Behavior of neurons in the abducens nucleus of the alert cat—II. Internuclear neurons

The activity of 43 antidromically identified abducens internuclear neurons with conduction velocities ranging from 14 to 54 m/s was analyzed in alert cats during spontaneous and vestibular induced eye movements. The discharge rate of internuclear neurons significantly increased with successive adducting positions of the contralateral eye. Slopes of rate-position (k) relationships ranged from 3.1 to 17.9 spikes/deg (mean 12.01 ± 3.1). Threshold ranged from −19° to +3°. Frequency saturation was never observed for any internuclear neuron within the oculomotor range. Although straight lines were selected to illustrate the rate-position relationships, exponential curves always provided the best statistical fit demonstrating that an enhancement in frequency potentiation (k) must accompany more eccentric fixations in the on direction. Internuclear neurons showed a low variability in firing rate (<3.0%) for fixations less than 1 s. Variability increased with both longer and repeated fixations of the same eye position. Discharge rates were found to depend upon both the direction of the preceding eye movement and the animal's level of alertness. Separate regression lines of rate-position relations following saccades in the on and off directions differed significantly in slope (100%), but not threshold. The observed static hysteresis in an identified non-motoneuron shows this property to be in a central neural circuit prior to the extraocular motoneuron. The slopes (k) of rate-position plots for all internuclear neurons decreased significantly (100%) when level of alertness changed from “alert” (1 ± 0.2saccades/s) to “drowsy” (0.5 ± 0.2saccades/s). Thresholds, however, were not significantly altered. Discharge rate of abducens internuclear neurons increased abruptly 10.4 ± 2.5 ms preceding saccades in the on direction, and decreased 20.5 ± 7.8ms before saccades in the off direction. Internuclear neuronal activity was not affected by pure vertical saccades. During on direction saccades, firing frequency did not saturate, but increased with velocity in a linear fashion. Exponential functions often fit the data better due to the difference in slopes of rate-velocity plots for on vs off direction saccades. Slopes (rs) of rate-velocity regression lines during spontaneous saccades ranged from 0.99 to 4.10 spikes/s/deg/s (mean 2.16 ± 0.93). During saccades in the off direction activity always decreased, but it seldom ceased. Rate-velocity regression lines measured during the fast phase of vestibular nystagmus (rsv = 2.09 ± 0.88) showed no significant differences from rs slopes in 82% of the cases. However, both rs and rsv values were significantly different from slow phase velocity during vestibular nystagmus (rv = 2.52 ± 1.1) in six out of 10 neurons. The intraburst firing rate during saccades showed a continuously changing profile similar to that described for extraocular motoneurons in the cat and monkey. Exponential-like changes in firing rate were observed near the end of on and off direction saccades. Compared to abducens motoneurons, internuclear neurons showed a more irregular firing rate, higher k and r values (1.4 and 1.9 times, respectively), lower thresholds for recruitment and higher firing frequency at a given eye position. Given the increases in k and r, time constants (Ts = rs/k) were also 1.4 times higher for internuclear neurons than abducens motoneurons. All parameters were predictably modified by the animal's level of alertness and exhibited a directional hysteresis. Discharge rate was correlated to vergence movements in 5/13 neurons. During monocular eye movements of the contralateral eye, internuclear neurons responded with k and r values not significantly different from those during conjugate eye movement. There was, however, no response for monocular movements of the ipsilateral eye. These data suggest differences in synaptic organization between abducens and medial rectus motoneurons during vergence and conjugate gaze. Sinusoidal rotation in the dark at 1 Hz produced a phase lead of the instantaneous firing frequency envelope with respect to eye position in the orbit of 42.6. The time constant (T0) calculated from the transfer function between firing frequency of internuclear neurons and contralateral eye position was found to be 152.9 ± 43.1 ms. This value significantly differed from those calculated from individual Ts = rs/k (174.9 ± 66.6), Tsv = rsv/k (169.1 ± 63.4) and Tv = rv/k (201.2 ± 67.0) quotients; but not from that described for abducens motoneurons (143.1 ± 35.1). Every internuclear neuron participated in all horizontal conjugate eye movements, except vergence. The relationship between eye position thresholds and the value (k) for rate-position relationships showed the same slope of 0.3 described for abducens motoneurons. These findings argue that the distribution of cell sizes may be ordered alike in both populations. By contrast, the relation between the coefficient k and conduction velocity exhibited a higher slope for abducens internuclear neurons than for motoneurons. This difference suggests that internuclear neurons, as a pool, are more excitable than motoneurons which correlates well with their different roles in horizontal eye movement.