(361c) Computational High-Throughput Screening of Modified RNA Interactions with Proteins

The modulation of RNA-protein interactions by modified RNAs have been linked to human diseases including neurological disorders and cancers¹. There are over 150 types of known highly dynamic and reversible chemical modifications that occur in RNAs. However, until recently, the field of RNA modifications had been in near dormant state², and little is known about the many protein-RNA modification interactions and how these interactions may regulate protein function. We developed the first computational protocol that characterizes interactions between proteins and RNAs containing post-transcriptional modifications³. Given a RNA-protein structure and a modifiable RNA nucleoside position, the protocol identifies RNA modifications that are energetically favorable to interact with the binding site of a target protein. The protocol operates in two-stages using molecular dynamics (MD) simulations and energy calculations in CHARMM. In the first stage, a decision-making tool consisting of short MD simulations and interaction energy calculations performs a high-throughput search through a list of RNA modifications, categorized into trees according to their structural and physicochemical properties, and selects a subset of RNA modifications prone to interact favorably with the target protein. In the second stage, multi-ns explicit-solvent MD simulations and free energy calculations are performed for the target protein in complex with RNAs containing the selected RNA modifications resulting in detailed 3D structures of the target protein in complex with RNA modifications, and the identification of RNA modifications that bind to the target protein with high affinity. We implemented and experimentally validated this CHARMM-based protocol through a test case involving the study of RNA modifications in complex with the E. coli polynucleotide Phosphorylase (PNPase)³ protein. Our studies uncover the favorable interactions between PNPase and 8-oxo-7,8-dihydroguanosine and reveal that PNPase strongly binds a diverse set of RNA modifications³. Our protocol showed reasonably high correlation between computational and experimental results, using electrophoretic mobility shift assays, and thus constitutes a promising stepping-stone towards a platform broadening our understanding of protein interactions with RNA modifications in key biological systems.

References:

1. Mihailovic MK, Chen A, Gonzalez-Rivera JC, Contreras LM. Defective Ribonucleoproteins, Mistakes in RNA Processing, and Diseases. Biochemistry. 2017;56(10):1367-1382.
2. Helm M, Motorin Y. Detecting RNA modifications in the epitranscriptome: predict and validate. Nat Rev Genet. 2017;18(5):275-291.
3. Orr AA, Gonzalez-Rivera JC, Wilson M, Bhikha PR, Wang D, Contreras LM, Tamamis P. A high-throughput and rapid computational method for screening of RNA post-transcriptional modifications that can be recognized by target proteins. Methods. 2018. doi: 10.1016/j.ymeth.2018.01.015.