Nonlinear Changes of Transmembrane Potential During Defibrillation Shocks
Role of Ca2+ Current
Abstract—Defibrillation shocks induce complex nonlinear changes of transmembrane potential (ΔVm). To elucidate the ionic mechanisms of nonlinear ΔVm, we studied the effects of ionic channel blockers on ΔVm in geometrically defined myocyte cultures. Experiments were carried out in cell strands with widths of 0.2 mm (narrow strands) and 0.8 mm (wide strands) produced using a technique of directed cell growth. Uniform-field shocks were applied across strands during the action potential (AP) plateau, and the distribution of shock-induced ΔVm was measured using an optical mapping technique. Nifedipine and 4-aminopyridine were applied to inhibit the L-type calcium current (ICa) and the transient outward current (Ito), respectively. In control conditions, the distribution of ΔVm across cell strands was highly asymmetrical with a large ratio of negative to positive ΔVm (ΔV−m/ΔV+m) measured at the opposite strand borders. Application of nifedipine caused a large increase of ΔV+m and a decrease of ΔV−m/ΔV+m, indicating involvement of ICa in the asymmetrical ΔVm, likely as a result of the outward flow of ICa when Vm exceeded the ICa reversal potential. ΔV−m decreased in the narrow strands but remained unchanged in the wide strands, indicating that the changes of ΔV−m were caused by electrotonic interaction with an area of depolarization. 4-Aminopyridine did not change ΔV−m/ΔV+m. These results provide evidence that (1) the asymmetry of shock-induced ΔVm during the AP plateau is due to outward flow of ICa in the depolarized portions of the strands, (2) Ito is not involved in the mechanism of ΔVm asymmetry, and (3) the effects of drugs on ΔVm are modulated by the tissue geometry.
Electric shocks are routinely used to terminate fibrillation in patients, yet the mechanisms of defibrillation are not well understood. Particularly, it is not yet known how the shock field changes the transmembrane potential of cardiac cells (Vm). Because shocks are applied in the extracellular space, both positive and negative changes of Vm are produced by the shocks. It is believed that Vm changes during shocks can be due to several factors, including nonuniformity of the electric field1 2 3 4 and nonuniformity of the tissue structure.5 6 Combination of these 2 factors can fully predict the magnitude and the distribution pattern of Vm changes in the linear model systems.7 Particularly, ΔVm in a linear system is directly proportional to the strength of the shock field. In real cardiac tissue, however, Vm changes are strongly nonlinear, suggesting the involvement of some additional factors.
Three types of nonlinear ΔVm were reported. First, unlike in linear systems, strong shocks applied during the plateau phase of the action potential (AP) in isolated guinea pig papillary muscles,8 9 in cultured strands of neonatal rat myocytes,10 11 and in isolated guinea pig myocytes12 induced asymmetrical changes with negative ΔVm (ΔV−m) larger than positive ΔVm (ΔV+m). Second, with increasing shock strength the amplitude of ΔVm in cardiac tissue did not increase proportionally but reached a saturation level.8 9 11 Third, large negative ΔVm in cultured cell monolayers11 and in isolated myocytes12 exhibited a nonmonotonic behavior whereby a strong initial hyperpolarization was followed by a positive shift of Vm.
The mechanisms of these nonlinear Vm changes are not known. The ΔVm asymmetry with ΔV−m>ΔV+m reflects an increase of the net outward current. We hypothesized that such ΔVms are related to modulation of activity of ionic channels. During the early plateau phase of AP, the main currents in neonatal rat myocytes are the calcium current (ICa) and the transient outward current (Ito). To understand the role of these currents in the ΔVm asymmetry, we investigated the effect of respective channel blockers on ΔVm. Because Vm changes are dependent on the geometry of cardiac tissue and electric field, experiments were performed with uniform-field shocks in geometrically defined strands of cultured myocytes. Also, because electrotonic interaction between the areas depolarized and hyperpolarized by shocks can affect the Vm changes,11 conditions with both strong and weak electrotonic interaction were studied using narrow (width<λ, the electrotonic space constant) and wide (width>λ) strands, respectively.
Materials and Methods
Ventricular myocytes from 2-day-old Wistar rats were grown in cell strands (Figure 1A⇓) using the technique of directed cell growth.11 13 The widths of the strands (Figure 1B⇓) were 0.2 and 0.8 mm, below or above the space constant (0.35 mm).14 Strands were first incubated in medium M199 (Life Technologies) containing Earl’s salts, 20 mmol/L HEPES, 10% FBS (Sigma), 20 μg/mL vitamin B12, and antibiotics at 37°C in a humidified atmosphere of 5% CO2. Starting on the second day, a serum-free medium, UltraCulture (BioWhittaker), was used. To stimulate cell beating, the medium was supplemented with 2 μmol/L norepinephrine.
Optical Measurements of ΔVm
After 4 to 6 days in culture, monolayers were transferred into a perfusion bath and superfused with Hanks solution (Life Technologies) with pH 7.4 and temperature 35°C to 36°C. Cells were stained for 3 to 6 minutes with 2.5 μmol/L of the Vm-sensitive dye RH-237 (Molecular Probes). Optical measurements were performed at areas without visible anatomic discontinuities in monolayer structure. The dye fluorescence was measured using a 16×16-photodiode array (Hamamatsu) and a microscopic mapping system described previously.11 The bandwidth of the system was 1 kHz. With 10×, 20×, and 40× objectives used, the spatial resolution (center-to-center interdiode distance) was 110, 55, and 27.5 μm, respectively, and the sampling rate was 10 kHz per channel.
The cells were paced at a cycle length of 500 ms. Rectangular uniform-field shocks (duration 10 ms) were applied via 2 large platinum electrodes positioned at the bath ends (Figure 1A⇑). The field strength, E, was measured in the bath using a bipolar silver electrode (wire diameter 0.1 mm, interelectrode distance 1.1 mm). Delivery of shocks was synchronized with stimulation pulses. The delay between AP upstroke and onset of the shocks was 9 to 18 ms. Two channel blockers were applied: ICa inhibitor nifedipine (2 μmol/L) and Ito inhibitor 4-aminopyridine (4-AP; 2 mmol/L). Because nifedipine is photosensitive, a fresh solution was prepared in the dark for each day of use.
Shock-induced ΔVm was measured as the difference between a linear fit of the plateau phase and the magnitude of the shock response 5 ms after shock onset (Figure 1C⇑). The fit was determined using linear regression of the signal during the 5-ms interval before shock application. Comparison with a recording without shock shows that the fitting procedure allowed the reduction of errors in ΔVm associated with changes in fluorescence intensity due to factors other than the shock. The ΔVm was normalized by the AP amplitude (APA). The degree of ΔVm asymmetry was characterized by the ratio of ΔV−m and ΔV+m measured at the opposite strand edges. ΔVms at the edges were spatially averaged including only those signals that had a constant slope before shock application. Data were expressed as mean±SD. Differences were compared using the 2-tailed paired t test. All statistical probability values are expressed for differences from control conditions. Results were considered statistically significant if P<0.05.
To ensure uniformity of cell properties, only those strands were selected for analysis that exhibited no major discontinuities in activation spread and spatial distribution of ΔVm and had conduction velocity >20 cm/s and normalized maximal upstroke rate of rise >80 V/s. Experiments were performed in a total of 37 strands from 26 cell monolayers produced in 7 cell cultures.
Effect of Nifedipine on ΔVm
Figure 2⇓ presents results of a typical experiment in a strand 0.22 mm in width. Figure 2A⇓ compares the Vm traces obtained in control conditions, after 5 minutes of perfusion with 2 μmol/L of nifedipine, and 15 minutes after washout from a horizontal row of photodiodes shown in Figure 2B⇓. The strength of the shock field was 16.4, 16.2, and 16.1 V/cm, respectively. The shocks were applied during the early plateau phase of the AP, ≈12 ms after the AP upstroke. In control conditions, the shock induced depolarization on the left side of the strand and hyperpolarization on the right side with a linear transition in between (Figure 2C⇓). The spatial distribution of ΔVm across the strand was rather uniform (Figure 2B⇓) and strongly asymmetrical with a much larger portion of the strand being hyperpolarized than depolarized; the asymmetry ratio was 1.96. Such asymmetrical Vm changes were described previously as nonlinear ΔVms of type II.11
Application of nifedipine increased the amplitudes of ΔV+m and decreased the amplitudes of ΔV−m at all sites (Figure 2A⇑, thick traces), preserving the uniformity (Figure 2B⇑) and linearity (Figure 2C⇑) of the ΔVm distribution. The overall effect of nifedipine on ΔVm was a positive shift of ΔVm without changing the slope of the spatial ΔVm gradient (Figure 2C⇑). As a result, the 0 isoline in the map of ΔVm distribution (Figure 2B⇑) shifted toward the center of the strand, reflecting the reduction in the asymmetry of ΔVm distribution. The maximal ΔV+m increased from 94 to 119%APA, and maximal ΔV−m decreased from −184 to –156%APA; the ΔVm asymmetry ratio decreased from 1.96 to 1.31. At the same time, the sum of ΔV+m and ΔV−m remained nearly constant. After washout from nifedipine for 15 minutes, Vm changes were similar to control (thin gray traces).
Similar results were obtained in 8 strands measuring 0.2 mm in width. The data from these experiments are summarized in Figure 2D⇑. The mean shock strength was 15.3±1 V/cm. The maximal ΔV+m was 66±8%APA in control, and it increased during nifedipine application to 88±16%APA (P<0.001), which represents a change of 34%. On washout, the ΔV+m returned to the control value (67±9%APA). Nifedipine caused a reversible decrease of the mean maximal ΔV−m from 139±14% to 126±20%APA (P<0.05). As a result, the ΔVm asymmetry ratio decreased from 2.14±0.4 to 1.5±0.4 (P<0.001) and returned to the control value (2.07±0.3, NS) after washout.
On average, nifedipine induced a small (≈5%) and reversible increase of the sum of ΔV+m and ΔV−m from 204±13 to 214±16%APA (P<0.05). In narrow strands (width≪λ), ΔV+m+ΔV−m should remain constant (see Appendix). Therefore, the small increase in the optically measured ΔV+m+ΔV−m can be attributed to a decrease of APA caused by nifedipine.15 Also, nifedipine slightly decreased the conduction velocity (37±5 versus 34±6 cm/s, P<0.05), whereas changes in the maximal upstroke rate of rise (100±15 versus 95±13 V/s) were not statistically significant.
To test whether the effect of nifedipine was different for small and symmetrical ΔVms described previously,11 measurements were carried out at lower shock strengths. Figures 3A⇓ and 3B⇓ illustrate a typical measurement, and Figure 3C⇓ presents statistical data from 6 strands. The shock strength in these measurements was 2.3±0.1 V/cm inducing nearly equal maximal ΔV+m and ΔV−m (16.8±3.0 and 18.7±3.2%APA, NS). Nifedipine caused small increases in both ΔV+m and ΔV−m (to 21.3±2.7 and 22.2±4.3%APA) without changing the asymmetry ratio (1.13±0.2 versus 1.05±0.2%APA, NS).
To test whether the strand width modulated the effects of nifedipine on ΔVm, measurements were performed in strands 0.8 mm in width. The shock strength in these experiments was adjusted to produce asymmetrical ΔVm with magnitudes similar to the ones observed in the narrow strands. Figures 4A⇓ through 4C⇓ present a typical example of ΔVm caused by 10.8 V/cm (control) and 10.6 V/cm (nifedipine) shocks. The spatial distribution of ΔVm was uniform (Figure 4B⇓) and nonlinear (Figure 4C⇓), as expected from ΔVm in strands with width larger than λ. The maximal amplitudes of ΔV+m and ΔV−m were 57 and −200%APA, respectively, resulting in the asymmetry ratio of 3.5. As in the narrow strands, application of nifedipine caused a strong increase of the ΔV+m (100%APA). Contrary to the narrow strands, however, the maximal ΔV−m slightly increased (–207%APA). The increase of ΔV+m reduced the asymmetry ratio to 2.1.
Similar results were obtained in 8 strands measuring 0.8 mm in width (Figure 4D⇑). The mean shock strength was 9.3±0.8 V/cm. In the control, the magnitudes of maximal ΔV+m and ΔV−m were 83±8 and 156±17%APA, respectively. Nifedipine caused a large and reversible increase of ΔV+m to 107±11%APA (P<0.001), which represents a change of 70%, and a small increase of ΔV−m to 171±20%APA (P<0.001). Because of the increase of ΔV+m, the ΔVm asymmetry was reduced from 2.52±0.5 to 1.62±0.2 (P<0.001).
Effect of 4-AP on ΔVm
The effects of 4-AP (2 mmol/L) on shock-induced ΔVm were determined in 7 narrow strands. In 6 of 7 strands, 4-AP caused a small reduction in ΔV+m and ΔV−m, whereas in 1 strand they were slightly increased. Figure 5A⇓ presents an example of the Vm recordings during application of a 15.6 V/cm shock in which ΔVm was decreased. For simplicity, only the traces measured at the strand edges with maximal ΔV+m and ΔV−m are shown. The relative changes in ΔV+m and ΔV−m were similar, which is reflected in the decrease of the slope of the spatial profile of ΔVm with only a slight shift of the 0 point (Figure 5B⇓), reflecting no change in the ΔVm asymmetry. Figure 5C⇓ presents data of ΔVm measurements in the 6 strands in which ΔVm was decreased. 4-AP caused small reductions in the average ΔV+m, ΔV−m and ΔV+m+ΔV−m, but no change in the asymmetry ratio in these strands or in the remaining strand. The decrease of ΔV+m+ΔV−m without change of the ΔV−m/ΔV+m can be attributed to a decrease of the APA and not to changes in ΔVm per se.
Effect of Shocks on Diastolic Vm
It has been previously suggested that ΔVm asymmetry is caused by membrane electroporation.16 Membrane electroporation is typically manifested by reduction of diastolic potential after shock application.17 To test whether shocks induced such changes of Vm in cell cultures, 500-ms recordings of Vm were carried out in 4 narrow strands (shock strength 18.7±2 V/cm) and 4 wide strands (shock strength 9.8±0.9 V/cm). Figure 6⇓ compares Vm recordings with and without a shock in a typical case with asymmetrical ΔVm (Figure 6A⇓) caused by 9.9 V/cm shock in a 0.8-mm strand. The repolarization phase of the AP (Figure 6B⇓), although being compromised by the motion artifact, was not significantly different in traces with and without shock. Most importantly, there were no changes in the diastolic level of Vm, indicating an absence of electroporation. Similar results were obtained in all 8 strands.
Understanding the mechanisms of Vm changes induced by an electric field is essential for understanding the mechanism of defibrillation. In the present work, we used ionic channel blockers to evaluate the contribution of Ca2+ and potassium ionic currents in the nonlinear Vm changes in cultured monolayers of neonatal rat myocytes. The most important findings of this work are that (1) blocking the L-type Ca2+ current with nifedipine caused an increase in the shock-induced ΔV+m and a reduction in the ΔVm asymmetry, (2) application of the Ito blocker 4-AP did not reduce the ΔVm asymmetry, and (3) effects of the drugs on ΔVm were modulated by the tissue geometry.
Roles of K+ Currents in Nonlinear ΔVm
The asymmetry in the shock-induced ΔVm with larger negative than positive Vm changes reflects an increase in the net outward current. Therefore, it could be expected that the mechanism of the ΔVm asymmetry was related to the flow of an outward K+ current and application of potassium channel blockers should reduce the degree of the ΔVm asymmetry. In a previous study, we tested the effects of barium chloride and dofetilide, inhibitors of inward rectifier and delayed rectifier potassium currents.11 Neither of these channel blockers reversed the asymmetrical distribution of ΔVm. Another possible candidate for such an outward current is Ito, which is prominent during the early plateau phase.15 We found that application of Ito blocker 4-AP reduced both ΔV−m and ΔV+m without change in the ΔVm asymmetry ratio (Figure 5⇑). Part of the changes in the optically measured ΔVm caused by 4-AP might be attributed to the effect of 4-AP on the APA. An increase of APA by 4-AP is indicated by the decrease of ΔV−m+ΔV+m (Figure 5C⇑). Because APA was used to normalize ΔVm measurements, an increase of APA by 4-AP can explain the apparent decrease of ΔVm. Importantly for the questions addressed in this work, however, 4-AP did not change the asymmetry ratio, which excludes Ito as the outward current responsible for the mechanism of ΔVm asymmetry. On the basis of these results, as well as the results of the previous study,11 it can be concluded that the outward potassium currents are not responsible for the asymmetrical nonlinear dynamics of ΔVm in myocyte cultures.
Role of Ca2+ Current in Nonlinear ΔVm
Quite unexpectedly, it was found that the asymmetrical nonlinear behavior of ΔVm was reversed by application of Ca2+ channel blocker nifedipine. As seen in Figures 2⇑ and 4⇑, nifedipine caused a strong increase in the magnitude of the maximal positive ΔVm, but either no change (wide strands) or a decrease (narrow strands) of the negative ΔVm. In both cases, the ΔVm asymmetry was strongly reduced.
As we argue below, the effect of nifedipine on the ΔV−m in the narrow strands was not direct but was mediated via electrotonic interaction with the area of positive ΔVm. Thus, the key effect of nifedipine was the increase of the ΔV+m. In general, 2 different explanations for this effect are possible. First, nifedipine could directly affect the ΔV+m. Second, similarly to the effect of 4-AP, nifedipine could affect APA, thus changing the optically measured ΔV+m. Although it is known that nifedipine can slightly reduce APA in neonatal rat myocytes,15 this effect does not explain the available data. Indeed, the degree of the APA change can be inferred from the change in the sum of optical ΔV+m and ΔV−m in the narrow strands. This sum equals E*L/APA, where E and L are the field strength and the strand width, respectively (see Appendix). As shown in Figure 2D⇑, this parameter changed in average by only ≈5%. The average change in ΔV+m was 6-fold larger in the narrow strands (≈33%) and >10-fold larger in the wide strands (≈73%). Another argument against this interpretation is that nifedipine strongly reduced the asymmetry ratio, which is independent of APA. Thus, the APA changes cannot account for the measured changes in the optical ΔVm.
More likely, nifedipine caused a direct effect on the shock-induced ΔV+m. The fact that nifedipine increased ΔV+m indicates that the flow of Ca2+ current in control conditions prevented the positive changes of Vm or, in other words, ICa was directed outward. Normally, ICa is inward but it changes direction when the Vm is above the reversal potential. This line of reasoning leads to a conclusion that the ΔVm asymmetry was caused by the outward flow of ICa in the depolarized portions of strands when the Vm levels exceeded the ICa reversal potential. According to patch-clamp studies, the ICa reversal potential in rat and rabbit myocytes is 45 to 50 mV.18 19 The magnitudes of the nifedipine-sensitive ΔV+m caused by strong shocks were higher in this study, which is consistent with the idea that ICa was outwardly directed in these portions of the strands. This explanation of the ΔVm asymmetry is also supported by measurements using weak shocks, which showed no sensitivity of the ΔVm ratio to nifedipine when ΔVms were ≈20%APA. Assuming that the APA was 100 mV and the plateau level was 20 mV, this level is close to the ICa reversal potential.
In addition to the effect of ΔVm asymmetry, involvement of Ca2+ current might also explain the saturation of positive ΔVm observed during strong shocks.9 11 Previously, the effect of ΔVm saturation was linked to a nonspecific increase in membrane conductance caused by membrane electroporation.16 However, saturation of positive ΔVm was observed in cardiac tissue at ΔVm levels of ≈100 mV,9 11 whereas electroporation in voltage-clamp experiments typically occurred at ΔVm >300 mV.17 Furthermore, shocks caused no change in the diastolic potential (Figure 6⇑), which is expected from electroporation.17 Therefore, results of this study indicate that the ionic properties of the cell membrane reduce the likelihood of reaching very high Vm levels, thus protecting cardiac cells from the detrimental effects of strong electric shocks.17
Modulation of Drug Effects by Tissue Structure
An interesting feature of the effects of drugs on ΔVm is that they were strongly dependent on the tissue structure. Thus, nifedipine caused reduction of the negative ΔVm in the narrow strands (Figure 2⇑) but not in the wide strands (Figure 4⇑). The decrease of ΔV−m in the narrow strands was likely not due to the effects of nifedipine on the channel properties at these locations but was rather due to the electrotonic interaction between the depolarized and hyperpolarized areas. Because of a strong electrotonic interaction in the narrow strands, any charge entering intracellular space on one side of the strand should quickly redistribute across the strand, causing equivalent shifts of Vm at all sites, as in Figure 2⇑. When the area of hyperpolarization is “buffered” from the area of depolarization, the ΔV−m was much less affected by the shock (Figure 4⇑).
Overall, these results indicate that the effects of shocks on Vm in cardiac tissue are a result of dynamic interaction among several factors, including active properties of cell membrane, myocardial structure, and passive tissue properties, as well as applied electric field.
Extrapolation of the results of this study to the intact adult myocardium can be potentially limited by differences in the ultrastructural and ionic channel properties between neonatal and adult myocytes. Particularly, the cellular distribution of gap junctions is different in cells of these 2 types.20 21 22 This difference might affect distribution of the shock-induced ΔVm on the subcellular scale, but it should not play a significant role in ΔVm measured on a scale larger than cell size. In the context of this study, a more important factor is the species- and age-dependent differences in properties of ionic currents.23 24 25 26 27 Nevertheless, the fact that the same types on nonlinear ΔVm were observed both in neonatal rat10 11 and in adult guinea pig8 9 12 preparations suggests that response of Vm to defibrillation shocks shares common mechanisms.
Magnitude of (ΔV+m+ΔV−m) in Narrow Strands
When strand width, L, is much smaller than the electrotonic space constant, λ, the intracellular potential, Vi, is constant across the strand. Therefore, any change in Vm (Vm=Vi−Ve, where Ve is extracellular potential) across a strand is due to changes in Ve, or The change in Ve across a strand, ΔV+e+ΔV−e, is equal to E*L, where E is field strength. Therefore, and for optically measured ΔVm,
This work was supported by a grant from the Whitaker Foundation (to V.G.F.) and NIH Grant HL-42760 (to R.E.I.). We thank Windy Jones for her help with preparation of the cell cultures.
- Received June 30, 2000.
- Revision received July 21, 2000.
- Accepted July 21, 2000.
- © 2000 American Heart Association, Inc.
Wikswo J. Tissue anisotropy, the cardiac bidomain, and the virtual electrode effect. In: Zipes DP, Jalife J, eds. Cardiac Electrophysiology: From Cell to Bedside. Philadelphia: WB Saunders Co Inc; 1995:348–361.
Zhou X, Smith W, Rollins D, Ideker R. Transmembrane potential changes caused by shocks in guinea pig papillary muscle. Am J Physiol. 1996;271:H2536–H2546.
Gillis AM, Fast VG, Rohr S, Kléber AG. Effects of defibrillation shocks on the spatial distribution of the transmembrane potential in strands and monolayers of cultured neonatal rat ventricular myocytes. Circ Res. 1996;79:676–690.
Fast VG, Rohr S, Ideker RE. Non-linear changes of transmembrane potential caused by defibrillation shocks in strands of cultured myocytes. Am J Physiol. 2000;278:H688–H697.
Cheng DK, Tung L, Sobie EA. Nonuniform responses of transmembrane potential during electric field stimulation of single cardiac cells. Am J Physiol. 1999;277:H351–H362.
Rohr S, Schölly DM, Kléber AG. Patterned growth of neonatal rat heart cells in culture: morphological and electrophysiological characterization. Circ Res. 1991;68:114–130.
Darrow BJ, Laing JG, Lampe PD, Saffitz JE, Beyer EC. Expression of multiple connexins in cultured neonatal rat ventricular myocytes. Circ Res. 1995;76:381–387.
Darrow BJ, Fast VG, Kléber AG, Beyer EC, Saffitz JE. Functional and structural assessment of intercellular communication: increased conduction velocity and enhanced connexin expression in dibutyryl cAMP-treated cultured cardiac myocytes. Circ Res. 1996;79:174–183.
Osaka T, Joyner RW. Developmental changes in calcium currents of rabbit ventricular cells. Circ Res. 1991;68:788–796.
Huynh TV, Chen F, Wetzel GT, Friedman WF, Klitzner TS. Developmental changes in membrane Ca2+ and K+ currents in fetal, neonatal, and adult rabbit ventricular myocytes. Circ Res. 1992;70:508–515.
Wahler GM. Developmental increases in the inwardly rectifying potassium current of rat ventricular myocytes. Am J Physiol. 1992;262:C1266–C1272.