We stained the eye discs of the third-instar eyFLP;Rab6 mutant larvae with anti-24B10, anti-Elav, anti-Senseless (labels R8), and anti-Prospero (labels R7). The staining patterns all appear normal, indicating that

there are no differentiation defects in Rab6 mutant eyes (data not shown). We then labeled R7 cells with Pan-R7 Gal4 and assessed the R7 targeting GDC-0068 molecular weight phenotypes of Rab6 mutants. 18.8% ± 3% (n = 265) of the R7 cells in Rab6 mutants (versus 19.7% ± 3% [n = 268] for rich1 mutants) fail to target to the correct layers in the medulla ( Figures 7C, 7C′, and 7D). The similarity in phenotype between Rab6 and rich mutant PR suggests that they function in the same pathway. To test this hypothesis, we removed one copy of Rab6 in the

eyFLP; rich1 or eyFLP; rich2 mutant animals and quantified the R7 targeting defects in the adult medulla. Rab6 heterozygous animals do not have R7 targeting defects, while the hypomorphic allele rich2 only has weak R7 targeting defects (8.5% ± 2%, n = 329; Figures 7F and 7M). We reasoned that loss of one copy of Rab6 in the partial loss of function mutant rich2 would enhance the phenotype. Indeed, loss of a copy of Rab6 enhanced the R7 targeting defects in eyFLP; rich2 animals (17.0% ± 3%, n = 483) similar to the rich1 mutant phenotype (compare Figures 7E and 7F to 7G; Figure 7M). However, loss of a copy of Rab6 does not modify the R7 targeting defects of the null allele rich1 (19.0% ± 3%; n = 308; compare Figures 7E to 7H; Figure 7M). Hence, the phenotypic and genetic data strongly argue this website that rich and Rab6 indeed function in a common pathway. To provide additional evidence that Rich and Rab6 function together,

we overexpressed a constitutively active form of Rab6 (Rab6CA) in rich1 mutant cells using the MARCM system. This greatly suppressed the R7 targeting much defects caused by loss of rich. Moreover, overexpression of a dominant negative form of Rab6 (Rab6DN) did not suppress or enhance the rich1 phenotype, indicating that Rich positively regulates Rab6 activity, possibly as a GEF ( Figures 7I–7L and 7N). To determine if Rab6 physically interacts with Rich, we overexpressed the WD40 domain (AA8–AA521) and RIC1 domain (AA711–AA995) of Rich in S2 cells and performed GST pull-down experiments using GST fusion proteins containing either wild-type Rab6 (Rab6WT), Rab6CA, or Rab6DN. A typical GEF protein tends to bind to the nucleotide-free form (e.g., Rab6WT) or GDP-bound form (e.g., Rab6DN). Consistent with this, both the WD40 domain and the RIC1 domain have a strong affinity for Rab6WT and Rab6DN but not Rab6CA, suggesting that Rich directly regulates Rab6 (Figure 7O). To directly assess the GEF activity of Rich, we overexpressed Rich in S2 cells and performed GTP exchange assay with the cell lysate. In this experimental setting, a positive control, Tiam1, showed proper GTP exchange activity for Rac1.