Burst reset and frequency control of the neuronal oscillators in the cardiac ganglion of the crab, Portunus sanguinolentus.

1. The five large and four small neurones in the cardiac ganglion of the crab, Portunus, are electrotonically coupled and behave as a single relaxation oscillator, exhibiting periodic bursting activity in vitro. Recorded from the large neurone somata, this activity consists of 200-400 ms slow depolarizations called 'driver potentials' (Tazaki & Cooke, 1979a), accompanied by attenuated action potentials and EPSP's from small neurone input. 2. There is a strong positive correlation between the duration of the driver potential and the duration of the following interburst interval in the spontaneously active ganglion. This correlation is preserved during prolonged depolarization and hyperpolarization. 3. When a driver potential is prematurely terminated by an injected current pulse, the following interburst interval is shortened in direct proportion to the decrease in driver potential duration. 4. When a driver potential or a burst of high-frequency action potential activity is evoked by a depolarizing current pulse, the cardiac oscillator resets to the point of maximum hyperpolarization of the burst cycle, and the following interburst interval is of normal duration. Resetting following an evoked driver potential is complete. Partial resetting occurs only after short, evoked action potential bursts in the absence of a driver potential. 5. Reset of the oscillator causes phase shifts in the subsequent cycles of activity, which vary with the phase of application and duration of the injected current pulse. Response curves have been constructed for a comprehensive range of durations and intensities of hyperpolarizing and depolarizing current pulses applied at all phases of the oscillator cycle. 6. The phase shifts are composed of contributions from the duration of the current pulse, from the premature initiation of the slow depolarizing pacemaker potential due to early termination of the burst, and from the change in interburst interval correlated with truncation of the driver potential. 7. Considering the cardiac ganglion as a relaxation oscillator, frequencey control by entrainment to periodically applied current pulses was quantitatively predicted from the phase-response curves and experimentally confirmed. 8. A high concentration (10(-5) M) of octopamine can inhibit driver potential activity in the large neurones. This was used to examine possible frequency modulating effects of electrotonic feedback from the large neurone driver potentials onto the small neurone pacemaker activity. 9. The observations are discussed in relation to the ionic model for driver potentials and slow pacemaker potential activity in the cardiac ganglion, as proposed by Tazaki & Cooke (1979a, b).