A salt bridge stabilizes the helix formed by isolated C-peptide of RNase A (mechanism of protein folding/side-chain interaction/enthalpy of folding/code for helix formation)

C-peptide, which contains the 13 NH2-terminal residues of RNase A, shows partial helix formation in water at low temperature (1?C, pH 5, 0.1 M NaCl), as judged by CD spectra; the helix is formed intramolecularly [Brown, J. E. & Klee, W. A. (1971) Biochemistry 10, 470-476]. We find that helix stability depends strongly on pH: both a protonated histidine (residue 12) and a deprotonated glutamate (residue 9 or 2 or both) are required for optimal stability. This information, together with model building, suggests that the salt bridge Glu-9* * * His-12+ stabilizes the helix. Formation ofthe helix is enthalpy driven [van't Hoff AH, -16 kcal/ mol (1 cal = 4.18 J)] and the helix is not observed above 30?C. Proton NMR data indicate that several side chains adopt specific conformations as the helix is formed. These results have two implications for the mechanism of protein folding. First, they indicate that short a-helices, stabilized by specific side-chain interactions within the helix, can be stable enough in water to function as folding intermediates. Second, they suggest that similar experiments with peptides of controlled amino acid sequence could be used to catalogue the intrahelix interactions that stabilize or destabilize a-helices in aqueous solution. These data might provide the code relating amino acid sequence to the locations of a-helices in proteins. Short a-helices, of the size range usually found in globular proteins (6-20 residues), are highly unstable in water in the absence of specific stabilizing interactions, according to data obtained with random copolymers by the "host-guest" technique (1). For short helices (n 99% pure) and were used to correct the CD spectra of the lactone. In NMR experiments, the resonances of the lactone and carboxylic acid forms are well resolved at low temperatures for several resonances, including the ones studied here. C-peptide concentrations were determined by the ninhydrin method (5), using leucine as a standard. To minimize interconversion between the two forms of C-peptide, the stock solutions of C-peptide lactone and carboxylic acid were kept at pH 2 and pH 10, respectively. HPLC. The lactone and carboxylic acid forms of C-peptide were analyzed in a Waters HPLC system (model 273 with a U6K injector, model 730 data module, and model 441 detector) using a reverse-phase column (Waters ,uBondapak C18) with detection at 214 nm. Separation was done isocratically at 25?C using a mobile phase of 6% isopropanol/10 mM ammonium acetate, pH 5.3. * Present address: Institute of Biochemistry and Biophysics, Polish Academy of Sciences, Ul. Rakowiecka 36, 02-532 Warsaw, Poland.