PROJECT SUMMARY/ABSTRACT Because neurons integrate and process information via modulation of their membrane potential, the ability to monitor voltage is critical to understanding how single and groups of neurons compute. Genetically encoded voltage indicators (GEVIs) —fluorescent proteins that report voltage dynamics as changes in brightness— are emerging as a preferred recording method because they can track voltage transients with high spatiotemporal resolution and cell type specificity. Particularly sought after are GEVIs that perform well under two-photon (2P) microscopy, the method of choice for imaging neural activity in highly scattering tissue such as the rodent brain. We have recently demonstrated that a combination of the 2P optical recording method ULoVE and the indicator JEDI-2P enable sustained (> 30 min), fast (> 1 kHz), deep-tissue (< 400 µm) monitoring of voltage dynamics in individual neuronal somas in awake behaving mice. However, ULoVE is fundamentally unable to record from cells and structures located in different focal planes. This is a critical limitation as neuronal computations typically involve cells or neurites located at different depths. The goal of this proposal is to address this technology gap and enable three-dimensional optical voltage recordings in awake-behaving mice. We propose to optimize 3D-CASH, a new method that enables the three-dimensional recording of calcium dynamics but whose lower signal-to-noise ratio prevents reliable voltage recordings. We propose several complementary but independent approaches to improving the signal-to-noise ratio of voltage recordings, including hardware-based strategies for efficiently exciting cells/structures while minimizing motion artifacts and neuropil background fluorescence (Aim 1). We also propose a new generation of GEVIs that improve the detectability of subthreshold and spikes (Aim 2) and optimized methods for subcellular localization of GEVIs to increase the signal from specific structures of interest such as dendrites or somas while reducing background contamination (Aim 3). We anticipate that this project will produce improved GEVIs of general utility for neuroscience applications and a new optical approach that enables three-dimensional voltage recordings in vivo. These new technologies will allow the neuroscience community to ask questions that are currently technically infeasible, paving the way for a more detailed understanding of dendritic integration and neural network computations in living animals.