Optimization of protein force-field parameters with the Protein Data Bank

Quantum chemical studies on hydrogen bonds in helical secondary structures

We present a brief review of our recent computational studies of hydrogen bonds (H-bonds) in helical secondary structures of proteins, α-helix and 3₁₀-helix, using a Negative Fragmentation Approach with density functional theory. We found that the depolarized electronic structures of the carbonyl oxygen of the ith residue and the amide hydrogen of the (i + 4)th residue cause weaker H-bond in an α-helix than in an isolated H-bond. Our calculations showed that the H-bond energies in the 3₁₀-helix were also weaker than those of the isolated H-bonds. In the 3₁₀-helices, the adjacent N-H group at the (i + 1)th residue was closer to the C=O group of the H-bond pair than the adjacent C=O group in the 3₁₀-helices, whereas the adjacent C=O group at the (i + 1)th residue was close to the H-bond acceptor in α-helices. Therefore, the destabilization of the H-bond is attributed to the depolarization caused by the adjacent residue of the helical backbone connecting the H-bond donor and acceptor. The differences in the change in electron density revealed that such depolarizations were caused by the local electronic interactions in their neighborhood inside the helical structure and redistributed the electron density. We also present the improvements in the force field of classical molecular simulation, based on our findings.

Supplementary Information

The online version contains supplementary material available at 10.1007/s12551-022-01034-5.

Protein structure predictions by enhanced conformational sampling methods

In this Special Festschrift Issue for the celebration of Professor Nobuhiro Gō's 80th birthday, we review enhanced conformational sampling methods for protein structure predictions. We present several generalized-ensemble algorithms such as multicanonical algorithm, replica-exchange method, etc. and parallel Monte Carlo or molecular dynamics method with genetic crossover. Examples of the results of these methods applied to the predictions of protein tertiary structures are also presented.

Determination of the structural ensemble of the molten globule state of a protein by computer simulations

We have used computer simulations to investigate the structural nature of the molten globule (MG) state of canine milk lysozyme. To sample the conformational space efficiently, we performed replica-exchange umbrella sampling simulations with the radius of gyration as a reaction coordinate. We applied the Weighted Histogram Analysis Method to the trajectory of the simulations to obtain the potential of mean force, from which we identified representative structures corresponding to local minima in the free energy surface. The representative structures obtained in this way are in accord with the characteristics of the MG state reported previously by experimental studies. We conjecture that the MG state comprises a series of partially structured states undergoing relatively fast conformational interchange.

Salt effects on hydrophobic-core formation in folding of a helical miniprotein studied by molecular dynamics simulations

We have investigated effects of salt ions on folding events of a helical miniprotein chicken villin headpiece subdomain HP36. Low concentrations of ions alter electrostatic interactions between charged groups of a protein and can change the populations of conformers. Here, we compare two data sets of folding simulations of HP36 in explicit water solvent with or without ions. For efficient sampling of the conformational space of HP36, the multicanonical replica-exchange molecular dynamics method was employed. Our analyses suggest that salt alters salt-bridging nature of the protein at later stages of folding at room temperature. Especially, more nonnative, nonlocal salt bridges are formed at near-native conformations in pure water. Our analyses also show that such salt-bridge formation hinders the fully native hydrophobic-core packing at the final stages of folding.

Amino-acid-dependent main-chain torsion-energy terms for protein systems

Many commonly used force fields for protein systems such as AMBER, CHARMM, GROMACS, OPLS, and ECEPP have amino-acid-independent force-field parameters for main-chain torsion-energy terms. Here, we propose a new type of amino-acid-dependent torsion-energy terms in the force fields. As an example, we applied this approach to AMBER ff03 force field and determined new amino-acid-dependent parameters for ψ (N-C(α)-C-N) and ζ (C(β)-C(α)-C-N) angles for each amino acid by using our optimization method, which is one of the knowledge-based approach. In order to test the validity of the new force-field parameters, we then performed folding simulations of α-helical and β-hairpin peptides, using the optimized force field. The results showed that the new force-field parameters gave structures more consistent with the experimental implications than the original AMBER ff03 force field.

Coulomb replica-exchange method: handling electrostatic attractive and repulsive forces for biomolecules

We propose a new type of the Hamiltonian replica-exchange method (REM) for molecular dynamics (MD) and Monte Carlo simulations, which we refer to as the Coulomb REM (CREM). In this method, electrostatic charge parameters in the Coulomb interactions are exchanged among replicas while temperatures are exchanged in the usual REM. By varying the atom charges, the CREM overcomes free-energy barriers and realizes more efficient sampling in the conformational space than the REM. Furthermore, this method requires only a smaller number of replicas because only the atom charges of solute molecules are used as exchanged parameters. We performed Coulomb replica-exchange MD simulations of an alanine dipeptide in explicit water solvent and compared the results with those of the conventional canonical, replica exchange, and van der Waals REMs. Two force fields of AMBER parm99 and AMBER parm99SB were used. As a result, the CREM sampled all local-minimum free-energy states more frequently than the other methods for both force fields. Moreover, the Coulomb, van der Waals, and usual REMs were applied to a fragment of an amyloid-β peptide (Aβ) in explicit water solvent to compare the sampling efficiency of these methods for a larger system. The CREM sampled structures of the Aβ fragment more efficiently than the other methods. We obtained β-helix, α-helix, 3(10)-helix, β-hairpin, and β-sheet structures as stable structures and deduced pathways of conformational transitions among these structures from a free-energy landscape.

PROTEIN FORCE-FIELD PARAMETERS OPTIMIZED WITH THE PROTEIN DATA BANK II: COMPARISONS OF FORCE FIELDS BY FOLDING SIMULATIONS OF SHORT PEPTIDES

In Paper I of this series, the formulations of the optimization method of existing force-field parameters for protein systems have been presented. We then applied it to five sets of force-field parameters, namely, AMBER parm94, AMBER parm96, AMBER parm99, CHARMM version 22, and OPLS-AA. In order to test the validity of these force fields, the folding simulations of α-helical and β-hairpin peptides have been performed with each of the original and optimized force-field parameters. We found that all five modified force-field parameters gave both α-helical and β-hairpin structures more consistent with the experimental implications than the original force fields.

PROTEIN FORCE-FIELD PARAMETERS OPTIMIZED WITH THE PROTEIN DATA BANK I: FORCE-FIELD OPTIMIZATIONS

We optimized five existing sets of force-field parameters for protein systems by our recently proposed method. The five force fields are AMBER parm94, AMBER parm96, AMBER parm99, CHARMM version 22, and OPLS-AA. The method consists of minimizing the sum of the square of the force acting on each atom in the proteins with the structures from the Protein Data Bank (PDB). We selected the partial-charge and backbone torsion-energy parameters for this optimization, and 100 molecules from the PDB were used. We gave detailed comparisons of the optimized force fields and found that there is a tendency of convergence towards the same function for the torsion-energy term.

Ab initio prediction of protein-ligand binding structures by replica-exchange umbrella sampling simulations

We have developed a prediction method for the binding structures of ligands with proteins. Our method consists of three steps. First, replica-exchange umbrella sampling simulations are performed along the distance between a putative binding site of a protein and a ligand as the reaction coordinate. Second, we obtain the potential of mean force (PMF) of the unbiased system using the weighted histogram analysis method and determine the distance that corresponds to the global minimum of PMF. Third, structures that have this global-minimum distance and energy values around the average potential energy are collected and analyzed using the principal component analysis. We predict the binding structure as the global-minimum free energy state on the free energy landscapes along the two major principal component axes. As test cases, we applied our method to five protein-ligand complex systems. Starting from the configuration in which the protein and the ligand are far away from each other in each system, our method predicted the ligand binding structures in excellent agreement with the experimental data from Protein Data Bank.

Balance between alpha and beta structures in ab initio protein folding

Despite initial successes in folding of proteins by molecular simulation, it is becoming increasingly evident that current energy functions (force fields) tend to favor either alpha or beta secondary structure, such that the choice of force field is governed by the structural class of the protein. Here, we study the folding of peptides with either predominantly alpha (Trp cage) or beta (GB1 hairpin) structure with a modified version of the Amber ff03 force field, optimized to reproduce structural propensity in a helix-forming peptide. Using extensive replica exchange molecular dynamics simulations starting from completely unfolded configurations, we obtain the correct folded structure for each peptide, in close agreement with the experimental native structure (<1.5 A all-atom root-mean-square deviation). We obtain converged equilibrium distributions, with folded populations at standard conditions (approximately 300 K), in remarkable accord with experiment. Further comparison to experimental data from NMR spectroscopy and FRET suggests that although the folded structures are accurately reproduced, the unfolded state remains too structured and compact. Our results suggest that the backbone correction results in a force field that is transferable to the folding of proteins from different structural classes.

A test on peptide stability of AMBER force fields with implicit solvation

We used replica exchange molecular dynamics (REMD) simulations to evaluate four different AMBER force fields and three different implicit solvent models. Our aim was to determine if these physics-based models captured the correct secondary structures of two alpha-helical and two beta-peptides: the 14-mer EK helix of Baldwin and co-workers, the C-terminal helix of ribonuclease, the 16-mer C-terminal hairpin of protein G, and the trpzip2 miniprotein. The different models gave different results, but generally we found that AMBER ff96 plus the implicit solvent model of Onufriev, Bashford, and Case gave reasonable structures, and is fairly well-balanced between helix and sheet. We also observed differences in the strength of ion pairing in the solvent models, we but found that the native secondary structures were retained even when salt bridges were prevented in the conformational sampling. Overall, this work indicates that some of these all-atom physics-based force fields may be good starting points for protein folding and protein structure prediction.

Consistent free energy landscapes and thermodynamic properties of small proteins based on a single all-atom force field employing an implicit solvation

We have attempted to improve the PARAM99 force field in conjunction with the generalized Born (GB) solvation model with a surface area correction for more consistent protein folding simulations. For this purpose, using an extended alphabeta training set of five well-studied molecules with various folds (alpha, beta, and betabetaalpha), a previously modified version of PARAM99/GBSA is further refined, such that all native states of the five training species correspond to their lowest free energy minimum states. The resulting modified force field (PARAM99MOD5/GBSA) clearly produces reasonably acceptable conformational free energy surfaces of the training set with correct identifications of their native states in the free energy minimum states. Moreover, due to its well-balanced nature, this new force field is expected to describe secondary structure propensities of diverse folds in a more consistent manner. Remarkably, temperature dependent behaviors simulated with the current force field are in good agreement with the experiment. This agreement is a significant improvement over the existing standard all-atom force fields. In addition, fundamentally important thermodynamic quantities, such as folding enthalpy (DeltaH) and entropy (DeltaS), agree reasonably well with the experimental data.

Cooperative folding mechanism of a β-hairpin peptide studied by a multicanonical replica-exchange molecular dynamics simulation

G-peptide is a 16-residue peptide of the C-terminal end of streptococcal protein G B1 domain, which is known to fold into a specific beta-hairpin within 6 micros. Here, we study molecular mechanism on the stability and folding of G-peptide by performing a multicanonical replica-exchange (MUCAREM) molecular dynamics simulation with explicit solvent. Unlike the preceding simulations of the same peptide, the simulation was started from an unfolded conformation without any experimental information on the native conformation. In the 278-ns trajectory, we observed three independent folding events. Thus MUCAREM can be estimated to accelerate the folding reaction more than 60 times than the conventional molecular dynamics simulations. The free-energy landscape of the peptide at room temperature shows that there are three essential subevents in the folding pathway to construct the native-like beta-hairpin conformation: (i) a hydrophobic collapse of the peptide occurs with the side-chain contacts between Tyr45 and Phe52, (ii) then, the native-like turn is formed accompanying with the hydrogen-bonded network around the turn region, and (iii) finally, the rest of the backbone hydrogen bonds are formed. A number of stable native hydrogen bonds are formed cooperatively during the second stage, suggesting the importance of the formation of the specific turn structure. This is also supported by the accumulation of the nonnative conformations only with the hydrophobic cluster around Tyr45 and Phe52. These simulation results are consistent with high phi-values of the turn region observed by experiment.

Importance of the Single Amino Acid Potential in Water for Secondary and Tertiary Structures of Proteins

The structure of a protein molecule is considered to be primarily determined by the inter-amino-acid nonbonded interactions, such as hydrogen bonds. However, the conformational space of the polypeptide chain should be simultaneously restricted by the intrinsic conformational preferences of the individual amino acids. We present here precise single amino acid potential (SAAP) surfaces for glycine (For-Gly-NH(2)) and alanine (For-Ala-NH(2)) in water (epsilon = 78.39) and ether (epsilon = 4.335), which were calculated at the HF/6-31+G(d,p) level applying the self-consistent isodensity polarizable continuum model (SCIPCM) reaction field with geometry optimization in the corresponding solvents. The obtained Ramachandran potential surfaces in water showed distinct potential wells in the alpha- and beta-regions. The profiles were in almost perfect agreement with the Ramachandran plots of glycine and alanine residues in folded proteins, suggesting the Boltzmann distributions on the SAAP surfaces. Molecular simulations of polyalanines (For-Ala(n)-NH(2); n = 3-5) by using the SAAP force field equipped with the SCIPCM potentials revealed that the polyalanines readily form 3(10)-helical structures in water but not in vacuo. In ether (hydrophobic environments), the helical structures were relatively stable, but the most stable structure was assigned to a different one. These results indicated that the intrinsic conformational preferences of the individual amino acids (i.e., the SAAPs) in water are of significant importance not only for describing conformations of a polypeptide chain in the random coil state but also for understanding the folding to the secondary and tertiary structures.

Strike a balance: optimization of backbone torsion parameters of AMBER polarizable force field for simulations of proteins and peptides

Based on the AMBER polarizable model (ff02), we have re-optimized the parameters related to the main-chain (Phi, Psi) torsion angles by fitting to the Boltzmann-weighted average quantum mechanical (QM) energies of the important regions (i.e., beta, P(II), alpha(R), and alpha(L) regions). Following the naming convention of the AMBER force field series, this release will be called ff02pol.rl The force field has been assessed both by energetic comparison against the QM data and by the replica exchange molecular dynamics simulations of short alanine peptides in water. For Ace-Ala-Nme, the simulated populations in the beta, P(II) and alpha(R) regions were approximately 30, 43, and 26%, respectively. For Ace-(Ala)(7)-Nme, the populations in these three regions were approximately 24, 49, and 26%. Both were in qualitative agreement with the NMR and CD experimental conclusions. In comparison with the previous force field, ff02pol.rl demonstrated good balance among these three important regions. The optimized torsion parameters, together with those in ff02, allow us to carry out simulations on proteins and peptides with the consideration of polarization.

Free energy surfaces of miniproteins with a ββα motif: Replica exchange molecular dynamics simulation with an implicit solvation model

Designed miniproteins with a betabetaalpha motif, such as BBA5, 1FSD, and 1PSV can serve as a benchmark set to test the validity of all-atom force fields with computer simulation, because they contain all the basic structural elements in protein folding. Unfortunately, it was found that the standard all-atom force fields with the generalized Born (GB) implicit solvation model tend to produce distorted free energy surfaces for the betabetaalpha proteins, not only because energetically those proteins need to be described by more balanced weights of the alpha- and beta-strands, but also because the GB implicit solvation model suffers from overestimated salt bridge effects. In an attempt to resolve these problems, we have modified one of the standard all-atom force fields in conjunction with the GB model, such that each native state of the betabetaalpha proteins is in its free energy minimum state with reasonable energy barriers separating local minima. With this modified energy model, the free energy contour map in each protein was constructed from the replica exchange molecular dynamics REMD simulation. The resulting free energy surfaces are significantly improved in comparison with previous simulation results and consistent with general views on small protein folding behaviors with realistic topology and energetics of all three proteins.

Molecular mechanism for stabilizing a short helical peptide studied by generalized-ensemble simulations with explicit solvent

We study the folding mechanism of an analog of the C-peptide of ribonuclease A in explicit water by a replica-exchange multicanonical molecular dynamics simulation based on all-atom models. The multicanonical weight factor was determined by the combined use of the multicanonical replica-exchange method and the replica-exchange multicanonical algorithm. Using statistically reliable data thus obtained, we have examined the free-energy landscape of the peptide system. The global-minimum free-energy state in the landscape at room temperature has an alpha-helix structure with a distortion near the N-terminus. The state also has a salt bridge between Glu(-)-2 and Arg(+)-10 and an aromatic-aromatic interaction between Phe-8 and His(+)-12, both of which have been observed in x-ray and other experimental measurements. Principal component analysis clearly shows the different roles of these side-chain interactions in the peptide folding. The side-chain interaction between Phe-8 and His(+)-12 greatly enhances the stability of helical structure toward the C-terminal end, whereas the salt bridge between Glu(-)-2 and Arg(+)-10 mainly works as a restraint to prevent the alpha-helix structure from extending to the N-terminus. The free-energy landscape of C-peptide reveals a funnel-like shape where all of these interactions consistently exist only in the global-minimum state. This is the major reason why the native structure of the short helical peptide shows significant stability at low temperatures.