The disparate nature of “intergenic” polyadenylation sites

Publisher Website
Google Scholar
Abstract
The termination of mature eukaryotic mRNAs occurs at specific polyadenylation sites located downstream from stop codons in the 3′-untranslated region (UTR). An accurate delineation of these sites is essential for the study of 3′-UTR-based gene regulation and for the design of pertinent probes for transcriptome analysis. Although typical poly(A) sites are located between 0 and 2 kb from the stop codon, EST sequence analyses have identified sites located at unexpectedly long ranges (5–10 kb) in a number of genes. Here we perform a complete mapping of EST and full-length cDNA sequences on the mouse and human genome to observe putative poly(A) sites extending beyond annotated 3′-ends and into the intergenic regions. We introduce several quality parameters for poly(A) site prediction and train a classification tree to associate P-values to predicted sites. We observe a higher than background level of high-scoring sites up to 12–15 kb past the stop codon, both in human and mouse. This leads to an estimate of about 5000 human genes having unreported 3′-end extensions and about 3500 novel polyadenylated transcripts lying in present “intergenic” regions. These high-scoring, long-range poly(A) sites corresponding to novel transcripts and gene extensions should be incorporated into current human and mouse gene repositories.