More

Abstract: RNA-binding proteins (RBPs) interact with RNA to affect every step of RNA metabolism. RNA binding sites (RBSs) constitute the molecular basis of RNA-protein interaction. The RBSs can be identified by liquid chromatography and tandem mass spectrometry (LC-MS/MS) analyses of the protein-RNA conjugates created by UVC-crosslinking. However, RBS mapping remains highly challenging due to the complexity of the RNA adducts generated by UVC-crosslinking. To address the limitation, here I developed RBS-ID, a method that uses hydrofluoride (HF) to fully cleave RNA into mono-nucleosides, thereby drastically enhancing the coverage and reaching single amino acid resolution. Moreover, the simple mono-nucleoside adducts offer a confident and quantitative measure of direct RNA-protein interaction. Using RBS-ID, I profiled ~2,000 human RBSs to discover novel RNA-binding residues. I analyzed the protein-, domain-, and amino acid-level characteristics of the identified RBSs. Also, I discussed potential roles of RBS post-translational modification (PTM) in modulation of RNA binding activity. In addition, I applied RBS-ID to in-depth discover RBSs in Streptococcus pyogenes Cas9(spCas9) complexed with its single guide RNA (sgRNA). The proximity of 84 identified RBSs to the RNA-binding interface was structurally verified. Among the RBSs, I found two residues functionally important for the gene editing activity of spCas9. Next, I expanded the methodology of RBS-ID to further develop an updated version called pRBS-ID. Here, RBSs are more comprehensively profiled using photoactivatable ribonucleosides (PARs). UVA-crosslinking allowed specific capture of the U or G-interacting RBSs in human RBPs, respectively. This enabled the comparison of base-specific features underlying RNA-protein interaction. Also, the interrogation of ribonucleoprotein (RNP) structural data allowed me to distinguish globular and disordered regions in proteins that interact with RNA. I further developed and applied an upgraded search pipeline to analyze the datasets from UVC-based RBS-ID experiments, altogether offering a comprehensive list of human RBSs with high coverage (3,077 RBSs in 532 proteins in total). Overall, RBS-ID and pRBS-ID are amenable platforms for investigating the dynamic RNA-protein interactions and the molecular basis of posttranscriptional gene regulation.