Optical Transmembrane Potential Measurements During Defibrillation-Strength Shocks in Perfused Rabbit Hearts
Abstract To study the optical transmembrane potential change (ΔF) induced during shocks, optical recordings were obtained in 15 isolated perfused rabbit hearts treated with the potentiometric dye di-4-ANEPPS and diacetyl monoxime. Shock electrodes were sutured on the right and left ventricles. A laser beam 30 μm in diameter was used to optically excite di-4-ANEPPS. Fluorescence from a region 150 μm in diameter was recorded during a shock. In the macroscopic study (six animals), there were nine recording spots that were 3 mm apart between the two shock electrodes. In the microscopic study, there were three recording regions that were 3 mm away from either shock electrode and midway between them, with nine recording spots that were 30 μm (three animals), 100 μm (three animals), and 300 μm (three animals) apart in each region. After 20 S1 stimuli, a 10-ms truncated exponential S2 shock of defibrillation-threshold strength was given during the plateau of the last S1 action potential. In the microscopic study, shocks were also given during diastole, with ΔF recordings at the three recording regions. Shocks of both polarities were tested. ΔF during the shock was expressed as a percentage of the fluorescence change during the S1 upstroke action potential amplitude (the S1 Fapa), ie, ΔF/Fapa%. In the macroscopic study, the magnitudes of ΔF/Fapa% from recording spots 1 to 9, numbered from the left to the right ventricular electrodes, were 77±41%, 46±32%, 32±27%, 28±20%, 37±25%, 24±20%, 33±22%, 37±25%, and 59±29%, respectively (P<.05 among the nine spots). Depolarization or hyperpolarization could occur near either shock electrode with both shock polarities, but the magnitude of hyperpolarization was 1.8±0.9 times that of depolarization at the same recording spot when the shock polarity was reversed (P<.01). In the microscopic study, the change in ΔF/Fapa% varied significantly over the microscopic regions examined. The maximum values of ΔF/Fapa% for hyperpolarizing shocks during diastole reached only 7±10% of those for shocks during the plateau (P<.01). During diastole, the time until a new action potential occurred after the beginning of the shock was shorter when the membrane was depolarized (1.1±0.5 ms) than when it was hyperpolarized (12.8±9.1 ms, P<.01). Conclusions are as follows: (1) A shock can induce either hyperpolarization or depolarization. (2) Hyperpolarization or depolarization during a shock can occur near either the anodal or cathodal shock electrode. (3) Variation of ΔF/Fapa% exists within a microscopic region. (4) The effects of a shock during an action potential plateau are different from those during diastole. (5) The different responses of the ΔF during a shock affect the excitation latency during diastole.
Delivering a strong electrical shock across the heart has long been used to terminate ventricular fibrillation in clinical and experimental practice, yet the mechanism by which the shock affects the heart is still not completely known. Although much work has been performed to study the changes of action potentials following an electrical field stimulus, such as changes of action potential duration and refractory period after a shock,1 2 3 4 5 6 7 only a few studies have used optical recording techniques to study cellular effects during the electrical field stimulus itself. In single isolated cells, these studies showed depolarization at the end of the cell facing the cathode and hyperpolarization at the other end of the cell facing the anode.8 9 In isolated rabbit hearts, depolarization or hyperpolarization of the membrane potential has previously been recorded at a single recording spot3 5 10 and at 64 spots.11
Several mathematical models have been developed to predict the relation between the shock strength and the responses of the transmembrane potential, but some of the results of these models are conflicting and have not yet been verified experimentally.12 13 14 15 For example, some models predict that the change in transmembrane potential over space during a shock exhibits a “sawtooth” pattern, with each sawtooth corresponding to a single cell or group of cells; ie, one end of a cell or group of cells is depolarized, and the other end is hyperpolarized.13 14 If the sawtooth pattern exists and the magnitude of depolarization equals that of hyperpolarization, the transmembrane potential changes during a shock should be summed to zero when a single averaged measurement of transmembrane potential is made from many cells over a large space. However, only depolarization of the transmembrane potential was observed during the shock, when a large single optical recording spot with a diameter of 750 μm and a depth of several hundred micrometers was recorded by Dillon,3 5 suggesting that the sawtooth pattern either (1) is not present, (2) is asymmetrical, or (3) is superimposed on another pattern of transmembrane potential changes. Recently, Knisley et al11 and Neunlist and Tung16 have shown that the optical transmembrane potential, called V̅m since it represents a weighted average over the volume of illumination,16 can be either depolarized or hyperpolarized near an anodal point–stimulating electrode in myocardium. This differs from the traditional concept that only depolarization occurs near the cathode and only hyperpolarization occurs near the anode.17 18 However, this result is consistent with another family of models of the myocardium called bidomain models.12 19
Optical recording techniques have an advantage over conventional electrical recordings in that they do not record the shock or stimulus artifacts. Optical recording has recently been used to investigate the V̅m changes caused by either point or field electrical stimulation in isolated cardiac cells9 20 or in the heart.3 5 10 11 16 One of the disadvantages of optical recording is the large illumination area in the heart, which results in recordings from many cells rather than a single cell. Only a few studies avoided this disadvantage by recording optical signals in either isolated cardiac cells8 9 or cultured cardiac cell strands.21 In an attempt to reduce the influence of optical volume recording, the present study used an illumination area of 30-μm diameter and recorded the optical signals in a microscopic region by moving this optical spot in 30-, 100-, or 300-μm steps between adjacent recording spots.
Cellular excitation and alteration in cardiac action potentials that occur after a defibrillation shock are thought to result from changes in the transmembrane potential caused by the shock. Dillon3 5 has used optical recording techniques at a single cardiac site to show the effects on action potential duration of electrical shocks delivered during different phases of the action potential. More information about the spatial distribution of transmembrane potential changes during a shock is needed. Also, understanding the mechanism of defibrillation requires knowledge of the transmembrane potential changes during the shock in addition to action potential changes following the shock. Therefore, the purpose of the present study is to determine the V̅m changes at different recording sites during shocks of defibrillation-threshold strength in perfused rabbit hearts.
Materials and Methods
Fifteen New Zealand White rabbits weighing 3.5 to 4.5 kg were injected intravenously with sodium nembutal (30 mg/kg body wt) and with 2000 U heparin. The hearts were rapidly excised through a median sternotomy and immersed in cold Tyrode’s solution. The hearts were then mounted on a Langendorff perfusion system, and the coronary arteries were continuously perfused via a cannula in the aortic root with Tyrode’s solution under a constant pressure of 45 mm Hg. Both ventricles were vented. The heart sat in a silicon cradle, and the posterior surface of the heart from which the recordings were made was exposed to room temperature air. The sinus node was removed to reduce the heart rate so that the ventricles could be regularly paced by S1 stimuli. The Tyrode’s solution had the following formula (mmol/L): NaCl 123, CaCl2 1.8, MgCl2 1.1, KCl 4.5, Na2HPO4 1, NaHCO3 20, and glucose 11. The solution was continuously gassed with a mixture of 95% O2/5% CO2, giving a pH of 7.35 to 7.40. Solution temperature was controlled with a thermostatic water bath, and the left ventricular endocardial temperature was monitored by a thermistor and maintained in the range of 36°C to 38°C. The coronary effluent was collected and returned to the perfusion reservoir. A total volume of 4 L Tyrode’s solution was continuously recirculated throughout the experiment.
Two round platinum-mesh shock electrodes (10-mm diameter) were sutured on the right and left ventricles (Fig 1⇓). The distance between these two electrodes was ≈30 mm. Recordings of V̅m were made from the posterior portion of the ventricles between the two patch electrodes. A pair of stainless steel wires was inserted into the wall of the left ventricle to record ventricular electrograms and to monitor the ventricular rhythm. Another pair of stainless steel wires was inserted into the ventricular apex to pace the ventricles and to induce ventricular fibrillation. To measure the surface extracellular potential gradient during the shock, two tungsten electrodes of 0.1-mm diameter were held 4 mm apart in a small piece of silicon rubber. These electrodes were placed immediately adjacent to each laser recording spot as shocks were given. The two electrodes were positioned so that the imaginary line connecting the two electrodes was superimposed on the imaginary line connecting the two shock electrodes. The potential gradient was calculated as the peak difference between the two potential recordings during the shock divided by the 4-mm distance between the two recording electrodes.
At the beginning of each experiment, the defibrillation threshold of the heart was determined for a 10-ms truncated exponential monophasic shock generated by a defibrillator (Ventritex HVS-02) delivered through two patch-shock electrodes. Episodes of ventricular fibrillation were induced with 60-Hz alternating current delivered for 2 seconds to the ventricular apex via two pacing electrodes. Ventricular fibrillation was detected by (1) the rapid activation rate of ventricular electrograms and (2) the loss of synchronized ventricular contraction. Defibrillation shocks were given after 15 seconds of fibrillation. Initial shock intensity was 60 V. The intensity of subsequent shocks was modified on the basis of the success or failure of the previous defibrillation shock. The shock intensity was increased by 10 V if the shock failed to defibrillate or was decreased by 10 V if the shock succeeded. The lowest voltage resulting in successful defibrillation was designated as the defibrillation threshold. Unsuccessful defibrillation shocks were followed by a salvage shock. There was a 4-minute interval between each fibrillation episode. During ventricular fibrillation, coronary perfusion was discontinued by interrupting aortic perfusion. Coronary artery perfusion was resumed immediately after defibrillation. The defibrillation threshold for both shock polarities was determined. The higher defibrillation threshold was used for the next test if the defibrillation thresholds were different.
After measurement of the defibrillation threshold, diacetyl monoxime was added to the Tyrode’s solution at a concentration of 20 mmol/L to eliminate ventricular contraction. Lack of contraction was confirmed by a CCD video camera. A voltage-sensitive dye, di-4-ANEPPS, was dissolved in ethanol,9 added to Tyrode’s solution, and used to record V̅m from spots between the two patch-shock electrodes. The final concentration of di-4-ANEPPS in the Tyrode’s solution was 0.005 mmol/L. Our pilot experiments indicated that there was only a small difference in defibrillation threshold before and after administering di-4-ANEPPS and diacetyl monoxime (P=NS, authors’ unpublished data, 1994). Experimental trials were begun after 10 minutes of exposure of the heart to the di-4-ANEPPS Tyrode’s solution. The heart was paced by a WPI stimulator with stimulus isolator (WPI Pulsemaster A300). Twenty S1 stimuli were given at twice diastolic threshold strength and 2-ms duration with a 300-ms S1-S1 interval. A 10-ms truncated exponential monophasic S2 shock of defibrillation-threshold strength was given from the defibrillator during the plateau of the last S1 action potential. Since the time required for activation to propagate from the pacing site to the optical recording site varied with the location of the different recording sites, the S1-S2 interval was adjusted during each experiment in order to deliver the S2 shock during the plateau, ≈20 to 30 ms after the upstroke of the last S1 action potential at the recording site. This S1-S2 interval was chosen because the cells are refractory at this time. Therefore, the change in V̅m during the shock can be observed and measured without being contaminated by the upstroke of a shock-induced new action potential.9 11
Shocks of each polarity were tested at each recording spot. For the microscopic study described below, after shocks were administered during the plateau of the action potential, shocks with either polarity were also given during the diastolic period at some recording spots. These recordings were made in sequence at each site. First, the last S1-paced action potential was recorded as a control action potential without an S2 shock. Second, the last S1 action potential was recorded with an S2 shock delivered during the plateau of the action potential or during diastole. Third, the last S1 action potential was recorded with an S2 shock with the same timing and strength but of opposite polarity to the previous S2 shock. Shock polarity was tested in random order. The first set of recordings was made from the site closest to the left ventricular shock electrode. After each set of three optical recordings, the recording spot was moved to the next recording spot from the left ventricle toward the right ventricle. The shock potential gradient at each recording spot was calculated by the shock potential recorded by the two tungsten electrodes adjacent to the recording spot and the distance between these two electrodes.
The distance between adjacent spots ranged from macroscopic to microscopic. In the macroscopic study, which used six animals, recordings were made from nine spots with 3-mm spacing between adjacent recording spots (Fig 1⇑). In the microscopic study, which involved a total of nine animals, there were three regions from which nine recordings in a line were made. Two regions were 3 mm away from each shock electrode, and the third was midway between the two shock electrodes, corresponding to spots 1, 5, and 9 (L, M, and R in Fig 1⇑). V̅m was optically recorded sequentially from nine equispaced recording spots in each of the three regions, with the S2 shock given during the action potential plateau. There were three subgroups in the microscopic study, with three perfused hearts studied in each subgroup. Subgroup 1 had 30-μm spacing between adjacent recording spots. Subgroup 2 had 100-μm spacing between adjacent recording spots. Subgroup 3 had 300-μm spacing between adjacent recording spots. In all animals in the microscopic study, shocks with both polarities were given during the action potential plateau at each recording site and were also given during diastole at one of nine recording spots at each region, ie, 3 mm apart from the shock electrodes and midway between two shock electrodes. The time required for all recordings during each experiment was <2 hours.
A 514-nm argon laser beam passed through and was focused by an objective lens (model 170 10/0.25, Leitz Wetzlar; numerical aperture, 0.25). A region of the epicardium ≈30 μm in diameter was illuminated with ≈0.4 mW or 56.6 W/cm2 of light. The 10-mm diameter of the 30-μm laser beam shown on a monitor by a CCD video camera was calibrated by a reticle. The CCD video camera was also used to monitor the heart surface to ensure that there was minimal motion of the heart during recordings in the microscopic group. The illumination light power was measured after the experiment by a digital power meter probe (model 815, Newport) placed in the position previously occupied by the heart. The laser illumination was turned on just before and turned off just after each trial to reduce photobleaching of the dye and the detrimental effect of the laser beam on the cells. The total illuminating and optical recording interval during each S1-S2 trial was 600 ms.
Fluorescence light was collected by the objective lens and passed upward through a dichroic beam splitter with a cutoff wavelength of 520 nm to separate the reflected illumination from the fluorescence. The fluorescence light passed through an eyepiece (Bausch and Lomb; focal length, 1 cm) and then through a long-pass filter with a cutoff wavelength of 590 nm. An avalanche photodiode (model TIED69, Texas Instruments) with an active diameter of 1.5 mm was positioned above the eyepiece and filter with a micromanipulator. The reverse voltage was held constant with an isolated battery supply and was kept below the breakdown voltage to ensure linearity of the photodiode response. Magnification at the photodiode was ×10. Therefore, fluorescence light was collected from a 150-μm-diameter region of heart. This larger collection area, 150 μm in diameter compared with the 30-μm illumination area, was used to ensure the collection of all fluorescence and to reduce noise. The fluorescence signal was low-pass–filtered at −3 dB and a frequency of 1063 Hz. The optical system responded in ≈1 ms to a step increase in fluorescence. The recordings from the laser system, the ventricular electrogram, and the shock potential recordings were monitored and stored on a digitizing oscilloscope with 12-bit accuracy (model 3001A, Norland) and computer hard disk (Macintosh IIx, Apple). The sampling rate for the recordings was 2000 Hz, with which the transmembrane potential changes during the shock were the same as those when a sampling rate of 8000 was used (authors’ unpublished data, 1994).
Analysis of Results
The changes in fluorescence corresponding to amplitudes of paced action potentials and V̅m changes induced during S2 shocks were expressed as a percentage of the background fluorescence level. The optical signal could not be interpreted to give a direct indication of the absolute transmembrane voltage. Fapa was represented by the change in fluorescence between the maximum depolarization potential reached by that action potential and the minimum repolarization potential reached at the end of the previous action potential. ΔF during an S2 shock was calculated as the difference between the peak fluorescence deflection (either depolarization or hyperpolarization) during the shock and the fluorescence just before the shock. ΔF was measured as follows: (1) The beginning of a shock was determined from the signal recorded by the shock potential recording electrodes. (2) A baseline, determined as the mean value during a 2-ms interval just before the beginning of the shock, was obtained. (3) Peak fluorescence change within the interval from the beginning to 10 ms after the beginning of the shock was determined with respect to the baseline. ΔF during the shock was normalized by dividing it by the last S1 Fapa and multiplying by 100 to represent it as a percentage, ie, ΔF/Fapa%. Depolarization of the membrane potential during a shock was defined as a more positive membrane potential during the shock than just before the shock; hyperpolarization was defined as a membrane potential more negative during than just before the shock. The maximum dFapa/dt was defined as the maximum time derivative of the Fapa upstroke from the last S1 action potential. APD80 was defined as the interval from the maximum time derivative of the Fapa upstroke to the time for Fapa to decline to 20% of its amplitude. The shock APD80 was normalized by dividing it by the control S1 APD80 and multiplying by 100. For a shock given during diastole, excitation latency was measured as the interval between the beginning of the S2 shock and the maximum time derivative during the upstroke of the S2-induced optical action potential.
ΔF and the action potential duration were expressed as a percentage to compensate partially for variations (1) in staining by the dye, (2) in dye washout, and (3) among different animals and different S1-S2 episodes. A linear subtraction was performed in which a line passing through the baseline before S1 and the baseline 450 ms after S1, ie, after repolarization occurred, was subtracted from the fluorescence recordings to compensate for decreases in fluorescence introduced by bleaching of the dye by the laser illumination. The beat-to-beat variation in fluorescence for the last S1 Fapa at the same recording site was also measured to see if the fraction of fluorescence changes during different trials at the same site was stable. The variation in fluorescence from beat to beat at the same recording site was determined by dividing the Fapa corresponding to the upstroke of the S1 optical action potential during which the shock was given by the upstroke of the previous S1 optical action potential and then multiplying by 100, ie, ΔFapa%. The maximum variation of the changes in fluorescence from beat to beat was obtained from each subgroup of the microscopic group. The repeatability of ΔF was determined for six recording sites in which shocks with the same strength and polarity were repeated while recordings were made at the same site. The ΔF caused by these shocks was determined and compared to determine the difference in ΔF caused by two sequential shocks. The data were analyzed by ANOVA (Student-Newman-Keuls test) and the paired t test. A value of P<.05 was interpreted as significant. Values are given as the mean±1 SD.
Magnitude of Fluorescence Change During the Upstroke of the Last S1 Action Potential
Fapa, which indicates the change in the fluorescence corresponding to the last S1 action potential amplitude, was determined as a fraction of the baseline fluorescence before the last S1 pacing stimulus. Fapa in the present study was 6.7±1.3%. The amount of repolarization of the last S1 action potential occurring during the interval between the two measurement times (just before the shock and at the peak fluorescence change during the shock) was 1.5±1.8% of Fapa. The changes in Fapa, ie, ΔFapa, of the last S1 action potential from five consecutive recordings in the microscopic group were 0.0±0.0%, −0.5±4.7%, −0.8±4.0%, −0.4±5.9%, and −1.4±6.6% (P>.05 compared with the first recording), where 0.0±0.0% represents the first S1 Fapa of the five trials. The changes in maximum dFapa/dt of the upstroke of the last S1 action potential from these five consecutive recordings in the microscopic group were 100±0%, 103±22%, 99±16%, 103±23%, and 102±19% (P>.05 compared with the first recording), where 100±0% was taken as control from the first recording of the five trials. Lack of significant changes in ΔFapa and maximum dFapa/dt for the five consecutive recordings indicated that the optical recordings at the same spot were reproducible and that the dye and the optical excitation had negligible effects on the tissue.22 The defibrillation threshold for the anode on the left ventricle and the cathode on the right ventricle was 3.9±7.4% lower than the defibrillation threshold for the reversed polarity (P=.055). The lack of significant difference between the two shock polarities might be due to both shock electrodes being on the ventricles.
ΔF During Shock in Macroscopic Study
Shocks equal in strength to the defibrillation threshold induced significant changes in V̅m when given during the plateau of the action potential. Fig 2⇓ shows the mean change in ΔF/Fapa% during the S2 shocks. These changes included either depolarization or hyperpolarization, depending on the shock polarity. When the shock polarity was positive to the right ventricular electrode and negative to the left ventricular electrode, the mean change in V̅m was depolarization (mean value above the zero line in the top of Fig 2⇓) across the heart at all nine spots. When the shock polarity was the opposite, ie, negative in the right ventricle and positive in the left ventricle, the mean change was hyperpolarization (mean value below the zero line in the top of Fig 2⇓) at all spots.
The large standard deviations of the mean values (Fig 2⇑) were caused by the fact that although depolarization was caused primarily by one shock polarity and hyperpolarization by the other polarity, exceptions did occur. When the shock polarity was positive to the right ventricular electrode and negative to the left ventricular electrode, depolarization occurred near the left ventricular shock electrode (spot 1) in five of six shocks, and hyperpolarization occurred only once (Table 1⇓). Near the right ventricular shock electrode (spot 9), depolarization occurred in four of six shocks, and hyperpolarization occurred in the other two cases with the same shock polarity. With the opposite polarity, ie, negative to the right ventricular electrode and positive to the left ventricular electrode, depolarization was induced near the left ventricular shock electrode (spot 1) in only one of six shocks, whereas the other five cases showed hyperpolarization. Depolarization was induced near the right ventricular shock electrode (spot 9) in two of six shocks, whereas the other four cases showed hyperpolarization. When hyperpolarization and depolarization at each recording site were separated into individual groups regardless of the shock polarity, the amount of hyperpolarization during the shock was 1.8±0.9 times the amount of depolarization at the same recording spot when the shock polarity was reversed (P<.01). Reversing shock polarity usually reversed the polarity of the transmembrane potential change, ie, from depolarization to hyperpolarization or from hyperpolarization to depolarization. Thus, whereas either shock polarity induced either depolarization or hyperpolarization, one or the other predominated for each shock polarity.
As expected, the extracellular potential gradient field created during the shock was largest nearest the shock electrodes (spots 1 and 9) and decreased to a minimum in the region midway between the two shock electrodes (spots 4, 5, and 6), as shown in the bottom of Fig 2⇑. The magnitudes of the mean value in ΔF/Fapa% for both polarities were 77±41%, 46±32%, 32±27%, 28±20%, 37±25%, 24±20%, 33±22%, 37±25%, and 59±29% for spots 1 through 9, respectively. ΔF near the shock electrodes (spots 1 and 9) was greater than that farther away from the shock electrodes (P<.05), consonant with the higher potential gradients at these two spots. However, ΔF/Fapa was not statistically different among spots 2 through 8 (P=NS). The correlation between ΔF during the shock (TP) and the potential gradient (PG) was weak (r=.45), with a regression equation of TP=0.3+3.3PG.
Fig 3⇓ shows V̅m recordings during all of the shocks in one rabbit heart. Shocks with negative polarity to the left ventricular electrode and positive polarity to the right ventricular electrode induced depolarization at all recording spots except spot 3, where hyperpolarization occurred. Shocks with the opposite polarity induced hyperpolarization at all recording spots except spot 3, where a small depolarization occurred. In the top tracings, depolarization occurred near both anodal and cathodal shock electrodes (spots 1 and 9), whereas hyperpolarization occurred near both electrodes in the bottom tracings, with shock polarity reversed. The responses near the shock electrodes (Fig 3⇓) were larger than those at the center, and the degree of hyperpolarization was greater than the degree of depolarization at the same recording spot. A large variation in ΔF/Fapa% could occur within a space of 3 mm, as illustrated by the large changes between spots 2 and 3 and between spots 3 and 4 (Fig 3⇓).
Prolongation of Action Potential Duration by Shock
Action potential duration in the macroscopic study was significantly prolonged by shocks at defibrillation-threshold strength at spots 1, 2, 3, 8, and 9 (Fig 2⇑) (P<.05). At spots 1 and 9, which were 3 mm away from the shock electrodes, the normalized action potential duration was 126±9% and 120±13%, respectively. At the midpoint between the two shock electrodes (spot 5), the normalized action potential duration was only 101±12% compared with the control S1 action potential duration of 100% (P=NS). Thus, those recording spots where the action potential duration was significantly prolonged were close to the shock electrodes and received higher potential gradients (Fig 2⇑) than the spots (spots 4, 5, 6, and 7) where action potential duration was not significantly prolonged. The correlation between normalized APD80 and ΔF during the shock (TP) was APD80=105+0.17TP (r=.4, P<.0001). The weak correlation is indicated by the small r value. For all recording spots in the macroscopic study, the normalized APD80 for shocks that induced hyperpolarization was 110±14%, which was not significantly different from that for shocks that induced depolarization, 113±13%.
ΔF During Shock in Microscopic Study
Since a large variation in V̅m could occur 3 mm apart (Fig 3⇑), we measured the V̅m changes during the shock in three microscopic regions with a distance of 30, 100, or 300 μm between the nine recording spots at each of the three regions. The potential gradients generated by the shocks in the three recording regions (L, M, and R in Fig 1⇑) were 18±2, 7±2, and 19±3 V/cm. V̅m changed at the nine recording spots in each region for all experiments (Fig 4⇓). For each of the three distances between recording spots, the responses predominately observed during the shock were similar to those seen in the macroscopic study: (1) The ΔF/Fapa% near the shock electrodes (left and right ventricular [L and R regions]) was greater than that at the midpoint (M region). (2) Depolarization or hyperpolarization during the shock could occur near either shock electrode with either shock polarity; however, depolarization was more common when the left ventricular electrode was the cathode, and hyperpolarization was more common when it was the anode. (3) Hyperpolarization was usually greater than depolarization at the same recording spot when the polarity was reversed. Hyperpolarization occurred in 78% of the shock episodes near the anodal shock electrode and 28% near the cathodal electrode; depolarization occurred in 22% of shock episodes near the anode and 72% near the cathode. Hyperpolarization was 78±15%, 28±2%, and 46±23% of the last S1 action potential amplitude at spots L, M, and R, respectively, which was greater than depolarization at the same spots (47±18%, 15±10%, and 24±23%) (P<.01). The change in V̅m varied within a microscopic region, since the values for each of nine spots did not form a horizontal straight line (Fig 4⇓). The variation of ΔF/Fapa% in Table 2⇓ was calculated for each recording region with the same spacing between recording spots by subtracting the minimum ΔF/Fapa% from the maximum for each curve in Fig 4⇓. Table 2⇓ shows the existence of variation of ΔF/Fapa% during the shock within a microscopic region with a maximum variation of 23±10% to 59±34% of Fapa. Of the 54 curves in Fig 4⇓, 11 crossed zero, indicating that the V̅m change varied from depolarization to hyperpolarization or vice versa in the microscopic region. Table 2⇓ also shows the maximum beat-to-beat variation of Fapa for the last S1-induced action potential upstrokes at recording sites in each microscopic group. The mean values for such maximum variation (ΔFapa%) was significantly lower than the mean value of the maximum variation of ΔF/Fapa% caused by the shocks in each subgroup (P<.01), suggesting that the variation in ΔF/Fapa% during the shock existed over space. In addition, in six recording sites where ΔF caused by a shock was determined repeatedly for the same recording site, the mean value of difference in ΔF/Fapa% caused by two sequential shocks was 4.6±2.3%, which was significantly smaller than the mean value of the maximum variation of ΔF/Fapa% caused by the shocks over space in each microscopic subgroup (P<.01). Thus, the variation in ΔF/Fapa% in each microscopic subgroup was caused mainly by the V̅m changes over space during a shock, not by the lack of repeatability of the recordings.
Fig 5⇓ shows the optically recorded action potentials from a region 3 mm from the left ventricular shock electrode in a perfused heart in which the distance between adjacent recording spots was 100 μm. The potential gradient in this recording region was 20 V/cm. When the shock polarity was negative to the left ventricular electrode, the shock induced depolarization of the optical transmembrane potential. The amplitude of depolarization became smaller as the recording spots were moved from left to right in 100-μm steps until the zero point was crossed and hyperpolarization occurred at spots 8 and 9. For shocks of the opposite polarity, ie, positive to the left ventricular electrode, hyperpolarization became smaller in amplitude as the recording spots were moved from left to right. Thus, Fig 5⇓ illustrates a large change in the V̅m during a shock within an 800-μm region.
ΔF Caused by Shock During Diastole
Optical recordings were also performed during diastolic shocks in the nine isolated perfused rabbit hearts in the microscopic study. The shock was delivered during both the plateau of the action potential and diastole at three recording spots (L, M, and R in Fig 1⇑). Hyperpolarization caused by the shocks during diastole reached only 3±4%, 3±3%, and 3±5% of the last S1 action potential amplitude at L, M, and R regions, respectively, before a new action potential occurred. This degree of hyperpolarization was only 7±10% of the amplitude of hyperpolarization when the shock was given during the action potential plateau (P<.05). For shocks during diastole, excitation latency, ie, the time after the beginning of the 10-ms shock when a new action potential occurred, was shorter when V̅m was depolarized (1.1±0.5 ms) than when it was hyperpolarized (12.8±9.1 ms, P<.01). Because of the short latency, depolarization caused by the shock could not be quantified because a new action potential was almost immediately superimposed on it. Thus, the changes in V̅m caused by shocks delivered during the plateau of the action potential differ from those during diastole, and the response of V̅m varies with shock polarity.
Fig 6⇓ shows optical recordings in which the shock was delivered during the plateau of action potential and during diastole. In this example, the recordings were made at a spot 3 mm away from the left ventricular shock electrode. The potential gradient at this recording spot was 20 V/cm. For the shocks delivered during the action potential plateau, depolarization occurred when the shock polarity was negative to the left ventricular electrode (L−R+ in Fig 6A⇓); hyperpolarization occurred with the opposite polarity (L+R− in Fig 6A⇓). During diastole, a much smaller amount of hyperpolarization was observed (L+R− in Fig 6B⇓) with the same polarity and strength shock that caused hyperpolarization during the plateau of the action potential (L+R− in Fig 6A⇓). The initiation of the action potential occurred nearly at the beginning of the shock that induced depolarization (L−R+ in Fig 6B⇓), whereas it occurred 16 ms after the onset of the shock that induced hyperpolarization (L+R− in Fig 6B⇓), which is 6 ms after the end of the shock.
The present study examined the responses of the cardiac transmembrane potential of the perfused heart to shocks of defibrillation strength. To pursue this goal, V̅m recordings were used to avoid the shock artifact, which often occurs during conventional microelectrode recordings and which may obscure the transmembrane potential during the shock. The main results of the present study are that (1) a shock can induce either hyperpolarization or depolarization, (2) hyperpolarization or depolarization during the shock can occur near either the anodal or cathodal shock electrode, (3) variation of ΔF during the shock exists within a microscopic region, (4) the effects of a shock during the plateau of an action potential are different from those during diastole, and (5) the different responses of V̅m during a shock affect the excitation latency during diastole.
ΔF values in the present study confirm and extend the results of previous experiments in isolated cells and hearts using optical recording techniques.3 5 7 8 9 10 11 16 Shock field stimulation can induce either depolarization or hyperpolarization, depending on the shock polarity as reported by others.9 10 The magnitude of ΔF is greatest near the shock electrode and decreases with increased distance from the electrode (Figs 2⇑ and 3⇑). Our findings of both depolarization and hyperpolarization during a shock differ from the results of Dillon,3 5 who found only depolarization during a shock. This difference may have arisen because Dillon (1) used a different shocking electrode configuration than we did, (2) gave shocks of only a single polarity, and (3) recorded from only a single spot. This single spot may have been in a region that was always depolarized by the shock electrode location and polarity used in the Dillon study.
Theoretical models for a one-dimensional cable predict that during a stimulation pulse the transmembrane potential is depolarized near the cathodal electrode and hyperpolarized near the anodal electrode.17 18 However, the present study demonstrated that either hyperpolarization or depolarization during the shock can occur near either anodal or cathodal shock electrodes (Table 1⇑). These results are consistent with the study of Knisley et al,11 in which stimulation produced hyperpolarization in a region a few millimeters away from a cathodal electrode in the perfused left ventricle with prefrozen endomyocardium and midmyocardium, and with the study of Neunlist and Tung,16 23 which indicated regional depolarization of cardiac muscle adjacent to an epicardial anode. This phenomenon may be explained by bidomain models of two-dimensional myocardium, which predict that although point stimulation with an extracellular cathodal electrode produces depolarization in the tissue directly beneath the electrode and in tissue from the electrode in the direction across fibers, it produces regions of hyperpolarization a short distance away from the electrode along fibers because of different anisotropies of the intracellular and extracellular conductances in the tissue.11 24 25 The distance from the cathode at which hyperpolarization first occurs is thought to vary, depending on fiber orientation with respect to the electrodes. Thus, small differences in distance to the electrode and in fiber orientation might explain why depolarization was recorded in five hearts and hyperpolarization was recorded in one heart at recording spot 1 (≈3 mm from the left ventricular electrode) when it was the cathode and why depolarization was recorded in two hearts and hyperpolarization was recorded in four hearts at recording spot 9 (≈3 mm from the right ventricular electrode) when it was the cathode (Table 1⇑).
Our results demonstrate that when a shock is given during the plateau of the action potential, the shock-induced hyperpolarization obtained with one shock polarity is 1.8±0.9 times the shock-induced depolarization obtained with the other shock polarity at the same recording spot. These results are consistent with the results observed in the optical recordings of transmembrane potential for electrical field stimulation given during the action potential plateau in isolated ventricular myocytes9 and in perfused rabbit hearts11 as well as in intracellular recordings for pulses of either polarity given during the action potential plateau in isolated ventricular fibers.26 If the only effect of a shock field on the transmembrane potential was to charge the membrane and the membrane had no rectifying properties, then the magnitude of depolarization during a shock should be the same as the magnitude of hyperpolarization when the shock polarity is reversed.17 This difference in the magnitude of hyperpolarization and depolarization may be caused by the rectifying properties of the ion channels, with a higher impedance to current flow in one direction versus the other across the cell membrane.27 Another possibility is that in addition to charging the membrane, a shock may alter the active processes in the membrane, such as the conductivities of the ionic channels, leading to larger hyperpolarization than depolarization.
The results of the present study show experimentally for the first time the existence of spatial variation in ΔF caused by a shock within a microscopic region from 0.24 to 2.4 mm in length (Fig 4⇑). Such changes varied from 23±10% to 59±34% of the action potential amplitude (Table 2⇑). ΔF values during the shock could change from depolarization to hyperpolarization in a region as small as 0.8 mm (Fig 5⇑).
This phenomenon is difficult to explain by a mathematical model predicting that the oscillation of the transmembrane potential changes over space during a shock exhibits a sawtooth pattern, with each sawtooth corresponding to a single cell or group of cells.14 According to the single-cell form of this model, with the exception of cells very near the shock electrodes, the part of each cell nearest the cathodal electrode is depolarized, and the part nearest the anodal electrode is hyperpolarized. The optical recording spot size in the present study may encompass several cells to hundreds of cells. Thus, we should have recorded a value near a weighted mean of the transmembrane potential changes over all of the cells within the recording spot. According to the sawtooth model, all cells except those close to the shock electrodes have similar spatial patterns of depolarization and hyperpolarization with the magnitude of transmembrane potential directly proportional to the extracellular potential gradient generated by the shock.13 Since the potential gradient changed slowly over small distances in the heart (Fig 2⇑; and authors’ unpublished data, 1994), the sawtooth model would predict only small variation in the transmembrane potential changes during the shock over a microscopic region instead of the large variation and occasional changes in polarity that were observed (Fig 4⇑). The sawtooth model predicts a strong linear relation between the potential gradient and the change in transmembrane potential over space.13 14 Only a weak correlation (r=.45, Fig 2⇑) was observed.
The changes can be better explained if the sawtooth model is modified so that each sawtooth occurs over a bundle of cells ≈2000 μm long and 200 μm wide instead of over a single cell.13 In this case, each individual sawtooth is larger than our recording spot size, so that large changes in transmembrane potential during the shock could be observed at adjacent recording sites. Other models can also explain these large variations. The presence of connective tissue septae or blood vessels may create secondary sources that cause large variations in transmembrane potential changes near them.25 Different forms of the bidomain model also predict such large variations in the transmembrane potential changes caused by the shock.12 28
One possible reason for the observed result that a shock given during diastole produced smaller ΔF values than the same strength shock given during the action potential plateau may be the higher membrane conductance during diastole than during the action potential plateau.10 29 A second possible reason may be a combination of passive and active processes during a shock delivered during diastole. The large hyperpolarization that was observed when the shock was given during the action potential plateau could have been counteracted by an active process, such as ionic channel activity inducing inward currents caused by the same strength shock given during diastole. This explanation is supported by another result obtained in the present study when the shock was given during diastole. Excitation latency was shorter when the transmembrane potential was depolarized (1.1±0.5 ms) than when it was hyperpolarized (12.8±9.1 ms, P<.01). When shocks, which induced hyperpolarization during the action potential plateau, were given during diastole, the upstroke of the shock-induced action potential could sometimes occur during the shock interval (<10 ms after the beginning of the shock pulse), suggesting the opening of sodium channels leading to an inward current during the shock. This inward current may counteract the hyperpolarization effect of a shock, resulting in the smaller transmembrane potential changes (Fig 6⇑). On the other hand, the hyperpolarization caused by a shock may have delayed the initiation of a new action potential so that the excitation latency was longer when the transmembrane potential was hyperpolarized by this shock.
Previous studies have reported that action potential prolongation by a shock increases when shock strength or the S1-S2 coupling interval is increased.1 3 4 5 6 In the present study, at recording spots that were near the shock electrodes and therefore exposed to higher potential gradients, APD80 was significantly prolonged by the defibrillation-threshold shock given during the action potential plateau,4 30 whereas at spots midway between the two shock electrodes and exposed to low potential gradients, APD80 was not obviously affected.4 30 Action potential prolongation may have occurred at spots midway between the two shock electrodes if longer S1-S2 coupling intervals had been used in the present study, since the action potential is prolonged significantly more when the shock is given later during the action potential.1 3 4 5 6 Although prolongation of action potential duration is thought to be important for both defibrillation and the initiation of arrhythmias,1 2 3 4 5 7 the mechanism for prolongation is still not definitely known. It is not clear how the prolongation of action potential duration is directly related to the transmembrane potential changes during the shock, since action potential prolongation occurs in both the depolarized and hyperpolarized regions (Fig 2⇑) and the correlation between APD80 and ΔF is not strong (r=.4).
Limitations of the Optical Recording System
Optical recording techniques provide advantages over microelectrode recording techniques for determining transmembrane potential changes during electric field stimulation. Optical recordings with voltage-sensitive dye can be recorded from the whole heart and are not directly affected by the extracellular shock field. But, optical recordings of transmembrane potential changes during field stimulation also have limitations. One limitation is the photobleaching effect, which can reduce the baseline fluorescence during intense laser illumination. To reduce the effects of photobleaching, the laser illumination in the present study was performed only during the recording for each shock trial, and decreases in baseline fluorescence during recordings were eliminated by linear subtractions (see “Materials and Methods”). Also, since the fractional fluorescence change corresponding to a given transmembrane voltage change is not constant for all spots, only the relative changes in the optical transmembrane potentials can be estimated with respect to the last S1 action potential upstroke. Therefore, changes in V̅m caused by a shock were expressed as a percentage of the last S1 optical action potential upstroke. Another limitation of the optical recording system is the large volume of myocardium from which fluorescence is collected. The 30-μm area illuminated by each spot may cover only a few cells. But, the depth of the optical recording volume may be hundreds of microns.3 As a result, a single recording is probably a weighted mean from up to hundreds of cells instead of a recording from a single cell.16 Thus, the optical recording volume may cause some information to be missed if rapid spatial changes in transmembrane potential occur within the recording volume. The action potential plateau was gradually repolarizing during the interval when shocks were given; hence, the hyperpolarization and depolarization measurements may include this normal repolarization, which was 1.5±1.8% of the action potential amplitude.
Effects of Diacetyl Monoxime on Transmembrane Potentials
Diacetyl monoxime (20 mmol/L) was used in the present study to eliminate the motion artifact during the optical recording.9 This concentration of diacetyl monoxime has been reported to reduce the contractile force to zero, but this effect is reversible.9 31 There are no major effects of diacetyl monoxime on the membrane electrical properties. For example, in a study using guinea pig papillary muscles, 20 mmol/L diacetyl monoxime decreased the action potential amplitude from a control value of 128 mV in normal Tyrode’s solution to 123 mV, whereas it slightly shortened the action potential duration.31 Thus, there should have been no major influence of diacetyl monoxime on the amplitude of ΔF during a shock in the present study.9 Also, action potential prolongation and refractory period prolongation occurring in the presence of 20 mmol/L diacetyl monoxime are similar to the prolongation without diacetyl monoxime.7 Thus, the action potential prolongation observed in the present study is probably not due to the effects of diacetyl monoxime.
An optical recording technique was used in the present study to determine the change in transmembrane potential during shocks of defibrillation-threshold strength in isolated perfused rabbit hearts. These shocks induced significant depolarization or hyperpolarization and significant variation of ΔF values within a microscopic region. ΔF values for shocks during diastole were smaller than for shocks during the action potential plateau. For shocks during diastole, excitation latency was shorter when the transmembrane potential was depolarized by the shock than when it was hyperpolarized. Both hyperpolarization and depolarization caused by shocks in the plateau of the action potential resulted in prolongation of action potential duration. The relation between the extracellular potential gradient caused by the shock, ΔF during a shock, and the ensuing changes in action potential after the shock are complex, suggesting that the mechanisms by which a shock defibrillates are also likely to be complex.
Selected Abbreviations and Acronyms
|V̅ m||=||optical transmembrane potential|
|ΔF||=||V̅m change caused during the shock|
|Fapa||=||optical action potential amplitude|
|ΔFapa||=||changes in Fapa|
|APD80||=||action potential duration at 80% repolarization|
This study was supported in part by National Institutes of Health research grants HL-33637, HL-42760, and HL-44066; National Science Foundation Engineering Research Center grant CDR-8622201; American Heart Association, North Caroline Affiliate, Inc, grants NC-93-GS-19 and NC-91-G-14; North Carolina Biotechnology Center grant 9113-ARG-0612; and a grant from the Whitaker Foundation.
- Received May 10, 1994.
- Accepted May 8, 1995.
- © 1995 American Heart Association, Inc.
Zhou X, Knisley SB, Wolf PD, Rollins DL, Smith WM, Ideker RE. Prolongation of repolarization time by electric field stimulation with monophasic and biphasic shocks in open-chest dogs. Circ Res. 1991;68:1761-1767.
Swartz JF, Jones JL, Jones RE, Fletcher R. Conditioning prepulse of biphasic defibrillator waveforms enhances refractoriness to fibrillation wavefronts. Circ Res. 1991;68:438-449.
Dillon SM. Optical recordings in the rabbit heart show that defibrillation strength shocks prolong the duration of depolarization and the refractory period. Circ Res. 1991;69:842-856.
Knisley SB, Smith WM, Ideker RE. Effect of field stimulation on cellular repolarization in rabbit myocardium: implications for reentry induction. Circ Res. 1992;70:707-715.
Dillon SM. Synchronized repolarization after defibrillation shocks: a possible component of the defibrillation process demonstrated by optical recordings in rabbit heart. Circulation. 1992;85:1865-1878.
Zhou X,Wolf PD, Rollins DL, Afework Y, Smith WM, Ideker RE. Effects of monophasic and biphasic shocks on action potentials during ventricular fibrillation in dogs. Circ Res. 1993;73:325-334.
Knisley SB, Hill BC. Optical recordings of the effect of electrical stimulation on action potential repolarization and the induction of reentry in two-dimensional perfused rabbit epicardium. Circulation. 1993;88:2402-2414.
Knisley SB, Blitchington TF, Hill BC, Grant AO, Smith WM, Pilkington TC, Ideker RE. Optical measurements of transmembrane potential changes during electric field stimulation of ventricular cells. Circ Res. 1993;72:255-270.
Dillon SM, Wit AL. Use of voltage sensitive dyes to investigate electrical defibrillation. In: Harris G, Walker C, eds. Proceedings of the Annual International Conference of the IEEE Engineering in Medicine and Biology Society. Piscataway, NJ: Institute of Electrical and Electronics Engineers Inc; 1988:0215-0216.
Trayanova NA, Pilkington TC. The use of spectral methods in bidomain studies. In: Pilkington TC, Loftis B, Thompson JF, Woo S-L-Y, Palmer TC, Budinger TF, eds. High Performance Computing in Biomedical Research. Boca Raton, Fla: CRC Press Inc; 1993:403-425.
Hodgkin AL, Rushton WAH. The electrical constants of a crustacean nerve fibre. Proc R Soc Lond B Biol Sci. 1946;133:444-479.
Jack JJB, Noble D, Tsien RW. Electric Curent Flow in Excitable Cells. London, England: Oxford University Press; 1983:413-422.
Trayanova NA, Pilkington TC, Henriquez CS. A periodic bidomain model for cardiac tissue. In: Nagel JH, Smith WM, eds. Proceedings of the 13th Annual International Conference of the IEEE Engineering in Medicine and Biology Society. Piscataway, NJ: Institute of Electrical and Electronics Engineers Inc; 1991:502-503.
Windisch H, Ahammer H, Schaffer P, Müller W, Hartbauer M. Nonuniform response of the field stimulated cardiomyocyte: a study by optical micromapping. In: Szeto AYJ, Rangayyan RM, eds. Proceedings of the 15th Annual International Conference of the IEEE Engineering in Medicine and Biology Society. Piscataway, NJ: Institute of Electrical and Electronics Engineers Inc; 1993:850.
Fast BG, Kléber AG. Microscopic conduction in cultured strands of neonatal rat heart cells measured with voltage-sensitive dyes. Circ Res. 1993;73:914-925.
Tung L, Neunlist M. Regional depolarization of cardiac muscle adjacent to an epicardial stimulating anode. Am Heart J. 1992:124:834. Abstract.
Wikswo JP Jr, Wisialowski TA, Altemeier WA, Balser JR, Kopelman HA, Roden DM. Virtual cathode effects during stimulation of cardiac muscle: two-dimensional in vivo experiments. Circ Res. 1991;68:513-530.
Trayanova N, Roth BJ. Cardiac tissue in an electric field: a study of electrical stimulation. In: Murray A, Arzbaecher R, eds. Proceedings of Computers in Cardiology. Los Alamitos, Calif: IEEE Computer Society Press; 1992:695-698.
Cranefield PF, Hoffman BF. Propagated repolarization in heart muscle. J Gen Physiol. 1958;41:633-649.
Kass RS. The ionic basis of electrical activity in the heart. In: Sperelakis N, ed. Physiology and Pathophysiology of the Heart. Boston, Mass: Martinus-Nijhoff; 1984:83-96.
Weidmann S. Effect of current flow on the membrane potential of cardiac muscle. J Physiol (Lond). 1951;115:227-236.
Dillon SM, Mehra R. Prolongation of ventricular refractoriness by defibrillation shocks may be due to additional depolarization of the action potential. J Cardiovasc Electrophysiol. 1992;3:442-456.
Li T, Sperelakis N, Teneick RE, Solaro RJ. Effects of diacetyl monoxime on cardiac excitation-contraction coupling. J Pharmacol Exp Ther. 1985;232:688-695.