Protein Evolution

The human genome contains about 25,000 protein-coding genes, which give rise to about 50,000 different functional proteins. Only a small fraction of those has been studied experimentally, and the number of proteins considered as 'well understood' is even smaller. Thus, even 10 years after the 'decoding' of the human genome, we are still far from understanding what our genes are all about.

In our group, we try to address the gene/protein function problem by both bioinformatical and experimental methods. In particular, we make use of the fact that most eukaryotic proteins are not monolithic structures, but rather have a modular architecture consisting of multiple building blocks. Most of these blocks have a specific unit functionality, and the combination of these so-called 'functional domains' determines the function of the whole protein. Over the course of evolution, the functional modules have been re-used, duplicated, mutated, deleted, reassorted, or any combination thereof. The reconstruction of the evolutionary history of a domain type yields important insights into the domain's function, its binding partners, and other important issues. firefox

What we are doing

The bioinformatical branch of our group uses computational methods to discover new domain types or to detect new instances of known domains in poorly characterized proteins. The analysis of a domain's conservation pattern or its phyletic distribution often allows a precise prediction of its function, which is amenable to experimental validation. By combining information on multiple domains in a protein - and by integrating other data sources obtained from a wide variety of model organisms - it is possible to make useful predictions on the function and properties of novel proteins, even if they have never been studied experimentally.

The experimental branch of our group will generate some of the data sources necessary for a successful protein functional prediction, one example being a domain-domain interaction map of major model organisms. In addition, we will complement the bioinformatical screens for protein functions by suitable experimental screens and validate some of the bioinformatical findings. The feasibility of this approach has been documented in several case studies that can be looked up in our publication list. firefox

Ubiquitin and protein homeostasis

The methods summarized in the previous paragraphs can be applied to all kinds of proteins. A particular biological focus of our group are components of intracellular protein degradation pathways. On the one hand. the proper regulation of protein degradation is of crucial importance for the cell; errors in this process are causally involved in diseases like cancer, Alzheimer, Parkinson, and many others. On the other hand, protein degradation is regulated by a highly complex network of >1000 interacting components. A key role is held by the small protein ubiquitin, which can be attached to target proteins, labeling them for later degradation by the proteasome. Over the last few years, evidence has accumulated that ubiquitin attachment also regulates proteasome-independent processes, many of them also involved in protein degradation (e.g. endocytosis, autophagy). Besides ubiquitin, a number of ubiquitin-like proteins (e.g. Sumo, Nedd8, Isg15, Fat10, Urm1) can also be attached to proteins, adding another layer of complexity. The fact that most regulators of the ubiquitin system have a highly modular architecture makes them excellent subjects for studies in protein evolution. .

Created using the CSS template Higherground