Analyzing Co-transcriptional Processing by Nanopore Sequencing

Abstract

Human genes contain many long introns with degenerate sequence information at splice sites, requiring sophisticated mechanisms to locate and coordinate the excision of multiple introns within the same pre-mRNA transcript. Fundamental aspects of this process remain unexplored due to a lack of quantitative approaches that monitor pre-mRNA splicing as transcripts are produced in living cells. In this dissertation, I describe an approach to directly sequence long nascent RNA molecules and the insights I learned from observing the evolution of RNA isoforms as they are transcribed and processed. To get this approach to work, I first revealed how combining two strategies for purifying nascent RNA, 4sU metabolic labeling and cellular fractionation, significantly improves the enrichment of unspliced nascent transcripts. Second, I adapted direct RNA nanopore sequencing for long nascent RNAs. Applying this technique to human cells revealed that most co-transcriptional splicing occurs after RNA polymerase II transcribes at least four kilobases of pre-mRNA, suggesting that mammalian splicing occurs distally from the transcription machinery. By contrast, I found that splicing in Drosophila cells typically occurs within two kilobases of transcribing short introns, indicating differential splicing kinetics across species and intron classes. Inhibition of the branch-site recognition complex SF3B globally abolished co-transcriptional splicing in both species demonstrating the potency of splicing inhibitors and validating the ability of this method to observe rapid changes in co-transcriptional splicing dynamics. Sequencing long isoforms of nascent RNA also reveals the order and coordination of intron splicing. I found that splicing order does not strictly follow the order of transcription and is influenced by cis-regulatory elements, especially in Drosophila. In human cells, introns with delayed splicing tend to neighbor alternative exons and are associated with RNA-binding factors, emphasizing the importance of trans-regulatory factors in mammalian splicing order. Moreover, neighboring introns in human cells tend to be spliced concurrently, implying that splicing occurs cooperatively. Through sequencing long nascent RNAs in their native context, this work unveils the organizational complexity of metazoan RNA processing and provides a new lens into the kinetics and order of co-transcriptional splicing on a global scale.