[PhD Thesis Presentation_Zoom]-Masakazu Taira- Investigation of serotonergic regulation of reward-based behaviors
Presenter: Masakazu Taira
Supervisor: Kenji Doya
Title: Investigation of serotonergic regulation of reward-based behaviors
Serotonin is an important neuromodulator in reward-driven learning and decision making. Particularly, serotonin neurons in the dorsal raphe nucleus (DRN) send dense projections throughout the brain and its implication in reward-based behaviors has been examined using various types of behavioral tasks. However, how DRN serotonin affects the computational processes of decision making remains unclear. Reinforcement learning is a theoretical framework to describe decision making process. Previous study based on the RL framework proposed hypotheses on the role of serotonin in decision making, such as temporal discounting and model-based value computation. The overall aim of this thesis is to behaviorally examine these hypotheses to understand the role of serotonin in reward-based behaviors.
Particularly I examined two hypotheses. First hypothesis is that in serotonin would control relative importance of future rewards. Previous behavioral studies showed that serotonin activation/suppression enhance/decrease patience to wait for future rewards. However, how serotonin regulate patience to act for future rewards remains unknown. In the first part of my thesis, I trained mice to perform a free-operant lever-pressing task, in which motor action rather than stationary waiting is required, and tested the effect of optogenetic activation and inhibition of DRN serotonin neurons on sustained motor actions for future rewards. I found that optogenetic manipulation of DRN serotonin neurons did affect neither persistent motor actions for future rewards nor response vigor, suggesting different role of DRN serotonin neurons in motor actions for future rewards compared to stationary waiting.
Second hypothesis is that serotonin would modulate model-based decision making. In model-based decision making, agents use their own internal model of action-outcome relationship to plan forward and select actions. Previous computational studies proposed the regulation of model-based decision making by serotonin neurons. However, direct behavioral evidence of how serotonin regulate the process is still limited. The two-step decision making task is an established behavioral task to understand model-based decision making not only in human subject but also in rodents. In the second half of my thesis, I trained mice to perform the two step decision making task and tested how optogenetic inhibition of serotonin neurons affect mice model-based decision making. I found the tendency that optogenetic inhibition of DRN serotonin neurons decreased the weight of model-based decision making in action selection.
These results revealed underlying computational role of DRN serotonin neurons regulating reward-based behaviors.