Immobilization and the conformational study of phospholipid and phospholipid-protein vesicles

Materials Science and Engineering C 29 (2009) 1480–1485

Contents lists available at ScienceDirect

Materials Science and Engineering C

j ourna l homepage: www.e lsev ie r.com/ locate /msec

Immobilization and the conformational study of phospholipid andphospholipid-protein vesicles

Tapanendu Kamilya, Prabir Pal, Mrityunjoy Mahato, G.B. Talapatra ⁎Department of Spectroscopy, Indian Association for the Cultivation of Science, Jadavpur, Kolkata-700032, India

⁎ Corresponding author. Tel.: +91 33 24734971; fax:E-mail address: [email protected] (G.B. Talapatra).

0928-4931/$ – see front matter © 2008 Elsevier B.V. Adoi:10.1016/j.msec.2008.12.003

a b s t r a c t

Article history:

Keywords:Protein-lipid vesicleAggregationDrop cast film

ates the immobilization of phospholipids and phospholipids-protein vesicles onsolid substrate by simple drop-cast technique. We have also studied the structural and conformationaltransition of protein (OVA) molecules when they are loaded in immobilized vesicles of DPPC. High-resolutionFE-SEM imaging is used to study the structural aspects of the immobilized vesicles of DPPC and OVA loadedDPPC. FTIR analyses of amide bands are being used to inspect the extent of the conformational transitions ofβ-sheet to α-helix of OVA in immobilized DPPC vesicles. Immobilized OVA-DPPC vesicles provide thestructure of individual OVA molecule attached with DPPC, without aggregation amongst them.

© 2008 Elsevier B.V. All rights reserved.

1. Introduction

Phospholipids vesicles are closed bilayer aggregates that haverecognized a rapidly developing interest in both basic and appliedsciences. Immobilization and stabilization of vesicles onto solidsubstrates have been utilized as models for cell membranes andcomprise compartments as nano/micro-container due to their smalland controllable diameters that can be used to encapsulate, such asproteins, DNA, enzymes, various drug molecules as well as forbiosensing devices. [1–7]. When biomolecules or other chemicalreactants are loaded into the biocompatible container, conformationaltransitions of protein and chemical reaction can be performed inside it[8–10]. Various methods for immobilization and stabilization of vesicleson solid supports have been described in the literature, including theLangmuir–Blodgett (LB), spin coating and the electro-staticallyassembled layer-by-layer (LbL) films by using polymer matrix [11–14].Our main aim is the immobilization/stabilization of lipid, protein-lipidvesicles on solid substrate without using any prefabricated polymermatrix and to study the conformational transitions of proteinmoleculeswhen they are loaded in vesicles. This work is concerned with drop castof vesicular solution on hydrophilic substrate, above the gel transitiontemperature of lipid, to immobilize phospholipids (DPPC)-ovalbumin(OVA) vesicles onto solid substrate. High-resolution FE-SEM imaging isused to study the structural aspects of the immobilized DPPC and OVAmixed DPPC vesicles. FTIR investigations of amide bands are beingutilized to inspect the extent of the conformational transitions of OVA inimmobilizedDPPCvesicles. Results are discussed in the eventof protein-

+91 33 24732805.

ll rights reserved.

lipidmixed vesicles preparation and their immobilization, in addition toaggregation and conformational change of protein in vesicle immobi-lized on solid substrate.

2. Experimental section

2.1. Vesicle preparation and attachment of vesicle

1, 2-dipalmitoyl-sn-glycero-3-phosphocholine (DPPC) and ovalbu-min (OVA) were purchased from Sigma Chemical Co. Ethanol fromMerck and freshly prepared double distilled water, deionized with aMilli-Q water purification system from Millipore (USA) were used invesicle preparation. The pH and resistivity of this water as produced are6.8 and 18.2 MΩ cm respectively. The hydrophilic glass substrates andsilicon wafers were cleaned very carefully, using the process, describedin our earlier article [15].

DPPCvesicleswerepreparedby theethanol injectionmethod [16,17].A solution of 20 mg of DPPC dissolved in 500 μl of ethanol was rapidlyinjected into 20 ml of deionized water at 45 °C i.e. above gel transitiontemperature (41.5 °C) ofDPPC [16,17] by using amicro-liter syringe. Afterinjection, the mixture was kept for 1 h at 45 °C. A drop of this vesicularsolution was dropped on previously cleaned hydrophilic slides by drop-cast method and dried under dry nitrogen atmosphere in a desiccator[18]. To produce vesicles attached with protein, a quantity of 0.1 mg/mlof OVA solution was preheated to 45 °C. 1 ml of this solution was thenadded to 45 °C preheated 3 ml DPPC vesicle solution (DPPC:OVA was30:1 by weight) and kept at 45 °C for 1 h [15]. A drop of this OVA-DPPCmixed vesicular solution above gel transition temperature was droppedon previously cleaned hydrophilic slides and dried under dry nitrogenatmosphere in a desiccator. Finally, the samevesicular solutionwas takenat room temperature (26 °C), below gel transition temperature of

Fig. 1. Panel A represents the FE-SEM image of a bare glass slide. Panels B and C represent the micrometer scale FE-SEM images of DPPC vesicles attached on the glass with varyingmagnifications. In panel B, the long and the short arrows show a large and a small vesicle, respectively. Arrows in panel C indicate the flat plated objects formed by vesicle fusion.Panel D represents the micrometer scale FE-SEM images of OVA-DPPC mixed vesicles attached on a glass slide.

Fig. 2. Schematic diagram of formation of bilayer of DPPC from vesicle on a solidsubstrate. A: spontaneous rupture of larger vesicle to lipid bilayer. B: the fusion andeventual rupture of neighboring smaller vesicles to lipid bilayer.

1481T. Kamilya et al. / Materials Science and Engineering C 29 (2009) 1480–1485

DPPC and a drop of this solution was dropped on previously cleanedhydrophilic slides and dried under dry nitrogen atmosphere in adesiccator.

2.2. Surface morphology measurement

The surface morphology of all the drop cast films was studied byhigh-resolution field emission scanning electronmicroscope (FE-SEM,model no: JEOL JSM-6700 F) with use range: 0.5–30 kV with a lateralresolution in the range of 2.2 to 1.2 nm.

2.3. Measurement of FTIR spectra

FTIR spectra of OVA cast film and OVA loaded vesicles of DPPC filmon silicon wafer substrate were recorded at room temperature on aMagna-IR (Model No 750 spectrometer, series II), Nicolet, USA. In allthe cases, the datawere averaged over 100 scans. The resolution of theinstrument is 4 cm−1.

3. Results and discussion

3.1. Surface morphology of immobilized DPPC and OVA-DPPC vesicles

Vesicles of DPPC were attached on hydrophilic slide by drop castmethod [18] from the vesicular solution of DPPC, kept above the geltransition temperature. Here it is important to note that the drop castmethod, below the gel transition temperature ofDPPCdoes not formanyvesicular structure (will be discuss later). Panel A of Fig. 1 shows the FE-SEM image of a bare glass slide to compare the surfacemorphologywith

other films. The low-resolution FE-SEM images of these dehydratedvesicles with different magnification are presented in panels B and C.Fromthese images, it is evident that there is a heterogeneous populationof vesicles with diameters ranging from nanometer (~100 nm) tomicrometer (~2–2.5 μm)range. Inpanel B, the long and the short arrowsindicate a large and a small vesicle respectively. One can also see frompanel B that the nearly spherical structure of a vesicle, in the absence of

Fig. 3. Panel A represents the FE-SEM image of an OVA-DPPC mixed large vesicle in micrometer scale. This is the high scale magnification image of panel D of Fig. 1. Panels B, C and Drepresent the nanometer scale resolution FE-SEM images of panel Awith differentmagnifications. The inset of panel C shows themonomeric structure of OVA (obtained fromprotein databank). The arrow indicates the single monomer in the FE-SEM image.

Fig. 4. The micrometer and nanometer scale FE-SEM images of the OVA-DPPC membraneattached onglass slides bydrop castmethod (below thegel transition temperature ofDPPC).

1482 T. Kamilya et al. / Materials Science and Engineering C 29 (2009) 1480–1485

water, confirms the self-stabilizing character of the assemblies [18].With a higher magnification, some oblate structured vesicles areobserved as shown in panel C. A similar observation was found in thecase of azobenzine modified dendrimer giant vesicle [18]. Our FE-SEMimages of vesicles are similar with the Cryo-TEM observation of DPPCvesicle reported by Cabaleiro-Lago et al. [19] and Cryo-HRSEMobservation of POPC vesicles reported by Menger et al. [20,21]. Panel Cshows immobilized vesicles in higher magnification. By close observa-tion, one can see flat plated objects (indicated by arrows) together withimmobilized vesicles.

We believe that these flat plated objects on the glass substrate arelipid bilayer formeddue to fusion aswell as rupture of vesicles. There areseveral causes for the formation of the lipid bilayer by vesicle fusion. Theschematic diagram for the formation of lipid bilayer is indicated in Fig. 2.When vesicular solution is dropped on the substrate, their hydrophilichead group attaches the vesicles on the substrate and some adsorbedlarger size vesicles rupture spontaneously, driven by the substrate-induced deformation (Fig. 2A). [22]. The fusion and eventually therupture of neighboring attached vesiclesmay be another probable causeof the formation of the bilayer for smaller sized vesicles (Fig. 2B)[12,22,23]. Moreover, the resulting supported bilayer induces the rup-ture of an adjacent vesicle. Propagation of such a cascaded rupture eventacross several neighboring vesicles leads to the formation of extendedbilayer patches.

Vesicles of OVA-DPPC were attached on hydrophilic slide by dropcast method [18] from the vesicular solution of OVA-DPPC, kept abovethe gel transition temperature. Panel D of Fig. 1 represents the imageof the OVA attached DPPC vesicles in micrometer scale resolution. Theheterogeneous populations of vesicles with diameters from nan-ometer to micrometer were also observed as observed in panel B ofFig. 1. Therefore, the apparent structure of OVA-DPPC and pure DPPC

1483T. Kamilya et al. / Materials Science and Engineering C 29 (2009) 1480–1485

vesicles are almost similar. The diameter of the largest vesicle in thecase of pure DPPC is 2–2.5 μm. Whereas, in the case of OVA attachedDPPC, the diameter of the largest vesicle is almost ~6 μm. Therefore,the average diameters of OVA attached DPPC vesicles are found to belarger than the pure DPPC vesicles, and implies that the incorporationof protein increases the size of vesicles. It is known that, when proteinis injected in lipid vesicles, a hydrophobic interaction between the acylchains of lipid and the hydrophobic moieties of protein takes place. Ahydrophobic mismatch [24–27] between the acyl chains of lipid andthe hydrophobic moieties of protein occurs in such a way that theyorient among themselves to minimize the hydrophobic mismatchbetween them. We witnessed similar observations in the case of OVApenetration into the DPPC monolayer at air water interface [15]. Thishydrophobicmismatch adjustment togetherwith the inclusion of OVAin preformed DPPC vesicle might be responsible for higher vesiculardimension of protein-infested vesicles.

Panel A of Fig. 3 shows the surface morphology of OVA mixed DPPCsingle vesicle immobilized on hydrophilic substrate. The dehydratedvesicle has some how retained its structure with some cracks on it.Panels B to D in Fig. 3 show the surfacemorphology of OVAmixed DPPCvesicle in nanometer scale resolution with a variation in magnification.FE-SEM image shows that the whole surface is covered with whitishnanometer sized granules of almost equal dimensions with that of theOVA monomeric crystal structure (schematic of OVA monomer isshown in the inset of panel C) [28]. FE-SEM analysis shows that thediameters of these granules are in the range of 8–10 nm, more or lessmatched with the dimension of OVA molecule as it has an ellipsoidalshape with dimensions of 7×4.5×5 nm [28–30]. Therefore, thesegranules are expected to be the individual protein molecules. Thus, wefound the individual structure of OVA without larger aggregates ofproteins among themselves in immobilized OVA-DPPC vesicles.

Fig. 5. FTIR spectra of amide I and amide II band. A: OVA cast film. B: OVA-DPPC immobilized vsecond derivative spectrum of amide I band of C. D: multi-peak fitting curves of normalized

We mentioned earlier that the drop cast technique below the geltransition temperature of DPPC does not form any vesicular structure.Panel-A of Fig. 4 shows the FE-SEM image of OVA-DPPC film, producedbelow the gel transition temperature of DPPC. The observed whitishpatches are actually protein-infested membrane of DPPC. The high-resolution image as shown in panel B confirms this observation. Inhigh resolution, we observed a fine texture of whitish small globulesthroughout the film without any aggregation among OVA molecules,similar to the FE-SEM image of OVA-DPPC vesicles in nanometer scale.

3.2. FTIR study

FTIR analysis of amide bands is a useful tool [31] and it is being usedto inspect the extent of the conformational transitions of OVA inimmobilized DPPC vesicles. The unfolding, intra- and intermolecularassociations of protein were studied by monitoring the peak positionsand width of amide bands within a fixed range. To study theconformational change of OVA in vesicular thin film more accurately,we have introduced the FTIR spectrum of OVAdrop cast film. Fig. 5A andB represent the FTIR absorption spectra, in the 1500–1700 cm−1 region;of the film prepared by OVA drop cast and OVA-DPPC vesiclerespectively. In this region two main bands, amide I and amide II ofOVAare observed. The amide I band (1700–1600 cm−1) ismainly due toCfO stretching modes of peptide linkages, while the amide II band(1600–1500 cm−1) is assigned to the coupling of the N–H in-planebending and C–N stretching modes [31]. The vibrational energies of thecarboxyl group depend in reality on the different conformations of theprotein, such as α-helix, β-sheet, β-turns and intra and intermolecularaggregates. The determination and the assignment of the spectralcomponents of the amide I band can then give the information on theprotein secondary structure [32–33]. A Gaussian multiple-peak-fitting

esicle. C: multi-peak fitting curves of normalized amide I band of A. Inset of C shows theamide I band of B. The inset shows the second derivative spectrum of amide I band of D.

Fig. 6. Bar diagram of the percentage area of conformers in OVA cast and OVA-DPPCimmobilized vesicles obtained from fitting results, as indicated in Table 1. The insetshows the ratio of the percentage area of β/α components in OVA cast and OVA-DPPCimmobilized vesicles respectively.

1484 T. Kamilya et al. / Materials Science and Engineering C 29 (2009) 1480–1485

procedure has been employed to the amide I band of FTIR spectra byusing Microcal Origin 7.5 software after baseline correction. The qualityof thefittingwas evaluated based on theχ2 values (on theorder of 10−6)and the square of the correlation coefficient (R2) values 0.999. Themultiple peaks resulting from the deconvolution will provide us theconformations of OVA in different conditions and to identify its com-ponent and, in particular, to determine the corresponding peakfrequencies. The percentage area of the deconvoluted peaks gives therelative area of the components. It is worth noting that, in all the spectraconsidered in the present work, the maximum number N of thecomponents which can be safely identified in the deconvoluted amide Iband does not exceed N=5 to have a meaningful fitting. However, incase of immobilized OVA-DPPC vesicles the best fit is achieved forN=3.The amide I band centered on 1654 cm−1 is the characteristic of the α-helical structure (α-component),whereas thebandat around1638 cm−1

is typical for the β-sheet structure (β-component). The 1618 cm−1 and1683 cm−1 bands are assigned to inter- (A1 component) and intra-molecular aggregates (A2 component) respectively. The 1666 cm−1

component can be ascribed to the vibrationmodes originated by turns inthe β-sheet structures (T-component) [34–37]. We have targeted thedifferent peak values for different components to fit the band. We havealso tried tomake a secondderivative of the amide I band tofindboth theposition and the number of the peaks. Thefitted secondderivative curves(shown in inst of Fig. 5C and D) show that our curve fitting resultsresembles the second derivative spectrum in terms of number of peaksand positions.

The fitting results of normalized amide-I peak of OVA drop castfilm (Fig. 5C) and DPPC-OVA immobilized vesicle (Fig. 5D) arereported in Table 1. The inset of (Fig. 5C) and the inset of (Fig. 5D)show the second derivative of the amide I band. The fitting results andthe second derivative of the amide I band show the same result. Thesummary of the fitting result is indicated in Fig. 6 by bar diagram.

The summary of fitting results implies that the OVA cast film iscomposed of high amount of β-component (43.76%) and of a smallamount of aggregates. Since the β/α ratio (1.16) is greater than unity,some α helix may be converted into β sheet. Small values of A1 (2.00%)and A2 (2.78%) indicate a negligible amount of very large intra andintermolecular aggregates, and confirm the observation of not findinglarger aggregates in FE-SEM images. The figure of OVA cast film is notincorporated into the text as we reported it in our earlier literature [15].It is interesting to observe that the immobilized OVA-DPPC vesicle iscomposed of a high amount ofα-component (61.74%)without any inter-and intra-molecular aggregates (no A1 and A2 components). The β/αratio is 0.42, less than unity implies no conversion ofαhelix intoβ sheet.Zhang et al. [38] reported the increase of α-helix in vesicle medium inthe case of β-lactoglobuline, rich in β-structure. We believe that asimilar phenomenon is operating in this case. Moreover, someelectrostatic [15] and subsequent hydrophobic interactions may leadto destabilization of the β-structure and allow the segment into theregion of acyl chains of DPPC, and promote the formation of helicalsegments [38,39]. The increment of α-helix due to a hydrophobicinteraction in the vesicle matrix preserves the folding of the OVAmolecules and prevents the formation of larger aggregates [38,39].

Table 1Fitting parameters of amide-I band for differently fabricated thin film.

Conformers Area (%) Position (cm−1) FWHM (cm−1)

A B A B A B

A1 2.00 1615.7 11.94β 43.76 26.11 1636.8 1634.0 22.43 24.55α 37.58 61.74 1658.6 1656.3 21.06 30.07T 13.88 12.14 1677.8 1682.3 17.11 20.53A2 2.78 1689.1 10.73

A: cast film of OVA, B: OVA-DPPC immobilized vesicle. Area (%)=100 represents thetotal area under the curve. FWHM is full width at half maximum of a peak.

These arguments are also in support of the observation of whitishgranules (plausibly being the individual protein molecules) in FE-SEMimage.

4. Conclusion

Vesicle immobilization above the gel transition temperature of DPPCshows the self-stabilizing character of the assemblies with some oblatestructured vesicles in the absence of water, with 100 nm to fewmicrometers in diameter. The average diameter of OVA attached DPPCvesicles are larger in diameter than the pure DPPC vesicles due toincorporation of protein that increases the size of vesicles, resulting forthehydrophobicmismatch. Proteinmolecules are attachedwith vesiclesvia allowing their segment into the region of acyl chains of DPPC byelectrostatic and hydrophobic interactions. Protein-lipid mixed vesiclesprovide the individual structure of OVA molecules without aggregationamongst themselves. A conformational transition of the β-sheet toan α-helix of OVA occurs into the protein-lipid mixed vesicles.

The present FTIR study along with the FE-SEM images clearlyprovides us the organization of the protein-lipid mixed vesicularprotein film consisting of individual structure of OVA moleculeswithout aggregation amongst themselves.

Acknowledgements

We thank DST, Government of India (Project No.- SR/S2/CMP-0051/2006) for partial financial support. MM also thanks CSIR,Government of India for providing the CSIR-NET fellowship.

References

[1] L. Zhang, S. Granick, Nano Lett. 6 (2006) 694.[2] V. Vamvakaki, D. Fournier, N.A. Chaniotakis, Biosens. Bioelectron. 21 (2005) 384.[3] C.Oja, P. Tardi,M.P. Schutze-Redelmeier, P.R. Cullis, Biochem. Biophys.1468 (2000) 31.[4] A. Sharma, U.S. Sharma, Int. J. Pharm. 154 (1997) 123.[5] A. Pantos, D. Tsiourvas, C.M. Palcos, G. Nounesis, Langmuir 21 (2005) 6696.[6] D.P. Nikolelis, T. Hianik, U.J. Krull, Electroanalysis 11 (1999) 7.[7] D.A. LaVan, T. McGuire, R. Langer, Nat. Biotechnol. 21 (2003) 1184.[8] M. Yoshimoto, S.Q. Wang, K. Fukunaga, M. Treyer, P. Walde, R. Kuboi, K. Nakao,

Biotechnol. Bioeng. 85 (2004) 222.[9] E. Rhoades, E. Gussakovsky, G. Haran, Proc. Natl. Acad. Sci. U. S. A. 100 (2003) 3197.[10] I.A. Chen, K. Salchi-Ashtiani, J.W. Szostak, J. Am. Chem. Soc. 127 (2005) 13213.[11] L. Wang, M. Schönhoff, H. Möhwald, J. Phys. Chem. B 106 (2002) 9135.[12] K. Ariga, J.P. Hill, M.V. Lee, A. Vinu, R. Charvet, S. Acharya, Sci. Technol. Adv. Mater. 9

(2008) 014109.[13] M.L. Moraes, M.S. Baptista, R. Itri, V. Zucolotto, O.N.O. liveira Jr., Mater. Sci. Eng. C 28

(2008) 467.[14] Y. Ishizuka-Katsura, T. Wazawa, T. Ban, K. Morigaki, S. Aoyama, J. Biosci. Bioeng. 105

(2008) 527.

1485T. Kamilya et al. / Materials Science and Engineering C 29 (2009) 1480–1485

[15] T. Kamilya, P. Pal, G.B. Talapatra, J. Phys. Chem. B 11 (2007) 1199.[16] R.R.C.New(Ed.), Vesicles: APracticalApproach,OxfordUniversity, Oxford,1990, p. 63.[17] L. Stryer, Biochemistry, Freeman, New York, 1998 p 271.[18] G. Dol, K. Tsuda, J. Weener, M. Bartels, T. Asavei, T. Gensch, J. Hofkens, L. Latterini,

Angew. Chem. Int. Ed. 40 (2001) 1710.[19] C. Cabaleiro-Lago, L. Gracia-Rio, P. Herves, J. Perez-Juste, J. Phys. Chem. B 110

(2006) 8524.[20] F.M. Menger, K.L. Caran, P. Apkarian, Langmuir 16 (2000) 98.[21] F.M. Menger, N. Balachande, J. Am. Chem. Soc. 114 (1992) 5862.[22] I. Reviakine, A. Brission, Langmuir 16 (2000) 1806.[23] R.P. Richter, R. Berat, A.R. Brission, Langmuir 22 (2006) 3497.[24] F. Dumas, C.M. Lebrun, J. Tocanne, FEBS Lett. 458 (1999) 271.[25] A. Killan, Biochim. Biophys. Acta 1376 (1998) 401.[26] M.E. Bloom, O.G. Mouritsen, Q. Rev. Biophys. 24 (1991) 293.[27] O.G. Mouritsen, R.L. Biltonen, in: A. Watts (Ed.), Protein–Lipid Interactions, 1993, p.1.

[28] Protein Data Bank (PDB 1OVA).[29] P.E. Stein, A.G. Leslie, J.T. Finch, R.W. Carrell, J. Mol. Biol. 221 (1991) 941.[30] Y. Mine, Trends Food Sci. Technol. 6 (1995) 225.[31] Protein–Ligand Interactions: Structure and Spectroscopy; A Practical Approach

Edited by: S.E. Harding, B.Z. Chowdhry, Oxford University Press.[32] M. Carbonaro, P. Maselli, P. Dore, A. Nucara, Food Chem. 108 (2008) 361.[33] M. Jackson, H.H. Mantsch, Crit. Rev. Biochem. Mol. Biol. 30 (1995) 95.[34] C. Bhattacharjee, S. Saha, A. Biswas,M. Kundu, L. Ghosh, K.P. Das, Protein J. 24 (2005) 27.[35] G.T. Meng, C.Y. Ma, Int. J. Biol. Macromol. 29 (2001) 287.[36] J. Bandekar, Biochem. Biophys. Acta 1120 (1992) 123.[37] K. Murayama, M. Tomida, Biochemistry 43 (2004) 11526.[38] X. Zhang, T.A. Keiderling, Biochemistry 45 (2006) 8444.[39] J. Li, Y. Du, P. Boullanger, L. Jiang, Thin Solid Films 352 (1999) 213.