Protein structure and folding – simplicity within complexity

As in any challenging science problem, a sensible approach is to consider the hints provided by nature. In spite of their diversity, globular proteins share some amazing commonalities: they fold rapidly and reproducibly (Anfinsen, 1973); their folding is driven by the affinity or lack of certain side chains and the backbone to the surrounding water; they are able to catalyze reactions and speed them up by many orders of magnitude; the folded forms of proteins are comprised of helices and sheets – the scaffolding which is provided by hydrogen bonds (Pauling & Corey, 1951; Pauling, Corey, & Branson, 1951); and proteins seem to have a generic propensity to aggregate and form amyloid (Dobson, 2003), which, in turn, is implicated in debilitating diseases. A remarkable experimental observation is that the total number of distinct folds seems to be of the order of just a few thousand (Chothia & Finkelstein, 1990). More than 70 years ago, Bernal wrote a remarkable paper entitled, “Structure of Proteins” (Bernal, 1939), whose abstract reads: The structure of proteins is the major unsolved problem on the boundary of chemistry and biology to-day. We have not yet found the key to the problem, but in recent years a mass of new evidence and new lines of attack have enabled us to see it in a far more concrete and precise form, and to have some hope that we are near to solving it. Bernal goes on to suggest: Any effective picture of protein structure must provide at the same time for the common character of all proteins as exemplified by their many chemical and physical similarities, and for the highly specific nature of each protein type. A striking feature of proteins is that a large number of sequences adopt one of a relatively small number of putative protein folds (Chothia & Finkelstein, 1990). This has led to thinking about how one may determine the structure of a given sequence from knowledge of the menu of the folds. A crucial point to consider is the nature of the unfolded state of a protein (Rose, 2002). Does a given sequence already have the address of its native-state structure encoded in its nascent conformation? Coarse-grained models of proteins have been studied over the years (Miyazawa & Jernigan, 1985) with varying degrees of approximation to determine generic and transferable effective interactions between pairs of amino acids that ensure that the energy of a sequence in its native state is lower than in accessible competing conformations. Such knowledge-based approaches in their simplest form (Jha, Vishveshwara, & Banavar, 2010) approximate a protein native-state structure as being made up of a gas of amino acids and study the propensity for specific pairs of amino acids to be spatially near each other compared to what one would expect from random considerations. More refined calculations involving the side-chain atoms show that a protein native-state structure is a delicate packing of all the constituent atoms respecting steric constraints (Ramachandran, Ramakrishnan, & Sasisekharan, 1963), yielding an emergent connected community arising through short-range interactions screened by the surrounding water molecules (Deb, Vishveshwara, & Vishveshwara, 2009). Ben-Naim (2012) points out that the Gibbs energy landscape may be characterized by many minima (see Figure 5 of his paper), some of which may be rarely visited and others which may be inaccessible because of high barriers. The experimental observation that a given protein or sequence of amino acids folds reproducibly and rapidly into a specific native fold, is encapsulated in the folding funnel picture (Dill & Chan, 1997; Wolynes, Onuchic, & Thirumalai, 1995). Were there multiple competing and accessible minima in the Gibbs energy landscape, one would not observe reproducible folding. Were there not a smooth and accessible funnel-like basin of attraction for the native fold, one would not observe