RNA transcripts that encode truncated proteins are eliminated from eukaryotic cells by a process known as mRNA surveillance. In yeast and Caenorhabditis elegans, classical genetic approaches have identified some of the genes required for this surveillance. A puzzle, however, is why mRNA surveillance exists at all. What are the natural targets? Mitrovich and Anderson provide some possible answers.

The authors reasoned that, in a C. elegans mRNA surveillance mutant ( smg-2 ), RNA transcripts that are targets for mRNA surveillance will tend to be present at higher levels, relative to wild-type strains. This was the basis for an RNA subtraction experiment and, sure enough, transcripts were identified that were enriched in the surveillance mutant.

Gene transcripts for four ribosomal proteins were repeatedly represented in the selected pool and, in all cases, the transcript was aberrant, containing part of an intron that would normally have been removed by splicing. The intronic sequence also contained in-frame stop codons. Overall, it seems that these ribosomal protein genes produce two types of transcript — a productive and an aberrant transcript — and the aberrant transcript is normally removed by mRNA surveillance.

Mitrovich and Anderson found that the intronic sequences present in the aberrant transcripts are highly conserved among a group of nematodes, suggesting that the sequences have functional significance, despite not encoding protein. The authors speculate that this function helps to ensure the tight regulation of ribosomal protein levels. When ribosomal protein levels rise, one way of reducing new protein synthesis is to bias splicing towards the aberrant transcripts, and the authors provide data to support this view. But then the cell has a problem — the aberrant transcript could encode a truncated, and possibly deleterious, protein. At that point, the surveillance mechanism steps in and removes the offending transcripts.