VISUALIZING SYNAPTIC COMPETITION: MOLECULAR AND STRUCTURAL REMODELING IN THE DEVELOPING VISUAL SYSTEM

Abstract

Throughout development, neural circuits undergo plasticity as synapses are formed and pruned by molecular and genetic programs in response to changing spontaneous and sensory-driven activity. However, the mechanisms by which activity shapes synaptic development remain poorly understood. In the developing visual system, proper wiring of eye-specific inputs is essential for binocular vision. Eye-specific segregation of the dorsal lateral geniculate nucleus (dLGN) is a model system for studying this process. At birth, the mouse dLGN receives overlapping inputs from retinal ganglion cells (RGCs) of both eyes, whereas in adults, contralateral inputs dominate, and ipsilateral projections occupy a small, distinct region with minimal overlap between eye-specific inputs. This eye-specific segregation occurs during the first postnatal week and is sensitive to changes in spontaneous retinal activity. The mechanisms driving activity-dependent refinement are difficult to study due to the diffraction limit of conventional light microscopy and challenges in identifying immature retinogeniculate synapses with electron microscopy (EM). To address these limitations, we optimized super-resolution microscopy and EM for characterizing synapses in neonatal mice and employed these techniques to study retinofugal synapse development.In Chapter 2, I present a detailed protocol for volumetric STochastic Optical Reconstruction Microscopy (STORM), including troubleshooting and recommendations. In Chapter 3, I combine STORM with mass spectrometry to examine synaptic development in the suprachiasmatic nucleus and show that increased presynaptic protein abundance after eye opening reflects a rise in synapse number, not presynaptic size or extrasynaptic protein levels. In Chapter 4, I present a new transgenic mouse line that expresses mitochondria-targeted dAPEX2 in ipsilaterally projecting RGCs, enabling the identification of retinogeniculate synapses in neonates by EM. Using volumetric STORM and EM, we found that during the first postnatal week, retinogeniculate boutons form multi-active-zone synapses, with the dominant eye developing more numerous and larger contacts. Genetic disruption of retinal activity reduces overall synapse formation and clustering, revealing eye-specific mechanisms of refinement. Finally, in Chapter 5, I pharmacologically disrupted retinal waves to bias eye-specific competition in the dLGN and found that axons from the “losing” eye grow exuberantly, likely seeking appropriate postsynaptic partners to stabilize their connections.