These data indicate that Rich regulates CadN in a cell type specific manner. Loss of CadN leads to defects in cartridge formation in lamina and mistargeting of R7 cells ( Lee et al., 2001), similar to the defects we observed in rich and Rab6 mutants, suggesting that Rich and Rab6 function in a common pathway to regulate CadN trafficking. To test this, we removed one copy of CadN from eyFLP; rich1 or eyFLP; rich2 mutant animals. Loss of one copy of CadN greatly enhanced the targeting phenotype of the hypomorphic allele rich2 but not the of the null allele rich1, providing further evidence that rich and CadN function in a common pathway

( Figures 8I–8L). Moreover, homozygous double mutants for rich and CadN exhibit Selleck MK 2206 very similar phenotypes to CadN mutants. The data therefore indicate that Rich and Rab6 regulate CadN trafficking to affect axon target selection in the eye. To assess the specificity of this genetic interaction we also tested weather rich interacted genetically with other genes, including DLAR, liprin

α, or Jeb. We did not observe any interactions between rich and DLAR, liprin α, or Jeb ( Figure S6B), Small Molecule Compound Library consistent with our previous data. CadN is broadly expressed in the fly CNS and also plays important roles in determining synaptic specificity of olfactory receptor neurons (ORNs) (Hummel and Zipursky, 2004 and Zhu and Luo, 2004). To test whether rich and Rab6 mutants exhibit similar phenotypes in other neurons we focused on the ORNs. In Drosophila, around 1500 ORNs are present in the antenna and the maxillary palps. The ORNs send their axons into the antennal lobe (AL), where they form around 50 highly organized neuropilar structures, the glomeruli ( Laissue et al., 1999). The axons of

ORNs in CadN mutant typically target the appropriate region of the AL but fail to converge on a single glomerulus and instead spread out on the surface of different glomeruli ( Hummel and Zipursky, 2004 and Zhu and Luo, 2004). We generated mosaic flies with rich or Rab6 mutant ORN and wild-type AL targets using the MARCM system. eyFLP was used to induce mitotic recombination in the ORN progenitor cells but not their targets ( Hummel and Zipursky, 2004). To distinguish different subclasses of ORNs, different olfactory receptor Gal4s were used to label the mutant ORNs. We tested three different subclasses of ORNs, including two from the crotamiton antenna (OR22a, Or47b) and one from maxillary palps (Or46a). These three ORNs were previously shown to require CadN to establish proper connections with their targets ( Hummel and Zipursky, 2004). In both rich and Rab6 mutants, the AL Or47b and Or46a neurons fail to converge their axons into a single glomerulus, very similar to CadN mutants, indicating that rich, Rab6, and CadN regulate a common process in the antenna ( Figure S8C). However, the Or22a neurons do not have any obvious defects when rich or Rab6 is lost ( Figure S8C), in contrast to CadN mutants.