Circuit-specified channels of the entopeduncular nucleus during conflict between reward-seeking and threat-avoidance

About This Grant

Project Summary Animals select actions based on behavioral context. The ability to appropriately select actions from a set of relevant options is dependent on the basal ganglia, which receive inputs at the striatum and produce outputs via the entopeduncular nucleus (EP) and substantia nigra pars reticulata (SNr). Foraging mice exploring novel environments are a robust model for action selection for the following reasons: 1) Water-deprived mice learn to suppress escape if doing so maximizes the chance for a water reward, an effect known as learned suppression of escape (LSE). 2) Threat and reward encoding dopamine neurons from the midbrain project to separate subregions of the striatum, the Tail of the Striatum (TS) and caudal Ventrolateral Striatum (cVLS) respectively. 3) Somatostatin(Sst)+ EP neurons signal the outcome of reward-seeking and threat-avoidant actions and receive partially overlapping but largely distinct input from TS and cVLS. 4) Sst+ EP neurons in turn project exclusively to the lateral habenula (LHb), a structure which encodes a negative reward prediction error and is implicated in learning of aversive action outcomes (Sst+ LHb-projecting EP neurons will throughout this proposal be referred to simply as EPLHb neurons). The striatum and EPLHb neurons are likely implicated in the shaping of future action selections through experience with threat and reward during behavior, but how these structures cooperate in this function has not been explored. Central Hypothesis: EPLHb neurons are topographically organized by monosynaptic TS or cVLS input into subcircuits which signal threat or reward signals and which are necessary for different aspects of LSE. Aim1: Map the strength and topographical organization of direct synaptic connections between striatum and EPLHb neurons. Unique subpopulations of EPLHb neurons receive projections from either threat-related TS or reward-related cVLS. It has not been confirmed whether EPLHb subpopulations receive monosynaptic GABAergic input from TS and cVLS as canonical basal ganglia models would suggest. In Aim1 I will use optical excitation of TS/cVLS axons coupled with ex-vivo patch clamp electrophysiology of EPLHb neurons throughout the EPLHb population to test for the presence, strength, density, and spatial locations of monosynaptic GABAergic inputs to EPLHb from the TS/cVLS. Aim2: Determine the activity patterns and necessity of striatal input-defined EPLHb neurons in guiding threat/reward conflict behavior. Silencing EPLHb neurons with tetanus toxin accelerates LSE, confirming expectations that these neurons shape action selection under threat/reward conflict conditions. EPLHb neurons have traditionally been studied in vivo using methods which do not distinguish between neuron subpopulations identified by input source. In Aim2a I will leverage fiber photometry of TSEP or cVLSEP neurons in vivo to distinguish their signaling of threat and reward as mice undergo LSE. In Aim2b I will silence synaptic release from TSEP or cVLSEP neurons to determine whether these are uniquely necessary for typical LSE.