Force field impact and spin-probe modeling in molecular dynamics simulations of spin-labeled T4 lysozyme

Several attempts have been made to compute electron paramagnetic resonance (EPR) spectra of biomolecules, using motional models or simulated trajectories to describe dynamics. Ideally, the simulated trajectories should capture fast (picosecond) snapshots of spin-probe rotations accurately, while being lengthy enough to ensure a proper Fourier integration of the time-domain signal. It is the interplay of the two criteria that poses computational challenges to the method. In this context, an analysis of the spin-probe and protein conformational sampling and equilibration, with different force fields and with explicit solvent, may be a useful attempt. The present work reports a comparative study of the effect of the molecular dynamics (MD) force field on conformational sampling and equilibration in two spin-labeled T4 lysozyme (T4L) variants, N40C and K48C. Ensembles of 10× 3 ns-trajectories per variant and per force field (OPLS/AMBER and AMBER99) are analyzed for a reliable assessment of convergence and sampling. It is found that subtle site-dependent differences in spin-probe rotations and torsions are more readily captured in the AMBER99 trajectories than in the OPLS/AMBER simulations. On the other hand, sampling and equilibration are found to be better with the OPLS/AMBER force field at equal trajectory lengths.Figure: Left panel: The spin-probe R1 ring and the spin-probe Euler angles , and . Middle panel: Illustration of the diffusion in a cone model for the spin-probe motion: snapshots of helix B and of the R1 ring in N40C, taken at 0.3 ps intervals from AMBER trajectory 1. Right panel: The N40C mutant with the spin label (solid mode), solvated in a cubic box.     

Membrane proteins: molecular dynamics simulations

Molecular dynamics simulations of membrane proteins are making rapid progress, because of new high-resolution structures, advances in computer hardware and atomistic simulation algorithms, and the recent introduction of coarse-grained models for membranes and proteins. In addition to several large ion channel simulations, recent studies have explored how individual amino acids interact with the bilayer or snorkel/anchor to the headgroup region, and it has been possible to calculate water/membrane partition free energies. This has resulted in a view of bilayers as being adaptive rather than purely hydrophobic solvents, with important implications, for example, for interaction between lipids and arginines in the charged S4 helix of voltage-gated ion channels. However, several studies indicate that the typical current simulations fall short of exhaustive sampling, and that even simple protein-membrane interactions require at least ca. 1mus to fully sample their dynamics. One new way this is being addressed is coarse-grained models that enable mesoscopic simulations on multi-mus scale. These have been used to model interactions, self-assembly and membrane perturbations induced by proteins. While they cannot replace all-atom simulations, they are a potentially useful technique for initial insertion, placement, and low-resolution refinement.     

Analysis of ligand binding to proteins using molecular dynamics simulations

This work aims to explore theoretically the molecular mechanisms of ligand binding to proteins through the use of molecular dynamics simulations. The binding of sodium dodecyl sulfate (SDS) to cobra cardio toxin A3 (CTX A3) and thiourea (TOU) to lysozyme have been chosen as the two model systems. Data acquisitions were made by Gromacs software. To begin with, the collisions of ligand molecules with every residue of CTX A3 and lysozyme were evaluated. With this information in hand, the average numbers of collisions with each residue was defined and then assessed. Next, a measure of the affinity of a residue, P_i, referred to as conformational factor, toward a ligand molecule was established. Based on the results provided, all site-making residues for CTX A3 and lysozyme were identified. The results are in good agreement with the experimental data. Finally, based on this method, all site-making residues of bovine carbonic anhydrase (BCA) toward the SDS ligand were predicted.     

Application of molecular dynamics simulations to design a dual-purpose oligopeptide linker sequence for fusion proteins

Proteins are often monitored by combining a fluorescent polypeptide tag with the target protein. However, due to the high molecular weight and immunogenicity of such tags, they are not suitable choices for combining with fusion proteins such as immunotoxins. In this study, we designed a polypeptide sequence with a dual role (it acts as both a linker and a fluorescent probe) to use with fusion proteins. Two common fluorescent tag sequences based on tetracysteine were compared to a commonly used rigid linker as well as our proposed dual-purpose sequence. Computational investigations showed that the dual-purpose sequence was structurally stable and may be a good choice to use as both a linker and a fluorescence marker between two moieties in a fusion protein.     

OmpA: Gating and dynamics via molecular dynamics simulations

Outer membrane proteins (OMPs) of Gram-negative bacteria have a variety of functions including passive transport, active transport, catalysis, pathogenesis and signal transduction. Whilst the structures of ∼ 25 OMPs are currently known, there is relatively little known about their dynamics in different environments. The outer membrane protein, OmpA from Escherichia coli has been studied extensively in different environments both experimentally and computationally, and thus provides an ideal test case for the study of the dynamics and environmental interactions of outer membrane proteins. We review molecular dynamics simulations of OmpA and its homologues in a variety of different environments and discuss possible mechanisms of pore gating. The transmembrane domain of E. coli OmpA shows subtle differences in dynamics and interactions between a detergent micelle and a lipid bilayer environment. Simulations of the crystallographic unit cell reveal a micelle-like network of detergent molecules interacting with the protein monomers. Simulation and modelling studies emphasise the role of an electrostatic-switch mechanism in the pore-gating mechanism. Simulation studies have been extended to comparative models of OmpA homologues from Pseudomonas aeruginosa (OprF) and Pasteurella multocida (PmOmpA), the latter model including the periplasmic C-terminal domain.     

Driven to near-experimental accuracy by refinement via molecular dynamics simulations

Protein model refinement has been an essential part of successful protein structure prediction. Molecular dynamics simulation-based refinement methods have shown consistent improvement of protein models. There had been progress in the extent of refinement for a few years since the idea of ensemble averaging of sampled conformations emerged. There was little progress in CASP12 because conformational sampling was not sufficiently diverse due to harmonic restraints. During CASP13, a new refinement method was tested that achieved significant improvements over CASP12. The new method intended to address previous bottlenecks in the refinement problem by introducing new features. Flat-bottom harmonic restraints replaced harmonic restraints, sampling was performed iteratively, and a new scoring function and selection criteria were used. The new protocol expanded conformational sampling at reduced computational costs. In addition to overall improvements, some models were refined significantly to near-experimental accuracy.     

A statistical approach to the interpretation of molecular dynamics simulations of calmodulin equilibrium dynamics

A sample of 35 independent molecular dynamics (MD) simulations of calmodulin (CaM) equilibrium dynamics was prepared from different but equally plausible initial conditions (20 simulations of the wild-type protein and 15 simulations of the D129N mutant). CaM's radius of gyration and backbone mean-square fluctuations were analyzed for the effect of the D129N mutation, and simulations were compared with experiments. Statistical tests were employed for quantitative comparisons at the desired error level. The computational model predicted statistically significant compaction of CaM relative to the crystal structure, consistent with the results of small-angle X-ray scattering (SAXS) experiments. This effect was not observed in several previously reported studies of (Ca2+)(4)-CaM, which relied on a single MD run. In contrast to radius of gyration, backbone mean-square fluctuations showed a distinctly non-normal and positively skewed distribution for nearly all residues. Furthermore, the D129N mutation affected the backbone dynamics in a complex manner and reduced the mobility of Glu123, Met124, Ile125, Arg126, and Glu127 located in the adjacent alpha-helix G. The implications of these observations for the comparisons of MD simulations with experiments are discussed. The proposed approach may be useful in studies of protein equilibrium dynamics where MD simulations fall short of properly sampling the conformational space, and when the comparison with experiments is affected by the reproducibility of the computational model.     

Electrostatic screening in molecular dynamics simulations.

The screened Coulombic potential has been shown to describe satisfactorily equilibrium properties like pK shifts, the effects of charged groups on redox potentials and binding constants of metal ions. To test how well the screening of the electrostatic potential describes the dynamical trajectory of a macromolecular system, a series of comparative simulations have been carried out on a protein system which explicitly included water molecules and a system in vacuo. For the system without solvent the results of using (i) the standard potential form were compared with results of (ii) the potential where the Coulomb term was modified by the inclusion of a distance dependent dielectric, epsilon (r), to model the screening effect of bulk water, and (iii) standard potential modified by reducing the charge on ionized residue side chains. All molecular dynamics simulations have been carried out on bovine pancreatic trypsin inhibitor. Comparisons between the resulting trajectories, averaged structures, hydrogen bonding patterns and properties such as solvent accessible surface area and radius of gyration are described. The results show that the dynamical behaviour of the protein calculated with a screened electrostatic term compares more favourably with the time-dependent structural changes of the full system with explicitly included water than the standard vacuum simulation.     

Convergence of molecular dynamics simulations of membrane proteins

The central question in evaluating almost any result from a molecular dynamics simulation is whether the calculation has converged. Unfortunately, assessing the ergodicity of a single trajectory is very difficult to do. In this work, we assess the sampling of molecular dynamics simulations of the membrane protein rhodopsin by comparing the results from 26 independent trajectories, each run for 100 ns. By examining principal components and cluster populations, we show that rhodopsin's fluctuations are not well described by 100 ns of dynamics, and that the sampling is not fully converged even for individual loops. The results serve as a reminder of the caution required when interpreting molecular dynamics simulations of macromolecules.     

Homology modeling and molecular dynamics simulations of lymphotactin

We have modeled the structure of human lymphotactin (hLpnt), by homology modeling and molecular dynamics simulations. This chemokine is unique in having a single disulfide bond and a long C-terminal tail. Because other structural classes of chemokines have two pairs of Cys residues, compared to one in Lpnt, and because it has been shown that both disulfide bonds are required for stability and function, the question arises how the Lpnt maintains its structural integrity. The initial structure of hLpnt was constructed by homology modeling. The first 63 residues in the monomer of hLpnt were modeled using the structure of the human CC chemokine, RANTES, whose sequence appeared most similar. The structure of the long C-terminal tail, missing in RANTES, was taken from the human muscle fatty-acid binding protein. In a Protein Data Bank search, this protein was found to contain a sequence that was most homologous to the long tail. Consequently, the modeled hLpnt C-terminal tail consisted of both alpha-helical and beta-motifs. The complete model of the hLpnt monomer consisted of two alpha-helices located above the five-stranded beta-sheet. Molecular dynamics simulations of the solvated initial model have indicated that the stability of the predicted fold is related to the geometry of Pro78. The five-stranded beta-sheet appeared to be preserved only when Pro78 was modeled in the cis conformation. Simulations were also performed both for the C-terminal truncated forms of the hLpnt that contained one or two (CC chemokine-like) disulfide bonds, and for the chicken Lpnt (cLpnt). Our MD simulations indicated that the turn region (T30-G34) in hLpnt is important for the interactions with the receptor, and that the long C-terminal region stabilizes both the turn (T30-G34) and the five-stranded beta-sheet. The major conclusion from our theoretical studies is that the lack of one disulfide bond and the extension of the C-terminus in hLptn are mutually complementary. It is very likely that removal of two Cys residues sufficiently destabilizes the structure of a chemokine molecule, particularly the core beta-sheet, to abolish its biological function. However, this situation is rectified by the long C-terminal segment. The role of this long region is most likely to stabilize the first beta-turn region and alpha-helix H1, explaining how this chemokine can function with a single disulfide bond.     

Conformational analysis and molecular dynamics simulations of maltose

Constrained conformational energy minimizations have been used to calculate an adiabatic (Φ, ψ) potential energy surface for the disaccharide β-maltose. The inclusion of molecular flexibility in the conformational energy analysis of the disaccharide was found to significantly lower the barriers to conformational transitions, as has been observed previously for other systems. Several low energy wells were identified on the adiabatic surface which differ in energy by small amounts and with low absolute barriers separating them, indicating the possibility of a non-negligible equilibrium population distribution in each well. If such a distribution of conformations existed in the physical system, the conformation observed by NMR NOE measurements would thus be a “virtual” conformation. Molecular dynamics simulations of the motions of this molecule in vacuum were also conducted and indicate that the rate of relaxation of the molecule to the adiabatic surface may be slower than the typical timescale of conformational fluctuations. This effect is apparently due to an unphysical persistence of hydrogen bond patterns in vacuum which does not occur in aqueous solution. Trajectories undergoing transitions between wells were calculated and the effects of such conformational transitions upon the ensemble mean structure, such as might be observed in an NMR experiment, were demonstrated.     

Integrating experimental data with molecular simulations to investigate RNA structural dynamics

Conformational dynamics is crucial for ribonucleic acid (RNA) function. Techniques such as nuclear magnetic resonance, cryo-electron microscopy, small- and wide-angle X-ray scattering, chemical probing, single-molecule Förster resonance energy transfer, or even thermal or mechanical denaturation experiments probe RNA dynamics at different time and space resolutions. Their combination with accurate atomistic molecular dynamics (MD) simulations paves the way for quantitative and detailed studies of RNA dynamics. First, experiments provide a quantitative validation tool for MD simulations. Second, available data can be used to refine simulated structural ensembles to match experiments. Finally, comparison with experiments allows for improving MD force fields that are transferable to new systems for which data is not available. Here we review the recent literature and provide our perspective on this field.     

Preparation and characterization of spin-labeled oligonucleotides for DNA hybridization.

Sequence-specific spin-labeled oligodeoxynucleotides with conformation-sensitive electron paramagnetic resonance (EPR) signals are synthesized and examined as solution-phase nucleic acid hybridization probes. Either a proxyl or tempo ring linked to the C(5) position of deoxyuridine (dU) by a nonrigid two-atom methylamino tether is incorporated within 15-mers by phosphotriester chemistry yielding stable spin-labeled probes with distinctive EPR specific activity (AEPR) values. The AEPR is greater for a proxyl-labeled than for a tempo-labeled probe and is consistent with EPR data of enzymatically labeled 26-mers [Bobst, A. M., Pauly, G. T., Keyes, R. S., and Bobst, E. V. (1988) FEBS Lett. 228, 33-36], after normalizing for percent labeling. The spectral characteristics of the free probes and the probe/target complexes are similar to those of enzymatically spin-labeled nucleic acids containing a different nonrigid two-atom-tethered spin label [Bobst, A. M., Kao, S.-C., Toppin, R. C., Ireland, J. C., and Thomas, I. E. (1984) J. Mol. Biol. 173, 63-70]. The presence of target DNA is detected in solution by EPR spectroscopy and the assay is based on the characteristic line-shape change associated with hybridization. The EPR spectra of free and bound probe reflect little interference from changes in global dynamics of the probe, and the line-shape change upon complexation results primarily from a change in local base dynamics. The presence or absence of hybridization can be detected in a loop-gap resonator with about 1 pmol of spin-labeled 15-mer within minutes.     

Atomistic modeling of apatite-collagen composites from molecular dynamics simulations extended to hyperspace

The preparation of atomistic models of apatite-collagen composite mimicking enamel at length scales in the range of 1–10 nanometers is outlined. This bio-composite is characterized by a peculiar interplay of the collagen triplehelix and the apatite crystal structure. Structural coherence is however only obtained after drastic rearrangements, namely the depletion of protein-protein hydrogen bonds and the incorporation of calcium triangles which are stabilized by salt-bridges with the collagen molecule. Starting from an isolated collagen triple helix and a single-crystalline apatite structure, a composite model is obtained by gradually merging the two components via an additional (hyperspace) coordinate. This approach allows smooth structural relaxation of both components whilst avoiding singularities in potential energy due to atomic overlap.     

Predicting thymine dimerization yields from molecular dynamics simulations

It was recently shown that thymine dimers in the all-thymine oligonucleotide (dT)₁₈ are fully formed in <1 ps after ultraviolet excitation. The speed and low quantum yield of this reaction suggest that only a small fraction of the conformers of this structurally disordered oligonucleotide are in a position to react at the instant of photon absorption. In this work, we explore the hypothesis that conventional molecular dynamics simulations can be used to predict the yield of cyclobutane pyrimidine dimers in DNA. Conformations obtained from simulations of thymidylyl-(3′-5′)-thymidine in various cosolvents were classified as dimerizable or nondimerizable depending on the distance between the C5-C6 double bonds of the adjacent thymine bases and the torsion angle between them. The quantum yield of cyclobutane pyrimidine dimer formation was calculated as the number of dimerizable conformations divided by the total number of conformations. The experimental quantum yields measured in the different solvents were satisfactorily reproduced using physically reasonable values for the two parameters. The mean dimerizable structure computed by averaging all of the dimerizable cis-syn conformations is structurally similar to the actual cis-syn dimer. Compared to the canonical B-form TT step, the most important structural property of a dimerizable conformation is its reduced helical twist angle of 22°.     

Protein flexibility using constraints from molecular dynamics simulations

Proteins are held together in the native state by hydrophobic interactions, hydrogen bonds and interactions with the surrounding water, whose strength as well as spatial and temporal distribution affects protein flexibility and hence function. We study these effects using 10 ns molecular dynamics simulations of pure water and of two proteins, the glutamate receptor ligand binding domain and barnase. We find that most of the noncovalent interactions flicker on and off over typically nanoseconds, and so we can obtain good statistics from the molecular dynamics simulations. Based on this information, a topological network of rigid bonds corresponding to a protein structure with covalent and noncovalent bonds is constructed, with account being taken of the influence of the flickering hydrogen bonds. We define the duty cycle for the noncovalent interactions as the percentage of time a given interaction is present, which we use as an input to investigate flexibility/rigidity patterns, in the algorithm FIRST which constructs and analyses topological networks.     

Time-averaged order parameter restraints in molecular dynamics simulations

A method is described that allows experimental \(S^2\) order parameters to be enforced as a time-averaged quantity in molecular dynamics simulations. The two parameters that characterize time-averaged restraining, the memory relaxation time and the weight of the restraining potential energy term in the potential energy function used in the simulation, are systematically investigated based on two model systems, a vector with one end restrained in space and a pentapeptide. For the latter it is shown that the backbone N–H order parameter of individual residues can be enforced such that the spatial fluctuations of quantities depending on atomic coordinates are not significantly perturbed. The applicability to realistic systems is illustrated for the B3 domain of protein G in aqueous solution.     

Non-B DNA conformations analysis through molecular dynamics simulations

BackgroundNon-B DNA conformations are molecular structures that do not follow the canonical DNA double helix. Mutagenetic instability in nuclear and mitochondrial DNA (mtDNA) genomes has been associated with simple non-B DNA conformations, as hairpins or more complex structures, as G-quadruplexes. One of these structures is Structure A, a cloverleaf-like non-B conformation predicted for a 93-nt (nucleotide) stretch of the mtDNA control region 5′-peripheral domain. Structure A is embedded in a hot spot for the 3′ end of human mtDNA deletions revealing its importance in influencing the mutational instability of the mtDNA genome.MethodsTo better characterize Structure A, we predicted its 3D conformation using state-of-art methods and algorithms. The methodologic workflow consisted in the prediction of non-B conformations using molecular dynamics simulations. The conservation scores of alignments of the Structure A region in humans, primates, and mammals, was also calculated.ResultsOur results show that these computational methods are able to measure the stability of non-B conformations by using the level of base pairing during molecular dynamics. Structure A showed high stability and low flexibility correlated with high conservation scores in mammalian, more specifically in primate lineages.ConclusionsWe showed that 3D non-B conformations can be predicted and characterized by our methodology. This allowed the in-depth analysis of the structure A, and the main results showed the structure remains stable during the simulations.General significanceThe fine-scale atomic molecular determination of this type of non-B conformation opens the way to perform computational molecular studies that can show their involvement in mtDNA cellular mechanisms.     

Retinal release from opsin in molecular dynamics simulations

The crystal structures of opsin in the ligand-free and the G-protein-interacting states showed two inter-helical openings between transmembrane (TM) helices TM1 and TM7 and between TM5 and TM6 near the extracellular side that were thought to serve as the retinal uptake and release gates. However, it is unclear which opening is for 11-cis-retinal uptake or all-trans-retinal release although speculations have been proposed based on the structural features of opsin and retinal. In this work, we simulated the exit process of all-trans-retinal from the ligand-free opsin structure by the classical molecular dynamics (MD) and random acceleration molecular dynamics (RAMD). In the 64 ns classical MD simulation, retinal remained in the receptor but moved significantly toward the TM5-TM6 opening and almost inserted into the opening after 50 ns. Complete exit was observed in 114 out of 160 RAMD trajectories with the TM5-TM6 opening being the predominant egress gate while egress from the TM1-TM7 opening was observed in only a few trajectories when relatively large acceleration was applied and large structural alteration of the protein resulted. These results suggest that photolyzed all-trans-retinal is likely released through the TM5-TM6 opening. Based on the unidirectional mechanism of retinal exchange suggested by experiment, we speculate that the TM1-TM7 opening serves as the 11-cis-retinal uptake gate. The spatial occupancy maps of retinal computed from the 160 RAMD trajectories further indicated that retinal experienced significant interactions with the receptor during the exit process. The implications of these findings for disease mechanisms of rhodopsin mutants are discussed.