Relatively recent re-evaluations of the behavioral deficits of patients with basal ganglia dysfunction have revealed significant and striking cognitive impairments. These results were particularly surprising because for some time, the basal ganglia have been thought of as an extension of the motor control system. In an attempt to determine the precise cognitive consequence of basal ganglia (in particular striatum) compromise, animal researchers have focused on the contribution of the basal ganglia to learning and memory in rat and monkey. The rodent lesion literature suggests a selective role for the caudate stimulus-response or egocentric forms of learning (dissociating it from the context-based learning performed by rodent hippocampus), while the primate electrophysiological evidence supports a dual role for caudate in both stimulus-response and context-dependent learning. Here we present a potential resolution to the apparent discrepancy between response learning and contextual interpretations of the caudate nucleus. We propose that the caudate provides a response reference system to help define the goals of future actions that are deemed appropriate for the current context regardless of whether the specific task to be learned is stimulus-response in nature or involves the more flexible processing of context-dependent learning. In spatial learning, we argue that while the hippocampal sensory context system is important to allow the animal to organize sensory information in different ways, the caudate response referent system allows animals to quickly redefine response options as environmental demands change, the latter of which can play fundamental roles in many types of behavior. To distinguish between a strictly response theory and a response reference theory of caudate, the behavioral correlates of caudate single unit activity characterized as rats perform either a spatial context task, an egocentric response task, or a random forced choice task on a radial maze or a T-maze. Hippocampal place cell activity will be recorded simultaneously with the caudate activity to allow for direct analysis of the relative contributions of the two structures to different forms of learning. The possibility that dopamine regulates the learning-induced unit changes in caudate will be tested by looking at the effects of a dopamine neurotoxin unit-behavioral correlates. In this way, we not only address what unique role caudate has in different forms of learning, but also how the caudate may contribute to a larger neural circuit that allows organisms to ultimately have maximal flexibility in terms of sensory interpretations and response options.
Showing the most recent 10 out of 45 publications
