ABSTRACT: The lateral habenula (LHb) impacts motivated behavior through dense direct and indirect projections to midbrain dopaminergic and serotonergic neurons. Some LHb neurons project directly to midbrain dopaminergic or serotonergic neurons; others project indirectly through one of multiple types of midbrain GABAergic neurons. Similar to dopaminergic neurons, the LHb encodes reward prediction error (RPE) - the discrepancy between expected and actual value – a powerful computation that guides learning from environmental feedback. However, there is a large amount of molecular and anatomical heterogeneity in the LHb, and it is unclear how reward encoding varies among LHb neurons with different projection targets, subnuclear location, and molecular profiles. Here we propose to develop methods that would link in vivo information encoding, projection target, subnuclear location, and molecular expression properties of individual LHb neurons – thus delineating functional, multimodal classes of LHb neurons to better understand the organization of the LHb. We hypothesize that a subset of functional classes of LHb neurons encodes RPE, and more generally, that distinct in vivo activity patterns map onto neurons with different projection targets, subnuclear location, and molecular profiles. In Aim 1, we will optimize our GRIN lens imaging methods to increase the number of LHb neurons that can be imaged simultaneously. We will also develop the use of a non-toxic rabies virus system for functionally imaging LHb neurons with different cell-type-specific midbrain projections in mice. In Aim 2, we will validate methods for identifying individual LHb neurons in brain slices that were imaged in vivo. These methods will allow us to specify the molecular class and subnuclear location of neurons that were functionally imaged in behaving mice, and together with Aim 1, link in vivo information encoding, projection target, subnuclear location, and molecular expression properties of individual LHb neurons. Moreover, the experiments proposed in Aim 2 would be proof- of-principle for other studies that aim to understand the function of different cell types (especially cell types that are defined by multiple molecular markers) – a critical step for making sense of the complexity of the nervous system.