Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Sodium channels in excitable tissues can be modified by lipid-soluble toxins produced by a variety of venomous animals and poisonous plants (Honerjäger, 1982; Hille, 1992; Wang and Wang, 1998). Included within this group are pyrethrins, which are insecticidal esters isolated from Chrysanthemum species, and pyrethroids, their synthetic derivatives. The synthetic compounds are divided into two groups on the basis of differences in chemical structure (Gammon et al., 1981; Valentine, 1990). Type I pyrethroids have no -cyano-3-phenoxybenzyl group and appear to act principally on sensory nerves; the type II pyrethroids have an -cyano-3-phenoxybenzyl group and appear to preferentially affect motor nerves. Pyrethroid modification of insect neuronal Na⁺ channels evokes prominent afterdepolarizations that prolong the action potential duration (Song and Narahashi, 1996; Lee et al., 1999). These actions cause paralysis, leading ultimately to insecticidal activity (Gammon et al., 1981).

It has been calculated that pyrethroids account for roughly 25% of global insecticide sales (Williamson et al., 1996), and environmental exposure to these widely used compounds is quite common. It is also known that pyrethroid toxicity encompasses aquatic vertebrates and mammals that may be accidentally exposed. As in insects, the primary mode of pyrethroid action in vertebrates involves the modification of neuronal Na⁺ channels. During depolarizations, pyrethroid-modified neuronal Na⁺ channels carry a slowly inactivating Na⁺ current (I_Na) that increases in amplitude during pyrethroid exposure. Pyrethroid modification also gives rise to a slowly declining inward Na⁺ tail current upon membrane repolarization (Chinn and Narahashi, 1986; deWeille and Leinders, 1989; Holloway et al., 1989; Song and Narahashi, 1996; Vais et al., 2000). In addition, both voltage-dependent activation and inactivation of Na⁺ channels are shifted to hyperpolarized potentials by pyrethroids (Narahashi et al., 1995; Trainer et al., 1997). By these mechanisms, pyrethroid-modified Na⁺ channels underlie the prominent neuronal afterdepolarizations that produce typical toxic effects (Song and Narahashi, 1996; Lee et al., 1999).

The selectivity and relative toxicity of pyrethroids in neuronal tissue from different species has been rather well researched (Valentine, 1990; Narahashi, 1996; Tabarean and Narahashi, 1998; Motomura and Narahashi, 2000), but the potential cardiotoxicity of this group of compounds appears to have been little studied. It is known that brain tissue contains multiple Na⁺ channel isoforms encoded by different genes (e.g., SCN1A-3A) with still other isoforms present in the peripheral nervous system (for review, see Clare et al., 2000). There is also a specific cardiac Na⁺ channel isoform (encoded by SCN5A), which exhibits distinct voltage-dependent kinetic properties and pharmacology (Krafte et al., 1991; Fozzard and Hanck, 1996; Balser, 1999). The cardiac channel has important clinical significance in that its mutations are implicated in malignant cardiac arrhythmias (Wang et al., 1995; Chen et al., 1998). However, despite the importance of this cardiac Na⁺ channel, research into its interactions with pyrethroids has been almost entirely lacking. Studies in whole cardiac tissue showed that pyrethroids are positively inotropic, possibly as a result of depolarization of myocardial sympathetic nerve terminals and the resulting effects of released catecholamines (Forshaw and Bradbury, 1983; Berlin et al., 1984). However, an additional inotropic effect was attributed to Na⁺ loading of the cardiac tissue, possibly resulting from pyrethroid modification of native cardiac Na⁺ channels. In support of this finding are limited data suggesting that single cardiac Na⁺ channels are susceptible to modification by the type II pyrethroid deltamethrin (Grant et al., 1993). However, no detailed electrophysiological characterization of the effects of pyrethroids on macroscopic cardiac Na⁺ current has so-far been carried out. In the present study therefore, we have evaluated the actions of pyrethroids on isolated mammalian ventricular cells and whole hearts. Our results reveal that these agents can prolong action potentials in ventricular myocytes and, under voltage clamp, prolong the duration of I_Na. In spontaneously beating isolated hearts, pyrethroids increased the contractile variability, consistent with considerable arrhythmogenic potential for this class of compounds.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Isolated Rat Ventricular Myocytes. Ventricular myocytes were isolated from hearts of male Wistar rats (~200 g) killed by intraperitoneal injection of pentobarbitone sodium (200 mg/kg). The isolated heart was perfused at 37°C on Langendorff apparatus and myocytes were enzymatically isolated (Spencer et al., 2000). Briefly, this involved 5 min of perfusion with a nominally Ca²⁺-free solution (see below) containing 0.25 mg/ml collagenase (Worthington type II, 215 units/mg), 0.05 mg/ml protease (Pronase), and 0.3 mg/ml bovine serum albumin. Chopped ventricular fragments were gently agitated in perfusion solution for approximately 30 min. Dissociated myocytes were decanted at 5-min intervals and refrigerated at 4°C until use (<8 h).

Electrophysiology of Rat Ventricular Myocytes. All myocytes were whole cell voltage-clamped at 40 mV using an Axopatch 200A amplifier and single patch electrodes with resistances of 2.5 to 5 M. Coupling to pCLAMP software (version 6.0) was used to observe "signature currents" evoked by a voltage-clamp step to 90 mV for 10 ms followed by a linear membrane potential ramp to +70 mV (Spencer et al., 2000). In current-clamp mode, action potentials were evoked by brief suprathreshold current pulses determined individually for each myocyte. All electrophysiological experiments were performed at room temperature (20-25°C) and a basic stimulation frequency of 0.33 Hz was used throughout. Continuous data recordings were made during 3 to 5 min of superfusion with pyrethroids followed by a comparable washout period (unless voltage control was lost). Bath solutions were completely exchanged within 2 min. All signals were initially digitized at 10 kHz and stored on digital audiotape. In off-line analysis, current and voltage were sampled at 1 kHz and low-pass filtered at 500 Hz.

Solutions for Rat Myocytes. Unless stated, all chemicals were purchased from Sigma-Aldrich Chemical Co. (St. Louis, MO). Cell isolation solution contained the following: 135.0 mM NaCl, 5.4 mM KCl, 1.0 mM MgCl₂, 0.3 mM NaH₂PO₄, and 10.0 mM HEPES (neutralized to pH 7.2 with NaOH). For initial removal of blood from the heart and for storage of the myocytes, 0.2 mM Ca²⁺ was added to this solution. Isolated myocytes were superfused with modified Tyrode's solutions containing 145.0 mM NaCl, 4.0 mM KCl, 1.0 mM MgCl₂, 2.0 mM CaCl₂, 10.0 mM D-glucose, and 10.0 mM HEPES (neutralized to pH 7.4 with NaOH). Pyrethroids (Reidel-de Haen, St. Louis, MO) were dissolved in dimethyl sulfoxide to make 50 mM stock solutions. Aliquots (<0.05% by volume) were added to superfusion solutions immediately before use to give a final concentration of 10 µM. The effects of two type I pyrethroids (tefluthrin and tetramethrin) and two type II pyrethroids (fenpropathrin and -cypermethrin) were studied. Readers are referred to the Sigma-Aldrich on-line catalog (www.sigma-aldrich.com) for the full chemical names of these compounds. Cells were dialyzed with pipette solution that had the following composition: 75.0 mM K-glutamate; 30.0 mM K-piperazine-N,N'-bis[2-ethanesulfonic acid] (PIPES), neutralized to pH 7.1 with KOH; 20.0 mM KCl, 0.5 mM MgCl₂, 0.05 mM K₂-EGTA, 10.0 mM Mg-ATP, 5.0 mM Tris-phosphocreatine, 0.1 mM Tris-GTP, and 5.0 mM pyruvic acid.

Isolated Guinea Pig Myocytes. Myocytes were also isolated from male guinea pigs (~400 g). Hearts were digested using collagenase and protease enzymes in an enzymatic dispersion method broadly similar to that described above (Levi and Issberner, 1996). The guinea pig cells were stored at room temperature in solution containing 1 mM Ca²⁺ until use. Cells remained viable for up to 8 h after isolation.

Guinea Pig Myocyte Electrophysiology. Isolated ventricular myocytes were superfused at 37°C in ramp experiments; for selective I_Na recordings, experiments were conducted at room temperature (20-22°C). Patch pipettes (Corning 7052 glass; A-M Systems, Everett, WA) were pulled (P-87; Sutter Instrument Co., Novato, CA) and polished (Narishige MF-83 microforge) to resistances of between 2 and 3 M for voltage-ramp experiments. Whole cell voltage-clamp recordings were made using an Axopatch 200B amplifier and cell capacitance was measured by either analyzing the charging transients elicited by a 5-mV voltage step or reading the capacitance value from the dial on the amplifier after compensating for series resistance and cell capacitance. These methods have been shown previously to give similar values for cell capacitance (Hancox et al., 1993). Series resistance values were in the range of 2 to 4 M. Typically, 75 to 80% of series resistance could be compensated.

Solutions for Guinea Pig Myocytes. Myocytes were initially superfused with modified Tyrode's solution containing 140.0 mM NaCl, 4.0 mM KCl, 2.5 mM CaCl₂, 1.0 mM MgCl₂, 10.0 mM D-glucose, and 5 mM HEPES (adjusted to pH 7.45 with NaOH). Signature currents were evoked as described above using a K⁺-based pipette (internal) solution, which contained 10.0 mM NaCl, 113.0 mM KCl, 0.4 mM MgCl₂, 5.0 mM K₂ATP, 5.0 mM D-glucose, 10.0 mM HEPES, and 5.0 mM BAPTA (adjusted to pH 7.2 with KOH). For quantitation of I_Na, the pipette solution contained 130.0 mM CsCl, 10.0 mM NaCl, 0.4 mM MgCl₂, 5.0 mM Mg-ATP, 5.0 mM glucose, 10.0 mM HEPES, and 5.0 mM BAPTA (adjusted to pH 7.3 with CsOH). The external solution for these experiments contained 130.0 mM CsCl, 10.0 mM NaCl, 1.2 mM MgCl₂, 1.0 CaCl₂ mM, 11.0 mM D-glucose, and 20.0 mM HEPES (adjusted to pH 7.3 with CsOH). Pyrethroids were treated as described earlier. TTX (Tocris, Bristol, UK) was dissolved in deionized water (3 mM) and added directly to superfusion solutions as necessary to give a final concentration of 30 µM.

Data Analysis and Statistics. Voltage- and current-clamp protocols were generated using the program Winwcp (version 1.7a; written and supplied by John Dempster, Strathclyde University, Glasgow, UK) via a Digidata 1200B interface (Axon Instruments, Foster City, CA). Data were recorded on-line at 2 kHz, except I_Na data for which a recording frequency of 10 kHz was used. Data were stored on the hard disk of an IBM compatible PC, and analyzed using Winwcp. Figures were constructed using FigP (Biosoft, Cambridge, UK), and statistical analysis performed using EXCEL (Microsoft, Redmond, WA).

Isolated Perfused Rat Hearts. A limited series of experiments was performed to examine the effects of pyrethroids on Langendorff-perfused isolated rat heart preparations. In these experiments, the hearts were retrogradely perfused at 38°C with a modified Krebs-Henseleit solution, equilibrated with O₂, CO₂ (95%, 5%), containing 118.5 mM NaCl, 25.0 mM NaHCO₃, 1.2 mM MgSO₄, 1.2 mM KH₂PO₄, 3.0 mM KCl, 2.5 mM CaCl₂, 11.1 mM D-glucose, pH 7.4. Pyrethroids were added to this perfusion solution at 10 µM as described for the myocyte experiments. Control experiments were also performed in which 10 µM propranolol was included in the perfusion solution to eliminate catecholamine release from myocardial sympathetic terminals. In all experiments, each heart was allowed to stabilize for a 10-min period before pyrethroid perfusion was started. Ventricular contractions, recorded via a strain gauge, were digitized at 400 Hz using proprietary software (Chart; AD Instruments, NSW, Australia). Peak contractile force was output in ASCII files and analyzed using in-house software (Borg et al., 2001).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Rat Ventricular Myocytes. Initial characterization of pyrethroid effects on rat ventricular cells was achieved by examining the profiles of action potentials (APs) before and after adding pyrethroids (at 10 µM) to the superfusion solution. Each panel of Fig. 1 shows two examples of superimposed APs for each of four pyrethroids studied. Single APs recorded under control conditions and in the presence of fenpropathrin (Fig. 1a), -cypermethrin (Fig. 1b), tefluthrin (Fig. 1c), or tetramethrin (Fig. 1d) are compared. With the exception of tetramethrin, all pyrethroids markedly prolonged the late phase of AP repolarization. APs often had a secondary upward voltage deflection after the initial rapid repolarization phase (Fig. 1b, right; and c). In addition, early afterdepolarizations frequently occurred (Fig. 1a, left; b; and c, right). The timing of single or multiple afterdepolarizations with respect to the AP upstroke appeared to vary randomly. Consistent changes in resting membrane potential were absent, and occasional APs without afterdepolarizations were still observed during pyrethroid superfusion.

View larger version (14K): [in this window] [in a new window]   Fig. 1.   Effects of superfused pyrethroids on rat ventricular action potential morphology. Superimposed representative APs recorded in control, prior to superfusion with pyrethroids, and in the presence of fenpropathrin (a), -cypermethrin (b), tefluthrin (c), and tetramethrin (d). Results from separate individual cells are shown in left and right panels. In all panels, APs recorded in the presence of the pyrethroids are indicated with asterisks (*). Note the variation in control AP morphology from cell to cell, but that this variation had little influence over the AP changes attributable to the effects of pyrethroids.

View larger version (19K): [in this window] [in a new window]   Fig. 2.   Plots showing the time course of the effects of pyrethroids on action potential duration. Data from representative rat ventricular cells are presented. Action potential duration is expressed as APD90 (see text). a, APD90 after adding 10 µM fenpropathrin to the superfusion solution, at time = 0, monitored for 5 min followed by 2 min of washout as indicated above the plot; b-d, identical plots to a for the other pyrethroids as indicated.

View larger version (14K): [in this window] [in a new window]   Fig. 3.   Signature currents recorded during 500-ms linear voltage ramps from 90 to +70 mV, before and during superfusion with pyrethroids. In each panel, the control response prior to pyrethroid superfusion is displayed on the left. On the right is a typical signature current obtained after 3 or more minutes of superfusion with the compound indicated above. Asterisks indicate that pyrethroids prolonged the component previously identified as INa (Spencer et al., 2000).

Guinea Pig Ventricular Myocytes. To elucidate the effects of pyrethroids on cardiac I_Na, a series of experiments was performed to quantify the voltage- and time-dependent properties of this current. Selective recordings of I_Na were made using conventional "square pulse" voltage commands using recording conditions recently validated for guinea pig ventricular myocytes (Yuill et al., 2000). Selective I_Na measurements in the present study therefore focused on myocytes isolated from this species. Initial experiments used voltage ramps (at velocities between 0.32 and 0.8 mV/ms) to obtain signature currents in guinea pig cells for comparison with the results from rat myocytes presented above. In these experiments, performed at 37°C to confirm that pyrethroids retained their efficacy at mammalian body temperature, standard external solution was used and then one of three pyrethroid agents (fenpropathrin, tefluthrin, or tetramethrin) was added. For each condition, net currents were measured and then 30 µM TTX was applied. This allowed us to measure the TTX-sensitive (TTX-S) current during the ramp before and during exposure to each compound. Sample data for fenpropathrin are shown in Fig. 4. Similar to our observations from rat myocytes, I_Na was visible as a large and fast downward deflection during the ramp (Fig. 4a), which was blocked by 30 µM TTX (Fig. 4b). In the presence of 10 µM fenpropathrin, this current component became wider (Fig. 4c), while retaining its sensitivity to TTX (Fig. 4d). Recordings displayed on a faster time scale show the TTX-S current for standard extracellular solution and for fenpropathrin (Fig. 5a). TTX-S current was prolonged by fenpropathrin. Similar results were observed in nine cells to which ramp protocols were applied. In 10 other cells, 10 µM tefluthrin also prolonged the TTX-S current component during applied voltage ramps (Fig. 5b), while in five cells, tetramethrin exerted little effect (Fig. 5c). In summary therefore, pyrethroids produced qualitatively similar effects on I_Na in guinea pig and rat ventricular myocytes.

View larger version (12K): [in this window] [in a new window]   Fig. 4.   Effects of 10 µM fenpropathrin on guinea pig whole cell signature current during an applied voltage ramp. a, current under control conditions, elicited from a holding potential of 50 mV by a ramp from 100 mV to +60 mV. b, 30 µM TTX abolished the sodium current component. c, application of 10 µM fenpropathrin prolonged the sodium current component during the ramp. d, shows that 30 µM TTX also inhibited the sodium current in the presence of fenpropathrin. The time scale bar refers to all panels.

View larger version (13K): [in this window] [in a new window]   Fig. 5.   Effects of fenpropathrin, tefluthrin, and tetramethrin on TTX-sensitive currents elicited by voltage ramps. a, left, shows a TTX-sensitive current under control conditions, elicited by a voltage-ramp protocol. Fenpropathrin (10 µM) prolonged the TTX-sensitive current (right). b, 10 µM tefluthrin also prolonged the TTX-sensitive current. c, left, shows a TTX-sensitive current elicited under control conditions. Tetramethrin (10 µM) had comparatively little effect on the current (c, right). Time scale bar refers to all panels.

(1)

View larger version (15K): [in this window] [in a new window]   Fig. 6.   Effect of fenpropathrin, tefluthrin, and tetramethrin on sodium currents recorded selectively. a, left, shows a rapidly activating and inactivating, TTX-sensitive INa, elicited under control conditions by a step depolarization (top) from 140 to 30 mV. a, right, current amplitude shows a slight increase and a slowing of inactivation time course after application of 10 µM fenpropathrin. b, left, shows a TTX-sensitive INa under control conditions. b, right, 10 µM tefluthrin produced an increase in current amplitude and slowed inactivation time course. Tetramethrin (10 µM) had no visible effect on INa (c, left and right). Note, time bases shown in the top panels (voltage protocols) also refer to the bottom panels.

(2)

View larger version (20K): [in this window] [in a new window]   Fig. 7.   Effect of fenpropathrin on the current-voltage relationship of INa recorded selectively. a, families of INa under control conditions, elicited by step depolarizations (below) from a prepulse potential of 140 mV to a range of more positive test potentials, and b, in the presence of 10 µM fenpropathrin. c, the mean I-V relationships for peak INa (open symbols: control, data fitted by eq. 2 using V0.5 = 37.55 ± 0.33 mV, k = 5.14 ± 0.29 mV; closed symbols: fenpropathrin, V0.5 = 43.67 ± 0.18 mV, k = 6.57 ± 0.15 mV). d, the fractional increase in peak INa after application of fenpropathrin plotted against test potential. Although the largest increase was observed at 60 mV, this could not be plotted because control INa was close to zero, making the proportionate increase in current with fenpropathrin extremely high.

(3)

View larger version (18K): [in this window] [in a new window]   Fig. 8.   Effect of fenpropathrin on the voltage dependence of INa inactivation. a, left, INa records elicited under control conditions, by step depolarizations to 30 mV from a range of more negative prepulse potentials (b). a, right, the effect of 10 µM fenpropathrin. c, the inactivation curve under control conditions (data fitted by eq. 3; , V0.5 = 92.07 ± 0.33 mV, k = 7.94 ± 0.29) and after application of 10 µM fenpropathrin (, V0.5 = 97.11 ± 0.40 mV, k = 7.23 ± 0.35).

Pyrethroids Affect the Cardiac I_Na Window. The V_0.5 and k values obtained from the I-V fits (Fig. 7c) and inactivation curves (Fig. 8c) for I_Na were used to estimate the I_Na window as shown in Fig. 9. The inactivation parameters were simulated at 2-mV intervals using eq. 3 and activation parameters were simulated using the following equation:

(4)

where "activation variable" at any test potential (V_m) occurs within the range 0 to 1; V_0.5 and k have similar meanings to those in eq. 2. For the control situation, the small area of overlap in Fig. 9 denotes the I_Na window region: a potential range over which a small persistent Na⁺ entry would be anticipated. In the simulated presence of fenpropathrin, in which both activation and inactivation curves were shifted 5 mV to negative potentials, the I_Na window was larger than under control conditions. This suggests that over the window voltage range, steady-state Na⁺ entry via I_Na could have been increased by fenpropathrin. Due to alterations in the time and voltage dependence of I_Na, it is therefore likely that an increased window current contributes to toxic effects in the presence of pyrethroids. In total, these changes in I_Na appear to be involved in AP prolongation and the generation of afterdepolarizations in cardiac cells.

View larger version (13K): [in this window] [in a new window]   Fig. 9.   Effects of fenpropathrin on the simulated INa window. Experimental data from Figs. 7 and 8 were used together with eqs. 3 and 4 (see Results) to calculate activation and inactivation variables at 2-mV intervals between 140 mV and +40 mV. At each voltage (Vm), control activation variable = 1/(1 + exp [(37.55  Vm)/5.14]), fenpropathrin activation variable = 1/(1 + exp [(43.67  Vm)/6.57 ]), control inactivation variable = 1  (1/(1 + exp [(92.07  Vm)/7.94])), fenpropathrin inactivation variable = 1  (1/(1 + exp [(97.11  Vm)/7.23])). The resulting simulated activation and inactivation curves were then overlaid and the area of overlap selected and shown at a high magnification to illustrate the INa window in control () and in fenpropathrin (). The potential of the peak of the INa window shifted to a more negative potential in the presence of the pyrethroid and the integrated window current increased by ~43%.

Isolated Perfused Rat Hearts. To examine the possible effects of circulating pyrethroids at the whole heart level, we determined the force of contraction (FOC) of isolated perfused rat hearts beating spontaneously. The resting heart rate of these preparations was 325 ± 10 beats per min (n = 22). A continuous recording from a one such heart is shown in Fig. 10a. Data were recorded for an approximately 10-min control period followed by an equivalent period of perfusion by 10 µM tefluthrin. Peak FOC, displayed in arbitrary units, was fairly stable throughout the control period. A switching artifact, appearing as a transient dip in FOC, invariably accompanied the exchange of control and pyrethroid-containing perfusion solutions. In the example shown in Fig. 10a, after approximately 5 min of pyrethroid perfusion, noise in the peak FOC trace dramatically increased. This was found to be attributable to periodical episodes of intense variation in contractile amplitude. Figure 10b shows individual contractions during one of these episodes. The variations in contractile amplitude were associated with irregularities in the intervals between heartbeats. Periods in which beats were absent became randomly interspersed in the recordings, and therefore, it appears that periods of contractile variability involved episodic cardiac asystoly. However, no significant differences in overall heart rate were observed between control and pyrethroid perfusion periods for any of the compounds studied. The appearance of irregular activity during perfusion with pyrethroids suggests that these compounds have considerable proarrhythmic potential.

View larger version (30K): [in this window] [in a new window]   Fig. 10.   Effects of pyrethroids on contractility in isolated Langendorff-perfused rat heart preparations. a, continuous plot derived from a recording of peak FOC before and during perfusion of a representative preparation with 10 µM tefluthrin. The composition of the perfusion solution is indicated above the double-headed arrows. A dip in FOC after changing solutions was probably an artifact of the fact that the pyrethroid-containing solution had been static in the perfusion line prior to switching. b, individual contractions of the same unstimulated preparation recorded during the control period (left) and during pyrethroid perfusion (right). c, histograms showing mean coefficients of variation in contractile amplitude during control and pyrethroid perfusions for each of the compounds indicated below the horizontal axis. Asterisks indicate statistically significant differences compared with control values.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The principal findings of the present study at the cellular level are that the type I pyrethroid tefluthrin and the type II pyrethroids fenpropathrin and -cypermethrin 1) prolonged ventricular action potentials and evoked afterdepolarizations; 2) modified the time course of I_Na by altering the relative proportions of fast and slowly inactivating current; and 3) altered the voltage dependence of I_Na. At the whole heart level, these effects corresponded with a pyrethroid-induced increase in the variability of contractile force, suggestive of proarrhythmic activity. Several aspects of these findings merit detailed consideration.

Na⁺ Channel Targets of Pyrethroid Modification. Previous molecular studies have established conclusively that the insecticidal activity of pyrethroids is uniquely determined by the locus coding for the para neuronal Na⁺ channel (Miyazaki et al., 1996; Williamson et al., 1996; Lee et al., 1999). Indeed, point mutations in para Na⁺ channels suffice to confer pyrethroid resistance (and cross-resistance to DDT). In both vertebrate and invertebrate neurons, Na⁺ channel modification by pyrethroids produces a steady-state Na⁺ current and a large, slowly declining tail current after repolarization. These effects appear to be attributable to the suppression of both voltage-dependent inactivation and deactivation of Na⁺ channels (Vais et al., 2000). The steady-state current that is produced appears to underlie the prominent neuronal afterdepolarizations that characterize pyrethroid poisoning.

Modifiers of Inactivation in Cardiac Na⁺ Channels. Fast inactivation gating in mammalian Na⁺ channels is particularly prone to interference by various lethal toxins and mutations. Hydrophobic toxins such as veratridine or batrachotoxin bind via the channel pore and, as well as negatively shifting the threshold voltage for I_Na activation by up to 50 mV, eliminate almost all fast inactivation (Honerjäger, 1982; Wang and Wang, 1998, 1999; Chattou et al., 2000). Neuronal Na⁺ channels also appear to be stabilized in the open state by pyrethroids that eliminate fast inactivation (Chinn and Narahashi, 1986; deWeille and Leinders, 1989; Motomura and Narahashi, 2000). The present findings indicate a subtler mode of pyrethroid action for the agents we applied to cardiac cells. The partitioning of rapidly inactivating I_Na into components described by fast and slow time constants (Yuill et al., 2000) was clearly altered in our experiments by both fenpropathrin and tefluthrin. During superfusion with control solutions, more than 95% of fast-inactivating current was described by an 1-ms time constant (), the remaining 5% by  of 10 ms. The proportion of current described by the faster time constant fell to about 64% in the presence of 10 µM fenpropathrin, and as low as 35% in the presence of 10 µM tefluthrin or 50 µM fenpropathrin. Although the origin of these two phases of inactivation is at present unresolved, they probably reflect pyrethroid-sensitive conformational influences on charge movements within the channel protein.

Na⁺ Current and Action Potential Prolongation in Heart. Toxins that prolong the open state of cardiac sodium channels elicit both AP prolongation and afterdepolarizations (Honerjäger, 1982). The underlying mechanisms appear to be distinct for each, but ultimately stem from imbalances in the normal sequence of inward and outward currents that maintain the AP plateau. An exceedingly delicate balance seems to exist between inward and outward ionic currents during the AP plateau such that the net transmembrane current is close to zero. Toxins and mutations that increase Na⁺ currents appear to upset this balance in favor of prolonged net inward current with concomitant AP prolongation (Bennett et al., 1995; Wang et al., 1995). AP prolongation, in turn, may directly or indirectly provoke Ca²⁺ influx into the cell. First, reactivation of sarcolemmal L-type Ca²⁺ currents may occur during the prolonged AP, directly increasing cytosolic Ca²⁺ concentration (Makielski and January, 1998). Second, cytosolic Na⁺ loading during prolonged I_Na almost certainly leads to increased Ca²⁺ influx via the Na⁺-Ca²⁺ exchanger mechanism. Most lipid-soluble Na⁺ channel toxins appear to be positively inotropic via this mechanism (Honerjäger, 1982; Berlin et al., 1984). Thus, Ca²⁺ overload of the sarcoplasmic reticulum could prime the cell for mistimed Ca²⁺ releases, evoking afterdepolarizations. However, it is worth considering that in the present experiments, prolongation of APD often exceeded 1 s (Fig. 1) yet the duration of I_Na was more modestly prolonged (Fig. 3). Therefore, if this AP prolongation involved Ca²⁺-dependent currents, the underlying SR Ca²⁺ overload should have been indicated by positive inotropic effects. Surprisingly, the contractions of isolated spontaneously beating rat hearts were not increased by pyrethroids.

Mammalian Pyrethroid Toxicity. Pyrethroid poisoning after accidental exposure of humans and domestic animals is characterized mostly by symptoms related to neuronal hyperexcitability (Valentine, 1990; Narahashi et al., 1995; Motomura and Narahashi, 2000). Nevertheless, pyrethroid insecticides are rather selective poisons. Reasons for this selectivity toward insect neurons over vertebrate neurons have been extensively discussed by Narahashi et al. (1995), Song and Narahashi (1996), and Narahashi (1996). First, the temperature coefficient for pyrethroid actions is <1, so efficacy is greater at the relatively cold body temperatures of invertebrates (Motomura and Narahashi, 2000). Second, pyrethroids associate with cell membrane lipids with greater avidity in insects than in mammals, and the greater surface area/volume ratio in small creatures increases the likelihood that the compounds will reach their cellular targets after being ingested (Song and Narahashi, 1996). Thus, due to basic physiology, pyrethroid toxicity is likely to be higher in insects than in mammals. Effective concentration ranges for neurons also differ between mammals and insects. In mammals, K_d values <5 µM have been measured for tetramethrin (type I; Song and Narahashi, 1996) and deltamethrin (type II; Tabarean and Narahashi, 1998). In insects, the sensitivity could be an order of magnitude higher (Vais et al., 2000). The fact that in the present study, 1 µM fenpropathrin affected only slightly the inactivation parameters of I_Na in mammalian cardiac myocytes suggests that cardiac Na⁺ channels may generally display a lesser sensitivity to pyrethroids than those in neurons. Consistent with this, we observed that tetramethrin had low potency against cardiac I_Na, whereas in neuronal cells, this compound has high potency (Gammon et al., 1981; Narahashi et al., 1995; Song and Narahashi, 1996; Motomura and Narahashi, 2000). It appears likely that the lack of cardiac potency of this compound could be related to structural constraints such as stereospecificity.