Liposome Structure Imaging by Atomic Force Microscopy

Ovidiu Stoica, Apichai Tuanyok, Xiaowei Yao, and Manfred H. Jericho , David Pink , Terry J. Beveridge. Langmuir 2003 19 (26), 10916-10924. Abstract | ...
1 downloads 0 Views 319KB Size
Download PDF
Langmuir 2002, 18, 6513-6520

6513

Liposome Structure Imaging by Atomic Force Microscopy: Verification of Improved Liposome Stability during Adsorption of Multiple Aggregated Vesicles O. Teschke* Nano-Structure Laboratory, IFGW/UNICAMP, 13081-970, Campinas, SP, Brazil

E. F. de Souza Instituto de Cieˆ ncias Biolo´ gicas e Quı´mica, Pontifı´cia Universidade Cato´ lica de Campinas, 13020-904, Campinas, SP, Brazil Received March 1, 2002. In Final Form: May 21, 2002 Atomic force microscopy has enabled direct visualization of the liposome structure supported on mica surfaces in air and silanized mica surfaces in aqueous media. The images display distinct patterns of adhered liposomes: multiple and single vesicle liposomes and flat supported bilayers. The multiple vesicle liposome structure is not visible by optical microscopy since the vesicles forming the liposome have diameters as small as 20 nm. Molecularly resolved force versus distance curves displaying the organization of hydrocarbon chains (mono- or bilayers) corroborate the presence of distinct adsorbed structures observed by scanning the surface. The high resolution of the observed liposome images allows the visualization of the aggregation of the multiple vesicles forming liposomes which were shown to have their origin in the liposome formation process and not during adsorption. Since most of the observed liposomes are aggregated vesicles, this aggregated structure has a substantially larger stability during adsorption than the single vesicle structure and consequently a larger resistance in maintaining its shape and function as a carrier of cosmetics, food additives, and drugs. This observation also has some important consequences in the liposomes’ selective permeability when they are used as carriers.

Introduction Liposomes are spherical, bilayer vesicles that form spontaneously when certain phospholipids are dispersed in water. Liposomes made up of phospholipids have been considered as models of biomembranes. Various functions associated with biomembranes, such as aggregation, fusion, and selective permeability, which depend on the hydrophilic-lipophilic balance of the contents of liposomes are taken into consideration when developing liposomes. When they are applied to the field of drug delivery systems, a controlled release is one of the most important functions in the design of liposomes for delivery. By the use of optical techniques such as reflection interference contrast microscopy, vesicle (diameter g 1 µm) adsorption phenomena have been investigated by Radler et al.1 Shape transformations of vesicles of DMPC and POPC in ion-free water were induced by changing the area-to-volume ratio via temperature variations as reported by Kas and Sackmann.2 The influence of a phospholipid transmembrane redistribution on the shape of nonspherical flaccid vesicles was investigated at a fixed temperature using optical microscopy by Farge and Devaux.3 The first images revealing the molecular structure of supported phospholipid films were reported in 1990.4 Puu et al.5 showed that liquid bilayer structures, partly fused liposomes, and membrane particles are observed by atomic force microscopy (AFM). Since then, * To whom correspondence should be addressed. UNICAMP/ IFGW/DFA, Caixa Postal 6165, 13083-970, Campinas, SP, Brasil. Fax: (5519) 788-5376. E-mail: [email protected]. (1) Radler, J.; Strey, H.; Sackmann, E. Langmuir 1990, 11, 4539. (2) Puu, G.; Artursson, E.; Gustafson, I.; Lundstrom, M.; Jass, J. Biosens. Bioelectron. 2000, 15, 31. (3) Kas, J.; Sackmann, E. Biophys. J. 1991, 60, 825.

a broad spectrum of AFM applications have emerged in model membrane biophysics, allowing the study of the structure and function of biomembranes and biological processes such as molecular recognition, enzymatic activity, and cell adhesion processes. Although the forces between lipid membranes have been widely studied using the osmotic stress method,6 the surface force apparatus,7 and the micropipet technique,8 direct information at high lateral resolution was inaccessible. AFM offers the unique opportunity to probe local physical properties and interaction forces of lipid bilayers with nanoscale lateral resolution, thereby providing new insight into the molecular mechanisms of cell adhesion and membrane fusion. Recently, Reviakine and Brisson9 studied the early growth stages of supported zwitterionic phospholipid bilayers by vesicle fusion on mica. The topography change of liposomes from a vesicle shape to a bilayer through their direct adhesion on a mica surface was recorded in situ by AFM.10 The 1iposomes adhered immediately on the mica surface, and the spherica1 shape of the vesicles ruptured spontaneously and was deformed to produce a flat supported film on the mica. Previous force versus distance measurements by AFM using tips made of silicon nitride (Si3N4) and a mica substrate have been reviewed by B. Capella and G. (4) Erger, M.; Ohnesorge, F.; Weisenhorn, A. L.; Heyn, S. P.; Drake, B.; Prater, C. B.; Gould, S. A. C.; Hansma, P. K.; Gaub, H. E. J. Struct. Biol. 1990, 103, 89. (5) Farge, E.; Devaux, P. F. Biophys. J. 1992, 61, 347. (6) Le Neveu, D. M.; Rand, R. P.; Parsegian, V. A. Nature 1976, 259, 601. (7) Marra, J.; Israelachvili, J. Biochemistry 1985, 24, 4608. (8) Evans, E. A. Biophys. J. 1980, 31, 425. (9) Reviakine, I.; Brisson, A. Langmuir 2000, 16, 1806. (10) Egawa, H.; Furusawa, K. Langmuir 1999, 15, 1660.

10.1021/la025689v CCC: $22.00 © 2002 American Chemical Society Published on Web 07/20/2002

6514

Langmuir, Vol. 18, No. 17, 2002

Diether.11 Calibration, noise, and systematic errors are discussed in detail.12 Results were obtained first by Butt,13 then by Weisenhorn et al.,14 and later by Atkins and Pashley.15 To eliminate the problem of the unknown shape of the tip, various studies11 have used modified cantilevers with tips of known geometry. With the advent of Si3N4 supertips, the tip/substrate interaction configuration was radically altered. The details of the interaction are described in our previous publication.16 In this work, we have investigated the structure of liposomes adsorbed on mica. Liposome/substrate interaction was modified by using various surface treatment processes resulting in a variable strength substrate/ liposome interaction. Adsorbed vesicles were observed initially on untreated mica and then on silanized mica substrates. Liposomes were also observed in water after drying liposome solutions on mica surfaces. The visualization of the adsorbed liposome allows us to determine if they are isolated in aqueous solution or clustered to form multiply aggregated structures. Experiment Materials. Phosphatidylcholine from egg yolks 99.9% (PPC) and cholesterol 99% (Ch) were purchased from Aldrich, and a chloroform analytical reagent grade was purchased from Merck. All chemicals were used without further purification. We used distilled and deionized water obtained by a Milli-Q system. Preparation of Liposomes. Large vesicles were made by reverse-phase evaporation17-19 using a mixture of 2.25 × 10-4 M phosphatidylcholine-cholesterol (2:1 molar ratio). The PPC and Ch were mixed in 150 mL of chloroform and evaporated to dryness on a rotary evaporator (Fisaton) to form a dry organic film. The film was kept under vacuum for at least 3 h. A 150-mL volume of deionized water at 328 K was added to the PPC and Ch mixture. This mixture was then sonicated for 10 min using a bath-type sonicator (Odontobras 1440D, 20 kHz). To remove the largest particles and hence obtain a more homogeneous liposome population, the suspension was extruded through cellulose acetate membranes of 0.45 µm pore size in a Millipore filtration cell. Photon Correlation Spectroscopy. The average diameter size and electrophoretic mobility were measured in a ZetaPlus Instrument (Brookhaven Instruments) with Bi-MAS software and a solid-state laser (15 mW, λ ) 670 nm, scattering angle of 90°) as the radiation source.20 The liposome samples were diluted with 2.5 mL of filtered 10-3 M aqueous potassium chloride in order to give a suitable scattering intensity. Atomic Force Microscopy. The AFM observations were performed with a TopoMetrix TMX2000 AFM. All the images were obtained with the contact mode using triangular cantilevers (Si3N4) which were 100 nm long and had a spring constant of 0.03 N/m. Muscovite mica was chosen as a solid substrate and used immediately after cleavage in a clean atmosphere. The initial surface (cleaved mica surface) characterization was performed in water by using a fluid cell.21 During this characterization experiment, the probe and cantilever were immersed completely in the water solution. The liposome on mica suspension was dried from air (65% humidity) for 3 h. (11) Capella, B.; Diether, G. Surf. Sci. Rep. 1999, 34, 1. (12) Teschke, O.; Ceotto, G.; de Souza, E. F. Phys. Rev. E 2001, 64, 11605. (13) Butt, H. J. Biophys. J. 1991, 60, 1438. (14) Weisenhorn, A. L.; Maivald, P.; Butt, H. J.; Hansma, P. K. Phys. Rev. B 1992, 45, 11226. (15) Atkins, D. T.; Pashley, R. M. Langmuir 1993, 9, 2232. (16) Teschke, O.; de Souza, E. F. Appl. Phys. Lett. 1999, 74, 1755. (17) Magin, R. L.; Meisman, M. R. Chem. Phys. Lipids 1984, 34, 245. (18) Magin, R. L.; Meisman, M. R. Cancer Drug Delivery 1984, 1, 109. (19) Hsiao-Chang, C.; Magin, R. L. J. Pharm. Sci. 1989, 4, 311. (20) Monteiro, V. A. R.; de Souza, E. F.; Azevedo, M. M. M.; Galembeck, F. J. Colloid Interface Sci. 1999, 217, 237. (21) Teschke, O.; Douglas, R. A.; Prolla, T. A. Appl. Phys. Lett. 1997, 70, 1977.

Teschke and de Souza

Figure 1. Force vs distance curve indicated by 9, for mica surfaces immersed in a liposome solution. The repulsive deviation from the horizontal line (∼10 nm from contact) is followed by an attractive regime (indicated by WI). A large repulsion component is measured when the tip further approaches the interface, followed by an attractive regime (indicated by WII, which corresponds to the thickness of a bilayer). The control experiment was performed using mica substrates immersed in water (curves O). Each curve presented was registered using at least five different substrates and three different tips with the various approach velocities averaged by using measurements at different points of the sample. The best signal-to-noise ratio curves were used. In situ vesicle adsorption was performed as follows: a freshly cleaved piece of mica was installed in the microscope, and after allowing the microscope to equilibrate for a minimum of 60 min, a freshly sonicated liposome suspension diluted to 8 µg/mL in water was injected. The contact mode fluid cell was washed extensively with water, then with 95% ethanol, and again with water before each experiment. Mica plates (15 mm in diameter) were glued to the metal disks using epoxy glue. Images were recorded in the constant-force mode using sharpened silicon nitride tips mounted on cantilevers with nominal force constants of 0.03 N/m. The scan angle was 90°. Force was kept at the lowest possible value by continuously adjusting the set point during imaging. Silanization of the Mica Surface. Surface modification was accomplished by a liquid-phase deposit of silane in an organic solvent. Immediately after cleavage, the mica samples were placed in a Petri dish previously filled with 10 mL of a 2% w/w aminopropylthiethoxysilane (APTES) solution in dry toluene. The solvent was evaporated at 60 °C, and the Petri dish was transferred to a stove and kept at 120 oC for 3 h.

Results and Discussion Control experiment curves were performed using Si3N4 tips and mica substrates immersed in the Milli-Q Plus water and scanning mica surfaces. One of the force versus separation control curves is shown by curve O in Figure 1. No structure is observed in the force versus distance curve for a distance close (∼5 nm) to the mica/water interface. Initially, we tried to image liposomes on mica surfaces immersed in a liposome solution by scanning various regions of 1 × 1 cm2 sample substrates. The scanning of the mica surface after an immersion period of 60 min does not show any adsorbed liposome. Force versus distance curves were then measured for this surface (Figure 1). Results show a uniformly covered surface. The repulsion deviation from the horizontal line, starting at ∼10 nm from contact, is followed by an attraction regime, when

Multiple Aggregated Vesicle Liposomes

Langmuir, Vol. 18, No. 17, 2002 6515

Figure 2. Force vs distance curve for mica surfaces immersed in water using the tip previously employed in measuring the force vs distance curve shown in Figure 1.

Figure 3. The average force vs distance curve for mica surfaces immersed in Ca2(PO4) solutions. The bilayer rupture force value (indicated by WI) is ∼0.7 nN.

the force achieves ∼0.2 nN (indicated by WI). A large repulsion component is observed when the tip further approaches the interface, which is followed by an attraction regime (indicated by WII, which corresponds to the thickness of a bilayer). Figure 1 then shows that in liposome solutions there are two attraction component regions indicated by WI and WII in the measured force versus distance curves. The liposome solution was removed from the liquid cell, and a new mica piece was immersed in clean water. Instead of using a new cantilever, we utilized the one that had previously measured the force versus distance curves in a liposome solution (shown in Figure 1). The measured force versus distance curve is shown in Figure 2. At ∼8.0 nm from the surface, there is a rapid change in force with a small change in tip/surface separation, and at about 5.0 nm the tip is attracted to the mica surface (indicated by WI). This attraction force has a range that corresponds to a bilayer thickness (∼5 nm) and a rupture force of ∼0.1 nN. Since there are no adsorbed layers on the mica surface, this layer is then adsorbed on the tip surface. By comparing Figures 1 and 2, we may identify the regions of the force versus distance curve shown in Figure 1 as follows: (a) the region measured when the tip and the mica substrate are 10 nm apart and showing a rupture force of ∼0.2 nN corresponds to the bilayer adsorbed on the tip; (b) the region starting at 5 nm from the interface and showing a much larger rupture force corresponds to the bilayer adsorbed at the mica surface. This adsorbed structure results from the strong liposome/mica interfacial charge interaction; consequently, there is a rupture of the liposome spherical structure and the formation of bilayers which spread over the surface. No liposomes were observed on mica substrates immersed in liposome solutions. The strong interaction on the liposome/mica surface that resulted in the liposome destruction and the formation of adsorbed bilayers in aqueous solutions has to be modified in order to image adsorbed liposomes on mica in aqueous solutions. We then prepared mica substrates by immersion in Ca2(PO4) solutions for 15 min. Substrates were removed from the solution, rinsed successively with Milli-Q water, and then immersed in a liposome solution. The sample was scanned 60 min after the mica immersion in the solution and, as observed in untreated mica substrates, no liposomes were observed on the surface. The average measured force versus distance curve is shown in Figure 3. Curves with shapes similar to the ones shown in Figure 1 are depicted. The noticeable difference is the bilayer

rupture force value (indicated by WI), now ∼0.7 nN, a factor of 2 lower than the value shown in Figure 1. This is an indication that the interaction energy layer/substrate was decreased by a factor of ∼2. To further decrease the interaction between the substrate and the liposomes in solution, silanized mica surfaces were used as substrates and the observations were undertaken in an aqueous environment. The substrate and liposome interaction was still too strong, resulting in the rupture of the liposome and the formation of a bilayer absorbed on the mica surface. Therefore, we decided that before scanning the sample, the solution deposited on the substrate should be dried from air in a 60% humidity atmosphere. The reason for this extra step in the liposome fixation process is that during evaporation the decrease in the water volume present in the liposome/ substrate interaction region results in an increase in the counterion concentration and consequently a decrease in the double-layer width. When contact is established between the substrate and the liposome, both surfaces are approximately neutral and the electrostatic interaction is substantially decreased. The deposited structure thickness was then measured in water using force versus distance curves (shown in Figure 4). Three bilayers, characterized by a repulsion and followed by an attraction force acting on the tip, are present in the tip/substrate interaction region, as previously described22 and shown in Figure 1: The first structure starts at 20 nm and ends at 15 nm, the second starts at 15 nm and ends at 8.5 nm, and the third starts at 8.5 nm and ends at the substrate surface. The first region in the force versus distance curve (indicated by 0-2) corresponds to the compression (0-1) and removal (1-2) of the adsorbed layer from the tip surface since the repulsive force value is close to the rupture force obtained for the removal of the tip-adsorbed layer, shown in Figure 2. The second measured structure in the interaction region (indicated by 2-4) starts at 15 nm from the interface; this structure is totally removed from the interaction region for a tip-applied force of 0.1 nN. The third bilayer present in the tip/substrate interaction region is only removed from this region for a tipapplied force of 0.5 nN. Samples dried from air show a smaller adsorbed bilayer rupture force than the force measured on wet (not dried) samples (compare Figure 4 with Figure 1). This is related (22) Teschke, O.; Ceotto, G.; de Souza, E. F. Chem. Phys. Lett. 2001, 344, 429.

6516

Langmuir, Vol. 18, No. 17, 2002

Figure 4. Force vs distance curve measured on a substrate dried in air in a 60% humidity atmosphere. Three bilayers are present in the tip/substrate interaction region: the first structure starting at 20 nm and ending at 15 nm, the second starting at 15 nm and ending at 8.5 nm, and the third starting at 8.5 nm and ending at the substrate surface.

to the neutralization of surface charges at the liposome and substrate surfaces during the drying process. In regions 2-3 and 4-5 shown in Figure 4, the liposomes respond elastically to small stress but undergo plastic deformation when stresses are severe (close to points 3

Teschke and de Souza

and 5). The tangled polymer chain provides the membrane with considerable flexibility when liposomes are stressed. The elastic contribution during the compression of the top bilayer is indicated by 2-3 in Figure 4, and it starts at 16 nm from the interface. The bilayer and the water inside the liposome are compressed up to the tip/substrate distance of 11 nm when the top bilayer is ruptured. The tip is then attracted to the region inside the vesicle. The range of attraction force is ∼4 nm which corresponds to the bilayer width. In the tip/substrate interaction region, only the bilayer is left adsorbed at the mica/solution interface; this layer is compressed from 7 nm up to 3.8 nm when there is the rupture of the membrane and the tip is attracted to the mica surface (∼3.8 nm corresponds to the bilayer width). The rupture force necessary to perforate the bilayer adsorbed at the mica interface is 0.35 nN, and this value is approximately 3 times larger than the rupture force of first bilayer perforated by the tip, which corresponds to the top of the adsorbed liposome starting at 16 nm away from the interface. The rupture force (indicated by 1) of the bilayer adsorbed on an uncharged tip is shown in Figure 4 and is equal to ∼0.02 nN. The rupture force of the top bilayer of the liposome is 0.1 nN, and finally the one corresponding to the adsorbed bilayer on the mica is ∼0.35 nN. The rupture force decreases as the bilayers are further apart from the mica substrate. A substantial part of the rupture force is then associated with the electric field generated by the mica interfacial charges that decays exponentially as it separates from the interface.23

Figure 5. (a) AFM image of liposomes adsorbed from a suspension dried in air on a silanized mica substrate registered in an aqueous solution. A conformational deformation from a spherical vesicle to a flattened liposome is shown. (b) The vertical profile of the image. The height (H) of the liposome is ∼20 nm. Point A corresponds to the height where the force vs distance curve shown in Figure 4 was measured. (c) Three-dimensional image of the scanned structure. The displayed vesicle is formed by the aggregation of two liposomes.

Multiple Aggregated Vesicle Liposomes

Langmuir, Vol. 18, No. 17, 2002 6517

Figure 6. (a) Image of a liposome adsorbed on silanized mica and from a suspension dried previously from an aqueous solution for observation. (b) The vertical profile image of the liposome showing three vesicles indicated by 1, 2, and 3. (c) Three-dimensional image of the three-vesicle aggregated liposome.

The sample used to measure force versus distance curves was then imaged. From the AFM images shown in Figure 5a, we observe that a conformational deformation from a spherical vesicle to a flattened liposome has occurred on the mica surface. The vertical profile of the image is shown in Figure 5b; the height of the liposome is 18 nm, which clearly indicates that the interaction with the substrate resulted in a flattened structure since the measured diameter of the average-sized structure is ∼200 nm. Vesicle deformation can be also recognized from the force versus distance curves as previously discussed (see Figure 4). The height of the liposome where the force versus distance curves were measured is 16 nm which corresponds to point A in Figure 5b. Figure 5c shows a threedimensional image of the scanned structure. It is also possible to observe that the displayed liposome is formed by the aggregation of two vesicles. Figure 6a shows an image of another liposome adsorbed on silanized mica and dried previously from an aqueous solution for observation. The image suggests that this liposome is formed by the aggregation of three vesicles indicated by 1, 2, and 3 in the vertical profile image (Figure 6b). The three-dimensional image is shown in Figure 6c, and again the three liposome aggregations are clearly shown. Probably, there is one adsorbed liposome indicated by 1 (Figure 6b) at the bottom, in contact with the mica surface and with ∼4 nm thickness, and two on its top (indicated by 2 and 3).

We decided then to investigate the structure of aggregated liposomes by observing dried samples deposited on mica and then scanned in air. The sample was prepared by depositing 3 mL of liposome solution on a 1 cm2 area of a mica surface. An overview of the dried structure is depicted in Figure 7a-c, showing successively larger amplifications. These images show that the liposomes after the drying process have adhered to the mica surface. Some of spherical vesicles ruptured spontaneously and were deformed to produce a flat supportive film on the mica surface (see 1 in Figure 7c). The image of the adsorbed dry liposomes shows the following characteristics: The outer diameter of the deposited structure is formed by a largediameter adsorbed layer which has a thickness that corresponds to one bilayer indicated by 1 or two bilayers indicated by 2. Deposited on these bilayers are flattened structures formed by 30 nm diameter liposomes (indicated by 3), by 40 nm diameter liposomes (indicated by 4), by 50 nm diameter liposomes (indicated by 5), and so on up to 220 nm diameter (indicated by 6) adsorbed liposomes. The average liposome diameter measured in a Zeta Plus instrument is ∼200 nm, as shown by the diameter distribution curve depicted in Figure 8. It is also possible to observe that even smaller diameter liposomes are formed by the aggregation of two or more vesicles. A (23) Teschke, O.; Ceotto, G.; de Souza, E. F. Phys. Rev. E 2001, 64, 11605.

6518

Langmuir, Vol. 18, No. 17, 2002

Teschke and de Souza

Figure 7. Liposome adsorbed on mica from a suspension dried previous to observation in air. Images a-c show successively larger amplifications. Some of the spherical vesicles ruptured spontaneously and were deformed to produce a flat supportive film (see 1 in (c)). The two-bilayer type is indicated by 2. Deposited on these bilayer flattened structures are 30 nm diameter liposomes (indicated by 3), 40 nm diameter liposomes (indicated by 4), 50 nm diameter liposomes (indicated by 5), and 220 nm diameter liposomes (indicated by 6).

Figure 8. Size distribution for vesicles in suspension measured in a Zeta Plus instrument. The measured vesicle average diameter is ∼200 nm.

detailed view of a multiple aggregated vesicle liposome structure is shown in Figure 9. Other details of the deposited structure are shown in Figure 10a, where an image of a structure formed by a one-vesicle liposome deposited at the edge of two deposited bilayers is displayed. Figure 10b displays the vertical profile of the image. The profile indicated by 2 corresponds to a flattened region where the two bilayers with thicknesses of ∼5 nm are

Figure 9. AFM image of the structure of a multiple aggregated vesicle liposome.

deposited, and the profile indicated by 1 corresponds to the one-vesicle liposome deposited on these bilayers. Figure 11a displays the image of a liposome formed by the aggregation of various vesicles; its vertical profile image is displayed in Figure 11b. Single Vesicle and Multiple Aggregated Vesicle Liposomes. Both single-bilayer disks and vesicles could

Multiple Aggregated Vesicle Liposomes

Langmuir, Vol. 18, No. 17, 2002 6519

Figure 10. (a) Details of the deposited structure formed by one vesicle deposited at the edge of two deposited bilayers. (b) Image of the vertical profile corresponding to a two-bilayer region with thicknesses of ∼5 nm and on its top a one-vesicle liposome.

Figure 11. (a) Liposome formed by the aggregation of various vesicles. (b) Image of the vertical profile.

be visualized by AFM when the solution was allowed to dry on the mica surface (Figure 7). Single-bilayer disks have a constant height of 5 ( 1 nm above the mica substrate and vary in size, while vesicles varied in size as well as height (Figures 10 and 11). The size distribution for liposomes in suspension is shown in Figure 8. The mean radius of liposomes was found to be 200 nm. Most of the observed liposomes in Figure 7a-c show a structure that has an outer deformed layer probably associated with a flattened spherical vesicle and central vesicles that did not rupture. The most common structure corresponds to a liposome formed by the aggregation of multiple vesicles (Figures 7c, 9, and 11). Aggregation is used to signify the formation of a larger vesicle from two or more smaller vesicles, that is, refers to the joining of two or more smaller vesicles. We propose the following formation mechanism for the deposited liposomes on the mica surface. The process is schematically shown in Figure 12a for a one-vesicle liposome and in Figure 12b for a three-vesicle liposome. During the adsorption process, the liposome formed by the aggregation of various vesicles has the lower part of its structure in contact with the mica substrate. Due to mica/liposome interaction, this structure is ruptured. This results in the formation of a flattened bilayer that adheres to the surface. Since most of the liposomes are formed by the aggregation of multiple vesicles, the contact liposome/ substrate structure is formed by a flat bilayer; the other aggregated vesicles of the liposome are shown at the top as flattened spherical vesicles.

Figure 12. Schematic diagram of the liposome adsorption process. (a) Single vesicle liposome (where the symbol ) indicates a bilayer). (b) Three-vesicle liposome.

The average size of the liposomes in solution measured by photon correlation spectroscopy is ∼200 nm. Starting with a population of 200 nm mean diameter liposomes in solution, aggregation during the drying process would result in larger diameter liposomes. However, these larger diameter liposomes were not observed; therefore, probably most of the aggregation process takes place during the preparation of the liposomes and not during the drying process where the substrate/liposome interaction is probably stronger than the liposome/liposome interaction.

6520

Langmuir, Vol. 18, No. 17, 2002

A second point that deserves attention is that since most of the observed liposomes are formed by aggregation of multiple vesicles, this is evidence that adsorbed aggregated vesicle liposomes have a substantially larger stability than single liposomes. Liposomes in solution described in this work are then formed by multiple vesicles, and the visualization of this aggregated structure is only possible by AFM imaging because the size of the aggregated vesicles forming the liposome is typically ∼20 nm (see Figure 9). The aggregated liposome structure has to be considered when developing liposomes to be applied to the field of drug delivery systems since the selective permeability is substantially modified in aggregated vesicles when compared to single spherical ones. Conclusions Molecularly resolved force versus distance curves displaying the organization of hydrocarbon chains differentiate adsorbed bilayers from liposomes (see Figure 4). High-resolution AFM has enabled direct visualization of liposomes supported on mica in air and on silanized mica in an aqueous medium, resolving multiple aggregated vesicles forming the liposome. These details are not visible by optical microscopy since the vesicles forming the liposome have diameters as small as ∼20 nm. To image vesicle liposomes in an aqueous environment, the liposome/substrate interaction has to be weak enough to prevent the rupture of the liposome structure but strong enough to keep adhering during scanning by the tip. Liposomes formed by the aggregation of multiple vesicles satisfy these requirements because only the vesicle that first got in contact with the mica substrate was flattened and resulted in the formation of a bilayer spread over the

Teschke and de Souza

mica substrate; the other vesicles forming liposomes appear as flattened spherical vesicles on the top of this bilayer. Multiple vesicle liposomes on mica observed in liquids and air show the same average diameter as the ones measured by photon correlation spectroscopy in aqueous solution, indicating that their size was unchanged in the drying process. The average-sized liposome imaged by AFM shows that the liposomes are formed by the aggregation of a few vesicles; consequently, the liposome is formed by smaller vesicles that aggregate to form ∼200 nm diameter vesicles as measured by photon correlation spectroscopy. This observation has some important consequences in the permeation rate of cosmetics, food additives, and drugs where liposomes are used as carriers because aggregated vesicles with a smaller average diameter have a substantially different permeation rate than single spherical vesicles with the same diameter. Our results also show that when liposomes interact with a charged substrate (like mica) the structure formed by aggregation of various vesicles has a greater stability (understood as resistance to deforming their shape) than the one formed by a single spherical vesicle; consequently, there is greater resistance to rupture that destroys the liposome’s function as a carrier of cosmetics, food additives, and drugs. Acknowledgment. The authors are grateful to J. R. Castro and L. O. Bonugli for technical assistance and acknowledge funding support from CNPq Grant 523.268/ 95-5. LA025689V