Estimations of the size of nucleation regions in globular proteins

Folding of many single-domain proteins has been described using the nucleation-collapse (NC) mechanism. According to NC, folding (formation of secondary structures and tertiary interactions) and chain collapse occur synchronously upon formation of native-like structures involving a critical number of residues. Using simple nucleation theory together with structure-based thermodynamic data, the average size of the most probable nucleus N(R)*, for single-domain proteins, is estimated to be between 15 and 30 residues. We argue that finite-sized fluctuations in this estimate can be large so that nearly half of the residues of a 100 residue protein can be part of the folding nucleus. Inclusion of surface area changes in the folded and unfolded states are important in the determination of N(R)*.