MD Simulation Data Research Articles

Inversion of the Side‐Chain Stereochemistry of Indvidual Thr or Ile Residues in a Protein Molecule: Impact on the Folding, Stability, and Structure of the ShK Toxin

AbstractShK toxin is a cysteine‐rich 35‐residue protein ion‐channel ligand isolated from the sea anemone Stichodactyla helianthus. In this work, we studied the effect of inverting the side chain stereochemistry of individual Thr or Ile residues on the properties of the ShK protein. Molecular dynamics simulations were used to calculate the free energy cost of inverting the side‐chain stereochemistry of individual Thr or Ile residues. Guided by the computational results, we used chemical protein synthesis to prepare three ShK polypeptide chain analogues, each containing either an allo‐Thr or an allo‐Ile residue. The three allo‐Thr or allo‐Ile‐containing ShK polypeptides were able to fold into defined protein products, but with different folding propensities. Their relative thermal stabilities were measured and were consistent with the MD simulation data. Structures of the three ShK analogue proteins were determined by quasi‐racemic X‐ray crystallography and were similar to wild‐type ShK. All three ShK analogues retained ion‐channel blocking activity.

Learning reduced kinetic Monte Carlo models of complex chemistry from molecular dynamics.

We propose a novel statistical learning framework for automatically and efficiently building reduced kinetic Monte Carlo (KMC) models of large-scale elementary reaction networks from data generated by a single or few molecular dynamics simulations (MD). Existing approaches for identifying species and reactions from molecular dynamics typically use bond length and duration criteria, where bond duration is a fixed parameter motivated by an understanding of bond vibrational frequencies. In contrast, we show that for highly reactive systems, bond duration should be a model parameter that is chosen to maximize the predictive power of the resulting statistical model. We demonstrate our method on a high temperature, high pressure system of reacting liquid methane, and show that the learned KMC model is able to extrapolate more than an order of magnitude in time for key molecules. Additionally, our KMC model of elementary reactions enables us to isolate the most important set of reactions governing the behavior of key molecules found in the MD simulation. We develop a new data-driven algorithm to reduce the chemical reaction network which can be solved either as an integer program or efficiently using L1 regularization, and compare our results with simple count-based reduction. For our liquid methane system, we discover that rare reactions do not play a significant role in the system, and find that less than 7% of the approximately 2000 reactions observed from molecular dynamics are necessary to reproduce the molecular concentration over time of methane. The framework described in this work paves the way towards a genomic approach to studying complex chemical systems, where expensive MD simulation data can be reused to contribute to an increasingly large and accurate genome of elementary reactions and rates.

Unveiling the Mechanism of Arginine Transport through AdiC with Molecular Dynamics Simulations: The Guiding Role of Aromatic Residues

Commensal and pathogenic enteric bacteria have developed several systems to adapt to proton leakage into the cytoplasm resulting from extreme acidic conditions. One such system involves arginine uptake followed by export of the decarboxylated product agmatine, carried out by the arginine/agmatine antiporter (AdiC), which thus works as a virtual proton pump. Here, using classical and targeted molecular dynamics, we investigated at the atomic level the mechanism of arginine transport through AdiC of E. coli. Overall, our MD simulation data clearly demonstrate that global rearrangements of several transmembrane segments are necessary but not sufficient for achieving transitions between structural states along the arginine translocation pathway. In particular, local structural changes, namely rotameric conversions of two aromatic residues, are needed to regulate access to both the outward- and inward-facing states. Our simulations have also enabled identification of a few residues, overwhelmingly aromatic, which are essential to guiding arginine in the course of its translocation. Most of them belong to gating elements whose coordinated motions contribute to the alternating access mechanism. Their conservation in all known E. coli acid resistance antiporters suggests that the transport mechanisms of these systems share common features. Last but not least, knowledge of the functional properties of AdiC can advance our understanding of the members of the amino acid-carbocation-polyamine superfamily, notably in eukaryotic cells.

Thermal transport across carbon nanotube-graphene covalent and van der Waals junctions

Carbon nanotubes and graphene are promising materials for thermal management applications due to their high thermal conductivities. However, their thermal properties are anisotropic, and the radial or cross-plane direction thermal conductivity is low. A 3D Carbon nanotube (CNT)-graphene structure has previously been proposed to address this limitation, and direct molecular dynamics simulations have been used to predict the associated thermal conductivity. In this work, by recognizing that thermal resistance primarily comes from CNT-graphene junctions, a simple network model of thermal transport in pillared graphene structure is developed. Using non-equilibrium molecular dynamics, the resistance across an individual CNT-graphene junction with sp2 covalent bonds is found to be around 6×10−11 m2K/W, which is significantly lower than typical values reported for planar interfaces between dissimilar materials. In contrast, the resistance across a van der Waals junction is about 4×10−8 m2K/W. Interestingly, when the CNT pillar length is small, the interfacial resistance of the sp2 covalent junction is found to decrease as the CNT pillar length decreases, suggesting the presence of coherence effects. To explain this intriguing trend, the junction thermal resistance is decomposed into interfacial region and boundary components, and it is found that while the boundary resistance has little dependence on the pillar length, the interfacial region resistance decreases as the pillar length decreases. This is explained by calculating the local phonon density of states (LDOS) of different regions near the boundary. The LDOS overlap between the interfacial region and the center region of CNT increases as the pillar length decreases, leading to the decrease of interfacial region resistance. The junction resistance Rj is eventually used in the network model to estimate the effective thermal conductivity, and the results agree well with direct MD simulation data, demonstrating the effectiveness of our model.

Chain end mobilities in polymer melts--a computational study.

The Rouse model can be regarded as the standard model to describe the dynamics of a short polymer chain under melt conditions. In this contribution, we explicitly check one of the fundamental assumptions of this model, namely, that of a uniform friction coefficient for all monomers, on the basis of MD simulation data of a poly(ethylene oxide) (PEO) melt. This question immediately arises from the fact that in a real polymer melt, the terminal monomers have on average more intermolecular neighbors than the central monomers, and one would expect that exactly these details affect the precise value of the friction coefficient. The mobilities are determined by our recently developed statistical method, which provides detailed insights into the local polymer dynamics. Moreover, it yields complementary information to that obtained from the mean square displacement (MSD) or the Rouse mode analysis. It turns out that the Rouse assumption of a uniform mobility is fulfilled to a good approximation for the PEO melt. However, a more detailed analysis reveals that the underlying microscopic dynamics are highly affected by different contributions from intra- and intermolecular excluded volume interactions, which cannot be taken into account by a modified friction coefficient. Minor deviations occur only for the terminal monomers on larger time scales, which can be attributed to the presence of two different escape mechanisms from their first coordination sphere. These effects remain elusive when studying the dynamics with the MSD only.

Structural investigation of the interaction between the tandem SH3 domains of c-Cbl-associated protein and vinculin.

c-Cbl-associated protein (CAP) is an important cytoskeletal adaptor protein involved in the regulation of adhesion turnover. The interaction between CAP and vinculin is critical for the recruitment of CAP to focal adhesions. The tandem SH3 domains (herein termed SH3a and SH3b) of CAP are responsible for its interaction with vinculin. However, the structural mechanism underlying the interaction between CAP and vinculin is poorly understood. In this manuscript, we report the solution structure of the tandem SH3 domains of CAP. Our NMR and ITC data indicate that the SH3a and SH3b domains of CAP simultaneously bind to a long proline-rich region of vinculin with different binding specificities. Furthermore, the crystal structures of the individual SH3a and SH3b domains complexed with their substrate peptides indicate that Q807(SH3a) and D881(SH3b) are the critical residues determining the different binding specificities of the SH3 domains. Based on the obtained structural information, a model of the SH3ab-vinculin complex was generated using MD simulation and SAXS data.

A Site Density Functional Theory for Water: Application to Solvation of Amino Acid Side Chains.

We report a site density functional theory (SDFT) based on the conventional atomistic models of water and the universality ansatz of the bridge functional. The excess Helmholtz energy functional is formulated in terms of a quadratic expansion with respect to the local density deviation from that of a uniform system and a universal functional for all higher-order terms approximated by that of a reference hard-sphere system. With the atomistic pair direct correlation functions of the uniform system calculated from MD simulation and an analytical expression for the bridge functional from the modified fundamental measure theory, the SDFT can be used to predict the structure and thermodynamic properties of water under inhomogeneous conditions with a computational cost negligible in comparison to that of brute-force simulations. The numerical performance of the SDFT has been demonstrated with the predictions of the solvation free energies of 15 molecular analogs of amino acid side chains in water represented by SPC/E, SPC, and TIP3P models. For theTIP3P model, a comparison of the theoretical predictions with MD simulation and experimental data shows agreement within 0.64 and 1.09 kcal/mol on average, respectively.

The Effect of Neutral Ion Aggregate Formation on the Electrical Conductivity of an Ionic Liquid and its Mixtures with Chloroform

We have studied the electrical conductivity of mixtures of the ionic liquid (IL) [C(2)mim][NTf(2)] with chloroform for various concentrations at a temperature of 303 K by both experiments and MD simulations. The molar conductivities exhibit an initial increase with decreasing IL concentration, which again sharply decreases in the region of highly diluted IL concentrations. This behavior is according to a competition between 1) the decreasing viscosity with increasing chloroform concentration due to fluidizing the mixture, and 2) the decreasing amount of charge carriers due to the formation of neutral IL aggregates. The simulated molar conductivities and their concentration dependence are found to be in reasonable agreement with the experimentally determined values, suggesting that essential correlations between the solvated ionic species in the solution are well represented. Both experiments and simulations show that the Nernst-Einstein relation drastically overestimates the conductivity of the solutions at low IL concentrations. According to our MD simulation data, this observation is due to the extensive formation of neutral aggregates in the low-concentration regime.

Molecular simulation and macroscopic modeling of the diffusion of hydrogen, carbon monoxide and water in heavy n-alkane mixtures

The self-diffusion coefficient of hydrogen (H(2)), carbon monoxide (CO) and water (H(2)O) in n-alkanes was studied by molecular dynamics simulation. Diffusion in a few pure n-alkanes (namely n-C(8), n-C(20), n-C(64) and n-C(96)) was examined. In addition, binary n-C(12)-n-C(96) mixtures with various compositions as well as more realistic five- and six-n-alkane component mixtures were simulated. In all cases, the TraPPE united atom force field was used for the n-alkane molecules. The force field for the mixture of n-alkanes was initially validated against experimental density values and was shown to be accurate. Moreover, macroscopic correlations for predicting diffusion coefficient of H(2), CO and H(2)O in n-alkanes and mixtures of n-alkanes were developed. The functional form of the correlation was based on the rough hard sphere theory (RHS). The correlation was applied to simulation data and an absolute average deviation (AAD) of 5.8% for pure n-alkanes and 3.4% for n-alkane mixtures was obtained. Correlation parameters vary in a systematic way with carbon number and so they can be used to provide predictions in the absence of any experimental or molecular simulation data. Finally, in order to reduce the number of adjustable parameters, for the n-alkane mixtures the "pseudo-carbon number" approach was used. This approach resulted in relatively higher deviation from MD simulation data (AAD of 18.2%); however, it provides a convenient and fast method to predict diffusion coefficients. The correlations developed here are expected to be useful for engineering calculations related to the design of the Gas-to-Liquid process.

Quantum clustering and network analysis of MD simulation trajectories to probe the conformational ensembles of protein–ligand interactions

In this article, we present a novel application of a quantum clustering (QC) technique to objectively cluster the conformations, sampled by molecular dynamics simulations performed on different ligand bound structures of the protein. We further portray each conformational population in terms of dynamically stable network parameters which beautifully capture the ligand induced variations in the ensemble in atomistic detail. The conformational populations thus identified by the QC method and verified by network parameters are evaluated for different ligand bound states of the protein pyrrolysyl-tRNA synthetase (DhPylRS) from D. hafniense. The ligand/environment induced re-distribution of protein conformational ensembles forms the basis for understanding several important biological phenomena such as allostery and enzyme catalysis. The atomistic level characterization of each population in the conformational ensemble in terms of the re-orchestrated networks of amino acids is a challenging problem, especially when the changes are minimal at the backbone level. Here we demonstrate that the QC method is sensitive to such subtle changes and is able to cluster MD snapshots which are similar at the side-chain interaction level. Although we have applied these methods on simulation trajectories of a modest time scale (20 ns each), we emphasize that our methodology provides a general approach towards an objective clustering of large-scale MD simulation data and may be applied to probe multistate equilibria at higher time scales, and to problems related to protein folding for any protein or protein-protein/RNA/DNA complex of interest with a known structure.

Size of the Dynamic Bead in Polymers

Presented analysis of neutron, mechanical, and MD simulation data available in the literature demonstrates that the dynamic bead size (the smallest subchain that still exhibits the Rouse-like dynamics) in most of the polymers is significantly larger than the traditionally defined Kuhn segment. Moreover, our analysis emphasizes that even the static bead size (e.g., chain statistics) disagrees with the Kuhn segment length. We demonstrate that the deficiency of the Kuhn segment definition is based on the assumption of a chain being completely extended inside a single bead. The analysis suggests that representation of a real polymer chain by the bead-and-spring model with a single parameter C∞ cannot be correct. One needs more parameters to reflect correctly details of the chain structure in the bead-and-spring model.

Structural Quality and MD Simulations of Homology Models of Major Facilitator Superfamily (MFS) Transporter Proteins

Title: Structural Quality and MD Simulations of Homology Models of Major Facilitator Superfamily (MFS) Transporter ProteinsThe MFS is a large group of transporter proteins that catalyze uniport; solute:cation symport; and/or solute:cation or solute:solute antiport. Currently, relatively few experimental structures are known for these membrane transport proteins. Homology modeling enables us to extrapolate from X-ray structures of (bacterial) MFS proteins. However, quality assessment of homology models is crucial, especially given the relatively low resolution of some of the template structures. Furthermore, although structurally conserved, the MFS transporters do not share high sequence identity (i.e. less than 30% identity is observed). In this work, homology models of MFS were constructed based on crystal structures of lactose permease (LacY, PDB ID 1PV7), GlpT (PDB ID 1PW4) and the multidrug transporter EmrD (PDB ID 2GFP). These MFS model structures, as well as MFS crystal structures, were subjected to atomistic and coarse grained (CG) MD simulations. Quality assessments and measures of conformational “stability” of the simulated structures were compared for homology models and X-ray structures. It was concluded that local stereochemical quality of MFS protein structures does not predict conformational “stability” of the protein in MD simulations. An extensive stereochemical analysis based on known membrane protein crystal structures and corresponding MD simulation data was used to assess the MFS model structures. It was shown that MD stereochemical profiles can be used to distinguish between models constructed based on high quality template structures, models constructed based on a low quality template structures, and ‘decoys’(i.e. models constructed based on a incorrectly chosen templates).

Jan van Es

The mitochondrial Tim chaperones are responsible for the transport of membrane proteins across the inter-membrane space to the inner and outer mitochondrial membranes. TIM9·10, a hexameric 70 kDa protein complex...

Ion spatial distributions at the liquid–vapor interface of aqueous potassium fluoride solutions

X-Ray photoemission spectroscopy operating under ambient pressure conditions is used to probe ion distributions throughout the interfacial region of a free-flowing aqueous liquid micro-jet of 6 M potassium fluoride. Varying the energy of the ejected photoelectrons by carrying out experiments as a function of X-ray wavelength measures the composition of the aqueous-vapor interfacial region at various depths. The F(-) to K(+) atomic ratio is equal to unity throughout the interfacial region to a depth of 2 nm. The experimental ion profiles are compared with the results of a classical molecular dynamics simulation of a 6 M aqueous KF solution employing polarizable potentials. The experimental results are in qualitative agreement with the simulations when integrated over an exponentially decaying probe depth characteristic of an APPES experiment. First principles molecular dynamics simulations have been used to calculate the potential of mean force for moving a fluoride anion across the air-water interface. The results show that the fluoride anion is repelled from the interface, consistent with the depletion of F(-) at the interface revealed by the APPES experiment and polarizable force field-based molecular dynamics simulation. Together, the APPES and MD simulation data provide a detailed description of the aqueous-vapor interface of alkali fluoride systems. This work offers the first direct observation of the ion distribution at an aqueous potassium fluoride solution interface. The current experimental results are compared to those previously obtained for saturated solutions of KBr and KI to underscore the strong difference in surface propensity between soft/large and hard/small halide ions in aqueous solution.

The improvement of the algorithm for order parameter calculation ( S2) from molecular dynamics simulation using the correlation motion function

The generalized order parameter, S 2, calculated from MD simulation trajectory using time-dependent internal Correlation Motion Function (CMF) agrees well with NMR derived S 2 processed with the extended model-free analysis approach. However, the former lies considerably lower comparing to simple model-free derived data from NMR experiments. In the present study we analyze possible reasons of such disagreement. In the general case we propose to use preexponential factors from expression for internal CMF rather than ordinary S 2 values. Particularly, in case of the simple model-free S 2 experimental values we suggest comparing them with S eff 2 = 1 + S 2 − S f 2 computed from MD simulation data. We show that the S eff 2 values are in a good agreement with NMR derived S 2 values obtained using the simple model-free analysis.

Completed Three-Dimensional Model of Human Coagulation Factor Va. Molecular Dynamics Simulations and Structural Analyses

Factor Va is the critical cofactor for prothrombinase assembly required for timely and efficient prothrombin activation. In the absence of a complete crystal structure for the cofactor, Pellequer et al. [(2000) Thromb. Haemostasis 84, 849-857] proposed an incomplete homology model of factor Va (it lacks 46 amino acids from the carboxyl terminus of the heavy chain), which is a static model in a vacuum. A recently published X-ray structure of activated protein C (APC) inactivated bovine factor Va(i) (without the A2 domain) suggests a completely new arrangement of the C1 and C2 domains as compared with the previously published structure of the recombinant C1 and C2 domains. Our aims were (a) to exchange the C1 and C2 domains of the homology model with the modified bovine C1 and C2 domains using the X-ray structure as a template, (b) to determine by computation the three-dimensional model for the carboxyl-terminal peptide of the factor Va heavy chain (Ser(664)-Arg(709)) and incorporate it into the incomplete model, (c) to obtain a complete model of the cofactor folded in solution that might account for its physiological functions and interactions with other components of prothrombinase, and (d) to use the model in order to understand the mechanism of factor Va inactivation by APC. In the first step a sequence alignment of the human and bovine C1 and C2 domains was performed followed by amino acid changes in the three-dimensional structure where the sequences were not identical. The new model of the C1 and C2 domains was then attached to the homology model. The analysis of the MD simulation data revealed that several domains of the cofactor were significantly displaced during simulation. Using our completed model of human factor Va, we are also demonstrating for the first time that cleavage of membrane-bound normal factor Va as well as membrane-bound factor V(LEIDEN) by APC at Arg(306) is required for the dissociation of the A2 domain from the rest of the molecule. Thus, differences in the inactivation rates of the two cofactor molecules are due to differences in the rate of cleavage at Arg(306). The data demonstrate that our model represents the foundation for the establishment of a complete prothrombinase complex model, which might be successful in describing accurately the ternary protein-protein interaction and thus accounts for experimental observations.

Equations of state for fluids containing alternating heteronuclear chain molecules

Statistical associating fluid theory-dimer (SAFT-D) and monomer–dimer mixture theory equations of state of linear homonuclear chain are extended to fluids containing alternating heteronuclear chain molecules separately. The proposed models account for the appropriate site–site correlation functions at contact. The modified equations of state show a good agreement with generalized Flory dimer theory and MD simulation data for small and medium size ratio of hard sphere diameters.

Statistical fluid theory for associating fluids containing alternating heteronuclear chain molecules

Statistical associating fluid theory of homonuclear dimerized chain fluids and homonuclear monomer-dimer mixture chain fluids are extended to fluids containing alternating heteronuclear chain molecules separately. The proposed models account for the appropriate site-site correlation functions at contact. The modified equations of state show a good agreement with generalized Flory dimer theory and MD simulation data for small and medium size ratio of hard sphere diameters.

Liquid water (D 2O): a dynamic model emerging from near-infrared DO-D stretching overtone studies

The extended dynamic model of liquid water which considers five different hydrogen-bonded states (including three dominant states) and two major positions of the water molecule in equilibrium, is presented based mostly onthe analysis of near-infrared overtone spectra for the O-D stretching vibration of pure liquid D 2 O, obtained and evaluated within the temperature range of 278 - 358 K (5 - 85 °C). In agreement with the earlier data of Luck and Ditter (1969), and Angell and Rodgers (1984), it was found that the integrated intensity (but not the shape) of the specified (composite) overtone band, situated within the spectral range of 4250 - 5750 cm - 1 , is almost independent of temperature, and displays a quasi-isosbestic point at ca. 5100 cm - 1 . Moreover, the density-corrected difference spectra taken with intervals of 10, 20 and 40 K clearly indicate that the temperature effect is determined by the intensity redistribution between two major components (sub-bands) of the total of five forming the composite overtone contour. These two sub-bands according to their center-of-mass positions, and on the basis of published structural (Bosio et al., 1983), static spectroscopic (Berglund et al., 1978, Walrafen et al., 1986, 1996), and MD simulation (Scortino et al., 1990-1992, Sato and Hirata, 1999) data, can be attributed to the normal and temperature-distorted (T-bifurcated, but not totally broken) hydrogen-bonded states of O-D oscillators, respectively. In addition, the global Gaussian-contour analysis of obtained composite spectra clearly revealed the existence of O-D oscillator states corresponding to the fraction of presumably pressure-responsive (P-bifurcated) hydrogen bonds promoted by formation of additional rigid five-coordinated water configurations (furnished also by forming of quasi-linear O-O-O trimers within the hexagonal Ih ice-like lattice). The integrated intensity due to this fraction of hydrogen-bonded states does not display any change with temperature, over the whole range of existence of liquid water, as indicated earlier by Luck, Angell, and their co-workers. The whole variety of structural, static spectroscopic, and MD simulation data, together with the results of dielectric relaxation temperature dependence studies (Ronne et al., 1997, 1999, Buchner et al., 1999) can be readily explained in the framework of dynamic displacement model first formulated by Agmon (1996). This model implies the temperature-dependent flip-flop equilibrium between two major positions for a given water molecule (corresponding to the tetrahedral four-neighbor and the interstitial five-neighbor configurations), the exchange between which is accomplished by the breakage and formation of a net normal and a net T-bifurcated (and vice versa) hydrogen bond, respectively. The extended van't Hoff analysis indicate that the transition between two states of water at ca. 313 K (40 °C) takes place with almost zero equilibrium free energy, but significant counterbalanced enthalpy and entropy contributions (3.3 kcal mol - 1 and 10.5 cal mol - 1 K - 1 , respectively). In contrast, the effect of high pressure, according to the published data, can be viewed as leading to the displacements thanks to the highly-cooperative process of formation of P-bifurcated hydrogen bonds only, without altering the quasi-equilibrium between normal and T-bifurcated fractions.

Role of Competitive PEO−Water and Water−Water Hydrogen Bonding in Aqueous Solution PEO Behavior

The behavior of aqueous solutions of poly(ethylene oxide) (PEO) is studied theoretically by applying a mean field-like approach which includes the effect of the competition of PEO and water as proton acceptors in hydrogen bond formation. Accounting for this effect is of crucial importance for a correct description of all solution properties. We calculate the temperature and concentration dependence of the average fraction of hydrogen bonds between PEO and water and find very good agreement between our predictions and experimental or MD simulation data. We also make predictions concerning the temperature behavior of the second virial coefficient A2 and the effective interaction parameter χeff and compare it with experimental data. We found that the decrease of A2 with temperature is caused by the delicate balance of the opposing effects of water−PEO and water−water hydrogen bonding. The phase diagram for PEO of different molecular weights in water is calculated using experimentally reported data for the en...