O-GlcNAc regulates gene expression by controlling detained intron splicing

Abstract

<jats:title>ABSTRACT</jats:title><jats:p>Intron detention in precursor RNAs serves to regulate expression of a substantial fraction of genes in eukaryotic genomes. How detained intron (DI) splicing is controlled is poorly understood. Here we show that a ubiquitous post-translational modification called O-GlcNAc, which is thought to integrate signaling pathways as nutrient conditions fluctuate, controls detained intron splicing. Using specific inhibitors of the enzyme that installs O-GlcNAc (O-GlcNAc transferase, or OGT) and the enzyme that removes O-GlcNAc (O-GlcNAcase, or OGA), we first show that O-GlcNAc regulates splicing of the highly conserved detained introns in <jats:italic>OGT</jats:italic> and <jats:italic>OGA</jats:italic> to control mRNA abundance in order to buffer O-GlcNAc changes. We show that <jats:italic>OGT</jats:italic> and <jats:italic>OGA</jats:italic> represent two distinct paradigms for how DI splicing can control gene expression. We also show that when DI splicing of the O-GlcNAc-cycling genes fails to restore O-GlcNAc homeostasis, there is a global change in detained intron levels. Strikingly, almost all detained introns are spliced more efficiently when O-GlcNAc levels are low, yet other alternative splicing pathways change minimally. Our results demonstrate that O-GlcNAc controls detained intron splicing to tune system-wide gene expression, providing a means to couple nutrient conditions to the cell’s transcriptional regime.</jats:p>

CRISPR-Cas9-mediated depletion ofO-GlcNAc hydrolase and transferase for functional dissection ofO-GlcNAcylation in human cells

“…OGA knockout resulted in a substantial reduction of OGT expression, while disruption of OGT expression led to undetectable OGA levels. Previously it was shown that prolonged OGT inhibition leads to fewer productive OGA transcripts 32 , consistent with the CRISPR-Cas9 data presented in this manuscript which shows that OGA expression is abolished by OGT knockdown.…”

“…OGA contains two introns between exons 10 and 12 split by exon 11 and these introns can undergo retention. However, contrary to OGT regulation, transient retention of these introns during splicing ultimately leads to a functional transcript and higher OGA expression in conditions of elevated O-GlcNAc modification, while exon 11 skipping and introduction of a premature stop-codon occurs at low O-GlcNAc levels leading to decreased OGA expression 32 .…”

“…OGA-null and OGT-depleted cell lines generated by CRISPR-Cas9 could help to define the mechanism of homeostatic regulation between OGA and OGT. It has been shown that intron retention between exons 4 and 5 (which encompasses a premature stop-codon) of the OGT transcript negatively regulates its expression in response to increased O-GlcNAc modification 32,33 .…”

“…Similarly, transcriptional regulation of OGA expression via retained introns was reported as part of a feedback regulation loop 32 . OGA contains two introns between exons 10 and 12 split by exon 11 and these introns can undergo retention.…”

“…Importantly, the retained intron regions in OGA and OGT are conserved in vertebrates, implying a common mechanism 32 . Cell lines described here could help to identify the factor, or multiple factors (such as O-GlcNAc sensors), responsible for the regulation of OGA/OGT splicing and expression.…”