Research in this laboratory is devoted to understand the molecular mechanisms that direct neuronal connections during development, using the Drosophila visual system as a model. We currently focus on the first order connectivity, specifically the connections made by the R7 class of photoreceptor neurons to their brain target neurons. We utilize genetic analysis coupled with confocal microscopy imaging and behavioral assays to explore the mechanisms by which cadherin-mediated adhesion regulates the formation of specific connections. Fly vision is mediated by three classes of photoreceptor neurons (R-cells), each of which responds to a specific spectrum of light and connects to a specific layer in the brain. This wiring principle, named layer-specific targeting, is commonly observed in complex nervous systems, including vertebrate brains. This wiring pattern might facilitate information processing. We have chosen to study the connections made by R7 class of photoreceptor neurons to the M6 layer in the medulla ganglion because (a) they are essential for visual behavior, (b) they can be easily visualized by histology and be manipulated by genetic means. Using the UV/VIS choice test, we have previously identified 5 loci that affect R7 connectivity. These include N-cadherin, receptor phosphatase LAR, milton and two novel loci. Mosaic analyses indicated that these genes function cell-autonomously in R7s (presynaptic termini) as removing any of them in R7 neurons results in targeting defects. Comparing the expressivity of different mutant phenotypes suggests that N-cadherin and LAR are the major determinants of R7 target selection. To gain insight into the mechanism of action of N-cadherin and LAR, we performed developmental analysis coupled with high-resolution image analysis. We found that N-cadherin is required for initial targeting of R7 axons. Removing N-cadherin in single R7 axons results in approximately 15% of mutant R7 axons failing to project into the presumptive R7-receipent layer. While the mutant growth cones may reach the R8-layer (or further into the intermediate layer) and extend their filopodia into the presumptive R7-reciepent layer, the growth cone proper fail to move into the correct layer as seen in the wild-type. We propose that N-cadherin mutant R7 growth cones are able to sample, but fail to recognize their targets. Alternatively, N-cadherin might be involved in translating target-derived signal into directional movement of growth cone proper. By 25% pupal development, about 50% of R7 axons retract out of R7-receipent layer, suggesting that N-cadherin plays a role in the maintenance of R7-target interaction. Furthermore, the N-cadherin mutant R7 growth cones exhibit various morphological defects. They fail to fully expand once reaching the correct layer, and some expand prematurely before reaching the target layer. We speculate that N-cadherin-mediated adhesion induces the remodeling of actin-based cytoskeleton which in turns, modulates growth cone morphology. In contrast, the initial targeting of LAR mutant R7 axons is normal. No growth cones morphological defect was observed at the early stage. The LAR mutants R7 growth cones remain in the R7-recipient layer up to 25% pupal development. We have previously shown that the LAR mutant R7 axons retract at the later stage (~35% pupal development). Thus, LAR is required for stabilizing the interaction between R7 growth cones and their targets but not for initial targeting. Removing both N-cadherin and LAR in R7 axons results in the initial targeting defects seen in N-cadherin mutants. In addition, N-cadherin and LAR double mutant R7s show higher expressivity of targeting defects at adult stage than that of N-cadherin or LAR single mutants. Based on differential onset of mutant phenotype and the enhanced phenotype in the double mutants, we suggest that N-cadherin and LAR have distinct functions in R7 target selection. We are currently analyzing the other loci that affect R7-dependent behavior. We found that three loci, including one known gene, milton, affect the morphology of R7 synaptic termini. Milton is associated with kinesin and its function is required for axonal transport of mitochondria to synapses. The role of mitochondria in neuronal target selection is not known and is a subject of future study. We are currently performing developmental analysis on milton and other mutants to understand their roles in R7 target selection.
Showing the most recent 10 out of 11 publications
