Over the years, classical genetic analysis has identified key regulators involved in metazoan development. With the advent of high-throughput approaches, the pathways are beginning to be assembled into networks, the topology of which provides further insights into how development is orchestrated. Two recent studies published in Genes & Development used ChIP-on-chip experiments to elucidate transcriptional regulatory networks in the early development of Drosophila melanogaster, focusing on a few key transcription factors involved in the specification of mesoderm and the dorsoventral (DV) axis. They revealed unexpected crosstalk between biological processes and hitherto unappreciated modes of regulation.

Sandmann et al. focused on the central regulator of mesoderm development in D. melanogaster — Twist. To identify Twist-bound cis-regulatory motifs, the authors performed ChIP-on-chip analysis at two consecutive time periods in early fly development, combined with expression profiling of twist loss- and gain-of-function embryos. The difference in the binding of Twist to cis elements at these time points revealed the dynamic nature of Twist-mediated regulation. Furthermore, computational approaches followed by ChIP experiments found evidence for extensive combinatorial binding of Dorsal and Twist at the earlier time point and Tinman and Twist at the later time point.

The analysis unexpectedly revealed that Twist also regulates genes that are essential for proliferation and transcriptional regulation. The authors combined the temporal binding map of Twist with in vivo binding data for Myocyte enhancing factor 2 (MEF2), Dorsal and Tinman — three key mesodermal transcription factors. The resulting core network of transcriptional regulation of early mesoderm development reveals extensive combinatorial binding and frequent feed-forward regulation. The topology of this network argues against the hierarchical model of a master regulator function. Instead of regulating only a few factors, each of which has a handful of their own direct targets, Twist directly affects multiple levels in this developmental network.

Zeitlinger et al. also used ChIP-on-chip analysis but their focus was on genome-wide discovery of the targets of the DV determinants: Dorsal and its two earliest targets, Twist and Snail. Their ChIP-on-chip analysis was performed on embryos in which Toll receptor has been constitutively activated, which leads to high levels of Dorsal, Twist and Snail, and makes all of the embryo's cells adopt a mesodermal fate. As expected from the genome-wide scale of this work, new DV enhancers were identified. But the discovery that many segmentation genes contain binding sites for Dorsal, Twist and/or Snail uncovered previously unanticipated links between the DV and anterioposterior (AP) networks. Moreover, binding sites for these transcription factors were also found at genes that encode components of signal transduction such as Decapentaplegic (DPP), Epidermal growth factor (EGF) and Notch pathways.

Together, the studies show how ChIP-on-chip analysis can extend the core sets of regulators that are known from genetic studies. In doing so, they reveal previously unanticipated insights into the architecture and the extent of connectivity of early developmental networks.