Cholesterol-Induced Structural Changes in Saturated Phospholipid

Apr 6, 2017 - molar concentration (χc) up to which the lipid−cholesterol miscibility gradually increases and then further increase in the concentra...
0 downloads 0 Views 7MB Size
Download PDF
Article pubs.acs.org/JPCB

Cholesterol-Induced Structural Changes in Saturated Phospholipid Model Membranes Revealed through X‑ray Scattering Technique Rajendra P. Giri, Abhijit Chakrabarti, and Mrinmay K. Mukhopadhyay* Saha Institute of Nuclear Physics, HBNI, 1/AF, Bidhannagar, Kolkata 700064, India ABSTRACT: Lateral and out-of-plane organization of cholesterol and its effect on regulating the physicochemical properties of zwitterionic phospholipid model membranes have been investigated by a pressure− area isotherm study from the Langmuir monolayer, atomic force microscopy (AFM), and X-ray reflectivity (XRR) measurements from supported binary monolayer films. The systematic isotherm studies on the Langmuir monolayer of phospholipids and the subsequent extraction of excess Gibbs free energy (ΔGexc) revealed the mechanism of cholesterol interaction and the molecular cooperativeness for different arrangements in the phospholipid model membranes. We have found a critical cholesterol molar concentration (χc) up to which the lipid−cholesterol miscibility gradually increases and then further increase in the concentration leads to an inhomogeneous structure formation similar to raft structures. The thickening in the lipid acyl chain and the subsequent lowering of the lipid head group thickness up to χc are also evident from the XRR study. Beyond χc, large-sized domains are observed in the AFM images from the deposited monolayer. χc has also been observed to depend on the phase of the monolayer, in particular, ∼25 molar % in the gel phase and ∼40 molar % in the fluid phase, wherein a regular distribution has been found with the highest separation between the cholesterol molecules. The extracted isothermal compressibility coefficient (CS) and ΔGexc from the monolayer isotherms indicate that the molecular arrangement at χc are the most stable configurations of the monolayer. Our study provides direct evidence into cholesterol-induced evolution in phase behavior and the consequent model on the structure at different phases in the phospholipid Langmuir monolayers.

■

INTRODUCTION The phase behavior and structural properties of the model biomembranes are of considerable scientific and practical interest as they provide a basic platform for understanding the membrane functions in living cells.1,2 The fluid mosaic model3 that treats the biomembranes as a homogeneous mixture of cellular components has been seen no longer to be crudely valid in the eukaryotic cell membranes. The modern concept of the cell membrane structure envisions the presence of an ordered structure of cholesterol-enriched micro-domain assemblies, known as lipid rafts, in a relatively disordered lipid matrix.4,5 These assemblies are believed to be involved in a number of cellular processes, including drug delivery, protein sorting, signal transduction, and so forth.1,2,6 Cholesterol, an amphipathic molecule very often treated as a rigid cylinder, is an abundant class of lipids regulating the physicochemical properties of the mammalian cell membranes.7 Even a small cholesterol fraction in the model lung surfactant can offer control over the surfactant spreadability by reducing the viscosity of the surfactant interface by orders of magnitude.8 In addition, the phospholipid acyl chains, in the physiologically relevant liquid crystalline state, show a higher degree of orientational and conformational order in the presence of cholesterol.9,10 It is believed to have a strong effect on the structure and functions of proteins residing in the membrane. © 2017 American Chemical Society

The chemical structure of a cholesterol molecule includes a tetracyclic fused ring in a trans-configuration along with a flexible hydrocarbon side chain that helps to closely attract the phospholipid acyl chains toward it with van der Waals forces acting between them. This fact leads to a laterally more condensed membrane with a higher packing density of phospholipids making it less permeable. Thus, cholesterol affects the conformational order and membrane permeability and regulates the lateral organization of the membrane components.11−19 There are some existing mechanical models, namely, the condensed complex model,20,21 the super lattice model,22,23 and the umbrella model,9,23 which are being used to explain the physical mechanism of the phospholipid−cholesterol interaction.24 It was reported earlier25 that the miscibility of cholesterol increases significantly in the saturated phosphatidylcholine (PC) lipids in comparison to that in the unsaturated one. Previous reports concerning the PC−cholesterol interaction both in the Langmuir monolayer26−28 and in the model bilayer22,24,29−32 have witnessed the cholesterol-induced structural changes in the phospholipid model membranes. Despite vivid research with model biomembranes, the phospholipid phase dependency of lateral and out-of-plane organization of cholesterols as a function Received: December 14, 2016 Revised: March 14, 2017 Published: April 6, 2017 4081

DOI: 10.1021/acs.jpcb.6b12587 J. Phys. Chem. B 2017, 121, 4081−4090

Article

The Journal of Physical Chemistry B

Figure 1. Schematic of the laboratory setup for XRR measurement from supported monolayer membranes.

(LE) to the liquid-condensed (LC) phase induced by cholesterol.37−40 XRR measurements have been performed from the phospholipid monolayers deposited on solid substrates to investigate the structural changes and precise location of the cholesterol molecules in the phospholipid membranes. Reflectivity analysis demonstrates that at a relatively low molar %, cholesterol produces a homogeneous mixture with the phospholipids in the membrane. The condensing effect of cholesterol provides more ordered arrangement in the phospholipid acyl chains subsequently leading to the lowering of the electron density in the head group region by inducing tilt in the lipid head group. Our observation also suggests the existence of a critical concentration of cholesterol (χc), at which lipid− cholesterol miscibility reaches a maximum depending on the monolayer phase. Our experimental study demonstrates that different theoretically predicted models for the lipid−cholesterol interaction work together at different strengths as the cholesterol molar concentration varies.

of concentration is still not achieved satisfactorily and needs significant effort to understand the interaction mechanism. The coarse-grained techniques used by previous researchers explored the overall effect on the thermotropic phase behavior and predicted the structure of the phospholipid model membranes. Owing to a high spatial resolution, the X-ray reflectivity (XRR) technique is more specific to the mesoscopic structural arrangement of biomolecules like cholesterol in a phospholipid membrane and can predict the model more accurately. The interaction strength varies significantly due to chain length variation because of the hydrophobic mismatch between the cholesterol and lipid acyl chains. Thus, the 1,2-dimyristoyl-snglycero-3-phosphocholine (DMPC) and 1,2-dipalmitoyl-snglycero-3-phosphocholine (DPPC) monolayer membranes having different chain lengths would serve as excellent model systems to investigate their structural analogues to the theoretically existing models. In the present work, we have studied the effect of cholesterol on the aminophospholipid membranes to understand its orientation and localization in modulating the physicochemical properties of the plasma membrane by primarily controlling the phase behavior of the membrane phospholipids,33 which plays a pivotal role in membrane organization, function, sorting, viability, and cell proliferation34,35 and regulates the activity of receptors and enzymes residing in the membrane. Cholesterol intercalation into the saturated aminophospholipid-containing bilayers gives rise to an intermediate phase called the “liquid ordered phase”,36 which is ordered in terms of the conformational flexibility of the lipid chains but is disordered from the point of view of the lateral position of the molecules.36 We have carried out a systematic measurement on the Langmuir monolayers and solid-supported phospholipid monolayers for a wide range of cholesterol concentrations. The pressure−area (Π−A) isotherm measurements from the Langmuir monolayers of DMPC at subphase temperature both above and below the nominal lipid chain melting temperature (Tm) and DPPC only below Tm show the cholesterol molar % and the subphase temperature-dependent phase behavior of the monolayer. In addition, the isothermal compressibility coefficient and excess Gibbs free energy extracted from the isotherms reveal the evolution of the monolayer phases from the liquid-expanded

■

MATERIALS AND METHODS DMPC, DPPC, cholesterol, and chloroform were purchased from Sigma-Aldrich and were used without further purification. The silicon (100) substrate with a polished surface on one side was purchased from Vin Corolla. Π−A Isotherm Measurement. The surface pressure−area (Π−A) isotherm measurements are performed using a computer-controlled KSV-NIMA medium-size trough equipped with two symmetrical barriers and a Wilhelmy-plate pressure sensor. Different molar concentrations of cholesterol and lipid solutions in chloroform were prepared and then spread on a water subphase. The concentration of the lipid−cholesterol mixture in chloroform was 0.2 mg/mL. There was a waiting time of 20 min for the complete evaporation of the solvent, after which we started compressing the monolayer at a speed of ∼1 Å2/ molecule/min, and the surface pressure was recorded as a function of the mean molecular area. The subphase temperature was controlled by the JULABO temperature controller. The silicon wafers were treated with a UV/O3 cleaner before the monolayer deposition to make them hydrophilic. The monolayer films were then transferred onto the wafers using the Langmuir− Schaefer method at a rate of 2 mm/min at a constant monolayer 4082

DOI: 10.1021/acs.jpcb.6b12587 J. Phys. Chem. B 2017, 121, 4081−4090

Article

The Journal of Physical Chemistry B surface pressure of 30 mN/m. In the case of the DMPC monolayer, the subphase temperatures were maintained at 15 and 30 °C to keep the monolayer in the gel and fluid phases, respectively. For the DPPC monolayer, the subphase temperature was maintained at 25 °C, which corresponds to the monolayer gel phase. To parameterize the physical mechanisms and molecular cooperativeness associated with the phase behavior of the Langmuir monolayer films, we calculated the compressibility coefficient (CS) and excess Gibbs free energy (ΔGexc) from the isotherms using the following formulas.28,41 CS = −

1 ⎛⎜ dA ⎞⎟ A ⎝ dΠ ⎠T

ΔGexc =

∫0

(1)

Π

[A12 − (χ1 A1 + χ2 A 2 )] dΠ

(2)

where A12 is the mean molecular area in the binary monolayer at the desired surface pressure of 30 mN/m. χ1, χ2, and A1, A2 are the mole fractions and mean molecular areas of the respective single-component monolayer at the same surface pressure. XRR Measurement and Analysis. XRR measurement from the supported monolayers were carried out with the Rigaku Smartlab diffractometer equipped with a 2 kW sealed Coolidge tube Cu Kα X-ray generator that corresponds to an X-ray wavelength of 1.54 Å. The schematic of the laboratory XRR setup has been sketched in Figure 1. In this instrument, the sample was placed horizontally and kept fixed on the goniometer base, whereas the source arm and detector arm rotate in a synchronized manner to collect the specular beam reflected from the sample. A high-resolution two-dimensional detector (HYPIX100) was used to collect the scattered beam from the sample. The high dynamic range and low background noise of the detector help us to collect the reflectivity from the sample much better than the conventional laboratory XRR setup, which is particularly suitable for thin-film characterization. The experimentally obtained XRR profiles from the supported monolayers were normalized with respect to the incident intensity. The profiles were fitted assuming a simple model of the monolayer structure. In this model, the monolayer is divided into a number of layers along the depth of the film where the layers differ by the average electron density. The reflectance from the interface j − 1, j following the recursive technique in Parratt’s formalism,42 can be written as rj − 1, j = exp( −2ikz , j − 1dj − 1)

Figure 2. XRR analysis scheme: Experimental XRR profile from the supported DPPC monolayer as well as the model fitting using Parratt’s formalism. The inset figure shows the EDP (symbols) and its differentiation (line) relative to the film depth with two different y axes, one for showing the electron density and the other for its differentiation. The cartoon in the inset shows the fine slices of the individual layers of constant electron density considered for model fitting of the XRR profiles.

thickness of the individual layers from the conventional plotting of the EDP profiles. For that reason, we have differentiated the EDP relative to the film depth

with the conventional EDP in the inset of Figure 2. Peaks in Δρ Δd curves are expected to appear at the interface between the layers where there is a change in the electron density. The difference between the peak positions correspond to the layer thickness, whereas the height and width of the peak are related to the roughness at the interface between the two layers. Such a protocol has been used to compare the fitting of different reflectivity profiles, which will be discussed in the XRR analysis section. Atomic Force Microscopy (AFM). The surface morphology of the deposited films was studied by AFM measurement using NanoScope IV MMAFM, Veeco in the tapping mode. The deposited monolayer samples were kept in a vacuum desiccator overnight at room temperature before the measurement. The reproducibility of the observed features has been checked by collecting images at three different positions on the sample.

rj , j + 1 + f j − 1, j 1 + rj , j + 1f j − 1, j

■

(3)

EXPERIMENTAL RESULTS Pressure−Area (Π−A) Isotherm Measurement. The Π− A isotherm measurement is an efficient technique to study the phase behavior of the Langmuir monolayers and hence the molecular-level thermodynamic interaction among the monolayer constituents.8,41,43−45 Isotherms of the DMPC monolayer with different molar concentrations of cholesterol at two different subphase temperatures, one below and one above Tm (∼23 °C), in particular, 15 and 30 °C, respectively, are shown in Figure 3a,b. The isotherm corresponding to the DPPC monolayer having different cholesterol molar % and at a subphase temperature of 25 °C, that is, in the monolayer gel phase, is shown in Figure 4. The isotherm corresponding to pure cholesterol has been observed to go through a sharp transition at a molecular area of ∼37 Å2, which is consistent with the previous studies.41,46 All other isotherms corresponding to the binary films

where f j − 1, j =

kz , j − 1 − kz , j kz , j − 1 + kz , j

( ΔΔρd ) and plotted the same along

(4)

In a stratified homogeneous media consisting of n number of layers, j = 0 represents the surrounding media and j = (n + 1) represents the substrate. The effective reflectivity from the sample as a whole can be calculated by summing up the reflectance from the interface between two adjacent layers starting from the substrate. However, the experimental reflectivity from an arbitrary sample takes the interfacial roughness into account.42 One such fitting of the reflectivity profile along with the roughness convoluted EDPs for the DPPC monolayer on a silicon substrate is shown in Figure 2. Owing to the roughness convolution, it is very difficult to visualize the 4083

DOI: 10.1021/acs.jpcb.6b12587 J. Phys. Chem. B 2017, 121, 4081−4090

Article

The Journal of Physical Chemistry B

have been observed to pass through different phases characterized by the change in the slope at different mean molecular areas depending on the conformation and orientation of the hydrophobic acyl chains of lipid.37,38,47,48 At the initial stage of compression, the monolayer remains in the gaseous phase, and further compression causes the film to pass through the LE phase, a transition region, known as the LE−LC phase coexistence region, followed by the LC phase.37,38 The coexistence of the LE and LC phases are observed in the monolayers of both the lipids, DMPC and DPPC used here, at subphase temperatures below their respective Tm. This is evident from the compressibility coefficient49,50 calculated using eq 1 and displayed in the inset of Figures 3a and 4. In addition, the variation of the mean molecular area with cholesterol concentration at different monolayer phases is displayed in Figure 5. Excess Gibbs free energy for all of the monolayers has also been calculated and plotted in the inset of Figure 5.

Figure 3. Pressure−area isotherms from the DMPC Langmuir monolayer with different molar % of cholesterol: (a) at T = 15 °C and (b) at T = 30 °C. Variation of isothermal compressibility with surface pressure at different cholesterol molar % is shown in the inset of each figure. χn represents n molar % of cholesterol associated to the monolayer. We have used the same color scheme for the monolayer isotherm as well as the corresponding isothermal compressibility at a particular cholesterol concentration.

Figure 5. Variation of the mean molecular area with molar % of cholesterol at a constant surface pressure of 30 mN/m. Excess Gibbs free energy calculated from the isotherms is shown in the inset for both the lipids.

In the binary film, the width of the LE−LC coexistence region has been observed to depend on the cholesterol molar fraction (χchol) as well as the hydrophobic chain length of the lipids. The peaks in the compressibility curves indicate highest intermolecular cooperativeness.39,40 From the compressibility curves in the inset of Figures 3 and 4, the LE−LC phase coexistence is clearly evident at a surface pressure of ∼30 and ∼12 mN/m in the gel phase of the DMPC and DPPC monolayers, respectively. In contrast, such a coexistence region is clearly missing in the fluid phase of the DMPC monolayer. In the compressibility curves, the height, width, and position of the peaks exhibit the thermodynamics and, hence, the molecularlevel interaction in the monolayer. On comparing Figure 3a,b it is evident that peaks are found only in case of the monolayer in the gel phase, and no such clear peak is found in the compressibility curves corresponding to the monolayer in the fluid phase at any cholesterol concentration. In the case of isotherms from DMPC in the gel phase (shown in Figure 3a), the peaks appear for 0 and 5% and do not show any such peak at 25 molar % of cholesterol in the lipid monolayer. Interestingly, the peak reappears at a cholesterol concentration of 40 molar %. The peak heights decrease with the increasing cholesterol molar %, although there

Figure 4. Pressure−area isotherms from the Langmuir monolayer of DPPC with different molar % of cholesterol at 25 °C, that is, in the monolayer gel phase. Inset shows the variation of isothermal compressibility with surface pressure at different cholesterol molar %. χn represents n molar % of cholesterol. The same color is used for a particular cholesterol concentration in both the isotherm and the corresponding compressibility.

4084

DOI: 10.1021/acs.jpcb.6b12587 J. Phys. Chem. B 2017, 121, 4081−4090

Article

The Journal of Physical Chemistry B

The experimental XRR data from the supported DMPC and DPPC monolayers in the gel phase with varying molar % of cholesterol along with the fit are shown in Figure 6a. XRR data

is almost a negligible shift in the peak position for 0 and 5 molar % of cholesterol but a little shift for 40 molar % of cholesterol in the lipid matrix. The appearance and change in the LE−LC coexistence region is also similar in the case of the DPPC monolayer in the gel phase (Figure 4), but here no peak appeared in the compressibility curve of the monolayer with 40 molar % of cholesterol. The height, position, and width of the peaks are different compared to those of the DMPC monolayer in the same phase, although the collapse pressure of both the monolayers are quiet similar. We could not distinguish the LE−LC coexistence phase at any cholesterol concentration higher than 40 molar % irrespective of the subphase temperature for the DPPC monolayer. Thus, the molecular cooperativeness not only depends on the phospholipid hydrophobic chain length but also on the monolayer phases. Hydrophobic mismatch between the guest and host molecules is believed to play a dominant role in determining the monolayer phase behavior. Cholesterol significantly reduces the orientational degrees of freedom of the lipid acyl chains, thus making the monolayer less compressible. In the case of a pure cholesterol monolayer, the surface pressure started increasing when the film as a whole transformed into the LE phase.51 Further compression made the cholesterol molecules in the monolayer untilted, which led to the incompressible phase known as the LC phase.51 In this phase, the monolayer surface pressure changed very rapidly until it reached the collapse pressure. Figure 5 shows the variation of the mean molecular area of the monolayer constituents with the cholesterol mole fraction at the surface pressure of 30 mN/m. Excess Gibbs free energy (ΔGexc) corresponding to each monolayer in both the gel and the fluid phases is calculated following eq 2, and their variation with molar % of cholesterol has been shown in the inset of Figure 5. Excess Gibbs free energy is the difference between the area per molecule present in the binary film and the fractional sum of the mean molecular area in their respective single component monolayers. Thus, a negative value of the Gibbs free energy indicates lowering in the mean area of a molecule while present in the binary monolayer compared to that of the same molecule present in the pure monolayer. The negative Gibbs free energy stands for the condensing effect of the guest molecules. From the polynomial fitting of the calculated Gibbs free energy, two dips are found in each curves, one at just below 25 molar % and the other at ∼75 molar % of cholesterol, corresponding to both the phospholipids in the gel phase. In addition, there is only one dip found at ∼40 molar % of cholesterol in the case of DMPC in the fluid phase. These dips are indicative of some special molecular arrangement in the binary monolayer driven by cholesterol at those particular mole fractions, which is discussed in the succeeding section. XRR and AFM Results from the Supported Monolayer. XRR is a nondestructive surface-sensitive technique used for structural characterization of thin films and multilayers.52−54 A highly intense X-ray beam is allowed to be incident on the sample, and the specularly scattered beam is collected as a function of the wave vector transfer, qz = 4π sin θ/λ, which contains the laterally averaged electron density profile (EDP) along the surface normal. Specular reflectivity provides us information about the thickness, electron density, and roughness of the films at a mesoscopic length scale.52,54 In this report, we present XRR measurements from the solid supported monolayer films to gain insight into the understanding of the cholesterol− phospholipid interaction from a structural perspective.

Figure 6. Experimentally obtained XRR profiles (symbols) with model fittings (lines) from the monolayers of DMPC and DPPC associated with varying molar fraction of cholesterol deposited on the silicon substrate at a surface pressure of 30 mN/m. (a) XRR from the DMPC and DPPC monolayers deposited in their gel phase at the subphase temperature of 15 and 25 °C, respectively. χn is used to represent molar % of cholesterol. (b) XRR from the DMPC monolayers deposited in the fluid phase at the subphase temperature of 30 °C. The portion marked in black circle where the first dips have appeared in the XRR profiles has been enlarged in the inset for clear visualization of monolayer thickness evolution in its fluid phase due to increasing cholesterol molar % in the monolayer.

for the DMPC monolayer in the fluid phase have been shown in Figure 6b. Each profile has been given a vertical shift by a factor of 100 in Figure 6a and 10 in Figure 6b for clear visualization. As it is known that the position of the first dip in the XRR curves is directly related to the total thickness oscillation of the monolayer, we have enlarged the portion in which the first dip appeared in Figure 6b, as displayed in the inset. It shows that the dip moves toward lower qz with increasing molar % of cholesterol up to 40%, after which it again moves toward higher qz. Shifting of the first dip toward lower qz actually indicates a thickening of the ⎛ 1 ⎞ monolayer as dtotal is inversely proportional to qz ⎜d total ∝ (q ) ⎟. ⎝ z dip ⎠ The analysis scheme for the XRR profiles has already been discussed in the XRR Measurement and Analysis section. The 4085

DOI: 10.1021/acs.jpcb.6b12587 J. Phys. Chem. B 2017, 121, 4081−4090

Article

The Journal of Physical Chemistry B differentiated EDPs corresponding to the DMPC and DPPC monolayers in the gel phase and the DMPC monolayer in the fluid phase are shown in Figure 7a,b, respectively. For clarity,

Figure 8. Variation in the hydrophobic tail thickness (dtail) and total thickness (dtotal) of the supported monolayers of DMPC and DPPC with varying mole fractions of cholesterol: DMPC monolayers were deposited at the subphase temperature of 30 °C, that is, in the fluid phase of the lipid, and the DPPC monolayers were deposited at the subphase temperature of 25 °C, that is, in the gel phase of the lipid. Surface pressure was maintained at 30 mN/m during deposition of all of the monolayers.

electron density. In contrast, the layer apart from the head group has a thickness of 9 Å and a relatively lower electron density of 0.214 e Å−3 due to the hydrophobic mismatch and the subsequent voids created by the cholesterol. Thus, the thickening of the hydrophobic tail portion by 2.5 Å, which corresponds to ∼19% of thickness change from the pure monolayer due to cholesterol incorporation, is evident. Moreover, the lowering in the head group thickness by 0.5 Å, that is, ∼6% change, has also been observed. Tails of the pure DMPC monolayer in the monolayer gel phase possess ∼36° of tilting, whereas this decreases to ∼14° at 25 molar %. EDPs corresponding to the DPPC monolayers deposited in the gel phase are depicted in Figure 7a. From the EDPs, it has been observed that the pure monolayer has a tail thickness of 16.2 Å. The binary mixture of the DPPC monolayer shows a hydrophobic tail thickness of 17.4 Å for 5 molar % and 17.6 Å for 25 molar % of cholesterol. Thus, cholesterol association makes the tail region thicker by ∼9% compared to that of the pure DPPC monolayer. It further evidences the monotonic decrease of the phospholipid head group size with increasing cholesterol concentration up to 25 molar %. The pure PC monolayer has a head group thickness of 9.8 Å, and the numbers are 9 and 8.7 Å for 5 and 25 molar % of cholesterol, respectively. These consistent variations in thickness of the different parts of the film convincingly illustrate the incorporation of cholesterol and its effect on the structure of the phospholipid monolayers. In addition, the XRR analysis suggests an exactly opposite effect in the lipid head group and tail thicknesses beyond 25 molar % of cholesterol compared to the observation below that concentration. Calculation also shows that the lipid chain tilting attains a minimum value at 25 molar % of cholesterol. The EDPs corresponding to the DMPC monolayers in the fluid phase in Figure 7b also show similar changes in the hydrophobic thickness as well as the total thickness (dtotal) due to cholesterol incorporation. The thickness variations can be seen from Figure 8. A close look into this figure shows that the slope of the tail thickness variation is slightly larger than that of the total

Figure 7. Extracted EDPs corresponding to (a) DMPC and DPPC in the gel phase and (b) DMPC in the fluid phase with a variety of cholesterol concentrations (χn) from the fitting of the experimentally obtained XRR profiles using Parratt’s formalism with a proper vertical shift for clear visualization.

each profile has been given a vertical shift of 0.1 e Å−3. The region indicated between the two dotted arrows for a color correspond to the head portion of the molecules. Thus, the cholesterolinduced thickness changes in the tail and head regions can be extracted from the positions of the arrows. Figure 8 displays the variation in the thickness of the hydrophobic tail and the total thickness of the monolayers with χchol. The lipid acyl chain tilting has also been calculated from the formula: cos θ = dtail/dmax tail with 55 dmax tail ≃ 16 Å for DMPC. A quantitative comparison of the pure DMPC monolayer and the monolayer with 25 molar % of cholesterol in the gel phase has been made. The pure DMPC monolayer has head group and tail thicknesses of ∼8.5 and 13 Å, respectively, with their respective electron densities of 0.443 and 0.27 e Å−3. In the film containing cholesterol, the head group region is ∼8 Å thick with an electron density of 0.426 e Å−3. Note that there exists two adjacent layers in the acyl chain region with significant electron density contrast in the monolayer containing cholesterol. One is adjacent to the head group region having a thickness of 6.5 Å and an electron density of 0.346 e Å−3. This layer contains the bulkier hydrophobic part of cholesterol that leads to a higher value of 4086

DOI: 10.1021/acs.jpcb.6b12587 J. Phys. Chem. B 2017, 121, 4081−4090

Article

The Journal of Physical Chemistry B

the head group region up to χc can be explained by the lipid head group tilting. Such a situation can be visualized from the schematic of molecular arrangement in Figure 9, wherein we

thickness variation. This fact corroborates with the phospholipid head group thickness variation (dhead), wherein dhead has been seen to decrease with increasing cholesterol molar %. XRR data from the fluid phase of the DMPC monolayers further demonstrate the lipid chain tilting to be ∼44, 40, 32, 23, 41, and 51°, respectively, for 0, 15, 25, 40, 60, and 75 molar % of cholesterol-incorporated monolayer. Thus, lipid chain tilting reduced upon increasing the cholesterol molar % up to 40 molar %, beyond which tilting again increased. The existence of χc at which lipid−cholesterol miscibility is maximum is further verified from the AFM images from the supported monolayers. The AFM images obtained from the pure DMPC monolayer and the monolayer with 25 molar % of cholesterol deposited in the gel phase are shown in Figure 10a,b. The pure DMPC monolayer formed a uniform layer, whereas the monolayer with cholesterol had height variations in different parts of the film. The AFM images collected from the DMPC monolayers deposited in the fluid phase at a 30 mN/m surface pressure are shown in Figure 11. It can be inferred from the obtained images that a uniformity in the binary monolayer is maintained up to χc, that is, 40 molar % in the fluid phase. Some inhomogeneities are observed even for the monolayer associated with 40% of cholesterol (the whitish domain area in Figure 11d). The height and the number of such domains increased in the monolayer with increasing concentrations of cholesterol beyond χc.

Figure 9. Sketch of the molecular arrangement predicted from the XRR result from the pristine DMPC monolayer (right) and the monolayer associated with cholesterol (left) deposited at the surface pressure of 30 mN/m and at the subphase temperature of 15 °C, that is, in the lipid gel phase. It indicates the cholesterol-induced structural changes in the monolayer. Cholesterol molecules are shown by the green ovals with brown tails. The lipid head groups are shielding the nonpolar part of cholesterol from direct exposure to water by tilting and stretching themselves. The presence of cholesterol has increased the thickness of the lipid acyl chain region by ordering them.

■

have compared the pure and cholesterol-added DMPC monolayer structure observed from the XRR data. This phenomenon can be described by the umbrella model initially proposed to explain the maximum solubility of the cholesterol molecules in the lipid matrix.9,23 The thickening in the lipid tail portion and, hence, the reduction in the lipid chain tilting up to χc can be described by the cholesterol-induced ordering in the acyl chain region. This is known as the condensing effect of cholesterol. Observations are similar in the case of the DMPC monolayer in the fluid phase. The only difference found is the value of χc. The thickening in the lipid tail portion up to 40 molar % of cholesterol and the thinning beyond that concentration suggests that χc in this case is 0.4. The lowest value of ΔGexc found at this concentration and the inhomogeneity found beyond χc in the AFM images in Figure 11 are also in agreement with this observation. In the case of the DPPC monolayer in the gel phase, χc is found to be 0.25. Beyond this concentration of cholesterol, the lipid acyl chain tilting became very high, which might be due to the downward movement of cholesterol toward the head group, as was observed by Ivankin et al.26 In a recent article by Ali et al.24 it has been reported that the presence of dips in the Gibbs free energy curves are actually suggestive of the existence of superlattice structures.22,56 Thus, the Gibbs free energy as shown in the inset of Figure 5 indicates the existence of a regular distribution of cholesterol at those particular concentrations in the binary mixture. This is known as the superlattice model. This model is completely a geometrical one that maximizes the separation between two adjacent cholesterols from each other.57,58 The difference in the cross-sectional area between the cholesterol and the lipid along with their hydrophobic mismatch causes a long-range repulsive force among the cholesterol moieties that actually drives the molecules to form a superlattice distribution. In agreement with the experimental findings reported by Ivankin et al.,26 we have also observed that at higher cholesterol concentrations (higher than χc), the cholesterol molecules form a phase-segregated domain structure and favor to reside near the air−water interface.

DISCUSSION All of the phospholipid monolayers corresponding to different cholesterol molar % clearly exhibit a phase change from gas to LE, followed by the LE−LC phase coexistence, and then the shift into the LC phase upon increasing the surface pressure. Cholesterol essentially possesses quite a different phase behavior compared to that of the pure DMPC and DPPC monolayers, as can be seen from Figures 3 and 4. Thus, cholesterol alters the miscibility and shows condensing properties in the binary monolayers.9 At the gaseous phase and at the gas−LC phase coexistence region, the cholesterol monolayer surface pressure remained very low. Further compression caused the entire monolayer to enter into the untilted LC phase, indicated by a rapid increase in the monolayer surface pressure.51 Figure 3a illustrates the effect of cholesterol in the phase behavior of the DMPC monolayer in the gel phase. The evolution in the phase behavior has further been quantified as a function of cholesterol molar % by the evaluation of the isothermal compressibility coefficient (CS), as shown in the inset of Figure 3a, and the excess Gibbs free energy (ΔGexc), as shown in the inset of Figure 5. The observation of peaks in compressibility curves indicates that in the region of the surface pressure between 25 and 35 mN/m the intermolecular cooperativeness is the highest. The appearance of peaks also depend on the cholesterol concentration. Moreover, the negative value of ΔGexc in the entire range of cholesterol concentrations convincingly represent the condensing effect of cholesterol. Two dips observed in ΔGexc curve represent the monolayer stoichiometry with the strongest attractive interaction. We denote the cholesterol molar % corresponding to such a stoichiometry to be the critical concentration (χc). Thus, 25 molar % is the value of χc for DMPC in the gel phase. Moreover, the addition of cholesterol up to χc provides extra stability to the monolayers. The XRR study has provided direct structural insight into the monolayers at different molar % of cholesterol. The thinning and subsequent lowering in the electron density in 4087

DOI: 10.1021/acs.jpcb.6b12587 J. Phys. Chem. B 2017, 121, 4081−4090

Article

The Journal of Physical Chemistry B The XRR study convincingly illustrates the predominant effect of cholesterol in the lipid acyl chain region than that in the head group portion. It can be seen from the percentage change in the thickness of different portions of the monolayer. The structure of the cholesterol molecule is suggested to be responsible for such an effect. The lack of sufficient hydrophilicity in the polar region causes the cholesterol molecule to interact favorably only with the lipid acyl chains. Moreover, the rigid ring structure attached to cholesterol attracts the lipid hydrophobic chains toward it. Thus, increasing the cholesterol concentration from zero to a certain value primarily tries to build a short-range ordering in the acyl chains of phospholipids. This phenomenon is described by the condensing effect of cholesterol where a lipid−cholesterol complex is formed with a stoichiometry having lower Gibbs free energy. At the ordering maxima, cholesterol produces a regular distribution in the monolayer, thus giving rise to the superlattice structure. Further increase in cholesterol makes the monolayer highly unstable. In the monolayer membrane, cholesterol molecules are covered by the polar head groups of the neighboring phospholipids to prevent the loss of free energy by the exposure of the hydrophobic part of cholesterol to water. This is because the lipid head groups are already in a state of maximal stretching. Excess cholesterols form a cluster,59 which leads to a phase-segregated structure as shown by the AFM images in Figures 10b and 11d−f. Note that we cannot rule out

Figure 11. AFM images obtained from the supported DMPC monolayer associated with different molar % of cholesterol: (a) pristine, (b) 15%, (c) 25%, (d) 40%, (e) 60%, and (f) 75%. Monolayers were deposited at the subphase temperature of 30 °C on the Si substrate at a surface pressure of 30 mN/m.

Figure 10. AFM images obtained from (a) the pristine DMPC monolayer and (b) the DMPC monolayer associated with 25 molar % of cholesterol deposited at the subphase temperature of 15 °C on the Si substrate at a surface pressure of 30 mN/m.

unfavorable. The umbrella model then comes into play by introducing a tilt in the lipid head group. This model tries to maximize the cholesterol solubility in the phospholipid monolayer and prevent the cholesterol from forming a cluster. At the ordering maxima, that is, at χc, we have found a superlattice structure in the binary film. At a sufficiently high concentration, cholesterol has been found to form a cluster or rather rafts that cannot be explained by this model. The umbrella model as well as the condensed-complex model are, therefore, believed to be predominant in the region of lower cholesterol concentrations. We have modeled the evolution of the molecular arrangement in the supported monolayer with varying concentrations of cholesterol, specifically at χ = χ0, χ < χc, χ = χc, and χ > χc, the schematic of which has been illustrated in Figure 12.

the possibility of cholesterol monohydrate formation by those excess cholesterols.9,10 The high value of lipid chain tilting even in the presence of cholesterol at higher concentrations may not necessarily suggest the actual tilting in the lipid acyl chain. This may appear due to the phase segregation in the monolayer. X-ray provides lateral averaged information from the sample as it is not sensitive to the local fluctuations. Thus, the average thickness (davg) obtained from the XRR data can be assumed as davg = [dLRALR + dCRACR]/[ALR + ACR], where dLR, dCR, and ALR, ACR are, respectively, the thickness and area covered by lipid-rich and cholesterol-rich domains. The difference in the total thickness of the lipid-rich and cholesterol-rich domains leads to the formation of a complex structure and, hence, lateral inhomogeneity. This fact can be primarily thought to be the origin of the increasing tilt angle of the lipid acyl chains beyond χc. Our observations clearly suggest that below χc, the condensing effect of cholesterol and the umbrella model work together. The condensing effect helps the phospholipid molecules to form a tightly packed structure with increasing cholesterol molar %. But this increment in concentration may lead the cholesterol molecule to come close to water, which is energetically highly

■

CONCLUSIONS The experimental observation depicts the direct structural insight into the interaction of cholesterol with zwitterionic saturated phospholipids. The Π−A isotherm, XRR, and AFM measurements have evidenced the presence of mechanical models in the binary monolayer depending on the cholesterol concentration and the phase of the monolayer. At low concentrations, the umbrella model along with the cholesterol 4088

DOI: 10.1021/acs.jpcb.6b12587 J. Phys. Chem. B 2017, 121, 4081−4090

Article

The Journal of Physical Chemistry B

(9) Huang, J.; Feigenson, G. W. A microscopic interaction model of maximum solubility of cholesterol in lipid bilayers. Biophys. J. 1999, 76, 2142−2157. (10) Huang, J.; Bublotz, J. T.; Feigenson, G. W. Maximum solubility of cholesterol in phosphatidylcholine and phosphatidylethanolamine bilayers. Biochim. Biophys. Acta, Biomembr. 1999, 1417, 89−100. (11) McMullen, T. P. W.; Lewis, R. N. A. H.; McElhaney, R. N. Cholesterol-phospholipid interactions, the liquid-ordered phase and lipid rafts in model and biological membranes. Curr. Opin. Colloid Interface Sci. 2004, 8, 459−468. (12) McMullen, T. P. W.; McElhaney, R. N. Physical Studies of Cholesterol-phospholipid Interactions. Curr. Opin. Colloid Interface Sci. 1996, 1, 83−90. (13) Parker, A.; Miles, K.; Cheng, K. H.; Huang, J. Lateral Distribution of Cholesterol in Dioleoylphosphatidylcholine Lipid Bilayers: Cholesterol-Phospholipid Interactions at High Cholesterol Limit. Biophys. J. 2004, 86, 1532−1544. (14) Vist, M. R.; Davis, J. H. Phase Equilibria of Cholesterol/ Dipalmitoylphosphatidylcholine Mixtures: 2H Nuclear Magnetic Resonance and Differential Scanning Calorimetry. Biochemistry 1990, 29, 451−464. (15) Knoll, W.; Schmidt, G.; Ibel, K.; Sackmann, E. Small-Angle Neutron Scattering Study of Lateral Phase Separation in Dimyristoylphosphatidylcholine-Cholesterol Mixed Membranes. Biochemistry 1985, 24, 5240−5246. (16) Albrecht, O.; Gruler, H.; Sackmann, E. Pressure-Composition Phase Diagrams of Cholesterol/Lecithin, Cholesterol/Phosphatidic Acid, and Lecithin/Phosphatidic Acid Mixed Monolayers: A Langmuir Film Balance Study. J. Colloid Interface Sci. 1981, 79, 319−338. (17) Lentz, B. R.; Barrow, D. A.; Hoechli, M. CholesterolPhosphatidylcholine Interactions in Multilamellar Vesicles. Biochemistry 1980, 19, 1943−1954. (18) McMullen, T. P.; Vilcheze, C.; McElhaney, R. N.; Bittman, R. Differential Scanning Calorimetric Study of the Effect of Sterol Side Chain Length and Structure on Dipalmitoylphosphatidylcholine Thermotropic Phase Behavior. Biophys. J. 1995, 69, 169−176. (19) Thewalt, J. L.; Bloom, M. Phosphatidylcholine: cholesterol phase diagrams. Biophys. J. 1992, 63, 1176−1181. (20) Radhakrishnan, A.; McConnell, H. M. Condensed Complexes of Cholesterol and Phospholipids. Biophys. J. 1999, 77, 1507−1517. (21) Radhakrishnan, A.; McConnell, H. M. Condensed Complexes of Cholesterol and Phospholipids. Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 12662−12666. (22) Chong, P. L. Evidence for regular distribution of sterols in liquid crystalline phosphatidylcholine bilayers. Proc. Natl. Acad. Sci. U.S.A. 1994, 91, 10069−10073. (23) Huang, J. Exploration of Molecular Interactions in Cholesterol Superlattices: Effect of Multibody Interactions. Biophys. J. 2002, 83, 1014−1025. (24) Ali, M. R.; Cheng, K. H.; Huang, J. Assess the nature of cholesterol-lipid interactions through the chemical potential of cholesterol in phosphatidylcholine bilayers. Proc. Natl. Acad. Sci. U.S.A. 2007, 104, 5372−5377. (25) Brzustowicz, M. R.; Cherezov, V.; Caffrey, M.; Stillwell, W.; Wassall, S. R. Molecular organization of cholesterol in polyunstaturated membranes: Microdomain formation. Biophys. J. 2002, 82, 285−298. (26) Ivankin, A.; Kuzmenko, I.; Gidalevitz, D. CholesterolPhospholipid Interactions: New Insights from Surface X-Ray Scattering Data. Phys. Rev. Lett. 2010, 104, No. 108101. (27) Demel, R. A.; van Deenen, L. L. M.; Pethica, B. A. Monolayer interactions of phospholipids and cholesterol. Biochim. Biophys. Acta, Biomembr. 1967, 135, 11−19. (28) Jurak, M. Thermodynamic Aspects of Cholesterol Effect on Properties of Phospholipid Monolayers: Langmuir and Langmuir Blodgett Monolayer Study. J. Phys. Chem. B 2013, 117, 3496−3502. (29) Kessel, A.; Ben-Tal, N.; May, S. Interactions of Cholesterol with Lipid Bilayers: The Preferred Configuration and Fluctuations. Biophys. J. 2001, 81, 643−658.

Figure 12. Sketch of the molecular arrangements in the DMPC monolayer associated with a variety of cholesterol molar %: (a) pristine DMPC monolayer, (b) at cholesterol concentrations below χc, (c) at cholesterol concentrations around χ c , and (d) at cholesterol concentrations beyond χc.

condensing effect give satisfactory results. At critical concentrations, the cholesterol molecules arrange in a way to maximize the solubility in the lipid matrix, leading to the most ordered configuration, known as the superlattice structure. Beyond the critical concentrations, cholesterol produces inhomogeneous complexes with the lipid molecules. Thus, in summary, our study suggests that instead of one particular model, all of the explanatory models work together with different strengths at different cholesterol molar % for the saturated zwitterionic phospholipid model membranes, irrespective of the monolayer phases.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Mrinmay K. Mukhopadhyay: 0000-0001-8050-8661 Notes

The authors declare no competing financial interest.

■

REFERENCES

(1) Simons, K.; Ikonen, E. Functional Rafts in Cell Membranes. Nature 1997, 387, 569−572. (2) Lingwood, D.; Simons, K. Lipid rafts as a membrane-organizing principle. Science 2010, 327, 46−50. (3) Singer, S. J.; Nicolson, G. L. The Fluid Mosaic Model of the Structure of Cell Membranes. Science 1972, 175, 720−731. (4) Epand, R. M.; Maekawa, S.; Yip, C. M.; Epand, R. F. ProteinInduced Formation of Cholesterol-Rich Domains. Biochemistry 2001, 40, 10514−10521. (5) Binder, W. H.; Barragan, V.; Menger, F. M. Domains and Rafts in Lipid Membranes. Angew. Chem., Int. Ed. 2003, 42, 5802−5827. (6) Jacobson, K.; Mouritsen, O. G.; Anderson, R. G. W. Lipid rafts: at a crossroad between cell biology and physics. Nat. Cell Biol. 2007, 9, 7−14. (7) Ohvo-Rekilä, H.; Ramstedt, B.; Leppimaki, P.; Slotte, J. P. Cholesterol interactions with phospholipids in membranes. Prog. Lipid Res. 2002, 41, 66−97. (8) Kim, K.; Choi, S. Q.; Zell, Z. A.; Squires, T. M.; Zasadzinski, J. A. Effect of cholesterol nanodomains on monolayer morphology and dynamics. Proc. Natl. Acad. Sci. U.S.A. 2013, 110, E3054−E3060. 4089

DOI: 10.1021/acs.jpcb.6b12587 J. Phys. Chem. B 2017, 121, 4081−4090

Article

The Journal of Physical Chemistry B

(55) Bayerl, T. M.; Thomas, R. K.; Penfold, J.; Rennie, A.; Sackmann, E. Specular reflection of neutrons at phospholipid monolayers: changes of monolayer structure and headgroup hydration at the transition from the expanded to the condensed phase state. Biophys. J. 1990, 57, 1095− 1098. (56) Somerharju, P.; Virtanen, J.; Cheng, K. H.; Hermansson, M. The superlattice model of lateral organization of membranes and its implications on membrane lipid homeostasis. Biochim. Biophys. Acta 2009, 1788, 12−23. (57) Virtanen, J.; Ruonala, M.; Vauhkonen, M.; Somerharju, P. Lateral Organization of Liquid-Crystalline Cholesterol-Dimyristoylphosphatidylcholine Bilayers. Evidence for Domains with Hexagonal and Centered Rectangular Cholesterol Superlattices. Biochemistry 1995, 34, 11568−11581. (58) Tang, D.; Van Der Meer, B. W.; Chen, S. Y. S. Evidence for a Regular Distribution of Cholesterol in Phospholipid Bilayers from Diphenylhexatriene Fluorescence. Biophys. J. 1995, 68, 1944−1951. (59) Bach, D.; Wachtel, E. Phospholipid/cholesterol model membranes: formation of cholesterol crystallites. Biochim. Biophys. Acta 2003, 1610, 187−197.

(30) Pan, J.; Mills, T. T.; Tristram-Nagle, S.; Nagle, J. F. Cholesterol Perturbs Lipid Bilayers Nonuniversally. Phys. Rev. Lett. 2008, 100, No. 198103. (31) Sengupta, P.; Singh, R. R. P.; Cox, D. L.; Slepoy, A. Lateral organization of cholesterol molecules in lipid-cholesterol assemblies. Phys. Rev. E: Stat., Nonlinear, Soft Matter Phys. 2004, 70, No. 021902. (32) Tabaei, S. R.; Jackman, J. A.; Kim, S.-O.; Liedberg, B.; Knoll, W.; Parikh, A. N.; Cho, N. Formation of Cholesterol-Rich Supported Membranes Using Solvent Assisted Lipid Self-Assembly. Langmuir 2014, 30, 13345−13352. (33) Finegold, H.; von, L. Cholesterol in Membrane Models; CRC Press: Boca Raton, 1988; Vol. 37, pp 519−624. (34) Simons, K.; Ikonen, L. How Cells Handle Cholesterol. Science 2000, 290, 1721−1726. (35) Mouritsen, O. G.; Zuckermann, M. J. What’s So Special About Cholesterol. Lipids 2004, 39, 1101−1113. (36) Mouritsen, O. G. The Liquid-ordered State Comes of Age. Biochim. Biophys. Acta 2010, 1798, 1286−1288. (37) Kaganer, V. M.; Mohwald, H.; Dutta, P. Structure and Phase Transitions in Langmuir Monolayers. Rev. Mod. Phys. 1999, 71, 779− 819. (38) Gaines, G. L. Insoluble Monolayers at Liquid-Gas Interfaces; WileyInterscience: New York, 1966. (39) Yu, Z.-W.; Jin, J.; Cao, Y. Characterization of the Liquid-Expanded to Liquid-Condensed Phase Transition of Monolayers by Means of Compressibility. Langmuir 2002, 18, 4530−4531. (40) Ihalainen, P.; Peltonen, J. Miscibility of Lipids in Monolayers Investigated through Adsorption Studies of Antibodies. Langmuir 2003, 19, 2226−2230. (41) Sabatini, K.; Mattila, J.-P.; Kinnunen, P. K. J. Interfacial Behavior of Cholesterol, Ergosterol, and Lanosterol in Mixtures with DPPC and DMPC. Biophys. J. 2008, 95, 2340−2355. (42) Parratt, L. G. Surface Studies of Solids by Total Reflection of XRays. Phys. Rev. 1954, 95, 359−369. (43) Giri, R. P.; Mukhopadhyay, M. K.; Mitra, M.; Chakrabarti, A.; Sanyal, M. K.; Ghosh, S. K.; Ma, Y.; Sinha, S. K.; Bera, S.; Lurio, L. B. Structural configuration of adsorbed membrane skeletal protein in phospholipid membranes. Communicated. (44) Hwang, M.-J.; Kim, K. Poly(ethylenimine) as a subphase stabilizer of stearic acid monolayers at the air/water interface: surface pressurearea isotherm and infrared spectroscopy study. Langmuir 1999, 15, 3563−3569. (45) Datta, A.; Sanyal, M. K.; Dhanabalan, A.; Major, S. S. Formation of Highly Condensed Ferric Stearate Monolayers at the Air-Water Interface. J. Phys. Chem. B 1997, 101, 9280−9286. (46) Seoane, R.; Minones, J.; Conde, O.; Minones, J.; Casas, M.; Iribarnegaray, E. Thermodynamic and Brewster Angle Microscopy studies of fatty acid/cholesterol mixtures at the air/water interface. J. Phys. Chem. B 2000, 104, 7735−7744. (47) Möhwald, H. Phospholipid and phospholipd-protein monolayer at the air/water interface. Annu. Rev. Phys. Chem. 1990, 41, 441−476. (48) McConnell, H. M. Structures and transitions in lipid monolayers at air-water interface. Annu. Rev. Phys. Chem. 1991, 42, 171−195. (49) Birdi, K. S. Lipid and Biopolymer Monolayers at Liquid Interfaces; Plenum Press: New York, 1989; Chapter 2. (50) Birdi, K. S. Self-Assembly Monolayer Structures of Lipids and Macromolecules at Interfaces; Kluwer Academic: New York, 1999; Chapter 2. (51) Cadena-Navaa, R. D.; Martin-Mironesa, J. M.; VazquezMartineza, E. A.; Rocab, J. A.; Ruiz-Garcia, J. Direct observations of phase changes in Langmuir films of Cholesterol. Rev. Mex. Fis. S 2006, 52, 32−40. (52) Dietrich, S.; Haase, A. Scattering of X-rays and neutrons at interfaces. Phys. Rep. 1995, 260, 1−138. (53) Sanyal, M. K.; Hazra, S.; Basu, J. K.; Datta, A. Extraction of density profile for near perfect multilayers. Phys. Rev. B 1998, 58, No. R4258. (54) Basu, J. K.; Sanyal, M. K. Ordering and growth of LangmuirBlodgett films: X-ray scattering studies. Phys. Rep. 2002, 363, 1−84. 4090

DOI: 10.1021/acs.jpcb.6b12587 J. Phys. Chem. B 2017, 121, 4081−4090