Background: We tested the hypothesis that action potential duration (APD) restitution of normal ventricular muscle cells is different during double premature stimuli (S3) compared with a single premature stimulus (S2). We propose a possible ionic mechanism for such a difference.
Methods and results: Action potentials and isometric tension were recorded simultaneously from isolated canine right ventricular trabeculae (2 x 2 x 10 mm) (n = 35). APD and tension restitution curves (APD) and peak tension versus diastolic interval [DI] of S2 and S3 were constructed by the extrastimulus method during pacing at 1,500 msec. The following results were obtained. 1) The APD restitution curve of S2 was different from that of S3. During the restitution of S2, an early biphasic upward hump was present at short DIs. In contrast, a smooth exponential rise was consistently seen during S3 restitution. 2) Peak tension remained significantly (p less than 0.001) lower during the restitution of S2 than during S3 restitution at all DIs tested. 3) The variation of APD during the initial 100 msec of DI was significantly longer during S3 than S2 (22 +/- 5 msec versus 41 +/- 5 msec, p less than 0.001). 4) Caffeine (2 mM, n = 5) and ryanodine (10 microM, n = 5) blocked cyclic variations of tension, presumably by blocking cyclic variations of intracellular calcium ion concentrations ([Ca2+]i), and eliminated the differences in APD restitution between S2 and S3. 5) Nisoldipine at high (5 microM) but not at lower (2 microM, n = 5) concentration eliminated the differences in restitution of both APD and tension between S2 and S3. 6) BAY K 8644 (100 nM, n = 5) had no effect on this difference.
Conclusions: Greater variations of APD occur during the restitution of S3 than during S2 at short DIs. These differences appear to be caused by cyclic variations in tension and thus in [Ca2+]i. Calcium-sensitive outward currents could explain these differences in APD restitution.