The locus coeruleus (LC), a small brainstem nucleus, is the primary source of the neuromodulator norepinephrine (NE) in the brain. LC receives input from widespread brain regions and projects throughout the forebrain, brainstem, cerebellum, and spinal cord. LC neurons release NE tonically to regulate baseline arousal, and phasically in the context of a variety of sensory-motor and behavioral functions. However, despite its brain- wide effects, the conditions under which LC-NE neurons are phasically activated and the modes of NE action during behavior are poorly understood. One prevailing theory suggests that NE acts to control the gain of output circuits, thereby modulating task performance by enhancing or dampening responses to stimuli. However, another theory suggests that NE release in cortical output regions acts to reset network activity, enabling task- switching or learning of new rules. Neither of these theories adequately explains the many observed roles of the LC-NE system in learning and behavior. We propose a new hypothesis of LC function, that spatiotemporal dynamics and modular circuits enable dissociated roles for the LC in behavioral execution and reinforcement during learned behaviors. Here, we propose to examine multiple features of this hypothesis using innovative approaches combining optically-tagged recordings of specific neuronal populations, advanced 2-photon imaging of identified neurons and axons, optogenetic manipulation of LC neurons and subpopulations, and computational approaches to define encoding of task variables by neurons. In Aim 1, we will record and manipulate the activity of LC neurons in mice performing an instrumentally conditioned task in which they detect auditory tones of variable intensity, execute a response, and receive positive or negative reinforcement. Using targeted recordings from LC-NE neurons as well as newly discovered LC-GABA neurons, we will examine the hypothesis that subsets of LC-NE and LC-GABA neurons encode task execution signals or reinforcement signals. Using this information, we will use cell-type specific optogenetics to activate or inhibit LC-NE or LC-GABA neurons during specific task epochs while measuring the effects on behavior. In Aim 2, we will assess anatomical modularity of LC projections to motor cortex or the prefrontal cortex (PFC), and examine the hypothesis that neurons with execution or reinforcement responses project preferentially to motor cortex or PFC. Subsequently, we will modulate the activity of LC neurons projecting to these targets and measure the effects on execution and learning. In Aim 3, we will examine the hypothesis that differential integration of NE release in motor cortex and PFC facilitates task execution and learning, respectively. We will monitor the fast kinetics of NE release in motor cortex and PFC using a genetically encoded NE sensor, and measure the impact on behavior of silencing NE activity in these cortical targets using optogenetic silencin...