Uncovering the isoform-resolution kinetic landscape of nonsense-mediated mRNA decay with EZbakR

Gene expression is a product of RNA synthesis and degradation kinetics, which can vary among transcript isoforms of the same gene. Some isoforms contain premature termination codons (PTCs) and are degraded via nonsense mediated mRNA decay (NMD), while others lack these features and are more stable. However, probing the stability of individual isoforms remains challenging. Methods like nucleotide-recoding RNA-seq (NR-seq) enable insights into RNA kinetics, but existing bioinformatic tools are unable to provide robust, isoform-specific degradation rate constant estimates. We present a strategy to infer isoform-level kinetics from short-read NR-seq data, implemented in the EZbakR-suite, which allows us to identify high confidence NMD targets. We also show how custom annotations improve these analyses. We uncovered unexpected stability variability in transcripts with conserved PTC-containing exons and found that other decay pathways can supersede NMD. Our findings provide mechanistic insights into established NMD efficiency correlates and highlight transcript features promoting efficient decay. Overall design: HEK-293 cells were treated with DMSO or 1 µM SMG1i for 8 hours. For TimeLapse-Seq, indicated samples were treated with 100 µM 4-thiouridine for 2 hours. Total RNA was subjected to TimeLapse chemistry and used for library preparation and short-read sequencing. Unlabeled total RNA samples were also used for long-read sequencing (Iso-seq).