RESEARCH PROJECTS

Genes have to be turned on and off at the right time and at the right place. The development of multicellular organisms depends on precisely controlled changes in gene expression patterns. Likewise, the proper functioning of differentiated cells and tissues requires that certain genes are active while others are silent. Because disturbances in the control of gene expression lead to developmental abnormalities and disease, an understanding of the underlying mechanisms is not only of basic scientific interest but also of medical relevance. Improved treatment of autoimmune diseases, cancers and other disorders may result if we advance our understanding of how genes are activated and repressed, and how the repressed or activated state is maintained over many cell generations.

In our lab, we are particularly interested in two aspects of gene regulation, the control by hormones and the control by chromatin structure changes. We use the fruitfly Drosophila melanogaster in our studies. This widely used genetic model organism allows us to address questions on a whole animal level, taking advantage of a broad range of experimental tools. Many of these tools are not available, or at least much more difficult to use, in other model systems. The generation of mutant or transgenic fruitflies, for instance, is much easier and less time-consuming than generating mutant or transgenic mice. In addition, analyses of genetically altered flies can be conducted in a much more detailed and quantitative manner because they have a vastly larger progeny than mice.

How can a systemic signal, which is distributed by the circulatory system throughout the whole organism, elicit different responses in different cells and at different times in development? Steroid hormones and other lipophilic signaling molecules, such as retinoic acid or the thyroid hormones, are such systemic signals. They act by binding to and activating nuclear receptors, which in turn act as ligand-dependent transcription factors. To be hormone-responsive, a cell must contain receptor molecules. Different types (isoforms) of receptors can be present in different cells, but this usually does not fully explain the specific transcriptional responses of these cells. We study how the steroid hormone ecdysone (or more precisely, 20-hydroxyecdysone) directs the tissue-and developmental stage-specific expression of genes that play a role during metamorphosis of the Drosophila larva into an adult fruitfly. In particular, we want to understand how edysone controls the genes that are responsible for the destruction of larval tissues during metamorphosis, which is brought about by programmed cell death. Earlier studies were focused on understanding how ecdysone can first activate and then repress a group of genes (salivary gland secretion protein or Sgs genes) that are exclusively expressed in the salivary glands of mid- to late-third instar larvae. We defined a hormone response unit that transforms inputs provided by the ecdysone receptor and by the tissue-specific factors Fork head and SEBP3 (a protein complex that contains the products of the daughterless and dAP-4 genes) into a temporally and spatially restricted expression pattern. Moreover, we developed a model for how a key regulator of the ecdysone-controlled genetic hierarchies that direct metamorphosis, the Broad-Complex, switches the positive hormone response of our model genes into a negative response (Renault et al., 2001). Surprisingly, it does so by integrating the "homeotic" cell identity factor Fork head, which is required for hormone-induced transcription of the Sgs genes in the salivary glands, into a hormone-controlled hierarchy. The fork head gene is downregulated by the Broad-Complex and the Sgs genes are thus rendered unresponsive to the hormonal signal.

Genes are packaged into nucleosomes and are integrated into higher order chromatin structures. How do these structural constraints influence the expression of a gene, and how do changes in chromatin structure contribute to gene regulation? We have discovered a whole new family of nuclear proteins that, as we believe, control gene expression and other nuclear functions (such as mitosis) by modifying chromatin structure. The discovery of the family was made possible by our identification of a novel DNA-binding domain in the Drosophila Pipsqueak protein (Lehmann et al., 1998) . Pipsqueak acts together with the well-studied GAGA factor in both the activation and silencing of homeotic genes. In silencing, the two proteins seem to provide the long-sought anchors, at least at a subset of sites, that recruit complexes of Polycomb group proteins to their DNA target elements, referred to as Polycomb response elements. These elements have been shown to confer an epigenetic maintenance of gene silencing over many consecutive cell generations. Biochemical, genetic and cytogenetic studies conducted in our lab led to a model in which Pipsqueak and the GAGA factor form a core protein complex that binds not only to homeotic genes but also to hundreds of other loci in euchromatin as well as to certain regions of the heterochromatin (Schwendemann and Lehmann, 2002). We study the interaction of this core complex with other proteins to better understand how Pipsqueak and the GAGA factor can bring about such divergent responses like transcriptional activation and repression. The localization in heterochromatin, together with a multitude of data obtained for the GAGA factor, suggests a chromatin-modifying mechanism that underlies the actions of these two proteins. Several lines of evidence suggest that other members of the Pipsqueak protein family also employ such a mechanism (Siegmund and Lehmann, 2002). We are therefore not only interested in Pipsqueak itself but also in the mode of action and the functions of other family members. The results of our most advanced studies, which are on the piefke gene, are consistent with our model of Pipsqueak family proteins being important chromatin regulators.