The voltage dependence and the kinetics of block by verapamil of L-type calcium current (ICa) were investigated in ventricular myocytes from rat hearts using the whole-cell patch-clamp technique. ICa was elicited repetitively in response to depolarizing voltage pulses from -80 mV to 0 mV at different pulse intervals and durations. Verapamil reduced the magnitude of ICa in a frequency-dependent manner without tonic component. The time course of ICa remained unchanged suggesting that not open but inactivated channels were affected by the drug. The interaction of verapamil with inactivated channels was investigated by the application of twin pulses. In the presence of verapamil, the duration of the first pulse significantly determined the magnitude of ICa during the second pulse. Variation of the duration of the first pulse between 12 and 3000 ms, followed by a pulse interval of 100 ms, resulted into a gradual decrease of ICa during the second pulse (180 ms), described by concentration-dependent monoexponential decay curves (tau = 1060 +/- 138 ms at 0.3 microM (n = 3); tau = 310 +/- 24 ms at 1 microM (n = 6), and tau = 125 +/- 7 ms at 10 microM (n = 5); means +/- SEM). Under control conditions, the changes in ICa were comparably negligible. The recovery of ICa from block was analyzed by the application of a twin pulse protocol in which two depolarising voltage pulses at fixed length (1. pulse at 3 s and 2. pulse at 180 ms) were interrupted by variable pulse intervals (6 ms-60 s). Under control conditions, recovery from inactivation was fast (tau = 11 +/- 0.7 ms; means +/- SEM; n = 3). In the presence of verapamil, recovery from block was about 500 times slower than under control conditions, independent of the drug concentration (tau = 5.05 +/- 0.44 s at 0.3 microM (n = 3), tau = 6.7 +/- 0.69 s at 1 microM (n = 4), and tau = 6.02 +/- 0.9 s at 10 microM (n = 5); means +/- SEM). Since development of block was dependent on the concentration of verapamil, whereas recovery from block was independent from the drug concentration, it is assumed that the described time constants for block and unblock reflect voltage-dependent net binding (tau on) and unbinding (tau off), respectively, of verapamil at its receptor sites. A computer simulation, including the time constants of block development at 0 mV and of recovery from block at -80 mV, predicted reasonably well the observed frequency-dependent block of ICa by verapamil. The development of either measured or calculated block of ICa, using 180 ms depolarising voltage pulses from -80 mV to 0 mV, was fitted by identical monoexponential association curves (tau = 7 s each at 0.2 Hz and tau = 1.7 s each at 1 Hz). When Ba2+ was used as the charge carrier, which removes the calcium-dependent inactivation of the current, verapamil (3 microM) was less efficient: ICa was decreased by 57 +/- 6% (means +/- SEM; n = 6), whereas IBa was decreased by 24 +/- 4% (means +/- SEM; n = 5). It is proposed that verapamil binds to calcium channels in their inactivated state at more positive potentials and dissociates from the channels in the resting state at more negative potentials. In the proposed scheme of periodical drug binding and unbinding, dependent on the state of the channels, the development of frequency-dependent block of ICa by verapamil is adequately predicted by the construction of cumulative association/ dissociation curves which include the experimentally determined time constants of development and recovery from block at 0 mV and -80 mV, respectively.