Coarse-grained MD simulations of membrane protein-bilayer self-assembly

Lipid interaction sites on channels, transporters and receptors: Recent insights from molecular dynamics simulations

Lipid molecules are able to selectively interact with specific sites on integral membrane proteins, and modulate their structure and function. Identification and characterization of these sites are of importance for our understanding of the molecular basis of membrane protein function and stability, and may facilitate the design of lipid-like drug molecules. Molecular dynamics simulations provide a powerful tool for the identification of these sites, complementing advances in membrane protein structural biology and biophysics. We describe recent notable biomolecular simulation studies which have identified lipid interaction sites on a range of different membrane proteins. The sites identified in these simulation studies agree well with those identified by complementary experimental techniques. This demonstrates the power of the molecular dynamics approach in the prediction and characterization of lipid interaction sites on integral membrane proteins. This article is part of a Special Issue entitled: Biosimulations edited by Ilpo Vattulainen and Tomasz Róg.

Crystal structures reveal the molecular basis of ion translocation in sodium/proton antiporters

To fully understand the transport mechanism of Na(+)/H(+) exchangers, it is necessary to clearly establish the global rearrangements required to facilitate ion translocation. Currently, two different transport models have been proposed. Some reports have suggested that structural isomerization is achieved through large elevator-like rearrangements similar to those seen in the structurally unrelated sodium-coupled glutamate-transporter homolog GltPh. Others have proposed that only small domain movements are required for ion exchange, and a conventional rocking-bundle model has been proposed instead. Here, to resolve these differences, we report atomic-resolution structures of the same Na(+)/H(+) antiporter (NapA from Thermus thermophilus) in both outward- and inward-facing conformations. These data combined with cross-linking, molecular dynamics simulations and isothermal calorimetry suggest that Na(+)/H(+) antiporters provide alternating access to the ion-binding site by using elevator-like structural transitions.

The membranes of Gram-negative bacteria: progress in molecular modelling and simulation

Molecular modelling and simulations have been employed to study the membranes of Gram-negative bacteria for over 20 years. Proteins native to these membranes, as well as antimicrobial peptides and drug molecules have been studied using molecular dynamics simulations in simple models of membranes, usually only comprising one lipid species. Thus, traditionally, the simulations have reflected the majority of in vitro membrane experimental setups, enabling observations from the latter to be rationalized at the molecular level. In the last few years, the sophistication and complexity of membrane models have improved considerably, such that the heterogeneity of the lipid and protein composition of the membranes can now be considered both at the atomistic and coarse-grain levels of granularity. Importantly this means relevant biology is now being retained in the models, thereby linking the in silico and in vivo scenarios. We discuss recent progress in simulations of proteins in simple lipid bilayers, more complex membrane models and finally describe some efforts to overcome timescale limitations of atomistic molecular dynamics simulations of bacterial membranes.

MD Simulations of P-Type ATPases in a Lipid Bilayer System

Molecular dynamics (MD) simulation is a computational method which provides insight on protein dynamics with high resolution in both space and time, in contrast to many experimental techniques. MD simulations can be used as a stand-alone method to study P-type ATPases as well as a complementary method aiding experimental studies. In particular, MD simulations have proved valuable in generating and confirming hypotheses relating to the structure and function of P-type ATPases. In the following, we describe a detailed practical procedure on how to set up and run a MD simulation of a P-type ATPase embedded in a lipid bilayer using software free of use for academics. We emphasize general considerations and problems typically encountered when setting up simulations. While full coverage of all possible procedures is beyond the scope of this chapter, we have chosen to illustrate the MD procedure with the Nanoscale Molecular Dynamics (NAMD) and the Visual Molecular Dynamics (VMD) software suites.

A Coarse Grained Model for a Lipid Membrane with Physiological Composition and Leaflet Asymmetry

The resemblance of lipid membrane models to physiological membranes determines how well molecular dynamics (MD) simulations imitate the dynamic behavior of cell membranes and membrane proteins. Physiological lipid membranes are composed of multiple types of phospholipids, and the leaflet compositions are generally asymmetric. Here we describe an approach for self-assembly of a Coarse-Grained (CG) membrane model with physiological composition and leaflet asymmetry using the MARTINI force field. An initial set-up of two boxes with different types of lipids according to the leaflet asymmetry of mammalian cell membranes stacked with 0.5 nm overlap, reliably resulted in the self-assembly of bilayer membranes with leaflet asymmetry resembling that of physiological mammalian cell membranes. Self-assembly in the presence of a fragment of the plasma membrane protein syntaxin 1A led to spontaneous specific positioning of phosphatidylionositol(4,5)bisphosphate at a positively charged stretch of syntaxin consistent with experimental data. An analogous approach choosing an initial set-up with two concentric shells filled with different lipid types results in successful assembly of a spherical vesicle with asymmetric leaflet composition. Self-assembly of the vesicle in the presence of the synaptic vesicle protein synaptobrevin 2 revealed the correct position of the synaptobrevin transmembrane domain. This is the first CG MD method to form a membrane with physiological lipid composition as well as leaflet asymmetry by self-assembly and will enable unbiased studies of the incorporation and dynamics of membrane proteins in more realistic CG membrane models.

Influence of lipids on the hydrophobic barrier within the pore of the TWIK-1 K2P channel

Several recent ion channel structures have revealed large side portals, or ‘fenestrations’ at the interface between their transmembrane helices that potentially expose the ion conduction pathway to the lipid core of the bilayer. In a recent study we demonstrated that functional activity of the TWIK-1 K2P channel is influenced by the presence of hydrophobic residues deep within the inner pore. These residues are located near the fenestrations in the TWIK-1 structure and promote dewetting of the pore by forming a hydrophobic barrier to ion conduction. During our previous MD simulations, lipid tails were observed to enter these fenestrations. In this addendum to that study, we investigate lipid contribution to the dewetting process. Our results demonstrate that lipid tails from both the upper and lower leaflets can occupy the fenestrations and partially penetrate into the pore. The lipid tails do not sterically occlude the pore, but there is an inverse correlation between the presence of water within the hydrophobic barrier and the number of lipids tails within the lining of the pore. However, dewetting still occurs in the absence of lipids tails, and pore hydration appears to be determined primarily by those side-chains lining the narrowest part of the pore cavity.

Toward a Coarse-Grained Protein Model Coupled with a Coarse-Grained Solvent Model: Solvation Free Energies of Amino Acid Side Chains

Recently, we reported that molecular dynamics (MD) simulations using a coarse-grained (CG) peptide model coupled with a CG water model are able to reproduce many of the structural and thermodynamic features of short peptides with nonpolar side chains at 10(3) times the normal speed (JCTC, 2007, 3, 2146-2161). To further develop a CG protein model for MD simulations, we systematically parametrized the side chains of all 20 naturally occurring amino acids. We developed the parameters by fitting the dihedral potentials of 13 small molecules, the densities and self-solvation free energies of liquids of eight organic molecules, and the hydration free energies of 35 small organic molecules. In a set of 11 classes of compounds (105 in total) including alkanes, alcohols, ethers, ketones/aldehydes, amines, amides, aromatics, carboxylic acids, sulfides/thiols, alkyl ammoniums, and carboxylate ions, the average error in the calculated hydration free energies compared with experimental results is about 1.4 kJ/mol. The average error in the calculated transfer free energies of the 19 side-chain analogues of amino acids from cyclohexane to water is about 2.2 kJ/mol. These results are comparable to the results of all-atom models.

Membrane/Toxin Interaction Energetics via Serial Multiscale Molecular Dynamics Simulations

Computing free energies of complex biomolecular systems via atomistic (AT) molecular dynamics (MD) simulations remains a challenge due to the need for adequate sampling and convergence. Recent coarse-grained (CG) methodology allows simulations of significantly larger systems (∼10(6) to 10(8) atoms) over longer (μs/ms) time scales. Such CG models appear to be capable of making semiquantitative predictions. However, their ability to reproduce accurate thermodynamic quantities remains uncertain. We have recently used CG MD simulations to compute the potential of mean force (PMF) or free energy profile of a small peptide toxin interacting with a lipid bilayer along a 1D reaction coordinate. The toxin studied was VSTx1 (Voltage Sensor Toxin 1) from spider venom which inhibits the archeabacterial voltage-gated potassium (Kv) channel KvAP by binding to the voltage-sensor (VS) domains. Here, we re-estimate this PMF profile using (i) AT MD simulations with explicit membrane and solvent and (ii) an implicit membrane and solvent (generalized Born; GBIM) model where only the peptide was explicit. We used the CG MD free energy simulations to guide the setup of the corresponding AT MD simulations. The aim was to avoid local minima in the AT simulations which would be difficult over shorter AT time scales. A cross-comparison of the PMF profiles revealed a conserved topology, although there were differences in the magnitude of the free energies. The CG and AT simulations predicted a membrane/water interface free energy well of -27 and -23 kcal/mol, respectively (with respect to water). The GBIM model, however, gave a reduced interfacial free energy well (-12 kcal/mol). In addition, the CG and GBIM models predicted a free energy barrier of +61 and +96 kcal/mol, respectively, for positioning the toxin at the center of the bilayer, which was considerably smaller in the AT simulations (+26 kcal/mol). Thus, we present a framework for serially combining CG and AT simulations to estimate the free energy of peptide/membrane interactions. Such approaches for combining simulations at different levels of granularity will become increasingly important in future studies of complex membrane/protein systems.

The role of lipids in mechanosensation

The ability of proteins to sense membrane tension is pervasive in biology. A higher-resolution structure of the Escherichia coli small-conductance mechanosensitive channel MscS identifies alkyl chains inside pockets formed by the transmembrane helices (TMs). Purified MscS contains E. coli lipids, and fluorescence quenching demonstrates that phospholipid acyl chains exchange between bilayer and TM pockets. Molecular dynamics and biophysical analyses show that the volume of the pockets and thus the number of lipid acyl chains within them decreases upon channel opening. Phospholipids with one acyl chain per head group (lysolipids) displace normal phospholipids (with two acyl chains) from MscS pockets and trigger channel opening. We propose that the extent of acyl-chain interdigitation in these pockets determines the conformation of MscS. When interdigitation is perturbed by increased membrane tension or by lysolipids, the closed state becomes unstable, and the channel gates.

All-atom lipid bilayer self-assembly with the AMBER and CHARMM lipid force fields

This communication reports the first example of spontaneous lipid bilayer formation in unbiased all-atom molecular dynamics (MD) simulations. Using two different lipid force fields we show simulations started from random mixtures of lipids and water in which four different types of phospholipids self-assemble into organized bilayers in under 1 microsecond.

Organization and Dynamics of Receptor Proteins in a Plasma Membrane

The interactions of membrane proteins are influenced by their lipid environment, with key lipid species able to regulate membrane protein function. Advances in high-resolution microscopy can reveal the organization and dynamics of proteins and lipids within living cells at resolutions <200 nm. Parallel advances in molecular simulations provide near-atomic-resolution models of the dynamics of the organization of membranes of in vivo-like complexity. We explore the dynamics of proteins and lipids in crowded and complex plasma membrane models, thereby closing the gap in length and complexity between computations and experiments. Our simulations provide insights into the mutual interplay between lipids and proteins in determining mesoscale (20-100 nm) fluctuations of the bilayer, and in enabling oligomerization and clustering of membrane proteins.

Regulation of the Ca(2+)-ATPase by cholesterol: a specific or non-specific effect?

Like other integral membrane proteins, the activity of the Sarco/Endoplasmic Reticulum Ca(2+)-ATPase (SERCA) is regulated by the membrane environment. Cholesterol is present in the endoplasmic reticulum membrane at low levels, and it has the potential to affect SERCA activity both through direct, specific interaction with the protein or through indirect interaction through changes of the overall membrane properties. There are experimental data arguing for both modes of action for a cholesterol-mediated regulation of SERCA. In the current study, coarse-grained molecular dynamics simulations are used to address how a mixed lipid-cholesterol membrane interacts with SERCA. Candidates for direct regulatory sites with specific cholesterol binding modes are extracted from the simulations. The binding pocket for thapsigargin, a nanomolar inhibitor of SERCA, has been suggested as a cholesterol binding site. However, the thapsigargin binding pocket displayed very little cholesterol occupation in the simulations. Neither did atomistic simulations of cholesterol in the thapsigargin binding pocket support any specific interaction. The current study points to a non-specific effect of cholesterol on SERCA activity, and offers an alternative interpretation of the experimental results used to argue for a specific effect.

Crystal structure of the sodium-proton antiporter NhaA dimer and new mechanistic insights

A dimeric structure of the sodium–proton antiporter NhaA provides insight into the roles of Asp163 and Lys300 in the transport mechanism.

Sodium–proton antiporters rapidly exchange protons and sodium ions across the membrane to regulate intracellular pH, cell volume, and sodium concentration. How ion binding and release is coupled to the conformational changes associated with transport is not clear. Here, we report a crystal form of the prototypical sodium–proton antiporter NhaA from Escherichia coli in which the protein is seen as a dimer. In this new structure, we observe a salt bridge between an essential aspartic acid (Asp163) and a conserved lysine (Lys300). An equivalent salt bridge is present in the homologous transporter NapA, but not in the only other known crystal structure of NhaA, which provides the foundation of most existing structural models of electrogenic sodium–proton antiport. Molecular dynamics simulations show that the stability of the salt bridge is weakened by sodium ions binding to Asp164 and the neighboring Asp163. This suggests that the transport mechanism involves Asp163 switching between forming a salt bridge with Lys300 and interacting with the sodium ion. pK_a calculations suggest that Asp163 is highly unlikely to be protonated when involved in the salt bridge. As it has been previously suggested that Asp163 is one of the two residues through which proton transport occurs, these results have clear implications to the current mechanistic models of sodium–proton antiport in NhaA.

MemProtMD: Automated Insertion of Membrane Protein Structures into Explicit Lipid Membranes

There has been exponential growth in the number of membrane protein structures determined. Nevertheless, these structures are usually resolved in the absence of their lipid environment. Coarse-grained molecular dynamics (CGMD) simulations enable insertion of membrane proteins into explicit models of lipid bilayers. We have automated the CGMD methodology, enabling membrane protein structures to be identified upon their release into the PDB and embedded into a membrane. The simulations are analyzed for protein-lipid interactions, identifying lipid binding sites, and revealing local bilayer deformations plus molecular access pathways within the membrane. The coarse-grained models of membrane protein/bilayer complexes are transformed to atomistic resolution for further analysis and simulation. Using this automated simulation pipeline, we have analyzed a number of recently determined membrane protein structures to predict their locations within a membrane, their lipid/protein interactions, and the functional implications of an enhanced understanding of the local membrane environment of each protein.

•

A simulation pipeline for predicting the location of a membrane protein in a bilayer

•

A protocol for identifying novel membrane protein structures in the PDB

•

Analysis of lipid binding sites and local bilayer deformation by membrane proteins

•

Functional implications from enhanced understanding of local membrane environments

Alchembed: A Computational Method for Incorporating Multiple Proteins into Complex Lipid Geometries

A necessary step prior to starting any membrane protein computer simulation is the creation of a well-packed configuration of protein(s) and lipids. Here, we demonstrate a method, alchembed, that can simultaneously and rapidly embed multiple proteins into arrangements of lipids described using either atomistic or coarse-grained force fields. During a short simulation, the interactions between the protein(s) and lipids are gradually switched on using a soft-core van der Waals potential. We validate the method on a range of membrane proteins and determine the optimal soft-core parameters required to insert membrane proteins. Since all of the major biomolecular codes include soft-core van der Waals potentials, no additional code is required to apply this method. A tutorial is included in the Supporting Information.

The cytosolic tail of the tumor marker protein Trop2--a structural switch triggered by phosphorylation

Trop2 is a transmembrane signaling glycoprotein upregulated in stem and carcinoma cells. Proliferation-enhancing signaling involves regulated intramembrane proteolytic release of a short cytoplasmic fragment, which is later engaged in a cytosolic signaling complex. We propose that Trop2 function is modulated by phosphorylation of a specific serine residue within this cytosolic region (Ser303), and by proximity effects exerted on the cytosolic tail by Trop2 dimerization. Structural characterization of both the transmembrane (Trop2TM) and cytosolic regions (Trop2IC) support this hypothesis, and shows that the central region of Trop2IC forms an α-helix. Comparison of NMR structures of non-phosphorylated and phosphorylated forms suggest that phosphorylation of Trop2IC triggers salt bridge reshuffling, resulting in significant conformational changes including ordering of the C-terminal tail. In addition, we demonstrate that the cytosolic regions of two Trop2 subunits can be brought into close proximity via transmembrane part dimerization. Finally, we show that Ser303-phosphorylation significantly affects the structure and accessibility of functionally important regions of the cytosolic tail. These observed structural features of Trop2 at the membrane-cytosol interface could be important for regulation of Trop2 signaling activity.

The juxtamembrane regions of human receptor tyrosine kinases exhibit conserved interaction sites with anionic lipids

Receptor tyrosine kinases (RTKs) play a critical role in diverse cellular processes and their activity is regulated by lipids in the surrounding membrane, including PIP₂ (phosphatidylinositol-4,5-bisphosphate) in the inner leaflet, and GM3 (monosialodihexosylganglioside) in the outer leaflet. However, the precise details of the interactions at the molecular level remain to be fully characterised. Using a multiscale molecular dynamics simulation approach, we comprehensively characterise anionic lipid interactions with all 58 known human RTKs. Our results demonstrate that the juxtamembrane (JM) regions of RTKs are critical for inducing clustering of anionic lipids, including PIP₂, both in simple asymmetric bilayers, and in more complex mixed membranes. Clustering is predominantly driven by interactions between a conserved cluster of basic residues within the first five positions of the JM region, and negatively charged lipid headgroups. This highlights a conserved interaction pattern shared across the human RTK family. In particular predominantly the N-terminal residues of the JM region are involved in the interactions with PIP₂, whilst residues within the distal JM region exhibit comparatively less lipid specificity. Our results suggest that JM–lipid interactions play a key role in RTK structure and function, and more generally in the nanoscale organisation of receptor-containing cell membranes.

Molecular dynamics simulations of membrane proteins

Membrane protein structures are underrepresented in the Protein Data Bank (PDB) due to difficulties associated with expression and crystallization. As such, it is one area where computational studies, particularly Molecular Dynamics (MD) simulations, can provide useful additional information. Recently, there has been substantial progress in the simulation of lipid bilayers and membrane proteins embedded within them. Initial efforts at simulating membrane proteins embedded within a lipid bilayer were relatively slow and interactive processes, but recent advances now mean that the setup and running of membrane protein simulations is somewhat more straightforward, though not without its problems. In this chapter, we outline practical methods for setting up and running MD simulations of a membrane protein embedded within a lipid bilayer and discuss methodologies that are likely to contribute future improvements.

In pursuit of an accurate spatial and temporal model of biomolecules at the atomistic level: a perspective on computer simulation

Despite huge advances in the computational techniques available for simulating biomolecules at the quantum-mechanical, atomistic and coarse-grained levels, there is still a widespread perception amongst the experimental community that these calculations are highly specialist and are not generally applicable by researchers outside the theoretical community. In this article, the successes and limitations of biomolecular simulation and the further developments that are likely in the near future are discussed. A brief overview is also provided of the experimental biophysical methods that are commonly used to probe biomolecular structure and dynamics, and the accuracy of the information that can be obtained from each is compared with that from modelling. It is concluded that progress towards an accurate spatial and temporal model of biomacromolecules requires a combination of all of these biophysical techniques, both experimental and computational.

Dynamic scenario of membrane binding process of kalata b1

Kalata B1 (kB1), a cyclotide that has been used in medical applications, displays cytotoxicity related to membrane binding and oligomerization. Our molecular dynamics simulation results demonstrate that Trp19 in loop 5 of both monomeric and tetrameric kB1 is a key residue for initial anchoring in the membrane binding process. This residue also facilitates the formation of kB1 tetramers. Additionally, we elucidate that kB1 preferentially binds to the membrane interfacial zone and is unable to penetrate into the membrane. In particular, significant roles of amino acid residues in loop 5 and loop 6 on the localization of kB1 to this membrane-water interface zone are found. This study reveals the roles of amino acid residues in the bioactivity of kB1, which is information that can be useful for designing new therapeutic cyclotides with less toxicity.

Lipid clustering correlates with membrane curvature as revealed by molecular simulations of complex lipid bilayers

Cell membranes are complex multicomponent systems, which are highly heterogeneous in the lipid distribution and composition. To date, most molecular simulations have focussed on relatively simple lipid compositions, helping to inform our understanding of in vitro experimental studies. Here we describe on simulations of complex asymmetric plasma membrane model, which contains seven different lipids species including the glycolipid GM3 in the outer leaflet and the anionic lipid, phosphatidylinositol 4,5-bisphophate (PIP₂), in the inner leaflet. Plasma membrane models consisting of 1500 lipids and resembling the in vivo composition were constructed and simulations were run for 5 µs. In these simulations the most striking feature was the formation of nano-clusters of GM3 within the outer leaflet. In simulations of protein interactions within a plasma membrane model, GM3, PIP₂, and cholesterol all formed favorable interactions with the model α-helical protein. A larger scale simulation of a model plasma membrane containing 6000 lipid molecules revealed correlations between curvature of the bilayer surface and clustering of lipid molecules. In particular, the concave (when viewed from the extracellular side) regions of the bilayer surface were locally enriched in GM3. In summary, these simulations explore the nanoscale dynamics of model bilayers which mimic the in vivo lipid composition of mammalian plasma membranes, revealing emergent nanoscale membrane organization which may be coupled both to fluctuations in local membrane geometry and to interactions with proteins.

Cell membranes play important roles in vivo both in shielding the cell interior from the surrounding environment and in cell function through lipid components of the membrane having roles in controlling protein function, cell signaling etc. We employ molecular dynamics simulations to explore the behavior of biologically realistic membrane models. Our simulations reveal nano-domain clustering of the glycolipid GM3 and to a lesser extent of the anionic lipid phosphatidylinositol 4,5-bisphophate (PIP₂). When including transmembrane proteins we are able to observe preferential interactions of known regulatory lipids (e.g. GM3, PIP₂ and cholesterol) with the proteins. Membrane curvature is shown to be coupled to the local lipid composition, suggestive of a link between lipid nano-domains and membrane geometry.

Prediction of the structure of a silk-like protein in oligomeric states using explicit and implicit solvent models

We perform Replica Exchange Molecular Dynamics (REMD) simulations on a silk-like protein design with amino-acid sequence [(Gly-Ala)3-Gly-Glu]5 to investigate the stability of a single protein, a dimer, a trimer and a tetramer made up of these proteins starting from β-roll and β-sheet structures in both explicit (TIP3P) and implicit (GBSA) solvent models. Our simulation results for the implicit solvent model agree with those for the explicit solvent model for simulation times up to the longest tested, being 30 ns per replica. From this we infer that the implicit solvent model that we use is reliable, allowing us to reach much longer time scales (up to 200 ns per replica). We find that the self-assembly of fibers of these proteins in solution must be a nucleated process, involving nuclei made up of at least three monomers. We also find that the conformation of the protein changes upon assembly, i.e., there is a transition from a disordered globular state to an ordered β-sheet structure in the self-assembled state of aggregates containing more than two monomers. This indicates that autosteric effects must be important in the polymerization of this protein, reminiscent of what is observed for β-amyloids. Our findings are consistent with recent experimental results on a protein with an amino acid sequence similar to that of the protein we study.

Assembly and stability of Salmonella enterica ser. Typhi TolC protein in POPE and DMPE

In this work we assessed the suitability of two different lipid membranes for the simulation of a TolC protein from Salmonella enterica serovar Typhi. The TolC protein family is found in many pathogenic Gram-negative bacteria including Vibrio cholera and Pseudomonas aeruginosa and acts as an outer membrane channel for expulsion of drug and toxin from the cell. In S. typhi, the causative agent for typhoid fever, the TolC outer membrane protein is an antigen for the pathogen. The lipid environment is an important modulator of membrane protein structure and function. We evaluated the conformation of the TolC protein in the presence of DMPE and POPE bilayers using molecular dynamics simulation. The S. typhi TolC protein exhibited similar conformational dynamics to TolC and its homologues. Conformational flexibility of the protein is seen in the C-terminal, extracellular loops, and α-helical region. Despite differences in the two lipids, significant similarities in the motion of the protein in POPE and DMPE were observed, including the rotational motion of the C-terminal residues and the partially open extracellular loops. However, analysis of the trajectories demonstrated effects of hydrophobic matching of the TolC protein in the membrane, particularly in the lengthening of the lipids and subtle movements of the protein's β-barrel towards the lower leaflet in DMPE. The study exhibited the use of molecular dynamics simulation in revealing the differential effect of membrane proteins and lipids on each other. In this study, POPE is potentially a more suitable model for future simulation of the S. typhi TolC protein.

A hydrophobic barrier deep within the inner pore of the TWIK-1 K2P potassium channel

Recent X-ray crystal structures of the two-pore domain (K2P) family of potassium channels have revealed a unique structural architecture at the point where the cytoplasmic bundle-crossing gate is found in most other tetrameric K(+) channels. However, despite the apparently open nature of the inner pore in the TWIK-1 (K2P1/KCNK1) crystal structure, the reasons underlying its low levels of functional activity remain unclear. In this study, we use a combination of molecular dynamics simulations and functional validation to demonstrate that TWIK-1 possesses a hydrophobic barrier deep within the inner pore, and that stochastic dewetting of this hydrophobic constriction acts as a major barrier to ion conduction. These results not only provide an important insight into the mechanisms which control TWIK-1 channel activity, but also have important implications for our understanding of how ion permeation may be controlled in similar ion channels and pores.

E. coli outer membrane and interactions with OmpLA

The outer membrane of Gram-negative bacteria is a unique asymmetric lipid bilayer composed of phospholipids (PLs) in the inner leaflet and lipopolysaccharides (LPSs) in the outer leaflet. Its function as a selective barrier is crucial for the survival of bacteria in many distinct environments, and it also renders Gram-negative bacteria more resistant to antibiotics than their Gram-positive counterparts. Here, we report the structural properties of a model of the Escherichia coli outer membrane and its interaction with outer membrane phospholipase A (OmpLA) utilizing molecular dynamics simulations. Our results reveal that given the lipid composition used here, the hydrophobic thickness of the outer membrane is ∼3 Å thinner than the corresponding PL bilayer, mainly because of the thinner LPS leaflet. Further thinning in the vicinity of OmpLA is observed due to hydrophobic matching. The particular shape of the OmpLA barrel induces various interactions between LPS and PL leaflets, resulting in asymmetric thinning around the protein. The interaction between OmpLA extracellular loops and LPS (headgroups and core oligosaccharides) stabilizes the loop conformation with reduced dynamics, which leads to secondary structure variation and loop displacement compared to that in a DLPC bilayer. In addition, we demonstrate that the LPS/PL ratios in asymmetric bilayers can be reliably estimated by the per-lipid surface area of each lipid type, and there is no statistical difference in the overall membrane structure for the outer membranes with one more or less LPS in the outer leaflet, although individual lipid properties vary slightly.

Are current atomistic force fields accurate enough to study proteins in crowded environments?

The high concentration of macromolecules in the crowded cellular interior influences different thermodynamic and kinetic properties of proteins, including their structural stabilities, intermolecular binding affinities and enzymatic rates. Moreover, various structural biology methods, such as NMR or different spectroscopies, typically involve samples with relatively high protein concentration. Due to large sampling requirements, however, the accuracy of classical molecular dynamics (MD) simulations in capturing protein behavior at high concentration still remains largely untested. Here, we use explicit-solvent MD simulations and a total of 6.4 µs of simulated time to study wild-type (folded) and oxidatively damaged (unfolded) forms of villin headpiece at 6 mM and 9.2 mM protein concentration. We first perform an exhaustive set of simulations with multiple protein molecules in the simulation box using GROMOS 45a3 and 54a7 force fields together with different types of electrostatics treatment and solution ionic strengths. Surprisingly, the two villin headpiece variants exhibit similar aggregation behavior, despite the fact that their estimated aggregation propensities markedly differ. Importantly, regardless of the simulation protocol applied, wild-type villin headpiece consistently aggregates even under conditions at which it is experimentally known to be soluble. We demonstrate that aggregation is accompanied by a large decrease in the total potential energy, with not only hydrophobic, but also polar residues and backbone contributing substantially. The same effect is directly observed for two other major atomistic force fields (AMBER99SB-ILDN and CHARMM22-CMAP) as well as indirectly shown for additional two (AMBER94, OPLS-AAL), and is possibly due to a general overestimation of the potential energy of protein-protein interactions at the expense of water-water and water-protein interactions. Overall, our results suggest that current MD force fields may distort the picture of protein behavior in biologically relevant crowded environments.

Protein behavior is strongly affected by highly crowded and interaction-rich environments, i.e., typical conditions in both biologically relevant systems, such as the cellular interior, and solution-based structural experiments, including NMR and different spectroscopies. On the other hand, primarily because of limited computational power, molecular dynamics (MD) simulations, a premier high-resolution method for analyzing structure, dynamics and interactions of proteins, have been predominantly used to study individual proteins at infinite dilution. To fill this gap, we use MD simulations to study the behavior of wild-type (aggregation-resistant) and oxidatively damaged (aggregation-prone) forms of villin headpiece at high concentration, and reveal unexpected limitations and inaccuracies of modern-day MD force fields when it comes to modeling proteins at physiologically or experimentally relevant concentrations.

Flexible gates generate occluded intermediates in the transport cycle of LacY

The major facilitator superfamily (MFS) transporter lactose permease (LacY) alternates between cytoplasmic and periplasmic open conformations to co-transport a sugar molecule together with a proton across the plasma membrane. Indirect experimental evidence suggested the existence of an occluded transition intermediate of LacY, which would prevent leaking of the proton gradient. As no experimental structure is known, the conformational transition is not fully understood in atomic detail. We simulated transition events from a cytoplasmic open conformation to a periplasmic open conformation with the dynamic importance sampling molecular dynamics method and observed occluded intermediates. Analysis of water permeation pathways and the electrostatic free-energy landscape of a solvated proton indicated that the occluded state contains a solvated central cavity inaccessible from either side of the membrane. We propose a pair of geometric order parameters that capture the state of the pathway through the MFS transporters as shown by a survey of available crystal structures and models. We present a model for the occluded state of apo-LacY, which is similar to the occluded crystal structures of the MFS transporters EmrD, PepT_So, NarU, PiPT and XylE. Our simulations are consistent with experimental double electron spin–spin distance measurements that have been interpreted to show occluded conformations. During the simulations, a salt bridge that has been postulated to be involved in driving the conformational transition formed. Our results argue against a simple rigid-body domain motion as implied by a strict “rocker-switch mechanism” and instead hint at an intricate coupling between two flexible gates.

•

The transport mechanism of LacY is hypothesized to involve an intermediate “occluded” state.

•

Such a state is observed in computer simulations of the conformational transitions.

•

Simulation data are validated with experimental double electron–electron spin resonance measurements.

•

The structural gating elements of LacY are identified.

•

Occluded LacY is similar to known occluded structures of homologous proteins.

Defining the membrane disruption mechanism of kalata B1 via coarse-grained molecular dynamics simulations

Kalata B1 has been demonstrated to have bioactivity relating to membrane disruption. In this study, we conducted coarse-grained molecular dynamics simulations to gain further insight into kB1 bioactivity. The simulations were performed at various concentrations of kB1 to capture the overall progression of its activity. Two configurations of kB1 oligomers, termed tower-like and wall-like clusters, were detected. The conjugation between the wall-like oligomers resulted in the formation of a ring-like hollow in the kB1 cluster on the membrane surface. Our results indicated that the molecules of kB1 were trapped at the membrane-water interface. The interfacial membrane binding of kB1 induced a positive membrane curvature, and the lipids were eventually extracted from the membrane through the kB1 ring-like hollow into the space inside the kB1 cluster. These findings provide an alternative view of the mechanism of kB1 bioactivity that corresponds with the concept of an interfacial bioactivity model.

An Anisotropic Coarse-Grained Model for Proteins Based On Gay-Berne and Electric Multipole Potentials

Gay–Berne anisotropic potential has been widely used to evaluate the nonbonded interactions between coarse-grained particles being described as elliptical rigid bodies. In this paper, we are presenting a coarse-grained model for twenty kinds of amino acids and proteins, based on the anisotropic Gay–Berne and point electric multipole (EMP) potentials. We demonstrate that the anisotropic coarse-grained model, namely GBEMP model, is able to reproduce many key features observed from experimental protein structures (Dunbrack Library), as well as from atomistic force field simulations (using AMOEBA, AMBER, and CHARMM force fields), while saving the computational cost by a factor of about 10–200 depending on specific cases and atomistic models. More importantly, unlike other coarse-grained approaches, our framework is based on the fundamental intermolecular forces with explicit treatment of electrostatic and repulsion-dispersion forces. As a result, the coarse-grained protein model presented an accurate description of nonbonded interactions (particularly electrostatic component) between hetero/homodimers (such as peptide–peptide, peptide–water). In addition, the encouraging performance of the model was reflected by the excellent correlation between GBEMP and AMOEBA models in the calculations of the dipole moment of peptides. In brief, the GBEMP model given here is general and transferable, suitable for simulating complex biomolecular systems.

The pore of voltage-gated potassium ion channels is strained when closed

Voltage-gated potassium channels form potassium-selective pores in cell membranes. They open or close in response to changes in the transmembrane potential and are essential for generating action potentials, and thus for the functioning of heart and brain. While a mechanism for how these channels close has been proposed, it is not clear what drives their opening. Here we use free energy molecular dynamics simulations to show that work must be done on the pore to reduce the kink in the pore-lining (S6) α-helices, thereby forming the helix bundle crossing and closing the channel. Strain is built up as the pore closes, which subsequently drives opening. We also determine the effect of mutating the PVPV motif that causes the kink in the S6 helix. Finally, an approximate upper limit on how far the S4 helix is displaced as the pore closes is estimated.

Voltage-gated potassium channels open and close in response to changes in transmembrane potential, but their opening mechanism is poorly understood. Here, free energy molecular dynamics simulations show that strain accumulates as the pore closes, which subsequently drives opening.

Energetics of Multi-Ion Conduction Pathways in Potassium Ion Channels

Potassium ion channels form pores in cell membranes, allowing potassium ions through while preventing the passage of sodium ions. Despite numerous high-resolution structures, it is not yet possible to relate their structure to their single molecule function other than at a qualitative level. Over the past decade, there has been a concerted effort using molecular dynamics to capture the thermodynamics and kinetics of conduction by calculating potentials of mean force (PMF). These can be used, in conjunction with the electro-diffusion theory, to predict the conductance of a specific ion channel. Here, we calculate seven independent PMFs, thereby studying the differences between two potassium ion channels, the effect of the CHARMM CMAP forcefield correction, and the sensitivity and reproducibility of the method. Thermodynamically stable ion-water configurations of the selectivity filter can be identified from all the free energy landscapes, but the heights of the kinetic barriers for potassium ions to move through the selectivity filter are, in nearly all cases, too high to predict conductances in line with experiment. This implies it is not currently feasible to predict the conductance of potassium ion channels, but other simpler channels may be more tractable.

Fragment-based modeling of membrane protein loops: successes, failures, and prospects for the future

Membrane proteins (MPs) have become a major focus in structure prediction, due to their medical importance. There is, however, a lack of fast and reliable methods that specialize in the modeling of MP loops. Often methods designed for soluble proteins (SPs) are applied directly to MPs. In this article, we investigate the validity of such an approach in the realm of fragment-based methods. We also examined the differences in membrane and soluble protein loops that might affect accuracy. We test our ability to predict soluble and MP loops with the previously published method FREAD. We show that it is possible to predict accurately the structure of MP loops using a database of MP fragments (0.5-1 Å median root-mean-square deviation). The presence of homologous proteins in the database helps prediction accuracy. However, even when homologues are removed better results are still achieved using fragments of MPs (0.8-1.6 Å) rather than SPs (1-4 Å) to model MP loops. We find that many fragments of SPs have shapes similar to their MP counterparts but have very different sequences; however, they do not appear to differ in their substitution patterns. Our findings may allow further improvements to fragment-based loop modeling algorithms for MPs. The current version of our proof-of-concept loop modeling protocol produces high-accuracy loop models for MPs and is available as a web server at http://medeller.info/fread.

A two-domain elevator mechanism for sodium/proton antiport

Sodium/proton (Na⁺/H⁺) antiporters, located at the plasma membrane in every cell, are vital for cell homeostasis¹. In humans, their dysfunction has been linked to diseases, such as, hypertension, heart failure and epilepsy and they are well-established drug targets². The best understood model system for Na⁺/H⁺ antiport is NhaA from Escherichia coli^1,3, where both EM and crystal structures are available^4-6. NhaA is made up of two distinct domains, a Core domain and a Dimerisation domain. In the NhaA crystal structure a cavity is located between the two domains providing access to the ion-binding site from the inward-facing surface of the protein^1,4. Like many Na⁺/H⁺ antiporters, the activity of NhaA is regulated by pH, only becoming active above pH 6.5, where a conformational change is thought to occur⁷. To date, the only reported NhaA crystal structure is of the low pH inactivated form⁴. Here, we describe the active-state structure of a Na⁺/H⁺ antiporter, NapA from Thermus thermophilus at 3 Å resolution, solved from crystals grown at pH 7.8. In the NapA structure, the Core and Dimerisation domains are in different positions to those seen in NhaA and a negatively charged cavity has now opened to the outside. The extracellular cavity allows access to a strictly conserved aspartate residue thought to directly coordinate ion-binding^1,8,9, a role supported here by molecular dynamics simulations. To alternate access to this ion-binding site, however, requires a surprisingly large rotation of the Core domain, some 20° against the Dimerisation interface. We conclude that despite their fast transport rates of up to 1500 ions/sec³, Na⁺/H⁺ antiporters operate by a two-domain rocking bundle model, revealing themes relevant to secondary-active transporters in general.

Substrate versus inhibitor dynamics of P-glycoprotein

By far the most studied multidrug resistance protein is P-glycoprotein. Despite recent structural data, key questions about its function remain. P-glycoprotein (P-gp) is flexible and undergoes large conformational changes as part of its function and in this respect, details not only of the export cycle, but also the recognition stage are currently lacking. Given the flexibility, molecular dynamics (MD) simulations provide an ideal tool to examine this aspect in detail. We have performed MD simulations to examine the behaviour of P-gp. In agreement with previous reports, we found that P-gp undergoes large conformational changes which tended to result in the nucleotide-binding domains coming closer together. In all simulations, the approach of the NBDs was asymmetrical in agreement with previous observations for other ABC transporter proteins. To validate the simulations, we make extensive comparison to previous cross-linking data. Our results show very good agreement with the available data. We then went on to compare the influence of inhibitor compounds bound with simulations of a substrate (daunorubicin) bound. Our results suggest that inhibitors may work by keeping the NBDs apart, thus preventing ATP-hydrolysis. On the other hand, repeat simulations of daunorubicin (substrate) in one particular binding pose suggest that the approach of the NBDs is not impaired and that the structure would be still be competent to perform ATP hydrolysis, thus providing a model for inhibition or substrate transport. Finally we compare the latter to earlier QSAR data to provide a model for the first part of substrate transport within P-gp.

Latest developments in experimental and computational approaches to characterize protein-lipid interactions

Understanding the functional roles of all the molecules in cells is an ultimate goal of modern biology. An important facet is to understand the functional contributions from intermolecular interactions, both within a class of molecules (e.g. protein-protein) or between classes (e.g. protein-DNA). While the technologies for analyzing protein-protein and protein-DNA interactions are well established, the field of protein-lipid interactions is still relatively nascent. Here, we review the current status of the experimental and computational approaches for detecting and analyzing protein-lipid interactions. Experimental technologies fall into two principal categories, namely solution-based and array-based methods. Computational methods include large-scale data-driven analyses and predictions/dynamic simulations based on prior knowledge of experimentally identified interactions. Advances in the experimental technologies have led to improved computational analyses and vice versa, thereby furthering our understanding of protein-lipid interactions and their importance in biological systems.

Memoir: template-based structure prediction for membrane proteins

Membrane proteins are estimated to be the targets of 50% of drugs that are currently in development, yet we have few membrane protein crystal structures. As a result, for a membrane protein of interest, the much-needed structural information usually comes from a homology model. Current homology modelling software is optimized for globular proteins, and ignores the constraints that the membrane is known to place on protein structure. Our Memoir server produces homology models using alignment and coordinate generation software that has been designed specifically for transmembrane proteins. Memoir is easy to use, with the only inputs being a structural template and the sequence that is to be modelled. We provide a video tutorial and a guide to assessing model quality. Supporting data aid manual refinement of the models. These data include a set of alternative conformations for each modelled loop, and a multiple sequence alignment that incorporates the query and template. Memoir works with both α-helical and β-barrel types of membrane proteins and is freely available at http://opig.stats.ox.ac.uk/webapps/memoir.

Multiscale simulations reveal conserved patterns of lipid interactions with aquaporins

Interactions of membrane proteins with lipid molecules are central to their stability and function. We have used multiscale molecular dynamics simulations to determine the extent to which interactions with lipids are conserved across the aquaporin (Aqp) family of membrane proteins. Simulation-based assessment of the lipid interactions made by Aqps when embedded within a simple phospholipid bilayer agrees well with the protein-lipid contacts determined by electron diffraction from 2D crystals. Extending this simulation-based analysis to all Aqps of known structure reveals a degree of conservation of such interactions across the Aqp structural proteome. Despite similarities in the binding orientations and interactions of the lipids, there do not appear to be distinct, high-specificity lipid binding sites on the surface of Aqps. Rather Aqps exhibit a more broadly conserved protein/lipid interface, suggestive of interchange between annular and bulk lipids, instead of a fixed annular "shell" of lipids.

The structural basis of ZMPSTE24-dependent laminopathies

Mutations in the nuclear membrane zinc metalloprotease ZMPSTE24 lead to diseases of lamin processing (laminopathies), such as the premature aging disease progeria and metabolic disorders. ZMPSTE24 processes prelamin A, a component of the nuclear lamina intermediate filaments, by cleaving it at two sites. Failure of this processing results in accumulation of farnesylated, membrane-associated prelamin A. The 3.4 angstrom crystal structure of human ZMPSTE24 has a seven transmembrane α-helical barrel structure, surrounding a large, water-filled, intramembrane chamber, capped by a zinc metalloprotease domain with the catalytic site facing into the chamber. The 3.8 angstrom structure of a complex with a CSIM tetrapeptide showed that the mode of binding of the substrate resembles that of an insect metalloprotease inhibitor in thermolysin. Laminopathy-associated mutations predicted to reduce ZMPSTE24 activity map to the zinc metalloprotease peptide-binding site and to the bottom of the chamber.

How long is a piece of loop?

Loops are irregular structures which connect two secondary structure elements in proteins. They often play important roles in function, including enzyme reactions and ligand binding. Despite their importance, their structure remains difficult to predict. Most protein loop structure prediction methods sample local loop segments and score them. In particular protein loop classifications and database search methods depend heavily on local properties of loops. Here we examine the distance between a loop's end points (span). We find that the distribution of loop span appears to be independent of the number of residues in the loop, in other words the separation between the anchors of a loop does not increase with an increase in the number of loop residues. Loop span is also unaffected by the secondary structures at the end points, unless the two anchors are part of an anti-parallel beta sheet. As loop span appears to be independent of global properties of the protein we suggest that its distribution can be described by a random fluctuation model based on the Maxwell-Boltzmann distribution. It is believed that the primary difficulty in protein loop structure prediction comes from the number of residues in the loop. Following the idea that loop span is an independent local property, we investigate its effect on protein loop structure prediction and show how normalised span (loop stretch) is related to the structural complexity of loops. Highly contracted loops are more difficult to predict than stretched loops.

The simulation approach to lipid-protein interactions

The interactions between lipids and proteins are crucial for a range of biological processes, from the folding and stability of membrane proteins to signaling and metabolism facilitated by lipid-binding proteins. However, high-resolution structural details concerning functional lipid/protein interactions are scarce due to barriers in both experimental isolation of native lipid-bound complexes and subsequent biophysical characterization. The molecular dynamics (MD) simulation approach provides a means to complement available structural data, yielding dynamic, structural, and thermodynamic data for a protein embedded within a physiologically realistic, modelled lipid environment. In this chapter, we provide a guide to current methods for setting up and running simulations of membrane proteins and soluble, lipid-binding proteins, using standard atomistically detailed representations, as well as simplified, coarse-grained models. In addition, we outline recent studies that illustrate the power of the simulation approach in the context of biologically relevant lipid/protein interactions.

Multiscale molecular dynamics simulations of membrane proteins

The time and length scales accessible by biomolecular simulations continue to increase. This is in part due to improvements in algorithms and computing performance, but is also the result of the emergence of coarse-grained (CG) potentials, which complement and extend the information obtainable from fully detailed models. CG methods have already proven successful for a range of applications that benefit from the ability to rapidly simulate spontaneous self-assembly within a lipid membrane environment, including the insertion and/or oligomerization of a range of "toy models," transmembrane peptides, and single- and multi-domain proteins. While these simplified approaches sacrifice atomistic level detail, it is now straightforward to "reverse map" from CG to atomistic descriptions, providing a strategy to assemble membrane proteins within a lipid environment, prior to all-atom simulation. Moreover, recent developments have been made in "dual resolution" techniques, allowing different molecules in the system to be modeled with atomistic or CG resolution simultaneously.