The dynamics of gene regulatory networks induced by Notch activation

Cells communicate with each other and receive information about their environments through molecular signalling pathways. One such signalling pathway involves the Notch receptor. This pathway is highly conserved, is important for many aspects of building multicellular animals and is disrupted in several human diseases, including cancers. If we are to understand how Notch is able to alter the behaviour of cells we need to understand how its activity changes the expression of genes in the cell. New technologies, such as DNA microarrays, allow us to simultaneously analyse all of the genes encoded by an animal's genome. Thus we can ask how many genes are switched on or off when a cell receives the Notch signal. By looking at different times after Notch activation we can further find out how quickly and in what order different genes are switched on. We propose investigating these questions using cells from the fruit-fly Drosophila as our model. Not only was this the animal where Notch was first discovered (a slight defect in gene activity causes a 'notch' in the fly wing), but also it has a smaller and more simple genome than mammals (for example humans have multiple Notch-like receptors whereas Drosophila has one). However, even using Drosophila, our experiments will produce a very complicated set of data and to fully analyse and understand the relationship between the genes that are turned on at different times we will need to use mathematical approaches. Indeed we propose to go beyond this and use our data to build mathematical models of the network of gene responses to Notch activation. The hope is that, as the models become more sophisticated and accurate, we will be able to predict, from first principal, the behaviour of this biological system. Thus we hope to be able to simulate in the computer what would happen if the system is defective (for example if a gene is mutant) or if it is treated with a particular drug. We will be able to verify these predictions through experiments, to see how accurate our models are, and to improve them. This approach requires that biologists and mathematicians collaborate to bring together their expertise as we propose here. Our long term goal therefore is to develop an understanding of Notch signalling by building predictive mathematical models; models that will be constructed by analysing the data we generate in this study, and that can subsequently be useful for further studies, for drug design and for application to other signalling pathways.

Our goals are (1) to elucidate the temporal characteristics of the transcriptional response to Notch signalling and (2) to develop predictive models from these data that will inform our understanding of the intrinsic regulatory circuits. Characterising the regulatory circuits underpinning genomic responses to signalling and developing the computational tools necessary for the predictive modelling of these circuits requires experimentally tractable systems with well-developed genomics resources. Drosophila offers an excellent system for quantitative analysis of metazoan genomes because it is comparatively simple and has less genome complexity than vertebrate models. We therefore propose to exploit an ex vivo model in Drosophila cells for generating quantitative data on the cellular responses to Notch signalling. In our analysis of these data we will identify temporal relationships and patterns within the responding genes, which will be tested to validate these networks.