ABSTRACT / PROJECT SUMMARY Spatial navigation requires working memory for the ability to flexibly update an internal representation of position as one moves through the world, yet also stably hold “in mind” one’s position during periods of rest. Despite the critical importance of working memory for a wide range of cognitive processes, we currently lack basic understanding of how working memory circuits balance the fundamental tension between flexibility and stability. This gap is due to three major challenges: (1) defining a complete network that holds internal representations during working memory; (2) the ability to causally test how fluidly networks can transition between distinct representations; and (3) a conceptual framework for how transition probabilities are modulated at a biophysical level. This proposal will overcome these challenges by investigating how dopamine modulates the stability of internal spatial representations in a tractable experimental system: the central complex of the fruit fly, Drosophila. We have developed methods to measure how dopaminergic modulation shapes synaptic, cellular, and network dynamics of genetically identified neurons that code for spatial orientation. First, we will measure when dopamine modulates navigational circuits using whole-cell electrophysiology from the brains of flies walking in virtual reality. Then we will define how dopamine levels shape network dynamics by using optogenetics to explore how dopamine alters the ease of overwriting spatial representations. Finally, we will use cell-type specific perturbations of dopamine receptors with in vivo electrophysiology and calcium imaging to define how changes to synaptic and intrinsic properties shape network fluidity. The ultimate goal is a biophysical-level description of how neuromodulation shapes working memory processing online. Due to the difficulty of interpreting and perturbing population activity that is distributed across large mammalian brains, these experiments have been previously out of reach. By using Drosophila, we can focus on a compact navigational circuit comprised of only a few hundred neurons with known connectivity and unmatched genetic access. Although there are clear differences between flies and mammals, dopamine signaling and spatial coding properties (head direction networks) are strikingly conserved across species. These similarities argue that the principles we discover in the fruit fly will be relevant to cognitive processing in other animals. A mechanistic understanding of working memory fluidity is essential for the top-down design of therapeutic strategies to treat cognitive disorders.