Loss of the neural integrator of the oculomotor system from brain stem lesions in monkey

1. Eye movements were recorded from four juvenile rhesus monkeys (Macaca mulatta) before and after the injection of neurotoxins (kainate or ibotenate) in the region of the medial vestibular and prepositus hypoglossi nuclei, an area hypothesized to be the locus of the neural integrator for horizontal eye movement commands. Eye movements were measured in the head-restrained animal by the magnetic field/eye-coil method. The monkeys were trained to follow visual targets. A chamber implanted over a trephine hole in the skull permitted recordings to be made in the brain stem with metal microelectrodes. The abducens nuclei were located and used as a reference point for subsequent neurotoxin injections through cannulas. The effects of these lesions on fixation, vestibuloocular and optokinetic responses, and smooth pursuit were compared with predicted oculomotor anomalies caused by a loss of the neural integrator. 2. Kainate and ibotenate did not create permanent lesions in this region of the brain stem. All the eye movements returned toward normal over the course of a few days to 2 wk. Histological examination revealed that the cannula tips were mainly located between the vestibular and prepositus hypoglossi nuclei, in their rostral 2 mm, bordered rostrally by the abducens nuclei. Dense gliosis clearly demarcated the cannula tracks, but for most injections there were no surrounding regions of neuronal loss. Thus the eye movement disorders were due to a reversible, not a permanent, lesion. 3. The time constant for the neural integrator was determined from the velocity of the centripetal drift of the eyes just after an eccentric saccade in total darkness. For intact animals this time constant was >20 s. Shortly after bilateral injections of neurotoxin, the time constant began to decrease and reached a minimum of 200 ms; every horizontal saccade was followed by a rapid centripetal drift with a time constant of ~200 ms. For vertical eye movements, in this acute phase, the time constant was ~2.5 s. 4. The vestibuloocular reflex (VOR) was drastically changed by the lesions. A step of constant head velocity in total darkness evoked a step change in eye position rather than in velocity. In the absence of the neural integrator, the step velocity command from the canal afferents was not integrated to produce a ramp of eye position (normal slow phases); rather this signal was relayed directly to the motoneurons and caused a step in eye position. The per- and postrotatory decay of the head velocity signal was decreased to 5-6 s indicating that vestibular velocity storage was also impaired. Loss of velocity storage, however, could be dissociated from integrator failure. 5. Optokinetic stimulation with a step in drum velocity also caused a step change in eye position. Just as with the VOR there was no integrator to generate a ramp eye-position signal from the step eye-velocity command. Optokinetic velocity storage was also damaged in a manner similar to that seen in the VOR. 6. Smooth pursuit was impaired by the neurotoxin injections. Within the first few hours after injection, the animals were unable to track small visual targets. With the assumption that the initial response to full-field stimulation with an optokinetic drum is due to pursuit, it was found again that a step of drum velocity produced a step in eye position. 7. These abnormalities developed simultaneously for each class of eye movement after every lesion and support the hypothesis that a final common integrator is used by all conjugate subsystems of the oculomotor system and that the integrator for horizontal eye movements is located in the region of the medial vestibular and prepositus hypoglossi nuclei.