Using molecular dynamics simulations to identify the key factors responsible for chiral recognition by an amino acid-based molecular micelle

Molecular dynamics (MD) simulations were used to investigate the binding of six chiral compounds to the amino acid-based molecular micelle (MM) poly-(sodium undecyl-(L)-leucine-leucine) or poly(SULL). The MM investigated is used as a chiral selector in capillary electrophoresis. The project goal was to characterize the chiral recognition mechanism in these separations and to move toward predictive models to identify the best amino acid-based MM for a given separation. Poly(SULL) was found to contain six binding sites into which chiral compounds could insert. Four sites had similar sizes, shapes, and electrostatic properties. Enantiomers of alprenolol, propranolol, 1,1′-bi-2-naphthyl-2,2′-diyl hydrogen phosphate, 1,1′-bi-2-naphthol, chlorthalidone, or lorazepam were separately docked into each binding pocket and MD simulations with the resulting intermolecular complexes were performed. Solvent-accessible surface area calculations showed the compounds preferentially associated with binding sites where they penetrated into the MM core and shielded their non-polar atoms from solvent. Furthermore, with five of the six compounds the enantiomer with the most favorable free energy of MM association also experienced the most favorable intermolecular hydrogen bonding interactions with the MM. This result suggests that stereoselective intermolecular hydrogen bonds play an important role in chiral discrimination in separations using amino acid-based MMs.GRAPHICAL ABSTRACT

本研究采用分子动力学模拟（Molecular dynamics simulations），探究六种手性化合物与基于氨基酸的分子胶束（molecular micelle, MM）聚（十一烷基-(L)-亮氨酸-亮氨酸钠）[poly(SULL)]的结合作用。该分子胶束常作为手性选择剂应用于毛细管电泳分离中。本项目的研究目标为表征此类分离过程中的手性识别机制，并构建预测模型，以针对特定分离需求筛选最优的氨基酸基分子胶束。研究发现poly(SULL)含有六个可供手性化合物嵌入的结合位点，其中四个位点在尺寸、形状及静电性质上较为相近。研究分别将阿普洛尔（alprenolol）、普萘洛尔（propranolol）、1,1'-联-2-萘酚-2,2'-二基磷酸氢酯（1,1′-bi-2-naphthyl-2,2′-diyl hydrogen phosphate）、1,1'-联-2-萘酚（1,1′-bi-2-naphthol）、氯噻酮（chlorthalidone）及劳拉西泮（lorazepam）的对映异构体分子对接至每个结合口袋中，并对得到的分子间复合物开展MD模拟。溶剂可及表面积（Solvent-accessible surface area）计算结果显示，目标化合物优先结合于可使其嵌入MM核心、并将非极性原子与溶剂隔离的位点。此外，六种化合物中的五种，其与MM结合自由能最优的对映异构体，同时也与MM之间存在最有利的分子间氢键相互作用。这一结果表明，立体选择性分子间氢键在基于氨基酸基MM的分离过程的手性分辨中发挥着重要作用。图文摘要