Slow dynamics in protein fluctuations revealed by time-structure based independent component analysis: the case of domain motions

Protein dynamics on a long time scale was investigated using all-atom molecular dynamics (MD) simulation and time-structure based independent component analysis (tICA). We selected the lysine-, arginine-, ornithine-binding protein (LAO) as a target protein and focused on its domain motions in the open state. A MD simulation of the LAO in explicit water was performed for 600 ns, in which slow and large-amplitude domain motions of the LAO were observed. After extracting domain motions by rigid-body domain analysis, the tICA was applied to the obtained rigid-body trajectory, yielding slow modes of the LAO's domain motions in order of decreasing time scale. The slowest mode detected by the tICA represented not a closure motion described by a largest-amplitude mode determined by the principal component analysis but a twist motion with a time scale of tens of nanoseconds. The slow dynamics of the LAO were well described by only the slowest mode and were characterized by transitions between two basins. The results show that tICA is promising for describing and analyzing slow dynamics of proteins.     

Possible involvement of collective domain movement in the catalytic reaction of soluble epoxide hydrolase

The hydrolysis of 1S,2S‐trans‐methylstyrene oxide by soluble epoxide hydrolases is studied by a 4‐ns molecular dynamics (MD) simulation. An analysis of the extent of correlated motions in the active site was carried out. Based on the calculated cross correlation coefficients form the covariance matrix, a new correlation parameter, termed the supercorrelation coefficient, is introduced. The supercorrelation coefficient indicates the extent to which two amino acid residues move in a correlated manner with respect to all other residues in the protein. The resulting map of the supercorrelation coefficients was used to identify segments of the protein that may show collective domain movements. Interestingly, an anti‐correlated motion is located across the active site, involving the catalytic triad and the tyrosines. This may suggest that if a link exists between enzyme dynamics and catalysis, it may be through an anti‐correlated collective domain movement that compresses the active site, thus initiating the conversion of E–NAC to E–TS. © 2004 Wiley Periodicals, Inc. Int J Quantum Chem, 2004     

Dynamics of proton transfer and vibrational relaxation in dilute hydrofluoric acid

The molecular mechanisms in both vibrational relaxation and proton transfer (PT) associated with infrared (IR)-induced PT in a dilute hydrofluoric acid solution at ambient temperature are studied by molecular dynamics (MD) simulations with the multistate empirical valence bond model. To investigate the solvation dynamics, a collective solvent coordinate and its perpendicular bath modes are defined from the diabatic energy gap and their motions are examined by the generalized Langevin equation (GLE) formalism. The GLE analysis using the equilibrium MD simulation shows that the major solvent reorganizations in the PT are represented by the libration and hindered translation. In particular, the libration gives the stronger coupling to the solvent reorganization and the faster relaxation. The nonequilibrium MD simulation demonstrated that both the HF stretching vibration and the solvent reorganization relax on a similar time scale and thus compete in the PT. It also supported the "presolvation mechanism" for the PT in this system.     

CAMERRA: An analysis tool for the computation of conformational dynamics by evaluating residue–residue associations

A computational method which extracts the dominant motions from an ensemble of biomolecular conformations via a correlation analysis of residue–residue contacts is presented. The algorithm first renders the structural information into contact matrices, then constructs the collective modes based on the correlated dynamics of a selected set of dynamic contacts. Associated programs can bridge the results for further visualization using graphics software. The aim of this method is to provide an analysis of conformations of biopolymers from the contact viewpoint. It may assist a systematical uncovering of conformational switching mechanisms existing in proteins and biopolymer systems in general by statistical analysis of simulation snapshots. In contrast to conventional correlation analyses of Cartesian coordinates (such as distance covariance analysis and Cartesian principal component analysis), this program also provides an alternative way to locate essential collective motions in general. Herein, we detail the algorithm in a stepwise manner and comment on the importance of the method as applied to decoding allosteric mechanisms. © 2018 Wiley Periodicals, Inc.     

General framework for studying the dynamics of folded and nonfolded proteins by NMR relaxation spectroscopy and MD simulation

A general framework is presented for the interpretation of NMR relaxation data of proteins. The method, termed isotropic reorientational eigenmode dynamics (iRED), relies on a principal component analysis of the isotropically averaged covariance matrix of the lattice functions of the spin interactions responsible for spin relaxation. The covariance matrix, which is evaluated using a molecular dynamics (MD) simulation, is diagonalized yielding reorientational eigenmodes and amplitudes that reveal detailed information about correlated protein dynamics. The eigenvalue distribution allows one to quantitatively assess whether overall and internal motions are statistically separable. To each eigenmode belongs a correlation time that can be adjusted to optimally reproduce experimental relaxation parameters. A key feature of the method is that it does not require separability of overall tumbling and internal motions, which makes it applicable to a wide range of systems, such as folded, partially folded, and unfolded biomolecular systems and other macromolecules in solution. The approach was applied to NMR relaxation data of ubiquitin collected at multiple magnetic fields in the native form and in the partially folded A-state using MD trajectories with lengths of 6 and 70 ns. The relaxation data of native ubiquitin are well reproduced after adjustment of the correlation times of the 10 largest eigenmodes. For this state, a high degree of separability between internal and overall motions is present as is reflected in large amplitude and collectivity gaps between internal and overall reorientational modes. In contrast, no such separability exists for the A-state. Residual overall tumbling motion involving the N-terminal beta-sheet and the central helix is observed for two of the largest modes only. By adjusting the correlation times of the 10 largest modes, a high degree of consistency between the experimental relaxation data and the iRED model is reached for this highly flexible biomolecule.     

Correlated atomic motions in liquid deuterium fluoride studied by coherent quasielastic neutron scattering

The collective dynamics of liquid deuterium fluoride are studied by means of high-resolution quasielastic and inelastic neutron scattering over a range of four decades in energy transfer. The spectra show a low-energy coherent quasielastic component which arises from correlated stochastic motions as well as a broad inelastic feature originating from overdamped density oscillations. While these results are at variance with previous works which report on the presence of propagating collective modes, they are fully consistent with neutron diffraction, nuclear magnetic resonance, and infrared/Raman experiments on this prototypical hydrogen-bonded fluid.     

Coarse-graining protein structures with local multivariate features from molecular dynamics

A multivariate statistical theory, local feature analysis (LFA), extracts functionally relevant domains from molecular dynamics (MD) trajectories. The LFA representations, like those of principal component analysis (PCA), are low dimensional and provide a reduced basis set for collective motions of simulated proteins, but the local features are sparsely distributed and spatially localized, in contrast to global PCA modes. One key problem in the assignment of local features is the coarse-graining of redundant LFA output functions by means of seed atoms. One can solve the combinatorial problem by adding seed atoms one after another to a growing set, minimizing a reconstruction error at each addition. This allows for an efficient implementation, but the sequential algorithm does not guarantee the optimal mutual correlation of the sequentially assigned features. Here, we present a novel coarse-graining algorithm for proteins that directly minimizes the mutual correlation of seed atoms by Monte Carlo (MC) simulations. Tests on MD trajectories of two biological systems, bacteriophage T4 lysozyme and myosin II motor domain S1, demonstrate that the new algorithm provides statistically reproducible results and describes functionally relevant dynamics. The well-known undersampling of large-scale motion by short MD simulations is apparent also in our model, but the new coarse-graining offers a major advantage over PCA; converged features are invariant across multiple windows of the trajectory, dividing the protein into converged regions and a smaller number of localized, undersampled regions. In addition to its use in structure classification, the proposed coarse-graining thus provides a localized measure of MD sampling efficiency.     

Exploring multiple timescale motions in protein GB3 using accelerated molecular dynamics and NMR spectroscopy

Biological function relies on the complex spectrum of conformational dynamics occurring in biomolecules. We have combined Accelerated Molecular Dynamics (AMD) with experimental results derived from NMR to probe multiple time-scale motions in the third IgG-binding domain of Protein G (GB3). AMD is shown to accurately reproduce the amplitude and distribution of slow motional modes characterized using residual dipolar couplings, reporting on dynamics up to the millisecond timescale. In agreement with experiment, larger amplitude slower motions are localized in the beta-strand/loop motif spanning residues 14-24 and in loop 42-44. Principal component analysis shows these fluctuations participating in the primary mode, substantiating the existence of a correlated motion traversing the beta-sheet that culminates in maximum excursions at the active site of the molecule. Fast dynamics were simulated using extensive standard MD simulations and compared to order parameters extracted from 15N relaxation. Notably 60 2-ns fully solvated MD simulations exploring the different conformational substates sampled from AMD resulted in better reproduction of order parameters compared to the same number of simulations starting from the relaxed crystal structure. This illustrates the inherent dependence of protein dynamics on local conformational topology. The results provide rare insight into the complex hierarchy of dynamics present in GB3 and allow us to develop a model of the conformational landscape native to the protein, appearing as a steep sided potential well whose flat bottom comprises multiple similar but discrete conformational substates.     

Harmonic analysis of large systems. III. Comparison with molecular dynamics

Atomic motions in bovine pancreatic trypsin inhibitor (BPTI), derived from molecular dynamics, harmonic analysis, and quasiharmonic analysis, are compared when a single protein model, energy parameters, and environment are employed. Molecular dynamics (MD) was carried out for 2 nanoseconds. An average structure was determined from the last nanosecond of the MD simulation, when no major structural changes were observed. This structure was used for several harmonic analysis calculations as well as for a reference structure for the quasiharmonic analysis, for both full basis and reduced basis sets. In contrast to the harmonic analysis results, the quasiharmonic reduced basis calculation using a spherical harmonics reduced basis provided good agreement with the full basis calculation, suggesting that when anharmonic effects are considered, BPTI can behave as a homogeneous object. An extensive analysis of the normal modes from a diverse set of 201 minimized MD simulation frames was performed. On only the sub-picosecond time scale were energy minima revisited after a transition to another state. This analysis shows that the dynamics average structure is not representative of the simulation frames in terms of energy and vibrational frequencies. For this model of BPTI, 42% of the motion (mean-squared fluctuation) can be attributed to harmonic limit behavior. A spectral analysis of the correlation function of deformation for a particular normal mode or quasiharmonic mode can be used to determine the time scales of motions which correspond to harmonic vibration, large-scale drift, or sharp transitions between local substrates. © 1995 John Wiley & Sons, Inc.     

Hierarchical structure of the energy landscape of proteins revisited by time series analysis. I. Mimicking protein dynamics in different time scales

Time series models, which are constructed from the projections of the molecular-dynamics (MD) runs on principal components (modes), are used to mimic the dynamics of two proteins: tendamistat and immunity protein of colicin E7 (ImmE7). Four independent MD runs of tendamistat and three independent runs of ImmE7 protein in vacuum are used to investigate the energy landscapes of these proteins. It is found that mean-square displacements of residues along the modes in different time scales can be mimicked by time series models, which are utilized in dividing protein dynamics into different regimes with respect to the dominating motion type. The first two regimes constitute the dominance of intraminimum motions during the first 5 ps and the random walk motion in a hierarchically higher-level energy minimum, which comprise the initial time period of the trajectories up to 20-40 ps for tendamistat and 80-120 ps for ImmE7. These are also the time ranges within which the linear nonstationary time series are completely satisfactory in explaining protein dynamics. Encountering energy barriers enclosing higher-level energy minima constrains the random walk motion of the proteins, and pseudorelaxation processes at different levels of minima are detected in tendamistat, depending on the sampling window size. Correlation (relaxation) times of 30-40 ps and 150-200 ps are detected for two energy envelopes of successive levels for tendamistat, which gives an overall idea about the hierarchical structure of the energy landscape. However, it should be stressed that correlation times of the modes are highly variable with respect to conformational subspaces and sampling window sizes, indicating the absence of an actual relaxation. The random-walk step sizes and the time length of the second regime are used to illuminate an important difference between the dynamics of the two proteins, which cannot be clarified by the investigation of relaxation times alone: ImmE7 has lower-energy barriers enclosing the higher-level energy minimum, preventing the protein to relax and letting it move in a random-walk fashion for a longer period of time.     

Revealing time bunching effect in single-molecule enzyme conformational dynamics

In this perspective, we focus our discussion on how the single-molecule spectroscopy and statistical analysis are able to reveal enzyme hidden properties, taking the study of T4 lysozyme as an example. Protein conformational fluctuations and dynamics play a crucial role in biomolecular functions, such as in enzymatic reactions. Single-molecule spectroscopy is a powerful approach to analyze protein conformational dynamics under physiological conditions, providing dynamic perspectives on a molecular-level understanding of protein structure-function mechanisms. Using single-molecule fluorescence spectroscopy, we have probed T4 lysozyme conformational motions under the hydrolysis reaction of a polysaccharide of E. coli B cell walls by monitoring the fluorescence resonant energy transfer (FRET) between a donor-acceptor probe pair tethered to T4 lysozyme domains involving open-close hinge-bending motions. Based on the single-molecule spectroscopic results, molecular dynamics simulation, a random walk model analysis, and a novel 2D statistical correlation analysis, we have revealed a time bunching effect in protein conformational motion dynamics that is critical to enzymatic functions. Bunching effect implies that conformational motion times tend to bunch in a finite and narrow time window. We show that convoluted multiple Poisson rate processes give rise to the bunching effect in the enzymatic reaction dynamics. Evidently, the bunching effect is likely common in protein conformational dynamics involving in conformation-gated protein functions. In this perspective, we will also discuss a new approach of 2D regional correlation analysis capable of analyzing fluctuation dynamics of complex multiple correlated and anti-correlated fluctuations under a non-correlated noise background. Using this new method, we are able to map out any defined segments along the fluctuation trajectories and determine whether they are correlated, anti-correlated, or non-correlated; after which, a cross correlation analysis can be applied for each specific segment to obtain a detailed fluctuation dynamics analysis.     

Hydration dynamics and time scales of coupled water-protein fluctuations

We report experimental and theoretical studies on water and protein dynamics following photoexcitation of apomyoglobin. Using site-directed mutation and with femtosecond resolution, we experimentally observed relaxation dynamics with a biphasic distribution of time scales, 5 and 87 ps, around the site Trp7. Theoretical studies using both linear response and direct nonequilibrium molecular dynamics (MD) calculations reproduced the biphasic behavior. Further constrained MD simulations with either frozen protein or frozen water revealed the molecular mechanism of slow hydration processes and elucidated the role of protein fluctuations. Observation of slow water dynamics in MD simulations requires protein flexibility, regardless of whether the slow Stokes shift component results from the water or protein contribution. The initial dynamics in a few picoseconds represents fast local motions such as reorientations and translations of hydrating water molecules, followed by slow relaxation involving strongly coupled water-protein motions. We observed a transition from one isomeric protein configuration to another after 10 ns during our 30 ns ground-state simulation. For one isomer, the surface hydration energy dominates the slow component of the total relaxation energy. For the other isomer, the slow component is dominated by protein interactions with the chromophore. In both cases, coupled water-protein motion is shown to be necessary for observation of the slow dynamics. Such biologically important water-protein motions occur on tens of picoseconds. One significant discrepancy exists between theory and experiment, the large inertial relaxation predicted by simulations but clearly absent in experiment. Further improvements required in the theoretical model are discussed.     

Localized and Collective Motions in HET‐s(218‐289) Fibrils from Combined NMR Relaxation and MD Simulation

Nuclear magnetic resonance (NMR) relaxation data and molecular dynamics (MD) simulations are combined to characterize the dynamics of the fungal prion HET‐s(218‐289) in its amyloid form. NMR data is analyzed with the dynamics detector method, which yields timescale‐specific information. An analogous analysis is performed on MD trajectories. Because specific MD predictions can be verified as agreeing with the NMR data, MD was used for further interpretation of NMR results: for the different timescales, cross‐correlation coefficients were derived to quantify the correlation of the motion between different residues. Short timescales are the result of very local motions, while longer timescales are found for longer‐range correlated motion. Similar trends on ns‐ and μs‐timescales suggest that μs motion in fibrils is the result of motion correlated over many fibril layers.     

Collective motions of myosin head derived from backbone molecular dynamics and combination with X‐ray solution scattering data

Collective modes of myosin head, S1, were derived from molecular dynamics trajectories of a simplified protein model, backbone model. Because the model requires only the positions of C_α‐atoms, large proteins are tractable, even when the amino acid sequence and the side‐chain orientations are unknown. The S1 is a large molecule and only the C_α‐atomic positions have been experimentally determined. In the simulation, large flipping motions of the extended α‐helical C‐terminal tail of S1 were found only in the largest and second largest amplitude collective modes, which were approximately perpendicular to each other. A few collective modes other than the first and second largest ones largely deformed the actin binding site of S1. The pair‐distance distribution, obtained from the X‐ray solution scattering, suggests a large conformational change of S1, isolated in solution without binding to actin filament, during the ATP hydrolysis. The modeling of the large conformational change was done with using the collective modes, and showed that the second mode mainly contributed to the large conformational change. The modeling procedure, introduced here, can be easily generalized for various experimental data, and applicable to the collective modes obtained from an all‐atom simulation. © 2001 John Wiley & Sons, Inc. J Comput Chem 22: 1983–1994, 2001     

Theoretical study of the truncated hemoglobin HbN: exploring the molecular basis of the NO detoxification mechanism

Mycobacterium tuberculosis is the causative agent of human tuberculosis. The nitric oxide reaction with oxy-truncated hemoglobin N (trHbN) has been proposed to be responsible for the resistance mechanism by which this microorganism can evade the toxic effects of NO. In this work, we explore the molecular basis of the NO detoxification mechanism using a combination of classical and hybrid quantum-classical (QM-MM) simulation techniques. We have investigated the structural flexibility of the protein, the ligand affinity properties, and the nitric oxide reaction with coordinated O2. The analysis of the classical MD trajectory allowed us to identify Phe62 as the gate of the main channel for ligand diffusion to the active site. Moreover, the opening of the channel stems from the interplay between collective backbone motions and local rearrangements in the side chains of the residues that form the bottleneck of the tunnel. Even though the protein environment is not found to make a significant contribution to the heme moiety catalyzed reaction, the binding site influences the physiological function of the enzyme at three different levels. First, by isolating the intermediates formed in the reaction, it prevents nondesired reactions from proceeding. Second, it modulates the ligand (O2, NO) affinity of the protein, which can be ascribed to both distal and proximal effects. Finally, the stabilization of the Tyr33-Gln58 pair upon O2 binding might alter the essential dynamics of the protein, leading in turn to a mechanism for ligand-induced regulation.     

On the applicability of elastic network normal modes in small-molecule docking

Incorporating backbone flexibility into protein-ligand docking is still a challenging problem. In protein-protein docking, normal mode analysis (NMA) has become increasingly popular as it can be used to describe the collective motions of a biological system, but the question of whether NMA can also be useful in predicting the conformational changes observed upon small-molecule binding has only been addressed in a few case studies. Here, we describe a large-scale study on the applicability of NMA for protein-ligand docking using 433 apo/holo pairs of the Astex data sets. On the basis of sets of the first normal modes from the apo structure, we first generated for each paired holo structure a set of conformations that optimally reproduce its C(α) trace with respect to the underlying normal mode subspace. Using AutoDock, GOLD, and FlexX we then docked the original ligands into these conformations to assess how the docking performance depends on the number of modes used to reproduce the holo structure. The results of our study indicate that, even for such a best-case scenario, the use of normal mode analysis in small-molecule docking is restricted and that a general rule on how many modes to use does not seem to exist or at least is not easy to find.