WDR33 alternative polyadenylation is dependent on stochastic poly(a) site usage and splicing efficiencies

• Published in: RNA biology Vol. 21; no. 1; pp. 970 - 980
• Main Authors: Liu, Lizhi, Seimiya, Takahiro, Manley, James L.
• Format: Journal Article
• Language: English
• Published: United States Taylor & Francis 31.12.2024; Taylor & Francis Group
• Subjects: alternative polyadenylation; Alternative Splicing; CFIm25; CRISPR/Cas13; Exons; Humans; innate immune responses; innate immunity; Introns; minigene; Nuclear Proteins - genetics; Nuclear Proteins - metabolism; Poly A - genetics; Poly A - metabolism; Polyadenylation; polyadenylation site; protein isoforms; Protein Isoforms - genetics; Protein Isoforms - metabolism; Research Paper; RNA; RNA Splicing; RNA, Messenger - genetics; RNA, Messenger - metabolism; WDR33; alternative splicing; minigene; CFIm25; CRISPR/Cas13; polyadenylation site; innate immune responses; WDR33; alternative polyadenylation
• ISSN: 1547-6286; 1555-8584; 1555-8584
• DOI: 10.1080/15476286.2024.2408708

Transcripts from the human WDR33 gene, which encodes a central component of the mRNA polyadenylation (PA) machinery, are subject to alternative polyadenylation (APA) within promoter-proximal introns/exons. This APA, which itself involves usage of multiple PA sites, results in the production of two non-canonical protein isoforms, V2 and V3, that are functionally completely unrelated to the full-length protein, with roles in innate immunity. The mechanism and regulation of WDR33 APA are unclear. Here, we report that levels of the PA factor CFIm25 modulate V2 and V3 expression, and that PA site usage of both V2 and V3 varies in distinct immune responses. Using newly developed assays to measure splicing and PA site strength, we show that splicing of V2-associated intron 6 is inefficient, allowing V2 to be produced using weak PA sites. Usage of V3's strong PA sites, on the other hand, is relatively low, reflecting the high efficiency of intron 7 splicing coupled with dependency on usage of an alternative 3' splice site within the intron. Overall, our findings demonstrate that usage of WDR33 alternative PA sites is stochastic, dependent on a complex interplay between splicing and PA, and thus provide new insights into mechanisms underlying APA.