Calculations of Absolute Free Energies, Enthalpies, and Entropies for Drug Binding.

Computer simulation methods can aid in the rational design of drugs aimed at a specific target, typically a protein. The affinity of a drug for its target is given by the free energy of binding. Binding can be further characterized by the enthalpy and entropy changes in the process. Methods exist to determine exact free energies, enthalpies, and entropies that are dependent only on the quality of the potential model and adequate sampling of conformational degrees of freedom. Entropy and enthalpy are roughly an order of magnitude more difficult to calculate than the free energy. This project combines a replica exchange method for enhanced sampling, designed to be efficient for protein-sized systems, with free energy calculations. This approach, replica exchange with dynamical scaling (REDS), uses two conventional simulations at different temperatures so that the entropy can be found from the temperature dependence of the free energy. A third replica is placed between them, with a modified Hamiltonian that allows it to span the temperature range of the conventional replicas. REDS provides temperature-dependent data and aids in sampling. It is applied to the bromodomain-containing protein 4 (BRD4) system. We find that for the force fields used, the free energies are accurate but the entropies and enthalpies are not, with the entropic contribution being too positive. Reproducing the entropy and enthalpy of binding appears to be a more stringent test of the force fields than reproducing the free energy.

Effects of polarizability and charge transfer on water dynamics and the underlying activation energies.

A large number of force fields have been proposed for describing the behavior of liquid water within classical atomistic simulations, particularly molecular dynamics. In the past two decades, models that incorporate molecular polarizability and even charge transfer have become more prevalent, in attempts to develop more accurate descriptions. These are frequently parameterized to reproduce the measured thermodynamics, phase behavior, and structure of water. On the other hand, the dynamics of water is rarely considered in the construction of these models, despite its importance in their ultimate applications. In this paper, we explore the structure and dynamics of polarizable and charge-transfer water models, with a focus on timescales that directly or indirectly relate to hydrogen bond (H-bond) making and breaking. Moreover, we use the recently developed fluctuation theory for dynamics to determine the temperature dependence of these properties to shed light on the driving forces. This approach provides key insight into the timescale activation energies through a rigorous decomposition into contributions from the different interactions, including polarization and charge transfer. The results show that charge transfer effects have a negligible effect on the activation energies. Furthermore, the same tension between electrostatic and van der Waals interactions that is found in fixed-charge water models also governs the behavior of polarizable models. The models are found to involve significant energy-entropy compensation, pointing to the importance of developing water models that accurately describe the temperature dependence of water structure and dynamics.

The pH Response of a Peptoid Oligomer.

Polypeptoids are N-substituted glycine polymers, which differ from peptides in the placement of the side chain on the amide nitrogen rather than the Cα carbon. A peptoid with a chiral side chain containing both an aromatic group and carboxylic acid has a structure that responds to pH changes. All-atom molecular dynamics simulations using a force field specifically tuned for peptoids were carried out with an advanced sampling method for the peptoid (S)-N-(1-carboxy-2-phenylethyl)glycine in the high and low pH limits. The simulations show that the structure changes from mostly cis amide bonds at low pH to mostly trans bonds at high pH. The structural changes are driven by side chain-backbone hydrogen bonds at low pH and side chain repulsions and increased water contact at high pH.

Unraveling the Role of Charge Patterning in the Micellar Structure of Sequence-Defined Amphiphilic Peptoid Oligomers by Molecular Dynamics Simulations.

Electrostatic interactions play a significant role in regulating biological systems and have received increasing attention due to their usefulness in designing advanced stimulus-responsive materials. Polypeptoids are highly tunable N-substituted peptidomimetic polymers that lack backbone hydrogen bonding and chirality. Therefore, polypeptoids are suitable systems to study the effect of noncovalent interactions of substituents without complications of backbone intramolecular and intermolecular hydrogen bonding. In this study, all-atom molecular dynamics (MD) simulations were performed on micelles formed by a series of sequence-defined ionic polypeptoid block copolymers consisting of a hydrophobic segment and a hydrophilic segment in an aqueous solution. By combining the results from MD simulations and experimental small-angle neutron scattering data, further insights were gained into the internal structure of the formed polypeptoid micelles, which is not always directly accessible from experiments. In addition, information was gained into the physics of the noncovalent interactions responsible for the self-assembly of weakly charged polypeptoids in an aqueous solution. While the aggregation number is governed by electrostatic repulsion of the negatively charged carboxylate (COO-) substituents on the polypeptoid chain within the micelle, MD simulations indicate that the position of the charge on singly charged chains mediates the shape of the micelle through the charge-dipole interactions between the COO- substituent and the surrounding water. Therefore, the polypeptoid micelles formed from the single-charged series offer the possibility for tailorable micelle shapes. In contrast, the polypeptoid micelles formed from the triple-charged series are characterized by more pronounced electrostatic repulsion that competes with more significant charge-sodium interactions, making it difficult to predict the shape of the micelles. This work has helped further develop design principles for the shape and structure of self-assembled micelles by controlling the position of charged moieties on the backbone of polypeptoid block copolymers.

Simulation studies of polypeptoids using replica exchange with dynamical scaling and dihedral biasing.

Polypeptoids differ from polypeptides in that the amide bond can more frequently adopt both cis and trans conformations. The transition between the two conformations requires overcoming a large energy barrier, making it difficult for conventional molecular simulations to adequately visit the cis and trans structures. A replica-exchange method is presented that allows for easy rotations of the amide bond and also an efficient linking to a high temperature replica. The method allows for just three replicas (one at the temperature and Hamiltonian of interest, a second high temperature replica with a biased dihedral potential, and a third connecting them) to overcome the amide bond sampling problem and also enhance sampling for other coordinates. The results indicate that for short peptoid oligomers, the conformations can range from all cis to all trans with an average cis/trans ratio that depends on side chain and potential model.

Molecular dynamics simulations of the flexibility and inhibition of SARS-CoV-2 NSP 13 helicase.

The helicase protein of the Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) is both a good potential drug target and very flexible. The flexibility, and therefore its function, could be reduced through knowledge of these motions and identification of allosteric pockets. Using molecular dynamics simulations with enhanced sampling, we determined key modes of motion and sites on the protein that are at the interface between flexible domains of the proteins. We developed an approach to map the principal components of motion onto the surface of a potential binding pocket to help in the identification of allosteric sites.

Molecular dynamics simulations of allosteric motions and competitive inhibition of the Zika virus helicase.

The 2015 Zika outbreak sparked major global concern and emphasized the reality and dangers still posed by mosquito borne pathogens. While efforts have been made to develop a vaccine and other therapeutics, there is still a great demand for antiviral drugs targeting Zika and other flaviviruses. The non-structural protein 3 (NS3) helicase is a vital component of the viral replication complex, tasked with unwinding the viral dsRNA molecule into single strands. Given this critical function, the Zika virus helicase is a potential therapeutic target and the focus of many ongoing research efforts. Using a combination of drug docking and molecular dynamics simulations, we have identified a list of competitive helicase inhibitors targeting the ATP hydrolysis site and have discovered a potential allosteric site capable of distorting both of the protein's active sites.

Channelrhodopsin C1C2: Photocycle kinetics and interactions near the central gate.

Channelrhodopsins (ChR) are light-sensitive cation channels used in optogenetics, a technique that applies light to control cells (e.g., neurons) that have been modified genetically to express those channels. Although mutations are known to affect pore kinetics, little is known about how mutations induce changes at the molecular scale. To address this issue, we first measured channel opening and closing rates of a ChR chimera (C1C2) and selected variants (N297D, N297V, and V125L). Then, we used atomistic simulations to correlate those rates with changes in pore structure, hydration, and chemical interactions among key gating residues of C1C2 in both closed and open states. Overall, the experimental results show that C1C2 and its mutants do not behave like ChR2 or its analogous variants, except V125L, making C1C2 a unique channel. Our atomistic simulations confirmed that opening of the channel and initial hydration of the gating regions between helices I, II, III, and VII of the channel occurs with 1) the presence of 13-cis retinal; 2) deprotonation of a glutamic acid gating residue, E129; and 3) subsequent weakening of the central gate hydrogen bond between the same glutamic acid E129 and asparagine N297 in the central region of the pore. Also, an aspartate (D292) is the unambiguous primary proton acceptor for the retinal Schiff base in the hydrated channel.

Insights into the Thermal Response of a Poly(ethylene oxide)-poly(propylene oxide)-poly(ethylene oxide) Triblock Polymer in Water.

A thermal responsive block copolymer made up of ethylene oxide (EO) and propylene oxide (PO) blocks was simulated with optimized atomistic potentials and enhanced sampling methods over a range of temperatures. The results for the L42 pluronic polymer (EO)4(PO)22(EO)4, which is known to undergo a transition in this temperature range, and the similarly sized (EO)30 polymer, which does not, are compared. The thermal responsive L42 polymers in a dilute solution tend to aggregate, and this tendency gets stronger as temperature increases. The poly(ethylene oxide) polymer shows no such tendency. The aggregation is stabilized by the hydrophobic contact of the propylene oxide methyl groups, which outweighs a small loss in hydrogen bonds between the ether oxygens and water.

An Implementation of Replica Exchange with Dynamical Scaling for Efficient Large-Scale Simulations.

An implementation of the replica exchange with dynamical scaling (REDS) method in the commonly used molecular dynamics program GROMACS is presented. REDS is a replica exchange method that requires fewer replicas than conventional replica exchange while still providing data over a range of temperatures and can be used in either constant volume or constant pressure ensembles. Details for running REDS simulations are given, and an application to the human islet amyloid polypeptide (hIAPP) 11-25 fragment shows that the model efficiently samples conformational space.

Analysis of Atomistic Potentials for Poly(ethylene glycol) Ethers.

Two different potentials, the modified TraPPE-UA model of Fischer, J. [ J. Phys. Chem. B 2008, 112, 2388-2398] and the modified general AMBER force-field (GAFF) model of Barbosa, N. S. V. [ J. Mol. Model. 2017, 23, 194]are tested for a variety of temperature-dependent properties for neat and aqueous solutions for small poly(oxyethylene) (PEO) oligomers and larger polymers. A set of charges for PEO of arbitrary size is adapted for the modified GAFF model. Both models accurately reproduce experimental properties, but we find that the modified TraPPE-UA model is more accurate.

Polarizable MD simulations of ionic liquids: How does additional charge transfer change the dynamics?

Both experimental and computational evidence exist that Coulomb interactions between the molecular ions in ionic liquids are significantly damped by almost a factor of two. This circumstance is often used to justify charge scaling. However, as polarizable MD simulations are also capable of explaining the reduced Coulomb interaction between the ionic liquid ions [C. Schröder, Phys. Chem. Chem. Phys., 2012, 14, 3089], the question arises, if the reduced Coulomb interactions are due to a charge transfer between the molecules or due to an overall effect of induced dipolar interactions. We aim to contribute to this discussion using polarizable MD simulations of 1-ethyl-3-methylimidazolium tetrafluoroborate including a new model for treating charge transfer between the cations and anions. The diffusion time scales are not changed significantly with the inclusion of charge transfer, but individual ions show a strong dependence on charge transfer amounts. Ions which have transferred more charge, and have a charge with a smaller magnitude, diffuse slower. The charge transfer model shows a slightly larger conductivity, despite having smaller charges, and shows a much stronger contribution of the anions to the conductivity. With charge transfer, the anions become the dominant species for charge transport, while the polarizable models show a roughly equal contribution from the anions and the cations.

Coarse-Grained Models for Constant pH Simulations of Carboxylic Acids.

A model for carboxylic acids, in both the protonated and deprotonated states, is developed in which hydrogen interaction sites are not used and all interactions are short-ranged. A method for constant pH simulations, which exploits these features of the model, is developed. The constant pH method samples protonation states by making discrete Monte Carlo steps and is able to efficiently move between states in two steps. The method is applied to the polymer poly(methacrylic acid), a pH-responsive polymer that undergoes structural changes as a function of pH. The model is able to reproduce the structural changes induced by pH.

Water hydrogen degrees of freedom and the hydrophobic effect.

Hydrogen bonds are the key interaction that establishes the liquid and solvent properties of water. Nevertheless, it is possible to construct an accurate molecular model of water which does not include hydrogens or any orientational interactions. Using this model, we calculate the structural and thermodynamic properties for the hydration of methane and ethane. The addition of the hydrophobic solute leads to changes in structure, as can be seen in slightly enhanced tetrahedral geometries and slightly reduced Voronoi volumes of water near the solute. The entropy of hydration from the model is about half the experimental value, suggesting that what is left out of the model-the orientational or hydrogen response-contributes to about half the entropy. For the hydrophobic association of two methane molecules in water, the hydrogen degrees of freedom do not seem to play an important role and the entropy of association is similar to all-atom models.

Towards a coarse-grained model of the peptoid backbone: the case of N,N-dimethylacetamide.

In this study, a coarse-grained (CG) model for N,N-dimethylacetamide (DMA), which represents the polypeptoid backbone, is developed as a step towards establishing a CG model of the complex polypeptoid system. Polypeptoids or poly N-substituted glycines are a type of peptidomimetic polymers that are highly tunable, and hence an ideal model system to study self-assembly as a function of chemical groups in aqueous soft matter systems. The DMA CG model is parameterized to reproduce the structural properties of DMA liquid as well as a dilute aqueous solution of DMA using a reference all atom model, namely the OPLS-AA force-field. The intermolecular forces are represented by the Stillinger-Weber potential, that consists of both two- and three-body terms that are very short-ranged. The model is validated on thermodynamic properties of liquid and aqueous DMA, as well as the vapor-liquid interface of liquid DMA and the structure of a concentrated aqueous solution of DMA in water as well as a simple peptoid in water. Without long-ranged interactions and the absence of interaction sites on hydrogen atoms, the CG DMA model is an order of magnitude faster than the higher resolution all-atom (AA) model.

The influence of polarizability and charge transfer on specific ion effects in the dynamics of aqueous salt solutions.

The diffusion rates for water molecules in salt solutions depend on the identity of the ions, as well as their concentration. Among the alkali metal ions, cesium and potassium increase and sodium strongly decreases the diffusion constant of water. The origin of the difference can be understood by examining the simulation results using different potential models. In this work, aqueous solutions of salts are simulated with a variety of models. Commonly used non-polarizable models, which otherwise reproduce many experimental properties, do not capture the trend in the diffusion constant, while models which include polarization and/or charge transfer interactions do. For the non-polarizable models, the diffusion constant decreases too strongly with salt concentration. The changes in the water diffusion constant with increasing salt concentration match the diffusion constant of the ion. The ion diffusion constant is dependent on the residence time for water in the ion solvation shell. The non-polarizable models over-estimate the residence time, relative to the translational diffusion constant and so tend to under-estimate the ion and water diffusion constants.

Coarse-Grained Simulations of Aqueous Thermoresponsive Polyethers.

Thermoresponsive polymers can change structure or solubility as a function of temperature. Block co-polymers of polyethers have a response that depends on polymer molecular weight and co-polymer composition. A coarse-grained model for aqueous polyethers is developed and applied to polyethylene oxide and polyethylene oxide-polypropylene oxide-polyethylene oxide triblock co-polymers. In this model, no interaction sites on hydrogen atoms are included, no Coulombic interactions are present, and all interactions are short-ranged, treated with a combination of two- and three-body terms. Our simulations find that The triblock co-polymers tend to associate at temperatures above 350 K. The aggregation is stabilized by contact between The hydrophobic methyl groups on The propylene oxide monomers and involves a large, favorable change in entropy.

Keystrokes, Mouse Clicks, and Gazing at the Computer: How Physician Interaction with the EHR Affects Patient Participation.

BACKGROUND: Evidence is mixed regarding how physicians' use of the electronic health record (EHR) affects communication in medical encounters. OBJECTIVE: To investigate whether the different ways physicians interact with the computer (mouse clicks, key strokes, and gaze) vary in their effects on patient participation in the consultation, physicians' efforts to facilitate patient involvement, and silence. DESIGN: Cross-sectional, observational study of video and event recordings of primary care and specialty consultations. PARTICIPANTS: Thirty-two physicians and 217 patients. MAIN MEASURES: Predictor variables included measures of physician interaction with the EHR (mouse clicks, key strokes, gaze). Outcome measures included active patient participation (asking questions, stating preferences, expressing concerns), physician facilitation of patient involvement (partnership-building and supportive talk), and silence. KEY RESULTS: Patients were less active participants in consultations in which physicians engaged in more keyboard activity (b = -0.002, SE = 0.001, p = 0.02). More physician gaze at the computer was associated with more silence in the encounter (b = 0.21, SE = 0.09, p = 0.02). Physicians' facilitative communication, which predicted more active patient participation (b = 0.65, SE = 0.14, p < 0.001), was not related to EHR activity measures. CONCLUSIONS: Patients may be more reluctant to actively participate in medical encounters when physicians are more physically engaged with the computer (e.g., keyboard activity) than when their behavior is less demonstrative (e.g., gazing at EHR). Using easy to deploy communication tactics (e.g., asking about a patient's thoughts and concerns, social conversation) while working on the computer can help physicians engage patients as well as maintain conversational flow.

Coarse-Grained Models of Aqueous and Pure Liquid Alkanes.

A model for linear alkanes is presented in which interaction sites are only on the carbon atoms, and the range of the potential is reduced using the Stillinger-Weber potential. The model is optimized for aqueous and liquid alkane properties and can match thermodynamic and structural properties, including solvation free energies, liquid densities, and liquid/vapor and liquid/water surface tensions for alkanes over a range of lengths. The results for long alkanes indicate that such models can be useful as accurate, yet efficient, coarse-grained potentials for macromolecules in water and other environments.

Aggregation of cyclic polypeptoids bearing zwitterionic end-groups with attractive dipole-dipole and solvophobic interactions: a study by small-angle neutron scattering and molecular dynamics simulation.

Aggregation behavior of cyclic polypeptoids bearing zwitterionic end-groups in methanol has been studied using a combination of experimental and simulation techniques. The data from SANS and cryo-TEM indicate that the solution contains small clusters of these cyclic polypeptoids, ranging from a single polypeptoid chain to small oligomers, while the linear counterpart shows no cluster formation. Atomistic molecular dynamics simulations reveal that the driving force for this clustering behavior is due to the interplay between the effective repulsion due to the solvation of the dipoles formed by the charged end-groups in each polypeptoid chain and the attractive forces due to dipole-dipole interactions and the solvophobic effect.