Assessing the protein folding problem - theory and simulations

George Rose
Johns Hopkins University
Biophysics

The unfolded state of a protein is thought to involve an enormous, heterogeneous population of
conformers, including the native one. Protein folding is commonly viewed as an equilibrium
reaction, U <--> N, in which a single homogeneous population (N, the native state) is selectively
stabilized from this unfolded population (U, the unfolded state). As an example of this view,
consider a dilute solution of unfolded protein. Typically, such a solution might contain ~10^15
molecules. For a 100-residue protein in which each residue can occupy either of two possible
Ramachandran regions, the chain can visit 2^100 or approximately 10^30 conformers. Thus, under conditions that favor unfolding, it is quite possible that no two molecules will have the same conformation.

From this perspective, the central thermodynamic question is: how can the population overcome
conformational entropy (~30R ln (10), approximately 40 kcal/mol at room temperature from the backbone alone) in order to stabilize a single conformer uniquely? The central dynamic question is: as the population wends its way toward the native state, how can chains avoid getting stuck in adventitious, ~kT-sized traps?

Many investigators have provided answers to these questions. We offer yet another. Both explicit counting (Pappu et al., Proc. Natl. Acad. Sci. 2000, 97, 12565-70) and simple simulations using LINUS (Srinivasan & Rose, Proc. Natl. Acad. Sci. 1999, 96, 14258-14263) show that the size of conformational space is far smaller than previously thought. The number of
accessible conformations is winnowed by sterics. This effect was first noted in the alanine
dipeptide by Ramachandran, Ramakrishnan and Sasisekharan forty years ago (e.g.,
Ramachandran et al., J. Mol. Biol. 1963, 7, 95-9). However, systematic, local steric interactions
(Fitzkee and Rose, Protein Science, 13: 633-639) and specific solvation effects (Proteins, in
press) exert an influence beyond the dipeptide, and they serve to reduce the size of accessible
conformational space and organize the chain.

Presentation (PowerPoint File)

Back to Workshop III: Structural Proteomics