UCSF

Humans are often described as visual creatures as we heavily rely on our vision to perceive and navigate the world around us. Proper visual system function requires that each individual neuron to precisely identify and connect to its respective synaptic partners out of the billions of other neurons in the nervous system. Here, I will present my thesis work using the mouse visual system to understand the developmental organization of synaptic networks and the function of the neural circuits that underlie sight. First, I mapped circuits formed from retinal ganglion cells (RGCs) to the superior colliculus (SC) using a novel sequencing-based circuit discovery technique. The technique, named Trans-Seq, is able to anterogradely map neural circuits from genetically defined starter neurons. I used Trans-Seq to classify SC neuron types innervated by genetically defined RGC types. Alpha-RGCs (αRGCs), but not On-Off Direction-Selective Ganglion Cells (ooDSGCs), were found to connect to Nephronectin-positive wide-field neurons (NPWFs). This connection was validated using genetic labeling, electrophysiology, and retrograde tracing. I then identified the extracellular matrix protein Nephronectin as a determinant for selective synaptic choice from αRGC to NPWFs via binding to Integrin-α8β1. Experiments deleting Npnt in the SC and Itgα8 in RGCs suggested that Itgα8β1 expressed by αRGCs mediates proper αRGC axonal targeting and synapse formation by binding to Npnt deposited in the SC. Next, to generate quantitative molecular circuit maps of retinotectal connectivity, I have built upon Trans-Seq and superimposed anterograde tracing information on top of transcript information from multiplexed error-robust fluorescence in situ hybridization (MERFISH). I have established a limited retinotectal map linking genetically defined RGC types to SC neuron types by revealing the spatial connectivity patterns across multiple retinotectal pairs. Gene imputation additionally revealed rich information surrounding the molecular cues responsible for forming these circuits, offering candidates to test for future mechanistic inquiries. Taken together, these data reveal the architecture of retinotectal circuits and how molecular cues ensure wiring specificity in the visual system.