Hydrostatic Pressure Promotes Domain Formation in Model LipidRaft MembranesDavid L. Worcester†,§ and Michael Weinrich*,†,‡

†NIST Center for Neutron Research, National Institute of Standards and Technology, Gaithersburg 20899, Maryland, United States‡Eunice Kennedy Shriver National Center of Child Health and Human Development, National Institutes of Health, 31 Center Drive,Bethesda 20892, Maryland, United States§Department of Physiology and Biophysics, University of California, Irvine 92697, California, United States

*S Supporting Information

ABSTRACT: Neutron diffraction measurements demonstrate that hydro-static pressure promotes liquid-ordered (Lo) domain formation in lipidmembranes prepared as both oriented multilayers and unilamellar vesiclesmade of a canonical ternary lipid mixture for which demixing transitionshave been extensively studied. The results demonstrate an unusually largedependence of the mixing transition on hydrostatic pressure. Additionally,data at 28 °C show that the magnitude of increase in Lo caused by 10 MPapressure is much the same as the decrease in Lo produced by twiceminimum alveolar concentrations (MAC) of general anesthetics such ashalothane, nitrous oxide, and xenon. Therefore, the results may provide aplausible explanation for the reversal of general anesthesia by hydrostaticpressure.

Inhomogeneity in biological membranes has gained consid-erable interest with the observations that cholesterol-rich

regions of plasma membranes selectively incorporate certainmembrane proteins and thereby regulate membrane processessuch as transport and signaling.1 Model ternary lipidmembranes of unsaturated and saturated phospholipids mixedwith cholesterol exhibit coexisting liquid ordered (Lo) anddisordered (Ld) domains and are thought to resembleproperties of mammalian plasma membranes. The formationand nature of the domains have been extensively studied toestablish domain sizes and phase diagrams based oncomposition and temperature.2 The demixing transitionsoccur over an unusually broad range of temperatures, spanning10 °C or more. This is in distinct contrast with the sharp maintransition (gel to Ld) of single phospholipids. We have recentlyshown that hydrophobic compounds, including halothane,nitrous oxide, and xenon, which are used for inhalationanesthesia, promote mixing and shift the demixing transition byseveral degrees to lower temperatures.3,4 Thus, inhalationanesthetics increase lipid mixing and decrease the amount of Lo(raft) phase.While much work has focused on inhalation anesthetics as

ligands which bind directly to protein ion channels,5 there isstill no satisfactory explanation or consensus for themechanisms of general anesthesia and the remarkable effectof hydrostatic pressure in reversing general anesthesia.Pressures above 10 MPa (∼100 atm) reverse general anesthesiain tadpoles6 and mice;7 however, changes of enzyme activitiesor protein structural changes require pressures at least an orderof magnitude greater.8,9 Because hydrostatic pressure reverses

anesthesia, we investigated the effects of hydrostatic pressureon a canonical Lo/Ld domain forming lipid mixture. The resultsshow that hydrostatic pressure is very effective in producing Lo

domain formation and reversing lipid mixing. The Clapeyronequation dP/dT = ΔH/TΔV (where P is pressure, T istemperature, ΔH is the change in enthalpy, and ΔV is thechange in volume) can be used to describe pressure changes ofthe mixing transition. The results demonstrate surprisingly largevalues of dT/dP over the broad temperature range of thetransition.Neutron diffraction was used to study oriented multilayers of

dipalmitoylphosphatidylcholine (DPPC)/dioleoylphosphatidyl-choline (DOPC)/cholesterol (2/2/1 molar ratio) in D2O atpressures up to 50 MPa. At atmospheric pressure, twodiffraction peaks, corresponding to Lo and Ld domains, wereeasily measured for both first and second orders of diffraction attemperatures at or below 31 °C, as in previous work.3,4 Sampleswere fully hydrated in D2O, which provided high contrast withthe undeuterated lipids and large structure factors for the firstand second orders. Higher diffraction orders are reduced byD2O and intensity was insufficient to readily observe these.After allowing the sample to anneal overnight, the temperaturewas gradually raised until the two first order peaks became onepeak at 34 °C. Hydrostatic pressure of 10.3 MPa resulted inbroadening of this first-order diffraction peak, with a small

Received: September 25, 2015Accepted: October 22, 2015

Letter

pubs.acs.org/JPCL

© XXXX American Chemical Society 4417 DOI: 10.1021/acs.jpclett.5b02134J. Phys. Chem. Lett. 2015, 6, 4417−4421

shoulder, but gave clear separation of second-order diffractionpeaks. With increasing pressure, both the first- and second-order diffraction peaks became increasingly separated (Figures1a,b). This was immediately reversible upon lowering thepressure. Effects of pressure were measured at 23, 28, 31, and34 °C to span the temperature range of the mixing transition.For small changes in the unit cell dimension, the change in

ratio of first-order peak areas (Lo/Ld) approximates the changein the ratio of the mass in the two domains and is decreased byboth increasing temperature and application of anesthetics.3,4

As illustrated in Figure 1c,d, ratios for both first- and second-order peaks decrease with temperature but increase monotoni-cally with increasing pressure at all temperatures. For both first-and second-order ratios, comparing equal ratios at differenttemperatures gives (dT/dP) ≈ − (0.2 °C/MPa).Another way to evaluate the effects of pressure on raft

forming lipids is by comparison with the effects of inhalationanesthetics. The linear fits to the first order ratios at 28 and 31°C in Figure 1c give an increase of 17.6 and 25% in Lo/Ld for a10 MPa change in pressure, respectively. This is much the same(in magnitude) as the decrease in Lo/Ld given by our previousdata3,4 for the effects of twice a minimum alveolarconcentration of the general anesthetics halothane, nitrousoxide, and xenon. For example, xenon and nitrous oxideproduced 15 and 17% reductions in Lo/Ld for 2 MAC,respectively, at 28 °C (from figure 3a,b in ref 4).At each temperature and pressure, d spacings were

determined by linear fits using first- and second-order peakpositions and the Bragg equation for at least two separatemeasurements. To place the d spacings in context, they areplotted in Figure 2a together with data redrawn from earlierneutron diffraction pressure studies on DPPC-cholesterol andegg phosphatidylcholine (EPC) multilamellar vesicles.10 In theprevious work, the d spacings for DPPC-cholesterol and EPCincrease by ∼0.1 Å per 10 MPa. The pressure-induced changesin d spacing for the DPPC/DOPC/cholesterol system arelarger (0.2 to 0.4 Å per 10 MPa), which reflects changes in thecomposition of the domains (demixing) that was not present inthe one- and two-component samples of the previous work.The d spacing for the Ld domain in the present work actuallydecreases (−0.04 to −0.2 Å per 10 MPa) with pressure.Because pure Ld bilayers increase in thickness with pressure10 itis concluded that the observed decrease in d spacing for Lddomains in ternary mixtures reflects the transfer of certainlipids, DPPC, and cholesterol, out of the Ld domains.Because d spacings as well as peak intensities were affected by

pressure, membrane density profiles were calculated by Fouriersynthesis for each temperature and pressure to determine thebilayer thicknesses of the domains.11 Figure 2b displays thesebilayer thicknesses as a function of pressure. The slopes are verysimilar to the slopes for d spacings. Increase in bilayer thicknesswith increasing pressure is a distinctive feature of many lipidbilayer membranes. In these cases, bilayer compressibilitynormal to the membrane is negative and results from large,positive in-plane compressibility. Lipid molecular areas decreasewith pressure, and closer molecular packing therefore results inincreased bilayer thickness. In ternary lipid mixtures such asstudied here, demixing and domain formation or growth alsooccurs, resulting in large thickness increase for liquid-ordereddomains and a distinctive thickness decrease (rather than justsmaller increase) for liquid-disordered domains. Demixingeffects are therefore dominant and the symmetry between theLo and Ld data in Figure 2a,b largely reflects the reciprocal

Figure 1. (a) First-order Bragg diffraction peaks from DOPC/DPPC/cholesterol (2/2/1 molar ratio) multilayers on silicon substrate in D2Oat 34 °C. Neutron wavelength is 5 Å. The momentum transfer (q) isnormal to the plane of the multilayers. Intensity is total neutroncounts. Pressures are relative to atmospheric pressure (1 MPa ≈ 10×atmospheric pressure). There is only one peak at atmospheric pressureand 34 °C. As pressure increases, the peak broadens until clearseparation occurs at 20 MPa. The first peak corresponds to the liquidordered phase and the second peak corresponds to the liquiddisordered phase. (b) Second-order Bragg diffraction peaks for thesame system. Increasing separation between the peaks with increasing

The Journal of Physical Chemistry Letters Letter

DOI: 10.1021/acs.jpclett.5b02134J. Phys. Chem. Lett. 2015, 6, 4417−4421

4418

compositional changes for the two domains, which result fromtransfer of certain lipids from one domain to the other. Thisconclusion is confirmed by the SANS experiment describedbelow.Measurements of Bragg peaks corresponding to different

compositional domains in multilayers require that 2D domainsin individual bilayers stack in a domain-homogeneous way

normal to the bilayer planes.12 This is a remarkable feature athigh hydration and must be due to interbilayer interactions thatdepend on area per lipid so that 2D domains of likecomposition (e.g., Lo and Lo) will stack and form homogeneous3D domains.13 Diffraction techniques will not readily detectdifferent domains without such stacking, and this could be thecase when domains become small.To investigate whether the data from multilayers reflect a

physiologically relevant effect that can occur in singlemembranes, we used small-angle neutron scattering (SANS)to measure the effect of hydrostatic pressure on mixing/demixing transitions in small unilamellar vesicles (SUVs). Todetect domains in the SUVs we employed the contrastmatching technique first developed and employed by Pencer14

for SANS studies of lipid domains. The DPPC in the lipidmixture had both fatty acid chains deuterated, and the D2O/H2O mixture was selected to have the same neutron scatteringlength density as the lipid mixture when fully mixed at 33 °C sothat the scattering intensity at this temperature is nearlyminimum; however, domains smaller than ∼10 nm are hard todetect by SANS,15 and physiological domains may be smallerthan this. As demixing occurs at lower temperatures, contrastwith both types of domains increases, giving increasedscattering. Figure 3a displays the raw intensity data obtainedfrom vesicles at 25 °C. As pressure is increased fromatmospheric pressure to 31 MPa, there is a correspondingincrease in scattering, indicating the growth of two distinctdomains with different scattering length densities.Quantification of the domain separation is best accomplished

using the scattering invariant Q,14,16,17 Q = ∫ Iq2 dq, and undermatch conditions is a function of the relative proportions of thetwo domains.17 The relationship between temperature,pressure, and the scattering invariant Q is illustrated in Figure3b.At atmospheric pressure and 31 °C, detection of separate

domains is barely visible and the signal grows progressivelylarger as temperature is decreased, corresponding well toprevious SANS measurements on this mixture.16 In fact, ourmeasurements reproduced the data of ref 16 almost exactly. At25 °C measurements at different pressures demonstrate thesame antagonism between temperature and pressure, as seen inthe multilayer experiment. There is a nonlinear increase in thescattering invariant Q up to 10 MPa. But at 21 and 31 MPa wesee a fairly constant relationship between temperature andpressure: − 1 °C ≈ 4 MPa, according to which, increasing thepressure by 4 MPa produces the same change in the invariantscattering function as a 1 °C reduction in temperature. Thus,measurements on the mixing transition of both the multilayerand unilamellar vesicles give values of dT/dP of −0.2 to −0.3°C/MPa. These values are in the same range as those observedfor the gel to liquid crystal transition for DPPC18 and othersaturated phospholipids.19 We chose the lipid mixture for thecurrent study because it is widely used as a model lipid raftmixture and because it has a broad mixing transition withcoexisting Lo and Ld phases through a wide temperature range.

2

Values of dT/dP for the gel to liquid crystal transition insaturated phospholipids change very little with the acyl chainlength,19 and the mixing transitions of different ternary raftforming mixtures are qualitatively similar between the differentmixtures, albeit with different transition midpoints.20 By theClapeyron equation, the similar values of dT/dP mean that theΔV/ΔS values are similar, as expected if there are similar

Figure 1. continued

pressure is more evident than in the first order peaks. (c) Ratios of theareas of first-order Bragg peaks as a function of pressure at fourtemperatures. Dotted lines are drawn to guide the eye. Solid lines arelinear fits (Supporting Information). At 34 °C and atmosphericpressure, the ratio of Lo to Ld is zero. (d) Ratios of the areas of thesecond-order Bragg peaks as a function of pressure at fourtemperatures.

Figure 2. (a) d spacing for the Lo and Ld phases as a function ofpressure for different temperatures. Error bars are smaller than theheights of the symbols. Solid lines are linear fits. Dotted lines aredrawn to guide the eye. Dashed lines are redrawn from ref 10 for aDPPC/cholesterol mixture and egg phosphatidylcholine (EPC), asindicated on the graph. Note that the d spacing for the Lo phaseincreases with increasing pressure more steeply than that of theDPPC/cholesterol binary mixture, while the spacing for the Ld phaseactually decreases with increasing pressure. The nonlinear portion ofthe trace at 34 °C makes clear that this unusual decrease in d spacing ismatched by a corresponding increase in the Lo phase, reflecting atransfer of lipid between these two phases. (b) Widths of the Lo and Ldbilayer lipid phases from Fourier reconstructions as a function ofpressure and temperature.

The Journal of Physical Chemistry Letters Letter

DOI: 10.1021/acs.jpclett.5b02134J. Phys. Chem. Lett. 2015, 6, 4417−4421

4419

volume/disorder changes in the hydrocarbon regions in bothcases.Previous investigators used differential scanning calorimetry

and electron spin resonance to examine the main transitions ofpure, saturated phospholipids18 and binary mixtures ofsaturated phospholipids with cholesterol21 as well as mixingtransitions in binary mixtures of saturated phospholipids whichdiffer in fatty acid chain lengths by at least four carbons.22

Studies also examined the antagonistic effects of temperatureand pressure on membrane fluidity, for which dT/dP is thechange in temperature required to offset the effects of a changein pressure. Such studies of synaptic and myelin fractions ofgoldfish brain23 and liquid crystalline lipid bilayers24 gave dT/dP values in the range of 0.13 to 0.21 °C/MPa. Severalgroups25−28 previously observed antagonistic effects of pressureand anesthetics on transitions between gel and liquid crystalstates in these model systems. Extending these studies to morephysiological systems requires unsaturated as well as saturatedphospholipids and in place of gel states, Lo domains.Our demonstrations that hydrostatic pressure of physiolog-

ically relevant magnitude reverses domain mixing in bothmultilayers and unilamellar vesicles of a ternary lipid raftmixture, suggest that anesthetic effects found for lipid

domains3,4 have physiological significance. There is strongevidence of the association of signaling proteins, includingsome ion channels, with lipid rafts,1,29 and lipid rafts have beenshown to be involved in the organization of neurotransmitterreceptor complexes on postsynaptic membranes,30−33 includingthose involved in nociception.33 Rapid trafficking of gluta-matergic receptors in and out of the region of post synapticdensity is well established and appears crucial for the regulationof excitatory synapses.34−36 One implication of the pressureresults reported here, combined with our3,4 and other’s37

demonstrations of the effects of anesthetics on liquid-ordered(raft) domains, is that inhalation anesthetics could dispersepostsynaptic receptors, which are normally gathered at synapseswhere they are available to released neurotransmitters. Bydispersal, receptors are effectively inhibited because lessneurotransmitter is available to them outside the synapticregion. Receptor dispersal would act in addition to receptorinhibition by direct binding of anesthetic to receptors. Thus,these two mechanisms may work together to diminish synaptictransmission. The former is expected to be responsible forpressure reversal of anesthesia because pressure reversal of thelatter has been difficult to demonstrate38 (see also Note inSupporting Information). It is concluded that a plausibleexplanation of the phenomenon of pressure reversal of generalanesthesia is provided by lipid Lo domain formation and that Lodomain dissolution is a mechanism of anesthetic action thatworks in addition to receptor inhibition by direct anestheticbinding. This is based on the physical properties of complexlipid mixtures that are significant for the lipid/protein mixturesof biological membranes, especially at synapses.

■ ASSOCIATED CONTENT*S Supporting InformationThe Supporting Information is available free of charge on theACS Publications website at DOI: 10.1021/acs.jpclett.5b02134.

Detailed materials and methods, coefficients to the linearfits to Figures 1 and 2, phase diagram, and notes. (PDF)

■ AUTHOR INFORMATIONCorresponding Author*Phone 301-402-4201. E-mail mw287k@nih.gov.NotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTSWe thank the National Science Foundation and the NISTCenter for Neutron Research for access to neutron scatteringfacilities and Juscellino Leao for providing and helping withpressure equipment. We thank Boualem Hammouda and JunLiu for help with SANS. Support to D.L.W. was provided byU.S. National Institutes of Health Grant GM 86685 (toStephen H. White). We thank Dr. Sergey Bezrukov for usefuldiscussions. The identification of any commercial product ortrade name does not imply any endorsement or recommenda-tion by the National Institute of Standards and Technology.

■ REFERENCES(1) Lingwood, D.; Simons, K. Lipid Rafts as a Membrane-OrganizingPrinciple. Science 2010, 327, 46−50.(2) Feigenson, G. W. Phase Diagrams and Lipid Domains inMulticomponent Lipid Bilayer Mixtures. Biochim. Biophys. Acta,Biomembr. 2009, 1788, 47−52.

Figure 3. (a) Small-angle neutron scattering intensity versus q forD62-DPPC/DOPC/cholesterol (2/2/1 molar ratio) unilamellarvesicles in D2O/H2O (45%) at 25 °C matched to the scatteringlength for vesicles at atmospheric pressure at 33 °C: ▼, 0 MPa; ▲, 10MPa; ●, 21 MPa; ■, 31 MPa. Detector at 5 m, neutron wavelength 6Å. (b) Scattering invariant Q (∫ Iq2 dq) for the SANS data as afunction of pressure and temperature. Dotted line is drawn to guidethe eye through the points at 0 MPa. The arrow indicates ΔT, whileΔP is the MPa value at 25 °C and the ratio of these values gives dT/dP, assuming linearity. The dT/dP values thus obtained are 0.26, 0.24,and 0.20 °C/MPa, from top to bottom.

The Journal of Physical Chemistry Letters Letter

DOI: 10.1021/acs.jpclett.5b02134J. Phys. Chem. Lett. 2015, 6, 4417−4421

4420

(3) Weinrich, M.; Nanda, H.; Worcester, D. L.; Majkrzak, C. F.;Maranville, B. B.; Bezrukov, S. M. Halothane Changes the DomainStructure of a Binary Lipid Membrane. Langmuir 2012, 28, 4723−8.(4) Weinrich, M.; Worcester, D. L. Xenon and Other VolatileAnesthetics Change Domain Structure in Model Lipid RaftMembranes. J. Phys. Chem. B 2013, 117, 16141−7.(5) Franks, N. P.; Lieb, W. R. Molecular Mechanisms of GeneralAnaesthesia. Nature 1982, 300, 487−93.(6) Johnson, F. H.; Flagler, E. A. Hydrostatic Pressure Reversal ofNarcosis in Tadpoles. Science 1950, 112, 91−2.(7) Lever, M. J.; Miller, K. W.; Paton, W. D.; Smith, E. B. PressureReversal of Anaesthesia. Nature 1971, 231, 368−71.(8) Boonyaratanakornkit, B. B.; Park, C. B.; Clark, D. S. PressureEffects on Intra- and Intermolecular Interactions Within Proteins.Biochim. Biophys. Acta, Protein Struct. Mol. Enzymol. 2002, 1595, 235−49.(9) Girard, E.; Marchal, S.; Perez, J.; Finet, S.; Kahn, R.; Fourme, R.;Marassio, G.; Dhaussy, A. C.; Prange, T.; Giffard, M.; Dulin, F.;Bonnete, F.; Lange, R.; Abraini, J. H.; Mezouar, M.; Colloc’h, N.Structure-Function Perturbation and Dissociation of Tetrameric UrateOxidase by High Hydrostatic Pressure. Biophys. J. 2010, 98, 2365−73.(10) Braganza, L. F.; Worcester, D. L. Structural Changes in LipidBilayers and Biological Membranes Caused Hydrostatic Pressure.Biochemistry 1986, 25, 7484−8.(11) Worcester, D. L.; Franks, N. P. Structural Analysis of HydratedEgg Lecithin and Cholesterol Bilayers. II. Neutron Diffraction. J. Mol.Biol. 1976, 100, 359−78.(12) Mills, T. T.; Tristram-Nagle, S.; Heberle, F. A.; Morales, N. F.;Zhao, J.; Wu, J.; Toombes, G. E.; Nagle, J. F.; Feigenson, G. W.Liquid-Liquid Domains in Bilayers Detected by Wide Angle X-RayScattering. Biophys. J. 2008, 95, 682−90.(13) Tayebi, L.; Ma, Y.; Vashaee, D.; Chen, G.; Sinha, S. K.; Parikh,A. N. Long-Range Interlayer Alignment Of Intralayer Domains InStacked Lipid Bilayers. Nat. Mater. 2012, 11, 1074−80.(14) Pencer, J.; Mills, T.; Anghel, V.; Krueger, S.; Epand, R. M.;Katsaras, J. Detection of Submicron-Sized Raft-Like Domains inMembranes by Small-Angle Neutron Scattering. Eur. Phys. J. E: SoftMatter Biol. Phys. 2005, 18, 447−58.(15) Petruzielo, R. S.; Heberle, F. A.; Drazba, P.; Katsaras, J.;Feigenson, G. W. Phase Behavior and Domain Size in Sphingomyelin-Containing Lipid Bilayers. Biochim. Biophys. Acta, Biomembr. 2013,1828, 1302−13.(16) Masui, T.; Urakami, N.; Imai, M. Nano-Meter-Sized DomainFormation in Lipid Membranes Observed by Small Angle NeutronScattering. Eur. Phys. J. E: Soft Matter Biol. Phys. 2008, 27, 379−89.(17) Pencer, J.; Mills, T. T.; Kucerka, N.; Nieh, M. P.; Katsaras, J.Small-Angle Neutron Scattering to Detect Rafts and Lipid Domains.Methods Mol. Biol. 2007, 398, 231−44.(18) Albon, N.; Sturtevant, J. M. Nature of the Gel to Liquid CrystalTransition of Synthetic Phosphatidylcholines. Proc. Natl. Acad. Sci. U.S. A. 1978, 75, 2258−60.(19) Ichimori, H.; Hata, T.; Matsuki, H.; Kaneshina, S. BarotropicPhase Transitions and Pressure-Induced Interdigitation on BilayerMembranes of Phospholipids with Varying Acyl Chain Lengths.Biochim. Biophys. Acta, Biomembr. 1998, 1414, 165−74.(20) Veatch, S. L.; Keller, S. L. Separation of Liquid Phases in GiantVesicles of Ternary Mixtures of Phospholipids and Cholesterol.Biophys. J. 2003, 85, 3074−3083.(21) Shimshick, E. J.; McConnell, H. M. Lateral Phase Separations inBinary Mixtures of Cholesterol and Phospholipids. Biochem. Biophys.Res. Commun. 1973, 53, 446−51.(22) Mabrey, S.; Sturtevant, J. M. Investigation of Phase Transitionsof Lipids and Lipid Mixtures by Sensitivity Differential ScanningCalorimetry. Proc. Natl. Acad. Sci. U. S. A. 1976, 73, 3862−6.(23) Chong, P. L.; Cossins, A. R.; Weber, G. A Differential PolarizedPhase Fluorometric Study of the Effects of High Hydrostatic PressureUpon the Fluidity of Cellular Membranes. Biochemistry 1983, 22,409−15.

(24) Chong, P. L.; Weber, G. Pressure Dependence of 1,6-Diphenyl-1,3,5-Hexiatriane Flourescence in Single-Component Liposomes.Biochemistry 1983, 22, 5544−5550.(25) Kamaya, H.; Ueda, I.; Moore, P. S.; Eyring, H. AntagonismBetween High Pressure and Anesthetics In The Thermal Phase-Transition of Dipalmitoyl Phosphatidylcholine Bilayer. Biochim.Biophys. Acta, Biomembr. 1979, 550, 131−7.(26) Mountcastle, D. B.; Biltonen, R. L.; Halsey, M. J. Effect ofAnesthetics and Pressure on the Thermotropic Behavior of Multi-lamellar Dipalmitoylphosphatidylcholine Liposomes. Proc. Natl. Acad.Sci. U. S. A. 1978, 75, 4906−10.(27) Trudell, J. R.; Hubbell, W. L.; Cohen, E. N.; Kendig, J. J.Pressure Reversal of Anesthesia: The Extent of Small-MoleculeExclusion From Spin-Labeled Phospholipid Model Membranes.Anesthesiology 1973, 38, 207−11.(28) Trudell, J. R.; Payan, D. G.; Chin, J. H.; Cohen, E. N. TheAntagonistic Effect of an Inhalation Anesthetic and High Pressure onthe Phase Diagram of Mixed Dipalmitoyl-Dimyristoylphosphatidylcho-line Bilayers. Proc. Natl. Acad. Sci. U. S. A. 1975, 72, 210−3.(29) Head, B. P.; Patel, H. H.; Insel, P. A. Interaction of Membrane/Lipid Rafts with the Cytoskeleton: Impact on Signaling and Function:Membrane/Lipid Rafts, Mediators of Cytoskeletal Arrangement andCell Signaling. Biochim. Biophys. Acta, Biomembr. 2014, 1838, 532−45.(30) Delint-Ramirez, I.; Fernandez, E.; Bayes, A.; Kicsi, E.;Komiyama, N. H.; Grant, S. G. In Vivo Composition of NMDAReceptor Signaling Complexes Differs Between Membrane Sub-domains and is Modulated By PSD-95 and PSD-93. J. Neurosci. 2010,30, 8162−70.(31) Delint-Ramirez, I.; Salcedo-Tello, P.; Bermudez-Rattoni, F.Spatial Memory Formation Induces Recruitment of NMDA Receptorand PSD-95 to Synaptic Lipid Rafts. J. Neurochem. 2008, 106, 1658−68.(32) Swulius, M. T.; Farley, M. M.; Bryant, M. A.; Waxham, M. N.Electron Cryotomography of Postsynaptic Densities During Develop-ment Reveals a Mechanism of Assembly. Neuroscience 2012, 212, 19−29.(33) Pristera, A.; Okuse, K. Building Excitable Membranes: LipidRafts and Multiple Controls on Trafficking of Electrogenic Molecules.Neuroscientist 2012, 18, 70−81.(34) Choquet, D.; Triller, A. The Dynamic Synapse. Neuron 2013,80, 691−703.(35) Huganir, R. L.; Nicoll, R. A. AMPARs and Synaptic Plasticity:The Last 25 Years. Neuron 2013, 80, 704−17.(36) Renner, M.; Choquet, D.; Triller, A. Control of the PostsynapticMembrane Viscosity. J. Neurosci. 2009, 29, 2926−37.(37) Gray, E.; Karslake, J.; Machta, B. B.; Veatch, S. L. Liquid GeneralAnesthetics Lower Critical Temperatures in Plasma MembraneVesicles. Biophys. J. 2013, 105, 2751−9.(38) Franks, N. P. Molecular Targets Underlying GeneralAnaesthesia. Br. J. Pharmacol. 2006, 147 (Suppl 1), S72−81.

The Journal of Physical Chemistry Letters Letter

DOI: 10.1021/acs.jpclett.5b02134J. Phys. Chem. Lett. 2015, 6, 4417−4421

4421