Whole-brain voltage imaging in larval zebrafish using light-field microscopy

Abstract

Monitoring pan-neuronal electrical activity in awake vertebrates is a yet-to-be-achieved goal in neuroscience that will facilitate our efforts to understand how the brain processes information and mediates behaviors. Here we present Sparse Decomposition Light-Field Microscopy (SDLFM), a light-field microscopy method with improved resolution for volumetric neural activity imaging, and computational and experimental evaluations of its capacity for whole-brain voltage imaging in larval zebrafish. The work in this thesis includes three parts. First, we designed and implemented SDLFM and demonstrated its cellular-resolution ability in larval zebrafish expressing pan-neuronal GCaMP6. Second, we simulated the voltage imaging and the computation processes of SDLFM for zebrafish brains and found that in order to achieve a >0.9 correlation coefficient between the extracted SDLFM signals and the real voltage signals with our current hardware and voltage indicators, the proportion of the neurons randomly labeled with fluorescence in a zebrafish brain must be reduced to below 20%. Finally, we imaged live brains of zebrafish larvae expressing pan-neuronal Voltron525-ST and transient SomArchon at a volume rate of 307 Hz with a volumetric field-of-view of ~700 [mu]mx~400 [mu]mx~300 [mu]m. Although spontaneous spikes were traced in these recordings, we speculate some signals might result from noise and artifacts. In summary, we have developed SDLFM, which allows high-speed whole-brain imaging at cellular resolution in larval zebrafish. New camera sensors with higher quantum efficiency and larger dynamic range, and new voltage indicators with increased brightness and sensitivity are needed to achieve the signal-to-noise ratio required to record the voltage activity from a densely labeled fish brain.