Anti-cooperativity and cooperativity in hydrophobic interactions: Three-body free energy landscapes and comparison with implicit-solvent potential functions for proteins

Data Mining Meets Machine Learning: A Novel ANN-based Multi-Body Interaction Docking Scoring Function (MBI-Score) based on Utilizing Frequent Geometric and Chemical Patterns of Interfacial Atoms in Native Protein-Ligand Complexes

Accurate prediction of binding poses is crucial to structure-based drug design. We employ two powerful artificial intelligence (AI) approaches, data-mining and machine-learning, to design artificial neural network (ANN) based pose-scoring function. It is a simple machine-learning-based statistical function that employs frequent geometric and chemical patterns of interacting atoms at protein-ligand interfaces. The patterns are derived by mining interfaces of "native" protein-ligand complexes. Each interface is represented by a graph where nodes are atoms and edges connect protein-ligand interfacial atoms located within certain cutoff distance of each other. Applying frequent subgraph mining to these interfaces provides "native" frequent patterns of interacting atoms. Subsequently, given a pose for a protein-ligand complex of interest, the pose-scoring function (the information-processing unit or neuron) calculates the degree of matching between the interaction patterns present at the pose's interface and the native frequent patterns. The pose-scoring function takes into account the frequency of occurrence of the matching native patterns, the size of the match, and the degree of geometrical similarity between pose-specific and matching native frequent patterns. This novel "multi-body interaction" pose-scoring function (MBI-Score) was validated using two databases, PDBbind and Astex-85, and it outperformed seven commonly used commercial scoring functions. MBI-Score is available at www.khashanlab.org/mbi-score.

An ABSINTH-Based Protocol for Predicting Binding Affinities between Proteins and Small Molecules

The core task in computational drug discovery is to accurately predict binding free energies in receptor-ligand systems for large libraries of putative binders. Here, the ABSINTH implicit solvent model and force field are extended to describe small, organic molecules and their interactions with proteins. We show that an automatic pipeline based on partitioning arbitrary molecules into substructures corresponding to model compounds with known free energies of solvation can be combined with the CHARMM general force field into a method that is successful at the two important challenges a scoring function faces in virtual screening work flows: it ranks known binders with correlation values rivaling that of comparable state-of-the-art methods and it enriches true binders in a set of decoys. Our protocol introduces innovative modifications to common virtual screening workflows, notably the use of explicit ions as competitors and the integration over multiple protein and ligand species differing in their protonation states. We demonstrate the value of modifications to both the protocol and ABSINTH itself. We conclude by discussing the limitations of high-throughput implicit methods such as the one proposed here.

Engineering Order and Cooperativity in a Disordered Protein

Structural disorder in proteins arises from a complex interplay between weak hydrophobicity and unfavorable electrostatic interactions. The extent to which the hydrophobic effect contributes to the unique and compact native state of proteins is, however, confounded by large compensation between multiple entropic and energetic terms. Here we show that protein structural order and cooperativity arise as emergent properties upon hydrophobic substitutions in a disordered system with non-intuitive effects on folding and function. Aided by sequence-structure analysis, equilibrium, and kinetic spectroscopic studies, we engineer two hydrophobic mutations in the disordered DNA-binding domain of CytR that act synergistically, but not in isolation, to promote structure, compactness, and stability. The double mutant, with properties of a fully ordered domain, exhibits weak cooperativity with a complex and rugged conformational landscape. The mutant, however, binds cognate DNA with an affinity only marginally higher than that of the wild type, though nontrivial differences are observed in the binding to noncognate DNA. Our work provides direct experimental evidence of the dominant role of non-additive hydrophobic effects in shaping the molecular evolution of order in disordered proteins and vice versa, which could be generalized to even folded proteins with implications for protein design and functional manipulation.

Free Energy Calculations Based on Coupling Proximal Distribution Functions and Thermodynamic Cycles

Techniques to calculate the free energy changes of a system are very useful in the study of biophysical and biochemical properties. In practice, free energy changes can be described with thermodynamic cycles, and the free energy change of an individual process can be computed by sufficiently sampling the corresponding configurations. However, this is still time-consuming especially for large biomolecular systems. Previously, we have shown that by utilizing precomputed solute-solvent correlations, so-called proximal distribution functions (pDF), we are capable of reconstructing the solvent environment near solute atoms, thus estimating the solute-solvent interactions and solvation free energies of molecules. In this contribution, we apply the technique of pDF-reconstructions to calculate chemical potentials and use this information in thermodynamic cycles. This illustrates how free energy changes of nontrivial chemical processes in aqueous solution systems can be rapidly estimated.

Experimental and theoretical investigation of intramolecular cooperativity in cyclic benzene trimer motif

A series of new symmetrical tripodal molecules 1a–4b with a central benzene scaffold substituted with methyl/ethyl groups and three benzimidazolyl units having a bithiophene/biphenyl/5-alkylthiophene motif at the 2-position via a –CH₂– unit were synthesized and characterized by elemental analysis, HR-MS, and NMR spectroscopy. NMR spectral data reveal that all molecules adopt a cyclic benzene trimer (CBT) using three benzimidazolyl units. Intramolecular cooperative edge-to-face C–H⋯π interactions stabilize the CBT motif in solution and are strong in ethyl substituted molecules (1b–4b) compared to methyl substituted (1a–4a) ones. However, the strength of the CBT unit in the tripodal molecule is independent of the length of the substituent at the 2-position of the benzimidazolyl unit. The relative ¹H NMR chemical shift calculated at the MPW1PW91/6-311+G(d,p) level of theory corroborates the experimental values, and the calculations predict the distribution of the structures into syn isomers. The relative change in the NMR chemical shift is justified by the relative change in the magnitude of the (3,+3) critical point (CP) in the molecular electrostatic potential (MESP) topography. Also, a linear correlation of the intramolecular C–H⋯π interactions evaluated at M062X/6-311+G(d,p) with the relative NMR chemical shift suggest the latter as a measure of intramolecular cooperativity.

A family of biaryl/alkylthiophene (R–R) benzimidazolyl-based tripodal molecules with cyclic benzene trimer (CBT) motif was synthesized and studied by NMR spectroscopy and MPW1PW91/6-311+G(d,p) theory.

Transferable Coarse-Grained Models of Liquid-Liquid Equilibrium Using Local Density Potentials Optimized with the Relative Entropy

Bottom-up coarse-grained (CG) models are now regularly pursued to enable large length and time scale molecular simulations of complex, often macromolecular systems. However, predicting fluid phase equilibria using such models remains fundamentally challenging. A major problem stems from the typically low transferability of CG models beyond the densities and/or compositions at which they are parametrized, which is necessary if they are to describe distinct structural and thermodynamic properties unique to each phase. CG model transferability is compounded by the representation of the inherently multibody coarse interactions using pair potentials that neglect higher order effects. Here, we propose to construct transferable single site CG models of liquid mixtures by supplementing traditional CG pair interactions with local density potentials, which constitute a computationally inexpensive mean-field approach to describe many-body effects, in that site energies are modulated by the local solution environment. To illustrate the approach, we use intra- and interspecies local density potentials to develop CG models of benzene-water solutions that show impressive transferability in structural metrics (pair correlation functions, density profiles) throughout composition space, in contrast to pair-only CG representations. While further refinement may be necessary to represent more complex thermodynamic properties, like the liquid-liquid interfacial tension, the generality and improvement offered by the local density approach are highly encouraging for enabling complex phase equilibrium modeling using CG models.

Discovery of Inhibitors of Four Bromodomains by Fragment-Anchored Ligand Docking

The high-throughput docking protocol called ALTA-VS (anchor-based library tailoring approach for virtual screening) was developed in 2005 for the efficient in silico screening of large libraries of compounds by preselection of only those molecules that have optimal fragments (anchors) for the protein target. Here we present an updated version of ALTA-VS with a broader range of potential applications. The evaluation of binding energy makes use of a classical force field with implicit solvent in the continuum dielectric approximation. In about 2 days per protein target on a 96-core compute cluster (equipped with Xeon E3-1280 quad core processors at 2.5 GHz), the screening of a library of nearly 77 000 diverse molecules with the updated ALTA-VS protocol has resulted in the identification of 19, 3, 3, and 2 μM inhibitors of the human bromodomains ATAD2, BAZ2B, BRD4(1), and CREBBP, respectively. The success ratio (i.e., number of actives in a competition binding assay in vitro divided by the number of compounds tested) ranges from 8% to 13% in dose-response measurements. The poses predicted by fragment-based docking for the three ligands of the BAZ2B bromodomain were confirmed by protein X-ray crystallography.

A self-consistent structural perturbation approach for determining the magnitude and extent of allosteric coupling in proteins

Elucidating the extent of energetic coupling between residues in single-domain proteins, which is a fundamental determinant of allostery, information transfer and folding cooperativity, has remained a grand challenge. While several sequence- and structure-based approaches have been proposed, a self-consistent description that is simultaneously compatible with unfolding thermodynamics is lacking. We recently developed a simple structural perturbation protocol that captures the changes in thermodynamic stabilities induced by point mutations within the protein interior. Here, we show that a fundamental residue-specific component of this perturbation approach, the coupling distance, is uniquely sensitive to the environment of a residue in the protein to a distance of ∼15 Å. With just the protein contact map as an input, we reproduce the extent of percolation of perturbations within the structure as observed in network analysis of intra-protein interactions, molecular dynamics simulations and NMR-observed changes in chemical shifts. Using this rapid protocol that relies on a single structure, we explain the results of statistical coupling analysis (SCA) that requires hundreds of sequences to identify functionally critical sectors, the propagation and dissipation of perturbations within proteins and the higher-order couplings deduced from detailed NMR experiments. Our results thus shed light on the possible mechanistic origins of signaling through the interaction network within proteins, the likely distance dependence of perturbations induced by ligands and post-translational modifications and the origins of folding cooperativity through many-body interactions.

Real-time monitoring of hydrophobic aggregation reveals a critical role of cooperativity in hydrophobic effect

The hydrophobic interaction drives nonpolar solutes to aggregate in aqueous solution, and hence plays a critical role in many fundamental processes in nature. An important property intrinsic to hydrophobic interaction is its cooperative nature, which is originated from the collective motions of water hydrogen bond networks surrounding hydrophobic solutes. This property is widely believed to enhance the formation of hydrophobic core in proteins. However, cooperativity in hydrophobic interactions has not been successfully characterized by experiments. Here, we quantify cooperativity in hydrophobic interactions by real-time monitoring the aggregation of hydrophobic solute (hexaphenylsilole, HPS) in a microfluidic mixer. We show that association of a HPS molecule to its aggregate in water occurs at sub-microsecond, and the free energy change is −5.8 to −13.6 kcal mol⁻¹. Most strikingly, we discover that cooperativity constitutes up to 40% of this free energy. Our results provide quantitative evidence for the critical role of cooperativity in hydrophobic interactions.

Hydrophobic interactions occur between nonpolar molecules in water and their experimental quantification can help the understanding of biological self-assembly. Here Jiang et al. examine the kinetics and thermodynamics of hydrophobic aggregation in a bulk environment and characterize its cooperativity.

Perplexing cooperative folding and stability of a low-sequence complexity, polyproline 2 protein lacking a hydrophobic core

The burial of hydrophobic side chains in a protein core generally is thought to be the major ingredient for stable, cooperative folding. Here, we show that, for the snow flea antifreeze protein (sfAFP), stability and cooperativity can occur without a hydrophobic core, and without α-helices or β-sheets. sfAFP has low sequence complexity with 46% glycine and an interior filled only with backbone H-bonds between six polyproline 2 (PP2) helices. However, the protein folds in a kinetically two-state manner and is moderately stable at room temperature. We believe that a major part of the stability arises from the unusual match between residue-level PP2 dihedral angle bias in the unfolded state and PP2 helical structure in the native state. Additional stabilizing factors that compensate for the dearth of hydrophobic burial include shorter and stronger H-bonds, and increased entropy in the folded state. These results extend our understanding of the origins of cooperativity and stability in protein folding, including the balance between solvent and polypeptide chain entropies.

Coarse-grained models using local-density potentials optimized with the relative entropy: Application to implicit solvation

Bottom-up multiscale techniques are frequently used to develop coarse-grained (CG) models for simulations at extended length and time scales but are often limited by a compromise between computational efficiency and accuracy. The conventional approach to CG nonbonded interactions uses pair potentials which, while computationally efficient, can neglect the inherently multibody contributions of the local environment of a site to its energy, due to degrees of freedom that were coarse-grained out. This effect often causes the CG potential to depend strongly on the overall system density, composition, or other properties, which limits its transferability to states other than the one at which it was parameterized. Here, we propose to incorporate multibody effects into CG potentials through additional nonbonded terms, beyond pair interactions, that depend in a mean-field manner on local densities of different atomic species. This approach is analogous to embedded atom and bond-order models that seek to capture multibody electronic effects in metallic systems. We show that the relative entropy coarse-graining framework offers a systematic route to parameterizing such local density potentials. We then characterize this approach in the development of implicit solvation strategies for interactions between model hydrophobes in an aqueous environment.

Nonpolar Solvation Free Energy from Proximal Distribution Functions

Using precomputed near neighbor or proximal distribution functions (pDFs) that approximate solvent density about atoms in a chemically bonded context one can estimate the solvation structures around complex solutes and the corresponding solute-solvent energetics. In this contribution, we extend this technique to calculate the solvation free energies (ΔG) of a variety of solutes. In particular we use pDFs computed for small peptide molecules to estimate ΔG for larger peptide systems. We separately compute the non polar (ΔG_vdW) and electrostatic (ΔG_elec) components of the underlying potential model. Here we show how the former can be estimated by thermodynamic integration using pDF-reconstructed solute-solvent interaction energy. The electrostatic component can be approximated with Linear Response theory as half of the electrostatic solute-solvent interaction energy. We test the method by calculating the solvation free energies of butane, propanol, polyalanine, and polyglycine and by comparing with traditional free energy simulations. Results indicate that the pDF-reconstruction algorithm approximately reproduces ΔG_vdW calculated by benchmark free energy simulations to within ∼ kcal/mol accuracy. The use of transferable pDFs for each solute atom allows for a rapid estimation of ΔG for arbitrary molecular systems.

Study on the Application of the Combination of TMD Simulation and Umbrella Sampling in PMF Calculation for Molecular Conformational Transitions

Free energy calculations of the potential of mean force (PMF) based on the combination of targeted molecular dynamics (TMD) simulations and umbrella samplings as a function of physical coordinates have been applied to explore the detailed pathways and the corresponding free energy profiles for the conformational transition processes of the butane molecule and the 35-residue villin headpiece subdomain (HP35). The accurate PMF profiles for describing the dihedral rotation of butane under both coordinates of dihedral rotation and root mean square deviation (RMSD) variation were obtained based on the different umbrella samplings from the same TMD simulations. The initial structures for the umbrella samplings can be conveniently selected from the TMD trajectories. For the application of this computational method in the unfolding process of the HP35 protein, the PMF calculation along with the coordinate of the radius of gyration (R_g) presents the gradual increase of free energies by about 1 kcal/mol with the energy fluctuations. The feature of conformational transition for the unfolding process of the HP35 protein shows that the spherical structure extends and the middle α-helix unfolds firstly, followed by the unfolding of other α-helices. The computational method for the PMF calculations based on the combination of TMD simulations and umbrella samplings provided a valuable strategy in investigating detailed conformational transition pathways for other allosteric processes.

Salting-Out of Methane in the Aqueous Solutions of Urea and Glycine-Betaine

We have studied the hydrophobic association and solvation of methane molecules in aqueous solutions of urea and glycine betaine (GB). We have calculated the potentials of mean force (PMFs) between methane molecules in water, aqueous GB, aqueous urea and aqueous urea-GB mixtures. The PMFs and equilibrium constants indicate that both urea and GB increase the hydrophobic association of methane. Calculation of thermodynamic parameters shows that the association of methane is stabilized by entropy whereas solvation is favored by enthalpy. In the case of the water-urea-GB mixture, both hydrophobic association and solvation are stabilized by entropy. From the investigation of radial distribution functions, running coordination numbers and excess coordination numbers, we infer that both urea and GB are preferentially excluded from methane surface in the mixtures of osmolytes and methane is preferentially solvated by water molecules in all the mixtures. The favorable exclusion of both urea and GB from the methane surface suggests that both urea and GB increase the interaction between methane molecules, i.e., salting-out of methane. We observe that addition of both urea and GB to water enhances local water structure. The calculated values of diffusion constants of water also suggest enhanced water-water interactions in the presence of urea and GB. The calculated free energies of methane in these mixtures show that methane is less soluble in the mixtures of urea and GB than in water. The data on solvation free energies support the observations obtained from the PMFs of methane molecules.

The application of the integral equation theory to study the hydrophobic interaction

The Wertheim's integral equation theory was tested against newly obtained Monte Carlo computer simulations to describe the potential of mean force between two hydrophobic particles. An excellent agreement was obtained between the theoretical and simulation results. Further, the Wertheim's integral equation theory with polymer Percus-Yevick closure qualitatively correctly (with respect to the experimental data) describes the solvation structure under conditions where the simulation results are difficult to obtain with good enough accuracy.

Effects of desolvation barriers and sidechains on local–nonlocal coupling and chevron behaviors in coarse-grained models of protein folding

Coarse-grained protein chain models with desolvation barriers or sidechains lead to stronger local–nonlocal coupling and more linear chevron plots.

Water 16-mers and hexamers: assessment of the three-body and electrostatically embedded many-body approximations of the correlation energy or the nonlocal energy as ways to include cooperative effects

This work presents a new fragment method, the electrostatically embedded many-body expansion of the nonlocal energy (EE-MB-NE), and shows that it, along with the previously proposed electrostatically embedded many-body expansion of the correlation energy (EE-MB-CE), produces accurate results for large systems at the level of CCSD(T) coupled cluster theory. We primarily study water 16-mers, but we also test the EE-MB-CE method on water hexamers. We analyze the distributions of two-body and three-body terms to show why the many-body expansion of the electrostatically embedded correlation energy converges faster than the many-body expansion of the entire electrostatically embedded interaction potential. The average magnitude of the dimer contributions to the pairwise additive (PA) term of the correlation energy (which neglects cooperative effects) is only one-half of that of the average dimer contribution to the PA term of the expansion of the total energy; this explains why the mean unsigned error (MUE) of the EE-PA-CE approximation is only one-half of that of the EE-PA approximation. Similarly, the average magnitude of the trimer contributions to the three-body (3B) term of the EE-3B-CE approximation is only one-fourth of that of the EE-3B approximation, and the MUE of the EE-3B-CE approximation is one-fourth that of the EE-3B approximation. Finally, we test the efficacy of two- and three-body density functional corrections. One such density functional correction method, the new EE-PA-NE method, with the OLYP or the OHLYP density functional (where the OHLYP functional is the OptX exchange functional combined with the LYP correlation functional multiplied by 0.5), has the best performance-to-price ratio of any method whose computational cost scales as the third power of the number of monomers and is competitive in accuracy in the tests presented here with even the electrostatically embedded three-body approximation.

Implicit Solvation Parameters Derived from Explicit Water Forces in Large-Scale Molecular Dynamics Simulations

Implicit solvation is a mean force approach to model solvent forces acting on a solute molecule. It is frequently used in molecular simulations to reduce the computational cost of solvent treatment. In the first instance, the free energy of solvation and the associated solvent–solute forces can be approximated by a function of the solvent-accessible surface area (SASA) of the solute and differentiated by an atom–specific solvation parameter σ_i^SASA. A procedure for the determination of values for the σ_i^SASA parameters through matching of explicit and implicit solvation forces is proposed. Using the results of Molecular Dynamics simulations of 188 topologically diverse protein structures in water and in implicit solvent, values for the σ_i^SASA parameters for atom types i of the standard amino acids in the GROMOS force field have been determined. A simplified representation based on groups of atom types σ_g^SASA was obtained via partitioning of the atom–type σ_i^SASA distributions by dynamic programming. Three groups of atom types with well separated parameter ranges were obtained, and their performance in implicit versus explicit simulations was assessed. The solvent forces are available at http://mathbio.nimr.mrc.ac.uk/wiki/Solvent_Forces.

Scoring protein interaction decoys using exposed residues (SPIDER): a novel multibody interaction scoring function based on frequent geometric patterns of interfacial residues

Accurate prediction of the structure of protein-protein complexes in computational docking experiments remains a formidable challenge. It has been recognized that identifying native or native-like poses among multiple decoys is the major bottleneck of the current scoring functions used in docking. We have developed a novel multibody pose-scoring function that has no theoretical limit on the number of residues contributing to the individual interaction terms. We use a coarse-grain representation of a protein-protein complex where each residue is represented by its side chain centroid. We apply a computational geometry approach called Almost-Delaunay tessellation that transforms protein-protein complexes into a residue contact network, or an undirectional graph where vertex-residues are nodes connected by edges. This treatment forms a family of interfacial graphs representing a dataset of protein-protein complexes. We then employ frequent subgraph mining approach to identify common interfacial residue patterns that appear in at least a subset of native protein-protein interfaces. The geometrical parameters and frequency of occurrence of each "native" pattern in the training set are used to develop the new SPIDER scoring function. SPIDER was validated using standard "ZDOCK" benchmark dataset that was not used in the development of SPIDER. We demonstrate that SPIDER scoring function ranks native and native-like poses above geometrical decoys and that it exceeds in performance a popular ZRANK scoring function. SPIDER was ranked among the top scoring functions in a recent round of CAPRI (Critical Assessment of PRedicted Interactions) blind test of protein-protein docking methods.

Hydrophobic interactions in the formation of secondary structures in small peptides

Effects of the attractive and repulsive parts of hydrophobic interactions on α helices and β sheets in small peptides are investigated using a simple atomic potential. Typically, a physical spatial range of attraction tends to favor β sheets, but α helices would be favored if the attractive range were more extended. We also found that desolvation barriers favor β sheets in collapsed conformations of polyalanine, polyvaline, polyleucine, and three fragments of amyloid peptides tested in this study. Our results provide insight into the multifaceted role of hydrophobicity in secondary structure formation, including the α to β transitions in certain amyloid peptides.

Solvation Effect on the Conformations of Alanine Dipeptide: Integral Equation Approach

We present an implicit solvent model based on the extended reference interaction site model (XRISM) integral equation theory, which is a molecular theory of solvation. The solvation free energy is composed of additive potentials of mean force (PMF) of various functional groups. The XRISM theory is applied to determine the PMF of each group in water and NaBr electrolyte solutions. The method has been coupled to Brownian dynamics (BD) and is illustrated here on alanine dipeptide. The results of the method are compared with those obtained by explicit water simulations and other popular implicit solvent models for detailed discussion. The comparison of our model with other methods indicates that the intramolecular correlation and the solvation structure influence the stability of the P(II) and alpha(R) conformers. The results of NaBr electrolyte solutions show that the concentration of electrolyte also has a substantial effect on the favored conformations.

Hydrophobic interactions in model enclosures from small to large length scales: non-additivity in explicit and implicit solvent models

The binding affinities between a united-atom methane and various model hydrophobic enclosures were studied through high accuracy free energy perturbation methods (FEP). We investigated the non-additivity of the hydrophobic interaction in these systems, measured by the deviation of its binding affinity from that predicted by the pairwise additivity approximation. While only small non-additivity effects were previously reported in the interactions in methane trimers, we found large cooperative effects (as large as -1.14 kcal mol(-1) or approximately a 25% increase in the binding affinity) and anti-cooperative effects (as large as 0.45 kcal mol(-1)) for these model enclosed systems. Decomposition of the total potential of mean force (PMF) into increasing orders of multi-body interactions indicates that the contributions of the higher order multi-body interactions can be either positive or negative in different systems, and increasing the order of multi-body interactions considered did not necessarily improve the accuracy. A general correlation between the sign of the non-additivity effect and the curvature of the solute molecular surface was observed. We found that implicit solvent models based on the molecular surface area (MSA) performed much better, not only in predicting binding affinities, but also in predicting the non-additivity effects, compared with models based on the solvent accessible surface area (SASA), suggesting that MSA is a better descriptor of the curvature of the solutes. We also show how the non-additivity contribution changes as the hydrophobicity of the plate is decreased from the dewetting regime to the wetting regime.

A coarse-grained protein force field for folding and structure prediction

We have revisited the protein coarse-grained optimized potential for efficient structure prediction (OPEP). The training and validation sets consist of 13 and 16 protein targets. Because optimization depends on details of how the ensemble of decoys is sampled, trial conformations are generated by molecular dynamics, threading, greedy, and Monte Carlo simulations, or taken from publicly available databases. The OPEP parameters are varied by a genetic algorithm using a scoring function which requires that the native structure has the lowest energy, and the native-like structures have energy higher than the native structure but lower than the remote conformations. Overall, we find that OPEP correctly identifies 24 native or native-like states for 29 targets and has very similar capability to the all-atom discrete optimized protein energy model (DOPE), found recently to outperform five currently used energy models.

Expanding the realm of ultrafast protein folding: gpW, a midsize natural single-domain with alpha+beta topology that folds downhill

All ultrafast folding proteins known to date are either very small in size (less than 45 residues), have an alpha-helix bundle topology, or have been artificially engineered. In fact, many of them share two or even all three features. Here we show that gpW, a natural 62-residue alpha+beta protein expected to fold slowly in a two-state fashion, folds in microseconds (i.e., from tau = 33 micros at 310 K to tau = 1.7 micros at 355 K). Thermodynamic analyses of gpW reveal probe dependent thermal denaturation, complex coupling between two denaturing agents, and differential scanning calorimetry (DSC) thermogram characteristic of folding over a negligible thermodynamic folding barrier. The free energy surface analysis of gpW folding kinetics also produces a marginal folding barrier of about thermal energy ( RT) at the denaturation midpoint. From these results we conclude that gpW folds in the downhill regime and is close to the global downhill limit. This protein seems to be poised toward downhill folding by a loosely packed hydrophobic core with low aromatic content, large stabilizing contributions from local interactions, and abundance of positive charges on the native surface. These special features, together with a complex functional role in bacteriophage lambda assembly, suggest that gpW has been engineered to fold downhill by natural selection.

Quantifying intrinsic specificity: a potential complement to affinity in drug screening

We report here the investigation of a novel description of specificity in protein-ligand binding based on energy landscape theory. We define a new term, intrinsic specificity ratio (ISR), which describes the level of discrimination in binding free energies of the native basin for a protein-ligand complex from the weaker binding states of the same ligand. We discuss the relationship between the intrinsic specificity we defined here and the conventional definition of specificity. In a docking study of molecules with the enzyme COX-2, we demonstrate a statistical correspondence between ISR value and geometrical shapes of the small molecules binding to COX-2. We further observe that the known selective (nonselective) inhibitors of COX-2 have higher (lower) ISR values. We suggest that intrinsic specificity ratio may be a useful new criterion and a complement to affinity in drug screening and in searching for potential drug lead compounds.

Water properties and potential of mean force for hydrophobic interactions of methane and nanoscopic pockets studied by computer simulations

We consider model systems consisting of a methane molecule and hemispherical pockets of subnanometer radii whose walls are made of hydrophobic material. The potential of mean force for process of translocation of the methane molecule from bulk water into the pockets' interior is obtained, based on an explicit solvent molecular dynamics simulations. Accompanying changes in water density around the interacting objects and spatial distribution of solvent's potential energy are analyzed, allowing for interpretation of details of hydrophobic interactions in relation to hydrophobic hydration properties. Applicability of surface area-based models of hydrophobic effect for systems of interest is also investigated. A total work for the translocation process is not dependent on pocket's size, indicating that pocket desolvation has little contribution to free energy changes, which is consistent with the observation that solvent density is significantly reduced inside "unperturbed" pockets. Substantial solvent effects are shown to have a longer range than in case of a well investigated methane pair. A desolvation barrier is present in a smaller pocket system but disappears in the larger one, suggesting that a form of a "hydrophobic collapse" is observed.

Pressure and temperature dependence of hydrophobic hydration: Volumetric, compressibility, and thermodynamic signatures

The combined effect of pressure and temperature on hydrophobic hydration of a nonpolar methanelike solute is investigated by extensive simulations in the TIP4P model of water. Using test-particle insertion techniques, free energies of hydration under a range of pressures from 1 to 3000 atm are computed at eight temperatures ranging from 278.15 to 368.15 K. Corresponding enthalpy, entropy, and heat capacity accompanying the hydration process are estimated from the temperature dependence of the free energies. Partial molar and excess volumes calculated using pressure derivatives of the simulated free energies are consistent with those determined by direct volume simulations; but direct volume determination offers more reliable estimates for compressibility. At 298.15 K, partial molar and excess isothermal compressibilities of methane are negative at 1 atm. Partial molar and excess adiabatic (isentropic) compressibilities are estimated to be also negative under the same conditions. But partial molar and excess isothermal compressibilities are positive at high pressures, with a crossover from negative to positive compressibility at approximately 100-1000 atm. This trend is consistent with experiments on aliphatic amino acids and pressure-unfolded states of proteins. For the range of pressures simulated, hydration heat capacity exhibits little pressure dependence, also in apparent agreement with experiment. When pressure is raised at constant room temperature, hydration free energy increases while its entropic component remains essentially constant. Thus, the increasing unfavorability of hydration under raised pressure is seen as largely an enthalpic effect. Ramifications of the findings of the authors for biopolymer conformational transitions are discussed.

Hydrophobic interaction and hydrogen-bond network for a methane pair in liquid water

We employ fully quantum-mechanical molecular dynamics simulations to evaluate the force between two methanes dissolved in water, as a model for hydrophobic association. A stable configuration is found near the methane-methane contact separation, while a shallow second potential minimum occurs for the solvent-separated configuration. The strength and shape of the potential of mean force are in conflict with earlier classical force-field simulations but agree well with a simple hydrophobic burial model which is based on solubility experiments. Examination of solvent dynamics reveals stable water cages at several specific methane-methane separations.

Effects of lengthscales and attractions on the collapse of hydrophobic polymers in water

We present results from extensive molecular dynamics simulations of collapse transitions of hydrophobic polymers in explicit water focused on understanding effects of lengthscale of the hydrophobic surface and of attractive interactions on folding. Hydrophobic polymers display parabolic, protein-like, temperature-dependent free energy of unfolding. Folded states of small attractive polymers are marginally stable at 300 K and can be unfolded by heating or cooling. Increasing the lengthscale or decreasing the polymer-water attractions stabilizes folded states significantly, the former dominated by the hydration contribution. That hydration contribution can be described by the surface tension model, DeltaG = gamma(T)DeltaA, where the surface tension, gamma, is lengthscale-dependent and decreases monotonically with temperature. The resulting variation of the hydration entropy with polymer lengthscale is consistent with theoretical predictions of Huang and Chandler [Huang DM, Chandler D (2000) Proc Natl Acad Sci USA 97:] that explain the blurring of entropy convergence observed in protein folding thermodynamics. Analysis of water structure shows that the polymer-water hydrophobic interface is soft and weakly dewetted, and is characterized by enhanced interfacial density fluctuations. Formation of this interface, which induces polymer folding, is strongly opposed by enthalpy and favored by entropy, similar to the vapor-liquid interface.

Comparison of solvation-effect methods for the simulation of peptide interactions with a hydrophobic surface

In this study we investigated the interaction behavior between thirteen different small peptides and a hydrophobic surface using three progressively more complex methods of representing solvation effects: a united-atom implicit solvation method [CHARMM 19 force field (C19) with Analytical Continuum Electrostatics (ACE)], an all-atom implicit solvation method (C22 with GBMV), and an all-atom explicit solvation method (C22 with TIP3P). The adsorption behavior of each peptide was characterized by the calculation of the potential of mean force as a function of peptide-surface separation distance. The results from the C22/TIP3P model suggest that hydrophobic peptides exhibit relatively strong adsorption behavior, polar and positively-charged peptides exhibit negligible to relatively weak favorable interactions with the surface, and negatively-charged peptides strongly resist adsorption. Compared to the TIP3P model, the ACE and GBMV implicit solvent models predict much stronger attractions for the hydrophobic peptides as well as stronger repulsions for the negatively-charged peptides on the CH(3)-SAM surface. These comparisons provide a basis from which each of these implicit solvation methods may be reparameterized to provide closer agreement with explicitly represented solvation in simulations of peptide and protein adsorption to functionalized surfaces.

Cooperativity and the origins of rapid, single-exponential kinetics in protein folding

The folding of naturally occurring, single-domain proteins is usually well described as a simple, single-exponential process lacking significant trapped states. Here we further explore the hypothesis that the smooth energy landscape this implies, and the rapid kinetics it engenders, arises due to the extraordinary thermodynamic cooperativity of protein folding. Studying Miyazawa-Jernigan lattice polymers, we find that, even under conditions where the folding energy landscape is relatively optimized (designed sequences folding at their temperature of maximum folding rate), the folding of protein-like heteropolymers is accelerated when their thermodynamic cooperativity is enhanced by enhancing the nonadditivity of their energy potentials. At lower temperatures, where kinetic traps presumably play a more significant role in defining folding rates, we observe still greater cooperativity-induced acceleration. Consistent with these observations, we find that the folding kinetics of our computational models more closely approximates single-exponential behavior as their cooperativity approaches optimal levels. These observations suggest that the rapid folding of naturally occurring proteins is, in part, a consequence of their remarkably cooperative folding.

Atom-by-atom analysis of global downhill protein folding

Protein folding is an inherently complex process involving coordination of the intricate networks of weak interactions that stabilize native three-dimensional structures. In the conventional paradigm, simple protein structures are assumed to fold in an all-or-none process that is inaccessible to experiment. Existing experimental methods therefore probe folding mechanisms indirectly. A widely used approach interprets changes in protein stability and/or folding kinetics, induced by engineered mutations, in terms of the structure of the native protein. In addition to limitations in connecting energetics with structure, mutational methods have significant experimental uncertainties and are unable to map complex networks of interactions. In contrast, analytical theory predicts small barriers to folding and the possibility of downhill folding. These theoretical predictions have been confirmed experimentally in recent years, including the observation of global downhill folding. However, a key remaining question is whether downhill folding can indeed lead to the high-resolution analysis of protein folding processes. Here we show, with the use of nuclear magnetic resonance (NMR), that the downhill protein BBL from Escherichia coli unfolds atom by atom starting from a defined three-dimensional structure. Thermal unfolding data on 158 backbone and side-chain protons out of a total of 204 provide a detailed view of the structural events during folding. This view confirms the statistical nature of folding, and exposes the interplay between hydrogen bonding, hydrophobic forces, backbone conformation and side-chain entropy. From the data we also obtain a map of the interaction network in this protein, which reveals the source of folding cooperativity. Our approach can be extended to other proteins with marginal barriers (less than 3RT), providing a new tool for the study of protein folding.

Application of a two-length-scale field theory to the solvation of neutral and charged molecules

We develop a continuous self-consistent theory of solute-water interactions that allows determination of the hydrophobic layer around a solute molecule of any geometry, with an explicit account of solvent structure described by its correlation function. We compute the mean solvent density profile n(r) surrounding the solute molecule as well as its solvation free energy deltaG. We compare the two-length-scale field theory to the numerical data of Monte Carlo simulations found in the literature for spherical molecules and discuss the possibility of self-consistent adjustment of the free parameters of the theory. In the framework of this approach, we compute the solvation free energies of alkane molecules and the free energy of interaction of two spheres of radius R separated by the distance D. We describe the general setting of the self-consistent account of electrostatic interactions in the framework of our model where the water is considered not as a continuous medium but as a gas of dipoles. We analyze the limiting cases where the proposed theory coincides with the electrostatics of a continuous medium.

Solvation and desolvation effects in protein folding: native flexibility, kinetic cooperativity and enthalpic barriers under isostability conditions

As different parts of a protein chain approach one another during folding, they are expected to encounter desolvation barriers before optimal packing is achieved. This impediment originates from the water molecule's finite size, which entails a net energetic cost for water exclusion when the formation of compensating close intraprotein contacts is not yet complete. Based on recent advances, we extend our exploration of these microscopic elementary desolvation barriers' roles in the emergence of generic properties of protein folding. Using continuum Gō-like C(alpha) chain models of chymotrypsin inhibitor 2 (CI2) and barnase as examples, we underscore that elementary desolvation barriers between a protein's constituent groups can significantly reduce native conformational fluctuations relative to model predictions that neglected these barriers. An increasing height of elementary desolvation barriers leads to thermodynamically more cooperative folding/unfolding transitions (i.e., higher overall empirical folding barriers) and higher degrees of kinetic cooperativity as manifested by more linear rate-stability relationships under constant temperature. Applying a spatially non-uniform thermodynamic parametrization we recently introduced for the pairwise C(alpha) potentials of mean force, the present barnase model further illustrates that desolvation is a probable physical underpinning for the experimentally observed high intrinsic enthalpic folding barrier under isostability conditions.

Predicting the energetics of osmolyte-induced protein folding/unfolding

A primary thermodynamic goal in protein biochemistry is to attain predictive understanding of the detailed energetic changes that are responsible for folding/unfolding. Through use of recently determined free energies of side-chain and backbone transfer from water to osmolytes and Tanford's transfer model, we demonstrate that the long-sought goal of predicting solvent-dependent cooperative protein folding/unfolding free-energy changes (m values) can be achieved. Moreover, the approach permits dissection of the folding/unfolding free-energy changes into individual contributions from the peptide backbone and residue side chains.

Recent advances in the development and application of implicit solvent models in biomolecule simulations

Advances have recently been made in the development of implicit solvent methodologies and their application to the modeling of biomolecules, particularly with regard to generalized Born approaches, dielectric screening function formulations and models based on solvent-accessible surface areas. Interesting new developments include more refined non-polar solvation energy estimators, and implicit methods for modeling low-dielectric and heterogeneous environments such as membrane systems. These have been successfully applied to molecular dynamics simulations, the scoring of protein conformations, and the calculation of binding affinities and folding free energy landscapes.

Chevron behavior and isostable enthalpic barriers in protein folding: successes and limitations of simple Gō-like modeling

It has been demonstrated that a "near-Levinthal" cooperative mechanism, whereby the common Gō interaction scheme is augmented by an extra favorability for the native state as a whole, can lead to apparent two-state folding/unfolding kinetics over a broad range of native stabilities in lattice models of proteins. Here such a mechanism is shown to be generalizable to a simplified continuum (off-lattice) Langevin dynamics model with a Calpha protein chain representation, with the resulting chevron plots exhibiting an extended quasilinear regime reminiscent of that of apparent two-state real proteins. Similarly high degrees of cooperativity are possible in Gō-like continuum models with rudimentary pairwise desolvation barriers as well. In these models, cooperativity increases with increasing desolvation barrier height, suggesting strongly that two-state-like folding/unfolding kinetics would be achievable when the pairwise desolvation barrier becomes sufficiently high. Besides cooperativity, another generic folding property of interest that has emerged from published experiments on several apparent two-state proteins is that their folding relaxation under constant native stability (isostability) conditions is essentially Arrhenius, entailing high intrinsic enthalpic folding barriers of approximately 17-30 kcal/mol. Based on a new analysis of published data on barnase, here we propose that a similar property should also apply to a certain class of non-two-state proteins that fold with chevron rollovers. However, several continuum Gō-like constructs considered here fail to predict any significant intrinsic enthalpic folding barrier under isostability conditions; thus the physical origin of such barriers in real proteins remains to be elucidated.

Geometric cooperativity and anticooperativity of three-body interactions in native proteins

Characterizing multibody interactions of hydrophobic, polar, and ionizable residues in protein is important for understanding the stability of protein structures. We introduce a geometric model for quantifying 3-body interactions in native proteins. With this model, empirical propensity values for many types of 3-body interactions can be reliably estimated from a database of native protein structures, despite the overwhelming presence of pairwise contacts. In addition, we define a nonadditive coefficient that characterizes cooperativity and anticooperativity of residue interactions in native proteins by measuring the deviation of 3-body interactions from 3 independent pairwise interactions. It compares the 3-body propensity value from what would be expected if only pairwise interactions were considered, and highlights the distinction of propensity and cooperativity of 3-body interaction. Based on the geometric model, and what can be inferred from statistical analysis of such a model, we find that hydrophobic interactions and hydrogen-bonding interactions make nonadditive contributions to protein stability, but the nonadditive nature depends on whether such interactions are located in the protein interior or on the protein surface. When located in the interior, many hydrophobic interactions such as those involving alkyl residues are anticooperative. Salt-bridge and regular hydrogen-bonding interactions, such as those involving ionizable residues and polar residues, are cooperative. When located on the protein surface, these salt-bridge and regular hydrogen-bonding interactions are anticooperative, and hydrophobic interactions involving alkyl residues become cooperative. We show with examples that incorporating 3-body interactions improves discrimination of protein native structures against decoy conformations. In addition, analysis of cooperative 3-body interaction may reveal spatial motifs that can suggest specific protein functions.

Desolvation is a likely origin of robust enthalpic barriers to protein folding

Experimental data from global analyses of temperature (T) and denaturant dependence of the folding rates of small proteins led to an intrinsic enthalpic folding barrier hypothesis: to a good approximation, the T-dependence of folding rate under constant native stability conditions is Arrhenius. Furthermore, for a given protein, the slope of isostability folding rate versus 1/T is essentially independent of native stability. This hypothesis implies a simple relationship between chevron and Eyring plots of folding that is easily discernible when both sets of rates are expressed as functions of native stability. Using experimental data in the literature, we verify the predicted chevron-Eyring relationship for 14 proteins and determine their intrinsic enthalpic folding barriers, which vary approximately from 15 kcal/mol to 40 kcal/mol for different proteins. These enthalpic barriers do not appear to correlate with folding rates, but they exhibit correlation with equilibrium unfolding enthalpy at room temperature. Intrinsic enthalpic barriers with similarly high magnitudes apply as well to at least two cases of peptide-peptide and peptide-protein association, suggesting that these barriers are a hallmark of certain general and fundamental kinetic processes during folding and binding. Using a class of explicit-chain C(alpha) protein models with constant elementary enthalpic desolvation barriers between C(alpha) positions, we show that small microscopic pairwise desolvation barriers, which are a direct consequence of the particulate nature of water, can act cooperatively to give rise to a significant overall enthalpic barrier to folding. This theoretical finding provides a physical rationalization for the high intrinsic enthalpic barriers in protein folding energetics. Ramifications of entropy-enthalpy compensation in hydrophobic association for the height of enthalpic desolvation barrier are discussed.

Molecular Origin of Anticooperativity in Hydrophobic Association

The structuring of water molecules in the vicinity of nonpolar solutes is responsible for hydrophobic hydration and association thermodynamics in aqueous solutions. Here, we studied the potential of mean force (PMF) for the formation of a dimer and trimers of methane molecules in three specific configurations in explicit water to explain multibody effects in hydrophobic association on a molecular level. We analyzed the packing and orientation of water molecules in the vicinity of the solute to explain the effect of ordering of the water around nonpolar solutes on many-body interactions. Consistent with previous theoretical studies, we observed cooperativity, manifested as a reduction of the height of the desolvation barrier for the trimer in an isosceles triangle geometry, but for linear trimers, we observed only anticooperativity. A simple mechanistic picture of hydrophobic association is drawn. The free energy of hydrophobic association depends primarily on the difference in the number of water molecules in the first solvation shell of a cluster and that in the monomers of a cluster; this can be approximated by the molecular surface area. However, there are unfavorable electrostatic interactions between the water molecules from different parts of the solvation shell of a trimer because of their increased orientation induced by the nonpolar solute. These electrostatic interactions make an anticooperative contribution to the PMF, which is clearly manifested for the linear trimer where the multibody contribution due to changes in the molecular surface area is equal to zero. The information theory model of hydrophobic interactions of Hummer et al. also explains the anticooperativity of hydrophobic association of the linear trimers; however, it predicts anticooperativity with a qualitatively identical distance dependence for nonlinear trimers, which disagrees with the results of simulations.