Received 2002-02-27 Accepted 2002-03-15

This study was supported in part by research grants 9930254N (Y.F.X.) from the American Heart Association, and DK38165 & HL62284 (A.L.) and DA12762 (J.P.M) from the National Health Institute of the U.S. Public Health Service.

Corresponding author.

Phone: (617)667-8649；

Fax: (617)667-4833；

Email: yxiao@caregroup.harvard.edu

Acta Physiologica Sinica

Aug. 2002, 54 (4), 271～281

Research Paper

Effects of polyunsaturated fatty acids on cardiac voltage-activated K+ currents in adult ferret cardiomyocytes

Yong-Fu XIAO1,2,*, James P. MORGAN1, Alexander LEAF 2

1Stem Cell Research Laboratory, Beth Israel Deaconess Medical Center;

2Massachusetts General Hospital; Department of Medicine, Harvard Medical School, Boston, MA 02215, USA

Abstract: This study was carried out in adult ferret cardiomyocytes to investigate the effects of the n-3 polyunsaturated fatty acids (PUFAs) on voltage-gated K+ currents. We report that the two outward K+ currents: the transient outward K+ current (Ito) and the delayed rectifier K+ current (IK), are both inhibited by the n-3 PUFAs, while the inwardly rectifying K+ current (IK1) is unaffected by the n-3 PUFAs. Docosahexaenoic acid (C22:6n-3, DHA) produced a concentration-dependent suppression of Ito and IK in adult ferret cardiomyocytes with an IC50 of 7.5 and 20 μmol/L, respectively; but not IK1. In addition, eicosapentaenoic acid (C20:5n-3, EPA) had the effects on the three K+ channels similar to DHA. Arachidonic acid (C20:4n-6, AA) at 5 or 10 μmol/L, after an initial inhibitory effect on IK, caused an activation of IK,AA which was prevented by pretreatment with indomethacin, a cyclooxygenase inhibitor. Monounsaturated and saturated fatty acids, which are not antiarrhythmic, lack the effects on these K+ currents. Our results demonstrate that the n-3 PUFAs inhibit cardiac Ito and IK with much less potency compared to their effects on cardiac Na+ and Ca2+ currents as we reported previously. This inhibition of the cardiac ion currents by the n-3 PUFAs may contribute to their antiarrhythmic actions.

Key words: cardiomyocytes; potassium channels; arrhythmia; polyunsaturated fatty acids

多不饱和脂肪酸对成年雪貂心肌钾通道的作用

萧永福1,2,*， James P. MORGAN1, Alexander LEAF 2

１哈佛医学院医学系贝斯以色列迪肯尼斯医学中心干细胞实验室；２麻省总医院，美国波士顿

摘要:

本研究是在成年雪貂的心肌上研究多不饱和脂肪酸(PUFA)对电压门控钾通道的效应。我们观察到，n-3 PUFA能抑制短时性外向钾电流（I to）和延迟整流钾电流（I K），而对内向整流钾电流（I K1）则没有明显影响。二十二碳六烯酸（DHA）对I to和I k能产生浓度依赖性的抑制作用，其IC50分别为7.5 和20 μmol/L, 但不影响I K1。二十碳五烯酸（EPA）对这三种钾通道的作用与DHA 相似。花生四烯酸（5 或10 μmol/L）先引起IK的抑制，然后引起IK,AA的激活；用环氧合酶抑制剂消炎痛可以阻断花生四烯酸激活IK,AA的作用。不具有抗心律失常作用的单不饱和脂肪酸和饱和脂肪酸都不明显影响这些钾通道的活性。上述实验结果证明，n-3 PUFA能抑制心肌细胞的Ito和IK, 但和我们以前报道的PUFA对心肌钠电流和钙电流的作用相比，其对I to和I k抑制作用的效能较低。n-3 PUFA的抗心律失常效应可能与它们抑制心肌钠、钙、钾通道的作用有关。

关键词: 心肌细胞; 钾通道; 心律失常; 多不饱和脂肪酸

学科分类号: Q463; Q689

Several studies now have reported that the dietary n-3 polyunsaturated fatty acids primarily from marine sources, prevent ischemia-induced malignant cardiac arrhythmias in animal[1-4] and probably in humans[5-7]. This protective effect of ingestion of fish or fish oils is attributed to their eicosapentaenoic acid (C20: 5n-3, EPA) and docosahexaenoic acid (C22:6n-3, DHA) content, as demonstrated by administering a highly concentrated fish oil preparation of the free n-3 PUFAs intravenously just prior to the ischemic challenge and preventing the fatal arrhythmias[4]. But studies with isolated cultured neonatal rat cardiac myocytes showed that PUFAs of both essential fatty acids, n-3 (available primarily from fish in today's diets) and n-6 (predominantly from vegetable oils), are anti-arrhythmic, whereas monounsaturated and saturated fatty acids lack this effect[8]. Arachidonic acid (C20:4n-6, AA), however, is anomalous in its effects. In one-third of the experiments it was arrhythmogenic. This proarrhythmic action was due not to the free AA but rather to cyclooxygenase eicosanoids of AA[8].

The presence of the free PUFAs alters the electrophysiology of the myocytes in a manner to increase their stability[9]. The PUFAs accomplish this by slightly, but significantly, hyperpolarizing the resting or diastolic membrane potential of the myocytes and raising the voltage threshold for the opening of the fast sodium channel, necessitating some 50% increase in the strength of a depolarizing stimulus to elicit an action potential. The PUFAs also prolong the relative refractory period by some 150%, thus prolonging phase 4 of the cardiac electrical cycle despite shortening the duration of the action potential. These two actions, affecting each myocyte in the heart, would account for the increased stability of the myocytes and the slowing of the beating rate or pulse seen regularly with administration of fish oils.

Important contributors to these electrophysiologic effects are the inhibition by the PUFAs of the voltage-dependent fast sodium currents[10] and the L-type voltage-gated calcium currents in cardiomyocytes[11]. The effect seems to result from a direct, noncovalent bonding of the PUFAs to the protein of the sodium channels[12]. Only the antiarrhythmic PUFAs have these effects on the INa; monounsaturated and saturated fatty acids are without the effect.

It is well known that Ito, IK, and IK1, are critical potassium channels for the movement of K+ across the cell membrane. The present study was undertaken to determine the effect of the PUFAs on the voltage-dependent potassium currents in mammalian cardiomyocytes. We report that the antiarrhythmic PUFAs do inhibit Ito and IK, but not IK1, in adult ferret cardiomyocytes. These results may help to interpret the novel antiarrhythmic actions of the n-3 PUFAs.

1 MATERIALS AND METHODS

1.1 Isolation of ferret cardiomyocytes The experimental protocol was approved by the Animal Care Committee of Beth Israel Deaconess Medical Center and performed according to the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH Publication No. 85-23, revised 1996). The methods used to isolate single left ventricular myocytes of adult ferrets (male, 8 to 12 weeks) were similar to those previously described[13]. Briefly, the heart was rapidly excised from a chloroform (CHCl3) anesthetized ferret and the aorta was quickly connected to a modified Langendorff system. This perfusion system had a hydrostatic pressure of 80 cm and a flow rate of 8-10 ml/min. The heart was perfused for 6 min with oxygenated 37℃ Ca2+-free Tyrode's solution containing (in mmol/L): NaCl 137, KCl 5, MgCl2 1, HEPES 10, and glucose 10 with pH 7.4. After this, 50 ml Ca2+-free Tyrode's solution containing 45 to 50 mg collagenase (CLS 2, Worthington Biomedical Corporation, Freehold, NJ, USA), 1 to 2 mg protease Type ⅩⅣ, and 0.1% bovine serum albumin (Sigma Chemical Company, St. Louis, MO) was recirculated for 38 to 45 min. The heart was then sequentially washed with 50 ml 0.2 mmol/L Ca2+ and 50 ml 0.4 mmol/L Ca2+ Tyrode's solution plus 0.1% bovine serum albumin. When the enzymatic solution was completely washed out, several pieces of atrial and left ventricular tissue were cut off and separately placed into two petri dishes (60×15 mm) containing 0.4 mmol/L Ca2+ Tyrode's solution plus 0.1% bovine serum albumin. The cardiac tissue was further cut into finer pieces and gently shaken for 1 to 2 min. Atrial or left ventricular myocytes in the culture dishes were kept at room temperature (22-23℃) for 1 h before patching them.

1.2 Recording of transient outward (Ito), delayed rectifier (IK), and inwardly rectifying (IK1) K+ currents A small amount (15-30 μl) of solution containing freshly isolated ferret cardiomyocytes was transferred to a chamber containing 0.5 ml of 2 mmol/L Ca2+ Tyrode's solution mounted on an inverted microscope (Nikon). Myocytes were continuously superfused (2-3 ml/min) with 2 mmol/L Ca2+ Tyrode's solution. Glass recording pipettes filled with the internal solution had -1 MΩ resistance and were connected via an Ag-AgCl wire to an Axopatch 1D amplifier (Axon Instrumemts, Foster City, CA, USA). Single quiescent atrial or ventricular myocytes with clear striation were chosen for patch clamping. The pipette was carefully advanced to the surface of the cell membrane, and a slow and gentle suction was continuously applied until a gigaseal was achieved[14]. After compensating the electrode capacitance, the whole-cell configuration was achieved by an additional pulse suction. The whole-cell membrane capacitance was routinely measured by a method described elsewhere[13]. Correction of cell capacitance and series resistance was then performed before running an experimental voltage-clamp protocol. Myocytes were equilibrated with the bath and the electrode solutions for 5 to 10 min after forming the whole-cell configuration before data were collected. In order to inactivate the voltage-activated Na+ current, the holding potential was set at －40 mV for IK and IK1 recording. Most of recordings of IK and IK1 were made in the presence of 1 μmol/L nifedipine in the bath solution to block the L-type Ca2+ channel. This concentration of nifedipine did not alter IK and IK1, as well as the PUFA's actions on IK and IK1. To exclude the possibility that the effects of PUFAs on IK were secondary to the block of Na+ channels, therefore, in some experiments 5 μmol/L tetrodotoxin (TTX) was added to the bath solution before the application of PUFAs. We found that 5 μmol/L TTX did not affect the effects of PUFAs. Ito was recorded with a holding potential at －80 mV and in the presence of 1 μmol/L nifedipine and 5 μmol/L TTX. Current recordings were made from the same myocyte before, during, and after exposure to fatty acids. Currents were filtered at 5 kHz and stored on a hard disk of a IBM compatible computer running the PCLAMP 8.2 programs (Axon Instruments, CA, USA). Different concentrations of PUFAs were applied by a fast puffing system[10]. In some experiments after control recordings the ferret ventricular myocytes were pre-superfused with 20 μmol/L indomethacin for 10 min to block cyclooxygenase activity or with 0.1 μmol/L staurosporine for 10 min to suppress protein kinase C (PKC) activity. Effects of DHA or AA were assessed in the ferret myocytes pretreated with staurosporine or indomethacin, respectively. All experiments were performed at room temperature of 22-23℃.

1.3 Materials Fatty acids (obtained from Sigma) were dissolved weekly in 100% ethanol at a concentration of 10 mmol/L and stored under a nitrogen atmosphere at －20℃ before use. The experimental concentration of fatty acids was obtained by dilution of the stocks and the diluted solution contained negligible ethanol which alone had no effect on Ito, IK, or IK1. Tetrodotoxin was obtained from Research Biochemicals Incorporated (Natick, MA, USA) and dissolved in deionized water at 3 mmol/L concentration and stored at 4℃. Nifedipine (Sigma) was dissolved in dimethyl sulfoxide (DMSO) at a concentration of 10 mmol/L. The stock solution with protection from light was stored at 4℃. The pipette solution for recording IK, Ito, and IK1 contained (in mmol/L): KCl 80, KOH 60, MgCl2 1, CaCl2 1, EGTA 10, HEPES 10, MgATP 5, pH=7.3. The 2 mmol/L Ca2+ Tyrode's solution plus 5 μmol/L TTX with or without fatty acid was applied in all the experiments as the bath solution.

1.4 Data analysis The amplitude of IK was measured at the plateau of the current near the end of the voltage pulse (Fig.1). The peak amplitude of Ito was automatically picked up with spike analysis of the Clampfit software (Axon Instruments, CA, USA). The inactivation time constant of Ito was calculated by least-squares fitting of the signal exponential with the Clampfit software. The amplitude of IK1 was measured at a point near the end of voltage pulses. Current density, pA/pF, of single atrial and left ventricular myocytes of IK and Ito was calculated by dividing the amplitude of the current by the cell membrane capacitance. The relative current of IK or Ito after the treatment of PUFAs was calculated as I(PUFAs)/I(Control) from the same cell. Data were analyzed by One Way Analysis of Variance (ANOVA) or by paired or un-paired Student's t test. P<0.05 was considered as statistically significant difference. All data are presented as mean±SEM.

2 RESULTS

2.1 The K+ currents in single adult ferret atrial and left ventricular myocytes

In single atrial and left ventricular myocytes enzymatically isolated from adult ferret hearts we could record all three voltage-activated K+ currents, the transient outward K+ current (Ito), the inwardly rectifying K+ current (IK1), and the delayed rectifier K+ current (IK), which were blocked by 4-aminopyridine, Cs+, and tetraethylammonium (Sigma), respectively. Although the size of single left ventricular myocytes (209±5 pF, n=98, P<0.0001) is significantly larger than that of single atrial myocytes (82±7 pF, n=7), the current densities of IK, calculated from the currents generated by －40 to 70 mV pulses, are not statistically different (P>0.05), 1.9±0.1 and 2.4±0.3 pA/pF for ventricular and atrial cells, respectively. Changing the voltage step of pulses from －40 to 70 mV to －80 to 70 mV did not affect the amplitude of IK, 398±62 pA vs 413±71 pA (P>0.05, n=11), measured at a point near the end of 4000 ms pulses (Fig.1A). Changing the holding potential from －40 to －80 mV also did not alter the current-voltage relation and the reversal potential (around －80 mV) of IK. The difference we found was that Ito activated by depolarizing pulses was more prominent at the holding potential of －80 mV. However, more than 50% of the left ventricular myo-

Fig.1. Inhibition of the delayed rectifier K+ current (IK) by DHA in ferret left ventricular myocytes. A: The superimposed traces of IK were recorded from a ferret ventricular myocyte. The currents were elicited by 4000 ms pulses from a holding potential of －40 mV to 60 mV with 10 mV increments (see the protocol in panel A) every 10 s. B: IK recorded from the same cell as in A was markedly suppressed by extracellular application of 10 μmol/L DHA. The horizontal dotted lines in A and B are the 0 current level. The vertical dotted lines are the places where the late portions of IK were measured. C: The averaged current-voltage relationships of IK in adult ferret cardiomyocytes were plotted for control (○) and 10 μmol/L DHA (●, n=6). D: Concentration-dependent suppression of IK was shown. The IC50 of DHA was 20 μmol/L. The relative current was calculated by IK(DHA)/IK(control). IK was elicited by a 4000-ms pulse from a holding potential of －40 to 70 mV and measured at the time point as in panels A and B. Each bar represents the mean±SEM of at least 6 cells. *P<0.05， **P<0.01， ***P<0.001 vs control.

cytes patched in this study had significant Ito when the myocytes were depolarized from a holding potential of －40 mV.

2.2 Suppressant effects of DHA on IK

Ferret left ventricular myocytes were voltage-clamped by 4000 ms depolarizing pulses from a holding potential of －40 mV. Figure 1 shows that depolarizing steps evoked a family of outward currents. Extracellular application of 10 μmol/L DHA induced a significant reduction of the amplitude of the currents and reached the maximal effect within 3 min. This inhibitory effect of DHA on IK was effective on all the currents activated by the various voltages (Fig.1). The DHA-induced suppression of IK did not alter the current-voltage relationship (Fig.1C). IK returned toward the control level after washing off PUFAs from the myocytes with 2 mg/ml delipidated bovine serum albumin added to the bath. The DHA-induced inhibition (10 μmol/L) of IK was also found in atrial myocytes in which IK (elicited by single step pulses with 4000 ms duration from －40 mV to 70 mV) was suppressed by 62±4% (P<0.05, n=7). This effect is very similar to the suppression of IK in the left ventricular myocytes by the same concentration of DHA (58±6% of the control, P<0.01, n=8, Fig.1).

In order to test the possibility that different holding potential might affect the DHA-induced inhibition of IK, we did another group of experiments to hold the myocytes at －80 mV instead of －40 mV. Under control condition the amplitude of IK elicited by pulses to 70 mV was similar for －40 or －80 mV. Extracellular application of 5 μmol/L DHA produced similar suppression of IK, 69±3% (n=12, P<0.01, vs control) and 63±6% (n=7, P<0.01, vs control) for the holding potentials at －40 and －80 mV, respectively. Therefore, changing the holding potential did not affect the DHA-induced inhibition of IK in ferret ventricular myocytes.

The DHA-induced inhibition of IK is concentration-dependent in adult ferret ventricular myocytes. Figure 1D indicates that at a concentration as low as 0.2 μmol/L of DHA IK was significantly suppressed and the inhibition increased when the concentration was raised. In the ventricular myocytes treated with 50 μmol/L DHA IK was suppressed by 61±4% (P<0.001, n=11). The apparent IC50 of DHA on IK was around 20 μmol/L in ferret ventricular myocytes.

In view of the reports of regulation of IK involving protein kinase C (PKC)[15] and the inhibition of INa and ICa by PUFAs also involving PKC[16], we did another series of experiments to test a possible role of PKC in the PUFA-induced suppression of IK. Extracellular perfusion with the PKC inhibitor staurosporine (0.1 μmol/L, a concentration reported to produce a maximal inhibition of PKC[16]) for 10 min had no effect on IK, but 10 μmol/L DHA produced a significant suppression (65±6%, P<0.05, n=5). The time course to reach the maximal inhibition of IK was similar to that for those myocytes perfused with DHA alone (data not shown).

2.3 Effects of other polyunsaturated and saturated fatty acids on IK

Table 1 summarizes the effects of other fatty acids, such as eicosapentaenoic acid (EPA), eicosatetraynoic acid (ETYA), linolenic acid (LNA), linoleic acid (LA), oleic acid (OA), and stearic acid (SA), on IK of ferret left ventricular myocytes. EPA, ETYA, LNA, and LA at 5 and 10 μmol/L produced significant inhibition on IK which is similar to the effects of DHA on IK. In contrast, extracellular application of 5 and 10 μmol/L of OA, a monounsaturated fatty acid, and SA, a saturated fatty acid, did not have any significant effects on IK (Table 1).

Table　１. Effects of fatty acids on IK and Ito in ferret left ventricular myocytes

Fatty acid	Structure	% inhibition of IK		% inhibition of Ito
Fatty acid	Structure	5 μmol/L	10 μmol/L	10 μmol/L
DHA	22:6 (n-3)	31±3a (12)	42±6c (8)	57±7c (7)
EPA	20:5 (n-3)	26±4a (6)	40±3c (8)	67±5b (4)
LNA	18:3 (n-3)	22±3b (7)	46±4c (8)	49±8a (4)
LA	18:2 (n-6)	18±3a (6)	28±2b (9)	65±9a (4)
ETYA	20:4	22±3b (6)	46±4c (6)	64±10a (4)
OA	18:1 (n-9)	9±11 (9)	8±10 (7)	---
SA	18:0	5±3 (6)	2±7 (6)	4±2 (4)

Values are expressed as the mean percent inhibition±SEM. Numbers in parentheses are the numbers of individual cells treated with fatty acids. Statistical significance was tested between the peak amplitudes of IK and Ito suppressed by fatty acids and their corresponding controls. DHA, docosahexaenoic acid; EPA, eicosapentaenoic acid; LNA, α-linolenic acid; LA, linoleic acid; ETYA, eicosatetraynoic acid; OA, oleic acid; SA, stearic acid. aP<0.05; bP<0.01; cP<0.001; versus their corresponding controls.

2.4 Activation of an outwardly rectifier current (IK,AA) by AA

As shown in Fig.2A and 2B the effect of AA on the potassium current was biphasic with an initial inhibition of IK and late activation of IK,AA. The inhibition occurred within 30 s after application of 10 μmol/L AA. The activation was observed between 1 to 4 min after the myocytes were treated with 10 μmol/L AA and the average time for the activation was 106±16 s (n=10). During the AA-induced activation of the outward current we monitored the effect of AA on Ito (Fig.2C) to see if AA still had its expected effects as a free PUFA to inhibit Ito.

Fig.2. Effects of arachidonic acid (AA) on IK in ferret left ventricular myocytes. A: Time-course of 10 μmol/L AA on IK of a ventricular myocyte. The methods for recording and measurement of IK were the same as described in Fig.1C (opened box), control; AA (solid box), 10 μmol/L arachidonic acid; W (hatched box), washout of AA. The interruption of X-axis represents the interval when the measurement of Ito and the current-voltage relation of IK were made during the application of AA. B: The current-voltage relation curves were plotted according to the currents recorded for the control (○), 10 μmol/L AA (●), and 3 min after washout of AA (△). The inset is the original current traces recorded by 4000 ms single-step pulses from a holding potential of －40 mV to 70 mV.

Although the steady-state holding current was increased, Ito was significantly inhibited during the activation by AA (Fig.2C, Middle trace).

The activated outward current is concentration-related. Although 5 μmol/L AA did not cause a significant initial inhibition of IK (17±2%, n=5, P>0.05), the late activation of the outward current in the same cells was significant (136±10% of the control, n=5, P<0.05). 10 μmol/L AA, however, produced significant effects on IK for both inhibition (29±4%, n=10, P<0.01) and activation (210±20% of the control, n=10, P<0.01) in the same cells.

It has been reported[8] that AA alone could cause fibrillation of cultured rat myocytes and indomethacin blocked the fibrillation. In the present experiment ETYA, a non-metabolized acetylenic analog of AA, which blocks all the enzymes which oxygenate AA[17], produced only an inhibitory effect on IK in ferret ventricular myocytes. To test further whether the AA-activated outward current was due to its metabolites, pre-perfusion of 20 μmol/L indomethacin, a cyclooxygenase inhibitor, for 10 min prevented the AA-activated outward currents in all ventricular myocytes tested with 10 μmol/L AA and only produced an AA-induced suppression of IK (42±8%, P<0.05, n=7). Perfusion with 20 μmol/L indomethacin alone for 10 min did not cause significant change of IK (99±2% of the control, P>0.05).

2.5 Effects of PUFAs on Ito

As shown in Fig.3A, 10 μmol/L DHA suppressed superimposed transient K+ currents, Ito, recorded from a ferret ventricular myocyte and the suppression was fully reversible after washout of DHA (Fig.3B). The slope of the current-voltage relationship of Ito elicited by pulses from a holding potential of －80 mV down to －90 mV and up to 60 mV with 15-mV increments for each step was reduced (Fig.3C). Within 10 s after fast superfusion of 10 μmol/L DHA solution Ito began to decline and achieved a new steady-state low level within 2 min (Fig.3B). Also within 2 min Ito returned to the pre-drug level after washout of DHA with the bath solution containing 2 mg/ml of delipidated bovine serum albumin. The DHA-induced inhibition of Ito is concentration-dependent with an IC50 of 7.5 μmol/L (Fig.3D). Fig.3A and Fig.4 (A and B) show that 10 μmol/L DHA significantly enhanced the inactivation process of Ito in ferret cardiomyocytes. The inactivation time constant of Ito elicited by 400-ms single voltage step pulses from －80 mV to 60 mV was significantly reduced, from 77±7 ms for control to 36±8 ms for 10 μmol/L DHA.

Fig.3. DHA-induced suppression of Ito in ferret ventricular myocytes. A: Superimposed current traces of Ito were elicited by 300 ms single-step pulses from a holding potential of －80 mV to 60 mV in the absence (a, control; c, washout) and presence (b) of 10 μmol/L DHA. B: The time course of the inhibition of Ito after extracellular perfusion of 10 μmol/L DHA solution (open bar). The symbols of a, b and c represented the time when the original currents in panel A were recorded. C: The effects of DHA on the current-voltage relationship of Ito (○, control; ●, 10 μmol/L DHA; □, washout). The membrane holding potential was －80 mV. Ito was elicited by a group of voltage commands with 300 ms duration from －90 to 60 mV in 15-mV increments every 10 s. D: Concentration-dependent inhibition of Ito. Each bar represents the mean±SEM of peak Ito of 4 to 7 cells. Ito was elicited by 300-ms single-voltage steps from a holding potential of －80 mV to 60 mV every 10 s. The IC50 of DHA is approximately 7.5 μmol/L DHA. *P<0.05; ***P<0.001 vs control.

(n=7, P<0.01). Figure 4 also shows that while Ito was significantly inhibited by 10 μmol/L DHA, the voltage dependence of the steady-state inactivation was not affected (Fig.4C and 4D). Currents were elicited with a double-pulse protocol which consisted of a 250 ms test pulse to 60 mV following a 500 ms conditioning pulse varying from －120 mV to 30 mV with 10-mV increments. The V1/2 of the steady-state inactivation was －41±1 mV for control and －38±2 mV for DHA (P>0.05, n=6).

Other fatty acids, such as EPA, ETYA, LNA, LA, and LA, at 10 μmol/L concentration (Table 1) also produced a significant inhibition of Ito. In contrast, extracellular application of the same concentration of SA (stearic acid), a saturated fatty acid, had no significant effect on Ito.

2.6 Effects of PUFAs on IK1

Since we observed the suppressant effect of PUFAs on IK and Ito (except for the late stimulation by AA of IK,AA), it was of interest to evaluate whether PUFAs had any action on the inwardly rectifier K+ current (IK1), activated by hyperpolarization, in ferret left ventricular myocytes. DHA at 10 μmol/L altered neither the slope nor the amplitude of IK1 significantly (Fig.5). Specifically, IK1 elicited by 200 ms pulses with a single voltage step pulse from －40 to －140 mV was 102±4% (P>0.05, n=6) and 98±6% (P>0.05, n=5) of the control for 5 and 10 μmol/L of DHA, respectively. Furthermore, we even did not find any inhibition of IK1 in the presence of 20 μmol/L DHA (n=2). Also in Fig.4A and 4B, whereas DHA clearly inhibited Ito, IK1, activated during the conditioning prepulses, was not affected. The effects of other fatty acids, including EPA, ETYA, LNA, LA, AA, and SA, were consistent with DHA and had no effects on IK1 in ferret ventricular myocytes (data not shown).

Fig.4. Effects of DHA on the steady-state inactivation of Ito. A and B are the superimposed current traces recorded from a ventricular myocyte in the absence (A) and presence (B) of 10 μmol/L DHA. The cell was held at －80 mV and paired command voltage steps were applied. The first of these was a 500-ms prepulse which varied in magnitude from －120 to 30 mV in 10-mV increments every 10 s. The second one was a 250-ms testing pulse from the various voltages of the prepulses to 60 mV. C: Normalized mean peak currents (n=6) were plotted against the voltages of prepulses for control (○) and 10 μmol/L DHA (△). The dotted vertical lines in A and B are the places where peak Ito was measured. D: Relationship of prepulse voltage and DHA-induced suppression (n=6) of peak Ito in the absence (○) and presence (△) of 10 μmol/L DHA.

Fig.5. Insensitivity of the inward rectifier K+ current (IK1) to extracellular application of DHA in a representative ferret ventricular myocyte. Panel A shows the superimposed traces of IK1 in the absence (○) and presence (△) of 10 μmol/L DHA. IK1 was elicited by 200 ms pulses from a holding potential of －40 mV to －190 mV with increments of －10 mV every 10 s. B: The current-voltage relation curves were plotted according to the values measured at the vertical dotted lines of the superimposed traces of IK1 in pane A.

3 DISCUSSION

A major reason for this study was to explain the electrophysiologic effects of PUFAs on arrhythmias. As reported[9], they increase the resting or diastolic membrane potential, increase the threshold voltage for the gating of the fast Na+ channels, and markedly prolong the relative refractory period in cultured neonatal rat cardiomyocytes, despite shortening the duration of the action potential. In the unanesthetized dogs they also shorten the QT interval (the action potential duration) and slow the pulse rate[4], similar, in both respects, to their effects on the cultured myocytes[8]. Although it seemed earlier perhaps possible to explain the electrophysiologic effects of PUFAs on the basis of their blocking voltage-gated Na+ and Ca2+ currents[10,11,19-21], it was important to determine their actions on other ion currents. In fact, inhibition of the outward K+ currents, Ito and IK, by PUFAs acts in the wrong direction to explain the observed electrophysiologic effects of PUFAs. Blocking Ito and IK would both act to prolong rather than shorten the duration of the action potential. Indeed, we have reported the effects of PUFAs on the L-type Ca2+, ICa,L,[11] as well as on the voltage-gated Na+ currents, INa[10]. PUFAs inhibit both ICa,L and INa with much greater potency than the potassium currents. The IC50s for ICa,L, INa, Ito, and IK are 0.8, 4.8, 7.5, and 20 μmol/L, respectively. It seems that the inhibitory effect of PUFAs on the ICa,L and INa would dominate over the weaker inhibition of the outward K+ currents to explain the observed small, but significant, reduction in the action potential duration of cultured cardiomyocytes and the electrocardiographic QT interval in dogs.

The finding that only PUFAs with a free carboxyl group blocked the Na+ currents in isolated neonatal rat cardiomyocytes[18] suggested that their negatively charged carboxyl group is important for the PUFAs' action. The present results are consistent with some earlier isolated studies on the repolarizing K+ currents (Ito) in rat cardiac myocytes[22,23] and on IA and IK in dissociated pinealocytes[24]. We find that only the same PUFAs which are antiarrhythmic in isolated cardiomyocytes in vitro and in dogs in vivo have the inhibitory effects on IK and Ito, but no effect on IK1. The lack of an effect of PUFAs on IK1 may result from the difference of channel structure among IK, Ito, and IK1[25]. Honore et al.[23] have reported that DHA and AA blocked the major cardiac delayed rectifier K+ channel (Kv1.5) when the later was expressed in a Chinese hamster ovarian cell line, as well as IK in cultured mouse and rat cardiac myocytes. They reported that the inhibition occurred when DHA was applied extracellularly and not when included in the patch electrode.

Several studies show that IK1 is essential for maintaining different resting membrane potential in rabbit pacemaker or non-pacemaker cardiomyocytes[26] and in rat microglia[27]. During myocardial ischemia intracellular K+ leaks out due to membrane damage, which causes membrane depolarization and arrhythmias. A lower resting potential in ischemic (－70 mV) than in nonischemic (－78 mV) human ventricular myocytes has been reported[28]. In addition, anesthetic thiopental depolarized resting membrane potential by depression of IK1 and increased the incidence of ventricular arrhythmias[29]. Here, we demonstrate that the n-3 PUFAs inhibit IK and Ito, but not IK1. These effects result in a decrease in the efflux of K+ without a depolarization of cardiac resting membrane potential, which is consistent with our previous finding of slight hyperpolarization of resting cardiomyocytes in the presence of PUFAs[9]. Therefore, the effects of the n-3 PUFAs on K+ channels plus on Ca2+ and Na+ channels may protect the heart from arrhythmias during myocardial ischemia.

The effects of the PUFAs on the K+ currents (with the exception of the late stimulatory effect of AA on an outward K+ current) seem to be a direct effect of the fatty acids, as we have shown to be the case for their inhibitory action on the fast voltage-dependent Na+ channel[12] and on the L-type Ca2+ channels[30]. A possible effect of the PUFAs via stimulation of protein kinase C seems excluded by the failure of staurosporine, a potent PKC inhibitor, to block the control IK or the inhibitory effect of DHA on IK. Indomethacin, a potent PKC inhibitor of cyclooxygenase blocked the late stimulatory action of AA on IK,AA but not the initial inhibitory effect. Furthermore, ETYA, an inhibitor of the cyclooxygenase, lipoxygenase, and epoxygenase enzymes[17], that produce the active oxygenated metabolites of AA, was itself only an inhibitor of IK (Table 1).

Only the free polyunsaturated fatty acids which we previously identified to be antiarrhythmic in the cultured spontaneously beating neonatal rat cardiomyocytes[8] affect the K+ currents, as they do the Ca2+ and Na+ currents. Of the dietary PUFAs tested, both the n-3 and n-6 are antiarrhythmic, only AA (C20:4n-6) was anomalous[8]. It frequently induced arrhythmias when added alone to the cultured spontaneously and rhythmically beating neonatal rat myocytes. This arrhythmic effect, however, proved not to be due to the free AA, but rather to an oxygenated metabolite of AA. When AA was added to the myocytes together with an inhibitor of cyclooxygenase, indomethacin (20 μmol/L), AA behaved like the n-3 PUFAs in being only antiarrhythmic[8]. It is of interest, therefore, that of the PUFAs tested only AA again was anomalous, producing a stimulation of an outward K+ current in the cardiomyocytes following an initial inhibitory effect on IK (Fig.2). Furtheremore, this stimulatory effect on IK,AA by AA was also prevented by the cyclooxygenase inhibitor, indomethacin. Whether the stimulation of IK,AA by AA is related to the arrhythmogenic action of AA, we do not now know, but the correlation suggests such a relationship.

In conclusion, polyunsaturated fatty acids of both the n-3 and n-6 classes produce a concentration-dependent inhibition of two major voltage-dependent potassium currents, IK and Ito, in adult ferret cardiomyocytes. The PUFAs had no effect on IK1 which is critical for maintaining normal resting membrane potential[27]. AA, unlike the other PUFAs, produced a late stimulation of an outward potassium current following the expected initial inhibition of IK. This effect was not due to the free AA but to a cyclooxygenase metabolite, which may have relevance to the frequent arrhythmogenic effect of AA on cultured neonatal rat cardiomyocytes, which also results from a cyclooxygenase metabolite of AA, not from the free AA itself.

REFERENCES

[1] McLennan, PL, Abeywardena MY, Charnock JS. Influence of dietary lipids on arrhythmias and infarction after coronary artery ligation in rats. Can J Physiol Pharmacol, 1985,63:1411-1417.

[2] McLennan PL, Bride TM, Abeywardena MY, Charnock JS. Dietary lipid modulation of ventricular fibrillation threshold in the marmoset monkey. Am Heart J, 1992,123:1555-1561.

[3] Hock CE, Beck LD, Bodine LC, Reibel DK. Influence of dietary n-3 fatty acids on myocardial ischemia and re-perfusion. Am J Physiol (Heart and Circ Physiol), 1990,259:H1518-H1526.

[4] Billman GE, Kang JX, Leaf A. Prevention of sudden cardiac death by dietary pure omega-3 polyunsaturated fatty acids in dogs. Circulation, 1999,99(18):2452-2457.

[5] Burr ML, Fehily AM, Gilbert JF, Rogers S, Holliday RM, Sweetnam PW, Elwood PC, Deadman NM. Effects of changes in fat, fish, and fibre intakes on death and myocardial reinfarction: diet and reinfarction trial (DART). Lancet, 1989, 334:757-761.

[6] de Lorgeril M, Renaud S, Mamelle N, Salen P, Martin JL, Monjaud I, Guidollet J, Toubout P, Delaye J. Mediterranean alpha-linolenic acid-rich diet in secondary prevention of coronary heart disease. Lancet,1994,343:1454-1459.

[7] Siscovick DS, Raghunathan TE, King I, Siscovick DS, Raghunathan TE, King I, Weinmann S, Wicklund KG, Albright J, Bovbjerg V, Arbogast P, Smith H, Kushi LH. Dietary intake and cell membrane levels of long-chain n-3 polyunsaturated fatty acids and the risk of primary cardiac arrest. J Am Med Assoc, 1995,274:1363-1367.

[8] Kang JX, Leaf A. Effects of long-chain polyunsaturated fatty acids on the contraction of neonatal rat cardiac myocytes. Proc Natl Acad Sci (USA),1994,91:9886-9890.

[9] Kang JX, Xiao YF, Leaf A. Free, long-chain, polyunsaturated fatty acids reduce membrane electrical excitability in neonatal rat cardiac myocytes. Proc Natl Acad Sci (USA), 1995,92:3997-4001.

[10] Xiao YF, Kang JX, Morgan JP, Leaf A. Blocking effects of polyunsaturated fatty acids on Na+ channels of neonatal rat ventricular myocytes. Proc Natl Acad Sci (USA), 1995,92:11000-11004.

[11] Xiao YF, Gomez AM, Morgan JP, Lederer WJ, Leaf A. Suppression of voltage-gated L-type Ca2+ currents by polyunsaturated fatty acids in adult and neonatal rat ventricular myocytes. Proc Natl Acad Sci (USA), 1997,94:4182-4187.

[12] Kang JX, Leaf A. Evidence that free polyunsaturated fatty acids modify Na+ channels by directly binding to the channel proteins. Proc Natl Acad Sci (USA), 1996,93:3542-3546.

[13] Xiao YF, McArdle JJ. Elevated density and altered pharmacologic properties of myocardial calcium current of the spontaneously hypertensive rat. J Hypertens, 1994,12:783-790.

[14] Hamill OP, Marty A, Neher E, Sakmann B, Sigworth FJ. Improved patch-clamp techniques for high resolution current recording from cells and cell-free membrane patches. Pfluegers Arch, 1981,391:85-100.

[15] Tohse N, Kameyama M, Sekiguchi K, Shearman MS, Kanno M. Protein kinase C activation enhances the delayed rectifier potassium current in guinea-pig heart cells. J Mol Cell Cardiol, 1990,22:725-734.

[16] Linden DJ, Routtenberg A. Cis-fatty acids, which activate protein kinase C, attenuate Na+ and Ca2+ currents in mouse neuroblastoma cells. J Physiol (Lond), 1989,419:95-119.

[17] Tobias LD, Hamilton JG. The effect of 5,8,11,14-eicosatetraynoic on lipid metabolism. Lipids, 1978,14:181-193.

[18] Kang JX, Leaf A. Protective effects of All-trans-retinoic acid against cardiac arrhythmias induced by isoproterenol, lysophosphatidylcholine or ischemia and reperfusion. J Cardiovasc Pharmacol, 1995, 26:943-948.

[19] Xiao YF, Wright SN, Wang GK, Morgan JP, Leaf A. Fatty acids suppress voltage-gated Na+ currents in HEK293t cells transfected with the α-subunit of the human cardiac Na+ channel. Proc. Natl Acad Sci (USA), 1998,95:2680-2685.

[20] Xiao YF, Wright SN, Wang GK, Morgan JP, Leaf A. Coexpression with the ?1 subunit modifies the kinetics and fatty-acid block of hH1? Na+ channels. Am J Physiol (Heart and Circ. Physiol.), 2000,279:H35-H46.

[21] Xiao YF, Ke Q, Wang SY, Auktor K, Yang Y, Wang GK, Morgan JP, Leaf A. Single point mutations affect fatty acid block of human myocardial sodium channel alpha subunit Na+ channels. Proc Natl Acad Sci (USA), 2001,98:3606-3611.

[22] Bogdanov KY, Spurgeon HA, Vinogradova TM, Lakatta EG. Modulation of the transient outward current in adult rat ventricular myocytes by polyunsaturated fatty acids. Am J Physiol (Heart and Circ. Physiol.), 1998,274:H571-H579.

[23] Honore E, Barhanin J, Attali B, Lesage F, Lazdunski M. External blockade of the major cardiac delayed-rectifier K+ (Kv1.5) by polyunsaturated fatty acids. Proc Natl Acad Sci (USA), 1994,91:1937-1941.

[24] Poling JS, Karanian JW, Salem N Jr, Stefano V. Time- and voltage-dependent block of delayed rectifier potassium channels by docosahexaenoic acid. Mol Pharmacol, 1995, 47:381-390.

[25] Catterall WA. Structure and function of voltage-sensitive ion channels. Science, 1989, 242:50-61.

[26] Noma A, Nakayama T, Kurachi Y, Irisawa H. Resting K conductances in pacemaker and non-pacemaker heart cells of the rabbit. Jpn J Physiol, 1984, 34:245-254.

[27] Chung S, Jung W, Lee MY. Inward and outward rectifying potassium currents set membrane potentials in activated rat microglia. Neurosci Lett, 1999,262:121-124.

[28] Mubagwa K, Flameng W, Carmeliet E. Resting and action potentials of nonischemic and chronically ischemic human ventricular muscle. J Cardiovasc Electrophysiol, 1994,5:659-671.

[29] Pancrazio JJ, Frazer MJ, Lynch C. Barbiturate anesthetics depress the resting K+ conductance of myocardium. J Pharmacol Exp Ther, 1993,265:358-365.

[30] Hallaq H, Smith T W, Leaf A. Modulation of dihydropyridine-sensitive calcium channels in heart cells by fish oil fatty acids. Proc Natl Acad Sci (USA), 1992,89:1760-1764.