The new view of proteins sees protein folding and binding as diffusion-like processes, pictorially described as gliding down the slopes of funnel-shaped energy landscapes . These energy landscapes are not smooth. Rather, they are scattered by energy barriers of various heights, which give them a rugged or rough appearance. Rough potentials generally occur in frustrated systems – systems which are unable to satisfy all their interactions in any given arrangement . In proteins, frustration arises from multiple, sequence-specific competing interactions and from topological constraints which are imposed by their polymeric nature . It was realized almost 20 years ago that roughness can be an important factor in determining mechanisms and rates of protein folding and binding. This is because diffusion in rough potentials can become extremely slow owing to trapping in local minima that traverse the surface. Rough energy landscapes have been demonstrated in simulations of small model proteins, including sequences designed to have a smooth folding potential. However, direct experimental measurements of protein energy landscape roughness have not yet been made.

Using a recent theory (Hyeon, C. & Thirumalai, D. Proc. Natl. Acad. Sci. USA 100, 10249 (2003)), we derived the energy landscape roughness of a bi-molecular protein complex from single-molecule dynamic force spectroscopy measurements conducted at different temperatures. The results yield an overall energy scale for roughness of over 5 kBT. indicating that fluctuations in the potential surface can be quite substantial. This work represents the first ever measurement of roughness of protein energy landscapes.