Colloids and Surfaces B: Biointerfaces 52 (2006) 57 75

Transcription

1 Colloids and Surfaces B: Biointerfaces 52 (2006) Mode of interaction of ganglioside Langmuir monolayer originated from echinoderms: Three binary systems of ganglioside/dppc, ganglioside/dmpe, and ganglioside/cholesterol Kazuki Hoda a, Yuriko Ikeda b, Hideya Kawasaki c, Koji Yamada b, Ryuichi Higuchi b, Osamu Shibata a, a Division of Biointerfacial Science, Graduate School of Pharmaceutical Sciences, Kyushu University, Maidashi, Higashi-ku, Fukuoka , Japan b Division of Natural Products Chemistry, Graduate School of Pharmaceutical Sciences, Kyushu University, Maidashi, Higashi-ku, Fukuoka , Japan c Department of Chemistry, Graduate School of Sciences, Kyushu University, Hakozakii, Higashi-ku, Fukuoka , Japan Received 2 May 2006; received in revised form 20 June 2006; accepted 11 July 2006 Available online 14 July 2006 Abstract The surface pressure (π) area (A), the surface potential ( V) A, and the dipole moment (µ ) A isotherms were obtained for monolayers made from a ganglioside originated from echinoderms [Diadema setosum ganglioside (DSG-1)], dipalmitoylphosphatidylcholine (DPPC), dimyristoylphosphatidylethanolamine (DMPE), cholesterol (Ch), and their combinations. Monolayers spread on several different substrates were investigated at the air/water interface by the Wilhelmy method, ionizing electrode method, fluorescence microscopy (FM) and atomic force microscopy (AFM). Surface potentials ( V) of pure components were analyzed using the three-layer model proposed by Demchak and Fort [R.J. Demchak, T. Fort, J. Colloid Interface Sci. 46 (1974) ]. The new finding was that DSG-1 was stable and showed a liquid-expanded film and that its monolayer behavior of V was sensitive for the change of the NaCl concentration in the subphase. Moreover, the miscibility of DSG-1 and three major lipids in the two-component monolayers was examined by plotting the variation of the molecular area and the surface potential as a function of the DSG-1 molar fraction (X DSG-1 ), using the additivity rule. From the A X DSG-1 and V m X DSG-1 plots, partial molecular surface area (PMA) and apparent partial molecular surface potential (APSP) were determined at the discrete surface pressure. The PMA and APSP with the mole fraction were extensively discussed for the miscible system. The miscibility was also investigated from the two-dimensional phase diagrams. Furthermore, a regular surface mixture, for which the Joos equation was used for the analysis of the collapse pressure of two-component monolayers, allowed calculation of the interaction parameter (ξ) and the interaction energy ( ε) between them. The observations using fluorescence microscopy and AFM image also provide us the miscibility in the monolayer state Elsevier B.V. All rights reserved. Keywords: Ganglioside; Dipalmitoylphosphatidylcholine (DPPC); Dimyristoylphosphatidylethanolamine (DMPE); Cholesterol (Ch); Surface pressure (π) surface area (A) isotherm; Surface potential ( V) surface area (A) isotherm; Fluorescence microscopy (FM); Atomic force microscopy (AFM) 1. Introduction Glycosphingolipids (GSLs) are typical constituents of various cell membranes in a wide variety of organisms [1,2]. In particular, it is known that GSLs have numerous physiological functions due to variations in the sugar chain in spite of Corresponding author. Tel.: ; fax: address: shibata@phar.kyushu-u.ac.jp (O. Shibata). URL: kaimen/. very small presence of the constituents. GSLs are classified into cerebrosides, sulfatides, ceramide oligohexosides, globosides, and gangliosides based on the constituent sugars. The simplest glycolipids are monoglycosylceramides (cerebrosides) in which there is only one sugar residue (either glucose or galactose), but most complex GSLs contain one or more N-acetyl-neuramic (sialic) acid residues possessing a net negative charge. Gangliosides are especially enriched in the brain and nervous tissues and they constitute 5 10% of the total lipid mass in nerve cells [3]. Simons and Ikonen [4,5] proposed a model for mem /$ see front matter 2006 Elsevier B.V. All rights reserved. doi: /j.colsurfb

2 58 K. Hoda et al. / Colloids and Surfaces B: Biointerfaces 52 (2006) brane structure that the membrane contains microdomain or rafts enriched in sphingomyelin, glycosphingolipid and cholesterol; proteins can selectively be included or excluded from these microdomains and also a model of lipid rafts has a function as platforms for signal transduction that is triggered by tropic factor receptors [6]. Membrane domains are made up by rafts of enriched in cholesterol, sphingomyelin (SM) and GSLs, and these domains have received increasing attention in cell biology [7]. Disturbances in molecular composition, cellular distribution and/or formation of such domains might thus have implications in pathophysiological events. GSLs are involved in the regulation of many cellular functions, including cell recognition [8 10], cell differentiation [11,12], signal transduction [13,14], apoptosis [15], receptors for virus [16], and progress of Alzheimer s disease (AD) [17 20]. Recently, a number of GSLs have been isolated from marine invertebrates such as echinoderms, poriferans, and mollusks [21 25]. Concerning surface physicochemical properties, we have also investigated biologically active GSLs from echinoderms to elucidate the structure-function relationships of GSLs and to develop novel medicinal resources [26,27]. Although it has been shown that gangliosides inhibit the activity of phospholipases A 2 and C through a surface-mediated mechanism [28,29], gangliosides used in these studies almost originated from mammals. It is hopeful that gangliosides from marine invertebrates have higher bioactivity than those from mammals. The interactions of gangliosides from mammals with lipid substrates have been extensively studied using different experimental methods [30 32]. An effective way to investigate the in-plane interactions between cholesterol and other lipids is to use Langmuir film balance approaches [33]. An important advantage of studying lipid lipid interactions using the monolayer approach is that the range of molecular areas known to occur in membrane systems can be investigated systematically while avoiding the mesomorphic changes that often occur in bulk hydrated dispersions (e.g. bilayer vesicles) as lipid composition is varied. Also, the in-plane lipid lipid interactions can be isolated from the influences of changing transbilayer compositional distributions that may occur in bilayer systems when cholesterol mole fractions are varied [34,35]. For these reasons, monolayer investigations of sphiongolipid cholestrol interactions have proven useful [36 42]. In these earlier studies, interactions between cholesterol and different sphingolipids were evaluated by measuring directly the average cross-sectional molecular areas at various mixing ratios and surface pressures. The extent to which the average molecular area in a mixed composition film decreased below that predicted by summing the known molecular areas of the pure lipids (apportioned by mole fraction) provided a measure of the area condensation induced by cholesterol. The classic monolayer approach was originally used to show the marked condensing effect that cholestrol has on fluid-phase phosphoglycerides [33 and references therein]. As a result, the extent of lipid monolayer condensation induced by cholesterol has been used often as an indicator of affinity or interaction strength that cholesterol has for various phosphoglycerides or sphigolipids [36,38,39,43]. For example, Bordi et al. investigated mono- (GM1, GM2, and GM3) and di-sialogangliosides (GD1a and GD1b) inserted in a DPPC matrix [44]. Various kinds of multi-components monolayers originating from bovine brain spread at the air/water interface have been studied [44 52]. However, the molecular mechanisms underlying the gangliosides from marine invertebrates have not been investigated, yet. Here we have focused on characterizing the Langmuir monolayer behavior of pure ganglioside from echinoderms (DSG-1), phospholipids (DPPC and DMPE), cholesterol (Ch) and their two-component systems at the air/water interface. The surface pressure (π) area (A), surface potential ( V) area (A), and dipole moment (µ ) area (A) isotherms were obtained for the pure compounds and their two-component systems. The surface potentials were analyzed using the three-layer model proposed by Demchak and Fort [53]. The phase behavior of twocomponent monolayers was examined in terms of additivity of molecular surface area and surface potential. Furthermore, it was analyzed employing the partial molar molecular area (PMA) and the apparent partial molar surface potential (APSP). The molecular interaction between monolayer components was investigated using the Joos equation. Finally, the monolayers were examined by fluorescence microscopy (FM) and atomic force microscopy (AFM). 2. Experimental 2.1. Materials The sea urchin Diadema setosum (Gangaze in Japanese) is belonging to the diadematidae family, the diadematoida order, the echinoidea class, and the echinodermata phylum of animals. It was collected in the sea near Kouyagi in Nagasaki, Japan in The isolated material from the echinoderm of ganglioside (D. setosum ganglioside, DSG-1) was checked by 1 H and 13 C NMR spectra, FAB MS spectra, and GC MS spectra after purification by TLC and HPLC. The compositions of the hydrophobic acyl chain and long chain base (LCB) are given in Table 1. The chemical structure of isolated ganglioside DSG-1 was molecular species as shown in Fig. 1. The more detailed separation and purification will be reported elsewhere [54]. Table 1 Acyl chain compositions of DSG-1 Acyl chain Composition (%) C16:0 3.3 C18:0 6.9 C22: C22:0 9.8 C23: C23:0 6.4 C24: C24:0 6.7 LCB part C18:0 100

3 K. Hoda et al. / Colloids and Surfaces B: Biointerfaces 52 (2006) Fig. 1. Chemical structure of DSG-1. Dipalmitoylphosphatidylcholine (l- -1-palmitoyl-2-hydroxy-sn-glycero-3-phosphocholine; DPPC) and dimyristoylphosphatidylethanolamine (1,2-dimyristoyl-sn-glycero-3-phosphoethanolamine; DMPE) were purchased from Avanti Polar Lipids, Inc. (Birmingham, AL, USA), and cholesterol (Ch) was obtained from Sigma Chemical Company, USA. Their purity was >99% and used without further purification. Their stock solutions (0.5 mm) were prepared in chloroform/methanol mixture (2/1 v/v) for DMPE and in n-hexane/ethanol mixture (7/3 v/v) for the rests, and 90 L of each solution was spread at the air/aqueous solution interface. Chloroform, n-hexane and methanol were purchased from Cica-Merck (Uvasol). On the other hand, ethanol came from Nacalai Tesque. A period of time, 15 min, was needed to evaporate the spreading solvent. All the solvent purity was of spectroscopic or fluorometric grade. Several different substrates were prepared using thrice distilled water (surface tension, 72.7 mn m 1 at K; resistivity, 18 M cm). Sodium chloride (Nacalai Tesque) was roasted at 1023 K for 24 h to remove any surface active organic impurity [26,27] Surface pressure (π) area (A) and surface potential ( V) area (A) isotherms The surface pressure (π) was measured using an automated homemade Wilhelmy film balance. The surface pressure balance (Mettler Toledo, AG245) had a resolution of 0.01 mn m 1. The surface pressure measurement system was equipped with a filter paper (Whatman 541, periphery 4 cm). The trough (effective area 150 mm 480 mm) was made from Teflon-coated brass. The π A isotherms were recorded at ± 0.05 K. The monolayer was compressed at a speed of 0.14 nm 2 molecules 1 min 1. The standard deviations for area and surface pressure measurements were 0.01 nm 2 and 0.1 mn m 1, respectively. Surface potential ( V) was recorded upon compression of the monolayer spread on several different substrates at K. It was monitored using an ionizing 241 Am electrode placed 1 2 mm above the interface, while a reference electrode was dipped into the subphase. The standard deviations for surface potential measurements were ±2.5 mv. The other conditions were the same as described in previous papers [55]. The V A data were transferred to the surface dipole moment (µ )by computer as reported in the literature [55]. Before each experiment, the trough was rinsed and cleaned with acetone and chloroform, alternately. The absence of surface-active compounds in the subphase (0.15 M NaCl, about ph 6.5) was checked by reducing the available surface area to less than 4% of its original area after sufficient time was allowed for adsorption of possible impurities that might be present by trace amounts in the substrate. Only substrate that did not show changes of surface pressure above 0.5 mn m 1 and of surface potential above 50 mv on this procedure was used Fluorescence microscopy Fluorescence images were observed using an automated homemade Wilhelmy film balance equipped with a fluorescence microscope (BM-1000, U.S.I. system, Japan) [56]. It is possible to record simultaneously the surface pressure (π) area (A) and the surface potential ( V) A isotherms along with the monolayer images in order to correlate these properties of the same monolayer. A 300 W lamp (XL 300, Pneum) was used for the fluorescence excitation. A 546 nm band path filter (Mitutoyo) was used for excitation and a 590 nm cut-off filter (Olympus) for emission. The monolayer was observed using a 20 longdistance objective lens (Mitutoyo f = 200/focal length 20 mm). A xanthylium 3,6-bis(diethylamino)-9-(2-octadecyloxycarbonyl) phenyl chloride (R18, Molecular Probes) was used as an insoluble fluorescent probe. The probe has the absorbance and emission band maxima at 556 and 578 nm, respectively. The solution used in the fluorescence microscopy experiments contained 1 mol% of the fluorescent probe against insoluble materials. Fluorescence images were recorded with a CCD camera (757 JAI ICCD camera, Denmark) connected to the microscope and transferred directly into computer memory through an online image processor (VAIO PCV-R53, Sony: video capture soft). The entire optical set-up was placed on an active vibration isolation unit (Model-AY-1812, Visolator, Japan). Image analysis was performed using Scion image. The percentage in FM images indicates the ratio of ordered (or LC) domain to the screen area. The operation of the present microscopy was similar to the previous reports [56].

4 60 K. Hoda et al. / Colloids and Surfaces B: Biointerfaces 52 (2006) Atomic force microscopy Langmuir Blodgett (LB) film preparations were carried out with a homemade LB trough. Freshly cleaved mica (Okenshoji Co., Tokyo, Japan) was used as a supporting solid substrate for the film deposition. At selected surface pressures, a transfer velocity of 5 mm min 1 was used for film-forming materials on a 0.15 MNaCl subphase. AFM images were obtained using an SPA 400 instrument (Seiko Instruments Co., Japan) at ± 2 K in a tapping mode, which provided both a topographical image and a phase contrast one. Prior to use, the cantilevers were irradiated with ultraviolet light (8 mw/cm 2 at nm) for 20 min and then washed with distilled ethanol. The tapping mode images (512 points collected per line) were collected with scan rates of Hz, using silicon tips (Olympus Co., Japan) with a nominal spring constant of 1.8 N m 1 under the normal atmosphere. Transferred samples were checked for possible tip-induced deformation by zooming out after a region had been scanned. The domain size distribution was measured with the particle analysis function of Seiko Instruments software [56] Surface pressure (π) area (A), surface potential ( V) area (A), and dipole moment (µ ) area (A) isotherms of DSG-1 monolayer on different NaCl concentrations Charge effect of DSG-1 head group The ganglioside constituents are unique in that a sialic acid directly binds to the glucose of cerebroside, that they are mutually connected in tandem, and that some are located in the internal parts of the sugar chain. It became also apparent that sialic acid was indispensable for the neuritogenic activities. The aim of this approach was to investigate the influence that the polar head groups of the negative charge for the DSG-1 might have on the degree of the capture for sodium ion. Here, π A, V A, and µ A isotherms of DSG-1 monolayers on various different concentration substrates were shown in Fig. 3. The area-normalized surface potential ( V N = VA) asa function of the area per molecule [57] are shown in Fig. 5A C of Section 3.4. However, in an attempt to interpret the differences in V N for this compound, it should be kept in mind that this parameter has a different meaning between uncharged and 3. Results and discussion 3.1. Stability of the DSG-1 monolayer To check the stability of the monolayer, the relaxation of the surface pressure for DSG-1, DPPC, DMPE and Ch after compression up to ca. 35 mn m 1 was measured. A 0.15 M NaCl solution was chosen as the subphase in order to mimic a biomembrane-like environment. Fig. 2 shows that the surface pressure of DSG-1 decreased during the first 20 min and then remained constant at ca. 34 mn m 1. On the other hand, even though small decreasing in the surface pressure was observed for DPPC, DMPE, and Ch even after 1hr, their aqueous solubilities are so low that these species can hardly dissolve in bulk solution (substrate phase). Their surface pressures are in good agreement with literature value [44]. So they are stable enough to form an insoluble monolayer at the air/water interface. Fig. 2. Time dependence of the surface pressure (π). Langmuir monolayers were compressed up to 35 mn m 1 on 0.15 M NaCl at K. The π t measurement was then started. Fig. 3. (a) Surface pressure (π) area (A), (b) surface potential ( V) area (A), and (c) surface dipole moment (µ ) area (A) isotherms of DSG-1 monolayer on six different NaCl concentrations at K.

5 K. Hoda et al. / Colloids and Surfaces B: Biointerfaces 52 (2006) charged molecules, since in the latter an ionic double layer contributes to the total surface potential in the monolayer. Assuming a value of unity for the dielectric constant, the surface potential of a monolayer can be expressed [58] as V = µ ε 0 εa + ψ 0 (1) where ε 0 is the permittivity of a vacuum, ε the mean permittivity of the monolayer (which is assumed to be 1), A the molecular area (nm 2 ), ψ 0 the potential difference of the ionic double layer in ionized monolayers (for uncharged lipids ψ 0 = 0), µ the overall dipole moment in the direction perpendicular to the interface which results from the contribution of the dipole moments µ 1 from the orientated water molecules, µ 2 from the polar head group (saccharide), and µ 3 from the hydrocarbon chains (the terminal C H bond (the CH 3 group)) [58]. Thus, V is a complex quantity resulting from several independent and hardly separable contributions. However, since the chemical structure of the glycosphingolipids has a well-established pattern, comparisons between them are possible. For most of these compounds the hydrophobic portion of the molecule was, on average, identical [59], especially in the case where they were prepared from other components of the series. Therefore, the µ 3 contribution was assumed to be equal throughout. In addition, the contribution of the polar head groups of the different compounds can in every case be related to the compound that has an identical polar head except for the lack of a carbohydrate unit. Consequently, the hydration shell of the common part of the carbohydrate chain of two similar compounds so related were considered to be grossly equal, and the difference in the resultant µ was attributed to the added carbohydrate unit. Finally, the vertical component of surface dipole moment (µ ) was calculated from the measured V values connecting with Gouy Chapman theory. The double-layer contribution ψ 0 arises from the presence of counterions in the subphase opposite to the net charge carried by the molecular polar head group. Only DSG-1 molecule is negatively charged of all components used in this study, and therefore, it is possible to evaluate the counterion effects by changing the concentration of electrolyte. Here, we consider only the variation of the measured surface potential ( V) with the molecular surface area, although the theory does not take into account the specific adsorption of ionic species or the change of effective charge density needed to calculate the potential ψ 0, which results in the uncertainties in estimating the association constants of the different anionic or zwitterionic groups on the polar heads [48]. In Fig. 3, the π A, V A and µ A isotherms of pure DSG-1 system showed the significant concentration effect of subphases. When the concentration of sodium chloride increased in the subphase, the π A isotherms are shift to a larger area due to the effect of lipid head group charge. At more than 0.5 M NaCl, the π A isotherm of DSG-1 arrived at the saturated state. It was assured that the π A isotherm of DSG-1 did not change much, when any excess NaCl was added in the subphase. This phenomenon is explained that the carboxyl group of dissociated sialic acid captured sodium ions and then neutralized monolayer showed a definite π A isotherm profile of DSG-1. Generally, if Table 2 Surface potential data used for electrostatic potential (ψ 0 ) in interface relative to adjacent conducting phase (mv) evaluation NaCl solution (M) A (nm 2 ) V (mv) ψ 0 (mv) Molecular surface area (A) and surface potential ( V) were obtained by at 40 mn m 1. ψ 0 represents the double-layer contribution. the negative charge was neutralized, the electrostatic repulsion between the neutralized molecules might become less making their monolayer more condensed [60]. But the present result was inconsistent with the theory. It is suggested that large ions with a low charge density disrupted the water structure in such a way that monolayers spread on the surface were expanded [61]. From the surface potential V, the different ψ 0 values for the subphase used in this study are shown in Table 2. The ψ 0 value increases with increasing NaCl concentrations. It was indicated that the negative charge of DSG-1 head group was neutralized by Na + ions. Accordingly, the absolute value of V increased depending on NaCl concentration increment (Fig. 3b) Surface dipole moments (µ ) of DSG-1 We analyzed the surface potential of monolayers on 0.15 M NaCl using the three-layer model proposed by Demchak and Fort [53], which is based on the earlier model of Davies and Rideal [58]. This model postulates independent contributions of subphase (layer 1), polar head group (layer 2), and hydrophobic chain (layer 3). Independent dipole moments and effective local dielectric constants are attributed to each of the three layers. Other models such as the Helmholtz model and the Vogel Möbius model are also available [62]. These different models were reviewed [63]. The conclusion was that, despite its limitations, the Demchak and Fort model provides good agreement between the µ values estimated from the monolayer surface potentials and those determined from measurements on bulk material for various aliphatic compounds. The estimation of µ (the vertical components of the dipole moment to the plane of the monolayer) of polar head groups and hydrocarbon chains using the Demchak and Fort model assumes close-packed Langmuir monolayer [53,63]. Application of this model to mainly the DSG-1 liquid-expanded (LE) monolayer may lead to a rough estimation. However, if the closest-packed DSG-1 monolayer is applied to this model, it may lead to a useful estimation, which can help us to provide qualitative explanation of surface potential behavior. We have thus compared the experimental values of µ in the most condensed state of the monolayer with those calculated µ calc by the three-layer model equation: µ calc = µ 1 ε 1 + µ 2 ε 2 + µ 3 ε 3 (2)

6 62 K. Hoda et al. / Colloids and Surfaces B: Biointerfaces 52 (2006) Table 3 Surface potential data used for dipole moment evaluation Sample A (nm 2 ) V (mv) µ (D) Stearic acid (SA) DSG DPPC DMPE Ch A is the molecular surface area obtained by the close-packed high pressure portion of the π A isotherms and V is obtained at that point. µ is total dipole moment and the subphase was 0.15 M NaCl at K in all cases. where µ 1 /ε 1, µ 2 /ε 2 and µ 3 /ε 3 are the contributions of the subphase, polar head group and hydrophobic chain group, respectively. We want to determine the contribution of the hydrophilic group of DSG-1. Carboxylic and hydroxyl groups have already been determined by the Demchak and Fort model [53]. The initial set of values proposed by Demchak and Fort (µ 1 /ε 1 = D, ε 2 = 7.6, and ε 3 = 5.3) were determined for monolayers made from terphenyl derivatives and octadecyl nitrile. Another set of values were determined by Petrov, Polymerropoulos and Möhwald (µ 1 /ε 1 = D, ε 2 = 7.6, and ε 3 = 4.2) for monolayers of n-heptanol and 16-bromohexadecanol [64] and by Taylor and Oliveira (µ 1 /ε 1 = D, ε 2 = 6.4, and ε 3 = 2.8) for monolayers of halogenated fatty acids and amines [63]. To determine the set of the parameters fitting in our experimental condition, the selection of parameter values was done using those for the standard sample, stearic acid (SA). These data were listed in Table 3. The experimental values of surface dipole moment for SA used to determine the set of the parameters were as follows µ (SA) = µ 1 + µcooh 2 ε 1 ε 2 + µch3 3 = 0.16 D (3) ε 3 In the calculation, it was assumed that the C X dipole of terminal CH 2 X moiety (where X is a hydrogen) was inclined at half the tetrahedral angle (i.e ) with respect to the water surface as suggested by Bernett et al. [65] and that the group moments had the values given by Smyth [66]. In addition, it was assumed that the C H group moment was 0.4 D, the carbon being negatively charged [67]. So the contribution of the terminal methyl group was 0.33 D [53]. Here, we have used µ 2 (COOH- cis(cis) acid) = 0.82 D value, because these parameter values provided a good agreement between calculated values and experimental ones of dipole moments measured on a saline phase. We have used the combination of the set of values (µ 1 /ε 1 = D, ε 2 = 7.6, µ 3 = 0.33 D, and ε 3 = 2.8 for CH 3 ). Secondly, we evaluated the contribution of the head group of DSG-1. We have used the CH 3 group of D = 0.66 D for the following equation: µ (DSG-1) = µ 1 + µdsg-1 + ε 1 2ε µ 3 = 0.27 D (4) 2 ε 3 The contribution of head group of DSG-1 comes from the sialic acid and the -D-glucose. From Eq. (4), we could get µ DSG 1 2 = 0.75 D. Finally, to make sure the suitability of the above set of parameters, we rechecked the contribution of the polar head group to the dipole moment for DPPC and DMPE: µ (DPPC) = µ 1 + µpc 2 ε 1 ε 2 µ (DMPE) = µ 1 + µpe 2 ε 1 ε 2 + µch3 3 = 0.58 D (5) ε 3 + µch3 3 = 0.54 D (6) ε 3 Next for Ch molecules, we evaluated the contribution of the hydrophobic steroid segment by the following equation (7) (see Table 3). The contribution of µ 2 to surface dipole moment have been proposed for the different conformations of the OH group: µ 2 (OH-gauche) = 1.00 D, µ 2 (OH-trans)= 0.63 D, and µ 2 (OH-free) = 0.18 D. We have employed the above µ 2 (OHgauche) = 1.00 D, as many studies support the gauche conformation for condensed alkanol monolayers [53]: µ (Ch) = µ 1 + µoh 2 ε 1 ε 2 + µch 3 = 0.37 D (7) ε 3 Using the experimentally determined V values (see Table 3) and assuming the set of the values (µ 1 /ε 1 = D, ε 2 = 7.6, µ 3 = 0.33 D for DPPC and DPPE, and ε 3 = 2.8 for CH 3 ), the above three equations allowed us to obtain µ PC 2 = 3.11 D for PC, µ PE 2 = 2.80 D for PE, and µ Ch 3 = 0.85 D for Ch. These values for DPPC, DMPE and Ch are a little bit larger than those reported by Taylor et al. [63] (2.44 and 2.23 for DPPC and DPPE, respectively) and by Nakahara et al. [27] (0.74 for Ch). These differences may result from variation in experimental conditions such as substrate composition (electrolyte and ph), compression rate, and so forth Three binary systems of DSG-1 with DPPC, DMPE and Ch The π A, V A, and V N (= VA) A isotherms of monolayers made from DSG-1, phospholipids (DPPC and DMPE) and Ch on 0.15 M NaCl solution at K were shown in Fig. 4. The DSG-1 monolayer formed a typical liquid-expanded (LE) film and its extrapolated area was 0.89 nm 2. A cross-section area of DSG-1 was dominated by its large head group containing -NeuAc and -glucose. The DSG-1 monolayer collapsed at 44.5 mn m 1 (0.55 nm 2 ). The surface potential ( V) was a measure of the electrostatic field gradient perpendicular to the surface and thus varies considerably with the molecular surface density. The behavior of V A isotherms corresponds to the change of the molecular orientation upon compression. The surface potential ( V) of DSG-1 always showed positive and V value of 46.0 mv at high surface pressure was much smaller than those for other lipids. It indicated that its two hydrocarbon chains would not stand vertically at high surface pressure because the head group was fairly bulky. And the V A isotherm showed a hump at 1.3 nm 2. This hump comes from the conformational change

7 K. Hoda et al. / Colloids and Surfaces B: Biointerfaces 52 (2006) Fig. 4. (a) Surface pressure (π) area (A), (b) surface potential ( V) area (A), and (c) surface dipole moment (µ ) area (A) isotherms of pure DSG-1, DPPC, DMPE and Ch monolayers on 0.15 M NaCl at K. in the monolayer state. This change coincided with a phase transition from a gaseous phase to LE phase, which was observed as a morphological change by using fluorescence microscopy. A π A isotherm of DPPC presented a characteristic firstorder transition from the LE phase to the liquid-condensed (LC) phase. The transition pressure (π eq ) at K was 11.8 mn m 1, above which the surface pressure rose due to the orientational change. Its collapse pressure of DPPC was 55.2 mn m 1 (0.39 nm 2 ). Similar to DPPC, DMPE also had the LE/LC phase transition at 14.1 mn m 1 and its collapse pressure was 50.8 mn m 1 (0.36 nm 2 ). Ch formed an ordered LC monolayer, its collapse pressure was 42.0 mn m 1 (0.36 nm 2 ). Absolute values of V at the closed packing state showed 550 mv for DPPC, 565 mv for DMPE, and 392 mv for Ch. These values were very close to those previously reported [26,27,68]. Next, three combinations of two-component monolayer systems composed of the DSG-1 and three major lipids (DPPC, DMPE, and Ch) have been studied in order to clarify the interaction between two components and the miscibility in the monolayer state. For the above purpose, the π A, V A, and V N (= VA) A isotherms of DSG-1/DPPC, DSG-1/DMPE, and DSG-1/Ch system were measured at K on a 0.15 M NaCl subphase. The isotherms of these binary systems were shown in Fig. 5. Forπ A isotherms, all the curves of the twocomponent systems existed between those of the respective pure components and successively increased with the mole fraction of DSG-1. The π A isotherms in DSG-1/phospholipids (DPPC and DMPE) systems were shown in Fig. 5A and B, respectively. For X DSG-1 lower than 0.5, π A isotherms displayed phase transition pressure that increased almost linearly with increasing X DSG-1. The behavior suggested that DSG-1 had an ability to mix with DPPC or DMPE in the monolayer. On the other hand, Fig. 5C shows π A, V A and V N (= VA) A isotherms of binary DSG-1/Ch system. Pure Ch formed typical LC film in its π A isotherm, and all the isotherms of the binary system sit in the order of the mole fraction (X DSG-1 ) between those of both pure systems. The influence of X DSG-1 on the V A (b), and V N (= VA) A (c) isotherms is shown correspondently in Fig. 5A C. Both the surface potential ( V) and the area-normalized surface potential ( V N = VA) of the monolayers also clearly indicate that all the curves of the two-component systems exist between those of the respective pure components and that they successively change with the mole fraction. The interaction between DSG-1 and their lipid molecules was investigated by examining whether the variation of the mean molecular areas as a function of X DSG-1 satisfied the additivity rule [69]. A comparison between the experimental mean molecular areas and the molecular areas based on ideal mixing at four surface pressures (5, 15, 25, and 35 mn m 1 ) was shown in Fig. 6.Forπ = 5 and 35 mn m 1 in DSG-1/DPPC and DSG- 1/DMPE system, experimental values show a small negative deviation from the theoretical line, indicating attractive interactions between them. This may result from the fact that the interactions are mainly governed by the enhanced attractions between hydrophilic groups. At 15 and 25 mn m 1, positive deviations are observed, resulting from the LE/LC coexistence region of DPPC and DMPE in their π A isotherm. In DSG-1/Ch system, the A X DSG-1 shows a big negative deviation at all the surface pressures. This suggests that these negative deviations attribute to a good fit in size between the small polar head group (Ch) and big one (DSG-1) and to a good packing between the big steroid segment (Ch) and hydrophobic chain of DSG-1. An analysis of V for the binary monolayer in terms of the additivity rule is presented in Fig. 7.Forπ =5mNm 1 in DSG- 1/DPPC and DSG-1/DMPE systems, the variation almost obeys the additivity rule. For the other pressures, comparisons of the experimental data with calculated ones clearly indicate small negative deviations from the ideal line, which results from the conformational change of head group for both the DSG-1 saccharide and DPPC choline group and from change in packing density of hydrophobic groups for DSG-1 and DPPC at high surface pressures. In DSG-1/Ch system, the additivity of V almost indicates a good agreement with the ideal line for 5 and 15 mn m 1.ForX DSG-1 = 0 0.3, the additivity of V shows good linearity of the ideal line at higher surface pressures of 25 and 35 mn m 1. While for 0.3 < X DSG-1 < 1, the additivity of V shows a positive deviation, where Ch acts like a filler to promote more compact packing in the two-component monolayer state.

8 64 K. Hoda et al. / Colloids and Surfaces B: Biointerfaces 52 (2006) Fig. 5. (a) Surface pressure (π) area (A), (b) surface potential ( V) area (A), and (c) surface dipole moment (µ ) area (A) isotherms of DSG-1 with (A) DPPC, (B) DMPE, and (C) Ch as a function of DSG-1 molar fraction (X DSG-1 ) on 0.15 M NaCl at K. To clarify the miscibility of these two-component systems in more detail, we have applied mean molecular surface areas (A m ), partial molecular surface areas (PMA) and apparent partial molecular surface potentials (APSP). When PMA and APSP are evaluated, the molecular occupation and orientation behavior at the air/subphase interface can be more clearly seen Partial molar quantities (PMA and APSP) When π A isotherms of a given binary system are analyzed, it is necessary to examine whether the relation of mean molecular surface areas (A m ) with the mole fraction (X) satisfies the additivity rule or not, and if not, which negative or positive deviation is observed. The comparison between experimental mean molecular areas and calculated ones for ideal mixing at four surface pressures (5, 15, 25, and 35 mn m 1 ) are showed in Fig. 6. A binary system can show the ideal behavior by either forming ideally mixed monolayer or that the two components can not mix at all but can form the so-called patched film, where the additivity rule should show a linear relation as indicated by a broken line. The behavior of occupied surface area and surface potential can be seen more clearly if the partial molar quantities are evaluated, where they have been employed in previous studies [26,27]. Here, A m satisfies the following equation: A m = X 1 A 1 + X 2 A 2 (8) where X 1 and X 2 are the mole fractions of the components 1 and 2, respectively, and A 1 and A 2 are the partial molecular areas (PMA) in the two-component film at a definite surface pressure. When PMA is denoted as A 1 and A 2 for components 1 and 2, respectively, A 1 and A 2 values can be determined as the respective intercept value at X 2 = 0 and X 2 = 1 of a tangential line drawn at any given point on the A m X DSG-1 curve as shown in Fig. 6. A 1 and A 2 from the relation are given as A 1 = A m X 2 ( A m / X 2 ) T,π (9) A 2 = A m + (1 X 2 )( A m / X 2 ) T,π (10) where A i is defined as A i = ( A t / N i ) T,π (11) when N 1 plus N 2 molecules from a surface area A t (=N 1 A 1 + N 2 A 2 ), and 1 and 2 denote DSG-1 and one of three lipids, respectively. It is noted that examination of the PMA values for respective component molecules in binary systems as a

9 K. Hoda et al. / Colloids and Surfaces B: Biointerfaces 52 (2006) Fig. 6. Mean molecular area (A m ) of binary (a) DSG-1/DPPC, (b) DSG-1/DMPE, or (c) DSG-1/Ch systems as a function of the composition of DSG-1 at four different pressures. The dashed lines were calculated by assuming the additivity rule, and the solid circles represent experimental values. function of composition can make us see the interaction mode between the two different molecules that depends, in many cases, on mixing ratio. Correspondingly, the apparent partial molecular surface potential (APSP) can be obtained from the relationship between the average molecular surface potential and mole fraction from Fig. 7, which is the same as the above procedures: V 1 = V m X 2 ( V m / X 2 ) T,π (12) where V m was evaluated by dividing the measured surface potential ( V) by the number of molecules in the unit area. The surface potential ( V) is measured by an air electrode whose cross-sectional area is ca. 1 cm 2. Therefore, we assumed its dimension to be mv cm 2. The average molecular surface potential in mv molecule 1 unit can be obtained by dividing the V by the number of molecules per 1 cm 2 from the π A isotherm. When APSP is denoted as V 1 and V 2 for components 1 and 2, they are determined by the respective intercepts at X 2 = 0 and X 2 = 1 of a tangential line drawn at any given point on the V m X DSG-1 curve. PMA and APSP procedures were applied to three kinds of binary systems. The PMA X DSG-1 curves were shown in Fig. 8. It is noted that if the two-component systems are ideal mixing, the PMA and APSP should be parallel to the axis of

10 66 K. Hoda et al. / Colloids and Surfaces B: Biointerfaces 52 (2006) Fig. 7. Surface potential ( V) of binary (a) DSG-1/DPPC, (b) DSG-1/DMPE, or (c) DSG-1/Ch systems as a function of the composition of DSG-1 at four different pressures. The dashed lines were calculated by assuming the additivity rule, and the solid circles represent experimental values. X DSG-1 (the additivity rule). As is shown in Fig. 8a, the PMA at5mnm 1 of DSG-1 and DPPC exhibit a negative deviation, which markedly depends on the mole fraction. It suggests that the intermolecular interaction between two hydrophilic parts is stronger than the respective pure components themselves and that the extent of interaction is governed by molecular arrangement. At 15 mn m 1, both molecules have shown such characteristic PMA changes depending greatly much on the composition range. Here, it is noted that the positive PMA means that addition of a molecule to a given monolayer system causes a increase in total area compared with the value of individual PMA or reflects the effect of the LE/LC phase transition of DPPC molecule. At 25 mn m 1, PMA of two components show an almost linear. At 35 mn m 1, where a DPPC molecule completely forms a LC film, the DPPC contribution for the negative deviation of area is much larger when DPPC is a minor component. It is suggested that DSG-1 saccharide can assume native conformations of more preferable coexisting with DPPC on the monolayer state. The PMA for DSG-1/DMPE system indicates almost similar trend to the DSG-1/DPPC system. On the contrary, for DSG-1/Ch system, PMA s of DSG-1 and Ch show enormously different behavior depending on the mole fraction at all surface pressures. In the case where DSG-1 is smaller in the mole fraction, DSG-1 component makes its PMA change-

11 K. Hoda et al. / Colloids and Surfaces B: Biointerfaces 52 (2006) Fig. 8. Partial molecular surface area (PMA) of binary (a) DSG-1/DPPC, (b) DSG-1/DMPE, or (c) DSG-1/Ch systems as a function of the composition of DSG-1 at four different pressures. able very much, while that of Ch is almost constant. Negatively deviated PMA means that addition of other component molecule at given system causes a reduction in total area, accompanying the sacrifice of the molecular area by other component in the surface. Then, the interaction among molecules is enhanced. In contrast to PMA, the APSP X DSG-1 curves for three systems are shown in Fig. 9. The APSP for both DSG-1/DPPC and DSG-1/DMPE systems indicates the similar behavior at each surface pressure. It is found that DSG-1 molecules have almost similar orientation states at each surface pressure, although partner molecules are different in the length of hydrocarbon chains and their head groups. At 5 and 15 mn m 1, APSP of DSG-1 decreases with increasing mole fraction of DSG-1. For example, DPPC molecules at X DSG-1 = 0.9 are surrounded almost by the DSG-1 molecules in the binary DSG-1/DPPC system. In the monolayer state, DSG-1 has a minimum molecular area of about 0.89 nm 2 (Fig. 4), which is limited by the much large head group cross-sectional area. Then, the cross-sectional area of a closely packed, all trans, hydrocarbon chain is about 0.20 nm 2, and then the hydrocarbon portion of the double-chain molecules would occupy an area of A = nm 2 = 0.40 nm 2 [70].Itis assumed that the extrapolated area of DSG-1 is much larger than 0.40 nm 2, because DSG-1 has large head group. As indicated by decreasing its APSP, this mismatch results in a tilt angle

12 68 K. Hoda et al. / Colloids and Surfaces B: Biointerfaces 52 (2006) Fig. 9. Apparent partial molecular surface potential (APSP) of binary (a) DSG-1/DPPC, (b) DSG-1/DMPE, or (c) DSG-1/Ch systems as a function of the composition of DSG-1 at four different pressures. of aliphatic chains and a reduction in the attractive interactions between the DPPC and DSG-1 chains. The tilted chains are also accompanied by a decrease in the coherence of monolayer packing. Upon compression at 25 and 35 mn m 1, both components show an almost linear relation in regard to APSP versus X DSG-1. For DSG-1/Ch system, APSP indicates the similar behavior at different surface pressures as shown in Fig. 9c. It is found that APSP of Ch molecules remains almost constant with increasing the mole fraction of X DSG-1 from 0.1 to 0.3 and then increases with an increase in X DSG-1 > 0.3. It suggests that these positive deviations result from a good fit in size as mentioned above Two-dimensional phase diagrams From the π A isotherms in these two-component systems, their two-dimensional phase diagrams were constructed by the change of transition pressures (π eq ) and collapse pressures (π c ) at various mole fractions of DSG-1. The phase diagrams at K are shown in Fig. 10. In DSG-1/DPPC system, the transition pressures from disordered (gaseous or liquid-expanded) to ordered (liquidcondensed) phase are also plotted against a mole fraction of DSG-1 in Fig. 10a. For X DSG-1 from 0 to 0.5, π A isotherm displays a phase transition pressure (π eq ) that changes almost

13 K. Hoda et al. / Colloids and Surfaces B: Biointerfaces 52 (2006) Fig. 10. The transition pressure (π eq ) mole fraction and the collapse pressure (π c ) mole fraction of DSG-1 (X DSG-1 ) phase diagram of binary (a) DSG-1/DPPC, (b) DSG-1/DMPE, and (c) DSG-1/Ch systems on 0.15 M NaCl at K. The dashed line was calculated by Eq. (13) in the case of ξ =0. linearly with X DSG-1. Judging from the change of the transition pressure, the two components at all other mole fractions are miscible each other. This behavior is the first evidence of the miscibility for the two components within the monolayer states. This can be explained by the fact that film-forming molecules become denser by compression, decreasing the surface tension by the film-forming molecules. Then the resultant surface pressure increased. Increase in the transition pressure with mole fraction of DSG-1 means that the transition does take place when the film-forming molecules become denser with the mole fraction. These phenomena resemble the elevation of boiling point of liquid mixture. Assuming that the surface mixtures behave as a regular solution with a hexagonal lattice, the coexistence phase boundary between the ordered monolayer phase and the bulk phase can be theoretically simulated by the Joos equation, and the interaction parameter (ξ) was calculated from this deviation [71]: { l = x1 s γ1 exp (πm c πc 1 ) ω } 1 exp{ξ(x2 s kt )2 } { + x2 s γ2 exp (πm c πc 2 ) ω } 2 exp{ξ(x1 s kt )2 } (13) where x1 s and xs 2 denote the mole fraction of components 1 and 2 in the two component monolayer, respectively, and π1 c and πc 2 the corresponding collapse pressures of components 1 and 2, πm c the collapse pressure of the two-component monolayer at given composition of x1 s and xs 2, ω 1 and ω 2 the corresponding limiting molecular surface area at the collapse points, γ 1 and γ 2 the surface activity coefficients at the collapse point, ξ the interaction parameter, and kt is the product of the Boltzmann constant and the Kelvin temperature. The solid curve is made coincident with the experimental values by adjusting ξ. In these figures, M indicates a two-component mixed monolayer formed by DSG-1 and DPPC species, while Bulk denotes a solid phase of DSG-1 and DPPC ( bulk phase may be called solid phase ). The collapse pressure π c determined at each mole fraction is indicated by filled circles, where the dotted line shows the case where the interaction parameter (ξ) is zero. From the equation, the interaction parameter ξ for DSG- 1/DPPC system is This means that there is mutual interaction between two components in the two-component monolayer that is stronger than the mean of the interactions between pure component molecules themselves. As a result, they are completely mixing. The interaction energy ε can be calculated the following equation: ε = ξrt 6 (14) and the ε value is 574 J mol 1 for DSG-1/DPPC system. As the result, this system is the positive azeotropic type. In DSG-1/DMPE system, the phase diagram was also shown in Fig. 10b, which is the same procedure as the above case. The transition pressure (π eq ) is about 14.1 mn m 1 for pure DMPE, which is a bit larger than that of pure DPPC (about 11.8 mn m 1 ). Similar to the previous case, π A isotherms display π eq that change almost linearly with increasing X DSG-1. This system also produced a negative interaction parameter (ξ = 0.46 and 190 J mol 1 ). Thus, DSG-1 and DMPE are miscible in the expanded state as well as in the condensed state and the system behaves as a homogeneous mixture. Finally, the phase diagram in DSG-1/Ch system is constructed as shown in Fig. 10c, and the coexistence phase boundary between the monolayer phase and the bulk phase can be theoretically also simulated by the Joos equation. Just like the former two systems, the negative interaction parameter ξ was obtained (ξ = 0.96). Also the interaction energy is ε = 397 J mol 1. This means that the mutual interaction between two components

14 70 K. Hoda et al. / Colloids and Surfaces B: Biointerfaces 52 (2006) Fig. 11. Fluorescence micrographs of (a) DSG-1/DPPC binary monolayer (X DSG-1 = 0, 0.1, and 0.3), (b) DSG-1/DMPE mixed monolayer (X DSG-1 = 0, 0.1, and 0.3) and (c) DSG-1/Ch binary monolayer (X DSG-1 = 0, 0.1, and 0.3) observed on 0.15 M NaCl at K. The monolayer doped 1 mol% of fluorescent probe. The number in these images indicates the surface pressure and the LC domain percentage in the micrograph. Scale bar represents 100 m.

15 K. Hoda et al. / Colloids and Surfaces B: Biointerfaces 52 (2006) Fig. 11. (Continued ). is stronger than that between pure component molecules themselves for DSG-1. Consequently three binary systems are miscible each other and they show the positive azeotropic type in their phase diagrams Fluorescence microscopy In order to interpret the phase behavior on the π A isotherms, we investigated the morphology by fluorescence microscopy, which provides a direct image of the monolayers. The contrast is due to a difference in dye solubility between disordered (or LE) and ordered phases (or LC). Fluorescence micrographs (FMs) of each pure component and their combinations spread on 0.15 M NaCl at K are shown in Fig. 11. Essentially two phases of bright and dark contrast were observed for pure DPPC, corresponding to the LE matrix and LC domains, respectively [72 74]. It is known that the proportion of dark LC phase increases at the expense of the bright LE phase with surface pressure increase. The DPPC monolayer on the 0.15 M NaCl solution develops LC domain structures in a qualitatively similar fashion as reported [75,76]. Angular and grain-like domains first appeared at its transition pressure ( 11.8 mn m 1 ). Upon compression they grew to form distorted star-shaped domains with a blurred periphery that progressively formed a network. Above 21 mn m 1, the images start loosing their crispness and the visual impression becomes a progressive blurring of the domain boundaries. This blurring may be caused by the dissolution of the dye into the dye-depleted regions of the monolayer after the phase transition has been completed, i.e. when both the dye-enriched and the dye-depleted regions have become the same phase. Above 40 mn m 1,it is likely that some of probes were inserted in the dark phase areas and the intensity of the probe in the bright (fluid) domains decreased, suggesting that self-quenching has occurred because of probe molecules coming in close contact with one another. The percentage is the ratio of LC domain in each image to the total area as shown in Fig. 11a. Monolayers of DSG-1 used in this study did not form the LC domains in the monolayer state. As the result, FM showed the homogeneous LE image. As mentioned in the former section, DSG-1 was a molecular species whose hydrophobic parts were too bulky to be closely packed together. Independence of the bright patterns on surface pressure in the FM image was the evidence of forming such LE films. Firstly, the mole fraction dependence of the transition pressure for DSG-1/DPPC system is observed on the FM images in Fig. 11a. At low surface pressures (π < π eq ), DSG-1/DPPC system was uniformly fluoresced, showing apparently homogeneous LE phase without liquid-condensed (LC) phase of dark domains. Increasing the surface pressure, LC domains appear from X DSG-1 = 0.05 to 0.5. In each case, the LE/LC coexistence region is observed, and the transition pressure (π eq ) is higher than that of pure DPPC. This suggests that the observed dark domains in these figures would represent a condensed DPPCenriched phase. The increment in concentration of DSG-1 (at X DSG-1 = 0.3) makes the LE phase larger, and the dark domains of DPPC became small. At high surface pressure, the growth of

16 72 K. Hoda et al. / Colloids and Surfaces B: Biointerfaces 52 (2006) circular LC domains persisted in binary DSG-1/DPPC monolayers. As a result, this system is completely miscible each other. For pure DMPE (Fig. 11b), domain nucleation starts to occur at the kink in the π A isotherm at 14.1 mn m 1 and their domains were observed at the coexistence state of both LE phase and LC phases. With increasing surface pressure from 15 to 18 mn m 1, the percentage of LC phase in each image increased and complete LC domain appeared at 22 mn m 1. The FM images of binary DSG-1/DMPE system were shown in Fig. 11b. Similar to DSG-1/DPPC system, LC domains appear from X DSG-1 = 0.05 to 0.5. FM images of pure Ch (Fig. 11c) showed some large domains at zero surface pressure, where the dark regions were the gaseous phase, because fluorescence was completely quenched. The medium gray regions were the LC phase. In binary DSG-1/Ch monolayer, the morphologies of bright image do not change up to 50 mn m 1 in the FM images. The addition of small amount of DSG-1 to Ch induced the phase separation at zero surface pressure. Morphology of the separated phase was changed to homogeneous dark images by the film compression up to high surface pressure. It suggests that DSG-1 become miscible with Ch Atomic force microscopy Atomic force microscopy (AFM) is a surface imaging technique with a nanometer-scale resolution [77,78]. AFM for this study provided both topography and phase contrast (topology) images. The topography image reflects the sample topography, while the phase contrast image, which is originated from the energy loss of the oscillating AFM tip, shows the chemical structures of heterogeneous samples. Also on surfaces with local variations of mechanical properties such as biological samples, the AFM phase image provides the best contrast of fine morphological and nano-structural features. To investigate the effect of adding a small amount of DSG-1 molecules to DPPC and to Fig. 12. AFM images in both tapping mode and phase mode. (a) Topography and (b) corresponding topology of the pure DSG-1 monolayer transferred onto mica at 30 mn m 1 at the scan area of 500 nm 500 nm. (c) Topography and (d) corresponding topology for DSG-1/DPPC on X DSG-1 = 0.1 (5 m 5 m) and the zoomed-in images of the topography (c) and topology (d) were shown in (e) and (f) at the scan area of 500 nm 500 nm. (g) Topography and (h) corresponding topology for DSG-1/DMPE on X DSG-1 = 0.1 (300 nm 300 nm), respectively.

17 K. Hoda et al. / Colloids and Surfaces B: Biointerfaces 52 (2006) Fig. 12. (Continued ). clarify the miscibility between them, AFM images of monolayer for DSG-1 and two binary systems transferred to mica substrates at 30 mn m 1 on 0.15 M NaCl are shown in Fig. 12. The surface pressure value of 30 mn m 1 is reflected in biological membranes as reported papers [79]; that is, comparing the actions of phospholipases on erythrocyte membranes and on lipid monolayers, Demel et al. concluded that the pressure in the outer monolayer of the erythrocyte membrane is equivalent to surface pressure of mn m 1 of a monolayer. We have employed AFM in the tapping mode which has various advantages for imaging soft matter and biological sample compared with the contact mode [79,80]. Only single component of DPPC [81] and DMPE showed a uniformed pattern in the AFM image, while AFM images (500 nm 500 nm) of both topography (a) and topology (b) for DSG-1 were a little bit irregular pattern due to molecular species (Fig. 12). Fig. 12c and d shows AFM images (5 m 5 m) of a LB films containing small amounts of DSG-1 (X DSG-1 = 0.1) transferred onto mica at 30 mn m 1. The bright pattern of DSG-1 molecules dispersed randomly in dark DPPC surface in the topography image, and the cross-sectional profile indicates that the average height is about 1.2 nm higher than the surrounding, which almost coincided with the predicted chain length difference between the longest acyl chain (see Table 1) of DSG-1 monolayer and DPPC. Because non-tilted DPPC molecules is estimated about 2.4 nm in height without polar head group at the interface, while the longest acyl chain of DSG-1 is about 3.6 nm height by a computer simulation (CS ChemOffice UltraTM 5.0). The phase contrast image (Fig. 12d) shows more clearly the morphological character than the corresponding topography due to the difference in the surface physico-chemical property between DPPC and DSG-1. The matters above indicate that the patterns of Fig. 12c and d are considered to be a micro-phase separation. A similar phase-separation was found in AFM image for the ganglioside GM1/DPPC and GM3/DPPC monolayer [81 83]. The behavior of AFM images did not agreed with that of FM images in terms of the miscibility between these two components, as