To provide a method for screening a DNA-bonding protein by which the protein capable of bonding to a specific target sequence, e.g. a cancer gene promoter sequence such as an HRE-I sequence can be selected at a high efficiency and obtain both a plasmid used therefor and a peptide capable of specifically bonding to the HRE-I sequence.

This plasmid is preferably used for screening a peptide having properties of bonding to a target gene and comprises a base sequence capable of coding a peptide to be tested for evaluating the degree of bonding to the target gene, a promoter containing the target gene and a reporter gene bonded to the promoter so as to enable the operation. The promoter contains the target gene so as to control the expression of the reporter gene according to the degree of properties of bonding the peptide to the target gene contained therein. Furthermore, the peptide comprises 7 residues having a specific sequence.