It is now more than 150 years since Darwin published his "Origin of Species", in which he proposed natural selection as the major mechanism for adaptation. Selection acts on phenotypic variability within populations: those individuals who are better adapted to their environment survive longer and have more offspring than less adapted ones. However, we still lack a coherent view of how such variability arises during evolution and development, and how it reflects molecular variation in the genome. In other words, we lack a precise idea of what is being selected.

The relationship between genotype and phenotype is complex and non-linear. Traditional genetic and molecular experimental methods are limited in their ability to keep track of the many factors and regulatory interactions involved. For this reason, we need a systems-biology approach, based on mathematical modeling, to address this question. We are carrying out an integrative, comparative analysis of real evolving developmental gene regulatory networks using a novel reverse-engineering approach (the gene circuit method). Gene circuits are computational tools to extract regulatory information from quantitative spatial gene expression data (Fig. 1). This is achieved by fitting mathematical models of gene networks to data (model optimization). The resulting models give us the structure and dynamics of the regulatory network responsible for the observed patterns, which in turn predict regulatory mechanisms that can be tested experimentally.

We study the evolution of the following networks in dipteran insects (flies, midge, and mosquitoes): the gap gene network involved in pattern formation in the early embryo of dipterans, the gene network underlying muscle and heart development during organogenesis, and the thoracic patterning network responsible for the positioning of mechanosensory bristles on the dorsal cuticle of the animal (Fig. 2A). We are establishing three dipteran species—the fruit fly Drosophila melanogaster, the scuttle fly Megaselia abdita and the moth midge Clogmia albipunctata—as model systems to experimentally and quantitatively test hypotheses derived from systems-biology approaches to evolutionary developmental biology (Fig. 2B).

Our models allow us to infer the regulatory interactions necessary and sufficient to explain the observed expression patterns by fitting models to data. Models from different species can be compared to reveal which interactions are conserved and which have diverged during evolution. In addition, we study evolutionary transitions between species using an in silico evolution approach. We will test these predictions by using RNA interference (RNAi) in various species and reporter assays in Drosophila. Our approach provides an integrative view of network evolution across multiple levels, from the molecular to the phenotypic. To our knowledge, this has not yet been achieved for any real developmental system.

We are collaborating with research groups in Europe and the U.S.

John Reinitz, University of Chicago, USA

Urs Schmidt-Ott, University of Chicago, USA

Maria Samsonova, St. Petersburg State Polytechnical University, Russia

Jaap Kaandorp / Joke Blom, University of Amsterdam, The Netherlands

Julio Banga, Instituto de Investigaciones Marinas, CSIC, Vigo, Spain

... and are partners in the EraNET projects MODHEART (Co-ordinated by Laurent Perrin, Marseille, France) and MOPDEV (Co-ordinated by Jaap Kaandorp, Amsterdam, The Netherlands).