Sequence-Dependent & -Independent Effects of Intron-Mediated Enhancement (IME)

Abstract

Introns are ubiquitous features of eukaryotic genes, and their precise removal from pre-mRNA transcripts by the spliceosome is an essential step in gene expression. Genomic deletion of an intron from a gene tends to reduce its expression, and addition of an intron tends to increase it. This phenomenon, termed Intron-Mediated Enhancement (IME), has been observed in many organisms, genes, and introns. IME can act at multiple levels to increase transcription rate, processing rate, export efficiency, translational efficiency, and stability of the processed mRNA. These stimulatory effects range across orders of magnitude depending on the context, and also on the identity of the intron, as has been shown in Arabidopsis thaliana. Presently, little is known how intron sequence may determine the mode or magnitude of effect on gene expression output in animals. In this study we report the design and execution of several massively parallel reporter assays (MPRAs), interrogating the effect of tens of thousands of synthetic and natural intron sequences on gene expression in the human HEK293T and HeLa cell lines. We observe that even with random internal sequence, most of these introns splice well and trigger IME. In the primary tested context, the average intron stimulates an eight-fold increase in both mRNA and protein output over intronless controls, suggesting that the enhancement is largely at the level of mRNA accumulation. We analyze the sequence features associated with highly-enhancing introns and demonstrate that the poly-uridine (polyU) content of an intron is positively correlated with its impact on host gene mRNA and protein level. In a second library of natural intron sequences, we observe that U12-type introns do not stimulate IME, while U2-type introns universally do. Surprisingly, we observe in both MPRAs that the enhancement from random introns is similar to or greater than the enhancement from natural sequences. In sum, we have developed a robust experimental platform for interrogating the sequence-activity relationship of IME, and used it to uncover new insights into this unsung sculptor of eukaryotic gene expression.