Transient K current in the somatic membrane of cultured central neurons of embryonic rat.

1. Somatic K currents of cultured hippocampal, striatal, and spinal cord neurons of embryonic rat were recorded under voltage clamp in membrane spheres ("blebs") excised by means of a tight-seal pipette. 2. The somatic K current in blebs was subject to rapid and near complete inactivation during 300-ms depolarizations, whereas whole-cell K currents included a substantial maintained component. Size and kinetic properties of bleb and whole-cell currents were stable throughout the recording period. 3. The steady-state inactivation of somatic A current was steeply voltage dependent and complete near voltage levels that activated current, whereas peak conductances did not saturate during depolarizations up to +90 mV. Activation started with a delay. Half-times of activation decreased with depolarization, but half-times of inactivation varied little with depolarization. Recovery from inactivation followed a sigmoidal time course with half-times of approximately 50 ms. 4. Half-times of activation and inactivation varied over more than an order of magnitude between individual neurons. Midpoint potentials of inactivation and peak conductance varied over approximately 40 mV. The parameter ranges of hippocampal, striatal, and spinal cord neurons overlapped. 5. Individual soma membranes revealed signs of K channel heterogeneity in their 4-aminopyridine block, current fluctuations, and current kinetics. On the other hand, currents elicited after conditioning pulses that established varied degrees of steady-state inactivation or of recovery from full inactivation had superimposable time courses. 6. The described characteristics of the somatic A channels are compared with those reported for the RCK4, Raw3, and mShal products expressed in Xenopus oocytes. Whereas the ranges of voltage dependencies and of most kinetic characteristics are compatible among native and cloned channels, these three cloned channels recover much more slowly from inactivation. In addition, inactivation in native channels, unlike that in RCK4 and Raw3 channels, was stable after excision in a subcellular fragment.