Journal of Colloid and Interface Science 388 (2012) 162–169

Contents lists available at SciVerse ScienceDirect

Journal of Colloid and Interface Science

www.elsevier .com/locate / jc is

Interaction of human serum albumin with monofluorinatedphospholipid monolayers

Paula Toimil a,⇑, Gerardo Prieto a, José Miñones Jr. b, José M. Trillo b, Félix Sarmiento a

a Biophysics and Interfaces Group, Department of Applied Physics, Faculty of Physics, University of Santiago de Compostela, 15782 Santiago de Compostela, Spainb Department of Physical Chemistry, Faculty of Pharmacy, University of Santiago de Compostela, 15782 Santiago de Compostela, Spain

a r t i c l e i n f o a b s t r a c t

Article history:Received 22 June 2012Accepted 18 August 2012Available online 29 August 2012

Keywords:Protein monolayersFluorinated phospholipidsLangmuir monolayersMixed monolayersBAMRelative thickness

0021-9797/$ - see front matter � 2012 Elsevier Inc. Ahttp://dx.doi.org/10.1016/j.jcis.2012.08.035

⇑ Corresponding author.E-mail address: paula.toimil@usc.es (P. Toimil).

In this work the interaction between human serum albumin (HSA) and a monofluorinated phospholipid,1-palmitoyl-2-[16-fluoropalmitoyl-phosphatidylcholine] (F-DPPC), was studied by using Langmuirmonolayer and Brewster angle microscopy (BAM) techniques. Different amounts of F-DPPC were spreadon a previously formed HSA monolayer located at the air/water interface at 25 �C and the mixed mono-layers thus obtained showed the existence of a liquid expanded–liquid condensed (LE–LC) phase transi-tion (at 14 mN/m), attributed to the pure F-DPPC monolayer, coexisting with a second transition (at 22–24 mN/m) corresponding to the protein conformational change from an unfolded state to another in‘‘loops’’ configuration. Relative thickness measurements recorded during the compression of the mixedmonolayers showed the existence of an ‘‘exclusion’’ surface pressure (pexc), above which the protein issqueezed out the interface, but not totally. BAM images reveal that some protein molecules in a packed‘‘loops’’ configuration remain at the interface at surface pressures higher than the ‘‘exclusion’’ surfacepressure. The application of the Defay–Crisp phase rule to the phase diagram of the F-DPPC/HSA systemcan explain the existence of certain regions of surface pressure in which the mixed monolayer compo-nents are miscible, as well as those others that they are immiscible.

� 2012 Elsevier Inc. All rights reserved.

1. Introduction

Liposomes are spherical vesicles surrounded by a membraneconsisting of one or more phospholipid bilayers. Their high bio-compatibility and their special structure, which allows the encap-sulation of substances inside, makes the liposomes can be used inmany fields of science and technology. Thereby, liposomes havebeen used as model of membranes [1,2], as transport systems forcontrolled drug delivery [3,4], in cosmetic and food industry[5,6], in cancer therapy and gene delivery [7,8], etc. Namely, inthe field of pharmaceutical technology, it has been shown thatthe use of liposomes as delivery systems of biologically active com-pounds reduces the toxicity of anti-tumor agents, facilitates the re-lease of the drugs inside the cells and enhances the immuneresponse, being also possible their use as artificial vaccines [9,10].

However, conventional liposomes exhibit a poor stability dur-ing their storage and during the time they remain in the blood-stream, so that in recent years numerous research wereconducted in order to design new types of liposomes exhibiting aminor interaction with other substances and, therefore, a greaterstability [11]. Among the numerous strategies used to improvethe stability of liposomes, and taking into account the benefits of

ll rights reserved.

some characteristic properties of fluorine atom (as its high electro-negativity or its low affinity for the lipidic phases), much effortshas been focused on the design of new types of liposomes incorpo-rating fluorinated components into their structure. Different strat-egies have been used to obtain the so-called fluorinated liposomes.Most of them focus in the preparation of bilayers made up of mod-ified lipids where some hydrogen atoms in their chains have beenreplaced with fluorine atoms [12,13]. Another class of fluorinatedliposomes are those obtained by the combination of standard lipidswith fluorocarbon/hydrocarbon diblocks molecules within the li-pid bilayer [14,15]. As a result, fluorinated amphiphiles show ahigh stability in comparison to hydrogenated analogous [16,17].

Nevertheless, several problems as a low efficiency in the drugencapsulation, probably due to the high permeability to Na+ ions,or their toxicity, limit the therapeutical applications of the fluori-nated liposomes [18,19]. Therefore, it was proposed the use of par-tially fluorinated phospholipids as a new method to improve thestability and the inertness of conventional liposomes, avoidingthe possible damage which can cause the incorporation of fluori-nated elements to the bilayer. In addition, taking into account thatthe DPPC is the majority constituent of pulmonary surfactant, themajor component of cell membranes and the phospholipid oftenused in the preparation of liposomes [20,21], in this work a mono-fluorinated phospholipid derived from DPPC, where an hydrogenatom at the end of one of two hydrocarbon chains has been substi-

Scheme 1. Structure of F-DPPC.

P. Toimil et al. / Journal of Colloid and Interface Science 388 (2012) 162–169 163

tuted by a fluorine atom (F-DPPC) (see Scheme 1), was used tostudy the influence of this fluoride atom on the monolayer behav-ior of this compound, as well as on its interaction with other sub-stances that may affect its stability. Among these, previous papersfrom Seeger and other authors [22,23] have demonstrated thatsome serum proteins such as albumin, hemoglobin or fibrinogencan destabilize the bilayer of the liposomes provoking the leakageof their intern content. More specifically, Holm et al. [24,25]showed the ability of human serum albumin to penetrate lipidmonolayers and reduce the permeability and fluidity of liposomes,Thus, in this work we have studied the interaction between F-DPPCand HSA in mixed monolayers spread at the air–water interface inorder to check if the use of monofluorinated phospholipids in pres-ence of HSA may provide greater stability than the conventionalphospholipids and improve the liposome design for drug delivery.For this purpose, the Langmuir monolayer technique was used be-cause this method offers the possibility of controlling the molecu-lar organization of the membrane components in a two-dimensional structure. In this way it is possible to know the num-ber of protein and phospholipids molecules which are interactingat the interface and, as a result, to deepen the knowledge of theF-DPPC/HSA interaction.

2. Materials and methods

2.1. Materials

Human serum albumin (HSA, lyophilized powder, essentiallyprotease free, 96–99%) with molecular weight: 66.5 kDa and iso-electric point in water at 25 �C: 4.7) was purchased from Sigma.1-palmitoyl-2-[16-fluoropalmitoyl-phosphatidylcholine] (F-DPPC)was supplied by Avanti with purity higher than 99%. Chloroform,used as solvent of this phospholipid, was purchased from Merck(purity 99–99.4% via GC). All materials were used without furtherpurification.

Ultrapure water, used as the subphase, was obtained from aMilli-Ro, Milli-Q reverse osmosis system (Millipore Corp.) contain-ing two carbon- and two ion-exchange columns. Finally, the waterwas purified through a 0.22 lm zetapore filter. The resistivity ofthe purified water was 18 MX cm and the pH was 6.

2.2. Compression isotherms of pure components

Surface pressure (p) versus molecular area (A) isotherms werecarried out in a Langmuir trough NIMA 601(Coventry, UK) with aworking surface area of 500 cm2 placed on an anti vibrational ta-ble. Surface pressure was measured with the accuracy of±0.1 mN/m using a Wilhelmy plate made from chromatographypaper (Whatman Chr1) as the pressure sensor.

Unlike phospholipids, HSA is soluble in water, which causes addi-tional complications for obtaining stable monolayers at the air–water interface. So, in order to prevent its dissolution in the aqueoussubphase, the Trurnit’s method [26] was used to spread this proteinon the air/water interface. Specifically, a 130 lL aliquot of HSA solu-tion of 0.17 mg/mL concentration (to obtain a monolayer formed by2 � 1014 molecules) was dropped from the top of a glass rod (5 mmdiameter, 10 cm height) positioned above the air–water interface in

order to spreading uniformly the solution on the aqueous subphase.The number of spread protein molecules at the interface remainedconstant in all experiments. The protein monolayer was allowed toequilibrate for 3 h to ensure its full extension and then was com-pressed by moving the Teflon barrier. Moreover, in order to checkthe stability and the reproducibility of the HSA monolayers, twocompression–decompression cycles were performed on the mono-layer with an interval of 3 h between them. The obtained results(not shown in this work) showed the existence of two overlappingisotherms (with a difference in the area lower than 1%), evidencingthe reproducibility of the measurements and the no desorption ofprotein molecules into the aqueous subphase.

Stock phospholipid solutions of 0.6 mg/mL were prepared inchloroform and adequate volumes were deposited on the HSAmonolayer with a Gilson Microman micropipette, precise to within0.2 lL. Once spread, the solutions were left for 20 min in order to en-sure complete evaporation of the solvent and afterward the com-pression was initiated with a barrier speed of 15 cm2/min. Thesubphase temperature (25 �C) was controlled by a circulating watersystem from a Haake thermostat, with an accuracy of ±0.1 �C.

2.3. BAM images and relative thickness of monolayers

Brewster angle microscopy (BAM) images and ellipsometric mea-surements were carried out on the monolayer using a NFT BAM 2Plus (Göttingen, Germany) placed on the Nima film balance. TheBAM was equipped with a 30 mW laser emitting p-polarized lightwith a wavelength of 532 nm which was reflected off at the air/waterinterface at the Brewster angle (53.1�). Under such condition, thereflectivity of the beam was almost zero on the pure water interface.The reflected beam passed through a focal lens, into an analyzer at aknown angle of incident polarization, and finally to a CCD camera,which measures gray levels (GLs) instead of relative intensity (I).This light intensity at each point in the BAM image depends on thelocal thickness and film optical properties. At the Brewster angle:I = |Rp|2 = C d2, where I is the relative intensity or relative reflectivity(defined as the ratio of the reflected intensity (Ir) and the incidentintensity (I0), I = Ir/I0), Rp is the p-component of the light, C is a con-stant and d is the relative film thickness. To measure this relativethickness, a camera calibration is previously necessary in order todetermine the relationship between the gray levels and the relativereflectivity. The procedure used for this calibration was described inprevious articles [27–29], but currently with the BAM 2 Plus equip-ment the calibration is automatically performed for each measure-ment at different shutter speeds, providing the relative thicknessvalues of the monolayer as it is compressed. The lateral resolutionof the microscope was 2 lm, the shutter speed used was 1/50 sand the images were digitalized and processed to optimize imagequality. BAM images were obtained during the compression processand recorded along the isotherm.

3. Results

3.1. Pure F-DPPC monolayer

The surface pressure–area curve corresponding to the compres-sion of the F-DPPC monolayer is shown in Fig. 1. At the beginning

164 P. Toimil et al. / Journal of Colloid and Interface Science 388 (2012) 162–169

of compression, the monolayer exhibits a gas–liquid expandedphase transition in which the surface pressure remains constant(p–0 mN/m) as the monolayer is compressed. BAM images in thisregion show a diffuse cloud-like morphology evidencing the exis-tence of ‘‘islands’’ of liquid state co-existing with gas phases. Whenthis phase transition is exceeded and the area occupied by themonolayer reaches the value of 115 Å2/mol (lift-off area), the sur-face pressure progressively rises with the monolayer compressionto a value of 12.7 mN/m. The limiting area corresponding to this re-gion (liquid expanded state) is 110 Å2/mol. Above 12.7 mN/m, theisotherm shows the existence of a second plateau in which the sur-face pressure remains almost constant, with values between 12.7and 14.1 mN/m. In this plateau region the monolayer shows twosurface phases in equilibrium: liquid expanded (LE) and liquidcondensed (LC) phases, which are characteristics of phospholipidmonolayers below the temperature of the gel–liquid crystal transi-tion. BAM images taken at surface pressures near the LE–LC phasetransition show the existence of bright small circular condenseddomains which increase in size, adopting ovoid-like shapes asthe film is compressed along the plateau. A similar behavior wasobserved in other phospholipids for this LE–LC phase transition[30,31].

To determinate the physical state of the investigated film and toget information on the ordering of molecules in the monolayer, thecompressional modulus (C�1

s ) values were calculated according tothe expression:

C�1s ¼ �A

dpdA

� �T

where A denotes the area per molecule at a given surfacepressure p.

From the minimum value of this parameter can be determined(more accurately than from the isotherm) the surface pressure atwhich a monolayer phase transition occurs. Besides, changes incompressional modulus can be correlated with the physical stateof the monolayer [32].

Compressional modulus versus surface pressure values for theF-DPPC monolayer are presented in Fig. 2A. The highest values ofC�1

s (ca. 300 mN/m) evidence that the monolayer is in a liquid-con-densed (LC) state before reaching the collapse, with densely packedmolecules. A characteristic minimum appearing at p = 14 mN/m

20 40 60 80 100 120 140 160 180

0

10

20

30

40

50

60

70

T = 25 ºC

π (m

N/m

)

Area ( Å2/molec)

π = 0 mN/m

π = 17 mN/m

collapse

π = 11 mN/m

Fig. 1. Surface pressure (p)–area (A) isotherm and BAM images corresponding tothe F-DPPC monolayer formed by 2.88 � 1016 molecules spread on aqueoussubphase and compressed to a constant speed of 15 cm2/min (T = 25 �C). Scalebar: 10 lm.

corresponds to the plateau observed in the p–A isotherm of thiscompound. As it can be seen, the surface pressure correspondingto this minimum (LE–LC transition surface pressure) can be deter-mined more accurately than in the corresponding p–A isotherm.

At p < 14 mN/m, the maximum C�1s value obtained (ca. 40 mN/m)

is characteristic of liquid expanded (LE) monolayers. The packing ofmolecules in such films is intermediate between a closely packedin condensed monolayers and those widely separated in gaseousfilms, where the hydrophobic parts of the molecules are in arandom orientation, while their polar groups are anchored in thewater subphase.

The evolution of the relative thickness (Dd) versus surface pres-sure during the compression of the F-DPPC monolayer is presentedin Fig. 2B. Along the LE phase the relative thickness increases line-arly with the surface pressure. When the monolayer reaches theplateau region (about 14 mN/m), the relative thickness experiencesa sudden increase which is attributed to the orientation change ofthe aliphatic chains from a more or less tilted orientation in the LEstate to a more vertical one in the LC state. Finally, after the LE–LCtransition, when the monolayer is in a closed-packed state, the rel-ative thickness of the film remains practically constant, provingthat the orientation of the hydrocarbon chains at the air–waterinterface does not change with the compression of the monolayeruntil the collapse.

3.2. Pure HSA monolayer

Considering that a protein monolayer is well spread in the air/water interface when its limiting area is about 1 m2/mg [33], pre-vious studies were performed in order to determining the condi-tions for obtain a well spread protein monolayer with areproducible behavior. Some results of these studies are shownin Fig. 3, which correspond to the influence on the p–A isothermsof the number of HSA molecules spread at the interface (Fig. 3A),as well as to the effect of the time elapsed between the spreadingof the protein monolayer and the beginning of the compression(Fig. 3B). From these results, consistent with some earlier pub-lished [34], we concluded that a good protein spreading (limitingarea of 1 m2/mg) and reproducible results were achieved when2 � 1014 molecules were deposited at the air–water interface anda waiting time of 3 h was used.

The p–A isotherm showed in Fig. 4A was obtained at 25 �C un-der these conditions and displays a similar behavior to that foundfor other proteins [35]. This curve exhibits a pseudoplateau at sur-face pressures between 20 and 24 mN/m, approximately, attrib-uted to a conformational change of the protein from an unfoldedstate, in which the amino acid residues are located at the interface(trains conformation), to a coiled state in which the polar groups ofthe protein are immersed into the subphase and the hydrophobicregions are oriented to the air (loops conformation), as is illustratedin Scheme 2. BAM images at the beginning of the pseudoplateaushow the existence of very small bright circular domains alignedin rows, which are grouped along the plateau to originate thickbright strips at the end of the same. Due to the change in confor-mation of the protein along the plateau, the monolayer relativethickness increases sharply when it reaches the pressure of22 mN/m (Fig. 4B), i.e., when the protein adopts the loopsconformation.

3.3. F-DPPC/HSA mixed monolayers

3.3.1. p–A isotherms and compressional modulusTo get mixtures of proteins and lipids there is no an appropriate

solvent which allow to obtain a homogeneous solution of bothcomponents and, consequently, obtaining a mixed monolayer byspreading such solution at the air–water interface. For this reason,

0 10 20 30 40 50 600.0

0.5

1.0

1.5

Δd (n

m)

π (mN/m)

B

0 10 20 30 40 50 600

50

100

150

200

250

300

350A

Cs-1

(m

N/m

)

π (mN/m)

Fig. 2. Compressional modulus (C�1s ) (A) and relative thickness (B) versus surface pressure (p) curves of F-DPPC monolayer spread at the air/water interface at 25 �C.

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.80

5

10

15

20

25

30

π (m

N/m

)

Area (m2/mg)

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8

Area (m2/mg)

1x1014

2x1014

5x1014

Number of HSA molecules

A

0

5

10

15

20

25

30

35

10 minutes 30 minutes 1 hour 3 hours

π (m

N/m

)Waiting time

B

Fig. 3. Surface pressure (p)-area isotherms of HSA monolayers obtained by spreading different number of protein molecules (A) and after different waiting times (B).(T = 25 �C).

P. Toimil et al. / Journal of Colloid and Interface Science 388 (2012) 162–169 165

in this work the procedure used for obtaining a F-DPPC/HSA mixedmonolayer was the following: first, a constant number of proteinmolecules (2 � 1014) were spread at the interface and then, on thisHSA monolayer, adequate volumes of F-DPPC solution were depos-ited until reaching the desired lipid/protein ratio. This procedureensures the existence of the same number of protein molecules un-folded at the interface regardless of the amount of lipid present inthe mixed monolayer. However, the problem is that the initial sur-face pressure and the initial area occupied by the mixed monolay-ers varies as the number of lipid molecules is increased, i.e., as XF-

DPPC is increasing. Specifically, as it can be observed in Fig. 5A, inmonolayers with 0.7 and 0.8 lipid mass fractions the initial surfacepressure was not zero, but between 5 and 15 mN/m, respectively,due to the high number of phospholipid molecules required to ob-tain these particular lipid/protein ratios.

At surface pressures below 22–24 mN/m, the isotherms shift tosmaller areas as the amount of F-DPPC in the protein monolayer in-creases. Thus, the progressive addition of F-DDPC into the HSAmonolayer always provokes the same effect: only a film contrac-tion, remaining in isotherms the plateau corresponding to the for-mation of protein loops, as well as that of the LE–LC phasetransition observed in the F-DPPC monolayer. Therefore, thisbehavior proves the coexistence of both components (the proteinand the lipid) at the interface. However, at surface pressures above24 mN/m, it can be observed that the isotherms are near parallel

one to other, collapsing at the same surface pressure (similar tothat of the lipid monolayer), which suggests the squeezing-out ofthe protein molecules from the interface. In these conditions, thearrangement of the phospholipid hydrocarbon chains that remainin the films is practically the same regardless of the mixed filmcomposition.

Compressional modulus (C�1s )–surface pressure (p) plots for F-

DPPC/HSA mixed monolayers exhibit two maxima (Fig. 5B): thefirst, at a surface pressure about 10 mN/m (with values ofC�1

s � 40–50 mN/m), is characteristic of monolayers in liquid – ex-panded state, while the second, at p � 45 mN/m, denotes the exis-tence of films in liquid condensed state (C�1

s = 120–210 mN/m).Between them, two minima (L and P) are observed. They corre-spond to the inflections (plateaus) on the isotherms shown inFig. 5A. The L minimum, at p = 14 mN/m, is attributed to the LE–LC phase transition of F-DPPC monolayer, and the P minimum, atp = 22–24 mN/m, corresponds to the formation and packing ofthe protein in loops conformation. The surface pressure values cor-responding to both minima are practically independent of themixed films composition (Table 1), although in some mixturesone of the two minimum is not observed, due to the prevalenceof one component over the other masks the behavior of the minor-ity component in the monolayer.

In the Fig. 5B, the minima identified as L and P correspond tophospholipid and HSA phase transitions, respectively.

0.0 0.5 1.0 1.5 2.0 2.5

0

10

20

30

π (m

N/m

)

Area (m2/ mg)

A

π = 2.5 mN/m

π = 1.1 mN/m

π = 19.2 mN/m

π = 25.7 mN/m

100 20 30

0

1

2

3

4

5

Δd (n

m)

π (mN/m)

B

Fig. 4. (A) Surface pressure–area isotherm and BAM images of HSA monolayerspread on the air/water interface at 25 �C. Scale bar: 10 lm. (B) Relative thickness–surface pressure curve.

0.0 0.5 1.00

20

40

60A

0.1 0.3 0.5 0.7 0.8

XF-DPPC

π (m

N/m

)

Area (m2/mg)

0 10 20 30 40 50 600

20

40

60

80

100

120

140

160

180

200

220

240B

XF-DPPC

0.1 0.3 0.5 0.7 0.8

L P

Cs-1

(m

N/m

)

π (mN/m)

Fig. 5. Surface pressure–area isotherms (A) and compressional modulus-surfacepressure (B) curves for F-DPPC/HSA mixed monolayers spread at the air/waterinterface at 25 �C.

166 P. Toimil et al. / Journal of Colloid and Interface Science 388 (2012) 162–169

3.3.2. Relative thicknessRelative thickness (Dd) versus surface pressures curves, re-

corded during the compression of F-DPPC/HSA mixed monolayerswith different mass fractions of phospholipid (XF-DPPC = 0.3, 0.5and 0.8), are shown in Fig. 6. Moreover, Dd–p curves correspondingto pure components were included in order to compare the behav-ior of the mixtures with the single components. For brevity, the re-sults of mixtures with composition XF-DPPC 0.1 and 0.7 were notincluded because they do not provided any additional informationin relation to that obtained for the other mixed films studied.

As it can be seen, the relative thickness of mixed monolayers isgreater than that of the pure components, and increases as themonolayer is compressed. This increment occurs in all mixtures

Scheme 2. Views of the possible conformations of HSA molecules in the

and it is lower when the amount of phospholipid in the mixedmonolayer film is increased.

At a surface pressure about 22–24 mN/m takes plate a suddendecrease in the relative thickness of the monolayers, reaching a va-lue which remains practically constant during the remainder of thecompression. The surface pressure values corresponding to theabrupt decrease in the thickness of the monolayer (‘‘exclusion’’surface pressure, pexc) were very similar for all mixtures (Table 1).

Thickness decrease suggest the expulsion of the protein mole-cules outside the monolayer, although it should be noted thatthe ejection is not complete since the thickness values of the mixedmonolayers at p P pexc do not match the pure phospholipid, whatwould occurs if only this component remains in the monolayerafter complete removal of the protein. But this is not the case sinceDd–p curves show that the thickness of the mixed films at p P pexc

is greater than that of the pure phospholipid, decreasing this differ-

monolayer: ‘‘trains’’ conformation (A) and ‘‘loops’’ conformation (B).

Table 1Surface pressure values corresponding to phase transitions of F-DPPC/HSA mixedmonolayers spread on water at 25 �C: (pL) F-DPPC phase transition; (pP) HSA phasetransition; (pexc) ‘‘exclusion’’ surface pressure of the protein; (pcoll) F-DPPC collapse.The error in the determination of the surface pressures was ±0.5 mN/m.

XF-DPPC pL (mN/m) pP (mN/m) pexc (mN/m) pcoll (mN/m)

0 – 24.0 – –0.1 – 23.5 23.2 –0.3 14.1 22.3 23.3 52.40.5 14.3 22.0 23.3 52.50.7 14.3 22.2 22.3 52.50.8 – 23.5 22.3 52.21 14.6 – 51.9

P. Toimil et al. / Journal of Colloid and Interface Science 388 (2012) 162–169 167

ence as increases the content of F-DPPC in the mixed monolayers;thus, when XF-DPPC = 0.8 practically coincide both values. It is worthnoting that in this region numerous noise peaks were observed,evidencing the existence of regions with different reflectivityformed by non-expelled protein molecules and by phospholipidmolecules. Moreover, the number and intensity of these noisepeaks diminishes as the content of phospholipid in the mixed filmsis increased. Since in the procedure used in this work the numberof protein molecules spread at the interface remains constant, andtaking account that at surface pressures higher than 20 mN/m therelative thickness of the protein is greater than of the F-DPPCmonolayer (compare Fig. 2B with 4B), the obtained results suggesta progressive ejection of protein molecules from the interface bythe action of phospholipid molecules. The displacement from the

B

0

1

2

3

4

C

0 10 20 30 40 50 600

1

2

3

4X F-DPPC

A

0.3 0 1

Δd (n

m)

Δd (n

m)

π (mN/m)

0 10 20

π (m

πexc

Fig. 6. Relative thickness versus surface pressure curves for F-DPPC/HSA mixed monol

interface of protein molecules by phospholipids was suggestedand discussed by other authors [36].

3.3.3. Miscibility/immiscibility of film forming componentsWith data shown in table 1, the phase diagram of Fig. 7 was

obtained, which shows the existence of three plateaus where thesurface pressure corresponding to the different phase transitionsof F-DPPC/HSA mixed monolayers remains practically constantregardless the composition of the mixed films. From this diagram,the miscibility (or immiscibility) of film components at theair–water interface can be analyzed [37,38]. Indeed, according tothe Crisp–Defay surface phase rule [39,40], the number of surfacephases (Ps) in equilibrium, under constant temperature and exter-nal pressure, is given as Ps = Cs�F + 1, where Cs is the number ofcomponents at the surface (in our case 2: F-DPPC and HSA) and Fthe degrees of freedom of the mixed system. Along the horizontallines showed in Fig. 7, the surface pressure is maintained at aconstant value irrespective of the mixed film composition. Thus,F = 0 and, therefore, Ps = 3.

The existence of these three surface phases can be explainedaccording to the different orientation of the molecules of bothcomponents at the air–water interface. As described, the first pla-teau of Fig. 7 represents the LE–LC phase transition of the F-DDPCat pL = 14 mN/m. Below the plateau, the F-DDPC molecules, in theliquid expanded state, exhibit a tilted orientation (more or lesshorizontal with respect to the interface) forming a miscible system(M) with the protein molecules in trains conformation, also hori-zontally oriented at the interface. BAM images taken in this region

0

1

2

3

4X F-DPPC

0.5 0 1

πexc

X F-DPPC

0.8 0 1

πexc

Δd (n

m)

0 10 20 30 40 50 60

π (mN/m)

30 40 50 60

N/m)

ayers at 25 �C. The uncertainty in the relative thickness measurements was of 2%.

0.0 0.5 1.0

0

10

20

30

40

50

60

70

πL

πP

P(loops) + L(collapsed)

πcol

P(loops) + L(V)

P(trains) + L(V)

π (

mN

/m)

XF-DPPC

M

Fig. 7. Phase diagram for mixed monolayers of F-DPPC and HSA spread at the air/water interface at 25 �C.

168 P. Toimil et al. / Journal of Colloid and Interface Science 388 (2012) 162–169

for mixtures of different composition show the existence of homo-geneous mixed films (Fig. 8A–C), without phase separation or do-mains formation, implying miscibility of both components.During the LE–LC phase transition the phospholipid molecules un-dergo a change of orientation until a more upright position, caus-ing the appearance of two immiscible phases: one, formed bylipid molecules with more vertical orientation (Lv), and anotherby protein molecules in trains conformation (Ptrains). As a result,along this first plateau in the phase diagram there is three surfacephases in equilibrium: M, Lv and Ptrains. The immiscibility of compo-nents in this region (between 14 and 22 mN/m) was confirmed byBAM images. Indeed, some bright spots (corresponding to the pro-tein) sometimes have appeared in the field of observation of the

Fig. 8. BAM images of F-DPPC/HSA mixed monolayers with different lipid mass fractions(D) XF-DPPC = 0.5 (E) XF-DPPC = 0.7 (F) XF-DPPC = 0.7 (G) XF-DPPC = 0.3 (H) XF-DPPC = 0.5 (I) XF-D

microscope (Fig. 8D), while in others cases characteristic ovoid–like condensed domains of F-DPPC monolayer were observed(Fig. 8E). This suggests the existence at the interface of separated‘‘patches’’ of each component. A picture as in the Fig. 8F, in whichthe presence of separated domains of both components is ob-served, was obtained by chance in our experiences, evidencingthe separation of the components in the monolayer.

The miscibility or immiscibility of the components in mixedmonolayers as a result of their orientation in the air–water inter-face has been widely referenced in the literature. Thereby, misciblemonolayers formed by components with the same orientationhave been reported for mixed systems such as poly(vinylstearate)and single long-chain compounds [41]; cholesterol and myristicacid [42]; dipalmitoylphosphatidylcholine/cholesterol mixtures[43], poly(benzyl-methacrylate) and arachidic acid [44], or, morerecently, for amphotericin B and dipalmitoylphosphatidyl serinemixed monolayers [45]. In the same way, it has also been reportedthat the different orientation of monolayer components causestheir immiscibility at the interface, as occurs in systems formedby poly(methylmethacrylate) (PMMA) and stearic or myristic acid[46]; PMMA and poly(octadecylmethacrylate]) [47] and, betweenothers, 1-glycerol monooleoil and 1-glycerol monoestearoil mix-tures [48].

When the mixed monolayer reaches a surface pressure (pp) about22–24 mN/m, the protein adopts the loops conformation and after,with the compression, is squeezed out from the interface. However,according to the results discussed in the previous section, theejected protein rather does not dissolve in the aqueous subphase,but their polar groups remain in a ‘‘subsurface region’’ and the apolarresidues gather at the interface forming a new surface phaseseparated from the Lv lipid phase. Accordingly, above the secondplateau of Fig. 7, two immiscible phases in equilibrium coexist,which are formed by protein loops molecules (Ploops), partiallycollapsed, separated from F-DPPC vertically oriented molecules(Lv). BAM images corresponding to this region show the presenceof gray thick stripes of collapsed protein isolated by dark homoge-neous areas of phospholipid in liquid condensed state (Fig. 8G–I).

taken at different surface pressures. (A) XF-DPPC = 0.3 (B) XF-DPPC = 0.7 (C) XF-DPPC = 0.5PPC = 0.5. Scale bar: 20 lm.

P. Toimil et al. / Journal of Colloid and Interface Science 388 (2012) 162–169 169

Finally, at approximately 52 mN/m (third plateau), the F-DPPCmolecules are being ejected from the monolayer during its com-pression, forming a new phase (Lcollapsed) separated from HSA mol-ecules, some of which remain still at the interface forming loops.

4. Conclusions

Monofluorinated F-DPPC phospholipid with one fluorine atomin the hydrocarbon chain can be used to improve the stability ofliposomes reducing the toxicity produced by the use of highlyfluorinated phospholipids. On the other hand, given that proteinmolecules modify the stability and permeability of the liposomes,decreasing their efficiency as drug delivery, in this work the stabil-ity of F-DPPC monolayers in presence of HSA was studied by usingthe Langmuir monolayer technique. DPPC/HSA mixed films wereobtained by depositing different amounts of phospholipid mole-cules (those suitable to obtain mixtures of different composition)on the water surface previously covered by a monolayer of proteinwhich was allowed to equilibrate for 3 h to ensure its full exten-sion. This procedure ensures the existence of the same numberof protein molecules unfolded at the interface regardless of theamount of lipid present in the mixed monolayer. The p–A iso-therms and C�1

s � p curves have evidenced the existence of twophase transitions in the monolayers, both independent of themixed films composition: the first at a surface pressure of14 mN/m (corresponding to the LE–LC phase transition of theF-DPPC) and the second at 22–24 mN/m (corresponding to a con-formational change of the HSA from an unfolded state to a morepacked state, forming loops). These results suggest that F-DPPCand HSA are immiscible at the air–water interface.

Relative thickness measurements have showed the existence ofan abrupt decrease in the thickness of the monolayer at a surfacepressure of about 22 mN/m (‘‘exclusion’’ surface pressure). Thisthickness decrease is attributed to the expulsion of the proteinmolecules outside the monolayer, although the ejection is notcomplete since the thickness values of the mixed monolayers atp P pexc do not math the pure phospholipid, what would occursif only this component remains in the monolayer after completeremoval of the protein. Miscibility or immiscibility of film compo-nents at the air–water interface was analyzed by applying theCrisp–Defay phase rule to the phase diagram of the mixed mono-layers. From this analysis, three surface phases in equilibrium wereobtained along the horizontal lines of the phase diagram, and theircomposition was justified according to the different orientation ofthe molecules of both components at the air–water interface. BAMimages confirmed the existence of these surface phases.

Finally, comparing these results with the obtained for DPPC/HSA mixed monolayers [49], can be conclude that the presenceof a fluorine atom in the tail of the phospholipids does not hardlyaffects the surface behavior of the lipid–protein system, so theobtained results have showed that the mono fluorinated phospho-lipids may be useful to improve the design of liposomes used asdrug delivery since they are less toxic than the poly fluorinatedcompounds and more stable than the hydrogenated one.

Acknowledgments

This work was supported by the Spanish ‘‘Ministerio de Eco-nomía y Competitividad’’ (Project MAT2011-26330) and by the‘‘European Regional Development Fund (ERDF)’’and from ‘‘Xuntade Galicia’’ (Project INCITE08PXIB206030PR).

References

[1] S.J. Singer, G.L. Nicolson, Science 175 (1972) 720.[2] P.H. Fishman, J. Moss, R.L. Richards, R.O. Brady, C.R. Alving, Biochemistry 18

(1979) 2562.[3] M.J. Poznansky, R.L. Juliano, Pharmacol. Rev. 36 (1984) 277.[4] N.O. Dhoot, M.A. Wheatley, J. Pharm. Sci. 92 (2003) 679.[5] R.H. Mueller, R.D. Petersen, A. Hornmoss, J. Pardeike, Adv. Drug Deliv. Rev. 59

(2007) 522.[6] B.F. Gibbs, S. Kermasha, I. Alli, C.N. Mulligan, Int. J. Food Sci. Nutr. 50 (1999)

213.[7] D.D. Lasic, J. Controlled Release 48 (1997) 203.[8] T. Serikawa, A. Kikuchi, S. Sugaya, N. Suzuki, H. Kikuchi, K. Tanaka, J. Controlled

Release 113 (2006) 255.[9] D.D. Lasic, Liposomes from Physics to Application, Elsevier, Amsterdam,

1993.[10] D. Christensen, K.S. Korsholm, I. Rosenkrands, T. Lindenstrom, P. Andersen,

E.M. Agger, Exp. Rev. Vaccines 6 (2007) 785.[11] G. Gregoriadis, Liposomes as Drug Carriers: Recent Trends and Progress, Wiley,

Chichester, New York, 1988.[12] P. Vierling, C. Santaella, J. Greiner, J. Fluorine Chem. 107 (2001) 337.[13] S. Szönyi, H.J. Watzke, A. Cambon, Thin Solid Films 284–285 (1996) 769.[14] L. Trevino, F. Frezard, J.P. Rolland, M. Postel, J.G. Riess, Colloids Surf. A. 88

(1994) 223.[15] L. Trevino, F. Frézard, M. Postel, J.G. Riess, J. Liposome Res. 4 (1994) 1017.[16] J.G. Riess, Colloids Surf. A 84 (1994) 33.[17] J.G. Riess, M.P. Krafft, Chem. Phys. Lipids 75 (1995) 1.[18] F. Frezard, C. Santaella, M.J. Montisci, P. Vierling, J.G. Riess, Biochim. Biophys.

Acta 1194 (1994) 61.[19] M.P. Krafft, Adv. Drug Deliv. Rev. 47 (2001) 209.[20] O. Blanco, J. Perez-Gil, Eur. J. Pharmacol. 568 (2007) 1.[21] K. Egbaria, N. Weiner, Adv. Drug Deliv. Rev. 5 (1990) 287.[22] W. Seeger, C. Grube, A. Gunther, R. Schmidt, Eur. Respir. J. 6 (1993) 971.[23] K.M.W. Keough, C.S. Parsons, P.T. Phang, M.G. Tweeddale, Can. J. Physiol.

Pharmacol. 66 (1988) 1166.[24] B.A. Holm, R.H. Notter, J.N. Finkelstein, Chem. Phys. Lipids 38 (1985) 287.[25] B.A. Holm, G. Enhorning, R.H. Notter, Chem. Phys. Lipids 49 (1988) 49.[26] H.J. Trurnit, J. Colloid Sci. 15 (1960) 1.[27] M.N.G. de Mul, J.A. Mann, Langmuir 14 (1998) 2455.[28] J.M. Rodríguez-Patino, C.C. Sanchez, M.R. Rodríguez-Nino, Langmuir 15 (1999)

2484.[29] M. Miñones Conde, O. Conde, J.M. Trillo, J. Miñones Jr., Langmuir 27 (2011)

3424.[30] P. Toimil, G. Prieto, J. Miñones Jr., F. Sarmiento, Phys. Chem. Chem. Phys. 12

(2010) 13323.[31] A. Aroti, E. Leontidis, E. Maltseva, G. Brezesinski, J. Phys. Chem. B. 108 (2004)

15238.[32] J.T. Davies, S.E.K. Rideal, Interfacial Phenomena, second ed., Academic Press,

New York, 1968.[33] F. Macritchie, Adv. Colloid Interface Sci. 25 (1986) 341.[34] J. Sánchez-González, J. Ruiz-Garcı́a, M.J. Gálvez-Ruiz, J. Colloid Interface Sci.

267 (2003) 286.[35] X. Wang, Q. He, S. Zheng, G. Brezesinski, H. Mohwald, J. Li, J. Phys. Chem. B. 108

(2004) 14171.[36] J. Miñones Conde, J. Miñones Trillo, J.M. Rodriguez Patino, J. Phys. Chem. B. 110

(2006) 11582.[37] G.L. Gaines, Insoluble Monolayers at Liquid–Gas Interfaces, Interscience

Publishers, New York, 1966.[38] F.C. Goodrich, in: J.H. Schulman (Ed.), Proceedings of the Second Inter-

national Congress on Surface Activity, vol. I, Butterworth Press, London, 1957,p. p. 85.

[39] D.J. Crisp, Surface Chemistry (Supplement Research), Butterworth, London,1949.

[40] R. Defay, I. Prigogine, A.G. Bellemans, D.H. Everett, Surface Tension andAdsorption, Longmans, London, 1966.

[41] K. Fukuda, T. Kato, S. Machida, Y. Shimizu, J. Colloid Interface Sci. 68(1979) 82.

[42] K. Motomura, T. Terazono, H. Matuo, R. Matuura, J. Colloid Interface Sci. 57(1976) 52.

[43] O. Albrecht, H. Gruler, E. Sackmann, J. Colloid Interface Sci. 79 (1981) 319.[44] K.G.N. Borve, Colloid. Polym. Sci. 270 (1992) 140.[45] J. Miñones, J. Miñones, J.M. Rodriguez-Patino, O. Conde, E. Iribarnegaray, J.

Phys. Chem. B. 107 (2003) 4189.[46] G. Gabrielli, A. Maddii, J. Colloid Interface Sci. 64 (1978) 19.[47] G. Gabrielli, M. Puggelli, P. Baglioni, J. Colloid Interface Sci. 86 (1982) 485.[48] A. Niccolai, P. Baglioni, L. Dei, G. Gabrielli, Colloid. Polym. Sci. 267

(1989) 262.[49] P. Toimil, G. Prieto, J. Miñones Jr., J.M. Trillo, F. Sarmiento, Colloids Surf. B. 92

(2012) 64.