Surface topography of liposomes and phospholipid bilayers by scanning tunneling microscopy (STM)


Material Information

Surface topography of liposomes and phospholipid bilayers by scanning tunneling microscopy (STM)
Physical Description:
v, 75 leaves : ill. ; 29 cm.
Fowler, Kevin C., 1952-
Publication Date:


Subjects / Keywords:
Liposomes   ( mesh )
Lipid Bilayers   ( mesh )
Phospholipids   ( mesh )
Membrane Proteins -- ultrastructure   ( mesh )
Microscopy, Scanning Tunneling   ( mesh )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph. D.)--University of Florida, 1994.
Includes bibliographical references (leaves 72-74).
Statement of Responsibility:
by Kevin C. Fowler.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
oclc - 50445754
System ID:

Full Text








Hans Schreier has provided guidance and support in a

variety of ways during the author's graduate program and

dissertation phases. Dr. Larry Bottomley has graciously

allowed the use of the scanning tunneling microscope and

provided training on this instrument. Dr. Erdos has

participated in many helpful discussions regarding EM

technique and specimen preparation protocols and has provided

electron micrographs for use in the text. The remaining

members of the author's committee, Dr. Hartmut Derendorf, Dr.

Dinesh Shah, and Dr. Richard Hammer, have participated in a

valuable capacity in reviewing the manuscript and advising in

several phases of the research. The author's wife, Jacque

Fowler, has provided patient support throughout the graduate

study and is responsible for typing of the manuscript. Much

appreciation is expressed to the author's parents, and

especially the author's father, Clem Fowler, for providing

artistic renditions of liposome structure. Dr. Rao has

kindly assisted in performing the water evaporation procedure

and has provided these data. Dr. Carey Mobley has graciously

performed DSC analysis of liposome samples. Natasa Skalko

has kindly allowed the author to use a micrograph from her

cryofracture research.


ACKNOWLEDGMENTS ............................................ii

ABSTRACT ................................................... iv


1 INTRODUCTION ........................................... 1


3 PREVIOUS IMAGING OF LIPOSOMES .......................... 6

MICROSCOPY ............................................ 13


6 MEMBRANE PROTEIN MODEL ................................ 21



9 MATERIALS AND METHODS ................................. 33

10 DETAIL OF EXPERIMENTAL TECHNIQUE....................... 38

11 RESULTS AND DISCUSSION ................................ 45

12 CONCLUSIONS ........................................... 68

LIST OF REFERENCES ......................................... 72

BIOGRAPHICAL SKETCH ........................................ 75


Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy




April, 1994

Chairman: Hans Schreier, PhD
Major Department: Pharmaceutics

Liposomes were examined under the scanning tunneling

microscope (STM) to resolve details of the lipid bilayer.

Liposomes consisted of phosphatidylcholine and

phosphatidylglycerol, (9:1 molar ratio), or a blend

containing cholesterol. Since the STM requires a conductive

sample to establish the tunneling current, the liposomes were

impregnated with osmium tetroxide (0.5% w/v) or uranyl

acetate (1% w/v) for up to two weeks. The liposomes were

mounted on gold covered mica discs for examination in the

STM. Specimens were deposited in a droplet of buffer and

allowed to adsorb, then the tunneling current was established

and the liposomes were examined at room temperature and


Osmium impregnated specimens deposited on gold surfaces

provided the best images. Image topography analysis using

the graphics processor module shows a diameter of

approximately 100 nm, which conforms to the size measured by

laser light scattering. The curvature of the liposome is

relatively smooth and continuous, indicating good

preservation of the liposome structure. Minute subunits,

approximately 0.7 nm in diameter apparent in the scans of the

surface of the liposome, correspond to phosphate head groups.

On further desiccation of the bilayers, globular structures

10 to 30 nm in diameter are evident. These correspond to the

maximum curvature possible of a phospholipid bilayer.

In a second phase of the experiment, liposomes

containing proteins inserted into the bilayer were

successfully imaged. The data obtained corresponded with

those imaged using the scanning electron microscope.

This dissertation describes the first successful imaging

of hydrated and dehydrated liposome vesicles using STM. The

changes seen in the lipid bilayer from the intact hydrated

state to the desiccated state provide a valuable first step

in characterizing the phospholipid bilayer and the structural

changes which take place in desiccation. The STM imaging of

liposomes containing proteins provide valuable data on the

morphology of the protein/phospholipid bilayer system.


This treatise will present in detail the background,

mechanism and results of a new technique developed to

visualize the surface of phospholipid bilayers and liposomes

using the scanning tunneling microscope (STM). The advantage

of such a technique is that it may examine at a molecular

level the arrangement of the molecules which comprise the

surfaces of these structures. In addition, this technique

can be applied to liposomes in the fully hydrated state, as

well as liposomes which have been partially dehydrated and

those which have been completely dehydrated. Since the

specimens may be examined under ambient conditions, a visual

record which is more representative of the natural condition

may be obtained. The liposomes which have been visualized in

the most dehydrated state actually more precisely represent a

desiccated phospholipid bilayer since the overall structure

of the hydrated liposome vesicle has been lost during the

process of dehydration. The formulation of the liposomes was

carried out at the University of Florida, and the

visualization was carried out at the Georgia Institute of


Technology using the scanning tunneling microscope owned by

the Department of Chemistry and Biochemistry.


A liposome is a hollow vesicle, usually suspended in an

aqueous buffer and the interior of which is similarly

aqueous. The size may range from 20 nm to several microns in

diameter (1). The membrane of this vesicle, which surrounds

the internal cavity, is composed of a dipolar lipid material,

usually phospholipids which have hydrophilic head regions and

hydrophobic lipid hydrocarbon tails. These phospholipids are

arranged in such a manner so that the polar phosphate head

groups are oriented toward the interior space on the interior

of the bilayer and toward the surrounding aqueous medium on

the exterior of the bilayer. The lipid chains, then, are

oriented toward each other on the interior of the bilayer,

and the hydrophobic forces holding these lipid chains in

close proximity to each other are one of the primary factors

which stabilize the structure of the liposome. A wide

variety of phospholipids may form liposomes, and even

mixtures of phospholipids will form a stable liposome

structure. Since phospholipids are, as a whole, hydrophobic,

a liposome will form spontaneously when a phospholipid

monolayer or bilayer is suspended in aqueous media. This

characteristic was used to advantage in forming liposomes for

this study. Figure 2-1 shows an ideal representation of the

structure of a "unilammelar" liposome. The spheroid portions

represent the phosphate heads of individual phospholipid

molecules and the angular portions represent the fatty acid

tails which are attached to the phospholipid phosphate head

groups. Since these phospholipids are a primary component of

mammalian cell membranes, a liposome represents an ideal

empty cell model which may be used to participate in drug

delivery or in studies of lipid bilayer membranes. Although

the ideal liposome structure shown in figure 2-1 has only one

phospholipid bilayer surrounding the internal space, it is

possible to have a number of phospholipid bilayers. These

structures are called multilammelar vesicles (MLVs). These

vesicles may be formed in a wide range of sizes as small as

100 nm or as large as several microns in diameter. Since

this study is primarily concerned with liposomes which most

closely represent the biological structures seen in nature,

unilammelar vesicles will be studied here. The unilammelar

vesicles can be further broken down into two subgroups, the

small unilammelar vesicles (SUVs) and the large unilammelar

vesicles (LUVs). The SUVs are the smallest with a size range

of around 25 nm, up to around 100 nm. The LUVs range from

about 100 nm all the way up to approximately 5 microns in

diameter (1). New (2) has described a subclass of

intermediate liposomes (IUVs) of 100-250 nm diameter.

Figure 2-1. Diagram of Liposome Structure


A variety of techniques has been used in the past for

imaging liposomes at relatively high resolution and

magnification. Each of these methods, however, suffers from

one or more drawbacks which will be elaborated here. Figure

3-1 (Schreier and Erdos personal communication) shows a

micrograph of liposomes which have been chemically fixed and

dehydrated for examination under the scanning electron

microscope (SEM). The scanning electron microscope is

capable of relatively high resolution images of the surfaces

of a variety of a conductive materials; however, the

specimens must be contained within an evacuated chamber, and

therefore, must be completely dehydrated. In this case, the

specimens were dehydrated in a graded series of solvents

which has caused the fusion of the vesicles seen in this

micrograph. Additionally, since the vesicle structure is

quite fragile and subject to deterioration under the heat of

the electron beam, the conductivity was increased by coating

with a gold palladium alloy. While the overall surface

curvature of the liposomes has been preserved and imaged

here, the fusion of the vesicles is therefore seen to be an

artifact of specimen preparation and high resolution of the

surface topography of these structures is obscured by the

presence of the metallic coating. This method then is

suitable for gross examination but is incapable of resolving

fine surface detail or examining the shape of liposomes in

the fully hydrated state.

In an attempt to prevent the fusion of the vesicles seen

in figure 3-1, an alternate specimen protocol was developed

by which the liposomes were dehydrated using an abbreviated

solvent dehydration protocol followed by treatment with

hexamethyldisilazane as a transitional fluid to air drying.

This step was followed by quickly freezing in liquid

nitrogen. The vesicles could then be examined in the frozen

state using the scanning electron microscope without metallic

coating. The advantages of this method are first that

exposure to the graded series of solvents was minimized and

secondarily that the metallic coating which obscures the

surface topography was not used. The results of this

investigation are seen in figure 3-2. As can be seen here, a

wide range of sizes of liposomes are represented and the

spherical conformation has been well preserved. There is

also very little fusion of the vesicles and the surfaces have

not broken down under the heat of the electron beam.

Although thermal drift is a problem using this technique in

an ordinary SEM, cold stage SEMs are available which

alleviates this problem. The specimen still must be

completely dehydrated and visualized in an evacuated chamber;

however, and this, of course, precludes the possibility of

observing liposomes in the hydrated state. In addition, the

SEM is not capable of molecular resolution. Newer,

"environmental SEMs" have been built which may examine

biological specimens at 0.1 atmospheres, but it is likely

that this degree of negative pressure would prove disruptive

to the bilayer structure and the resolution of these

instruments is even more limited than the ordinary SEM.

Another technique which uses the transmission electron

microscope (TEM) is freeze fracture. In this technique, the

liposome structure is quickly frozen and then fractured using

a frozen knife at liquid nitrogen temperatures. The frozen

fractured surface, which may still reflect the hydrated state

in vitreous ice, is replicated by coating with carbon and

shadowing with platinum. The replica is then removed,

cleaned, and examined under the TEM. This has been used to

advantage in many samples of a biological nature to visualize

the fractured interior of the specimens, but resolution is

limited to the particle size of the carbon used to produce

the replica, and even when optimized cannot replicate

molecular topography. As applied to liposomes, this

technique is useful in visualizing the multiple lipid

bilayers in a multilammelar vesicle. Figure 3-3 (Schreier

and Erdos, personal communication) shows such a micrograph of

a cryofractured multilammelar liposome. As can be seen here,

quite a number of phospholipid bilayers are visualized;

however, this method suffers from the necessity that the

specimens must be replicated, which results in a loss of

resolution as discussed above. The transmission electron

microscope has also been used to visualize negative images of

liposomes and the negative stain technique can produce high

resolution images of whole liposomes. In this procedure, the

liposomes are deposited onto a thin film and an electron

opaque material, such as phosphotungstic acid, is allowed to

react with and surround the surface of the vesicles. This

produces an electron opaque region around the surface of the

liposome and shows the structure of the liposome as a

negative image, but again, molecular resolution is not

possible. A negative stain image of liposomes is shown in

figure 3-4 (Schreier and Erdos, personal communication). A

"cryo TEM" can be used to examine whole liposomes in vitreous

ice and can selectively image various parts by use of

subtractive energy filtering, but artifactual images,

possibly produced as a consequence of the thickness of the

spheres, can be a problem; and molecular resolution still is

not possible. From the discussion above, it is apparent that

although many techniques have been developed to visualize

liposomes under near physiological conditions, none exists

which will allow examination of these specimens at room

temperature and pressure while offering the simultaneous

advantage of molecular resolution.

w-mo momm

Figure 3-1. SEM of Liposomes

*~ "

lo I

L\ *x

V &

-1. 4


Figure 3-2. Cryo SEM of Liposomes

Figure 3-3. Cryofracture of MLV


F-' "'I


.. C

Figure 3-4. Negative Stain of Liposomes


The first scanning tunneling microscope was constructed

by Binnig et al. in 1982 (3). The images constructed using

this microscope were the atomic structure of calcium iridium

tin crystals and gold particles. Binnig's microscope was of

course a laboratory model, but since then, commercial models,

most notably those produced by Nanoscope, Inc., have been

produced and distributed worldwide. A number of authors have

examined inorganic material by STM, but organic material has

also been successfully imaged. DNA, in particular, has been

visualized by Lee et al.(4), Amrein et al.(5), Lindsay et

al.(6), and Cricenti et al.(7). Other organic molecules

visualized by STM by Dunlap and Bustamante include

tetranucleotide models of polyhydroxyadenylate (8). The

micrographs obtained by these authors clearly show the

positioning of the imidazole and pyrimidine rings. This is

evidence that the STM could be used to image subparts of DNA

or polynucleotide molecules with future improvement in

technique. Enzymes have also been visualized by STM.

Phosphorylase D and phosphorylase kinase have been imaged by

Edstrom et al. (9). They show linear arrays of dimers, as

well as butterfly-like images. Other research has produced

STM images of biological membranes. Fisher et al. (10) have

imaged the purple membranes isolated from Halobacterium

halobium, and the cell membrane of the oocyte of the clawed

frog Xenopus leavis has been examined under the STM by

Ruppersberg et al. (11). Although in the majority of these

cases the specimens were coated with carbon or a metallic

coating to enhance conductivity, the STM is clearly finding a

growing application in the biological arena for imaging of a

variety of materials at high resolution.

Phospholipid bilayers have been previously imaged by STM

in the desiccated state by Luo et al.(12), and the imaging of

liposomes and phospholipid bilayers described here has been

previously published in abbreviated form (13).


The liposome membrane is actually a bilayer composed of

phospholipid molecules (Fig. 2-1). For the first phase of

experiments, phosphatidylcholine and phosphatidylglycerol

were used as components of this membrane in a ratio of 90%

phosphatidylcholine to 10% phosphatidylglycerol. The

specific molecular structure of these two molecules is

important to an understanding of the overall liposome system

and the mechanism of the technique described here. Figure 5-

1, therefore, shows a diagram of a phosphatidylcholine (PC)

molecule. It consists of a polar head group which contains a

phosphate core. In the case of phosphatidylcholine, there is

a choline residue bearing a plus charge due to the presence

of the nitrogen atom. Since the phosphate portion of the

head group contains a negative charge, the molecule is

zwitterionic and a net zero charge is seen at the head group,

although there are some polar characteristics due to the

separation of the two charges. Attached to the phosphate

portion of the head group are the two hydrophobic fatty acid

tails. These hydrocarbon chains are usually not completely

saturated and may contain one or more double bonds depending

on the source of the molecule, the processing during

extraction, and the treatment of the phospholipid subsequent

to its purification since these hydrocarbon tails are easily

oxidized. The length of these fatty acids is also somewhat

variable, depending on the source, but is usually 16 to 18

carbons long for sources from mammalian systems (2).

Figure 5-2 shows a diagram of the phosphatidylglycerol

(PG) molecule. Although very similar in basic structure to

the phosphatidylcholine molecule, the glycerol head group is

a nonpolar structure without charge. Since the phosphate

portion of the molecule bears a negative charge and is not

influenced by any charge from the glycerol head group, the

overall charge of the phosphatidylglycerol molecule is

negative. The fatty acid chain links are similar to the

phosphatidylcholine molecule and one or more points of

unsaturation will likely be present.

Neither of these molecules is soluble in water, and upon

exposure to this solvent, the molecules will arrange

themselves in monolayers at the air-water interface or

bilayers with the hydrophobic fatty acid tails aligned away

from the aqueous phase. This gives rise, as stated

previously, to the overall structure of the liposome with the

fatty acid tails opposing each other in the center of the

bilayer and the more polar phosphate head groups facing the

interior aqueous cavity and the exterior aqueous phase.

The repulsion of the hydrophobic portions of the

phospholipid molecules away from water is important to the

stability of the liposome as a whole. Dill and Stigter(14)

have shown that the lateral repulsion between the

phosphatidylcholine head groups is quite pronounced since

they bear a similar charge. Although this is somewhat

moderated by the addition of phosphatidylglycerol to the

system, the hydrophylic nature and lateral repulsion of the

PC head groups results in an instability at the surface of

the liposome which can be expected to be increased with

increasing temperature according to the results obtained by

these authors. This mutual repulsion of the

phosphatidylcholine head groups should result in deep valleys

appearing between the phosphate head groups when viewed at

high resolution under the STM. According to the evidence

presented by Dill and Stigter, this valley effect would be

additionally enhanced by the prediction that the nitrogen

part of the dipolar head group will become increasingly

submerged in the hydrocarbon portion of the phospholipid

bilayer with increasing temperature. This should form easily

discernible gaps between phospholipid head groups in the

hydrated state at room temperature. The authors also make

the point that the interactions between

phosphatidylethanolamine molecules are much weaker and have

no significant temperature dependence. This is significant

in these experiments in that a mixed phospholipid array will

probably be more stable and more tightly packed than one

which is comprised primarily of similar molecules such as the

PC/PG system. Part of the experimental results which will be

presented in this paper include scanning tunneling microscopy

of mixed phospholipid liposomes which contain antigenic

proteins as a part of the membrane system. This mixed system

closely approximates a naturally occurring human

immunodeficiency virus cell membrane and is more suitable for

studies involving antigenic responses in natural systems.

The increased packing density of a mixed bilayer system of

this type as predicted by Dill and Stigter is supported by

experimental results reported by Dr. Rao (personal

communication). He measured the ability of water to

evaporate across a monolayer of packed phospholipids as a

part of this study to predict the stability of phospholipid

bilayers, and as shown below, the transmission of water

across a mixed phospholipid monolayer is less than that which

can be evaporated across a PC/PG phospholipid monolayer.



Phosphatidylcholine 6.7 x 10-4

Phosphatidylethanolamine 6.13 x 10-4

Sphingomyeline 6.1 x 10-4

Cholesterol 5.65 x 10-4

Phosphatidylserine 7.0 x 10-4

Phosphatidylglycerol 5.6 x 10-4

Mixed Lipid (Membrane Mimic) 5.65 x 10-4

PC + PG (9:1) 6.1 x 10-4

PC+PG(9:1)[50%]+Chole[50%] 6.7 x 10-4

PC + Chole (1:1) 6.0 x 10-4

Blank (No Monolayer) 7.2 x 10-4

0 U

I I I +
H H Me

Figure 5-1. Phosphatidylcholine Molecule





Figure 5-2. Phosphatidylglycerol Molecule



As was mentioned previously, part of the experimental

design was the examination of liposomes which have had

membrane proteins included as part of the system in an effort

to examine the topography and distribution of such membrane

proteins. The fact that cell membranes are selectively

permeable to some substances and that active transport can

occur across membrane surfaces which can operate against

concentration gradients has lent ample evidence to the fact

that operational proteins should be part of a cell membrane

system. Studies by Lenard and Singer (15) of optical

circular dichroism of membrane preparations in aqueous

dispersion have shown that, indeed, proteins are a component

of cell membrane systems. This has led to a number of

speculations about the actual positioning and confirmation of

these proteins within the phospholipid bilayer membrane.

Most of these models are a version of the so called fluid

mosaic model in which the proteins are either partially or

completely inserted within the bilayer, a portion of which

extends above the surface for recognition of the exterior

environment and another portion of which occupies spaces

between the phospholipid molecules of the bilayer. This

gives rise to a discontinuous model of the phospholipid

bilayer as found in the natural cell membrane. The

discontinuities are thought to be occupied by the relatively

large protein components. The fluid mosaic refers to the

fact that the phospholipid bilayer is relatively fluid,

allowing these proteins to migrate to some extent laterally

and vertically, depending on the portion of the protein which

is hydrophobic and the portion which is hydrophilic. The

mosaic characteristic of the model would refer to the

discontinuities introduced by the presence of these proteins.

Since a portion of the protein would extend down into the

hydrophobic portion of the phospholipid bilayer, this portion

of the protein must be similarly hydrophobic and would serve

as a membrane anchor, as well as a pathway for selective

transport of solutes and cell nutrients. It has been

theorized by Zahler and Weibel (16) that approximately 30% of

the central region of the phospholipid bilayer is occupied by

protein. The remainder of the protein would extend up above

the phosphate head groups of the phospholipid bilayer and

serve as recognition sites for the exterior environment.

Since these models are based on theoretical supposition

founded on indirect evidence, it is important to visualize

these structures directly in order to gain an appreciation of

the topography of the surface of the phospholipid bilayer

when it contains these protein units. Since the scanning

tunneling microscope is easily capable of the resolution


required to visualize these structures, visualization of

these proteins has become a goal of this study.


Cevc and Marsh (17) have approximated the dimensional

features of the phospholipid bilayer in some detail using

statistical averaging methods and transformational matrices.

They also make the point, however, that the temperature and

degree of hydration can substantially modify the dimensional

characteristics of phospholipid bilayers. The mathematical

theory that they used to generate the dimensions of

phospholipid molecules in the bilayer is, however, supported

by x-ray data and maps of electron density. These data show

that for phosphatidylcholine bilayers, the length of the

fatty acid chains is approximately 1.4 nm. The phosphate

head group would extend this dimension slightly in a vertical

direction. These authors also calculate the dimension of the

phosphate head group to be approximately 0.7 nm when viewed

from the top. The approximate 0.7 nm diameter of a phosphate

head group has additionally been confirmed by C. Luo et al.

in a scanning tunneling microscopy experiment visualizing

adsorbed, desiccated phosphatidylcholine bilayers (12).

Cevc and Marsh state that the dimensional spacing of the

phosphate head groups in a phospholipid bilayer is profoundly

affected by the degree of hydration, however (17). These

authors report that the general effect is that the lipid

packing density is loosened in the presence of water due to

the disruptive effect of the adsorbed water between the head

group/head group bonds. This, of course, would result in a

greater spacing between the phosphate head groups and would

produce an effect similar to the one described previously as

reported by Dill and Stigter (14). This loosening allows

more kinking and mobility of the fatty acid chains and

results in an overall thinning of the phospholipid bilayer

(17). These authors go on to say, however, that the precise

effects of hydration are too complex to be accurately

measured and predicted. These effects should therefore be

taken as general trends to be incorporated into the

structures expected to be seen under the scanning tunneling

microscope rather than support for or refutation of specific



The scanning tunneling microscope takes advantage of the

fact that electrons will tunnel through space between a

charged probe tip and the surface of a specimen which has

been grounded. This implies that the specimen must be, at

least to some degree, conductive. At the heart of the

imaging technique described herein is enhancement of the

conductivity of liposomes and phospholipid bilayers using

selected heavy metal fixatives. The magnitude of the charge

passing between the probe tip and the specimen is

proportional to the distance of the tip of the probe to the

sample surface. If the flowing charge (tunneling current)

passing between these two points is held constant, the tip of

the probe must move vertically to a degree which will

describe the topography of the specimen as the tip is scanned

across the surface. Since a stream of electrons is involved

in establishing this current, atomic resolution is possible

between the electron shells of the atoms in the sample

surface. Feedback controls move the tip in a constant

distance from the surface of the specimen by activating

piezoelectric crystals which drive the tip in X, Y, and Z

directions. The data output from the movement of this probe

tip consists initially of parallel raster scan lines showing

the orientation of the probe tip relative to an XY grid and

vertical tip displacement (Z direction). These scan data are

then electronically assembled into a topographical

representation of the specimen surface. Figure 8-1 shows a

diagrammatic view of the STM scan head assembly. For

simplicity, the X and Y piezoelectric crystals are not shown

in this diagram. Below the probe tip is an artistic

conception of a liposome resting on the specimen platform

which, as mentioned previously, is grounded. In order to

obtain good resolution, the probe tip must be extremely fine.

In the ideal case, this is monatomic at the very tip;

however, in actual situations, a somewhat blunter tip is

experienced. The tips are typically mechanically sharpened,

chemically etched or mechanically cut. For the results

presented here, the tips were cut using surgical scissors

held at an approximate 45 degree angle to the axis of the

probe wire. Figure 8-2 shows a diagrammatic representation

of the ideal probe tip of the type used in these experiments.

The angular portion represents the cut surface produced by

the surgical scissors, but as stated previously, the very tip

of this probe is only ideally monotonic, and in the actual

case, several atoms likely occupy the terminal end. This

probe wire is platinum iridium alloy; however, some

instruments also use tungsten. The distance between the

probe tip and the sample surface is typically 1 nm or less,

and the tunneling current is measured on the order of

nanoamperes. This makes the tunneling current and associated

parameters very critical to avoid specimen damage due to

collision of the scan tip with the specimen surface. The

magnitude of the tunneling current which is established

between the probe tip, as stated before, is proportional to

distance, and the following equation describes this tunneling


I( e-2kd

where k is a work function particular to the surface being

examined, d is the distance between the probe tip and the

specimen surface and I is the tunneling current. In the case

of these experiments, a gold specimen platform has been used.

This allows for good electrical contact between the specimen

and the surface, and since the gold is epitaxially deposited,

the topography of this surface is not excessively rough. The

instrument used for these studies was a commercially produced

Nanoscope II (Digital Instruments, Santa Barbara, CA).

Figure 8-3 shows a view of the scan head assembly, which

includes the chrome plated specimen platform and scanning

head; the probe tip is located within the hemispherical

access opening. Figure 8-4 shows an overall view of of the

scan head and the two computers used to construct and analyze

the image. The computer shown in the background is the one


used to control the movements of the scan head and establish

the tunneling parameters, thus constructing the image. The

foreground computer is used to display the image and

manipulate it using tilt, rotation, and color enhancement





Diagram of Scanning Apparatus

Figure 8-1.


PT/IR WIRE 80/20
ALLOY .010" DIA.

Figure 8-2. Diagram of Scanning Probe Tip

Figure 8-3. Scanning Head

Figure 8-4. Scanning Tunneling Microscope


For the initial phase of this study, liposomes composed

of 90% phosphatidylcholine and 10% phosphatidylglycerol

(molar ratio), at approximately 10 micromolar total lipid

concentration in 0.02 M Tris buffer, at pH 7.2, were produced

by an extrusion method using a filter housing connected

between two Hamilton microliter syringes (LiposoFast, Avestin

Incorporated, Ottawa, Canada), as described by MacDonald et

al.(18). In this method, the phospholipids are dissolved in

chloroform and then a dried film is produced on the walls of

a spherical flask by rotary evaporation. The film is

reconstituted in the aqueous buffer described above, which

results in spontaneous formation of large liposomes of a

range of sizes. The size was reduced to a more uniform

population by repeatedly passing the preparation through a

100 nm polycarbonate filter; in this case, 19 passes were


For most specimens, liposomes were impregnated with

osmium tetroxide (0.5% w/v) or uranyl acetate (1% w/v), for a

period of approximately 30 minutes, although periods of up to

two weeks showed no damage to the specimens. Following this

fixation/impregnation step, approximately 10 microliters of

the liposome preparation was withdrawn from the fixation tube

and deposited onto the gold covered mica specimen platform.

As stated previously, the gold was deposited epitaxially onto

freshly cleaved mica surfaces. By electron beam evaporating

99.999% pure gold onto mica surfaces heated to 480 degrees C,

a smooth surface containing gold plateaus of very flat

topography was obtained. In earlier experiments, adherence

of the liposomes to the surface of the gold was of some

concern. Derivatized gold was therefore used for the surface

according to the following technique.

First the discs were inspected under an optical

microscope and those found to be relatively free of scratches

and defects were selected for derivatization. These discs

were then immersed in fresh portions of four mM ethanolic

solutions of alkanedisulfides for 10 minutes. This was

repeated three times. This was followed by 10 immersions in

ethanol to rinse off the disordered assemblies of the

disulfide, and then each disc was air dried overnight. This

treatment resulted in an adsorbed monolayer of the

alkanethiols which would enhance the adherence of the

liposomes to the surface, however, as stated above, this was

found to be unnecessary and was abandoned after the early

phase of experimentation.

Following deposition of the liposome suspension onto the

gold surface, the preparation was allowed to remain in

contact with the gold surface for approximately five minutes.

The excess liquid was then withdrawn using a piece of cut

filter paper. This procedure left the hydrated liposomes

which had been adsorbed to the gold covered mica disc for

examination under the scanning tunneling microscope. These

hydrated specimens were scanned immediately upon withdrawal

of the excess fluid; however, for the dehydrated specimens,

approximately one hour of air drying at ambient temperature

was allowed. These desiccated specimens were then scanned.

The scanning tunneling microscope (Nanoscope II) used

for these studies was produced by Digital Instruments, Inc.

(Santa Barbara, CA). All the specimens were imaged at

ambient temperature and pressure which, as stated earlier, is

an advantage of this type of visualization procedure. In

order to minimize the possibility of contact of the probe tip

with the liposome preparations and the destruction of the

specimens due to this contact, a very small tunneling current

(0.2 nA) was typically used. The bias voltage, which is a

differential between the scan probe tip and the specimen

surface was also held relatively high (between -50 mV and 500

mV). This allowed the tunneling current to be established at

relatively great distance from the surface of the bilayer.

In order to improve the quality of the image, a scan rate of

1.93 Hz was set. All of the scanning for these results was

conducted in constant current height mode using the hand cut

80/20 platinum iridium alloy wire of 0.010" in diameter for

the probe tip. In order to confirm the expected size of the

initial preparations of the 90% PC 10% PG liposomes,

independent size measurements were performed using a Nicomp

model 320 laser particle sizer (Particle Sizing Systems,

Santa Barbara, CA).California.

For the part of the experiment involving liposomes which

had been modified with surface proteins, the method described

by Chander and Schreier (19) was used. Rather than using the

extrusion method described earlier, this method uses dialysis

for production of these liposomes. A blend of phospholipids

closely approximating the percentages found in viral

envelopes such as the human immunodeficiency virus-1 (HIV-1)

was used according to the concentrations listed below.

Phospholipids Mole %

Phosphatidylcholine (PC) 23.7

Phosphatidylethanolamine (PE) 22.6

Sphingomyelin (SM) 28.1

Phosphatidylserine (PS) 25.7

This blend of phospholipids was then mixed with

cholesterol to form a 50% phospholipid/50% cholesterol final

preparation. This preparation was dissolved in chloroform

and the resulting solution was added to a Tris buffer at pH

7.0 at a concentration of 2 mg per ml. Also included in this

buffer was sodium deoxycholate at a ratio of 45:1 (molar

ratio). This served to solubilize all of the lipids and

dialysis was carried out over a period of 48 hours using a

tubular dialysis membrane with a molecular weight cut-off of

1000 daltons. The resulting liposomes were then partially

solubilized using a solution of sodium deoxycholate in a

detergent lipid molar ratio of 5:1. This effectively relaxed

the packing of the bilayer to facilitate the insertion of the
protein. Then 100pl of a 1000 pg/ml solution of gpl60

protein (an HIV surface glycoprotein) was added to 3 mL of

liposome suspension and the preparation was incubated for one

hour at room temperature to allow the protein to insert into

the bilayer by natural hydrophobic affinity. This mixture

was then dialyzed for another 36 hours in order to remove the

detergent and repack the bilayer around the inserted protein.


As has been mentioned previously, one of the primary

aspects of this technique has been increasing the

conductivity of this relatively nonconductive biological

structure by fixation with osmium tetroxide or uranyl

acetate. Osmium tetroxide was used for the majority of the

specimen preparations, and this is particularly suitable for

a preparation of this type because it readily reacts with

points of unsaturation on fatty acid hydrocarbon chains.

Hayat (20) has described the reaction of osmium tetroxide

with these unsaturated hydrocarbons in some detail. He

describes the formation of a stable diester which involves

the reaction of a single osmium atom with two fatty acid

chains. This configuration is shown in figure 10-1, and as

can be seen in this diagram, not only does the osmium

tetroxide provide a bridge of conductivity within the lipid

region, but it also provides some mechanical stability due to

its cross linking diester oxidation behavior. This fixative,

then, is particularly suitable for an experimental technique

of this type since the liposome structure is subjected to

considerable mechanical stress after deposition onto the gold

surface and scanning with the STM. Since osmium tetroxide is

nonreactive with saturated hydrocarbons, the reaction was not

thought to be excessively disruptive to the phospholipid

bilayer. In order to investigate this more fully from a

thermodynamic standpoint, differential scanning calorimetry

was performed on liposomes which had been reacted with osmium

tetroxide as compared with those without osmium tetroxide.

As can be seen in figure 10-2, the curves for the

thermodynamic behavior of the two systems are identical.

This is further evidence that the fixation treatment with

osmium tetroxide, while cross bridging the hydrocarbon

lipophilic region to some extent, is not so extensive as to

disrupt the structure and thermodynamic behavior of the

system as a whole.

Uranyl acetate was also used as a fixative for certain

specimens in this study, particularly when higher resolution

images of the surface of the liposome were desired. This

fixative reacts primarily with the phosphate head groups and

provides a path of conductivity closer to the surface which

allows a closer approach of the STM probe tip and higher

resolution imaging of the phosphate head groups. Shah (21)

has described the reaction of the uranyl acetate ion with

these phosphate head groups by measuring the change in

electrical potential and has confirmed not only reaction with

these groups, but an increase in the stability of a monolayer

which has been reacted with uranyl acetate. Like osmium

tetroxide then, uranyl acetate has the potential to form a

polymeric lattice which not only increases conductivity of

the surface of liposomes through the phosphate head groups

intercalated with uranium atoms, but also an increase in

mechanical stability due to the influence of this ion between

the phosphate head groups. The product of this reaction is

most likely uranium phosphate, and since the diameter of the

uranium atom is only 1.2 to 2.0 angstroms in diameter,

depending on electron valence configuration, the size of this

ion is not thought to be a significant factor in imaging of

the liposome surface, although its presence could certainly

be detected in a micrograph of high resolution. Figure 10-3

shows diagrammatically the reaction of uranyl acetate

(uranium atoms), with the phosphate head groups in the

surface of a phospholipid bilayer.

Unfortunately, the combination of both of these

fixatives on the liposome structure proved to be destructive

to the system as a whole, therefore, they could not be used

together but were used separately to achieve an increase in

stability of the liposome and increased conductivity which is

critical to successful imaging of materials under the

scanning tunneling microscope at high resolution. Although

uranyl acetate provided the best imaging of the surface at

high resolution, this fixative was inferior to osmium

tetroxide for structural preservation. Therefore, as stated

above, osmium tetroxide was used for the majority of the

preparations, but uranyl acetate was used to advantage in at


least one case to examine the configuration of the phosphate

head groups in the hydrated state at high resolution.

Figure 10-1. Reaction with Osmium Tetroxide






\ \ s i

.0 0

.\ \ 8 r Q

0 I I I

L -L
?\ 8 Pc .S

(M) /AQ- I Y3H o
0 C '.

head group

Uranium Atom
bc i

head group

Figure 10-3. Reaction with Uranyl Acetate





The initial results shown here represent liposomes in

the hydrated state. As was stated before, these were scanned

immediately after deposition onto the surface of gold

plateaus, and figure 11-la shows such a liposome using the

scan line option of the image generating software. Figure

11-lb illustrates this image in cross section. The scan

lines seen represent the raster pattern of the probe tip as

it passes over the specimen and the specimen platform. Note

that the liposome appears as a hemisphere because the path of

the probe tip is such that it cannot probe underneath a

concavity or overhanging edge. The liposome is collapsed

under its own weight although the smooth contour indicates

that it is in the completely hydrated state.

Figure 11-2 shows the same liposome as seen in the

normal image mode of surface topographical observation. Note

in this case that the gold plateaus are quite different in

structure from the liposome indicating that this is indeed a

liposome structure and is not part of the gold plateau


The scanning tunneling microscope image manipulation

module is capable of generating two types of contour lines to

examine the contours of the surface of specimens. Figure 11-

3 shows the same liposome seen in figure 11-2 but with

contour line enhancement. Note in this case that the contour

lines are smooth and regularly spaced, confirming the well

hydrated state of the liposome. Again, the gold plateaus are

seen to have quite different topographical characteristics

than the liposome itself.

Figure 11-4 shows an overhead view of this same

liposome, and this is typically the initial view that is seen

when first scanning the specimen platform which has been

treated with a liposome preparation. The topographical

representation is shown in the scale to the right, and as can

be see here, the liposome protrudes significantly above the

gold plateau background as identified by its increased

density of white coloration.

The other type of contour lines which the scanning

tunneling microscope can assign to a specimen are seen in

figure 11-5. Here, the same overhead view as seen in figure

11-4 is shown, however, to enhance the vertical perspective

of the liposome, the thin contour lines have been assigned to

various areas of the specimen at constant height. Note that

the gray scale still indicates a dramatic relief of the

liposome, however, in this case, the contour lines further

enhance the vertical perspective of the structure. As can be

seen here, the contour lines are smooth and well spaced,

again indicating the round, fully hydrated characteristics of

the liposome.

In order to further confirm the fact that the structure

seen is indeed a liposome and not an anomaly in the gold

plateau surface, a transect was taken of the image as seen in

figures 11-4 and 11-5. Figure 11-6 shows this transect. The

lower left portion of the micrograph shows the positioning of

the transect line, and above this image is seen the cross

section as described by this transect. As can be seen in

this micrograph, the overall diameter of the liposome is

approximately 100 nm. This is confirmed by independent

analysis by Nicomp which shows the distribution of sizes of

liposomes from this same preparation. Note that the greatest

fraction of liposomes as analyzed by this independent method

is in the 108 nm category (figure 11-7). This corresponds

well with the diameter seen in figure 11-6 which is a

transect of the liposome seen in figures 11-1 through 11-5,

and it confirms again that the structure seen is actually due

to experimental deposition of the liposome preparation onto

the gold plateaus.

A lower magnification view of epitaxially deposited gold

plateaus with liposomes is seen in Figure 11-8. This is

typical of the initial views seen in the scanning tunneling

microscope before zooming in on a particular portion of the

specimen plateau for examination of individual liposomes.

In order to examine the phosphate head groups of the

surface of a hydrated liposome in greater detail, a uranyl

acetate fixed specimen was used. Figure 11-9 shows the high

resolution/high magnification view of this surface. Note in

this case that the bias has been increased to allow the

scanning probe tip to move at a greater distance from the

specimen surface, thereby minimizing damage. Within the

micrograph seen here is a small box which describes the area

which has been examined in greater detail in figure 11-10.

After manipulation of the image to convert the overhead view

seen in figure 11-9 to a surface view seen in figure 11-10,

it is evident that the surface topography shows a number of

hills and valleys corresponding to phosphate head groups and

possibly to intervening presence of some uranium atoms. Note

that the size of the head group seen in figure 11-10,

according to the scale, correlates well with that reported

previously by other authors (12, 17). Some evidence of the

relatively wide spacing as reported by these authors can also

be seen in this micrograph. A transect of this view seen in

figure 11-11 more clearly shows the size of the phosphate

head groups and confirms the correlation. Note in this

micrograph that the diameter of the head groups is

approximately 0.7 nm which is in good agreement with the

values as predicted by the authors cited previously.

One of the artifacts of specimen preparation which can

be seen in the scanning tunneling microscope is seen in

figure 11-12. This micrograph shows liposomes which have

fused on the specimen platform, presumably due to mechanical

pressures of being in close proximity to one another during

the withdrawal of the fluid within which they were suspended.

Note in this micrograph that although the contours of the

liposome surface are smooth, the irregular topography of the

lines in a lateral dimension shows evidence of fusion, and

the specimen to the right shows a more aggregated combination

of liposomes, the far side of which is not visible in the

micrograph. It was experimental results such as this which

allowed optimization of the concentration of liposomes in the

buffer and an eventual density or concentration of 10

micromolar phospholipid was arrived at, which provided the

best results.

As the liposomes were allowed to desiccate on the

specimen platform, a general collapse of the liposome

structure could be expected. Figure 11-13 shows a liposome

in the early stages of collapse. Note in this case that the

scan lines which describe the surface of the specimen show a

relatively flat apical surface corresponding to an initial

depression of the phospholipid bilayer since the water within

has partially evaporated. This figure shows that

predictably, the top of the liposome, which encounters the

greatest mechanical stress and which can deform with the

least compromise in the overall liposome structure will be

the first part to show physical evidence of dehydration.

Figure 11-14 shows liposomes in a slightly greater state of

dehydration. Here, the sides of the liposomes as shown by

the dislocation of the contour lines have begun to sink in

due to desiccation of the specimens. Figure 11-15a shows

liposomes in an even greater state of dehydration. This is

accompanied by an artist's conception of the structure in

cross section (figure 11-15b). Notice here that globular

structures have begun to appear around the sides of the

liposomes as described by the contour lines. This would be

the most efficient means of maintaining the overall spherical

confirmation with removal of interior water. The contour

lines in figure 11-16 show liposome structures even more

severely dehydrated. In this case, severe globular

characteristics can be seen around all edges of the liposomes

and some aggregation and fusion of the vesicles has taken

place. The gold plateaus of the specimen platform can still

be seen in this micrograph underneath the largely desiccated

liposome structures.

When a liposome is allowed to go to complete desiccation

at room temperature and pressure (approximately one hour),

the structure seen in figure 11-17a is seen. An artistic

rendition of a probable cross section is shown in figure 11-

17b for increased clarity. Note in this case that the center

of the liposome has completely caved in due to removal of the

water in the interior cavity. The exterior structure has

taken on a completely globular topography. Note that these

globules are approximately 20 nm in diameter which

corresponds with the smallest possible size of a vesicle due

to the interior constraints of the phosphate head groups

pressing against one another (2). Although this liposome can

be presumed to be almost completely dehydrated at a gross

level after one hour of exposure to ambient conditions, the

water of hydration would still be present in the hydrophobic

layer since this is adsorbed very strongly to the hydrocarbon

tails. Figure 11-18 shows the same image but with contour

line enhancement. These contour lines show more clearly the

size of the globular structure which, as mentioned before, is

approximately 20 nm in diameter. Although this dimension may

vary slightly depending on the temperature and degree of

hydration, it can be seen here to clearly correlate well with

the minimum size of phosphatidyl choline vesicles as reported

previously by Huang (22). Figure 11-19 shows another

desiccated liposome and further confirms that this globular

topography is generally seen when a specimen is allowed to

desiccate on the specimen platform under ambient conditions

for at least one hour.

Prior to an examination of liposomes containing

proteins, a brief imaging was performed on liposomes

containing the blend of phospholipids corresponding to a

viral membrane as described previously in materials and

methods. An image of such a liposome is shown in figure 11-

20. Note in this case that the smooth contour of the

topography shows that the structure is indeed capable of

withstanding the stresses of the procedure and that a fully

hydrated liposome of this type can be imaged using the STM.

After incorporation of the membrane proteins, according

to the dialysis procedure described in Materials and Methods,

great difficulty was encountered in obtaining an STM image of

an intact liposome. This difficulty is probably due to the

inherent instability of a phospholipid bilayer which has been

partially disrupted by secondary dialysis and the insertion

of relatively large protein molecules. The only instance in

which a relatively intact liposome could be imaged is shown

in figure 11-21. This micrograph shows the majority of the

surface of a liposome and, as can be seen here, parts of the

liposome surface show protrusions which would correspond to

parts of proteins projecting above the phospholipid bilayer.

This is in agreement with the fluid mosaic model discussed

previously, and has also been seen in freeze fracture

transmission electron microscopy by another investigator

(Natasa Skalko personal communication). Such a TEM image is

seen in Figure 11-22. Note the strong correlation between

the surface topography seen in this micrograph and the

surface topography seen in the STM image in Figure 11-21.

This serves as support that the image in micrograph 11-21 is

indeed an STM image of a liposome which contains antigenic

proteins as part of the phospholipid membrane. This

technique, then, could be useful in future studies to

visually further define the distribution and confirmation of

such proteins in a variety of liposomes and using a variety

of different proteins.

Figure 11-1. Hydrated Liposome

Figure 11-la. STM Image

Figure 11-lb. Artist's Conception

aXNa i%

Figure 11-2. STM of Hydrated Liposome

Figure 11-3. STM of Hydrated Liposome (Contour)

Figure 11-4. Overhead STM of Hydrated Liposome

Figure 11-5. Topographic STM of Hydrated Liposome

euiosodTq pa~21pAH J0o oesuJj, *9-I aFjn6i

U .. ...... ..... N



1IllI I I I
I -
o 0
1 0

no 0
o l


-NAUJ4ww "wrQ r -h 3

(D m X<

o 0

Figure 11-8. Liposomes on Gold Plateaus

Figure 11-9. Close View of Liposome Surface

Figure 11-10. Molecular Topography of Liposome Surface

Figure 11-11. Transect of Molecular Topography

Figure 11-12. Aggregated Liposomes

Figure 11-13. Partially Collapsed Liposome

Figure 11-14. Partially Desiccated Liposomes

Figure 11-15. Partially Desiccated Liposomes

Figure 1l-15a. STM Image

Figure 11-15b. Artist's Conception

Figure 11-16. Partially Desiccated Liposomes

Figure 11-17. Completely Desiccated Liposome

Figure 11-17a. STM Image

Figure 11-17b. Artist's Conception

<&p *t*

Figure 11-18. Completely Desiccated Liposome Topography

Figure 11-19. Completely Desiccated Liposome

Figure 11-20. Contour Image of Virosome Shell

Figure 11-21. STM of Virosome Surface

Figure 11-22. Freeze fractured liposomes
Courtesy Natasa Skalko (Center for Drug Delivery Research)


This dissertation illustrates that the scanning

tunneling microscope can be used to image whole liposomes in

the hydrated and desiccated state. Successful imaging of

these structures is dependent upon enhancement of

conductivity and possibly upon enhancement of structural

stability of the liposomes using fixatives such as osmium

tetroxide and uranyl acetate which have an affinity for

various parts of the liposome structure and increase

conductivity due to their heavy metal character. Since this

is the first application of these types of fixatives to

enhance the imaging capabilities of the scanning tunneling

microscope with biological structures of this type, a variety

of applications could be projected for future investigations.

One application of this technique could be to further

the investigation of the phospholipid films containing

cholesterol hemisuccinate. Talsma et al. (23) state that the

addition of this phospholipid to a preparation of liposomes

not only induces a charge, but stabilizes geometrical packing

and structural stability of the resulting liposomes. This

could be further investigated using the scanning tunneling

microscope after treatment via the methods described above.

Thus far, the authors base their conclusions on the size of

the vesicles formed rather than actual analysis of the

surface of the vesicles.

Another study conducted by Ondrias et al. (24), seeks to

confirm the effect of calcium channel blockers on the

phosphotidylcholine bilayer. These authors use spectroscopy

to describe the disordering effect of calcium channel

blockers on these types of membranes. They state that

concentration is an overall influence on the disordering

effect, however, the spectroscopic method is indirect and

also requires labeling at the 16 carbon position. A study of

this type could be enhanced by utilization of the techniques

described above since the scanning tunneling microscope could

easily identify disordering of the phospholipid bilayer.

Another study reported by Otoda et al. (25), elaborates

on the change in orientation of a glycopeptide in the lipid

bilayer membrane. This has been induced by lectin binding

and has been detected by a fluorescent probe attached to the

N terminal and a lactose unit at the C terminal. The authors

state that enhancement of the fluorescence intensity and

reduction of fluorescent quenching increases with the

addition of lectin and that these results indicate that the

peptide segment prefers a more perpendicular orientation to

the membrane upon association with lectin. Again, this

study, although presenting strong evidence, utilizes indirect

techniques. If such a preparation were synthesized and then

subjected to analysis by scanning tunneling microscopy

utilizing the methods described above, a visual record could

be obtained of the change in orientation of the G8.peptide

segment directly.

The interaction of proteins with the phospholipid

bilayer discussed above concerns those which are anchored to

some extent within the phospholipid bilayer, but which also

protrude above the surface of the membrane. A study has also

been reported that concerns the behavior of Staphlococcus

proteins. Tomita et al. (26) state that these form channels

within a phospholipid bilayer according to the theory of the

authors. These authors report on the leakage of the interior

contents of a liposome membrane after treatment with zinc,

cadmium, calcium, and magnesium ions, however, the method

used to study the behavior of these supposed channel forming

proteins is again indirect (i.e. the leakage of

carboxyfluorescein from the interior). Scanning tunneling

microscopy could confirm such a channel forming phenomenon

directly by visualizing the surface of the liposome membrane

after treatment with the Staphlococcus protein.

Another pore forming study elaborates on the pore

forming peptides from pathogenic and nonpathogenic species of

Entamoeba histolytica. The authors, Leippe et al. (27),

state that a pore forming peptide was isolated from a

nonpathogenic isolate. Although the authors project the pore

forming capabilities of the peptides based on structural

analysis of the molecues, an analysis by scanning tunneling

microscopy of the surface or the liposome membrane could

confirm the activity of these molecules beyond question.

It is evident from the foregoing that the scanning

tunneling microscope can be used to good advantage in

examination of phospholipid bilayers and intact liposomes;

even those in the completely hydrated state. By utilizing

the techniques described above, low resolution images of

liposomes as a whole and high resolution images of the

liposome surface may be obtained. The high resolution images

particularly may be of use in examining the arrangement and

spacing of the phospholipid molecues in the surface of the

phospholipid bilayer. This technique, then, could be of

considerable value in examination of liposomes which have

been modified by the inclusion of proteins as reported here

and other modification techniques as described above.


1 Crommelin, D.J.A. and Schreier, H. in: Colloidal Drug
Delivery Systems, J. Kreuter, ed., Marcel Dekkes, New
York, 1994, pp. 73-190.

2. New, R.R.C., Liposomes a Practical Approach, IRL Press,
Oxford University, Oxford, 1990, p. 23.

3. Binnig, G., Rohrer, H., Gerber, Ch., and Weibel, E.,
Surface studies by scanning tunneling microscopy, Phys.
Rev. Lett. 49:57-61(1982).

4. Lee, G, Arscott, P.G., Bloomfield, V.A., Evans, D.F.,
Scanning tunneling microscopy of nucleic acids, Science

5. Amrein, M., Durr, R., Stasiak, A., Gross, H., Travaglini,
G., Scanning tunneling microscopy of uncoated recA-DNA
complexes, Science 242:1708-1711(1989).

6. Lindsay, S.M., Nagahara, L.A., Thundat, T., Oden, P.,
Sequence, packing and nanometer scale structure in STM
images of nucleic acids under water, J. Biol. Struct.
Dvn. 7:289-299(1989).

7. Cricenti, A., Selci, A., Felici, A.C., Generosi, R.,
Gori, E., Djaczenko, W., Chiarotti, G., Molecular
Structure of DNA by scanning tunneling microscopy,
Science 245:1226-1227(1989).

8. Dunlap, D.D, Bustamante, C., Images of single-stranded
nucleic acids by scanning tunneling microscopy, Nature

9. Edstrom, R.D., Yang, S.R., Lee, G., Evans, D.F., Viewing
molecules with scanning tunneling microscopy and atomic
force microscopy, FASEB J. 4:3144-3151(1990).

10. Fisher, K.A., Whitfield, S.L., Thomson, R.E., Yanagimoto,
K.C., Gustafsson, M.G., Clarke, J., Measuring changes in
membrane thickness by scanning tunneling microscopy,
Biochim. Biophys. Acta 1023:325-334(1990).

11. Ruppersberg, J.P., Horber, J.K., Gerber, C., Binnig, G.,
Imaging of cell membraneous and cytoskeletal structures
with a scanning tunneling microscope, FEBS Lett. 257:460-

12. Luo, C., Zhu, C., Ruan, L., Huang, G., Dai, C., Cheng,
Z., Bai, C., Su, Y., Xu, S, Lin, K., Baldeschwieler,
J.D., Scanning tunneling microscopy of the
phosphatidylcholine bilayers, J. Vac. Sci. Technol. A

13. Fowler, K., Bottomley, L., Schreier, H, Surface
topography of phospholipid bilayers and vesicles
(liposomes) by scanning tunneling microscopy (STM), J.
Cont. Rel. 22:283-291 (1992)

14. Dill, K.A., Stigter, D., Lateral interactions among
phosphatidylcholine and phosphatidylethanolamine head
groups in phospholipid monolyers and bilayers,
Biochemistry 27:3446-3453(1988).

15. Lenard, J., Singer, S.J., in: Form and Function of
Phospholipids, Chap. 8, G.B. Ansell, J.N. Hawthorne,
R.M.C. Dawson, eds., Elsevier Scientific Publishing
Company, Amsterdam, 1973, p. 183.

16. Zahler, P, Weibel, E.R., in: Form and Function of
Phospholipids, Chap 8, G.B. Ansell, J.N. Hawthorne,
R.M.C. Dawson, eds., Elsevier Scientific Publishing
Company, Amsterdam, 1973, p. 183.

17. Cevc, G., Marsh, D., Phospholipid Bilayers Physical
Principles and Models, John Wiley & Sons, New York, 1987,
pp. 86-90, 334.

18. MacDonald, R.C., MacDonald, R.I., Menco, B.P.M.,
Takeshita, K., Subbarao, B.K., Hu, L., Small-volume
extrusion apparatus for preparation of large, unilamellar
vesicles, Biochim. Biophys. Acta 1061: 297-303(1991).

19. Chander, R., Schreier, H., Artificial Viral Envelopes
Containing Recombinant Human Immunodeficiency Virus (HIV)
gp 160, Life Sci. 50:481-489(1992).

20. Hayat, M.A., Principles and Techniques of Electron
Microscopy: Biological Applications, Van Nostrand
Reinhold Company, New York, 1970, 1: 40-41.

21. Shah, D.O., Interaction of Uranyl Ions with Phospholipid
and Cholesterol Monolayers, Colloid Interface Sci.

22. Huang, C., Mason, J.T. Geometric packing constraints in
egg phosphatidyl-choline vesicles. Proc. Natl. Acad. Sci.
USA 75:308-310(1978).

23. Talsma, H., Gooris, G, v.Steenbergen, M.J., Salomons,
M.A., Bouwstra, J., Crommelin, D.J., The influence of the
molar ratio of cholesterol hemisuccinate/ dipalmitoyl-
phosphatidylcholine on liposomee' formation after lipid
film hydration, Chem. Phys. Lipids 62:105-112(1992).

24. Ondrias, K., Ondriasova, E., Stasko, A., Perturbation
effect of eight calcium channel blockers on liposomal
membranes prepared from rat brain total lipids, Chem.
Phys.Lipids 62: 11-17(1992).

25. Otoda, K, Kimura, S., Imanishi, Y., Orientation change of
glycopeptide in lipid bilayer membrane induced by lectin
binding, Biochim. Biophys. Acta 1145:33-41(1993).

26. Tomita, T., Watanabe, M., Yarita, Y., Assembly and
channel-forming activity of a natural-occurring nicked
molecule of Staphylococcus aureus alpha-toxin, Biochim.
Biophys.Acta 1145:51-57(1993).

27. Leippe, M., Bahr, E., Tannich, E., Horstmann, R.D.,
Comparison of pore-forming peptides from pathogenic and
nonpathogenic Entamoeba histolytica, Mol. Biochem.
Parasitol. 59:101-109(1993).


The author was born in Atlanta, Georgia, on September

10, 1952. After receiving a B.S. in zoology in 1978 and an

M.S. in zoology in 1984 from the University of Georgia, he

worked for a number of years in various aspects of the

pharmaceutical industry. He enrolled in the Pharmaceutics

Department at the University of Florida in the fall of 1990.

This degree program led to a Ph.D. degree which will serve to

enhance his ability to contribute to the pharmaceutical

sciences. He was married in the fall of 1992 and anticipates

employment in pharmaceutical industry or academia.

I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality, as
a dissertation for the degree of Doct r of P osophy.

Ha 'h chreier, Chair
Associate Professor of

I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality, as
a dissertation for the degree of Doctor of philosophy.

Hartmut Derendorf, Cochair
Professor of Pharmaceutics

I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, /i scop ad quality, as
a dissertation for the degree of Dc r f h' o hy.

'Richard Hammer
Professor of Pharmaceutics

I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality, as
a dissertation for the degree of Doctor of Philosophy.

Dinesh 0. Shah
Professor of Chemical

I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and, qu ty as
a dissertation for the degree of Doc or of Phi shy.

Associ t pientist of
Cente fo Biotechnology

This dissertation was submitted to the Graduate Faculty
of the College of Pharmacy and to the Graduate School and was
accepted as partial fulfillment of the requirements for the
degree of Doctor of Philosophy. r ;

April, 1994 _______'_-_ ____ ____

Dean, Graduate School

3II I 1 11111262 08557 04621111111111
3 1262 08557 0462