Fluorinated proteins: from design and synthesis to structure and stability.

Fluorine is all but absent from biology; however, it has proved to be a remarkably useful element with which to modulate the activity of biological molecules and to study their mechanism of action. Our laboratory's interest in incorporating fluorine into proteins was stimulated by the unusual physicochemical properties exhibited by perfluorinated small molecules. These include extreme chemical inertness and thermal stability, properties that have made them valuable as nonstick coatings and fire retardants. Fluorocarbons also exhibit an unusual propensity to phase segregation. This phenomenon, which has been termed the "fluorous effect", has been effectively exploited in organic synthesis to purify compounds from reaction mixtures by extracting fluorocarbon-tagged molecules into fluorocarbon solvents. As biochemists, we were curious to explore whether the unusual physicochemical properties of perfluorocarbons could be engineered into proteins. To do this, we developed a synthesis of a highly fluorinated amino acid, hexafluoroleucine, and designed a model 4-helix bundle protein, α4H, in which the hydrophobic core was packed exclusively with leucine. We then investigated the effects of repacking the hydrophobic core of α4H with various combinations of leucine and hexafluoroleucine. These initial studies demonstrated that fluorination is a general and effective strategy for enhancing the stability of proteins against chemical and thermal denaturation and proteolytic degradation. We had originally envisaged that the "fluorous interactions", postulated from the self-segregating properties of fluorous solvents, might be used to mediate specific protein-protein interactions orthogonal to those of natural proteins. However, various lines of evidence indicate that no special, favorable fluorine-fluorine interactions occur in the core of the fluorinated α4 protein. This makes it unlikely that fluorinated amino acids can be used to direct protein-protein interactions. More recent detailed thermodynamic and structural studies in our laboratory have uncovered the basis for the remarkably general ability of fluorinated side chains to stabilize protein structure. Crystal structures of α4H and its fluorinated analogues show that the fluorinated residues fit into the hydrophobic core with remarkably little perturbation to the structure. This is explained by the fact that fluorinated side chains, although larger, very closely preserve the shape of the hydrophobic amino acids they replace. Thus, an increase in buried hydrophobic surface area in the folded state is responsible for the additional thermodynamic stability of the fluorinated protein. Measurements of ΔG°, ΔH°, ΔS°, and ΔCp° for unfolding demonstrate that the "fluorous" stabilization of these protein arises from the hydrophobic effect in the same way that hydrophobic partitioning stabilizes natural proteins.