Function and regulation of poly(A)-tail length

Abstract

Poly(A) tails are found at the 3' ends of nearly all eukaryotic messenger RNAs (mRNAs) and long non-coding RNAs. The presence of a poly(A) tail promotes translation and inhibits decay of an mRNA, with both effects mediated through poly(A)-binding protein. However, an understanding of the relationship between the length of a poly(A) tail and these aspects of mRNA metabolism has been limited, primarily because of the lack of a technology that provides high-resolution poly(A)-tail length measurements in a global manner. This dissertation describes a new, high-throughput-sequencing-based method (PAL-seq) that measures the tails of individual mRNA molecules by coupling a fluorescence-based readout of poly(A)-tail length with sequencing of the poly(A)-proximal region. Using PAL-seq, we have found that poly(A)-tail lengths exhibit a notably poor correlation with translational efficiency (as measured by ribosome profiling) across genes in nearly all systems we have examined. In contrast, early zebrafish and Xenopus laevis embryos display a striking correlation (Spearman R > 0.6) that disappears at gastrulation. This developmental uncoupling of tail length and translational efficiency explains the different outcomes of microRNA (miRNA)-mediated poly(A)-tail shortening in zebrafish embryos before and after gastrulation, with translational repression being the predominant effect before and mRNA destabilization after. We have also observed that poly(A)-tail lengths do not correlate positively with mRNA half-lives in mammalian cells, and that miRNAs do not promote any apparent tail shortening in this setting. Since these results could be explained by differences in deadenylation rates, we performed a kinetic analysis in which we captured newly-made mRNAs of different age ranges. The deadenylation rates that we calculated after measuring tails over time correlated strongly with mRNA half-lives (Spearman R < -0.6), reinforcing the notion that tail shortening leads to mRNA downregulation. When we repeated the timecourse with prior overexpression of a miRNA, we found that miRNAmediated tail shortening was generally modest, but of a magnitude not significantly different from that expected given the accompanying decreases in mRNA stability.