide formation on coronary flow during exercise in the dog. Cardiovasc Res 1994;28: l101 Kostic MM, Schrader J. Role of nitric oxide in reactive hyperemia of guinea pig heart Circ Res ~ Yamabe H, Oknrmura K, lshizaka H, Tsuchiya T, Yasue H. Role of endothelium-derived nitric oxide in myocardial reactive hyperemia. Am J Physiol 1992;263:H8-H14. l121 Guyton AC, Ross J, Carrier 0. Walker JR. Evidence for tissue oxygen demand as the major factor causing autoregulation. Circ Res 1~,15(Suppl1): Matini J, Honig CR. Direct measurement of intercapillary distance 1131

9 R. Tstutoda et al./cardiovascular Research 31 (1996) in beating rat heart in situ under various conditions of 0, supply. Microvasc Res 1%9;1: [14] Dole WP, Montville WJ, Bishop VS. Dependency of myocardial reactive hypemmia on coronary artery pressure in the dog. Am J Physiol 1981;24OzH70!3-H715. il.51 Giles RW, Wilcken DEL. Reactive hyperemia in the dog heart: evidence for a myogenic contribution. Cardiovasc Res 1977;11: Aversano T, Duyang P, Silverman H. Blockade of the ATP-sensitive potassium channel modulates reactive hyperemia in the canine coronary circulation. Circ Res 1991;69: [17] Kanatsuka H, Sekiguchi N, Sato K, Akai K, et al. Microvascular sites and mechanisms responsible for reactive hyperemia in the coronary circulation of the beating canine heart. Circ Res 1992;71: [18] Bern RM. Regulation of coronary blood flow. Physiol Rev 1964$4: l-29. [19] Saito D, Steinhart CR, Nixon DG, Ollson RA. ntracoronary adenosine deaminase reduces canine myocardial reactive hyperemia. Circ Res 198 1;49: [20] Kitakaze M, Hori M, Takashima S, et al. Superoxide dismutase enhances &hernia-induced reactive hypetemic flow and adenosine release in dogs: a role of nucleotidase activity. Circ Res 1992;71: Grygrewski RJ, Palmer RMJ, Moncada S. Superoxide anion is involved in the breakdown of endothelium-derived vascular relaxing factor. Nature 1986;320: [22] Rubanyi GM, Vanhoutte PM. Superoxide anions and hypoxia inactivate endothelium-derived relaxing factor. Am J Physiol :H822-H827. [23] Tsunoda R, Okumura K, shizaka H, Matsunaga T, Yasue H. nfluence of the inhibition of endothelium-derived nitric oxide formation to the effects of vasoconstrictor agents, neuropeptide Y, clonidine and ergonovine, on coronary vascular resistance. J Cardiovast Pharmacol [24] Coffman JD, Gregg DE. Reactive hyperemia characteristics of the myocardium. Am J Physiol 1960; 199: [25] Matsunaga T, Okumura K, shizaka H, Tsunoda R, Yasue H. Effect of cumulative doses of NC-nitro-L-arginine methyl ester on coronary flow of anesthetized and conscious dogs. Arch nt Pharmacodyn 1994;327: Bckenhoff JE, Hafkenschiel JH, Landmesser CM, Harmel M. Cardiac oxygen metabolism and control of the coronary circulation. Am J Physiol 1947; [27] Lamb FS, Webb RC. Vascular effects of free radicals generated by the electron stimulation. Am J Physiol 1984,247:H70!3-H714. [28] Linder V. Heinle H. Dose the xanthine-xanthine oxidase system alter contractile behaviour of vascular smooth muscle? Pfliigers Arch 1987;408: [29] Kuo L, Davis J, Chillian W.M. Endothelium-dependent, flow-induced dilation of isolated coronary arterioles. Am J Physiol 19!30;259:H1063-H1070. [30] Rubanyi GM, Romero JC, Vanhoutte PM. Flow-induced release of endothelium-derived relaxing factor. Am J Physiol 1986;250:H1145-H1149. [31 Hintze TH, Vamer SF. Reactive dilatation of large coronary artery in conscious dogs. Circ Res 1984,54: Todoki K, Okabe E, Kiyose T, Sekishita T, to H. Oxygen free radical-mediated selective endothelial dysfunction in isolated coronary artery. Am J Physiol 1992;267:H806-H812. [33] Sato K, Akaike T, Kohno M, Ando M, Maeda H. Hydroxyl radical production by H,O, plus Cu,Zn-superoxide dismutase reflects the activity of free copper released from the oxidatively damaged enzyme. J Biol Chem [34] Gmar BA, McCord JM. nterstitial equilibration of superoxide dismutase correlates with its protective effect in the isolated rabbit heart. J Mol Cell Cardiol 1991;23: