Trans-Axonal Signaling and Activity-Dependent Mechanisms Of Topographic Mapping and Refinement in Vivo

Brain connectivity and function depend on the precise formation of neuronal connections during development. In the central nervous system, most axonal projections are organized into topographic maps according to the spatial organization of the neurons they originate from or the type of stimulus they respond to. Topographic mapping is in large part established by interactions between axons and their target as well as by activity-dependent pruning of mistargeted or excessive arbors. Another important mechanism contributing to topographic map formation is pre-target axon sorting, where axons become pre-ordered en route to their destination. Both topographic mapping and pre-target sorting involve the pruning and refinement of mistargeted axons and arbors, however the mechanisms underlying these processes have yet to be elucidated. In the visual system, retinal axons are topographically sorted along the dorso-ventral axis in the optic tract before reaching the optic tectum. Previous studies in zebrafish have shown that optic tract sorting is achieved through the selective degeneration of missorted dorsal axons that have erroneously misrouted along the dorsal branch of the tract. Heparan sulfate (HS), a type of sugar chains carried by core proteins known as heparan sulfate proteoglycans (HSPGs), acts non-cell-autonomously along ventral axons to regulate the degeneration of these missorted dorsal axons. How HSPGs act to regulate pre-target sorting and the pruning of missorted axons still remains unknown. Here, we have identified the HSPG Glypican-3 (Gpc3) as specifically expressed in ventral retinal ganglion cells (RGCs) throughout development. Using CRISPR/Cas9 genome editing, we have generated several gpc3 mutant alleles encoding a truncated, non-functional protein. Analysis of retinal axon sorting in gpc3 mutants reveals that some dorsal retinal axons are missorted along the dorsal branch of the optic tract, demonstrating a novel function for Gpc3 in axon-axon interactions. Interestingly, Gpc3 seems to genetically interact with Tenm3, a transmembrane protein that is also expressed in the ventral retina. Tenm3 mutants also display missorted dorsal retinal axons as do embryos that are heterozygous for both gpc3 and tenm3. Overall, our study unravels a novel function for both Gpc3 and Tenm3 in trans-axonal signaling and developmental axon pruning during neural circuit formation. The refinement of mistargeted axonal arbors is another crucial step for topographic mapping and maturation. Unfortunately, our understanding of topographic map formation and refinement has been limited by our inability to observe these mechanisms directly in vivo. To overcome that challenge, we used Cre-mediated recombination of a new colorswitch reporter to generate a transgenic model that allows for the analysis of retinotopic map formation and maturation in vivo. We found that while the antero-posterior retinotopic map forms early in development, it remains dynamic with nasal and temporal retinal arbors expanding their projection domains over time. While temporal retinal axons seem to arborize directly at their proper target in the anterior tectum, nasal projections initially arborize in the anterior tectum as well as the posterior tectum. Over time, mistargeted nasal projections in the anterior tectum refine in an activity-dependent process, driving the sharpening of the antero-posterior map. With our novel line, we provide the first unbiased and quantitative analysis of topographic mapping and refinement in real time in vivo. Altogether, our studies unraveled novel mechanisms of axonal pruning during neural circuit development and provide a solid platform for futures studies to determine the signaling pathways and complex mechanisms governing these processes