Cotranscriptional splicing efficiencies differ within genes and between cell types

Pre-mRNA splicing is carried out by the spliceosome and involves splice site recognition, removal of introns, and ligation of exons. Components of the spliceosome have been shown to interact with the elongating RNA polymerase II (RNAPII), which is thought to allow splicing to occur concurrently with transcription. However, little is known about the regulation and efficiency of cotranscriptional splicing in human cells. In this study, we used Bru-seq and BruChase-seq to determine the cotranscriptional splicing efficiencies of 17,000 introns expressed across six human cell lines. We found that less than half of all introns across these six cell lines were cotranscriptionally spliced. Splicing efficiencies for individual introns showed variations across cell lines, suggesting that splicing may be regulated in a cell type–specific manner. Moreover, the splicing efficiency of introns varied within genes. The efficiency of cotranscriptional splicing did not correlate with gene length, intron position, splice site strengths, or the intron/neighboring exons GC content. However, we identified binding signals from multiple RNA-binding proteins (RBPs) that correlated with splicing efficiency, including core spliceosomal machinery components—such as SF3B4, U2AF1, and U2AF2 showing higher binding signals in poorly spliced introns. In addition, multiple RBPs, such as BUD13, PUM1, and SND1, showed preferential binding in exons that flank introns with high splicing efficiencies. The nascent RNA splicing patterns presented here across multiple cell types add to our understanding of the complexity in RNA splicing, wherein RNA-binding proteins may play important roles in determining splicing outcomes in a cell type– and intron-specific manner.