Neural Substrates of Choosing Actions and Motivational Drive, a Role for the Striatum

Abstract

Optimal decision making requires one to determine the best action among available alternatives as well as the most appropriate level of engagement for performance. While current research and models of decision making have largely focused on the former problem, or action selection, less is known about the latter problem of the selection of motivational drive. Thus, I designed a self-paced decision-making paradigm that aimed to dissociate both facets of selection in rats. First, I showed that the expected net value of potential options influenced rats' general motivation to perform: rats globally exhibited shorter latency to initiate trials in states of high net return than in states of low net return. In contrast, the relative value of options biased choice direction. To study the neural substrates underlying either process, I examined the role of the striatum, which is closely connected with cortex and dopamine neurons, acting as a major hub for reward-related information. In chapter 1, I show that selective lesions of the dorsomedial (DMS) but not ventral striatum (VS) impaired net value-dependent motivational drive but largely spared choice biases. Specifically, DMS lesions rendered animals' latency to initiate trials dependent on the absolute value of immediately preceding trial outcomes rather than on the net value of options. Accordingly, tetrode recordings in Chapter 2 showed that the DMS rather than VS predominantly encodes net value. In fact, net value representation in the DMS was stronger than either absolute or relative value representations during early trial epochs. Thus, the DMS flexibly encodes net expected return, which can guide the selection of motivational drive.