beta-hairpin-forming peptides; models of early stages of protein folding

Abstract

Formation of beta-hairpins is considered the initial step of folding of many proteins and, consequently, peptides constituting the beta-hairpin sequence of proteins (the beta-hairpin-forming peptides) are considered as models of early stages of protein folding. In this article, we discuss the results of experimental studies (circular-dichroism, infrared and nuclear magnetic resonance spectroscopy, and differential scanning calorimetry) of the structure of beta-hairpin-forming peptides excised from the B1 domain of protein G, which are known to fold on their own. We demonstrate that local interactions at the turn sequence and hydrophobic interactions between nonpolar residues are the dominant structure-determining factors, while there is no convincing evidence that stable backbone hydrogen bonds are formed in these peptides in aqueous solution. Consequently, the most plausible mechanism for folding of the beta-hairpin sequence appears to be the broken-zipper mechanism consisting of the following three steps: (i) bending the chain at the turn sequence owing to favorable local interactions, (ii) formation of loose hydrophobic contacts between nonpolar residues, which occur close to the contacts in the native structure of the protein but not exactly in the same position and, finally, (iii) formation of backbone hydrogen bonds and locking the hydrophobic contacts in the native positions as a hydrophobic core develops, sufficient to dehydrate the backbone peptide groups. This mechanism provides sufficient uniqueness (contacts form between residues that become close together because the chain is bent at the turn position) and robustness (contacts need not occur at once in the native positions) for folding a beta-hairpin sequence.

Figure 1

Plots of the molar ellipticities of a 16-residue peptide corresponding to a variant of the G-peptide at three different wavelengths (λ₁ = 190 nm, λ₂ = 220 nm, λ₃ = 229 nm) vs. temperature. The folding temperatures, estimated by using the two-state model [23], are T_{λ=190 nm} = 311 K, T_{λ=220 nm} = 303 K. Because of the absence of an inflection point in the respective curve, the folding temperature could not be determined for λ = 229 nm..

Figure 2

Plot of the chemical shifts of the δ and ε protons of the tyrosine residue in the 16-residue peptide corresponding to a variant of the G-peptide versus temperature. The folding temperatures, estimated by using the two-state model [23], are T_δ = 284 K and T_ε = 306 K.

Figure 3

The heat-capacity curve obtained by using the DSC technique for a 16-residue peptide corresponding to a variant of the G-peptide [23].

Figure 4

The heat-capacity curve for an 8-residue-long variant of the G-peptide recorded in water at pH = 6.57 with structures calculated by using NMR restraints determined at T = 283, 305, 313, and 323 K [25].

Figure 5

Mechanisms of structure formation of the G-harpin peptides: a) zipper mechanism [20,55,56]; b) hydrophobic-collapse mechanism [57]; c) broken-zipper mechanism [27]. Black circles represents nonpolar residues, thick horizontal bars represents hydrophobic interactions between nonpolar residues, thin horizontal lines represents hydrogen bonds, dotted line represents fragment of the sequence which has turn propensity, and arrows represents direction of folding/unfolding process.