The Role of Serotonin in Ischemic Cellular Damage and the Infarct Size-Reducing Effect of Sarpogrelate, a 5-Hydroxytryptamine-2

CORE Metadata, citation and similar papers at core.ac.uk

Provided by Elsevier - Publisher Connector

Journal of the American College of Cardiology Vol. 40, No. 7, 2002 © 2002 by the American College of Cardiology Foundation ISSN 0735-1097/02/$22.00 Published by Elsevier Science Inc. PII S0735-1097(02)02158-7 The Role of Serotonin in Ischemic Cellular Damage and the Infarct Size-Reducing Effect of Sarpogrelate, a 5-Hydroxytryptamine-2 Receptor Blocker, in Rabbit Hearts Yasuko Shimizu, MD,* Shinya Minatoguchi, MD,* Kazuaki Hashimoto, MD,* Yoshihiro Uno, MD,* Masazumi Arai, MD,* Ningyuan Wang, MD,* Xuehai Chen, MD,* Chuanjian Lu, MD,* Genzou Takemura, MD,* Masaaki Shimomura, MSC,† Takako Fujiwara, MD,† Hisayoshi Fujiwara, MD* Gifu and Kyoto, Japan

OBJECTIVES We aimed to clarify the relation between sarpogrelate (SG), a 5-hydroxytryptamine (5-HT)-2 receptor blocker, and myocardial interstitial serotonin or infarct size during ischemia and reperfusion. BACKGROUND In cardiac tissues serotonin is rich in vascular platelets, mast cells, sympathetic nerve endings, and the receptors are present in platelets and cardiomyocytes. METHODS The myocardial interstitial serotonin levels were measured using a microdialysis technique during 30-min ischemia with and without SG in in vivo as well as isolated rabbit hearts. Other rabbits underwent 30 min of ischemia and 48 h of reperfusion, and the effect of SG on the infarct size was investigated in the absence and presence of a selective protein kinase C (PKC) inhibitor, chelerythrine (5 mg/kg, intravenously), or a mitochondrial adenosine triphosphate sensitive potassium (KATP) channel blocker, 5-hydroxydecanoate (5-HD) (5 mg/kg, intravenously). In another series, the effect of SG on PKC isoforms in cytosol and membrane fraction was assessed after a 20-min global ischemia in isolated rabbit hearts. RESULTS Interstitial serotonin levels were markedly increased during 30-min ischemia in in vivo and isolated hearts, and the increases were inhibited by SG in each. The infarct size was reduced by SG (27 Ϯ 2% vs. 40 Ϯ 3% of control). This effect was blocked by chelerythrine and 5-HD, respectively. Sarpogrelate further enhanced the ischemia-induced translocation of PKC-⑀ to the membrane fraction. CONCLUSIONS Sarpogrelate reduces the myocardial infarct size by inhibiting the serotonin release followed ⑀ by enhancement of PKC- translocation and opening of the mitochondrial KATP channel in ischemic myocytes. (J Am Coll Cardiol 2002;40:1347–55) © 2002 by the American College of Cardiology Foundation

In general, platelets in the vascular beds are rich in serotonin creased in some patients with angina pectoris (9). It was (5-hydroxytryptamine [5-HT]) in the cytoplasm (1). The reported that blockade of 5-HT2 receptors by ketanserine 5-HT2A receptors are observed on the cell membranes of and cinanserine can protect the isolated rat heart against platelets and smooth muscle cells (2,3), and the binding ischemia in terms of cardiac function (10) and that cinan- with released serotonin can induce release of serotonin from serine has a protective effect on pacing-induced ischemia in platelet cytoplasm, platelet aggregation, and contraction of an in vivo model of dogs (11). Sarpogrelate (SG), a selective smooth muscle cells (4). These effects are inhibited by 5-HT2 receptor blocker, is the only drug, among many 5-HT2 receptor blockers. That is, 5-HT2 receptor blockers 5-HT2 receptor blockers, that is currently clinically used for are also inhibitors of serotonin release (5). On the other patients with arteriosclerotic obliterance. Sarpogrelate may hand, 5-HT1 receptors are present on the cell membrane of be protective against human angina pectoris through an endothelial cells, stimulation of which releases endothelial- increase in collateral circulation (12). However, the role of derived relaxing factor followed by relaxation of smooth serotonin and 5-HT2 receptor blockers in myocardial isch- muscle cells (1). In the cardiac tissues, serotonin has been emic cellular damage is unclear. Our hypothesis is as identiﬁed in platelets in the vascular beds (6), mast cells (7), follows. First, serotonin is released from platelets in the and sympathetic nerve endings (8). The 5-HT2 receptors vascular beds, mast cells, and sympathetic nerve endings at are not found in the mast cells or sympathetic nerve endings the area at risk during ischemia and reperfusion, as indicated but are found in the vascular platelets (1). It is postulated by the increase in myocardial interstitial serotonin. Second, that 5-HT receptors are present in cardiomyocytes, al- SG can inhibit the serotonin release and the bindings though the function is unclear. between serotonin and the 5-HT2 receptor of cardiomyo- The transcardiac plasma serotonin concentration is in- cytes, as indicated by the disappearance of the elevated

From the *Second Department of Internal Medicine, Gifu University School of myocardial interstitial serotonin levels after treatment with Medicine, Gifu and the †Kyoto Women’s University, Kyoto, Japan. Supported, in SG, and, thus, can reduce the infarct size. Therefore, the part, by a Grant-in-Aid for Scientiﬁc Research (C) (10670639) from the Japan aim of the present study was to clarify in rabbit hearts with Society for the Promotion of Science. Manuscript received October 24, 2001; revised manuscript received May 20, 2002, minimum collateral circulation: 1) whether the myocardial accepted June 27, 2002. interstitial serotonin level is increased during ischemia and 1348 Shimizu et al. JACC Vol. 40, No. 7, 2002 Myocardial Serotonin and Infarct Size October 2, 2002:1347–55

probe placement was confirmed at autopsy. The microdi- Abbreviations and Acronyms alysis probe was perfused with Ringer’s solution at a rate of ANOVA ϭ analysis of variance 10 ␮l/min. After a 60-min rest following the completion of EGTA ϭ ethyleneglycoltetraacetic acid the instrumentation, the dialysate was sampled during a ϭ 5-HD 5-hydroxydecanoate 30-min period before ischemia and during a 30-min period 5-HT ϭ 5-hydroxytryptamine HPLC ϭ high-performance liquid chromatography of ischemia and during two consecutive 30-min periods of ϭ KATP adenosine triphosphate sensitive reperfusion. Ischemia was induced by occluding the coro- potassium nary branch by pulling the snare, which was then fixed by ϭ LV left ventricle clamping the tube with a mosquito hemostat. Myocardial PKC ϭ protein kinase C PMSF ϭ phenylmethylsulfonyl fluoride ischemia was confirmed by regional cyanosis and electrocar- SDS-PAGE ϭ sodium dodecyl sulfate-polyacrylamide gel diographic change. Reperfusion was confirmed by the myo- electrophoresis cardial blush over the risk area after releasing the snare. SG ϭ sarpogrelate After the initial preparation but before the coronary occlu- ϭ Tris-HCl 2-amino-2-hydroxymethyl-1,3- sion, the animals were assigned to one of the two groups propanedial-HCl described in the following text. A total of 14 rabbits underwent 30-min ischemia and reperfusion, and, if the increase is blocked by SG; 2) 60-min reperfusion in the presence of saline (control group, whether SG can reduce the myocardial infarct size, and if n ϭ 7) or the 5-HT2 receptor blocker SG (SG group, n ϭ translocation of protein kinase C (PKC) and opening of the 7; 10 mg/kg/h, intravenous, starting 10 min before ischemia

adenosine triphosphate sensitive potassium (KATP) chan- until 20 min after reperfusion). In the preliminary study nels, which are the main pathways of the protective effect in using three rabbits, the plasma concentration of SG in ischemic preconditioning (13), are associated with the arterial blood assessed by high-performance liquid chroma- infarct size-reducing effect. tography (HPLC) was 588 Ϯ 169 ng/ml and 688 Ϯ 308 ng/ml at 10 min and 60 min, respectively, after the start of METHODS venous infusion of SG at a rate of 10 mg/kg/h. The concentration was reported to be adequate to block the Protocol 1: myocardial interstitial serotonin in in vivo 5-HT2 receptors (5). Therefore, the dose of 10 mg/kg/h and isolated beating hearts. IN VIVO STUDY. Surgical prep- was selected in the present study. aration was as previously reported (14). Briefly, rabbits were Isolated heart study. The 12 rabbits were anesthetized and anesthetized with sodium pentobarbital (30 to 40 mg/kg, ventilated as described above. The hearts were excised, intravenous), and additional doses were administered when immediately arrested in ice-cold buffer, mounted, and then required throughout the experiment. They were intubated and ventilated with room air supplemented with low-flow perfused at a constant pressure of 100 cm H2O with oxygen using a mechanical ventilator (tidal volume, 20 to 30 Krebs-Henseleit solution gassed with 95% O2 and 5% CO2 ml, respiratory rate, 20/min to 30/min; Model SN-480-5, at 37°C. After a 20-min stabilization period, the hearts were Shinano, Tokyo, Japan). Serial blood gas analysis was perfused with buffer for 30 min and then global ischemia performed, and ventilatory conditions were adjusted to keep was induced by stopping the flow of perfusate for 30 min, the arterial blood gas within the physiologic range. The left and then reperfused for 30 min without ischemia. Sarpogre- carotid artery and jugular vein were cannulated to monitor late was administered at a dose of 10 mg/kg/h for 10 min peripheral arterial pressure and to administer drugs or saline. before global ischemia in six hearts. Another six hearts The rabbits were then systematically heparinized (500 without SG were used as controls. A microdialysis probe for U/kg). A thoracotomy was performed at the third intercos- dialysate sampling was implanted in the myocardium along tal space, and the heart was exposed after excising the the axis of the ventricular fibers and reached from the pericardium. A 4-0 silk suture on a small curved needle was epicardial outer layer to the endocardial inner layer of the passed through the myocardium beneath a large marginal myocardium. The microdialysis probe was perfused with ␮ branch of the left circumflex coronary artery, which supplies Ringer’s solution at a rate of 10 l/min. After a 20-min rest much of the anterolateral and apical walls of the left following the completion of the instrumentation, the dialy- ventricle (LV). A small vinyl tube was passed into both ends sate was sampled during a 30-min period before ischemia, of the suture to make the snare. and during a 30-min period of ischemia, and during a A microdialysis probe (PNF 1700, 20 mm length, 0.31 30-min period of reperfusion. mm outside diameter, 0.2 mm inside diameter, transverse The dialysate samples obtained from both in vivo and type, 50,000 MW cut-off) for dialysate sampling was isolated hearts were frozen at Ϫ83°C until further analysis. implanted in the risk region of the myocardium, which was The dialysate serotonin concentrations were measured using served by the anterolateral coronary artery along the axis of a reversed-phase HPLC system (MCM column, MC Med- the ventricular fibers and reached from the epicardial outer ical, Tokyo, Japan) coupled with electrochemical detection layer to the endocardial inner layer of the myocardium. The (Coulochem II, ESA, Chemsford, Massachusetts). JACC Vol. 40, No. 7, 2002 Shimizu et al. 1349 October 2, 2002:1347–55 Myocardial Serotonin and Infarct Size

Table 1. Ventricular Fibrillation, Death, and Survival in the Experimental Groups Technical Occlusion VF Reﬂow VF Premature Survivors Groups No. Exclusion (Total No./No. Nonsurvivors) (Total No./No. Nonsurvivors) Death (%) Control 15 1 2/2 1/1 0 11/15 (73) SG (5 mg/kg/h) pre 13 1 0/0 1/1 0 11/13 (84) SG (10 mg/kg/h) pre 12 0 1/1 0/0 0 11/12 (91) SG (10 mg/kg/h) post 14 1 1/1 0/0 1 11/14 (78) Chelerythrine ϩ SG 15 1 1/1 1/1 1 11/15 (73) Chelerythrine 13 1 0/0 0/0 2 11/13 (84) 5-HD ϩ SG 14 1 1/1 0/0 1 11/14 (78) 5-HD 14 1 0/0 1/1 1 11/14 (78)

SG ϭ sarpogrelate; VF ϭ ventricular ﬁbrillation; 5-HD ϭ 5-hydroxydecanoate.

Protocol 2: infarct size. A total of 110 rabbits underwent Protocol 3: assessment of the subcellular distribution of a surgical preparation similar to that in protocol 1 except for PKC-⑀. The 20 rabbits were anesthetized and ventilated as the insertion of the microdialysis probe (Table 1). Two described in the preceding text. The hearts were excised, types of studies (pre-treatment and post-treatment) were immediately arrested in ice-cold buffer, mounted, and per- performed. All rabbits underwent 30 min of ischemia fused at a constant pressure of 100 cm H2O with Krebs- followed by 48-h reperfusion, and the rabbits were ran- Henseleit solution gassed with 95% O2 and 5% CO2 at domly assigned to each treatment group by lot. In the 37°C. After a 20-min stabilization period, the hearts were control group (n ϭ 15), 1 ml of placebo saline (0.9% NaCl, perfused with buffer containing either SG of 10 mg/kg/h pH 7.4) was injected 10 min before ischemia. In the SG (n ϭ 10) or an equivalent volume of saline (n ϭ 10) for 10 pretreated groups, each rabbit was intravenously infused min. Then, 20-min global ischemia was induced in a half of with SG at a rate of 5 or 10 mg/kg/h (n ϭ 13 and n ϭ 12, each group. Another half of each group (nonischemia respectively), starting 10 min before ischemia until 20 min groups) was perfused with buffer for 20 min. We selected 20 after reperfusion for 60 min. In the SG post-treated group min of ischemia because we previously demonstrated that (n ϭ 14), SG (10 mg/kg/h) was intravenously infused PKC-⑀ is maximally translocated at 20 min of ischemia in starting 10 min before reperfusion until 50 min after rabbit hearts (16). At the end of the 20-min ischemia, the reperfusion for 60 min. In the SG ϩ chelerythrine group ventricles were quickly separated, frozen in liquid nitrogen, (n ϭ 15), chelerythrine (a selective PKC inhibitor, 5 mg/kg) and stored at Ϫ83°C. was intravenously administered 10 min before SG infusion Frozen samples were weighed and homogenized in ﬁve (10 mg/kg/h ϫ 60 min) starting 10 min before ischemia. In volumes of buffer containing 300 mM sucrose, 4 mM the SG ϩ 5-hydroxydecanoate (5-HD) group (n ϭ 14), N-2-hydroxyethylpiperazine-N-2-ethanesulfonic acid, 2

5-HD (a mitochondrial KATP channel blocker, 5 mg/kg) mM ethyleneglycoltetraacetic acid (EGTA), 1 mM phenyl- was intravenously administered 10 min before SG infusion methylsulfonyl fluoride (PMSF), and 20 mM leupeptin (10 mg/kg/h ϫ 60 min) starting 10 min before ischemia. using a Polytron homogenizer at the maximum speed in five The chelerythrine group (n ϭ 13) or 5-HD group (n ϭ 14) 5-s bursts. The homogenates were centrifuged at 1,000 ϫ g was intravenously injected with chelerythrine (5 mg/kg) or for 10 min, after which the supernatant was referred to as 5-HD (5 mg/kg) 20 min before ischemia. After the exper- the cytosolic fraction. The pellet (membrane fraction) was iment the chest was closed, and the rabbits were allowed to dispersed in a buffer containing 20 mM 2-amino-2- recover from anesthesia for two days. hydroxymethyl-1,3-propanedial-HCl (Tris-HCl), 1 mM The postmortem examination was as previously reported EGTA, 1 mM PMSF, 20 ␮M leupeptin, and 0.5% Triton (14). At the end of the study, the rabbits were heparinized X-100. The protein concentration in samples was deter- (500 U/kg) and killed by an overdose of pentobarbital. The mined by the Bradford method (Bio-Rad protein assay kit, heart was excised and mounted on a Langendorff apparatus. California). The coronary branch was reoccluded, and monastral blue The PKC-⑀ present in the cytosolic and membrane dye (4%, Sigma Chemical Co., St. Louis, Missouri) was fractions was assessed using a standard sodium dodecyl injected from the aorta at 80 mm Hg. The LV was sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) sectioned into seven slices parallel to the atrioventricular Western immunoblotting technique. The proteins (30 mg) ring. Each slice was weighed, incubated in a 1% solution of present in each sample fraction were separated by SDS- triphenyl tetrazolium chloride at 37°C to visualize the PAGE on 8.0% gels and then electrophoretically transferred infarct area (15), and photographed. The area of the to 0.45 ␮m polyvinilidene difluoride membranes using a ischemic region and the infarcted myocardium were traced Semi-Dry transfer cell (Bio-Rad). The transfer buffer con- on each LV slice and multiplied by the slice’s weight, then tained 25 mM Tris-HCl and 192 mM glycine in 20% the infarct size was expressed as a fraction of the risk region methanol. The blots were initially blocked overnight with of LV for each heart. The measurement of the infarct size 5% milk in buffer containing 20 mM Tris-HCl (pH 7.4), was performed by two persons blinded to the treatment. 137 mM NaCl, 0.05% Tween-20, and then incubated for 1350 Shimizu et al. JACC Vol. 40, No. 7, 2002 Myocardial Serotonin and Infarct Size October 2, 2002:1347–55

Table 2. Hemodynamic Parameters (Protocol 1) After Before After Baseline Occlusion Reperfusion Reperfusion Systolic blood pressure (mm Hg) Control 113 Ϯ 890Ϯ 686Ϯ 782Ϯ 9 SG (10 mg/kg/h) pre 95 Ϯ 778Ϯ 594Ϯ 978Ϯ 4 SG (10 mg/kg/h) post 86 Ϯ 487Ϯ 772Ϯ 887Ϯ 2 Diastolic blood pressure (mm Hg) Control 90 Ϯ 973Ϯ 669Ϯ 768Ϯ 9 SG (10 mg/kg/h) pre 71 Ϯ 660Ϯ 675Ϯ 12 61 Ϯ 5 SG (10 mg/kg/h) post 62 Ϯ 168Ϯ 864Ϯ 564Ϯ 4 Mean blood pressure (mm Hg) Control 99 Ϯ 979Ϯ 673Ϯ 673Ϯ 9 SG (10 mg/kg/h) pre 80 Ϯ 767Ϯ 683Ϯ 11 67 Ϯ 5 SG (10 mg/kg/h) post 70 Ϯ 276Ϯ 874Ϯ 573Ϯ 3 Heart rate (beats/min) Control 231 Ϯ 10 240 Ϯ 7 220 Ϯ 7 198 Ϯ 9 SG (10 mg/kg/h) pre 246 Ϯ 13 239 Ϯ 2 232 Ϯ 6 183 Ϯ 7 SG (10 mg/kg/h) post 231 Ϯ 15 248 Ϯ 16 229 Ϯ 22 208 Ϯ 17

1.5 h with anti-PKC-⑀ antibody (1:100 dilution). The and basal heart rate among any group in protocol 1 (Table primary antibody was monoclonal raised against peptides 2) and protocol 2 (Table 3). Throughout the experiment the corresponding to amino acids 1 to 175 as mapped to the blood pressure and heart rate were similar among all groups. carboxyterminus of human PKC-⑀. After washing, the blots Myocardial interstitial serotonin levels. As shown in were incubated for1hatroomtemperature with a Figure 1A, the myocardial interstitial serotonin level in the peroxidase-linked, goat anti-mouse secondary antibody (1: in vivo rabbit model significantly increased during 30-min 5,000 dilution). The bound antibody was then visualized ischemia and during the first 30-min reperfusion (41.6 Ϯ using an enhanced chemiluminescence kit (Amersham). 6.7 ng/ml and 47.2 Ϯ 16.6 ng/ml vs. 16.3 Ϯ 3.4 ng/ml at Protein kinase C was quantified densitometrically using control), and decreased to below the control level during the suitable autoradiographs. second 30-min reperfusion (7.8 Ϯ 1.9 ng/ml). This increase Statistical analysis. Values are expressed as group means Ϯ in interstitial serotonin during 30-min ischemia and the first SEM. To compare the group means of the infarct sizes and 30-min reperfusion was blocked by the pretreatment with ⑀ quantitation of PKC- , one-way analysis of variance the 5-HT2 blocker SG. (ANOVA) was performed, and if the ANOVA results were In the isolated heart model, as shown in Figure 1B, the significant, a Scheffe´’s post-hoc test was performed. To com- myocardial interstitial serotonin level before ischemia (0.052 pare the group means of the hemodynamic parameters, inter- Ϯ 0.018 ng/ml) was quite low compared with that of the in stitial serotonin levels, a two-way repeated measures ANOVA vivo model (16.3 Ϯ 3.4 ng/ml), but was significantly followed by a Scheffe´’s post-hoc test was performed. A p value increased during ischemia (1.155 Ϯ 0.427 ng/ml). The Ͻ of 0.05 was considered to be statistically significant. increase disappeared during early reperfusion (0.243 Ϯ 0.103 ng/ml). The serotonin increases were inhibited by RESULTS SG. Mortality and animal exclusion. A total of 110 rabbits Myocardial infarct size. The mean percentages of the area Ϯ Ϯ Ϯ were initially enrolled in the infarct size study. There was no at risk (percentage of LV) were: 34 1%, 30 3%, 31 Ϯ Ϯ Ϯ Ϯ Ϯ significant difference in the number of animals assigned to 1%, 31 2%, 34 2%, 35 2%, 37 3%, and 37 2% each of the eight groups or in the incidence and time of in the control, SG pre-5 group, SG pre-10 group, SG post ventricular fibrillation and mortality (Table 1). Among group, SG ϩ chelerythrine group, chelerythrine group, SG these animals seven were excluded because of technical ϩ 5-HD group, and the 5-HD group, respectively. No problems, six had ventricular fibrillation during coronary significant difference was seen among these groups. occlusion, and four had ventricular fibrillation during reper- As shown in Figure 2, the infarct size as a percentage of fusion; all of these rabbits died. Six rabbits died after the the area at risk was 45 Ϯ 3% in the control group. The SG first day of the experiment. Thus, the experiments were pre-5 and SG pre-10 groups showed a significant reduction completed in the remaining 88 rabbits (11 in each of the in infarct size (31 Ϯ 3% and 27 Ϯ 2%, respectively, p Ͻ 0.05 eight groups), and the findings from these animals were vs. the control group). The SG post group did not show a used for the analysis. significant reduction in infarct size (36 Ϯ 2%). Treatment Hemodynamic parameters. There were no significant dif- with chelerythrine and 5-HD blocked the infarct size- ferences in the basal systolic and diastolic blood pressures reducing effect of SG, respectively (38 Ϯ 3% and 39 Ϯ 3%, JACC Vol. 40, No. 7, 2002 Shimizu et al. 1351 October 2, 2002:1347–55 Myocardial Serotonin and Infarct Size

Table 3. Hemodynamic Parameters (Protocol 2) After Before After Baseline Occlusion Reperfusion Reperfusion Systolic blood pressure (mm Hg) Control 132 Ϯ 11 118 Ϯ 11 112 Ϯ 12 108 Ϯ 16 SG (5 mg/kg/h) pre 123 Ϯ 10 111 Ϯ 10 109 Ϯ 14 102 Ϯ 14 SG (10 mg/kg/h) pre 124 Ϯ 11 114 Ϯ 10 110 Ϯ 11 109 Ϯ 13 SG (10 mg/kg/h) post 131 Ϯ 12 116 Ϯ 13 115 Ϯ 10 111 Ϯ 14 Chelerythrine ϩ SG 129 Ϯ 10 110 Ϯ 10 106 Ϯ 12 112 Ϯ 12 Chelerythrine 125 Ϯ 11 111 Ϯ 9 109 Ϯ 10 109 Ϯ 10 5-HD ϩ SG 127 Ϯ 11 114 Ϯ 14 105 Ϯ 10 110 Ϯ 13 5-HD 128 Ϯ 13 113 Ϯ 12 107 Ϯ 15 107 Ϯ 14 Diastolic blood pressure (mm Hg) Control 84 Ϯ 12 79 Ϯ 14 77 Ϯ 10 80 Ϯ 10 SG (5 mg/kg/h) pre 88 Ϯ 13 82 Ϯ 15 87 Ϯ 12 86 Ϯ 11 SG (10 mg/kg/h) pre 92 Ϯ 11 86 Ϯ 13 86 Ϯ 12 85 Ϯ 14 SG (10 mg/kg/h) post 88 Ϯ 10 78 Ϯ 13 80 Ϯ 10 81 Ϯ 10 Chelerythrine ϩ SG 90 Ϯ 11 83 Ϯ 15 87 Ϯ 12 86 Ϯ 10 Chelerythrine 86 Ϯ 12 78 Ϯ 13 80 Ϯ 10 79 Ϯ 8 5-HD ϩ SG 92 Ϯ 16 86 Ϯ 10 80 Ϯ 10 87 Ϯ 13 5-HD 87 Ϯ 10 80 Ϯ 10 87 Ϯ 12 81 Ϯ 10 Heart rate (beats/min) Control 222 Ϯ 21 229 Ϯ 22 228 Ϯ 12 231 Ϯ 10 SG (5 mg/kg/h) pre 228 Ϯ 14 227 Ϯ 17 233 Ϯ 17 225 Ϯ 10 SG (10 mg/kg/h) pre 227 Ϯ 13 224 Ϯ 12 230 Ϯ 18 218 Ϯ 13 SG (10 mg/kg/h) post 220 Ϯ 22 228 Ϯ 12 229 Ϯ 18 233 Ϯ 14 Chelerythrine ϩ SG 234 Ϯ 14 230 Ϯ 12 234 Ϯ 12 225 Ϯ 15 Cheleruthrine 222 Ϯ 20 231 Ϯ 18 235 Ϯ 10 238 Ϯ 10 5-HD ϩ SG 230 Ϯ 18 233 Ϯ 16 226 Ϯ 15 225 Ϯ 14 5-HD 232 Ϯ 12 221 Ϯ 19 233 Ϯ 19 232 Ϯ 12

After Occlusion ϭ 10 min after occlusion; After Reperfusion ϭ 20 min after reperfusion; Baseline ϭ immediately before occlusion; Before Reperfusion ϭ immediately before reperfusion; 5-HD ϭ 5-hydroxydecanoate; SG ϭ sarpogrelate. respectively). Chelerythrine or 5-HD alone did not affect leased serotonin induces a large amount of serotonin release the infarct size (38 Ϯ 2% and 41 Ϯ 3%, respectively). from platelet cytoplasm and more platelet aggregation. The Subcellular translocation of PKC isoform. Protein kinase presence of 5-HT2 receptors has not been reported in mast C-⑀ was detected as a single band with a molecular mass of cells or sympathetic nerve endings. The interstitial serotonin 98 kDa. The quantitative densitometric analyses illustrated increases were almost completely blocked by SG, a selective in Figure 3 showed that 20 min of ischemia caused 5-HT2 receptor blocker. These findings suggest that an translocation of PKC-⑀ from the cytosolic to the membrane increase in myocardial interstitial serotonin during ischemia fraction and that SG enhanced the ischemia-induced in- is merely due to serotonin release from platelet cytoplasm ⑀ crease in PKC- in the membrane fraction. Sarpogrelate did via the vicious cycle between platelet 5-HT2A receptor and ⑀ not cause any translocation of PKC- in the nonischemic serotonin rather than the release from mast cells and heart tissue. sympathetic nerve endings. In the present study, isolated hearts were perfused with DISCUSSION Krebs-Henseleit solution, which did not include blood or Release of serotonin into the myocardial interstitial space platelets. The myocardial interstitial serotonin level at the during ischemia and the blocking effect by SG. Although baseline was quite low, compared with that of the in vivo microdialysis technique has been used for the measurements hearts. This finding could be explained by platelet washout of adenosine and noradrenaline in the myocardial intersti- due to perfusion with Krebs-Henseleit solution before tium during ischemia (16,17,18), no study is published ischemia. The myocardial interstitial serotonin level of regarding serotonin, to our knowledge. isolated hearts was clearly increased during ischemia, al- The present study revealed an increase in serotonin in though the increase was small compared with that of the in myocardial interstitial tissues during ischemia using micro- vivo hearts. One of the possible mechanisms is the release of dialysis. In in vivo cardiac tissues, serotonin is rich in platelet serotonin from mast cells and/or sympathetic nerve endings. cytoplasm of the vascular beds, although mast cells and However, small amounts of platelets with serotonin would sympathetic nerve endings also have some serotonin (Fig. remain in a part of myocardial capillary beds despite the 4). Ischemia, indicating the lack of blood flow, induces washout effect by Krebs-Henseleit solution. The findings of platelet aggregation and release of serotonin initially. As the present study showed that the increase in myocardial platelets have 5-HT2A receptors, their binding with re- interstitial serotonin was abolished by SG. These findings 1352 Shimizu et al. JACC Vol. 40, No. 7, 2002 Myocardial Serotonin and Infarct Size October 2, 2002:1347–55

Figure 1. (A) Myocardial interstitial serotonin levels before, during 30-min ischemia, and during two consecutive 30-min reperfusions in the absence and presence of sarpogrelate in an in vivo model. Myocardial interstitial serotonin levels were significantly increased during ischemia and the first 30-min reperfusion. These increases were inhibited by sarpogrelate. Vertical axis: eluate concentration (ng/ml). *p Ͻ 0.05 versus baseline control. (B) Myocardial interstitial serotonin levels before, during 30-min ischemia, and during 30-min reperfusion in an isolated heart model with and without sarpogrelate. Myocardial interstitial serotonin levels were significantly increased during ischemia without sarpogrelate. Sarpogrelate completely abolished the increase in myocardial interstitial serotonin during ischemia. Vertical axis: eluate concentration (ng/ml). *p Ͻ 0.05 versus baseline control. suggest that the serotonin release from the remaining In in vivo hearts, the increase in interstitial serotonin was platelet via the vicious cycle plays an important role in the observed during early reperfusion and disappeared thereaf- increase of myocardial interstitial serotonin in isolated ter. On the other hand, the increase was not seen even hearts during ischemia. during early reperfusion in isolated hearts. This suggests JACC Vol. 40, No. 7, 2002 Shimizu et al. 1353 October 2, 2002:1347–55 Myocardial Serotonin and Infarct Size

Figure 2. Infarct size as a percentage of the area at risk (AAR). Sarpogrelate (SG) pretreated group: SG was infused at a rate of 5 or 10 mg/kg/h starting 10 min before ischemia until 20 min after reperfusion for 60 min. Sarpogrelate post-treated group: SG was infused at a rate of 10 mg/kg/h starting 10 min before reperfusion until 50 min after reperfusion for 60 min. Sarpogrelate (5 and 10 mg/kg/h) reduced the infarct size, which was blocked by a selective protein kinase C inhibitor, chelerythrine, or a mitochondrial adenosine triphosphate sensitive potassium channel blocker, 5-hydroxydecanoate (5-HD). Chelerythrine or 5-HD alone did not affect the infarct size. *p Ͻ 0.05 versus control. that the increase during early reperfusion in in vivo hearts Possible mechanisms of the infarct size-reducing effect depends on blood reperfusion with platelets. That is, aggre- by SG. Sarpogrelate significantly reduced the myocardial gation of new reperfused platelets and serotonin release infarct size in in vivo rabbit hearts. Because SG had no effect from these may be important. on heart rate or blood pressure, the infarct-size-reducing Translocation of myocardial PKC-⑀ and myocardial in- effect of SG was not due to the decrease in the double terstitial serotonin in isolated hearts. Protein kinase C product, an indicator of oxygen demand. In addition, the consists of various subtypes and is present in various types of infarct size-reducing effect of SG was independent of the cells, especially rich in blood cells. Furthermore, rapid regional blood flow because the collateral circulation in the freezing in liquid nitrogen is needed for the precise mea- rabbit heart was minimal (20). surement. Therefore, we used isolated hearts perfused with Cardiomyocytes are protected against ischemic cellular Krebs-Henseleit solution, but not in vivo hearts, for the damage via translocation of PKC-⑀ in isolated rabbit hearts, measurement of myocardial PKC, as previously reported followed by opening of the mitochondrial KATP channel (19). The findings of this study in isolated hearts that (13,21), which was confirmed in this study. The findings of PKC-⑀ was translocated from the cytosolic fraction to the this study showed an increase in myocardial interstitial membrane fraction during ischemia confirmed the previous serotonin during ischemia and reperfusion, which was observation (19). In the present study we found that SG, a inhibited by SG, a selective 5-HT2 receptor blocker, as 5-HT2 receptor blocker, enhanced translocation of myocar- discussed previously. Sarpogrelate enhanced further trans- dial PKC-⑀ and inhibited the increase in myocardial inter- location of PKC-⑀ and reduced infarct size. The infarct- stitial serotonin during ischemia in isolated hearts. This size-reducing effect of SG disappeared after treatment with suggests that increased translocation of myocardial PKC-⑀ chelerythrine, a selective PKC inhibitor, and with 5-HD, a may be associated with inhibition of myocardial interstitial selective inhibitor of the mitochondrial KATP channel. serotonin (Fig. 4). These findings suggest that increased myocardial interstitial 1354 Shimizu et al. JACC Vol. 40, No. 7, 2002 Myocardial Serotonin and Infarct Size October 2, 2002:1347–55

Figure 3. Quantitation of protein kinase C-⑀ in the cytosolic and membrane fractions in the control, control ϩ sarpogrelate, ischemia, and sarpogrelate ϩ ischemia groups. *p Ͻ 0.05.

Figure 4. Role of serotonin in ischemia and reperfusion and the possible mechanism of the infarct size-reducing effect by sarpogrelate. Serotonin (5-hydroxytryptamine [5-HT]) is released from cardiac tissues within the area at risk, such as platelets in the vascular beds, mast cells, and sympathetic nerve endings, in ischemia and reperfusion. The increase in myocardial interstitial serotonin induces the release of a large amount of serotonin from platelet cytoplasm and platelet aggregation via 5-HT2A receptor on the cell membrane of platelets (#), and the myocardial interstitial serotonin is further increased. Note the presence of a vicious cycle among the released serotonin, 5-HT2A receptors, on the cell membrane of platelets and rich serotonin in platelet cytoplasm. Sarpogrelate, 5-HT2 receptor blocker, inhibits the binding between released serotonin and 5-HT2A receptors on the cell membrane of platelets and blocks the increase in interstitial serotonin by breaking the vicious circle. Translocation of protein kinase C (PKC)-⑀ from the cytosolic to the membrane fraction followed by opening of the mitochondrial adenosine triphosphate sensitive potassium channel is an intrinsic protective mechanism of cardiomyocytes against ischemic damage. The released serotonin during ischemia inhibits the translocation of PKC-⑀ (*). Sarpogrelate further enhances translocation of PKC-⑀ followed by opening of the mitochondrial adenosine triphosphate sensitive potassium channel and reduces infarct size via inhibition of serotonin release. JACC Vol. 40, No. 7, 2002 Shimizu et al. 1355 October 2, 2002:1347–55 Myocardial Serotonin and Infarct Size serotonin can inhibit translocation of PKC-⑀ followed by pathway in vascular smooth muscle. J Pharmacol Exp Ther 1999;291: opening of the mitochondrial K channel, and SG 1179–87. ATP 4. Cerrito F, Lazzaro MP, Gaudio E, Arminio P, Aloisi G. 5-HT2 reduces infarct size by inhibiting an increase in serotonin receptors and serotonin release: their role in human platelet aggrega- release during ischemia and reperfusion (Fig. 4). tion. Life Sci 1993;53:209–15. Other possible mechanisms of the infarct size-reducing 5. Hara H, Osakabe M, Kitajima A, Tamao Y, Kikumoto R. MCI-9042, a new antiplatelet agent, is a selective S2-serotonergic receptor effect by SG are as follows. Stimulation of 5-HT2 receptors antagonist. Thromb Haemost 1991;65:415–20. by serotonin accumulates inositol phosphates through phos- 6. Stoltz JF. Uptake and storage of serotonin by platelets. In: Vanhoutte pholipase C-mediated hydrolysis of phosphoinositides (22), PM, editor. Serotonin and the Cardiovascular System. New York, NY: Raven Press, 1985, 37–42. which can release calcium from intracellular pools, and may 7. Parikh V, Singh M. Resident cardiac mast cells and the cardioprotec- cause calcium overload and damage the myocytes (23). tive effect of ischemic preconditioning in isolated rat heart. J Cardio- Serotonin induces vasoconstriction and thrombus formation vasc Pharmacol 1997;30:149–56. 8. McCulloch MW, Rand MJ, Rose A, Story DF. Uptake and release of in intramyocardial small vessels, and may cause the so-called 5-hydroxytryptamine by sympathetic nerve terminals in rat atria no-reflow phenomenon (24). The calcium overload and (abstr). Br J Pharmacol 1984;83:P384. so-called no-reflow phenomenon are associated with early 9. Van den Berg EK, Schmitz JM, Benedict CR, Malloy CR, Willerson JT, Dehmer GJ. Transcardiac serotonin concentration is increased in reperfusion injury. In this study serotonin was increased not selected patients with limiting angina and complex coronary lesion only during ischemia, but also during early reperfusion, and morphology. Circulation 1989;79:116–24. the increases were inhibited by SG. Therefore, the inhibi- 10. Grover GJ, Sargent GA, Dzwonczyk S, Normandin DE, Antonaccio MJ. Protective effect of serotonin (5-HT2) receptor antagonists in tion of calcium overload and the so-called no-reflow phe- ischemic rat hearts. J Cardiovasc Pharmacol 1993;22:664–72. nomenon may be included in the protective mechanism of 11. Grover GJ, Parham CS, Youssef S, Ogletree ML. Protective effect of SG. the serotonin receptor antagonist cinanserin in two canine models of pacing-induced myocardial ischemia. Pharmacology 1995;50:286–97. Clinical implications. Sarpogrelate is currently used for 12. Tanaka T, Fujita M, Nakae I, et al. Improvement of exercise capacity the treatment of patients with arteriosclerosis obliterance by sarpogrelate as a result of augmented collateral circulation in because of its vasodilating and antiplatelet actions. It has patients with effort angina. J Am Coll Cardiol 1998;32:1982–6. 13. Wang Y, Hirai K, Ashraf M. Activation of mitochondrial ATP- been reported that SG is beneficial for human angina sensitive K(ϩ) channel for cardiac protection against ischemic injury is pectoris through an increase in the collateral circulation dependent on protein kinase C activity. Circ Res 1999;85:731–41. (12). In this study we found a cardioprotective effect against 14. Minatoguchi S, Arai M, Uno Y, et al. A novel anti-diabetic drug, miglitol, markedly reduces myocardial infarct size in rabbits. Br J infarction, which was independent of collateral flow or Pharmacol 1999;128:1667–72. double products. The doses of SG used in this study were 5 15. Fishbein MC, Meerbaum S, Rit J, et al. Early phase acute myocardial or 10 mg/kg/h, and plasma concentrations of SG 10 and 60 infarct size quantification: validation of the triphenyl tetrazolium chloride tissue enzyme staining technique. Am Heart J 1981;101:593– min after infusion of 10 mg/kg/h of SG was 588 and 688 600. ng/ml, respectively. The clinical dose of SG is 100 mg ϫ 3 16. Van Wylen DGL. Effect of ischemic preconditioning on interstitial times/day, and the maximum plasma concentration of SG purine metabolite and lactate accumulation during myocardial ischemia. Circulation 1994;89:2283–9. after oral administration of 100 mg of SG was reported to 17. Akiyama T, Yamazaki T, Ninomiya I. Differential regional responses be 540 ng/ml in healthy volunteers (25), which is very of myocardial interstitial noradrenaline levels to coronary occlusion. similar to that of the present study. In the present study SG Cardiovasc Res 1993;27:817–22. reduced myocardial infarct size only when administered 18. Minatoguchi S, Uno Y, Arai M, et al. Modulation of interstitial noradrenaline levels during ischemic preconditioning through brady- before ischemia. Therefore, SG may be useful for reducing kinin B2 receptors and angiotensin II receptors in rabbits. J Cardiovasc infarct size when acute myocardial infarction occurs in the Pharmacol 1999;34 Suppl 4:S43–6. patients with angina pectoris and atherosclerosis obliterance 19. Arai M, Minatoguchi S, Kumada H, et al. Role of protein kinase C in the reduction of infarct size by N-methyl-1-deoxynojirimycin, an that have previously been being treated with SG. Further ␣-1,6-Glucosidasr inhibitor. Br J Pharmacol 2001;133:635–42. clinical investigations are warranted. 20. Harken AH, Simson MB, Haselgrove J, Westein L, Harden WR III, Barlow CH. Early ischemia after complete coronary ligation in the rabbit, dog, pig and monkey. Am J Physiol 1981;241:H202–10. Reprint requests and correspondence: Dr. Hisayoshi Fujiwara, 21. Liu GS, Cohen MV, Mochly-Rosen D, Downey JM. Protein kinase Second Department of Internal Medicine, Gifu University School of C-epsilon is responsible for the protection of preconditioning in rabbit Medicine, 40 Tsukasa Machi, Gifu 500, Japan. E-mail: gifuim-gif@ cardiomyocytes. J Mol Cell Cardiol 1999;31:1937–48. umin.ac.jp. 22. Hamamori Y, Yokoyama M, Yamada M, Akita H, Goshima K, Fukuzaki H. 5-Hydroxytryptamine induces phospholipase C-mediated hydrolysis of phosphoinositides through 5-hydroxytryptamine-2 receptors REFERENCES in cultured fetal mouse ventricular myocytes. Circ Res 1990;66:1474–83. 23. Nosek TM, Williams MF, Zeigler ST, Godt RE. Inositol triphos- 1. Vanhoutte PM. Serotonin, hypertension and vascular disease. Neth phate enhances calcium release in skinned cardiac and skeletal muscle. J Med 1991;38:35–42. Am J Physiol 1986;250:C807–11. 2. Spigset O, Mjorndal T. Serotonin 5-HT2A receptor binding in 24. Metais C, Li J, Simons M, Sellke FW. Serotonin-induced coronary platelets from healthy subjects as studied by [3H]-lysergic acid contraction increases after blood cardioplegia-reperfusion: role of diethylamide ([3H]-LSD): intra- and interindividual variability. Neu- COX-2 expression. Circulation 1999;100 SupplII:328–34. ropsychopharmacology 1997;16:285–93. 25. Isogai Y, Yokose T, Ikemoto T, Maeda T, Akiyaka M. Inhibitory 3. Banes A, Florian JA, Watts SW. Mechanisms of 5-hydroxytryptamine effect of sarpogrelate hydrochloride (MCI-9042) on platelet aggrega- (2A) receptor activation of the mitogen-activated protein kinase tion in healthy volunteers. J Clin Ther Med 1991;7:1227–33.