Collaborative Research: Connecting the sequence logic of RNA splicing to nuclear localization

New York University, New York NY

Investigators

Abstract

In eukaryotic cells, mature messenger RNAs (mRNA) are generated from newly transcribed pre-mRNA through an essential processing step called RNA splicing. During splicing, parts of the pre-mRNA (known as introns) are removed, and the remaining parts (exons) are joined together…but not every intron is removed from every pre-mRNA. This collaborative project aims to understand the sequences in pre-mRNA that determine which exons are removed in specific mRNAs. The PIs will deploy high-throughput assays where millions of sequences are tested in parallel, and analyzing the results using interpretable machine learning techniques. This project will also reveal novel mechanisms contributing to splicing decisions. Specifically, it will reveal the role that certain small bodies next to the nucleus, known as “nuclear speckles,” play in RNA splicing. In addition, the proposed research will provide substantial resources for the development and dissemination of an interdisciplinary curriculum for graduate and undergraduate education. The project will offer recruitment, involvement, and career development of students and scholars at different educational stages. It will increase the participation of studnets and trainees biological and computational research and education via the training of graduate and postdoctoral researchers. The project will also include outreach to high school students and undergraduates. RNA splicing occurs in a sequence-dependent manner following logic encoded in the “splicing code”. Besides the conserved sequences (5’ and 3’ splice sites, branchpoint) that are required for the splicing reaction, additional logic is encoded in splicing regulatory elements. These sequence elements are recognized by various splicing factor proteins, including SR proteins and heterogeneous nuclear ribonucleoproteins (hnRNPs). SR proteins and hnRNP proteins regulate splicing decisions in a sequence- and context-dependent manner. Specifically, SR proteins tend to promote splicing if their binding motifs are in exons and tend to repress splicing if their binding motifs are in introns; whereas hnRNP proteins have the opposite effect. However, the mechanisms underlying context-dependent regulatory effects remain to be fully revealed. Interestingly, splicing factors exhibit distinct subcellular locations relative to nuclear speckles; SR proteins are enriched in nuclear speckles, whereas hnRNP proteins are uniformly distributed in the nucleoplasm. These observations lead us to hypothesize that positioning of RNAs with respect to nuclear speckles serves as a bridge between sequence features and splicing outcomes. This project will test this hypothesis, and explore the mechanisms of splicing and the function of nuclear speckles by combining two complementary approaches: (1) Identifying the relationship between sequences and splicing outcomes using interpretable machine learning-based analysis of massively parallel reporter assays (MPRA); (2) Characterizing the relationship between sequences and positioning relative to nuclear speckles through imaging RNA substrates, spliceosomal components, and splicing factors. This research will reveal novel mechanisms contributing to splicing decision, specifically through sequence-encoded intra-speckle RNA positioning. This project is co-funded by the Cellular Dynamics and Function and Genetic Mechanisms programs of the Molecular and Cellular Biosciences Division. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.