Whole-genome ChIP-chip analysis of Dorsal, Twist, and Snail suggests integration of diverse patterning processes in the Drosophila embryo

Abstract

Genetic studies have identified numerous sequence-specific transcription factors that control development, yet little is known about their in vivo distribution across animal genomes. We determined the genome-wide occupancy of the dorsoventral (DV) determinants Dorsal, Twist, and Snail in the Drosophila embryo using chromatin immunoprecipitation coupled with microarray analysis (ChIP-chip). The in vivo binding of these proteins correlate tightly with the limits of known enhancers. Our analysis predicts substantially more target genes than previous estimates, and includes Dpp signaling components and anteroposterior (AP) segmentation determinants. Thus, the ChIP-chip data uncover a much larger than expected regulatory network, which integrates diverse patterning processes during development.

Figure 1.

Drosophila ChIP–chip identifies known enhancers of Dorsal, Twist, and Snail. Drosophila embryos, aged 2–4 h, from Toll^10b mothers were used to perform ChIP using antibodies against Dorsal, Twist, and Snail. Most enhancers known to be regulated by these factors were successfully identified, including type 1 target genes sna (A), mir-1 (B), and type 2 target genes rho (C) and brk (D). The graphs show unprocessed ChIP enrichment ratios (Y-axis) for Dorsal (red), Twist (blue), and Snail (green), across chromosomal regions (X-axis). Gene model (black arrow) and the known enhancer (red) are shown below.

Figure 2.

Enrichment of Dorsal-, Twist-, and Snail-binding motifs in bound regions. Sequences from DTS and TS regions with greater than fivefold ChIP enrichment were searched for the presence of the known Dorsal (red), Twist (blue), or Snail (red) motifs, and the evolutionary conservation of these motifs across the 12 sequenced Drosophila genomes was determined (as percent of branch length within the phylogenetic tree). The graph shows the fraction of regions at each conservation cutoff (0%, 20%, 40%, 60%, 80%, and 100%). As a control, the same analysis was performed with regions of identical length that were randomly distributed among intronic and intergenic regions. Using a χ² test, the motif enrichment and evolutionary conservation was highly significant for Dorsal (p < 10⁻¹⁴⁸), Twist (p < 10⁻²²⁸), and Snail (p < 10⁻³⁰⁶).

Figure 3.

Identification of novel DV enhancers. ChIP–chip data identified enhancers for the DV genes wntD (A), vnd (B–D), and tup (E). The left column shows the ChIP enrichment ratios of Dorsal (red), Twist (blue), and Snail (green). Bar, 2 kb. The tested enhancer (arrow) is shown with transcription factor motifs below (Dorsal site, red square; Twist site, blue triangle; Snail site, green circle; shading corresponds to their conservation across the 12 sequenced Drosophila genomes). The pattern of enhancer-driven lacZ expression in transgenic embryos (middle column) resembles that of the respective endogenous genes (right column). Unexpectedly, in addition to the previously known vnd enhancer (B), two regulatory regions (C,D) were identified that drive lacZ expression in subsets of the endogenous vnd expression pattern: The region in C drives early expression in the vNE, while the region in D drives later expression in the medial column (mc). Embryos are oriented anterior to the left, dorsal is up.

Figure 4.

Regulation of AP genes by Dorsal, Twist, and Snail. ChIP–chip binding data of Dorsal (red), Twist (blue), and Snail (green) at the loci of AP-regulated genes are shown in the left column (fold enrichment is indicated on the Y-axis): otd (A), tll (B), run (C), h (D), and kni (E). Bar, 2 kb. The presence of binding motifs and their conservation across the 12 sequenced Drosophila genomes at the occupied regions are indicated below the graphs (Dorsal site, red square; Twist site, blue triangle; Snail site, green circle). Many identified regions overlap with previously identified regulatory regions (light blue). Association with, and hence regulation by, Dorsal, Twist, and Snail may be responsible for the DV modulation exhibited by these AP genes, as shown in the right column (arrows). Embryos are oriented anterior to the left, dorsal is up.

Figure 5.

Integration of new putative target genes into the Dorsal network. Many newly identified putative target genes of Dorsal, Twist, and Snail are genes that were previously thought to be induced further downstream in the DV patterning network. Thus, Dorsal appears to regulate many target genes both directly and indirectly (a configuration also known as a feed-forward motif) in the suppression of dorsal ectodermal fate (A–C) and neurectodermal fate (D–F). Newly identified connections are marked by a star. Examples of such configurations are shown for target genes of Dpp signaling (A), which are also directly regulated by Dorsal; target genes of the transcription factor Zen (B), which are also directly regulated by Dorsal; target gene suppression by Dorsal (C), which can also be mediated by Snail; targets of EGF signaling (D), which are regulated by Snail at multiple levels; targets of Notch signaling (E), which are also regulated by Snail at multiple levels; and suppression of target genes by Snail or through induction of microRNAs (F). The regulatory relationship (activation, arrow; repression, orthogonal bar) of new connections was inferred based on the tissue-specific expression patterns of the target genes (Stathopoulos et al. 2002; Biemar et al. 2006).