Strategies for Modeling Ligand Docking to Natural and Engineered RNA Structures

Abstract

RNA is a multi-functional molecule, combining genetic information carrier's role with regulatory functions determined mostly by its three-dimensional structure. These characteristics make it a unique building material for designed nanoparticles, in which it provides structural scaffolds and active elements that can harness, for example, RNA interference pathway. An important aspect of this research and design is prediction of small molecule binding poses to RNA and their impact on the structure to affect therapeutic outcomes or control the shape and function of nanoparticles. Most computational small molecule docking programs have been optimized to perform well for protein targets. Predictions for RNA targets are much less accurate due to inherent differences between these two biomolecules. RNA structures are generally more flexible and dynamic than proteins, with more even distribution of charge on their surface and fewer natural docking pockets. In addition, structurally important divalent cations and complicated tertiary interactions may interfere with the docking target identification. RNA docking targets may be exposed transiently, and our approach to this problem starts with exploration of the dynamics of the target structure. It is followed by a docking screen against a sample of predicted dynamic states and selection of results informed by any available experimental information. Selected complexes undergo another round of dynamics simulations. Thus, we first increase the number of targeted conformations, and finally re-evaluate the stability of predicted docking poses using the state-of-the-art RNA and small molecule parameters that improve on the rough draft qualities of the docking scores. We present new predictions of small molecule docking poses for a priming loop target in the Hepatitis B virus pre-genomic RNA, the ENE triple helix of Kaposi's Sarcoma Herpesvirus, and a ligand-stabilized RNA nanoring structure. Funded in part by HHSN261200800001E.