Studies of certain model systems — bacteria, yeast, flies and worms — that can be manipulated genetically have been extremely useful in outlining fundamental molecular signalling pathways that are common to most organisms. But the genomes of more complex creatures, including mammals, can contain an order of magnitude more genes, many of which do not resemble those known from the model systems. Without a means of directly regulating the expression of these genes, it has been difficult to dissect their roles, particularly if they are involved in development- or tissue-specific processes. That looks set to change, however, thanks to an exciting new technique described by Laising Yen and colleagues elsewhere in this issue (Nature 431, 471–476; 2004).

The discovery of ribozymes — RNA sequences that can mediate their own cleavage, in the absence of any protein cofactor — showed that RNAs had the potential to act as regulatory molecules. This potential was realized with the identification of riboswitches, messenger RNAs containing specialized ribozyme sequences whose self-cleavage is prevented by the binding of a small molecule. Natural riboswitches that recognize ligands such as metabolic intermediates or nucleotides have been described in bacterial systems, where they function to regulate gene expression.

Yen and colleagues exploited this information to develop a way of regulating gene expression in mammalian cells (see illustration). The system consists of two parts: a ribozyme sequence that is both highly efficient and constitutively active in vivo, and a means of modulating the ribozyme's self-cleavage activity (using either a ligand or an oligonucleotide complementary to the ribozyme sequence). If a ribozyme-encoding DNA sequence is placed upstream of the coding region of a gene of interest, the gene's expression will be repressed unless the small inducer molecule or antisense oligonucleotide is provided.

Needless to say, a certain amount of fiddling was required to find a workable ribozyme–inducer pair. Once done, however, the system could regulate gene expression in many cultured cell types and in mice; expression was almost undetectable unless the inducer molecule was present. Proof of this principle was obtained with a transgene that the authors inserted into the cells.

In conjunction with other emerging technologies, such as aptamers, it may be possible to tailor gene-regulation systems to respond to virtually any small molecule or metabolite. If ribozyme sequences are incorporated into an endogenous gene, it should also be possible to monitor that gene's function during development or in specific tissues by supplying specific concentrations of intracellular molecules. And having different ribozyme sequences will allow several genes to be examined either singly or together.