We used the whole cell patch clamp technique to investigate the characteristics of modification of cardiac Na+ channel gating by the sea anemone polypeptide toxin anthopleurin-A (AP-A). Guinea pig ventricular myocytes were isolated enzymatically using a retrograde perfusion apparatus. Holding potential was - 140 mV and test potentials ranged from - 1 00 to + 40 mV (pulse duration 100 or 1000 ms). AP-A (50-100 nM) markedly slowed the rate of decay of Na+ current (I(Na)) and increased peak I(Na) conductance (g(Na)) by 38 +/- 5.5% (mean +/- SEM, P < 0.001, n = 12) with little change in slope factor (n = 12) or voltage midpoint of the g(Na)/V relationship after correction for spontaneous shifts. The voltage dependence of steady-state I(Na) availability (h(infinity)) demonstrated an increase in slope factor from 5.9 +/- 0.8 mV in control to 8.0 +/- 0.7 mV after modification by AP-A (P < 0.01, n = 14) whereas any shift in the voltage midpoint of this relationship could be accounted for by a spontaneous time-dependent shift. APA-modified I(Na) showed a use-dependent decrease in peak current amplitude (interpulse interval 500 ms) when pulse duration was 1 00 ms (- 15 +/- 2 %, P < 0.01, n = 17) but showed no decline when pulse duration was 100 ms (- 3 +/- 1 %). This use-dependent effect was probably the result of a decrease in the rate of recovery from inactivation caused by AP-A which had a small effect on the fast time constant of recovery (from 4.1 +/- 0.3 ms in control to 6.0 +/- 1.1 ms after AP-A, P < 0.05) but increased the slow time constant from 66.2 +/- 6.5 ms in control to 188.9 +/- 36.4 ms (P < 0.002, n = 19) after exposure to AP-A. Increasing external divalent cation concentration (either Ca2+ or Mg2+) to 10 mM abolished the effects of AP-A on the rate of I(Na) decay. These results demonstrate that modification of cardiac Na+ channels by AP-A markedly slowed I(Na) inactivation and altered the voltage dependence of activation; these alterations in gating characteristics, in turn, caused an increase in g(Na) presumably by increasing the number of channels open at peak I(Na). AP-A slows the rate of recovery of I(Na) from inactivation which is probably the basis for a use-dependent decrease in peak amplitude. Finally, AP-A binding is sensitive to external divalent cation concentrations. Thus, increasing [Mg2+]o or [Ca2+]o displaces AP-A from binding, suggesting that they share related binding sites on the external surface of the Na+ channel.