Transformation of value in the basal ganglia

Northwestern University At Chicago, Evanston IL

Investigators

Abstract

PROJECT SUMMARY/ABSTRACT Foraging is fundamental to the survival of all animals and requires choosing which resources to pursue and which to forgo. These choices involve integrating information about the costs and benefits associated with each resource and are affected by the overall abundance of the foraging environment. Despite the importance of this behavior, how cost and benefit information are signaled in the brain to affect choice remains poorly understood. Understanding the neural circuit computations underlying foraging requires methods for manipulating and monitoring neural activity with temporal and cell type specificity. Many such tools are available in mice, but tasks typically used with mice are not well-suited to disentangle how cost and benefit are encoded in the brain. Thus, we developed a new foraging task for mice in which they make sequential decisions to pursue or reject offers with costs (time they must hold a nose poke) and benefits (reward size) that vary independently across trials. Additionally, the quality of the foraging environment changes in blocks, which changes the choices mice make without changing the features of the offers. This allows us to dissociate value and choice, which are often confounded. These features will allow us to disambiguate possible value encoding schemes in the brain. We will test a model of the basal ganglia whereby the direct pathway computes information in favor of performing an action (e.g., benefit) while the indirect pathway computes information against performing an action (e.g., cost). This information converges in an output nucleus of the basal ganglia, the substantia nigra pars reticulata (SNr). Whichever pathway has accrued more evidence biases the level of activity in the SNr, which then drives acceptance or rejection of the offer via its projections to the thalamus and brainstem. Finally, we hypothesize that dopamine release in the striatum signals whether an offer is better or worse than the average reward rate (i.e., environmental value). Thus, dopamine release to an intermediate value offer would be larger in a scarce environment. To test this model, we will image activity in the two major cell types in the striatum, direct and indirect pathway spiny projection neurons, to test the hypothesis that they separately encode the benefit and cost of the offers. Additionally, we will optogenetically inhibit these neurons to test the hypothesis that direct pathway inhibition decreases the relationship between offer benefit and choice while inhibition of the indirect pathway decreases the relationship between offer cost and choice. In Aim 2, we will image activity in the SNr. Here, we expect neural activity to reflect both cost and benefit and to be better correlated with choice than value. In Aim 3 we will use fiber photometry to measure dopamine release in the striatum to determine how it is affected by offer and environmental value. Together, these experiments will provide substantial insight into how cost- benefit decision making is supported by the basal ganglia by revealing multi-area circuit mechanisms regulating foraging behavior.