1. The single-unit activity of neurons in the vestibular nucleus, the prepositus nucleus, and the abducens nucleus, whose activity was primarily related to horizontal eye movements, was recorded in alert squirrel monkeys that were trained to track a small visual target by generating smooth pursuit eye movements and to cancel their horizontal vestibuloocular reflex (VOR) by fixating a head stationary target. 2. The spiking behavior of each cell was recorded during 1) spontaneous eye movements, 2) horizontal smooth pursuit of a target that was moved sinusoidally ±20°/s at 0.5 Hz, 3) horizontal VOR evoked by 0.5-Hz sinusoidal turntable rotations ±40°/s (VOR(s)), and 4) voluntary cancellation of the VOR by fixation of a head-stationary target during 0.5-Hz sinusoidal turntable rotation at ±40°/s (VORC(s)). The responses of most (28/42) of the units were recorded during unpredictable 100-ms steps in head acceleration (400°/s2) that were generated while the monkey was fixating a target light. The acceleration steps were generated either when the monkey was stationary or when the turntable was already rotating (VOR(t) trials), and the monkey was canceling its VOR (VORC(t) trials). 3. The firing behavior of all 12 of the abducens neurons recorded was closely related to horizontal eye position and eye velocity during all of the behavioral paradigms used, although there was a small but significant increase in the eye position sensitivity of many of these units when the eye was moving (smooth pursuit) versus when the eye was stationary (fixation). 4. Many neurons in the prepositus nucleus and the medial vestibular nucleus (n = 15) were similar to abducens neurons, in that their firing rate was related primarily to horizontal eye position and eye velocity, regardless of the behavioral paradigm used. These cells were, on average, more sensitive to eye position and smooth pursuit eye velocity than were abducens neurons. 5. The firing rate of 15 other neurons in the prepositus and medial vestibular nucleus was related primarily to horizontal smooth pursuit eye movements. The tonic firing rate of all of these smooth pursuit (SP) cells was related to horizontal eye position, and the majority generated bursts of spikes during saccades in all directions but their off direction. Six of the SP neurons fired in phase with ipsilateral eye movements, whereas the remaining nine were sensitive to eye movements in the opposite direction. During VORC(s), the firing rate of SP neurons was modulated in phase with head movements that were in the same direction as their eye movement sensitivity, although their depth of modulation was usually much less than their sensitivity to eye velocity during smooth pursuit. The response of SP(s) neurons during the VOR(s) could not be explained in terms of the head and eye movement sensitivities estimated during VORC and SP paradigms, respectively. These data suggested that the responses of SP neurons during VORC(s) were related to smooth pursuit eye movements rather than to head velocity. 6. When an unpredictable step in head acceleration was generated in their on direction, SP cells did not respond until 80 ms after the step was initiated during both VOR1 and VORC1 trials. The latter observation also suggested that the modulation of SP neurons during head movements was related to smooth pursuit eye movements, rather than to vestibular nerve activity. 7. We conclude that the firing behavior of abducens neurons during the VOR and VORC can be explained if it is assumed that they receive synaptic inputs from secondary vestibular position-vestibular-pause (PVP) neurons, burst tonic neurons, and SP neurons. We suggest that the premotor inputs related to the nonvisual mechanism for VORC are carried by PVP and burst tonic neurons, whereas premotor inputs related to smooth pursuit mechanism for VORC are carried by SP and burst tonic neurons.
ASJC Scopus subject areas