Species-specific factors mediate extensive heterogeneity of mRNA 3′ ends in yeasts

Abstract

Most eukaryotic genes express mRNAs with alternative polyadenylation sites at their 3' ends. Here we show that polyadenylated 3' termini in three yeast species (Saccharomyces cerevisiae, Kluyveromyces lactis, and Debaryomyces hansenii) are remarkably heterogeneous. Instead of a few discrete 3' ends, the average yeast gene has an "end zone," a >200 bp window with >60 distinct poly(A) sites, the most used of which represents only 20% of the mRNA molecules. The pattern of polyadenylation within this zone varies across species, with D. hansenii possessing a higher focus on a single dominant point closer to the ORF terminus. Some polyadenylation occurs within mRNA coding regions with a strong bias toward the promoter. The polyadenylation pattern is determined by a highly degenerate sequence over a broad region and by a local sequence that relies on A residues after the cleavage point. Many dominant poly(A) sites are predicted to adopt a common secondary structure that may be recognized by the cleavage/polyadenylation machinery. We suggest that the end zone reflects a region permissive for polyadenylation, within which cleavage occurs preferentially at the A-rich sequence. In S. cerevisiae strains, D. hansenii genes adopt the S. cerevisiae polyadenylation profile, indicating that the polyadenylation pattern is mediated primarily by species-specific factors.