NUCLEUS ACCUMBENS PROCESSING OF REWARD-PREDICTIVE CUES

Project Summary When addicts encounter conditioned stimuli (CSs) associated with drugs, they often find themselves engaging in drug-seeking behavior, even after a period of abstinence. Understanding how CSs become able to promote drug- and reward-seeking behavior is thus of primary importance to understanding addiction and relapse. In this proposal, we focus on how conditioned approach behavior ? the conditioned locomotor response to the CS that often brings the subject closer to the predicted reward ? is learned. We will test three related hypotheses. First, previous studies show that NMDA receptor activation in the nucleus accumbens (NAc) is required for ac- quisition of the conditioned approach response, suggesting that synaptic plasticity within the NAc is responsible for this learning. Studies from our laboratory have shown that in animals that have already learned the response, many NAc neurons are excited by CSs, that the excitations precede approach movement initiation and predict its latency, and that they are causal to approach. Therefore, we hypothesize that synaptic plasticity within the NAc gives rise to CS-evoked excitations during task acquisition, and that this mechanism is causal to learning. Second, the NAc receives a prominent excitatory projection from the basolateral amygdala (BLA), which, in trained animals, is required for both conditioned approach behavior and for the CS-evoked excitations of NAc neurons. Previous studies suggest that many BLA neurons encode sensory salience in that they fire in response to prominent stimuli even before the animal has learned their predictive value. NAc neurons, on the other hand, do not begin to fire in response to CSs until the subject begins to learn the approach response to the CS. Therefore, we hypothesize that synaptic plasticity at the BLA-NAc synapse is required for emergence of the CS- evoked excitations of NAc neurons, and that this mechanism is causal to learning. Finally, NAc CS-evoked excitations also require dopamine, which is provided by the projection from the ventral tegmental area (VTA). Stimulation of dopamine neurons is strongly reinforcing, possibly because it facilitates the formation and maintenance of associations between stimuli and positive outcomes. We hypothesize that NAc cue-evoked excitations are a product of dopamine neuron-mediated plasticity, and that therefore stimulation of dopamine neurons is sufficient to maintain cue-evoked excitations (and hence cue-evoked approach behavior) on subsequent encounters with the cue. We will test these hypotheses with a unique and powerful combination of cutting-edge techniques. We will record the unit firing activity of neurons in the NAc and BLA throughout task acquisition learning, while simultaneously injecting an NMDA antagonist into the same structure from which we record. We will also use this method in combination with chemogenetic silencing of BLA terminals in the NAc by local microinjection of a DREADD agonist into the NAc. Finally, we will use optogenetic control over VTA dopamine neurons to examine how the VTA-NAc dopamine projection impacts the NAc neuronal activity that underlies learning.