The Cytoplasmic LIM Domain Protein Espinas Contributes to Photoreceptor Layer Selection in the Visual System

SIMPLE SUMMARY: One of the central questions in neurobiology is how neurons discriminate between one another during circuit assembly. A common strategy of many nervous systems is the organization of brain regions in layers, to facilitate that neurons encounter a limited repertoire of synaptic partners. The fly visual system, which is structured in layers like many regions of the vertebrate brain, is used to identify cell surface molecules that mediate recognition between neurons and allow them to extend to specific layers. However, little is known about the intracellular pathways that link cell surface molecules to the cytoskeleton to determine whether or not to stabilize in a layer. Flamingo and its vertebrate homologs CELSR1/2 are cell surface molecules with widespread roles in neurite growth. In the fly visual system in particular, Flamingo regulates layer selection of a particular neuronal type. Our data suggests that in this context, Flamingo signals to the cytoskeleton through the conserved cytoplasmic molecule Espinas/PRICKLE2. Given that Flamingo and Espinas, as well as their respective vertebrate homologs, are broadly expressed in the nervous system, elucidating the interactions between them can reveal conserved mechanisms and provide valuable insights into the assembly of neural circuits. ABSTRACT: During circuit assembly it is essential that neurons connect with their specific synaptic partners. To facilitate this process, a common strategy in many organisms is the organization of brain regions, including the fly visual system, in layers and columns. The atypical-cadherin Flamingo (Fmi) and the receptor Golden Goal (Gogo) were proposed to regulate both the temporary and final layer selection of the R8 photoreceptor, through the cytoplasmic domain of Gogo. Our data suggests that Fmi intracellular signaling is also relevant for R8 final layer selection. The LIM-domain cytoplasmic molecule Espinas (Esn) binds Fmi, and they cooperatively control dendritic self-avoidance in sensory neurons. We observed defects in R8 layer selection in esn mutants with axons overshooting the final target layer, and we demonstrated that the LIM domain is necessary for layer selection. fmi knockdown in photoreceptors results in most R8 axons stalling at the temporary layer, however, we also detected R8 axons projecting past the final-target layer, and showed that fmi and esn genetically interact. Based on the previously described physical and genetic interactions between Fmi/Esn and the findings presented here, we propose that Esn signals downstream of Fmi to stabilize R8 axons in their final target layer.