Depth distribution of neuronal activity related to a visual reaction time task in the monkey prefrontal cortex

1. The depth distributions of neurons with changes in activity during a visual reaction time task were investigated in the dorsolateral prefrontal cortex of the macaque monkey, using glass micropipettes. The task was initiated by the monkey pressing a lever and consisted of an initial waiting phase (3.0-s period); a warning phase (green lamp, a variable period of 1.5-3.5 s); a lever-release GO phase (red lamp); and a final reward phase. The locations of neurons, in terms of the cortical layer, whose activities were recorded during performance of the task, were estimated histologically by marks made during the recording session. Marks were made by passing a DC current (anodal, 10-20 microA, 10-20 s) through the tip of an electrode which contained carbon fibers. Manipulator readings during the experiments and measurements of the distance of the marks from the cortical surface for 28 electrode penetrations showed a discrepancy of 2.0 +/- 5.0%, indicating that the depths at which task-related neurons were located could be estimated with errors of less than 10%. 2. Out of 162 task-related neurons recorded during 31 electrode penetrations, 53 showed changes in activity only during the warning phase (W-type; 19 phasic, 10 phasic-tonic, and 24 tonic), 37 showed changes only during the GO phase (GO-type; 4 Cue-coupled, 27 Intermediate, and 6 Movement-coupled), 34 showed changes during both the warning and GO phases (WG-type; 9 phasic, 10 phasic-tonic, and 15 tonic), and 38 showed changes during the reward phase (RE-type; 22 phasic and 16 tonic). 3. The various task-related neurons were distributed differently in different layers. Most neurons were recorded from layers III through V. In layer I, no neurons were recorded. In layer II, only a small number of neurons, with changes during the warning phase, were recorded (n = 7, 4%). One-third of the neurons were recorded in layer III (n = 51, 32%); neurons with changes during the warning phase were the most numerous (n = 24) and were significantly more numerous than neurons with changes associated with other phases of the task. One-fourth of the neurons were recorded in layer IV (n = 43, 27%); neurons with changes during the reward phase were the most numerous (n = 19), and were significantly more numerous than neurons with changes during both the warning and GO phases and also more numerous than neurons with changes during the GO phase.(ABSTRACT TRUNCATED AT 400 WORDS)