The overall goal of our research is to understand how the information contained in animal genomes transforms into coordinated cellular behaviors that collectively represent development. It can be argued that the most direct manifestation of the genomic program is the tissue specific regulation of gene expression. Therefore, by describing the gene expression patterns in the context of cellular anatomy of the developing system, we take the necessary, first step towards understanding the information transfer from genome sequences to developmental processes. We combine wet-lab and informatics approaches working closely with experimentalists as well as the growing group of computational biologists in Dresden.

We focus on Drosophila embryogenesis as primary model system and insist on limiting our approaches to the ones that can be applied to many and potentially all genes in the genome. Initially we used high-throughput methods such as RNA in situ hybridization to visualize patterns of gene expression during embryonic development (APOGEE and FlyFish in collaboration with Eric Lecuyer and Henry Krause). More recently we developed the FlyFos system that allows systematic generation of faithful, fluorescently labeled, live  transgenic gene expression reporters. We develop novel imaging approaches based on Selective Plane Illumination Microscopy (SPIM) that will allow us to monitor the entire developing system, distinguishing all cells and associating gene activity reporters to them (see movie). This will result in the most qualitatively complete representation of patterns of gene expression and will lay down the observational foundation for systems biology analysis of development. These advanced imaging approaches necessitate new developments in quantitative 3D image analysis and to that end we pursue several projects that combine computer vision and biology (Fiji and CATMAID).

The observed complexity of gene expression regulation in embryogenesis arises, to large extent, from evolution of cis-regulatory sequences. The mechanisms of action of natural selection on non-coding sequences remain relatively poorly understood. We use the morphologically highly similar embryonic development of sequenced Drosophilid species to study the sequence determinants underlying divergence of gene expression patterns. We use high-throughput methods such as micro-arrays and deep sequencing, to uncover divergent gene expression profiles. In the course of this work we found unique, quantitative supporting evidence for a fundamental law of nature, the hourglass model of morphological evolution. We use the hourglass hypothesis as motivation to evaluate, using the tools we developed in D. melanogaster, the divergence of embryonic patterns in different species before, during and after the insect phylotypic stage. We hope that by linking the divergence in sequence, to the divergence in gene expression regulation and the embryo phenotype we will provide new insights into the evolutionary mechanisms shaping the gene regulatory networks in early embryogenesis.