It has been known that the mesolimbic dopamine system, mainly DA neurons in ventral tegemental area (VTA) and their efferent projections, is involved in the information processing related to natural and drug reward. Ungless et al. (2001) have shown that single exposure of cocaine can cause adaption of excitatory synaptic strength in the DA neurons of VTA. Furthermore, although nature reward such as sucrose also cause transient synaptic change in VTA DA neurons, the chronic cocaine induces persistent increases of excitatory synaptic strength in the VTA DA neurons (Chen et al., 2008). More importantly, the increased AMPA receptor-mediated fast synaptic transmission caused by acute or chronic cocaine treatment occludes the induction of LTP in vitro, which is considered as underlying cellular mechanism of learning and memory. On the other hand, Stuber et al. (2008) show that Pavlovian cue-reward association learning requires the involvement of the VTA and the learning experience can cause transient increase of the fast excitatory synaptic strength in VTA DA neurons. Furthermore, the increased synaptic strength occludes the induction of LTP in vitro. These studies raise the question whether the sustained increased of AMPA receptor transmission in VTA DA neurons may occlude the subsequent cue-reward association learning, which has been shown to involve VTA DA neurons. To better understand the role of synaptic plasticity in the midbrain DA neurons in the appetitive Pavlovian learning, I have been using single exposure of cocaine as a tool to increase the synaptic strength in VTA DA neurons and test how this treatment affects the performance of mice in cue-reward association learning. The result suggests that cocaine pre-exposure decreases the cue-approaching behavior. Due to the complex pharmacology effects of cocaine onto the whole brain, I plan to address the question more specifically by using optogenetic and mouse genetic approaches. The optogenetic approach is to inject a double-floxed AAV virus expressing channelrhodopsin (ChR2) to the VTA of the DAT-Cre mouse line to selectively express ChR2 in VTA DA neurons. With the implantation of optic fiber targeting to VTA, I can manipulate the neuronal activity of VTA DA neurons in vivo. Brown et al. (2010) have shown that activating VTA DA neurons optogenetically can induce similar potentiation of AMPAR transmission as single exposure of cocaine. I would like to adapt similar manipulation to test if the capability of the mice to learn cue-reward association task is affected after the saturation of the synaptic strength in VTA DA neurons by optogenetic manipulation. On the other hand, a mouse genetic approach will also be employed to answer the question. Thorase is an AAA+ ATPase and has recently been shown to controls AMPAR internalization, specific to GluR2-containing AMPARs. Knockout of Thorase has been shown to increase surface AMPA receptors and AMPAR currents (Zhang et al., 2011). Through the collaboration with Dr. Valina L. Dawsons laboratory, I am planning to generate Thorase conditional KO mice by crossing Thorase Flox/+ mice to DAT-Cre mouse to generate DA neurons-specific Thorase KO mice. Knockout of Thorase in DA neurons is supposed to increase surface AMPA receptor expression and thus can serve as another option of increasing AMPARs in DA neurons. Another question I would like to study is which afferent pathways to the VTA are activated and how they are modified during the appetitive learning. This question will be best answered with optogenetics approach. By injecting CaMKII-driven-ChR2 AAV to candidate brain regions projecting to the VTA, I will be able to isolate specific glutamatergic afferent pathway and in vitro slice electrophysiology will be used for quantitative analysis. The synaptic properties, including AMPA/NMDA ratio, RI and mini-EPSCs, of specific afferent pathway onto VTA DA neurons will be measured across the training sessions of appetitive reward learning. Once the critical afferent pathway and critical time window are identified, I will use CaMKII-driven-Halo 3.0 AAV to inhibit the identified pathway in vivo to understand the role of it during the appetitive learning.