Striatal Interneuron Contributions to Behavioral Control

About This Grant

Extensive work suggests the input nucleus of the basal ganglia, the striatum, is an essential site in filtering and amplifying behaviorally relevant inputs to shape motor output according to outcome. This filtering occurs via competitive control of striatal spiny projection neuron (SPN) activity by widespread excitatory projections from cortex, thalamus, and amygdala, which is reinforced by striatal dopaminergic signaling that encodes deviations from expected outcomes. Another understudied, yet potentially crucial circuit element in this initial input filtering are local striatal interneurons, which can be subdivided into subtypes dependent on synaptic connectivity, intrinsic properties, and molecular diversity. Despite their scarcity compared to SPNs, recent work suggests they can strongly modulate striatal SPN output and behavior. Our lab has previously discovered a role for striatal low threshold spiking interneurons (LTSIs), which target SPN dendrites, in controlling acquisition of early instrumental learning. However, functions outside of initial learning have not been investigated. Another striatal interneuron subtype, tyrosine hydroxylase-positive interneurons (THINs) - which despite their weak expression of this enzyme do not release dopamine – participate in local inhibitory control of LTSIs and SPNs in in vitro preparations and may maintain goal-oriented operant choice in vivo. Despite some recent successes, our understanding of how local striatal interneurons support complementary computational functions for motor control has been slowed by the absence of consistent behavioral frameworks in which to compare subtypes. Here we address this limitation by combining a modular, head-fixed value-based choice task with viral-genetic tools to systematically examine striatal LTSI and THIN contributions. We hypothesize that LTSIs and THINs encode and control distinct components of value-based choice via striatal interactions, including how choices are selected and executed. To test this, we first use population calcium imaging of each subtype together with behavioral modeling that captures latent value signals to understand what parts of value-based behavior each cell type represents. Next, we test the causal contributions of these subtype-specific activity patterns by using bi-directional optogenetic manipulations at two timepoints – in the pre-choice decision epoch and in the outcome epoch, when the results of actions are used to inform future choice. To map the synaptic connectivity of LTS and TH interneurons within the local striatal circuit, we use in vitro acute slices with optogenetic recruitment of specific interneurons together with paired subtype-specific SPN recordings. Finally, we combine in vivo optogenetic striatal interneuron manipulations together with extracellular recordings of SPNs to test the hypothesis that interneuron effects are in part determined by behavioral context dependent SPN interactions. Together, this proposal provides a comprehensive approach to understand how different striatal interneuron populations contribute to the filtering processes at the center of striatal control of motor behavior.