Comparative Electropharmacological Actions of Some Constituents from Ginkgo biloba Extract in Guinea-pig Ventricular Cardiomyocytes

Effects of the constituents from Ginkgo biloba extract (GBE) on the action potentials and the ionic currents in guinea pig ventricular cardiomyocytes were investigated using whole-cell and current-clamp techniques. The constituents, ginkgolides A, B, C and quercetin, had depressant effects at 0.1–3μM on the action potential configuration. Ginkgolide A (1–3 μM) prolonged the action potential (action potential duration: APD) at 75% and 90% repolarizations (APD75 and APD90). However, ginkgolides B and C at low concentrations prolonged APD, but at higher concentrations (>1 μM) shortened APD. Quercetin at 3 μM prolonged the APD, but not at the lower concentrations. These constituents also inhibited the Vmax. The resting potential was unaffected. In voltage-clamp experiments, ginkgolides A and B (0.1–3 μM) markedly and concentration-dependently increased the Ca2+ current (ICa) and the delayed rectifier K+ current (IK), and decreased the inwardly rectifying K+ current (IK1). On the other hand, ginkgolide C failed to affect the ICa but increased the IK by 14.0 ± 2.3% (n = 6, P < 0.05) at 1 μM. Quercetin inhibited ICa, and enhanced IK but decreased IK1. These responses to the constituents were almost reversible (80–90% of control) after a 10- to 20-min washout. These results indicate that even at acute administrations, these constituents produce the effective actions on the APD and the underlying ionic currents in cardiomyocytes. Each constituent does not exhibit a uniform response, although GBE acts as a net.

Effects of the constituents from Ginkgo biloba extract (GBE) on the action potentials and the ionic currents in guinea pig ventricular cardiomyocytes were investigated using whole-cell and current-clamp techniques. The constituents, ginkgolides A, B, C and quercetin, had depressant effects at 0.1-3 M on the action potential configuration. Ginkgolide A (1-3 M) prolonged the action potential (action potential duration: APD) at 75% and 90% repolarizations (APD 75 and APD 90 ). However, ginkgolides B and C at low concentrations prolonged APD, but at higher concentrations (Ͼ1 M) shortened APD. Quercetin at 3 M prolonged the APD, but not at the lower concentrations. These constituents also inhibited the V max . The resting potential was unaffected. In voltage-clamp experiments, ginkgolides A and B (0.1-3 M) markedly and concentration-dependently increased the Ca 2ϩ current (I Ca ) and the delayed rectifier K ϩ current (I K ), and decreased the inwardly rectifying K ϩ current (I K1 ). On the other hand, ginkgolide C failed to affect the I Ca but increased the I K by 14.0 Ϯ 2.3% (n ϭ 6, P Ͻ0.05) at 1 M. Quercetin inhibited I Ca , and enhanced I K but decreased I K1 . These responses to the constituents were almost reversible (80-90% of control) after a 10-to 20-min washout. These results indicate that even at acute administrations, these constituents produce the effective actions on the APD and the underlying ionic currents in cardiomyocytes. Each constituent does not exhibit a uniform response, although GBE acts as a net.
Keywords: action potential -ionic currents -ginkgolide A -ginkgolide B -ginkgolide C -Ginkgo biloba extract -guinea pig ventricular cardiomyocytes -quercetin For reprints and all correspondence: Hiroyasu Satoh, Department of Pharmacology, Division of Molecular and Cellular Biology, Nara Medical University, School of Medicine, Kashihara, Nara 634-8521, Japan. Tel: ϩ81-744-29-8831; Fax: ϩ81-744-25-7657; E-mail: hysat@naramed-u.ac.jp The online version of this article has been published under an open access model. Users are entitled to use, reproduce, disseminate, or display the open access version of this article provided that: the original authorship is properly and fully attributed; the Journal and Oxford University Press are attributed as the original place of publication with the correct citation details given; if an article is subsequently reproduced or disseminated not in its entirety but only in part or as a derivative work this must be clearly indicated.
number of its constituent substances (6,8). Unknown mechanisms of these constituents for the ionic channel currents still remain. The aim of the present experiments was to examine, by use of a patch-clamp technique, how these constituents of GBE, other than bilobalide, affect the action potentials and the underlying ionic currents. In addition, the cardiac electropharmacological actions of each constituent were compared.

Materials and Methods
All experiments were carried out according to the guidelines laid down by the Nara Medical University Animal Welfare committee, and also under the terms of the Declaration of Helsinki.

Cell Preparation
Cells were prepared from tissue taken from the ventricle muscle of guinea pig hearts, using methods similar to those described previously (14,16). Under sodium pentobarbital (30 mg/kg, i.p.) anesthesia, the chest was opened and the aorta was cannulated in situ. The heart was dissected out and perfused with normal Tyrode solution on the Langendorff apparatus. After a washout of blood, the heart was perfused with Ca 2ϩ -free Tyrode solution, and spontaneous beating ceased. Then, the perfusate was switched to low-Ca 2ϩ (30-60 M) Tyrode solution containing 0.4 mg/ml collagenase (Type I, Sigma Chemical, St Louis, MO) for about 20 min. The heart was washed out by high-K ϩ and low-Cl Ϫ solution (KB solution), and was dissected with scissors. The temperature of all solutions was maintained at 36ЊC.

Current-and Voltage-clamp Experiments
Current-clamp and whole-cell voltage-clamp recordings were performed using an Axopatch patch-clamp amplifier (Axon Instruments, Burlingame, CA) and standard techniques. Patch pipettes from borosilicate glass capillaries were fabricated using a two-stage puller, and had a resistance of 5-7 M⍀. The series resistance error was Ͻ3-7 mV, and no compensation was used. The liquid junction potential between the pipette solution and the external solution (Ͻ10 mV) was corrected for all membrane potential recordings. Experiments were carried out at a temperature of 36.5ЊC. The data were stored and analyzed on an IBM-AT microcomputer, using the PCLAMP analysis program (Axon Instruments). Current traces were filtered using a cut-off frequency of 1 kHz for plotting. The I Ca was measured as the difference between the peak current and the zero current, and the I K1 was the difference between the current at the end of a 1-s test pulse and the zero current. The I K was measured at the peak of the outward tail current. All values are given as mean Ϯ S.E.M. The differences of the mean values were analyzed by Student's t-test and ANOVA for paired data, and a P-value of Ͻ0.05 was considered significant.

Gingko biloba Effects on the Action Potentials
To examine the effects of each constituent on the action potential configuration, a current-clamp experiment was carried out. The isolated single cell was stimulated at 1 Hz. As shown in Fig. 1A, ginkgolide A (at Ͼ1 M) markedly prolonged the APD; at 3 M, 75% and 90% repolarizations of APD (APD 75 and APD 90 ) were by ϩ43.8 Ϯ 3.6% (n ϭ 8,
The maximum rate of deporization (V max ) significantly decreased by ~18-20% (n ϭ 8-9) at higher concentrations (1-3 M) of these constituents. Other action potential parameters were also unaffected in the presence of all the constituents. The modulation of the action potential configurations by the constituents is summarized in Fig. 3. A washout of the drugs for 15-20 min recovered to ~70-80% of the control value.

Constituents of Gingko biloba and Quercetin Exert Contrasting Effects on Ionic Currents
Whole-cell patch voltage-clamp experiments were performed to examine the effects of the constituents on the underlying ionic currents. Test pulses (1 s duration) were applied to Ϫ20 to 60 mV and Ϫ40 to Ϫ120 mV from a holding potential of Ϫ30 mV. The average capacitance was 86.1 Ϯ 2.0 pF (n ϭ 44). As shown in Fig. 4, ginkgolide A (1 M) enhanced both the Ca 2ϩ current (I Ca ), by 68.2 Ϯ 2.8% (n ϭ 7, P Ͻ 0.001), and the delayed rectifier K ϩ current (I K ) by 66.3 Ϯ 3.8% (n ϭ 7, P Ͻ 0.001). Each constituent (0.1-3 M) was cumulatively administrated to the bath solution. Ginkgolide B at 0.3 M also increased the I Ca at 10 mV by 58.1 Ϯ 2.8% (n ϭ 6, P Ͻ 0.001) (Fig. 5). Increasing the concentrations (1-3 M) still enhanced the I Ca , but the maximal response at 0.3 M decreased. The I K at 60 mV increased by 89.3 Ϯ 3.8% (n ϭ 6, P Ͻ 0.001) at 3 M, but the inwardly rectifying K ϩ current (I K1 ) was not affected significantly. Ginkgolide C had a smaller or no effect on the I Ca , but markedly increased the I K by 59.2 Ϯ 2.8% (n ϭ 7, P Ͻ 0.001) at 3 M (Fig. 6).
On the other hand, the application of quercetin (0.1-3 M) inhibited the I Ca (Fig. 7A and B). The I Ca at 10 mV decreased by 34.9 Ϯ 3.2% (n ϭ 8, P Ͻ 0.05) at 0.3 M and by 56.8 Ϯ 3.3% (n ϭ 8, P Ͻ 0.05) at 3 M. The responses were produced in a concentration-dependent manner. Simultaneously, the I K at 60 mV increased by 60.4 Ϯ 2.7% (n ϭ 8, P Ͻ 0.001) at 0.3 M and by 89.7 Ϯ 3.3% (n ϭ 8, P Ͻ 0.001) at 3 M. Quercetin simultaneously did not affect the I K1 to a significant extent (by~10-15%) at low concentrations, but at 3 M inhibited the I K1 by 12.4 Ϯ 2.1% (n ϭ 8, P Ͻ 0.05). The responses to the constituents for the I Ca and I K currents are summarized in Fig. 8. The responses to all the constituents were recovered to ~80-90% of the control value after a 20-min washout.

Discussion
The present experiments showed that: (i) ginkgolide A prolonged the APD, but ginkgolides B and C shortened APD; (ii) quercetin prolonged the APD; (iii) the constituents also inhibited the V max and the RP was unaffected; (iv) ginkgolides A and B increased I Ca and I K , and decreased I K1 ; (v) ginkgolide C failed to affect the I Ca but increased the I K ; (vi) quercetin inhibited I Ca , and enhanced I K but decreased I K1 ; and (vii) these responses were almost reversible after a washout. These results indicate that acute administrations of these constituents can produce the effective actions on cardiomyocytes. Each constituent never exhibits a uniform response.

APD Regulation
In experiments for the cardiovascular system, GBE does not affect heart rate and contractility, but produces a concentration-dependent increase in coronary flow (12,13). In our laboratory, however, GBE caused a negative chronotropic effect in rabbit SA nodal cells (unpublished data). Also, GBE inhibited I Ca and I K , but bilobalide (a constituent) increased them in guinea pig ventricular cardiomyocytes (14). Simultaneously, GBE prolonged the APD, whereas bilobalide shortened it. In the present experiments, a major action of the constituents was the alteration of APD involved with the effects on the ionic channels. The APD means a period for the repolarization of the membrane, making T wave on ECG. Thus, the APD of cardiac myocytes is clinically reflected directly to the QT interval. The QT interval is a reflection of the repolarization of action potential, which is mainly responsible for the modulation of I K . The APD prolongation increases the refractory period and simultaneously elevates the cellular Ca 2ϩ concentration (26)(27)(28). Therefore, the constituents would play an important role as ionic channel modulators of   cadiomyocytes, although the effects of each constituent on the APD and the ionic currents were not exhibited in a uniform direction. Quercetin, a kind of flavonoid, had a smaller or no effect on the action potentials, but caused the APD prolongation only at high concentrations. Also, quercetin inhibited I Ca and enhanced I K . Quercetin has been reported to possess many pharmacological effects (29,30); the modification of eicosanoid synthesis, prevention of platelet aggregation, and vasorelaxation due to the inhibition of PK-C.

Modulation of the Ionic Currents
Furthermore, these constituents also acted on the V max , although they did not affect other action potential parameters. In general, the V max is used as an activator of the Na ϩ channel current (I Na ). Thus, the constituents have an inhibitory action on the I Na , resulting in an inhibition of the conduction velocity and a suppression of excitability. In addition, the constituents simultaneously modulated the ionic channel currents such as I Ca , I K , I K1 and I Na . These effects might cause antiarrhythmic actions. In all the constituents, I K enhancement was finally produced, indicative of a cell protection due to an APD shortening and a decline of cellular Ca 2ϩ concentration. The channel activity of single myocytes might be caused a run-down, especially Ca 2ϩ channel and rapidly activating K ϩ channel. The cells not causing run-down were chosen and used for the experiments. However, there may be a limitation to patch-clamp experiments. The responses to the constituents were considered to be due to the effects of constituents, because of the reversible response. Therefore, these constituents would exert many helpful and protective actions upon cardiac cells.

Clinical Uses and Summary
GBE can be administered to patients with mild to moderate symptoms of cerebral insufficiency (1,2). The half-life of GBE is 2-3 h (11). In pharmacokinetic analysis, the bioavailabilities of ginkgolides A and B are practically high, but that of ginkgolide C is very low (31). Of the GBE-containing flavonoids, quercetin has the highest percentage (8.91%). With regard to the distribution of radioactivity in the cardiovascular system, the tissues in areas such as vein, heart and aorta are relatively higher 3-7 h after oral administration (32). Quercetin possesses a vasodilating action (33,34). Also, the histamine-induced contraction of isolated guinea pig intestine was inhibited by a mixture of flavonoids (35). In our laboratory, GBE as a mixture exhibited a potent vasodilatory action (34). Therefore, these responses are most likely produced by the modulation of ionic channel currents that have been demonstrated in this study.
It has been reported that a single dose of GBE does not produce potent pharmacological activities, and repeated doses are needed to produce beneficial effects over a long period of administration (36)(37)(38). In the present in vitro experiments, however, acutely single administrations produced marked actions on the ionic channel currents and the action potentials. Since GBE is a mixture of its constituents, the total response to GBE would be a result of complex interactions with the constituents. Therefore, the constituents from GBE would play an important role in modulating the strong APD prolongation by interacting with each other. Although relatively higher concentrations were used in this study, because the pharmacological effects on the ionic channels were so remarkable, GBE and its constituents would be helpful in the treatment of cerebral disorders due to their effects on central nervous system (CNS) neurons. Further experiments are needed to elucidate the detailed mechanisms of the cardiac actions of Ginkgo biloba.