Separating the wheat from the chaff in target identification for bioactive molecules
The search for bioactive small molecules and their cellular targets in drug discovery and chemical biology that effectively modulate disease-relevant signaling pathways is like looking for the proverbial needle in the haystack. Here, finding a bioactive small molecule is just the beginning, the search really gets serious with the identification and validation of the cellular target. The Waldmann group together with the Steinhilber group from Goethe University Frankfurt and the Schneider group from ETH Zurich now developed a composite strategy, where chemical proteomics and in silico target prediction were successfully combined to identify 5-Lipoxygenase as the target of the Wnt pathway inhibitor Lipoxygenin.
In the past plant extracts that contain natural products have been applied successfully in treating a plethora of human diseases, often without any prior knowledge of the bioactive compound and their molecular targets. Rapid developments especially in organic synthesis methodology but also in the isolation of chemical compounds have opened the doors to an enormous pool of potentially bioactive compounds. On the other side, there is a variety of potential targets in the cell: not only over 20,000 proteins, but also nucleic acids, carbohydrates and lipids. The identification of target molecules in chemical biology and drug discovery is performed via two strategies. In reverse chemical genetics, a well-known protein is addressed directly in a target-based screen. This approach provides hit compounds with known targets but their influence on the biological system is not known and remains to be confirmed in cellular assays. In forward chemical genetics the effect of compounds on biological system is investigated. This unbiased approach identifies bioactive small molecules causing a desirable phenotype and may uncover novel targets for a given cellular process. This approach, however, entails a challenging, time- and resource-intensive identification and confirmation of the cellular targets that often meets with failure.
Currently, there are three distinct but complementary method classes for target identification, divided into chemical proteomic, genetic and computational approaches. A generic methodology is lacking. The research group of Prof. Dr. Dr. h.c. Herbert Waldmann now combined two approaches and could thereby enhance the efficiency of the target identification and validation process.
In a first step the group performed a cell-based screen of more than 10,000 compounds for Wnt pathway inhibitors. Since misregulated Wnt signaling is linked to type II diabetes, neuronal and cardiovascular diseases, and to cancer development, it is of great interest especially for pharmaceutical research. Unfortunately, a drug for clinical use has not been approved so far, resulting in a high demand for novel inhibitors. The group was optimistic when they identified 3,5-substituted-2,4-dimethoxypyridines as a novel, promising Wnt inhibitor class. This hit molecule was further applied as a bait in affinity-based chemical proteomics, to fish out potential targets out of a protein extract. But the obtained targets could not be validated in additional experiments
To prevent further meaningless validation experiments the scientist developed a new composite strategy for target identification by complementing chemical proteomics methodology with innovative in silico target prediction. Here, the similarity between the hit molecule and already published bioactive compounds is evaluated regarding its structure and properties. Application of both methods for the novel Wnt inhibitor suggested several different targets but one target was common for both lists: arachidonate 5-lipoxygenase (5-LO). Enzyme inhibition studies confirmed arachidonate 5-LO as a target of the previously identified dimethoxyprimidine compound, which was therefore termed Lipoxygenin. The scientists uncovered that Lipoxygenin prevents nuclear localization of 5-LO and thereby reduces the nuclear amount of 5-LO and -catenin, a key player in Wnt signaling. Moreover, Lipoxygenin inhibited further developmental signaling pathways, such as Activin A, BMP, Hedgehog and TGF- signaling. These findings suggested an unknown role of 5-LO as a common regulator in developmental signaling.
“Our study emphasizes the effectiveness of combining experimental and computational approaches to guide target identification and validation. It could therefore help to conserve important resources in both chemical biology research and pharmaceutical industry” says Slava Ziegler, co-author and project group leader in the department of Prof. Dr. Dr. h.c. Herbert Waldmann.