INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSION REFERENCES

Surfactins are surface-active lipopeptides produced by Bacillus subtilis strains, which are cyclic heptapeptides containing a fatty acid chain (Kakinuma et al., 1969). They are attracting more and more attention in basic and applied research due to their high surface activity (Maget-Dana and Ptak, 1992a) and to their important biological properties, including antiviral, antibacterial, and hemolytic activities (Bernheimer and Avigad, 1970; Vollenbroich et al., 1997a,b). The biological activity of surfactin directly relies on its interactions with biomembranes; consequently, understanding the molecular interactions and mixing behavior of surfactin with phospholipids in thin films is of great importance. Interfacial properties measurements and molecular modeling approaches have provided valuable insight into the miscibility and molecular orientation of mixed surfactin/phospholipid monolayers; however, these methods do not provide direct information on the spatial organization of the monolayers.

Lipid monolayers prepared by the Langmuir-Blodgett (LB) technique are valuable models of biomembranes. In particular, transferring lipid films onto solid substrata offers the possibility to apply surface analysis techniques that cannot be used at the air/water interface or in solution. X-ray photoelectron spectroscopy (XPS) provides a direct chemical analysis of solid surfaces, with an analyzed depth of about 5 nm (Ratner and McElroy, 1986). Although XPS has been widely used to study the surface of materials, its application to probe the spatial organization of lipid LB monolayers has been limited (Solletti et al., 1996). During the last decade, atomic force microscopy (AFM) has emerged as a powerful tool for imaging the structure of supported lipid films (Egger et al., 1990; Weisenhorn et al., 1990; Zasadzinski et al., 1991; Hui et al., 1995; Mou et al., 1995; Solletti et al., 1996; Dufrêne et al., 1997). However, the use of AFM to study the organization of mixed surfactin/phospholipid monolayers has not yet been reported.

The aim of this paper is to gain insight into the spatial organization (miscibility, molecular orientation) of mixed surfactin/phosphatidylcholine monolayers. To this end, the morphology and chemical composition of the mixed monolayers transferred on mica are determined by AFM and XPS, respectively, and compared with the interfacial properties of the films at the air/water interface.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSION REFERENCES

Surfactin with a -hydroxy fatty acid chain of 15 carbon atoms (molecular weight, 1036) was used in this study. It was produced and purified as described previously (Razafindralambo et al., 1998). Primary structure and purity of the surfactin-C₁₅ (>95%) were ascertained by analytical RP-HPLC (Chromspher 5 µm C18 column, 1 × 25 cm, Chrompack, Middelburg, The Netherlands), amino acid analysis (Moore and Stein, 1951), and electrospray mass spectrometry (Finnigan MAT 900 ST) measurements.

LB monolayers were prepared at 20°C with an automated LB system (LFW2 3"5-Lauda, Königshofen, Germany). Surfactin and dipalmitoyl phosphatidylcholine (DPPC) purchased from Sigma Chemical Co. (St. Louis, MO) were dissolved at 1 mM in chloroform/methanol (2:1). Pure solutions and (0.1:0.9), (0.25:0.75), and (0.5:0.5) molar mixtures of surfactin and DPPC were spread on a milliQ water (Millipore Co., Milford, MA) subphase adjusted at pH 2.0 with HCl. After evaporation of the solvent, monolayers were compressed at a rate of 150 cm²/min. They were deposited at a constant surface pressure of 20 mN/m, i.e., well below the collapse pressure, by raising vertically freshly cleaved mica through the air-water interface at a rate of 10 mm/min. The transfer ratios were all close to 1:1. For determining the compression isotherm curves, films were compressed at a rate of 61.8 cm²/min. The difference between molecular areas of two independent sets of measurements was less than 2.5%.

AFM measurements were performed at room temperature (20°C) using a commercial optical lever microscope (Nanoscope III, Digital Instruments, Santa Barbara, CA). Contact mode topographic and friction images were recorded using oxide-sharpened microfabricated Si₃N₄ cantilevers (Park Scientific Instruments, Mountain View, CA) with typical radius of curvature of 20 nm and spring constants ranging from 0.01 N/m to 0.1 N/m. The imaging force was kept as low as possible (~1 nN) and the scan rate was 2 Hz.

XPS analyses were performed with an SSI X-Probe (SSX-100/206) photoelectron spectrometer from Fisons, interfaced with a Hewlett Packard 9000/310 computer allowing instrument control, data accumulation, and data treatment. The pressure during analysis was between 2.5 × 10⁶ Pa and 2.5 × 10⁷ Pa. The spectrometer used monochromatized Al K x-ray radiation (1486.6 eV). The irradiated zone was an elliptic spot, with a shorter axis of 1000 µm. The constant pass energy in the hemispherical analyzer was 150 eV or 50 eV. In these conditions, the resolution determined by the full width at half maximum of the Au_4f7/2 peak of a standard gold sample was about 1.5 and 1.0 eV, respectively. The flood gun energy was set to 8 eV and a nickel grid was placed 3 mm above the surface.

The following sequences of spectra were recorded: survey spectrum, C_1s, K_2p, O_1s, P_2p, Si_2p, Al_2p, N_1s, and finally C_1s again to check for the absence of sample degradation. Binding energies were determined by reference to the C_1s component due to carbon bound only to carbon and hydrogen, set at 284.8 eV. The background was subtracted linearly. Data treatment was performed with the ESCA 8.3 D software provided by the spectrometer manufacturer. Atom concentration ratios were calculated using the peak areas normalized on basis of acquisition parameters and of sensitivity factors proposed by the manufacturer (mean free path varying according to the 0.7th power of the photoelectron kinetic energy; Scofield cross-sections (Scofield, 1976); constant transmission function).

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSION REFERENCES

Fig. 1 shows the surface pressure-area (-A) isotherms, at the air-water interface, of pure surfactin and DPPC monolayers, and of mixed surfactin/DPPC monolayers at 0.1, 0.25, and 0.5 surfactin molar ratios. At 20 mN/m, DPPC and surfactin occupy 46 and 142 Å²/molecule, respectively. The mean area for mixed monolayers is always between those of pure monolayers.

View larger version (26K): [in this window] [in a new window]   FIGURE 1   Surface pressure-area (-A) isotherms, at the air-water interface, of pure surfactin and DPPC monolayers, and of mixed surfactin/DPPC monolayers at 0.1, 0.25, and 0.5 surfactin molar ratios recorded at 20°C with a water subphase at pH 2.0.

Fig. 2 shows AFM topographic and friction images of mixed surfactin/DPPC monolayers transferred on mica for 0.1, 0.25, and 0.5 surfactin molar ratios. All monolayers show phase separation, the shape and the size of the domains varying with the molar composition. For 0.1, 0.25 and 0.5 surfactin molar ratios, the area fraction covered by the lower (darker) domains in the topographic images is 26 ± 4%, 52 ± 6%, and 74 ± 3%, while the step height measured between the lower and higher domains is 1.2 ± 0.1 nm for the three molar ratios (mean values and standard deviations of height differences measured from cross-sections taken from three different images). The topographic and friction contrasts always show a negative correlation, higher friction being associated with the lower domains.

View larger version (141K): [in this window] [in a new window]   FIGURE 2   AFM topographic (A, C, E) and friction (B, D, F) images (2 × 2 µm) of mixed surfactin/DPPC monolayers at 0.1 (A, B), 0.25 (C, D), and 0.5 (E, F) surfactin molar ratios. The root-mean-square surface roughness of the two phases in the topographic images was ~0.1 nm (over 100 × 100 nm areas).

Fig. 3 presents the surface elemental composition determined by XPS for mixed surfactin/DPPC monolayers at 0.0, 0.1, 0.25, 0.5, and 1.0 surfactin molar ratios. Continuous changes of the surface composition occur with increasing surfactin molar ratios: the atom fractions of O_1s, N_1s, Si_2p, Al_2p, and K_2p increase concomitantly with a decrease in the C_1s and P_2p atom fractions. The concentrations of Si_2p, Al_2p, and K_2p are close to the 3:3:1 ratio expected for Muscovite mica.

View larger version (34K): [in this window] [in a new window]   FIGURE 3   Surface chemical composition determined by XPS for pure surfactin and DPPC monolayers on mica and mixed surfactin/DPPC monolayers at 0.1, 0.25, and 0.5 surfactin molar ratios on mica. Atom fractions (%) of C1s and O1s (A), N1s and P2p (B), and Si2p, Al2p, and K2p (C). Bars, difference between two independent sets of data.

Representative N_1s spectra for the pure and mixed monolayers, shown in Fig. 4, indicate that nitrogen is involved in different chemical functions depending on the surfactin molar ratio. For the pure surfactin monolayer, nitrogen appears at about 400 eV, which is attributable to unprotonated amine or amide functions (Gerin et al., 1995). In contrast, for the pure DPPC monolayer, nitrogen is observed at about 402 eV, which is typical of protonated amine (Gerin et al., 1995). The nitrogen peaks of mixed monolayers clearly show two components, attributed to unprotonated amine or amide functions (400 eV) and to protonated amine (402 eV), the contribution of the first component increasing with increasing surfactin molar ratios.

View larger version (21K): [in this window] [in a new window]   FIGURE 4   Representative N1s spectra of mixed surfactin/DPPC monolayers on mica for surfactin molar ratios (top to bottom) of 0.0, 0.1, 0.25, 0.5, and 1.0.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSION REFERENCES

Interfacial properties at the air/water interface

The -A isotherms of the pure surfactin and DPPC monolayers at the air-water interface (Fig. 1) are in agreement with those reported in previous studies (Marra, 1985; Maget-Dana and Ptak, 1992a,b). Their shape indicates that, at 20 mN/m, the DPPC monolayer is characterized by a two-dimensional solid-like organization, while surfactin has a two-dimensional liquid-like organization. The area of DPPC at 20 mN/m (46 Å²) reflects a vertical, or slightly tilted, orientation of the lipid molecules (Marra and Israelachvili, 1985). In contrast, the large area of surfactin (142 Å²) corresponds to an orientation in which the peptide ring is lying horizontally (Gallet et al., 1999). At higher surface pressure, the surfactin isotherm shows a horizontal plateau. Then, a sharp increase of surface pressure is observed at very low areas per molecule, corresponding to a condensed state in which the peptide cycles are probably perpendicular to the interface (Maget-Dana and Ptak, 1992a).

To gain insight into the miscibility of surfactin and DPPC in mixed monolayers, the mean molecular area observed for the mixed films at 20 mN/m may be plotted against the molar fraction of surfactin (Gaines, 1966; Maget-Dana et al., 1989; Maget-Dana and Ptak, 1992b). The plot, presented in Fig. 5, shows small positive deviations from additivity for the mean molecular area, which suggests incomplete miscibility of the two compounds with the formation of bidimensional domains (Maget-Dana et al., 1989).

View larger version (15K): [in this window] [in a new window]   FIGURE 5   Mean molecular area of mixed surfactin/DPPC monolayers as a function of the surfactin molar ratio. The line represents the additivity rule values.

The interactions between surfactin and DPPC in mixed films may be further assessed by calculating the excess free energy of mixing G_m^ex using the Goodrich relationship (Gaines, 1966; Maget-Dana and Ptak, 1995; Razafindralambo et al., 1997):

(1)

where A is the mean molecular area, x is the molar fraction, subscripts 1, 2, and 12 refer to pure compounds 1 and 2, and to their mixtures, respectively. G_m^ex values of 2.7, 3.1, and 3.4 are obtained for 0.1, 0.25, and 0.5 surfactin molar ratios, respectively. Thus, a positive excess of free energy of mixing is found for the three surfactin/DPPC monolayer systems, indicating that the interactions between lipopeptide and lipid molecules are weaker than the interactions between the pure compounds themselves. Accordingly, both the deviations from the additivity rule and the excess free energy of mixing suggest that surfactin and DPPC molecules have a tendency to form bidimensional domains in the mixed monolayers.

Surface morphology of mixed monolayers

AFM topographic images reveal phase separation for the three mixed surfactin/DPPC monolayers (Fig. 2). The area fraction occupied by the two domains may be compared with that expected for surfactin and DPPC in the mixed monolayer at the air-water interface (Fig. 1). Considering the surfactin molar ratio (X_s) and the areas occupied by the surfactin and DPPC molecules in pure monolayers (A_s and A_d), the fraction of the surface occupied by surfactin in mixed monolayers at the air-water interface () may be calculated as follows:

(2)

At 20 mN/m, i.e., for A_s = 142 Å² and A_d = 46 Å²,  is found to be 0.26, 0.51, and 0.76, for X_s = 0.1, 0.25, and 0.5, respectively. The area fractions occupied by the lower domains in the AFM topographic images (26, 52, and 74, for X_s = 0.1, 0.25, and 0.5, respectively) are in excellent agreement with the  values calculated by Eq. 2. This indicates that: 1) the lower and higher levels in the AFM topographic images can be assigned to surfactin and DPPC, respectively; 2) surfactin and DPPC are completely immiscible in the conditions investigated here; 3) the molecular packing of the two compounds within the mixed monolayers is not significantly affected by transferring the films from the air/water interface onto mica.

The observation of phase separation for mixed surfactin/DPPC monolayers may be of biological importance. Previous studies have shown that the interactions of surfactin with biological membranes determine its antibacterial and antiviral action and involve insertion into the lipid bilayers, permeability changes probably resulting from ion channel formation, and membrane disruption at high surfactin concentrations (Sheppard et al., 1991; Thimon et al., 1993; Maget-Dana and Ptak, 1995; Vollenbroich et al., 1997a,b). The tendency of surfactin to self-associate and form bidimensional aggregates was proposed to be involved in channel formation. Hence, the small bidimensional domains of surfactin observed here at low concentration (X_s = 0.1) may play an important role in determining the surfactin biological activity.

The step height measured between the surfactin and DPPC domains in the topographic images may be related to the orientation assumed by the two compounds within the films. The thickness of the DPPC and surfactin monolayers (t) may be estimated from the molecular mass (M_w) and specific mass (d) of the compounds, the Avogadro constant (N_av) and the area at the air-water interface at the deposition pressure of 20 mN/m (A):

(3)

Considering d = 1 g/cm³ gives a monolayer thickness of 2.6 nm and 1.2 nm for DPPC and surfactin, respectively. The thickness deduced for DPPC is in good agreement with that reported from refractive index measurements (Marra and Israelachvili, 1985). The step height measured by AFM, 1.2 ± 0.1 nm, is close to the difference between the computed thicknesses, 1.4 nm. Since the two compounds have fairly similar molecular lengths, the large difference in film thicknesses observed by AFM provides direct evidence for a difference in molecular configuration; while DPPC assumes a vertical (or slightly tilted) orientation (Marra and Israelachvili, 1985), the large peptide ring of surfactin is lying on the mica surface. This is consistent with a recent molecular modeling study which showed that the more stable structure for surfactin at a hydrophobic/hydrophilic interface corresponds to the peptide ring positioned in the plane of the interface with the fatty acid chain folded to interact with the aminoacids (Gallet et al., 1999).

A significant contrast in friction is observed (Fig. 2), despite the presence of alkyl chain ends for both compounds. The friction contrast may originate from a difference in molecular packing within the films: compared to the closely packed DPPC alkyl chains, the molecular disorder of the surfactin alkyl chains may give rise to higher friction. On the other hand, the higher friction on the surfactin domains may also reflect higher surface energy, compared to DPPC, due to the exposure of polar amino acids at the surface.

Spatial organization of pure and mixed monolayers

XPS brings further information on the spatial organization of the monolayers. Since both nitrogen and carbon are characteristic of the biological overlayers, but not of mica, N/C ratios determined by XPS for pure monolayers, (N/C)xps, may be compared with N/C ratios deduced from the stoichiometric composition, (N/C)sto, of surfactin (C₅₃O₁₃N₇H₉₃) and DPPC (C₄₀O₆N₁P₁H₈₀). For the pure surfactin monolayer, a good agreement is found between the two ratios: (N/C)xps = 0.14 and (N/C)sto = 0.13. Hence, the XPS data fit with a random overlayer model, i.e., a model in which surfactin is not vertically oriented (Fig. 6 A). This model is consistent with the molecular configuration deduced from the AFM images. The agreement between the experimental and stoichiometric N/C ratios also indicates that the concentration of carbon originating from organic contamination is negligible.

View larger version (41K): [in this window] [in a new window]   FIGURE 6   Geometric models considered for interpreting XPS data. (A) surfactin monolayer; (B) DPPC monolayer; (C) mixed surfactin/DPPC monolayer. The surfactin monolayer is modeled as a homogeneous layer. The DPPC monolayer is modeled as an inner polar layer of thickness tP and an outer hydrocarbon layer of thickness tHC. Mixed monolayers are modeled as combinations of the two above models; the fraction of the surface occupied by surfactin is .

In contrast, for the pure DPPC monolayer, the (N/C)xps ratio (0.016) is significantly lower than the (N/C)sto ratio (0.025) suggesting that the contribution of nitrogen is attenuated due to the presence of an overlayer. To gain further insight into the spatial organization of the film, the XPS data may be interpreted considering that the pure DPPC monolayer consists of a double layer, i.e., an inner polar layer of thickness t_P and an outer hydrocarbon layer of thickness t_HC (Fig. 6 B). The apparent N/C ratio expected from XPS is given by:

(4)

where i_C (1.00) and i_N (1.68) are the sensitivity factors for carbon and nitrogen, respectively; _C (1.0) and _N (1.8) are the Scofield cross-sections (Scofield, 1976);  represents the inelastic electron mean free path in the polar layer (_N^P = 3.2 nm and _C^P = 3.5 nm; Andrade, 1985) or in the hydrocarbon layer (_N^HC = 3.6 nm and _C^HC = 3.9 nm; Andrade, 1985); C_N^P and C_C^HC are the concentrations of nitrogen and carbon in the polar (C_N^P = 3.2 mmol cm³) and hydrocarbon (C_C^HC = 76.8 mmol cm³) layers;  is the take-off angle measured between the normal to the sample surface and the direction of photoelectron collection, i.e., 55°.

Fig. 7 shows the variation of the apparent N/C ratio calculated from Eq. 4 as a function of the polar (t_p) and hydrocarbon (t_HC) layer thicknesses. The experimental N/C ratio of 0.016 is compatible with different pairs of (t_P, t_HC) values. Considering a t_P value of 0.7 nm, estimated from the volume occupied by the DPPC head group, i.e., 0.324 nm³ (Marra and Israelachvili, 1985), and the area occupied at the air-water interface, gives a best value for t_HC of about 1.5 nm and thus a total thickness of 2.2 nm, which is in satisfactory agreement with the expected DPPC film thickness. This observation supports the validity of the double layer model for the DPPC film, i.e., a model of vertically oriented molecules in which the polar headgroups are in contact with mica (Fig. 6 B). It also points to the fact that the level of carbonaceous contamination at the surface is low. Note that a variation of 25% on t_HC or t_P values would cause a variation of about 25% on the apparent N/C ratio (Fig. 7), indicating that the sensitivity of the results to the data precision is appreciable.

View larger version (17K): [in this window] [in a new window]   FIGURE 7   Variation of the apparent N/C ratios computed for the pure DPPC monolayer (Eq. 4) as a function of tP for tHC of 1.0, 1.25, 1.5, 1.75, and 2.0 nm. The horizontal line shows the N/C ratio determined by XPS.

Mixed monolayers may be modeled as a combination of the two pure monolayers (Fig. 6 C), the apparent N/C ratio being computed as follows:

(5)

where  is the fraction of the surface occupied by surfactin; N_surfactin, N_dppc, C_surfactin, and C_dppc represent the apparent nitrogen and carbon atom fractions determined by XPS on pure surfactin and pure DPPC monolayers. Fig. 8 shows the apparent N/C ratios, computed from Eq. 5, as a function of . As can be seen, experimental values of N/C and  determined by XPS and AFM for surfactin molar ratios of 0.1, 0.25, and 0.5 fit well with the above combined model. Hence, the XPS data for the mixed systems are accounted for by a combination of the two pure monolayers, considering respective surface coverages that are in excellent agreement with those measured by AFM. In this work, LB monolayers were prepared at a low surface pressure (20 mN/m), which corresponds to surfactin peptide rings lying horizontally. In further studies, it could be interesting to focus on the spatial organization of mixed monolayers transferred at higher surface pressures (e.g., 50 mN/m), in relation with cell and liposome membrane properties.

View larger version (13K): [in this window] [in a new window]   FIGURE 8   Variation of N/C ratio as a function of the surfactin surface coverage () for the mixed monolayers. Open and closed circles are two independent sets of experimental data. The line shows the apparent N/C ratios computed from Eq. 5.

	CONCLUSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSION REFERENCES

Although the miscibility and the interactions of biologically active lipopeptides with lipids have been studied for many years, direct information at high spatial resolution was not available. In this study, phase separation is directly visualized by AFM for mixed surfactin/DPPC monolayers on mica at 0.1, 0.25, and 0.5 surfactin molar ratios. The observation of bidimensional domains is consistent with the monolayer properties at the air-water interface. The fraction of the surface occupied by the two domains indicate complete immiscibility of the two compounds. AFM height measurements and modeling of the XPS data provide direct and independent pieces of evidence for the molecular orientation of the two compounds within the monolayers; while DPPC assumes a vertical orientation with the polar head groups in contact with mica, the large peptide ring of surfactin is lying on the mica surface. This work has promising applications in biophysics for the direct characterization of the spatial organization (domain formation, molecular orientation) of multicomponent monolayers and adsorbed phases.