A heteropolymer model study for the mechanism of protein folding

A heteropolymer model of randomly self-interacting chains in two dimensions is studied with numerical simulations in order to elucidate the folding mechanism of protein. We find that the model occasionally shows folding propensity depending on the sequence of random numbers given to the chain. We study the thermodynamic and kinematic roles in the folding mechanism by grouping the local energy minima found in the simulations into clusters according to the similarity of their conformations. It is suggested that the local minima to which some heteropolymers show a folding tendency are always the lowest energy states of the energy spectrum within a cluster, though which cluster is selected depends on the sequence. For the eight random sequences we study, we find that the energy gap between the ground state and excited states is little correlated with folding or nonfolding. We rather find that folding propensities are correlated with the global structure of the average energy surface, implying a dominant kinetic role in the folding mechanism, although thermal factors cannot be ignored as the mechanism of choosing the ground state within a cluster of states connected by small deformations. We suggest that a hierarchical cluster structure plays an important role in selecting a unique folded state out of the huge number of local minima of heteropolymers. © 1997 John Wiley & Sons, Inc.     

Topological properties of the energy landscape of small peptides

The topology of the potential energy landscape (PEL) underlying the dynamics of a two dimensional off-lattice model for a heteropolymer is analyzed for different sequences of amino-acids. A statistical characterization of the metastable minima and first-order saddles of the PEL highlights structural differences in the landscape of good and bad folding sequences and provides insight on the chain dynamics during folding.     

Free energy landscape and folding mechanism of a beta-hairpin in explicit water: a replica exchange molecular dynamics study

The free energy landscape and the folding mechanism of the C-terminal beta-hairpin of protein G is studied by extensive replica exchange molecular dynamics simulations (40 replicas and 340 ns total simulation time), using the GROMOS96 force field and the SPC explicit water solvent. The study reveals that the system preferentially adopts a beta-hairpin structure at biologically important temperatures, and that the helix content is low at all temperatures studied. Representing the free energy landscape as a function of several types of reaction coordinates, four local minima corresponding to the folded, partially folded, molten globule, and unfolded states are identified. The findings suggest that the folding of the beta-hairpin occurs as the sequence: collapse of hydrophobic core --> formation of H-bond --> formation of the turn. Identifying the folded and molten globule states as the main conformations, the free energy landscape of the beta-hairpin is consistent with a two-state behavior with a broad transition state. The temperature dependence of the folding-unfolding transition is investigated in some detail. The enthalpy and entropy jumps at the folding transition temperature are found to be about three times lower than the experimental estimates, indicating that the folding-unfolding transition in silico is less cooperative than its in vitro counterpart.     

Molecular basis of cooperativity in protein folding IV. Core: A general cooperative folding model

The cooperative nature of the protein folding process is independent of the characteristic fold and the specific secondary structure attributes of a globular protein. A general folding/unfolding model should, therefore, be based upon structural features that transcend the peculiarities of α-helices, β-sheets, and other structural motifs found in proteins. The studies presented in this paper suggest that a single structural characteristic common to all globular proteins is essential for cooperative folding. The formation of a partly folded state from the native state results in the exposure to solvent of two distinct regions: (1) the portions of the protein that are unfolded; and (2) the “complementary surfaces,” located in the regions of the protein that remain folded. The cooperative character of the folding/unfolding transition is determined largely by the energetics of exposing complementary surface regions to the solvent. By definition, complementary regions are present only in partly folded states; they are absent from the native and unfolded states. An unfavorable free energy lowers the probability of partly folded states and increases the cooperativity of the transition. In this paper we present a mathematical formulation of this behavior and develop a general cooperative folding/unfolding model, termed the “complementary region” (CORE) model. This model successfully reproduces the main properties of folding/unfolding transitions without limiting the number of partly folded states accessible to the protein, thereby permitting a systematic examination of the structural and solvent conditions under which intermediates become populated. It is shown that the CORE model predicts two-state folding/unfolding behavior, even though the two-state character is not assumed in the model. © 1993 Wiley-Liss, Inc.     

Microscopic reversibility of protein folding in molecular dynamics simulations of the engrailed homeodomain

The principle of microscopic reversibility states that at equilibrium the number of molecules entering a state by a given path must equal those exiting the state via the same path under identical conditions or, in structural terms, that the conformations along the two pathways are the same. There has been some indirect evidence indicating that protein folding is such a process, but there have been few conclusive findings. In this study, we performed molecular dynamics simulations of an ultrafast unfolding and folding protein at its melting temperature to observe, on an atom-by-atom basis, the pathways the protein followed as it unfolded and folded within a continuous trajectory. In a total of 0.67 micros of simulation in water, we found six transient denaturing events near the melting temperature (323 and 330 K) and an additional refolding event following a previously identified unfolding event at a high temperature (373 K). In each case, unfolding and refolding transition state ensembles were identified, and they agreed well with experiment on the basis of a comparison of S and Phi values. On the basis of several structural properties, these 13 transition state ensembles agreed very well with each other and with four previously identified transition states from high-temperature denaturing simulations. Thus, not only were the unfolding and refolding transition states part of the same ensemble, but in five of the seven cases, the pathway the protein took as it unfolded was nearly identical to the subsequent refolding pathway. These events provide compelling evidence that protein folding is a microscopically reversible process. In the other two cases, the folding and unfolding transition states were remarkably similar to each other but the paths deviated.     

All-atom Monte Carlo simulation of GCAA RNA folding

We report a detailed all-atom simulation of the folding of the GCAA RNA tetraloop. The GCAA tetraloop motif is a very common and thermodynamically stable secondary structure in natural RNAs. We use our simulation methods to study the folding behavior of a 12-base GCAA tetraloop structure with a four-base helix adjacent to the tetraloop proper. We implement an all-atom Monte Carlo (MC) simulation of RNA structural dynamics using a Go potential. Molecular dynamics (MD) simulation of RNA and protein has realistic energetics and sterics, but is extremely expensive in terms of computational time. By coarsely treating non-covalent energetics, but retaining all-atom sterics and entropic effects, all-atom MC techniques are a useful method for the study of protein and now RNA. We observe a sharp folding transition for this structure, and in simulations at room temperature the state histogram shows three distinct minima: an unfolded state (U), a more narrow intermediated state (I), and a narrow folded state (F). The intermediate consists primarily of structures with the GCAA loop and some helix hydrogen bonds formed. Repeated kinetic folding simulations reveal that the number of helix base-pairs forms a simple 1D reaction coordinate for the I-->N transition.     

The folding pathway of ubiquitin from all-atom molecular dynamics simulations

The folding (unfolding) pathway of ubiquitin is probed using all-atom molecular dynamics simulations. We dissect the folding pathway using two techniques: first, we probe the folding pathway of ubiquitin by calculating the evolution of structural properties over time and second, we identify the rate determining transition state for folding. The structural properties that we look at are hydrophobic solvent accessible surface area (SASA) and Calpha-root-mean-square deviation (rmsd). These properties on their own tell us relatively little about the folding pathway of ubiquitin; however, when plotted against each other, they become powerful tools for dissecting ubiquitin's folding mechanism. Plots of Calpha-rmsd against SASA serve as a phase space trajectories for the folding of ubiquitin. In this study, these plots show that ubiquitin folds to the native state via the population of an intermediate state. This is shown by an initial hydrophobic collapse phase followed by a second phase of secondary structure arrangement. Analysis of the structure of the intermediate state shows that it is a collapsed species with very little secondary structure. In reconciling these observations with recent experimental data, the transition that we observe in our simulations from the unfolded state (U) to the intermediate state (I) most likely occurs in the dead-time of the stopped flow instrument. The folding pathway of ubiquitin is probed further by identification of the rate-determining transition state for folding. The method used for this is essential dynamics, which utilizes a principal component analysis (PCA) on the atomic fluctuations throughout the simulation. The five transition state structures identified in silico are in good agreement with the experimentally determined transition state. The calculation of phi-values from the structures generated in the simulations is also carried out and it shows a good correlation with the experimentally measured values. An initial analysis of the denatured state shows that it is compact with fluctuating regions of nonnative secondary structure. It is found that the compactness in the denatured state is due to the burial of some hydrophobic residues. We conclude by looking at a correlation between folding kinetics and residual structure in the denatured state. A hierarchical folding mechanism is then proposed for ubiquitin.     

Identifying the protein folding nucleus using molecular dynamics

Molecular dynamics simulations of folding in an off-lattice protein model reveal a nucleation scenario, in which a few well-defined contacts are formed with high probability in the transition state ensemble of conformations. Their appearance determines folding cooperativity and drives the model protein into its folded conformation. Amino acid residues participating in those contacts may serve as "accelerator pedals" used by molecular evolution to control protein folding rate.     

Long-range electrostatic effects on peptide folding.

The reversible folding/unfolding of a short peptide in solution is studied by molecular dynamics simulations. The effects of long-range electrostatic interactions are examined and found to be important both for the equilibrium between folded and unfolded states and the dynamics of the folding process. The neglect of long-range electrostatics leads to an increased population of unfolded states and increased structural fluctuations. When such interactions are taken into account, the peptide unfolds and folds to the experimentally determined structure several times during a 25 ns trajectory, with approximately equal populations of folded and unfolded states in the neighborhood of its proposed melting temperature. The effect of using spherical boundary conditions rather than periodic ones does not appear to have any major effect on the folding dynamics.     

Molecular basis of co-operativity in protein folding.

The folding/unfolding transition of proteins is a highly co-operative process characterized by the presence of very few or no thermodynamically stable partially folded intermediate states. The purpose of this paper is to present a thermodynamic formalism aimed at describing quantitatively the co-operative folding behavior of proteins. In order to account for this behavior, a hierarchical algorithm aimed at evaluating the folding/unfolding partition function has been developed. This formalism defines the partition function in terms of multiple levels of interacting co-operative folding units. A co-operative folding unit is defined as a protein structural element that exhibits two-state folding/unfolding behavior. At the most fundamental level are those structural elements that behave co-operatively as a result of purely local interactions. Higher-order co-operative folding units are formed through interactions between different structural elements. The hierarchical formalism utilizes the crystallographic structure of the protein as a template to generate partially folded conformations defined in terms of co-operative folding units. The Gibbs free energy of those states and their corresponding statistical weights are then computed using experimental energetic parameters determined calorimetrically. This formalism has been applied to the case of myoglobin. It is shown that the hierarchical partition function correctly predicts the presence, energetics and co-operativity of the heat and cold denaturation transitions. The major contribution to the co-operative folding behavior arises from the solvent exposure of non-polar residues located in regions complementary to those that have undergone unfolding. This entropically uncompensated and energetically unfavorable solvent exposure characterizes all partially folded states but not the unfolded state, thus minimizing the population of partially folded intermediates throughout the folding/unfolding transition.     

Computational assessment of folding energy landscapes in heterodimeric coiled coils

The coiled coil structural motif consists of alpha helices supercoiling around each other to form staggered knobs‐into‐holes packing. Such structures are deceptively simple, especially as they often can be described with parametric equations, but are known to exist in various conformations. Even the simplest systems, consisting of 2 monomers, can assemble into a wide range of states. They can form canonical as well as noncanonical coiled coils, be parallel or antiparallel, where helices associate with different degrees of shift, tilt, and rotation. Here, we investigate the energy landscape of heterodimeric coiled coils by carrying out de novo folding simulations starting from amino acid sequence. We folded a diverse set of 22 heterodimers and demonstrate that the approach is capable of identifying the atomic details in the experimental structure in the majority of cases. Our methodology also enables exploration of alternative states that can be accessible in solution beyond the experimentally determined structure. For many systems, we observe folding energy landscapes with multiple energy minima and several isoenergetic states. By comparing coiled coils from single domains and those extracted from larger proteins, we find that standalone coiled coils have deeper energy wells at the experimentally determined conformation. By folding the competing homodimeric states in addition to the heterodimers, we observe that the structural specificity towards the heteromeric state is often small. Taken together, our results demonstrate that de novo folding simulations can be a powerful tool to characterize structural specificity of coiled coils when coupled to assessment of energy landscapes.     

Application of the multiensemble sampling to the equilibrium folding of proteins

MOTIVATION: Conventional Monte Carlo and molecular dynamics simulations of proteins in the canonical ensemble are of little use, because they tend to get trapped in states of energy local minima at low temperatures. One way to surmount this difficulty is to use a non-Boltzmann sampling method in which conformations are sampled upon a general weighting function instead of the conventional Boltzmann weighting function. The multiensemble sampling (MES) method is a non-Boltzmann sampling method that was originally developed to estimate free energy differences between systems with different potential energies and/or at different thermodynamic states. The method has not yet been applied to studies of complex molecular systems such as proteins. RESULTS: MES Monte Carlo simulations of small proteins have been carried out using a united-residue force field. The proteins at several temperatures from the unfolded to the folded states were simulated in a single MC run at a time and their equilibrium thermodynamic properties were calculated correctly. The distributions of sampled conformations clearly indicate that, when going through states of energy local minima, the MES simulation did not get trapped in them but escaped from them so quickly that all the relevant parts of conformation space could be sampled properly. A two-step folding process consisting of a collapse transition followed by a folding transition is observed. This study demonstrates that the use of MES alleviates the multiple-minima problem greatly. AVAILABILITY: Available on request from the authors.     

Ten-microsecond molecular dynamics simulation of a fast-folding WW domain

All-atom molecular dynamics (MD) simulations of protein folding allow analysis of the folding process at an unprecedented level of detail. Unfortunately, such simulations have not yet reached their full potential both due to difficulties in sufficiently sampling the microsecond timescales needed for folding, and because the force field used may yield neither the correct dynamical sequence of events nor the folded structure. The ongoing study of protein folding through computational methods thus requires both improvements in the performance of molecular dynamics programs to make longer timescales accessible, and testing of force fields in the context of folding simulations. We report a ten-microsecond simulation of an incipient downhill-folding WW domain mutant along with measurement of a molecular time and activated folding time of 1.5 microseconds and 13.3 microseconds, respectively. The protein simulated in explicit solvent exhibits several metastable states with incorrect topology and does not assume the native state during the present simulations.     

The contribution of the residues from the main hydrophobic core of ribonuclease A to its pressure-folding transition state

The role of hydrophobic interactions established by the residues that belong to the main hydrophobic core of ribonuclease A in its pressure-folding transition state was investigated using the Phi-value method. The folding kinetics was studied using pressure-jump techniques both in the pressurization and depressurization directions. The ratio between the folding activation volume and the reaction volume (beta p-value), which is an index of the compactness or degree of solvation of the transition state, was calculated. All the positions analyzed presented fractional Phi f-values, and the lowest were those corresponding to the most critical positions for the ribonuclease A stability. The structure of the transition state of the hydrophobic core of ribonuclease A, from the point of view of formed interactions, is a relatively, uniformly expanded form of the folded structure with a mean Phi f-value of 0.43. This places it halfway between the folded and unfolded states. On the other hand, for the variants, the average of beta p-values is 0.4, suggesting a transition state that is 40% native-like. Altogether the results suggest that the pressure-folding transition state of ribonuclease A looks like a collapsed globule with some secondary structure and a weakened hydrophobic core. A good correlation was found between the Phi f-values and the Deltabeta p-values. Although the nature of the transition state inferred from pressure-induced folding studies and the results of the protein engineering method have been reported to be consistent for other proteins, to the best of our knowledge this is the first direct comparison using a set of mutants.     

Elimination of a misfolded folding intermediate by a single point mutation

We present an analysis of the folding behavior of the 159-residue major birch pollen allergen Bet v 1. The protein contains a water-filled channel running through it. Consequently, the protein has a hydrophobic shell, rather than a hydrophobic core. During the folding of the protein from either the urea-, pH-, or SDS-denatured state, Bet v 1 transiently populates a partially folded intermediate state. This state appears to be misfolded, since it has to unfold at least partially to fold to the native state. The misfolded intermediate is not, however, a result of the water-filled channel in Bet v 1. The intermediate completely disappears in the mutant Tyr --> Trp120, in which the channel is still present. Tyr120 appears to behave as a "negative gatekeeper" which attenuates efficient folding. The close structural homologue, the apple allergen Mal d 1, also folds without any detectable folding intermediates. However, the position of the transition state on the reaction coordinate, which is a measure of its overall compactness relative to the denatured and native states, is reduced dramatically from ca. 0.9 in Bet v 1 to around 0.5 in Mal d 1. We suggest that this large shift in the transition state structure is partly due to different local helix propensities. Given that individual mutations can have such large effects on folding, one should not a priori expect structurally homologous proteins to fold by the same mechanism.     

Multiscale Coarse-Graining of the Protein Energy Landscape

A variety of coarse-grained (CG) models exists for simulation of proteins. An outstanding problem is the construction of a CG model with physically accurate conformational energetics rivaling all-atom force fields. In the present work, atomistic simulations of peptide folding and aggregation equilibria are force-matched using multiscale coarse-graining to develop and test a CG interaction potential of general utility for the simulation of proteins of arbitrary sequence. The reduced representation relies on multiple interaction sites to maintain the anisotropic packing and polarity of individual sidechains. CG energy landscapes computed from replica exchange simulations of the folding of Trpzip, Trp-cage and adenylate kinase resemble those of other reduced representations; non-native structures are observed with energies similar to those of the native state. The artifactual stabilization of misfolded states implies that non-native interactions play a deciding role in deviations from ideal funnel-like cooperative folding. The role of surface tension, backbone hydrogen bonding and the smooth pairwise CG landscape is discussed. Ab initio folding aside, the improved treatment of sidechain rotamers results in stability of the native state in constant temperature simulations of Trpzip, Trp-cage, and the open to closed conformational transition of adenylate kinase, illustrating the potential value of the CG force field for simulating protein complexes and transitions between well-defined structural states.     

The protein folding network

The conformation space of a 20 residue antiparallel beta-sheet peptide, sampled by molecular dynamics simulations, is mapped to a network. Snapshots saved along the trajectory are grouped according to secondary structure into nodes of the network and the transitions between them are links. The conformation space network describes the significant free energy minima and their dynamic connectivity without requiring arbitrarily chosen reaction coordinates. As previously found for the Internet and the World-Wide Web as well as for social and biological networks, the conformation space network is scale-free and contains highly connected hubs like the native state which is the most populated free energy basin. Furthermore, the native basin exhibits a hierarchical organization, which is not found for a random heteropolymer lacking a predominant free-energy minimum. The network topology is used to identify conformations in the folding transition state (TS) ensemble, and provides a basis for understanding the heterogeneity of the TS and denatured state ensemble as well as the existence of multiple pathways.     

Structured pathway across the transition state for peptide folding revealed by molecular dynamics simulations

Small globular proteins and peptides commonly exhibit two-state folding kinetics in which the rate limiting step of folding is the surmounting of a single free energy barrier at the transition state (TS) separating the folded and the unfolded states. An intriguing question is whether the polypeptide chain reaches, and leaves, the TS by completely random fluctuations, or whether there is a directed, stepwise process. Here, the folding TS of a 15-residue β-hairpin peptide, Peptide 1, is characterized using independent 2.5 μs-long unbiased atomistic molecular dynamics (MD) simulations (a total of 15 μs). The trajectories were started from fully unfolded structures. Multiple (spontaneous) folding events to the NMR-derived conformation are observed, allowing both structural and dynamical characterization of the folding TS. A common loop-like topology is observed in all the TS structures with native end-to-end and turn contacts, while the central segments of the strands are not in contact. Non-native sidechain contacts are present in the TS between the only tryptophan (W11) and the turn region (P7-G9). Prior to the TS the turn is found to be already locked by the W11 sidechain, while the ends are apart. Once the ends have also come into contact, the TS is reached. Finally, along the reactive folding paths the cooperative loss of the W11 non-native contacts and the formation of the central inter-strand native contacts lead to the peptide rapidly proceeding from the TS to the native state. The present results indicate a directed stepwise process to folding the peptide.     

Essential dynamics of reversible peptide folding: memory-free conformational dynamics governed by internal hydrogen bonds

A principal component analysis has been applied on equilibrium simulations of a beta-heptapeptide that shows reversible folding in a methanol solution. The analysis shows that the configurational space contains only three dense sub-states. These states of relatively low free energy correspond to the "native" left-handed helix, a partly helical intermediate, and a hairpin-like structure. The collection of unfolded conformations form a relatively diffuse cloud with little substructure. Internal hydrogen-bonding energies were found to correlate well with the degree of folding. The native helical structure folds from the N terminus; the transition from the major folding intermediate to the native helical structure involves the formation of the two most C-terminal backbone hydrogen bonds. A four-state Markov model was found to describe transition frequencies between the conformational states within error limits, indicating that memory-effects are negligible beyond the nanosecond time-scale. The dominant native state fluctuations were found to be very similar to unfolding motions, suggesting that unfolding pathways can be inferred from fluctuations in the native state. The low-dimensional essential subspace, describing 69% of the collective atomic fluctuations, was found to converge at time-scales of the order of one nanosecond at all temperatures investigated, whereas folding/unfolding takes place at significantly longer time-scales, even above the melting temperature.