Water in protein structure prediction

Abstract

Proteins have evolved to use water to help guide folding. A physically motivated, nonpairwise-additive model of water-mediated interactions added to a protein structure prediction Hamiltonian yields marked improvement in the quality of structure prediction for larger proteins. Free energy profile analysis suggests that long-range water-mediated potentials guide folding and smooth the underlying folding funnel. Analyzing simulation trajectories gives direct evidence that water-mediated interactions facilitate native-like packing of supersecondary structural elements. Long-range pairing of hydrophilic groups is an integral part of protein architecture. Specific water-mediated interactions are a universal feature of biomolecular recognition landscapes in both folding and binding. W ater is intimately involved in protein folding (1-4). That proteins denature both on heating and cooling strongly implicates the involvement of water degrees of freedom. Kauzmann (5) correctly inferred from thermodynamics the hydrophobic layering characteristic of protein structure before protein structures were determined crystallographically. The kinetics of water exclusion is often considered in discussing mechanisms of protein folding, but again it is the avoidance of water in the final folded structure that is emphasized (1). Hydrophobicity patterns have long been a dominant consideration in predicting protein structure by using sequence data (6) and are basic in synthetic protein design (7). Nevertheless, the structured character of water has not been a paramount factor in most existing algorithms for structure prediction (8). These usually rely on effective pair potentials (9) or buried surface area terms to account for the free energy of burying hydrophobic residues (10).In this article, we hypothesize that specific water-mediated interactions help guide the folding process even before native contacts form. Using this idea we develop a bioinformatic, nonpairwise-additive interaction model accounting for water and show that it greatly improves the efficiency and accuracy of structure prediction for ␣-helical proteins. Analysis of folding trajectories with this potential strongly implicates the guiding role of long-range water-mediated interactions. Interestingly, we find here that long-range hydrophilic interactions, as distinct from hydrophobic interactions, also take center stage.The bioinformatic route to water-mediated potentials is difficult in several ways (for more directly physical approaches see ref. 11). Although bound water is visible in structures, localizing waters is more difficult than localizing main-chain atoms. Monomeric protein structures also have relatively few visible watermediated interactions. Our path to a water-mediated potential started with an energy landscape analysis of protein-protein interactions and a bioinformatic survey of interfaces in dimer structures (12, 13). We found that the often-used contact potentials (9) worked well to describe hydrophobic binding interfaces; however, hydrophilic interf...

Sparsely populated folding intermediates of the Fyn SH3 domain: Matching native-centric essential dynamics and experiment

“…Our finding of sparsely populated intermediates, indicating early formation of the central ␤-sheet during the folding process, is in general agreement with previous simulation results ( [45][46][47]. The present approach is limited in that it uses only a reduced chain representation, solvation (13,14,45,48,49) is not explicitly treated, and aspects of kinetic cooperativity (15) are yet to be addressed. Nonetheless, the simplicity of our model allows for broad conformational sampling and energy covariance and thermodynamic analyses over many cycles of reversible folding/ unfolding transitions, which are currently difficult to achieve in higher-resolution models.…”

Protein structure prediction using basin-hopping

“…25,26 This model is sometimes termed the associative memory contact ͑AMC͒ model to distinguish it from the associative memory water ͑AMW͒ model, which uses nonadditive water mediated interactions. 14,27 Since this model has been described in detail before, 15,28 we will only summarize its form here. We employ a version of the coarse-grained model where the 20 letter amino acid code has been reduced to four, and the number of atoms per residue is limited to three ͑C ␣ , C ␤ , and O͒, except for glycine.…”

Hydrophobicity of protein surfaces: Separating geometry from chemistry

“…Although several studies of the role of water in protein folding have been undertaken (7) and this role has been shown to be important (22), the prevalence of cavitation in the process remains an open question (8). Numerous studies have shown, however, that the details of the water-surface interaction are critical in determining the observation of cavitation (see, e.g., ref.…”