Translation of Small Open Reading Frames within Unannotated RNA Transcripts in Saccharomyces cerevisiae

High-throughput gene expression analysis has revealed a plethora of previously undetected transcripts in eukaryotic cells. In this study, we investigate >1,100 unannotated transcripts in yeast predicted to lack protein-coding capacity. We show that a majority of these RNAs are enriched on polyribosomes akin to mRNAs. Ribosome profiling demonstrates that many bind translocating ribosomes within predicted open reading frames 10–96 codons in size. We validate expression of peptides encoded within a subset of these RNAs and provide evidence for conservation among yeast species. Consistent with their translation, many of these transcripts are targeted for degradation by the translation-dependent nonsense-mediated RNA decay (NMD) pathway. We identify lncRNAs that are also sensitive to NMD, indicating that translation of noncoding transcripts also occurs in mammals. These data demonstrate transcripts considered to lack coding potential are bona fide protein coding and expand the proteome of yeast and possibly other eukaryotes. [Display omitted] •RNAs identified in yeast associate with polyribosomes analogously to mRNA•Ribosome profiling defines short open reading frames 10–96 codons in size•RNAs are sensitive to translation-dependent nonsense-mediated RNA decay•Sensitivity of noncoding RNA to NMD is conserved in mammals Genome-wide gene expression analyses have revealed that eukaryotes express a wide range of previously unidentified RNA transcripts, many of which are considered to lack protein-coding capacity. In this study, Smith et al. identify hundreds of unannotated RNA transcripts in yeast, many of which are shown to be engaged by the translational machinery and harbor short open reading frames. These findings provide evidence for translation of predicted noncoding RNA and unveil additional coding capacity within the yeast genome.