A strategy for the de novo design of helical proteins with stable folds

This paper describes peptide analogs and the design strategy that were used to facilitate the final construction of a de novo-designed protein (ALIN) whose stable tertiary fold has been determined recently by NMR spectroscopy. Previous studies have suggested that the main problem in the de novo design of proteins is the attainment of a protein with a defined fold. To effectively overcome this mainchain multiconformation problem, three related steps, with experimental evaluation of the design hypotheses for each step, were pursued in the design process. Firstly, 15-residue sequences with experimentally verified high helicites were selected for the helical regions. Secondly, hydrophobic and electrostatic interhelical interactions as well as an interhelical disulfide bridge were designed to favor an antiparallel configuration of the helix axis. Finally, a loop with sufficient flexibility was inserted to stabilize the helices in the desired orientation. To assess the design strategy, peptides corresponding to each design step were synthesized and their structures verified experimentally by far-UV CD. As anticipated, ALIN was the most helical, and the SS-bridged dimeric peptides were more helical than their monomeric counterparts. The van't Hoff enthalpy change for ALIN computed from the CD denaturation curve and assuming a two-state model was 50 kJ/mol, a value close to that observed for helical coiled-coils. Overall, this report shows that small, simple proteins can be built using the current knowledge of protein structures.