Effect of Premature Depolarization on the Duration of Action Potentials in Purkinje and Ventricular Fibers of the Moderator Band of the Pig Heart

Transcription

1 Effect of Premature Depolarization on the Duration of Action Potentials in Purkinje and Ventricular Fibers of the Moderator Band of the Pig Heart ROLE OF PROXMTY AND THE DURATON OF THE PRECEDNG ACTON POTENTAL By L. S. Get+es, N. Morehouse, and B. Surawicz ABSTRACT We compared premature and nonpremature action potentials in Purkinje and ventricular fibers from the moderator band of the pig heart to determine if the duration of premature action potentials depended on factors other than preceding cycle length. n Purkinje fibers, the duration of premature action potentials was cycle-length dependent in responses originating more than 100 msec after the preceding repolarization, but the duration of earlier responses was less than the cycle-length-dependent duration. This cycle-length-independent shortening of premature responses increased with greater proximity to the preceding repolarization and increasing duration of the preceding action potential. n ventricular fibers, the duration of premature action potentials was greater than the cycle-length-dependent duration. This cycle-length-independent lengthening increased as the duration of the preceding action potential increased; it also depended on proximity, being greatest when proximity ranged between 26 and 275 msec. The difference between the durations of simultaneously recorded Purkinje and ventricular action potentials decreased as prematurity increased, but the earliest premature Purkinje action potential was consistently shorter than the simultaneously recorded ventricular action potential. Thus, premature stimulation produced different effects in Purkinje and ventricular fibers. However, in both fibers, the deviation of the duration of premature action potentials from the cycle-length-dependent duration was determined, at least in part, by the duration of the preceding action potential and proximity. KEY WORDS repolarization arrhythmia reentry B The steady-state duration of action potentials in Purkinje and ventricular fibers varies with cycle length (1-5). However, the duration of the first action potential following an abrupt change in cycle length may differ from From the Cardiovascular Division, Department of Medicine, University of Kentucky College of Medicine, Lexington, Kentucky This study was supported by a Kentucky Heart Association research grant, a University of Kentucky general research support grant, a Kentucky Heart Association summer student research fellowship program, and National Heart and Lung nstitute Research Grant HE dispersion ionic currents extrasystoles the steady-state duration (3-7). Moreover, previous reports indicate that the effects of a sudden change in cycle length may differ in Purkinje and ventricular fibers and may even vary within fibers of the same type (3-9). The factors responsible for these reported differences have not been elucidated. However, it has been suggested that the duration of the The work was done during Mrs. Morehouse's appointment as a summer research fellow of the Kentucky Heart Association. Received February 24, Accepted for publication October 11,

2 56 GETTES, MOREHOUSE, SURAWCZ first response following a sudden change of cycle length (or premature response) depends in some way on preceding events, i.e., the past history or "memory" effect (4, 7). t has also been postulated that the ionic currents generated during the preceding response may affect the duration of the subsequent response (10). n this study, we have compared the durations of premature and steady-state action potentials preceded by cycles of the same length. We have analyzed the discrepancies between these durations in terms of the duration of the action potential and the diastolic interval immediately preceding the premature response. We have termed this diastolic interval "proximity". Since, with greater proximity, the premature response originates during repolarization of the preceding response, it was also necessary to consider the membrane potential at the onset of depolarization, the "takeoff potential". We postulated that such an analysis would: (1) help to reconcile the differences apparent in earlier reports, (2) allow us to predict with greater accuracy the effects of premature depolarization on the duration of action potentials, and (3) provide greater insight into the factors responsible for the known association between early premature responses and ventricular fibrillation (11). Methods Action potentials from Purkinje and ventricular fibers were recorded from the perfused moderator band of the pig heart with flexibly mounted 3M KCl-filled glass microelectrodes and photographed from an Electronics-for-Medicine recorder as previously described (12). The perfusion solution contained 143 TM Na, 4.8 m\i K, 2.5 rrim Ca, and 1.2 HM Mg in 1 liter, and the temperature of the solution was maintained between 35 and 37 C. The preparation was driven by stimuli 2 5 msec in duration and 1.5 times diastolic threshold strength from a Grass S-4 stimulator and isolation unit. The stimuli were delivered through a pair of stainless steel hook electrodes inserted into the septal end of the preparation. n all experiments, both Purkinje and ventricular action potentials were recorded more than 2 cm from the stimulating electrodes. Premature and nonpremature Purkinje action potentials were compared only when they were recorded from the same cell. Ventricular action potentials were compared when they were recorded from the same area of the preparation, because we were usually unable to record from the same muscle fiber throughout the experiment. The preparation was stimulated at a basic driving rate maintained at /sec, and action potentials were recorded as progressively earlier premature responses were induced. Each premature response was preceded by at least six nonpremature responses, and the records were not accepted for analysis unless the duration of the nonpremature action potential had returned to its steady-state value. The basic driving rate was then increased in steps so that at each step cycle length equaled that which preceded a premature response. The driving rate was maintained at each step until the duration of the action potential became stable. This usually required 2-3 minutes. Duration of an action potential was determined at the 90% level of repolarization, because this point could be determined more accurately than the point at which repolarization was complete. We accepted such records for analysis only when the durations of ten consecutive action potentials were equal. The records of premature action potentials were accepted for analysis when the amplitude of the preceding nonpremature action potential exceeded 110 mv in the Purkinje fiber and 100 mv in the ventricular fiber. By comparing the duration of premature and nonpremature action potentials preceded by identical cycles, we were able to determine whether the duration of the premature action potential was greater or less than the steady-state cycle-length dependent duration (CL-independent lengthening or shortening, respectively). We assessed the effects of premature depolarization on the relationship between the durations of the action potentials of Purkinje and ventricular fibers when such action potentials were recorded simultaneously at a basic driving rate of 1/sec. Proximity was measured from the 90% repolarization point in the nonpremature action potential to the onset of the subsequent premature action potential. Proximity was less than zero when the premature action potential arose before the fiber was 90% repolarized. Statistical evaluation was performed using the Student f-test. Results PURKNJE FBERS The durations of premature and nonpremature action potentials preceded by identical cycles were compared in 17 experiments. A representative experiment is illustrated in Figures 1 and 2. n Figure 1, the premature

3 DURATON OF PREMATURE ACTON POTENTALS 57 CL " OO MB K K 540 APD 90KREP0L CL : 625 APO 90%REP0L CL k"k S U. LU U. W U NONCLDEPENDENT 3HORTENM8 FGURE 1 Action potentials recorded from the same Purkinje fiber. CL = cycle length, APD = action potential duration determined at the 90% repolarization level. Top: The basic driving rate is 0.9/sec (CL = 1100 msec), and progressively earlier premature stimuli are introduced. Middle: The driving rate is increased until the cycle length equals that preceding the premature action potentials in the top row. Bottom: Superimposed tracings from the top row (broken lines) and middle row (solid lines). The numbers below the superimposed tracings are the differences between the durations of nonpremature and premature action potentials preceded by cycles of the same length (CL-independent shortening). responses occur progressively earlier, and proximity changes from 275 msec in A to 10 msec in E. The premature action potentials arise from the resting potential in Figure 1A and B, and their durations are equal to those of nonpremature action potentials preceded by identical cycles. The premature action potentials in Figure 1C and D also arise from the resting potential, but their duration is 15 and 35 msec less, respectively, than expected on the basis of cycle length. We attribute the CL-independent shortening to the progressive increase in proximity, since the other factors which might have influenced the duration of the premature action potential, i.e., takeoff potential and the duration of the preceding action potential, are the same as in Figure 1A and B. n Figure E, the premature action potential arises from a lower takeoff potential, and CL-independent shortening has increased to 175 msec. t is impossible to distinguish the effect of greater proximity from that of lower takeoff potential on the duration of this premature action potential,

4 58 GETTES, MOREHOUSE, SURAWCZ soo 90 > E SO (275) NON PREMATURE AP O PREMATURE AP BASC CL 1100 () PROXMTY (90% REPOL TO ONSET PREMATURE APlmwc B NON CL DEPENDENT SHORTENNG PREC. CYCLE LENGTH msec BOO MOO 1200 FGURE 2 Graphic representation of the experiment shown in part in Figure 1. Durations of premature and nonpremature action potentials and takeoff potentials (TOP) of the premature action potentials are plotted against length of the preceding cycle. The shaded area represents the CL-independent shortening of the premature response. The numbers in parentheses represent the proximity of the premature response to the nonpremature response (msec). APD = action potential duration; CL = cycle length. since both factors change simultaneously. Figure 2 presents a more detailed analysis of the same experiment and shows that when proximity ranges from 85 to 40 msec, CL-independent shortening of up to 40 msec occurs in premature action potentials arising from the resting potential. However, when proximity ranges from 40 to 10 msec, CL-independent shortening is associated with a progressive decrease in takeoff potential. n the experiment illustrated in Figures 1 and 2, the duration of the action potential preceding the premature response was constant and therefore could not be a factor contributing to the CL-independent shortening. To evaluate this factor, we compared the CL-independent shortening of premature action potentials which followed action potentials of different durations, but arose with the same proximity to the preceding action potential and from the same takeoff potential. A representative experiment is illustrated in Figure 3. The duration of the nonpremature action potential is 410 msec in A and 290 msec in B. This difference is due to the longer basic cycle length in A. The premature action potentials in A and B each arise from the same takeoff potential ( 65 mv) and with the same proximity to the preceding action potential ( 5 msec). The preceding cycle length and the duration of the premature action potential are shorter in B than in A, but the CLindependent shortening is 40 msec greater in A than in B. This experiment shows that the CL-independent shortening is increased by the longer duration of the preceding action potential. n Figure 4, we have analyzed our results by relating the CL-independent shortening observed in all experiments to the duration of the action potential preceding the premature response and to the proximity of the premature response. This analysis shows that in each of the groups determined by proximity, CLindependent shortening tended to increase as the duration of the preceding action potential increased, and in each of the groups determined by the duration of the preceding action potential, CL-independent shortening increased as proximity increased. When proximity exceeded 100 msec, the duration of the premature action potential equaled the CLdependent duration and therefore CL-independent shortening was absent. When proximity ranged from 76 to 100 msec, CL-independent shortening was present and tended to be greater when the duration of preceding action potentials was longer. However, the differences were not significant at the P< 0.05 level. When proximity ranged from 1 to 75 msec, CLindependent shortening was significantly greater (P<0.05) in two groups with longer preceding action potentials than in the group with the shortest preceding action potentials. When proximity ranged from 24 to 0 msec, significant differences existed between each of the groups. n each of the groups determined by the duration of preceding action potentials, statistically significant differences (P < 0.05) in CLindependent shortening existed between the proximity ranges of 24 to 0, 1 to 25, and 26

5 DURATON OF PREMATURE ACTON POTENTALS 59 CL OO ; APD % REPOL ! -29o -29O-W CL \ APD 90% REPOL CL NON CL DEPENDENT SHORTENNG FGURE m sec. 50mV Action potentials of premature and nonpremature responses recorded from the same Purkinje fiber. Top: The basic driving rate is 0.9/sec (CL 1100 msec) in A and 2.1/sec (CL = 487 msec) in B. Abbreviations are the same as in Figure 1. to 50 msec. n the remaining groups, statistically significant differences between the increments in proximity could be demonstrated only when the duration of preceding action potentials was msec, most likely because of the larger sample size of this group. n all fibers studied, the first lowering of the takeoff potential occurred when proximity ranged from 30 to 50 msec. Thus, when proximity ranged from 24 to 50 msec, takeoff potential was lower, and the effects of proximity on the duration of premature action potentials and CL-independent shortening could not be distinguished from the possible effects of the lower takeoff potential. However, when proximity ranged from 51 to 100 msec, takeoff potential was not lowered. Our results show that when a premature response originates in an incompletely repolarized fiber small differences in proximity or takeoff potential or both cause large differences in the duration of premature action potentials when Purkinje fibers with slightly different action potential durations are simultaneously depolarized by the same premature stimulus. Figure 5 illustrates an experiment in

6 GETTES, MOREHOUSE, SURAWCZ [ TOP DECREASED TOP NOT DECREASED 0 PREC. APD msec PREC. APD msec 0 PREC. APD msec 5 PROXMTY msec FGURE 4 The mean value of CL-independent shortening it1 Purkinje fibers plus the standard error is shown as a function of proximity and the duration of the prececling action potential (Prec. APD). TOP = takeof potential. which such action potentials were recorded. The duration of the nonpremature action potential is 15 msec greater in the upper trace. Therefore, the premature action potential in the upper trace arises in closer proximity to the preceding nonpremature action potential and from a lower takeoff potential than it does in the lower trace. This results in a 50-msec difference in the duration of the premature action potentials and causes the reversal of the relationship between the durations of the action potentials in these two fibers. VENTRCULAR FBERS The durations of premature and nonpremature action potentials preceded by identical cycles were compared in 11 experiments. A representative experiment is shown in Figure 6. The proximity of the premature response changes progressively from 250 msec in A to -20 msec in D. The premature action potentials in A and B arise from fully repolarized fibers and therefore from the resting potential. n C and D, the premature action potentials arise from incompletely repolarized fibers and therefore from decreased takeoff potentials. The duration of the premature action poten- tial in A is 5 msec longer than that of the preceding action potential and 45 msec longer than the CL-dependent duration. n B-D, the duration of the premature action potentials progressively decreases and approaches the CL-dependent duration so that in D the difference between the premature and nonpremature action potentials preceded by identical cycles, i.e., the CL-independent lengthening, has decreased to 20 msec. Figure 7 presents a more detailed analysis of the same experiment and shows that in the proximity range of 450 to 150 msec the duration of the premature action potential remained nearly constant while the duration of the action potentials preceded by identical cycles decreased from 400 to 350 msec. As a result, CL-independent lengthening increased from 0 to 50 msec. Within the proximity range of 150 to 50 msec, the duration of the premature action potential decreased from 400 to 340 msec. This decrease paralleled the decrease in the steady-state t APD % REPOL. FGURE 5 Action potentiah recorded sinlultaneorrsly froin two diferent Pttrkinje fibers. The records hacje been traced and remounted for greater clarity. The basic driving rate is 0.87/sec (CL = 1150 msec), and the premature action potential is elicited 555 msec after the basic response. Abbrecjiations are the same as in Figure 1. Circulation Research, Vol. XXX, junuary 1972

7 DURATON OF PREMATURE ACTON POTENTALS 61 CL OO APD % REPOL. CL APD 90% REPOL CL 420 NON CL DEPENDENT LENGTHENNG FGURE msec Ventricular action potentials recorded from the same area of the preparation. Top: The basic driving rate is 0.9/sec (CL = 1100 msec), and progressively earlier premature stimuli are introduced. Middle: The driving rate is increased until the cycle length equals that preceding the premature action potentials in the top row. Bottom: Superimposed tracings from the top row (broken line) and middle row (solid line). The numbers below the tracings are the differences between the durations of premature and nonpremature action potentials preceded by identical cycles (CL-independent lengthening). n A and B, the premature action potentials arise from the resting potential, but in C and D, the premature action potentials arise from a lower takeoff potential. CL-independent lengthening is greatest in C; the amplitude of the premature action potential in B is slightly lower than that of the preceding nonpremature action potential. duration (from 350 to 290 msec). Therefore the magnitude of CL-independent lengthening remained constant. Within the proximity range of 60 to 20 msec, the duration of the premature action potential decreased from 340 to 265 msec. This decrease exceeded the decrease in the steady-state duration of action potentials preceded by identical cycles (from 290 to 245 msec). As a result, the magnitude of the CL-independent lengthening decreased to 20 msec. n the entire group of experiments, the durations of premature action potentials arising within 150 msec of the preceding repolarization were always shorter than the duration of the preceding nonpremature action potential. As illustrated in Figure 6B, this shortening of action potential duration was associated with a slight decrease in action potential amplitude even in premature action potentials arising from the resting potential.

8 g CR5) (150) PREC. CYCLE LENGTH msec BOO 900 FGURE 7 NON PREMATURE AP O PREMATURE AP BASC CL 1100 () PROXMTY (90% REPOL TO ONSET PREMATURE AP) «c NON CL DEPENDENT LENGTHENNG Graphic representation of the experiment shown in part in Figure 6. Duration of premature and nonpremature action potentials and takeoff potentials (TOP) of the premature action potentials are plotted against length of the preceding cycle. The shaded area represents the CL-independent lengthening of the premature action potential, and the numbers in parentheses represent the proximity of the premature response to the nonpremature response (msec). CL-independent lengthening increases and then decreases as the proximity of the premature action potential increases. n Figure 8, we have analyzed our results by relating CL-independent lengthening of the premature response to the duration of the preceding action potential and to the proximity of the premature response. Within each proximity range, CL-independent lengthening tended to increase as the duration of the premature action potential increased. The CLindependent lengthening was greater within the intermediate proximity ranges (26 to 275 msec) than within the 24 to 25 msec or the 276 to 475 msec ranges. n the entire group of experiments, the first lowering of takeoff potential occurred when proximity ranged from 20 to 50 msec. Thus, when proximity ranged from 25 to 50 msec, takeoff potential was lower and the effects of proximity on the duration of premature action potentials and CL-independent lengthening could not be distinguished from the possible effect of the lower takeoff potential. However, when proximity ranged between 51 and 475 msec, takeoff potential was not lowered. GETTES, MOREHOUSE, SURAWCZ EFFECT OF PREMATURE DEPOLARZATON ON SMULTANEOUSLY RECORDED PURKNJE AND VENTRCULAR ACTON POTENTALS Premature action potentials from Purkinje and ventricular fibers were recorded simultaneously in 12 experiments. The results of a representative experiment are shown in Figure 9. n this experiment, the nonpremature Purkinje action potential was 90 msec longer than the ventricular action potential. As the premature response occurred progressively earlier, the difference between the duration of the premature Purkinje and ventricular action potentials progressively decreased. When preceding cycle length was 540 msec, the takeoff potential of the premature action potential was at the resting level in both types of fibers, and the durations of the premature action potentials were equal. When preceding cycle length was less than 540 msec, the takeoff potential of the premature Purkinje action potential decreased, and the duration of the premature Purkinje action potential became shorter than the duration of the simultaneously recorded ventricular action potential. The duration of the earliest premature Purkinje 75 ^ S r TOP DECREASEO i JL TOPNOT OECREASED PREC. APD mstc n PREC. APO msec PROXMTY msec FGURE 8 Graphic representation of the relationship between CL-independent lengthening, proximity, and the duration of the preceding action potential. The mean value of CL-independent lengthening plus the standard error is shown. The number of comparisons in each group is shown within the bars. Abbreviations are the same as in Figure 4.

9 DURATON OF PREMATURE ACTON POTENTALS 63 O * PURKNJE FBER AP X * VENTRCULAR FBER AP PREC. CYCLE LENGTH n yo-o-o- XMcfrr** X X- FGURE 9 Graphic representation of an experiment in which action potentials of Purkinje (o) and ventricular (x) fibers were simultaneously recorded as progressively earlier premature responses were elicited. The experiment was terminated when the earliest premature Purkinje response was determined. Premature action potential duration (APD) and takeoff potential (TOP) are plotted against the length of the preceding cycle. The boxed symbols are the values for the nonpremature responses (basic driving rate = 1/sec). the results of Greenspan et al. (9). The duration of premature action potentials originating less than 100 msec from the preceding repolarization was less than the CL-dependent duration. The magnitude of this CL-independent shortening increased with: (1) increasing duration of the preceding action potential, (2) greater proximity to the preceding repolarization, and (3) lower takeoff potential. Although such CL-independent shortening has not been previously reported, its presence might have been suspected from the studies of refractory period by Moore et al. (5) and Janse (7) and from the action potentials computed by Noble (10). Moore et al. (5) did not observe CLindependent shortening in canine Purkinje fibers. However, Figure 8 of their report illustrated as much as 100 msec of CLindependent shortening of the refractory action potential, recorded when preceding cycle length was 460 msec, was 100 msec shorter than the duration of the simultaneously recorded ventricular action potential. The change in the relationship of Purkinje and ventricular action potentials associated with the earliest premature Purkinje response in another experiment is illustrated in Figure 10. n the 12 experiments, the duration of the nonpremature action potential in the Purkinje fiber was msec greater than in the ventricular fiber. However, the duration of the earliest premature Purkinje action potential was msec less than the duration of the simultaneously recorded ventricular action potential. Discussion n this study, the duration of premature Purkinje action potentials originating more than 100 msec from the preceding 90% repolarization level equaled the steady-state CL-dependent duration. This result confirms m sec. FGURE 10 Action potentials recorded simultaneously from a Purkinje fiber (top) and ventricular fiber (bottom). The records have been traced and remounted for greater clarity. The first two action potentials are recorded when the basic driving rate is 1/sec. The third action potential is the earliest premature Purkinje response. The numbers below the action potentials are the differences (msec) between the Purkinje and ventricular action potential durations (P-V). n the nonpremature responses, the duration of the Purkinje action potential is 190 msec longer than the duration of the ventricular action potential, but in the premature response the duration of the Purkinje action potential is 41 msec shorter than the duration of the ventricular action potential.

10 64 GETTES, MOREHOUSE, SURAWCZ period. The proximity of these premature responses, estimated by determining the difference between preceding cycle length ( msec) and the refractory periods of the preceding responses ( msec) approximated 0 msec. These results are consistent with our findings and suggest that CLindependent shortening occurred in some of their fibers. Janse (7) studied changes in refractory period in the bundle branches of isolated perfused dog hearts. He stated that in the first beat following an abrupt decrease in cycle length from 600 to 300 msec, the refractory period of the bundle branches "... shortened to nearly the steady state level." However, Figure 28 of his thesis shows the following: (1) When cycle length was 600 msec, the duration of the refractory period in the left bundle (235 msec) was longer than in the right (225 msec). Thus, when cycle length was abruptly changed to 300 msec, the first response occurred 65 msec after the end of the refractory period in the left bundle and 75 msec after the end of the refractory period in the right. (2) n the first response after the change in cycle length, the refractory period was shorter in the left bundle than in the right, and in both the refractory period of the first response was shorter than that of subsequent responses. These results are consistent with our observation and suggest to us that CL-independent shortening might occur in the intact heart. Our results may be explained by the characteristics of the ionic currents responsible for the action potential. The effects of the duration of the preceding action potentials and proximity may be attributed to the slow repolarizing (13) and slow depolarizing currents (14-18). The possible effects of lower takeoff potential may be attributed to the interaction of rapid and slow depolarizing currents (18-20). We have shown that proximity and the duration of the preceding action potentials were also important determinants of the duration of the premature action potential in ventricular fibers. An appreciation of the effect of proximity helps to clarify divergent results in earlier reports. n our study, the duration of premature ventricular action potentials was not shorter than that of preceding nonpremature action potentials until the response arose within 150 msec of the preceding repolarization. This result confirms the observations of Hoffman and Suckling (3), Gibbs and Johnson (4), and Greenspan et al. (9). n responses originating within 150 msec of the preceding repolarization, the duration of the premature action potential was shorter than the preceding action potential and approached the CL-dependent duration. This observation is consistent with those of Moore et al. (5), who, in Figure 9 of their paper, illustrated premature action potentials arising less than 100 msec from the preceding 90% repolarization level. Janse et al. (6, 7) stated that the shortening of the refractory period in the first response after an abrupt change in cycle length from 600 to 300 msec was "... about 27% of the total shortening." The proximity of the first response after this decrease in cycle length ranged from approximately 30 to 90 msec (Fig. 4 of ref. 7). However, when cycle length was changed from 600 to 225 or 220 msec, the proximity of the first response was approximately 0 or 20 msec, respectively, and approximately 50% of the total shortening of the refractory period occurred (Figs. 15 and 16 of ref. 7). These results are consistent with the decrease in CLindependent lengthening which we observed when proximity ranged between 24 and 25 msec. We have shown that CL-independent shortening in Purkinje fibers and CL-independent lengthening in ventricular fibers increased as the duration of the preceding action potential increased. This difference between Purkinje and ventricular fibers may reflect differences in membrane resistance during the plateau and in the behavior of the ionic currents responsible for the plateau (18, 21-24). We could not eliminate the possibility that electrotonic interaction contributed to our results. However, it is unlikely that these

11 DURATON OF PREMATURE ACTON POTENTALS 65 effects were different in premature and nonpremature responses, since the slope of terminal repolarization in these action potentials was similar. n addition, we did not observe action potentials with sharp spikes and concave repolarization slopes as did Mendez et al. (25) and Sasyniuk and Mendez (26). The results obtained from simultaneously recorded Purkinje and ventricular action potentials were expected because of the results obtained in the individual fibers. These results confirm Janse's observations of refractory periods (7) and supplement his findings by demonstrating that, in early premature responses, Purkinje action potentials may be of shorter duration than the simultaneously recorded ventricular action potentials. Our observations may be pertinent to the genesis of ventricular reentrant arrhythmias. We (27) previously correlated the onset of ventricular fibrillation in the isolated rabbit heart with the slowed depolarization and shortened duration of the premature ventricular action potentials arising from incompletely repolarized fibers. Our present study indicates that the change in the duration of premature action potentials was a function of proximity rather than of lower takeoff potential. Our studies also suggest that the increased temporal dispersion of the recovery of excitability in early premature responses, described by Han and Moe (28), may be related to small differences in the proximity of premature responses to preceding nonpremature action potentials at various myocardial sites. CL-independent shortening in Purkinje fibers may change the sequence of repolarization in fibers having action potentials of different durations and may reverse the recovery of excitability in Purkinje and ventricular fibers. A reversal in the sequence of repolarization has been shown by others in the right and left bundle branches (7) and in central and peripheral canine false tendons (25, 29). The reversal in the sequence of repolarization in Purkinje and ventricular fibers might contribute to the genesis of ventricular arrhythmias by abolishing the protection afforded ventricular fibers by the normally longer duration of action potentials in Purkinje fibers. At slow heart rates, the duration of the nonpremature Purkinje action potential would be prolonged, CL-independent shortening would be exaggerated, and the associated effects on the sequence of repolarization would be more pronounced. These factors may account, in part, for the increased incidence of ventricular arrhythmias associated with bradycardia (30). Acknowledgment The authors acknowledge with thanks the technical assistance of Miss Virginia Sandquist and Mr. Jack Buchanan. References 1. THAUTWEN, W., AND DUDEL, J.: Aktionspotential und Mechanogram des Warmbliiterherzmuskels als Funktion der Schlagfrequenz. Pfluegers Arch 260:24-39, CAHMELET, E.E., AND LACQUET, L.: Duree du potentiel d'action ventriculaire de grenouille en fonction de la frequence: nfluence des variations ioniques de K + et de Na +. Arch nt Physiol Biochim 66:1-21, HOFFMAN, B.F., AND SUCKLNC, E.E.: Effect of heart rate on cardiac membrane potentials and the unipolar electrogram. Am J Physiol 179: , GBBS, C.L., AND JOHNSON, E.A.: Effect of changes in frequency of stimulation upon rabbit ventricular action potential. Circ Res 9: , MOORE, E.N., PRESTON, J.B., AND MOE, G.K.: Durations of transmembrane action potentials and functional refractory periods of canine false tendon and ventricular myocardium: Comparisons in single fibers. Circ Res 17: , JANSE, M.J., VAN DER STEEN, A.B.M., VAN DAM, R.TH., AND DUHRER, D.: Refractory period of the dog's ventricular myocardium following sudden changes in frequency. Circ Res 24: , JANSE, M.J.: Effect of Changes in Heart Rate on the Refractory Period of the Heart. Amsterdam, Mondell-Offsetdrukkerij, MENDEZ, C, GRUHZT, C.C., AND MOE, G.K.: nfluence of cycle length upon refractory period of auricles, ventricles, and A-V node in the dog. Am J Physiol 184: , GREENSPAN, K., EDMANDS, R.E., AND FSCH, C: Effects of cycle-length alterations on canine cardiac action potentials. Am J Physiol 212: , 1967.

12 66 GETTES, MOREHOUSE, SURAWCZ 10. NOBLE, D.: Modification of the Hodgkin-Huxley equations applicable to Purkinje fiber action and pacemaker potential. ] Physiol (Lond) 160: , SMRK, F.H., AND PALMER, D.G.: A myocardial syndrome: With particular reference to the occurrence of sudden death and of premature systoles interrupting antecedent T waves. Am ] Cardiol 6: , GETTES, L., AND SURAWCZ, B.: Effects of low and high concentrations of potassium on the simultaneously recorded Purkinje and ventricular action potentials of the perfused pig moderator band. Circ Res 23: , NOBLE, D., AND TSEN, R.W.: Outward membrane currents activated in the plateau range of potentials in cardiac Purkinje fibres. J Physiol (Lond) 200: , DECK, K.A., AND TRAUTWEN, W.: onic currents in cardiac excitation. Pfluegers Arch 280:65-80, MCALLSTER, R.F., AND NOBLE, D.: Time and voltage dependence of the slow outward current in cardiac Purkinje fibres. J Physiol (Lond) 186: , REUTER, H.: Strom-Spannungsbeziehungen von Purkinje-Fasern bei verschiedenen extracellularen Calcium-Konzentrationen und unter Adrenalineinwirkung. Pfluegers Arch 287: , REUTER, H.: Dependence of slow inward current in Purkinje fibres on the extracellular calcium concentration. J Physiol (Lond) 192: , REUTER, H.: Slow inactivation of currents in cardiac Purkinje fibres. J Physiol (Lond) 197: , MASCHER, D., AND PEPER, K.: TWO components of inward current in myocardial muscle fibers. Pfluegers Arch 307: , WEDMANN, S.: Effect of the cardiac membrane potential on the rapid availability of the sodium-carrying system. J Physiol (Lond) 127: , WEDMANN, S.: Electrical constants of Purkinje fibers. J Physiol (Lond) 118: , JOHNSON, E.A., AND TLLE, J.: nvestigations of the electrical properties of cardiac muscle fibers with the aid of intracellular double barrelled electrodes. J Gen Physiol 44: , GEBSCH, G., AND WEDMANN, S.: Membrane currents in mammalian ventricular heart muscle fibres using a "voltage clamp" technique. Helv Physiol Pharmacol Acta 25: , BEELER, G.W., JR., AND REUTER, H.: Membrane calcium currents in ventricular myocardial fibres. J Physiol (Lond) 207: , MENDEZ, C., MUELLER, W.J., MERDETH, J., AND MOE, G.K.: nteraction of transmembrane potentials in canine Purkinje fibers and at Purkinje fiber-muscle junctions. Circ Res 24: , SASYNUK, B.., AND MENDEZ, C.: Mechanism for re-entry in canine ventricular tissues. Circ Res 28:3-15, SURAWCZ, B., GETTES, L.S., AND PONCE-ZUMNO, A.: Relation of vulnerability of ECG and action potential characteristics of premature beats. Am J Physiol 212: , HAN, J., AND MOE, G.K.: Non-uniform recovery of excitability of ventricular muscle. Circ Res 14:44-60, MVERBERG, R.J., STEWART, J.W., AND HOFFMAN, B.F.: Electrophysiological properties of the canine peripheral A-V conducting system. Circ Res 26: , HAN, J.: Mechanisms of ventricular arrhythmias associated with myocardial infarction. Am J Cardiol 24: , 1969.