Conformational consequences of i, i + 3 cystine linkages: nucleation for alpha-helicity?

Methods to introduce specific secondary structural elements into peptides and proteins are vital for the rational design of peptide and non-peptide drug candidates as well as in the de novo design of proteins. Here the incorporation of a disulfide linkage between cysteine residues spaced three amino acids apart (i, i + 3) as a method to induce helicity is examined. Two dodecamer peptides, A-V-S-E-C-Q-L-C-H-D-K-G-NH2, differing in the chirality of the cysteine at the fifth position (the i position), have been synthesized and conformationally studied both in the linear and cyclized form. This peptide sequence, derived from the N-terminal sequence of parathyroid hormone related protein, does not form helices, even as part of the 1-34 fully active domain of the protein. The four analogs (two cyclic and two linear) were analyzed both in aqueous solution and in the presence of sodium dodecyl sulfate micelles. In aqueous solution the linear peptides display no evidence for secondary structure, while the cyclization induces a turn centered about the cysteine residues. In the presence of micelles the linear form of the peptides adopts bent conformations, containing turns, but results from both NMR and CD provide no evidence of helices. The oxidized L,L-peptide in the micellar solution does not present a well defined conformation, although the presence of one helical turn is evident. The cyclic D,L analog adopts a helical structure (not an alpha-helix) extending from residue 2 to 9, with non-standard phi, psi values caused by the presence of the D-amino acid. These results clearly illustrate that the ability of D-Cys(i), Cys(i + 3) cyclization to initiate helix formation depends greatly on the solvent used. Therefore, any drug-design principle utilizing this modification for helix nucleation must keep the environment in which the peptide is biologically active in mind.