Effect of n-alkanes on membranes in lipid-water systems

The changes in leaflet thickness and surface area per lipid molecule as a function of added amount of $\text{n-}\text{alkane}$ solvents have been studied in a number of lipid water systems by low angle X-ray diffraction techniques. The most probable site of accumulation of the $\text{n-}\text{alkane}$ is identified as the middle of the leaflet as opposed to intercalation with lipid hydrocarbon chains. Adsorption of $\text{n-}\text{alkane}$ depends on alkyl chain length and the organisation of the lipid hydrocarbon chains in the lipid water phases. Attention is drawn to the possible relationship between these results, the effect of $\text{n-}\text{alkanes}$ on isolated lipid bilayers and their effects as anaesthetics.     

Water permeation through the lipid bilayer membrane. Test of the liquid hydrocarbon model

According to the liquid hydrocarbon model, the lipid bilayer is viewed simply as a thin slice of bulk hydrocarbon liquid. This allows the water permeability of the bilayer to be calculated from bulk properties. In this paper the prediction of the liquid hydrocarbon model is compared with the known water permeability coefficient of the glycerol monoolein/n-hexadecane bilayer (Fettiplace, R. (1978) Biochim. Biophys. Acta 513, 1–10). As the alkyl chain of glycerol monoolein is equivalent to 8-heptadecene, the water permeability coefficient of 8-heptadecene/n-hexadecane mixtures was measured for temperatures between 20 and 35°C. The mole fraction of n-hexadecane in the bulk liquid was chosen at each temperature to match the known mole fraction of n-hexadecane in the bilayer (White, S. (1976) Nature 262, 421–422). The predicted water permeability coefficient agrees with the measured value at 32°C but is 40% above the measured value at 20°C. The apparent activation energy predicted by the liquid hydrocarbon model is 9.0 ± 0.3 kcal/mol, while the measured value is 14.2 ± 1.0 kcal/mol. The failure of the liquid hydrocarbon model probably results from a different molecular organization of the hydrocarbon chains in the bilayer and in the bulk liquid.     

The thickness of monoolein lipid bilayers as determined from reflectance measurements

Measurements of the reflectance of monoolein $\text{n-}\text{alkane}$ and monoolein/squalene lipid bilayers have been made. The total thickness of the bilayer was calculated from the dependence of reflectance on the refractive index of the aqueous salt or sucrose solution surrounding the bilayer. The total thickness was then compared to the thickness of the hydrocarbon chain region as determined from capacitance measurements. From this comparison, we found that the thickness of each polar region of the bilayers in salt solutions was $\text{0.5 ± 0.1}\text{nm}$, independent of the hydrocarbon solvent used. When the aqueous solutions contained sucrose, each polar region was approx. 0.9 nm thick. When $\text{n-}\text{tetradecane}$ and $\text{n-}\text{hexadecane}$ were used as solvents, microlenses of solvent trapped in the monoolein bilayer increased the reflectance. After about one hour, the coalescence of microlenses into larger lenses allowed the reflectance of the bilayer alone to be measured. The use of reflectance to measure the thickness of monoolein bilayers appears to be consistent with other methods and to give useful information about the structure of lipid bilayers.     

The structure of a model membrane in relation to the viscoelastic properties of the red cell membrane

The molecular arrangement within a lamellar structure composed of human erythrocyte lipids is determined. The 45 A thick lipid layer, in water, is filled in the interior with a liquid-like configuration of the hydrocarbon chains of phospholipid molecules and is covered on both sides by their hydrophilic polar groups. Cholesterol is located so that part of its steroid nucleus is between the polar groups of the phospholipid molecules while the rest of the molecule extends into the inner hydrocarbon layer. This lipid leaflet would be expected to have the mechanical properties of a purely liquid surface, as other authors have shown for the "black" lipid membranes. Data are presented which demonstrate that the intact erythrocyte membrane is a tough viscoelastic substance with a Young''s modulus of 10⁶–10⁸ dynes/cm² and a viscosity of 10⁷–10¹⁰ poises. The parameters and the kinetics of membrane breakdown are incompatible with the model system of pure lipid. Caution must be exercised in applying various data on the model systems to intact membranes.     

Voltage-induced capacitance relaxation of lipid bilayer membranes Effects of membrane composition

The specific capacity of black lipid membranes of different phospholipids dissolved in n-alkanes was investigated. The hydrocarbon thickness of these membranes, as calculated from the electric capacity with a dielectric constant of 2.1, was in most cases close to 5 nm. It was found that the specific capacity is not constant with time after blackening. It shows a linear time dependence characteristic for each lipid/solvent system.The influence of a transmembrane potential on the capacity of the membranes was measured. It was shown that the extent of the capacity change, obtained 10 s after applying a voltage, was strongly dependent on the lipid composition as well as the solvent content of the membranes. The capacity change of the membranes seems to be cause mainly by a thickness change and not by an area increase of the membranes.     

Lateral tensions and pressures in membranes and lipid monolayers

The effects of lateral tension on the properties of membranes are often explained in comparison with analogous experiments on monolayers, which yield more detailed data. To calculate the effects of changes in tension on the composition of, or incorporation of amphiphiles into membranes we examine (i) the fidelity of the monolayer analogy, (ii) the range of possible tensions in a membrane, and the way in which tensions arise and (iii) the equilibrium partitioning of amphiphiles between aqueous solution and a bilayer under tension. We argue that, at the same areas per molecule, a monolayer at an $\text{n-}\text{alkane}\text{/}\text{water}$ interface is a closer analogy of the lipid bilayer than a monolayer at an air/water interface. Next, we show from a thermodynamic argument that changes in membrane tension can affect the absorption of very large amphiphiles such as proteins, but that physiological tensions are unlikely to affect the absorption of lipids or drugs. Finally we consider the possibility that the measured bulk tension in a complicated membrane such as that of the erythrocyte may be larger than the local tension in the fluid mosaic portions, and suggest a model which explains the ability of the erythrocyte membrane to withstand much higher tensions than other biological membranes and lipid bilayers.     

Effects of n-alkanes on the morphology of lipid bilayers A freeze-fracture and negative stain analysis

The effect of $\text{n-}\text{alkanes}$ on the ultrastructure of lipid bilayers has been investigated using freeze-fracture and negative stain electron microscopy. It has been found that the morphology of bilayers containing the long alkane tetradecane is quite different from bilayers containing the short alkane hexane. The smooth fracture faces of gel and liquid crystalline state bilayers are unmodified by tetradecane. However, hexane dramatically alters the hydrophobic bilayer interior, producing large (20 to 50 nm) mounds and depressions in the fracture faces. The fracture steps in these multilayer preparations containing hexane are variable in thickness and often considerably wider than the corresponding fracture steps in multilayers which contain tetradecane or are solvent-free. Alkanes also modify the structure of the P_β′ or ‘banded’ phase of phosphatidylcholine bilayers. The incorporation of tetradecane removes the banded structure from both the bilayer's hydrophilic surface, as viewed by negative staining, and the bilayer's hydrophobic interior, as viewed by the freeze-fracture technique. These results are consistent with X-ray diffraction data which imply that long alkanes are primarily located between adjacent lipid hydrocarbon chains in each monolayer of the bilayer, while short alkanes can partition into the geometric center of the bilayer between apposing monolayers.     

Hydrocarbons of Rhodnius prolixus, a Chagas disease vector

The surface hydrocarbons of the blood-sucking insect, Rhodnius prolixus, a major Chagas disease vector in Venezuela, Colombia and Central America, were characterized by capillary gas chromatography coupled to mass spectrometry (CGC-MS). A total of 54 single or multicomponent peaks of saturated, straight-chain and methyl-branched hydrocarbons were identified. Major n-alkanes were n-C27, n-C29, n-C31 and n-C33 hydrocarbons. In the branched fraction, methyl groups were at positions 3, 5, 7, 11, 13, 15 and 17- for monomethyl isomers, and separated by three or five methylene groups for the trimethyl or tetramethyl derivatives. For the higher molecular weight components of 37, 39 and 41 atoms in the carbon skeleton, the di-, tri- and tetramethyl branches were usually separated by three or five, and sometimes 7, 11 or 13, methylene groups. The internal hydrocarbon pool contained larger amounts of the higher molecular weight methyl-branched components. Qualitative differences among epicuticular and internal hydrocarbon compositions were detected, both in adult and nymphal stages. No significant sexual dimorphism was detected, but a significant shift in the major n-alkane components was evident from the nymphal to the adult stage, differing also in the relative amounts of the higher molecular weight methyl-branched chains. Comparison of the hydrocarbon components to that of other Chagas disease vectors is discussed.     

Initial membrane reaction in peptidoglycan synthesis. Perturbation of lipid-phospho-N-acetylmuramyl-pentapeptide translocase interactions by n-Butanol

$\text{Phospho}\text{-N-}\text{acetylmuramyl-pentapeptide translocase}$, the initial membrane enzyme in the biosynthesis of peptidoglycan, requires a lipid microenvironment for function. $\text{n-}\text{Butanol}$ was reversibly intercalated into membranes to perturb the hydrophobic interactions in this microenvironment in order to define further the role of lipid. In the concentration range for maximal stimulation of enzymic activity (0.12–0.18 M), $\text{n-}\text{butanol}$ causes a 40% decrease in the fluorescence emission of the dansylated product, undecaprenyl $\text{diphosphate}\text{-(N}{}^{\mathrm{?}}\text{-}\text{dansyl)pentapeptide}$. Since no change in emission maximum occurs below 22°C in the presence of 0.12 M $\text{n-}\text{butanol}$, it is concluded that intercalation of this alkanol causes an increase in fluidity. Above 22°C this concentration of $\text{n-}\text{butanol}$ causes both a decrease in the fluorescence emission and a red shift in the emission maximum. It is concluded that a polarity change as well as fluidity change occurs above 22°C. $\text{n-}\text{Butanol}$ also causes a significant change in the phase transition experienced by the dansylated lipid product. Thus, it is possible with $\text{n-}\text{alkanols}$, e.g. $\text{n-}\text{butanol}$, to perturb lipid-translocase interactions resulting in an increase in fluidity in the microenvironment of the enzyme. This change in fluidity correlates with a stimulation of enzymic activity.     

The water permeability of the monoolein/triolein bilayer membrane

The water permeability of the lipid bilayer can be used as a probe of membrane structure. A simple model of the bilayer, the liquid hydrocarbon model, views the membrane as a thin slice of bulk hydrocarbon liquid. A previous study (Petersen, D. (1980) Biochim. Biophys. Acta 600, 666–677) showed that this model does not accurately predict the water permeability of the monoolein/n-hexadecane bilayer: the measured activation energy for water permeation is 50% above the predicted value. From this it was inferred that the hydrocarbon chains in the lipid bilayer are more ordered than in the bulk hydrocarbon liquid. The present study tests the liquid hydrocarbon model for the monoolein/triolein bilayer, which has been shown to contain very little triolein in the plane of the membrane (Waldbillig, R.C. and Szabo, G. (1979) Biochim. Biophys. Acta 557, 295–305). Measurements of the water permeability coefficient of the bilayer are compared with predictions of the liquid hydrocarbon model based on measurements of the water permeability coefficient of bulk 8-heptadecene. The predicted and measured values agree quite closely over the temperature range studied (15–35°C): the predicted activation energy is 11.1±0.2 kcal/mol, whereas the measured activation energy for the bilayer is 9.8±0.7 kcal/mol. This close agreement is in contrast with the monoolein/n-hexadecane results and suggests that, insofar as water permeation is concerned, the liquid hydrocarbon model quite closely represents the monoolein/triolein bilayer.     

Planar supported bilayer polymers formed from bis-diene lipids by Langmuir-Blodgett deposition and UV irradiation

Substrate-supported lipid bilayers have been prepared from bis-diene functionalized phosphorylcholine (PC) lipids and polymerized by UV irradiation. The overall bilayer structure is largely preserved upon removal from water, although significant loss of material occurs from the upper leaflet of the bilayer, likely due to desorption at the air/water interface. The morphology and surface structure of the bilayer, as observed by AFM, indicate a substantially different arrangement of the lipids in the hydrated and dehydrated states, presumably due to the loss of water from the near surface region. These changes have been correlated with infrared spectral shifts sensitive to the conformation of the hydrocarbon chains. Protein adsorption studies show that rehydrated, polymerized bilayers retain a degree of resistance to BSA adsorption intermediate between model hydrophobic and fluid PC lipid bilayer surfaces. The degree of protein adsorption is correlated with desorption of material from the upper leaflet of the bilayer upon drying, which produces voids at which hydrophobically driven protein adsorption occurs.     

Acyl-chain methyl distributions of liquid-ordered and -disordered membranes

A central feature of the lipid raft concept is the formation of cholesterol-rich lipid domains. The introduction of relatively rigid cholesterol molecules into fluid liquid-disordered (L_d) phospholipid bilayers can produce liquid-ordered (L_o) mixtures in which the rigidity of cholesterol causes partial ordering of the flexible hydrocarbon acyl chains of the phospholipids. Several lines of evidence support this concept, but direct structural information about L_o membranes is lacking. Here we present the structure of L_o membranes formed from cholesterol and dioleoylphosphatidylcholine (DOPC). Specific deuteration of the DOPC acyl-chain methyl groups and neutron diffraction measurements reveal an extraordinary disorder of the acyl chains of neat L_d DOPC bilayers. The disorder is so great that >20% of the methyl groups are in intimate contact with water in the bilayer interface. The ordering of the DOPC acyl chains by cholesterol leads to retraction of the methyl groups away from the interface. Molecular dynamics simulations based on experimental systems reveal asymmetric transbilayer distributions of the methyl groups associated with each bilayer leaflet.     

Phospholipid phase transitions. Effects of n-alcohols, n-monocarboxylic acids, phenylalkyl alcohols and quatenary ammonium compounds

The interactions of a series of alcohols, acids and quaternary ammonium salts with a phosphatidylcholine-water model biomembrane (dipalmitoyl phosphatidylcholine) system have been studied using differential scanning calorimetry. In particular the effects of these molecules upon the lipid endothermic phase transitions were investigated over a range of concentrations. A variety of effects was observed. (a) Those molecules which shift or broaden the main lipid transition can also remove the pretransition endotherm. (b) $\text{n-}\text{Alcohols}$ and $\text{n-}\text{monocarboxylic}$ acids containing the same number of carbon atoms have very similar effects at molar concentrations up to 40%. Those molecules containing 12 or more carbon atoms raise the main lipid phase transition whilst those molecules containing 10 or less carbon atoms lower this transition temperature. (c) The phase diagram of stearoyl alcohol in the phosphatidylcholine-water system shows the formation of lipid-alcohol complexes. (d) Alkyl trimethyl ammonium bromides showed behaviour which differs considerably from $\text{n-}\text{alcohols}$ and $\text{n-}\text{carboxylic}$ acids of the same chain length. (e) Other alkyltrialkyl and tetraalkylammonium bromides show that a variety of effects on the lipid phase transition can be obtained. (f) With the homologous series of phenylalkyl alcohols from benzyl alcohol to 4-phenyl butanol increasing the number of methylenes between the terminal OH and the benzene ring leads to greater interaction between solute and bilayer.The range of different effects obtained with the compounds studied offers a means for introducing various degrees and types of perturbation into membrane systems.     

Modulation of folding and assembly of the membrane protein bacteriorhodopsin by intermolecular forces within the lipid bilayer.

Three different lipid systems have been developed to investigate the effect of physicochemical forces within the lipid bilayer on the folding of the integral membrane protein bacteriorhodopsin. Each system consists of lipid vesicles containing two lipid species, one with phosphatidylcholine and the other with phosphatidylethanolamine headgroups, but the same hydrocarbon chains: either L-alpha-1, 2-dioleoyl, L-alpha-1,2-dipalmitoleoyl, or L-alpha-1,2-dimyristoyl. Increasing the mole fraction of the phosphatidylethanolamine lipid increases the desire of each monolayer leaflet in the bilayer to curve toward water. This increases the torque tension of such monolayers, when they are constrained to remain flat in the vesicle bilayer. Consequently, the lateral pressure in the hydrocarbon chain region increases, and we have used excimer fluorescence from pyrene-labeled phosphatidylcholine lipids to probe these pressure changes. We show that bacteriorhodopsin regenerates to about 95% yield in vesicles of 100% phosphatidylcholine. The regeneration yield decreases as the mole fraction of the corresponding phosphatidylethanolamine component is increased. The decrease in yield correlates with the increase in lateral pressure which the lipid chains exert on the refolding protein. We suggest that the increase in lipid chain pressure either hinders insertion of the denatured state of bacterioopsin into the bilayer or slows a folding step within the bilayer, to the extent that an intermediate involved in bacteriorhodopsin regeneration is effectively trapped.     

Molecular Dynamics Simulations of Phospholipid Bilayers

Abstract

Molecular dynamics (MD) simulations at 37°C have been performed on three phospholipid bilayer systems composed of the lipids DLPE, DOPE, and DOPC. The model used included 24 explicit lipid molecules and explicit waters of solvation in the polar head group regions, together with constant-pressure periodic boundary conditions in three dimensions. Using this model, a MD simulation samples part of an infinite planar lipid bilayer. The lipid dynamics and packing behavior were characterized. Furthermore, using the results of the simulations, a number of diverse properties including bilayer structural parameters, hydrocarbon chain order parameters, dihedral conformations, electron density profile, hydration per lipid, and water distribution along the bilayer normal were calculated. Many of these properties are available for the three lipid systems chosen, making them well suited for evaluating the model and protocols used in these simulations by direct comparisons with experimental data. The calculated MD behavior, chain disorder, and lipid packing parameter, i.e. the ratio of the effective areas of hydrocarbon tails and head group per lipid (a_t/a_h), correctly predict the aggregation preferences of the three lipids observed experimentally at 37°C, namely: a gel bilayer for DLPE, a hexagonal tube for DOPE, and a liquid crystalline bilayer for DOPC. In addition, the model and conditions used in the MD simulations led to good agreement of the calculated properties of the bilayers with available experimental results, demonstrating the reliability of the simulations. The effects of the cis unsaturation in the hydrocarbon chains of DOPE and DOPC, compared to the fully saturated one in DLPE, as well as the effects of the different polar head groups of PC and PE with the same unsaturated chains on the lipid packing and bilayer structure have been investigated. The results of these studies indicate the ability of MD methods to provide molecular-level insights into the structure and dynamics of lipid assemblies.     

Structure of molecular aggregates of 1-(3-sn-phosphatidyl)-l-myo-inositol 3,4-bis(phosphate) in water

The molecular organization of 1-(3-sn-phosphatidyl)-l-myo-inositol 3,4-bis-(phosphate)/water systems is investigated over a wide range of lipid concentrations using X-ray diffraction, calorimetry, analytical ultracentrifugation, densitometry and viscometry.At high lipid concentrations, the lipid molecules are found to form a lamellar phase. The repeat distance increases from 60 to 120 Å with increasing water content to 70 wt% and the surface area per lipid molecule increases from 41.7 Ų to a limiting value of 100 Ų.On the other hand, at very low lipid concentrations the molecules are found to form not vesicles but micelles, the total molecular weight of which takes a value of 93 000.This finding revises the prevalent view that lipids containing two (or more) hydrocarbon chains form extended bilayers or vesicles, whereas single chained lipids form micelles. (Tanford, C.(1972) J. Phys. Chem. 76, 3020–3024).     

Phase diagram of lipid A from Salmonella minnesota and Escherichia coli rough mutant lipopolysaccharide

We have reported here on the structural polymorphism of lipid A, the “endotoxic principle” of bacterial lipopolysaccharide. For lipid A of rough mutant lipopolysaccharide from Salmonella minnesota and Escherichia coli, the three-dimensional supramolecular structures were determined with x-ray diffraction utilizing synchrotron radiation. The investigations were performed in the water concentration range 10 to 95% by weight, at [lipid A]:[Mg²⁺] molar ratios from 1:0 to 0.1:1, and in the temperature range from 20 to 70°C. These data were correlated with measurements of the β→α phase behaviour which was monitored with differential scanning calorimetry and Fourier-transform infrared spectroscopy. We found that the transition temperature of the acyl chains ranges—in the absence of Mg²⁺—from 45°C at high to 56°C at low water content, and—at an equimolar content of Mg²⁺—from 52°C at high to 59°C at low water concentrations. In the gel phase—in which the lipid A acyl chains are more disordered than those from saturated phospholipids—cubic phases are adopted at high water content (>60%) and at high [lipid A):[Mg²⁺] molar ratios. At low water contents, lamellar states are assumed exclusively. In the liquid crystalline state of lipid A, the hexagonal H_II, state is adopted under all conditions. The structural variability of lipid A is highest at high water concentrations, and structural changes may be induced by only slight changes in temperature, water content, and Mg²⁺ concentration. Under physiological conditions, however, the lipid A assemblies exhibit a strong preference to cubic structures.     

Simulation study of a gramicidin/lipid bilayer system in excess water and lipid. I. Structure of the molecular complex

This paper reports on a simulation of a gramicidin channel inserted into a fluid phase DMPC bilayer with 100 lipid molecules. Two lipid molecules per leaflet were removed to insert the gramicidin, so the resulting preparation had 96 lipid molecules and 3209 water molecules. Constant surface tension boundary conditions were employed. Like previous simulations with a lower lipid/gramicidin ratio (Woolf, T. B., and B. Roux. 1996. Proteins: Struct., Funct., Genet. 24:92-114), it is found that tryptophan-water hydrogen bonds are more common than tryptophan-phospholipid hydrogen bonds. However, one of the tryptophan NH groups entered into an unusually long-lived hydrogen bonding pattern with two glycerol oxygens of one of the phospholipid molecules. Comparisons were made between the behavior of the lipids adjacent to the channel with those farther away. It was found that hydrocarbon chains of lipids adjacent to the channel had higher-order parameters than those farther away. The thickness of the lipid bilayer immediately adjacent to the channel was greater than it was farther away. In general, the lipids adjacent to the membrane had similar orientations to those seen by Woolf and Roux, while those farther away had similar orientations to those pertaining before the insertion of the gramicidin. A corollary to this observation is that the thickness of the hydrocarbon region adjacent to the gramicidin was much thicker than what other studies have identified as the "hydrophobic length" of the gramicidin channel.