The current density and the kinetics of voltage-sensitive sodium channels during neural differentiation were examined in the isolated, cleavage-arrested blastomere of ascidian embryos which contains presumptive neural regions. The macroscopic sodium current were measured with the two-microelectrode voltage-clamp technique and the single sodium channel currents were recorded with the patch-clamp technique under the cell-attached configuration. 2. The entire time course of sodium channel development could be divided into three phases from the current density and channel gating properties. 3. In the first phase, from fertilization to about 40 h, the density of the sodium channel current was from 8 to 50 microA cm-2. The channel gating properties were similar to those of the sodium channel in the egg cell except for a negative shift in the voltage dependence of the peak inward current, the steady-state inactivation, and the decay time constant. The sodium channels in this phase were classified as 'type-I' channels. 4. In the second phase (40-60 h after fertilization), the density of the sodium channel current increased from 20 to 800 microA cm-2. The curves of the I-V relationship and of the steady-state inactivation shifted in the positive direction by 5-10 mV. 5. At 45-55 h, when the rate of increase in the sodium current was greatest, as much as 40 microA cm-2 h-1, the decay time course of the sodium current became slowest. The time for the current to decline from the peak to the one-tenth of the peak (t 1/10) increased to about five times that in the first phase. After 55 h t 1/10 gradually decreased. 6. In this phase, steady-state inactivation curves showed two inflexion points at different levels of membrane potential and were fitted with a sum of two Boltzmann distribution curves with distinct parameters. The relative contribution of the component with its voltage dependence shifted in the positive direction tended to decrease with development. 7. On examining single-channel recordings, two types of sodium channel were identified in this phase. One type (type-II) showed frequent repetitions of open-to-shut states throughout a voltage step. The ensemble current of the type-II channel showed a slow decay, suggesting that this type of channel may underlie the markedly slow decay of the macroscopic current in this phase. The second type (type-III) had more late openings than the type-I channel but fewer than the type-II channel.(ABSTRACT TRUNCATED AT 400 WORDS)
