The Substrates of Nonsense-Mediated mRNA Decay in Caenorhabditis elegans

Nonsense-mediated mRNA decay (NMD) is a conserved pathway that strongly influences eukaryotic gene expression. Inactivating or inhibiting NMD affects the abundance of a substantial fraction of the transcriptome in numerous species. Transcripts whose abundance is altered in NMD-deficient cells may represent either direct substrates of NMD or indirect effects of inhibiting NMD. We present a genome-wide investigation of the direct substrates of NMD in Caenorhabditis elegans. Our goals were (i) to identify mRNA substrates of NMD and (ii) to distinguish those mRNAs from others whose abundance is indirectly influenced by the absence of NMD. We previously demonstrated that Upf1p/SMG-2, the central effector of NMD in all studied eukaryotes, preferentially associates with mRNAs that contain premature translation termination codons. We used this preferential association to distinguish direct from indirect effects by coupling immunopurification of Upf1/SMG-2 with high-throughput mRNA sequencing of NMD-deficient mutants and NMD-proficient controls. We identify 680 substrates of NMD, 171 of which contain novel spliced forms that (i) include sequences of annotated introns and (ii) have not been previously documented in the C. elegans transcriptome. NMD degrades unproductively spliced mRNAs with sufficient efficiency in NMD-proficient strains that such mRNAs were not previously known. Two classes of genes are enriched among the identified NMD substrates: (i) mRNAs of expressed pseudogenes and (ii) mRNAs of gene families whose gene number has recently expanded in the C. elegans genome. Our results identify novel NMD substrates and provide a context for understanding NMD's role in normal gene expression and genome evolution.