Elements of non-random structure in the unfolded states of proteins: location and possible implications for protein folding mechanisms

Abstract

Summary — This study introduces a simple computational procedure to search protein sequences for the segments with above average propensity to adopt non-random structures (which includes the native-like structure) in the unfolded state. The procedure consists of systematical conformational analysis of all overlapping hexapeptide segments in the protein sequence. The main aim of the computational approach is to determine the 3D structure most preferable for a given residue in the protein sequence, as determined by local interactions within the set of hexapeptides featuring the particular residue under consideration. Specifically, this study focuses on four types of “template” 3D structures that may be adopted by a hexapeptide, namely β-strand, α-helix, β-turn and the native-like structure of the folded state (assumed to be known). The study discusses also the possible importance of such segments for the different molecular mechanism of folding of the two prototypical proteins, namely the 65-residue barley chymotrypsin inhibitor 2 (CI2) and the 110-residue ribonuclease from Bacillus amyloliquefaciens (barnase). The computational results suggest that dynamic equilibrium in the unfolded state for the continuous fragment 6—27 in CI2 will likely prefer the native-like structure that may be preserved during folding. For barnase, on the contrary, dynamic equilibrium preferring the native-like structure most likely will occur in the unfolded state only at several small separate fragments, so the large non-native non-random segments of the unfolded state have to be restructured during folding.