Experimental Progress Zebrafish visual behaviors: We have begun recording from large populations of neurons in the larval zebrafish brain during visually guided behaviors. We are focused on mapping the response properties of neurons to prey-like stimuli in our virtual reality environment using fish larvae expressing calcium indicators in all neurons in the brain. In parallel, we have almost finished acquiring a 3D electron microscopy dataset of the entire brain of a larval zebrafish. We have collected this volume at a resolution that will allow us to map the projections between brain regions. Neurogenesis in the olfactory bulb: We are in the midst of examining the connectivity between adult born neurons and projection neurons in the mouse olfactory bulb using a combination of optogenetic and electrophysiological approaches. In parallel, we have collected an electron microscopy volume from an olfactory bulb glomerulus and are validating that we can reliably trace neurons and identify synapses using our recently developed extracellular space preservation protocol. EM analysis of retinal circuitry: We collaborated with several research groups over the past year and contributed EM data and analysis to studies of a variety of circuits in the mouse retina. Collaborations with Dr. Fred Rieke and Dr. Joshua Singer resulted in two publications;a collaborative publication with Dr. Gautam Awatramani is under review. We are also in the process of preparing a manuscript following up on our previous work on the direction-selectivity circuit in collaboration with Dr. Rob Smith. Technical Progress Development of custom serial-block face microtome: The microtome design was completed and we are successfully using the instrument to collect large 3D electron microscopy datasets. 3D EM data analysis: We have pursued two avenues to aid the automated analysis of 3D EM data. On the experimental side, we conducted a study examining the benefits of preserving the extracellular space during chemical tissue fixation that is normally lost. We noted multiple benefits including: 1) increased penetration of stains and antibodies, 2) improved the detection of electrical synapses, and 3) significantly reduced error rates in automated segmentations of EM data. This work generated a manuscript that is submitted for publication. On the data analysis side, we have continued to improve the performance of neural networks we have implemented to automate reconstruction. We are in the process of segmenting large image stacks with these networks and are quantifying the error free path length along which we are able to automatically reconstruct axons and dendrites. Web based Proofreading Tool: We have begun using our web-based proofreading tool internally to both label training data for neural networks and to proofread automated segmentation results. We have demonstrated that the consensus among members of the lab leads to improved quality of training data compared to any single user. We are in the process of making this website externally accessible and will begin recruiting students to help with data analysis.
