Molecular systems biology of gene regulation

Our goal is to tackle the central question of how genes are regulated with a new research approach that can be described as molecular systems biology.  To achieve this, we collect systemic data that characterize the molecular phenotype of cells in certain states and under defined conditions. We then identify molecular models that explain the systemic data, and find correlations between structural and systemic observations. This way we obtain new testable hypotheses that suggest further structural, functional, or systemic experiments. An iterative application of this approach can lead to a satisfactory understanding of gene transcription and its regulation in eukaryotes. In the following we provide examples for projects in this area.

Eukaryotic transcription regulation relies on coactivator complexes such as the Mediator or SAGA complexes. The Mediator is required for the regulation of mRNA transcription by Pol II. The mechanism of how Mediator enables regulated transcription is unknown, mainly due to the limited structural information. Mediator is very large, comprising 25 subunits in yeast, with a total molecular weight of one mega-Dalton. In the past, we developed co-expression strategies to assemble and purify Mediator subcomplexes. Using this strategy we could obtain structural information for one third of the Mediator subunits and their complexes. With the use of structure-system correlations, we demonstrated that structural Mediator submodules exist that serve gene-specific functions. To obtain mechanistic insights, we complement structural information with biochemical data. Assays to study transcription initiation, elongation, and termination in vitro allow us to probe structural features by site-directed mutagenesis and to test mechanistic models. To investigate which mechanisms are important in living cells, we use yeast genetics and phenotyping in vivo.

To study gene regulation and the mechanisms determining mRNA levels in the cell, we use and develop functional genomics approaches and correlate the resulting systemic data with structural information. One approach is to map gene-regulatory proteins over the genome in living cells. Chromatin immuno-precipitation coupled to high-resolution tiling microarrays (ChIP-chip) provides such occupancy profiles for RNA polymerases and their associated factors over the entire yeast genome. Another approach is to measure the rates of mRNA synthesis at all genes. To do this, we developed “dynamic transcriptome analysis” or DTA, which is based on non-perturbing metabolic RNA labeling coupled to microarray analysis. This method also provides the mRNA half-lives and thus enables studies of post-transcriptional regulatory events such as mRNA decay. The large datasets obtained with these functional genomics methods are analyzed with computational biology tools, often in collaboration with experts. Current projects aim at unraveling the rate-determining steps for genomic regulation in vivo and an identification of the underlying molecular assemblies.