Decoding global RNP topologies in splicing regulation - Alternative splicing (AS) is a major mechanism that generates the vast transcriptome and proteome diversity from the limited genome. Spatial and temporal regulation of AS contributes to cell differentiation and lineage determination. Mutations that disrupt splicing cause a wide range of diseases including neurodegeneration, muscular dystrophies, and cancer. Therefore, understanding and targeting splicing is essential to the future of precision medicine. In the past 3 decades, AS studies have focused on cis elements and their associated trans factors, such as RNA binding proteins (RBPs). The laws of thermodynamics dictate that RNAs fold into low free energy structures in the context of RNA-protein complexes (RNPs), however, very little is known about how pre-mRNA structures in large RNPs control splicing, and various RNA processing events in general. Our lab invented several chemical crosslink-ligation based methods that enabled direct analysis of transcriptome-wide protein-independent RNA 2D and 3D structures in vivo (PARIS and SHARC). In this proposal, in Aim 1, we develop and benchmark a high throughput technology, SHARCLIP, to directly capture all RNA structures, RNA-RNA, RNA-protein, and protein-protein interactions together. SHARCLIP is conceptually similar to methods that analyze protein-mediated chromatin conformations, e.g., ChIA-PET and hiChIP, and will bring 1D RNP interaction studies to higher dimensions. In aim 2, applying PARIS and SHARC to chromatin associated RNAs, and the new method SHARCLIP to key splicing regulators, we will build transcriptome-wide RNA-structure and RNA-protein interaction models and deconvolve their dynamics in 2 ENCODE cell lines HepG2 and K562, as well as 6 brain cell lineages from a human iPS differentiation system. In aim 3, further integrating cell type specific AS programs and disease variants implicated in splicing, we will test whether mutations act by altering the structures. With the structure models as a guide, we will use a combination of CRISPR genome editing and structure-perturbing antisense oligos to test the roles of specific structures in regulating splicing in neuronal differentiation. Together, this proposal will produce a new technology that simultaneously capture multi-valent RNA-protein interactome and RNA structurome in vivo, establish a structure-based splicing code, provide an important resource to enable mechanistic studies of RNA processing, and pave the way for future therapeutic targeting of splicing.
