Suppression of voltage-gated L-type Ca²⁺ currents by polyunsaturated fatty acids in adult and neonatal rat ventricular myocytes

Abstract

Our recent data show that in cardiac myocytes polyunsaturated fatty acids (PUFAs) are antiarrhythmic. They reduce I_Na, shorten the action potential, shift the threshold for excitation to more positive potentials, and prolong the relative refractory period. In this study we use patch-clamp techniques in whole-cell mode and confocal Ca²⁺ imaging to examine the effects of PUFAs on the voltage-gated L-type Ca²⁺ current (I_Ca,L), elementary sarcoplasmic reticulum Ca²⁺-release events (Ca²⁺-sparks), and [Ca²⁺]_i transients in isolated rat ventricular myocytes. Extracellular application of eicosapentaenoic acid (EPA; C20:5 n − 3) produced a prompt and reversible concentration-dependent suppression of I_Ca,L. The concentration of EPA to produce 50% inhibition of I_Ca was 0.8 μM in neonatal rat heart cells and 2.1 μM in adult ventricular myocytes. While the EPA induced suppression of I_Ca,L, it did not significantly alter the shape of the current–voltage relation but did produce a small, but significant, negative shift of the steady-state inactivation curve. The inhibition of I_Ca,L was voltage- and time-dependent, but not use- or frequency-dependent. Other PUFAs, such as docosahexaenoic acid, arachidonic acid, linolenic acid, linoleic acid, conjugated linoleic acid, and eicosatetraynoic acid had similar effects on I_Ca,L as EPA. All-trans-retinoic acid, which had been shown to suppress induced arrhythmogenic activity in rat heart cells, also produced a significant inhibition of I_Ca,L. The saturated stearic acid and the monounsaturated oleic acid had no effect on I_Ca,L. Because both I_Ca,L and sarcoplasmic reticulum Ca²⁺-release underlie many cardiac arrhythmias, we examined the effects of EPA on I_Ca,L and Ca²⁺-sparks. While EPA suppressed both, it did not change the temporal or spatial character of the Ca²⁺-sparks, nor did it alter the ability of I_Ca,L to trigger Ca²⁺-sparks. We conclude that PUFAs may act as antiarrhythmic agents in vivo in normal and Ca²⁺-overloaded cells principally because they reduce Ca²⁺ entry by blocking I_Ca,L. Furthermore, PUFAs act directly to decrease I_Na and I_Ca,L, but indirectly to reduce the [Ca²⁺]_i transients and [Ca²⁺]_i-activated membrane current. Although a negative inotropic action is associated with application of PUFAs, it is clear that by reducing I_Ca,L, I_Na and Ca²⁺-sparks, PUFAs can reduce spontaneous extrasystoles in the heart. The mechanisms by which PUFAs act are discussed.

Several studies now have reported that the n-3 polyunsaturated marine oils prevent ischemia-induced malignant cardiac arrhythmias in animals (1–4) and probably in humans (5–7). This protective antiarrhythmic effect of ingestion of fish or fish oil has been shown to be attributable to their eicosapentaenoic acid (EPA; C20:5 n − 3) and docosahexaenoic acid (C22:6 n − 3) content as demonstrated by administering a highly concentrated fish oil preparation of free n − 3 PUFAs intravenously just before the ischemic challenge and preventing the fatal arrhythmias (4). But studies with isolated cultured neonatal rat cardiac myocytes showed that polyunsaturated fatty acids, both n − 6 and n − 3, are antiarrhythmic during Ca²⁺ overload (8). Arachidonic acid (AA; C20:4 n − 6), however, is anomalous in its effects. Although, as a free fatty acid, AA is also antiarrhythmic, its cyclooxygenase eicosanoids cause arrhythmias (8). The monounsaturated fatty acid, oleic acid (C18:1 n − 9), and saturated fatty acids have no antiarrhythmic effects.

Further the antiarrhythmic effect occurs quickly but only with the free fatty acid form of the PUFAs; the ethyl ester or triglyceride are not promptly antiarrhythmic (8). Presumably, the free PUFAs exert their effect by partitioning into the hospitable, hydrophobic environment of the cell membranes among the acyl chains of the phospholipids of cardiomyocytes, but a receptor-mediated signaling pathway cannot be ruled out. In any case, the effect is quickly reversed when the free PUFAs are extracted from the cells by adding delipidated BSA to the bathing medium. This observation suggests that the PUFAs are neither fully incorporated into membrane phospholipids nor covalent bound to any constituents of the myocyte to produce the antiarrhythmic effect (8).

The presence of the free PUFAs in the extracellular medium alters the electrophysiology of the cardiac myocytes and can restore electrical stability to Ca²⁺-overloaded cells. This apparently beneficial action of the PUFAs is associated with a reduction in cellular excitability due to a more positive threshold for the gating of I_Na, hyperpolarization of the resting potential, and prolongation of the relative refractory period despite shortening of the action potential (9) and a reduced I_Na (10). In this study we examine the role of I_Ca,L and sarcoplasmic reticulum (SR) Ca²⁺ release in the action of PUFAs on cardiac myocytes. We observe clear reductions in I_Ca,L and SR Ca²⁺ release, both important in Ca²⁺ overload arrhythmias (11–15). We find that the direct action of PUFAs responsible for the reduced SR Ca²⁺ release is on I_Ca,L. We discuss the possible molecular actions of PUFAs that can account for their reductions of I_Na, potassium currents, and I_Ca,L.

MATERIALS AND METHODS

Materials.

Fatty acids (obtained from Sigma) were dissolved in 100% ethanol at a concentration of 10 mM and stored under a nitrogen atmosphere at −20°C before use. The experimental concentration of fatty acids was obtained by dilution of the stocks and contained negligible ethanol, which alone had no effect on I_Ca. The pipette solution for recording I_Ca,L contained (in mM): CsCl 100, CsOH 40, MgCl₂ 1, CaCl₂ 1, EGTA 11, Hepes 10, MgATP 5 (pH, 7.3) or contained (whenever [Ca²⁺]_i was measured) CsCl 130, fluo-3 0.1, Hepes 10, MgCl₂ 0.33, tetraethylammonium chloride 20, MgATP 4 (pH, 7.2). The bathing solution contained (in mM): NaCl 137, CsCl 5, MgCl₂ 1, CaCl₂ 2, Hepes 10, glucose 10 (pH, 7.4) for the measurements in neonatal heart cells. The experiments in adult myocytes used external solutions containing (in mM) NaCl 135, MgCl₂ 1, CsCl₂₀, glucose 10, Hepes 10, 4-aminopyridine 3, CaCl₂ 1 (pH, 7.4). In some cases, N-methyl d-glucamine (120 mM) replaced the NaCl (noted in text).

Culture of Neonatal Rat Ventricular Myocytes.

Primary cultures of ventricular myocytes were prepared from 1-day-old neonatal rats with a commercial isolation kit (Worthington). This kit uses purified enzyme preparations to produce separated cells, which were seeded onto glass coverslips (3 mm diameter) and beat spontaneously. Cells were incubated at 37°C in air with 5% CO₂ added and 98% relative humidity (model 3123, Forma Scientific, Marietta, OH). The culture medium was changed every other day. Ventricular myocytes used for patch-clamp experiments were 3 to 5 days in culture.

Isolation of Adult Rat Ventricular Myocytes.

The methods used to isolate single left ventricular myocytes of adult rats (males with 200 to 300 g of body weight, Sprague–Dawley; Charles River Breeding Laboratories) was similar to those previously described (13, 16). Briefly, the heart was rapidly excised from an anesthetized rat (using inhaled ethyl ether or i.p. pentobarbital), and the aorta was quickly connected to a modified Langendorf system and perfused through the coronary arteries with modified Tyrode’s solutions containing very low levels of [Ca²⁺] to remove blood, and then enzymes (including collagenase (CLS 2, Worthington) and protease (Sigma) were added to the perfusion solution. After enzymatic treatment (10–25 min), the hearts were washed in an enzyme-free solution, the heart was cut into small pieces, and cells were shaken loose. Extracellular Ca²⁺ was increased to 1 mM, and cells were kept at room temperature until use.

Recording of I_Ca,L, the Ca²⁺ Current.

Standard patch-clamp methods in whole-cell configuration were used (14, 17, 19) to record membrane current in cultured neonatal and adult rat cardiac myocytes. Electrodes had uncompensated resistance of between 0.5 and 4 MΩ. For adult cells we used only the lower resistance electrodes. Current recordings were made from the same myocyte before, during, and after exposure to fatty acids using an Axopatch 1D or 200A patch clamp amplifier (Axon Instruments, Foster City, CA). Currents were filtered at 2 kHz and analyzed using PClamp software. Various concentrations of PUFAs were applied by a fast puffing system or a rapid change bath. All experiments were performed at a room temperature of 21–23°C.

Ca²⁺-Sparks, Ca²⁺ Imaging [Ca²⁺]_i Measurements.

To a standard pipette solution, we added 100 μM fluo-3 (see above). An MRC 600 confocal microscope coupled to a Nikon Diaphot microscope with a Zeiss 63× 1.25 NA oil immersion lens was used to image cardiac myocytes while simultaneously patch-clamping the cells in whole-cell mode. Ca²⁺-sparks were seen as local increases of [Ca²⁺]_i from the resting level of about 100 nM to a peak level of about 300 nM and were imaged using line-scan image acquisition at rates between 2–6 ms per line (13, 14, 17, 18). The time-to-peak [Ca²⁺]_i for the observed Ca²⁺-sparks was similar to those reported earlier (10 ms), and the half-time of decay was also about the same (20 ms) as was the size of the Ca²⁺-spark at the half-peak (about 2 μm).

Data Analysis. Ca²⁺ current.

The amplitude of I_Ca,L was measured as the difference between the peak inward current and the current at the end of the depolarizing pulse. Current density, pA/pF, of I_Ca,L of single ventricular myocytes was calculated by dividing the amplitude of the peak current by the cell membrane capacitance. The relative current of I_Ca,L after the treatment of PUFAs was calculated as I_(PUFAs)/I_(Control) from the same cell. Data were analyzed by ANOVA or by paired or unpaired Student’s t test. P < 0.05 was considered as statistically significant difference. All data are presented as mean ± SEM.

Ca²⁺-sparks.

Ca²⁺-sparks were identified by their size, elevation of [Ca²⁺]_i above the background signal, and kinetics (see refs. 13, 15, and 17). All image processing used IDL software (Research System, Boulder, CO).

RESULTS

Voltage-Gated L-Type Ca²⁺ Currents, I_Ca,L, Are Suppressed by EPA The actions of EPA on I_Ca,L were examined in adult rat ventricular myocytes (see Fig.

1) and in cultured neonatal rat ventricular myocytes. In adult ventricular myocytes, the following voltage protocol was applied. From a holding potential at the normal resting potential of −80 mV (to maintain normal intracellular [Ca²⁺]_i), a slow ramp depolarization to −60 mV was applied to inactivate I_Na, followed by test depolarizations of 200 ms to elicit I_Ca,L. The application of EPA (1.5 μM) reduced I_Ca,L density to about half of its control level, and during the washout partial recovery was achieved. While the shape of the current–voltage relationship was not obviously changed by EPA nor was the apparent reversal potential (around +60 mV), a small (3.33 ± 0.4 mV) statistically significant negative shift of the I_Ca,L inactivation curve was observed. This shift was reversed by the washout of EPA.

EPA application to cultured neonatal rat ventricular myocytes at 1.0 μM concentration produced results similar to those seen in Fig. 1 despite the established differences in adult and neonatal cardiac myocyte morphology and biochemistry. Compared with the brick-like shape of the adult rat ventricular myocytes, cultured neonatal rat ventricular myocytes are thin, tentacled round and unstriated cells with reduced volume and no transverse tubules. Despite the smaller surface area, I_Ca,L density is greater (19). As is well-established in the literature, the resting potential in the neonatal ventricular myocyte is more positive and the cells are spontaneously active. The spontaneous activity depends largely on the interactions of inward I_Na and I_Ca,L and outward currents (dominated by several K currents) with little or no “pacemaker current,” I_f. The average capacitance of neonatal rat single ventricular myocytes was 51.8 ± 2.7 pF (n = 51), which is consistent with our previous report (10). The current density of peak I_Ca,L elicited by single-step pulse from −40 to 0 mV was 6.6 ± 0.4 pA/pF (n = 51) in the neonatal cardiomyocytes bathed in 2 mM Ca²⁺ Tyrode solution. The current-voltage relation and the steady-state inactivation of I_Ca,L in adult rat ventricular myocytes were the same as those observed in neonatal rat ventricular myocytes.

Figure 1.

Suppressant effect of EPA on the voltage-activated L-type Ca²⁺ current in adult rat ventricular myocytes. (A) Superimposed current density traces were elicited by 200-ms pulses with 10-mV increments from a potential of −50 mV to 50 mV. From a holding potential of −80 mV, brief depolarizations were applied to maintain a constant intracellular Ca²⁺ load. To elicit a test pulse, a slow ramp from −80 to −60 mV was applied, and the membrane potential was held at −60 mV for 50 ms before the test depolarization was applied. Current records under control conditions (open circles), 1.5 μM EPA (filled circles), and re-control (crosses) are shown (Upper) from a typical cell while averaged normalized current density-voltage relationships are shown (Lower) for control (n = 11), 1.5 μM EPA (n = 11) and washout (n = 5) are shown. The current density values were normalized to the maximum level observed under control conditions for each experiment. (B) Effects of EPA on the steady-state inactivation curve for I_Ca,L adult rat ventricular myocytes. Holding potential of −80 mV with prepulses to maintain constant SR Ca²⁺ load as in A. A slow depolarizing ramp was applied from −80 to −60 mV to inactivate I_Na. The membrane potential was immediately changed to the test potential for 1 s to produce a steady-state inactivation of I_Ca before stepping to 0 mV for 200 ms to assess I_Ca. (Left) Superimposed original Ca²⁺ current density records for cells exposed to no EPA (control, open circles) and for cells exposed to 10 μM EPA (closed circles). (Right) Normalized steady-state inactivation of peak I_Ca showing control (open circles) with fitting parameters of V_0.5 = −31.97 ± 1.1 mV and K = 5.2 ± 0.26 (n = 11), 10 μM EPA (filled circles) with fitting parameters of V_0.5 = −34.53 ± 1.1 mV and K = 5.11 ± 0.15 (n = 11) and washout (crosses) with fitting parameters of V_0.5 = −31.86 ± 0.9 mV and K = 4.9 ± 0.13 (n = 5). Control and EPA curves are significantly different (P < 0.05).

The time course of EPA-induced suppression of I_Ca,L in cultured neonatal rat ventricular myocytes with various EPA concentrations ranging from 0.1 μM to 40 μM is rapid, occurring in seconds. The time-course of the washout is also rapid. Fig. 2 summarizes the concentration dependence of the EPA-induced suppression of I_Ca,L in cultured neonatal rat ventricular myocytes. The concentration of EPA to produce 50% inhibition of I_Ca,L was 0.8 μM in these neonatal cells and 2.1 in adult cardiac myocytes. In neonatal heart cells I_Ca,L was almost completely inhibited when the concentration of EPA was above 5 μM.

Figure 2.

Inhibition of I_Ca,L by EPA in cultured neonatal rat ventricular myocytes. (A) Time course of the inhibitory effect of EPA on the peak I_Ca,L. The dotted short lines marked with 0.1, 5, and 40 represent the periods of extracellular puffing 0.1, 5, and 40 μM EPA. C, control; EPA, application of EPA; W, washout of EPA with the bath solution alone. (B) The concentration-dependent curve of EPA suppression of I_Ca,L. The number in parentheses represents individual ventricular myocytes treated with various concentrations of EPA. The IC₅₀ of EPA is 0.8 μM.

Effect of Holding Potential on Inhibition of I_Ca,L by EPA in Na⁺-Free Medium.

A series of experiments was done with the holding potential at either −40 mV or −80 mV in cultured neonatal rat ventricular myocytes. In these experiments 137 mM Na⁺ in the bath solution was replaced with 120 mM N-methyl-d-glucamine to block current through the sodium channel from contaminating I_Ca,L. Changing the holding potential from −40 to −80 mV did not alter the current-voltage relation I_Ca,L in the Na⁺-free bath solution. Compared with the amplitude of peak I_Ca,L evoked by single-step pulses from −40 to 0 mV, I_Ca,L elicited by pulses from −80 to 0 mV was insignificantly enhanced by 5.4 ± 3.5% (n = 6, P > 0.05). Extracellular application of 1 μM EPA produced a significantly greater suppression of I_Ca,L when elicited from the holding potential −40 to 0 mV than from −80 to 0 mV (53 ± 6% compared with 40 ± 5% inhibition, respectively; n = 6, P < 0.05). As changing the holding potential did affect the results, the EPA-induced suppression of I_Ca is voltage-dependent in cultured neonatal ventricular myocytes (Fig. 3A).

Figure 3.

Voltage- and time-dependent suppression of Ca²⁺ currents by EPA in cultured neonatal rat ventricular myocytes. (A) x axis represents the holding potentials. y axis represents the relative peak Ca²⁺ currents evoked by depolarizing pulses from −40 or −80 mV to 0 mV. (B) Time-dependent inhibition of Ca²⁺ currents in cultured neonatal rat ventricular myocytes. Ca²⁺ currents were elicited by a train of depolarizing pulses from a holding potential of −40 mV to 0 mV after a 1- to 2-min rest. Relative current was calculated by I_Ca,L(EPA)/I_{Ca,L(control)} recorded from the same myocyte. ▵, control, 1 Hz, n = 4; ○, 1 μM EPA, 1 Hz, n = 4; □, 1 μM EPA, 0.2 Hz, n = 3; ▪, 1 μM EPA, 0.1 Hz, n = 5.

Time-Course of the Inhibitory Actions of EPA on I_Ca,L.

To test whether the EPA-induced suppressant effect of I_Ca,L takes time to develop, as we observed previously in the study of the blocking effect of EPA on I_Na (10), we examined I_Ca,L after the addition of EPA. Additionally, we examined the effect of changing the frequency of depolarizing pulses on the action of EPA to suppress I_Ca,L. Ca²⁺ currents were elicited by a train of depolarizing pulses from a holding potential of −40 to 0 mV starting at time 0, after a 1- to 2-min rest. Fig. 3B shows that the peak I_Ca,L was decreased by 9 ± 10% (n = 4, P > 0.05) of the control after a 90-pulse train stimulation at 1 Hz. Application of a train of depolarizing pulses at frequencies of 0.2 or 0.1 Hz also had no significant effects on I_Ca,L in cultured neonatal cardiomyocytes (data not shown). Extracellular puffing 1 μM EPA produced a maximal 50 to 55% suppression of I_Ca,L in neonatal rat ventricular myocytes with a train stimulation at 0.1 to 1 Hz. The time required to reach the same maximal level of inhibition of I_Ca,L was well superimposed after application of EPA (Fig. 3B). Thus, the suppression of I_Ca,L by EPA developed over time but was not frequency- or use-dependent.

Effects of Other Unsaturated and Saturated Fatty Acids on I_Ca,L in Cultured Neonatal Rat Ventricular Myocytes.

Table 1 summarizes the effects of other fatty acids, such as docosahexaenoic acid, eicosatetraynoic acid, linolenic acid, linoleic acid, oleic acid, and stearic acid on I_Ca,L of cultured neonatal rat ventricular myocytes. Docosahexaenoic acid, linolenic acid, AA, and linoleic acid at 5 μM concentration produced a significantly inhibitory effect on I_Ca,L, which is similar to the effect of EPA. Eicosatetraynoic acid (5 μM) significantly inhibited I_Ca,L, but its potency on I_Ca,L was much less (P < 0.01) than that of EPA. All these PUFAs had similar effects on the steady-state inactivation of the calcium channel as EPA, which caused 3 to 5 mV shift to negative potentials at the V_1/2 point. In contrast, extracellular application of 5 μM of oleic acid, a monounsaturated fatty acid or the saturated fatty acid, stearic acid, had neither a significant effect on I_Ca,L nor on the steady-state inactivation of the channel. It is interesting that a conjugated linoleic acid had a similar effect on I_Ca,L, but its potency was significantly less than that of EPA (Table 1).

Table 1.

Comparison of the suppressant effects of fatty acids on Ca²⁺ channels in cultured neonatal rat ventricular myocytes

Fatty acid	n	μM	% inhibition	P
EPA	5	5	83  ±  6	<0.01
DHA	6	5	62  ±  6	<0.01
LNA	5	5	77  ±  8	<0.01
AA	6	5	76  ±  7	<0.05
LA	6	5	76  ±  7	<0.01
CLA	5	5	44  ±  4	<0.01
ETYA	5	5	45  ±  6	<0.01
SA	7	5	13  ±  6	>0.05
OA	6	5	13  ±  10	>0.05
RA	6	10	55  ±  7	<0.05

Values represent the mean percent inhibition of I_Ca ± SEM. Currents were elicited by depolarizing pulses from a holding potential of −40 mV to 0 mV every 5 s. Statistical significance was tested between the peak amplitudes of I_Ca recorded in the absence and presence of fatty acid. n, the number of individual myocytes exposed to each fatty acid. DHA, docosahexaenoic acid; LNA, linolenic acid; LA, linoleic acid; CLA, conjugated linoleic acid; ETYA, eicosatetraynoic acid; SA, stearic acid; OA, oleic acid; RA, all-trans-retinoic acid. 

It has been shown that all-trans-retinoic acid has protective effects against cardiac arrhythmias induced by isoproterenol, lysophosphatidylcholine, or ischemia and reperfusion (20). In the present study we evaluated the possibility that all-trans-retinoic acid has the same effects on I_Ca,L as other PUFAs. The observed inhibitory action on I_Ca,L may contribute importantly to its antiarrhythmic effect. Table 1 indicates that although less potent than EPA (P < 0.01), the amplitude of peak I_Ca,L was significantly reduced after exposure of the myocytes to all-trans-retinoic acid.

Suppression of I_Ca,L by PUFAs in Adult Rat Ventricular Myocytes.

We examined the effects of EPA, AA, and stearic acid on I_Ca,L of left ventricular myocytes freshly isolated enzymatically from adult rat hearts. The relative potency of EPA and AA to reduce I_Ca,L in adult cardiac ventricular myocytes is shown in Fig. 4. Extracellular application of 1 or 5 μM EPA suppressed I_Ca,L to 57 ± 6% (P < 0.01, n = 5) or to 47 ± 5% (P < 0.01, n = 8) of the control level, respectively. The concentration of EPA to produce 50% inhibition of I_Ca,L in the adult rat ventricular myocytes was 2.1 μM, which is 2.6-fold higher than that in cultured neonatal rat cardiomyocytes (0.8 μM). Again, the saturated fatty acid, stearic acid, had no significant effects on I_Ca,L in adult rat ventricular myocytes.

Figure 4.

Effects of fatty acids on I_Ca in adult rat ventricular myocytes. I_Ca was elicited by 100-ms pulses of single voltage-step from a holding potential of −40 mV to 0 mV. Relative current was calculated by I _Ca,L(EPA)/I _{Ca,L(control)} recorded from the same myocyte. Each fatty acid was added to the bath solution. SA, stearic acid. ∗∗, P < 0.01; vs. the control. The number of ventricular myocytes treated with an individual fatty acid is indicated in parentheses.

Because the reported effects of AA on I_Ca,L of adult cardiac myocytes are conflicting (21–23), we assessed the effects of AA on I_Ca,L in adult rat ventricular myocytes. Extracellular application of 5 μM AA produced 44 ± 3% inhibition of peak I_Ca,L (from 2,142 ± 359 pA of the control to 1,206 ± 219 pA, P < 0.01) in five ventricular myocytes (Fig. 4). In the presence of AA, additional application of 1 μM isoproterenol did not antagonize the suppressant effect of AA. The averaged amplitude of peak I_Ca,L was 1,105 ± 218 pA (n = 5) recorded at 3 to 5 min after exposure of the myocytes to AA and isoproterenol.

Actions of EPA on Ca²⁺-Sparks in Adult Cardiac Myocytes.

Ca²⁺ signaling in adult cardiac myocytes underlies excitation-contraction coupling (19) by a mechanism known as calcium-induced Ca²⁺-release (CICR) (14). When a heart cell depolarizes, the brief opening of the sarcolemmal L-type Ca²⁺ channels permits [Ca²⁺]_i around the channel itself to become quite high. This local [Ca²⁺]_i increase triggers Ca²⁺-release from the SR through the SR Ca²⁺-release channels (also known as ryanodine receptors or RyRs). Thus the stochastic activation of the L-type Ca²⁺ channels by depolarization is responsible for triggering stochastically the RyRs to release Ca²⁺ by “local” CICR (14, 17, 18). The local Ca²⁺-release by a functional unit of excitation-contraction coupling can be observed as a Ca²⁺-spark in heart cells loaded with the fluorescent Ca²⁺-sensitive indicator, fluo-3, and viewed with a confocal microscope (13). When heart cells become “overloaded” with Ca²⁺, they become arrhythmogenic (see refs. 11 and 13) and produce arrhythmogenic I_TI current and waves of elevated [Ca²⁺]_i that propagate within the heart cell (15). Furthermore during Ca²⁺ overload the RyRs become more sensitive to the triggering process, produce an increased number of spontaneous Ca²⁺-sparks, and produce propagating waves of elevated Ca²⁺, all of which can be viewed with the confocal microscope while measuring membrane current. Thus Ca²⁺-sparks and their relationship to I_Ca,L can be used to examine the subcellular links between I_Ca,L, the SR, and cellular Ca²⁺ signaling.

Fig. 5 A and B show that the presence of EPA (1.5 and 15 μM, respectively) markedly reduced the calcium transients induced by I_Ca,L with the higher concentration producing a greater reduction. To examine the subcellular nature of the actions of EPA, we signal-averaged Ca²⁺-sparks from control and EPA-treated single adult cardiac myocytes. The images in Fig. 5B show the line-scan images of the [Ca²⁺]_i transient in the absence (top, ○) and presence (bottom, •) of 15 μM EPA. The reduction in SR Ca²⁺ release is clear from this. By contrast the examination of the individual sparks shown in Fig. 5C reveals no change, i.e., the time-constant of decay (τ) of the calcium sparks are 19.6 ± 1.4 ms (n = 46) for no EPA and 18.3 ± 1.7 ms (n = 33) after EPA application, indicating no significant change in the kinetics of Ca²⁺ sparks. There was also no significant difference in spatial spread of the Ca²⁺-sparks (see Fig. 5C). The decrease in the [Ca²⁺]_i transient could, however, also reflect the efficacy of coupling between the L-type Ca²⁺ channels in the sarcolemma and the RyRs in the SR. We investigate this by examining the number of Ca²⁺-sparks per unit I_Ca,L (see e.g., ref. 17) as is shown in Fig. 5D. The solid curve (control) overlies the dashed curve (EPA) indicating that EPA reduces SR Ca²⁺-release only by decreasing the I_Ca,L that triggers SR Ca²⁺ release.

Figure 5.

Effects of EPA on [Ca²⁺]_i transients and Ca²⁺-sparks in adult rat ventricular myocytes. Voltage protocol as in Fig. 1. (A) [Ca²⁺]_i transient measured as peak fluorescence (F) at the test potential compared with fluorescence at −80 mV (F_o), F/F_o. Control (open circles, solid line); 1.5 μM EPA (filled circles, dashed line). (B) Same as A but 15 μM EPA. (Right) Line-scan image of a [Ca²⁺]_i transient under control conditions at 0 mV (Upper) and after EPA (Lower). Individual Ca²⁺-sparks can be visualized readily after the application of EPA. (C) Signal-averaged Ca²⁺-spark under control conditions (Upper) τ =19.6 ± 1.4 ms (n = 46) and after the application of 15 μM EPA (Lower) τ =18.3 ± 1.7 ms (n = 33). (D) Ca²⁺-spark number normalized by the magnitude of I_Ca (EC coupling “gain function”) under control conditions, solid line (n = 9) and after 15 μM EPA, dashed line (n = 10). Significance for A and B: ∗∗, P < 0.01; ∗, P < 0.05.

DISCUSSION

The finding that only PUFAs with a free carboxyl group blocked the Na⁺ channels in isolated neonatal rat cardiac myocytes (10) suggested that they might be interacting with the positively charged amino acids in the S4 voltage sensor of the α-subunit of Na⁺ channels (24). Because this peptide segment is largely conserved in Ca²⁺ channels (24), it seemed likely that the PUFAs also may affect Ca²⁺ channels.

A reason for this study was to help explain the electrophysiologic effects of the antiarrhythmic PUFAs on cardiac myocytes. Although it had seemed possible to explain the electrophysiologic effects of the PUFAs on the basis of their blocking effects on Na⁺ channels (10), it was important to determine their actions on other ion channels before accepting such a simple unified action. Indeed the blocking effects PUFAs on the outward K⁺ currents, I_to and I_k, by the PUFAs (Y.-F.X., J.P.M., and A.L., unpublished data) are in the wrong direction to explain the observed electrophysiologic effects of the PUFAs (9). Blocking I_to and I_k would both act to prolong rather than shorten the duration of the action potentials. Also the apparent IC₅₀ for inhibition of I_k is 20 μM compared with 4.8 μM, the IC₅₀ for inhibition of I_Na (10). These findings suggest that effects of the PUFAs on K⁺ channels do not affect the basic electrophysiologic actions of the PUFAs on cardiomyocytes. By contrast, the sensitivity of I_Ca,L in ventricular myocytes to the PUFA is some 5 to 6 times greater even than on I_Na, suggesting a major role for inhibition of I_Ca,L in the electrophysiologic and antiarrhythmic actions of the PUFAs. The “vectorial” sum of the effects of PUFA on all the ion channels must account for the electrophysiologic effects observed on the myocytes.

Only the free polyunsaturated fatty acids that have been previously identified to be antiarrhythmic in the cultured spontaneously beating neonatal rat cardiomyocytes (8) affect the Ca²⁺ channels, as they do the Na⁺ channels (10). In addition, conjugated trans-linoleic acid, which has become of interest because of potential antimitotic actions (26) inhibits I_Ca,L. All-trans-retinoic acid, which, like the PUFAs, is specifically antiarrhythmic and shares with the PUFAs their structural requirements for an anti-arrhythmic agent (20), namely a long chain hydrocarbon, two or more C=C double bonds and a free carboxyl group at one end (8), also inhibits I_Ca,L, but less potently than EPA. The inhibitory effect of eicosatetraynoic acid on I_Ca,L, an analog of AA, but which inhibits the cyclooxygenase, lipoxygenase, and epoxygenase enzymes that oxygenated AA to form potent cell messengers (27), indicates that such AA metabolites are not playing a role in the antiarrhythmic effects of the n − 6 PUFAs.

The effects of PUFAs on the Ca²⁺ channels that we have observed appear to be essentially the same in the adult rat ventricular myocytes and in the neonatal rat ventricular myocytes, though the latter is more sensitive to the PUFA. The neonatal myocyte preparation has been important in our investigations because it is useful as a model for certain cardiac arrhythmias. Although this preparation consists of ventricular myocytes, which are not spontaneously active in the adult, the neonatal ventricular myocytes are intrinsically rhythmic in culture. Increased spontaneous rates are observed when I_Ca,L increases or when diverse potassium currents decrease. This sensitivity of the spontaneous rate is similar to the sensitivity of occurrence of “early after-depolarizations,” a kind of arrhythmia used in many model investigations (28).

We have also shown that the action of EPA to reduce I_Ca,L produces a reduction of the [Ca²⁺]_i transient. While there is a reduction of the probability that Ca²⁺-sparks are triggered by a depolarization, the relationship between I_Ca,L and Ca²⁺-sparks is unchanged. Similarly there is no change in the kinetics of the Ca²⁺-spark, the elementary unit of SR Ca²⁺ release. From this we conclude there is no direct action of EPA on the SR Ca²⁺ release process or on SR Ca²⁺ re-uptake (29).

These effects of the free PUFAs on the I_Ca,L of isolated rat myocytes, as is the case with I_Na and I_K, demonstrate that, in the absence of neuronal or hormonal stimuli, the PUFAs have a primary direct effect on membrane ion channel currents.

Kang and A.L. (30) have reported that the anti-arrhythmic PUFAs will prevent or terminate the tachyarrhythmias induced by a β-adrenergic agonist, isoproterenol, or its intracellular messenger, c-AMP. Petit-Jacques and Hartzell (31) recently have shown that the inhibitory effect of AA (C20:4 n − 6) on I_Ca,L was even further down stream of channel phosphorylation by protein kinase A in frog ventricular cells previously stimulated by the β-adrenergic agent isoproterenol. They concluded that the AA inhibits I_Ca,L by a mechanism which involves, in part, stimulation of protein phosphatase activity (29). In this study we show in isolated rat cardiomyocytes, however, that in the absence of β-adrenergic stimulation, PUFAs were still potent inhibitors of I_Ca,L. This is consistent with our hypothesis that the PUFAs actions result from the direct electrical consequence of their effects on ion channel proteins, not on the activity of intracellular messengers.

Pepe et al. (32) confirmed that in adult rat cardiomyocytes an earlier finding that n − 3 PUFAs inhibited the action of dihydropyridine agonists and antagonists on the L-type Ca²⁺ channels of neonatal rat ventricular myocytes (33). But using their whole cell patch clamp recordings, reported no inhibitory effect of the n − 3 PUFAs on the L-type Ca²⁺ currents per se. The finding of an inhibition of L-type Ca²⁺ currents in the present study is more consistent with the earlier report that n − 3 PUFAs modulate the actions of dihydropyridine agonists and antagonists in cardiomyocytes (33). The ability of the n − 3 PUFAs, shown in that study, to displace [³H]nitrendipine from its binding site at the extracellular pole of the calcium channel (34), suggests that the effects of PUFAs on I_Ca,L result from noncovalent binding or interaction with the protein of the calcium channel as found with the voltage-sensitive sodium channel.

ABBREVIATIONS

I_Ca,L

voltage-gated L-type Ca²⁺ current

EPA

eicosapentaenoic acid

AA

arachidonic acid

PUFA

polyunsaturated fatty acid

SFA

saturated fatty acid

SR

sarcoplasmic reticulum