ABSTRACT Enormous cell type diversity occurs in the mammalian central nervous system, but how neuronal cell diversity is created remains a major question. Cell type specification is largely regulated by gene expression where mRNAs play a central role. While cell type-specific transcription is essential for gene expression, post-transcriptional modifications and clearance of mRNAs are equally important for gene regulation. Understanding cell type- specific mRNA dynamics, especially epitranscriptomic modifications, mRNA transport and degradation, is key to understanding mechanisms of neuronal cell diversity and pathogenesis of RNA binding protein mediated neurological disorders. Current crosslinking and immunoprecipitation-based methods have led to significant insights about protein-RNA interaction, but they are not applicable to rare cells or profiling hundreds of cell types in the brain. This proposal outlines a radically new approach, PREDIT, that enables high throughput detection of protein-RNA interaction at single cell resolution. The method labels RNAs by creating nucleobase changes that can be identified by single cell full-length RNA sequencing together with cell type information. In this proposal, we first describe three major phases for method development: (1) screening RNA editing enzymes and optimizing their editing activity; (2) evaluating the efficiency and specificity of PREDIT with innovations in both molecular designs and computation analyses; and (3) applying the method to picogram level RNA input and single cell detection of protein-RNA interaction. We further describe applications of PREDIT to address the following questions: (1) whether N6-methyladenosine is differentially recognized by different brain cell types? (2) how do different cell types degrade their mRNAs? (3) what mRNAs are tagged for transport? and (4) what are the cell type-specific mRNA substrates for disease-causing RNA binding proteins? PREDIT makes it easy to track protein-RNA interaction in cell lines and intact tissues, and makes it possible for the first time to study protein-RNA interaction among diverse brain cell types at unprecedented throughput. Together, these experiments will establish a convenient and powerful tool that is widely applicable to problems in fundamental RNA biology, developmental neurobiology and neurological disorders, leading to a more complete picture of neuronal cell diversity. 1

NARRATIVE RNA binding proteins are increasingly associated with neurological diseases, but how to pinpoint their RNA targets among hundreds of neuronal cell types is a challenge. This project outlines the development and applications of a novel method, PREDIT, which enables panoramic views of RNA substrates across diverse brain cell types. This tool will create a new paradigm to study neuronal cell diversity and neurological disorders by high throughput detection of cell type-specific protein-RNA interaction. 1