AKABORI AND NAGLE VOL. 8 ’ NO. 4 ’ 3123–3127 ’ 2014

www.acsnano.org

3123

April 14, 2014

C 2014 American Chemical Society

Comparing Lipid Membranesin Different EnvironmentsKiyotaka Akabori and John F. Nagle*

Department of Physics, Carnegie Mellon University, Pittsburgh, Pennsylvania 15213, United States

Biomembranes are one of biology'skey nanomaterials (∼5 nm thick),and they are essential to all cellular

life. Manipulation of membranes and oftheir constitutive core, lipid bilayers, is ne-cessary to engineer systems toward applica-tions.1�4 It is important to characterize howsuch manipulations and the resulting envi-ronment perturb the structure and to buildsystems that minimize unnatural perturba-tions.5�8 The study by Watkins et al.9 de-scribed in this issue ofACSNanohas focusedon a detailed characterization of single lipidbilayers supported on a solid substrate.Remarkable differences were found com-pared to lipid monolayers, for which thesupported bilayers were constructed by theLangmuir�Blodgett/Langmuir�Schaefer(LB/LS) technique. Some data in the litera-ture suggest that differences also occur fortraditional systems that consist of multi-lamellar arrays of lipid bilayers,10�13 but

there is some concern that those resultswere not well-documented.9

The primary experimental quantity exhib-iting the differences found byWatkins et al.9

is the wide-angle X-ray scattering (WAXS)chain packing reflections in the gel phaseof dipalmitoylphosphatidylcholine (DPPC)lipid bilayers.10�13 Each lipid has a pair ofhydrophobic saturated hydrocarbon chains2 nm long capped on one end by a hydro-philic headgroup in contact with water,thereby forming a bilayer often thought ofas two back-to-back monolayers. The chainstilt with an angle θ relative to the bilayernormal because the interfacial area occupiedby the headgroup is greater than the areaof the pair of chains.14 Remarkably, thechains in each monolayer tilt in the sameazimuthal (φ) direction even though thereare no covalent bonds between the mono-layers.10,11,15 As the number of X-ray reflec-tions is far smaller than occurs in polyethy-lene crystals or in the subgel phase of lipidbilayers,16 the classical interpretation ofthe gel phase WAXS data is that the chainsare disordered with respect to rotationsabout their long axis, suggesting thatthey can be modeled as tilted cylinderspacked together. However, this classicalinterpretation was refined to accommodatean apparent discrepancy that occurred inmultilamellar arrays between the value ofthe tilt angle θ inferred from infrared (IR)spectroscopy and that inferred just from thepositions of the WAXS peaks.17 Infrared is

The study by Watkins et al.

described in this issue of ACS

Nano has focused on a

detailed characterization of

single lipid bilayers supported

on a solid substrate.

* Address correspondence tonagle@cmu.edu.

Published online10.1021/nn501499t

ABSTRACT When engineering lipid membranes for applications, it is essential to characterize them to

avoid artifacts introduced by manipulation and the experimental environment. Wide-angle X-ray scattering is

a powerful structural characterization tool for well-ordered lipid systems. It reveals remarkable differences in

rotational order parameters for samples prepared in different ways. New data and perspectives are presented

here for multilamellar systems that support and extend the characterization work on unilamellar systems

that is reported by Watkins et al. in this issue of ACS Nano.

PERSPECTIV

E

AKABORI AND NAGLE VOL. 8 ’ NO. 4 ’ 3123–3127 ’ 2014

www.acsnano.org

3124

sensitive not only to the tilt anglebut also to the rotational angle γof the all-trans C�C 3 3 3C�C planeabout the chain long axis relativeto the plane defined by the tiltdirection and the bilayer normal(see Figure 1). A two-dimensionalrotational order parameter, g =Æ2 cos2 γ � 1æ, is the appropriatestatistically averaged quantity todescribe this rotational disorder.Complete disorder has all γ equallyprobable and g = 0, whereas com-plete rotational order correspondsto |g| = 1 (g = �1 has the zigzagchain plane perpendicular to the tiltplane). A value g = �0.3 resolved

the IR/WAXS difference for multi-lamellar arrays.17

Wide-angle X-ray scattering datacontain information in addition tothe peak positions in reciprocalspace that suffice to obtain thechain tilt θ. The additional informa-tion is the relative intensity R of theWAXS in-plane reflections, usuallycalled the (20) peak in the literature,and the out-of plane (11) peak. Theclassical model of tilted cyclindersgives R = 1. Even before the currentpaper of Watkins et al.,9 it has beenknown that R depends on the rota-tional order parameter g, but a sim-plemodel for the electron density of

the chains suggested that R wouldonly be reduced by 40% for com-plete rotational order compared tothe value R = 1, so it was previouslyignored when interpreting WAXS.15

In contrast, the new model ofWatkins et al. suggests that R ismore sensitive to g,9 so careful mea-surement of R opens a new windowon the structure of lipid bilayer sys-tems. For supported lipid bilayers,the Watkins paper reports quitesmall experimental values for R ran-ging downward from less than 0.4,giving values of g varying from�0.5to �1.0 as the monolayer pressureat deposition was increased. In strik-ing contrast, g was reported to beessentially zero for monolayers, cor-responding to R= 1 and the classicalcylindrical model. In view of theseconsiderable differences betweenmonolayers and supported bilayers,it is of interest to compare R andg to the more traditional multi-lamellar lipid bilayer systems. Whilethe literature has some suggestivefigures,10�13 values of R were notgiven quantitatively, so it is timelyto viewhigher quality data that con-firm R ≈ 1.

Figure 2 shows grazing incidence(GIXD) WAXS scattering from amultilamellar stack of ∼2000 DMPClipid bilayers supported on a Si sub-strate. The lamellar repeat spacingD = 5.8 nm obtained from the la-mellar orders on the left indicatedthat the stack was nearly fully hy-drated from water vapor with anestimated RH greater than 98%.The intense peak in the upper rightcenter is the out-of-plane (11) re-flection. Directly below that is thefirst satellite on the (11) Bragg rod10

that is also seen in unoriented

Figure 1. Geometry of the hydrocarbon chains. (a) Normal N is perpendicular tothe plane of the bilayer, and the long axis along the all-trans chain is tilted by θ. (b)Looking down the chain axis, the carbons in the zigzag chain are shown by darkcircles that define a planeperpendicular to the paper that is rotatedbyγ relative tovector V that is in the chain tilt plane shown in (a).

Figure 2. Grazing incidence wide-angle X-ray scattering from an oriented stack of1,2-dimyristoyl-sn-glycero-3-phosphocholine (DMPC) bilayers; white is most in-tense. The beam is located at the x near the center left, and lamellar repeat ordersh=2�7 and h=9 are on themeridian above the beam. The bright (11) peak is locatedat (qr,qz) = (13.0 nm�1, 8.2 nm�1). The (1,�1) peak is strongly attenuated by thesubstrate.Most of the (20) peak, subsequently shown tobe centered at (14.8 nm�1, 0),is evenmore strongly attenuated by the substrate and the sample itself. A satellite onthe (1,1) Bragg rod occurs below the (1,1) peak. Temperature was 10 �C.

It is of interest to

compare R and g to the

more traditional

multilamellar lipid

bilayer systems.

PERSPECTIV

E

AKABORI AND NAGLE VOL. 8 ’ NO. 4 ’ 3123–3127 ’ 2014

www.acsnano.org

3125

multilamellar vesicles.15 The (1,�1)reflection, though strongly attenu-ated by having to go through the thin(35 μm) substrate, can be seen in thelower right center.Only the tail of the(20) reflection is visible. The centralpart of the (20) reflection cannot beseen because, for a positive incidentangleω, those scattered X-rays musttravel long distances through thesubstrate as well as the sample itself.When ω is decreased, more of the(20) reflection can be seen,13,18 but

its intensity still appears somewhatweaker than the (11) reflection. Notethat, historically, the (20) peak wasoften not observed and there waseven concern that it did not exist.While the grazing incidence ex-

perimental geometry in Figure 2might be taken to indicate R < 1,that is wrong. Figure 3 shows scat-tering from the same sample buttilted with respect to the beam byω = 45�. As was described previ-ously,11 this moves the reflections

around on the area detector, butmost importantly, the (20), (1,1), and(1,�1) reflections traverse nearlyequal thickness of the thin sub-strate, so the differential attenua-tion is negligible.Intensity profiles through the or-

ders in Figure 3 are shown in Figure 4.The in-plane (20) peak is narrowwithgreater maximum intensity than theout-of-plane (11) and (1,�1) peaks,so the integrated intensity is aboutthe same for all peaks, consistentwith the conventional view thatR = 1 for multilamellar systems.(The (11) peak width is instrumen-tally resolved here. The (20) widthwas even narrower when the instru-mental resolutionwas greater,15 butinstrumental resolution does notaffect R.) However, for a differentexposure of the same sample, R

was about 0.8, more consistent withthe g = �0.3 obtained from usingIR and WAXS.17 Nevertheless, asWatkins et al. suppose,9 there is aconsiderable difference comparedto the much smaller value of R

they obtain for supported singlebilayers.As supported single lipid bilayers

are used for many purposes, it isimportant that the method of form-ing them and subsequent distor-tions be documented. It could beargued that the most relevant lipidbilayer system to be studied is thefluid phase that does not have arotational order parameter. How-ever, obtaining fluid phase structureis relatively difficult and differenceseven more so. Another advantageof the gel phase for this kind ofstudy is that it is relatively rigid com-pared to fluid phases and would beexpected to be more robust underthe stresses of deposition and theeffect of a substrate. Detection ofeven subtle changes in the gelphase, like the rotational order pa-rameter, would indicate stressesthat would then likely deform a fluidphase even more, albeit undetect-ably so, but not necessarily withoutconsequences for studies employ-ing supported bilayers. Concernsabout the substrate have, indeed,

Figure 3. Transmission wide-angle X-ray scattering from the sample in Figure 1with the beam incident atω=45�, so the threemain labeled reflections are rotatedin the laboratory frame according to formula11 that also gives the q values in thecaption to Figure 2. The first satellites of the (1,1) and (1,�1) are visible, as are theoverlapping second satellites15 centered on the equator (qr = 0), which follows atrajectory from the beam toward the (20) reflection. The weak ring closer to thebeam than the WAXS reflections is from a Mylar window on the sample chamberand occurs because of imperfect subtraction of the background obtained from asubstrate with no lipid. The intensity of the beam on the charge-coupled devicewas highly attenuated by an absorber downstream from the sample, and there areno lamellar reflections for large incidence angles ω.

Figure 4. Integrated intensity versus pixel number in the direction perpendicularto the elongated reflections on the charge-coupled device in Figure 3.

PERSPECTIV

E

AKABORI AND NAGLE VOL. 8 ’ NO. 4 ’ 3123–3127 ’ 2014

www.acsnano.org

3126

led to devising many ways to sup-port single bilayers.1�8,12

The result for supported singlebilayers that R ∼ 0.3 from the studyof Watkins et al.9 together with theresult that R = 1 for multilamellarstacks makes it clear that there is asubstantial difference. As the multi-lamellar stacks have an ample 2 nmwater cushion between adjacentbilayers, it is unlikely that stackingaffects the properties of the indivi-dual bilayers; this is supported bythe absence of three-dimensionalWAXS reflections, indicating no cor-relation of the in-plane structure be-tween adjacent bilayers.10 Differenceswith the results from supported bi-layers would then be attributedeither to interactions with the sub-strate or to the deposition method.Because Watkins et al. obtain thesame small R using a different de-position method, the focus shifts tothe substrate. While multilamellarstacks are also deposited on sub-strates and the closest bilayerwouldlikely have similar properties tothe single supported bilayer, WAXSoverwhelmingly comes from theother ∼2000 bilayers that do notdirectly interact with the substrate.A concern for the results of

Watkins et al.9might have been thatthey did not perform transmissionWAXS, so their R value from GIXDmight have been small for the samereason that it has been artifactuallysmall for multilamellar samples.One difference is that they usehigher energy X-rays to make watermore transparent, but that is sim-ilar to having air above multilayerstacks. Their substrate was quartz,which is 20 times as absorptive aswater at their X-ray energy, so thatis similar to previous studies wherethe substrate is also much moreabsorptive than air. Crucially, theyperformed an experiment thatallays this concern; namely, theyfound that monolayers on waterand on quartz both have R ∼ 1. Alikely explanation for R smaller than1 for GIXD of oriented stacks is thatthe (20) scattering is attenuatedfrom those bilayers deep within

the 10 μm multilamellar stack be-cause they have to go fartherthrough the upper bilayers thanthe (11) scattering. This explanationcould be tested by examiningmulti-layer stacks with higher energyX-rays, which have much longerattenuation lengths in lipids.While it would be attractive to

suppose that the substrate alonecauses the difference in R, this isrefuted by the finding of Watkinset al. that monolayers also haveR = 1 when deposited on a quartzsubstrate.9 A difference betweenmonolayers and bilayers, both onthe substrate, is the interactionsbetween the monolayers in a bi-layer that manifests itself in parallelalignment of the chains in bothmonolayers.9�11 Watkins et al.

therefore suggest that obtaining asmall R and rotational ordering re-quires both the substrate interac-tion and the interactions betweenthe two monolayers in a bilayer.They also carefully document theinteresting and important variationof structural parameters in mono-layers and supported lipid bilayersas the surface pressure in the LB/LSdeposition method is varied. TheirFigure 6 shows that the chain tilt θof the supported bilayer is largerthan the monolayer from whichit was formed. That would be ex-pected to make the area per mole-cule larger for the bilayer, but theopposite occurs because the origi-nal packing density reflected in theinverse chain area Achain is consider-ably smaller for the monolayer. Thisis additional evidence supportingthe perspective that bilayers arenot simply two back-to-back mono-layers and that, instead of usingone system to elucidate the other,efforts should focus on the differ-ences between them.19 These newdata nicely do so.9

Methods for the Data in the IncludedFigures. Samples were prepared onthin (35 μm) Si wafers as previouslydescribed.13 Experiments were per-formed at the G1 station at CHESS.The X-ray energy was 10.5 keVat which the absorption length of

lipid is 2.6 mm. The W/B4C multi-layer monochromator had 1.2% en-ergy dispersion. The beam size was0.2 mm � 0.2 mm, and measureddivergences were less than 0.0001radians. Reflections recorded on the1024 � 1024 CCD detector (FingerLakes Instrumentation, Lima, NY)were broadened by the X-ray en-ergy dispersion and divergence; ad-ditionally, they were geometricallybroadened in Figure 2 by the beamwidth and the 25 mm length ofthe sample along the beam and inFigure 3 by the beam width andheight. The pixel size of the detectorwas 71.13 μm, and the distance fromsample to detector was 175 mm. Atthese distances, there were scatter-ing rings from the Mylar windowsof the hydration chamber.13 Highlyaccurate background for Figure 2was obtainedby rotating the sampleto negative angle to block scatteringfrom the sample.13 For the back-ground for Figure 3, the sample hadtobe replacedbyadifferent bare thinSi substrate, resulting in incompletesubtraction of the Mylar ring.

Conflict of Interest: The authors de-clare no competing financial interest.

Acknowledgment. The new data weretaken for a project supported by NIHGrant R01GM44976 at the Cornell HighEnergy Synchrotron Source (CHESS) sup-ported by the National Science Founda-tion (NSF) and the NIH/NIGMS under NSFAward DMR-0936384.

REFERENCES AND NOTES1. Branton, D.; Deamer, D. W.; Marziali,

A.; Bayley, H.; Benner, S. A.; Butler, T.;Di Ventra, M.; Garaj, S.; Hibbs, A.;Huang, X. H.; et al. The Potentialand Challenges of Nanopore Se-quencing. Nat. Biotechnol. 2008,26, 1146–1153.

2. Gupta, S.; Dura, J. A.; Freites, J. A.;Tobias, D. J.; Blasie, J. K. StructuralCharacterization of the Voltage-Sensor Domain and Voltage-GatedKþ-Channel Proteins VectoriallyOriented within a Single BilayerMembrane at the Solid/Vapor andSolid/Liquid Interfaces via NeutronInterferometry. Langmuir 2012, 28,10504–10520.

3. Hsieh, W. T.; Hsu, C. J.; Capraro, B. R.;Wu, T. T.; Chen, C. M.; Yang, S.;Baumgart, T. Curvature Sortingof Peripheral Proteins on Solid-Supported Wavy Membranes. Lang-muir 2012, 28, 12838–12843.

PERSPECTIV

E

AKABORI AND NAGLE VOL. 8 ’ NO. 4 ’ 3123–3127 ’ 2014

www.acsnano.org

3127

4. Chung, M.; Koo, B. J.; Boxer, S. G.Formation and Analysis of Topogra-phical Domains between LipidMembranes Tethered by DNA Hy-brids of Different Lengths. FaradayDiscuss. 2013, 161, 333–345.

5. Hughes, L. D.; Boxer, S. G. DNA-Based Patterning of Tethered Mem-brane Patches. Langmuir 2013, 29,12220–12227.

6. Jablin, M. S.; Dubey, M.; Zhernenkov,M.; Toomey, R.; Majewski, J. Influ-ence of Lipid Membrane Rigidityon Properties of Supporting Poly-mer. Biophys. J. 2011, 101, 128–133.

7. Budvytyte, R.; Valincius, G.; Niaura,G.; Voiciuk, V.; Mickevicius, M.;Chapman, H.; Goh, H. Z.; Shekhar,P.; Heinrich, F.; Shenoy, S.; et al.Structure and Properties of TetheredBilayer Lipid Membranes with Unsa-turated Anchor Molecules. Langmuir2013, 29, 8645–8656.

8. Fragneto, G.; Charitat, T.; Daillant, J.Floating Lipid Bilayers: Models forPhysics and Biology. Eur. Biophys. J.2012, 41, 863–874.

9. Watkins, E. B.; Miller, C. E.; Liao, W.-P.;Kuhl, T. L. Equilibrium or Quenched:Fundamental Differences betweenLipidMonolayers, Supported Bilayers,and Membranes. ACS Nano 2014,DOI: 10.1021/nn4052953.

10. Smith, G. S.; Sirota, E. B.; Safinya, C. R.;Plano, R. J.; Clark, N. A. X-ray Struc-tural Studies of Freely SuspendedOrdered Hydrated DMPC Multi-membrane Films. J. Chem. Phys.1990, 92, 4519–4529.

11. Tristram-Nagle, S.; Zhang, R.; Suter,R. M.; Worthington, C. R.; Sun, W. J.;Nagle, J. F. Measurement of ChainTilt Angle in Fully Hydrated Bilayersof Gel Phase Lecithins. Biophys. J.1993, 64, 1097–1109.

12. Ziblat, R.; Leiserowitz, L.; Addadi, L.Crystalline Lipid Domains: Charac-terization by X-ray Diffraction andTheir Relation to Biology. Angew.Chem., Int. Ed. 2011, 50, 3620–3629.

13. Mills, T. T.; Toombes, G. E. S.; Tristram-Nagle, S.; Smilgies, D. M.; Feigenson,G. W.; Nagle, J. F. Order Parametersand Areas in Fluid-Phase OrientedLipid Membranes Using Wide AngleX-ray Scattering. Biophys. J. 2008, 95,669–681.

14. Nagle, J. F. Theory of the Main LipidBilayer Phase-Transition. Annu. Rev.Phys. Chem. 1980, 31, 157–195.

15. Sun, W. J.; Suter, R. M.; Knewtson,M. A.; Worthington, C. R.; Tristram-Nagle, S.; Zhang, R.; Nagle, J. F. Orderand Disorder in Fully HydratedUnoriented Bilayers of Gel PhaseDipalmitoylphosphatidylcholine.Phys. Rev. E 1994, 49, 4665–4676.

16. Blaurock, A. E.;Mcintosh, T. J. Structureof the Crystalline Bilayer in the SubgelPhase of Dipalmitoylphosphatidyl-glycerol. Biochemistry 1986, 25, 299–305.

17. Nagle, J. F. Evidence for PartialRotational Order in Gel Phase DPPC.Biophys. J. 1993, 64, 1110–1112.

18. Tristram-Nagle, S.; Liu, Y. F.; Legleiter,J.; Nagle, J. F. Structure of Gel PhaseDMPC Determined by X-ray Diffrac-tion. Biophys. J.2002, 83, 3324–3335.

19. Nagle, J. F.; Tristram-Nagle, S. Struc-ture of Lipid Bilayers. Biochim. Bio-phys. Acta, Rev. Biomembr. 2000,1469, 159–195.

PERSPECTIV

E