Engineering DNA-Binding Proteins Based on the Helix-Turn-Helix Motif

The goal of this work was to study the highly specific nature of protein-DNA recognition, using a "rational design approach", i.e. by attempting to construct artificial DNA-binding proteins. The N-terminal domain of the 434 bacteriophage cl repressor, a five-helix domain with a helix-tum-helix motif (HTH) was used as an example. As most known DNA-binding proteins are dimers or contain multiple DNAbinding modules, we decided to design covalently dimerized bipartite DNA-binding proteins using elements of the 434 repressor. Specific DNA-binding was assayed by biochemical methods and protein-DNA interactions studied by biophysical te.chniques using circular dichroism (CD) difference spectroscopy. The new polypeptides were designed on the basis of the 3D structure of the 434 repressor, usmg computer graphics and conformational energy minimization/molecular dynamics techniques. The designed peptides were prepared by solid phase peptide synthesis or by recombinant DNA techniques. In one set of experiments (based on a "minimalist approach"), the HTH motif and its flanking elements were covalently .dimerized with synthetic linkers. The resulting molecules (P4 to P6) were prepared by solid phase peptide synthesis, in the form of dyad-symmetric branched peptides, connected through their C-termini. However, none of these molecules showed specific DNA binding in the presence of competitor DNA. In another set of experiments, the entire N-terminal domain was dimerized either in a dyad-symmetric fashion, using peptide synthesis (P7), or as direct sequence repeats produced with recombinant DNA techniques (RR69). In P7, a symmetric linker was used that consisted of two 6-aminohexanoic acid residues connected to both amino groups of a central lysine molecule. RR69 was prepared by recombinant DNA methods, and the two N-terminal domains were connected with the internal 69- 89 sequence of the 434 repressor. In gel-mobility shift experiments, both of these new molecules showed specific DNA binding, and they also showed the expected sequence specificity in footprinting, methylation interference and in vitro transcription inhibition experiments. These results show that dimeric architecture is crucial for achieving native-like propenies in artificial DNA-binding proteins. The conformational changes of the peptides occurring on DNA binding were studied using CD difference spectroscopy. Under the experimental conditions used, the monomeric N-terminal domain (Pl) does not seem to interact with DNA since ahelical induction was not detected. In contrast, both P7 and RR69 showed strong increase in their a-helical content when exposed to equimolar amounts of cognate DNA. Interestingly, a similar (about 30 % weaker) induction was caused by noncognate DNA. These results indicate that, in contrast to current views, conformational change is a part of specific DNA recognition by the HTH modules, and may also be crucial in scanning the DNA molecule in search of the specific binding sites.