Enhanced Sampling of LNCRNA Conformational Space for Defining Ensembles of Structures Used in Ensemble Docking and Virtual Screening of RNA-Focused Small Molecules

Abstract

RNA remains an underexplored class of drug targets owing to the limited number of RNA-focused experimental and computational screening assays. Importantly, the highly charged phosphate backbone and inherent flexibility unique to RNA macromolecules present computational challenges to virtual screening techniques involving conformational sampling and molecular docking. We have developed a computational methodology to examine docking of small molecules to an ensemble of RNA conformations. A pipeline is developed to target a critical triple helix in the oncogenic MALAT1, a long noncoding RNA that promotes metastasis in a large number of cancer types. We use a conformational FRET assay to show that the MALAT1 triple helix (M1TH) samples a large conformational space consisting of both native states at physiological conditions and inhibitory states when solutions conditions are altered. Preliminary Small Angle X-ray Scattering (SAXS) experiments show a change in radius of gyration (Rg) consistent with conformational FRET data. Replica Exchange Molecular Dynamics are being used to calculate two ensemble sets (i) native-like conformations and (ii) inhibitory conformations) of M1TH structures. Structural states are clustered based on Rg consistent with SAXS data. The relatively small size of the 76 nucleotide M1TH simulation structure allows rigorous sampling of conformational space. We hypothesize that small molecule induced disruption of this RNA triple helix is a novel approach for therapeutic intervention to inhibit progression of multiple cancer types. M1TH ensembles will be used in ensemble docking of compounds identified through a shape-based virtual screening methodology to identify novel M1TH-specific lead compounds that reduce cancer proliferation.