Identifying DNA-Binding Sites and Analyzing DNA-Binding Domains Using a Yeast Selection System

We describe genetic methods using yeast to analyze a DNA-binding protein to determine (i) the sequence of the DNA sites to which the protein binds and (ii) the location of the domain and specific amino acid residues in the protein responsible for DNA binding. These methods take advantage of the fact that a hybrid protein consisting of a particular DNA-binding domain and a transcriptional activation domain activates expression of a reporter gene that contains binding sites for the DNA-binding domain. We describe two applications of these methods. First, DNA fragments that contain binding sites for the DNA-binding protein of interest can be recovered from a library of fragments by their ability to mediate transcriptional activation of a reporter gene. If enough DNA fragments are identified, the consensus sequence of the DNA-binding site can usually be recognized. In addition, some of the DNA fragments may be derived from actual target genes regulated by the DNA-binding protein, and therefore these fragments might be used to identify such target genes. Second, a reporter gene whose expression inhibits cell growth and whose promoter contains binding sites for the DNA-binding protein can be used to select mutants defective in the DNA-binding domain. This procedure allows one to localize the DNA-binding domain within the protein and to identify amino acids important for DNA binding. The mutations that inactivate the DNA-binding domain are highly informative, since the method avoids the recovery of "uninteresting" mutations that simply destabilize the protein or prevent its synthesis. In principle, the methods we describe can be applied to any DNA-binding protein.