Immunohistochemistry and microscopy

Embryos were fixed in two different ways. For staining with anti-Crumbs (Cq4) and anti-Armadillo (N27A1) antibodies (both from the Developmental Studies Hybridoma Bank), we fixed the dechorionated embryos by a two second heat treatment at 93.4 °C, basically as described in Miller, Field & Alberts (1989). For the anti-Coracle (C566.9) and anti-Fasciclin 3 (7G10) stainings (antibodies also from the Developmental Studies Hybridoma Bank), we fixed dechorionated embryos according to Krahn et al. (2010). The rest of the protocol was according to Miller, Field & Alberts (1989). Anti-Crumbs was used at a 1:20 dilution, anti-Coracle at a1:200, anti-Fasciclin 3 at 1:200, and anti-Armadillo at a1:20. Secondary antibody was anti-mouse conjugated with Cy-3 (Invitrogen) at 1:1,000. For nuclei, we used Sytox-Green (1:3,000) according to the manufacturer’s instructions (Invitrogen). Embryos were mounted in Vectashield (Vector Laboratories), and imaged using a Zeiss 780 confocal microscope with 25× and 63× objectives. Images were acquired with Zeiss software, and manipulated using ImageJ (NIH). We used stage 13–14 embryos for the Crumbs, Coracle, and Armadillo antibody stainings; for the Fasciclin 3 staining, we used stages 15–16 embryos, except for Fig. S1, where stages 13–14 embryos were used. Finally, in Figs. 1C and ​and22 stage 16 embryos were used.

(A) Both cora² and chem¹ as heterozygotes significantly enhance homozygous mutant conditions of the other allele as embryos. The double homozygote has an intermediate phenotype between chem¹and cora². n is written to the left of each genotype column. (B) In chem mutants of stages 13–14 of embryogenesis, the lateral epithelial architecture is disrupted (compare control panel y w, to chem¹, chem³ and chem⁵ panels). For each genotype the top views (1–4) are projections of confocal stacks showing Cora staining in a head on view, with the left panels (1^′–4^′) also showing a nuclear Sytox-Green staining. Bottom panels show XZ projections of the stacks, to localize the Cora and Cora together with nuclei (Sytox) channels taken from the same stacks as above. Representative examples are shown, and an n of 8 embryos per genotype was imaged. The white arrows show details of the Cora staining. Notice basally mislocalized Cora staining in chem³ and chem⁵. Scale bar is 5 micrometers throughout. (C) In chem mutants of stage 16 of embryogenesis stained with Cora antibodies, there is a dramatic reduction of Cora expression. As in (B), for each genotype the top views (1–4) are projections of confocal stacks showing Cora staining in a head on view, with the left panels (1^′–4^′) also showing a nuclear Sytox-Green staining. Bottom panels show XZ projections of the stacks, to localize the Cora and Cora together with nuclei (Sytox) channels taken from the same stacks as above. Representative examples are shown, and an n of 8–9 embryos per genotype was imaged. The white arrows show details of the Cora staining. Scale bar is 5 micrometers throughout. (D) Quantification of Cora staining of experiments in (B and C). Notice significant reduction of Cora staining in all chem mutant embryos as stage 16 compared to stages 13–14. n is shown to the left of each column.

(A) 1–4 show confocal optical sections for the Cora channel (red); arrows point to Cora staining in all panels. 1^′–4^′ show the corresponding optical section with the Sytox (nuclei) channel added. In 1^′ the left white bracket illustrates the length measurement used to calculate the “partial length of cell” measurement, from the apical, lumenal site of Cora staining to the base of the salivary glands nuclei underneath. The smaller white right bracket in the same panel illustrates the Cora staining measurements used for Fig. 2C, showing the extension of the Cora staining in the cells. Scale bar is 10 micrometers. (B) Shows the quantitation of the relative Coracle staining using an area of 12.53 square microns for each measurement from optical sections as the ones illustrated in panels 1–4. The reduction of staining in chem³ is significant. n is written to the left of each genotype column. (C) The extension of Cora staining in relation to the cell length. In order to control for cell size differences, we measured in optical sections like the ones depicted in panels 1^′–4^′ the length of the cell from the apical, lumenal side of the Cora staining to the base of the underlying salivary gland nuclei, and we compared these measurements to the extension of the Cora staining. We graphed the percentage of the “partial cell length” thus measured to the Cora staining extension. chem⁵ extension is significantly reduced compared to the control. n is written to the left of each genotype column.