Expression of motor learning in the response of the primate vestibuloocular reflex pathway to electrical stimulation.

1. The vestibuloocular reflex (VOR) undergoes long-term adaptive changes in the presence of persistent retinal image motion during head turns. Previous experiments using natural stimuli have provided evidence that the VOR is subserved by parallel pathways, including some that are modified during learning and some that are not. We have used electrical stimulation of the vestibular labyrinth to investigate the temporal properties of the signals that are transmitted through the modified pathways. 2. Electrodes were implanted chronically in the superior semi-circular canal, the horizontal canal, or the vestibule for electrical activation of the vestibular afferents. Learning was induced by fitting the monkeys with spectacles that magnified or miniaturized vision. Before, during, and after motor learning, we measured the eye movements evoked by electrical stimulation of the labyrinth as well as the gain of the VOR, defined as eye speed divided by head speed during natural vestibular stimulation in the dark. 3. Trains of pulses applied to the labyrinth caused the eyes to move away from the side of stimulation with an initial rapid change in eye velocity followed by a steady-state plateau. Changes in the gain of the VOR caused large changes in the trajectory and magnitude of eye velocity during the plateau, showing that our stimulating electrodes had access to the modified pathways. 4. A single, brief current pulse applied to the labyrinth evoked an eye movement that had a latency of 5 ms and consisted of a pulse of eye velocity away from the side of the stimulation followed by a rebound toward the side of stimulation. To quantify the effect of motor learning on these eye movements, we pooled the data across different VOR gains and computed the slope of the relationship between eye velocity and VOR gain at each millisecond after the stimulus. We refer to the slope as the "modification index." 5. In comparison with the evoked eye velocity, the modification index took longer to return to baseline and showed a large peak at the time of the rebound in eye velocity. Increases in stimulus current increased both the amplitude and the duration of the modification index and revealed several later peaks. These observations suggest that the full expression of motor learning requires activation of multisynaptic pathways and recruitment of primary vestibular afferents with higher thresholds for electrical stimulation. 6. The modification index was almost always positive during the initial deflection in eye velocity, and the latency of the first change in the modification index was usually the same as the latency of the evoked eye movement.(ABSTRACT TRUNCATED AT 400 WORDS)