INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

OPTICAL MAPPING TECHNIQUES are now commonly used in the study of ventricular fibrillation (VF) and defibrillation. A benefit of these techniques is that they provide information on both the depolarization and the repolarization of cardiac action potentials. This information is crucial to the understanding of reentrant wave fronts (spiral waves), wave breaks, and VF. However, a major technical difficulty in applying optical mapping techniques to the study of cardiac activation is the vigorous contraction of the myocardium that distorts the optical signals by the introduction of motion artifacts. One way to overcome this technical difficulty is to use an excitation-contraction (E-C) uncoupler, such as diacetyl monoxime (DAM, or 2,3-butanedione monoxime, BDM) (1, 9, 10, 14, 20, 28), to immobilize the tissue. However, DAM also alters the membrane ionic currents and action potential properties (3, 8, 26). Recently, Riccio et al. (30) showed that DAM prevents the induction of VF and converts existing VF into a stable periodic rhythm by reducing the slope of the action potential duration (APD) restitution (APDR) curve. However, in that study, no mapping was performed, leaving the effects of DAM on patterns of activation undefined. The purpose of the present study was to perform optical mapping studies in swine right ventricles (RV) to determine the effects of DAM on the patterns of activation during VF at concentrations routinely used to immobilize the tissue. We also sought to determine whether or not these changes occur in parallel with changes in the slope of ADPR curve. A third aim was to study whether or not cytochalasin D (Cyto D), an E-C uncoupler without significant effects on APD in canine heart (3, 38), alters the slope of the APDR curve or the patterns of activation during VF.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Tissue preparation and optical mapping. The details of this swine model are similar to those reported elsewhere (19). Briefly, the chest was opened, and the heart was quickly removed. The right coronary artery was cannulated, and the RV was isolated and continuously perfused with oxygenated Tyrode solution (37 ± 0.5°C, pH 7.4 ± 0.5) for the study of DAM (n = 6) or Cyto D (n = 6). The composition of the Tyrode solution was as follows (in mM): 125.0 NaCl, 4.5 KCl, 0.5 MgCl₂, 0.54 CaCl₂, 1.2 NaH₂PO₄, 24.0 NaHCO₃, and 5.5 glucose and included 50 mg/l albumin. The RV was paced at 400-ms cycle lengths. The RVs were stained for 20 min with di-4-ANEPPS (20 min; 1 µmol/l) through coronary perfusion. After the dye staining was completed, VF was induced by rapid electrical stimulation, and the patterns of activation were mapped during VF. Simultaneous single cell transmembrane potential (TMP) recordings were performed using standard glass electrodes. The RV was then electrically defibrillated. TMP and pseudoelectrocardiograms were recorded during dynamic pacing through a pair of bipolar electrodes and during sustained VF. The dynamic pacing protocol was as follows (35). We first paced the RV at a 400-ms pacing interval using twice the diastolic threshold current for eight beats, which was then followed immediately by eight beats of pacing at 350-, 300-, 280-, 260-, 240-, 230-, 220-, 210-, 200-, 190-, 180-, 170-, and 150-ms cycle lengths (CL). TMP recordings obtained during the dynamic pacing protocol were used to construct the APDR curve. Afterward, the perfusate was changed to Tyrode solution containing increasing concentrations of DAM (5, 10, 15, and 20 mmol/l) or Cyto D (10, 20, and 40 µmol/l). The experimental protocol was then repeated after each concentration of the uncouplers and after washout.

Signal processing. The optical signals were temporally filtered and spatially averaged to reduce noise. For temporal filtering, we applied a five-point time median filter to each pixel. We take the original first five data points (we call them frames 1, 2, 3, 4, and 5) and then find the median value of those points and use that as the new value for point 3 (i.e., frame 3). We then take the next original five points (i.e., frames 2, 3, 4, 5, and 6) and find the median value of those and use that as the new value for point 4. We continue this exercise until the end of the data. We then take the tracing, invert the data, and bring the baseline down to zero, which is defined by the average of the five lowest fluorescent values recorded by that pixel. Afterward, we range normalized each pixel. We find the five lowest and five highest points and take the average of those numbers. We then adjust each fluorescent value of the pixel by the same amount so that the highest pixel value will be 255 and the lowest will be 0. Next, for each pixel on the frame, we average the fluorescent values of the pixel and its eight surrounding pixels. We use this average as the new value for the pixel. At completion of those averaging procedures, we repeated the procedure for a second time. At that time, we rezeroed the signal by bringing the baseline down to zero, defined by the average of the five lowest points of each pixel. We then range normalized the signals again. Temporal resolution is unaffected by moving median filters. Spatial resolution was reduced by the 3 × 3 moving spatial average filter, applied twice, by a factor ×2.5 from 0.27 to 0.68 mm.

Transmembrane potential, degree of disorder, and dynamic restitution. We determined peaks of TMP, interactivation interval, APD at 90% repolarization (APD₉₀), and diastolic interval (DI) using methods published elsewhere (19). The APDR curve was created by plotting APD₉₀ against the preceding DI. The restitution curve was generated by exponential fit using ORIGIN 5.0 (Microcal Software, Northampton, MA). Poincaré plots were constructed by plotting each successive interactivation interval against its previous value (12).

Statistical analysis. All data were presented as means ± SD. Student's t-tests were used to compare the means. ANOVA with Newman-Keuls test was used when multiple comparisons were performed. Pearson correlation was used to compare the means of the maximum slope of APDR curve and the KE. The generation of restitution curves and all statistical analyses were performed with the use of ORIGIN (Microcal Software). A P value <0.05 was considered significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Effects of DAM and Cyto D on APD₉₀. DAM and Cyto D had differential effects on transmembrane APD₉₀. Figure 1 shows an example of DAM causing a concentration-dependent shortening of APD₉₀, which was reversible upon washout. In contrast, Cyto D did not significantly affect APD₉₀ at concentrations that varied from 10 to 40 µmol/l. Table 1 shows the results of all RV studied. With concentrations of 5, 10, 15, and 20 mmol/l, DAM reduced APD₉₀ by 37.7%, 38.3%, 45%, and 48.3%, respectively (P < 0.001 for all comparisons). The effects were reversible upon washout. In contrast, Cyto D had no significant effect on the APD₉₀ at any of concentrations studied (Table 1).

View larger version (12K): [in this window] [in a new window]   Fig. 1.   Transmembrane potentials recorded from the epicardium of swine right ventricles (RV). Diacetyl monoxime (DAM) causes concentration-dependent, reversible action potential duration measured to 90% repoloarization (APD90) shortening (A). Cytochalasin D (Cyto D) has little effect on the APD90 (B).

Effects of DAM and Cyto D on slopes of APDR curves. At baseline, the mean maximal slopes of the APDR curves without DAM and Cyto D were 1.14 ± 0.38 and 1.29 ± 0.45, respectively (P = not significant, NS). With increasing concentrations, DAM progressively decreased the maximum slope of APDR, whereas Cyto D showed no such effect (Fig. 2). DAM reduced the maximal slopes of APDR curves in a concentration-dependent manner (Table 1). In contrast, there was no significant change in the slope of APDR curves in any of concentrations of Cyto D studied (Table 1).

View larger version (19K): [in this window] [in a new window]   Fig. 2.   Effects of increasing concentrations of DAM (A) and Cyto D (B) on the dynamic APD restitution curves determined from a swine RV epicardium. DI, diastolic interval.

Effects of DAM and Cyto D on spatiotemporal complexity of VF. Figure 3 summarizes the effects of DAM and Cyto D on the spatiotemporal complexity of VF. According to our previous studies (12, 19), Poincaré plots have four patterns corresponding to four different degrees of complexity in VF: formless, cloudlike appearance (highest in complexity, black); signet, ringlike appearance (higher complexity, crosshatch); cluster (lower complexity, parallel lines); and point (lowest complexity, white) (Fig. 3C). The Poincaré plots had either a formless, cloudlike appearance or signet ringlike appearance at baseline. With increasing concentrations of DAM (Fig. 3A), there was a progression from a higher complexity to lower complexity. At 20 mmol/l of DAM, the VF always converted to VT and Poincaré plots showed point patterns (lowest complexity). This pattern was reversible upon washout. On the other hand, a higher complexity persisted for all concentrations of Cyto D (Fig. 3B). Figure 3, D and E, shows typical pseudoelectrocardiograms and TMP recordings with their associated Poincaré plots shown in Fig. 3C. During baseline VF, pseudoelectrocardiograms and TMP recordings show highly variable amplitudes and interactivation intervals. These patterns were converted to periodic rhythm with higher concentrations of DAM.

View larger version (68K): [in this window] [in a new window]   Fig. 3.   Effects of DAM (A) and Cyto D (B) on spatiotemporal complexities of ventricular fibrillation (VF). C: Poincaré plots of 4 patterns corresponding to 4 different degrees of complexity in VF. D and E: typical pseudoelectrocardiogram and transmembrane potential (TMP) recordings, respectively. See text for details.

Kolmogorov entropy. The results of our KE calculations are shown in Table 1 by the two methods (17, 19). In both methods, entropy decreased significantly with increasing DAM and then sharply increased at washout. We also applied two statistical tests. First, we dropped the last (equals washout) point from each data set and tested the remaining six points for correlation. In every data set, by both methods, there was a statistically significant (P < 0.001) negative correlation with sequence position. In other words, the KE decreased significantly with increasing DAM. Second, to test the significance of the washout, for each of the two methods of calculation of KE, we considered the six KE values for the washout point and compared them with the six average values for the prewashout data by using Student's t-tests. In both methods, the prewashout KE values were significantly higher than the washout KE values. There were also excellent correlations between the maximum slope of the APDR curve and the KE method A (r = 0.93, P = 0.007) and KE method B (r = 0.98, P = 0.001). Whereas, in the Cyto D group, KE values did not show statistically significant changes from baseline to increasing Cyto D concentrations to a 60-min period of superfusion with Cyto D-free Tyrode solution (Table).

Effects of DAM and Cyto D on number of wavelets. All RV in VF at baseline had multiple wavelets within the mapped region (3.4 ± 0.6 in DAM group and 3.4 ± 0.5 in Cyto D group). With increasing concentrations of DAM, the mean number of wavelets progressively decreased (Table 1). In contrast, Cyto D had no such effect (Table 1). Figure 4A shows the number of wavelets at baseline, after 20 mmol/l of DAM perfusion, and after washout in a typical experiment. Figure 4, B and C, shows corresponding fluorescent signals and TMP recordings, respectively. With 20 mmol/l of DAM, the optical map usually showed a repetitive monomorphic planar wave front propagating from outside of the optical field to the other side of the field except in one episode, which showed a stationary spiral wave (Fig. 5) that persisted for >30 s.

View larger version (79K): [in this window] [in a new window]   Fig. 4.   Effects of DAM on wavelet number and TMP recordings. A: snap shots of membrane potential determined by optical mapping. We used color spectrum to represent membrane potentials. Black colors represent repolarized tissue just before the upstroke of the action potential. The color turns red as the membrane is depolarized. When the action potential reaches its peak, the color turns yellow. The color then gradually changes from yellow to green, then blue, and eventually back to black during repolarization. The wavelets were considered to be different if they had different directions of propagation and were separated from each other by blue or black areas. At 20 mmol/l of DAM, monomorphic tachycardia was induced in all preparations. There were multiple wavelets during VF at baseline (wavelet number = 4) and after washout (wavelet number = 3). However, with 20 mmol/l of DAM, there was a single large wave front that repetitively invade the mapped region from left lower corner of the mapped tissue. B and C: corresponding fluorescent signals and single cell TMPs, respectively, of the activation maps shown in A. depol, Depolarization; repol, repolarization.

View larger version (67K): [in this window] [in a new window]   Fig. 5.   Stationary spiral wave induced after 20 mmol/l of DAM perfusion. A: isochrone map of the activation sequences shown in B (time = 0 to 135 ms). Red is the first activated site and purple is the last.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In this study, we found that DAM results in concentration-dependent reduction of APD₉₀ and of the slope of APDR curves in isolated swine RV. These changes are associated with progressively reduced numbers of VF wave fronts and lesser dynamic complexity of VF as measured by KE. At high concentrations, DAM converts VF to VT. In contrast, Cyto D has no effect on APDR curves or the dynamics of VF. These results are compatible with the restitution hypothesis of VF and suggest that DAM may be unsuitable as an E-C uncoupler for optical mapping studies of VF.

Restitution hypothesis of VF. The restitution hypothesis (36) states that APDR is a critical factor in the generation and maintenance of VF and that targeting cardiac restitution could lead to effective antifibrillatory intervention. The argument that a steeply sloped APDR curve creates instabilities during rapid pacing was first made by Nolasco and Dahlen (27). Subsequently, many investigators demonstrated that APDR is important in the generation of alternans, which lead to the development of ventricular arrhythmias, including VF [see a recent review by Weiss et al. (36)]. A quantitative measure of APDR is through the construction of an APDR curve and the calculation of the maximum slope of that curve. A slope 1 is needed to create the dynamic instability that promotes the generation of wavebreak and VF. As a result, chemical agents that flatten the APDR curve (such as verapamil, DAM, or bretylium) decrease VF complexity (6, 30), prevent the induction of VF, and convert existing VF into a periodic rhythm (13, 30).

DAM and Cyto D. DAM is a commonly used chemical agent to induce E-C uncoupling during optical mapping studies (1, 9, 10, 14, 20, 28). Liu et al. (26) previously reported that DAM (5-20 mmol/l) decreased APD₉₀ and the refractory period in a concentration-dependent manner. Riccio et al. (30) expanded these observations and showed the DAM may convert VT to VF at concentrations commonly used for optical mapping studies. Both DAM and Cyto D reduce contraction. Biermann et al. (3) found that, in isolated canine right ventricular trabeculae, DAM at 10 mmol/l and Cyto D at 80 µmol/l were equally effective in reducing peak isometric force to 10 ± 3% and 8 ± 1%, respectively. Wu et al. (38) studied different concentrations of Cyto D on contraction in a canine left ventricular wedge preparation. The concentration used in that study was 10, 15, 20, 25, 30, 35, and 40 µmol/l. They showed that all the concentration of Cyto D could reduce arterial pulses to 1/e of its initial value. The authors suggested that Cyto D concentration between 20 and 30 µmol/l in the perfusate represented the range of the most usage-efficient dose for effective immobilization of ventricular muscle. Our study showed that, whereas both Cyto D (10-40 µmol/l) and DAM (5-20 mmol/l) reduce contraction, only DAM has significant effects on APDR curve and APD.

Alternative mechanisms for VF to VT transition. In addition to the restitution hypothesis, two other hypotheses have been proposed to explain the mechanisms of VF. One is the wavelength hypothesis (29), which states that the length of excitation wave (the product of refractory period and conduction velocity) is an important determinant for reentry. Lengthening of the wavelength is antiarrhythmic, while shortening the wavelength is proarrhythmic. However, in the present study we found that DAM shortens APD but converts VF to VT. Samie et al. (32) showed that verapamil, a calcium channel blocker that shortens APD, is also effective in converting VF to VT. One possible explanation of these findings is that calcium is a major contributor to TMP at sites of slow propagation (22), for example, near the core of reentry when the wave front curvature is steep. Because of this action, calcium channel blockade enlarges the core of reentry, lengthens the cycle length, and converts VF to VT (32). This mechanism could explain the effects of DAM on VF observed in the present study.

Implications on VF mapping. The present study also has significant implications on optical mapping of VF. Chen and colleagues (5) reported that reentrant wave fronts in VF are short-lived, lasting on average only a few rotations before they terminate spontaneously or by wave-wave interaction (4). Similarly, reentrant wave fronts during Wiggers' stage II VF are also short-lived, with an average life span of 3.2 rotations in canine ventricles (23). Rogers et al. (31) later reported even an shorter life span (1.5 rotations) for the reentrant wave fronts on the epicardium of the swine ventricles. Because of its very short life span and uncommon occurrence, Rogers et al. (31) suggested that sustained reentry is transmural or that mechanisms governing sustained reentry are relatively unimportant to the dynamics of VF.

Limitations of the study. Whereas Cyto D does not alter APD or flatten the APDR curve, it does reduce peak sodium currents in ventricular myocytes (34), which may in turn reduce conduction velocity and decrease the wavelength. Although Wu et al. (38) previously reported that Cyto D at a steady concentration of 10 µmol/l did not alter transmural conduction velocity, it is unclear whether conduction velocity is reduced when a higher concentration of Cyto D is used. It would be interesting to study the effects of Cyto D on conduction velocity at the concentrations used in this study. However, Frasier et al. (11) studied transmural activation patterns during epicardial and endocardial pacing. They found that activation in both cases is primarily endocardial to epicardial because of rapid endocardial conduction. These data indicate that the impulse propagation from one point to another on the epicardium may be much longer than the distance between these two points. Because optical mapping techniques cannot be used to study three-dimensional transmural propagation of activation, we were not able to accurately determine the conduction velocity. This is a limitation of the study.