Functional properties of corticotectal neurons in the monkey's frontal eye field

We examined the activity of identified corticotectal neurons in the frontal eye field of awake behaving rhesus monkeys (Macaca mulatta). Corticotectal neurons were antidromically excited using biphasic current pulses passed through monopolar microelectrodes within the superior collicullus. The activity of single corticotectal neurons was studied while the monkey performed behavioral tasks designed to test the relation of the neuron's discharge to visual and oculomotor events. Fifty-one frontal eye field corticotectal neurons were examined in two monkeys. Current thresholds for antidromic excitation ranged from 6 to 1,200 μA, with a mean of 330 αA. Antidromic latencies ranged from 1.2 to 6.0 ms, with a mean of 2.25 ms. Fifty-three percent of the identified corticotectal neurons were classified as having movement-related activity. They had little or no response to visual stimuli, but very strong activity before both visually guided and memory-guided saccades. An additional 6% of corticotectal neurons had visuomovement activity, combining both a visual- and a saccade-related response. In each case, visuomovement neurons antidromically excited from the superior colliculus had movement-related activity, which was much stronger than the visual component of their response. Twenty-two percent of the corticotectal neurons were primarily responsive to visual stimulation of the fovea. These included both neurons responding to the onset and neurons responding to the disappearance of a light flashed on the fovea. The remaining 20% of the corticotectal neurons were a heterogeneous group whose activity could not be classified as movement, visuomovement or foveal. Their responses included postsaccadic, anticipatory, and reward-related activity, as well as activity modulated during certain directions of smooth-pursuit eye movement. One neuron was unresponsive during all of the behavioral tasks used. There were no corticotectal neurons that could be classified as primarily responsive to peripheral visual stimuli. Histological reconstructions of electrode penetrations localized corticotectal neurons to layer V of the frontal eye field. For 22 corticotectal neurons tested, each had its minimum threshold for antidromic excitation within the superior colliculus, as judged by either histological confirmation, or surrounding neuronal responses recorded through the stimulation microelectrode. The majority of these neurons had minimum threshold sites within the intermediate layers; a few minimum threshold sites were located within the superficial or deep collicular layers. The lowest thresholds for antidromic excitation were obtained when the optimal saccade vectors associated with the frontal eye field recording and collicular stimulation sites were closely matched. There was a strong correlation between a measure of the difference between saccades associated with recording and stimulation sites and the log of threshold for antidromic excitation. This relationship was such that small increases in the vector difference between frontal eye field and collicular saccades were accompanied by large increases in threshold. In comparison to the entire population of frontal eye field neurons examined by Bruce and Goldberg, we conclude that there is a selective enrichment within the population of corticotectal projection neurons for neurons with eye movement-related activity and neurons with foveal visual activity, and a paucity of neurons with peripheral visual receptive fields and postsaccadic activity. These data suggest that the visual activity prevalent within the frontal eye field is likely to help generate the activity of movement-related neurons, but it is not a source for visual activity within the superior colliculus. The frontal eye field's projection to the superior colliculus provides several messages relevant for oculomotor performance, including information pertinent to the maintenance and release of fixation, and targeting information regarding an intended saccade.