Two-Dimensional Miscibility Studies of Alamethicin and Selected Film-Forming Molecules

Two-Dimensional Miscibility Studies of Alamethicin and Selected Film-Forming Molecules

Marcin Broniatowski,*,† Nuria Vila-Romeu,‡ and Patrycja Dynarowicz-Łatka†

Faculty of Chemistry, Jagiellonian UniVersity, Ingardena 3, 30-060 Krakow, Poland, and Faculty of Sciences,Department of Physical Chemistry, UniVersity of Vigo, Campus Ourense, As Lagoas s/n, 32004 Ourense, Spain

ReceiVed: January 10, 2008; ReVised Manuscript ReceiVed: April 22, 2008

Alamethicin (ALM), a 20-amino acid antibiotic peptide (peptaibol) from fungal sources, was mixed in Langmuirmonolayers with six different surfactants: semifluorinated (F6H18, F10H19, F8H10OH, F6H10SH) andhydrogenated (C18SH and DODAC), aimed at finding appropriate molecules for ALM incorporation fornanodevice construction. Alamethicin-containing mixed monolayers were investigated by means of surfacemanometry (π-A isotherms) and Brewster angle microscopy (BAM). Our results show that only semifluorinatedalkanes can serve as an appropriate material since they form miscible and homogeneous monolayers withALM within the whole concentration range. All the remaining surfactants, possessing polar groups, werefound to demix with ALM. This effect was explained as being due to the existence of strong polar interactionsbetween vertically oriented surfactant molecules, which tend to separate from horizontally oriented R-helicesof the peptide. On the contrary, semifluorinated alkanes, lacking any polar group in their structure and bearinga large dipole moment, interact with ALM, also possessing a huge cumulative dipole moment. Thesedipole-dipole interactions between ALM and SFAs are more attractive than those between SFA moleculesin their pure monolayers, causing the large ALM molecule, situated parallel to the interface, to be surroundedby SFA molecules in perpendicular orientation, leading to the formation of a highly organized binary mixedmonolayer. BAM images of the ALM monolayer indicate that this peptide collapses with the nucleation andgrowth mechanism, like the majority of surfactants, which contradicts the model of ALM collapse by desorption,previously published in the literature.

Introduction

Many organisms, from bacteria to human beings, produceshort amphiphilic peptides, which exhibit antibiotic, hemolytic,virucidal, and tumoricidal activities by interacting with lipidmembranes of living cells.1 Among them, an important groupof molecules, widely researched for the last four decades, arepeptaibols, i.e. 5-20 amino acid peptides, isolated from fungalsources, containing in their molecules a noncoded amino acid:R-aminoisobutyric acid (Aib)..2–4 Alamethicin (ALM)5,6 was thefirst peptaibol isolated, in 1967, from the culture of fungusTrichiderma Viridae. Because of a plethora of scientific papersdevoted to this peptide, it is treated as an archetypical compoundof the whole peptaibol family.6 Alamethicin was named takinginto account both its antibiotic activity (the ending “cin” of itsname) and the high content of methylalanine (an alternativename of Aib) in its structure, as evidenced in its sequence:

Ac-Aib-Pro-Aib-Ala-Aib-Ala-Gln-Aib-Val-Aib-Gly-Leu-Aib-Pro-Val-Aib-Aib-Glu-Gln-Phol

The fact that both ends of the peptide are blocked led to amisleading conclusion that the ALM molecule is cyclic;5,7

however, later experiments corroborated its linear structure.8

High content of Aib (8 residues for 20 amino acids) isresponsible for the R-helical structure of alamethicin,9 whichis slightly distorted by the Pro 14 residue.10 The crystal structureof ALM was resolved in 1983.11,12 Alamethicin is known toform highly voltage-dependent channels in lipid bilayers;13

however, their selectiveness is low, i.e. they can conduct both

anions and mono- and divalent cations.13 On the basis of thevoltage-gating of the ALM pore, different models of the ALMchannel have been suggested, and the most accepted one is theso-called barrel-stave model.11 According to this approach, theALM R-helices span the lipid bilayer perpendicularly, organizingthemselves in the membrane in a structural motif resemblingstaves of a barrel. The number of the “staves” (i.e., ALMR-helices) ranges from 4 to 12.1,13–15 It was suggested that thenumber of R-helices forming a channel depends on the voltage,which can explain the voltage-gating of the ALM pore.16

The self-assembly of ALM was investigated in solutions,17,18

artificial lipid bilayers,1,19–22 and living cell cultures23 as wellas modeled theoretically with the molecular dynamics.24–26 Inpolar solvents ALM exists in the form of monomers, whereaswith the lowering of the solvent polarity some aggregation takesplace. It was proved that ALM can span bilayer vesicles formednot only by phospholipids but also by vesicles from amphiphilichigh molecular mass polymers, which is of importance in thefield of artificial membrane construction.27

Alamethicin is a fascinating biomolecule because of itsspontaneous self-assembly in hydrophobic media, channelformation, and interesting electrical behavior. This peptide hasan amphipatic character and is capable of stable Langmuirmonolayer formation at the air/water interface, as originallydescribed by Ionov et al. in 2000.28 Although alamethicin hasbeen a subject of investigations for nearly 40 years now, whichresulted in more than 500 publications devoted just to thispeptaibol,6onlyfivearticlesdescribeitsbehaviorinmonolayers.28–32

In all these papers, mixed monolayers were investigated, aimedat finding a matrix suitable for alamethicin’s incorporation.Because of its properties, alamethicin can be applied, forexample, for alternating the permeability of artificial membranes

* Corresponding author. Tel. +48-12-6632082, fax: +48-12-6340515,e-mail: [email protected].

† Jagiellonian University.‡ University of Vigo.

J. Phys. Chem. B 2008, 112, 7762–77707762

10.1021/jp800234k CCC: $40.75 2008 American Chemical SocietyPublished on Web 06/10/2008

and controlling ion fluxes; however, first a suitable artificialmembrane has to be found, in which ALM is distributedhomogenously. Several surfactants have been tested in thisrespect. For example, Ionov et al. investigated mixed monolayersof ALM with selected synthetic phospholipids28 and octadecy-lamine,29 El Abed et al.31 discussed the interactions of ALMwith semifluorinated alkanes, Volinsky et al.30 researched thebehavior of ALM in ternary systems containing phospholipidsand polydiacetylene polymers, and Haefele et al.32 studied theinteractions of ALM with large amphiphilic block copolymers.Unfortunately, none of the above-mentioned compounds wasfound to fulfill requirements for ALM immobilization becauseof the observed immiscibility (or partial immiscibility) and phaseseparation.

In searching for a suitable material to mix with alamethicin,we have chosen six different model surfactants and investigatedtheir miscibility with ALM in Langmuir monolayers. Inparticular, our attention is directed to partially fluorinatedsurfactants. First, these molecules, because of the presence ofthe CF2CH2 bond, bear a large dipole moment and thereforeseem to be promising in stabilizing monolayers frompolypeptides 33,34 as a result of strong dipole-dipole interac-tions. Second, semifluorinated alkanes have already beentested regarding their miscibility with ALM,31 and aninteresting model of vertical phase separation developed,which we think is worth verification. In this paper we extendthe experimental materials and test different moleculespossessing a semifluorinated fragment. Apart from choosingsemifluorinated alkanes (1,1,1,2,2,3,3,4,4,5,5,6,6-tridecaf-luorotetracosane (F6H18) and 1,1,1,2,2,3,3,4,4,5,5,6,6,7,7,-8,8,9,9,10,10-henicosafluorononacosane (F10H19)), we haveselected semifluorinated amphiphiles with different polargroups,namely11,11,12,12,13,13,14,14,15,15,16,16,17,17,18,18,18-heptadecafluorooctadecan-1-ol (F8H10OH) and 11,11,12,12,-13,13,14,14,15,15,16,16,16-tridecafluorohexadecane-1-thiol(F6H10SH). The latter molecule, because of the presence ofan SH group, could serve as a good candidate for preparingorganized films on metal substrates35 and was compared withits hydrogenated analogue (octadecane-1-thiol, C18SH) asit regards the applicability for ALM incorporation. Finally,the film-forming, membrane-mimeting quaternary alkylam-monium salt, i.e. dioctadecyldimethylammonium chloride(DODAC), was used to broaden the studies of the interactionsof ALM with cationic surfactants, which were originated byIonov et al.,29 who studied the mixtures of ALM withoctadecylamine. In contrast to octadecylamine, the chargeof which depends strongly on the aqueous subphase pH,DODAC is an example of a cationic surfactant having thepositive charge localized on the nitrogen atom at a very widerange of pH values. Moreover, some fraction of ALM(Rf30ALM) is negatively charged,12 and therefore one canexpect the interactions with cationic surfactants to befavorable.

In our studies we applied classical methods, i.e., π-A isothermregistration for pure and mixed monolayers. In all the experi-ments the water/air interface was visualized by Brewster anglemicroscopy. The results were interpreted according to thethermodynamic theories of mixed two-dimensional solutions.

Experimental Section

Materials. Alamethicin (>90%) was purchased from Sigmaand used as received. C18OH (>98%) and DODAC (99%) weredelivered by Aldrich, whereas all the fluorinated compounds(F6H10SH, F8H10OH, F6H18, and F10H19) were synthesized

by one of us (M.B.) according to well-known procedures,described in detail in previous papers.36–38 Chloroform andabsolute ethanol (both of spectroscopic grade) were purchasedfrom Aldrich and used as received. Ultrapure Milli-Q water ofthe resistivity higher than 18.2 MΩ was used as a subphase.

Langmuir Experiments. The experiments were carried outon the 601-BAM double barrier NIMA Langmuir trough(NIMA, Coventry, UK). All the investigated film-formingmolecules were dissolved in 9:1 chloroform/ethanol mixture inthe concentration of 0.3-0.5 mg/mL. The stock solutions weremixed in appropriate proportions and dropped with a Hamiltonmicrosyringe on the surface of water. Ten minutes were allowedfor the evaporation of the spreading solvent, after which themonolayers were compressed with the speed of 15 cm2/min,which turned out to be appropriate as the compression speedup to 40 cm2/min was found not to affect the experimentalresults. The surface pressure was monitored continuously by aWilhelmy electronic microbalance with an accuracy of 0.1 mN/m, using filter paper made of Whatman ashless chromatographicpaper as the pressure sensor. The Langmuir trough wasthermostatted with a Julabo circulating water bath with anaccuracy of 0.1 °C. All the experiments were carried out at 20°C. All the π-A isotherms presented here are averages of atleast three independent experiments.

Brewster Angle Microscopy. A Brewster angle microscope,BAM 2 plus (NFT, Germany), was used for microscopicobservation of the monolayer structure. It is equipped with a50 mW laser, emitting p-polarized light of 532 nm wavelength,which is reflected off the air-water interface at approximately53.15° (Brewster angle). The lateral resolution of the microscopewas 2 µm. The images were digitized and processed to obtainthe best quality of the BAM pictures. Each image correspondsto a 770 µm × 570 µm monolayer fragment. All the appliedequipment was placed on an antivibration table.

Results

One-Component Monolayers. Before showing the isothermsfor binary mixtures with ALM let us first present the resultsfor pure Langmuir monolayers of the investigated chemicals.Figure 1 is devoted to pure alamethicin monolayers. Theisotherm agrees well with that reported by other authors 28–30

Although a nonzero value of surface pressure is already observed

Figure 1. Characterization of the pure alamethicin monolayer: π-Aisotherm, compression modulus, π dependence (inset), and representa-tive BAM images. Monolayer conditions at which the BAM imageswere taken are indicated by arrows.

Two-Dimensional Miscibility Studies of Alamethicin J. Phys. Chem. B, Vol. 112, No. 26, 2008 7763

at molecular areas as large as 5 nm2/molecule, a steep rise of πoccurs at ca. 3.8 nm2/molecule. The monolayer collapses at thearea of ca. 3.4 nm2/molecule and at πc ∼ 29 mN/m. Furthercompression does not lead to any surface pressure increase, anda plateau region in the π-A isotherm is observed. The plot ofthe compression modulus (CS

-1 ) -Adπ/dA) vs π is seen inthe inset of Figure 1. A maximum value of CS

-1 ) 250 mN/mis attained for the ALM monolayer, which proves its solidcharacter.39 BAM images were taken simultaneously with filmcompression (Figure 1). Even at high molecular areas, ALMorganizes in large butterfly-like domains, originally reported byVolinsky et al.,30 which are separated from each other (imagesa, b). The reduction of molecular area upon compressiondiminishes the distances between the domains, and at thebeginning of the steep rise of surface pressure they come inclose contact (images d, e). Further on, at molecular areascorresponding to the linear increase of surface pressure themonolayer is homogeneous (image f), whereas at collapse tinybright spots appear, indicating the nucleation of the 3D phasefrom the 2D film of ALM. Further reduction of the area availablefor ALM molecules leads to the increase of the 3D nuclei andto their organization into white stripes (see image h). Our BAMimages taken at the collapse of ALM do not support the modelproposed by El Abed et al.,31 according to which at the collapsepoint, ALM starts to desorb or is squeezed into the bulk water.It seems that the collapse of ALM can rather be correlated withthe nucleation and growth of the 3D phase at the water/airinterface (as it is observed for many other surfactants,40,41 andnot with the peptide desorption from the interface.

Figure 2 compiles the results characterizing pure monolayersof the film-forming molecules investigated in this study. Let usstart with C18SH (Figure 2I). There are only very few papersdescribing the behavior of thiols in Langmuir monolayers, incontrast to a large number of contributions on thiols in self-assembled monolayers. Although the formation of Langmuirmonolayers from C18SH has been reported in some papers 42–45

the monolayers have never been visualized. Upon compressionof the C18SH film, the surface pressure begins to increase atca. 0.25 nm2/molecule, similarly to octadecanol;43 however, the

monolayer collapses already at 15 mN/m. As it can be noticedin image Ib, liquid-expanded domains are formed at slightlyhigher molecular areas than the lift-off area. Some bright whitepoints can be noticed on the gray background of the liquid-expanded domains, indicating that even at zero surface pressuresome fraction of C18SH is incorporated into 3D domains. Atthe beginning of the steep surface pressure rise, the image isuniformly gray, indicating homogeneous coverage of theinterface. However, as visible in image Ic, the gray backgroundis encrusted with white spots, proving that some molecules ofC18SH are incorporated into 3D aggregates.

The isotherm of the investigated semifluorinated thiol,F6H10SH, is shown in Figure 2II. The surface pressure startsto increase at ca. 0.45 nm2/molecule, and the film collapses atca. 0.18 nm2/molecule at a relatively high surface pressure (50mN/m). A characteristic kink is visible in the course of theisotherm at ca. 0.3 nm2/molecule and 15 mN/m. Because thecross-section of the perfluorinated chain is 0.285 nm2, it seemsthat at this particular point of compression, the transition frommonolayer to bilayer takes place. This idea is corroborated withBAM images showing homogeneous Langmuir monolayerbelow the kink, whereas in the vicinity of the kink and at highersurface pressures, large flocks of small bright spots indicate thepresence of 3D objects, probably domains of bilayer coverage.

The π-A isotherm of DODAC is shown together with therepresentative BAM images in Figure 2III. The obtained resultsare in agreement with the literature;46 the monolayer has a liquid-expanded character up to ca. 10 mN/m, and at this particularpressure more condensed domains of snowflake-like shapeappear. They grow upon compression, and at the collapse-pointbrighter dots appear at the edges of the domains, indicating the3D phase formation.

The surface behavior of semifluorinated alkanes (SFAs) andsemifluorinated alcohols was a subject of our previous papers.36,37

Here we present only the isotherms of compounds (Figure 2IV)without BAM images, since the monolayers of these fluorinatedsurfactants were found to be homogeneous and structureless upto their collapse points. SFAs are devoid of any polar headgroup;their monolayers are stabilized mainly by van der Waals

Figure 2. Characterization of the pure monolayers of the surfactants together with representative BAM images: (I) C18SH; (II) F6H10SH; (III)DODAC; (IV) π-A isotherms of the three remaining surfactants: (1) F6H18; (2) F10H19; (3) F8H10OH. Monolayer conditions at which the BAMimages were taken are indicated by arrows.

7764 J. Phys. Chem. B, Vol. 112, No. 26, 2008 Broniatowski et al.

interactions, and because of this fact, the collapse pressures ofSFAs are much lower than those registered for semifluorinatedalcohol monolayers.

Binary Systems. Let us now proceed to the presentation ofthe results registered for mixed binary films of alamethicin andthe film-forming molecules investigated in this study. Insubsequent figures we present isotherms obtained for the binarysurface mixture of the following mole fraction of alamethicin,XALM ) 0.05, 0.1, 0.2, 0.3, 0.5, and 0.7. It seems that the firsttwo mole fractions (0.05 and 0.1) are most interesting, becausesuch systems can be treated as a surfactant matrix, in whichthe peptide is scattered, resembling the situation of ALM innatural lipid bilayers. Therefore, the isotherms for these molefractions were separated from the others, and together with theπ-A isotherm of the pure surfactant (for the purpose ofcomparison) we present them in distinct panels of the followingfigures. The surface manometry results are complemented withappropriate BAM images.

In Figure 3, the isotherms of mixed films of ALM withF8H10OH (Figure 3a,c) and DODAC (Figure 3b,d), i.e. classicalsurfactants, forming stable Langmuir monolayers, are presented.Regarding F8H10OH, the isotherm registered for the molefraction of 0.05 differs profoundly from the other isotherms, asthe collapse pressure (πc) for that particular mole fraction is57.1 mN/m, which is even slightly higher than for pureF8H10OH monolayer (53.4 mN/m). The remaining molefractions, πc, oscillate around 29.5 mN/m, which is a valuecharacteristic of pure ALM monolayers. BAM photos taken formixtures of ALM with F8H10OH are gathered in Figure 4a-h.The monolayers for the mixed system of XALM) 0.05 werehomogeneous up to ca. 20 mN/m; however, above this pressurevalue, 3D domain formation begins, and at 30 mN/m (corre-sponding to the collapse pressure of pure ALM monolayer) amultitude of small nuclei of the 3D phase can be observed.Therefore, it can be inferred that ALM is squeezed out fromthe mixed film at 30 mN/m, even though this fact is not reflectedin the course of the π-A isotherm. For XALM ) 0.1, 3D domainsappear at lower surface pressure as compared to the mixture ofXALM ) 0.05. At a higher proportion of ALM, the butterfly-like domains (typical for pure ALM monolayers) are clearlyseen, indicating the phase separation between ALM andF8H10OH.

Regarding the mixtures of ALM and DODAC, the situationis similar to the discussed above in the aspect that the isothermsfor the low mole fractions of ALM of 0.05 and 0.1 distort alsofrom the remaining isotherms registered for higher ALMcontent. The collapse pressure of pure DODAC is 54.8 mN/m,of the 0.95/0.05 mixture with ALM is 52.4 mN/m, and of the0.9/0.1 mixture is 41.8 mN/m, whereas the remaining mixturesoscillate around 29.5 mN/m. The π-A isotherms measured forthe lowest investigated proportion of ALM have a characteristicplateau region, which is less pronounced for the 0.95/0.05mixture as compared to the 0.9/0.1 mixed film. BAM imagesof the DODAC/ALM films were collected upon film compres-sion. The results for the mixtures 0.95/0.05 and 0.9/0.1 arevirtually identical; therefore, in Figure 4i-l we show only theresults for the former case. Up to a pressure of ca. 7 mN/m, thefilm is homogeneous, whereas at ca. 7 mN/m, the first smallgray domains start to appear, which grow upon film compres-sion, forming snow-flake-like shapes, characteristic of pureLangmuir monolayer from DODAC. At pressures above 29 mN/m, a few bright points can be observed at the edges of thedomains, proving some surface crystallization. Because thispressure is characteristic of ALM collapse, it seems that up tothis pressure value ALM is miscible with DODAC; however,at the pressure of pure ALM collapse, the peptide is expelledfrom the mixture and forms self-3D aggregates. At higher molefractions of ALM, both types of domains (i.e., large butterflyshaped and small snowflake-shaped) are visible in BAM images(data not shown here), suggesting phase separation.

Let us now present the results obtained for the binary systemsof ALM and the investigated thiols. The isotherms of themixtures C18SH/ALM and F6H10SH/ALM are gathered inFigure 5, parts a and b, respectively, whereas the isotherms forthe mole fraction of ALM of 0.05 and 0.01 together with theisotherms of pure thiols are presented in Figure 5, parts c andd. For both mixed systems the collapse pressures do not exceedthe collapse pressure of pure ALM monolayers, even for thelowest ratios of ALM. The isotherm registered for the mixture0.95C18SH/0.05ALM has a characteristic kink at ca. 7 mN/m,which is a pressure twice as low as the collapse pressure ofpure C18SH monolayer. This kink is less visible, althoughnoticeable, in the isotherm of the mixture containing 0.1 molratio of ALM. Interestingly, these kinks are not discernible inthe isotherms of the mixed films of ALM with F6H10SH at themole ratio of ALM of 0.05 and 0.1 (Figure 5b). BAM imagesof the discussed systems are shown in Figure 6. As far as C18SHis concerned, gray circular domains are visible for both 0.95/0.05 and 0.9/0.1 mixtures. They organize upon compression intoa complex texture of chain-like motifs, the brightness of whichrises with the proceeding compression. Finally the chain-likedomains merge, forming large, bright domains. It seems thatthe visible objects, even at low surface pressure, have at leastbilayer organization, and their thickness grows progressivelyupon compression.

The situation is different for mixtures containing semifluori-nated thiol for the lowest investigated mole ratios of ALM.Namely, up to ca. 20 mN/m, the monolayers are homogeneous;however, at higher pressure fast nucleation takes place in themixed film. At the mole ratios of ALM equal or higher than0.2, butterfly-shaped domains of ALM are visible in BAMimages, indicating the phase separation of the surfactantmolecules.

Finally, in Figure 6 and Figure 7, the π-A isotherms formixtures of alamethicin-containing semifluorinated alkanes(F6H18 and F10H19) are presented. Here, the behavior is similar

Figure 3. π-A isotherms for the mixtures of F8H10OH and DODACwith alamethicin: (a) the system F8H10OH/ALM; (b) the systemDODAC/ALM; (c) the system F8H10OH/ALM: (1) pure F8H10OH,(2) XALM ) 0.05, (3) XALM ) 0.1; (d) the system DODAC/ALM: (1)pure DODAC, (2) XALM ) 0.05, (3) XALM ) 0.1.


as already observed for mixtures of ALM with F8H10OH andDODAC, i.e. the collapse pressures of the mixtures containing0.05 and 0.1 mol ratios of alamethicin are higher versus collapsepressures of pure components (alamethicin and SFA). However,in the case of the investigated SFAs this phenomenon is muchmore pronounced, because the collapse pressures of SFAs arelower as compared to ALM, and significantly much lower thanthe collapse pressures of F8H10OH and DODAC monolayers.The representative BAM images for the mixed systems F6H18/ALM and F10H19/ALM are gathered in Figure 8. It isinteresting that the monolayers of these particular mixtures werehomogeneous up to the collapse pressure, where white domainsof the 3D phase appear, evolving upon further compression intolong, white stripes. Moreover, it turned out that for greater ALMmolar fractions (from 0.2 to 0.7), the observed images were

completely different from those observed for pure ALM films.No butterfly-like domains were observed, but instead small grayirregular structures were seen, which (similarly to pure ALMmonolayer) were visible already at zero surface pressure andlarge molecular areas. Here we present the picture registeredfor XALM ) 0.7 for both mixtures with F6H18 and F10H19.Upon compression, small gray domains come closer, but neitherappreciable growth nor changes of the domain shapes wereobserved. At the collapse point, some gray domains mergedand transformed into larger bright islets.

Discussion

To have a deeper insight into the phenomena occurring inmixed Langmuir monolayers of ALM and different investigated

Figure 4. Representative BAM images for (a-d) the system F8H10OH/ALM, XALM ) 0.05; (e-h) the system F8H10OH/ALM, XALM ) 0.1; (i-l)the system DODAC/ALM, XALM ) 0.05.

Figure 5. π-A isotherms for the mixtures of C18SH and F6H10SH with alamethicin: (a) the system C18SH/ALM; (b) the system F6H10SH/ALM;(c) the system C18SH/ALM: (1) pure C18SH, (2) XALM ) 0.05, (3) XALM ) 0.1; (d) the system F6H10SH/ALM: (1) pure F6H10SH, (2) XALM )0.05, (3) XALM ) 0.1.


surfactants, the thermodynamic approach has been applied.Similarly to 3D systems, also for 2D surface film, the excessthermodynamic functions can be defined for two-dimensionalmixtures. One of them is the two- dimensional excess freeenergy of mixing ∆G exc defined as follows:47

∆Gexc )∫0

πAexcdπ (1)

where Aexc denotes the excess area per molecule in a binarymixture, defined as:

Aexc )A12 - (A1x1 +A2x2) (2)where A12 is the molecular area observed at a given π valueupon the compression of the monolayer, whereas x1 and x2 are

mole fractions of surfactant 1 and surfactant 2 in the monolayer.For ideal 2D solutions, Aexc and ∆Gexc equal 0. Such a situationis rare, and usually mixed monolayers do not behave ideallyand ∆Gexc has a nonzero value. Very important is the sign of∆Gexc. The negative sign denotes more attractive interactionsbetween the unlike molecules in a mixed film than between likemolecules in their pure monolayers. On the contrary, a positivesign of ∆Gexc means that the interactions in a mixed filmbetween unlike molecules have a repulsive or at least lessattractive character than between like molecules in pure mono-layers of both components. The sign and value of ∆Gexc aloneis not a decisive parameter regarding the question of miscibility

Figure 6. Representative BAM images for (a-d) the system C18SH/ALM, XALM ) 0.05; (e-h) the system C18SH/ALM, XALM ) 0.1; (i-l) thesystem F6H10SH/ALM, XALM ) 0.05; (m-p) the system F6H10SH/ALM, XALM ) 0.1.

Figure 7. π-A isotherms for the mixtures of F6H18 and F10H19 with alamethicin: (a) the system F6H18/ALM; (b) the system F10H19/ALM; (c)the system F6H18/ALM: (1) pure F6H18, (2) XALM ) 0.05, (3) XALM ) 0.1; (d) the system F10H19/ALM: (1) pure F10H19, (2) XALM ) 0.05, (3)XALM ) 0.1.


between film components, but only together does the analysisof the collapse pressure vs film composition complemented withthe microscopic visualization of the film structure provide thecomplete characteristics of the 2D system. As regards πc, if thepresence of two independent collapses are observed in the courseof an isotherm from two film-forming molecules, correspondingto collapse pressure values for pure components, it is evidenceof their immiscibility in a monolayer.47 On the contrary, whenboth components mix in a monolayer, only one collapse in seenin the isotherm, the value of which varies with mixed filmcomposition.

Plots of the calculated ∆Gexc vs the mole ratio of ALM areshown in Figure 9. At first sight it is clear that ∆Gexc is evidentlynegative only for semifluorinated alkanes in the whole rangeof ALM concentrations, whereas for the remaining chemicalsin this study the value of ∆Gexc is positive. Only for the lowestALM proportion in the mixed films with DODAC is ∆Gexc

slightly negative.

∆Gexc attains the highest positive values for both investigatedthiols. This, together with the fact that πc is invariant with XALM,complemented with BAM images showing multilayer domainsof C18SH and small aggregates for F6H10SH, leads to theconclusion that ALM and both studied thiols do not mix inLangmuir monolayers within the whole concentration range.These results are rather discouraging regarding the idea ofpreparing alamethicin-containing layers on metal surfaces. Theknowledge on the behavior of thiols in Langmuir monolayersis very fragmentary, and practically so far only one compound(C18SH) has been tested 42–45 Recently, our group has starteda program on various amphiphilic thiols synthesis, and we expectto find a better film-forming material as compared to C18SHand F6H10SH.

The third compound studied, which seems to be immisciblewith ALM in binary Langmuir monolayers, is the semifluori-nated alcohol F8H10OH. This compound was found to bemiscible with the antibiotic gramicidin A34 and was promising

Figure 8. Representative BAM images for the systems: (a-d) F6H18/ALM, XALM ) 0.05, (a) below collapse, (b) at collapse; XALM ) 0.1, (c)below collapse, (d) at collapse; (e-h) F10H19/ALM, XALM ) 0.05, (e) below collapse, (f) at collapse; XALM ) 0.1, (g) below collapse, (h) atcollapse; (i,j) F6H18/ALM, XALM ) 0.7, (i) π ) 10 mN/m, (j) π ) 25 mN/m; (k,l) F10H19/ALM, XALM ) 0.7, (k) π ) 10 mN/m, (l) π ) 25 mN/m.

Figure 9. Excess free energy of mixing (∆Gexc) vs mole fraction of alamethicin for the investigated systems: (a) F8H10OH/ALM; (b) DODAC/ALM; (c) F6H18/ALM; (d) F10H19/ALM; (e) C18SH/ALM; (f) F6H10SH/ALM.


as an application in sensor construction and metal electrodefunctionalization, as it is easily transferable on different solidsubstrates. Unfortunately, in mixtures with ALM the results aredifferent. The positive sign of ∆Gexc together with BAM imagesprove that at least in mixtures of XALM g 0.2, F8H10OH andALM are immiscible. However, for the lowest mole ratio ofALM, the π-A isotherm rises smoothly without the appearanceof any kinks, πC is high, and BAM images are homogeneousup to ca. 30 mN/m. Thus, it can be inferred that bothcomponents are miscible, although ∆Gexc is slightly positive.Nevertheless, F8H10OH cannot stabilize ALM in the monolayerat surface pressures higher than πC of pure ALM, as thenucleation of the 3D phase was observed in BAM images undersuch conditions.

In summary, the above-discussed three compounds are notgood candidates for the construction of ALM-containing nan-odevices. On the basis of this background, DODAC seems tobe more suitable, as for the lowest ALM mole ratios it seemsto be miscible with ALM. The ∆Gexc values for such acomposition are slightly negative and BAM images for binaryfilms of such a composition show the presence of snowflake-like domains, characteristic of pure DODAC monolayers. Here,for the first time ALM was mixed with a typical cationicsurfactant possessing a permanent charge on the central nitrogenatom. As it was already mentioned, previously mixed mono-layers of ALM with octadecyl amine were investigated,29

corroborating the immiscibility of ALM with this surfactant inbinary films. Although octadecylamine was called in that papera cationic surfactant, it can acquire the positive charge on itsnitrogen atom only when it is protonated, i.e. in acidic pH.DODAC has an advantage of being used instead of octadecy-lamine, since the experiments can be led on neutral pH, whichis more favorable in peptide research, as the peptide can bedenaturized when spread on a strongly acidic subphase.Moreover, in contrast to octadecylamine, DODAC has two longoctadecyl chains, the presence of which can maximize thehydrophobic interactions between the surfactant and ALM.

In our opinion, the most valuable results have been observedfor mixtures of semifluorinated alkanes and ALM. For bothinvestigated SFAs, ∆Gexc values are negative for all ALM molefractions. Moreover, it was proved with BAM images that evenat a high alamethicin content (XALM ) 0.7), SFAs influence theorganization of the peptide molecule, as it forms small, irregulardomains in the presence of SFA in contrast to large butterfly-like domains, observed either in pure ALM monolayers or inbinary immiscible films with ALM. For low mole fractions ofALM of 0.05 and 0.1, the mixed Langmuir monolayers of SFAs/ALM are homogeneous, even at surface pressures higher thanthe collapse pressure of pure ALM monolayer. It seemstherefore, that the selected SFAs stabilize the ALM moleculesin an artificial semifluorinated matrix, which is perfectlyorganized at least up to the threshold of BAM resolution. Thisconclusion disagrees with the model of mixed ALM/F8H18films suggested by El Abed et al.31 According to these authors,ALM and the investigated SFA are immiscible and separatevertically. This means that in the mixture, first the ALM domainsare compressed up to the collapse, whereas SFA is scattered atthe water/air interface. Since the molecular areas correspondingto the ALM molecules are much larger compared to tinymolecules of SFA, no response from these molecules can bedetected in the course of the π-A isotherm, which resemblesthat registered for pure ALM monolayer. Later on, the film ofALM collapses in such a way that the water/air interface issaturated with the hydrophobic backbones of the ALM-building

amino acids and the ALM molecules are squeezed from theinterface and desorb into the subphase. SFA locates on thehydrophobic surface of the collapsed ALM, and then when themolecular area corresponds to its geometrical dimension, itforms a condensed film on the top of the collapsed ALMmonolayer, which is reflected in the further steep increase insurface pressure. However, our results prove that the mixedLangmuir monolayers of ALM and SFAs can differ from themodel described above. It was unambiguously observed by usthat ALM collapses by the nucleation and growth mechanism,like the majority of known surfactants, and not by desorbinginto bulk water. The evidence for this is the multitude of bright,growing 3D nuclei of ALM appearing at the collapse pressure,clearly visible in BAM images. The desorption model wouldbe probable if the BAM images were homogeneous after thecollapse, at the plateau region of the isotherm; however, asdiscussed above, this is not the case. Therefore, after the collapseof ALM monolayer, the water/air interface starts to have a veryirregular coverage; the growing ALM nuclei, differing probablybetween each other in the number of ALM layers, are inequilibrium with the regions of monomolecular coverage.Therefore, we think that the model of a highly organized layerof SFA on the top of ALM is not legitimized. Moreover, ifSFAs were immiscible with ALM, ∆Gexc values would bepositive as they are in the case of the other investigatedsurfactants. Much more probable is the model of a completelymiscible monolayer, in which long ALM helices lying flat onthe water-air interface are surrounded by the SFA moleculesand separated from each other. Thus, they can not self-assembleand form large condensed domains, as reflected in BAM imagesand much lower compression modulus values than thoseobserved for pure ALM monolayer.

It would be tempting to give some explanation of the observedinteractions. It seems that the key to the elucidation of theobserved trends in the studied binary mixtures is hidden in thestructure of the investigated surfactants. Four of the investigatedfilm-forming molecules contain a polar headgroup in theirmolecules, whereas the SFAs do not possess this structural motif.Such compounds like alcohols or ammonium salts interactstrongly with water molecules as well as with each other becauseof their polarity and because the dipole or the charge is situatedmainly on the polar headgroup. All the investigated surfactants,apart from ALM, take upright orientation in respect to theinterface, whereas ALM molecules were proved to lay flat onthe water surface.29,30 The long ALM R-helix disturbs theinteractions between the molecules of the other surfactants at aconsiderably long distance and it seems that for the purpose ofavoiding such disturbances, surfactant molecules tend to formseparate domains and do not mix with ALM. In the case ofSFAs the situation is different because of the lack of any polarheadgroup in these molecules. Monolayers from SFAs aremainly stabilized by the van der Waals forces,36 especially bypermanent dipole-permanent dipole interactions. The dipolemoment of SFAs originates mainly from the CH2CF2 junctionand has a value of ca. 2.8 D. ALM has the cumulative dipolemoment of ca. 75 D, located on many amino acid residues.30 Itseems probable that the dipole-dipole interactions betweenALM and SFAs are significantly more attractive than thosebetween SFA’s molecules in their pure monolayers, and this isthe reason why the large ALM molecule, situated parallel towardthe interface, is surrounded by SFA molecules in a perpendicularorientation, leading to the formation of a highly organized binarymixed monolayer.


Conclusions

In this paper it was proved that it is possible to find anartificial matrix in which ALM is miscible and distributedhomogenously. It seems that the previously suggested modelof mixed ALM/SFAs films can be incorrect and that no verticalphase separation takes place in such systems. ALM does notmix with classical surfactants, having a polar headgroup and ahydrophobic tail, because of the unfavorable orientation of themolecules at the surface and the presence of relatively stronginteractions between surfactant molecules, leading to a phaseseparation. This problem can be overcome when SFAs are used,and such binary mixtures seem to have potential application inthe construction of nanodevices containing ALM; however, theproper conditions of transferring such mixtures on appropriatesubstrates have to be elaborated, and this will be a subject ofour forthcoming studies.

Acknowledgment. This work was supported by Ministeriode Ciencia y Tecnologıa (Grant CTQ2006-04085) and Xuntade Galicia (Grant PGIDT06PXIB383004PR).

References and Notes

(1) Bechinger, B. J. Membr. Biol. 1997, 156, 197–211.(2) Whitmore, L.; Chugh, J. K.; Snook, C. F.; Wallace, B. A. J. Pept.

Sci. 2003, 9, 663–665.(3) Grigoriev, P. A.; Schlegel, B.; Kronen, M.; Berg, A.; Hartl, A.;

Grafe, U. J. Pept. Sci. 2003, 9, 763–768.(4) Degenkolb, T.; Kirschbaum, J.; Bruckner, H. Chem. BiodiVers.

2007, 4, 1052–1067.(5) Meyer, C. E.; Reusser, F. Experientia 1967, 23, 85–86.(6) Leitgeb, B.; Szekres, A.; Manczinger, L.; Vagvolgyi, C.; Kedics,

L. Chem. BiodiVers. 2007, 4, 1027–1051.(7) Payne, J. W.; Jakes, R.; Hatley, B. S. Biochem. J. 1970, 117, 757–

766.(8) Pandey, R. C.; Cook, J. C.; Rinehart, K. L., Jr. J. Am. Chem. Soc.

1977, 99, 8469–8483.(9) Peggion, C.; Jost, M.; Baldini, C.; Formaggio, F.; Toniolo, C. Chem.

BiodiVers. 2007, 4, 1183–1199.(10) Jacob, J.; Duclohier, H.; Cafiso, D. S. Biophys. J. 1999, 76, 1367–

1376.(11) Fox, R. D.; Richards, F. M. Nature 1982, 300, 325–330.(12) Duclohier, H.; Alder, G.; Kociolek, K.; Leplawy, M. T. J. Pept.

Sci. 2003, 9, 776–783.(13) Duclohier, H.; Wroblewski, H. J. Membr. Biol. 2001, 184, 1–12.(14) Mak, D. O. D.; Webb, W. W. Biophys. J. 1995, 69, 2323–2336.(15) Cantor, R. S. Biophys. J. 2002, 82, 2520–2525.(16) Luchian, T.; Mereuta, L. Bioelectrochemistry 2006, 69, 94–98.(17) Schwarz, G.; Savko, P. Biophys. J. 1982, 39, 211–219.(18) Marsh, D.; Jost, M.; Peggion, C.; Toniolo, C. Chem. BiodiVers.

2007, 4, 1269–1274.

(19) Bezrukov, S. M.; Rand, R. P.; Vodyanoy, I.; Parsegian, V. A.Faraday Discuss. 1998, 111, 173–183.

(20) Dave, C.; Billington, E.; Pan, Y. L.; Straus, S. K. Biophys. J. 2005,89, 2434–2442.

(21) Mottamal, M.; Lazaridis, T. Biophys. Chem 2006, 122, 50–57.(22) Woolley, G. A. Chem. BiodiVers. 2007, 4, 1323–1337.(23) Guihard, G.; Falk, S.; Vachon, V.; Laprade, R.; Schwartz, J. L.

Biochemistry 1999, 38, 6164–6170.(24) Biggin, P. C.; Sansom, M. S. P. Biophys. Chem. 1999, 76, 161–

183.(25) Tieleman, D. P.; Sansom, M. S. P. Berendsen, H. J. C. Biophys. J.

1999, 76, 40–49.(26) Kessel, A.; Cafiso, D. S.; Ben-Tal, N. Biophys. J. 2000, 78, 571–

583.(27) Vijayan, K.; Discher, D. E.; Lal, J.; Janmey, P.; Goulian, M. J.

Phys. Chem. B 2005, 109, 14356–14364.(28) Ionov, R.; El-Abed, A.; Angelova, A.; Goldmann, M.; Peretti, P.

Biophys. J. 2000, 78, 3026–3035.(29) Ionov, R.; El-Abed, A.; Goldmann, M.; Peretti, P. J. Phys. Chem.

B 2004, 108, 8485–8488.(30) Volinsky, R.; Kolusheva, S.; Berman, A. Jelinek, R. Langmuir 2004,

20, 11084–11091.(31) El Abed, A. I.; Ionov, R.; Daoud, M.; Abillon, O. Phys. ReV. E

2004, 70, 051607.(32) Haefele, T.; Kita-Tokarczyk, K.; Meier, W. Langmuir 2006, 22,

1164–1172.(33) Broniatowski, M.; Nieto-Suarez, M.; Vila-Romeu, N.; Dynarowicz-

Łatka, P. J. Phys. Chem. B 2006, 110, 19450–19455.(34) Broniatowski, M.; Obidowicz, K.; Vila-Romeu, N.; Broniatowska,

E.; Dynarowicz-Łatka, P. J. Colloid Interface Sci. 2007, 313, 600–607.(35) Barriet, D.; Lee, T. R. Curr. Opin. Colloid Interface Sci. 2003, 8,

236–242.(36) Broniatowski, M.; Sandez-Macho, I.; Minones, J., Jr.; Dynarowicz-

Łatka, P. J. Phys. Chem. B 2004, 108, 13403–13411.(37) Broniatowski, M.; Dynarowicz-Łatka, P. J. Colloid Interface Sci.

2006, 301, 315–322.(38) Peroche, S.; Parrot-Lopez, H. Tetrahedron Lett. 2003, 44, 241–

245.(39) HarkinsW. D. The Physical Chemistry of Surface Films; Reinhold:

New York, 1952; p 107.(40) Vollhardt, D.; Retter, U. J. Phys. Chem. 1991, 95, 3723–-3727.(41) Vollhardt, D. AdV. Colloid Interface Sci. 2006, 123 (126), 173–

188.(42) Itaya, A.; Van der Auweraer, M.; De Schryver, F. C. Langmuir

1989, 5, 1123–1126.(43) Bilewicz, R.; Majda, M. Langmuir 1991, 7, 2794–2802.(44) Zhao, W.; Kim, M. W.; Wurm, D. B.; Brittain, S. T.; Kim., Y. T.

Langmuir 1996, 12, 386–391.(45) Gupta, R. K.; Suresh, K. A.; Guo, R.; Kumar, S. Anal. Chim. Acta

2006, 568, 109–118.(46) Hc-Wydro, K.; Wydro, P.; Dynarowicz-Łatka, P. J. Colloid

Interface Sci. 2005, 286, 504–510.(47) Dynarowicz-Łatka, P.; Kita, K. AdV. Colloid Interface Sci. 1999,

79, 1–17.

JP800234K


Documents

Two-Dimensional Miscibility Studies of Alamethicin and Selected Film-Forming Molecules