Molecular Design Lab, Pharmaceutical & Medicinal Chemistry, Institute of Pharmacy, Freie Universität Berlin, Königin-Luise-Straße 2+4, Berlin, 14195, Germany.
Mol Inform. 2019 May;38(5):e1800141. doi: 10.1002/minf.201800141. Epub 2019 Feb 6.
For drug design projects it is essential to rationally induce and explain selectivity. In this context shape complementarity as well as protein and ligand flexibility represent important factors. Currently available tools for the analysis of protein-ligand interactions focus mainly on electrostatic complementarity and/or static structures. Here we address the shortcomings of available methods by presenting two new tools: The first one can be used to assess steric complementarity in flexible protein-ligand complexes in order to explain selectivity of known ligands. It further allows to determine ligand atoms with especially good or bad shape-fit which can be of use in lead optimization projects. The second tool was designed to detect differences in protein flexibility in similar proteins along with their exploitation for virtual screening. Both tools yield interesting results when applied to data of protein tyrosine phosphatase 1B (PTP1B): The case of PTP1B has proven especially difficult in terms of selectivity, due to a closely related phosphatase connected to severe undesired effects. With our tool for steric complementarity assessment we were able to explain previously undisclosed causes of moderate selectivity of selected PTP1B ligands. The second tool allowed us to find differences of flexibility in the two highly similar proteins and give directions for exploitation in virtual screening.
对于药物设计项目来说，合理地诱导和解释选择性是至关重要的。在这方面，形状互补以及蛋白质和配体的柔性是重要的因素。目前用于分析蛋白质-配体相互作用的工具主要集中在静电互补和/或静态结构上。在这里，我们通过提出两个新的工具来解决现有方法的缺点：第一个工具可用于评估柔性蛋白质-配体复合物中的空间互补性，以解释已知配体的选择性。它还可以确定配体原子的形状拟合特别好或特别差的原子，这在先导优化项目中可能会有用。第二个工具旨在检测相似蛋白质中蛋白质柔性的差异，并将其用于虚拟筛选。当将这两个工具应用于蛋白酪氨酸磷酸酶 1B（PTP1B）的数据时，它们都产生了有趣的结果：由于与严重不良反应相关的密切相关的磷酸酶，PTP1B 的选择性问题尤其难以解决。使用我们的空间互补性评估工具，我们能够解释先前未公开的选定 PTP1B 配体中度选择性的原因。第二个工具使我们能够发现两个高度相似的蛋白质之间的灵活性差异，并为虚拟筛选中的利用提供方向。
