SHORT REVIEW Structured Disorder and Conformational Selection Chung-Jung Tsai, 1 Buyong Ma, 2 Yuk Yin Sham, 2 Sandeep Kumar, 2 and Ruth Nussinov 1,3 * 1 Intramural Research Support Program—Science Application International Corporation (SAIC), Laboratory of Experimental and Computational Biology, NCI—Frederick, Frederick, Maryland 2 Laboratory of Experimental and Computational Biology, NCI—Frederick, Frederick, Maryland 3 Sackler Institute of Molecular Medicine, Department of Human Genetics and Molecular Medicine, Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel ABSTRACT Traditionally, molecular disorder has been viewed as local or global instability. Mol- ecules or regions displaying disorder have been considered inherently unstructured. The term has been routinely applied to cases for which no atomic coordinates can be derived from crystallized mol- ecules. Yet, even when it appears that the molecules are disordered, prevailing conformations exist, with population times higher than those of all alternate conformations. Disordered molecules are the out- come of rugged energy landscapes away from the native state around the bottom of the funnel. Rug- gedness has a biological function, creating a distri- bution of structured conformers that bind via confor- mational selection, driving association and multimolecular complex formation, whether chain- linked in folding or unlinked in binding. We classify disordered molecules into two types. The first type possesses a hydrophobic core. Here, even if the native conformation is unstable, it still has a large enough population time, enabling its experimental detection. In the second type, no such hydrophobic core exists. Hence, the native conformations of mol- ecules belonging to this category have shorter popu- lation times, hindering their experimental detec- tion. Although there is a continuum of distribution of hydrophobic cores in proteins, an empirical, sta- tistically based hydrophobicity function may be used as a guideline for distinguishing the two disor- dered molecule types. Furthermore, the two types relate to steps in the protein folding reaction. With respect to protein design, this leads us to propose that engineering-optimized specific electrostatic in- teractions to avoid electrostatic repulsion would reduce the type I disordered state, driving the mol- ten globule (MG) 3 native (N) state. In contrast, for overcoming the type II disordered state, in addition to specific interactions, a stronger hydrophobic core is also indicated, leading to the denatured 3 MG 3 N state. Proteins 2001;44:418 – 427. © 2001 Wiley-Liss, Inc. Key words: disorder; molten globule; stability; popu- lations; folding binding; conformational selection INTRODUCTION Previously, we have used the protein energy landscape concept to rationalize binding mechanisms and protein function. 1–4 We have focused on the energy landscape at and near the bottom of the funnels around the native conformation and on enzyme catalysis. Here we illustrate that the protein energy landscape theory is capable of accounting for protein function around the native state and in native disordered unfolded states. In either case, structured conformers bind via selection, with population shifts, rationalizing the rugged, rough folding funnels observed for disordered proteins. Furthermore, whether in ordered or disordered states, given structured conforma- tions prevail, albeit with (very) different population times, driving protein binding and complex formation through conformational selection. Here we argue that this is the function that dictates the extent of ruggedness. Hence, the ruggedness in regions away from the native folded states yields information on the type of function, such as in nucleic acid-binding domains or in proteins requiring cations. The protein energy landscape theory, or the folding funnel concept, 5–19 suggests that the most realistic model of a protein is a minimally frustrated heteropolymer with a rugged funnel-like landscape biased toward the native structure. 20 Here we focus on the rugged nature of the protein energy landscape away from native conformations. A rugged protein energy landscape has certain immediate Grant sponsor: BSF; Grant number 95-00208; Grant sponsor: Israel Science Foundation (Israel Academy of Sciences); Grant sponsor: Magnet; Grant sponsor: Ministry of Science (Israel); Grant sponsor: Tel Aviv University; Grant sponsor: Center of Excellence (Israel Academy of Sciences); Grant sponsor: National Cancer Institute (National Institutes of Health); Grant number: NO1-CO-56000. *Correspondence to: Ruth Nussinov, NCI—Frederick Building 469, Room 151, Frederick, MD 21702. E-mail: ruthn@ncifcrf.gov. Received 8 December 2000; Accepted 1 May 2001 PROTEINS: Structure, Function, and Genetics 44:418 – 427 (2001) © 2001 WILEY-LISS, INC.