With more than 125 000 protein crystal structures in the Protein Data Bank (PDB) and many more unpublished in-house structures, abundant structural information has become available to successfully pursue structure-based design. Besides this essential information, another pillar of structure-based ligand design is conformational analysis, aided by more than 875 000 small molecule structures in the Cambridge Structural Database (CSD). The third pillar is the in-depth understanding of weak intermolecular interactions, including the role of solvent. An emerging fourth pillar is the simultaneous combination of structural design work with physico-chemical and ADME properties, provided by big-data mining. All these topics are addressed in today's symposium.<br/>    <br/>We pursue since the 1980ies a multi-dimensional, highly collaborative approach towards deciphering and quantifying weak intermolecular interactions in chemical and biological systems. Experimental study in this research involves the investigation of protein-ligand interactions, synthetic host-guest complexation, and dynamic processes in designed unimolecular model systems, such as molecular torsion balances. It is complemented by computational analysis and data base mining in the CSD and the PDB. Examples of intermolecular interactions quantified by this approach are orthogonal dipolar interactions, organofluorine interactions, stacking on peptide bonds, cation-<i>π</i> interactions, halogen bonding, and chalcogen bonding. Enantioselective complexation based solely on shape complementarity and dispersion interactions is investigated with new chiral cage compounds, validating the 55% rule for optimal cavity filling established by Mecozzi and Rebek. Extensive protein and small-molecule crystallographic work is essential to all study. We also explore the energetics of the replacement of conserved water molecules in protein co-crystal structures by ligand parts. This work established that drug designers can learn a lot from host-guest model systems, as Martin Stahl stated in a recent F1000 recommendation. In fact, many of the weak interactions of current interest could only be quantified in model systems. The multi-dimensional approach is illustrated in examples taken from our structure-based drug design projects. Lessons learned are directly applicable to ligand design and optimization in drug discovery and crop protection research, but equally to the assembly of synthetic supramolecular systems.

40 Years of structure-based design: What have we learned?

254th National Meeting (2017)

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.