A new approach to solvent determination in QM/MM-based X-ray crystallographic refinement

点击翻译按钮获取中文摘要

DESCRIPTION (provided by applicant): Success in structure based drug design (SBDD) and fragment based drug design are ultimately largely dependent upon the quality of the three-dimensional (3D) structure of protein:ligand and protein:protein complexes that is the core structural technique. Both nuclear magnetic resonance (NMR) and X-ray crystallography are used to determine experimental models pertaining to these structures. Through a previous project, QuantumBio Inc. improved the quality of the X-ray refinement through integration of the Company's quantum mechanics (QM) based DivCon Discovery Suite with the PHENIX crystallographic package. This natural synergy brings the power and accuracy of quantum mechanics to the field of X-ray refinement as it plays to the core strengths of QM methods (e.g. no atom types, support for more "exotic" chemical systems, metals, and so on). This early success has led to an expanded commercial offering that was released in February 2014. The present proposal is focused on a further improvement the accuracy of the X-ray refinement protocol by incorporating improved explicit solvent structure determination. It is quite clear - an there is a growing body of evidence that indicates - that the influences of solvent have significant impact on the ligand and receptor structure as well as on the energetics of the binding. Very often exploration of an active site - such as in lead discovery and optimization - is a question of whether or not additional "unseen" waters are mediating the interactions between ligand, protein, cofactor, and so on. An intrinsic problem in macromolecular X-ray crystallography is that only a partial number of solvent molecules can be unambiguously revealed due to the resolution limitations. Unlike approaches such as WaterMap, conventional 3D-RISM, and SZMAP, which are used to predict waters regardless of their agreement with experiment, the key innovation of this method is through the use of an advanced explicit water determination algorithm to filter crystallographic data and generate the complete, experimental solvent structure within the macromolecular complex. In preliminary studies performed with our partners, the evidence that this approach is applicable to the problem at hand is quite compelling. Specifically, the results for the 2.5 A lysozyme crystal structure 2EPE have shown that the application of the new solvation methodology leads to twice as many waters or a 100% improvement in the hydration shell for the low-resolution lysozyme structure accompanied by the improvement of the overall crystallographic statistics. The method also successfully found key, crystallographic "bridging" waters along with active site pocket stabilization waters when executed on the protein:ligand complex represented in PDBid:3ERQ. Together, these preliminary results are quite encouraging, and completion of this SBIR will allow us to completely generalize and validate the method and prepare it for commercial deployment.