Behavior of eye-movement-related cells in the vestibular nuclei during combined rotational and translational stimuli.

1. Secondary position-vestibular-pause (PVP) neurons in the vestibuloocular reflex (VOR) pathway of adult rhesus monkeys were studied during combined semicircular canal and otolith stimulation. The head was rotated at 0.5 Hz with the axis of rotation centered between the otolith organs (on-axis, ON) and with the axis of rotation 23 cm in front of the otoliths (off-axis, OFF). Both conditions were tested with two different vergence angles by the use of 14-cm (near target, NT) and 100-cm (far target, FT) targets. 2. The tangential translational stimulus to the otoliths in the OFF trials should result in a compensatory eye movement that is opposite in direction to that resulting from the angular stimulus to the canals. The otolith stimulus should be great enough to reverse the eye movement response in the NT OFF trials according to geometric calculations. This reversal in eye movement direction occurred as expected although the latency of the reversal (70 ms) was somewhat greater than expected and the magnitude of the reversal was less than predicted solely on the basis of geometric considerations. 3. The responses of the PVP neurons were corrected for eye position sensitivity to investigate the head movement response components. The amplitude of the response in 22 of 24 PVP cells was reduced in the NT OFF condition compared with the FT OFF condition. This difference was not sufficient in itself to explain the observed reversal in eye movement response. 4. The average sensitivities of the neurons to rotation during the FT and NT ON trials were 1.38 and 1.41 spikes.s-1.deg-1.s-1, respectively. This is too small an increase to account for the increase in the angular VOR gain with near targets (approximately 25%); therefore cells other than PVP neurons must be responsible. 5. The average sensitivities of the PVP neurons to translational accelerations obtained from the FT and NT OFF trials were 305 and 484 spikes.s-1.g-1, respectively, which is higher than most otolith afferent sensitivities reported for 0.5-Hz stimuli in the literature. The otolith component is modified by ocular convergence (59% increase in sensitivity), but this increase is too small to account for the change in the translational VOR gain between the two conditions. 6. Although recordings were only obtained from seven eye-head-velocity cells, the results indicate that these neurons may provide the additional signals not present in the PVP cells. These neurons exhibited large differences between ON and OFF rotations and were found to substantially increase their modulation during the NT conditions compared with that observed during the FT conditions.

Behavior of cells without eye movement sensitivity in the vestibular nuclei during combined rotational and translational stimuli.

A total of 74 neurons that lacked eye movement sensitivity were recorded within the confines of the rostral medial and medial lateral vestibular nuclei. Of these, 36 had response characteristics that were consistent with combined canal and otolith inputs (CAOT neurons), 18 received canal inputs only (CA neurons), and 20 had otolith inputs only (OT neurons). Responses were measured during both rotational and combined rotational and translational stimuli at 0.5 and 3.0 Hz. The otolith signal was found to lag acceleration markedly at both frequencies. Indeed, one subset of CAOT neurons had otolith responses that led translational velocity by only 12 degrees at 0.5 Hz. All translation-responsive neurons decreased their phase lag with respect to acceleration when the stimulus frequency was increased and exhibited a large increase in sensitivity. As these cells have response dynamics that lie between those seen in otolith afferents and those required to drive the motoneurons during the translational VOR, they may represent an intermediate stage in the signal processing.

Eye position signals in the vestibular nuclei: consequences for models of integrator function.

Recordings from neurons in the vestibular nuclei indicate that the cells that carry eye position signals encode the position of a single eye (either ipsilateral or contralateral) during both conjugate and vergence eye movements. The fact that the vestibular nuclei are aware of the positions of each eye is not surprising as the otolith-based linear vestibulo-ocular reflex is known to change its behaviour as a function of uniocular eye position. This result suggests that the signal coming from the oculomotor velocity-to-position integrator specifies the position of each eye during vergence movements and thus must receive a vergence velocity input along with its conjugate velocity inputs. As there is no vergence system in laterally eyed animals, we have proposed two possible models of integrator arrangement that could have developed from conjugate directional (rather than uniocular) integrators in lower animals without frontally mounted eyes. Both of these models explain the existence of near-response cells and produce the required bidirectional gaze paretic nystagmus following unilateral lesions of one integrator. The models also make specific and different predictions concerning the effects of unilateral integrator lesions on the behaviour of the vergence system and thus make suggestions for further experiments.