Stability and flexibility of neural representations in the ventral striatum

The olfactory tubercle (OT) is a ventral striatal brain region implicated in reward processing and motivated behaviors including addiction. It receives olfactory sensory input, as well as strong dopaminergic innervation from the ventral tegmental area (VTA). Learning-induced changes in neural activity can occur rapidly and differentially in distinct types of neurons on the OT, but it is not clear how stable these changes in representations are. Recent studies in many brain regions have revealed a surprising instability in sensory representations, even when the animal’s behavior is stable. Unstable representations are hypothesized to potentially arise from ongoing synaptic plasticity, as might occur with ongoing activity or further learning. Whether such “representational drift” is present in striatal areas including the OT is unknown and will be a subject of inquiry in this proposal. The PIs will ask whether OT neural representation of previously learned associations is stable when new associations are learned, and whether the representations are altered when old associations are unlearned or degraded. The PIs will use deep brain multiphoton imaging at subcellular resolution to monitor activity in identified subtypes of OT neurons over weeks as mice learn and maintain odor-outcome associations. Using our collective expertise in behavior, physiology, and imaging, the PIs will tackle the following Aims. Aim 1: To determine whether neural representations of cue-outcome associations in the OT are stable over several days. The PIs will train mice to learn the arbitrarily assigned valence of several odors and track the activity of two cardinal types of neurons in the OT in behaving mice during and after task learning over multiple weeks. The PIs hypothesize that continuous experience and reinforcement will stabilize representations in the OT, which can degrade with pauses in experience. Aim 2: To determine whether and how existing representations in the OT change when new cue-outcome associations are learned. After mice have stably associated several odors with specific outcomes, the PIs will introduce new odors to be learned, reverse or extinguish familiar associations, as specific OT neural populations are tracked. The PIs hypothesize that new learning will lead to greater representational drift for older, familiar associations. Aim 3: To determine the role of dopaminergic signaling in maintaining stability and flexibility of representations in the OT. Whether continued dopamine signaling in the OT helps maintain stable representations through repetition or perturbs stability by promoting additional synaptic plasticity is unknown. The PIs will perturb dopaminergic signaling to the OT in mice that have learned cue-outcome associations and investigate how neural representations in the OT evolve. The PIs hypothesize that perturbing dopamine signaling will degrade representational stability even with continued experience. Our studies, in addition to answering questions about the function of the OT, will also shed light more broadly on exciting new ideas about continual learning.