Author + information
- Received March 9, 1998
- Revision received June 26, 1998
- Accepted July 24, 1998
- Published online November 15, 1998.
- Konstantin Alexiou, MDa,
- Thomas Dschietzig, MDa,
- Oliver Simscha,
- Michael Laule, MDa,
- Johannes Hundertmarka,
- Gert Baumann, MDa and
- Karl Stangl, MDa,* ()
- ↵*Address for correspondence: Dr. Karl Stangl, Medizinische Klinik und Poliklinik I, Charité, Humboldt-Universität zu Berlin, Schumannstrasse 20/21, 10098 Berlin, Germany
Objectives. We investigated whether endogenous pulmonary big endothelin has arrhythmogenic properties under normal conditions and in heritable hyperlipidemia.
Background. Endothelin (ET), one of the most potent vasoconstrictors, is known to induce ventricular arrhythmias. It is unclear, however, whether its precursor, big endothelin, released from the lung, contributes to arrhythmogenesis.
Methods. In a lung-heart model in which a Langendorff heart is serially perfused with the effluent from the isolated lung of the same animal, we evaluated arrhythmias in control and in Watanabe heritable hyperlipidemic (WHHL) rabbits.
Results. In both controls (n = 12) and WHHL (n = 8), serial perfusion evoked a decrease in coronary flow (controls, −11 ± 3%; WHHL, −25 ± 6%) and a fourfold increase of ventricular extrasystoles (VES) (controls, 40.7 ± 8; WHHL, 40.2 ± 5 VES/40 min, p < 0.05). However, WHHL developed more and longer nonsustained ventricular tachycardias (VT) compared with controls (incidence, 1.38 ± 1.1 vs. 0.33 ± 0.5 VT/40 min, p < 0.05; length, 14.36 ± 3.1 vs. 7.25 ± 1.5 beats/VT, p < 0.05). Arrhythmias were not ischemia-induced because corresponding mechanical flow reduction had no arrhythmogenic effect (n = 6 in controls and WHHL). Although vasoconstriction disappeared entirely, arrhythmias were only partly suppressed by ETAantagonists (BQ-123, 2 μmol/liter; A-127722, 20 μmol/liter). The ET-converting enzyme inhibitor phosphoramidon (50 μmol/liter) completely suppressed arrhythmias and vasoconstriction. The ETBantagonists (IRL-1038, 4 μmol/liter; IRL-1025, 5 μmol/liter) had no effect (n = 6).
Conclusions. Endogenous pulmonary big ET produces arrhythmogenic effects that are aggravated in heritable hyperlipidemia. These effects, requiring coronary conversion of big ET into ET, are partly ETA-mediated and ETB-independent.
Endothelin (ET), one of the most potent vasoconstricting and cardiotonic peptides (1), has attracted clinical interest owing to the fact that elevated plasma concentrations have been found in many pathophysiological situations, such as congestive heart failure (2,3)and pulmonary hypertension (4). Elevations in ET and its precursor, big endothelin, have also been described in acute myocardial infarction (5). The potential relevance of this finding has been examined in several animal models. These studies have demonstrated that exogenous ET causes coronary vasoconstriction and, at higher doses, ventricular arrhythmias, including ventricular fibrillation (6). These arrhythmogenic properties have been attributed to myocardial ischemia following intense ET-mediated coronary vasoconstriction. However, recent electrophysiological experiments conducted on cardiac tissues suggest that ET may also possess direct arrhythmogenic properties (7,8). Despite the demonstrated arrhythmogenic effects of the mature peptide, little is known about how its precursor, big ET, contributes to these properties.
By means of a new lung-heart model, we have recently shown (9)that a basal luminal pulmonary release of big ET occurs in isolated perfused rabbit lungs, unaccompanied by relevant amounts of the mature peptide, ET. Pulmonary big ET, locally converted into ET during coronary passage, causes an ETAreceptor-mediated decrease in coronary flow by 11% in controls. Among Watanabe heritable hyperlipidemic (WHHL) rabbits, in contrast, we determined that an identical release of pulmonary big ET caused a significantly stronger decrease in coronary flow by 25% of baseline levels.
In this study we investigated whether this endogenous pulmonary big ET has arrhythmogenic properties, and, if so, whether these effects are enhanced in heritable hyperlipidemia.
Male New Zealand White (NZW) rabbits and male WHHL rabbits (14 weeks old, Charles River), weighing 2.0 to 2.3 kg, were used for these experiments, which were performed in accordance with protocols approved by the institutional Animal Care and Use Committee. The study conforms to the guidelines of the American Heart Association on Research Animal Use adopted on November 11, 1984.
Watanabe rabbits are frequently used as an atherosclerosis model: owing to a heritable low density lipoprotein receptor defect, these animals develop hyperlipidemia from birth on (10), with the ensuing atherosclerotic process closely resembling that observed in familial hypercholesterolemia in humans (11).
For excision of heart and lung from the same animal, the rabbits were anesthetized with thiopental sodium (intraperitoneally 40 to 80 mg/kg body weight [BW]) after anticoagulation with heparin (1500 U/kg BW). Hearts and lungs were removed for isolated extracorporeal perfusion as previously described (9).
A tracheotomy permitted ventilation with a Siemens Servo 910 respirator (Siemens-Elema) at 50 breaths/min, with a tidal volume of 10 to 12 ml/kg BW, and 1 mm Hg positive end-expiratory pressure. The inspired gas mixture consisted of 95% O2and 5% CO2.
A median sternotomy was performed, and cannulas were placed into the pulmonary artery and the aorta. The heart was rapidly and carefully excised and perfused retrogradely through the aorta (Langendorff technique) with modified Krebs-Henseleit buffer (37.5°C). The pH was adjusted to 7.35 to 7.40, and the buffer oxygenated with a mixture of 95% and 5% CO2. Left and right ventricular pressures (LVP, RVP) were measured by means of a polyethylene catheter with a small latex balloon at its top. This device was inserted into the ventricular cavity and filled with saline solution until end-diastolic LVP and RVP stabilized between 1 to 3 mm Hg (0.13 to 0.40 kPa). Coronary flow rates were determined by an ultrasonic flowprobe (Transonic Systems). We accepted only those hearts that achieved a contractile performance of more than 70 mm Hg (9.46 kPa) LVPmax, more than 1000 mm Hg/s (135.35 kPa/s) rate of pressure development dp/dtmaxand more than 1000 mm Hg/s (135.35 kPa/s) rate of pressure decline, dp/dtmin. These criteria correspond to the 10% percentiles obtained by analyzing data from preceding experiments.
Meanwhile, the pulmonary vasculature was flushed in a nonrecirculatory mode with Krebs-Henseleit buffer (same composition, pH = 7.35 to 7.40) to remove residual blood elements. Lungs were suspended in a temperature-equilibrated humidified chamber (37.5°C) from a force transducer that monitored changes in lung weight. Pulmonary flow was gradually increased to 100 ml/min of nonrecirculatory perfusion. Only those lungs were selected for the present study that showed a constant pulmonary arterial pressure (6 to 10 mm Hg [0.81 to 1.35 kPa], zero-referenced at hilus), constant peak inflation pressures of 7 to 10 mm Hg [0.95 to 1.35 kPa], no weight gain (<0.5 g/h), and no signs of hemorrhage, edema or atelectasis.
The apparatus used in our investigation was manufactured by Hugo Sachs Electronics. Modification of the device allows rapid switching from separate to serial organ perfusion. Once switching has taken place, the amount of lung effluent that equals coronary flow is rapidly conveyed (2 s) to the heart apparatus. There it is oxygenated prior to becoming perfused, at constant pressure, through the coronary vessels.
Hearts and lungs were permitted to stabilize over 40 min, during which time both endothelium-dependent and -independent relaxation of the coronary arteries were assessed with the use of acetylcholine (2 μmol/liter), substance P (200 pmol/liter), sodium nitroprusside (4 μmol/liter) and nitroglycerin (4 μmol/liter). The lungs were rinsed and controlled for pressure and weight constancy. Following this initial period of separate organ perfusion, the systems were switched to serial perfusion for an additional 40 min. To determine ET and big ET, samples of pulmonary and coronary effluent were taken simultaneously at the end of separate perfusion and at 50, 500 and 1500 s during serial perfusion. We have recently reported this subset of results (9).
For all groups, registration of epicardial electrocardiogram (ECG) took place throughout the control and the serial periods, by using two stainless steel electrodes attached directly to the right atrium and the base of the left ventricle. The signal was digitalized and sampled with a rate of 500 Hz.
Arrhythmias were classified according to the following characteristics: Isolated ventricular extrasystoles (VES) were defined as discrete and identifiable premature QRS complexes, preceded by a shortened beat-to-beat interval. Ventricular tachycardia was defined as a run of four or more consecutive morphologically similar ventricular complexes. Ventricular fibrillation was defined as a signal for which individual QRS deflections could not be distinguished from one another (implying morphological instability) and for which the rate could no longer be measured (12). Both the incidence and the types of arrhythmias were determined and reported for the entire interval of 40 min.
Mechanical flow reduction
Six isolated hearts of both NZW and WHHL rabbits underwent mechanical flow reduction (i.e., volume-controlled perfusion) to determine whether it was possible to attribute arrhythmogenic events to alterations in coronary flow. Mechanical flow reduction mimicked the average coronary flow response during serial perfusion.
Administration of exogenous big ET and ET
In an attempt to reproduce the results obtained in the serial perfusion experiments, and to compare them with the effects of exogenous vasoconstrictors, isolated hearts of both NZW (n = 6) and Watanabe rabbits (n = 6) received a 40-min coronary infusion of human ET-1 (5 pmol/liter) or human big ET-1 (5 pmol/liter).
The following reagents were applied prior to switching from separate to serial perfusion, with application of each reagent to different subgroups (n = 4) of both the control and the WHHL rabbits: DesArg9-[Leu]8-bradykinin (5 μmol/liter), a bradykinin-1 receptor antagonist (13); Hoe-140 (1 μmol/liter), a bradykinin-2 receptor antagonist (13)and captopril (5 μmol/liter), an angiotensin-converting enzyme inhibitor (14). Infusion took place into coronary circulation. Meclofenamic acid (10 μmol/liter), a cyclooxygenase inhibitor (15), was applied into pulmonary artery.
Furthermore, we tested the following substances for their capability of influencing the effects observed during serial perfusion (n = 6 per subgroup): the nonselective endothelin antagonist PD-145065 (10 μmol/liter) (16), two different ETAreceptor antagonists, BQ-123 (2 μmol/liter) (17)and A-127722 (20 μmol/liter) (18), the ETBantagonists IRL-1038 (4 μmol/liter) (19)and IRL-1025 (5 μmol/liter) (19), as well as the ECE (endothelin-converting enzyme) inhibitor phosphoramidon (50 μmol/liter) (20).
The following were purchased from Sigma Chemical: acetylcholine, nitroglycerin, substance P (Arg-Pro-Lys-Pro-Gln-Gln-Phe-Phe-Gly-Leu-Met-NH2), sodium nitroprusside, meclofenamic acid (2-[(2,6-dichloro-3-methylphenyl)-amino]benzoic acid), captopril ([2S]-1-[3-mercapto-2-methyl-propionyl]-l-proline), desArg9-[Leu8]-bradykinin, BQ-123 (cyclo[d-Asp-Pro-d-Val-Leu-d-Trp]), human ET-1, human big ET-1 and phosphoramidon (N-(α-rhamnopyranosyloxyhydroxyphosphinyl)-Leu-Trp sodium salt). Parke Davis and Abbott generously provided PD-145065 (Ac-(D-2-(10,11-dihydro-5H-dibenzo[a,d]cyclohepten-5-yl))Gly-l-Leu-l-Asp-l-Ile-l-Trp disodium salt, lot B) and A-127722 ([2R,3R,4S]-2-(4-methoxyphenyl)-4-(1,3-benzodioxol-5-yl)-1-[[N,N-dibutyl-amino)-carboxyl]- methyl]pyrrolidine-3-carboxylic acid), respectively. Both IRL-1038 ([Cys11-Cys15]-endothelin-1(11-21)) and IRL-1025 ([Cys11-Leu15-endothelin-1(11-21)) were gifts from Ciba International Research Laboratories. Hoechst supplied Hoe-140 (d-Arg[hydroxyproline3,beta-thienylalanine5,d-Tic7,octahydroindol-2-yl-carbonylresidue8]-bradykinin).
Data are presented as mean values ± SD, unless otherwise indicated. Differences between groups as function of time were analyzed using a nonparametric ANOVA for repeated measures (21). After global testing, a multiple comparison procedure using Bonferroni-Holm adjustment of p was carried out (22). An error probability of p < 0.05 was regarded as significant. Lower levels of p are indicated.
Means of pulmonary arterial pressure (7.0 to 7.4 mm Hg), coronary flow (30.6 to 32.9 ml/min), contractile performance (LVPmax, 83 to 90 mm Hg) and heart rate at baseline (179 to 189 beats/min) were nearly identical in all groups observed.
In both NZW and WHHL groups, serial perfusion caused a coronary constriction, which was accompanied by decreases in mechanical performance. These mechanical alterations corresponded to the degree of flow reduction (9).
During the separate perfusion period, we found in all groups only ventricular extrasystoles (VES); the incidence of VES did not differ between the groups (Fig. 1).
Switching from separate to serial perfusion resulted in a significant increase in the number of VES in both the NZW and the WHHL groups (Fig. 1). But in WHHL, the incidence of monomorphic, nonsustained ventricular tachycardia (VT) was significantly increased in comparison to the NZW group (1.38 ± 1.06 vs. 0.33 ± 0.49 VT/40 min, p < 0.05). The VT could be observed in 6 of 8 hearts in the WHHL group, with a rate of 250 ± 12 beats/min, and in only 4 of 12 hearts in the NZW group (240 ± 10 beats/min). In addition, VT in WHHL was significantly longer than in NZW (14.36 ± 3.07 vs. 7.25 ± 1.50 beats/VT, p < 0.05).
The occurrence of both VES and VT throughout the entire period of serial perfusion was not time-dependent.
Administration of exogenous big ET and ET
In NZW hearts, coronary infusion of ET-1 and big ET-1, at a concentration of 5 pmol/liter, produced a drop in coronary flow by 14 ± 2% (baseline, 32.2 ± 2.5; ET, 27.7 ± 2.1 ml/min) and 9 ± 4% (baseline, 31.5 ± 2.9; big ET, 28.7 ± 1.9 ml/min), respectively. The corresponding flow decreases in WHHL hearts, measured at −29 ± 5% for ET-1 (baseline, 30.9 ± 3.0; ET, 21.9 ± 1.5 ml/min) and −23 ± 4% for big ET-1 (baseline, 31.0 ± 2.2; big ET, 23.9 ± 1.8 ml/min), were significantly steeper than those observed in NZW hearts. The effects of exogenous ET and big ET on ventricular arrhythmias were comparable to those observed in the NZW and WHHL serial perfusion groups (Table 1).
ET antagonists and inhibition of ECE produced identical results in controls and WHHL (Fig. 2): Flow reduction, the increase in VES and the occurence of VT were completely eliminated by coronary infusion of phosphoramidon (50 μmol/liter), an ECE inhibitor. Infusion of the nonselective ET antagonist PD-145065 (10 μmol/liter) as well as the ETAantagonists BQ-123 (2 μmol/liter) and A-127722 (20 μmol/liter) led to significant but not complete suppression of VES, whereas VT and the coronary flow reduction were completely eliminated by these agents. The ETBantagonists IRL-1038 (4 μmol/liter) and IRL-1025 (5 μmol/liter) had no effect on coronary flow or ventricular arrhythmias.
In both NZW and WHHL groups, coronary infusion of the following agents did not influence ventricular arrhythmias during serial perfusion (data not shown): captopril (5 μmol/liter), an angiotensin-converting enzyme inhibitor, desArg9-[Leu]8-bradykinin (5 μmol/liter), a bradykinin-1 receptor antagonist and Hoe-140 (1 μmol/liter), a bradykinin-2 receptor antagonist. Similarly, pulmonary infusion of the cyclooxygenase inhibitor meclofenamic acid (10 μmol/liter) did not influence the decrease in coronary flow and the incidence of VES in either group.
Big endothelin-induced arrhythmia: direct versus ischemia-/flow-dependent effects
In this project, we investigated whether pulmonary big ET is directly arrhythmogenic and, if so, whether this effect is dependent on coronary conversion to ET. Until now, conflicting results have appeared concerning the arrhythmogenic effects of ET and the underlying mechanisms. Previous studies conducted in pigs (6), rats (23)and dogs (24,25)suggested ischemia-dependent proarrhythmic effects of intracoronary ET administration. In contrast, the view that ET may function as a direct mediator of arrhythmogenesis is supported by findings revealing a higher proarrhythmic potency of ET compared with mere coronary artery occlusion (26), a sustentation of ET-induced arrhythmias despite restoration of coronary flow (27), and a dependency on the site of ET administration of ventricular arrhythmias (27). In addition, Yorikane et al. (7,8)reported a prolongation of the action potential duration by ET (200 nmol/liter), followed by the development of early afterdepolarizations.
Assessment of the direct electrophysiologic effects of ET, however, is hampered by the fact that the studies cited used very high doses of ET. Endothelin can be detected in the plasma of normal humans in concentrations ranging from 0.5 to 5 pmol/liter, and the lowest ET concentrations found to increase coronary tone, in rabbits, ranged between 1 and 10 pmol/liter (28). Consequently, it is unclear whether endogenous ET/big ET plays a role in the physiological modulation of myocardial rhythmogenesis.
ECE-dependent proarrhythmic effect of endogenous pulmonary big ET
Our findings demonstrate that the basal pulmonary release of big ET, determined at 5.5 to 5.7 pmol/liter in both groups (9), may induce ventricular arrhythmias and coronary vasoconstriction. Ventricular arrhythmias were not accompanied by ECG signs of ischemia, although coronary flow decreased by 11 ± 3% in control rabbits and by 25 ± 6% in the Watanabe group. In addition, mechanical reduction of coronary flow comparable to the big ET-induced vasoconstriction did not affect the incidence or severity of ventricular arrhythmias. The arrhythmogenic properties of pulmonary big ET can, therefore, hardly be explained by flow-dependent mechanisms. The data obtained in the present study point to a direct arrhythmogenic property of big ET or ET in our model. To determine whether this arrhythmogenic effect is due to the conversion from big ET to ET, or whether big ET itself possesses arrhythmogenic properties, we intracoronarily infused phosphoramidon, an ECE inhibitor. Phosphoramidon led to complete elimination of arrhythmias, which confirms that the proarrhythmic effect of big ET was exclusively caused by local coronary conversion to ET.
Studies conducted by Brunner (14,29)have revealed proarrhythmic effects of endogenous cardiac ET during reperfusion after low-flow ischemia in isolated rat hearts. We, in contrast, observed only a mild decrease in coronary flow evoked by endogenous pulmonary big ET. Therefore, though we cannot entirely exclude flow redistribution by pulmonary big ET leading to relative hypoperfusion of certain areas, we did not study myocardial ischemia in our experiments. This fact may explain why we were not able to suppress arrhythmias by coronary application of captopril, which contradicts the findings of Brunner (14), who demonstrated an angiotensin-endothelin proarrhythmic axis functioning during ischemia/reperfusion in rat hearts. On the other hand, it is even uncertain whether or not there are distinct pathways of ET-mediated arrhythmogenesis during ischemia/reperfusion and in our experimental setting. This item remains to be examined in further experimental studies.
Role of endothelin receptor subtypes in arrhythmogenesis
The question arises as to which receptor mediated these arrhythmogenic effects. In rats (29,30)and dogs (31), ET-induced ventricular arrhythmias proved sensitive to ETAblockade, whereas a contribution of ETBreceptors could be excluded. In the present study, the nonselective ET antagonist PD-145065 as well as the ETAantagonists BQ-123 and A-127722 significantly but not entirely hindered ventricular arrhythmias, although these agents completely prevented coronary constriction. The ETBantagonists IRL-1038 and IRL-1025 had no effect. Hence, in our model, the proarrhythmic action of ET is ETB-independent and partly ETA-mediated. Another still unknown receptor or mechanism seems likely, as postulated by Kasai et al. (32), for the ET-mediated positive inotropic effects in rabbit hearts.
Aggravated arrhythmia in heritable hyperlipidemia: possible role of endothelial dysfunction
As a key finding, the Watanabe group disclosed a significantly enhanced severity of arrhythmias, as indicated by a higher incidence of nonsustained VT. Despite the 2.5-fold stronger coronary vasoconstrictor effect of pulmonary big ET in this group, there was no evidence of ischemia-induced ventricular arrhythmias—an identical mechanical reduction of coronary flow produced no increase in arrhythmias.
We have recently shown that the Watanabe rabbits employed in our investigation suffered from coronary endothelial dysfunction (9). Impairment of endothelial function exists even before histological findings of atherosclerosis become apparent (33,34). The endothelial dysfunction affects both the nitric oxide (35)and the prostacyclin (36)pathway. Studies in dogs (37)and rats (38)have implicated nitric oxide as an endogenous mediator of preconditioning against severe ventricular arrhythmias. Similarly, it has been suggested that prostacyclin represents an antiarrhythmic substance (39,40). Hence, deterioration of ET function probably leads to higher susceptibility to arrhythmia, which could account for the difference observed between controls and WHHL.
With respect to the release of big ET observed, the question might arise whether the perfusate characteristics—that is, the absence of blood cells and plasma proteins—that is, the absence of blood cells and plasma proteins—may have influenced this key finding. Brunner and co-workers (41), however, have demonstrated that binding of ET to albumin does not affect the vasoactivity of ET. Furthermore, circulating blood has been excluded as a major site of big ET conversion (42)and clearance (43). As a result, the release of big ET described in this study is in all probability largely independent of the artificial conditions applied.
Concerning the enhanced severity of arrhythmia found in WHHL, this study does not provide direct proof that endothelial dysfunction represents the underlying mechanism. Although this explanation seems reasonable in the light of numerous studies revealing the antiarrhythmic effects of nitric oxide and prostacyclin, we cannot exclude additional arrhythmogenic properties inherent in WHHL isolated hearts.
Potential clinical implications
With respect to the potential clinical implications of our findings, attention should be directed to the severe ventricular arrhythmias that are characteristic of advanced chronic heart failure. Plasma concentrations of big ET (44), as well as pulmonary synthesis of ET (45), are known to increase with the severity of the disease. This increment in ET production is accompanied by an elevated expression of ET receptors (46). Furthermore, an important percentage of chronic heart failure develops on the basis of ischemic heart disease, in which endothelial dysfunction is common. In light of these clinical facts and our experimental findings showing enhanced severity of big ET-induced arrhythmias in heritable hyperlipidemia, we hypothesize that ET, in general, and pulmonary big ET, in particular, may act as an important mediator of arrhythmogenesis in chronic heart failure.
☆ This study was supported in part by departmental funds (95-020, pulmonary circulation).
- endothelin-converting enzyme
- New Zealand White rabbits
- ventricular extrasystole
- ventricular tachycardia
- Watanabe heritable hyperlipidemic rabbits
- Received March 9, 1998.
- Revision received June 26, 1998.
- Accepted July 24, 1998.
- American College of Cardiology
- Cody R.J,
- Haas G.J,
- Binkley P.F,
- Capers Q,
- Kelley R
- Wei C.M,
- Lerman A,
- Rodeheffer R.J,
- et al.
- Ezra D,
- Goldstein R.E,
- Czaja J.F,
- Feuerstein G.Z
- Yorikane R,
- Koike H,
- Miyake S
- Stangl K,
- Dschietzig T,
- Laule M,
- Alexiou K,
- Wernecke K.D,
- Baumann G
- Kondo T,
- Watanabe Y
- Walker M.J.A,
- Curtis M.J,
- Hearse D.J,
- et al.
- Brunner F,
- Kukovetz W.R
- Wu-Wong J.R,
- Dixon D.B,
- Chiou W.J,
- Opgenorth T.J
- Hisaki K,
- Matsumura Y,
- Maekawa H,
- Fujita K,
- Takaoka M,
- Morimoto S
- Bortz J,
- Lienert G.A,
- Boehnke K
- Holm S
- Raschack M,
- Unger L,
- Riechers H,
- Klinge D
- Clozel J.P,
- Clozel M
- Kusumoto K,
- Awane Y,
- Fujiwara S,
- Watanabe T
- Brunner F,
- Opie L.H
- Teerlink J.R,
- Breu V,
- Sprecher U,
- Clozel M,
- Clozel J.P
- Kasai H,
- Takanashi M,
- Takasaki C,
- Endoh M
- Reddy K.G,
- Nair R.N,
- Sheehan H.M,
- Hodgson J.M
- Zeiher A.M,
- Drexler H,
- Wollschläger H,
- Just H
- Pabla R,
- Curtis M.J
- Parratt J.R
- Mest H.J,
- Schneider S,
- Riedel A
- Watanabe Y,
- Naruse M,
- Monzen C,
- et al.
- Tsutamoto T,
- Wada A,
- Maeda Y,
- Adachi T,
- Kinoshita M
- Miyauchi T,
- Sakai S,
- Ihara M,
- et al.