The electrostatics of solvent and membrane interfaces and the role of electronic polarizability

High-Dimensional Parameter Search Method to Determine Force Field Mixing Terms in Molecular Simulations

Molecular dynamics (MD) force fields for lipids and ions are typically developed independently of one another. In simulations consisting of both lipids and ions, lipid-ion interaction energies are estimated using a predefined set of mixing rules for Lennard-Jones (LJ) interactions. This, however, does not guarantee their reliability. In fact, compared to the quantum mechanical reference data, Lorentz-Berthelot mixing rules substantially underestimate the binding energies of Na⁺ ions with small-molecule analogues of lipid headgroups, yielding errors on the order of 80 and 130 kJ/mol, respectively, for methyl acetate and diethyl phosphate. Previously, errors associated with mixing force fields have been reduced using approaches such as "NB-fix" in which LJ interactions are computed using explicit cross terms rather than those from mixing rules. Building on this idea, we derive explicit lipid-ion cross terms that also may implicitly include many-body cooperativity effects. Additionally, to account for the interdependency between cross terms, we optimize all cross terms simultaneously by performing high-dimensional searches using our ParOpt software. The cross terms we obtain reduce the errors due to mixing rules to below 10 kJ/mol. MD simulation of the lipid bilayer conducted using these optimized cross terms resolves the structural discrepancies between our previous simulations and small-angle X-ray and neutron scattering experiments. These results demonstrate that simulations of lipid bilayers with ions that are accurate up to structural data from scattering experiments can be performed without explicit polarization terms. However, it is worth noting that such NB-fix cross terms are not based on any physical principle; a polarizable lipid model would be more realistic and is still desired. Our approach is generic and can be applied to improve the accuracies of simulations employing mixed force fields.

Molecular Dynamics Simulations Based on Polarizable Models Show that Ion Permeation Interconverts between Different Mechanisms as a Function of Membrane Thickness

Different mechanisms have been proposed to explain the permeation of charged compounds through lipid membranes. Overall, it is expected that an ion-induced defect permeation mechanism, where substantial membrane deformations accompany ion movement, should be dominant in thin membranes but that a solubility-diffusion mechanism, where ions partition into the membrane core with large associated dehydration energy costs, becomes dominant in thicker membranes. However, while this physical picture is intuitively reasonable, capturing the interconversion between these two permeation mechanisms in molecular dynamics (MD) simulations based on atomic models is challenging. In particular, simulations relying on nonpolarizable force fields are artificially unfavorable to the solubility-diffusion mechanism, as induced polarization of the nonpolar hydrocarbon is ignored, causing overestimated free energy costs for charged molecules to enter into this region of the membrane. In this study, all-atom MD simulations based on nonpolarizable and polarizable force fields are used to quantitatively characterize the permeation process for the arginine side chain analog methyl-guanidinium through bilayer membranes of mono-unsaturated phosphatidylcholine lipids with and without cholesterol, resulting in thicknesses spanning from ∼24 to ∼42 Å. With simulations based on a nonpolarizable force field, ion translocation can take place solely through an ion-induced defect mechanism, with free energy barriers increasing linearly from 14 to 40 kcal/mol, depending on the thickness. However, with simulations based on a polarizable force field, ion translocation is predominantly dominated by an ion-induced defect mechanism in thin membranes, which progressively converts to a solubility-diffusion mechanism as the membranes get thicker. The transition between the two mechanisms occurs at a thickness of ∼29 Å, with lipid tails of 22 or more carbon atoms. This situation appears to represent the upper limit for ion-induced defect permeation within the current polarizable models. Beyond this thickness, it becomes energetically preferable for the ion to dehydrate and partition into the membrane core-a phenomenon that cannot be captured using the nonpolarizable models. Induced electronic polarizability therefore leads not just to a shift in permeation energetics but to an interconversion between two strikingly different physical mechanisms. The result highlights the importance of induced polarizability in modeling lipid membranes.

Interfacial water and ion distribution determine ζ potential and binding affinity of nanoparticles to biomolecules

The molecular features that dictate interactions between functionalized nanoparticles and biomolecules are not well understood. This is in part because for highly charged nanoparticles in solution, establishing a clear connection between the molecular features of surface ligands and common experimental observables such as ζ potential requires going beyond the classical models based on continuum and mean field models. Motivated by these considerations, molecular dynamics simulations are used to probe the electrostatic properties of functionalized gold nanoparticles and their interaction with a charged peptide in salt solutions. Counterions are observed to screen the bare ligand charge to a significant degree even at the moderate salt concentration of 50 mM. As a result, the apparent charge density and ζ potential are largely insensitive to the bare ligand charge densities, which fall in the range of ligand densities typically measured experimentally for gold nanoparticles. While this screening effect was predicted by classical models such as the Manning condensation theory, the magnitudes of the apparent surface charge from microscopic simulations and mean-field models are significantly different. Moreover, our simulations found that the chemical features of the surface ligand (e.g., primary vs. quaternary amines, heterogeneous ligand lengths) modulate the interfacial ion and water distributions and therefore the interfacial potential. The importance of interfacial water is further highlighted by the observation that introducing a fraction of hydrophobic ligands enhances the strength of electrostatic binding of the charged peptide. Finally, the simulations highlight that the electric double layer is perturbed upon binding interactions. As a result, it is the bare charge density rather than the apparent charge density or ζ potential that better correlates with binding affinity of the nanoparticle to a charged peptide. Overall, our study highlights the importance of molecular features of the nanoparticle/water interface and underscores a set of design rules for the modulation of electrostatic driven interactions at nano/bio interfaces.

Computational Modeling of Realistic Cell Membranes

Cell membranes contain a large variety of lipid types and are crowded with proteins, endowing them with the plasticity needed to fulfill their key roles in cell functioning. The compositional complexity of cellular membranes gives rise to a heterogeneous lateral organization, which is still poorly understood. Computational models, in particular molecular dynamics simulations and related techniques, have provided important insight into the organizational principles of cell membranes over the past decades. Now, we are witnessing a transition from simulations of simpler membrane models to multicomponent systems, culminating in realistic models of an increasing variety of cell types and organelles. Here, we review the state of the art in the field of realistic membrane simulations and discuss the current limitations and challenges ahead.

Evaluating the Effects of Geometry and Charge Flux in Force Field Modeling

We apply a model for analyzing the importance of conformational charge flux to 11 molecules with the R-(CH₂) _n-R structure (R = Cl, F, OH, SH, COOH, CONH₂, and NH₂ and n = 4-6). Atomic charges were obtained by fitting to results from density functional theory calculations using the HLY procedure, and their geometry dependence is decomposed into contributions from changes in bond lengths, bond angles, and torsional angles. The torsional degrees of freedom are the main contribution to the conformational dependence of atomic charges and molecular dipole moments, but indirect effects due to changes in bond distances and angles account for ∼15% of the variations. While the magnitude of charge flux and geometry effects have been found to be independent of the number of internal degrees of freedom, the nature of the R- group has a moderate influence. The indirect effects are comparable for all of the R-groups and are approximately one-half the magnitude of the corresponding effects in peptide models. However, the magnitudes are different, yet the relative importance of geometry and charge flux effects are completely similar to those of the peptide models, which suggests that modeling the charge flux effects for changes in bond lengths, bond angles, and torsional angles should be considered for developing improved force fields.

Counting charges on membrane-bound peptides

Quantifying the number of charges on peptides bound to interfaces requires reliable estimates of (i) surface coverage and (ii) surface charge, both of which are notoriously difficult parameters to obtain, especially at solid/water interfaces. Here, we report the thermodynamics and electrostatics governing the interactions of l-lysine and l-arginine octamers (Lys₈ and Arg₈) with supported lipid bilayers prepared.

Quantifying the number of charges on peptides bound to interfaces requires reliable estimates of (i) surface coverage and (ii) surface charge, both of which are notoriously difficult parameters to obtain, especially at solid/water interfaces. Here, we report the thermodynamics and electrostatics governing the interactions of l-lysine and l-arginine octamers (Lys₈ and Arg₈) with supported lipid bilayers prepared from a 9 : 1 mixture of 1,2-dimyristoyl-sn-glycero-3-phosphocholine (DMPC) and 1,2-dimyristoyl-sn-glycero-3-phospho-(1′-rac-glycerol) (sodium salt) (DMPG) from second harmonic generation (SHG) spectroscopy, quartz crystal microbalance with dissipation monitoring (QCM-D) and nanoplasmonic sensing (NPS) mass measurements, and atomistic simulations. The combined SHG/QCM-D/NPS approach provides interfacial charge density estimates from mean field theory for the attached peptides that are smaller by a factor of approximately two (0.12 ± 0.03 C m^–2 for Lys₈ and 0.10 ± 0.02 C m^–2 for Arg₈) relative to poly-l-lysine and poly-l-arginine. These results, along with atomistic simulations, indicate that the surface charge density of the supported lipid bilayer is neutralized by the attached cationic peptides. Moreover, the number of charges associated with each attached peptide is commensurate with those found in solution; that is, Lys₈ and Arg₈ are fully ionized when attached to the bilayer. Computer simulations indicate Lys₈ is more likely than Arg₈ to “stand-up” on the surface, interacting with lipid headgroups through one or two sidechains while Arg₈ is more likely to assume a “buried” conformation, interacting with the bilayer through up to six sidechains. Analysis of electrostatic potential and charge distribution from atomistic simulations suggests that the Gouy–Chapman model, which is widely used for mapping surface potential to surface charge, is semi-quantitatively valid; despite considerable orientational preference of interfacial water, the apparent dielectric constant for the interfacial solvent is about 30, due to the thermal fluctuation of the lipid–water interface.

An averaged polarizable potential for multiscale modeling in phospholipid membranes

A set of average atom-centered charges and polarizabilities has been developed for three types of phospholipids for use in polarizable embedding calculations. The lipids investigated are 1,2-dimyristoyl-sn-glycero-3-phosphocholine, 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine, and 1-palmitoyl-2-oleoyl-sn-glycerol-3-phospho-L-serine given their common use both in experimental and computational studies. The charges, and to a lesser extent the polarizabilities, are found to depend strongly on the molecular conformation of the lipids. Furthermore, the importance of explicit polarization is underlined for the description of larger assemblies of lipids, that is, membranes. In conclusion, we find that specially developed polarizable parameters are needed for embedding calculations in membranes, while common non-polarizable point-charge force fields usually perform well enough for structural and dynamical studies. © 2017 Wiley Periodicals, Inc.

Permeability across lipid membranes

Molecular permeation through lipid membranes is a fundamental biological process that is important for small neutral molecules and drug molecules. Precise characterization of free energy surface and diffusion coefficients along the permeation pathway is required in order to predict molecular permeability and elucidate the molecular mechanisms of permeation. Several recent technical developments, including improved molecular models and efficient sampling schemes, are illustrated in this review. For larger penetrants, explicit consideration of multiple collective variables, including orientational, conformational degrees of freedom, are required to be considered in addition to the distance from the membrane center along the membrane normal. Although computationally demanding, this method can provide significant insights into the molecular mechanisms of permeation for molecules of medical and pharmaceutical importance. This article is part of a Special Issue entitled: Biosimulations edited by Ilpo Vattulainen and Tomasz Róg.

Effects of Lipid Composition on Bilayer Membranes Quantified by All-Atom Molecular Dynamics

Biological bilayer membranes typically contain varying amounts of lamellar and nonlamellar lipids. Lamellar lipids, such as dioleoylphosphatidylcholine (DOPC), are defined by their tendency to form the lamellar phase, ubiquitous in biology. Nonlamellar lipids, such as dioleoylphosphatidylethanolamine (DOPE), prefer instead to form nonlamellar phases, which are mostly nonbiological. However, nonlamellar lipids mix with lamellar lipids in biomembrane structures that remain overall lamellar. Importantly, changes in the lamellar vs nonlamellar lipid composition are believed to affect membrane function and modulate membrane proteins. In this work, we employ atomistic molecular dynamics simulations to quantify how a range of bilayer properties are altered by variations in the lamellar vs nonlamellar lipid composition. Specifically, we simulate five DOPC/DOPE bilayers at mixing ratios of 1/0, 3/1, 1/1, 1/3, and 0/1. We examine properties including lipid area and bilayer thickness, as well as the transmembrane profiles of electron density, lateral pressure, electric field, and dipole potential. While the bilayer structure is only marginally altered by lipid composition changes, dramatic effects are observed for the lateral pressure, electric field, and dipole potential profiles. Possible implications for membrane function are discussed.

Membrane permeation of a peptide: it is better to be positive

A joint experimental and computational study investigates the translocation of a tryptophan molecule through a phospholipid membrane. Time dependent spectroscopy of the tryptophan side chain determines the rate of permeation into 150 nm phospholipid vesicles. Atomically detailed simulations are conducted to calculate the free energy profiles and the permeation coefficient. Different charging conditions of the peptide (positive, negative, or zwitterion) are considered. Both experiment and simulation reproduce the qualitative trend and suggest that the fastest permeation is when the tryptophan is positively charged. The permeation mechanism, which is revealed by molecular dynamics simulations, is of a translocation assisted by a local defect. The influence of long-range electrostatic interactions, such as the membrane dipole potential on the permeation process, is not significant.

The Role of Broken Symmetry in Solvation of a Spherical Cavity in Classical and Quantum Water Models

Insertion of a hard sphere cavity in liquid water breaks translational symmetry and generates an electrostatic potential difference between the region near the cavity and the bulk. Here, we clarify the physical interpretation of this potential and its calculation. We also show that the electrostatic potential in the center of small, medium, and large cavities depends very sensitively on the form of the assumed molecular interactions for different classical simple point-charge models and quantum mechanical DFT-based interaction potentials, as reflected in their description of donor and acceptor hydrogen bonds near the cavity. These differences can significantly affect the magnitude of the scalar electrostatic potential. We argue that the result of these studies will have direct consequences toward our understanding of the thermodynamics of ion solvation through the cavity charging process.

Modeling peptide binding to anionic membrane pores

Peptide-induced pore formation in membranes can be dissected into two steps: pore formation and peptide binding to the pore. A computational method is proposed to study the second step in anionic membranes. The electrostatic potential is obtained from numerical solutions to the Poisson-Boltzmann equation and is then used in conjunction with IMM1 (implicit membrane model 1). A double charge layer model is used to incorporate the effects of the membrane dipole potential. Inhomogeneity of the charge density in the pore, characterized by explicit membrane simulations of toroidal pores, is included in the model. This approach was applied to two extensively studied peptides, magainin and melittin. In agreement with previous work, binding to toroidal pores is more favorable than binding to the flat membrane. The dependence of binding energy on anionic content exhibits different patterns for the two peptides, in correlation with the different lipid selectivity that has been observed experimentally.

Polarizable force fields

This chapter provides an overview of the most common methods for including an explicit description of electronic polarization in molecular mechanics force fields: the induced point dipole, shell, and fluctuating charge models. The importance of including polarization effects in biomolecular simulations is discussed, and some of the most important achievements in the development of polarizable biomolecular force fields to date are highlighted.

Charge equilibration force fields for molecular dynamics simulations of lipids, bilayers, and integral membrane protein systems

With the continuing advances in computational hardware and novel force fields constructed using quantum mechanics, the outlook for non-additive force fields is promising. Our work in the past several years has demonstrated the utility of polarizable force fields, those based on the charge equilibration formalism, for a broad range of physical and biophysical systems. We have constructed and applied polarizable force fields for lipids and lipid bilayers. In this review of our recent work, we discuss the formalism we have adopted for implementing the charge equilibration (CHEQ) method for lipid molecules. We discuss the methodology, related issues, and briefly discuss results from recent applications of such force fields. Application areas include DPPC-water monolayers, potassium ion permeation free energetics in the gramicidin A bacterial channel, and free energetics of permeation of charged amino acid analogs across the water-bilayer interface. This article is part of a Special Issue entitled: Membrane protein structure and function.

Molecular dynamics simulation of hydrated DPPC monolayers using charge equilibration force fields

We present results of molecular dynamics simulations of a model DPPC-water monolayer using charge equilibration (CHEQ) force fields, which explicitly account for electronic polarization in a classical treatment of intermolecular interactions. The surface pressure, determined as the difference between the monolayer and pure water surface tensions at 323 K, is predicted to be 22.92 ±1.29 dyne/cm, just slightly below the broad range of experimental values reported for this system. The surface tension for the DPPC-water monolayer is predicted to be 42.35 ±1.16 dyne/cm, in close agreement with the experimentally determined value of 40.9 dyne/cm. This surface tension is also consistent with the value obtained from DPPC monolayer simulations using state-of-the-art nonpolarizable force fields. The current results of simulations predict a monolayer-water potential difference relative to the pure water-air interface of 0.64 ±0.02 Volts, an improved prediction compared to the fixed-charge CHARMM27 force field, yet still overestimating the experimental range of 0.3 to 0.45 Volts. As the charge equilibration model is a purely charge-based model for polarization, the current results suggest that explicitly modeled polarization effects can offer improvements in describing interfacial electrostatics in such systems.

Influence of the membrane dipole potential on peptide binding to lipid bilayers

The implicit membrane model IMM1 is extended to include the membrane dipole potential and applied to molecular dynamics simulations of the helical peptides alamethicin, WALP23, influenza hemagglutinin fusion peptide, HIV fusion peptide, magainin, and the pre-sequence of cytochrome c oxidase subunit IV (p25). The results show that the orientation of the peptides in the membrane can be influenced by the dipole potential. The binding affinity of all peptides except for the hemagglutinin fusion peptide decreases upon increase of the dipole potential. The changes in both orientation and binding affinity are explained by the interaction of the dipole potential with the helix backbone dipole and ionic side-chains. In general, peptides that tend to insert the N-terminus in the membrane and/or have positively charged side chains will lose binding affinity upon increase of the dipole potential.

The role of membrane thickness in charged protein-lipid interactions

Charged amino acids are known to be important in controlling the actions of integral and peripheral membrane proteins and cell disrupting peptides. Atomistic molecular dynamics studies have shed much light on the mechanisms of membrane binding and translocation of charged protein groups, yet the impact of the full diversity of membrane physico-chemical properties and topologies has yet to be explored. Here we have performed a systematic study of an arginine (Arg) side chain analog moving across saturated phosphatidylcholine (PC) bilayers of variable hydrocarbon tail length from 10 to 18 carbons. For all bilayers we observe similar ion-induced defects, where Arg draws water molecules and lipid head groups into the bilayers to avoid large dehydration energy costs. The free energy profiles all exhibit sharp climbs with increasing penetration into the hydrocarbon core, with predictable shifts between bilayers of different thickness, leading to barrier reduction from 26 kcal/mol for 18 carbons to 6 kcal/mol for 10 carbons. For lipids of 10 and 12 carbons we observe narrow transmembrane pores and corresponding plateaus in the free energy profiles. Allowing for movements of the protein and side chain snorkeling, we argue that the energetic cost for burying Arg inside a thin bilayer will be small, consistent with recent experiments, also leading to a dramatic reduction in pK(a) shifts for Arg. We provide evidence that Arg translocation occurs via an ion-induced defect mechanism, except in thick bilayers (of at least 18 carbons) where solubility-diffusion becomes energetically favored. Our findings shed light on the mechanisms of ion movement through membranes of varying composition, with implications for a range of charged protein-lipid interactions and the actions of cell-perturbing peptides. This article is part of a Special Issue entitled: Membrane protein structure and function.

Coarse grained model for exploring voltage dependent ion channels

The relationship between the membrane voltage and the gating of voltage activated ion channels and other systems have been a problem of great current interest. Unfortunately, reliable molecular simulations of external voltage effects present a major challenge, since meaningful converging microscopic simulations are not yet available and macroscopic treatments involve major uncertainties in terms of the dielectric used and other key features. This work extends our coarse grained (CG) model to simulations of membrane/protein systems under external potential. Special attention is devoted to a consistent modeling of the effect of external potential due to the electrodes, emphasizing semimacroscopic description of the electrolytes in the solution regions between the membranes and the electrodes, as well as the coupling between the combined potential from the electrodes plus the electrolytes and the protein ionized groups. We also provide a clear connection to microscopic treatment of the electrolytes and thus can explore possible conceptual problems that are hard to resolve by other current approaches. For example, we obtain a clear description of the charge distribution in the entire electrolyte system, including near the electrodes in membrane/electrodes systems (where continuum models do not seem to provide the relevant results). Furthermore, the present treatment provides an insight on the distribution of the electrolyte charges before and after equilibration across the membrane, and thus on the nature of the gating charge. The different aspects of the model have been carefully validated by considering problems ranging for the simple Debye-Huckel, and the Gouy-Chapman models to the evaluation of the electrolyte distribution between two electrodes, as well as the effect of extending the simulation system by periodic replicas. Overall the clear connection to microscopic descriptions combined with the power of the CG modeling seems to offer a powerful tool for exploring the balance between the protein conformational energy and the interaction with the external potential in voltage activated channels. To illustrate these features we present a preliminary study of the gating charge in the voltage activated Kv1.2 channel, using the actual change in the electrolyte charge distribution rather than the conventional macroscopic estimate. We also discuss other special features of the model, which include the ability to capture the effect of changes in the protonation states of the protein residues during the close to open voltage induced transition. This article is part of a Special Issue entitled: Membrane protein structure and function.

Effect of membrane tension on the electric field and dipole potential of lipid bilayer membrane

The dipole potential of lipid bilayer membrane controls the difference in permeability of the membrane to oppositely charged ions. We have combined molecular dynamics (MD) simulations and experimental studies to determine changes in electric field and electrostatic potential of 1,2-dioleoyl-sn-glycero-3-phosphocholine (DOPC) lipid bilayer in response to applied membrane tension. MD simulations based on CHARMM36 force field showed that electrostatic potential of DOPC bilayer decreases by ~45mV in the physiologically relevant range of membrane tension values (0 to 15dyn/cm). The electrostatic field exhibits a peak (~0.8×10(9)V/m) near the water/lipid interface which shifts by 0.9Å towards the bilayer center at 15dyn/cm. Maximum membrane tension of 15dyn/cm caused 6.4% increase in area per lipid, 4.7% decrease in bilayer thickness and 1.4% increase in the volume of the bilayer. Dipole-potential sensitive fluorescent probes were used to detect membrane tension induced changes in DOPC vesicles exposed to osmotic stress. Experiments confirmed that dipole potential of DOPC bilayer decreases at higher membrane tensions. These results are suggestive of a potentially new mechanosensing mechanism by which mechanically induced structural changes in the lipid bilayer membrane could modulate the function of membrane proteins by altering electrostatic interactions and energetics of protein conformational states.

On the role of anionic lipids in charged protein interactions with membranes

We investigate the role of anionic lipids in the binding to, and subsequent movement of charged protein groups in lipid membranes, to help understand the role of membrane composition in all membrane-active protein sequences. We demonstrate a small effect of phosphatidylglycerol (PG) lipids on the ability of an arginine (Arg) side chain to bind to, and cross a lipid membrane, despite possessing a neutralizing charge. We observe similar membrane deformations in lipid bilayers composed of phosphatidylcholine (PC) and PC/PG mixtures, with comparable numbers of water and lipid head groups pulled into the bilayer hydrocarbon core, and prohibitively large ~20 kcal/mol barriers for Arg transfer across each bilayer, dropping by just 2-3 kcal/mol due to the binding of PG lipids. We explore the causes of this small effect of introducing PG lipids and offer an explanation in terms of the limited membrane interaction for the choline groups of PC lipids bound to the translocating ion. Our calculations reveal a surprising lack of preference for Arg binding to PG lipids themselves, but a small increase in interfacial binding affinity for lipid bilayers containing PG lipids. These results help to explain the nature of competitive lipid binding to charged protein sequences, with implications for a wide range of membrane binding domains and cell perturbing peptides.

Electrostatics of deformable lipid membranes

It was recently demonstrated that significant local deformations of biological membranes take place due to the fields of charged peptides and ions, challenging the standard model of membrane electrostatics. The ability of ions to retain their immediate hydration environment, combined with the lack of sensitivity of permeability to ion type or even ion pairs, led us to question the extent to which hydration energetics and electrostatics control membrane ion permeation. Using the arginine analog methyl-guanidinium as a test case, we find that although hydrocarbon electronic polarizability causes dramatic changes in ion solvation free energy, as well as a significant change (approximately 0.4 V) in the membrane dipole potential, little change in membrane permeation energetics occurs. We attribute this to compensation of solvation terms from polar and polarizable nonpolar components within the membrane, and explain why the dipole potential is not fully sensed in terms of the locally deformed bilayer interface. Our descriptions provide a deeper understanding of the translocation process and allow predictions for poly-ions, ion pairs, charged lipids, and lipid flip-flop. We also report simulations of large hydrophobic-ion-like membrane defects and the ionophore valinomycin, which exhibit little membrane deformation, as well as hydrophilic defects and the ion channel gramicidin A, to provide parallels to membranes deformed by unassisted ion permeation.