Stage-Specific Alternative Polyadenylation During Human Neural Differentiation Revealed by Integrated Long- and...

Human embryonic stem cell (hESC) neural differentiation involves extensive APA; however, reliance on short-read sequencing in prior studies has offered only a limited view of its complexity and dynamic regulation. Here, we integrated Oxford Nanopore (ONT) long-read sequencing with Illumina short-read sequencing to systematically map the APA landscape during early hESC neural differentiation. Our hybrid approach uncovered remarkable transcriptomic complexity, identifying 20,823 novel transcripts and 8241 previously unannotated poly(A) sites (PASs). We characterized distinct dynamic patterns of 3' UTR-APA across differentiation and pinpointed SOX11 as a key APA-regulated target. Furthermore, we observed stage-specific enrichment of intronic APA in NPCs, as exemplified by SLC1A3, and performed a comprehensive, large-scale identification of high-confidence exon APA events. These results substantially expand the catalog of PASs during human neural differentiation and provide new insights into how APA-mediated post-transcriptional regulation contributes to cell-fate decisions.