Fig. 3. SO acts upon the soAE motif in vivo and in S2 cells. (A-C) 4xsoAE is ectopically induced in wing discs by EYA and EYA+SO protein but not by SO protein alone. (A) dpp-GAL4 driving UAS-so does not induce 4xsoAE-lacZ in wing discs. (B) Ectopic expression of 4xsoAE-lacZ is induced in spots along the AP boundary by dpp-GAL4:UAS-eya. (C) Co-expression of so enhances eya-mediated reporter gene activity. (D) Ocellus-specific expression of so10-soAE is lost in so² mutant flies (arrow). so² is a regulatory mutant that displays an ocelliless phenotype. (E) EYA protein is detectable in the ocellar region of so² mutant flies (arrow). (F-I) so is necessary for soAE activation: in so³ clones, 4xsoAE-lacZ expression is lost although EYA is present in the cells. Clones are delimited by a dashed line. (F) EYA expression is shown in red and marks the non-dying cells within the clone. (G) so³ clones are negatively marked by the absence of GFP expression (in green). (H) 4xsoAE-lacZ reporter gene expression (blue) is lost in so^–/–cells. (I) overlay of G and H. (J) ß-Galactosidase reporter assays in Drosophila S2 cells. Reporter plasmids containing the lacZ gene under the control of different enhancer fragments were transfected into the cells. Co-transfection of so or eya alone together with the 10xsoAE-lacZ reporter plasmid does not exceed basal activity (lanes 2, 3). By contrast, co-expression of so+eya with the 5xsoAE or 10xsoAE reporter leads to a strong induction of ß-galactosidase (lanes 4, 5). The mutated versions of the 5xsoAE reporter (5xmutHD, 5xmutGATA) still show some amount of induction when co-transfected with so+eya (lanes 6, 7). ß-Galactosidase values were normalized by co-transfecting 5 ng of plasmid expressing luciferase as an internal standard. The results represent an average ß-galactosidase activity taken from transfections done in triplicates (±s.d.) and are illustrated as the X-fold activation over the basal activity found for the reporter plasmid alone.