Selecting appropriate actions to optimize outcomes is crucial for animals’ survival in their environment. In this project, we are particularly interested in positive and negative reinforcement strategies, which are examples of the action selection processes based on different values of the outcomes where animals select an action to either obtain a positive outcome or avoid a negative outcome.

In previous studies, various brain regions have been shown to be possibly involved in such processes through encoding the values of outcomes (the orbitofrontal cortex, basolateral amygdala), and also action selection to optimize outcomes (the medial prefrontal cortex, Nucleus Accumbens). Thus, in this project, we are interested in which cortical and subcortical brain areas participate in the positive and negative reinforcement strategies. To achieve our research goal, we have modeled the process of positive and negative reinforcement using an operant behavioral control task. In the task, the animals are guided by two different tonal cues: when hearing the 3kHz tonal cue, the animals need to press the left lever to obtain a sucrose reward (positive reinforcement); when hearing the 11kHz tonal cue, the animals need to press the right lever to avoid an electric footshock (negative reinforcement). In later phases of this task, intermediate tones between 3kHz and 11kHz are played to create a scenario where the animals need to discriminate and choose between positive and negative reinforcement strategies.  

In this project, we have also applied a technique known as TRAP (Targeted Recombination in Active Populations), which allows for permanent labeling of active neuronal populations through immediate early gene (IEG) Fos in FosTRAP / Ai14 mice. With FosTRAP, we labeled the active neuronal populations across the whole brain under 4 different conditions respectively: 1) Positive reinforcement, 2) Negative reinforcement 3)Discrimination between positive and negative reinforcement strategies 4)Homecage (control). We then conducted 3D reconstruction of mouse brain and cell quantification in different brain regions to compare their activity.

Our preliminary cell quantification results show tendencies in areas such as Dorsal Striatum (Dors Str), Globus Pallidus Internal (GPi), Medial Septal Nucleus (Med Sept Nucl), Lateral Habenula (LH), as well as lateral (LA) and posterior amygdala (PA) to be more involved in negative reinforcement. Tha- lamic regions show tendencies in being more involved in discriminating between positive and negative reinforcement strategies. Cortical regions such as the Prelimbic (PreL) and Infralimbic (IL) cortex as well as subcortical regions such as Nucleus Accumbens (NAc), Pallidum and Lateral Septal Nuclues (Lat Sept Nucl) demonstrate tendencies in their involvement in both positive reinforcement and discrimination. Cortical regions such as the Orbitofrontal Cortex (OFC), Anterior Cingulate Cortex (ACC) and Insula tend to be more involved in both negative rein- forcement and discrimination. The hippocampal regions as well as the basolateral amygdala (BLA) and basomedial amygdala (BMA) do not show tendencies in their involvement among positive reinforcement, negative reinforcement, and discrimination. Overall, this project will help us reveal potential target areas in the brain that are involved in action selection, which provides a basis for further manipulation or recordings of neuronal activities.