However, their functions have not been studied in these species. The protein encoded by CG9063 is predicted to have WD40 domains as well as a RIC1 domain. The RIC1 domain, but not the WD40 domain, selleck chemical is present in a yeast protein RIC1p that interacts with Rgp1p. Together they form a GEF complex for Ypt6p, a homolog of Rab6 in yeast ( Siniossoglou et al., 2000).

Since the RIC1 domain is highly conserved across species and the protein encoded by CG9063 is the only protein containing a RIC1 domain in the fly, we named CG9063 rich, for ric1 homolog. We renamed 3L61 as rich1 and 3L62 as rich2. A genomic fragment that only contains rich rescues both the lethality and the targeting phenotypes ( Figure 1) of both alleles, indicating that rich is the gene responsible for the mutant phenotypes. To assess the expression pattern of Rich, we generated antibodies

against the N and C termini of the protein. As all antibodies performed poorly in all assays, we generated a genomic rescue construct with a triple HA tag inserted at the N terminus and introduced it into the rich1 mutant background. This construct rescues both the lethality and targeting phenotypes of rich1 mutants, indicating that the tagged protein is functional. At the third-instar larval stage ( Figures 4A and 4B), Rich is detected in most cells of the optic lobes. The enrichment in the lamina and ABT-199 cost medulla neuropil is consistent with its role in Astemizole PR cell axon targeting. At the pupal stage, Rich is expressed in PR cells, as well as the postsynaptic cells that form the lamina plexus and medulla ( Figures 4C and 4D). The expression pattern of Rich substantially overlaps with that of CadN in the optic lobe ( Figures S3A and S3B). In summary, Rich is present at the proper time and in the proper cells to account for the observed targeting defects. Since Rich is present in the PR cells as well as the post-synaptic cell types and the eyFLP system generates mutant clones in

both cell types, it is not obvious whether Rich is required cell autonomously in the PR cells. To determine if Rich is required in PR cells, we used ey3.5FLP ( Chotard et al., 2005) in combination with MARCM ( Lee and Luo, 1999) to generate clones of GFP-labeled rich1 homozygous mutant PR cells. In the medulla, 74% of the mutant R7 cells (n = 175) fail to target to the M6 layers showing that Rich is required in R cells for targeting specificity ( Figures 5A, 5B, 5A′, 5B′, and 5C). We also generated single R7 cell mutant clones using the GMR FLP system ( Lee et al., 2001). These clones are induced late in PR development and only affect 15% of the R7 cells. Interestingly, we did not observe any mistargeting phenotype, possibly because of perdurance ( Figure S4). We therefore assessed the phenotype of single mutant R7 cells surrounded by wild-type R7s and R8s in the ey3.5 FLP mutant animals. As shown in Figure 5D, 62.