Molecular and circuit mechanisms of learning supported by heterogeneous dopaminergic neurons

Project Summary/Abstract Dopaminergic neurons (DANs) are a molecularly, anatomically and functionally heterogeneous neuron group that is essential for learning across animal phyla. In the midbrain, distinct populations of DANs are responsible for memories with different valence or stability. Thus, the dopamine system comprises parallel subsystems, each of which operates as a qualitatively distinct learning system. This raises two important questions: 1. How does the heterogeneity of DANs impact synaptic plasticity to form distinct types of memories in each subsystem? 2. How are the signals from parallel subsystems integrated to ultimately trigger a unified behavior? Answers to these questions are required to understand the logic that governs the parallel memory systems. The mushroom body (MB), the major associative learning center in the Drosophila brain, is an excellent model to tackle these questions because it comprises dopamine subsystems, each of which is clearly defined by a unique set of DANs and MB output neurons (MBONs). These individual MB compartments support distinct types of memories that vary in valence and stability, properties shared with mammalian dopamine subsystems. However, in both invertebrate and vertebrate brains, it remains an open question whether the diversity of memory properties is derived from intrinsic characteristics of DANs or from an extrinsic circuit architecture. Aim 1 will test the hypothesis that combinations of DAN cotransmitters define compartment-specific rules of synaptic plasticity and thereby determine the memory properties. By identifying novel DAN cotransmitters and their physiological and behavioral roles, the causal relationship between plasticity rules and memory properties will be tested. In Aim 2, integration mechanisms of different types of memories will be determined by identifying neurons that pool input from multiple MBONs. Synaptic integration, behavioral roles and activity changes after learning will be determined in these integrator neurons. In this project, cell-type-specific transcriptome and the comprehensive connectome data available in the field will guide our molecular and circuit interrogation by in vivo electrophysiology, calcium imaging and behavioral assays. Collectively, this project will address fundamental questions regarding the heterogeneous organization of the dopamine systems and pioneer the circuit motif that is currently inaccessible in vertebrates.