Before we execute a given motor act, certain preparatory neural activity is thought to take place in areas that interact with the motor cortex. What are the characteristics of this activity and where does it occur? The basal ganglia — subcortical structures that are reciprocally connected with the motor cortex — have been a favourite focus of attention in relation to this question. Now, reporting in Journal of Neurophysiology, Lee and Assad studied the preparatory activity in the posterior putamen, comparing it during cued and self-initiated movements.

Using an elegant experimental design, the authors trained monkeys to guide a spot of light using a joystick under two different conditions. In some trials, they presented a visual cue that signalled the moment in which the monkeys had to start the movement; in others (self-timed movements), no cue was presented, but the monkeys still had to wait for a couple of seconds before moving the joystick. Lee and Assad found that the movement-related activity of neurons in the posterior putamen was similar for cued and self-timed movements; in both cases there was a marked increase in firing shortly before movement generation. But in the case of self-timed actions, this increase built up slowly, starting about 600 ms before movement onset. By contrast, this build-up was absent in cued trials, and activity simply rose sharply after presentation of the cue. Interestingly, in cued trials, the level of activity immediately before cue presentation correlated with the reaction time of the monkeys. In other words, the reaction time was shortest in trials in which the level of activity before the cue was highest.

The authors propose that exceeding a threshold level of activity in basal ganglia circuits might be necessary for movement initiation. In the case of self-timed movements, this threshold is slowly reached as a result of the activity build-up, whereas in the case of cued movements, the abrupt increase triggered by the cue is sufficient to reach such a threshold.

It will be interesting to explore whether similar patterns of activity occur in other basal ganglia nuclei and in other premotor regions. Similarly, using the experimental design of Lee and Assad to revisit the preparatory activity of the anterior putamen, which shows significant differences to the findings reported in this paper, might prove quite informative.