Regulation of Neuronal Dendrite Development and Migration by the Atypical Cadherin Fat3

Abstract

Neuronal shape and position are critical to the formation and function of neuronal circuits. Although neurons develop axons and dendrites cell-autonomously in vitro, how extracellular cues in vivo direct neurite specification and placement remains poorly understood. The role of atypical cadherin Fat3 in amacrine cell development illustrates how the same extracellular cue can guide both dendrite formation and migration. In the mouse retina, amacrine cells have a bipolar morphology during their migration. Upon reaching the nascent inner plexiform layer (IPL), they elaborate one neurite into the IPL and retract the other. Loss of Fat3 leads amacrine cells to develop an extra dendrite outside the IPL as well as errors in migration. We found Fat3 protein is concentrated at the IPL throughout amacrine cell development, suggesting Fat3 detects a directional signal. Here we investigated the signaling pathways upstream and downstream of Fat3 that mediate its role in amacrine cell development. In Drosophila, Fat’s ligand is Dachsous. Fat and Dachsous binding is modulated by the kinase Four-jointed. Our analysis of mutant retinas determined that Fat3 and vertebrate Four-jointed genetically interacted. However knockout studies of vertebrate Dachsous homologues suggested they are not relevant Fat3 ligands in the retina. Instead, analysis of retinas missing retinal ganglion cells suggested Fat3-mediated homophilic adhesion between amacrine cells may be important. Sparse loss of Fat3 from amacrine cells also led to extraneous neurites, suggesting Fat3 acts cell autonomously. Ex vivo live imaging revealed both migration and neurite dynamics were less directed in Fat3 mutant amacrine cells. We hypothesized Fat3 acts to target asymmetric localization of cytoskeletal regulators to the leading neurite. To identify downstream Fat3 effectors, we performed a pulldown assay using the Fat3 intracellular domain. We identified several cytoskeletal regulators as candidate binding partners. We focused on the actin regulators Ena/VASP and demonstrated a direct interaction with the Fat3 intracellular domain. In the retina, Ena/VASP localized with Fat3 to the IPL, and loss of Fat3 changed Ena/VASP distribution. Furthermore, forcing uniform membrane recruitment of Ena/VASP in developing amacrine cells phenocopied loss of Fat3. Together these results suggest Fat3 polarizes the activity of cytoskeletal effectors to help direct amacrine cell migration and dendrite placement.