UCSF

The retinotectal projection has long served as an experimentally accessible model for the study of topographic map formation and function. Retinal ganglion cells (RGCs) project to their midbrain target, the optic tectum, such that neighbor relationships are preserved between pre- and postsynaptic cells.

Organization of the map relies on first establishing positional information in the retina and tectum. Differentiating cells in the developing eye must be coded with positional information, giving them a molecular identity along the dorsal-ventral (DV), and anterior posterior (AP) axis. The DV relies on a cascade of secreted signalling factors which establish domains of expression for dorsal-specific and ventral-specific transcription factors. Through a behavioral forward genetic screen, we have identified a mutant, s327, with an aberrant retinotectal projection. The unusual projection arises as a result of defective retinal patterning, specifically lacking expression of the dorsal marker tbx5. Cloning of the disrupted gene in s327 mutants has identified radar as a necessary and sufficient component for dorsal retinal specification.

Along the anterior-posterior axis, this process relies on repellant interactions between ephrin-A ligands and EphA receptors. Chemorepellant interactions alone, however, cannot explain the observed map. It has therefore been suggested that competitive interactions between RGCs cause posterior projection of RGC axons, and are necessary for correct topographic map development. Using larval zebrafish and a novel experimental paradigm, I have tested the role of proposed axon-axon competition during retinotectal projection, and found that RGC-RGC interactions do not play a significant role in determining distal target location in the retina, but do affect axon arbor morphology by restricting arbor branch number and location.