Fifteen years ago, the ability to look at complex molecule landscapes inside cells using computers only just about existed in science fiction movies. Nowadays, science fiction has become reality thanks to bioinformaticians, who have developed better algorithms for analysing the huge amount of data from biochemical experiments and predicting three-dimensional molecule structures. Junior professor Dr. Stefan Günther from the Institute for Pharmaceutical Sciences at the University of Freiburg has a long-standing interest in computer sciences and is now developing software to model three-dimensional protein structures. This enables him to predict interaction mechanisms between therapeutic substances and cellular enzymes. But this is not all. Information about potential mechanisms of action of potential pharmaceutical compounds can already be found somewhere in the literature. But the question that arises is how can the “needle” be found in the literature haystack?
Stefan Günther was born in Gelsenkirchen in 1972 and grew up in Essen. He was very interested in the natural sciences and chose to start studying biology in Bochum in 1994 because he was interested in the broad range of different topics covered by the subject. Following his pre-diploma in 1997, he moved to Kiel and opted to study informatics as a minor subject. His brother was one of the first people to own a C64 and showed Stefan Günther very early on how programming knowledge can be useful. In Kiel, Stefan Günther started to model ecosystems, amongst other things. "I have always been interested in biology-related areas," said the junior professor of pharmaceutical bioinformatics at the Institute for Pharmaceutical Sciences at the University of Freiburg. He received his diploma in 2002 and decided to follow special training in computer sciences, which is why he chose to continue his studies at the Beuth University of Technology in Berlin, where he received a master's degree in computer sciences in 2004. His master's thesis dealt with interactions between DNA and proteins such as transcription factors. He developed special software programmes to predict the three-dimensional structure of molecules.
"This is still part of my current work," said Günther who moved to the Charité in Berlin to do his doctoral thesis on the specific bindings and molecular interactions between biological molecules. After a three-year period as scientific assistant and post-doc, Günther moved to Freiburg in 2008 where he was appointed junior professor at the University of Freiburg and where he established his own pharmaceutical bioinformatics group. While he was working at the Charité, Günther also developed software programmes to predict interaction mechanisms between proteins, as well as between proteins and small molecules, which include many pharmaceutically interesting compounds. It is important to know how biological molecules interact with each other in cells in order to be able to manipulate these interactions. Nowadays, Günther and his team use their know-how to find new therapeutic compounds or analyse their effect on several levels, including on the molecular level.
The initial step in the search for new therapeutic compounds is the functional description of microbial genomes. The genomes of bacteria or unicellular fungi harbour information about interesting metabolic pathways that can be used to synthesise antimicrobial substances. It is envisaged that this information will at some stage be used by the pharmaceutical industry to develop antibiotics. "Genome annotation is the initial step when describing a genome and making it useable for further research," said Günther whose team is developing software that is able to find coding and non-coding regions of microbial DNA sequences and to ascribe potential functions to genes. The researchers' goal is to decipher exotic metabolic pathways. Which enzymes are involved? How are pharmaceutically interesting metabolites synthesised? Where is it possible to interfere in order to obtain high enough yields of pharmaceutically active compounds?
Some of the aforementioned questions can be answered using the algorithms developed by Günther’s team. In one project, the Freiburg bioinformaticians worked with pharmaceutical biologists, medical doctors and chemists from the University of Freiburg and were able to describe the function of several interesting genes of a Streptomyces bacterium. Other questions still need to be answered. “In future, we will work with mathematicians and systems biologists from Freiburg,” said Günther going on to add “in order to model metabolic pathways and find the parameters that can be altered in order to control the yield of interesting compounds.”Predictions on the interaction mechanisms between biological macromolecules and small molecules, the annotation of microbial genes, three-dimensional structural analyses of enzymes – this is only one side of the coin. In many cases it is already possible to carry out comprehensive literature searches to discover proteins that are targeted by potential drugs. Numerous online literature databases list studies that investigate a certain drug for its application in a broad range of different applications. Günther and his team have developed a tool that enables the search for useful information from thousands of scientific articles stored in online databases. The Compounds in Literature (CL) is a web server for the fast and efficient assessment of the medical effect of new compounds. A beta version is available free of charge and comes with a user-friendly surface to scan texts and look for molecule pairs that frequently appear in one sentence. “If for example, the word acetylsalicylic acid is followed by the word prostaglandinic acid, then there is a high probability that the published study found that the latter interacts with the former,” said Günther. Researchers or students interested in further details can then take a more detailed look at the papers where the two words were mentioned.“What I find most exciting about my work is that I can pursue so many different approaches and themes,” said Günther. “As a biologist and software programmer I am able to talk with people from a broad range of life science disciplines. I might not understand every detail, but I can understand where the problem lies and discuss it with my interdisciplinary cooperation partners in order to find solutions.”
Further information:Jun.-Prof. Dr. Stefan GüntherUniversity of FreiburgInstitute for Pharmaceutical SciencesPharmaceutical BioinformaticsHermann-Herder-Strasse 979104 Freiburg i. Br.Tel.: +49-(0)761/203-4871E-mail: stefan.guenther(at)pharmazie.uni-freiburg.de