Fig. 7. Loss of SG1 and SG2 results in regions of salivary tube closure. (A) Df(3R)Exel6216 deletes only 16 genes, eight of which are the previously characterized PH4 genes (asterisks), including SG1 and SG2 (Abrams and Andrew, 2002). (B) Df(3R)Exel6216 salivary glands have abnormal lumenal morphology. (C) Confocal images of salivary glands stained with CrebA (blue), ß_HSpectrin (red) and Spectrin (green) revealed regions of tube dilation, constriction and apparent closure (asterisks) in the Df(3R)Exel6216 embryos (the images -1 to -4 are different focal planes from the same glands). (D) Composite images of 50-80 0.3 µm sections of salivary glands stained with CrebA (blue) and Crb (red) revealed uniform lumen diameters wild-type embryos (left panel) but variable lumenal diameters in the Df(3R)Exel6216 embryos (right panel). (E) Expression of either SG1 (left panels) or SG2 (right panels) driven by a fkh-GAL4 driver rescues the salivary gland lumenal irregularities of Df(3R)Exel6216 embryos. Although all of the embryos of the genotypes shown in the lower two sets of panels in E were expected to express either SG1 or SG2, expression of both proteins was variable, with undetectable expression in some embryos (middle panels). Detectable SG1 and SG2 expression correlated with rescue of the lumenal defects (lower panels). Df(3R)Exel6216 is balanced over TM6B, Ubx-lacZ, allowing us to distinguish the heterozygous from homozygous deficiency embryos by also staining with an -ßgal antiserum. ßgal staining (red) can be seen in the heterozygous embryos in the top panels in E near the distal end of the salivary gland.