Investigation of RNA 3D structure and small molecule interactions by a multidisciplinary approach
Presented By LSA Biophysics
Elizabeth D Tidwell1,2; Anna Anders3; Varun Gadkari3; Brandon T Ruotolo3, Aaron T Frank1, and Markos Koutmos1,2,3
1 University of Michigan Department of Biophysics, 2 University of Michigan Chemical Biology Interface Training Program 5T32GM132046-02, 3 University of Michigan Department of Chemistry.
Structured RNAs regulate many key processes in pathogens like bacteria and viruses; yet RNA remains under-explored as a drug target. Visualizing the structures and structural transitions of RNAs are important for RNA therapeutic development; however, there is limited structural data, dynamic information, and incomplete understanding of RNA interactions with small molecules (SM). We are developing a pipeline that combines automated and high-throughput analytical, structural, and synthetic biology tools with molecular modeling and machine learning for rapid and large-scale exploration of RNA structure and RNA:SM interactions. We selected riboswitches to train our method—beginning with the flavin mononucleotide riboswitch (FMN-RS) due to the availability of high-resolution structures with and without its cognate FMN substrate and other previously identified ligands. We performed rigid body docking simulations to identify and rank structurally distinct SM on their potential to recapitulate the ligand interaction between FMN and FMN-RS. Then, we optimized and altered two commonly used in vitro screening methods for protein:SM interactions—ion mobility mass spectrometry (IM-MS) and biolayer interferometry (BLI)—for use with RNA. We have successfully optimized RNA sample preparation and data collection for IM-MS using FMN-RS and mitochondrial tRNA leucine. Using collision induced unfolding, we have determined the unfolding states of FMN-RS in the presence and absence of ligand. Additionally, we have successfully used the Octet Red BLI system to measure the FMN-RS interaction with SM. We obtained kinetic binding information with identified ligands of FMN-RS and screened a subset of the predicted SM from our simulations. The results from each in vitro experiment will be used to alter and improve the search criteria for the simulations and illuminate properties of the SM:RNA interactions. Our interdisciplinary methodology will be further optimized to streamline identification of conformationally selective RNA SM binders and potentially improve drug discovery.