Exploring the parameter space of the coarse-grained UNRES force field by random search: selecting a transferable medium-resolution force field

Deciphering the Mystery in p300 Taz2-p53 TAD2 Recognition

Intrinsically disordered proteins (IDPs) engage in various fundamental biological activities, and their behavior is of particular importance for a better understanding of the verbose but well-organized signal transduction in cells. IDPs exhibit uniquely paradoxical features with low affinity but simultaneously high specificity in recognizing their binding targets. The transcription factor p53 plays a crucial role in cancer suppression, carrying out some of its biological functions using its disordered regions, such as N-terminal transactivation domain 2 (TAD2). Exploration of the binding and unbinding processes between proteins is challenging, and the inherently disordered properties of these regions further complicate the issue. Computer simulations are a powerful tool to complement the experiments to fill gaps to explore the binding/unbinding processes between proteins. Here, we investigated the binding mechanism between p300 Taz2 and p53 TAD2 through extensive molecular dynamics (MD) simulations using the physics-based UNited RESidue (UNRES) force field with additional Go̅-like potentials. Distance restraints extracted from the NMR-resolved structures were imposed on intermolecular residue pairs to accelerate binding simulations, in which Taz2 was immobilized in a native-like conformation and disordered TAD2 was fully free. Starting from six structures with TAD2 placed at different positions around Taz2, we observed a metastable intermediate state in which the middle helical segment of TAD2 is anchored in the binding pocket, highlighting the significance of the TAD2 helix in directing protein recognition. Physics-based binding simulations show that successful binding is achieved after a series of stages, including (1) protein collisions to initiate the formation of encounter complexes, (2) partial attachment of TAD2, and finally (3) full attachment of TAD2 to the correct binding pocket of Taz2. Furthermore, machine-learning-based PathDetect-SOM was used to identify two binding pathways, the encounter complexes, and the intermediate states.

AlphaFold Accurately Predicts the Structure of Ribosomally Synthesized and Post-Translationally Modified Peptide Biosynthetic Enzymes

Ribosomally synthesized and post-translationally modified peptides (RiPPs) are a growing class of natural products biosynthesized from a genetically encoded precursor peptide. The enzymes that install the post-translational modifications on these peptides have the potential to be useful catalysts in the production of natural-product-like compounds and can install non-proteogenic amino acids in peptides and proteins. However, engineering these enzymes has been somewhat limited, due in part to limited structural information on enzymes in the same families that nonetheless exhibit different substrate selectivities. Despite AlphaFold2's superior performance in single-chain protein structure prediction, its multimer version lacks accuracy and requires high-end GPUs, which are not typically available to most research groups. Additionally, the default parameters of AlphaFold2 may not be optimal for predicting complex structures like RiPP biosynthetic enzymes, due to their dynamic binding and substrate-modifying mechanisms. This study assessed the efficacy of the structure prediction program ColabFold (a variant of AlphaFold2) in modeling RiPP biosynthetic enzymes in both monomeric and dimeric forms. After extensive benchmarking, it was found that there were no statistically significant differences in the accuracy of the predicted structures, regardless of the various possible prediction parameters that were examined, and that with the default parameters, ColabFold was able to produce accurate models. We then generated additional structural predictions for select RiPP biosynthetic enzymes from multiple protein families and biosynthetic pathways. Our findings can serve as a reference for future enzyme engineering complemented by AlphaFold-related tools.

Simple and Effective Conformational Sampling Strategy for Intrinsically Disordered Proteins Using the UNRES Web Server

Intrinsically disordered proteins (IDPs) contain more charged amino acids than folded proteins, resulting in a lack of hydrophobic core(s) and a tendency to adopt rapidly interconverting structures rather than well-defined structures. The structural heterogeneity of IDPs, encoded by the amino acid sequence, is closely related to their unique roles in biological pathways, which require them to interact with different binding partners. Recently, Robustelli and co-workers have demonstrated that a balanced all-atom force field can be used to sample heterogeneous structures of disordered proteins ( Proc. Natl. Acad. Sci. U.S.A. 2018, 115, E4758-E4766). However, such a solution requires extensive computational resources, such as Anton supercomputers. Here, we propose a simple and effective solution to sample the conformational space of IDPs using a publicly available web server, namely, the UNited-RESidue (UNRES) web server. Our proposed solution requires no investment in computational resources and no prior knowledge of UNRES. UNRES Replica Exchange Molecular Dynamics (REMD) simulations were carried out on a set of eight disordered proteins at temperatures spanning from 270 to 430 K. Utilizing the latest UNRES force field designed for structured proteins, with proper selections of temperatures, we were able to produce comparable results to all-atom force fields as reported in work done by Robustelli and co-workers. In addition, NMR observables and the radius of gyration calculated from UNRES ensembles were directly compared with the experimental data to further evaluate the accuracy of the UNRES model at all temperatures. Our results suggest that carrying out the UNRES simulations at optimal temperatures using the UNRES web server can be a good alternative to sample heterogeneous structures of IDPs.

Insight into the Initial Stages of the Folding Process in Onconase Revealed by UNRES

The unfolded state of proteins presents many challenges to elucidate the structural basis for biological function. This state is characterized by a large degree of structural heterogeneity which makes it difficult to generate structural models. However, recent experiments into the initial folding events of the 104-residue ribonuclease homologue onconase (ONC) were able to identify the regions in the protein that participate in the initial folding of this protein. Therefore, to gain additional structural insight into the unfolded state of proteins, this study utilized molecular dynamics simulations using the UNited-RESidue (UNRES) force field to evaluate whether there is a good agreement between the experimentally determined initial structures and the structures identified by computer simulations along a folding pathway. Indeed, these UNRES simulations accurately identified the two regions experimentally observed to form the initial native structure along the folding pathway of ONC. In addition, these regions are determined to be chain folding initiation sites (CFIS) according to methods developed previously. Subsequent self-organization maps (SOM) analysis has revealed key structural states involved in these early folding events.

Probing Protein Aggregation Using the Coarse-Grained UNRES Force Field

Protein aggregation is the cause of many, often lethal, diseases, including the Alzheimer's, Parkinson's, and Huntington's diseases, and familial amyloidosis. Theoretical investigation of the mechanism of this process, including the structures of the oligomeric intermediates which are the most toxic, is difficult because of long time scale of aggregation. Coarse-grained models, which enable us to extend the simulation time scale by three or more orders of magnitude, are, therefore, of great advantage in such studies. In this chapter, we describe the application of the physics-based UNited RESidue (UNRES) force field developed in our laboratory to study protein aggregation, in both free simulations and simulations of aggregation propagation from an existing template (seed), and illustrate it with the examples of Aβ-peptide aggregation and Aβ-peptide-assisted aggregation of the peptides derived from the repeat domains of tau (TauRD).

Residue-Level Contact Reveals Modular Domain Interactions of PICK1 Are Driven by Both Electrostatic and Hydrophobic Forces

PICK1 is a multi-domain scaffolding protein that is uniquely comprised of both a PDZ domain and a BAR domain. While previous experiments have shown that the PDZ domain and the linker positively regulate the BAR domain and the C-terminus negatively regulates the BAR domain, the details of internal regulation mechanisms are unknown. Molecular dynamics (MD) simulations have been proven to be a useful tool in revealing the intramolecular interactions at atomic-level resolution. PICK1 performs its biological functions in a dimeric form which is extremely computationally demanding to simulate with an all-atom force field. Here, we use coarse-grained MD simulations to expose the key residues and driving forces in the internal regulations of PICK1. While the PDZ and BAR domains do not form a stable complex, our simulations show the PDZ domain preferentially interacting with the concave surface of the BAR domain over other BAR domain regions. Furthermore, our simulations show that the short helix in the linker region can form interactions with the PDZ domain. Our results reveal that the surface of the βB-βC loop, βC strand, and αA-βD loop of the PDZ domain can form a group of hydrophobic interactions surrounding the linker helix. These interactions are driven by hydrophobic forces. In contrast, our simulations reveal a very dynamic C-terminus that most often resides on the convex surface of the BAR domain rather than the previously suspected concave surface. These interactions are driven by a combination of electrostatic and hydrophobic interactions.

Atomistic detailed free-energy landscape of intrinsically disordered protein studied by multi-scale divide-and-conquer molecular dynamics simulation

Calcineurin (CaN) is a eukaryotic serine/threonine protein phosphatase activated by both Ca²⁺ and calmodulin (CaM), including intrinsically disordered region (IDR). The region undergoes folding into an α-helix form in the presence Ca²⁺ -loaded CaM. To sample the ordered structure of the IDR by conventional all atom model (AAM) molecular dynamics (MD) simulation, the IDR and Ca²⁺ -loaded CaM must be simultaneously treated. However, it is time-consuming task because the coupled folding and binding should include repeated binding and dissociation. Then, in this study, we propose novel multi-scale divide-and-conquer MD (MSDC-MD), which combines AAM-MD and coarse-grained model MD (CGM-MD). To speed up the conformation sampling, MSDC-MD simulation first treats the IDR by CGM to sample conformations from wide conformation space; then, multiple AAM-MD in a limited area is initiated using the resultant CGM conformation, which is reconstructed by homology modeling method. To investigate performance, we sampled the ordered conformation of the IDR using MSDC-MD; the root-mean-square distance (RMSD) with respect to the experimental structure was 2.23 Å.

Evaluation of the scale-consistent UNRES force field in template-free prediction of protein structures in the CASP13 experiment

The recent NEWCT-9P version of the coarse-grained UNRES force field for proteins, with scale-consistent formulas for the local and correlation terms, has been tested in the CASP13 experiment of the blind-prediction of protein structure, in the ab initio, contact-assisted, and data-assisted modes. Significant improvement of the performance has been observed with respect to the CASP11 and CASP12 experiments (by over 10 GDT_TS units for the ab initio mode predictions and by over 15 GDT_TS units for the contact-assisted prediction, respectively), which is a result of introducing scale-consistent terms and improved handling of contact-distance restraints. As in previous CASP exercises, UNRES ranked higher in the free modeling category than in the general category that included template based modeling targets. Use of distance restraints from the predicted contacts, albeit many of them were wrong, resulted in the increase of GDT_TS by over 8 units on average and introducing sparse restraints from small-angle X-ray/neutron scattering and chemical cross-link-mass-spectrometry experiments, and ambiguous restraints from nuclear magnetic resonance experiments has also improved the predictions by 8.6, 9.7, and 10.7 GDT_TS units on average, respectively.

Reoptimized UNRES Potential for Protein Model Quality Assessment

Ranking protein structure models is an elusive problem in bioinformatics. These models are evaluated on both the degree of similarity to the native structure and the folding pathway. Here, we simulated the use of the coarse-grained UNited RESidue (UNRES) force field as a tool to choose the best protein structure models for a given protein sequence among a pool of candidate models, using server data from the CASP11 experiment. Because the original UNRES was optimized for Molecular Dynamics simulations, we reoptimized UNRES using a deep feed-forward neural network, and we show that introducing additional descriptive features can produce better results. Overall, we found that the reoptimized UNRES performs better in selecting the best structures and tracking protein unwinding from its native state. We also found a relatively poor correlation between UNRES values and the model’s Template Modeling Score (TMS). This is remedied by reoptimization. We discuss some cases where our reoptimization procedure is useful.

Dependence of the Formation of Tau and Aβ Peptide Mixed Aggregates on the Secondary Structure of the N-Terminal Region of Aβ

One of the hallmarks of Alzheimer's disease is the formation of aggregates of the tau protein, a process that can be facilitated by the presence of fibrils formed by the amyloid β peptide (Aβ). However, the mechanism that triggers tau aggregation is still a matter of debate. The effect of Aβ₄₀ fibrils on the aggregation of the repeat domain of tau (TauRD) is investigated here by employing coarse-grained molecular dynamics simulations. The results indicate that the repeat domain of tau has a high affinity for Aβ₄₀ fibrils, with the ₂₆₁GSTENLK₂₆₇ fragment of tau driving TauRD toward the ₁₆KLVFFA₂₁ fragment in Aβ₄₀. Monomeric Aβ₄₀, in which the ₁₆KLVFFA₂₁ fragment is rarely found in an extended conformation (as in the fibril), has a low affinity for the TauRD, indicating that the ability of Aβ₄₀ fibrils to bind to the TauRD depends on the ₁₆KLVFFA₂₁ fragment of Aβ adopting an extended conformation.

UNRES server for physics-based coarse-grained simulations and prediction of protein structure, dynamics and thermodynamics

A server implementation of the UNRES package (http://www.unres.pl) for coarse-grained simulations of protein structures with the physics-based UNRES model, coined a name UNRES server, is presented. In contrast to most of the protein coarse-grained models, owing to its physics-based origin, the UNRES force field can be used in simulations, including those aimed at protein-structure prediction, without ancillary information from structural databases; however, the implementation includes the possibility of using restraints. Local energy minimization, canonical molecular dynamics simulations, replica exchange and multiplexed replica exchange molecular dynamics simulations can be run with the current UNRES server; the latter are suitable for protein-structure prediction. The user-supplied input includes protein sequence and, optionally, restraints from secondary-structure prediction or small x-ray scattering data, and simulation type and parameters which are selected or typed in. Oligomeric proteins, as well as those containing D-amino-acid residues and disulfide links can be treated. The output is displayed graphically (minimized structures, trajectories, final models, analysis of trajectory/ensembles); however, all output files can be downloaded by the user. The UNRES server can be freely accessed at http://unres-server.chem.ug.edu.pl.

Prediction of protein structure with the coarse-grained UNRES force field assisted by small X-ray scattering data and knowledge-based information

A new approach to assisted protein-structure prediction has been proposed, which is based on running multiplexed replica exchange molecular dynamics simulations with the coarse-grained UNRES force field with restraints derived from knowledge-based models and distance distribution from small angle X-ray scattering (SAXS) measurements. The latter restraints are incorporated into the target function as a maximum-likelihood term that guides the shape of the simulated structures towards that defined by SAXS. The approach was first verified with the 1KOY protein, for which the distance distribution was calculated from the experimental structure, and subsequently used to predict the structures of 11 data-assisted targets in the CASP12 experiment. Major improvement of the GDT_TS was obtained for 2 targets, minor improvement for other 2 while, for 6 target GDT_TS deteriorated compared with that calculated for predictions without the SAXS data, partly because of assuming a wrong multimeric state (for Ts866) or because the crystal conformation was more compact than the solution conformation (for Ts942). Particularly good results were obtained for Ts909, in which use of SAXS data resulted in the selection of a correctly packed trimer and, subsequently, increased the GDT_TS of monomer prediction. It was found that running simulations with correct oligomeric state is essential for the success in SAXS-data-assisted prediction.

Shielding effect in protein folding

One of the most important interactions responsible for protein folding and stability are hydrogen bonds between peptide groups. There is a constant competition between the water molecules and peptide groups in a hydrogen bond formation. Also side-chains take part in this process by reducing hydration of peptide group (shielding effect) that promotes the protein folding. In this paper, a new approach to take into account a shielding effect is presented. A modification of the energy function is derived and incorporated into the UNited RESidue (UNRES) force field. Canonical Molecular Dynamics and Replica Exchange Molecular Dynamics with UNRES force field is applied to study the influence of this effect on protein structure, folding kinetics and free energy landscapes. The results of test calculations suggest that even small contribution of this effect into energy function changes force field behavior as well as speeds up the folding process significantly.

Elucidating Important Sites and the Mechanism for Amyloid Fibril Formation by Coarse-Grained Molecular Dynamics

Fibrils formed by the β-amyloid (Aβ) peptide play a central role in the development of Alzheimer's disease. In this study, the principles governing their growth and stability are investigated by analyzing canonical and replica-exchange molecular dynamics trajectories of Aβ_(9-40) fibrils. In particular, an unstructured monomer was allowed to interact freely with an Aβ fibril template. Trajectories were generated with the coarse-grained united-residue force field, and one- and two-dimensional free-energy landscapes (FELs) along the backbone virtual-bond angle θ and backbone virtual-bond-dihedral angle γ of each residue and principal components, respectively, were analyzed. Also, thermal unbinding (unfolding) of an Aβ peptide from the fibril template was investigated. These analyses enable us to illustrate the entire process of Aβ fibril elongation and to elucidate the key residues involved in it. Several different pathways were identified during the search for the fibril conformation by the monomer, which finally follows a dock-lock mechanism with two distinct locking stages. However, it was found that the correct binding, with native hydrogen bonds, of the free monomer to the fibril template at both stages is crucial for fibril elongation. In other words, if the monomer is incorrectly bound (with nonnative hydrogen bonds) to the fibril template during the first "docking" stage, it can remain attached to it for a long time before it dissociates and either attempts a different binding or allows another monomer to bind. This finding is consistent with an experimentally observed "stop-and-go" mechanism of fibril growth.

Role of the sulfur to α-carbon thioether bridges in thurincin H

Thurincin H is a small protein produced by Bacillus thuringiensis SF361 with gram-positive antimicrobial properties. The toxins produced by B. thuringiensis are widely used in the agriculture as, e.g. natural preservatives in dairy products. The structure of thurincin H possesses four covalent sulfur to [Formula: see text]-carbon bonds that involve the cysteine side-chains; these bonds are probably responsible for the shape and stability of the protein and, thereby, for its antimicrobial properties. To examine the influence of the formation of the sulfur-carbon bonds on the folding pathways and stability of the protein, a series of canonical and multiplexed replica-exchange simulations with the coarse-grained UNRES force field was carried out without and with distance restraints imposed on selected S-C[Formula: see text] atom pairs. It was found that the order of the formation and breaking of the S-C[Formula: see text] thioether bonds significantly impacts on the foldability and stability of the thurincin H. It was also observed that thioether bridges play a major role in stabilizing the global fold of the protein, although it significantly diminishes the entropy of the system. The maximum foldability of thurincin H was observed in the presence of the optimal set of three out of four thioether bridges. Thus, the results suggest that the presence of ThnB enzyme and other agents that catalyze the formation of thioether bridges can be essential for correct folding of thurincin H and that the formation of the fourth bridge does not seem to facilitate folding; instead, it seems to rigidify the loop and prevent proteolysis.

Replica-exchange-with-tunneling for fast exploration of protein landscapes

While the use of replica-exchange molecular dynamics in protein simulations has become ubiquitous, its utility is limited in many practical applications. We propose to overcome some shortcomings that hold back its use in settings such as multi-scale or explicit solvent simulations by integrating ideas of hybrid MC/MD into the replica-exchange protocol. This Replica-Exchange-with-Tunneling method is tested by simulating the Trp-cage protein, a system often used in molecular biophysics for testing sampling techniques.

Optimization of a Nucleic Acids united-RESidue 2-Point model (NARES-2P) with a maximum-likelihood approach

Coarse-grained models are useful tools to investigate the structural and thermodynamic properties of biomolecules. They are obtained by merging several atoms into one interaction site. Such simplified models try to capture as much as possible information of the original biomolecular system in all-atom representation but the resulting parameters of these coarse-grained force fields still need further optimization. In this paper, a force field optimization method, which is based on maximum-likelihood fitting of the simulated to the experimental conformational ensembles and least-squares fitting of the simulated to the experimental heat-capacity curves, is applied to optimize the Nucleic Acid united-RESidue 2-point (NARES-2P) model for coarse-grained simulations of nucleic acids recently developed in our laboratory. The optimized NARES-2P force field reproduces the structural and thermodynamic data of small DNA molecules much better than the original force field.

Physics-Based Potentials for Coarse-Grained Modeling of Protein-DNA Interactions

Physics-based potentials have been developed for the interactions between proteins and DNA for simulations with the UNRES + NARES-2P force field. The mean-field interactions between a protein and a DNA molecule can be divided into eight categories: (1) nonpolar side chain–DNA base, (2) polar uncharged side chain–DNA base, (3) charged side chain–DNA base, (4) peptide group–phosphate group, (5) peptide group–DNA base, (6) nonpolar side chain–phosphate group, (7) polar uncharged side chain–phosphate group, and (8) charged side chain–phosphate group. Umbrella-sampling molecular dynamics simulations in explicit TIP3P water using the AMBER force field were carried out to determine the potentials of mean force (PMF) for all 105 pairs of interacting components. Approximate analytical expressions for the mean-field interaction energy of each pair of the different kinds of interacting molecules were then fitted to the PMFs to obtain the parameters of the analytical expressions. These analytical expressions can reproduce satisfactorily the PMF curves corresponding to different orientations of the interacting molecules. The results suggest that the physics-based mean-field potentials of amino acid–nucleotide interactions presented here can be used in coarse-grained simulation of protein–DNA interactions.

Optimization of Analytical Potentials for Coarse-Grained Biopolymer Models

The increasing trend in the recent literature on coarse grained (CG) models testifies their impact in the study of complex systems. However, the CG model landscape is variegated: even considering a given resolution level, the force fields are very heterogeneous and optimized with very different parametrization procedures. Along the road for standardization of CG models for biopolymers, here we describe a strategy to aid building and optimization of statistics based analytical force fields and its implementation in the software package AsParaGS (Assisted Parameterization platform for coarse Grained modelS). Our method is based on the use and optimization of analytical potentials, optimized by targeting internal variables statistical distributions by means of the combination of different algorithms (i.e., relative entropy driven stochastic exploration of the parameter space and iterative Boltzmann inversion). This allows designing a custom model that endows the force field terms with a physically sound meaning. Furthermore, the level of transferability and accuracy can be tuned through the choice of statistical data set composition. The method-illustrated by means of applications to helical polypeptides-also involves the analysis of two and three variable distributions, and allows handling issues related to the FF term correlations. AsParaGS is interfaced with general-purpose molecular dynamics codes and currently implements the "minimalist" subclass of CG models (i.e., one bead per amino acid, Cα based). Extensions to nucleic acids and different levels of coarse graining are in the course.

Performance of protein-structure predictions with the physics-based UNRES force field in CASP11

Participating as the Cornell-Gdansk group, we have used our physics-based coarse-grained UNited RESidue (UNRES) force field to predict protein structure in the 11th Community Wide Experiment on the Critical Assessment of Techniques for Protein Structure Prediction (CASP11). Our methodology involved extensive multiplexed replica exchange simulations of the target proteins with a recently improved UNRES force field to provide better reproductions of the local structures of polypeptide chains. All simulations were started from fully extended polypeptide chains, and no external information was included in the simulation process except for weak restraints on secondary structure to enable us to finish each prediction within the allowed 3-week time window. Because of simplified UNRES representation of polypeptide chains, use of enhanced sampling methods, code optimization and parallelization and sufficient computational resources, we were able to treat, for the first time, all 55 human prediction targets with sizes from 44 to 595 amino acid residues, the average size being 251 residues. Complete structures of six single-domain proteins were predicted accurately, with the highest accuracy being attained for the T0769, for which the CαRMSD was 3.8 Å for 97 residues of the experimental structure. Correct structures were also predicted for 13 domains of multi-domain proteins with accuracy comparable to that of the best template-based modeling methods. With further improvements of the UNRES force field that are now underway, our physics-based coarse-grained approach to protein-structure prediction will eventually reach global prediction capacity and, consequently, reliability in simulating protein structure and dynamics that are important in biochemical processes.

AVAILABILITY AND IMPLEMENTATION

Freely available on the web at http://www.unres.pl/ CONTACT: has5@cornell.edu.

Molecular dynamics of protein A and a WW domain with a united-residue model including hydrodynamic interaction

The folding of the N-terminal part of the B-domain of staphylococcal protein A (PDB ID: 1BDD, a 46-residue three-α-helix bundle) and the formin-binding protein 28 WW domain (PDB ID: 1E0L, a 37-residue three-stranded anti-parallel β protein) was studied by means of Langevin dynamics with the coarse-grained UNRES force field to assess the influence of hydrodynamic interactions on protein-folding pathways and kinetics. The unfolded, intermediate, and native-like structures were identified by cluster analysis, and multi-exponential functions were fitted to the time dependence of the fractions of native and intermediate structures, respectively, to determine bulk kinetics. It was found that introducing hydrodynamic interactions slows down both the formation of an intermediate state and the transition from the collapsed structures to the final native-like structures by creating multiple kinetic traps. Therefore, introducing hydrodynamic interactions considerably slows the folding, as opposed to the results obtained from earlier studies with the use of Gō-like models.

My 65 years in protein chemistry

This is a tour of a physical chemist through 65 years of protein chemistry from the time when emphasis was placed on the determination of the size and shape of the protein molecule as a colloidal particle, with an early breakthrough by James Sumner, followed by Linus Pauling and Fred Sanger, that a protein was a real molecule, albeit a macromolecule. It deals with the recognition of the nature and importance of hydrogen bonds and hydrophobic interactions in determining the structure, properties, and biological function of proteins until the present acquisition of an understanding of the structure, thermodynamics, and folding pathways from a linear array of amino acids to a biological entity. Along the way, with a combination of experiment and theoretical interpretation, a mechanism was elucidated for the thrombin-induced conversion of fibrinogen to a fibrin blood clot and for the oxidative-folding pathways of ribonuclease A. Before the atomic structure of a protein molecule was determined by x-ray diffraction or nuclear magnetic resonance spectroscopy, experimental studies of the fundamental interactions underlying protein structure led to several distance constraints which motivated the theoretical approach to determine protein structure, and culminated in the Empirical Conformational Energy Program for Peptides (ECEPP), an all-atom force field, with which the structures of fibrous collagen-like proteins and the 46-residue globular staphylococcal protein A were determined. To undertake the study of larger globular proteins, a physics-based coarse-grained UNited-RESidue (UNRES) force field was developed, and applied to the protein-folding problem in terms of structure, thermodynamics, dynamics, and folding pathways. Initially, single-chain and, ultimately, multiple-chain proteins were examined, and the methodology was extended to protein-protein interactions and to nucleic acids and to protein-nucleic acid interactions. The ultimate results led to an understanding of a variety of biological processes underlying natural and disease phenomena.

Physics-based potentials for the coupling between backbone- and side-chain-local conformational states in the UNited RESidue (UNRES) force field for protein simulations

The UNited RESidue (UNRES) model of polypeptide chains is a coarse-grained model in which each amino-acid residue is reduced to two interaction sites, namely, a united peptide group (p) located halfway between the two neighboring α-carbon atoms (Cαs), which serve only as geometrical points, and a united side chain (SC) attached to the respective Cα. Owing to this simplification, millisecond molecular dynamics simulations of large systems can be performed. While UNRES predicts overall folds well, it reproduces the details of local chain conformation with lower accuracy. Recently, we implemented new knowledge-based torsional potentials (Krupa et al. J. Chem. Theory Comput. 2013, 9, 4620–4632) that depend on the virtual-bond dihedral angles involving side chains: Cα···Cα···Cα···SC (τ(1)), SC···Cα···Cα···Cα (τ(2)), and SC···Cα···Cα···SC (τ(3)) in the UNRES force field. These potentials resulted in significant improvement of the simulated structures, especially in the loop regions. In this work, we introduce the physics-based counterparts of these potentials, which we derived from the all-atom energy surfaces of terminally blocked amino-acid residues by Boltzmann integration over the angles λ(1) and λ(2) for rotation about the Cα···Cα virtual-bond angles and over the side-chain angles χ. The energy surfaces were, in turn, calculated by using the semiempirical AM1 method of molecular quantum mechanics. Entropy contribution was evaluated with use of the harmonic approximation from Hessian matrices. One-dimensional Fourier series in the respective virtual-bond-dihedral angles were fitted to the calculated potentials, and these expressions have been implemented in the UNRES force field. Basic calibration of the UNRES force field with the new potentials was carried out with eight training proteins, by selecting the optimal weight of the new energy terms and reducing the weight of the regular torsional terms. The force field was subsequently benchmarked with a set of 22 proteins not used in the calibration. The new potentials result in a decrease of the root-mean-square deviation of the average conformation from the respective experimental structure by 0.86 Å on average; however, improvement of up to 5 Å was observed for some proteins.

WeFold: a coopetition for protein structure prediction

The protein structure prediction problem continues to elude scientists. Despite the introduction of many methods, only modest gains were made over the last decade for certain classes of prediction targets. To address this challenge, a social-media based worldwide collaborative effort, named WeFold, was undertaken by 13 labs. During the collaboration, the laboratories were simultaneously competing with each other. Here, we present the first attempt at "coopetition" in scientific research applied to the protein structure prediction and refinement problems. The coopetition was possible by allowing the participating labs to contribute different components of their protein structure prediction pipelines and create new hybrid pipelines that they tested during CASP10. This manuscript describes both successes and areas needing improvement as identified throughout the first WeFold experiment and discusses the efforts that are underway to advance this initiative. A footprint of all contributions and structures are publicly accessible at http://www.wefold.org.

A unified coarse-grained model of biological macromolecules based on mean-field multipole-multipole interactions

A unified coarse-grained model of three major classes of biological molecules—proteins, nucleic acids, and polysaccharides—has been developed. It is based on the observations that the repeated units of biopolymers (peptide groups, nucleic acid bases, sugar rings) are highly polar and their charge distributions can be represented crudely as point multipoles. The model is an extension of the united residue (UNRES) coarse-grained model of proteins developed previously in our laboratory. The respective force fields are defined as the potentials of mean force of biomacromolecules immersed in water, where all degrees of freedom not considered in the model have been averaged out. Reducing the representation to one center per polar interaction site leads to the representation of average site–site interactions as mean-field dipole–dipole interactions. Further expansion of the potentials of mean force of biopolymer chains into Kubo’s cluster-cumulant series leads to the appearance of mean-field dipole–dipole interactions, averaged in the context of local interactions within a biopolymer unit. These mean-field interactions account for the formation of regular structures encountered in biomacromolecules, e.g., α-helices and β-sheets in proteins, double helices in nucleic acids, and helicoidally packed structures in polysaccharides, which enables us to use a greatly reduced number of interacting sites without sacrificing the ability to reproduce the correct architecture. This reduction results in an extension of the simulation timescale by more than four orders of magnitude compared to the all-atom representation. Examples of the performance of the model are presented.

Figure

Revised Backbone-Virtual-Bond-Angle Potentials to Treat the l- and d-Amino Acid Residues in the Coarse-Grained United Residue (UNRES) Force Field

Continuing our effort to introduce d-amino-acid residues in the united residue (UNRES) force field developed in our laboratory, in this work the Cα ··· Cα ··· Cα backbone-virtual-bond-valence-angle (θ) potentials for systems containing d-amino-acid residues have been developed. The potentials were determined by integrating the combined energy surfaces of all possible triplets of terminally blocked glycine, alanine, and proline obtained with ab initio molecular quantum mechanics at the MP2/6-31G(d,p) level to calculate the corresponding potentials of mean force (PMFs). Subsequently, analytical expressions were fitted to the PMFs to give the virtual-bond-valence potentials to be used in UNRES. Alanine represented all types of amino-acid residues except glycine and proline. The blocking groups were either the N-acetyl and N',N'-dimethyl or N-acetyl and pyrrolidyl group, depending on whether the residue next in sequence was an alanine-type or a proline residue. A total of 126 potentials (63 symmetry-unrelated potentials for each set of terminally blocking groups) were determined. Together with the torsional, double-torsional, and side-chain-rotamer potentials for polypeptide chains containing d-amino-acid residues determined in our earlier work (Sieradzan et al. J. Chem. Theory Comput., 2012, 8, 4746), the new virtual-bond-angle (θ) potentials now constitute the complete set of physics-based potentials with which to run coarse-grained simulations of systems containing d-amino-acid residues. The ability of the extended UNRES force field to reproduce thermodynamics of polypeptide systems with d-amino-acid residues was tested by comparing the experimentally measured and the calculated free energies of helix formation of model KLALKLALxxLKLALKLA peptides, where x denotes any d- or l- amino-acid residue. The obtained results demonstrate that the UNRES force field with the new potentials reproduce the changes of free energies of helix formation upon d-substitution but overestimate the free energies of helix formation. To test the ability of UNRES with the new potentials to reproduce the structures of polypeptides with d-amino-acid residues, an ab initio replica-exchange folding simulation of thurincin H from Bacillus thuringiensis, which has d-amino-acid residues in the sequence, was carried out. UNRES was able to locate the native α-helical hairpin structure as the dominant structure even though no native sulfide-carbon bonds were present in the simulation.

Kinks, loops, and protein folding, with protein A as an example

The dynamics and energetics of formation of loops in the 46-residue N-terminal fragment of the B-domain of staphylococcal protein A has been studied. Numerical simulations have been performed using coarse-grained molecular dynamics with the united-residue (UNRES) force field. The results have been analyzed in terms of a kink (heteroclinic standing wave solution) of a generalized discrete nonlinear Schrödinger (DNLS) equation. In the case of proteins, the DNLS equation arises from a C(α)-trace-based energy function. Three individual kink profiles were identified in the experimental three-α-helix structure of protein A, in the range of the Glu16-Asn29, Leu20-Asn29, and Gln33-Asn44 residues, respectively; these correspond to two loops in the native structure. UNRES simulations were started from the full right-handed α-helix to obtain a clear picture of kink formation, which would otherwise be blurred by helix formation. All three kinks emerged during coarse-grained simulations. It was found that the formation of each is accompanied by a local free energy increase; this is expressed as the change of UNRES energy which has the physical sense of the potential of mean force of a polypeptide chain. The increase is about 7 kcal/mol. This value can thus be considered as the free energy barrier to kink formation in full α-helical segments of polypeptide chains. During the simulations, the kinks emerge, disappear, propagate, and annihilate each other many times. It was found that the formation of a kink is initiated by an abrupt change in the orientation of a pair of consecutive side chains in the loop region. This resembles the formation of a Bloch wall along a spin chain, where the C(α) backbone corresponds to the chain, and the amino acid side chains are interpreted as the spin variables. This observation suggests that nearest-neighbor side chain-side chain interactions are responsible for initiation of loop formation. It was also found that the individual kinks are reflected as clear peaks in the principal modes of the analyzed trajectory of protein A, the shapes of which resemble the directional derivatives of the kinks along the chain. These observations suggest that the kinks of the DNLS equation determine the functionally important motions of proteins.

PRIMO: A Transferable Coarse-grained Force Field for Proteins

We describe here the PRIMO (PRotein Intermediate Model) force field, a physics-based fully transferable additive coarse-grained potential energy function that is compatible with an all-atom force field for multi-scale simulations. The energy function consists of standard molecular dynamics energy terms plus a hydrogen-bonding potential term and is mainly parameterized based on the CHARMM22/CMAP force field in a bottom-up fashion. The solvent is treated implicitly via the generalized Born model. The bonded interactions are either harmonic or distance-based spline interpolated potentials. These potentials are defined on the basis of all-atom molecular dynamics (MD) simulations of dipeptides with the CHARMM22/CMAP force field. The non-bonded parameters are tuned by matching conformational free energies of diverse set of conformations with that of CHARMM all-atom results. PRIMO is designed to provide a correct description of conformational distribution of the backbone (ϕ/ψ) and side chains (χ1) for all amino acids with a CMAP correction term. The CMAP potential in PRIMO is optimized based on the new CHARMM C36 CMAP. The resulting optimized force field has been applied in MD simulations of several proteins of 36-155 amino acids and shown that the root-mean-squared-deviation of the average structure from the corresponding crystallographic structure varies between 1.80 and 4.03 Å. PRIMO is shown to fold several small peptides to their native-like structures from extended conformations. These results suggest the applicability of the PRIMO force field in the study of protein structures in aqueous solution, structure predictions as well as ab initio folding of small peptides.

Extension of UNRES force field to treat polypeptide chains with D-amino-acid residues

Coarse-grained force fields for protein simulations are usually designed and parameterized to treat proteins composed of natural L-amino-acid residues. However, D-amino-acid residues occur in bacterial, fungal (e.g., gramicidins), as well as human-designed proteins. For this reason, we have extended the UNRES coarse-grained force field developed in our laboratory to treat systems with D-amino-acid residues. We developed the respective virtual-bond-torsional and double-torsional potentials for rotation about the C α · · · C α virtual-bond axis and two consecutive C α · · · C α virtual-bond axes, respectively, as functions of virtual-bond-dihedral angles γ. In turn, these were calculated as potentials of mean force (PMFs) from the diabatic energy surfaces of terminally-blocked model compounds for glycine, alanine, and proline. The potential-energy surfaces were calculated by using the ab initio method of molecular quantum mechanics at the Møller-Plesset (MP2) level of theory and the 6-31G(d,p) basis set, with the rotation angles of the peptide groups about [Formula: see text] and [Formula: see text] used as variables, and the energy was minimized with respect to the remaining degrees of freedom. The PMFs were calculated by numerical integration for all pairs and triplets with all possible combinations of types (glycine, alanine, and proline) and chirality (D or L); however, symmetry relations reduce the number of non-equivalent torsional potentials to 13 and the number of double-torsional potentials to 63 for a given C-terminal blocking group. Subsequently, one- (for torsional) and two-dimensional (for double-torsional potentials) Fourier series were fitted to the PMFs to obtain analytical expressions. It was found that the torsional potentials of the x-Y and X-y types, where X and Y are Ala or Pro, respectively, and a lowercase letter denotes D-chirality, have global minima for small absolute values of γ, accounting for the double-helical structure of gramicidin A, which is a dimer of two chains, each possessing an alternating D-Tyr-L-Tyr sequence, and similar peptides. The side-chain and correlation potentials for D-amino-acid residues were obtained by applying the reflection about the [Formula: see text] plane to the respective potentials for the L-amino-acid residues.

Determination of effective potentials for the stretching of C(α) ⋯ C(α) virtual bonds in polypeptide chains for coarse-grained simulations of proteins from ab initio energy surfaces of N-methylacetamide and N-acetylpyrrolidine

The potentials of mean force (PMF's) for the deformation of the C(α) ⋯ C(α) virtual bonds in polypeptide chains were determined from the diabatic energy surfaces of N-methylacetamide (modeling regular peptide groups) and N-acetylpyrrolidine (modeling the peptide groups preceding proline), calculated at the Møller-Plesset (MP2) ab initio level of theory with the 6-31G(d,p) basis set. The energy surfaces were expressed in the C(α) ⋯ C(α) virtual-bond length (d) and the H-N-C(α) ⋯ C' improper dihedral angle (α) that describes the pyramidicity of the amide nitrogen, or in the C(α)-C'(O)-N-C(α) dihedral angle (ω) and the angle α. For each grid point, the potential energy was minimized with respect to all remaining degrees of freedom. The PMF's obtained from the (d, α) energy surfaces produced realistic free-energy barriers to the trans-cis transition (10 kcal/mol and 13 kcal/mol for the regular and proline peptide groups, respectively, compared to 12.6 - 13.9 kcal/mol and 17.3 - 19.6 kcal/mol determined experimentally for glycylglycine and N-acylprolines, respectively), while those obtained from the (ω, α) energy maps produced either low-quality PMF curves when direct Boltzmann summation was implemented to compute the PMF's or too-flat curves with too-low free-energy barriers to the trans-cis transition if harmonic extrapolation was used to estimate the contributions to the partition function. An analytical bimodal logarithmic-Gaussian expression was fitted to the PMF's, and the potentials were implemented in the UNRES force field. Test Langevin-dynamics simulations were carried out for the Gly-Gly and Gly-Pro dipeptides, which showed a 10(6)-fold increase of the simulated rate of the trans-cis isomerization with respect to that measured experimentally; effectively the same result was obtained with the analytical Kramers theory of reaction rate applied to the UNRES representation of the peptide groups. Application of Kramers' theory to compute the rate constants from the all-atom ab initio energy surfaces of the model compounds studied resulted in isomerization rates close to the experimental values, which demonstrates that the increase of the isomerization rate in UNRES simulations results solely from averaging out the secondary degrees of freedom.

Velocity-scaling optimized replica exchange molecular dynamics of proteins in a hybrid explicit/implicit solvent

We propose a scheme for replica exchange molecular dynamics of proteins in explicit solvent that minimizes the number of required replicas using velocity rescaling. Our approach relies on a hybrid method where the protein evolves at each temperature in an explicit solvent, but replica exchange moves utilize an implicit solvent term. The two terms are coupled through the velocity rescaling. We test the efficiency of this approach for a common test case, the trp-cage protein.

A study of the α-helical intermediate preceding the aggregation of the amino-terminal fragment of the β amyloid peptide (Aβ(1-28))

The β amyloid (Aβ) peptide aggregates to form β-rich structures that are known to trigger Alzheimer's disease. Experiments suggest that an α-helical intermediate precedes the formation of these aggregates. However, a description at the molecular level of the α-to-β transition has not been obtained. Because it has been proposed that the transition might be initiated in the amino-terminal region of Aβ, we studied the aggregation of the 28-residue amino-terminal fragment of Aβ (Aβ(1-28)) using molecular dynamics and a coarse-grained force field. Simulations starting from extended and helical conformations showed that oligomerization is initiated by the formation of intermolecular β-sheets between the residues in the N-terminal regions. In simulations starting from the α-helical conformation, forcing residues 17-21 to remain in the initial (helical) conformation prevents aggregation but allows for the formation of dimers, indicating that oligomerization, initiated along the nonhelical N-terminal regions, cannot progress without the α-to-β transition propagating along the chains.

Coarse-grained force field: general folding theory

We review the coarse-grained UNited RESidue (UNRES) force field for the simulations of protein structure and dynamics, which is being developed in our laboratory over the last several years. UNRES is a physics-based force field, the prototype of which is defined as a potential of mean force of polypeptide chains in water, where all the degrees of freedom except the coordinates of α-carbon atoms and side-chain centers have been integrated out. We describe the initial implementation of UNRES to protein-structure prediction formulated as a search for the global minimum of the potential-energy function and its subsequent molecular dynamics and extensions of molecular-dynamics implementation, which enabled us to study protein-folding pathways and thermodynamics, as well as to reformulate the protein-structure prediction problem as a search for the conformational ensemble with the lowest free energy at temperatures below the folding-transition temperature. Applications of UNRES to study biological problems are also described.

Simple physics-based analytical formulas for the potentials of mean force of the interaction of amino-acid side chains in water. VI. Oppositely charged side chains

The two-site coarse-grained model for the interactions of charged side chains, to be used with our coarse-grained UNRES force field for protein simulations proposed in the accompanying paper, has been extended to pairs of oppositely charged side chains. The potentials of mean force of four pairs of molecules modeling charged amino-acid side chains, i.e., propionate-n-pentylamine cation (for aspartic acid-lysine), butyrate-n-pentylamine cation (for glutamic acid-lysine), propionate-1-butylguanidine (for aspartic acid-arginine), and butyrate-1-butylguanidine (for glutamic acid-arginine) pairs were determined by umbrella-sampling molecular dynamics simulations in explicit water as functions of distance and orientation, and the analytical expression was fitted to the potentials of mean force. Compared to pairs of like-charged side chains discussed in the accompanying paper, an average quadrupole-quadrupole interaction term had to be introduced to reproduce the Coulombic interactions, and a multistate model of charge distribution had to be introduced to fit the potentials of mean force of all oppositely charged pairs well. The model reproduces all salt-bridge minima and, consequently, is likely to improve the performance of the UNRES force field.

Simple physics-based analytical formulas for the potentials of mean force of the interaction of amino-acid side chains in water. V. Like-charged side chains

A new model of side-chain-side-chain interactions for charged side-chains of amino acids, to be used in the UNRES force-field, has been developed, in which a side chain consists of a nonpolar and a charged site. The interaction energy between the nonpolar sites is composed of a Gay-Berne and a cavity term; the interaction energy between the charged sites consists of a Lennard-Jones term, a Coulombic term, a generalized-Born term, and a cavity term, while the interaction energy between the nonpolar and charged sites is composed of a Gay-Berne and a polarization term. We parametrized the energy function for the models of all six pairs of natural like-charged amino-acid side chains, namely propionate-propionate (for the aspartic acid-aspartic acid pair), butyrate-butyrate (for the glutamic acid-glutamic acid pair), propionate-butyrate (for the aspartic acid-glutamic acid pair), pentylamine cation-pentylamine cation (for the lysine-lysine pair), 1-butylguanidine cation-1-butylguanidine cation (for the arginine-arginine pair), and pentylamine cation-1-butylguanidine cation (for the lysine-arginine pair). By using umbrella-sampling molecular dynamics simulations in explicit TIP3P water, we determined the potentials of mean force of the above-mentioned pairs as functions of distance and orientation and fitted analytical expressions to them. The positions and depths of the contact minima and the positions and heights of the desolvation maxima, including their dependence on the orientation of the molecules were well represented by analytical expressions for all systems. The values of the parameters of all the energy components are physically reasonable, which justifies use of such potentials in coarse-grain protein-folding simulations.

Mechanism of fiber assembly: treatment of Aβ peptide aggregation with a coarse-grained united-residue force field

The growth mechanism of β-amyloid (Aβ) peptide fibrils was studied by a physics-based coarse-grained united-residue model and molecular dynamics (MD) simulations. To identify the mechanism of monomer addition to an Aβ(1-40) fibril, we placed an unstructured monomer at a distance of 20 Å from a fibril template and allowed it to interact freely with the latter. The monomer was not biased towards fibril conformation by either the force field or the MD algorithm. With the use of a coarse-grained model with replica-exchange molecular dynamics, a longer timescale was accessible, making it possible to observe how the monomers probe different binding modes during their search for the fibril conformation. Although different assembly pathways were seen, they all follow a dock-lock mechanism with two distinct locking stages, consistent with experimental data on fibril elongation. Whereas these experiments have not been able to characterize the conformations populating the different stages, we have been able to describe these different stages explicitly by following free monomers as they dock onto a fibril template and to adopt the fibril conformation (i.e., we describe fibril elongation step by step at the molecular level). During the first stage of the assembly ("docking"), the monomer tries different conformations. After docking, the monomer is locked into the fibril through two different locking stages. In the first stage, the monomer forms hydrogen bonds with the fibril template along one of the strands in a two-stranded β-hairpin; in the second stage, hydrogen bonds are formed along the second strand, locking the monomer into the fibril structure. The data reveal a free-energy barrier separating the two locking stages. The importance of hydrophobic interactions and hydrogen bonds in the stability of the Aβ fibril structure was examined by carrying out additional canonical MD simulations of oligomers with different numbers of chains (4-16 chains), with the fibril structure as the initial conformation. The data confirm that the structures are stabilized largely by hydrophobic interactions and show that intermolecular hydrogen bonds are highly stable and contribute to the stability of the oligomers as well.