RESUMO
Affinity ranking of structurally diverse small-molecule ligands is a challenging problem with important applications in structure-based drug discovery. Absolute binding free energy methods can model diverse ligands, but the high computational cost of the current methods limits application to data sets with few ligands. We recently developed MELD-Bracket, a Molecular Dynamics method for efficient affinity ranking of ligands [ JCTC 2022, 18 (1), 374-379]. It utilizes a Bayesian framework to guide sampling to relevant regions of phase space, and it couples this with a bracket-like competition on a pool of ligands. Here we find that 6-competitor MELD-Bracket can rank dozens of diverse ligands that have low structural similarity and different net charges. We benchmark it on four protein systemsâPTB1B, Tyk2, BACE, and JAK3âhaving varied modes of interactions. We also validated 8-competitor and 12-competitor protocols. The MELD-Bracket protocols presented here may have the appropriate balance of accuracy and computational efficiency to be suitable for ranking diverse ligands from typical drug discovery campaigns.
Assuntos
Simulação de Dinâmica Molecular , Proteínas , Ligação Proteica , Teorema de Bayes , Proteínas/química , LigantesRESUMO
The composition dependent local environment of three organic dyes in binary mixtures of a room temperature ionic liquid (1-methyl-3-pentylimidazolium bromide, [pmim][Br]) and water is studied by fluorescence correlation spectroscopy (FCS) and molecular dynamics (MD) simulations. We used three dyes-neutral coumarin 480 (C480), anionic coumarin 343 (C343), and highly hydrophobic 4-(dicyanomethylene)-2-methyl-6-(p-dimethyl-aminostyryl)-4H-pyran (DCM)-to probe different environments in the binary mixtures. The heterogeneity of the [pmim][Br]-water mixture leads to multiple values (i.e., distribution) of diffusion coefficients (Dt). In these binary mixtures, the effective viscosity (ηeff, obtained from FCS) and the local concentration of the [pmim][Br] around the three dyes (revealed by MD simulations) are found to be quite different than that in bulk. The viscosity experienced by the C480 and C343 dyes is almost twice as large as that experienced by DCM dye. Through rigorous MD simulation, we show that in the vicinity of the less hydrophobic coumarin dyes (C480 and C343) compared to DCM dye, the local concentration of the [pmim][Br] is â¼3-7 times larger than that in bulk. In the case of the most hydrophobic dye, DCM, the local concentration of [pmim][Br] is almost similar to bulk-like. Further analysis reveals the formation of hydrogen bond between the imidazolium ring of [pmim][Br] and the carbonyl oxygen atom of the coumarin dyes (C-H[pmim][Br]â¯O=CDye). Finally, computer simulation indicates a slow component of solvation dynamics in the [pmim][Br]-water mixture in the time scale of â¼100-200 ps, which is similar to the experimental observation.
RESUMO
Effect of a room temperature ionic liquid (RTIL, [pmim][Br]) on the structure and dynamics of the protein, lysozyme, is investigated by fluorescence correlation spectroscopy (FCS) and molecular dynamic (MD) simulation. The FCS data indicate that addition of the RTIL ([pmim][Br]) leads to reduction in size and faster conformational dynamics of the protein. The hydrodynamic radius (rH) of lysozyme decreases from 18 Å in 0 M [pmim][Br] to 11 Å in 1.5 M [pmim][Br] while the conformational relaxation time decreases from 65 µs to 5 µs. Molecular origin of the collapse (size reduction) of lysozyme in aqueous RTIL is analyzed by MD simulation. The radial distribution function of water, RTIL cation, and RTIL anion from protein clearly indicates that addition of RTIL causes replacement of interfacial water by RTIL cation ([pmim](+)) from the first solvation layer of the protein providing a comparatively dehydrated environment. This preferential solvation of the protein by the RTIL cation extends up to â¼30 Å from the protein surface giving rise to a nanoscopic cage of overall radius 42 Å. In the nanoscopic cage of the RTIL (42 Å), volume fraction of the protein (radius 12 Å) is only about 2%. RTIL anion does not show any preferential solvation near protein surface. Comparison of effective radius obtained from simulation and from FCS data suggests that the "dry" protein (radius 12 Å) alone diffuses in a nanoscopic cage of RTIL (radius 42 Å). MD simulation further reveals a decrease in distance ("domain closure") between the two domains (alpha and beta) of the protein leading to a more compact structure compared to that in the native state.
Assuntos
Proteínas Aviárias/química , Proteínas do Ovo/química , Líquidos Iônicos/química , Muramidase/química , Temperatura , Água/química , Animais , Galinhas , Feminino , Hidrodinâmica , Simulação de Dinâmica Molecular , Conformação Proteica , Soluções , Solventes/química , Espectrometria de Fluorescência , ViscosidadeRESUMO
Some proteins are conformational switches, able to transition between relatively different conformations. To understand what drives them requires computing the free-energy difference ΔGAB between their stable states, A and B. Molecular dynamics (MD) simulations alone are often slow because they require a reaction coordinate and must sample many transitions in between. Here, we show that modeling employing limited data (MELD) x MD on known endstates A and B is accurate and efficient because it does not require passing over barriers or knowing reaction coordinates. We validate this method on two problems: (1) it gives correct relative populations of α and ß conformers for small designed chameleon sequences of protein G; and (2) it correctly predicts the conformations of the C-terminal domain (CTD) of RfaH. Free-energy methods like MELD x MD can often resolve structures that confuse machine-learning (ML) methods.
RESUMO
Recently, predicting the native structures of proteins has become possible using computational molecular physics (CMP)âphysics-based force fields sampled with proper statisticsâbut only for small proteins. Algorithms with better scaling are needed. We describe ML x MELD x MD, a molecular dynamics (MD) method that inputs residue contacts derived from machine learning (ML) servers into MELD, a Bayesian accelerator that preserves detailed-balance statistics. Contacts are derived from trRosetta-predicted distance histograms (distograms) and are integrated into MELD's atomistic MD as spatial restraints through parametrized potential functions. In the CASP14 blind prediction event, ML x MELD x MD predicted 13 native structures to better than 4.5 Å error, including for 10 proteins in the range of 115-250 amino acids long. Also, the scaling of simulation time vs protein length is much better than unguided MD: tsim â¼ e0.023N for ML x MELD x MD vs tsim â¼ e0.168N for MD alone. This shows how machine learning information can be leveraged to advance physics-based modeling of proteins.
Assuntos
Simulação de Dinâmica Molecular , Dobramento de Proteína , Teorema de Bayes , Biologia Computacional/métodos , Aprendizado de Máquina , Conformação ProteicaRESUMO
Clathrate hydrate forms when a hydrophobic molecule is entrapped inside a water cage or cavity. Although biomolecular structures also have hydrophobic patches, clathrate-like water is found in only a limited number of biomolecules. Also, while clathrate hydrates form at low temperature and moderately higher pressure, clathrate-like water is observed in biomolecular structure at ambient temperature and pressure. These indicate presence of other factors along with hydrophobic environment behind the formation of clathrate-like water in biomolecules. In the current study, we presented a systematic approach to explore the factors behind the formation of clathrate-like water in biomolecules by means of molecular dynamics simulation of a model protein, maxi, which is a naturally occurring nanopore and has clathrate-like water inside the pore. Removal of either confinement or hydrophobic environment results in the disappearance of clathrate-like water ordering, indicating a coupled role of these two factors. Apart from these two factors, clathrate-like water ordering also requires anchoring groups that can stabilize the clathrate-like water through hydrogen bonding. Our results uncover crucial factors for the stabilization of clathrate-like ordering in biomolecular structure which can be used for the development of new biomolecular structure promoting clathrate formation.
Assuntos
Pressão , Proteínas/química , Temperatura , Terpenos/química , Água/química , Ligação de Hidrogênio , Interações Hidrofóbicas e Hidrofílicas , Simulação de Dinâmica Molecular , Conformação ProteicaRESUMO
Among several denaturants, urea and guanidinium chloride (GdmCl) are the two strong and extensively used denaturants in unfolding experiments. However, the sequences of events in terms of secondary structure melting of several proteins in these two solvents have generated diverse results. A clear molecular understanding has still not been drawn to address the differential secondary structure specificity between urea and GdmCl. Here, we present a way to predict the possible unfolding route of α/ß proteins, in terms of secondary structure melting, based on the observation of relative solvent exposure of the alpha-helix and beta-sheet of protein in water. We find that while beta-sheet always melts first in urea, a diverse preference is evident for GdmCl. Larger solvent exposure of the backbone of the beta-sheet helps the beta-sheet to make favorable hydrogen bonds with urea which in turn causes the initial melting of the beta-sheet of alpha/beta proteins in urea. On the other hand, because GdmCl has hydrophobic weakening ability through preferential interaction, the region which has comparatively more hydrophobic solvent exposure melts first in GdmCl. Urea- and GdmCl-induced initial unfolding pathways of the alpha/beta protein is thus determined by the relative solvent exposure of the alpha-helix and beta-sheet of protein in water. Therefore, detailed knowledge of relative solvent exposures in water can provide a hint of the possible unfolding pathway provided the mode of action of the solvent is known.
Assuntos
Guanidina/química , Proteínas/química , Ureia/química , Humanos , Simulação de Dinâmica Molecular , Conformação Proteica , Desnaturação Proteica , Solventes , TermodinâmicaRESUMO
The primary driving force for protein folding is the formation of a well-packed, anhydrous core. However, recently, the crystal structure of an antifreeze protein, maxi, has been resolved where the core of the protein is filled with water, which apparently contradicts the existing notion of protein folding. Here, we have performed standard molecular dynamics (MD) simulation, replica exchange MD (REMD) simulation, and umbrella sampling using TIP4P water at various temperatures (300, 260, and 240 K) to explore the origin of this unusual structural feature. It is evident from standard MD and REMD simulations that the protein is found to be stable at 240 K in its unusual state. The core of protein has two layers of semi-clathrate water separating the methyl groups of alanine residues from different helical strands. However, with increasing temperature (260 and 300 K), the stability decreases as the core becomes dehydrated, and methyl groups of alanine are tightly packed driven by hydrophobic interactions. Calculation of the potential of mean force by an umbrella sampling technique between a pair of model hydrophobes resembling maxi protein at 240 K shows the stabilization of second solvent-separated minima (SSM), which provides a thermodynamic rationale of the unusual structural feature in terms of weakening of the hydrophobic interaction. Because the stabilization of SSMs is implicated for cold denaturation, it suggests that the maxi protein is so designed by nature where the cold denatured-like state becomes the biologically active form as it works near or below the freezing point of water.
Assuntos
Proteínas Anticongelantes/química , Simulação de Dinâmica Molecular , Água/química , Proteínas Anticongelantes/metabolismo , Interações Hidrofóbicas e Hidrofílicas , Conformação Proteica , Desnaturação Proteica , Estabilidade Proteica , Solventes/química , TemperaturaRESUMO
In this review, we give a brief overview on how the interaction of proteins with ionic liquids, alcohols and dimethyl sulfoxide (DMSO) influences the stability, conformational dynamics and function of proteins/enzymes. We present experimental results obtained from fluorescence correlation spectroscopy on the effect of ionic liquid or alcohol or DMSO on the size (more precisely, the diffusion constant) and conformational dynamics of lysozyme, cytochrome c and human serum albumin in aqueous solution. The interaction of ionic liquid with biomolecules (e.g. protein, DNA etc.) has emerged as a current frontier. We demonstrate that ionic liquids are excellent stabilizers of protein and DNA and, in some cases, cause refolding of a protein already denatured by chemical denaturing agents. We show that in ethanol-water binary mixture, proteins undergo non-monotonic changes in size and dynamics with increasing ethanol content. We also discuss the effect of water-DMSO mixture on the stability of proteins. We demonstrate how large-scale molecular dynamics simulations have revealed the molecular origin of this observed phenomenon and provide a microscopic picture of the immediate environment of the biomolecules. Finally, we describe how favorable interactions of ionic liquids may be utilized for in situ generation of fluorescent gold nano-clusters for imaging a live cell.
RESUMO
The hydrophobic effect appears to be a key driving force for many chemical and biological processes, such as protein folding, protein-protein interactions, membrane bilayer self-assembly, and so forth. In this study, we calculated the potential of mean force (PMF) using umbrella sampling technique between different model hydrophobes (methane-methane, cyclobutane-cyclobutane, and between two rodlike hydrophobes) at lower than ambient temperatures (300, 260, and 240 K). We find the appearance of a second solvent-separated minimum at â¼1.0 nm apart from the usual contact and first solvent-separated minimum in the PMF profile of the methane pair at low temperature. In the PMF between both cyclobutane and the rodlike hydrophobe pairs, the second solvent-separated pair (SSSP) becomes even more stable than the first solvent-separated pair (FSSP) at 240 K. Analysis of the water structure shows that, at 240 K, the core water of SSSP for the rodlike hydrophobe pair is more strongly hydrogen bonded and more tetrahedrally oriented than that of the FSSP. Strongly hydrogen-bonded ordered water molecules implicate strong water-water interactions, which are responsible for stabilization of SSSP at low temperature. This weakening of hydrophobic interactions through stabilization of SSSP may play a key role in the cold denaturation of protein.
RESUMO
Despite their routine use as protein denaturants, the comprehensive understanding of the molecular mechanisms by which urea and guanidinium chloride (GdmCl) disrupts proteins' structure is still lacking. Here, we use steered molecular dynamics simulations along with the umbrella sampling technique to elucidate the mechanism of unfolding of chicken villin headpiece (HP-36) in these two denaturants. We find that while urea denatures protein predominantly by forming hydrogen bonds with the protein backbone, GdmCl commences unfolding by weakening of the hydrophobic interactions present in the core. The potential of mean force calculation indicates the reduction of hydrophobic interactions between two benzene moieties in 6 M GdmCl as compared to 6 M urea. We observe a near parallel orientation between the guanidinium cation and aromatic side chains of the HP-36 suggesting π-cation type stacking interactions which play a crucial role in weakening of the hydrophobic interaction. We use QM/MM optimization calculations to estimate the energetics of this π-cation interaction. Additionally, the consistency of the unfolding paths between high temperature (400 K) unfolding simulations and steered molecular dynamics simulations strengthens the proposed molecular mechanism of unfolding further.