DEV02251 Supplementary Material

Files in this Data Supplement:

• Supplemental Figure 1 - Fig. S1. Clusters of expression of segmentation genes revealing the high temporal resolution of expression data. Maternal genes, gap genes and segmental polarity genes are found in three main clusters, whereas pair-rule genes (primary and secondary) are split in the last two clusters. Clustering was made on the mean values of the triplicate experiments for each time-point. The colour scale is the same as in Figs 1 and 2.

• Supplemental Figure 2 - Fig. S2. Lipid droplets phenotypes.(A,A¢). RNAi to CG7428/halo inhibits the clearing of the embryo at the onset of cellularisation. The insets correspond to the green box. The cytoplasm basal to the nuclei (orange arrows) is darker in halo RNAi (A¢) than in control (A) embryos. (B,C) RNAi to CG1624/dpld impairs the ‘clouding’ of the epithelium at the end of cellularisation. The cytoplasm between the nuclei and the invagination front is clearer in dpldRNAi embryos. The insets (C) correspond to the green box. The cytoplasm (orange arrows) is clearer in dpld than in control embryos. (D) Lipid droplets distribution (marked with Bodipy, green) in control and in dpld RNAi embryos. PatJ (red) highlights the invagination front. The apical movement of lipid droplets is inhibited at the end of cellularisation and during gastrulation (clouding). (E) RT-PCR experiments show reduced expression of dpld in dpld RNAi embryos compared with water-injected embryos. (F) RNAi to dpld was performed in control (OreR), EP1050 and EP2291 strains overexpressing dpld under the GAL4 system during cellularisation. The brightness of the epithelium was assessed at the beginning of gastrulation and a value was attributed for each embryo, from 1 (dark epithelium, i.e. normal clouding) to 4 (very bright epithelium, i.e. absence of clouding, similar to klarsicht phenotype). dpldRNAi phenotype expressivity and penetrance is indistinguishable in OreR and maternal Gal4 embryos.

• Supplemental Figure 3 - Fig. S3. bitesize (btsz) controls gastrulation. (A) In control embryos the epithelium elongates along the anteroposterior axis and folds back dorsally during gastrulation (the orange arrow head points to the posterior tip of the extending epithelium). (B) In a btsz RNAi embryo, gastrulation arrests as the epithelium fails to elongate (white arrowheads) and becomes irregular (white arrows).

• Supplemental Figure 4 - Fig. S4. Characterization of the char^EY07696 mutant. Two different transcripts that only differ in the 5¢UTR sequence before the ATG are expressed. The P-element is inserted in the intronic sequence about 800 nucleotides 5¢ of the acceptor splice site of the CG5175-RA transcript. Semi-quantitative RT-PCR shows that the CG5175-RA transcript is absent in char^EY07696 homozygous mutants, whereas the RB transcript is normally expressed. RT-PCRs against actin42A transcripts are used to normalize amounts of cDNA products in the RT-PCR. Arrows indicate the position of PCR primers AF, BF and CR (see details in Materials and methods).

• Supplemental Figure 5 - Fig. S5. Morphology of nuclei labelled with and an anti Lamin/Dm0 antibody. Control (A) and transheterozygous Dfchar/ char^EY07696 mutant (B) embryos. There are defects in apical anchoring of the nuclei viewed from the side (B, top) compared with control embryos (A, top). In addition, sections of nuclei from the top show that the nuclei do not display deep infolding of the NE present in controls (compare A with B, bottom).

• Supplemental Figure 6 - Fig. S6. GST pull-down assay of HA-Char on GST and GST-Char beads. The lysate of Drosophila S2 cells expressing HA-Char is used. HA-Char is detected by western blot with an anti HA antibody.

• Supplemental Table 1A (Microsoft Excel)
• Supplemental Table 1B (Microsoft Excel)
• Supplemental Table 1C (Microsoft Excel) -

  Table. S1. Affymetrix microarray data. (A) Expression values for each experiment, i.e. three experiments for each time point (T0 to T4), are shown in the ‘signal’ columns. These values can be compared for a given gene at different time points but not between genes. The presence (P) or absence (A) of expression of a gene is shown in the ‘detection’ column. The detection is said marginal (M) when the presence or absence cannot be determined with certainty. Signal Log Ratio (SLR) for an experiment versus (vs) another one gives the log2 fold-change of expression of a gene between these two experiments (e.g. T1_1vsT0_2_Signal Log Ratio of –2 for a gene says that its expression decreases by 4 from experiment T0_2 to experiment T1_1). The change in expression is given in the ‘change’ column as a decrease (D), an increase (I), a marginal decrease (MD), a marginal increase (MI) or no change (NC). (B) Mean detection and signal log ratio (SLR) values for each time point. (C) This table indicates the occurrence of the following parameters for each time point and each gene: P, i.e. whether the gene is present or not; SLR>2, whether the log2 signal ratio is >2 between two time points; and ChangeI/ChangeD, whether the expression of gene increases (I) or decreases (D) between two time points. P, Present. If a ‘P in T0-count’ equals 2, it means that the corresponding gene is detected ‘Present’ two times out of the three experiments for the T0 time point. SLR>2, signal log ratio>2 for one experiment versus another one. If a ‘SLR>2 T1vsT0-count’ equals 1, it means that the signal log ratio of its expression in T1 versus T0 is superior or equal to two in one out of the three comparisons. ChangeI/ChangeD. If ‘ChangeI T3vsT1-count’ equals 0 for a gene, it means that the change in its expression from T1 to T3 was found increased in none of the three comparisons.

• Supplemental Table 2 (Microsoft Excel) - Table S2. The sheet ‘Overview’ shows the number of genes expressed at any of the T0-T4 stages. The list of genes specifically expressed at T0, T1, T2, T3 or T4 only is indicated in the sheets T0, T1, T2, T3 and T4, respectively.

• Supplemental Table 3 (Microsoft Excel) - Table S3. Details of the results of RNAi screen for the 57 selected genes. The genes showing a phenotype by RNAi are highlighted in yellow and the penetrance is indicated in parenthesis. The hatching rate (measured 36-48 hours after RNAi) is also shown together with the number of embryos tested for each experiment. The genes for which RNAi produced a lethality greater than 90% (hatching rate<10%) are shown in red. RNAi against CG7343 did not give rise to a btsz phenotype although CG14858 and CG7343 have both been assigned to the btsz locus. We found in fact that the dsRNA probe that was used against CG7343 targeted a non-expressed putative exon within the btsz locus (based on RT-PCR, northern blots) (see Serano and Rubin, 2003) (F.P. et al., unpublished). By constrast, RNAi against CG14858 recognizes all btsz isoforms in early embryos (F.P. et al., submitted). The phenotype observed with RNAi to CG11582 is due to interference against its close ortholog twisted gastrulation. A deficiency removing CG11582 did not give rise to gastrulation phenotype.