Origins and Implications of Intron Retention Quantitative Trait Loci in Human Tissues

Abstract

Intron retention is a type of alternative splicing in which introns remain unspliced in mature RNA transcripts. In order to explore the landscape and consequences of genetically regulated intron retention, we perform an intron retention quantitative trait loci (irQTL) analysis in 49 human tissues across 838 individuals. We identify 8,624 unique intron retention events associated with genetic polymorphisms. 1,369 irQTLs (16%) are also associated with genome-wide association study (GWAS) traits, suggesting that genetically regulated intron retention may play a causal role in the regulation of these GWAS traits. 1,999 irQTLs (23%) colocalize with eQTLs to their respective gene, suggesting that intron retention and steady-state gene expression levels are regulated by shared sets of genetic variants. Through a colocalization analysis of irQTL:eQTL events, we demonstrate that irQTLs are sufficient to generate eQTLs when one of the alternatively spliced transcripts is preferentially targeted by the nonsense mediated decay (NMD) pathway. Surprisingly, for intron retention events whose potential NMD effects can be confidently predicted based on their positions within known gene annotations, we find that 58.8% (923/1570) of the colocalized irQTL and eQTL pairs show effect size directions that are discordant with the NMD model. Moreover, we find that irQTLs are significantly more likely to occur in the same gene with the same effect size direction as compared to exon skipping QTLs. Through mathematical modeling and analysis of experimental perturbation data, we provide evidence that eQTLs are able to generate irQTLs by altering the steady state ratios of spliced and unspliced transcripts, and we postulate that this mechanism may partially underlie the widespread intron retention observed previously in various biological conditions. Taken together, these results show that intron retention and steady state gene expression levels are closely intertwined to regulate phenotypic traits.