The subthalamic nucleus was once regarded as no more than an inhibitory node in an indirect circuit of the basal ganglia. It came to prominence because it is a target for effective therapeutic electrical stimulation in Parkinson’s disease. This, and its anatomical prime position in receipt of direct cortical inputs, raises obvious questions about whether it does more than control tremors.
It has been suggested that STN is important for ‘holding the horses’ pending a go signal from cortex when response conflict has been resolved (see Frank et al. (2007) Science. 318: 1309–12). I will present evidence from two-forced choice reaction time performance in rats that supports this suggestion, but goes further in providing evidence that this role is not a passive ‘brake’. Rather, the STN actively inhibits response initiation pending internal response conflict resolution but does not contribute to inhibitory control when response choice is externally disambiguated.
Inactivation of the STN also results in apparent changes in attention and we have recently shown that these rats do not form an attentional set. It is not immediately obvious how the inhibition of response initiation (a motor control ‘braking’ process) relates to having an attentional set (a perceptual control ‘classifying’ process). While it is possible that these are two distinct deficits (and people have suggested that the STN has motor and cognitive parallel ‘circuits’), it is also possible (and perhaps parsimonious to assume) that they are not.