- Email: [email protected]
water interface
Colloids and Surfaces B: Biointerfaces 41 (2005) 63–72
Interactions between the ganglioside GM1 and hexadecylphosphocholine (miltefosine) in monolayers at the air/water interface Isabel Rey G´omez-Serranillosa , Jos´e Mi˜nones Jr.a , Patrycja Dynarowicz-Ł˛atkab,∗ , Eduardo Iribarnegaraya , Matilde Casasa a
b
Department of Physical Chemistry, Faculty of Pharmacy, University of Santiago de Compostela, Spain Jagiellonian University, Faculty of Chemistry, Department of General Chemistry, Ingardena 3, 30-060 Krak´ow, Poland Accepted 1 November 2004 Available online 16 December 2004
Abstract The ganglioside, GM1 , was studied as Langmuir monolayers at the air/water interface with surface pressure-area measurements in addition to Brewster angle microscopy. A characteristic plateau transition, observed on aqueous subphases of pH 2 and 6, 20 ◦ C, at the surface pressure of ca. 20 mN/m, was attributed to the reorientation of GM1 polar group upon film compression. This transition was found to disappear at alkaline subphases (pH 10) due to the hydration of fully ionized polar group, hindering its reorientation. The interactions between GM1 and hexadecylphosphocholine (miltefosine) were investigated in mixed monolayers and analyzed with the mean molecular areas, excess areas of mixing and the excess free energy of mixing versus film composition plots. The monolayers stability, quantified by the collapse pressure values, as well as the strength of interaction was found to diminish in the following order: pH 6 > pH 2 > pH 10. The strongest interaction occurs for mixed films of miltefosine molar fraction, XM = 0.7–0.8, especially at low pressure region, and are explained as being due to the surface complex formation of 3:1 or 4:1 (miltefosine:ganglioside) stoichiometry (XM = 0.75 or 0.8, respectively). © 2004 Elsevier B.V. All rights reserved. Keywords: Langmuir monolayers; Hexadecylphosphocholine; Ganglioside GM1 ; Interactions; Air/water interface
1. Introduction Hexadecylphosphocholine (miltefosine) is a representative of a new generation of anti-cancer drugs based on a phospholipid-like structure [1]. Its pharmacological action is believed to be related to the disturbance of membrane function by this “false” phospholipid-analogue, which results in the inhibition of cells proliferation [2,3]. For a better understanding of the mode of miltefosine action, we have undertaken studies on its interaction with three kinds of membrane lipids (sterols, phospholipids and glycosphingolipids), which are representative for each side of the cellular membrane. Since both, miltefosine itself and the membrane lipids ∗ Corresponding author. Tel.: +48 12 6336377x2236; fax: +48 12 6340515. E-mail address: [email protected] (P. Dynarowicz-Ł˛atka).
0927-7765/$ – see front matter © 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.colsurfb.2004.11.004
are capable of stable monolayer formation at the air/water interface [4–6], we have successfully applied the Langmuir monolayer technique [7] to study the intermolecular interactions. One of our previous works [8] describes miltefosine interactions with a typical membrane sterol – cholesterol – which is placed in-between both sides of the biomembrane, while our recent paper [9] presents the investigations on the interaction with a typical phospholipid forming the inner part of the membrane [10], namely OPPE (-oleoyl-␥-palmitoyl l-␣-phosphatidylethanolamine). This work is aimed at investigating the interactions with a ganglioside (GM1 ), which is found exclusively in the outer half of the bilayer [1]. Gangliosides belong to the group of complex lipids (socalled glycolipids) containing a carbohydrate residue. All the glycolipids in animal cells are glycosphingolipids (GSLs). Their apolar (aliphatic) part (so-called ceramide) is constituted by sphingosine (an 18-carbon atom, dihydric alcohol,
64
I.R. G´omez-Serranillos et al. / Colloids and Surfaces B: Biointerfaces 41 (2005) 63–72
which contains an amino group at C17 ) and a long-chain fatty acid in amide linkage (Scheme 1). The bulky polar part, on the other hand, is formed by carbohydrate moieties attached to hydroxymethyl group of ceramide. Such compounds are known as cerebrosides. The gangliosides contain, in addition to sugar residues, N-acetylneuraminic acid (NANA), also called sialic acid. Due to their structure, the gangliosides display amphiphilic properties, and are negatively charged at neutral pH. According to the number and position of sugar residues and sialic acid, different types of gangliosides are identified, according to a specific nomenclature. For example, letter G stands for a ganglioside in general; the letters M, D, T, etc. identify the number of sialic acid residues (1, 2, 3, etc.), while the subscript 1, 2, etc. is an arbitrary number assigned according to the chromatographic mobility. In this work, we have investigated the ganglioside GM1 , the chemical structure of which is shown in Scheme 1. The gangliosides, first described by Klenk [11] in 1935 and Blix [12] in 1938, are inserted in the extracellular part of synaptic nerve cell (neuron) membranes of vertebrates, where they constitute about 6% of the total lipid mass [13]. They are involved in intermolecular interactions, differentiation, transformation, ontogenesis and apoptosis [14–17]. Some of them are highly antigenic [18] while others – like GM1 – act as a cell-surface receptor for the bacterial cholera toxin [19]. Currently, gangliosides are investigated for their efficiency to develop an anti-cancer vaccine, especially for epithelial cancer of breast.
2. Experimental Both miltefosine (99%) and the investigated ganglioside (monosialoganglioside palmitoyl-GM1 , 95%) were supplied by Sigma. The compounds were stored in a refrigerator without the access of light. Spreading solutions were prepared by dissolving each compound in a 4:1 mixture of chloroform:alcohol (absolute ethanol for miltefosine; methanol for GM1 ) with a typical concentration of 0.2–0.3 mg/ml. Mixed
Scheme 1. Chemical structure of GM1 . Glc, Glu and GalNAc denote galactose, glucose and N-acetylglucosamine, respectively.
solutions were prepared from respective stock solutions of both compounds, and were dropped onto the subphase with a Microman Gilson microsyringe, precise to ±0.2 l. The number of molecules deposited on the water subphase was kept constant in all experiments (3.2 × 1016 molecules). Ultrapure water (produced by a Nanopure water purification system coupled to a Milli-Q water purification system, resistivity = 18.2 M cm) was used as a subphase. The subphase temperature was controlled to within 0.1 ◦ C by a circulating water system from Haake. HCl or NaOH was added into water to adjust the pH value of the subphase. Experiments were carried out with a NIMA 601 (Coventry, U.K.) trough (total area = 525 cm2 ), placed on an anti-vibration table. Surface pressure was measured with the accuracy of ±0.1 mN/m using a Wilhelmy plate made from ashless chromatography paper (Whatman Chr1). After spreading, monolayers were left for 10 min to ensure solvent evaporation, and afterwards the compression was initiated with a barrier speed of 50 cm2 /min ˚ 2 /(molecule min)). (15.6 A Brewster angle microscopy images and ellipsometric measurements were performed with BAM-ELLI 2000 (NFT, Germany) equipped with a 30 mW laser emitting p-polarized light at 690 nm wavelength which was reflected off the air/water interface at approximately 53.1◦ (Brewster angle). In measuring the relative reflectivity of the film, a camera calibration was necessary as described elsewhere [20]. At Brewster angle: I = |Rp|2 = Cd 2
(1)
where I is the relative reflectivity, C is a constant, d is the film thickness, and Rp is the p-component of the light. The lateral resolution of the microscope was 2 m, and the images were digitized in order to obtain high quality BAM images.
3. Results 3.1. Pure monolayers A thorough Langmuir monolayer characteristic of miltefosine is described in our former paper [4]. Miltefosine was found to form liquid-expanded monolayers that collapse at ca. 40 mN/m. The film stability was found to be satisfactory [4], and the monolayer behavior was hardly influenced by such experimental conditions as the compression speed, number of deposited molecules, subphase temperature (within the range of 10–25 ◦ C) and subphase pH (within the range of 2–12). The surface pressure–area (π–A) isotherms of the investigated ganglioside at 20 ◦ C spread at aqueous subphases of pH 2, 6 and 11 are shown in Fig. 1. At pH 6, the monolayer shows a gas-liquid expanded transition region at molec˚ 2 /molecule, where the surface ular areas larger than 120 A pressure is maintained constant at ca. 0.5 mN/m. At areas ˚ 2 /molecule and at low surface pressures, the below 120 A
I.R. G´omez-Serranillos et al. / Colloids and Surfaces B: Biointerfaces 41 (2005) 63–72
Fig. 1. Surface pressure (π)–area (A) isotherm of GM1 spread at 20 ◦ C on aqueous subphases of pH 2, 6 and 10. Inset: compressional modulus (Cs −1 )–surface pressure (π) dependencies.
monolayer shows a liquid-expanded state, characterized by a compression modulus (Cs −1 ) value of 36 mN/m and a limiting area (determined by extrapolating of the linear fragment of the isotherm preceding the collapse to π = 0) of about ˚ 2 /molecule. Luckham et al. [21] reported a similar 101.7 A value for the limiting area of GM1 spread on water of pH 5.6, ˚ 2 /molecule) was obtained on however, a lower value (88.6 A 145 mM aqueous NaCl subphase, and was attributed to the screening of the negative charges on the ganglioside headgroups (due to the ionized sialic acid at neutral pH [22]) by the presence of salt, thus reducing the electrostatic repulsion between the individual molecules in the monolayer. At ˚ 2 /molecule, a phase tranapproximately 20 mN/m and 74 A sition shoulder is reached and upon further compression, the monolayer shows a liquid-condensed state of low compressibility (Cs −1 = 154 mN/m). Finally, the film collapse occurs at 65.9 mN/m, proving high film stability. Fidelio et al. [23] obtained for GM1 monolayers a collapse pressure value of about 60 mN/m at lower temperatures. At acidic pH (pH = 2), the monolayer occupies smaller area at the interface as compared to pH 6, and the limiting molecular areas in the liquid-expanded and liquid-condensed ˚ 2 /molecule and 59.4 A ˚ 2 /molecule, respecphase are 90.3 A tively. The phase transition appears at ca. 20 mN/m, similarly as it was observed at pH 6. The film collapse is reached at ˚ 2 /molecule at the same surface pressure (65.9 mN/m) 38.8 A as at pH 6. Luckham et al. [21] reported a similar value ˚ 2 /molecule) for the molecular area at the collapse. (38 A At pH 10, the monolayer is most expanded as compared to all the investigated subphases, without visible plateau transition, and the collapse occurs at much lower surface pressure (ca. 51 mN/m). The Cs −1 –π curves, corresponding to the isotherms recorded at the three studied pHs, are shown in the inset of Fig. 1. These dependencies clearly indicate the existence of a
65
liquid expanded–liquid condensed (LE–LC) phase transition at pH 2 and 6, which appears as a pronounced minimum that separates the expanded and condensed phase at the surface pressure of about 20 mN/m. At basic subphase pH, there is no clear minimum, but rather a plateau region of Cs −1 values ranging from 60 to 70 mN/m, which are typical for the liquid-expanded state of a monolayer. Brewster angle microscopy has been applied in addition to the surface pressure–area isotherm to characterize the monolayer from GM1 . The images, which are shown in Fig. 2, were recorded at pH 6, 20 ◦ C. At low surface pressure region, where the monolayer is in a liquid-expanded state, the image is completely homogeneous (Fig. 2A, taken at π = 4 mN/m) and does not change upon film compression until post-transition region. At higher surface pressures (above 25 mN/m), small bright domains appear (Fig. 2 B, taken at 35 mN/m), which correspond to the condensed phase. These domains fuse together, forming compact agglomerates, seen as stripes, at surface pressures corresponding to the film collapse (Fig. 2C). The relative reflectivity (I) versus time (t) curve is shown together with the π–t isotherm in Fig. 2. Along the liquidexpanded phase and the transition region, the relative reflectivity increases monotonically upon film compression, similarly to surface pressure, until the value of 1.1 × 10−5 (arbitrary units) is reached at the end of this phase. The increase in the relative reflectivity from the start of compression (I = 1 × 10−6 ) is thus 11-fold, which corresponds to the increase in monolayer thickness by 3.5 times. Along the liquidcondensed state, above 25 mN/m, the relative reflectivity is maintained practically constant, indicating that the monolayer thickness is no longer changing. Only at the monolayer collapse, a small increase in reflectivity signal can be observed. An important feature of the I versus t dependence is the absence of noise peaks in the liquid-expanded region, which accounts for the completely homogeneous BAM images observed in this region. Along the liquid-condensed phase, on the other hand, some noises can be noticed, which are due to the presence of small condensed-phase domains, visible in the image B, Fig. 2. The increase in subphase temperature (Fig. 3) causes film expansion and shifts the LE–LC transition to higher surface pressures. These results are in good agreement with those obtained by Luckham et al. [21]. At 26 ◦ C they reported the transition to occur at the pressure of 25 mN/m ˚ 2 /molecule), while the results presented herein, cor(at 71 A responding to 25 ◦ C, show the transition at π = 26.3 mN/m ˚ 2 /molecule. Film expansion caused by corresponding to 73 A the temperature increase is expected due to the weakening of intermolecular forces between apolar tails of neighboring molecules in the monolayer. However, an anomalous tendency was reported by Beitinger et al. [24] for a monolayer spread on 5 mM buffer solution of pH 7.4, where the temperature increase provoked film condensation to smaller molecular areas. The authors, however, did not comment on such a peculiar behavior.
66
I.R. G´omez-Serranillos et al. / Colloids and Surfaces B: Biointerfaces 41 (2005) 63–72
Fig. 2. Time evolution of the relative reflectivity, I () and surface pressure, π () of GM1 monolayer spread on water, pH 6 at 20 ◦ C, together with BAM images taken at the points designated on the isotherm.
The compression modulus curves (inset of Fig. 3) confirm film expansion with the temperature increase. Firstly, the position of the minimum (M, indicated by arrows), which is referred to the LE–LC transition, undergoes displacement towards higher surface pressures with temperature. Secondly, with the increase of temperature, Cs −1 values at the first maximum increase, while those corresponding to the second max-
imum diminish. This finally leads to the formation of one monolayer phase (one maximum) at high subphase temperature. This means that the two physical states of the monolayer, which can be clearly distinguished at low temperature, fuse into one expanded state. The collapse pressure varies with temperature (Fig. 3), i.e. within the range of 10–30 ◦ C, the values of πc are comprised between 62 and 66 mN/m. These values are higher than those obtained by Fidelio et al. [23] (45 mN/m for GM1 monolayer spread at 30 ◦ C, and 55 mN/m for temperature interval between 20 and 25 ◦ C). This can be due to different experimental conditions, i.e. in ref. [23] the monolayer was spread on 145 mM aqueous NaCl solution and was compressed with lower speed (ca. 7 cm2 /min.). 3.2. Mixed monolayers of miltefosine and GM1
Fig. 3. Surface pressure (π)–area (A) isotherm of GM1 spread on aqueous subphases of pH 6 at different temperatures. Inset: compression modulus (Cs −1 )–surface pressure (π) dependencies.
The surface pressure–area isotherms for mixed miltefosine-GM1 monolayers, spread at different subphase pHs (2, 6 and 10) at 20 ◦ C, are shown in Fig. 4 A–C. As it can be seen, the behavior of mixed films is generally similar at pHs 2 and 6, i.e. mixtures richer in ganglioside (XM = 0.1–0.3) exhibit similar behavior to pure ganglioside, showing a transition (LE–LC) at nearly identical surface pressure as pure GM1 (ca. 20 mN/m; see Table 1), which, however, becomes progressively less pronounced as miltefosine proportion in the monolayer increases, and
I.R. G´omez-Serranillos et al. / Colloids and Surfaces B: Biointerfaces 41 (2005) 63–72
67
Fig. 4. Surface pressure (π)–area (A) isotherms of GM1 , miltefosine and their mixtures spread on aqueous subphase of pH 2 (A), pH 6 (B) and pH 10 (C) at 20 ◦ C.
finally disappears for a mixed monolayer of XM = 0.5, which shows an intermediate behavior between pure miltefosine and ganglioside, i.e. is expanded at surface pressures below 20 mN/m, and condensed at higher pressures, without a plateau transition. Mixed monolayers rich in miltefosine (XM = 0.9) look almost identical as pure miltefosine, however, they are slightly shifted towards smaller areas and collapse at a higher surface pressures. At basic pH, the GM1 monolayer is expanded, without visible phase transition, and mixed films are of expanded type as well, within the whole range of composition. When the milte-
fosine content exceeds 0.5 mole fraction, the mean molecular area occupied by film molecules is smaller as compared to both pure components. The value of the collapse pressure is related to the monolayer stability, i.e. the higher it is, the more stable the monolayer is. As it can be seen in Table 1, the collapse pressure changes with the incorporation of miltefosine into GM1 monolayer at different subphase pHs. The trend observed at pH 2 and 6 is similar; the stability of mixed films increases upon increasing the proportion of miltefosine, reaching a maximum (corresponding to XM = 0.5), and decreases after-
Table 1 Compression modulus (Cs −1 , mN/m), transition surface pressure (πt , mN/m) and collapse pressure (πc , mN/m) values for monolayers of GM1 , miltefosine and their mixtures at different subphase pH at 20 ◦ C Monolayer composition
GM1 XM = 0.1 XM = 0.3 XM = 0.5 XM = 0.7 XM = 0.9 Miltefosine
pH 2 −1
pH 6 −1
−1
pH 10 −1
Cs 1st maximum
Cs 2nd maximum
πt
πc
Cs 1st maximum
Cs 2nd maximum
πt
πc
Cs −1 1st maximum
πc
51 46 47 42 56 47 43
154 166 132 – – – –
20.0 17.0 17.1 – – – –
65.9 67.1 68.5 71.7 65.0 39.5 36.5
51 47 47 42 56 48 47
154 166 132 – – – –
20.0 16.7 17.1 – – – –
66.4 67.0 68.4 71.9 72.8 39.6 39.5
– 59 59 56 65 50 43
51.5 55.6 61.5 60.8 59.3 44.4 37.5
68
I.R. G´omez-Serranillos et al. / Colloids and Surfaces B: Biointerfaces 41 (2005) 63–72
wards. The monolayer stability at pH 10 is lower as compared to acidic and neutral subphases, no matter what the film composition is. This is quite logical since the monolayer of pure ganglioside also collapses at much lower pressure at pH 10 as compared to other investigated subphase pHs. Analyzing the collapse pressure values it may be concluded that the stability of mixed films obey the following order: pH 6 > pH 2 > pH 10. The surface pressure–area isotherms shown in Fig. 4 clearly indicate condensation of monolayers at all the investigated pH values and surface pressure regions. The extent of contraction of mixed monolayers in respect to pure components is a measure of interaction between film-forming molecules, and can be visualized on the diagrams of the mean molecular area versus film composition (A12 = f(XM )). Such dependencies are presented in Fig. 5 for different subphase pHs in both low (5 mN/m) and high (20 mN/m, 35 mN/m) surface pressures. The dotted lines illustrate the additive re-
lationship for the ideal system: A12 = A1 X1 + A2 X2 (A12 is the mean molecular area of mixed monolayer, X1 and X2 denote the mole fractions of components 1 and 2; A1 and A2 stand for molecular areas of pure components at the same surface pressure as A12 was determined). Within the whole range of the investigated pHs and mole fraction, negative deviations from the ideality can be observed both in low and high surface pressure regions. As regards the magnitude of the discussed deviations, stronger occur at lower surface pressures (5 mN/m). Moreover, when the subphase pH is 6, the films condensation is stronger as compared to pH 2 or 10. Thus it may be concluded that the interactions between miltefosine and GM1 , responsible for the observed deviations, diminish in the following order: pH 6 > pH 2 > pH 10. One may quantify the interactions by the excess functions, for example the excess areas of mixing (Aexc ) and excess free energy of mixing ( Gexc ). Values of Aexc can be calculated
Fig. 5. Dependence of the mean molecular area (A12 ) on the molar fraction of miltefosine (XM ) in mixed films with GM1 , spread at aqueous subphases of pH 2 (A), pH 6 (B) and pH 10 (C) at 20 ◦ C.
I.R. G´omez-Serranillos et al. / Colloids and Surfaces B: Biointerfaces 41 (2005) 63–72
by comparing the area of mixed monolayer (A12 ) with that for unmixed, one-component monolayers (A1 ; A2 ), according to the equation: Aexc = A12 − (A1 X1 + A2 X2 ). If the mixture is ideal, or its components are immiscible, Aexc is zero and the dependence Aexc = f(Xi ) is linear. Deviations from these conditions indicate miscibility and non-ideality. From Fig. 6A–C, which presents the results of Aexc calculated for the investigated system, one may distinguish more clearly the differences in mixed films behavior at various pHs. At low pressures (5 mN/m), the Aexc –XM dependencies exhibit a clear minimum for XM = 0.8 (pH 2) and 0.7 (pH 6 and 10). While on subphases of pH 2 and 10 at high surface pressures (20 and 30 mN/m) the extent of contraction is similar for all mixtures within the range of XM = 0.1–0.8, on pH 6 mixtures of XM = 0.6–0.7 show the highest condensation, which is visible as a pronounced minimum. It is thus clear that on pH 6, the values of excess area of mixing are
69
most negative, independently on pressure region and film composition. A more detailed analysis of the intermolecular interactions is based on the excess free energy of mixing ( Gexc ) calculations, which can be determined from the following relation derived by Goodrich [25] and Pagano and Gershfeld [26]: π Gexc = N (A12 − X1 A1 − X2 A2 )dπ (2) 0
Fig. 7A–C presents Gexc values at three different pHs at 20 ◦ C as a function of miltefosine composition in mixed monolayers. The results are very similar to those discussed above, i.e. on both pH 2 and 10 the curves Gexc –XM show very little differences. On the contrary, on pH 6 a clear minimum is seen at XM ∼ 0.8 (at high surface pressures) and at XM = 0.9 (for π = 5 mN/m). It may thus be concluded
Fig. 6. Plots of the excess area of mixing (Aexc ) vs. molar fraction of miltefosine (XM ) in mixed films with GM1 , spread at aqueous subphase of pH 2 (A), pH 6 (B) and pH 10 (C) at 20 ◦ C.
70
I.R. G´omez-Serranillos et al. / Colloids and Surfaces B: Biointerfaces 41 (2005) 63–72
Fig. 7. Excess free energy ( Gexc ) as a function of mole fraction of miltefosine (XM ) in mixed films with GM1 , spread at aqueous subphase of pH 2 (A), pH 6 (B) and pH 10 (C) at 20 ◦ C.
that there are attractive interactions between miltefosine and GM1 , which are strongest at pH 6, while on pH 2 and 10 they are nearly of the same order.
4. Discussion The ganglioside, GM1 , has already been investigated in monolayers at the air/water interface by other authors, however, the results are rather difficult to compare since the experimental conditions were different as compared to those applied in the present work. Namely, GM1 was studied on a big variety of different subphases, containing various ions, for example on 145 mM aqueous NaCl [27], 5 mM triethanolamine/HCl buffer of pH 7.4 [15,16,24], 3.5 mM phosphate buffer [17], etc. Crowe et al. [28] published values of molecular areas at 20 and 30 mN/m on aqueous sub-
phases, however, did not show the complete π–A isotherm, while Parker et al. [29] presented the isotherm between 0 and 30 mN/m, but did not specified the subphase temperature. Only the results presented by Luckham et al. [21] can be useful for comparison with the present data, although the subphase pH was different (5.6 and 1). Nevertheless, the reported values are in a good agreement with those presented herein. The phase transition (LE–LC), which appears on the π/A isotherm of GM1 as an inflection at π = 20 mN/m, was attributed by Luckham et al. [21] to the electrostatic interactions between charged groups of sialic acid (which form, together with sugar molecules, polar headgroup of the ganglioside molecule), causing the increase in surface pressure observed above the transition in the condensed state of the monolayer. This interpretation, of course, does not explain the existence of interactions at pH 1, where sialic acid molecules
I.R. G´omez-Serranillos et al. / Colloids and Surfaces B: Biointerfaces 41 (2005) 63–72
are almost completely protonated (pKa = 2.6) [17,22]. Moreover, the electrostatic interactions are expected to be stronger at pH 10, where sialic acid molecules are completely ionized, and therefore if the interpretation given by the authors [21] was correct, the phase transition in the π–A isotherms should be very much pronounced. In fact, the results shown in Fig. 1 show the disappearance of this transition on alkaline subphases, where the monolayer becomes liquid-expanded along the full compression. Other authors [15,24] interpret the existence of the phase transition as being due to the reorientation of bulky polar group of GM1 , based on the fact that the monolayer properties of gangliosides, in particular mean molecular area, depend – nearly solely – on the respective polar groups, which was shown by Maggio et al. [27]. In view of this fact, it is logical to attribute the phase transition to the change of polar group conformation in the subphase. This change in orientation does not affect the arrangement of apolar tails of the molecule, since the size of the polar group is sufficiently large to hinder the intermolecular attractions between the tails, preventing the formation of molecular aggregates, but causing, on the other hand, strong monolayer resistance to the compression as the molecules become closely packed. This explains the absence of any compact domains in BAM images, and the decrease in monolayer compressibility (increase of its rigidity) above the phase transition. The absence of any inflection (corresponding to a transition) in the π–A isotherms at 20 mN/m for monolayers spread on aqueous subphase of pH 10, can be due to the hydration of polar groups, caused by their complete ionization. It may be assumed that water molecules surrounding polar groups hinder, or make impossible, their reorientation upon compression. In consequence, the monolayer on alkaline subphases is expanded, without any transition, throughout the full compression. Upon compression, water molecules solvating polar groups of the ganglioside are squeezed out from the monolayer into bulk phase, which causes the decrease of the mean molecular area accompanied by an increase in surface pressure. Such a behavior resembles a soaked sponge, which is squeezed out from water. A lower collapse pressure of GM1 monolayer on pH 10 as compared to other investigated subphase pHs can be due to the fact that presence of hydration shell surrounding polar groups increases the distances between hydrocarbon tails of neighboring molecules, lessening in this way van der Waals attraction forces and causing the decrease of film stability. As the mixed monolayers are concerned, their higher collapse surface pressure values evidence for the presence of stronger attractive forces between molecules in the mixed film as compared to monolayers from pure components. The difference between experimental πc value and those calculated on the basis of the additivity rule is a measure of the strength of interactions between film components, which are
71
found to diminish in the following order: pH 6 > pH 2 > pH 10. Similar trend is observed when analyzing the results of the mean molecular areas (Fig. 5), as well as the free energy of mixing values (Fig. 7). Since miltefosine was found to be insensitive to the variation in subphase pH [4], it is evident that the ganglioside is responsible for the observed effect. In this sense, one may expect that at pH 6 the attractive electrostatic interactions which are established between negatively charged sialic acid of the ganglioside polar group and positively charged trimethylammonium groups of miltefosine are responsible (in addition to the van der Waals attraction forces between apolar tails) for the considerable increase in mixed films stability in relation to monolayers from pure components. Possible repulsive forces between the ganglioside and the phosphate group of miltefosine would give less contribution because of the strong hydration of the phosphate ion. At pH 2, when sialic acid is uncharged, the electrostatic interactions in the mixed film cannot be established, however, other kinds of interaction can occur, like for example ion–dipole interactions between the polar groups of miltefosine (positively charged at this pH) and those of the ganglioside. Differences observed at pH 6 and 2 are due to the different energy of the ion-zwitterion (pH 6) and ion–dipole interactions. At pH 10, both polar groups of ganglioside and miltefosine are negatively charged, and therefore repulsive electrostatic interactions could be expected, resulting in the decreased stability of the mixed monolayer. Therefore, the deviations from ideal behavior should be positive, contrary to the experimental results. This can be explained by the fact that at this subphase pH the attractive van der Waals forces between hydrocarbon tails predominate over the electrostatic repulsions between polar groups, which are significantly weakened by the hydrated water molecules. Moreover, it should be mentioned that between the monosaccharides (forming in part the bulky polar group of the ganglioside) and the polar group of miltefosine, various types of attractive interactions can be established, for example dipole–dipole, ion–dipole, hydrogen bonds, etc., which contribute significantly to the stability of mixed monolayers. Although these interactions also occur at other subphase pHs, however, especially at pH 10 their contribution is prevailing. A bigger polar group of the ganglioside, as compared to miltefosine, allows us to understand why the strongest contraction occurs for mixed films of XM = 0.7–0.8, especially at low surface pressure region, where the monolayer is expanded. In this situation, the molecular area of the monolayer is determined by the bulky polar group of the ganglioside, and the molecules of miltefosine can penetrate between neighboring molecules of GM1 , forming surface complexes of 3:1 or 4:1 (miltefosine:ganglioside) stoichiometry (XM = 0.75 or 0.8, respectively). Therefore, mixtures of high proportion of miltefosine are all involved in the stable complex formation, and none of them remain free, which accounts the observed highest interactions and stability.
72
I.R. G´omez-Serranillos et al. / Colloids and Surfaces B: Biointerfaces 41 (2005) 63–72
Acknowledgements This financial support of Xunta de Galicia (project PGIDT99PXI20302B) is gratefully acknowledged. References [1] C. Unger, E.A.M. Fleer, D.J. Kim, A. Breiser, P. Hilgard, J. Engel, G.A. Nagel, G.H. Eibl, Acta Oncol. 28 (1989) 213. [2] K. Boggs, C.O. Rock, S. Jackowski, Biochim. Biophys. Acta 1389 (1998) 1. [3] Y.Z. Wang, Y.B. Chang, C.J. Xing, Tissue React. 20 (1998) 1017. [4] I. Rey G´omez-Serranillos, J. Mi˜nones Jr., P. Dynarowicz-Ł˛atka, E. Iribarnegaray, M. Casas, Phys. Chem. Chem. Phys. 6 (2004) 1580. [5] J.R. Seoane, J. Mi˜nones, O. Conde, J. Mi˜nones Jr., M. Casas, E. Iribarnegaray, J. Phys. Chem. B 104 (2000) 7735. [6] J. Mi˜nones Jr., J.M. Rodriguez Patino, O. Conde, C. Carrera, R.J. Seoane, Coll. Surf. A 203 (2002) 273. [7] G. Gainess Jr., in: I. Prigogine (Ed.), Insoluble Monolayers at LiquidGas Interface, Interscience, New York, 1966. [8] I. Rey G´omez-Serranillos, J. Mi˜nones Jr., P. Dynarowicz-Ł˛atka, J. Mi˜nones, E. Iribarnegaray, Langmuir 20 (2004) 928. [9] I. Rey G´omez-Serranillos, J. Mi˜nones Jr., P. Dynarowicz-Ł˛atka, J. Mi˜nones, O. Conde, Langmuir (2004), in press. [10] M.I. Guirr, A.T. James, Lipid Biochemistry, third ed., University Press, Cambridge, 1980. [11] E.Z. Klenk, Z. Physiol. Chem. Hoppe-Seyler’s 235 (1935) 24. [12] G. Blix, Scand. Arch. Physiol. 80 (1938) 46.
[13] B. Alberts, D. Bray, J. Lewis, M. Raff, K. Roberts, J.D. Watson, Molecular Biology of the Cell, Garland, New York, 1983. [14] B. Maggio, T. Ariga, J.M. Sturtenat, R.K. Yu, Biochim. Biophys. Acta 818 (1985) 1. [15] W. Probst, D. M¨obius, H. Rahmann, Cell. Mol. Neurobiol. 4 (1984) 157. [16] F. Schifferer, H. Beitinger, H. Rahmann, D. M¨obius, FEBS Lett. 233 (1988) 158. [17] H. Beitinger, V. Vogel, D. M¨obius, H. Rahmann, Biochim. Biophys. Acta 984 (1989) 293. [18] F. Schifferer, W. Cordorch, H. Beitinger, D. M¨obius, H. Rahmann, J. Biochem. 109 (1991) 622. [19] C.H. Wynn, B. Robson, J. Theor. Biol. 123 (1986) 221. [20] J. Rodriguez Patino, C.C. Sanchez, M.R. Rodriguez Ni˜no, Langmuir 15 (1999) 2484. [21] P. Luckham, J. Wood, S. Frogatt, R. Swart, J. Colloid Interface Sci. 156 (1993) 164. [22] M.B. Abramson, R.K. Yu, V. Zaby, Biochim. Biophys. Acta 280 (1972) 365. [23] G.D. Fidelio, B. Maggio, F.A. Cumar, Biochim. Biophys. Acta 854 (1986) 231. [24] H. Beitinger, W. Probst, D. M¨obius, H. Rahmann, J. Biochem. 102 (1987) 963. [25] F.C. Goodrich, Proceedings of the Second International Congress on Surface Activity, vol. 1, Butterworth, London, 1957, p. 85. [26] R.E. Pagano, N.L. Gershfeld, J. Phys. Chem. 76 (1972) 1238. [27] B. Maggio, F.A. Cumar, R. Capputto, Biochem. J. 171 (1978) 559. [28] J.H. Crowe, M.A. Whittam, D. Chapman, L.M. Crowe, Biochim. Biophys. Acta 769 (1984) 151. [29] J.L. Parker, J. Colloid Interface Sci. 135 (1990) 571.
Interactions between the ganglioside GM1 and hexadecylphosphocholine (miltefosine) in monolayers at the air/water interface Isabel Rey G´omez-Serranillosa , Jos´e Mi˜nones Jr.a , Patrycja Dynarowicz-Ł˛atkab,∗ , Eduardo Iribarnegaraya , Matilde Casasa a
b
Department of Physical Chemistry, Faculty of Pharmacy, University of Santiago de Compostela, Spain Jagiellonian University, Faculty of Chemistry, Department of General Chemistry, Ingardena 3, 30-060 Krak´ow, Poland Accepted 1 November 2004 Available online 16 December 2004
Abstract The ganglioside, GM1 , was studied as Langmuir monolayers at the air/water interface with surface pressure-area measurements in addition to Brewster angle microscopy. A characteristic plateau transition, observed on aqueous subphases of pH 2 and 6, 20 ◦ C, at the surface pressure of ca. 20 mN/m, was attributed to the reorientation of GM1 polar group upon film compression. This transition was found to disappear at alkaline subphases (pH 10) due to the hydration of fully ionized polar group, hindering its reorientation. The interactions between GM1 and hexadecylphosphocholine (miltefosine) were investigated in mixed monolayers and analyzed with the mean molecular areas, excess areas of mixing and the excess free energy of mixing versus film composition plots. The monolayers stability, quantified by the collapse pressure values, as well as the strength of interaction was found to diminish in the following order: pH 6 > pH 2 > pH 10. The strongest interaction occurs for mixed films of miltefosine molar fraction, XM = 0.7–0.8, especially at low pressure region, and are explained as being due to the surface complex formation of 3:1 or 4:1 (miltefosine:ganglioside) stoichiometry (XM = 0.75 or 0.8, respectively). © 2004 Elsevier B.V. All rights reserved. Keywords: Langmuir monolayers; Hexadecylphosphocholine; Ganglioside GM1 ; Interactions; Air/water interface
1. Introduction Hexadecylphosphocholine (miltefosine) is a representative of a new generation of anti-cancer drugs based on a phospholipid-like structure [1]. Its pharmacological action is believed to be related to the disturbance of membrane function by this “false” phospholipid-analogue, which results in the inhibition of cells proliferation [2,3]. For a better understanding of the mode of miltefosine action, we have undertaken studies on its interaction with three kinds of membrane lipids (sterols, phospholipids and glycosphingolipids), which are representative for each side of the cellular membrane. Since both, miltefosine itself and the membrane lipids ∗ Corresponding author. Tel.: +48 12 6336377x2236; fax: +48 12 6340515. E-mail address: [email protected] (P. Dynarowicz-Ł˛atka).
0927-7765/$ – see front matter © 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.colsurfb.2004.11.004
are capable of stable monolayer formation at the air/water interface [4–6], we have successfully applied the Langmuir monolayer technique [7] to study the intermolecular interactions. One of our previous works [8] describes miltefosine interactions with a typical membrane sterol – cholesterol – which is placed in-between both sides of the biomembrane, while our recent paper [9] presents the investigations on the interaction with a typical phospholipid forming the inner part of the membrane [10], namely OPPE (-oleoyl-␥-palmitoyl l-␣-phosphatidylethanolamine). This work is aimed at investigating the interactions with a ganglioside (GM1 ), which is found exclusively in the outer half of the bilayer [1]. Gangliosides belong to the group of complex lipids (socalled glycolipids) containing a carbohydrate residue. All the glycolipids in animal cells are glycosphingolipids (GSLs). Their apolar (aliphatic) part (so-called ceramide) is constituted by sphingosine (an 18-carbon atom, dihydric alcohol,
64
I.R. G´omez-Serranillos et al. / Colloids and Surfaces B: Biointerfaces 41 (2005) 63–72
which contains an amino group at C17 ) and a long-chain fatty acid in amide linkage (Scheme 1). The bulky polar part, on the other hand, is formed by carbohydrate moieties attached to hydroxymethyl group of ceramide. Such compounds are known as cerebrosides. The gangliosides contain, in addition to sugar residues, N-acetylneuraminic acid (NANA), also called sialic acid. Due to their structure, the gangliosides display amphiphilic properties, and are negatively charged at neutral pH. According to the number and position of sugar residues and sialic acid, different types of gangliosides are identified, according to a specific nomenclature. For example, letter G stands for a ganglioside in general; the letters M, D, T, etc. identify the number of sialic acid residues (1, 2, 3, etc.), while the subscript 1, 2, etc. is an arbitrary number assigned according to the chromatographic mobility. In this work, we have investigated the ganglioside GM1 , the chemical structure of which is shown in Scheme 1. The gangliosides, first described by Klenk [11] in 1935 and Blix [12] in 1938, are inserted in the extracellular part of synaptic nerve cell (neuron) membranes of vertebrates, where they constitute about 6% of the total lipid mass [13]. They are involved in intermolecular interactions, differentiation, transformation, ontogenesis and apoptosis [14–17]. Some of them are highly antigenic [18] while others – like GM1 – act as a cell-surface receptor for the bacterial cholera toxin [19]. Currently, gangliosides are investigated for their efficiency to develop an anti-cancer vaccine, especially for epithelial cancer of breast.
2. Experimental Both miltefosine (99%) and the investigated ganglioside (monosialoganglioside palmitoyl-GM1 , 95%) were supplied by Sigma. The compounds were stored in a refrigerator without the access of light. Spreading solutions were prepared by dissolving each compound in a 4:1 mixture of chloroform:alcohol (absolute ethanol for miltefosine; methanol for GM1 ) with a typical concentration of 0.2–0.3 mg/ml. Mixed
Scheme 1. Chemical structure of GM1 . Glc, Glu and GalNAc denote galactose, glucose and N-acetylglucosamine, respectively.
solutions were prepared from respective stock solutions of both compounds, and were dropped onto the subphase with a Microman Gilson microsyringe, precise to ±0.2 l. The number of molecules deposited on the water subphase was kept constant in all experiments (3.2 × 1016 molecules). Ultrapure water (produced by a Nanopure water purification system coupled to a Milli-Q water purification system, resistivity = 18.2 M cm) was used as a subphase. The subphase temperature was controlled to within 0.1 ◦ C by a circulating water system from Haake. HCl or NaOH was added into water to adjust the pH value of the subphase. Experiments were carried out with a NIMA 601 (Coventry, U.K.) trough (total area = 525 cm2 ), placed on an anti-vibration table. Surface pressure was measured with the accuracy of ±0.1 mN/m using a Wilhelmy plate made from ashless chromatography paper (Whatman Chr1). After spreading, monolayers were left for 10 min to ensure solvent evaporation, and afterwards the compression was initiated with a barrier speed of 50 cm2 /min ˚ 2 /(molecule min)). (15.6 A Brewster angle microscopy images and ellipsometric measurements were performed with BAM-ELLI 2000 (NFT, Germany) equipped with a 30 mW laser emitting p-polarized light at 690 nm wavelength which was reflected off the air/water interface at approximately 53.1◦ (Brewster angle). In measuring the relative reflectivity of the film, a camera calibration was necessary as described elsewhere [20]. At Brewster angle: I = |Rp|2 = Cd 2
(1)
where I is the relative reflectivity, C is a constant, d is the film thickness, and Rp is the p-component of the light. The lateral resolution of the microscope was 2 m, and the images were digitized in order to obtain high quality BAM images.
3. Results 3.1. Pure monolayers A thorough Langmuir monolayer characteristic of miltefosine is described in our former paper [4]. Miltefosine was found to form liquid-expanded monolayers that collapse at ca. 40 mN/m. The film stability was found to be satisfactory [4], and the monolayer behavior was hardly influenced by such experimental conditions as the compression speed, number of deposited molecules, subphase temperature (within the range of 10–25 ◦ C) and subphase pH (within the range of 2–12). The surface pressure–area (π–A) isotherms of the investigated ganglioside at 20 ◦ C spread at aqueous subphases of pH 2, 6 and 11 are shown in Fig. 1. At pH 6, the monolayer shows a gas-liquid expanded transition region at molec˚ 2 /molecule, where the surface ular areas larger than 120 A pressure is maintained constant at ca. 0.5 mN/m. At areas ˚ 2 /molecule and at low surface pressures, the below 120 A
I.R. G´omez-Serranillos et al. / Colloids and Surfaces B: Biointerfaces 41 (2005) 63–72
Fig. 1. Surface pressure (π)–area (A) isotherm of GM1 spread at 20 ◦ C on aqueous subphases of pH 2, 6 and 10. Inset: compressional modulus (Cs −1 )–surface pressure (π) dependencies.
monolayer shows a liquid-expanded state, characterized by a compression modulus (Cs −1 ) value of 36 mN/m and a limiting area (determined by extrapolating of the linear fragment of the isotherm preceding the collapse to π = 0) of about ˚ 2 /molecule. Luckham et al. [21] reported a similar 101.7 A value for the limiting area of GM1 spread on water of pH 5.6, ˚ 2 /molecule) was obtained on however, a lower value (88.6 A 145 mM aqueous NaCl subphase, and was attributed to the screening of the negative charges on the ganglioside headgroups (due to the ionized sialic acid at neutral pH [22]) by the presence of salt, thus reducing the electrostatic repulsion between the individual molecules in the monolayer. At ˚ 2 /molecule, a phase tranapproximately 20 mN/m and 74 A sition shoulder is reached and upon further compression, the monolayer shows a liquid-condensed state of low compressibility (Cs −1 = 154 mN/m). Finally, the film collapse occurs at 65.9 mN/m, proving high film stability. Fidelio et al. [23] obtained for GM1 monolayers a collapse pressure value of about 60 mN/m at lower temperatures. At acidic pH (pH = 2), the monolayer occupies smaller area at the interface as compared to pH 6, and the limiting molecular areas in the liquid-expanded and liquid-condensed ˚ 2 /molecule and 59.4 A ˚ 2 /molecule, respecphase are 90.3 A tively. The phase transition appears at ca. 20 mN/m, similarly as it was observed at pH 6. The film collapse is reached at ˚ 2 /molecule at the same surface pressure (65.9 mN/m) 38.8 A as at pH 6. Luckham et al. [21] reported a similar value ˚ 2 /molecule) for the molecular area at the collapse. (38 A At pH 10, the monolayer is most expanded as compared to all the investigated subphases, without visible plateau transition, and the collapse occurs at much lower surface pressure (ca. 51 mN/m). The Cs −1 –π curves, corresponding to the isotherms recorded at the three studied pHs, are shown in the inset of Fig. 1. These dependencies clearly indicate the existence of a
65
liquid expanded–liquid condensed (LE–LC) phase transition at pH 2 and 6, which appears as a pronounced minimum that separates the expanded and condensed phase at the surface pressure of about 20 mN/m. At basic subphase pH, there is no clear minimum, but rather a plateau region of Cs −1 values ranging from 60 to 70 mN/m, which are typical for the liquid-expanded state of a monolayer. Brewster angle microscopy has been applied in addition to the surface pressure–area isotherm to characterize the monolayer from GM1 . The images, which are shown in Fig. 2, were recorded at pH 6, 20 ◦ C. At low surface pressure region, where the monolayer is in a liquid-expanded state, the image is completely homogeneous (Fig. 2A, taken at π = 4 mN/m) and does not change upon film compression until post-transition region. At higher surface pressures (above 25 mN/m), small bright domains appear (Fig. 2 B, taken at 35 mN/m), which correspond to the condensed phase. These domains fuse together, forming compact agglomerates, seen as stripes, at surface pressures corresponding to the film collapse (Fig. 2C). The relative reflectivity (I) versus time (t) curve is shown together with the π–t isotherm in Fig. 2. Along the liquidexpanded phase and the transition region, the relative reflectivity increases monotonically upon film compression, similarly to surface pressure, until the value of 1.1 × 10−5 (arbitrary units) is reached at the end of this phase. The increase in the relative reflectivity from the start of compression (I = 1 × 10−6 ) is thus 11-fold, which corresponds to the increase in monolayer thickness by 3.5 times. Along the liquidcondensed state, above 25 mN/m, the relative reflectivity is maintained practically constant, indicating that the monolayer thickness is no longer changing. Only at the monolayer collapse, a small increase in reflectivity signal can be observed. An important feature of the I versus t dependence is the absence of noise peaks in the liquid-expanded region, which accounts for the completely homogeneous BAM images observed in this region. Along the liquid-condensed phase, on the other hand, some noises can be noticed, which are due to the presence of small condensed-phase domains, visible in the image B, Fig. 2. The increase in subphase temperature (Fig. 3) causes film expansion and shifts the LE–LC transition to higher surface pressures. These results are in good agreement with those obtained by Luckham et al. [21]. At 26 ◦ C they reported the transition to occur at the pressure of 25 mN/m ˚ 2 /molecule), while the results presented herein, cor(at 71 A responding to 25 ◦ C, show the transition at π = 26.3 mN/m ˚ 2 /molecule. Film expansion caused by corresponding to 73 A the temperature increase is expected due to the weakening of intermolecular forces between apolar tails of neighboring molecules in the monolayer. However, an anomalous tendency was reported by Beitinger et al. [24] for a monolayer spread on 5 mM buffer solution of pH 7.4, where the temperature increase provoked film condensation to smaller molecular areas. The authors, however, did not comment on such a peculiar behavior.
66
I.R. G´omez-Serranillos et al. / Colloids and Surfaces B: Biointerfaces 41 (2005) 63–72
Fig. 2. Time evolution of the relative reflectivity, I () and surface pressure, π () of GM1 monolayer spread on water, pH 6 at 20 ◦ C, together with BAM images taken at the points designated on the isotherm.
The compression modulus curves (inset of Fig. 3) confirm film expansion with the temperature increase. Firstly, the position of the minimum (M, indicated by arrows), which is referred to the LE–LC transition, undergoes displacement towards higher surface pressures with temperature. Secondly, with the increase of temperature, Cs −1 values at the first maximum increase, while those corresponding to the second max-
imum diminish. This finally leads to the formation of one monolayer phase (one maximum) at high subphase temperature. This means that the two physical states of the monolayer, which can be clearly distinguished at low temperature, fuse into one expanded state. The collapse pressure varies with temperature (Fig. 3), i.e. within the range of 10–30 ◦ C, the values of πc are comprised between 62 and 66 mN/m. These values are higher than those obtained by Fidelio et al. [23] (45 mN/m for GM1 monolayer spread at 30 ◦ C, and 55 mN/m for temperature interval between 20 and 25 ◦ C). This can be due to different experimental conditions, i.e. in ref. [23] the monolayer was spread on 145 mM aqueous NaCl solution and was compressed with lower speed (ca. 7 cm2 /min.). 3.2. Mixed monolayers of miltefosine and GM1
Fig. 3. Surface pressure (π)–area (A) isotherm of GM1 spread on aqueous subphases of pH 6 at different temperatures. Inset: compression modulus (Cs −1 )–surface pressure (π) dependencies.
The surface pressure–area isotherms for mixed miltefosine-GM1 monolayers, spread at different subphase pHs (2, 6 and 10) at 20 ◦ C, are shown in Fig. 4 A–C. As it can be seen, the behavior of mixed films is generally similar at pHs 2 and 6, i.e. mixtures richer in ganglioside (XM = 0.1–0.3) exhibit similar behavior to pure ganglioside, showing a transition (LE–LC) at nearly identical surface pressure as pure GM1 (ca. 20 mN/m; see Table 1), which, however, becomes progressively less pronounced as miltefosine proportion in the monolayer increases, and
I.R. G´omez-Serranillos et al. / Colloids and Surfaces B: Biointerfaces 41 (2005) 63–72
67
Fig. 4. Surface pressure (π)–area (A) isotherms of GM1 , miltefosine and their mixtures spread on aqueous subphase of pH 2 (A), pH 6 (B) and pH 10 (C) at 20 ◦ C.
finally disappears for a mixed monolayer of XM = 0.5, which shows an intermediate behavior between pure miltefosine and ganglioside, i.e. is expanded at surface pressures below 20 mN/m, and condensed at higher pressures, without a plateau transition. Mixed monolayers rich in miltefosine (XM = 0.9) look almost identical as pure miltefosine, however, they are slightly shifted towards smaller areas and collapse at a higher surface pressures. At basic pH, the GM1 monolayer is expanded, without visible phase transition, and mixed films are of expanded type as well, within the whole range of composition. When the milte-
fosine content exceeds 0.5 mole fraction, the mean molecular area occupied by film molecules is smaller as compared to both pure components. The value of the collapse pressure is related to the monolayer stability, i.e. the higher it is, the more stable the monolayer is. As it can be seen in Table 1, the collapse pressure changes with the incorporation of miltefosine into GM1 monolayer at different subphase pHs. The trend observed at pH 2 and 6 is similar; the stability of mixed films increases upon increasing the proportion of miltefosine, reaching a maximum (corresponding to XM = 0.5), and decreases after-
Table 1 Compression modulus (Cs −1 , mN/m), transition surface pressure (πt , mN/m) and collapse pressure (πc , mN/m) values for monolayers of GM1 , miltefosine and their mixtures at different subphase pH at 20 ◦ C Monolayer composition
GM1 XM = 0.1 XM = 0.3 XM = 0.5 XM = 0.7 XM = 0.9 Miltefosine
pH 2 −1
pH 6 −1
−1
pH 10 −1
Cs 1st maximum
Cs 2nd maximum
πt
πc
Cs 1st maximum
Cs 2nd maximum
πt
πc
Cs −1 1st maximum
πc
51 46 47 42 56 47 43
154 166 132 – – – –
20.0 17.0 17.1 – – – –
65.9 67.1 68.5 71.7 65.0 39.5 36.5
51 47 47 42 56 48 47
154 166 132 – – – –
20.0 16.7 17.1 – – – –
66.4 67.0 68.4 71.9 72.8 39.6 39.5
– 59 59 56 65 50 43
51.5 55.6 61.5 60.8 59.3 44.4 37.5
68
I.R. G´omez-Serranillos et al. / Colloids and Surfaces B: Biointerfaces 41 (2005) 63–72
wards. The monolayer stability at pH 10 is lower as compared to acidic and neutral subphases, no matter what the film composition is. This is quite logical since the monolayer of pure ganglioside also collapses at much lower pressure at pH 10 as compared to other investigated subphase pHs. Analyzing the collapse pressure values it may be concluded that the stability of mixed films obey the following order: pH 6 > pH 2 > pH 10. The surface pressure–area isotherms shown in Fig. 4 clearly indicate condensation of monolayers at all the investigated pH values and surface pressure regions. The extent of contraction of mixed monolayers in respect to pure components is a measure of interaction between film-forming molecules, and can be visualized on the diagrams of the mean molecular area versus film composition (A12 = f(XM )). Such dependencies are presented in Fig. 5 for different subphase pHs in both low (5 mN/m) and high (20 mN/m, 35 mN/m) surface pressures. The dotted lines illustrate the additive re-
lationship for the ideal system: A12 = A1 X1 + A2 X2 (A12 is the mean molecular area of mixed monolayer, X1 and X2 denote the mole fractions of components 1 and 2; A1 and A2 stand for molecular areas of pure components at the same surface pressure as A12 was determined). Within the whole range of the investigated pHs and mole fraction, negative deviations from the ideality can be observed both in low and high surface pressure regions. As regards the magnitude of the discussed deviations, stronger occur at lower surface pressures (5 mN/m). Moreover, when the subphase pH is 6, the films condensation is stronger as compared to pH 2 or 10. Thus it may be concluded that the interactions between miltefosine and GM1 , responsible for the observed deviations, diminish in the following order: pH 6 > pH 2 > pH 10. One may quantify the interactions by the excess functions, for example the excess areas of mixing (Aexc ) and excess free energy of mixing ( Gexc ). Values of Aexc can be calculated
Fig. 5. Dependence of the mean molecular area (A12 ) on the molar fraction of miltefosine (XM ) in mixed films with GM1 , spread at aqueous subphases of pH 2 (A), pH 6 (B) and pH 10 (C) at 20 ◦ C.
I.R. G´omez-Serranillos et al. / Colloids and Surfaces B: Biointerfaces 41 (2005) 63–72
by comparing the area of mixed monolayer (A12 ) with that for unmixed, one-component monolayers (A1 ; A2 ), according to the equation: Aexc = A12 − (A1 X1 + A2 X2 ). If the mixture is ideal, or its components are immiscible, Aexc is zero and the dependence Aexc = f(Xi ) is linear. Deviations from these conditions indicate miscibility and non-ideality. From Fig. 6A–C, which presents the results of Aexc calculated for the investigated system, one may distinguish more clearly the differences in mixed films behavior at various pHs. At low pressures (5 mN/m), the Aexc –XM dependencies exhibit a clear minimum for XM = 0.8 (pH 2) and 0.7 (pH 6 and 10). While on subphases of pH 2 and 10 at high surface pressures (20 and 30 mN/m) the extent of contraction is similar for all mixtures within the range of XM = 0.1–0.8, on pH 6 mixtures of XM = 0.6–0.7 show the highest condensation, which is visible as a pronounced minimum. It is thus clear that on pH 6, the values of excess area of mixing are
69
most negative, independently on pressure region and film composition. A more detailed analysis of the intermolecular interactions is based on the excess free energy of mixing ( Gexc ) calculations, which can be determined from the following relation derived by Goodrich [25] and Pagano and Gershfeld [26]: π Gexc = N (A12 − X1 A1 − X2 A2 )dπ (2) 0
Fig. 7A–C presents Gexc values at three different pHs at 20 ◦ C as a function of miltefosine composition in mixed monolayers. The results are very similar to those discussed above, i.e. on both pH 2 and 10 the curves Gexc –XM show very little differences. On the contrary, on pH 6 a clear minimum is seen at XM ∼ 0.8 (at high surface pressures) and at XM = 0.9 (for π = 5 mN/m). It may thus be concluded
Fig. 6. Plots of the excess area of mixing (Aexc ) vs. molar fraction of miltefosine (XM ) in mixed films with GM1 , spread at aqueous subphase of pH 2 (A), pH 6 (B) and pH 10 (C) at 20 ◦ C.
70
I.R. G´omez-Serranillos et al. / Colloids and Surfaces B: Biointerfaces 41 (2005) 63–72
Fig. 7. Excess free energy ( Gexc ) as a function of mole fraction of miltefosine (XM ) in mixed films with GM1 , spread at aqueous subphase of pH 2 (A), pH 6 (B) and pH 10 (C) at 20 ◦ C.
that there are attractive interactions between miltefosine and GM1 , which are strongest at pH 6, while on pH 2 and 10 they are nearly of the same order.
4. Discussion The ganglioside, GM1 , has already been investigated in monolayers at the air/water interface by other authors, however, the results are rather difficult to compare since the experimental conditions were different as compared to those applied in the present work. Namely, GM1 was studied on a big variety of different subphases, containing various ions, for example on 145 mM aqueous NaCl [27], 5 mM triethanolamine/HCl buffer of pH 7.4 [15,16,24], 3.5 mM phosphate buffer [17], etc. Crowe et al. [28] published values of molecular areas at 20 and 30 mN/m on aqueous sub-
phases, however, did not show the complete π–A isotherm, while Parker et al. [29] presented the isotherm between 0 and 30 mN/m, but did not specified the subphase temperature. Only the results presented by Luckham et al. [21] can be useful for comparison with the present data, although the subphase pH was different (5.6 and 1). Nevertheless, the reported values are in a good agreement with those presented herein. The phase transition (LE–LC), which appears on the π/A isotherm of GM1 as an inflection at π = 20 mN/m, was attributed by Luckham et al. [21] to the electrostatic interactions between charged groups of sialic acid (which form, together with sugar molecules, polar headgroup of the ganglioside molecule), causing the increase in surface pressure observed above the transition in the condensed state of the monolayer. This interpretation, of course, does not explain the existence of interactions at pH 1, where sialic acid molecules
I.R. G´omez-Serranillos et al. / Colloids and Surfaces B: Biointerfaces 41 (2005) 63–72
are almost completely protonated (pKa = 2.6) [17,22]. Moreover, the electrostatic interactions are expected to be stronger at pH 10, where sialic acid molecules are completely ionized, and therefore if the interpretation given by the authors [21] was correct, the phase transition in the π–A isotherms should be very much pronounced. In fact, the results shown in Fig. 1 show the disappearance of this transition on alkaline subphases, where the monolayer becomes liquid-expanded along the full compression. Other authors [15,24] interpret the existence of the phase transition as being due to the reorientation of bulky polar group of GM1 , based on the fact that the monolayer properties of gangliosides, in particular mean molecular area, depend – nearly solely – on the respective polar groups, which was shown by Maggio et al. [27]. In view of this fact, it is logical to attribute the phase transition to the change of polar group conformation in the subphase. This change in orientation does not affect the arrangement of apolar tails of the molecule, since the size of the polar group is sufficiently large to hinder the intermolecular attractions between the tails, preventing the formation of molecular aggregates, but causing, on the other hand, strong monolayer resistance to the compression as the molecules become closely packed. This explains the absence of any compact domains in BAM images, and the decrease in monolayer compressibility (increase of its rigidity) above the phase transition. The absence of any inflection (corresponding to a transition) in the π–A isotherms at 20 mN/m for monolayers spread on aqueous subphase of pH 10, can be due to the hydration of polar groups, caused by their complete ionization. It may be assumed that water molecules surrounding polar groups hinder, or make impossible, their reorientation upon compression. In consequence, the monolayer on alkaline subphases is expanded, without any transition, throughout the full compression. Upon compression, water molecules solvating polar groups of the ganglioside are squeezed out from the monolayer into bulk phase, which causes the decrease of the mean molecular area accompanied by an increase in surface pressure. Such a behavior resembles a soaked sponge, which is squeezed out from water. A lower collapse pressure of GM1 monolayer on pH 10 as compared to other investigated subphase pHs can be due to the fact that presence of hydration shell surrounding polar groups increases the distances between hydrocarbon tails of neighboring molecules, lessening in this way van der Waals attraction forces and causing the decrease of film stability. As the mixed monolayers are concerned, their higher collapse surface pressure values evidence for the presence of stronger attractive forces between molecules in the mixed film as compared to monolayers from pure components. The difference between experimental πc value and those calculated on the basis of the additivity rule is a measure of the strength of interactions between film components, which are
71
found to diminish in the following order: pH 6 > pH 2 > pH 10. Similar trend is observed when analyzing the results of the mean molecular areas (Fig. 5), as well as the free energy of mixing values (Fig. 7). Since miltefosine was found to be insensitive to the variation in subphase pH [4], it is evident that the ganglioside is responsible for the observed effect. In this sense, one may expect that at pH 6 the attractive electrostatic interactions which are established between negatively charged sialic acid of the ganglioside polar group and positively charged trimethylammonium groups of miltefosine are responsible (in addition to the van der Waals attraction forces between apolar tails) for the considerable increase in mixed films stability in relation to monolayers from pure components. Possible repulsive forces between the ganglioside and the phosphate group of miltefosine would give less contribution because of the strong hydration of the phosphate ion. At pH 2, when sialic acid is uncharged, the electrostatic interactions in the mixed film cannot be established, however, other kinds of interaction can occur, like for example ion–dipole interactions between the polar groups of miltefosine (positively charged at this pH) and those of the ganglioside. Differences observed at pH 6 and 2 are due to the different energy of the ion-zwitterion (pH 6) and ion–dipole interactions. At pH 10, both polar groups of ganglioside and miltefosine are negatively charged, and therefore repulsive electrostatic interactions could be expected, resulting in the decreased stability of the mixed monolayer. Therefore, the deviations from ideal behavior should be positive, contrary to the experimental results. This can be explained by the fact that at this subphase pH the attractive van der Waals forces between hydrocarbon tails predominate over the electrostatic repulsions between polar groups, which are significantly weakened by the hydrated water molecules. Moreover, it should be mentioned that between the monosaccharides (forming in part the bulky polar group of the ganglioside) and the polar group of miltefosine, various types of attractive interactions can be established, for example dipole–dipole, ion–dipole, hydrogen bonds, etc., which contribute significantly to the stability of mixed monolayers. Although these interactions also occur at other subphase pHs, however, especially at pH 10 their contribution is prevailing. A bigger polar group of the ganglioside, as compared to miltefosine, allows us to understand why the strongest contraction occurs for mixed films of XM = 0.7–0.8, especially at low surface pressure region, where the monolayer is expanded. In this situation, the molecular area of the monolayer is determined by the bulky polar group of the ganglioside, and the molecules of miltefosine can penetrate between neighboring molecules of GM1 , forming surface complexes of 3:1 or 4:1 (miltefosine:ganglioside) stoichiometry (XM = 0.75 or 0.8, respectively). Therefore, mixtures of high proportion of miltefosine are all involved in the stable complex formation, and none of them remain free, which accounts the observed highest interactions and stability.
72
I.R. G´omez-Serranillos et al. / Colloids and Surfaces B: Biointerfaces 41 (2005) 63–72
Acknowledgements This financial support of Xunta de Galicia (project PGIDT99PXI20302B) is gratefully acknowledged. References [1] C. Unger, E.A.M. Fleer, D.J. Kim, A. Breiser, P. Hilgard, J. Engel, G.A. Nagel, G.H. Eibl, Acta Oncol. 28 (1989) 213. [2] K. Boggs, C.O. Rock, S. Jackowski, Biochim. Biophys. Acta 1389 (1998) 1. [3] Y.Z. Wang, Y.B. Chang, C.J. Xing, Tissue React. 20 (1998) 1017. [4] I. Rey G´omez-Serranillos, J. Mi˜nones Jr., P. Dynarowicz-Ł˛atka, E. Iribarnegaray, M. Casas, Phys. Chem. Chem. Phys. 6 (2004) 1580. [5] J.R. Seoane, J. Mi˜nones, O. Conde, J. Mi˜nones Jr., M. Casas, E. Iribarnegaray, J. Phys. Chem. B 104 (2000) 7735. [6] J. Mi˜nones Jr., J.M. Rodriguez Patino, O. Conde, C. Carrera, R.J. Seoane, Coll. Surf. A 203 (2002) 273. [7] G. Gainess Jr., in: I. Prigogine (Ed.), Insoluble Monolayers at LiquidGas Interface, Interscience, New York, 1966. [8] I. Rey G´omez-Serranillos, J. Mi˜nones Jr., P. Dynarowicz-Ł˛atka, J. Mi˜nones, E. Iribarnegaray, Langmuir 20 (2004) 928. [9] I. Rey G´omez-Serranillos, J. Mi˜nones Jr., P. Dynarowicz-Ł˛atka, J. Mi˜nones, O. Conde, Langmuir (2004), in press. [10] M.I. Guirr, A.T. James, Lipid Biochemistry, third ed., University Press, Cambridge, 1980. [11] E.Z. Klenk, Z. Physiol. Chem. Hoppe-Seyler’s 235 (1935) 24. [12] G. Blix, Scand. Arch. Physiol. 80 (1938) 46.
[13] B. Alberts, D. Bray, J. Lewis, M. Raff, K. Roberts, J.D. Watson, Molecular Biology of the Cell, Garland, New York, 1983. [14] B. Maggio, T. Ariga, J.M. Sturtenat, R.K. Yu, Biochim. Biophys. Acta 818 (1985) 1. [15] W. Probst, D. M¨obius, H. Rahmann, Cell. Mol. Neurobiol. 4 (1984) 157. [16] F. Schifferer, H. Beitinger, H. Rahmann, D. M¨obius, FEBS Lett. 233 (1988) 158. [17] H. Beitinger, V. Vogel, D. M¨obius, H. Rahmann, Biochim. Biophys. Acta 984 (1989) 293. [18] F. Schifferer, W. Cordorch, H. Beitinger, D. M¨obius, H. Rahmann, J. Biochem. 109 (1991) 622. [19] C.H. Wynn, B. Robson, J. Theor. Biol. 123 (1986) 221. [20] J. Rodriguez Patino, C.C. Sanchez, M.R. Rodriguez Ni˜no, Langmuir 15 (1999) 2484. [21] P. Luckham, J. Wood, S. Frogatt, R. Swart, J. Colloid Interface Sci. 156 (1993) 164. [22] M.B. Abramson, R.K. Yu, V. Zaby, Biochim. Biophys. Acta 280 (1972) 365. [23] G.D. Fidelio, B. Maggio, F.A. Cumar, Biochim. Biophys. Acta 854 (1986) 231. [24] H. Beitinger, W. Probst, D. M¨obius, H. Rahmann, J. Biochem. 102 (1987) 963. [25] F.C. Goodrich, Proceedings of the Second International Congress on Surface Activity, vol. 1, Butterworth, London, 1957, p. 85. [26] R.E. Pagano, N.L. Gershfeld, J. Phys. Chem. 76 (1972) 1238. [27] B. Maggio, F.A. Cumar, R. Capputto, Biochem. J. 171 (1978) 559. [28] J.H. Crowe, M.A. Whittam, D. Chapman, L.M. Crowe, Biochim. Biophys. Acta 769 (1984) 151. [29] J.L. Parker, J. Colloid Interface Sci. 135 (1990) 571.