19970
J. Phys. Chem. B 2005, 109, 19970-19979
Influence of the Saturation Chain and Head Group Charge of Phospholipids in the Interaction of Hepatitis G Virus Synthetic Peptides S. Pe´ rez,†,‡ J. Min˜ ones, Jr.,§ M. Espina,‡ M. A. Alsina,‡ I. Haro,† and C. Mestres*,‡ Associated Unit CSIC “Peptides & Proteins: physicochemical studies” Physicochemistry Department, Faculty of Pharmacy, AV. Joan XXIII s.n. 08028 Barcelona, Spain, Department of Peptide & Protein Chemistry, IIQAB-CSIC, Jordi Girona Salgado 18-26, 08034 Barcelona, Spain, and Department of Physical Chemistry, Faculty of Pharmacy, UniVersity of Santiago de Compostela, 15706 Santiago de Compostela, Spain ReceiVed: March 30, 2005; In Final Form: June 27, 2005
Using the Langmuir technique, we have studied the properties at the air/water interface and the interaction of the hepatitis G virus synthetic peptide E1(53-66) and its palmitoyl derivative with membrane phospholipids. These phospholipids had different characteristics referring to the net charge and saturation of the acyl chain. The palmitoyl derivative was more stable at the air/water interface and in the kinetic at constant area measurements showed higher incorporation to the monolayer. The interaction was higher for saturated phospholipids and those with a negative net charge. When the peptides were in the subphase, they produced changes in the miscibility of mixed monolayers composed of DPPC/DPPG or DOPC/DOPG. It can be deduced from the results obtained that electrostatic interactions play a major role, but when the peptide is derivatized with the palmitoyl chain, hydrophobic interactions are added to the former ones. The interaction is also influenced by the saturation of the acyl chain.
Introduction Hepatitis G virus (HGV), also known as GBV-C, was simultaneously identified by Simons1 and Linnen2 in 1996, from the sera with chronic liver disease of unknown cause. This virus belongs to the Flaviviridae family and has a single stranded positive orientated genome of approximately 9400 nucleotides in length.3 It codifies structural (E1 and E2) and nonstructural (NS2, NS3, NS4, and NS5) proteins. Persistent GBV-C/HGV viremia is common, with 0.9-3% of health U. S. blood donors and approximately 20-30% of patients with HCV infection persistently infected with GBVC/HGV.4,5 Following infection, about 80% of people clear their viremia, concomitantly developing an antibody to the GBV-C/ HGV E2 protein.6,7 By now, there is no evidence that this virus is responsible for liver disease, but it has been recently related with HIV regarding the inhibition of progression to AIDS.8 The impact of coinfection with hepatitis B or C virus on the outcome of AIDS patients is associated with an impaired survival, and these patients often die from liver failure instead of AIDS. On the contrary, previous studies have suggested that people coinfected with HIV and GBV-C/HGV have delayed HIV disease, indicating a beneficial influence of GBV-C/HGV on HIV infection. The definitive mechanism of how GBV-C/HGV could inhibit the progression to AIDS is at present unknown. A recent publication by the group of Prof. Stapleton has shown that GBV-C/HGV induces HIV-inhibitory chemokines and reduces expression of the HIV coreceptor CCR5 in vitro.9 Elucidation of the mechanism of the fusion of enveloped viruses such as HIV or GBV-C/HGV to target membranes has attracted considerable attention because of its relative simplicity and * Corresponding author. Phone: +34934024553. Fax: +34934035987. E-mail:
[email protected]. † Associated Unit CSIC “Peptides & Proteins: physicochemical studies” Physicochemistry Department. ‡ Department of Peptide & Protein Chemistry, IIQAB-CSIC. § University of Santiago de Compostela.
potential clinical importance. Apart from the functions of viral binding to target membranes and the activation of viral fusion proteins, usually only one viral protein is responsible for the actual membrane fusion step. However, the nature of the interaction of viral fusion proteins with membranes and the mechanism by which these proteins accelerate the formation of membrane fusion intermediates are poorly understood.10 Concerning the interactions between viral fusion proteins and the lipids of the target membrane, synthetic peptides corresponding to defined regions of the proteins offer a unique approach for studying molecular interactions, mainly due to the higher degree of homogeneity and the better chemical definition of synthetic products compared to the purified or recombinant proteins. Moreover, lipidation can be used pharmacologically to bring specific binding sites of ligands to membranes. This can be accomplished by chemically synthesizing the appropriate lipopeptide. Although there is a large variety of lipid molecules in cell membranes, there are only relatively few lipids that are found covalently attached to peptides or proteins.11 Furthermore, lipopeptides are currently under intensive investigation because they can generate comprehensive immune responses without the use of adjuvants.12 The most widely used approach to get information about the potential immunoresponse of the synthetic peptides is the analysis of their physicochemical properties by using model membranes. The use of a simple model such as monolayers gives important information about how different types of molecules interact with membranes. We can obtain information on how the peptides alter the structure of a film composed of cell membrane phospholipids and if differences occur depending on the characteristics of these phospholipids (net charge, fatty acid saturation). The behavior of the peptides at the air/water interface is also studied through these techniques. Moreover, the use of Brewster angle microscopy (BAM) ensures a higher knowledge
10.1021/jp0516240 CCC: $30.25 © 2005 American Chemical Society Published on Web 10/01/2005
Interaction of Hepatitis G Virus Synthetic Peptides of the characteristics of the peptides at the interface, giving information about the two-dimensional organization of the monolayer (size and shape of domains and heterogeneity).13 In this work, we describe and discuss the physicochemical characterization, at the air/water interface, of the synthetic peptide E1(53-66) (AGLAVRPGKSAAQL) and its lipidated palmitoyl derivative, PalmE1(53-66), belonging to the nonwell-studied structural 1 (E1) GBV-C/HGV protein. The interaction of both peptides with lipid monolayers has been also studied. We have chosen four synthetic phospholipids differing both in the degree of saturation of the acyl chains and in the net charge of the headgroupsdipalmitoyl phosphatidylcholine (DPPC), dioleoyl phosphatidylcholine (DOPC), dipalmitoyl phosphatidylglycerol (DPPG), and dioleoyl phosphatidylglycerol (DOPG)sto obtain information on how the net charge and the saturation of the fatty acids of these phospholipids influence the incorporation of peptides into the films. Materials and Methods Chemicals. N-R-Fluorenylmethoxycarbonylamino acids and Rink amide resin were obtained from Novabiochem (Nottingham, U.K.). Dimethylformamide (DMF), dichloromethane (DCM), and 20% piperidine/DMF were purchased from Fluka. Washing solvents such as acetic acid and diethyl ether were obtained from Merck (Poole, Dorset, U.K.). Trifluoroacetic acid (TFA) was supplied by Fluka (Buchs, Switzerland). NHydroxybenzotriazole (HOBt), N,N′-diisopropylcarbodiimide (DIPCDI), and 2-(1H-benzotriazole-1-yl)-1-3-3-tetramethyluroniumtetrafluoroborate (TBTU) coupling reagents were obtained from Novabiochem. N,N-diisopropylethylamine (DEIA) was obtained from Merck. 1,8-Diazabicyclo[5.4.0]-undec-7-ene was purchased from Aldrich. Palmitic acid was obtained from Sigma. 1,2-Dipalmitoyl-sn-glycero-3-phosphocholine, 1,2-dipalmitoyl-sn-glycero-3-[phospho-rac-(1-glycerol)] (sodium salt), 1,2dioleoyl-sn-glycero-3-phosphocholine, and 1,2-dioleoyl-snglycero-3-[phospho-rac-(1-glycerol)] (sodium salt) were purchased from Avanti Lipids. Throughout this study, the aqueous subphase was phosphate buffered saline (PBS) (pH 7.4, conductivity 15.4 mS/cm, 313.35 mOsm/kg). Peptide Synthesis. The synthesis was carried out on a Rink amide resin (0.65 mequiv/g) by a solid phase methodology as previously described.14 Briefly, a 9-fluoromethylmethoxycarbonyl-tert-butyl (Fmoc/tBu) strategy by means of N,N′-diisopropylcarbodiimide/1-hydroxybenzotriazole (DIPCDI/HOBt) activation was used. For difficult couplings, 2-(1H-benzotriazole1-yl)-1-3-3-tetramethyluroniumtetrafluoroborate (TBTU) and N,N-diisopropylethylamine (DEIA) agents were used. Side protection was effected by the following: 2,2,5,7,8-pentamethylchroman-6-sulfonyl (Pmc) for Arg, tert-butyl (tBu) for Gln and Ser, and t-butoxycarbonyl (Boc) for Lys. Threefold molar excesses of Fmoc amino acids were used throughout the synthesis. The stepwise addition of each residue was assessed by Kaiser’s (ninhydrin) test.15 Deprotection was performed in 20% piperidine/DMF. For difficult deprotections, 1,8-diazabicyclo[5.4.0]-undec-7-ene was used. To cleave the peptide from the resin and to remove the sidechain blocking groups, the resin was treated with TFA solution containing appropriate scavengers such as TFA/H2O/EDT in a ratio of 95:2.5:2.5. The palmitoyl derivative was obtained also working in the solid phase by using palmitic acid and DIPCDI/HOBt coupling reagents in dimethylformamide/dichloromethane, as previously described for other synthetic peptides.16 Briefly, dry peptide resin
J. Phys. Chem. B, Vol. 109, No. 42, 2005 19971 TABLE 1: Analytical Data for Synthetic Peptidesa amino acid analysisb L(2) 1.9, Q(1) 1.1, A(4) 4.0, S(1) 1.1, K(1) 1.17, G(2) 2.0, P(1) 1.0, R(1) 0.60, V(1) 0.6
HPLC (retention time, tR)c
ESI-MS
E1(53-66)
k′ ) 3.6
M+ ) 1338.2
PalmE1(53-66)
k′ ) 5.8
M+ ) 1577.8
a Eluents: A, H2O (0.05% TFA); B, ACN (0.05%). b Theoretical values in parentheses. Ser, Thr, and Met are known to undergo some degradation during 6 M HCl hydrolytic treatment. c A, water (0.05% TFA); B, acetonitrile (0.05% TFA). Flow: 1 mL/min. Detection: 215 and 280 nm. Gradient elution: 95 A to 5% A in 30 min.
was swollen in DMF for 1 h and the solvent decanted off. Palmitic acid was dissolved in a minimum amount of DMF, followed by the addition of DIPCD/HOBt reagents. The mixture was set aside at room temperature and swirled occasionally. The reaction was complete in 3 h, as judged by the Kaiser test. The resin was filtered and washed sequentially with DMF, 2-methyl-2-butanol, acetic acid, 2-methyl-2-butanol, and diethyl ether and dried in a vacuum. The identity of the synthetic peptides was confirmed before by mass spectrometry, amino acid analysis, and HPLC (Table 1). Monolayers. All monolayer experiments were carried out by the Wilhelmy plate method, using a Langmuir balance (KSV LB5000) as it was described. Surface pressure was measured to a precision of (0.05 mN/m. Surface ActiVity Measurements. A mini Teflon trough, cylindrical in shape with a capacity of 70 mL, was used to study the surface activity of peptides and their insertion into lipid monolayers. The subphase was PBS with a pH of 7.4. To determine the surface activity, increasing volumes of concentrated solutions of peptides were injected beneath the surface and pressure increases were recorded. For kinetics at constant area, a lipid solution in chloroform was spread in the surface to achieve the initial surface pressure desired. After stabilization of the monolayer, the peptidic solution was injected in the subphase and pressure increases recorded. Compression Isotherms. Surface pressure area measurements were performed on a rectangular Teflon trough (850 mL) of 81 000 mm2. Lipid monolayers were spread from methanol/ chloroform solutions on a subphase containing pH 7.4 PBS or PBS with the peptides. The organic solvent was allowed to evaporate for 20 min before the film was compressed at a rate of 20 mm/min. In all of the experiments, the subphase temperature was 21 ( 1 °C. BAM Experiments. Brewster angle microscopy images and ellipsometric measurements were performed with a BAM 2 Plus instrument (NFT, Go¨ttingen, Germany) equipped with a 30 mW laser emitting p-polarized light at a 690 nm wavelength which was reflected off the air/water interface at approximately 53.1° (Brewster angle), as described elsewhere.17 The shutter speed used was 1/50. In measuring the relative reflectivity of the film, a camera calibration was necessary. The expected signal at a certain angle is given by the Fresnel equations18 and may be calculated easily. A comparison of this theoretical intensity Ip with the measured gray value G should allow the generation of a calibration curve Ip(G), relating any measured G value to its corresponding Ip value. Provided that the camera and digitizer response is linear, this is a straight line:
I ) |Rp|2 ) Cd2
(1)
where C is a constant, d is the film thickness, and Rp is the
19972 J. Phys. Chem. B, Vol. 109, No. 42, 2005
Pe´rez et al.
Figure 1. Profiles for the E1 protein of the hepatitis G virus by four different scales, with the upper left corresponding to the accessibility scale by Janin, the upper right corresponding to the antigenicity scale by Welling, the lower left a β turn scale by Chou and Fasman, and the lower right corresponding to the hydrophilicity scale by Hopp and Woods. The segment in the region of residues 53-66 was selected according to these predictions.
p-component of the light. The reflected light intensity may be calculated from a single layer optical model as a function of the film thickness (d) and the refractive index (n). This is the concept: calibration
opt. model, n
G 98 Ip 98d The lateral resolution of the microscope was 2 µm, and the images were digitalized and processed to optimize image quality; those shown below correspond to 768 × 572 pixels. To measure the relative film thickness, we followed the procedures described by Rodriguez-Patin˜o et al.19 Results The selection of putative antigenic domains of the GBV-C/ HGV E1 protein was performed by the alignment of several published sequences of virus isolates. The consensus sequences obtained by means of a comparative study of the GBV-C/HGV isolates using the Clustalw program (www.ebi.ac.uk/clustalw) were studied by computerized prediction of antigenicity after analyzing the hydrophilicity and accessibility profiles of the proteins according to Janin,20 Welling,21 and Chou and Fasman.22 These characteristics have been considered as good predictors for defining antigenic sites within proteins. The sequence E1(53-66) showed theoretically suitable antigenicity and accessibility profiles and the probable presence of β turns. Figure 1 illustrates the hydrophilicity and accessibility plots that correspond to the whole structural E1 protein. The location of the selected peptide fragment containing the potential active epitope is indicated. Another method for displaying the distribution of hydrophobicity within the E1 protein was used. The selected method was a Predict Protein algorithm from Swiss-Protein. The profile of the 190 aa (E1) was evaluated by using PHD (profile fed neural network systems from HeiDelberg).23 PHDhtm24 predicts helical
transmembrane regions. The analysis of the E1 protein by using this algorithm showed a transmembrane region relatively next to sequence 53-66, between residues 75-102. The PiMohtm algorithm also confirms the presence of a transmembrane helix at this position. In our work, with the aim of resembling the hydrophobic environment of the sequence E1(53-66), a palmitic acid moiety was introduced at the N-terminus of the parent peptide E1(53-66). Accordingly, the syntheses of E1(53-66) and PalmE1(5366) were performed and accomplished by an Fmoc-based solid phase methodology, as described in the Materials and Methods section. As shown in Table 1, the corresponding selected peptides described in the Materials and Methods section were successfully synthesized and characterized by amino acid analysis, HPLC, and electrospray mass spectrometry. Yields based on peptidyl-resin weight increase were almost quantitative. The purity of the samples after HPLC purification was checked by analytical HPLC, in all cases being higher than 95%. Electrospray mass spectra showed the only presence of the expected molecular peaks. Pure Peptide Monolayers. The pure E1(53-66) peptide sequence and its palmitoyl derivative were spread at the air/ water interphase from chloroform solutions and a compression isotherm was recorded (Figure 2). E1(53-66) shows an isotherm that reaches a maximum pressure of 27 mN/m with a limiting area (A0) of 0.02 nm2/ resid and an elasticity (inset of Figure 2) compatible with those of peptides.25 However, it has to be noted that this is a small area per molecule. This low value is probably due to the high hydrophility of the peptide that is solubilized and ejected to the subphase through compression. Brewster angle microscopy images taken for E1 and PalmE1 peptides along the full monolayer compression at 20 °C on phosphate buffered solution (PBS) at pH 7.4 are presented in Figures 3 and 4. At high molecular areas, the images obtained are very similar in both
Interaction of Hepatitis G Virus Synthetic Peptides
J. Phys. Chem. B, Vol. 109, No. 42, 2005 19973
Figure 2. Surface pressure (π)-area (A) compression isotherms for E1(53-66) and PalmE1(53-66). BAM images of Figures 3 and 4 are taken at the points designated on the isotherm. Inset: Compressional modulus-surface pressure (Cs-1-π) plots for E1(53-66) and PalmE1(53-66) spreads on PBS at pH 7.4.
Figure 3. BAM images corresponding to different stages of compression for E1(53-66) monolayers spread on PBS at pH 7.4 and (A) at 0 mN/m, (A′) at 0 mN/m, (B) at 0 mN/m, (C) at 0 mN/m, (D) at 5 mN/ m, and (E) at 25 mN/m.
Figure 4. BAM images corresponding to different stages of compression for PalmE1(53-66) monolayers spread on PBS at pH 7.4 and (F) at 0 mN/m, (G) at 1 mN/m, (H) at 2.5 mN/m, (I) at 20 mN/m, (I′) at 20 mN/m, and (J) at 37 mN/m.
peptides (images A and F) for E1 and PalmE1, respectively. The images correspond to the presence of a gas phase (dark
areas) coexisting with an expanded phase, forming oval domains. Below 1 mN/m, with the analyzer rotated at 60°, the domains
19974 J. Phys. Chem. B, Vol. 109, No. 42, 2005
Figure 5. Time evolution of surface pressure (π) and film thickness (d) during a compression for a E1(53-66) and PalmE1(53-66) monolayer spreads on PBS at pH 7.4.
change their brightness, as shown in the photograph of the E1 peptide (image A′). This situation is due to the different tilt azimuthal orientations of peptide molecules, which are reflected in different polarizations. The domains having a tilt azimuth of the molecules in the p-plane look bright, whereas those containing molecules with a tilt azimuth orthogonal to the p-plane look dark when the analyzer is in the p-plane. The optical anisotropy is also observed in the PalmE1 peptide at high molecular areas. Images B and C show a more packed expanded phase when the monolayer of the E1 peptide is compressed, denoting a similar behavior to the PalmE1 derivative (images G and H). Upon film compression, the behaviors of the peptide and the lipopeptide derivative are different. In the case of the E1 peptide, when the monolayer is compressed (π ) 5 mN/m), BAM images are completely homogeneous (image D); these homogeneous images correspond to the expanded phase, and only at surface pressures close to the monolayer collapse (π ) 25 mN/m), a homogeneous image sprinkled with bright nuclei of condensation is observed (image E), which denotes the beginning of the monolayer collapse via a nucleation mechanism. When the monolayer of the lipopeptide is compressed, the image obtained is homogeneous with irregular-shaped domains of considerable size (image I). The image corresponds to the presence of irregular condensed domains of palmitic acid, immersed in an expanded phase. These condensed domains have a high intensity and merge as the monolayer is compressed. Along this region (π ) 10-30 mN/m), the coexistence of black and bright zones proves that optical anisotropy still exists at high surface pressure values (image I′). The bright spots observed at π ) 37 mN/m (image J) may be considered to be the evidence of the nucleation of the 3D structures. d-t vs π-t Curves. The time evolution of thickness (d) and surface pressure (π) during the compression of E1 and the derivative PalmE1 peptide monolayer, spread on phosphate buffered solution (PBS) at pH 7.4, is presented in Figure 5. Upon compression, in the transition gas-expanded monolayer region (at π ∼ 20 mN/m), the thickness is practically constant with some reflectivity peaks, due to the pass through the reflectivity detector of a domain with condensed structure. The value of lipopeptide is 2.3 nm as compared to peptide E1, which is 0.6 nm. The thickness is small, but it increases markedly upon film compression, similarly to surface pressure, until its
Pe´rez et al.
Figure 6. Surface activity vs concentration of PalmE1(53-66). Inset: Effect of PalmE1(53-66) concentration on surface pressure. Surface pressure values were determined after 30 min of peptide injection in the subphase.
maximum value is attained for E1 (4.5 nm) and for the derivative (5.5 nm), at maximum surface pressure (25 and 40 mN/m, respectively). The thickness-time curves for the compression show the presence of some noise peaks due to the coexistence of round-shaped domains dispersed in the homogeneous expanded phase. This effect decrease in the PalmE1 due to the condensed effect that provokes the palmitic acid in the peptide. The presence of salts from phosphate buffered saline (PBS) used like substrate provokes the presence of a major number of noise peaks along the compression in both cases. Taking all of the above into consideration, it may be concluded that the expanded character of the E1 peptide monolayer, confirmed with both compressibility modulus values and homogeneous BAM images, is due to the tilted structure tail, which remains inclined also at high surface pressures. In PalmE1, the presence of palmitic acid produces a little condensing effect in the monolayer, causing an increment in the thickness values and presence in the BAM images of irregularshaped condensed domains. In the lipopeptide monolayer appears a continuous transition and the thickness values changes, at the beginning and at the end of transition, from 3 to 5.2 nm, respectively. This transition is not a true, first-order phase transition, due to the fact that the surface pressure does not remain constant. The increase in the surface pressure during the transition may be due to a reorganization of the peptide from 2D to 3D. Adsorption of the Peptides to the Air/Water Interface. The adsorption to the air/water interface of both peptides was measured after injecting increasing volumes of aqueous solutions in the pH 7.4 PBS subphase and recording the pressure increase. E1(53-66) showed very little surface activity; the higher concentration studied (2.7 µM) produced only pressure increases of 0.75 mN/m. These data are in agreement with the hydrophilic profile of the peptide. However, the lipopeptide showed a higher surface activity (Figure 6) with a pressure increase of 31.7 mN/m for a concentration of 0.90 µM. It can be observed in the inset of Figure 6 that between 0.45 and 0.90 µM the surface pressure increases are similar, corresponding to the saturation of the adsorption. As shown in Figure 6, the adsorption of the peptide to the interface is quite fast, reaching the maximum pressure in about 15 min. The number of peptide molecules per square meter (Γ) was calculated by applying Gibbs equation (eq 2), and the area per
Interaction of Hepatitis G Virus Synthetic Peptides
J. Phys. Chem. B, Vol. 109, No. 42, 2005 19975
TABLE 2: Γ Values for the PalmE1(53-66) Peptide concn (µM)
Γ (1018 molecules/m2)
area (nm2/molecule)
0.045 0.091 0.226 0.453 0.906
1.94 2.24 1.88 1.29 0.78
0.51 0.44 0.53 0.76 1.27
molecule at the interface was calculated by applying eq 3
Γ)
1 ∆Π RT ∆ ln c
(2)
1 ΓN
(3)
A)
where R is 8.31 J/(K mol), T is the temperature (294 K), c is the peptide concentration in the subphase, ∆π is the pressure increase, and N is Avogadro’s number. The results obtained are shown in Table 2. Variations in the area per molecule can be observed. These changes in the area values are probably due to changes in the peptide orientation as molecules reach the interface. Insertion in Lipid Monolayers. E1(53-66) penetration in DPPC, DPPG, DOPC, and DOPG monolayers was very low ( DPPC and for DOPG < DOPC. Studying the monolayer exclusion density, the same pattern was observed: DPPG (6.14 nm2/molecule) > DPPC (2.34 nm2/molecule) > DOPG (1.53 nm2/molecule) > DOPC (1.23 nm2/molecule). Therefore, it can be clearly seen that the net charge of the lipid is important and that the unsaturation of the fatty acids of the phospholipids reduces the interaction. Other authors have also found that Amphotericin interacts in a lesser degree with phospholipids containing unsaturated fatty acids.26 It seems that the presence of unsaturated bonds in the hydrophobic tails of the phospholipid reduces the possibility of these molecules to attain a vertical orientation at the air/water interface, making the headgroup less accessible. However, in all cases, the peptide inserts at pressures g30 mN/ m, probably indicating a good insertion in plasmatic membranes.27 Mixed Monolayers. To obtain deeper knowledge of the interactions between E1(53-66) and PalmE1(53-66) with the lipids studied before, we evaluated the effect of the peptides on the miscibility of DPPC/DPPG and DOPC/DOPG insoluble monolayers using compression isotherms. After performing these experiments, it was possible to calculate the excess Gibbs energy (∆GEX M ), corresponding to the interactions between the lipids assayed in the absence and presence of peptides. The calcula-
Figure 7. Penetration kinetics measured at different initial pressures of DPPC monolayers: 5, 10, 20, and 32 mN/m induced by PalmE1(5366).
Figure 8. Pressure increases recorded after the injection of PalmE1(5366) under lipid monolayers, spread at initial surface pressures of 5, 10, 20, and 32 mN/m.
tions were carried by using eq 4.28,29
∆GEX M )
π π π A12 dπ - N1∫πf0A1 dπ - N2∫πf0A2 dπ ∫πf0
(4)
A12 is the mean area per molecule in the mixed film, A1 and A2 are the areas per molecule in the pure films, N1 and N2 are the molar fractions, and π is the surface pressure (millinewtons per meter). Numerical data were calculated from the compression isotherms according to the mathematical method of Simpson. Mixed DPPC/DPPG Monolayers on the PBS Subphase. Figure 9a shows the π-A isotherms of DPPC/DPPG mixtures spread on the pH 7.4 PBS subphase. Practically all mixed monolayers showed area per molecule values higher that those of pure components, with all of them showing phase transitions. The collapse of the mixed monolayers occurs at the same pressure for all of the films, indicating that both components are immiscible at the collapse.30 Figure 10a shows the area per molecule at 20 mN/m as a function of the lipid composition of the film. Mixtures showed a nonideal behavior with positive deviations. ∆GEX M values (Table 3) are >0, suggesting that there is a tendency for the formation of aggregates between molecules of the same kind. However, these values are
