Transcriptome-wide detection and characterization of pseudouridine mRNA modifications across diverse human cell lines

Abstract

Pseudouridine (Ψ) is among the most abundant RNA modification in the human transcriptome and plays crucial roles in modulating mRNA stability, enhancing translation efficiency, and optimizing mRNA structure. The accurate detection and characterization of Ψ on mRNA is hindered by several factors, including the low abundance of mRNA, tissue-specific and cell-to-cell heterogeneity, and technical limitations. Nanopore direct RNA sequencing (DRS) enables full-length transcript sequencing while preserving post-transcriptional modifications in their native context. This technology allows for transcriptome-wide, amplification-free characterization of modifications at single nucleotide resolution. Differential analysis across different cell types can enhance our understanding of how Ψ influences gene regulation. We previously developed Mod-p ID, leveraging the systematic U-to-C basecalling error caused by the presence of Ψ on a native RNA strand to detect and map relative Ψ occupancy across the human transcriptome by comparing to an unmodified transcriptome. This tool can answer fundamental questions regarding the prevalence, conservation, and cell type-specific expression of Ψ across diverse tissue types.In this dissertation, we 1) develop a universal negative control as a community resource to provide a foundation for transcriptome-wide detection and characterization of RNA modifications, and 2) apply this resource for a comprehensive Ψ analysis across diverse human cell types using Mod-p ID.We first generated and characterized a multicellular in vitro transcribed (IVT), unmodified transcriptome (pan-IVT) from six immortalized human cell lines derived from diverse tissue types to provide a comprehensive negative control for public use within the RNA modification field. IVT libraries consist of paired transcripts assembled from canonical nucleotides using poly-A selected RNA. We show that combining multiple cell lines increases transcriptome coverage, helping to enrich data sets used for RNA modification analysis.We then sequenced the native transcriptomes of six human cell lines using DRS and used the Mod-p ID algorithm to catalog the relative positional occupancy of Ψ in each cell line. We show that Ψ sites can be conserved across cell lines, suggesting the presence of Ψ is important for normal cellular function or is cell-type specific, indicating the Ψ may also be consequential for cell-specific function.The transcriptome-wide annotated datasets constructed in this dissertation provide scientists across the RNA modification field with a negative control to conduct their own RNA modification analysis with the pan-IVT and highlight biologically interesting Ψ sites for robust functionalization studies.