If genes could post personal ads, they would read something like this: “Good-looking promoter seeking transcription factor for binding experience with promise of regulated gene expression.” Who wouldn't be interested in the reply? In the absence of such correspondence, biochemical screens have to reveal matches made between transcription factors (TFs) and promoters. In a recent article in Genome Research, Marian Walhout from the University of Massachusetts introduced a modification to the yeast one-hybrid (Y1H) assay, which opens the door to the high-throughput characterization of TF-promoter interactions (Deplancke et al., 2004).

In a traditional Y1H assay, a yeast strain is transformed with a bait vector containing multiple copies of short cis-regulatory elements followed by a reporter gene and a prey vector that expresses the TF. If the TF binds the regulatory sequence, expression of the reporter gene is initiated, and its product confers a growth advantage to the yeast strain. Using the Gateway system, which is based on homologous recombination rather than restriction enzyme digests, Walhout's team was able to greatly increase the speed of cloning so that they can now generate TF prey libraries and large numbers of bait clones. Also, instead of using cis-regulatory elements, their baits consist of whole promoter sequences. As Walhout points out, “This opens up a whole new level of possibility for systems biology....You can just take a promoter without knowing anything about it and try to identify transcription factors that can bind to that promoter.”

She chose C. elegans as a model system because the genome is very well annotated, but she sees no reason why this approach could not be used in other organisms as well. Initially, Walhout screened only four promoters against a TF library, thus validating the technique, but in the near future her group will apply this method to high-throughput analysis of many promoters. When asked for the motivation behind the study, she replied, “We want to understand how differential gene expression is controlled in space and time during development and, long term, also during homeostasis and in disease.” A tall order, but with the new Y1H assay it is certainly doable.