Scrutinizing molecular mechanics force fields on the submicrosecond timescale with NMR data

Caught between a ROCK and a hard place: current challenges in structure-based drug design

The discipline of structure-based drug design (SBDD) is several decades old and it is tempting to think that the proliferation of experimental structures for many drug targets might make computer-aided drug design (CADD) straightforward. However, this is far from true. In this review, we illustrate some of the challenges that CADD scientists face every day in their work, even now. We use Rho-associated protein kinase (ROCK), and public domain structures and data, as an example to illustrate some of the challenges we have experienced during our project targeting this protein. We hope that this will help to prevent unrealistic expectations of what CADD can accomplish and to educate non-CADD scientists regarding the challenges still facing their CADD colleagues.

Advanced computational approaches to understand protein aggregation

Protein aggregation is a widespread phenomenon implicated in debilitating diseases like Alzheimer's, Parkinson's, and cataracts, presenting complex hurdles for the field of molecular biology. In this review, we explore the evolving realm of computational methods and bioinformatics tools that have revolutionized our comprehension of protein aggregation. Beginning with a discussion of the multifaceted challenges associated with understanding this process and emphasizing the critical need for precise predictive tools, we highlight how computational techniques have become indispensable for understanding protein aggregation. We focus on molecular simulations, notably molecular dynamics (MD) simulations, spanning from atomistic to coarse-grained levels, which have emerged as pivotal tools in unraveling the complex dynamics governing protein aggregation in diseases such as cataracts, Alzheimer's, and Parkinson's. MD simulations provide microscopic insights into protein interactions and the subtleties of aggregation pathways, with advanced techniques like replica exchange molecular dynamics, Metadynamics (MetaD), and umbrella sampling enhancing our understanding by probing intricate energy landscapes and transition states. We delve into specific applications of MD simulations, elucidating the chaperone mechanism underlying cataract formation using Markov state modeling and the intricate pathways and interactions driving the toxic aggregate formation in Alzheimer's and Parkinson's disease. Transitioning we highlight how computational techniques, including bioinformatics, sequence analysis, structural data, machine learning algorithms, and artificial intelligence have become indispensable for predicting protein aggregation propensity and locating aggregation-prone regions within protein sequences. Throughout our exploration, we underscore the symbiotic relationship between computational approaches and empirical data, which has paved the way for potential therapeutic strategies against protein aggregation-related diseases. In conclusion, this review offers a comprehensive overview of advanced computational methodologies and bioinformatics tools that have catalyzed breakthroughs in unraveling the molecular basis of protein aggregation, with significant implications for clinical interventions, standing at the intersection of computational biology and experimental research.

On the Validation of Protein Force Fields Based on Structural Criteria

Molecular dynamics simulations depend critically on the quality of the force field used to describe the interatomic interactions and the extent to which it has been validated for use in a specific application. Using a curated test set of 52 high-resolution structures, 39 derived from X-ray diffraction and 13 solved using NMR, we consider the extent to which different parameter sets of the GROMOS protein force field can be distinguished based on comparing a range of structural criteria, including the number of backbone hydrogen bonds, the number of native hydrogen bonds, polar and nonpolar solvent-accessible surface area, radius of gyration, the prevalence of secondary structure elements, J-coupling constants, nuclear Overhauser effect (NOE) intensities, positional root-mean-square deviations (RMSD), and the distribution of backbone ϕ and ψ dihedral angles. It is shown that while statistically significant differences between the average values of individual metrics could be detected, these were in general small. Furthermore, improvements in agreement in one metric were often offset by loss of agreement in another. The work establishes a framework and test set against which protein force fields can be validated. It also highlights the danger of inferring the relative quality of a given force field based on a small range of structural properties or small number of proteins.

Unraveling motion in proteins by combining NMR relaxometry and molecular dynamics simulations: A case study on ubiquitin

Nuclear magnetic resonance (NMR) relaxation experiments shine light onto the dynamics of molecular systems in the picosecond to millisecond timescales. As these methods cannot provide an atomically resolved view of the motion of atoms, functional groups, or domains giving rise to such signals, relaxation techniques have been combined with molecular dynamics (MD) simulations to obtain mechanistic descriptions and gain insights into the functional role of side chain or domain motion. In this work, we present a comparison of five computational methods that permit the joint analysis of MD simulations and NMR relaxation experiments. We discuss their relative strengths and areas of applicability and demonstrate how they may be utilized to interpret the dynamics in MD simulations with the small protein ubiquitin as a test system. We focus on the aliphatic side chains given the rigidity of the backbone of this protein. We find encouraging agreement between experiment, Markov state models built in the χ1/χ2 rotamer space of isoleucine residues, explicit rotamer jump models, and a decomposition of the motion using ROMANCE. These methods allow us to ascribe the dynamics to specific rotamer jumps. Simulations with eight different combinations of force field and water model highlight how the different metrics may be employed to pinpoint force field deficiencies. Furthermore, the presented comparison offers a perspective on the utility of NMR relaxation to serve as validation data for the prediction of kinetics by state-of-the-art biomolecular force fields.

Accelerated Molecular Dynamics for Peptide Folding: Benchmarking Different Combinations of Force Fields and Explicit Solvent Models

Accelerated molecular dynamics (aMD) protocols were assessed on predicting the secondary structure of eight peptides, of which two are helical, three are β-hairpins, and three are disordered. Protocols consisted of combinations of three force fields (ff99SB, ff14SB, ff19SB) and two explicit solvation models (TIP3P and OPC), and were evaluated in two independent aMD simulations, one starting from an extended conformation, the other starting from a misfolded conformation. The results of these analyses indicate that all three combinations performed well on helical peptides. As for β-hairpins, ff19SB performed well with both solvation methods, with a slight preference for the TIP3P solvation model, even though performance was dependent on both peptide sequence and initial conformation. The ff19SB/OPC combination had the best performance on intrinsically disordered peptides. In general, ff14SB/TIP3P suffered the strongest helical bias.

An Imbalance in the Force: The Need for Standardized Benchmarks for Molecular Simulation

Force fields (FFs) for molecular simulation have been under development for more than half a century. As with any predictive model, rigorous testing and comparisons of models critically depends on the availability of standardized data sets and benchmarks. While such benchmarks are rather common in the fields of quantum chemistry, this is not the case for empirical FFs. That is, few benchmarks are reused to evaluate FFs, and development teams rather use their own training and test sets. Here we present an overview of currently available tests and benchmarks for computational chemistry, focusing on organic compounds, including halogens and common ions, as FFs for these are the most common ones. We argue that many of the benchmark data sets from quantum chemistry can in fact be reused for evaluating FFs, but new gas phase data is still needed for compounds containing phosphorus and sulfur in different valence states. In addition, more nonequilibrium interaction energies and forces, as well as molecular properties such as electrostatic potentials around compounds, would be beneficial. For the condensed phases there is a large body of experimental data available, and tools to utilize these data in an automated fashion are under development. If FF developers, as well as researchers in artificial intelligence, would adopt a number of these data sets, it would become easier to compare the relative strengths and weaknesses of different models and to, eventually, restore the balance in the force.

Collaborative Assessment of Molecular Geometries and Energies from the Open Force Field

Force fields form the basis for classical molecular simulations, and their accuracy is crucial for the quality of, for instance, protein-ligand binding simulations in drug discovery. The huge diversity of small-molecule chemistry makes it a challenge to build and parameterize a suitable force field. The Open Force Field Initiative is a combined industry and academic consortium developing a state-of-the-art small-molecule force field. In this report, industry members of the consortium worked together to objectively evaluate the performance of the force fields (referred to here as OpenFF) produced by the initiative on a combined public and proprietary dataset of 19,653 relevant molecules selected from their internal research and compound collections. This evaluation was important because it was completely blind; at most partners, none of the molecules or data were used in force field development or testing prior to this work. We compare the Open Force Field "Sage" version 2.0.0 and "Parsley" version 1.3.0 with GAFF-2.11-AM1BCC, OPLS4, and SMIRNOFF99Frosst. We analyzed force-field-optimized geometries and conformer energies compared to reference quantum mechanical data. We show that OPLS4 performs best, and the latest Open Force Field release shows a clear improvement compared to its predecessors. The performance of established force fields such as GAFF-2.11 was generally worse. While OpenFF researchers were involved in building the benchmarking infrastructure used in this work, benchmarking was done entirely in-house within industrial organizations and the resulting assessment is reported here. This work assesses the force field performance using separate benchmarking steps, external datasets, and involving external research groups. This effort may also be unique in terms of the number of different industrial partners involved, with 10 different companies participating in the benchmark efforts.

Integration of Experimental Data and Use of Automated Fitting Methods in Developing Protein Force Fields

The development of accurate protein force fields has been the cornerstone of molecular simulations for the past 50 years. During this period, many lessons have been learned regarding the use of experimental target data and parameter fitting procedures. Here, we review recent advances in protein force field development. We discuss the recent emergence of polarizable force fields and the role of electronic polarization and areas in which additive force fields fall short. The use of automated fitting methods and the inclusion of additional experimental solution data during parametrization is discussed as a means to highlight possible routes to improve the accuracy of force fields even further.

Molecular Dynamics Simulations of Transmembrane Cyclic Peptide Nanotubes Using Classical Force Fields, Hydrogen Mass Repartitioning, and Hydrogen Isotope Exchange Methods: A Critical Comparison

Self-assembled cyclic peptide nanotubes with alternating D- and L-amino acid residues in the sequence of each subunit have attracted a great deal of attention due to their potential for new nanotechnology and biomedical applications, mainly in the field of antimicrobial peptides. Molecular dynamics simulations can be used to characterize these systems with atomic resolution at different time scales, providing information that is difficult to obtain via wet lab experiments. However, the performance of classical force fields typically employed in the simulation of biomolecules has not yet been extensively tested with this kind of highly constrained peptide. Four different classical force fields (AMBER, CHARMM, OPLS, and GROMOS), using a nanotube formed by eight D,L-α-cyclic peptides inserted into a lipid bilayer as a model system, were employed here to fill this gap. Significant differences in the pseudo-cylindrical cavities formed by the nanotubes were observed, the most important being the diameter of the nanopores, the number and location of confined water molecules, and the density distribution of the solvent molecules. Furthermore, several modifications were performed on GROMOS54a7, aiming to explore acceleration strategies of the MD simulations. The hydrogen mass repartitioning (HMR) and hydrogen isotope exchange (HIE) methods were tested to slow down the fastest degrees of freedom. These approaches allowed a significant increase in the time step employed in the equation of the motion integration algorithm, from 2 fs up to 5–7 fs, with no serious changes in the structural and dynamical properties of the nanopores. Subtle differences with respect to the simulations with the unmodified force fields were observed in the concerted movements of the cyclic peptides, as well as in the lifetime of several H-bonds. All together, these results are expected to contribute to better understanding of the behavior of self-assembled cyclic peptide nanotubes, as well as to support the methods tested to speed up general MD simulations; additionally, they do provide a number of quantitative descriptors that are expected to be used as a reference to design new experiments intended to validate and complement computational studies of antimicrobial cyclic peptides.

Effect of the Force Field on Molecular Dynamics Simulations of the Multidrug Efflux Protein P-Glycoprotein

Molecular dynamics (MD) simulations have been used extensively to study P-glycoprotein (P-gp), a flexible multidrug transporter that is a key player in the development of multidrug resistance to chemotherapeutics. A substantial body of literature has grown from simulation studies that have employed various simulation conditions and parameters, including AMBER, CHARMM, OPLS, GROMOS, and coarse-grained force fields, drawing conclusions from simulations spanning hundreds of nanoseconds. Each force field is typically parametrized and validated on different data and observables, usually of small molecules and peptides; there have been few comparisons of force field performance on large protein-membrane systems. Here we compare the conformational ensembles of P-gp embedded in a POPC/cholesterol bilayer generated over 500 ns of replicate simulation with five force fields from popular biomolecular families: AMBER 99SB-ILDN, CHARMM 36, OPLS-AA/L, GROMOS 54A7, and MARTINI. We find considerable differences among the ensembles with little conformational overlap, although they correspond to similar extents to structural data obtained from electron paramagnetic resonance and cross-linking studies. Moreover, each trajectory was still sampling new conformations at a high rate after 500 ns of simulation, suggesting the need for more sampling. This work highlights the need to consider known limitations of the force field used (e.g., biases toward certain secondary structures) and the simulation itself (e.g., whether sufficient sampling has been achieved) when interpreting accumulated results of simulation studies of P-gp and other transport proteins.

The conformational and mutational landscape of the ubiquitin-like marker for autophagosome formation in cancer

Macroautophagy/autophagy is a cellular process to recycle damaged cellular components, and its modulation can be exploited for disease treatments. A key autophagy player is the ubiquitin-like protein MAP1LC3B/LC3B. Mutations and changes in MAP1LC3B expression occur in cancer samples. However, the investigation of the effects of these mutations on MAP1LC3B protein structure is still missing. Despite many LC3B structures that have been solved, a comprehensive study, including dynamics, has not yet been undertaken. To address this knowledge gap, we assessed nine physical models for biomolecular simulations for their capabilities to describe the structural ensemble of MAP1LC3B. With the resulting MAP1LC3B structural ensembles, we characterized the impact of 26 missense mutations from pan-cancer studies with different approaches, and we experimentally validated our prediction for six variants using cellular assays. Our findings shed light on damaging or neutral mutations in MAP1LC3B, providing an atlas of its modifications in cancer. In particular, P32Q mutation was found detrimental for protein stability with a propensity to aggregation. In a broader context, our framework can be applied to assess the pathogenicity of protein mutations or to prioritize variants for experimental studies, allowing to comprehensively account for different aspects that mutational events alter in terms of protein structure and function.Abbreviations: ATG: autophagy-related; Cα: alpha carbon; CG: coarse-grained; CHARMM: Chemistry at Harvard macromolecular mechanics; CONAN: contact analysis; FUNDC1: FUN14 domain containing 1; FYCO1: FYVE and coiled-coil domain containing 1; GABARAP: GABA type A receptor-associated protein; GROMACS: Groningen machine for chemical simulations; HP: hydrophobic pocket; LIR: LC3 interacting region; MAP1LC3B/LC3B microtubule associated protein 1 light chain 3 B; MD: molecular dynamics; OPTN: optineurin; OSF: open software foundation; PE: phosphatidylethanolamine, PLEKHM1: pleckstrin homology domain-containing family M 1; PSN: protein structure network; PTM: post-translational modification; SA: structural alphabet; SLiM: short linear motif; SQSTM1/p62: sequestosome 1; WT: wild-type.

Toward the solution of the protein structure prediction problem

Since Anfinsen demonstrated that the information encoded in a protein's amino acid sequence determines its structure in 1973, solving the protein structure prediction problem has been the Holy Grail of structural biology. The goal of protein structure prediction approaches is to utilize computational modeling to determine the spatial location of every atom in a protein molecule starting from only its amino acid sequence. Depending on whether homologous structures can be found in the Protein Data Bank (PDB), structure prediction methods have been historically categorized as template-based modeling (TBM) or template-free modeling (FM) approaches. Until recently, TBM has been the most reliable approach to predicting protein structures, and in the absence of reliable templates, the modeling accuracy sharply declines. Nevertheless, the results of the most recent community-wide assessment of protein structure prediction experiment (CASP14) have demonstrated that the protein structure prediction problem can be largely solved through the use of end-to-end deep machine learning techniques, where correct folds could be built for nearly all single-domain proteins without using the PDB templates. Critically, the model quality exhibited little correlation with the quality of available template structures, as well as the number of sequence homologs detected for a given target protein. Thus, the implementation of deep-learning techniques has essentially broken through the 50-year-old modeling border between TBM and FM approaches and has made the success of high-resolution structure prediction significantly less dependent on template availability in the PDB library.

Validating the CHARMM36m protein force field with LJ-PME reveals altered hydrogen bonding dynamics under elevated pressures

The pressure-temperature phase diagram is important to our understanding of the physics of biomolecules. Compared to studies on temperature effects, studies of the pressure dependence of protein dynamic are rather limited. Molecular dynamics (MD) simulations with fine-tuned force fields (FFs) offer a powerful tool to explore the influence of thermodynamic conditions on proteins. Here we evaluate the transferability of the CHARMM36m (C36m) protein force field at varied pressures compared with NMR data using ubiquitin as a model protein. The pressure dependences of J couplings for hydrogen bonds and order parameters for internal motion are in good agreement with experiment. We demonstrate that the C36m FF combined with the Lennard-Jones particle-mesh Ewald (LJ-PME) method is suitable for simulations in a wide range of temperature and pressure. As the ubiquitin remains stable up to 2500 bar, we identify the mobility and stability of different hydrogen bonds in response to pressure. Based on those results, C36m is expected to be applied to more proteins in the future to further investigate protein dynamics under elevated pressures.

NMR refinement and peptide folding using the GROMACS software

Nuclear magnetic resonance spectroscopy is used routinely for studying the three-dimensional structures and dynamics of proteins and nucleic acids. Structure determination is usually done by adding restraints based upon NMR data to a classical energy function and performing restrained molecular simulations. Here we report on the implementation of a script to extract NMR restraints from a NMR-STAR file and export it to the GROMACS software. With this package it is possible to model distance restraints, dihedral restraints and orientation restraints. The output from the script is validated by performing simulations with and without restraints, including the ab initio refinement of one peptide.

How to assess the structural dynamics of transcription factors by integrating sparse NMR and EPR constraints with molecular dynamics simulations

We review recent advances in modeling structural ensembles of transcription factors from nuclear magnetic resonance (NMR) and electron paramagnetic resonance (EPR) spectroscopic data, integrated with molecular dynamics (MD) simulations. We focus on approaches that confirm computed conformational ensembles by sparse constraints obtained from magnetic resonance. This combination enables the deduction of functional and structural protein models even if nuclear Overhauser effects (NOEs) are too scarce for conventional structure determination. We highlight recent insights into the folding-upon-DNA binding transitions of intrinsically disordered transcription factors that could be assessed using such integrative approaches.

Patterns in Protein Flexibility: A Comparison of NMR "Ensembles", MD Trajectories, and Crystallographic B-Factors

Proteins are molecular machines requiring flexibility to function. Crystallographic B-factors and Molecular Dynamics (MD) simulations both provide insights into protein flexibility on an atomic scale. Nuclear Magnetic Resonance (NMR) lacks a universally accepted analog of the B-factor. However, a lack of convergence in atomic coordinates in an NMR-based structure calculation also suggests atomic mobility. This paper describes a pattern in the coordinate uncertainties of backbone heavy atoms in NMR-derived structural “ensembles” first noted in the development of FindCore2 (previously called Expanded FindCore: DA Snyder, J Grullon, YJ Huang, R Tejero, GT Montelione, Proteins: Structure, Function, and Bioinformatics 82 (S2), 219–230) and demonstrates that this pattern exists in coordinate variances across MD trajectories but not in crystallographic B-factors. This either suggests that MD trajectories and NMR “ensembles” capture motional behavior of peptide bond units not captured by B-factors or indicates a deficiency common to force fields used in both NMR and MD calculations.

Recent advances in atomic molecular dynamics simulation of intrinsically disordered proteins

Intrinsically disordered proteins (IDPs) play important roles in cellular functions. The inherent structural heterogeneity of IDPs makes the high-resolution experimental characterization of IDPs extremely difficult. Molecular dynamics (MD) simulation could provide the atomic-level description of the structural and dynamic properties of IDPs. This perspective reviews the recent progress in atomic MD simulation studies of IDPs, including the development of force fields and sampling methods, as well as applications in IDP-involved protein-protein interactions. The employment of large-scale simulations and advanced sampling techniques allows more accurate estimation of the thermodynamics and kinetics of IDP-mediated protein interactions, and the holistic landscape of the binding process of IDPs is emerging.

Conformational Preferences of an Intrinsically Disordered Protein Domain: A Case Study for Modern Force Fields

Molecular simulations of intrinsically disordered proteins (IDPs) are challenging because they require sampling a very large number of relevant conformations, corresponding to a multitude of shallow minima in a flat free energy landscape. However, in the presence of a binding partner, the free energy landscape of an IDP can be dominated by few deep minima. This characteristic imposes high demands on the accuracy of the force field used to describe the molecular interactions. Here, as a model system for an IDP that is unstructured in solution but folds upon binding to a structured interaction partner, the transactivation domain of c-Myb was studied both in the unbound (free) form and when bound to the KIX domain. Six modern biomolecular force fields were systematically tested and compared in terms of their ability to describe the structural ensemble of the IDP. The protein force field/water model combinations included in this study are AMBER ff99SB-disp with its corresponding water model that was derived from TIP4P-D, CHARMM36m with TIP3P, ff15ipq with SPC/E_b, ff99SB*-ILDNP with TIP3P and TIP4P-D, and FB15 with TIP3P-FB water. Comparing the results from REST2-enhanced sampling simulations with experimental CD spectra and secondary chemical shifts reveals that the ff99SB-disp force field can realistically capture the broad and mildly helical structural ensemble of free c-Myb. The structural ensembles yielded by CHARMM36m, ff99SB*-ILDNP together with TIP4P-D water, and FB15 are also mildly helical; however, each of these force fields can be assigned a specific subset of c-Myb residues for which the simulations could not reproduce the experimental secondary chemical shifts. In addition, microsecond-timescale MD simulations of the KIX/c-Myb complex show that most force fields used preserve a stable helix fold of c-Myb in the complex. Still, all force fields predict a KIX/c-Myb complex interface that differs slightly from the structures provided by NMR because several NOE-derived distances between KIX and c-Myb were exceeded in the simulations. Taken together, the ff99SB-disp force field in the first place but also CHARMM36m, ff99SB*-ILDNP together with TIP4P-D water, and FB15 can be suitable choices for future simulation studies of the coupled folding and binding mechanism of the KIX/c-Myb complex and potentially also other IDPs.

Hybrid methods for combined experimental and computational determination of protein structure

Knowledge of protein structure is paramount to the understanding of biological function, developing new therapeutics, and making detailed mechanistic hypotheses. Therefore, methods to accurately elucidate three-dimensional structures of proteins are in high demand. While there are a few experimental techniques that can routinely provide high-resolution structures, such as x-ray crystallography, nuclear magnetic resonance (NMR), and cryo-EM, which have been developed to determine the structures of proteins, these techniques each have shortcomings and thus cannot be used in all cases. However, additionally, a large number of experimental techniques that provide some structural information, but not enough to assign atomic positions with high certainty have been developed. These methods offer sparse experimental data, which can also be noisy and inaccurate in some instances. In cases where it is not possible to determine the structure of a protein experimentally, computational structure prediction methods can be used as an alternative. Although computational methods can be performed without any experimental data in a large number of studies, inclusion of sparse experimental data into these prediction methods has yielded significant improvement. In this Perspective, we cover many of the successes of integrative modeling, computational modeling with experimental data, specifically for protein folding, protein-protein docking, and molecular dynamics simulations. We describe methods that incorporate sparse data from cryo-EM, NMR, mass spectrometry, electron paramagnetic resonance, small-angle x-ray scattering, Förster resonance energy transfer, and genetic sequence covariation. Finally, we highlight some of the major challenges in the field as well as possible future directions.

Benchmark assessment of molecular geometries and energies from small molecule force fields

Background: Force fields are used in a wide variety of contexts for classical molecular simulation, including studies on protein-ligand binding, membrane permeation, and thermophysical property prediction. The quality of these studies relies on the quality of the force fields used to represent the systems. Methods: Focusing on small molecules of fewer than 50 heavy atoms, our aim in this work is to compare nine force fields: GAFF, GAFF2, MMFF94, MMFF94S, OPLS3e, SMIRNOFF99Frosst, and the Open Force Field Parsley, versions 1.0, 1.1, and 1.2. On a dataset comprising 22,675 molecular structures of 3,271 molecules, we analyzed force field-optimized geometries and conformer energies compared to reference quantum mechanical (QM) data. Results: We show that while OPLS3e performs best, the latest Open Force Field Parsley release is approaching a comparable level of accuracy in reproducing QM geometries and energetics for this set of molecules. Meanwhile, the performance of established force fields such as MMFF94S and GAFF2 is generally somewhat worse. We also find that the series of recent Open Force Field versions provide significant increases in accuracy. Conclusions: This study provides an extensive test of the performance of different molecular mechanics force fields on a diverse molecule set, and highlights two (OPLS3e and OpenFF 1.2) that perform better than the others tested on the present comparison. Our molecule set and results are available for other researchers to use in testing.

Molecular dynamics simulations identify the regions of compromised thermostability in SazCA

The present study examined the structure and dynamics of the most active and thermostable carbonic anhydrase, SazCA, probed using molecular dynamics simulations. The molecular system was described by widely used biological force-fields (AMBER, CHARMM22, CHARMM36, and OPLS-AA) in conjunction with TIP3P water model. The comparison of molecular dynamics simulation results suggested AMBER to be a suitable choice to describe the structure and dynamics of SazCA. In addition to this, we also addressed the effect of temperature on the stability of SazCA. We performed molecular dynamics simulations at 313, 333, 353, 373, and 393 K to study the relationship between thermostability and flexibility in SazCA. The amino acid residues VAL98, ASN99, GLY100, LYS101, GLU145, and HIS207 were identified as the most flexible residues from root-mean-square fluctuations. The salt bridge analysis showed that ion-pairs ASP113-LYS81, ASP115-LYS81, ASP115-LYS114, GLU144-LYS143, and GLU144-LYS206, were responsible for the compromised thermal stability of SazCA.

Sampling Performance of Multiple Independent Molecular Dynamics Simulations of an RNA Aptamer

Using multiple independent simulations instead of one long simulation has been shown to improve the sampling performance attained with the molecular dynamics (MD) simulation method. However, it is generally not known how long each independent simulation should be, how many independent simulations should be used, or to what extent either of these factors affects the overall sampling performance achieved for a given system. The goal of the present study was to assess the sampling performance of multiple independent MD simulations, where each independent simulation begins from a different initial molecular conformation. For this purpose, we used an RNA aptamer that is 25 nucleotides long as a case study. The initial conformations of the aptamer are derived from six de novo predicted 3D structures. Each of the six de novo predicted structures is energy minimized in solution and equilibrated with MD simulations at high temperature. Ten conformations from these six high-temperature equilibration runs are selected as initial conformations for further simulations at ambient temperature. In total, we conducted 60 independent MD simulations, each with a duration of 100 ns, to study the conformation and dynamics of the aptamer. For each group of 10 independent simulations that originated from a particular de novo predicted structure, we evaluated the potential energy distribution of the RNA and used recurrence quantification analysis to examine the sampling of RNA conformational transitions. To assess the impact of starting from different de novo predicted structures, we computed the density of structure projection on principal components to compare the regions sampled by the different groups of ten independent simulations. The recurrence rate and dependence of initial conformation among the groups were also compared. We stress the necessity of using different initial configurations as simulation starting points by showing long simulations from different initial structures suffer from being trapped in different states. Finally, we summarized the sampling efficiency for the complete set of 60 independent simulations and determined regions of under-sampling on the potential energy landscape. The results suggest that conducting multiple independent simulations using a diverse set of de novo predicted structures is a promising approach to achieve sufficient sampling. This approach avoids undesirable outcomes, such as the problem of the RNA aptamer being trapped in a local minimum. For others wishing to conduct multiple independent simulations, the analysis protocol presented in this study is a guide for examining overall sampling and determining if more simulations are necessary for sufficient sampling.

pH-Induced Changes in Polypeptide Conformation: Force-Field Comparison with Experimental Validation

Microsecond-long all-atom molecular dynamics (MD) simulations, circular dichroism, laser Doppler velocimetry, and dynamic light-scattering techniques have been used to investigate pH-induced changes in the secondary structure, charge, and conformation of poly l-lysine (PLL) and poly l-glutamic acid (PGA). The employed combination of the experimental methods reveals for both PLL and PGA a narrow pH range at which they are charged enough to form stable colloidal suspensions, maintaining their α-helix content above 60%; an elevated charge state of the peptides required for colloidal stability promotes the peptide solvation as a random coil. To obtain a more microscopic view on the conformations and to verify the modeling performance, peptide secondary structure and conformations rising in MD simulations are also examined using three different force fields, i.e., OPLS-AA, CHARMM27, and AMBER99SB*-ILDNP. Ramachandran plots reveal that in the examined setup the α-helix content is systematically overestimated in CHARMM27, while OPLS-AA overestimates the β-sheet fraction at lower ionization degrees. At high ionization degrees, the OPLS-AA force-field-predicted secondary structure fractions match the experimentally measured distribution most closely. However, the pH-induced changes in PLL and PGA secondary structure are reasonably captured only by the AMBER99SB*-ILDNP force field, with the exception of the fully charged PGA in which the α-helix content is overestimated. The comparison to simulations results shows that the examined force fields involve significant deviations in their predictions for charged homopolypeptides. The detailed mapping of secondary structure dependency on pH for the polypeptides, especially finding the stable colloidal α-helical regime for both examined peptides, has significant potential for practical applications of the charged homopolypeptides. The findings raise attention especially to the pH fine tuning as an underappreciated control factor in surface modification and self-assembly.

Development of a Force Field for the Simulation of Single-Chain Proteins and Protein-Protein Complexes

The accuracy of atomistic physics-based force fields for the simulation of biological macromolecules has typically been benchmarked experimentally using biophysical data from simple, often single-chain systems. In the case of proteins, the careful refinement of force field parameters associated with torsion-angle potentials and the use of improved water models have enabled a great deal of progress toward the highly accurate simulation of such monomeric systems in both folded and, more recently, disordered states. In living organisms, however, proteins constantly interact with other macromolecules, such as proteins and nucleic acids, and these interactions are often essential for proper biological function. Here, we show that state-of-the-art force fields tuned to provide an accurate description of both ordered and disordered proteins can be limited in their ability to accurately describe protein-protein complexes. This observation prompted us to perform an extensive reparameterization of one variant of the Amber protein force field. Our objective involved refitting not only the parameters associated with torsion-angle potentials but also the parameters used to model nonbonded interactions, the specification of which is expected to be central to the accurate description of multicomponent systems. The resulting force field, which we call DES-Amber, allows for more accurate simulations of protein-protein complexes, while still providing a state-of-the-art description of both ordered and disordered single-chain proteins. Despite the improvements, calculated protein-protein association free energies still appear to deviate substantially from experiment, a result suggesting that more fundamental changes to the force field, such as the explicit treatment of polarization effects, may simultaneously further improve the modeling of single-chain proteins and protein-protein complexes.

Molecular Dynamics Ensemble Refinement of Intrinsically Disordered Peptides According to Deconvoluted Spectra from Circular Dichroism

We have developed a computational method of atomistically refining the structural ensemble of intrinsically disordered peptides (IDPs) facilitated by experimental measurements using circular dichroism spectroscopy (CD). A major challenge surrounding this approach stems from the deconvolution of experimental CD spectra into secondary structure features of the IDP ensemble. Currently available algorithms for CD deconvolution were designed to analyze the spectra of proteins with stable secondary structures. Herein, our work aims to minimize any bias from the peptide deconvolution analysis by implementing a non-negative linear least-squares fitting algorithm in conjunction with a CD reference data set that contains soluble and denatured proteins (SDP48). The non-negative linear least-squares method yields the best results for deconvolution of proteins with higher disordered content than currently available methods, according to a validation analysis of a set of protein spectra with Protein Data Bank entries. We subsequently used this analysis to deconvolute our experimental CD data to refine our computational model of the peptide secondary structure ensemble produced by all-atom molecular dynamics simulations with implicit solvent. We applied this approach to determine the ensemble structures of a set of short IDPs, that mimic the calmodulin binding domain of calcium/calmodulin-dependent protein kinase II and its 1-amino-acid and 3-amino-acid mutants. Our study offers a, to our knowledge, novel way to solve the ensemble secondary structures of IDPs in solution, which is important to advance the understanding of their roles in regulating signaling pathways through the formation of complexes with multiple partners.

Common Secondary and Tertiary Structural Features of Aptamer-Ligand Interaction Shared by RNA Aptamers with Different Primary Sequences

Aptamer selection can yield many oligonucleotides with different sequences and affinities for the target molecule. Here, we have combined computational and experimental approaches to understand if aptamers with different sequences but the same molecular target share structural and dynamical features. NEO1A, with a known NMR-solved structure, displays a flexible loop that interacts differently with individual aminoglycosides, its ligand affinities and specificities are responsive to ionic strength, and it possesses an adenosine in the loop that is critical for high-affinity ligand binding. NEO2A was obtained from the same selection and, although they are only 43% identical in overall sequence, NEO1A and NEO2A share similar loop sequences. Experimental analysis by 1D NMR and 2-aminopurine reporters combined with molecular dynamics modeling revealed similar structural and dynamical characteristics in both aptamers. These results are consistent with the hypothesis that the target ligand drives aptamer structure and also selects relevant dynamical characteristics for high-affinity aptamer-ligand interaction. Furthermore, they suggest that it might be possible to “migrate” structural and dynamical features between aptamer group members with different primary sequences but with the same target ligand.

Extensive tests and evaluation of the CHARMM36IDPSFF force field for intrinsically disordered proteins and folded proteins

Intrinsically disordered proteins (IDPs) have received increasing attention in recent studies due to their structural heterogeneity and critical biological functions. To fully understand the structural properties and determine accurate ensembles of IDPs, molecular dynamics (MD) simulation was widely used to sample diverse conformations and reveal the structural dynamics. However, the classical state-of-the-art force fields perform well for folded proteins while being unsatisfactory for the simulations of disordered proteins reported in many previous studies. Thus, improved force fields were developed to precisely describe both folded proteins and disordered proteins. Preliminary tests show that our newly developed CHARMM36IDPSFF (C36IDPSFF) force field can well reproduce the experimental observables of several disordered proteins, but more tests on different types of proteins are needed to further evaluate the performance of C36IDPSFF. Here, we extensively simulate short peptides, disordered proteins, and fast-folding proteins as well as folded proteins, and compare the simulated results with the experimental observables. The simulation results show that C36IDPSFF could substantially reproduce the experimental observables for most of the tested proteins but some limitations are also found in the radius of gyration of large disordered proteins and the stability of fast-folding proteins. This force field will facilitate large scale studies of protein structural dynamics and functions using MD simulations.

Comprehensive evaluation of the MM-GBSA method on bromodomain-inhibitor sets

MM-PB/GBSA methods represent a higher-level scoring theory than docking. This study reports an extensive testing of different MM-GBSA scoring schemes on two bromodomain (BRD) datasets. The first set is composed of 24 BRPF1 complexes, and the second one is a nonredundant set constructed from the PDBbind and composed of 28 diverse BRD complexes. A variety of MM-GBSA schemes were analyzed to evaluate the performance of four protocols with different numbers of minimization and MD steps, 10 different force fields and three different water models. Results showed that neither additional MD steps nor unfixing the receptor atoms improved scoring or ranking power. On the contrary, our results underscore the advantage of fixing receptor atoms or limiting the number of MD steps not only for a reduction in the computational costs but also for boosting the prediction accuracy. Among Amber force fields tested, ff14SB and its derivatives rather than ff94 or polarized force fields provided the most accurate scoring and ranking results. The TIP3P water model yielded the highest scoring and ranking power compared to the others. Posing power was further evaluated for the BRPF1 set. A slightly better posing power for the protocol which uses both minimization and MD steps with a fixed receptor than the one which uses only minimization with a fully flexible receptor-ligand system was observed. Overall, this study provides insights into the usage of the MM-GBSA methods for screening of BRD inhibitors, substantiating the benefits of shorter protocols and latest force fields and maintaining the crystal waters for accuracy.

Investigating targets for neuropharmacological intervention by molecular dynamics simulations

Medical research has identified over 500 brain disorders. Among these, there are still only very few neuropathologies whose causes are fully understood and, consequently, very few drugs whose mechanism of action is known. No FDA drug has been identified for major neurodegenerative diseases, such as Alzheimer's and Parkinson's. We still lack effective treatments and strategies for modulating progression or even early neurodegenerative disease onset diagnostic tools. A great support toward the highly needed identification of neuroactive drugs comes from computer simulation methods and, in particular, from molecular dynamics (MD). This provides insight into structure-function relationship of a target and predicts structure, dynamics and energetics of ligand/target complexes under biologically relevant conditions like temperature and physiological saline concentration. Here, we present examples of the predictive power of MD for neuroactive ligands/target complexes. This brief survey from our own research shows the usefulness of partnerships between academia and industry, and from joint efforts between experimental and theoretical groups.

High-throughput computational pipeline for 3-D structure preparation and in silico protein surface property screening: A case study on HBcAg dimer structures

Knowledge-based experimental design can aid biopharmaceutical high-throughput screening (HTS) experiments needed to identify critical manufacturability parameters. Prior knowledge can be obtained via computational methods such as protein property extraction from 3-D protein structures. This study presents a high-throughput 3-D structure preparation and refinement pipeline that supports structure screenings with an automated and data-dependent workflow. As a case study, three chimeric virus-like particle (VLP) building blocks, hepatitis B core antigen (HBcAg) dimers, were constructed. Molecular dynamics (MD) refinement quality, speed, stability, and correlation to zeta potential data was evaluated using different MD simulation settings. Settings included 2 force fields (YASARA2 and AMBER03) and 2 pKa computation methods (YASARA and H++). MD simulations contained a data-dependent termination via identification of a 2 ns Window of Stability, which was also used for robust descriptor extraction. MD simulation with YASARA2, independent of pKa computation method, was found to be most stable and computationally efficient. These settings resulted in a fast refinement (6.6-37.5 h), a good structure quality (-1.17--1.13) and a strong linear dependence between dimer surface charge and complete chimeric HBcAg VLP zeta potential. These results indicate the computational pipeline's applicability for early-stage candidate assessment and design optimization of HTS manufacturability or formulability experiments.

Adsorption of amino acids on graphene: assessment of current force fields

We compare the free energies of adsorption (ΔAads) and the structural preferences of amino acids on graphene obtained using the non-polarizable force fields-Amberff99SB-ILDN/TIP3P, CHARMM36/modified-TIP3P, OPLS-AA/M/TIP3P, and Amber03w/TIP4P/2005. The amino acid-graphene interactions are favorable irrespective of the force field. While the magnitudes of ΔAads differ between the force fields, the relative free energy of adsorption across amino acids is similar for the studied force fields. ΔAads positively correlates with amino acid-graphene and negatively correlates with graphene-water interaction energies. Using a combination of principal component analysis and density-based clustering technique, we grouped the structures observed in the graphene adsorbed state. The resulting population of clusters, and the conformation in each cluster indicate that the structures of the amino acid in the graphene adsorbed state vary across force fields. The differences in the conformations of amino acids are more severe in the graphene adsorbed state compared to the bulk state for all the force fields. Our findings suggest that the force fields studied will give qualitatively consistent relative strength of adsorption across proteins but different structural preferences in the graphene adsorbed state.

Characterizing Solvent Density Fluctuations in Dynamical Observation Volumes

Hydrophobic effects drive diverse aqueous assemblies, such as micelle formation or protein folding, wherein the solvent plays an important role. Consequently, characterizing the free energetics of solvent density fluctuations can lead to important insights into these processes. Although techniques such as the indirect umbrella sampling (INDUS) method can be used to characterize solvent fluctuations in static observation volumes of various sizes and shapes, characterizing how the solvent mediates inherently dynamic processes, such as self-assembly or conformational change, remains a challenge. In this work, we generalize the INDUS method to facilitate the enhanced sampling of solvent fluctuations in dynamical observation volumes, whose positions and shapes can evolve. We illustrate the usefulness of this generalization by characterizing water density fluctuations in dynamical volumes pertaining to the hydration of flexible solutes, the assembly of small hydrophobes, and conformational transitions in a model peptide. We also use the method to probe the dynamics of hard spheres.

Biased Agonist TRV027 Determinants in AT1R by Molecular Dynamics Simulations

Functional selectivity is a phenomenon observed in G protein-coupled receptors in which intermediate active-state conformations are stabilized by mutations or ligand binding, resulting in different sets of signaling pathways. Peptides capable of selectively activating β-arrestin, known as biased agonists, have already been characterized in vivo and could correspond to a new therapeutic approach for treatment of cardiovascular diseases. Despite the potential of biased agonism, the mechanism involved in selective signaling remains unclear. In this work, molecular dynamics simulations were employed to compare the conformational profile of the angiotensin II type 1 receptor (AT1R) crystal bound to angiotensin II, bound to the biased ligand TRV027, and in the apo form. Our results show that both ligands induce changes near the NPxxY motif in transmembrane domain 7 that are related to receptor activation. However, the biased ligand does not cause the rotamer toggle alternative positioning and displays an exclusive hydrogen-bonding pattern. Our work sheds light on the biased agonism mechanism and will help in the future design of novel biased agonists for AT1R.

Mechanistic insights into the Orai channel by molecular dynamics simulations

Highly Ca²⁺ selective channels trigger a large variety of cellular signaling processes in both excitable and non-excitable cells. Among these channels, the Orai channel is unique in its activation mechanism and its structure. It mediates Ca²⁺ influx into the cytosol with an extremely small unitary conductance over longer time-scales, ranging from minutes up to several hours. Its activation is regulated by the Ca²⁺ content of the endoplasmic reticulum (ER). Depletion of luminal [Ca²⁺]_ER is sensed by the STIM1 single transmembrane protein that directly binds and gates the Orai1 channel. Orai mediated Ca²⁺ influx increases cytosolic Ca²⁺ from 100 nM up to low micromolar range close to the pore and thereby forms Ca²⁺ microdomains. Hence, these features of the Orai channel can trigger long-term signaling processes without affecting the overall Ca²⁺ content of a single living cell. Here we focus on the architecture and dynamic conformational changes within the Orai channel. This review summarizes current achievements of molecular dynamics simulations in combination with live cell recordings to address gating and permeation of the Orai channel with molecular precision.

Analyzing Biomolecular Ensembles

Several techniques are available to generate conformational ensembles of proteins and other biomolecules either experimentally or computationally. These methods produce a large amount of data that need to be analyzed to identify structure-dynamics-function relationship. In this chapter, we will cover different tools to unveil the information hidden in conformational ensemble data and to guide toward the rationalization of the data. We included routinely used approaches such as dimensionality reduction, as well as new methods inspired by high-order statistics and graph theory.

Integrating Experiment and Theory to Understand TCR-pMHC Dynamics

The conformational dynamism of proteins is well established. Rather than having a single structure, proteins are more accurately described as a conformational ensemble that exists across a rugged energy landscape, where different conformational sub-states interconvert. The interaction between αβ T cell receptors (TCR) and cognate peptide-MHC (pMHC) is no exception, and is a dynamic process that involves substantial conformational change. This review focuses on technological advances that have begun to establish the role of conformational dynamics and dynamic allostery in TCR recognition of the pMHC and the early stages of signaling. We discuss how the marriage of molecular dynamics (MD) simulations with experimental techniques provides us with new ways to dissect and interpret the process of TCR ligation. Notably, application of simulation techniques lags behind other fields, but is predicted to make substantial contributions. Finally, we highlight integrated approaches that are being used to shed light on some of the key outstanding questions in the early events leading to TCR signaling.

On the Applicability of Force Fields To Study the Aggregation of Amyloidogenic Peptides Using Molecular Dynamics Simulations

Molecular dynamics simulations play an essential role in understanding biomolecular processes such as protein aggregation at temporal and spatial resolutions which are not attainable by experimental methods. For a correct modeling of protein aggregation, force fields must accurately represent molecular interactions. Here, we study the effect of five different force fields on the oligomer formation of Alzheimer's Aβ_16-22 peptide and two of its mutants: Aβ_16-22(F19V,F20V), which does not form fibrils, and Aβ_16-22(F19L) which forms fibrils faster than the wild type. We observe that while oligomer formation kinetics depends strongly on the force field, structural properties, such as the most relevant protein-protein contacts, are similar between them. The oligomer formation kinetics obtained with different force fields differ more from each other than the kinetics between aggregating and nonaggregating peptides simulated with a single force field. We discuss the difficulties in comparing atomistic simulations of amyloid oligomer formation with experimental observables.