Premature termination codons (PTCs) would spell disaster were it not for the quality control system of the cell, known as nonsense-mediated decay (NMD), which rapidly eliminates mRNAs with PTCs. Now, a report in the EMBO Journal by Izaurralde and colleagues shows that the NMD pathway in Drosophila seems to have elements of both the yeast and mammalian pathways.

A crucial step in NMD is the recognition of mRNAs that contain PTCs; so, how does a cell determine whether a stop codon is premature or legitimate? In mammalian cells, the presence of a downstream exon-junction complex (EJC) during translation termination provides warning of a premature stop codon. Components of the EJC interact with the UPF complex of proteins, which mediate NMD, triggering the rapid digestion (or degradation) of the offending transcript. In yeast, where there is usually no EJC, a cis-acting downstream sequence element (DSE) is the NMD trigger.

Previous studies have shown that three genes that are essential for NMD ( UPF1 , UPF2 and UPF3 ) are conserved in humans and yeast. However, other human genes ( SMG1 and SMG5–7) have no yeast orthologues.

To further investigate the molecular mechanism of NMD, Izaurralde and colleagues created a smart assay to determine whether or not specific genes were vital for NMD, based on Drosophila cell lines that constitutively express alcohol dehydrogenase ( adh ) mRNAs, into which PTCs had been introduced. If a gene was essential to NMD, then silencing using RNAi would increase levels of these reporter mRNAs. Using this technique, Gatfield et al. showed that Drosophila orthologues of UPF1–3, SMG1, SMG5 and SMG6 are essential for NMD. Intriguingly, no SMG7 orthologue was discovered.

Given the presence of components of the EJC in Drosophila, it was previously assumed that, as for mammalian cells, PTC detection would involve the EJC and exon–exon boundaries. However, Izaurralde and colleagues' reporter assay clearly showed that neither a downstream exon–exon boundary nor the EJC proteins that they tested were essential for NMD. So, against all expectations, Drosophila would seem to use an alternative — as yet unknown — method of PTC identification.

These studies provide us with a tantalizing glimpse into the evolution of the NMD mechanism, as Drosophila shares with yeast the ability to identify PTCs independently of exon boundaries, while at the same time requiring the products of SMG1, SMG5 and SMG6, in common with higher eukaryotes.