NONCLOCK BEHAVIOR OF INFERIOR OLIVE NEURONS - INTERSPIKE INTERVAL OF PURKINJE-CELL COMPLEX SPIKE DISCHARGE IN THE AWAKE BEHAVING MONKEY IS RANDOM

1. Complex spikes of cerebellar Purkinje cells recorded from awake, behaving monkeys were studied to determine the extent to which their discharge could be quantified as periodic. Three Rhesus monkeys were trained to perform up to five different tasks involving rotation of the wrist in relation to a visual cue. Complex spike activity was recorded during task performance and intertrial time. Interspike intervals were determined from the discharge of each of 89 Purkinje cells located throughout lobules IV, V, and VI. Autocorrelation and Fourier transform of the autocorrelation function were performed on the data. In addition, the activity from one cell was transformed so that the discharge occurred on the beat of a 10-Hz clock, and in a further transformation, on the beat of a noisy 10-Hz clock. These transformed data were then analyzed as described above. 2. Fourier transform of the autocorrelogram function of the data that had been transformed to a 10-Hz clock, and that of the noisy 10-Hz clock, both showed a prominent peak at 10 Hz. However, the autocorrelograms and the Fourier transforms of the autocorrelogram functions failed to reveal a prominent periodicity for the actual discharge of any of cells, at any frequency up to 100 Hz: the discharge appeared random with respect to the interspike interval. The discharge was not random with respect to behavior. Complex spike activity was commonly time locked to the start of wrist movement. We examined this discharge to see whether oscillatory discharge could be seen after alignment of the data on the start of wrist movement, or after alignment of the data on the complex spike occurring peri-start of wrist movement. No oscillation was seen for either alignment. 3. The inferior olive, which sends its climbing fibers to the cerebellum, has been implicated in such different activities as 1) pathological tremor of the soft palate, 2) physiological tremor, 3) the normal initiation of all bodily movement, and 4) motor learning. Previous work in pharmacologically or surgically treated animals has shown that, under some conditions, the discharge of these neurons is periodic and synchronous. This firing pattern has been interpreted to support a role in the first two activities. But measurements reported here in the awake monkey show just the opposite: the discharge is aperiodic to the extent of being random. As such, the inferior olive cannot be a ''motor clock'' in the general role that has been proposed. We propose as an alternate hypothesis that such random intervals are uniquely appropriate for stably maintaining the synaptic states that have previously been changed by increased climbing fiber discharge time locked to the pertinent behavior. In sum, the discharge pattern would be consistent with the motor learning theories of cerebellar cortex.