How do you deal with messenger RNAs that just don't know when to quit? Cells can create transcripts that contain no stop codons for the translation machinery to recognize, which can lead to the synthesis of abnormal and potentially harmful proteins. But according to two reports in Science, cells recognize these so-called 'nonstop' mRNAs and can destroy them using a mechanism called nonstop decay.

The cell has evolved a remarkable array of quality-control mechanisms to ensure gene expression occurs correctly, such as nonsense-mediated mRNA decay (NMD), which detects mRNAs that contain premature termination codons and stops the formation of truncated proteins.

But what happens in the opposite case, when no stop codon exists? Frischmeyer and colleagues constructed a nonstop-PGK1 construct — in which all in-frame termination codons were removed — and found that these 'nonstop' transcripts were just as unstable in Saccharomyces cerevisiae cells as a nonsense form of the gene.

However, the nonstop transcripts were still unstable in yeast mutants that lacked factors required for both the NMD pathway and for the destruction of normal mRNAs, yet required translation of the mRNA for the decay to occur (as shown by cycloheximide treatment, which inhibits protein synthesis, or depletion of charged tranfer RNAs) — so the destruction of nonstop transcripts had to be occurring through a different pathway.

The most obvious candidate is the exosome, a collection of proteins with a 3′–5′ exoribonuclease activity that processes RNAs, such as ribosomal RNA, small nuclear RNA and small nucleolar RNA. Van Hoof et al. showed that a yeast mutant for Ski4, one of the core exosome subunits that specifically disrupts cytoplasmic 3′–5′ degredation of mRNA, stabilized the nonstop-PGK1 at least sixfold. The nonstop transcripts were also stabilized by mutant forms of Ski7, which is related to the translation elongation factor EF1A and the translation temination factor eRF3, and is one of two other factors that are also required for exosome-mediated degradation. Transcriptional pulse–chase studies using transcripts with differing lengths of poly(A) tails suggested that the mechanism used to degrade the nonstop mRNAs starts at the 3′ end of the poly(A) tail.

Together, these data provide a model for how the cell detects and destroys nonstop mRNAs. When the ribosome reaches the end of the nonstop mRNA strand, the exosome homes in on the stalled ribosome. Ski7 associates with the cytoplasmic form of the exosome, which then degrades the mRNA, starting from the 3′ end of the poly(A) tail. Given the similarity of Ski7 to EF1A and eRF3 — which interact with the A site of ribosomes that contain a sense or nonsense codon, respectively — Ski7 might distinguish nonstop from normal mRNAs by binding to the empty A site of ribosomes that have reached the 3′ end of mRNAs.

Many questions remain but the authors say that nonstop decay might be a valuable and necesssary mechanism. It could be required for mRNAs containing a 3′-end processing signal located upstream of the termination codon (as is found in 1.2% of expressed sequence tags from S. cerevisiae), or the nonstop mRNAs might be created when 3′-end formation takes place upstream of the normal termination after RNA polymerase pausing, when ribosomes pause at rare codons or normal termination codons, or 3′–5′ decay is initiated on ribosome-bound mRNA. Nonstop decay is also initiated after manipulations that increase read-through of termination codons, as occurs after administration of aminoglycoside antibiotics.