Free-Energy Landscape and Rate Estimation of the Aromatic Ring Flips in Basic Pancreatic Trypsin Inhibitors Using Metadynamics.

Aromatic side chains (phenylalanine and tyrosine) of a protein flip by 180° around the Cß-Cγ axis (χ2 dihedral of the side chain), producing two symmetry-equivalent states. The study of ring flip dynamics with nuclear magnetic resonance (NMR) experiments helps to understand local conformational fluctuations. Ring flips are categorized as slow (milliseconds and onward) or fast (nanoseconds to near milliseconds) based on timescales accessible to NMR experiments. In this study, we investigated the ability of the infrequent metadynamics approach to estimate the flip rate and discriminate between slow and fast ring flips for eight individual aromatic side chains (F4, Y10, Y21, F22, Y23, F33, Y35, and F45) of the basic pancreatic trypsin inhibitor. Well-tempered metadynamics simulations were performed to estimate the ring-flipping free-energy surfaces for all eight aromatic residues. The results indicate that χ2 as a standalone collective variable (CV) is not sufficient to obtain computationally consistent results. Inclusion of a complementary CV, such as χ1(Cα-Cß), solved the problem for most residues and enabled us to classify fast and slow ring flips. This indicates the importance of librational motions in ring flips. Multiple pathways and mechanisms were observed for residues F4, Y10, and F22. Recrossing events were observed for residues F22 and F33, indicating a possible role of friction effects in ring flipping. The results demonstrate the successful application of infrequent metadynamics to estimate ring flip rates and identify certain limitations of the approach.

Flap Dynamics in Pepsin-Like Aspartic Proteases: A Computational Perspective Using Plasmepsin-II and BACE-1 as Model Systems.

The flexibility of ß hairpin structure known as the flap plays a key role in catalytic activity and substrate intake in pepsin-like aspartic proteases. Most of these enzymes share structural and sequential similarity. In this study, we have used apo Plm-II and BACE-1 as model systems. In the apo form of the proteases, a conserved tyrosine residue in the flap region remains in a dynamic equilibrium between the normal and flipped states through rotation of the χ1 and χ2 angles. Independent MD simulations of Plm-II and BACE-1 remained stuck either in the normal or flipped state. Metadynamics simulations using side-chain torsion angles (χ1 and χ2 of tyrosine) as collective variables sampled the transition between the normal and flipped states. Qualitatively, the two states were predicted to be equally populated. The normal and flipped states were stabilized by H-bond interactions to a tryptophan residue and to the catalytic aspartate, respectively. Further, mutation of tyrosine to an amino-acid with smaller side-chain, such as alanine, reduced the flexibility of the flap and resulted in a flap collapse (flap loses flexibility and remains stuck in a particular state). This is in accordance with previous experimental studies, which showed that mutation to alanine resulted in loss of activity in pepsin-like aspartic proteases. Our results suggest that the ring flipping associated with the tyrosine side-chain is the key order parameter that governs flap dynamics and opening of the binding pocket in most pepsin-like aspartic proteases.

How proteins modify water dynamics.

Much of biology happens at the protein-water interface, so all dynamical processes in this region are of fundamental importance. Local structural fluctuations in the hydration layer can be probed by 17O magnetic relaxation dispersion (MRD), which, at high frequencies, measures the integral of a biaxial rotational time correlation function (TCF)-the integral rotational correlation time. Numerous 17O MRD studies have demonstrated that this correlation time, when averaged over the first hydration shell, is longer than in bulk water by a factor 3-5. This rotational perturbation factor (RPF) has been corroborated by molecular dynamics simulations, which can also reveal the underlying molecular mechanisms. Here, we address several outstanding problems in this area by analyzing an extensive set of molecular dynamics data, including four globular proteins and three water models. The vexed issue of polarity versus topography as the primary determinant of hydration water dynamics is resolved by establishing a protein-invariant exponential dependence of the RPF on a simple confinement index. We conclude that the previously observed correlation of the RPF with surface polarity is a secondary effect of the correlation between polarity and confinement. Water rotation interpolates between a perturbed but bulk-like collective mechanism at low confinement and an exchange-mediated orientational randomization (EMOR) mechanism at high confinement. The EMOR process, which accounts for about half of the RPF, was not recognized in previous simulation studies, where only the early part of the TCF was examined. Based on the analysis of the experimentally relevant TCF over its full time course, we compare simulated and measured RPFs, finding a 30% discrepancy attributable to force field imperfections. We also compute the full 17O MRD profile, including the low-frequency dispersion produced by buried water molecules. Computing a local RPF for each hydration shell, we find that the perturbation decays exponentially with a decay "length" of 0.3 shells and that the second and higher shells account for a mere 3% of the total perturbation measured by 17O MRD. The only long-range effect is a weak water alignment in the electric field produced by an electroneutral protein (not screened by counterions), but this effect is negligibly small for 17O MRD. By contrast, we find that the 17O TCF is significantly more sensitive to the important short-range perturbations than the other two TCFs examined here.

The spatial range of protein hydration.

Proteins interact with their aqueous surroundings, thereby modifying the physical properties of the solvent. The extent of this perturbation has been investigated by numerous methods in the past half-century, but a consensus has still not emerged regarding the spatial range of the perturbation. To a large extent, the disparate views found in the current literature can be traced to the lack of a rigorous definition of the perturbation range. Stating that a particular solvent property differs from its bulk value at a certain distance from the protein is not particularly helpful since such findings depend on the sensitivity and precision of the technique used to probe the system. What is needed is a well-defined decay length, an intrinsic property of the protein in a dilute aqueous solution, that specifies the length scale on which a given physical property approaches its bulk-water value. Based on molecular dynamics simulations of four small globular proteins, we present such an analysis of the structural and dynamic properties of the hydrogen-bonded solvent network. The results demonstrate unequivocally that the solvent perturbation is short-ranged, with all investigated properties having exponential decay lengths of less than one hydration shell. The short range of the perturbation is a consequence of the high energy density of bulk water, rendering this solvent highly resistant to structural perturbations. The electric field from the protein, which under certain conditions can be long-ranged, induces a weak alignment of water dipoles, which, however, is merely the linear dielectric response of bulk water and, therefore, should not be thought of as a structural perturbation. By decomposing the first hydration shell into polarity-based subsets, we find that the hydration structure of the nonpolar parts of the protein surface is similar to that of small nonpolar solutes. For all four examined proteins, the mean number of water-water hydrogen bonds in the nonpolar subset is within 1% of the value in bulk water, suggesting that the fragmentation and topography of the nonpolar protein-water interface has evolved to minimize the propensity for protein aggregation by reducing the unfavorable free energy of hydrophobic hydration.

The geometry of protein hydration.

Based on molecular dynamics simulations of four globular proteins in dilute aqueous solution, with three different water models, we examine several, essentially geometrical, aspects of the protein-water interface that remain controversial or incompletely understood. First, we compare different hydration shell definitions, based on spatial or topological proximity criteria. We find that the best method for constructing monolayer shells with nearly complete coverage is to use a 5 Å water-carbon cutoff and a 4 Å water-water cutoff. Using this method, we determine a mean interfacial water area of 11.1 Å2 which appears to be a universal property of the protein-water interface. We then analyze the local coordination and packing density of water molecules in the hydration shells and in subsets of the first shell. The mean polar water coordination number in the first shell remains within 1% of the bulk-water value, and it is 5% lower in the nonpolar part of the first shell. The local packing density is obtained from additively weighted Voronoi tessellation, arguably the most physically realistic method for allocating space between protein and water. We find that water in all parts of the first hydration shell, including the nonpolar part, is more densely packed than in the bulk, with a shell-averaged density excess of 6% for all four proteins. We suggest reasons why this value differs from previous experimental and computational results, emphasizing the importance of a realistic placement of the protein-water dividing surface and the distinction between spatial correlation and packing density. The protein-induced perturbation of water coordination and packing density is found to be short-ranged, with an exponential decay "length" of 0.6 shells. We also compute the protein partial volume, analyze its decomposition, and argue against the relevance of electrostriction.

Can MM/GBSA calculations be sped up by system truncation?

We have studied whether calculations of the binding free energy of small ligands to a protein by the MM/GBSA approach (molecular mechanics combined with generalized Born and surface area solvation) can be sped up by including only a restricted number of atoms close to the ligand. If the protein is truncated before the molecular dynamics (MD) simulations, quite large changes are observed for the calculated binding energies, for example, 4 kJ/mol average difference for a radius of 19 Å for the binding of nine phenol derivatives to ferritin. The results are improved if no atoms are fixed in the simulations, with average and maximum errors of 2 and 3 kJ/mol at 19 Å and 3 and 6 kJ/mol at 7 Å. Similar results are obtained for two additional proteins, p38α MAP kinase and factor Xa. On the other hand, if energies are calculated on snapshots that are truncated after the MD simulation, all residues more than 8.5 Å from the ligand can be omitted without changing the energies by more than 1 kJ/mol on average (maximum error 1.4 kJ/mol). At the molecular mechanics level, the gain in computer time for such an approach is small. However, it shows what size of system should be used if the energies instead are calculated with a more demanding method, for example, quantum-mechanics. © 2017 Wiley Periodicals, Inc.

Prediction of binding poses to FXR using multi-targeted docking combined with molecular dynamics and enhanced sampling.

Advanced molecular docking methods often aim at capturing the flexibility of the protein upon binding to the ligand. In this study, we investigate whether instead a simple rigid docking method can be applied, if combined with multiple target structures to model the backbone flexibility and molecular dynamics simulations to model the sidechain and ligand flexibility. The methods are tested for the binding of 35 ligands to FXR as part of the first stage of the Drug Design Data Resource (D3R) Grand Challenge 2 blind challenge. The results show that the multiple-target docking protocol performs surprisingly well, with correct poses found for 21 of the ligands. MD simulations started on the docked structures are remarkably stable, but show almost no tendency of refining the structure closer to the experimentally found binding pose. Reconnaissance metadynamics enhances the exploration of new binding poses, but additional collective variables involving the protein are needed to exploit the full potential of the method.

Resolving the problem of trapped water in binding cavities: prediction of host-guest binding free energies in the SAMPL5 challenge by funnel metadynamics.

The funnel metadynamics method enables rigorous calculation of the potential of mean force along an arbitrary binding path and thereby evaluation of the absolute binding free energy. A problem of such physical paths is that the mechanism characterizing the binding process is not always obvious. In particular, it might involve reorganization of the solvent in the binding site, which is not easily captured with a few geometrically defined collective variables that can be used for biasing. In this paper, we propose and test a simple method to resolve this trapped-water problem by dividing the process into an artificial host-desolvation step and an actual binding step. We show that, under certain circumstances, the contribution from the desolvation step can be calculated without introducing further statistical errors. We apply the method to the problem of predicting host-guest binding free energies in the SAMPL5 blind challenge, using two octa-acid hosts and six guest molecules. For one of the hosts, well-converged results are obtained and the prediction of relative binding free energies is the best among all the SAMPL5 submissions. For the other host, which has a narrower binding pocket, the statistical uncertainties are slightly higher; longer simulations would therefore be needed to obtain conclusive results.

Ligand-Binding Affinity Estimates Supported by Quantum-Mechanical Methods.

One of the largest challenges of computational chemistry is calculation of accurate free energies for the binding of a small molecule to a biological macromolecule, which has immense implications in drug development. It is well-known that standard molecular-mechanics force fields used in most such calculations have a limited accuracy. Therefore, there has been a great interest in improving the estimates using quantum-mechanical (QM) methods. We review here approaches involving explicit QM energies to calculate binding affinities, with an emphasis on the methods, rather than on specific applications. Many different QM methods have been employed, ranging from semiempirical QM calculations, via density-functional theory, to strict coupled-cluster calculations. Dispersion and other empirical corrections are mandatory for the approximate methods, as well as large basis sets for the stricter methods. QM has been used for the ligand, for a few crucial groups around the ligand, for all the closest atoms (200-1000 atoms), or for the full receptor-ligand complex, but it is likely that with a proper embedding it might be enough to include all groups within ∼6 Šof the ligand. Approaches involving minimized structures, simulations of the end states of the binding reaction, or full free-energy simulations have been tested.

Converging ligand-binding free energies obtained with free-energy perturbations at the quantum mechanical level.

In this article, the convergence of quantum mechanical (QM) free-energy simulations based on molecular dynamics simulations at the molecular mechanics (MM) level has been investigated. We have estimated relative free energies for the binding of nine cyclic carboxylate ligands to the octa-acid deep-cavity host, including the host, the ligand, and all water molecules within 4.5 Å of the ligand in the QM calculations (158-224 atoms). We use single-step exponential averaging (ssEA) and the non-Boltzmann Bennett acceptance ratio (NBB) methods to estimate QM/MM free energy with the semi-empirical PM6-DH2X method, both based on interaction energies. We show that ssEA with cumulant expansion gives a better convergence and uses half as many QM calculations as NBB, although the two methods give consistent results. With 720,000 QM calculations per transformation, QM/MM free-energy estimates with a precision of 1 kJ/mol can be obtained for all eight relative energies with ssEA, showing that this approach can be used to calculate converged QM/MM binding free energies for realistic systems and large QM partitions. © 2016 The Authors. Journal of Computational Chemistry Published by Wiley Periodicals, Inc.

Flap Dynamics in Aspartic Proteases: A Computational Perspective.

Recent advances in biochemistry and drug design have placed proteases as one of the critical target groups for developing novel small-molecule inhibitors. Among all proteases, aspartic proteases have gained significant attention due to their role in HIV/AIDS, malaria, Alzheimer's disease, etc. The binding cleft is covered by one or two ß-hairpins (flaps) which need to be opened before a ligand can bind. After binding, the flaps close to retain the ligand in the active site. Development of computational tools has improved our understanding of flap dynamics and its role in ligand recognition. In the past decade, several computational approaches, for example molecular dynamics (MD) simulations, coarse-grained simulations, replica-exchange molecular dynamics (REMD) and metadynamics, have been used to understand flap dynamics and conformational motions associated with flap movements. This review is intended to summarize the computational progress towards understanding the flap dynamics of proteases and to be a reference for future studies in this field.

Binding affinities by alchemical perturbation using QM/MM with a large QM system and polarizable MM model.

The most general way to improve the accuracy of binding-affinity calculations for protein-ligand systems is to use quantum-mechanical (QM) methods together with rigorous alchemical-perturbation (AP) methods. We explore this approach by calculating the relative binding free energy of two synthetic disaccharides binding to galectin-3 at a reasonably high QM level (dispersion-corrected density functional theory with a triple-zeta basis set) and with a sufficiently large QM system to include all short-range interactions with the ligand (744-748 atoms). The rest of the protein is treated as a collection of atomic multipoles (up to quadrupoles) and polarizabilities. Several methods for evaluating the binding free energy from the 3600 QM calculations are investigated in terms of stability and accuracy. In particular, methods using QM calculations only at the endpoints of the transformation are compared with the recently proposed non-Boltzmann Bennett acceptance ratio (NBB) method that uses QM calculations at several stages of the transformation. Unfortunately, none of the rigorous approaches give sufficient statistical precision. However, a novel approximate method, involving the direct use of QM energies in the Bennett acceptance ratio method, gives similar results as NBB but with better precision, ∼3 kJ/mol. The statistical error can be further reduced by performing a greater number of QM calculations.

Coupled-cluster interaction energies for 200-atom host-guest systems.

We have developed a method to calculate interaction energies of large systems (such as host-guest or even protein-ligand systems) at the local coupled-cluster with singles, doubles, and perturbative triples level, and with extrapolation to the limit of a complete basis set. The method is based on the polarizable multipole interactions with supermolecular pairs molecular fractionation approach, which combines a pairwise quantum-mechanical evaluation of the short-range interactions with a polarizable multipole treatment of many-body effects. The method is tested for nine guest molecules binding to an octa-acid host (in total 198-207 atoms), as part of the SAMPL4 blind challenge. From the test calculations, the accuracy of the approach is found to be 10 kJ mol(-1) or better. Comparison with dispersion-corrected density functional theory reveals that the latter underestimates the dispersion contribution for this type of system, which leads to a difference in the ranking of the ligands.

Free-energy perturbation and quantum mechanical study of SAMPL4 octa-acid host-guest binding energies.

We have estimated free energies for the binding of nine cyclic carboxylate guest molecules to the octa-acid host in the SAMPL4 blind-test challenge with four different approaches. First, we used standard free-energy perturbation calculations of relative binding affinities, performed at the molecular-mechanics (MM) level with TIP3P waters, the GAFF force field, and two different sets of charges for the host and the guest, obtained either with the restrained electrostatic potential or AM1-BCC methods. Both charge sets give good and nearly identical results, with a mean absolute deviation (MAD) of 4 kJ/mol and a correlation coefficient (R (2)) of 0.8 compared to experimental results. Second, we tried to improve these predictions with 28,800 density-functional theory (DFT) calculations for selected snapshots and the non-Boltzmann Bennett acceptance-ratio method, but this led to much worse results, probably because of a too large difference between the MM and DFT potential-energy functions. Third, we tried to calculate absolute affinities using minimised DFT structures. This gave intermediate-quality results with MADs of 5-9 kJ/mol and R (2) = 0.6-0.8, depending on how the structures were obtained. Finally, we tried to improve these results using local coupled-cluster calculations with single and double excitations, and non-iterative perturbative treatment of triple excitations (LCCSD(T0)), employing the polarisable multipole interactions with supermolecular pairs approach. Unfortunately, this only degraded the predictions, probably because of a mismatch between the solvation energies obtained at the DFT and LCCSD(T0) levels.

Prediction of hydration free energies for the SAMPL4 data set with the AMOEBA polarizable force field.

Hydration free energy calculations are often used to validate molecular simulation methodologies and molecular mechanics force fields. We use the free-energy perturbation method together with the AMOEBA polarizable force field and the Poltype parametrization protocol to predict the hydration free energies of 52 molecules as part of the SAMPL4 blind challenge. For comparison, similar calculations are performed using the non-polarizable General Amber force field. Against our expectations, the latter force field gives the better results compared to experiment. One possible explanation is the sensitivity of the AMOEBA results to the conformation used for parametrization.

Prediction of SAMPL4 host-guest binding affinities using funnel metadynamics.

Accurately predicting binding affinities between ligands and macromolecules has been a much sought-after goal. A tremendous amount of resources can be saved in the pharmaceutical industry through accurate binding-affinity prediction and hence correct decision-making for the drug discovery processes. Owing to the structural complexity of macromolecules, one of the issues in binding affinity prediction using molecular dynamics is the adequate sampling of the conformational space. Recently, the funnel metadynamics method (Limongelli et al. in Proc Natl Acad Sci USA 110:6358, 2013) was developed to enhance the sampling of the ligand at the binding site as well as in the solvated state, and offer the possibility to predict the absolute binding free energy. We apply funnel metadynamics to predict host-guest binding affinities for the cucurbit[7]uril host as part of the SAMPL4 blind challenge. Using total simulation times of 300-400 ns per ligand, we show that the errors due to inadequate sampling are below 1 kcal/mol. However, despite the large investment in terms of computational time, the results compared to experiment are not better than a random guess. As we obtain differences of up to 11 kcal/mol when switching between two commonly used force fields (with automatically generated parameters), we strongly believe that in the pursuit of accurate binding free energies a more careful force-field parametrization is needed to address this type of system.

Comparison of MM/GBSA calculations based on explicit and implicit solvent simulations.

Molecular mechanics with generalised Born and surface area solvation (MM/GBSA) is a popular method to calculate the free energy of the binding of ligands to proteins. It involves molecular dynamics (MD) simulations with an explicit solvent of the protein-ligand complex to give a set of snapshots for which energies are calculated with an implicit solvent. This change in the solvation method (explicit → implicit) would strictly require that the energies are reweighted with the implicit-solvent energies, which is normally not done. In this paper we calculate MM/GBSA energies with two generalised Born models for snapshots generated by the same methods or by explicit-solvent simulations for five synthetic N-acetyllactosamine derivatives binding to galectin-3. We show that the resulting energies are very different both in absolute and relative terms, showing that the change in the solvent model is far from innocent and that standard MM/GBSA is not a consistent method. The ensembles generated with the various solvent models are quite different with root-mean-square deviations of 1.2-1.4 Å. The ensembles can be converted to each other by performing short MD simulations with the new method, but the convergence is slow, showing mean absolute differences in the calculated energies of 6-7 kJ mol(-1) after 2 ps simulations. Minimisations show even slower convergence and there are strong indications that the energies obtained from minimised structures are different from those obtained by MD.

Accurate Reaction Energies in Proteins Obtained by Combining QM/MM and Large QM Calculations.

We here suggest and test a new method to obtain stable energies in proteins for charge-neutral reactions by running large quantum mechanical (QM) calculations on structures obtained by combined QM and molecular mechanics (QM/MM) geometry optimization on several snapshots from molecular dynamics simulations. As a test case, we use a proton transfer between a metal-bound cysteine residue and a second-sphere histidine residue in the active site of [Ni,Fe] hydrogenase, which has been shown to be very sensitive to the surroundings. We include in the QM calculations all residues within 4.5 Å of the active site, two capped residues on each side of the active-site residues, and all charged groups that are buried inside the protein, which for this enzyme includes three iron-sulfur clusters, in total, 930 atoms. These calculations are performed at the BP86/def2-SV(P) level, but the energies are then extrapolated to the B3LYP/def2-TZVP level with a smaller QM system, and zero-point energy, entropy, and thermal effects are added. We test three approaches to model the remaining atoms of the protein solvent, viz., by standard QM/MM approaches using either mechanical or electrostatic embedding or by using a continuum solvation model for the large QM systems. Quite encouragingly, the three approaches give the same results within 14 kJ/mol, and variations in the size of the QM system do not change the energies by more than 8 kJ/mol, provided that the QM/MM junctions are not moved closer to the QM system. The statistical precision for the average over 10 snapshots is 1-3 kJ/mol.

Effect of Geometry Optimizations on QM-Cluster and QM/MM Studies of Reaction Energies in Proteins.

We have examined the effect of geometry optimization on energies calculated with the quantum mechanical (QM) cluster, combined QM and molecular mechanics (QM/MM), and big-QM approaches (very large single-point QM calculations taken from QM/MM-optimized structures, including all atoms within 4.5 Å of the minimal active site, all buried charged groups in the protein, and truncations moved at least three residues away from the active site). We studied a simple proton-transfer reaction between His-79 and Cys-546 in the active site of [Ni,Fe] hydrogenase and optimize QM systems of 50 different sizes (56-362 atoms). Geometries optimized with QM/MM are stable and reliable, whereas QM-cluster optimizations give larger changes in the structures and sometimes lead to large distortions in the active site if some hydrogen-bond partners to the metal ligands are omitted. Keeping 2-3 atoms for each truncated residue (rather than one) fixed in the optimization improves the results but does not solve all problems for the QM-cluster optimizations. QM-cluster energies in vacuum and a continuum solvent are insensitive to the geometry optimizations with a mean absolute change upon the optimizations of only 4-7 kJ/mol. This shows that geometry optimizations do not decrease the dependence of QM-cluster energies on how the QM system is selected; there is still a ∼60 kJ/mol difference between calculations in which groups have been added to the QM system according to their distance to the active site or based on QM/MM free-energy components. QM/MM energies do not show such a difference, but they converge rather slowly with respect to the size of the QM system, although the convergence is improved by moving truncations away from the active site. The big-QM energies are stable over the 50 different optimized structures, 57 ± 1 kJ/mol, although some smaller trends can be discerned. This shows that both QM-cluster geometries and energies should be interpreted with caution. Instead, we recommend QM/MM for geometry optimizations and energies calculated by the big-QM approach.

Locating binding poses in protein-ligand systems using reconnaissance metadynamics.

A molecular dynamics-based protocol is proposed for finding and scoring protein-ligand binding poses. This protocol uses the recently developed reconnaissance metadynamics method, which employs a self-learning algorithm to construct a bias that pushes the system away from the kinetic traps where it would otherwise remain. The exploration of phase space with this algorithm is shown to be roughly six to eight times faster than unbiased molecular dynamics and is only limited by the time taken to diffuse about the surface of the protein. We apply this method to the well-studied trypsin-benzamidine system and show that we are able to refind all the poses obtained from a reference EADock blind docking calculation. These poses can be scored based on the length of time the system remains trapped in the pose. Alternatively, one can perform dimensionality reduction on the output trajectory and obtain a map of phase space that can be used in more expensive free-energy calculations.