A COMPARATIVE STUDY OF NIOSOMES

(NON-IONIC SURFACTANT VESICLES)

AND LIPOSOMES: THEIR STABILITY

IN BIOLOGICAL ENVIRONMENTS.

A THESIS

submitted to

THE UNIVERSITY OF STRATIICLYDE

by

LISBETH R. HUME

in fulfilment of the requirements of the degree

of

DOCTOR OF PHILOSOPHY

October 1987.

DEPARTMENT OF PHARMACY

UNIVERSITY OF STRATHCLYDE

204, GEORGE STREET

GLASGOW GI IXW.
I would like to thank Dr. Alan J. Baillie, my supervisor and Professor Alexander

T. Florence, my head of Department, for their support, encouragement and helpful

discussions during the course of this work. I extend my sincere thanks to all my

friends and colleagues in the Department of Pharmaceutics.

I am greatly indebted to my wonderful parents, Tom and Jean, whom I love very

much. They have continually encouraged me and this thesis is dedicated to

them.

Finally, an extra special thanks to Mohan, my dearest friend.

 Vj "" Me,
aR m w-vac, e/zod
Xcdujt, 'ýk1 i6/ uwni&mF cv-ek
e, ck4 mad44t/tiecc,
ya" ?:!
ZA44
ýL 1682.
This thesis is divided into four main areas; Introduction, Experimental, Results and

Discussion and Conclusions and Future Work.

The work presented here is part of a major study on novel drug delivery systems

undertaken within the Department of Pharmaceutics, University of Strathclyde.

This project details the study of a new generation of surfactants in the

formation of vesicles suitable for drug delivery. The well-documented liposome

system was used as a model throughout this work.

 TABLE-OF CONTENTS

INTRODUCTION

1.1 PRINCIPLES OF DRUG

DELIVERY I
............ ...
Controlled ReleaseSystems ........ I
...
Drug Carriers 2
............ ...
Drug Targeting 3
........... ... 4
Biological Strategy ...
..........
1.2. APPLTCATIONS-OFDRUG
DELIVERY TECHNOLOGY ........ 5
...
Enzyme therapy 5
........... ... 6
Cancer Chemotherapy ...
.......... 7
Antifungal Agents ...
........... 9
Antimicrobials ...
............ 9
Diagnostic Radiology
........ ... ... 10
MiscellaneousUses ...
..........
1.3. UPOSOMES:
PRE ARATION. STABILITY.
INTERACTION S WITH BODY FLUI DS A N D
- CELLS 10
..... ...
1.3.1. PREPARATION OF LIPOSOMES..... 13
...
Multilamellar Vesicles 13
.......... ...
Small UnIlamellar Vesicles 13
........ ...
Large Unilamellar Vesicles 14
........ ....
1.3.2. SEPARATION OF FREE DRUG ..... is
...
Dialysis 15
............. ...
Centrifugation 16
............ ...
Gel Filtration 16
............ ...
Ultrafiltration 17
............ ...
1.3.3. COMPOSITIONS AND TISSUE
DISTRIBUTION OF LIPOSOMES 17
...
1.3.4. STABILITY OF LIPOSOMES 20
...... ...
Measurementof Stability In Biological
Fluids 23
.............. ... 23
FluorescentMarkers
.......... ...
1.3.5.LIPOSOME INTERACTIONS
WITH BLOOD PROTEINS 28
........ ....
Surface Potential and Its
Determination .**i,, 32
Measurementof the Zeta Potential
from Electrophoretic Mobility 35
....... .... 37
Calculation of Surface Charge ....
...... 39
Gel Electrophoresis ......... ....
1.4. INTERACTION- OF LIPOSOMES
WITH CELLS ............ 40
... 45
PhospholipaseA2 ........... ...
Carboxylic Ester Hydrolase ........ 47
...

1
1.5. NON-IONIC SURFACTA
VESTCLES 47
................
1.5.1.NIOSOMES:POXENTIAL DRUG
CARRIERS 50
............... 52
Cosmetic Application .............

AIMS OF PROJECT ............. 53

EXPERIMENTAL

2.1. MATERIALS 54
........... ...

2.2. METHODS 55
........... ...
2.2.1. PRODUCTION OF VESICLES...... 55
... 55
Ether Injection (E. I. ) Nlosomes ....... ... 55
Hand Shaken (H. S. ) Nlosomes ....... ...
H. S. Uposomes 56
........... ...
"Negativel y- Charged" Vesicles 56
....... ...
"Posi tivel y- Charged" Vesicles 56
........ ...
Purification CF 56
of ........... ...
CF Solution 57
............ ...
Buffered CF 57
............ ...
Separation of Free and Entrapped CF ...... 57
...

2.2.2. STABILITY OF VESICLES 58

..... .... 58
Leakage of CF ....
........... 58
Ef fect of pH ....
........... 59
Effect of temperature ......... ....
Preparation of Human Serum....... 59
.... 59
Effect of Plasma .......... ....
Effect of 10% BSA ......... 59
....
2.2.3. MEASUREMENT OF SURFACE
POTENTTAL .... 60
....
2.2.4. IDENTIFICATION OF
ADSORBED PROTEINS ........ 60
.... 60
Gel Electrophoresis ....
......... 61
Sample Preparation *,.. ....
... *,
Cleavageby Trypsin, EDTA, Urea and
IM NaOH 61
............ .... 61
Slab Gel Electrophoresis ....
........
Gels 62
Staining and Destaining of
...... .... 62
Drying Gels ....
........... 62
Photography of S.talned Protein Bands . ....
.... 63
Electroblotting ....
........... 63
Staining and Destaining of Blots ....
......
Immunological Detection of Proteins
63
on Nitrocellulose .......... ....

2.2.5. IN VIVO UPTAKE AND

DEGRADATION BY CELLS 64
....... ....
Effect of Enzymes on Vesicles Qq
65
, Uw ............. ....

ii
 RESULTSý

3.1. OF NIOS 66
-PRO12UCTION
Quantification of Entrapment. 68
Removal of Unentrapped CIF . 68

71
Buffer 71
pH Ef fects ....... 79
Temperature 86

89
Factors Affecting Plasma-Vesicle
Interaction 97
.......
Polyacrylamide Gel Electrophoresis 99
Electrophoresis 109
......
4
116

CONCLIISIONS AND FUTURE WORK

4.1. GENERAL CONCLUSIONS 133

..........
Specific Conclusions 134
............ ..
4.2. FUTURE WORK 136
.............
4.3. PEREECTIVES 137
.............
REFERENCES& APPENDI!QES

References 140
......... .......
Appendix I......... 161
.......
Appendix 2......... 162
.......
Appendix 3......... 164
.......

iii
.
 LIST OF ABBREVIATIONS

BSA bovine serum albumin

CF carboxyfluorescein

DCP dicetyl phosphate

DMPC dimyristoylphosphatidy1choline

DNA deoxyribonucleic acid

DPPC dipalmitoylphosphatidy1choline

DSPC distearoylphosphatidy1choline

DTPA diethylenetriaminepentaacetic acid

EDTA ethylenediaminetetraacetic acid

El ether injection

HS hand shaken

IgG immunoglobulin G

LUV large unilamellar vesicle

M. Wt. molecular weight

MLV multilamellar vesicle

NaCl sodium chloride

PAGE poylacrylamide gel electrophoresis

PC phosphatidy1choline

PG phosphatidylglycerol

PS phosphatidylserine

R* registered trade mark

RES reticuloendothelial system

SDS sodium dodecyl sulphate

SUV small unilamellar vesicles

Tc transition temperature
Submicron sized vesicles consisting of single and double chain non-ionic surfactant

mixtures were prepared by simple dispersion of surfactant dissolved in aqueous

medium, or alternatively, injecting the surfactant dissolved in organic solvent

into the aqueous phase.

Drug entrapment values were measured by using a fluorescent marker, 5,6-

Carboxyfluorescein, and drug release characteristics were evaluated in biological

media (serum and plasma) as a function of surfactant composition and in the

presence or absence of cholesterol. Surface charge measurements, zeta-potential,

as a function of pH, gel electrophoresis and immunoblotting (ELISA) were

performed in order to measure the interaction of components of the biological

fluid with the prepared vesicles. It was found that all vesicles carried a negative

charge and rapidly bound plasma protein, which included albumin and

immunoglobulin G, thus affecting the latency of the entrapped marker.

Uptake and degradation of niosomes (non-ionic surfactant vesicles) in a living,

unicellular, eukaryotic micro-organism was also investigated. It was found that the

rate of release of contents depended on the composition of the vesicles and was a

function of enzymatic degradation within these organisms rather than an

intracellular PH effect of the digestive organelle.

An identical protocol was carried out with the well- characterised liposome system

and their inherent stabilities under a variety of conditions directly compared with

niosomes.
 SECTION 1

INTRODUCTION
The past fifty years have witnessed major advances in the control of disease

brought about with the use of drugs. These advances are particularly apparent in

the treatment of infectious disease by means of vaccines and antibiotics. Some

successful anti-cancer drugs are available but the overall failure of cancer

chemotherapy has been an important stimulus to drug delivery research.

With accessible targets, it is possible to directly "titrate" the drug to the need of

the patient on the basis of biological response. Temporal administration of drug in

these situations is straight forward. However, when the target tissue is not

accessible, for example, a tumour, drug placement becomes difficult, especially

when more than a certain minimum drug concentration has to be maintained for a

significant time course. Many factors contribute to the complexity of drug

localisation at the target tissue; these include the behaviour of the drug in the

body as well as patient compliance.

Controlled Release Systems

Drug administration or delivery must continue at an appropriate rate until the

condition is cured or controlled with a minimum of side effects. In some

situations this might mean that the drug is delivered more promptly for short

periods of time and in other cases it would mean prolongation of drug levels. In

the latter category the terms, "sustained release" and "prolonged release" are

employed interchangeably; this designates only one aspect of controlled release,

namely, to produce adequate levels of drug in the body.

Previous efforts (Chien. 1980) have focussed on the use of polymers or pumps, to

provide "controlled" rather than "sustained" release. Controlled release systems are

designed to continuously deliver and maintain the drug concentration at a desired

I
level in the body. In general, a controlled release system utilises a polymer matrix

or pump as a rate controlling device to deliver the drug in a fixed, predetermined

pattern for a desired time period. Ideally, the use of this type of drug delivery

system should result in a steady drug concentration as a function of time, require

fewer and smaller dosages and cause fewer side effects. For the drug to be taken

up by the target tissue in the body, several events must occur:

1. The drug must first be released from the carrier,

2. It must then diffuse from the surface of the carrier through the surrounding

environment, and eventually,

It must accessthe target.

Drug Carriers

The concept of carrier mediated drug delivery has gained considerable interest in

the last decade or so. Directing drugs to specific target organs is an old idea

discussed by Paul Ehrlich (1906)in the beginning of the century.

To be efficient, the drug-carrier should fulfill several criteria:

1. The carrier should be pharmacologically inactive and the drug must be released

in an active form after interaction of the carrier with the target cells.

2. The drug carrier complex must be stable in plasma and extracellular spaces.

3. The carrier should have the ability to take the drug through those anatomical

barriers which separate the site of administration from the target; it must

also be specifically recognised by receptors or antigens present on the outer

membrane of the target cells.

4. The carrier should be non-toxic, non-immunogenic and biodegradable to avoid

cellular overload during long term repetitive treatment.

5. Production of the drug loaded carrier in the amounts and conditions required

2
 for clinical use, for example, sterility, apyrogenicity, should be easily achieved

and pharmaceutically acceptable in terms of stability and reproducibility.

Several systems for achieving some of these goals have been proposed. Liposomes

(Gregoriadis, 1983), resealed erythocytes (Ihler, 1983), microparticles (Sjoholm and

Edman, 1984), nanoparticles (Marty CLaL,1978,Oppenheim,1981) and albumin

microspheres (Tomlinson efgL, 1984) are examples of particulate colloid carriers

which are used as targetable drug delivery systems.

Drug Targeting

Drug targeting aims to limit the access of the pharmacological agent to selected

cells or tissues. In theory, such a strategy should decrease unwanted side effects

by decreasing drug levels in non-target cells and enhance therapeutic activity by

increasing the concentration of administered drug within target cells. Targeting

can be achieved in two, principally different, ways depending on the

characteristics of the carrier in the body, that is, by passive or active targeting

(Poste, 1983). Passive targeting utilizes the natural homing of the carrier in the

body after intravenous (i. v. ) administration.

It has been irrevocably shown (Poznansky and Juliano, 1984) that in fact most

carriers, irrespective of their material, nature and composition, will accumulate in

the reticuloendothelial system (RES) after i. v. injection. This innate behaviour can

be used to drug load the RES. In some tropical parasitic diseases which involve

the RES, this phenomenon may be used for real therapeutic benefit (Alving

CLaL,1978;Black CLd, 1977,* New cLaL1978). Active targeting implies an attempt to

circumvent the RES by some means and using, for example, monoclonal antibodies

with specificity for certain cells or receptors, they are intended to deliver the

drug to targets in a specific manner. A prerequisite for successful active targeting

3
is that the targets are in contact with the compartment to which the carrier is

administered.

Biological Strategy

Several drug delivery approaches make use of biological entities, such as

antibodies, red cells, and liposomes, as drug carriers. The biologically based

delivery systems share some of the characteristics of polymeric and prodrug

systems and can be used for similar purposes. Implicit in the work on biological

carriers is the goal of using highly specific physiological recognition mechanisms

as the basis of targeted delivery.

Carriers such as liposomes and red cells can be used to achieve sustained and

controlled drug administration. Drug loaded liposomes (Ostro, 1987) and

erythrocytes (DeLoach, 1986) can be introduced directly into the systemic

vasculature or for localised action into the appropriate organ or body cavity. Such

carriers can, in addition, perform some tasks which are beyond the capability of

the synthetic delivery systems. For example, liposomes, with their membrane

mimetic structure, can promote the cellular uptake of drugs that do not readily

penetrate cell membranes. This feature may be particularly important in cancer

chemotherapy where neoplastic cells often become drug resistant because

membrane alterations occur leading to reduced drug permeation.

Biological carriers can be used to deliver active agents which are of macro-

molecular dimensions, such as enzymes and nucleic acids; this property may be of

value in enzyme replacement therapy of inherited disease and in the genetic

manipulations of recombinant DNA technology. These carriers can, in theory, be

fitted with highly specific recognition ligands, for example antibodies, so that

4
they appear prime candidates for the eventual development of "targetable" drug

carrier complexes.

Controlled delivery is the desired effect of all drug delivery systems and presently

all sustained and prolonged drug delivery systems provide some degree of control,

albeit, incomplete. Thus, whereas second generation sustained release products

have made significant advances over their first generation counterparts, none of

the commercially available systems presently on the market is, in truth, a

controlled drug delivery system (Ihler, 1986).

This section deals only with liposome basedcarrier systems,and concentrateson

their potential use as drug delivery vehicles in midicine.

Enzyme therapy

The earliest proposal as carriers of therapeutic agents was made by Gregoriadis

and co-workers who studied the liposomal entrapment of amyloglucosidase

(E. C. 3.2.1.3) from Asgergillus nieer (Gregoriadis fW, 1971). They proposed that

liposomes might be ideal vehicles for introducing enzymes into deficient cells in

genetically inherited metabolic disorders, especially in glycogen storage diseases.

Although promising results were obtained in tissue culture, in experimental animals

and in a few preliminary clinical trials, the widespread applicability of liposomes

as enzyme carriers is limited. The major reason is the rapid uptake of liposomes

by the liver and spleen after intravenous injection. This would tend to eliminate

5
the use of liposomes in disorders affecting other tissues unless a means can be

found to direct them to such tissue. This applies in particular to diseases

affecting specialised cells, such as neurological tissues, which are unlikely to take

up liposomes. Considerable advances need to be made in liposome technology to

introduce the required degree of tissue specificity in vivo. However, drug delivery

technology may be potentially applicable to many other types of pharmacologic

agent and in different therapeutic situations, such as cancer chemotherapy.

Cancer Chemotherapy

Cancer chemotherapy is severely limited by the intrinsic toxicity of anticancer

drugs, hence it is important to control their pharmacokinetic behaviour and tissue

distribution. A great deal of research effort has been expended on the study of

liposomes, antibodies and other carriers for anticancer drugs. This topic has been

well reviewed (Gregoriadis, 1977). Many such drugs have been entrapped in

liposomes and these include; actinomycin D, vinblastin, daunomycin and cytosine

arabinoside (Juliano and Stamp. 1978), 5- fluoro-uracil (Gregoriadis, 1974) and the

closely related floxuridine (Mathias fLal, 1977), methotrexate (Kimelberg and

Tracy, 1978), 8-azaguanine and 6-mercaptopurine (Kano and Fendler, 1977),

bichloroethy1nitrosurea (Mayhew dat. 1976), cisdichlorobiscyclopentylamine platinum

(Deliconstantinos CW, 1977) and doxorubicin, also known as AdriamycinR* (Olson

CLal,1982).

Since drug tissue accumulation is a function of the concentration of free drug in

plasma, tissue toxicity of anticancer agents including cardiac, nephro- and

pulmonary toxicity may be reduced by confining the drug in a carrier. Peak

plasma concentrations and tissue accumulation can therefore be avoided and a

prolonged therapeutic plasma concentration achieved, due to "slow" drug release.

6
This concept has been employed successfully when doxorubicin was encapsulated

within liposomes.

Doxorubicin is the most widely used agent against various solid tumours and

1979). Despite the potency, its clinical usefulnessis

leukaemias(Blum CLgL,

hampered by delayed, irreversible cardiotoxicity (Ugoretz 1976) and acute nausea

and vomiting. Various researchers(Forssen and Tokes,198]:Shinozawa

CLA]98]: GabizonCUL,1982&1986.Mayhew CW, 1983:RahmanCLaL,1986) in a number

of laboratories have demonstratedin experimental animals that liposomal

doxorubicin is as effective as the free drug and is several times less toxic to the

heart. More recently investigators have shown that the encapsulateddrug also

causessignificantly fewer side effects in human studies (Gabizon CLgL,

1986).

These results may be attributed to a modification of organ distribution of drug;

accumulation of liposome associateddoxorubicin in cardiac tissue is greatly

reduced, while drug-containing liposomesaccumulatein the liver (Gabizon

fLAI, 1986,Mayhew f%gj,1983,Rahman CLjd,1986), a phenomenonknown as RES

loading. This leads to the working hypothesis that liver cells, following

sequestrationand degradation of liposomes,provide a "depot" for the drug that

gives rise to prolonged low plasmaconcentrationsand presumably, to the improved

efficacy of doxorubicin against liver metastases(due to high local concentrations).

Andfungal Agents

Systemic fungal infections are frequently seen in patients whose resistance is

depressed by disease or medications that supress the immune system. These

infections are a common cause of death in victims of acquired immune deficiency

syndrome, AIDS, and are common in cancer patients undergoing chemotherapy.

7
They often resist treatment because use of the antifungal drugs is limited by

toxicity.

In studies with mice liposomal amphotericin B, a powerful antifungal agent,

,
cured systemic fungal infections more effectively than the free drug (Lopez-

Berestein dd, 1983). Amphotericin B binds more strongly to ergosterol, the

primary fungal sterol, than to cholesterol, it's mammalian cell conterpart (Medoff

CLaLI 983; Hamillon -Miller, 1973;Chen and Bitiman, 1977,-Readio

dXj: 1982.Edwards, 1980). However, the clinically utilised formulation of

amphotericin B has a number of serious adverse side effects, most especially

severe nephrotoxicity (Medoff LA1983: Praft, 1977). The adverse actions of the
C.

drug in humans are likely to be due, at least partly, to the ability of

amphotericin B to interact with cholesterol and form pores in the cellular

membranes of the kidney and cardiovascular system. Amphotericin B binds readily

to the cell wall of fungi, is lipid soluble and therefore is incorporated into the

membrane of liposomes. It is postulated that the fungi "pulls" the drug out from

the liposome membrane, in a process similar to lipid exchange and are destroyed.

It has been shown that incorporation of amphotericin B into certain types of

liposomes can markedly reduce the toxicity of the drug without loss of antifungal

potency (Juliano CLal.1985). Studies in animals and in patients (Lopez-Berestein

r.t&j, 1983&l98J) support the idea that amphotericin B in multilamellar vesicles

composed of dimyristoylphosphatidy1choline (DMPQ and dimyristoylphosphatidyl-

glycerol (DMPG) remain therapeutically effective and significantly less toxic than

the free drug. The reduced toxicity in vivo may be due to various causes

including sustained or slow release of the drug, altered tissue kinetics and

distribution or effects on the immune system (Poznansky and Juliano, 1984;

Jullano, 1981).

8
Antimicrobials

The recognition that liposomes accumulate in the RES led to their use in the

treatment of certain parasitic diseaseswhere microbes reside within the

macrophages of the liver and other tissues of the RES. The access of free drug to

such sites is restricted, making infections there difficult to treat by conventional

means. Several workers have described the efficiency of liposomal drugs against

visceral leishmaniasis in animals (for a review, see Alving, 1986). Leishmaniasis is a

parasitic disease infecting many individuals throughout the world. The parasites

invade cells of the liver and spleen and can be lethal if the infection remains

untreated. Therapy involves multiple dosing with drugs (intra-muscularly), organic

antimonials, which are toxic and in sensitive individuals there is a risk of damage

to the heart, liver and kidney. Alving CLgI (1978) and Black tw (1977)

independently found that encapsulating antimonial drugs in liposomes reduced the

dosage required to treat leishmaniasis. This therapy takes advantage of the passive

targeting of the carrier to the liver and spleen.

Liposomes have also been used in the treatment of other infections that involve

macrophages, including brucellosis (Fountain CLOL,

1985), listeriosis (Bakker-

Woudenberg CLXL.1985) and acute salmonellosis (Desiderio and Campbell, I 983a; and

1983b).

Diagnostic Radiology

Because of the preferential delivery of liposomes to the liver and spleen in

diagnostic radiology, liposomes containing either radio-opaque water soluble dyes

or radio-op aque lipids have been used successfully for causing image enhancement

of liver and spleen by computed tomography (Ryan CLaL.1984;Caride cLaL.1984:

Seltzer daL, 1984).

9
MiscellaneousUses

Liposomeshave extensive use outside the medical field. One novel example is in

the maturation of cheeses by proteases contained within liposomes (Piard fLSL

1986).

1.3. LIPOSOMES: PREPARATION. STABTLITY.

In the early 1960's Bangharn and his collaborators at Cambridge made the

observation that phospholipids dispersed in water formed multilayered vesicles.

Each layer was formed from a bimolecular lipid membrane, and these layers

enclosed internal aqueous compartments (Bangham fLgL, 1965). It rapidly became

apparent that these artificial structures, which came to be known as liposomes,

closely resembled cellular membranes. These new model membranes soon became

popular tools for biochemists, cell physiologists and more recently, for

investigators concerned with drug therapy. Their various applications have been

extensively investigated (Gregoriadis. 1982;Kaye, 198]; Ryman CLaL,1980). A simplified

diagram of a liposome is shown in Figure 1.

Liposomes are of potential interest for the following reasons:

1. They are made of phospholipids which are natural constituents of all cell

membranes so that toxicity would not be expected.

2. Entrapped drugs are physically separated from the environment so long as the

liposomes remain intact. This can be used to protect drugs from enzymatic

degradation. In addition, toxic compounds, eg.doxorubicin, can be carried in the

bloodstream thus reducing cardiotoxicity (see page 7).

10
 MULTILAMELLAR UNILAMELLAR
VESICLE VESICLE

Figure 1: Schematic diagram of liposomes formed by mixing amphipathic lipids and

an aqueous solution can be multilamellar (left) or unilamellar (right). Multilamellar

vesicles have an "onion skin" structure in which concentric lipid bilayers are

separated by aqueous layers (blue areas). Unilamellar vesicles consist of lipid

bilayer surrounding an aqueous interior.

(Reproduced from reference, Ostro, 1987)

3. In vivo free and "liposome -entrapped" drugs are likely to be delivered to

different tissues or cell types after administration. While incorporated.

liposome -entrapped compounds will follow the fate of the liposome and they

will be released only at the site of liposome degradation. This allows newer

possibilities in site specific delivery of drugs, or drug targeting.

4. Liposomes can accomodate both hydrophilic and hydrophobic drugs, with little

or no need for chemical modifications of the drug. This is important for

product registration purposes since existing drugs can' be used in these novel

formulations.

5. Liposomes can be prepared with a variety of properties, for example,

differences in size, composition, surface charge, etc.

Therefore, it would seem possible to design an optimal carrier for different

applications by careful selection of these parameters.

In view of their intrinsic properties therefore, liposomes should improve existing

drug therapy by increasing efficacy and/or reducing side effects. Considerable

research efforts have been devoted to investigations on the use of liposomes as

drug carriers. Nearly 6500 articles and 139 patents have been published since 1971

and these numbers are increasing.

12
The literature hosts a large number of preparation procedures ranging from the

simple through the novel to the complex. Below are summarized a few of the

standard techniques.

Multilamellar Vesicles

The original preparation (Bangham CLaL,1965) of multilamellar vesicles (MLV) was

achieved by simple mechanical shaking of a thin dry film of egg phosphatidyl-

choline (PC) with an aqueous solution in a round-bottomed flask. This method

produces a heterogeneous mixture of closed multilamellar vesicles of varying size.

Most phospholipids studied so far have been found to produce such vesicles,

including purified preparations or lipid extracts from tissues. The major advantage

of MLV preparation is the simplicity of the procedure and the fact that it is

applicable to a wide variety of different lipid mixtures. The encapsulated volume

of MLV in terms of litres of aqueous space per mole of lipid is limited and

several-fold less than that of unilamellar vesicles of comparable size. A simple

procedure optimising entrapment levels for MLV has been described by Kirby and

Gregoriadis (1984). Preformed liposomes containing drug are flash-frozen as a thin

film in a round bottomed flask, freeze-dried under vacuum and rehydrated under

controlled conditions. The ability to freeze or rehydrate vesicles offers the added

advantage of long term storage.

Small Unilamellar Vesicles

In addition to the simple procedure used by Bangham, numerous alternative

methods of preparation have been described. Small unilamellar vesicles (SUV) can

be obtained by sonication of MLV preparations (Huang and Charlion. 1971), by

13
injecting an ethanol solution of the phospholipids in an aqueous phase (Balzri and

Korn. 1973). The particle size of MLV can also be reduced by extrusion at high

pressure through a French Press (Barenholz cLaL, 1979;Hamilton dd, 1980). The

yield of unilamellar vesicles and the vesicle size are dependent on the pressure .

Barenholz CW (1979) reported egg PC vesicles to be somewhat larger than that

of sonicated SUV (31.5-52.5 nm) but Hamilton CLIL (1980) found the size of

French Press SUV containing no cholesterol to be similar (15-30 nm) to that of

sonicated SUV.

Large Unifamellar Vesicles

Large unilamellar vesicles (LUV) are produced by dialysing a detergent containing

phospholipid solution (Milsmann CLaL,1978) by injecting phospholipids dissolved in

diethylether (Deamer and Bangham, 1976) or petroleum ether (Schieren, 1978) into

a warmed aqueous phase. Extrusion of MLV through polycarbonate membranes

(0.1jurn pore size) also results in the production of LUV with an average diameter,

0.9,um (Hope CLgL1985). However, most of these methods show fairly poor

entrapment yields for water soluble drugs, usually below ten percent.

The reversed phase evaporation technique (Szoka and Papahadjopoulos, 1978)

produces large unilamellar vesicles (0.1-0.914m diameter), abbreviated to REV,

which exhibit a high capture volume and have a high encapsulation efficiency. In

this method, the drug containing aqueous phase is emulsified in the presence of

phospholipids in diethylether. Evaporation of the solvent under vacuum forms large

unilamellar vesicles. Recently the Double Emulsion Technique was developed

(Schneider, 1977). In this method the drug-containing aqueous solution is dispersed

and sonicated in a non water-miscible organic solvent for example, dibutylether,

cyclohexane, etc. containing the phospholipids. The resulting water-in-oil emulsion

is then introduced under agitation into a second aqueous medium thereby forming

14
a double emulsion. Evaporation of the organic solvent(s) under a stream of

nitrogen, forms unilamellar vesicles with an average diameter of 0.15pm. These

LUV possessan encapsulation efficiency in the range thirty to eighty percent.

When hydrophobic drugs of appropriate structure are associated with liposomes by

inclusion in the bilayer phase, the degree of "encapsulation" is dependent upon

the saturation of the lipid phase. Under these circumstances, it is possible to

achieve degrees of encapsulation of over ninety percent, making it unneccesary to

remove the unbound drug. However in the case of water soluble drugs, the

encapsulated drug is only a fraction of the total drug used. An additional step is

required to remove the unbound drug from the drug-loaded liposomes. Dialysis,

centrifugation and gel filtration are some of the procedures used routinely for

this purpose.

Dialysis

Dialysis is routinely used (Olson CLUL1979) for the removal of unentrapped drug,

except when macromolecular compounds are involved. This technique requires

uncomplicated and inexpensive equipment and is conveniently scaled up. Dialysis is

effective in removing greater than ninety-five percent of all free drug with

sufficient number of changes of the dialysing medium. However this process is

slow. Typically, removal of this percentage of free drug in a liposomal dispersion

might require a minimum of three changes of the external medium over a ten to

twenty-four hour period at room temperature or in a cold room (40C). Care must

be taken to balance the osmotic strengths on either side of the membrane or the

volume of the liposomal dispersion will alter during dialysis. It is also possible

15
that the presenceof the external dialysing medium in equilibrium with the

vesicles might induce leakage of the encapsulated drug.

CentrIfugation

Centrifugation at various speeds is an effective means of isolating liposomes from

the free drug in the suspending medium. Two or more resuspensions and spin

stages are usually included to effect a complete removal of the free drug. The

centrifugal force required to pull liposomes down into a pellet is dependent on

the size of the liposomes. High speeds (>100,000g) and refridgerated conditions are

required for liposomes in the small to medium size ranges. The use of refrid-

gerated centrifuges operating at high speeds with large volumes (for large scale

preparations) of liposomal dispersions is energy intensive and expensive.

Therefore, it may not be suitable for the isolation of small liposomes For
.
relatively large liposomes, low centrifugation speeds around 10OOgoffers the

advantages of a short time scale of operation. However, the osmotic strength of

the resuspending medium must be equal to that of the original liposomal

dispersion to minimise osmotic shock and rupture of vesicles.

Gel Filtration

Gel permeation chromatography is used extensively both to separate liposomes

from unbound drug and to fractionate liposomal dispersions containing a mixture

of various sizes of vesicles. The technique is very effective and rapid under

laboratory conditions. Although gel filtration is used in the purification of

biological materials such as insulin on a large scale, the technique is slow and

laborious. Additionally, dilution of the liposomal suspension with the eluting

medium may necessitate another concentration step which could damage the

vesicles.

16
Ultraflitration

Ultrafiltration has been used (Rao, 1984) for the separation of non-associated

solutes from liposomes, but its use is not recommended since considerable amounts

of non-entrapped material often remains attached to the filtration membrane after

long periods of concentration and dilution. It has more potential for concen-

trating liposomes which have been diluted during column chromatography to

separate non-associated material.

1.3.3. COMPOSITIONS AND TISSUE

DISTRIBUTION OF LIPOSOMES

Besides the size and the structure, there are other parameters which can be

conveniently altered, most notably, the lipid composition, the surface charge and

the membrane fluidity. The most generally used phospholipid components include

egg phosphatidy1choline (PC), phosphatidylserine (PS), phosphatidylglycerol (PG)

and sphingomyelin (SM). Synthetic lecithins such as dipalmitoylphosphatidy1choline

(DPPQ, distearoylphosphatidy1choline (DSPQ and dimyristoylphosphatidy1choline

(DMPQ are also often used. In addition to phospholipids, large amounts of other

lipids can be incorporated into the liposome membrane. Cholesterol, which by

itself does not form vesicles, can be incorporated (upto 50 mol%) into the

phospholipid lamellae. Cholesterol has been extensively used to reduce the leakage

of entrapped water-soluble drugs and to increase the stability of liposomes in

plasma (see discussion, page 97). The net surface charge of liposomes can be

modified by addition of other long-chain amphiphatic molecules such as,

stearylamine (for positively charged liposomes) or compounds such as

phosphatidylserine, phosphatidic acid or dice tylphosphate for negatively charged

liposomes.

17
Another importantparameter,the phasetransition temperature(Tc) of the fatty

acyl groupsof phospholipidshasbeenshownto influencethe compositionand

permeabilityof liposomesand alter their interactionswith isolatedcells.At
temperaturesabovethe Tc, liposomesare designatedas "fluid" whereasbelowthis
temperaturethey are "solid".In the fluid form, the permeabilityis appreciably
higher than in the solid form. This Tc is characteristicfor eachphospholipid

species,is definedmostlyby the configurationof the acyl chain and alsothe

degreeof hydrationand chemistryof the headgroups(Williams and

Chapman,
1970).Thus, the presenceof cis-doublebonds,branchingand decreasing

chain length tend to reducethe Tc. Shownbelow is a schematicrepresentation

of
molecularorientationin phospholipidbilayersbelowand abovethe phase
transition.

0
4ý3A

41A

The head groups of the phospholipid molecules are represented by open circles

and fatty acids by solid lines. Tightly packed "solid" acyl chains showing inhibited

motion by NMR are represented by straight lines. Mobile (liquid) acyl chains are

represented by curved lines. The dimensions shown in the above scheme relate to

those reported with DPPC bilayers by Lesslauer t=1972). Permeability of

liposomes for various solutes increases markedly at Tc of the lipids (Fukuzawa

rLgL; 1979). An interesting practical implication of this phenomenon has been

suggested by Yatvin IW (1978) for selected drug delivery. Local hyperthermia

may be artificially induced to increase the specific release of an entrapped drug

18
at defined body areas, for example, tumours heated to a few degrees above body

temperature by microwave irradiation.

incorporation of cholesterol into the phospholipid membranes produces

considerable restiction of molecular motion in the region of the first eight to ten

carbon atoms of the acyl chain from the lipid-water interface. This effect is

accompanied by a condensation of the area per phospholipid molecule, a more

perpendicular orientation and a thicker membrane (Darke CLaj,1971).

I
SNOB ml 42.4

Phosphol0a Phoop
haIipid

Above is a schematic representation of the molecular architecture of a phospho-

lipid bilayer with and without cholesterol. The graphic representation for

phospholipid molecules are as previous, page 18. Cholesterol is represented by a

small circle (hydroxyl), a larger circular area (the four-ring steroid skeleton) and

a line (hydrocarbon tail). The dimensions are from X-ray data of Levine and

Wilkins (1971) with egg PC-Cholesterol multilayers. Addition of cholesterol (50

mol%) abolishes the Tc (Op den Kamp CLal: 1975) and diminishes liposomal

suseptibility to protein interactions (Tall and Lange, 1978:De Kruyff cLaL1974).

Values for Tc of the lipids and surfactant used in this study are included in

Appendix 1, page 160.

The tissue distribution and clearance kinetics of liposomes containing drugs are

affected by phospholipid composition, size and charge (Juliano and Stamp, 1975;

19
Rahman et.al, 1982). Smaller liposomes were shown to be better than larger

liposomes for specific delivery of their contents to calf thymus cells in vitro

(Machy and Leserman.1983). Sucesseshave been demonstrated for liposomes

including in vitro targeting through covalent attachment of proteins onto the

surface of liposomes (Leserman CLal,198]; Healh f.Lal, 1983) and more recentlyin

vivo targeting for hepatic parenchymal has been achieved using liposome bearing

galactose residues (Van Berkel CW, 1986) The interactions of liposomes with

cells in vivo occur in a complex biological milieu and the binding of proteins from

this milieu may also have important effects on liposome behaviour (Juliano and

Lin, 1980).

By using different lipids and modified preparations, a variety of liposomes can be

designed, varying in size, structure, surface charge, composition and membrane

fluidity. However further changes in these various parameters are obtained after

in vivo administration, resulting in anomalous distribution patterns and behaviour,

than originally predicted.

1.3.4. STABTLITY OF LIPOSOMES

A given liposomal formulation must have adequate stability over the time period

between its preparation and ultimate use. This includes the physical stability of

the liposomes in terms of the integrity of the encapsulated material and the size

parameters and the chemical stability of the component materials (lipids and

drugs). Thus a liposomal drug product must show an adequate "shelf life" under

normal conditions of storage. In addition, the liposomes should maintain their

integrity in vivo until entry into the target tissue or until their function of

sustained release is completed.

20
Studies on the stability and effective storage of liposome preparations are

necessary for their development as a drug delivery system. "Shelf life" stability

encompasses a number of parameters and the following aspects should be

examined:

a. maintenance of the vesicle size, including an examination for the aggregation

of units with time;

b. maintenanceof the structure of the lipids;

C. chemical stability of the lipids; and

d. influence of biological fluids on the integrity and permeability properties of

the liposome.

Liposomescan becomephysically unstable on storageeither due to leakageof the

encapsulateddrug into the suspendingmedium and/or due to aggregation or fusion

of the liposomesto form larger entities. It is important that a liposomal drug

product remains stable over a period of time, preferably under ambient conditions

of storage.The increasein size and drug leakageon storageof small unilamellar

liposomeswere found to be dependent upon the nature both of the lipid

componentsand of the encapsulatedmarker compounds(Frokjaer CL&L,

1982). Drug

retention behaviour of a range of liposomal compositions under various conditions

of storage has been cited (Szoka and Papahadjopoulos,1980).

The chemical stability of drugs incorporated in liposomes is outside the scope of

this discussion. However, the lipid components of a liposomal system are subject

to various chemical degradations on storage. The oxidative decomposition of egg

PC in liposomes can be retarded considerably by addition of alpha- tocopherol to

the lipid mixture (Hunt, 1982). Hence particular care is required when handling

unsaturated phospholipids subject to this type of oxidative degradation.

21
Unsaturated and saturated phospholipids are subject to hydrolysis in aqueous

media, resulting initially in the formation of the corresponding lysophospholipid

and fatty acid (Frokjaer CW, 1982). These reactions have implications both for the

integrity of the liposomes and the leakage of encapsulated drugs.

To optimise the potential of liposomal carriers, it is important to characterise

their stability in terms of controlled release of their contents in vitro, simulating

physiological conditions in vivo. The stability of various types of liposomes in the

presence of blood components has been reviewed (Gregoriadis f.Lal, 1981 ;

Scherphof cLaL,1981). The chemical nature of the phospholipid, the proportions of

cholesterol and the size (and lamellae) of the liposomes all appear to influence

their in vivo stability. It is therefore necessary to design and test liposomal

systems for adequate stability in vivo.

It is possible to conceive a liposomal drug carrier system for storage either as a

normal dispersion or in a form to be reconstituted before use. If the problem of

drug leakage can be resolved presentation as a ready to use dispersion has

obvious advantages due to simpler preparation techniques. A reconstitutable form

or kit, for example, a freeze-dried preparation will avoid both chemical and

physical stability problems in storige. Inclusion of trehalose or other such sugars,

has been shown to be effective in preventing fusion and leakage from liposomes

during freeze drying. Retention of upto 100% of an entrapped water soluble

marker and prevention of fusion has been reported (Madden cLxl, 1985,Strauss

ctxL, 1986).

With increasing therapeutic application and progression into human clinical trials,

it will be necessary to routinely prepare sterile batches of vesicles. Aseptic

techniques can be applied to most stages of these preparations. Antibiotics may be

added to the aqueous phase together with the therapeutic agent in liposome

22
preparation if necessary. Sterility of smaller vesicles can be achieved by passage

through 0.22pm sterile membrane filter, or alternatively, radiation sterilisation

may be employed (Gregoriadis. 1976). However it must be emphasised that sterility

must be maintained throughout the preparation if the liposomes are to pass

rigorous pyrogen testing protocols (Tyrrell and Ryman CLaL1976).

Measurement of Stability In Biological Fluids

The stability of the vesicles is defined in this context as their capability to retain

the structural integrity of the closed lipid bilayer and to prevent leakage of their

aqueous contents. The original method used to study permeability of lipid vesicles

by Bangham CLIJ (1965), involved the use of isotope tracers and has been used

for 22sod
iUM, 42potassium, 36chloride,
extensively for the study of efflux rates
86 14
rubidium and glucose. A modification of this method used by Kinsky gLaL

(1966) monitored the spectrophotometric reduction of nicotinamide dinucleotide.

Other, non-isotopic markers such as chromate ions have been used (Weissmann

CL,al, 1966) to study the permeability of liposomes.

Fluorescent Markers

A useful method to study cell-liposome interactions employs the use of water-

soluble markers (Pagano and Weinstein, 1978). The fluorescent dye 5,6-carboxy-

fluorescein (CF) was specially prepared (Weinstein CLaL,1977) as such a marker.

This highly water-soluble dye has since been used by a large number of

laboratories for many studies. CF has been widely used for studying liposome-cell

(Weinstein CLgL,1977: Blumenthal cLxL, 1977, Leung, ] 980: Van Reswoude and

Hoekstra, 1980), cell-cell (Dahl cW, 1981) and liposome-liposome (Wi1schut and

Papahadjopoulos, 1979) interactions. It has also been employed for monitoring

enzymatic attack on intact liposomes (Chen. 1977) or perturbation of the lipid

23
bilayer structure by serum proteins (Yaivin CLaL,1978;Guo cLaL,1980). The

concentration dependent self quenching of CF fluorescence permits leakage from

liposomes to be monitored continuously and hence a distinction can be made

between endocytosis of liposomes, binding of liposomes to the cell surface and

direct transfer of liposome contents to the cytoplasm (Weinstein cLaL,1977).

However this technique has its own limitations.

Structurally, CF resembles fluorescein but has an extra carboxyl group located at

the 5- or 6- position. It has an excitation maximum at approximately 492nm and

an emission maximum about 520nm, similar to fluorescein. The latency of CF

entrapped in liposomes is much greater than fluorescein due to the additional

carboxyl group which decreases the butanol-water partition coefficient that is,

it's lipophilicity is decreased, by about three orders of magnitude over a broad

range of concentrations and pH's (Grimes ftaL, 1982). This makes CF a better

alternative than fluorescein. The most-water soluble and fluorescent form of CF is

the trivalent anion predominant at neutral and alkaline pH's (Weinstein CLaL.1986),

as shown in figure 2. In practice, it is important to calibrate the fluorescent data

for the ambient pH, usually pH 7.4, and to be aware that the less soluble forms

present in acid medium have a greater tendency to bind to proteins, detergent

molecules and other organic materials (Lelkes and Tandeter. 1982). Hence when

experiments with protein solutions are conducted, the buffer should be at pH 7.0

or higher.

Typical of most aromatic fluorophores (Parker, 1968), the fluorescence of CF

decreases slightly as the temperature is raised. This decrease results primarily

from an increased non-radiative transfer of energy from the excited singlet states

of oxygen by intermolecular collision (Weinstein CW, 1984). CF is not a single

component dye but is a mixture of different ingredients that show at least six

24
 (at low PH)
H

HO
4.

Oeoooýý

1100
ii 0'
(at pll >6.5)

ý, \\, coop
0 00c.

-"=

Figure 2: The structure of Carboxyfluorescein (CF); molecular weight, 376.

25
spots on silica gel thin layer chromatography and exhibits at least six different

absorption peaks in the range 440-500nm.

CF can be purified by a method described by Ralston gill (1981). The

fluorescence emission of CF is sensitive to the presence of biological fluids.

Experiments have indicated that in a variety of sera, for example, mouse, horse,

bovine, foetal calf and human, the fluorescence emission is partially quenched

with a decrease in the excitation spectrum (Lelkes and Tandeter, 1982). However,

despite these drawbacks, experiments can be suitably controlled and CF has

proved to be a useful marker in many studies.

As an alternative marker, the fluorophore calcein has also been used (Allen and

Cleland, 1980) which is more water soluble at low p1l than CF. Calcein is

chemically a combination of fluorescein and ethylenediaminetetraacetic acid and it

shares properties of both these compounds (Diehl and Ellingboe, 1956). It was first

prepared by the interaction of fluorescein, formaldehyde and iminodiacetic acid

and it's structure determined (Wallach fLaL, 1959), as shown in figure 3. Calcein

carries a higher net negative charge at physiological pH and possessesa higher

molecular weight. It's rate of efflux, therefore, from liposomes is slower because

the overall negative charge reduces it's permeability through the membrane. Also,

calcein does not show a pH-dependent quenching in the physiological pH range

(Allen, 1984).

26
 ( at low pH)
Q
OOC COOH
II
qc." ý1ýH;.
+ý1,I
CH,
H H
H,
+/C -CO
*CH-COO
2. H

'3'o oc 0
coo
(at pH 7.0)
IV\ ý H,.
+MP

9
H
H,, 0 T'
+/C 7C
CHI-COT

Figure 3: The structure of Calcein; molecular weight, 623.

27
There have been a number of studies of the interaction of liposomes with proteins

and complex mixtures of proteins such as serum. The intent of these studies has

been

1. To examine the proteins capable of binding to liposomes.

2. To assessthe stability of liposomes in the biological environment.

The initial event, in vivo, to "foreign" materials and surfaces after contact of

blood is the adsorption of a protein layer which "conditions" the surface of the

material for further interactions (Baier and Dution. 1969). This layer is believed to

consist of proteins originating from plasma.

Proteins constitute the major part of the soluble material in blood, where their

concentration is normally 70-80 grams per litre. The number of different proteins

totals more than one hundred and fifty.

The most abundant protein found is serum albumin (70%) which has a relatively

low molecular weight, approximately 66,000-69,000 daltons and a plasma concen-

tration of 40 grams per litre. It acts as an important plasma expander and as a

carrier for substances such as hormones, fatty acids and vitamins (Spector and

Fletcher, 1977). It has a strong binding affinity for a number of other substances,

for example, certain drugs such as acetylsalicylic acid (Hawkins cLat, 1968). Other

compounds such as fluorescein also bind loosely (Andersson CLaL.1971) to albumin

mainly through hydrophobic and electrostatic interactions. The electrostatic

contribution is much less than would be expected; this is probably related to the

fact that fluorescein is a non -physiological compound and is "non-specifically"

bound to the albumin.

28
Immunoglobulin 0 (IgG) also binds to many surfaces exposed to blood or plasma

(Horbelt and Weathersby,1981). IgG has a concentration of 8-17 grams per litre in

normal adults and a molecular weight approximating 150,000 daltons. The molecule

comprises of two "heavy" chains (m. wt. 50,000 daltons) making up sixty-seven

percent of the molecule and two "light" chains (m. wt. 25,000 daltons) which

account for the remaining thirty-three percent of the molecule.

Fibronectin is a high molecular weight (220,000-250,000 daltons) glycoprotein

found in a soluble form in blood and other body fluids and in an insoluble form

in tissues. The concentration of fibronectin in plasma is around 300,ug/ml. Recent

reviews demonstrate the role of fibronectin (Mosher, 1984,Ouaissi and Cabron, 1985)

but there is only little information about it's interaction with membrane

phospholipid components. Rossi and Wallace (1983) has shown the binding of

fibronectin to phospholipid vesicles of a number of compositions. This binding of

fibronectin to liposomes can modulate liposome interaction with cells. 11suand

Juliano (1982) showed that fibronectin could enhance liposome uptake by

macrophages, presumably by interacting with cell surface fibronectin receptor

(Brown and Juliano, 1985).

Fibrinogen absorption from plasma or blood to a wide variety of surfaces has been

observed by a number of investigators (Horbell and Weathersby,1981,Horbett, 1981,

InIenfeld and Cooper, 1979, Vroman CLaL,1980). Fibrinogen is a high molecular

weight plasma protein (m. wt. 340,000 daltons) and is necessary for the formation

of blood clots. Fibrinogen has been shown to adsorb to glass (Brash cLaL,1984),

after contact with blood and appears to be an important constituent of the

adsorbed protein, even though it is known that the absolute surface concentration

is low. Previous data (Horbelt and Weathersby,1981) suggest that relative to other

blood proteins such as albumin, IgG and haemoglobin, fibrinogen is enriched on

29
the surface compared to the plasma phase by a factor of two or three depending

on the nature of the surface.

The use of liposomes for drug delivery necessitates the examination of the effects

of various biological fluids on the stability of liposomes. Liposomes recovered from

plasma were found to have altered electrophoretic mobilities (Black and

Gregoriadis. 1976), indicating possible adsorption of plasma proteins. The same

study showed that alpha-2 macroglobulin bound strongly to the surfaces of

liposomes, which had a composition of.

a. egg PC and cholesterol, molar ratio 7:2 (neutral),

b. inclusion of 10% phosphatidic acid, molar ratio 7:2: 1 (negative) and

C. with 10% stearylamine, molar ratio 7:2: 1 (positive).

This was the only protein found specifically bound to liposomes, irrespective of

size or charge on the vesicle. Alpha-2 macroglobulin was positively identified by

immunoelectrophoresis technique (Clarke and Freeman. 1968) against anti-(human

alpha-2 macroglobulin). Serum components such as IgG (Weissmann VQL. 1975) and

lipoproteins (Morrisett CL,

ýd,1977) also bind strongly to various types of liposomes.

Both neutral and positively charged liposomes have been shown (Juliano and Lin,

1980) to bind a group of high molecular weight (>200,000 daltons) polypeptides

which did not appear to be bound by negatively charged vesicles. Some of these

bound proteins could readily be identified as major blood components, such as,

albumin, IgG subunits, apoprotein Al, alpha-2 macroglobulin. whereas other bound

proteins were not readily identified and probably represented minor serum

components. The pattern of bound proteins was highly dependent on the chemical

and physical properties of the liposomes. For example, while negatively charged

liposomes bound an apparently random sample of serum proteins, positive and

neutral liposomes tended to bind selectively to a group of high molecular weight

(>200,000 daltons) protein components. Binding of these protein components was

30
very rapid (one minute at 370C), essentially irreversible, and the components

remained at the outer surface of the liposome and did not disrupt the vesicle

structure.

These workers (Juliano and Lin, 1980) also reported marked changes on the clotting

ability of plasma after exposure to either positive or negative liposomes. This

effect may be a result of the depletion of one or more clotting factor(s) by its

binding to the liposomes. Further studies, by the same group, showed that clotting

factor VII binds to liposomes irrespective of charge, where as factor VIII and

fibrinogen bind to liposomes, which carry a positive and negative charge.

Albumin can also interact with small unilamellar vesicles (SUV), bringing about a

release of entrapped markers and mediating transfer of phospholipid from the

vesicle to itself (Zborowski fLgL, 1977). The injection of liposomes on vivo or their

incubation with plasma in vitro leads to a major alteration in their size and

physical charecteristics (Krupp CLaL,1976).

Serum proteins also promote the efflux of the fluorescent dye calcein from SUV

(Allen and Cleland, 1980). As found by others, these vesicles could be stabilised

against protein disruption by the inclusion of cholesterol.

The stability of liposomes in vivo (blood circulation) and in vitro (in the presence

of serum, plasma or whole blood) is affected by the liposome cholesterol content

(Kirby CLAd.1980a). Incorporation of cholesterol into liposomes condenses the

phospholipid molecules above their Tc and reduces permeability to solutes (Demel

and DeKruft. 1976). This additional "compact packing" of phospholipids by

cholesterol can help prevent their removal by high density lipoproteins found in

plasma and preserve liposomal stability in serum (Gregoriadis CL&1,1981).

31
One implication of these studies is that, upon injection into the blood, liposomes

rapidly become irreversibly coated with a layer of serum proteins which alters

their surface properties. The nature of this layer (which differs for different

types of liposomes) is a main determinant of the clearance kinetics and tissue

distribution of injected liposomes. Active research is still being pursued in this

important area.

Surface Potential and Its Determination

Natural membranes bear net surface charges due to the presence of ionized groups

in their lipids and proteins (McLaughlind 977). The association between fixed

surface charges and the free ions in the surrounding aqueous medium, gives rise

to an interfacial electric field which determines surface potential (Avegard and

Haydon. 1973). A large number of biological phenomena, from cell adhesiveness to

membrane permeability are influenced by surface potential (McLaughlin, 1977:

Avegard and Haydon, 1973). As a consequence, many attempts have been made to

measure this parameter (Gaffney and Mich, I 976: Haydon and Myers, 1973). The

surface potential is an important parameter influencing liposomal behaviour. Jjj

vivo, surface charge density has been found to influence the distribution of

liposomes (Rahman CW, 1980) and in vitro, a high potential might contribute to

their physical stability by reducing aggregation and fusion (Frokjaer CW. 1982).

The charge that develops at the surface of a colloidal particle may arise from any

of several mechanisms, depending on the nature of the particle and its surround-

ing medium (Hunter. 1981). For particles dispersed in liquids, two of the most

important factors are the ionisation of groups on the particle surface and the

differential adsorption of ions of opposite charge from solution. The development

of a net charge at the particle surface affects the distribution of ions in the

surrounding interfacial region, resulting in an incresed concentration of counter

32
ions (that is, ions of charge opposite to that of the particle), close to the

surface. Thus, an electrical double layer is formed around each particle. The

double layer may be considered to consist of two parts: an inner region that

includes ions bound relatively strongly to the surface by adsorption, and an outer

or diffuse region in which ion distribution is determined by a balance of electro-

static forces and random thermal motion. In the region of the particle, the

potential decays with distance from the surface eventually reaching zero in the

bulk solution. An individual particle and its most closely associated ions move

through the solution as a unit and the potential at the boundary of this unit, that

is, at the surface of shear between the particle with it's ion atmosphere and the

surrounding medium, is known as the zeta Dotentlal.

In simplest terms, a vesicle with a net negatively charged surface, suspended in

an electrolyte solution in an electrophoresis apparatus, will attract ions of

opposite sign to form the so-called Stern laver around itself (see Figure 4). The

relatively fixed ions of the Stern layer will in turn attract counter ions, which

form the diffuse electrical double-layer. The concentrations of anions will be

higher than cations near the surface of the vesicle. At greater distances these

concentrations tend to become equal due to Brownian motion. The diffuse double-

layer of ions is therefore considered an ionic mixture, rather than in terms of a

number of concentric shells composed of distinct ionic species. As the vesicle

moves through its suspending fluid in the electrophoresis apparatus, part of its

diffuse double-layer moves with it and part stays behind with the bulk phase. The

plane between that part of the diffuse electrical double-layer moving with the cell

and that part remaining with the bulk phase of the suspending fluid, is the

hydrodynamic slip-plane.

33
 Solid
(particle) Liquid

+- +
++

++
+ ++
1+
Surface

Stem (Rigid) Layer

,
niffi co bulk of
L. 4yt: MI
solution
Slapping
lpp
laneg
plane
p
qjo
qJ6

Figure 4: The electrical double layer in the interfacial region to a charged surface

and the potentials in the surface region.

34
 Electrophoresis is, essentially, a measurement of the electrical potential between

an imaginary electrode located at the hydrodynamic slip-plane and one placed an

infinite distance away in the medium. This potential, the zeta potential, ý, is less

than the true surface potential, Oo, but the difference between the two is a matter

of conjecture (see later, page 38).

Electrophoretic mobilities allow rough approximates, within limits, of zeta-

potential to be made and even rougher estimates of surface potential, (Oo).

However, for vesicles of similar surface configuration, measured under similar,

nontraumatic environmental conditions, comparisons of average surface charge

density may be made from electrophoretic data, but absolute values of average

surface charge density are not exact.

Measurement of the Zeta Potential from Electrophoretic Mobility

Several phenomena, for example, electrophoresis, electroosmosis, streaming

potential and sedimentation potential can be exploited as a measure of zeta

potential and can be grouped under the heading of electrokinetic effects. They

involve the movement of a charged surface relative to its surrounding fluid, with

an electric potential either causing or resulting from the movement. The most

widely studied of these phenomena is electrophoresis, the movement of charged

particles suspended in a liquid medium under the influence of an applied electric

field (Shaw, 1968).

When an electric field is applied across an electrolyte, charged particles suspended

in the electrolyte are attracted toward the electrode of opposite charge. Viscous

forces acting on the particle tend to oppose this movement and, when an equili-

brium is reached between electrical attraction and viscous drag, the particle

moves with a constant velocity. The velocity is dependent on the strength of the

35
electric field or voltage gradient, the dielectric constant and the viscosity of the

medium, and the zeta potential. The velocity of a particle under unit electric field

is referred to as it's electrophoretic mobility. Zeta potential is related to electro-

phoretic mobility by the Henry equation:

cf f(K, )

where ju- mobility; e= dielectric constant; q- viscosity; a- particle radius and K

is the Debye Huckel parameter, which depends on the electrolyte concentration.

The units of K are reciprocal length, and I/K is often taken as a measure of the

"thickness" of the electrical double layer surrounding the particle.

Electrophoretic determinations of zeta potential are commonly made in high

electrolyte concentration, which results in a large value for K.. Under these

conditions, f(K. )- 1.5 and the Smoluchowski form of the Henry equation becomes

operative:

[21
'U =eC... ..........................
4r 17

which, in water at 250C, reduces to:

C- 12.8514 (mV) [3]

..............................

Thus, measurement of electrophoretic mobility leads to a simple calculation of

zeta potential. At low values of K. (for instance, sub-micron particles in a low

ionic strength medium), AK. )- 1.0, and an equally simple relationship results in

which the 4m7 factor of equation [21 is replaced by 61rq.

36
Various techniques exist for measuring electrophoretic mobility. Since the most

practical systems exhibit a range of electrophoretic mobilities for the several

distinctly different particles, for example, as is found in blood, the technique of

microelectrophoresis is generally preferred. The essenceof a microelectrophoresis

system is a cell, flat or cylindrical, at the ends of which are the two electrodes

between which the potential gradient is applied. The particles move toward the

appropriate electrode and their velocity is measured and expressed per unit field

strength as their mobility. Early methods of microelectrophoresis involved the

direct observation of individual particles using high resolution microscope

techniques with manual timing of their progress over a fixed measured distance.

This technique is particularly difficult in the case of poorly visible particles such

as those of small size or unfavourable refractive index. Problems in measurement

also arise with large particles where sedimentation occurs. A further limitation of

manual methods of microelectrophoresis is that particle concentration has to be

sufficiently reduced to allow an individual particle to be observed for the duration

of the measurement.

Alternatively, microelectrophoresis is performed using a low power laser and

modern day instruments display the mobilities of a number of particles in a given

sample in the form of a spectrum, from which zeta potentials are calculated and

displayed automatically, such as the Malvern Zetasizer lIcR*. The complete

measurement takes only a few minutes and the concentration of the sample does

not affect the results. The Zetasizer offers the added capability of submicron

particle size analysis to complement the electrophoretic measurement capability.

Calculation of Surface Charge

The net charge on the vesicles can be derived from the electrophoretic mobility,

using the Helmholtz-Smoluchowski equation, as shown overleaf-.

37
 Z- 17p'.............................. [41
io er

The zeta potentials are obtained by substituting the individual valuesof ju and e.,

the permittivity of free space, t) and e, the viscosity and the real part of the

relative permittivity of the suspendingelectrolyte, respectively. For aqueous

solutions of low ionic strengths, the physical constantsfor pure water can be

used without serious error. For water at 250C (298K), these values are as follows:
4NS.
rj= 8.3 X 10- M-2, er = 78.54 (Weast,1984). The potential, C, derived from

equation [4] correspondsto the potential at the hydrodynamic plane of shear, a

small distance beyond the outer membraneof the vesicle. No satisfactory theory

appearsto have been developedwhich accurately relates C to the actual surface

potential, Oo,of a cell or colloidal particle. Measurements(Pething CLfiL,1984)

using anionic free radicals to probe the surface charge of Ehrlich ascitescells
have provided a value of -5.7mV for the difference between the zeta potential and

the actual potential at the cell membranesurface. For the purposesof the work

described here, several assumptionsare made.

Assumption I: Oois equal to C.

Assumption 2: the vesicle wall is planar, and

Assumption 3: the chargesare distributed uniformly over the surface.

Then, for an electrolyte composed of mixed valency ions, the surface potential,

Oo, can be related to the surface charge density, a, by.-

a2= 2tocr kT Ej [
njb exp ( -zjqoo [5]
.................
kT

in equation [5), which is summed over all ionic species, j, of valency Zj ,k is

the Boltzmann constant, q is the absolute magnitude of the charge on an electron

19
(1.602 X 10- coulombs), T, is the absolute temperature and njb is the ionic

38
concentration (ions dm-3) in the bulk electrolyte away from the surface of the

vesicle.
If the electrolyte is not of mixed valency, but is a Z-Z electrolyte, then equation

[5] simplifies to:

c, (8nV. crkT)I. sin.h (ZCIOo) [61

...................
UT

Since nb NCb9 where N is the Avogadro constant and Cb is the bulk molar

concentration of the electrolyte, then numerical substitution into equation [61,

with e. " 78.54 and T- 298K gives :

a- 11.77 Cbl sin. h (Zqoo) [7]

. ...................
UT

In this equation, a, is in C/cm2, Cb is in mol. 1-1 and Oo is in volts. The

convention used in equation [7] is that the surface charge density is taken to be

negative for negative values of surface potential, Oo . Grahame (1947) adopted the

opposite convention, that is, a is positive for negative Oo, and vice versa, and so

a negative sign appears in his equivalent formula.

Gel Electrophoresis

Protein adsorption to vesicles is known to affect their surface charge (Alying

fLqL1978) and is an important parameter in their clearance rate from the

circulation (New CUL. 1978). It has been shown that liposomes of various

compositions have different surface charge characteristics and acquire different

arrays of bound protein upon exposure to blood; thus while all types of vesicle

acquire a net negative surface charge in the circulation, the identity of the bound

protein, rather than the surface charge, probably modulates the interaction of

liposomes with reticuloendothelial cells and thus their clearance from blood.

39
Since these bound proteins carry a net charge at any pH, other than their iso-

electric point, a useful method of studying their nature and composition makes

use of this charge. Proteins migrate at a rate which depends on the charge

density (the ratio of charge to mass); the higher the charge mass ratio, the faster

the molecule migrate. In theory, separation of different proteins as discrete zones

is readily achieved provided their relative mobilities are sufficiently different and

the distance allowed for migration is sufficiently large. Therefore, separation of

proteins using polyacrylamide gel electrophoresis (PAGE) depends on the charge

density of the proteins at the pH selected. Further, the use of gels prevents

convection currents, minimises diffusion and may also actively participate in the

separation process by interacting with the migrating particles. These gels can be

considered as porous media in which pore size is of the same order as that of the

protein molecule so that molecular sieving occurs and separation depends on

charge density and size. Polyacrylamide gel has become the medium of choice for

zone electrophoresis of most proteins since a wide range of pore size is readily

available. In addition, polyacrylamide gels have the advantages chemical inertness,

stability over a wide range of pH, temperature and ionic strength and trans-

parency which makes recording easy.

1.4. INTERACTION OF LIPOSOMESWITH

CELLS

The interaction of phospholipid vesicles with cells is a promising area in cell

biology since they have a potential for introducing new material into the

cytoplasm or the membrane of the cell. To date, growth- regulating nucleotides

(Papahadjopoulos CLaL,1974). drugs (Mayhew CfXL,1976), proteins (Weissmann CLQL

1977), nucleic acids (Mayhew CLaL,1977) and intact viruses (Wilson dAL. 1977) have

been introduced into cultured cells after encapsulation in vesicles. Lipid molecules

40
(Martin and McDonald, 1976) have been introduced into the red cell membrane and

water-soluble fluorescent molecules into cytoplasm of lymphocytes (Weinstein

1977;
CLqL, Blumenthal CL&1,1977)by interaction of vesicles containing these

compounds and the recipient cell. These studies have promoted the potential of

lipid vesicles as pharmacological tools. However, although the subject has been the

focus of considerable study (Pagano and Weinstein,1978;Kimelberg and

Mayhew. 1978), the mechanism of vesicle uptake by cells appears to be complex

and remains poorly understood. Various investigators have shown that adsorption,

phospholipid exchange, endocytosis and fusion are all possible pathways for the

interaction of vesicles with cells, as shown in figure 5. Adsorption would appear

to be the common initial interaction which precedes the subsequent mechanisms.

After adsorption to the cell (figure 5a), a liposome is likely to slowly release its

contents, some of which may enter the cell, depending on the nature of the

vesicle contents and the type of cell involved. An endocytosed liposome may be

processed (figure 5b), by a lysosome, an intracellular digestive organelle, after

which the lipid components of the liposome may be incorporated into the mem-

brane of the cell, whereas the aqueous solutes that escape lysosomal degradation

may be incorporated into the cytoplasm. A liposome that undergoes lipid exchange

(figure 5c) may take up lipid from the membrane of the cell and in return may

give up some lipid to the cell. When a liposome fuses with a cell (figure 5d), the

liposomal membrane merges with the cell membrane and the liposomal contents

become integrated with the cytoplasm of the cell.

A model for some of these interactions of liposomes with an in vitro cell system

is provided by cells of Tetrahvmena elliotti. T. elliotti is a free-living ciliated

protozoan, approximately seventy microns by thirty microns in size, as shown in

figure 6, as shown on page 43.

 a ADSORBED
I IPOSOML
AQUEOUS SO[ LJTL
0 b LiposomF
ýTTTrMTTT I [T I MTTTTT I rT UNDE RGOING
F NDOCY10SIS

.1
""
.
-
b:.

UNDIGESTED
PARTICLES
EXCRETED

4f
ý%
40'

PIDS EXCRETED
Y LYSOSOME

0 00

-0 0

TTTTiIIIIIIIIITIII
". . . -1-1-LLIJ-Li-Ill-LLI-LU 1111 111111111 1111110w> OW
Li r' 00
00
f USF DL IPOSOME C, IPOSOME UNDI H
GOING i lillt)
f XCHANG1

Figure 5: A schematic diagram showing the various Points of attachment of

liposomes with a cell. (see text for explanation).

(Reproduced from Ostro (1987).

42
 The ciliate protozoon, Tetrahymena. (a) A general view, showing extemal appearance.
(b) Diagrammatic cross section, showing main %tructural feature,, of the cell

(a) (b)

m
Jn Cylopharynx

Connec inn9
fibril
MiCronucleus
0
-----\LMacronucleus
ýý-Food
Basal granule
vacuoles
Contractile (8)
vacuole /I )
- \ý
Contractile -- Pellicle

vacuole pore

L- Cylop"Ct

147t2
It?
.

Figure 6: A diagramatic representation of Tetrahvmena elliotti.

43
Microscopic studies of ingestion in Tetrahvmena (Levy and Elliolt, 1968) have

shown that after entrapment, particulate material is invested with a membrane

within the oral apparatus and enters the cytoplasm as a food vacuole or

phagosome. Autophagic vacuoles (cytolysosomes) are also formed upon exposure to

non-nutrient media (Nilsson, 1970a) as well as during the stationary phase of

growth in nutrient media.(Ellioll and Clemmens,1966).

Nutrients in solution may also be taken up by the formation of pinocytic vacuoles

at specific sites along the pellicle, several of which coalesce to form the equi-

valent of a food vacuole (Elliott and Clemmens,1966). Degradation of engulfed

material begins when the phagosome or cytolysosome fuses with a primary

lysosome to form a phagolysosome (Elliott and Clemmens.1966). After fusion with

primary lysosomes both food vacuoles and autophagic vacuoles appear to be

processed in the same way, that is, digestion of the vacuolar contents with

permeation of the nutrient products into the cytosol and eventual egestion of the

indigestible residue via the cytoproct or cell anus.

Lysosomes are membrane -delimited structures containing characteristic hydrolytic

enzymes, most of which have acid pH optima. The lysosomes of Tetrahvmena, first

seen in electron micrographs (Elliolt, 1965) and sedimented from homogenates

(Muller rLa1,1966), are similar to those of many other cell types and contain a

number of hydrolases with acidic pH optima (Muller CLgL,1966). The known

constituents of these lysosomes include the following: hydrolytic enzymes,

structural lipids and proteins of the membrane, products of hydrolysis which

accumulate in the lysosomes and cellular parts and macromolecules which have yet

to be digested or which resist digestion.

44
Since pure lysosomes from T. elliotti cannot be obtained in sufficient quantities the

enzymes used in this work were all purchased commercially. However, the

hydrolytic enzymes within these lysosomes fall into the usual classification. A

number of lysosomal enzymes exist (for a detailed review, see Dingle, 1972) but the

action of two that are relevant to the intracellular processing of liposomes and

other related vesicles will be described here.

Phospholipase A2

Phospholipase A2 (EC 3.1.1.4.) belongs to the class of lipolytic enzymes that are

characterised by their ability to hydrolyse lipid substrates. The phospholipases are

classified according to the particular ester bond of the phosphoglyceride substrate

hydrolysed by the enzyme. Phospholipase A2, specifically, catalyses the hydrolysis

of the 2-acyl ester bonds of naturally occuring and synthetic phosphoglycerides

(van Deenan and de Haas,1964), producing a I-acyl-lysophosphoglyceride and fatty

acid.

SITE OF
ACTION 0
\2/ HC CHrRi 2HC O? Ri

Rs-C-OCH + H20 HOCH + R2COOH

ý
6 PhompholipaseA2
C&2+ 0
40 A
2HCOPOX 211COPOX
Noe Nop

X represents any of the polar head group moieties found in 3-sn-phospho-

glycerides, for example, H, choline, etc., R, and R2 are alkyl chains.

45
The enzyme is highly stereospecific and has an absolute requirement for calcium

ions that bind in a 1:1 molar ratio to the enzyme. The Ca2+ adds to the enzyme

before the substrate attaches to this complex (Nilsson, 1970b). This attachment

induces a conformational change in the enzyme-Ca2+-substrate complex. The fatty

acid is released from the enzyme, followed by the other product, lysophos-

phatidylcholine (lysolecithin) (Nilsson, 1970b). The stepwise reactions are

summarized below, where E represents phospholipase A2, S is the substrate, for

example, phosphatidy1choline and RCOOH is the released fatty acid.

S
ý-E-W+-S
E+ Ca2+---4 E-Ca2+

E-Ca2tlysolecithin-RCOOH
][-RCOOH

E-Ca2+-Iyso ecithin
t
lysolecithin + E-Ca2+

E+ Ca2+

An important and characteristic feature of most lipolytic enzymes is the strong

dependence of the catalytic properties of these enzymes on the physical state of

their substrates.

Liposomes are known to be poor substrates for pure pancreatic phospholipase A2

(de Haas eLaL1968), although, saturated long chain lecithins become more

susceptible to hydrolysis at their thermotropic phase transition (O'P den Kamp

CL.gL1975). Both below and above these temperatures (Tc), the more regular and
,
tighter packing of the phospholipid molecules may prevent formation of the

enzyme- interface complex. At the Tc, domains of molecules in the gel state exist

together with domains where the lipids are in the liquid-crystalline state, and

surface defects at the borders between these domains would allow the enzyme to

46
penetrate the interface and hydrolyse substrate molecules (Op den Kamp CLQL,

1975). Investigations with snake venom phospholipases show these enzymes are

more active on bilayer substrate systems, which are their natural substrates

(Slotboom CW, 1982). However, crude snake venom also includes a number of

other enzymes, as shown in Appendix 2, page 161.

Carboxylic Ester Hydrolase

Carboxylic ester hydrolase (EC 3.1.1.1.) catalyses the hydrolysis of a large number

of uncharged carboxylic esters. In contrast to lipases their action is generally

restricted to short chain fatty acid esters.

R-/ ESTERASE/ H20 R-/ + RI- OH

\0R, \
OH

ESTER FATTY ACID ALCOHOL

Unlike phospholipase A2, this enzyme does not require specific co-factors and is

thought to hydrolyse most, commonly encountered, ester linkages.

1.5. NON-IONIC SURFACTANT VESICLES

,

The formation of surfactant vesicles was first detected following the sonication of

a didodecyldimethyl bromide solution (Kunitake and Okahata. 1977). Since then a

large variety of synthetic amphiphiles have been shown to form membrane-like

molecular aggregates (Kunitake, 1986), such as bilayers, monolayers and vesicles, as

47
shown in figure 7, page 49. It is the hydrophobic effect (Tanford, 1980) which

provides the driving force for such aggregation, when surfactant molecules are

placed in an aqueous environment.

The type of aggregate formed and its properties, is defined by the structure of

the amphiphile. Ionic (Kunitake and Okahata, 1978), non-ionic and zwitter-ionic

dialkyl amphiphiles (Handjani-Vila CLgL,1979;Okahata CLqL,1981) are capable of

forming vesicles. Although the length of the chain is important for bilayer

formation, it is not essential for the hydrophobic part of the molecule to be a

double chain (Kunitake and Okahata, 1980). Many single chain ionic (Kunitake and

Okahata, 1980;Hargreaves. 1978) and non-ionic (Baillie rLgj, 1985) amphiphiles form

vesicles. In general, non-ionic surfactant molecules contain polyethoxy chains

which, together with the alkyl chain length, determine the aggregate morphology.

General
-Structure:
CH3(CH2)n-i OCH2
.....
CH 0 (CH2CH20)X H
... ... ...
CH3(CH2)n-1 OCH2
.....
hydrophobic hydrophilic

n- 12,14,16 and 18 units

x- 6-30

Some non-ionic surfactants have been shown to be relatively non toxic and to

enhance absorption of drugs across membranes (Kaneda efgL, 1974) by increasing

membrane permeability (Attwood and Florence, 1983) and in some cases, by

solubilising biomembrane components (Whitmore rLaL, 1979). This property of

amphiphiles, that is, their capacity to solubilise and hence alter the

characteristics of many drug molecules, potentiates their use in the field of drug

delivery.

48
 Qlf

wwtbr
R
c3 ::0
cyý

monoloyer sphericol rod-like

micelle micelle

reversed
micelle w/o microemulsion o/w microemulsion
ý --Zzzý
sonication

ý-30 rw, mRiRRRt,- ob

BLM single
black lipid compartment
membrane vesicle

multicompartment vesicle

Figure 7: A representation of organized structures formed from surfactants.

(photograph from, Fendler, J. li and Tundo, P (1984), "figure I").

49
 1.5.1.NIOSOMES:POTENTIAL DRUG CARRIERS

The discovery (Handjani-Vila cLAL,1979) that non-ionic surfactant molecules, such

as surfactant 1, as shown in figure 8, are capable of forming vesicles, niosomes,

entrapping an aqueous solution, was a lead to their potential use as drug carriers.

A system which could combine the advantages of liposomes with the ability to

increase membrane permeability displayed by the non-ionic surfactants would be of

great interest. An investigation to compare and contrast some of their relevant

properties with the apparently similar and well characterised liposome system

(Gregoriadis and Allison, 1980) is vital to the development of a niosomal drug

carrier system. A major prerequisite to the use of niosomes and other vesicles as

drug carriers is their integrity in biological fluids. These are, for example,

interstitial fluid (sub-cutaneous administration), synovial fluid (intra -articular

injection), contents of the stomach and intestine (oral route) and peritoneal cavity

(intra-peritoneal administration). However, the great majority of potential in vivo

applications involves intravenous administration so that stability in blood,

especially plasma has been studied extensively. In a carrier role, niosomes must be

able to both circulate in the body and retain drugs for significant periods of time

to optimise access to, and interaction with, target tissue and in appropriate

circumstances delivery of their contents to the interior of cells. An under-

standing of these processes which may affect niosome integrity in vivo is

essential to a study of "niosome -encapsulated drugs".

Preliminary work within these laboratories has shown (Baillie daL, 1984,1985). that

niosomes appear to be similar, in terms of their physical properties to liposomes.

Studies, in mice, have shown modified tissue distribution and excretion of

methotrexate entrapped in niosomes (Azmin CW, 1985&1986). Niosome formulation

caused methotrexate accumulation in the liver and enhanced level of the drug in

50
 Cle,
Hi,ý-O

H-0
Surfactant I
ItO H m. wt., 473
L

CýHý;
O
Surfactant Il
H, m. wt., 972
C,,Hgý H-0 CHICH-0
II H
(, HiOH
L J7

cre
00H
Cml-ý7 0 -C Hl- ClH
I Surfactant III
+C Hi--O- CH m. wt., 404
z
Cýb C 40 H H-OH A (92%)

c, 0-ý H HI-OH
C H-0-
2 H
H-OH B (8%)
H-OH

Figure 8: Three vesicle forming non-ionic surfactants synthesised by

Vanlerberghe gLqL (1978). 3 and 7 are number average values. Surfactant I and
11; the hydrophobic alkyl chain ether-linked to the hydrophilic polyglycerol head
group. Surfactant III is ester-linked and is a mixture of 2 isomers, A and B.
51
.
the brain after intravenous administration. Similar reports of brain accumulation

have been reported for methotrexate entrapped in liposomes after intravenous

injection into rats (Freise Vd, 1981). Niosomally- entrapped doxorubicin has also

been shown to have altered tissue distribution Intravenous injection of hand

.
shaken niosomes containing doxorubicin into mice (Rogerson 1986,Rogerson

f,LGL,1988) showed no apparent liver or spleen loading with drug but some

evidence of accumulation of doxorubicin in the lungs. These results may be a

direct consequence of the size distribution of the vesicles used, although the

lung-loaded perhaps indicates intravenous aggregation of the niosomes. Increased

anti- leishmanial activity after passive targeting of sodium stibogluconate to the

liver using niosomally-entrapped drug (Baillie eW, 1986) is further evidence of the

potential of drug-carrier role for niosomes. As part of an approach to the

optimisation of this drug-carrying potential of niosomes, it is important to

characterise their stability in terms of release of entrapped solute. The result of

a series of in vitro experiments, simulating physiological conditions in vivo, are

presented here to define the effects of blood proteins in niosome stability and in

addition the influence of pH and temperature on the integrity of various types of

non-ionic vesicle.

Cosmetic Application

NiosomeR' has recently been marketed commercially as a skin anti-aging system

by Lanc6me Laboratories, France. The "niosome system" is a complex of niosomes

believed to transport active elements within the skin. These active elements are

transported enclosed in the niosome sphere and are fully protected, so they can

be released intact where the skin needs them most. These niosomes are thought to

have an added integral reconstructive action because of their "biomimitism"

phenomenon. This gives them the potential to become incorporated into the skin's

52
intercellular support organisation to reconstruct it where it has become

disorganised.

Uposomes have set the standard in the field of drug carrier systems for

invetigating these new surfactants with a variety of composition effects. Nlosomes

may provide an alternative type of carrier system.

This thesis reDorts Dart of the work from ongoing research In our laboratories.

AIMS OF PROJECT

1. Preparation of niosomes from three non-ionic surfactants provided and

investigation of physical properties in direct comparison with liposomes.

2. Stability of vesicles in human plasma and serum, electrophoretic mobility

measurements and surface charge calculations.

3. Identification of proteins adsorbed to niosomes; a comparative study with

liposomes.

4. In vivo stability of vesicles in an eukaryotic cell, Tetrahvmena elliotto

uptake and degradation studies.

53
 SECTION 2

EXPERIMENTAL
Non-ionic surfactants, I, 11 and 111, were a gift from L'Or6al, France.

5,6-carboxyfluorscein (CF; Eastman, Kodak, Liverpool, England) was partially purified

(see page 56) over activated charcoal before use.

Cuprophan dialysis tubing (size 15mm diameter) was purchased from Medicell

International Ltd, London, England.

Acrylamide, ammonium persulphate (APS), amidoblack (napthol blue black), bis-

acrylamide, bovine serum albumin (BSA), bromophenol blue, calcein, carboxylic

ester hydrolase, cholesterol, 4-chloro-l-napthol, Coomassie blue R250, dicetyl-

phosphate (DCP), dipalmitoylphosphatidy1choline (DPPC), dimyristalphosphatidyl-

choline (DMPC), egg phosphatidy1choline (egg PC), glutaraldehyde, glycerol,

glycine, horse radish peroxidase, mercaptoethanol, phospholipase A2, sodium

dodecyl sulphate (SDS), snake venom from Vigera russeill (Russell's Viper),

stearylamine, thrombin and trypsin were purchased from Sigma Chemical Company,

Poole, England.

EDTA, N, N, N', N"-tetramethylethylenediamine (TEMED) and urea were purchased

from BDH Chemicals, Poole, England.

Rabbit antihuman IgG was purchased from Miles Laboratories, Stough, England.

Plasma, pooled citrated human, was obtained from the Royal Infirmary Hospital,

Glasgow, Scotland. Serum was prepared from citrated plasma by the addition of

thrombin and heat-inactivated.

54
Tetrahymenaelliotti strain 1630/1C(T. elliotti was obtained from the Culture

Centre of Algae and Protozoa, Cambridge, England.

Proteose peptone and liver extract were purchased from Oxoid, Basingstoke,

England.

All other reagents were of analytical grade.

2.2. METHODS

2.2.1. PRODUCTI N OF VESTCLES

-

Ether Injection (E. I. ) Niosomes

Surfactant or surfactant/cholesterol mixture (1.50 X 10-4M) was dissolved in

diethylether (20ml) and injected slowly (0.25ml min-1) through a needle (14 gauge)

into an aqueous solution of CF (5ml, 0.2M, pH7.4) maintained at 60oC.

Hand Shaken (H. S. ) Nlosomes

Surfactant or surfactant/cholesterol mixture (1.50 X 10-4M) was dissolved in

chloroform (10ml) in a round bottomed flask. The chloroform was removed at room

temperature (220C) under reduced pressure on a rotary evaporator to form a thin

film on the wall of the flask which was then hydrated with CF (5ml, 0.2M) at

500C. Gentle agitation for I hour completed the formation of niosomes.

55
H. S. Uposomes

Phospholipid/cholesterol mixture (1.50 X 10-4M) Was weighed into a round

bottomed flask and dissolved in chloroform (10ml). The chloroform was removed at

room temperature as above and the resulting thin film hydrated with CF solution

(5ml, 0.2M) at 500C under an atmosphere of nitrogen and agitated gently for I

hour.

"Negatively- Charged" Vesicles

Surfactant/cholesterol or phospholipid/cholesterol admixed with dicetylphosphate

(DCP, 10mol%, total molarity 1.50 X 10-4M) were dissolved in chloroform (10ml) in

a round bottomed flask. The chloroform was removed by evaporation at 220C

under reduced pressure (within a nitrogen atmosphere), leaving a thin layer on

the wall of the flask. Hydration with CF (5ml, 0.2M) for I hour at 500C with

gentle agitation resulted in the production of vesicles.

"Positively- Charged" Vesicles

Surfactant/cholesterol admixed with stearylamine (2 mol%, total molarity,

1.50 X 10-4M) was dissolved in chloroform and vesiclesformed as before. using

NaCi solution (2-0 X 10-3M) instead of CF.

Purif Ication of CF

CIF was purified by modifying the method of Ralston t1al (1981). Commercially

available CF (25g) was treated with activated charcoal (10g) in boiling ethanol

(300ml) contained in a round-bottomed flask (I litre). After refluxing for 30

minutes the mixture was cooled and filtered through Whatman No. 50 filter paper.

56
Cold distilled water (600ml) was addedand solids allowed to precipitate overnight

at OOC.The water was removed by filtration and the orange-coloured precipitate

was washedthoroughly (4 times, 50ml distilled water), dried at 500C in an airtight

container and stored in the dark at room temperature.

CIFSolution

Aliquots of this orange powder (0.2M) were weighed, made up to volume with the

medium (distilled water or phosphate buffered saline [PBS]) and adjusted to pH 7.4

with sodium hydroxide (NaOH, 4N).

Buffered CF

CF is usually made up in distilled water to the required molarity at pH 7.4 by

adding sodium hydroxide solution (NaOH, 4N). In *buffered CF", the distilled water

was replaced with phosphate buffered saline (PBS) and the requisite amount of

NaOH, to the final pH 7.4.

Separation of Free and Entrapped CF

Volumes of aqueous surfactant or phospholipid dispersions (5ml) from ether

injection or hand shaken techniques were exhaustively dialysed against PBS,

(1.30 X 10-3M, 0.9%w/v NaCI solution, pH 7.4). In experiments with T. elliotti,

NaCl solution was omitted from the buffer.

57
 2.2.2. STABILITY OF VESICLES

Leakage of CF

The CF entrapped within the vesicles is self-quenched at the working concen-

tration of 0.2M at the wavelenghts of measurement. Leakage and ensuing dilution

into the extra-vesicular bulk volume, increases the fluorescence of CF markedly,

which was measured (486nm excitation, 514nm analyser wavelengths) using a

Perkin ElmerR* 203 spectrofluorimeter.

Samples (2.5 X 10-2MI) were added to the test media at "time zero" to give a

final volume of 5ml. The fluorescence measured at these times were taken as zero

percent, although in practice 15 seconds elapsed before these readings could be

recorded. This amounted to 2-6% intensity of the total fluorescence. The samples

were gently agitated throughout the experiments and further readings obtained at

various time intervals. Maximum fluorescence (100%) for all niosomes and liposome

suspensions, Ftot, was measured after vesicle disruption by addition of propan-l-ol

(O.Iml), or Triton X-100 (O.Iml). The leakage of CF was corrected for background

fluorescence at "time zero", Fo. The percentage of CF released in each sample

was calculated as follows:

100 (Ft-Fo)
%Release -
Ftot

in which Ft - intensity at time "t".

Effect of pH

The leakage of CF was measured as described above, by challenging the vesicle

suspensions (2.5 X 10-2MI) with a variety of different buffers (McIlvaines citric

phosphatebuffer, pH 2.0 to 8.0) and incubating at 370C. All the solutions used

58
were of equal ionic strength (1.24) to that of the CF solution within the vesicle,

thus preventing osmotically driven leakage of the CF solution from the vesicles.

Effect of temperature

The efflux of CF from the vesicles was measured after incubation of a suspension

(2.5 X 10-2ml) in buffer (5ml, PBS at pH 7.4) at various temperatures (40C, 220C,

370C 0Q
and 50 for various time intervals.

Preparation of Human Serum

Serum was prepared from plasma by treatment with thrombin (20 NIH units ml-1)

at 370C for 10 minutes with gentle stirring, after which period the clot formed
0C
was removed. Serum was heat- inactivated by incubation at 51 for 30 minutes.

Any dilutions of plasma or serum were made by addition of PBS (pH 7.4) to the

correct volume.

Effect of Plasma

Leakage of CF from the vesicles was measured as above (page 58) after incubation

of a suspension (2.5 X 10-2ml) of vesicles in human plasma (100%) at 370C. Total

fluorescence (100%) was evaluated by disrupting all the vesicles using Triton

X-100 (O. Iml). Leakage of CF was also monitored in the same way in the presence

of serum and heat- inactivated serum.

Effect of 10% BSA

BSA (400mg) was dissolved in PBS(100ml, pH 7.4) at 370C. Leakageof vesicles

was measuredas above.

59
The electrophoretic mobility (ju) was measured as a function of pH in a laterally

placed flat cell micro-electrophoresis apparatus (Rank Brothers, Bottisham,

Cambridge, England) with an optical assembly and constant temperature bath at

0
25 C. The mobility was determined by measuring the time taken in seconds for

the vesicles to travel a pre-determined distance (usually 2cm) under the influence

of a constant known electric field (80 Volts). The vesicles were suspended in

solution (NaCl, 2X 10-3M), the pH of which was varied by the addition of dilute

hydrochloric acid (HCI) or sodium hydroxide (NaOH). The mobilities of at least 40

vesicles were measured at each pH and an average mobility obtained.

To measurethe effect of plasma(human, 50%) on electrophoretic mobility, vesicles

NaCl (2 X 10-3M) in (0.2M). These

were prepared entrapping glucose solution

vesicleswere incubated in plasma for 2 hours, centrifuged at 210g for 5 minutes,

washed twice and their electrophoretic mobility measuredas described above.

These measurements were used to calculate surface potentials.

2.2.4. OF ADSORBED
-IDENTIFICATION

Gel Electrophoresis

Electrophoresis in polyacrylamide gels containing sodium dodecyl sulphate (SDS;

0.1%) was carried out according to the method of Laemmli (1970) with minor

modifications, using a ShandonR*

vertical slab unit 200 (Bio-rad, Hertfordshire,

England).

60
Sample Preparation

Vesicle suspension (Iml) was incubated with plasma (2ml; 10% or 100%) for I hour,

2 hours or 24 hours. Samples were then diluted with PBS (10ml) and the vesicles

centrifuged at 10OOgfor 5 minutes. The pellet was resuspended, washed 3 times

and suspended in PBS (Iml). Measured quantities (2.5 X 10-2mg-7.0 X 10-2mg

protein per 2.0 X 10-2MI-5.0 X 10-2M, sample buffer) were dissolved in sample

buffer (6.25 X 10- 2M Tris-HCI, plus 2%SDS, 10%glycerol, 5%mercaptoethanol,

0.001%bromothymol blue, pH 6.8), heated (370C for 15 minutes) and cooled.

Cleavage by TryPsln, EDTA, Urea and 1M NaOH

Vesicleswere incubatedwith plasma(I hour, 370C).Thesepreparationswere

centrifugedand the resultingpelletswashedin buffer (PBS, X 3

1.3 10- M, p117.4)
Sampleswerereincubatedwith freshly preparedtrypsin (ImM),
and resuspended.
EDTA (5mM), urea(4M) and NaCI(IM) for 30 minutes.Thesesampleswere then
for analysison polyacrylamidegel as
by centrifugation,and resuspended
processed
previouslydescribedabove.

Slab Gel Electrophoresis

Samples were carefully loaded onto a 7.5% acrylamide gel and run normally at

200V until the dye entered the separating gel. The voltage was then reduced to

IOOV and left to run for 5 to 6 hours. The run was terminated Mien the dye

front reached 2cm from the bottom of the gel.

61
Staining and Destaining of Gels

Gel slabs were placed in a plastic tray containing Coomassie blue R250 (0.1

w/v in watermethanol:glacial acetic acid, 9:9:2) and agitated slowly. The gels were
fully stained within 4 to 6 hours at room temperature (220C). After staining was

completed, excess stain was removed to allow protein bands to be seen clearly.

Gels were destained in severalwashesof water:methanol:glacial acetic acid (9:9:2).

Drying Gels

Gels were dried using a Bio-radR* drier (model 224) apparatus. Two sheets of

filter paper (3mm) were placed on the stainless steel support screen of the dryer

and wetted with water. The slab gel was aligned onto this, care being taken not

to trap air bubbles under the gel since this would result in cracking during

drying. The gel was overlaid with a sheet of pre-wetted Saren WrapR* (cling

film), then a porous plastic sheet, and finally the silicon sheet (attached to the

apparatus) forming a leak-proof seal. Vacuum was supplied by a water aspirator or

vacuum pump fitted with a cold-finger water-trap and the heating block of the

dryer was turned on. The exact time for drying depended on the size and

concentration of the gel, generally 2 hours was sufficient. If air enters the

assembly before drying is complete the gel will crack. The resulting dried gel was

sandwiched between its filter paper and the protective Saren WrapR*.

Photography of Stained Protein Bands

Wet slab gels were laid directly onto an opal white screen illuminator, avoiding

air bubbles, and kept moist during photography by addition of 7% acetic acid. A

fine grain, panchromatic film was used and a medium-red filter to increaseband

contrast.

62
Electroblotting

This was carried out by the method of Towbin tLIJ (1979). Proteins were first

subjected to PAGE-SDS. Then the proteins were transferred to nitrocellulose

sheets. A sheet of nitrocellulose (0.45 prn pore size in roll form, MilliporeR) was

wetted with water and laid on a scouring pad (Scotch- B riteR*) which was

supported by a stiff plastic grid. The gel to be blotted was placed on the nitro-

cellulose sheet and any air bubbles were carefully removed. A second pad and

plastic grid were added and secured in position by two strong rubber bands. The

assembly was immersed vertically in an electrophoretic Trans-blot (Bio-RadR*)

with the nitrocellulose sheet facing the cathode (-ve). The electrode buffer was

25mM Tris 192mM glycine in 20% methanol (vol/vol), pH 8.3. The electrophoretic

run was at IOOmA, conveniently overnight.

Staining and Destaining of Blots

The blot was sectioned and stained with freshly prepared amidoblack (0.1% w/v

in 45% methanol/10% acetic acid) for 10 minutes. This blot was then immediately
0
destained by washing in several changes of hot acetic acid (2% v/v, 90 Q. This

manipulation was carried out in a ventilated fume hood.

Immunological Detection of Proteins on Nitrocellulose

Antigens (proteins) transferred from the gels to the nitrocellulose. paper were

detected by Enzyme Linked Immuno Sorbent Assay (ELISA) (Towbin cLal. 1979).

The electrophoretic blots (not stained with amido black) were soaked in 3%

BSA/5% GS (goat serum) in buffer (20mM Tris-HCl, l40mM NaCl solution, pH 7.4)

at room temperature, for I hour to block non-specific protein binding sites. They

were then washed in buffer containing Tween 20 (0.2% buffer/Tween) and

63
incubated with the antibody, horseradish peroxidase-conjugated rabbit antihuman

IgG for 2 hours at 40C. The antibody used was diluted I in 500 in BSA/GS/buffer

before use. To detect the antibody reaction the nitocellulose paper was incubated

in the dark with 4-chloro-l-napthol (6mg/ml in methanol) and buffer containing

0.01% H202 (6ml). The reaction was terminated within 10 minutes (indicated by

the developed dark colour) by washing the paper in a large volume of buffer or

deionised water. The blots were dried between filter papers (background staining

reduced on drying). These were stored protected from light.

2.2.5. IN VIVO UPTAKE AND DEGRADATI_QN

The ciliate, T. elliotti strain 1630/1C, was cultured axenically in proteose peptone

(2% w/v) and liver extract (0.1% w/v) at room temperature (220C). The cells were

harvested from these cultures after 4 days growth by centrifugation at I lOg for

15 minutes and the cell pellet was resuspended in phosphate buffer (100ml, p1l

7.4). The cells were resedimented at II Og for 15 minutes and the supernatent

decanted. The pellet was made up to volume (100ml) and left to recover from

centrifugation at room temperature for 24 hours. Samples (O.Iml) of niosome or

liposome suspension were mixed with this cell suspension (Iml) at time zero. At

suitable intervals, volumes (1.5 X 10-2MI) of this mixture were fixed in glutar-

aldehyde (2.5 X 10-2MI, 0.1% solution) and examined by epi- fluorescence

microscopy (PolyvarR*). Further evidence of vesicular integrity was obtained using

a camera mounted on the microscope and time-sequence photography (KodakR*

film, 160 ASA). At time intervals, the total fluorescence in each of 100 T. elliotti

cells was measured using a spectrophoto meter and an average percent of total

fluorescence obtained. Photographs at these various time intervals (0,1,2,3,4,5 and

64
6 hours) were taken and vacuole counts were recorded for over 40 cells at each

interval and vesicle type.

Effect of Enzymes on Vesicles (in vitro

Solutions of phospholipase A2, snake venom and carboxylic ester hydrolase were

all freshly prepared before use.

Phospholipase A2 activity has an absolute requirement for calcium ions (Ca2+) and

stock solutions were prepared by dissolving pure phospholipase (10mg) or crude

venom (12.5 mg) in Tris-HCI buffer (2-0 X 10-1 M, 100ml, pH 7.4), containing

CaCI2.2HiO (3.2g).

Carboxylic ester hydrolase (5mg) was reconstituted before use by dissolving in

Tris-HCI buffer (2.0 X 10-2 M, 1000ml, pH 7.4).

The leakage of CF (0.2M) from vesicles varying in their compositions was

measured, as described, page 58; and compared, when challenged in vitro with the

above enzyme solutions (2.5 X 10-2 ml). Leakage was measured over a time period

of 6 hours. These values were corrected for CF leakage into buffer, as follows:

100 (Fe-Fb)
% Release -
Ftot

where Fe fluorescence intensity in enzyme solution,

and Fb = intensity in buffer, both at time T.

Ftot maximum intensity of fluorescence.

65
 SECTION 3

RESULTS AND DISCUSSION

 3.1. PRODUCTION OF NIOSOMES

Vesicles of various compositions (table 1) were prepared for use in this study

using either the ether injection method (Deamer and Bangham, 1976) or, the

original hand-shaken method (Bangham cLaL1965). The former method results, as

reported for liposomes, in the formation of large unilamellar vesicles, with good

entrapment efficiencies and greater stability in terms of leakage (see later, figure

9, page 72). However, since not all the components required for vesicle production

were soluble in diethyl ether, the hand-shaken method was favoured for this

present study. This method produced multilamellar vesicles with comparable

entrapment efficiencies (table 2, page 68) but higher rates of leakage (figure 9).

The hand-shaken method was preferred for several reasons:

a. these vesicles were easier to produce and handle, with more reproducible

entrapment efficiencies between batches;

b. the overall stability of liposomes and niosomes were easily compared due to

their higher rates of leakage, a major aim of this work; and

C. the components required to prepare the vesicles can be dissolved in suitable

solvents, for example, chloroform, which are easily removed before the

hydration process, thus ensuring a quick, simple and non-hazardous procedure

in the production of every type of vesicle.

The entrapped volume of CF (ml mot-) was measured for 10 different types of

vesicles and is shown in table 2, page 68. Entrapment is greater for liposomes

than in all types of niosomes but does not appear to depend on the surfactant

used. These entrapment efficiencies are low but are comparable with previous

reports in the literature for liposomes (Deamer, 1984). However, these results

suggest that manipulation of compositions may increase the entrapment of

vesicles, example, addition of lipid (10%) with the original (100%) surfactant

composition.

66
 Table 1: VESICLES INVESTIGATED
-NOMENCLATURE -OF

100%Surfactant I Iloo
50% Surfactant I
50% Cholesterol ISOXHOLSO

60% Surfactant I
30% Cholesterol
10% DCP 160:CIIOL30: DCP10

68% Surfactant I
30% Cholesterol
2% Stearylamine 168:CIIOL30: SA2

50% Surfactant 11
50% Cholesterol IISOXIIOL50

60% Surfactant II
30% Cholesterol
10% DCP I160:CIIOL30: DCP10

100%Surfactant III Hiloo

50% Surfactant III

50% Cholesterol IIISOXIIOLSO

70% Surfactant III

20% Cholesterol
10% DCP 11170:CIIOL20: DCP10

100% DPPC DPPC100

50% DPPC
50% Cholesterol DPPCSO:CHOL50

70% DPPC
20% Cholesterol
10% DCP DPPC70: CIIOL20: DCPIO

50% Egg PC
50% Cholesterol Egg PCSO:CIIOL50

50% DMPC
50% Cholesterol DMPC50: CIIOLSO

67
Quantification of Entrapment

To avoid the misleading use of "percentage of starting material entrapped" as the

criterion for quantifying the substances in liposomes, it is advised that entrapment

is expressed as the concentration (moles) or weight (mg) of aqueous medium per

weight (mg) of liposomal lipid. The rationale for this has been discussed in detail

elsewhere (Tyrell and Heath fLOL.1976). The entrapment values reported here

(Table 2, page 69) were expressed in volume (ml) of CF per mole of surfactant. or

lipid mixture. The efficiencies were calculated as a percentage of the total CF

available for entrapment at hydration.

Removal of Unentrapped CF

Free drug or marker, can be removed after preparation of the vesicles in several

ways (see introduction, page 15). Gel chromatography has been used in many

laboratories and has proved an effective method of separation of free drug from

entrapped drug. However, this method takes a considerable period of time and

results in the collection of very dilute vesicle samples, therefore a concentration

step is usually required, for example, ultrafiltration or centrifugation. After

separation of vesicles from unentrapped CF, much of the free marker remains in

the column and this is difficult to remove. Hence, cleaning the column after every

use is time-consuming and the purchase of large quantities of column material

(SephadexR*) is not economically sound. The process of centrifugation provides a

quick, reliable method of separation requiring only a bench top centifuge, for

sedimenting medium to large vesicles, or an ultracentrifuge for smaller vesicles..

However, typical of many types of compact cells, centrifugal force causes rupture

and leakage of the vesicles. The method of choice in this study to separate free

from entrapped drug was dialysis. This was easily performed, conveniently over-

night, with several changes of eqi-osmolar buffer (for details see experimental,

68
 TABLE-2

Methods-of Production

Vesicle Hand Shaken Ether Injection

Type EntraDment Efficien cy Entranment Ef f icie ncy
(MI. Mol-1) (%) ( n*) (MI. MoI-1) (%) (n*)

I100 1670±260 6 6 1111±200 4.2 3

150:CHOL50 760±100 3 6 763±100 3.0 6

160:CHOL30: DCP10 700±99 3 6 601±108 2.3 3

II50: CHOL50 979±129 4 4 1569±120 5.9 6

1160:CHOL50: DCPIO 720±86 3 6 1212±99 4.5 3

III100 1590±63 6 6 1487±206 5.6 4

11150:CHOL50 650±52 3 6 503±86 1.9 4

II170: CHOL20: DCP10 320±68 1 6 355±46 1.3 3

DPPC100 2310±321 9 4 not determined

DPPC50: CHOL50 1140±90 5 2 not determined

(n*) - number of determinations.

The values in this table have been rounded to the nearest whole number.

69
page 57). Occasionally it was necessary to centrifuge the vesicles at low speeds

(<1000g) in the preparation of samples for polyacrlyamide slab gel electrophoresis

(see later, page 99), but here the leakage properties of the vesicles were not of

prime concern.

CF, obtained commercially, is a mixture of the 5- and 6- isomers of carboxylic

acid in unequal propotions. It also contains severalimpurities which can affect;

a. leakageof CF from the vesicles,

b. the stability of the vesicles.

The most effective method of purifying CF is already described (Ralston CUL,

1981) but alternative short cuts are also employed. In the experiments described

in this work, only two batches were purchased. In order to obtain at least 80%

yields, the CF was partially purified over activated charcoal and recrystallised as

described in the experimental (page 56). The CF was not purified completely since

column chromatography gave very poor yields (40%). Therefore, the CF used in

these experiments may contain a small number of impurities which can affect its

leakage through membranes. This may explain any discrepancies in the results

when compared with the literature values for liposomes. However, the same

material was used in all the vesicles studied, and a comparison was possible since

it was assumed that the impurities affect all types of vesicles in a similar manner.

Also the assayswere easy to perform and gave consistent values between

different batches of vesicles prepared. Ideally the purified CF should be checked

by high pressure liquid chromatography (HPLQ or by thin layer. chromatography

(TLC) for other products.

To determine the interference by Triton X-100, propan-l-ol or the vesicles on the

fluorescenceof CF, various concentrationsof each (0.5%-5.0%) were added to CF

solutions. The mixtures were incubated for 90 minutes at 370C and monitored

70
spectrofluorometrically. The detergent, alcohol or vesicles had no adverse effect

on the fluorescence of CF.

Buf fer

The rates of leakage from vesicles is dependent on the method of production, the

nature of the entrapped substance, external factors such as, temperature and p1l,

osmotic pressure (Baillie fLxl, 1985) and their composition. Figure 9 illustrates the

different rates of leakage from I50: CIIOL50 vesicles (a representative vesicle

preparation) produced by the ether injection (E. I. ) technique and the hand-shaken

(H. S.) method. The leakage rate was much higher with the latter method. In both

cases the entrapped substance was CF (0.2M, pH 7.4) and the external medium was
2 0
PBS (Sorensens,1.3 X 10- M, pH 7.4,22 Q. Leakage was measured as described in

the experimental section, page 58.

The inclusion of cholesterol in membranes produces restriction of molecular

motion in the region of the first 8 to 10 carbon atoms of the acyl chain from the

lipid-water interface, leaving the remainder of the chains relatively free. This

effect is accompanied by a condensation of the area per molecule, a more

perpendicular orientation and a thicker membrane (Hsia CW, 1971). Cholesterol

suppresses the gel-to-liquid crystal transition in liposomes, below, the Tc it acts

as a fluidising agent, while above the Tc it serves as a condensing agent (Demel

and De Kruyff, 1976). In yj= studies have shown that incorporation of cholesterol

into liposomal phospholipid bilayers reduces the permeability of a variety of

substances through the bilayers (Inoue, 1974) and reduces phospholipid loss to high

density lipoproteins (Kirby CUL, 1980b).

71
 CF Release from niosomes, 150:CH
groduced by her
et. injection and hand-
shaken methods at 22oC,

100

80

60
R
E
L 40
E
20

0
05 20 25
10 15
TIME (hours)
FIGURE 9
9 ETHER INJECTION * HAND SHAKEN

N, B, In the above graph the error bars

are standard error of mean (+),

"I71
Structural studies on phosphatidylcholine-cholesterol small unilamellar vesicles

suggest that the optimum percentage inclusion of cholesterol is 32 mol% for

vesicle formation. Beyond this, the vesicle is no longer spherical possessing a

small ill-defined asymmetry (Newman and Huang, 1975). However vesicle structures

are produced incorporating up to 50 mol% cholesterol into niosomes and it has

been reported (Vanlerberghe fLd]978) that 50 mol% cholesterol is required to

abolish the transition isotherm. Vesicle formation from mono-alkyl non-ionic

surfactant I, is an apparently novel observation but can be predicted on

theoretical grounds. It is probable that molecular interactions of the surfactants

with cholesterol are responsible for many physical characteristics of the resultant

vesicles.

Niosomes were produced containing various percentages of cholesterol (0-50 mot%)

and their entrapment values measured, as in figure 10. For all surfactants, IJI

and III, optimal CF entrapment was found with the inclusion of 40 mot%

cholesterol. Addition of DCP (10 mot%) significantly reduced entrapment values.

The results indicate, for surfactant I niosomes, that leakage of CF is also greatly

reduced with inclusion of 30-40 mot% cholesterol, as in figure 11.

Vesicles containing 100% surfactant have the highest entrapment efficiencies for

niosomes but are the least stable and readily release their contents with time.

Vesicles containing 50 mol% cholesterol are more stable in terms of CF efflux and

inclusion of small amounts of DCP (10 mol%) produces a subsequent increase in

leakage of CF. Increasing the mol-fraction of a charged lipid increases the

membrane fluidity (Papahadjopoulos CW, 1973) resulting in increased membrane

permeability.

Vesicles produced using surfactant II, had a minimum requirement of cholesterol

(10 mol%) for their formation. This surfactant appeared to have several molecules

73
 Entrapment of CF a tun.cItion ol
cholesterol content (mol",11 foý
0 1,1-3 SIto[[iýýý -itký(:
(,) fIIG-) týiriI 1_,
[ý 1(1111

1-000

F
1500

1000

500

0
10
U

CHOLESTEROL ADDED

FIGURE 10

VIVýAý-ý
tho cýntr, ýPmcýnt i: )
litres per mole

Wf
 Leakage ot CF from nýosomes preL)arexl
from Surfactant I cýnnkjýnlnýj v,,iry-l-n(j
amounts ol cholesterol

60

bO

40

,30

--- 20
E
10
i
23 4 5
TýME (hniir-))
FIGURE 11
None 20% "-iddod
10,t, + J&I-(I

IýIu, I(,' ý-ji" (-I(J(" vvjý3 I lilt 1(ý(,ý 1[1 to,ýIk ý[

(PBS, pHT4 at 22oC) and the cholestero
content i3 in r-nol %

75
of water present with it. This "gummy" material could be dried, at 100"C for

several hours, to a paste. However the structural integrity of this dried product

was no longer certain. Hence it was used as provided and 100% surfactant vesicles

could not be prepared.

Vesicles in buffer (PBS)leak their contents gradually with time and as stated

earlier, page 71, the stability increaseswith the inclusion of cholesterol and
decreaseswith the inclusion of DCP for every surfactant investigated, as in figure

12a,b,and c. Vesicles containing surfactants I and III did not differ significantly

from each other when challenged with buffer (pH 7.4,370C) as seen in figure 12,a

and c, over 24 hours. Surfactant II, however, produced vesicles which were more

resistant to leakage, figure 12,b. Inclusion of DCP (10 mol%) produced no

significant change in the rates of leakage of CF from vesicles produced from

surfactant H (1160:CHOL30: DCPIO). However, this could only be compared to

vesicles with cholesterol (IISO:CIIOL50). After 24 hours incubation in buffer, the

CF from within all niosomes effluxed to a greater extent (approximately two

times) than liposomes, the latter significantly more stable (50%) in buffer (figure

12,a,b, c).

No significant difference was found in the leakage from niosomes encapsulating

"buffered" or "non-bufferd" CF (figure 13, page 78) suggesting that buffered CF

solution is not essential for entrapment. Figure 13 further demonstrates that

incorporating 50 mol% cholesterol into the bilayer of vesicles, reduces leakage.

Since the fluorescence of CF is dependent on changes in temperature and pH

(Allen and Cleland, 1980), these were carefully controlled in all experiments. For

each measurement, at least 2 controls were used; a 0% standard, containing

"empty" vesicles (i. e. vesicles containing only PBS as entrapped substance)

suspended in PBS; and a 100% standard containing PBS and a sample of vesicles

76
 CF Relea,-o in Duftor(NK3, oW 4)
tor nlosomos 1100.150*CHOL50.
and IMCHOL30oMPIO

100
100

c 80 80
F
f3o flo -
R
0 40- - 40
1 tt
a
a 20- -

05 10 15 20 ý11111 o5 10 1 '.10 . 1,
15

TIME (hours) TIMF (tymins)

FIGURE 12* FIGURL I"

1100 lhocW)160
1130(A iot 10 DGPIn )PPCbO (AKA M)

llowj, ofor to t, ut for 0011110.) , I.., ý,.,,, ýh., h, t- I f, ", !"""1.1 ,ýI"
on sur factfinlo. cýornposl lion tind 'm !.., low, I'llWi, ,, "llp, r, i l1k)" wlý I
cond III ons (: Ofl, jl florls

CF in 1--ýtif
foi V. 4)
CHOL50 'ind
will) 111100.11150.
CHOL30; DCP10
11160;

A
100

80

60
I
40- -

20

0 Aý
0 5 10 15 20
TIMF (hot-jr,,-,)
FIGURE 12c
100
1() IWO
i
mi

ý" ():; ý)fIIfIý)f(ýý), IIý1tIý)rItIýIý, ý)(1( 111

addod tor comparison,

77
 ýrom niosomo,,,-),,Iloo , 1ýI(j
150:CHOL50 (,-,nntq-ýnln(---j"buffered"(R)
iiwl "normal"M (A(, i I 1ý1
()ý 1.
,, , )

0/
/0

c 80---

F
60-
E
L 40 --

E
A "o T

0
10 15 fl

TIME (hours)

FIGURE 13
1
100(N)
110,ý(Plý 11

1\ýu1(j tIIiI[ýI I
(j ,jI ýý ýIýý 1(j dI1 -1 lk, u
between the tw(-) pýt-ýpiit
rrespectýve, of th(, enclosed marker
completely disrupted using propan-l-ol or Triton X-100, resulting in a suitable

dilution of CF and maximum fluorescence for these conditions. Low levels of

fluorescence can be detected by increasing the sensitivity of the fluorimeter.

Readings at this very high sensitivity settings were discarded.

pH Effects

CF is a trivalent anion at neutral pH which becomes electrically neutral at acidic

pH; pKa's at 6.7,4.4 and 3.5 (Heiple and Taylor, 1982). Calcein is more strongly

charged than CF as a result of two methyliminodiacetic acid residues; carboxyl

pKa <4.0; methylimino pKa 10-12 (Wallach and Steck,1963). Calcein has been

reported to be more resistant to changes in pH especially over the range pH 6 to

8 (see introduction, page 26). Niosomes, ISO;CHOLSO, were produced entrapping 200

mM CF or 200 mM calcein and their efflux as a function of external p1l shown in

figure 14,a,b, c and d. Little significant difference in the profile of leakage

between these two markers was observed over the pH range 2.0-8.0. Therefore, CF

was considered to be an adequate marker for these experiments. In addition, other

experiments with various cell systems had been established in our laboratories

using CF as an intravesicular marker. Hence CF was used in all subsequent

experiments. Studies on the diffusion characteristics of weak acid molecules, like

CF, has implications for drug encapsulation and delivery in biological systems, for

example, methotrexate, used in chemotherapy, is a weak acid and is expected to

possess similar properties in terms of leakage at various pH values.

The effect of pH on CF leakage from 100% surfactant vesicles of all types is

shown in figure 15,a,b, c and d. For all vesicles examined, CF efflux was most

rapid at pH 4.0 and below. Rapid efflux at low pH (4.0 and below) is probably a

79
 Release from niosomes at oH 2.0 CF if i Calcein f, , ýl,- ----
-t -d
containinq CF and calcein as entracoed L)reýdwj tfuln ý3urldutýtlltI wilh jd(jud,
marker ý3, cholesterol at a oH of 4.0

ion % 100

m m
A ao A 80 -
R R
K K
60 E 60-- I
R
40-
40--
R

20--
ý'o

F C)--
0 10 15 20 25
10
JIM[ (hk)lv)
ý---FIGURE
14&
FIGURE 14b

ý, w;, j, l ý. r, 150: CHOL60 ifid ttm

I!. I ii
tomoof It uIo kopt at ýýý )' I ho fl'o! m(rm!; ii

contained cholestwol It Y) 11101%

Release, marker tr(jrn niosurnes Release

)f
. CE
and Calcein it L'Qnr, ýM of-, CF Calcoin
entrapoed woýi -.
jt -i DH ý)f §-0

'ý 100 loo

ý4 m
A 80 A Bo
R R
K K
E (30 E 60
R R

40- R 40 I

o 20

o
0
20 25 5 10
5 10 15 0
TIML (t)()urs)

FIGURE 14c FIGURE f4d

'''I'll !WA1,1 , ý, ý, 0. t, ýýIII, IIýI -, 160:CHOk .50 -. .ý-;, ý,

tt lo f 11wý L)fllf)s j nd tf tj (f) flipw ; it ul 4) )1 t, v th, ' hojw, lw, ýl
the experiment maintained at 370C.

IRO
 CF Release from niosorTies,110 LJ Jý"' lrýýIll jlQQ
and 111100at DH 2.0 arld 111100--_ýý_
gti_4,Q

1 ()() YýVVV -P qo---

100 ý-

80 80

F
(30 (30
R
40
e
1 40
e e
a a
3 20 9 20
e

-1 o -1 -- -II
05 10 15 20 25 0 10 16 2F
TIME (hours) IIMI (hollf-: )

FIGURE 16a FIOURE 16b

ln(l

I hofo mv, lotýll fulk'I'lo : 11 1111',

low pH from both swi la(Aiint I iwd
III vesicle typos

CF Relcaý3o from niosome,, 3. (21(-%lýýI.' If-, [II (-II(-'--;(-)[II(ýI

-Iý
1100 arid 111100arid fro 110Q111IJ
1111QQt[I(J
Ilosomes. DPPC100 at DH 6. . .

100
100

80 80

60 60

40 40

A
S 20 20
E

0 -1
0 5 10 16 20 25 () V)

TIME (hours)

FIGURE 16c FIGURF 15d

110() 1 lill()() + 111"ll''l,

III, ), Iý), )-ýý)fTiýII[)rI)[)IIIIIý )I II"! ý,, I,

'I
stn bio at t hi s pl iw hon LýO(T)[);Ir (id k)
the two niosomo preparations, addod lof cornonrison

81
result of protonation of the carboxyl moiety of CF (at high H" concentration)

which enhances CF diffusion across the bilayer to the external media. However,

inclusion of cholesterol in the bilayer appeared to confer a stabilising effect on

the vesicle at pH 4.0 and 6.0, as in figure 16,a and b, and inclusion of DCP

increased CF efflux, as in figure 16,c and d. At pH values between 6 and 8,

vesicles made from the dialkyl surfactant II showed a greater latency of CF than

those prepared with the monoalkyl surfactant III which in turn displayed a higher

retention than vesicles composed of the other single chain surfactant I, as in

figure 17, page 84 (see region pH 6.0-8.0). This presumably is a consequence of

membrane structure and the dialkyl surfactant II molecule might be expected to

pack more tightly resulting in an increased barrier to CF efflux. Since 100%

surfactant II vesicles could not be prepared, this comparison was only made with

vesicles containing 50 mol% cholesterol, as shown in figure 17 legend, page 84.

However, the highest latency for all vesicles studied was found near pH 7A.

minima of all graphs in figure 17. The pH of the CIF marker at the beginning of

the experiment was measured and set at 7.4 and it is envisaged that when the pH

of the external media is at 7.4, an equilibrium exists reducing loss of CIF from

within the vesicles. The CIF exists as the anion at this pH and charged particles

are most probably retained within the vesicle by weak bonding interaction with

the membrane.

To study the time course of the pH effect, small known volumes of vesicle

suspension were incubated in an appropriate buffer for 1,6 and 24 hours. As

expected the leakage at the lower pH's, 2.0-4.0, was much greater (maxima % CF

release in figure 18,a,b and c, page 85) than that at pH 6.0-8.0 for all the

incubation periods investigated (the minima in figure 18,a,b and c). After 6 hours

incubation, figure 18,b, leakage at all pH's was greater (approximately 30%, as in

figure 18,a versus 18,b) and this was more evident after 24 hours, as in

82
 OF Release in niosomes from surfactant
1,11 and III wlth added cholestero
(50 mol%) at 12H 4.0
,

1001 lu()
%
80--
C 80
F
60, 60--
R
40 - 40--

20

0-1

5 10 15 20 25 10 16 20 26

TIMF (hours) IIMI (hollfýý)

FIGURE ?#a FIGURE Itib

1,)o A i0l ý,O

C.holostorol stabilisos these vOslcl(), 3 At pli 60 tho vWild(I AM ()": If lý

Irom the effect of pH as explained all loss loakiof than fit pll 4
in the text, as montioriod In tho loxt

CF Rel(,, 3,,3(,, frQrTl rliOSOMC with F- i ý( jIT

jriýj Yl1with
1,11 d r or ii (i ýý-lj
--Adde
DCP at DH 4. At pH 6. ()
gddQd_DCP-_,

100
100

80 8n
c, c
F
(30 MO

40-- 40

20-- 20--

0
15 20 25 0 5 10 15 20
05 10
TIME (hours)

FIGURE 16c FIGURE idd

160: CHOL30: DCP10 k 1101 ')(' 1), P10
MOO: CWýl 50DCPlo

It)(It41( 11j)(if: 111111 11

ItiirtIk)iIf)IIiIýý)Iýrý

wrm sot at ý3ýo(, and tho vW;l(; IW3 fild

10 mol% DGP added to them, loss loakior at pH 6.0 than pH 40

83
 CF Release froin niosomes x-)d liposoMe$
over a pH range 2.0 to 8.0 fo[ j- vw iutý
ofvo -atlon,,-) ifter 24

100-

80--

60---

40

20

I
0--
1 4 10
0
pH RANGE

FIGURE 17
lb&CHOL5C L1k1L

IuIII (-Jt t ý--)

Iu \/VI
-I
comparison of vesicles with cholesto
is shown as a typical example.

8+
 CF Reloase from niosorne orai)aro Cj- Holoaa) Irom
from surfactant I comoared wit and lioosomo. DPPC5QiQtlQLN-. LL!tý.)i 0 hr
hoosome after I hour over gH 2-8 incubation uvur_"ti_raMg Qf 2-8

100 100

80-- 80

ao-- mo -
R
40-- 40--

20--

0
0 468 10 024
pH PAN(', I- pýl HAN(

FIGURE Ift FIGURE 10b

1 ýj W; Ild 11), 'o, llý It IM It', I

,II-, I )ITTW; NI'll 1)
I ikago kI no II cs oI on tr it ppiw (TIH I if Of II tI
Ba shown SU ILII)IU IITl)IO iis IJI1, *II

CF Reloriý,,(! from niosornes and lic)osomE,,

after 24 hour n(ýubation rit j frw(j(-ý
d
,. DH 2-8

100

80

60
R
E
L 40
E I
A
20
s
E
0 I
10
PI i RAr,,1(')E

FiGURE 18c

,,. -:, r

cholesterol waýi compared to liposomes

857
figure 18,c. Liposomes follow a similar trend but are more stable, as seen in

figure 18.a,b and c, page 85. These above experiments were all performed at 370C

and were measured at appropriate time intervals together with a control for CF

to account for any changes in the fluorescence characteristics. The retention of

entrapped CF, for all compositions investigated, were significantly less for

niosomes than for liposomes.

Temperature

The effects of various temperatures, 40C, 220C, 370C and 500 were investigated

using CF as the entrapped marker at pH 7.4 in eqi-osmolar PBS. All vesicles

examined showed a similar trend; that is, increased CF leakage with increased

temperature. In all cases although leakage at 40C and 220C (room temperature)
0C
were similar, a significant increase was apparent at 37 and SOOCand this was

most pronounced at lower incubation periods, for example, after I hour, as in

figure 19,a. Addition of cholesterol (50 mol%) significantly decreased temperature

induced leakage for both surfactant I and III vesicles, as in figure 19,b. This was

most apparent around the transition temperature for the surfactants, in agreement

with the findings of Vanlerberghe tW (1978). After 6 hours incubation or

greater, the leakage profiles do not correlate with temperature and are erratic,

irrespective of the composition (50 mol% cholesterol or 10 mol% DCP) of the

vesicles, as in figure 20,a and b, page 88.

In a clinical environment, it may be necessary to store suspensions of vesicles

over periods of time, for example, on shelves at room temperature (220C). It is

advantageous to determine the stability of these suspensions under these condi-

tions. It is also important to estimate the leakage, before administering these

suspensions in vivo. In some cases the composition of the vesicles may require

modification to optimise latency of the entrapped drug.

86
 100

80

(30

40

20

----4- -- I- -I--IiI
0 10 20 30 40 50

TEMPERATURF (o(, )
FIGURE We

mw; ornw; 11111)() filw"MY)W,

mosomes, 1100 and 111100

r(-)fTIrii L) iQ, ýjf LQr J_ hQU_r

incubation at vafiouý--te-mperature-v,
containing cholesterol

50

40

30
x
-

L 20
E
A
s 10
E

0 10 20 30 40 50 (30
TEMP[HAWN

FIGURE f9b

included Cholesterol 50 mol%

87
 CF Reloaso from niosomes contairun
at various temDerature
chnlf2ý-tQrol
houra incubation in L)uffor

506--

c 40--
F
30--
R
E
L 20
E
A 10--
s
E
0. iiiiiA
0 10 20 30 40 50 f3o
TEMPERATURE (OC)
1
FIOURE 2om

! '0 ( W)l "0 .II,., ý(I 1()l 1")

11"m ("Hol !, 0 . f)Pl'i ,, o (11( 11 1-

1 [141 (: 11ý)Ii '; hlf ý)l , 111w I 'N ,i,

")o M()I% if) ill th(I rliwýomw; wiwamd
and the butler was PBS.

CF Release from niosomes an

liposomes after 24 hQurs at
various te gorAt-urea-

40--
c
F
30--
R
E
L 20--
E
10 --
s
E

0 10 20 30 40 50
TEMPFRATURE (oC)

FIGURE 206
150 CHOI-50 ll')O('HOL! )()
W., I It-[ "t) M'i 1-

1ý, ý). Iý,,. ý, f, I 14 1, IIfI

ýI,; I, ýý.,, ),,.

oultof (PW; ) arl(j lp')""ofrm was

added for comparlson of stability,

88
To optimise the potential of niosomes as drug carriers, it is important to

correlate their stability in terms of release of their contents, in vitro, with

physiological conditions in vivo. Dye release was non-linear with time, when

plotted on Cartesian coordinates, figure 21,a,b, c and d, displaying a substantial

rapid component occuring during the first 30-60 minutes of incu- bation. This

initial high leakage rate induced by plasma may be due to disturbances in the

bilayers and breakdown of the least stable vesicles. Subsequent recovery of the

barrier function may result from formation of a protective outer coat by plasma

protein bound to the outermost membrane (see later, page 98). Similar reports on

liposomes (Guo CLXL,1980) suggest a rapid efflux of CF in the presence of plasma

for the first 30 minutes, until a plateau is reached. The leakage of CF from

vesicles challenged by buffer, 100% human plasma, serum and heat- inactivated

serum are compared in figure 21a,b, c and d. In these environments, niosomes,

150:CHOL50 and liposomes, DPPC50:CIIOL50 (suitable representatives), are

distinctly different.

When challenged with 100% human plasma, there was no change in the leakage

profile after 24 hours for niosomes, compared with buffer, although, the liposomes

had more than doubled their efflux of CF, as in figure 21,a and b. This increase

occurred during the first 30 to 60 minutes following incubation, attaining a

plateau for the remaining 23 hours measured, as in figure 21, b. The profile seen

with other niosomes followed a similar pattern, except those prepared from sur-

factant 11, see later page 95-96, when challenged with plasma, loosing

approximately 50% of their contents during the first hour of exposure, then

remaining stable against further leakage, as in figure 21, b.

89
 ,F Re! e3, ýe nio3ome3 and liposome niosomes and liD-Qaome
in Buffer (PBS) a, 37oC SLrla, inruhated in 100% Dlasma at 37oC
--t3-t
sa*, c-cal enresentati, e

100
'00 T
80
80

60
60

E 40
L 4 C)
E
A 20
S 2'
F

0 05 10 15 20 25
5 10 15 20 25 TIME (hours)
TIVE
FIGURE 21b
FIGURE 21s

tý. an t)quivaient liposome

Pee3ý,, -: niosome. 150: CHOL50 CF Release from niosomes and IiDosomes

anc liDoSome. DPPC, 50: CHOL50 3' 34ýCýr -, r :r heat-inactivate
blood serum serum at 3, oC

100

80

4
- --------------------------- 60
A F:
E
L 40-- 40
L
E E
A A
20-'- 20
S
EI E

05 10 15 20 25 05 10 15 20 2
TIME ýrcurs) TIME (ýiours)

FIGURE ýlc FIGURE 21d

150:CHOL50
OPPC50; CHOL50
cono, t, ons to iiposomes 'Of ! '(31 'C',

90
When challenged with serum (figure 21,c and d), a different response was

observed. Liposomes, DPPC50:CHOL50 displayed similar leakage kinetics to that

in buffer, as in figure 21,a, when challenged with both serum and heat- inactivated

serum and no increase in CF efflux was detected. This indicates that, for

liposomes, the major destabilising factor(s) is present in plasma (100%), as in

figure 21,b. With niosomes, 150:CHOL50, a similar CF leakage profile to that in

figure 21,b, was observed in serum; that is, an initial high burst of CF efflux

followed by a steadier lower rate. However, even after 24 hours, not all the CF

within these niosomes had migrated to the extra-vesicular bulk volume (plateau

region of figure 21, b, c and d), suggesting that the proteins in serum and plasma

had stabilised the membrane in some manner. Leakage due to the changes in

osmotic pressure can be ruled out because the osmotic pressure of plasma (average

value, 6.62 atm. equivalent to 5030 mm Hg at OOC,Documenta Geigy, 1962) is

always much higher than that of CF at the working concentration.

in the absence of plasma or serum components, the enclosed dye escaped slower,

hours instead of minutes, as in figure 21,a, in buffer, from these vesicles. This

slow release was both temperature and pH-dependant. Plasma and some of it's

constituents induced a more rapid release over the 30 to 60 minutes followed by a

more stable phase over a period of hours (up to 24 hours measured). For

liposomes, it has been suggested this initial high leakage of CF may be a result

of structural changes (Guo CW, 1980). It has also been speculated that the initial

burst of leakage and the subsequent recovery of the barrier function induced by

plasma may be due to disturbances of the bilayer followed by rearrangement of

membrane components (Yoshioka, 1984). However, details of this mechanism have

not been clarified.

Many reports in the literature (Damen CLaL.1982) suggest that the removal of

liposomal phospholipid with the lipids of high-density lipoprotein (HDL) causes the

91
release of the solute entrapped. Egg PC liposomes were shown to be transformed

into smaller particles, similar to HDL, a short time (5 minutes) after intravenous

administration into rats or after 5 minutes incubation with rat plasma (Krupp

CLal,1976). Similar results were reported (Scherphof CLRI,1978) after incubation of

liposomes with plasma HDL. The appearance of discoidal particles was observed

(Tall and Small, 1977) after incubation of DMPC liposomes with human HDL.

Purified bovine apolipoprotein A-1 (Apo-1) also forms "small" complexes with

DMPC liposomes containing up to 33 mol% cholesterol (Jonas and Krajnovich, 1978).

Serum apolipoproteins were found to be most potent liposome-disrupting agents of

serum in one study (Guo CW, 1980). Free apolipoproteins exist in native serum,

but the amounts are thought to be too small to account for the observed serum

activity. However, Apo-I is loosely associated with human HDL and may dissociate

in vitro under several mild conditions (Tall and Small. 1977). Reports on albumin

binding to liposomes (Zborowski daL, 1977,Law CW, 1986) may be explained by the

presence of small amounts of contaminating lipoproteins or apolipoproteins. Some

commercial preparations of BSA have been found to contain substantial amounts of

HDL that appear as lamellar particles in electron microscopy techniques

(Hamilton, 1978). Different commercial preparations of BSA may also contain

variable amounts of contaminants that induce the release of CF from such

vesicles.

Serum is produced from plasma by inducing a clot and removing it. The major

constituent removed from serum is fibrinogen, a protein found in blood and tissue

extracts which in the presence of thrombin is transformed into the insoluble

product, fibrin.

Thromboplas in + Ca2+
i
Prothrombin
...... ...........>Thrombin
Fibrinogen
.............. .......... >Fibrin

92
Several clotting factors, for example, factor VIII, are also removed in this

process. The results obtained using serum and heat- inactivated serum indicate that

the "complement system" has no influence on the leakage of CF from niosomes,

similar to results cited for liposomes (Mayhew fUL, 1980). However, heating serum

also inactivates a protein which appears to act in conjunction with high density

lipoprotein in transferring or exchanging phospholipids from liposomes (Damen

fLaL, 1981).

In addition to plasma lipoproteins, other plasma proteins are absorbed to lipo-

somes surfaces and may be involved in their destabilisation and removal from the

circulation (Juliano and Lin. 1980). Published data for liposomes (Guo fLd, 1980)

suggest that there is a rapid transfer of phospholipid from the vesicles to high

density lipoproteins in the initial incubation period with plasma until a state of

equilibrium has been reached (plateau region, similar to that in figure 21,b,c and

d, page 90). Other workers (Yoshioka =L, 1984) however dispute this. suggesting

that the initial burst of leakage may be due to reorganisation of membrane

structure. It has been shown (Agarwal eLgL,1986) that both stability in blood and

the clearance rate from the circulation can be modulated by structurally modifying

the ester linkages in the phospholipid component of liposomes.

Evidence has also been presented(Bosworth and Hunt,1982) which suggeststhat

the effect of blood componentson liposome permeability is saturable so that with

an increasing number of liposomesin a fixed volume of blood, the fraction of

entrapped material releaseddecreases.The typical effects of a 100%solution of

human plasma on a number of different niosomes,prepared from surfactant 1,11

and III, is shown in figure 22,a,b and c. This concentration of plasma (100%)

causedno obvious (within experimental error) increasein CIF efflux from

liposomes, DPPC50:CHOL50, figure 22a, or from niosomescontaining surfactant II

(1150:CHOL50 and I160:CHOL30:DCPIO), as in figure 22.b.

93
 ('F Release from nlo8omes containin CF Roloaso jr.QLfipjQjgMa§_j,, rUjUL"
surfactant I and IlDosome whe from
challenged with Dlaema (100% Witt) [)Iaama--(IQQlil

100
100-
80-
80

60-
(30 -

40-
L 40-

A 20
E
s A
s 20 -
E
01
05 10 15 20 t
TIME (hours)
FIGURE 22m

100 .............. I IUUM 2"

1ý1 N"
h"',
only ý--, h"I
on llflý1,J, optl Iflo
a(](]E)El ChOIGS(OfOl W88 50 MUM. MIN

CF Release from niosome-s L)tuvar(rd

from SurfaclýW III
with Dlasma (100%)

100

80 I

60

40

20

0
0 5 10 15 20

TIVI- (hours)
FIGURE 22c

Hlk,
(ý oý

Liý, osumt3ý3 t)j\, o riot t)t)Orj if,

this graph for clarity. The 3(lcod
cholesterol and DCP are 50 ano lorvol%

94
Vesicles composed entirely of surfactant, that is, I100 or III100 or with

cholesterol, I50: CHOL50 or 11150:0101,50, were rapidly destabilised or disrupted

after I to 2 hours and had released almost all the entrapped CF, as in figure 22,a

and c, page 94. Inclusion of cholesterol into these niosomes confer a stabilising

effect, figure 22a; vesicles composed of surfactant III (11150:CHOL50) resist the

initial destabilisation more than those containing surfactant 1, see figure 22,c

versus 22a. However, after 24 hours, both types of vesicles show similar leakage

kinetics. Addition of DCP to niosomes destabilises them further to the effect of

plasma (100%), as in figure 22b and c.

Several reports dealing with the interaction of albumin and liposomes have been

published (Law cLXL,1986;KimeIbergJ976; Sweet and Zull, 1970). The effect of a 10%

solution of bovine albumin (BSA) was measured to test it's destabilising effects on

the vesicles. The niosomes most stable to challenge with 100% plasma were

surfactant II vesicles, as in figure 22,b, page 94. Hence niosomes containing

surfactant II was compared with liposomes, DPPC50: CIIOLSO, as in figure 23, page

96. The liposomes were three to six times more stable than these niosomes in

both BSA (10%) solution, figure 23, page 96, and in buffer (PBS). as in figure

12,b, page 77. The addition of DCP to the niosomes, 1160:CIIOL30: DCPIO, decreased

stability, shown here, figure 23, page 96, by a two-fold increase in CF leakage.

This destabilising effect, of DCP on the niosome membrane increasing CF leakage,

was also observed in buffer, as in figure 12,b, page 77. The shapes of the leakage

curves of the said vesicles obtained with BSA (10%), as seen in figure 23, are

very different to those obtained with 100% human plasma, as in figure 22, b, page

94. This could mean that some other factor(s) are responsible for the initial high

leakage rates seen in 100% plasma. The adsorption of BSA onto negatively

charged, positively charged or neutral vesicles has been reported (Law fLfil. 1986)

not to differ quantitatively. Therefore. the charge characteristics of vesicles are

not factors affecting adsorption or release of marker. Hydrophobic interaction may

95
 CF Release from niosomes
from Surfact(, -ýril 11ijiid liposomes
lncur),,ýIed with BSA 10% ()v(ýý

100

0/
/0

80

60

40

20

C
(I) 5 10 15 20
TIME (hour)
FIGURE 23
11 bO: CHOLbO )I

1)PPG")0A "I iol

DPPC50: CHOL50,
example ý()i h)i , ()mparisnn,

96
dominate. As concentration increases, more albumin molecules are adsorbed and the

conformations on the vesicle surface rearranged into a denser molecular packing

or aggregation occurs on the surface. Most of the albumin molecules can also

penetrate the bilayer and bind to the hydrophobic region of the molecules by'

hydrophobic interaction which may be the dominant force for adsorption and its

subsequent effects.

Factors Affecting Plasma-Vesicle Interaction

Plasma protein-vesicle interaction is dependent on a number of factors which all

relate to the "quality" of the lipid-water interface. A smooth, planar, homogeneous

membrane cannot readily be penetrated by proteins; whereas any irregularity, such

as the existence of phase boundries (Wilshut CW, 1976&1978) or a high radius of

curvature of the bilayer (Wilshut dAL, 1978) may effectively facilitate pene-

tration. Structural defects are also introduced into the bilayer on sonication below

the Tc of the membrane components. This results in the formation of a vesicle

with a rough surface (Blaurock and Gamble,1979). These vesicles show high

leakage rates for encapsulated solute even in buffer since the defective structure

is more permeable to solute. At the phase transition, Tc, of the bilayer, liquid-

crystals and gel phase co-exist and the resultant phase boundaries provide bilayer

surface irregularities which facilitate protein penetration. A high degree of bilayer

curvature greatly influences lipid packing in vesicles (Scherphof and Damen

f,W, 1984), so that the susceptibility of liposomal phospholipid to lipoprotein

attack depends on vesicle size. It has been reported (Scherphof and Morselt, 1984)

that larger liposomes are less susceptible to the action of plasma than smaller

liposomes. In addition, multilamellar vesicles expose much less bilayer surface area

per mole of lipid than small unilamellar vesicles with their high specific surface

area, which like the strongly curved SUV bilayer increases susceptibility to

solubilisation by lipoproteins. Cholesterol has a stabilising influence on solute

97
release. The ability of cholesterol to condense the packing of molecules in the

bilayer (Demel and DeKrufl, 1976) abolishes the gel to liquid crystalline phase

transition TC and thereby stabilises the vesicles. Protein molecules apparently can

not then easily penetrate into the hydrophobic region.

The vesiclesinvestigated in this study are all multilamellar, thus reducing the

likelihood of leakagedue to bilayer stresscausedby the small radius of

curvature. The results substantiatethe role of cholesterol which increasesthe

stability of all the vesicles.However, to adequatelyexplain the leakageprofile of

vesicleschallengedwith plasma,two theories must be addressed.Albumin alone

may be responsiblefor the leakageprofile. Initial interaction of albumin with the

vesicle bilayer causesdestabilisationallowing rapid CF efflux. It is thought that

albumin moleculespenetrateinto the hydrophobic part of the membrane,at least,

partly and attach themselvesto this hydrophobic region of the bilayer. The

remainder of the albumin molecule forms a "coating" around the outside of the

vesicle resulting in the eventual formation of a "protective layer" and a decrease

in leakage.However, although albumin and similar proteins adsorb to the outside

of the vesiclesresulting in. a decreasein the leakagerates, the initial high ef flux

of entrapped substanceis a direct consequenceof high density lipoproteins (HDL)-

amphiphile exchange.Numerousreports of this are cited in the literature for

liposomes(Guo CW, 1980;Morrisett dQL,1977) and this may also be the the case

for niosomesdue to the high structural similarities between these surfactant

moleculesand the correspondingphospholipids. However, niosomesare apparently

more leakier than liposomes.These observations require further studies to identify

the actual mechanismsof leakage,for example, albumin insertion and/or HDL

exchange, from vesiclesin the presenceof plasma. Identification of these

processeswould allow optimal formulation of specific drug carrying vehicles.

98
Ideally to establish the degree of "stability" of any vesicle preparation in blood or

plasma, both the release of the lipid, or surfactant, and that of entrapped solute

should be monitored. The availability of radio-labelled surfactant would make this

feasible and complete these stability measurements.

Polyacrylamide Gel Efectrophoresis

Polyacrylamide gel electrophoresis was used to provide more detailed information

of the adsorbed proteins on vesicles after incubation with plasma. In this study

gel electrophoresis was performed to investigate the nature of proteins bound or

associated with different types of vesicle, under specific conditions. Three factors

were studied:

a. vesicle type and adsorbed protein;

b. effect of plasma concentration on the protein coat of vesicles; and

C. times of incubation.

Quantitative measurementswere not attempted. However, in every caseequal

volumes of vesicle suspensionwere used, and, assumingthe same number of

vesicleswere present in each volume, direct comparisonsof the results could be

A
made. variety of vesicle types were incubated with plasma(10% and 100%)at

various time intervals; 1,2 and 24 hours, centrifuged at low speeds(1000g) and

compared using gel electrophoresis.The gels were stained, dried and photographed

as described (seeexperimental, page 62).

The adsorption of plasma protein to 3 different formulations of niosomes is

illustrated in figure 24. This figure shows an increase in adsorption of proteins to

all 3 types of vesicles with increasing time of incubation. Thus protein uptake by

these vesicles appears to be time dependent. Careful examination of a number of

gels indicate that a greater amount of protein is associated with vesicles

99
 1. J

oft h

(,
i

Figure 24: Niosomes, 1150:CtIOL50 (1,2,3), 11150:CIIOI, 50 (4,5,6ýand 1150:('1101,50

(7,8,9) after incubation in IOOA)plasma for I hour (1,4,7), 2 hours (2,5,8) and 2.1

hours (3,6,9). Lane 10 are marker proteins.

a. soyabean trypsin inhibitor (20,000 daltons), b. carbonic anhydrase (30,000),

c. ovalbumin (43,000), d. bovine serum albumin (67,000), e. phosphor y lase B

(94,000).

100
containing surfactant III than those containing surfactant I or surfactant 11.

Proteins may more tightly bind to surfactant III vesicles and so are not removed

in the "washing procedure" or these vesicles have greater affinity for protein =

It. The protein adsorbed in greatest abundance to vesicles has been identified a's

serum albumin (m. wt. approximately 66,000, visible as the darkest band, d, in

figure 24, in the photographs). This protein was identified from its migration

distance through the gel compared to the marker proteins. The slope and intercept

of a plot of log molecular weight versus relative mobility for 5 known calibration

proteins were used to convert gel position into molecular weight corresponding to

that position, as shown in Appendix 3 page 163.

Figure 25 compares niosomes, III50: CHOL50, with liposomes, DPPCSO:CIIOL50, after

incubation in plasma(10%) for 24 hours. The bands appear at similar locations to

those observed previously indicating that the same types of proteins are adsorbed

to both these vesicles. However, for the same quantity of vesicles and plasma

proteins, niosomes, III50: CHOL50, show darker bands (lane 3 and 4, figure 25)

than liposomes (lanes I and 2, figure 25), indicating the presence of more protein

on the gel associated with the niosomes.

The effect of incubation times in plasma (10% and 100%) is shown in figures 26,

page 103 and 27, page 104, for two separate vesicles. Both types of niosomes,

ISOXHOL50 and IIISO: CHOL50, show an increase in adsorbed proteins with time

and amount of plasma, that is, 10% and 100%. Again IIISOXHOLSO vesicles, as in

figure 27, showed a greater amount of protein adsorbed than ISOXHOLSO vesicles,

figure 26, for reasons similar to above. There was an increase, in protein

concentration and the number of different proteins adsorbed with time. For

example, there was a greater number of protein bands for both types of vesicles

studied after 24 hours incubation, lanes 3 and 6, figure 26 and 27. The highest

number of bands was recorded when 100% plasma was used. However, again

101
 Nib

__-

134.

Figure 25: Incubation in 10% plasma for 24 hours.

Lanes I and 2 liposomes, type Dill"('50: ('1101,50

Lanes 3 and 4 niosomes, type 11150:('1101,50

Lane 5 10% plasma, Lane 6 marker.

102
 wwmm* '4v»

1
2,3 77

Figure 26: Effect of plasma concentration 10% (lanes 1,2,3) and 100% (1-anes 4,5,6)

and incubation times, I hour (lanes I and 4), 2 hours (lanes 2 and 5) and 24

hours (lanes 3 and 6) on niosomes 150:0101,50. Lane 7 is the marker.

103
 At
ILS, 4mmort
*I
zo-i
too

13 1-

Figure 27: Niosomes, 11150:CIIOL50 after incubation in 10% plasina (1,2,3) an(] 10011/o

plasma (4,5,6) for I hour (1,4), 2 hours (2,5) and 24 hours (3,6). Lane 7 show's the

marker, arrow denotes position of albumin.

104
surfactant III vesicles show a greater amount of protein adsorption than

I
surfactant vesicles, as seen by the more intense bands, lanes 3 and 6, in figure

27, page 104. As the concentration of plasma increases (10% to 100%), more

molecules of protein appear to be adsorbed. It has been shown for liposomes, that

there is an increase in the adsorption of BSA with increasing concentration until

a saturated region is attained (Law fLaL 1986). This saturated level and BSA

adsorption rate was found to be identical for positive, negative and "neutral"

liposomes. Although vesicles of different charge were not investigated using

PAGE, the increased adsorption to niosomes, prepared from surfactant III was

consistent throughout this study.

More accurate comparisons of band intensities can be made using a gel densito-

meter. This procedure involves the scanning of a gel using a laser beam at

wavelengths sensitive to the blue dye, Coomassie blue R250. This produces a

trace, measuring band intensity down the gel, with increasing peaks for increasing

intensity and the area under the peaks can be correlated to the amount of protein

on the gel. Therefore after measuring a standard of known protein concentration,

other peaks, or bands, on the gel can be quantified.

A Western blot and ELISA were also performed on some vesicles after 24 hours

incubation in 100% human plasma, in an attempt to tentatively identify a second

major band appearing on the acrylamide gels. This band was thought to represent

the large subunits of IgG (molecular weight approximately 50,000) and was present

on all gels. A second faint band corresponding to a molecular weight around

25,000 was frequently detected and may correspond to the light chain subunit of

IgG. Identification of these bands was attempted using a commercially produced

antibody for IgG and the results are shown in figure 28, overleaf. Unfortunately

this product was non-specific and bound to several other proteins on the blot,

especially albumin. This was a serious problem with this IgG antibody

105
 I . 40,1
bP-

I#

lp

1 2. 34L5
i. 48

Figure 28: Results from the Western blot after ELISA for IgG wILh vesicle types;

1. I160: CHOL30: DCPIO 2.1150: CIIOL50

3. Marker

4. DPPC70: CHOL20: DCPIO 5. DPPC50: ('1101,50

6.160: CHOL30: DCP10 7.1100

8.11170: CHOL20: DCPIO 9.1115 0: C 110 1,50.

106
and the Scottish Antibody Production Unit (SAPU), the manufacturers, also

confirmed the non-specific binding (personal communication with technical

service). Therefore it was decided not to repeat this part of the work.

Attempts were also made to remove bound proteins using various agents, such as

high ionic strength solutions eg. IM NaCl, chelating agents eg. EDTA, chaotropic

agents eg. 4M urea or proteases eg. trypsin (ImM for 30 minutes at 370C). No

significant change was observed with any of these agents. However, this experi-

ment was only performed once. Juliano and Lin (1980), reported binding of high

molecular weight, (HMW), >200,000 dalton protein components, dependent on

vesicle composition (eg.inclusion of cholesterol) and on the surface charge on the

vesicle. They also observed a time dependance in protein adsorption over the first

30 minutes in both serum or plasma and this adsorption was temperature sensitive;

protein binding occurs much faster at 370C than at 40C. When challenged with

low doses of protease, such as trypsin, cleavage and subsequent removal of the

proteins from the liposomes resulted. These HMW proteins appear to be loosely

attached to the outer surface of the vesicles but other lower molecular weight

serum proteins which are more tightly associated to liposomes were not displaced

by chelating or chaotropic agents or by high ionic strength solution. The apparent

lack of activity with trypsin and the vesicles prepared in this study may be due

to the relatively small quantity of high molecular weight protein attached to these

vesicle surface, as is apparent from the fewer bands on the gels in the HMW

region, figures 24-27, see previous pages; the upper regions of the gels.

The main purpose of these experiments was to identify the protein(s) which

adsorb to these vesicles and to study the effect of composition on type and

amount of protein bound. The results show that albumin is the most abundant

protein adsorbed under all test circumstances and suggest that IgG may also be

found associated with these vesicles. Fibrinogen is also known to bind to vesicles

107
but in amounts too small to be detectedunder these conditions. More sensitive

techniques are required. The abundanceor relative concentration of fibrinogen in

plasma has been shown to have little effect on adsorption tendency (Unlyal and
Brash,1982). It has been found (Vroman dAL, 1980) that fibrinogen adsorption from

plasma is highly time dependentand that initially adsorbed fibrinogen is' replaccd
by other proteins, notably high molecularweight kininogen (IIMWK), within a few

minutes of contact. Therefore, it is essentialto specify time of contact when

discussing protein adsorption. It is also possibleto quantify these experiments in

terms of a protein and/or CF evaluationto check the amounts present before gel

application. Alternatively, an assayof the surfactants used would quantify the

sample under study. However, such a test for these surfactants has not yet been

fully investigated.

Trivivo, the rate of elimination of particles from the vascular compartment is

controlled by factors such as size and surface charge. For instance, large

liposomes are removed more rapidly than small ones. This was anticipated from

the biphasic rate of clearance of liposomes of mixed sizes (Gregoriadis and

Ryman, 1972) and later confirmed (Juliano and Stamp, 1975). With surface charge, it

appeared that negative liposomes are removed more rapidly than those which are

made to bear positive or neutral charges (Tagesson ft&1,1977). This may be

related to the net negative charge on the surface of the vesicle following contact

with plasma proteins (Black and Gregoriadis, 1976) but the mechanism by which

such a charge controls the affinity of these vesicles to cells in vivo is unclear. It

is possible that the surface charge originally present in the vesicles modulates the

extent to, or even the manner in which plasma components bind onto vesicles and

thus cause association of the latter with cell surfaces. It seems likely, therefore,

that it may be possible to manipulate the interaction of liposomes with

reticuloendothelial cells and thus influence the clearance and distribution by

imparting an appropriate protein coat to the particles prior to injection.

108
Electrophoresis

To investigate the surface potential of vesicles, their electrophoretic mobilities

were measured. These vesicles were prepared entrapping NaCl (2.0 X 10-3M, pl,

5.4). The electrophoretic mobility (p) was measured as a function of pli as

described (see experimental, page 60).

The mobility was determined by measuringthe time taken for the vesiclesto

travel a distance of 2cm on the optical measuringgrid, under the influence of a

constant electric field. The field was produced by a constant current sourceof
2mA and its magnitude was calculatedfrom the conductivity of the solution.

During preliminary work it was found that vesicle mobility was dependenton

location, therefore all electrophoreticmeasurementswere made in the centre of

the cell. To ensure reproducible resultsshort, low current passagetimes and the

reversal of current direction through the cell was maintained. Heat effects were

minimised by these low currents. Data obtained for vesicle movement in either
direction did not differ significantly. The mobilities for at least 3 different

batchesof each vesicle type were recorded, each reading, an averageof 40

measurementsat the stationary level; 20 at the front wall and 20 at the back

wall, and a standard deviation of :t0.5 pm.s-1 per V. cm-1 was obtained.

Electrophoretic mobility, p, can be related to the zeta potential, f, using the

Helmholtz-Smoluchowski equation, see introduction, page 38.

Z- 17p (41
............................
£o Er

In aqueous solutions of low ionic strengths, the physical constants for pure water

apply. Assuming zeta potential, f, equals the surface potential, Oo and substitution

109
into equation [7], page 39, allows calculation of surface charge density, a. For a

working example, consider niosomes, ISOXHOL50. from table 3, page I 11;

Cb - bulk molar concentration of the electrolyte, (NaCI). 2.0 X 10-3 Mol 1,1.

z- valency of the electrolyte - 1.

qM absolute charge on the electron - 1.6 X 10-19 Coulomb.

T- absolute temperature - 298K (250C).

k- Boltzman constant - 1.38 X 10-23 A-1.

Substituting, ý- Oo - 13.6 X 10-3 V (from table 3, page I 11) into equation [7),

from page 39, we have,

a= 11.77 Cbl sin. h. (Zqoo / UT).

am 11.77(2.OXIO-3)1 sin.h (I X 1.6XIO'19X 13.6 X 10-3)

2X1.38XIO-23 X 298

a= -* 0.14 pC. CM-2.

- the minus sign originates from the negative electrophoretic mobility).

Values for ju, ý (0o) and a at pH 5.4 are presented in table 3, page I 11. The

"neutral" vesicles, I100,150: CHOL50, II50: CIIOLSO, 11150:CIIOL50 and

DPPCSO:CHOL50, all showed a small negative charge. A plausible explanation is

the polarisation in the head groups of the surfactants wherein the oxygen atom

carries a small negative charge causes the hydrogen atom to becolne slightly

positive. At the outer surface of the vesicles, this produces a small electropositive

force which attracts negatively charged hydroxyl groups from the external media.

resulting in the overall negative charge on the vesicle. It is possible to Otitrate"

the small negative charge on the "neutral* vesicles by inclusion of a positive

charge to the bilayer. This was achieved by the addition of small amounts of

110
 TABLEI

Vesicle Twe It E"-Ol a

(,um. s-I per V. cm-1) (V)Xlo-3 (,uC. cm-2)

Iloo -13.6 -0.14

I50: CHOL50 -13.6 -0.14

160:CHOL30: DCP10 -4.6 -60.0 -0.76

168:CHOL30: SA2 +0.02 +0.3 +0.003

1150:CHOL50 -1.2 - 16.2 -0.17

lilloo - 1.2 - 16.2 -0.17

11150:CHOL50
- 1.6 -20.6 -0.22

11170:CHOL20: DCP10 -4.6 -60.0 -0.76

DPPC50: CHOL50 - 1.3 -16.1 -0.17

DPPC70: CHOL20: DCPIO -5.0 -64.1 -0.84

III
stearylamine (2 mol%) to the surfactant/cholesterol mixture. These vesicles,

168: CHOL30: SA2, had a slight positive surface charge at pH 5.4 of a- +0.003

UC.CM-2. It should, however, be noted that at pH 10.0 (high hydroxyl ion

,

concentration) the charge on these positive vesicles became slightly negative

(a _ -0.001 pC. CM-2). Inclusion of DCP (10mol%) to the vesicles,

160: CHOL30: DCP10, II170: CHOL20: DCP10 and DPPC70: CIIOL20: DCPIO. introduces a

further negative charge into the bilayer and this is demonstrated by an increase

in electrophoretic mobility and a parallel increase in surface charge, table 3, page

Ill.

A representative curve of surface charge as a function of p1l is shown in figure

29, page 113. The change in pH to the external medium (NaCl electrolyte) is

effected by addition of a few drops of acid or base respectively. The graph shows

the relationship between pH and surface charge for niosomes, 150:0101,50, a

typical representative and liposomes. DPPCSO:CIIOL50. No differences in surface

charges were observed between these 2 types of vesicles when measured with

NaCl as the entrapped solution, the solid lines in figure 29, page 113. It appears

therefore, that the surface charge on the surfactant molecules in niosomes and

phospholipids in liposomes are affected to the same extent, irrespective of the

change in pH outside the vesicles. Similar results were obtained for all other

types of surfactant vesicles measured in this study. These other graphs were not

included for brevity. Also the addition of DCP to niosomes and liposomes was

shown to increase the surface charges in the same orders of magnitude.

This trend, for the said niosomes and liposomes, was also found when the vesicles

were first incubated in plasma (50% human) for 2 hours at 370C, washed and

resuspended in NaCI electrolyte solution. However. in this case the overall order

of surface charge was lower, as seen by the dotted lines in figure 29, overleaf.

112
 Changes in Surface charge wilh-pH
for niosomes and liposomes vvh(! ýi-,
NaCl and ri human plasma (50%) -ýý

s
0.6 t
u
R
F 0.4

0.2

0o

-0.2

-0.4
024 68 1() 1")

pH RANGF
FIGURE 29

Surf, I in NaCI + NaCl

f1 In Plasma Plasma

IUItu1 Vv
niosomes and liposomes and vesi(-, Ies were
incubated in plasma for 2 hrs at 37o(7

113
To measure the effects of plasma, vesicles were prepared entrapping NaCl electro-

lyte (2. OXIO-3M) in glucose solution (0.2M). This inclusion raised the internal

osmotic pressure of the vesicles which would otherwise disrupt when challenged

with plasma. The vesicles thus maintained a constant surface area and the surface

charge remained relatively insensitive to changes in the ionic strength of the

surrounding electrolyte. The effect of plasma on ju was measured. the surface

charge was calculated and is shown in table 4, page 115.

"Neutral" vesicles types, ISOXHOLSO, IISOXIIOL50 and DPPCSO.CIIOLSO and

"negatively charged" vesicles, 160:CHOL30: DCPIO and DPPC70:CIIOL20: DCPIO, in

plasma were not significantly different from each other. This effect is a direct

result of one or more component(s) of plasma adsorbing to the surface of the

vesicle resulting in the formation of a "protein barrier" surrounding each vesicle.

Measurements were increasingly difficult at pH 5.4 or less and for vesicle types,

168:CHOL30: DCP10, IIISOXHOLSO and II170: CIIOL20: DCPIO, measurements of

electrophoretic mobility in plasma were not feasible due to their random

movement within the field of view. The change in the external p1l has a very

marked effect on the "protein treated" vesicles, as can be seen from figure 29,

page 113. At pH range 6.0-8.0, their surface potential is much lower (smaller

positive number) than that due to untreated vesicles, further demonstrating the

"coating effect" of plasma proteins on the outer surface of these vesicles. This

coat protects the polar heads from the high external OH- environment thus

reducing their effect on the overall surface charge. In fact, at p1l 4.0 and below

(high HI environment), the surface charge is positive due to the measured positive

electophoretic values. The effect of surface charge due to the changes in p1l with

vesicles treated with plasma, must be therefore a direct effect of the charge on

the protein surrounding these vesicles.

114
Vesicle type Surface Charge, a (,UC. CM-2)

NaCl Plasma

I50: CHOL50 0.14 (0.17) 0.13

160:CHOL30: DCPIO 0.76 (0.87) 0.17

1150:CHOL50 OA 0.11

DPPCSO:CHOL50 0.17 0.11

DPPC70: CHOL20: DCP10 0.84 0.13

NB. The numbers in parentheses for vesicle types, 150:CIIOL50 and

160:CHOL30: DCPIO, in NaCI solution were calculated from direct measurements

using a Malvern Zetasizer IlcR*.

These numbers are in the same order as those obtained with the manual

method, using the Rgno* microelectrophoresis apparatus.

115
It is known that albumin is the major constituent of plasma proteins (70-80%).

Albumin, typical of other proteins can be represented as below:

R ---- HN ----- (CH2)n ----- COOK

It has an isoelectric point of 5.4. At low pH, that is, high hydrogen ion

concentration, the nitrogen atom (N) on the molecule. becomes protonated and

there is no charge on the carboxyl group (COOH). However, at high p1l (low

hydrogen ion concentration) the nitrogen atom looses it's proton and the COOII

group becomes ionised (that is, acquires a negative charge). The curve (surface

charge against pH, figure 29, page 113) cuts the X-axis at around 5.4 (within

experimental limits) which is the pH at which albumin exists as a "neutral"

molecule, that is, it's isoelectric point. This suggest that the major constituent of

blood adsorbed on the vesicles is probably serum albumin. All the other vesicles

investigated show similar trends and have profiles as shown in figure 29.

3.4. STUDIES WITH TETRAHYMENA ELLIOTTI

The rate of uptake and subsequent breakdown of niosomes containing surfactants

I, II and III and liposomes by the micro-organism T. elliotti were measured. The

rate of uptake is expressed as an average number of food vacuoles formed per

T. elliotti cell in a given time period. Although there are wide variations in the

numbers of ingested vesicles in the food vacuoles between individual cells, the

results show a trend and conform to a general pattern. T. elliotti cells used in this

study remained fully motile and active in the vesicle suspensions with little or no

signs of toxicity after 24 hours. This lack of toxicity was repeatedly noted with

every preparation.

116
The results showed little variation between the rates of uptake for the niosomes

of all types and liposomes, as seen in figure 30. This was expected since L

elliotti ingests most materials at a steady rate. The vesicles were broken down

within the food vacuoles resulting in the appearance of a diffuse intracellular

fluorescence, as in figure 30. This occured more readily for liposomes and

niosomes prepared from surfactant III, that is, 111100,11150:0101,50 and

II170: CHOL20: DCP10, than for vesicles with surfactants 1, for example, 1100,

150:CHOL50 and 160:CHOL30: DCP10, or surfactant type 11, such as. 1150:0101,50

and 1160:CHOL30: DCP10. An explanation for this observation is that 3urfactant III

is more susceptible to intracellular degradation than the ether linked surfactants I

and III. The efflux of vesicular CF into the cytoplasmic space of the T. elliotti cell

may also be a function of the low pH within the food vacuole which has been

approximated as pH 4.0 to 5.0 (Nuccitelli and Deamer.1982). Alternatively,

intracellular degradation above pH 4.0 may be caused by enzymes, normally used

to digest matter within the food vacuoles. These enzymes may be responsible for

the apparent vesicle degradation observed in vivo, or the CF leakage is a product

of the two processessimultaneously.

The uptake of vesicles and eventual release of entrapped CF from vesicles was

examined using epifluorescence microscopy. The uptake and breakdown of vesicles

was recorded by delayed time sequence photography and was seen clearly, using

the self-quenching properties of CF. Carboxyfluorescein entrapped within vesicles

was self-quenched and localised inside the food vacuoles after ingestion. These

vacuoles were visible as distinct fluorescent spheres, as seen in figure 31, page

119 for liposomes and figure 32, pages 120-122 for niosomes. Released vesicular

CF was diluted throughout the T. elliotti cell and hence produced a diffuse

intracellular fluorescence which increased with time. However the breakdown of

vesicles was also dependent on composition and the most rapid rate was seen with

117
 AD[Darent u take of niosomes lm(ý liposome
L)y Telliotti. cells ovet i giv(, n
tý mp,
,, []eriod,

A 12
c
*
U lo
0
L8

6
P
L4
P
2
E
Lo
L 20 40 60 80 100

TIME (mins)

FIGURE 30

L
ý-I 1 100 Counts
of cells, The Y-axis is th,,
fluorescent vacuoles p(, ý 100 cells

its
 ,R) P

ION,

MI( )NI

Figure 31: Apparent uptake by T. elliotti cells after 3 hours incubation ill "I

suspension containing liposomes, DPPC50: CIIOI, 50, under normal light (top) and

after ultraviolet illumination (bottom).

Magnification = 20OX.

119
 TOP
CL

M)II(M

rigure -iz: t-iuorescence pnotographs of I ijIlý)III at Various lillit's (I to 0 110111%.

a-f) incubation in a suspension containing niosomes, 150:0101.50 ( Photographs *-f

under normal illumination are to be found inside the back cover of' the the.%i%.
)

Magnification = 50OX.

1210
 I OP

(.

(I

M)II(

121
 Mll

(i

, ý( )I I( Y%l

122
 TABLE5

Vesicle Tvpj Time (in hours)

01123456

I50: CHOLSO --------

160:CHOL30: DCPIO

1150:CHOL50

1160:CHOL30: DCP10 ---

II150: CHOL50

11170:CHOL20-DCP10

DPPC50: CHOL50

DPPC70: CHOL20: DCP10

+- presence of diffuse intracellular fluorescence.

m no diffuse fluorescence observed.

123
liposomes since this resulted in the appearance of diffuse intracellular

fluorescence within an hour of incubation (table 5, page 123). With niosomes

containing the ester-linked molecule, surfactant III, types IIISOXIIOLSO and

11170:CHOL20: DCP10, release of CF was apparent only after I to 2 hours and with

surfactant II vesicles, II50: CHOLSO and I160:CHOL30: DCPIO, and surfactant 1, types

ISOXHOL50 and 160:CHOL30: DCPIO, after 2 and 6 hours respectively (table 5). In

general, inclusion of cholesterol in niosomes increases the stability of the vesicles

within the food vacuole, and the resultant diffuse fluorescence within the cell

body appears much later. This process was reversed when DCP (10 mol%) was

added to the niosomes and the vesicles appeared to release their contents at a

much faster rate. These observations are consistent with studies on liposomes

containing cholesterol and added DCP (table 5). The inclusion of DCP may attract

lysosomal enzymes to the outer membrane of liposomes hence increasing

destabilisation of the entire shell. This leakage is further enhanced by the effect

of the lower pH of the vacuole (around 4.0). Previous studies (see page 71) on

vesicle behaviour in buffer at pH 4.0, have shown a steady increase in leakage of

CF with time, as shown in figure 33a,b and c, page 125. Vesicles containing

surfactant III and cholesterol, 11150:CIIOL50, exhibit the fastest leakage at p1l 4.0,

figure 33c, followed by vesicles containing surfactant I and 11 with added

cholesterol respectively, figure 33a and b. In all cases inclusion of DCP (10 mol%)

into the vesicle membrane increases the rate of CF release. The trend shown in

figure 33a,b and c, was measured after incubation of the vesicles in buffer at p1l

4.0 for various time intervals at room temperature, 220C. These results suggest

that pH is an important factor in the intracellular destabilisation of niosomes.

whether within or outside the food vacuole of the T. elliotti.

The other possible destabilising factor is the presence of lysosomal enzymes, many

proteolytic in nature within the food vacuole. Two typical enzymes were chosen

to study their effects on vesicle stability; phospholipase A2, because the natural

124
 OF Release from niosomes creoare QF Rolease from nlooQmov jpvýLaLW
from Surfactant I after from Sur fauýjnj it J1
incubation at DH 4.0 In BUFFER iaH 4.0 In BUFf LR
ý

100
80 80

öo 80
R
E
40 40
L
E
A
s 20
E
0

0 25 o
10 is 20
TIME (hours) TImf

FIGURE 33m

ý"* IPOjl(A Wt,ý-Y-

a, o
i ý"i me'l
ir, io rnL)i%
j
"W, bars are slandioro orft)t It innonri

CF Release from niosomes prepare

from Sur 0 111, if
DH 4.0 in BUFFER

100
4'
80

60

40

20

0
0 5 10 15 20
TIME (hours)
FIGURE 33c
IllbO "'Hol ý)O jjjý'()

sanio ý)roportions is st, ito(i in ! ti(,,

tex t.

125
substrate for this enzyme is phospholipid and carboxylic ester hydrolase, since

this enzyme is known to break ester bonds. Carboxylic ester hydrolase produced

leakage of CF only in vesicles, IIISOXHOL50 and 11170:CHOL20: DCPIO, that is,

containing the ester linked surfactant III, resulting in approximately 40% loss of

entrapped CF after 6 hours. Liposomes, containing egg PC, that is, Egg

PC50: CHOL50, showed a delayed release of CF after 24 hours incubation where

40% of the contents had leaked, with this same enzyme, as seen in figure 34, page

127. Increasing the concentration of the enzyme produced no change in leakage

from vesicles containing surfactant III and did not reduce the measured 6 hours

lag period for liposomes, Egg PCSOXHOL50. However a five-fold increase in the

concentration of esterase resulted in a three-fold increase in the 24 hour release

of entrapped CF from Egg PC50;CHOL50 vesicles; this latter result is not shown

in figure 34.

As expected from the work of Slotboom jW (1982), pure porcine pancreatic

phospholipase A2 produced no increase in CF leakage for any vesicle types

measured even at five-fold concentration increases and excess calcium ions. This

agrees with the already published inability of pancreatic phospholipase to attack

long chain lecithins in the absence of added detergents or organic solvents

(Slotboom etal, 1982). However, snake venom from Russell's viper provides a good

source of phospholipase A2 (Russell, 1980; this venom also contains other enzymes,

see Appendix 2, page 161) and produced CF efflux from several types of vesicles.

The results are shown in figure 35, page 128; all liposome types examined and

niosomes composed of surfactant III, for example, DPPC100, DPPCSO:CIIOLSO,

DPPC70: CHOL20: DCP10, Egg PCSO:CHOL50, DMPC50: CIIOL50,11150: CIIOL50 and

11170:CHOL20: DCP10, displayed relatively high rates of leakage. Inclusion of

cholesterol decreased leakage and addition of DCP stimulated CF efflux similar to

the vesicles tested in vivo in Tetrahvmena, as shown in Table 5, page 123. No

increase in CF leakage from I150:CHOL50 vesicles was recorded during the

126
 Ef fect of carboxylic ester hydrolase
on CT Helec,
ise, fron) niosome and
liposome UreLý it, ýIk)i i,
ý,

100

80

60
R
E 40 J
L
1
E
A 20--
s
E0
5 10
TIME

(h our, (--))
FIGURE 34

Fg RC: flHflL &O

Sur factant III v( ý:ý,icles are t he mo,-, I

sensitive to lhi, -, enzyme. The lipo3otiw
preparation is detailed in the text,

127
 Eflect of snake venom r)hos holipase
A2 c)n CF ýtuw ome and
-n-i-o5
liposome LDreL)ariý,!
ý'()[iý,

100 4-

70

80
c
F
60
R
E
40

20 4

10

TIME (h our s)

FIGURE 35

11150:CHOL50

The niosomes used were prepared týom

Surfactant 111,fhe ester type and hii, i
varying of cholesterol add(--ý,j

129
24 hours observed and a very slight CF efflux was measured in vesicles, I100 and

150:CHOL50. Structural characteristics of the vesicles may govern their

suseptibility for enzymatic breakdown. It has been reported (Op den Kamp

rLaL1975) that the co-existence of solid and liquid phases in the bilayer create

irregularities in lipid packing at the border of these domains facilitating the

penetration of phospholipase A2.

The effect of temperature on enzymatic activity was measured using vesicles,

DMPCSO:CHOL50. This phospholipid has a TC of 230C (Ladbrooke CLXL,1969) which

is convenient to investigate the effect of temperature on CF leakage above and

below this Tc. Leakage of CF from these vesicles was measured at 370C and at
0
22 C. The results are shown in figure 36, page 130 and the CF leakage 'was lower

at 220C. Menashe and co-workers (1981) found that the time course of hydrolysis

exhibited a distinct lag period when they mixed pancreatic phospholipase A2 and

small uni- lamellar vesicles made from DPPC at or above the TC. However when

the enzyme was preincubated with the substrate for a short time below the Tc

and then assayed at high temperature, no lag period was observed. These results

support the hypothesis that a relatively slow substrate -enzyme organisational step

is required as part of an activation process and this is most rapid when the

substrate is in the gel state. However, the intrinsic enzymatic activity (after the

initial step of activation) is maximal when the substrate is in the liquid-

crystalline state. Hence, the lag time may reflect a penetration step of the

enzyme into the lipid-water interface (Langer, 1980:Langer and Peppas,1981). It has

been reported (Wi1schut CLgL1979) that phospholipase A2 can degrade the entire

pool of phospholipid in the outer half of a vesicle without releasing any entrapped

CF. Liposomes, DPPCSO:CHOL50, were produced by sonication as reported,

(Menashe fLgL1981) and incubated with pancreatic phospholipase A2 and calcium

ions. Sonicated vesicles have a high radius of curvature which may produce

irregularities in the surface so aiding enzyme penetration. CF efflux was

129
 CF Release from lir)osome above
(jtid
below the Tc of the lipid (Tc=23oC)
on addlflýon of snake venom_phosgbolipase

100

80

60
R
E
40
L
E
A
20
s
E 41
0
5 10 15 20

TIME (hours)
FIGURE 36
AP(-)VF Ir

DMPC50: CHOL50,
due to the low Tc M Iho
study the effecl m phospholipase A2

130
monitored over 24 hours and no substantial increase was observed at either 220C

or at 370C.

The effect of hydrolytic enzymes and pH were studied separately in the

laboratory. The leakage was greatest at pH 4.0, which is seen typically with

surfactant I and III vesicles, as shown in figure 33,a and c. page 125. Niosomes

produced using surfactant II and liposomes were more resistant to leakage at p1l

4.0. as in figure 33,b. These results indicate that pH alone was not responsible for

the in vivo release of CF in Tetrahvmena cells since surfactant I vesicles

appeared to be relatively stable within the lysosomes of T. ellootti, diffuse

intracellular fluorescence only after some 6 hours. However, niosomes composed of

surfactants 11 and III and liposomes, released their contents within Tetrahvmena

cells in less than 2 hours. In addition, on challenge with the enzymes,

phospholipase A2 and carboxylic ester hydrolase, surfactant I vesicles were

resistant and released little CF. The diffuse intracellular fluorescence apparent

after 6 hours in T. elliotti with these vesicles is presumably mediated by the action

of other lysosomal enzymes or probably low pH. The apparent delay in leakage

may also be caused by the presence of adsorbed protein (enzymes) on the vesicle

surface which form a "protective layer". Niosomes containing surfactants 11 and III

and liposomes are *pH resistant", yet release entrapped CF rapidly after ingestion

by Tetrahvmena. However, they are susceptible to challenge with phospholipase A2

and carboxylic ester hydrolase. The greatest efflux of CF was observed from

liposomes challenged by phospholipase A2. Thus, enzymatic activity may be a

major destabilising factor for niosomes containing surfactant II oir III in vitro in

Tetrahvmena cells, similar for liposomes. It is not clear why surfactant 11

niosomes should be destabilised in vitro by phospholipase A2. Presumably it is not

an enzyme phenomenon since there are no apparent points of attack, that is, no

susceptible bonds within the molecule for the "reagent enzyme" to attack.

131
However, snake venom also contains many other enzymes, as in Appendix 2, page

161 and the observed activity may be due to the other enzymes.

Enzymatic CF release appears to be more effective than that caused by low p1l.

This demonstrates that drug release and hence distribution at the subcellular level

can be tailored by suitable formulations of niosome preparations. This ability to

finely tune the release of active ingredient at the cellular level appear to be a

definite advantage of niosomes with respect to liposomes.

132
 SECTION 4

CLUSIONS AND FUTURE

 4.1. GENERAL CONCLUSIONS

The proposed aims at the outset of this project have been successfully completed.

A study of niosomes, prepared from three different non-ionic surfactants, have

been compared to liposomes. The niosomes have been fully characterised in terms

of their composition, stability and behaviour in a variety of systems, by analogy

to the well-documented liposome delivery system. The ability of these bodies to

interact with biological fluids have been studied thoroughly and attempts were

made to quantify this interaction by studying their electrophoretic mobilities and

changes in surface charges when added to blood plasma.

These vesicles were all found to bind to proteins in the blood and using poly-

acrylamide gel electrophoresis, the nature of the proteins adsorbed to their outer

surface were studied and in every case, albumin was found to be present. This

was not surprising since albumin constitutes the major protein in blood and is

expected to make the biggest contribution, as found in studies with liposomes.

However, such studies have never before been carried out with niosomes and

these findings reported here in should aid future formulation work with these

novel carrier systems.

The rate of survival of these vesicles in a living, eukaryotic cell, such as

T. elliotti, was a new approach to evaluate their stability within an in vitro

system. This organism, which served as a simple model, has a number of lytic

enzymes within it's digestive tract (food vacuoles) and the obseryption that

niosomes were more stable than liposomes in this environment was encouraging

for future development work for specific targeting related to the present model.

By formulating the chemotherapeutic agent within the desired niosome system,

their profile of breakdown within macrophages, containing living bacteria can be

133
evaluated. This ultimately will benefit the patient undergoing therapy by reducing

toxicity- the final goal of such targeting systems.

Specific Conclusions

The size and nature of the surfactant used in the preparation of niosomes

explains their stability in a given system. Irrespective of the composition, there

were little variations in their ability to encapsulate drugs. In this project, the

water soluble dye, CF was used while colleagues within our laboratories have used
drugs such as doxorubicin, methotrexate and the antimony salt, Pentostam R*

However, the ability to release the marker were different depending on the

surfactant under investigation. The most stable vesicles, as a general rule, over a

wide pH and temperature range, were those containing surfactant IL It appears

that the greater number of methylene units (-CH2-) in surfactant 11, increases the

lipophilicity of the molecule which packs this lamellar structure more tightly than

the other two types of surfactant vesicles studied. Addition of cholesterol into

the bilayer further stabilised these vesicles, irrespective of the types of surfactant

used, while addition of DCP destabilised them, in a variety of conditions and

environments.

The overall effect of blood proteins, on every type of vesicle prepared for this

study, remained the same and this has been shown to be due to the "protective

coat" of the albumin surrounding the outer membrane. As a consequence, their

behaviour in biological fluids was governed by the nature of this. external "coat",

which in turn affects the stability of entrapped drug. This effect was further

demonstrated when vesicles prepared from these three different types of

surfactants, which varied in their individual stabilities, electrophoretic mobilities

and charge densities were shown to have little difference when dispersed in

plasma. Since only a "qualitative" study of the influence of this "protective coat"

134
was undertaken in this study a detailed and *quantitative" estimate will identify

further differences which would help to optimise these niosome preparations as

drug delivery systems. Gel electrophoresis is an excellent technique for such

quantitative studies and is certainly a consideration for future work.

It was not surprising to find that niosomes containing surfactant III, an ester-

linked amphiphile, were the least stable when challenged with carboxylic ester

hydrolase, both in vitro (experiments using purified enzyme) and in vivo (within

the food vacuole in T. elliotti cells). The relative stability of these niosomes, when

compared to liposomes, to phospholipid A2 was an interesting observation. Due to

their phospholipid content, liposomes are more susceptible to degradation and

eventual release of contents. However niosomes containing the single chain

surfactant I were more stable under these conditions than those niosomes with

surfactant II. This must be a direct consequence of the nature of the surfactant

which affects the packingwithin the bilayer membrane.

The exact choice of non-ionic surfactant together with cholesterol produce

niosomes that provide some degree of protection against release of drug content

into the desired area. Further niosomes can be separated into unilamellar and

multilamellar vesicle types containing a drug of choice within the aqueous

compartment. Subjecting these vesicles to a systematic study, on a protocol

outlined in this thesis, can accurately determine their behaviour in a chosen

model. Together with a thorough knowledge of the biochemistry of the target

cell(s), the release of contents from such vesicles can be predicte4 and the

desired effects obtained. Such studies are part of the ongoing projects within our

laboratories.

135
 4.2. FUTURE WORK

The present investigations have been satisfactory in that the systematic study of

a novel drug delivery system has produced a number of new ideas which can be

further developed.

The most promising of these is the observation that although the niosome system

is anologous in many respects to the well-characterised liposomal system. their

subtle differences in properties can be exploited. In an attempt to acheive similar

and other specific systems, a number of controlled studies are suggested:

1. Separation and analysis of multilamellar and unilamellar vesicles after

preparation, by particle size extrusion to obtain niosomes of a particular size

range and study their individual characteristics in a chosen system of interest.

2. The use of radioactive drug within the aqueous compartment or a radiolabelled

surfactant incorporated into the membrane, for example. tritium or "Carbon

are ideal, relatively inexpensive radiotracers, will allow the progress of the

leakage in vivo to be monitored more closely in a system which can be

tailored to the biological environment under study. This alleviates the need of

markers which ultimately can only serve as an approximate model and may not

be specific. A better understanding of the interactions involved in drug

release, whether it is due to distortions in the outer membrane of the vesicle

or simple drug diffusion into the surrounding media, can be ascertained

accurately.

3. Quantitative gel electrophoresis to further amplify the protein-membrane

interactions, as mentioned previously, page 108.

4. A reappraisal of studies in which the liposomal drug delivery system has been

found to be non-specific or ineffective. In particular, applications in

immunology* a very "specific" area ideally suited for "tailor made" targeting.

136
 4.3. PERPECTIVES

Practical applications of liposomes have been explored in three main areas of

research, namely in model membrane studies, in controlled and targeted drug

delivery in vivo and in transfer of genetic and other materials into cells, in

culture. Of these, the drug delivery concept, including the immunological adjuvant

potential has attracted predominant interest. Promising results have been obtained

in specific applications in experimental therapy related to the RES as in, for

example, the accumulation of toxic metals in the liver (Rahman, 1980) and in

visceral leishmaniasis (Alving. 1986). The successful use of macrophage activating

factors within liposomes directed towards alveolar macrophages for the activation

of the eradication of lung metastasesin cancer chemotherapy by utilising the

liposome carrier with cytotoxic drugs, in laboratory animals have also been

reported (Rahman fLgL1986).

The use of liposomes as a model for biological membranes is confined to the

research laboratory, but their successful application for drug delivery and/or

genetic engineering will depend on several factors, such as a demonstration of the

superiority of the liposome carrier for the intended purpose, and also upon the

technical and economic feasibility of the system in practice. One of the

fundamental questions is whether it is possible to prepare, isolate and characterise

the particular liposomal system on an industrial scale with clearly defined and

reproducible properties. Furthermore, the physical and chemical stability of such

vesicles is of vital importance.

The future objectives of drug carrier systems can be considered as less ambitious

than they were a few years ago. These are no longer used to deliver highly

specific cytostatic agents to tumour cells. The emphasis is more on improving

therapeutic treatment of certain types of disease by combining some of the

137
 characteristics of liposome delivery systems such as preferential uptake by the

RES cells, slow release effect for the entrapped drug and eventually preferential

tissue localisation. The stimulation of macrophage mediated destruction of tumour

cells and invading micro-organisms as well as the reduction of some side-effects

of drugs are also areas of major therapeutic interest. There are numerous diseases

that involve pathological modifications of cellular blood constituents, for example,

sickle cell anaemia, or require active modification of cellular competance, as in

the oxygen carrying capacity or tumouricidal properties. In these cases, the

specific and/or non-specific interactions with vesicles carrying cellular effectors

either in their membranes and/or in their aqueous space could be used to modify

the recipient blood-cell physiology. For non-phagocytic and/or non-fusogenic

target cells like the erythrocytes, hydrophobic compounds could be incorporated

into the bilayer and transformed via exchange mechanisms into the plasma

membrane. Also, non-specific cellular absorption of the carrier with subsequent

release of it's contents could provide a circulating depot and/or create a locally

high concentration of drug in question in the vicinity of it's target, the blood

cells. To name but a few possibilities: Gersonde and Nicolau (1979) have shown

that the oxygen-carrying capacity of erythrocytes can be markedly enhanced ill

MiUr upon interaction with liposomes loaded with the allosteric effector, inositol

hexaphosphate. Circulating peripheral monocytes (Fidler CW. 1981) have been

shown to internalize liposomes containing muramyl dipeptide, with subsequent

activation of their turnouricidal properties. Gorecki CW (1980) have synthesised a

series of benzyl esters of aromatic or hydrophobic amino acids that are

potentially useful therapeutic agents for the treatment of sickle call disease but

cannot as yet be transferred effectively into the sickle erythrocytes (Gorecki

CLaL,1980). The action of such drugs might be improved if they were incorporated

into vesicular carriers that act either as depot systems with "sustained" survival

in the blood stream and/or could be "homed" specifically to surface determinants

on target cells contained in the circulation.

138
The vesicles described in this dissertation do not constitute a universal drug

carrier and their real possibilities are likely to be fairly limited. However, they

show very specific advantages for certain applications.

139
REFERENCES

APPENDICES
 REFERENCES

Agarwal, K., Bali, A and Gupta, C.M. (1986) Biochim. Biophys. Acta. UL 36-40.

Allen, T. M. and Cleland, L. G. (1980) Biochim. Biophys. Acta. 418-425.

Allen, T. M. (1984) In: Liposome Technology, Vol. 111,(Gregoriadis, G., ed.)

pp. 177-182, CRC Press, Boca Raton, Florida.

Alving, C. R., Steck, E. A., Chapman, G.M. Jr., and Hanson. W.L. (1978) Proc. Nail.

Acad. Sci. USA 75,2959-2963.

Alving, C. R. (1986) In: Int. Encycl. Pharm. Ther. Section 120, Methods in Drug

Delivery (Ihler, G.M., ed.) pp.281-300.

Andersson, L. O., Rehnstrom, A and Eaker, D. L. (1971) Eur. J. Biochem. 371-

380.

Attwood, D and Florence, A. T. (1983) Surfactant Systems: Their Chemistry,

Pharmacy and Biology. Chapman and Hall, London.

Avegard, R. and Haydon, D. A. (1973) In: An Introduction to the Principles of

Surface Chemistry. pp. 40-47, Cambridge University Press, London.

Amin, M. N., Florence A. T., Handjani-Vila, R.M., Stewart, J. S.B., Vanterberghe, G

and Whittaker, J. S. (1985) J. Pharm. Pharmacol. 17,237-242.

Amin, M. N., Florence A. T., Handjani-Vila, RX, Stewart, J. S.B., Vanlerberghe. G

and Whittaker, J. S. (1986) J. Microencapsulation. J, 95-100.

140
Baler, R. E and Dutton, R. C. (1969) J. Biomed. Maier. Res. 191-206.

Baillie, A. J., Florence, A. T.. Hume, L. R., Muirhead, G.T. and Rogerson, A. (1984) J.

Pharm. Pharmacol. Supp. 48P.

Baillie, A. J., Florence, A. T., Hume, L. R.. Muirhead, G-T and Rogerson, A. (1995) J.

Pharm. Pharmacol.17,863-868.

BaIllie, A. J.. Coombs, G.H., Dolan, T. F and Laurie, J. (1986) J. Pharm. Pharmacol.

M, 502-505.
.

Bakker- Woudenberg, I. A., Lokerse. A. F., Roerdink, F. H., Regis, D and Afichel, Af.F.

(1985) J. Infect. Dis. 151,917-924.

Bangham, A. D., Standish, M. M. and Watkins, J. C. (1965) J. Mol. Biol. LI, 238-252.

Barenhoiz, Y., Amselem, S and Lichtenberg, D. (1979) FEBS Lett. 22,210-214.

Batzri, S. and Korn, LD. (1973) Biochim. Biophys. Acta. 2M. 1015-1020.

Black, C. D. V. and Gregorladis, G. (1976) Biochem. Soc. Trans. 1,253-256.

Blaurock, A. E. and Gamble, R. C. (1979) J. Membr. Biol., IQ, 187-192.

Blum, R. H. and Carter, S. K. (1979) Ann. Intern. Med. N, 249-259.

Blumenthal, R., Weinstein, J. N., Sharrow. S.O. and Henkart, P. (1977) Proc. NatI.

Acad. Sci. USA 74,5603-5607.

141
Bosworth, M. E. and Hunt, C. A. (1992) J. Pharm. Sci.. 7L 100-104.

Brash, J. L., Szola, P and Thibodeau, J. A. (1984) Trans. Soc. Diomat. Z. 253-256.

Brown, P.J and Juliano, R. L. (1985) Science. M, 1448-1451.

Caride, V.J. and Sostman, H. D. (1994) In: Liposome Technology, Vol.11 (Gregoriadis.

G., ed.) pp. 107-124. CRC Press, Boca Raton.

Chen, R. F. (1977) Anal. Lett. I.Q. 787-795.

Chen, W.C. and Bittman, R. (1977) Biochemistry, Lk, 4145-4149.

Chien, Y. W. (1980) In: Drug delivery systems: Characteristics and biomedical

applications. (Juliano, R. L., ed.) pp. 11-83, New York, Oxford. Oxford University

Press.

Clarke, H. G. M and Freeman, T. (1969) Clin. Sci. 403-413.

Dahl, G., Azarnia. R. and Werner, R. (1981) Nature M, 683-685.

Damen, J., Regts, J and Scherphof, G. (1981) Biochim. Biophys. Acta., MI, 538-545

Damen, J., Regts, J and Scherphof. G. (1982) Biochim. Biophys. Acta., ZU, 444-452.

Darke, A., Finer, E. G., Flook, A. G and Phillips, M. C. (1971) FEBS Lett. LI. 326-

330.

142
De Kruyff, B., van Dijck, P. W. M.. Demel, R. A.,Schuijff, A., Brants, F and van

Deenen, L. L. M. (1974) Siochim. Biophys. Acta., IM 1-7.

De Haas, G. H., Postema, N. M., Nieuwenhuizen. W. and van Deenen, L. Af.M. (1968)

Biochim. Biophys. Acta., LIL 103-117.

Deamer, D. and Bangham, A. D. (1976) Biochim. Biophys. Acla., 629-634.

Deamer, D. (1984) In : Liposome Technology. Vol. 1 (Gregoriadis, G., ed. ) pp. 29-35.

CRC Press, Boca Raton, Florida.

Deliconstantinos, G., Gregoriadis, G., Abel, G., Jones, M. and Robertson, D. (1977)

Biochem. Soc. Trans. 1,1326-1328.

DeLoach, J. R. (1986) Med. Res. Rev. ff 4), 487-504.

Demel, R. A. and De Kruyff, B. (1976) Biochim. Biophys. Acla., JR. 109-132.

Desiderio, J. V. and Campbell, S.G. (1983a) J. Reliculoendothel. Soc. U. 279-287.

Desiderlo, J. V. and Campbell, S.G. (1983b) J. Infect. Dis. JýM, 563-570.

Diehl, H. and Ellingboe, J. L. (1956) Anal. Chem. 21,882-895.

Dingle, J. T. (1972) Lysosomes. North Holland Publishing Co., Amsterdam.

Documenta Geigy. (1962) p. 548.

143
Edwards, D. I. (1980) Antimicrobial Drug Action, University Park Press, Baltimore,

USA.

Ehrlich, P. (1906) In : Collected Studies on Immunity Vol 2. pp. 442-447., Wiley,

New York.

Elliott, A. M. (1965) Science L19, 640-641.

Elliott, A. M. and Clemmons, G. L. (1966) J. Protozool. 311-314.

Fendler, J. H and Tundo, P. (1984) Acc.Chem Res. LZ, 3-8.

Fidler, I. J., Sone, S., Fogler, W.E. and Barnes, Z. L. (1981) Proc. Nall. Acad. Sci.

USA 78,1680-1684.

Fildes, F. J. T. (1981) In: Liposomes: from Physical Structure to Therapeutic

Applications (Knight, C.G., ed.) pp-465-485. Elsevierl North Holland Biomedical

Press. Amsterdam.

Forssen, E. A. and Tokes, Z. A. (1981) Proc. Nall. Acad. Sci. USA ZL 1873-1877.

Fountain, M. W., Weiss, S.J., Fountain, A. G., Shen, A and Lenk, R. P. (1985) J.

Infect. Dis. 1,5L 529-535.

Freise, J., Muller, W., Broelsch, CH and Magersteds, P (1981) Hepato-

gastroenterol. 28,90-92.

Frokjaer, S., Hjorth. E. L. and Worts, 0. (1982) In: Optimisation of drug delivery

144
 (Bundgaard, H., Hansen, A. B. and Kofod. H., eds.) pp. 384-408, Munksgaard,

Copenhagen.

Fukuzawa, K., Ikeno, H., Tokumura, A and Tsukalani, H. (1979) Chem. Phys. Lipids

21,13-22.

Gabizon, A., Dagan, A., Goren, D., Barenholz, Y and Fuks, Z. (1982) Cancer Res.

ýLZ 4734-4739.

Gabizon, A., Meshover, A and Barenholz, Y. (1986) J Nall. Cancer Inst. ZZ, 459-

469.

Gaffney, B. J. and Mich, R.J. (1976) J. Am. Chem. Soc. 21,3044-3045.

Gersonde, K. and Nicolau, C. (1979) Blut. 19,1-7.

Gorecki, M., Votano, J. R. and Rich, A. (1980) Biochemistry LE, 1564-1568.

Grahame, D. C. (1947) Chem. Rev.: tL, 441-501.

Gregoriadis, G., Leathwood, P.D. and Ryman, B. E. (1971) FEBS Lett. 11,95-99.

Gregorladis, G. and Ryman, B. E. (1972) Biochem. J. M, 123-133.

Gregoriadis, G. (1974) Biochem. Soc. Trans. 2,117-119.

Gregoriadis, G. (1976) Meth. Enzymol. 44,698-709.

Gregorladis, G. (1977) Nature W, 407-409

145
 Gregoriadis, G. and Allison, A. C. (1990) Liposomes in hiological system.s. John

Wiley and Sons, Chichester, England.

Gregorladis, G., Kirby. C., Meehan, A. and Senior, J. (1981) In: Liposomes, drugs

and immunocompetent cell functions (Nicolau, C. and Paraf, A., eds.) pp.29-40,

Acad. Press. London.

Gregorladis, G. (1982) Drugs 24,261-266.

Gregorladis, G. (1983) Pharmacy International: 33-37.

Grimes, P. A., Stone, R.A., Laties, A.M. and Li, W. (1982) Arch. Ophthalmol. LO,

635-639.

Guo, L. S.S., Hamilton, R.L., Goerke. J., Weinstein, J. N. and Havel, R. J. (1980) J.

Lipid Res. 21,993-1003.

Hamilton, R. L. (1978) In: Disturbances in Lipid and Lipoprotein Metabolism.

(Dietschy, J. M., Gotto, A.M. Jr. and OnIko, J. A., eds.) pp. 155-171. American

Physiological Society, Bethesda, MD.

Hamilton -Miller, J. M. (1973) Bacteriol. Rev. 37,166-169.

Handjani-Vila, R. M., Ribier, A., Rondot, B. and Vanlerberghe, G. (1979) Int. J.

Cos- Sci., L 303-314.

Hargreaves, W. R. (1978) Biochem. L7,3759-3767.

Hawkins, D., Pinekard, R.N. and Farr, R.S. (1969) Science LN, 780-782.

146
 Haydon, D. A. and Myers, V. B. (1973) Biochim. Biophys. Acta. DL 429-443.

Heath, T. D., Montgomery, J. A., Piper, J. R. and Papahadjopoulos, D. (1983) Proc.

Nall. Acad. Sci. USA 0,1377- 1381.

Helple, J. M and Taylor, D. L (1992) In: Intracellular pH: its measurement, regulation

and utilisation in cellular functions. (Nuccitelli, R and Deamer, D. W., eds) pp. 21-

54. New York

Hope, M. J., Bally, M. B., Webb,G and Cullis. P.R. (1985) Biochim. Biophys. Acta.

M-, 55-65.

Horbett, T. A. (1981) J Biomed. Maier. Res. LI, 673-695.

Horbett, T. A and Weathersby, P.K. (1981) J. Biomed. Mater. Res. 11,403-423.

Hsla, J. C., Schneider, H. and Smith, I. C.P. (1971) Can. J. Biochem. lp_, 614-622.

Hsu, M. J and Juliano, R. L. (1992) Biochim. Biophys. Acta. ZX. 411-419.

Huang, C. H. and Chariton, H. P. (1971) J. Biol. Chem. 211,2555-2562.

Hunt, A. C. (1982) Biochim. Biophys. Acta. 719,450-463.

Hunter, R. J. (1981) Zeta Potential in Colloidal Science. Acad. Press, London.

IhIenfleld, J. V and Cooper, S. L (1979) J. Biomed. Maier. Res. U, 577-591.

IhIer, G. M. (1993) Pharmac. Ther. 29,151-169.

147
 IhIer, G. M. (1996) Methods in Drug Delivery. Int. Encycl. Pharm. Ther.. Section

120 and references therein.

Inoue, K. (1974) Biochim. Biophys. Acta. M, 390-402.

Jonas, A. and Krajnovich, D. J. (1978) J. Biol. Chem. M, 5758-5763.

Juliano, R. L. and Stamp, D. (1975) Biochim. Biophys. Res. Commun. U, 651-658.

Juliano, R. L. and Stamp, D. (1978) Biochem. Pharmacol. 2Z, 21-26.

Juliano, R. L. and Lin, G. (1980) In 'Liposomes and Immunobiology" (eds., Tom, B.

and Six, H. ) pp. 49-66, Elsevier, New York.

Juliano, R. L. (1981) In: Liposomes: from Physical Structure to Therapeutic

Application (Knight, C.G.,ed.) pp-21-32. Elseiver North Holland Biomedical Press,

Amsterdam.

Juliano, R. L., Lopez- Berestein, G., Hopfer. R.L., Mehta. R., Mehla, K and Mills, K.

(1985) Ann. N. Y. Acad. Sci. USA. 446,390-402.

Kaneda, A., Nishimura, K and Sezaki, H. (1974) Chem. Pharm. Bull. 21,523-528.

Kano, K and Fendler, J. H. (1977) Life Sci. 2Q. 1729-1734.

Kaye, S. B. (1991) Cancer Treat. Rev. 6., 27-50.

Kimelberg, H. K. (1976) Mol. Cell. Biochem. LQ, 171-190.

148
Kimelberg, H. K. and Mayhew, E. (1978) CRC Crit. Rev. Toxicol. §, 23-29.

Klmelberg, H. K., Tracy. T. F., Watson, R. E., Kung, D., Reiss, F. L. and Bourke, R.S.

(1978) Cancer Res. 706-715.

Kinsky, C.B., Haxby, J. A and Kinsky, S.C. (1966) Proc. Nail. Acad. Sci. USA a.

300-304.

Kirby, C., Clarke, J. and Gregoriadis, G. (1980s) Biochem. J. LU, 591-598.

Kirby, C., Clark, J. and Gregoriadis, G. (1980b) FEBS Lett. LW2), 324-328.

Kirby, C and Gregorladis, G. (1984) Biotechnology, No. LL, 979-984.

Krupp, L., Chobanian, A. V. and Brecher, P.I. (1976) Biochim. Biophys. Res.

Commun. Z2_,1251-1258.

Kunitake, T and Okahata, Y (1977) J. Am. Chem. Soc. 22,3860-3861.

Kunitake, T and Okahata, Y (1978) Bull. Chem. Soc. ipn. IL, 1877-1879.

Kunitake, T and Okahata, Y (1980) J. Am. Chem. Soc. JU, 549-553.

Kunitake, T. (1986) Ann. N. Y. Acad. Sci. USA EL 70-82.

Ladbrooke, B. D and Chapman, D. (1969) Chem. Phys. Lipids 1.304-309.

Laemmll, U. K. (1970) Nature 227,680-685.

149
Langer, R. (1980) Chem. Eng. Commun. §., 1-48.

Langer, R. and Peppas, N. A. (1981) Biomaterials Z 195-210.

Langer, R. and Peppas, N. A. (1983) J. Macromol. Sci. 4.247-252.

Law, S.L., Lo, W.Y., Pai, S.H., Teh. G.W. and Kou, F.Y. (1996) Int. J. Pharm., U,

237-241.

Lelkes, P.I. and Tandeter, H. B. (1982) Biochim. Blophys. Acla. ZM, 410-419.

Leserman, L. D., Machy, P. and Barbet, J. (1981) Nature W, 226-228.

Lesslauer, W., Cain, J. E and Blasie. J. L. (1972) Proc. Nall. Acad. Sci. U, 1499-

1506.

Leung, J. G. M. (1980) Biochim. Biophys. Acla. L9_7,427-432.

Levine, Y. K and Wilkins, M. H. F. (1971) Nature New Biol. M, 69-72.

Levy, M. and Elliott, A. M. (1968) J. Protozool. 11,208-212.

Lopez- Berestein, G., Mehta, R., Hopfer, R.L., Mills, K., Kasi, L., Afehta. K.,

Fainstein, V., Luna, M., Hersh, E.M and Juliano, R. L. (1993) J Ittfect. Dis. UZ,

939-945.

Lopez- Berestein, G., Fainstein, V., Hopfer. R., Sullivan, M., Rosenblum, Af., MAW,

R., Luna, M., Hersh. E.. Reuben, J., Mehta, K., Juliano, R.L and Bodey, G. (1985)

J. Infect. Dis. 151.704-710.

150
.
Machy, P. and Leserman, L. D. (1983) Biochim. Biophys. Acta. UQ, 313-320.

Madden, T. D., Bully, N. B., Hope, 3f. J., Cullis, P.R., Schieren, H. P. aqnd Janoll, AS

(1985) Biochim. Biophys. Acta. eL7,67-74.

Martin, F. J. and MacDonald, R. C. (1976) J. Cell Biol. ZO, 515-526.

Marty, J. M., Oppenheim, R.D. and Speiser, P. (1978) Pharm. Acla. Hely. 11,17-23.

Mathias, C., Thambi Dorai, D. and Bachhawal, B.K. (1977) Ind. J. Biochim. Biophys.

14,142-148.

Mayhew, E., Papahadjopoulos, D., Rustum, Y.M. and Dave, C. (1976) Cancer Res.

16,4406-4411.

Mayhew, E., Papahadjopoulos, D., O'Malley, J. A., Carter, W.A. and Vail, W.J. (1977)

Mol. Pharmacol. 13,488-495.

Mayhew, E., Gotfredsen, C., Schneider, YJ and Trouet, A. (1980) Biochem

Pharmacol. 29,877-886.

Mayhew, E., Rustum, Y and Vail, W.J. (1983) Cancer Drug Deliv. L, 43-58.

McLaughin, S. (1977) In: Current Topics in Membrane and Transport. Vol. 2,

(Bronner, F. and Kleinzeller, A., eds.) pp. 71-144, Acad. Press, New York.

Medoff, G., Braitburg, J., Kobayashi, G.S and Bolard, J (1983) Ann. Rev.

Pharmacol. Toxicol. 21,303-304.

151
Menashe, M., Lichtenberg, D., Gutierrez-Merino, C. and Billonen, R.L. (1981) J.

Biol. Chem. M, 4541-4543.

Milsmann, M. H. W., Schwendener, R.A. and Weder, H. G. (1978) Biochim. Biophys.

Acta. 512,147-154.

Morrisett, J. D., Jackson, R. L. and Gotto, A. M. (1977) Biochim. Biophys. Acta. IZZ,

93-103.

Mosher, D. S. (1984) Ann. Rev. Med. 561-575.

Muller, M., Baudhuin, P. and De Duve, C. (1966) J. Cell Physiol. 0.165-170.

New, R. B. C., Chance, M. L., Thomas, S.C. and Pelas, W. (1978) Nature IZI. 55-56.

Newman, G. C and Huang, C (1975) Biochem. L4,3363-3369.

Nilsson, J. R. (1970a) C.R. Trav. Lab. CarlsbergM, 87-90.

Nilsson, J. R. (1970b) C.R. Trav. Lab. Carlsberg la, 107-112.

Nucciteill, R and Deamer, D. W. (1982) Intracellular pH: its measurement, regulation

and utilisation in cellular functions, New York, A. R. Liss.

Okahata, Y., Tanamachi, S., Nagai, M and Kunitake, T (1991) J. Coll. Inter. Sci.

a2,401-417.

Olson, F., Hunt, C.A., Szoka, F. C., Vail, WJ and Papahadjopoulos, D. (1979)

Biochim. Biophys. Acta. lý_7,9-23.

152
Olson, F., Mayhew, E., Maslow, D.,Rustum, Y. and Szoka, F. (1992) Eur. J. Clin.

Oncol. L& 167-176.

Op den Kamp, J. A. F., Kauerz, M. T. and Van Deenen, L. L. M. (1975) Biochim.

Biophys. Acta. 406,169-177.

Oppenhenhelm, R. C. (1991) Int. J. Pharm. d, 217-234.

Ostro, M. (1987) Sci. Amer. M, 102-111

Pagano, R. E. and Weinstein, J. N. (1978) Ann. Rev. Blophys. Bloeng. Z, 435-468.

Papahadjopoulos, D., Jacobson, K., Nir, S. and Isac, T. (1973) Biochim. Blophys.

Acta. gl, 330-334.

Papahadjopoulos, D., Mayhew, E., Poste, G., Smith. S. and Vail. W.J. (1974) Nature

212,163-166.

Parker, C. A. (1968) Photoluminescence of Solutions. Elsevier, New York.

Pethig, R., GascOYne,P.R. C., McLaughlin, J. A. and S.-ent-Gyorgyi. A. (1984) Proc.

Nail. Acad. Sci. L, 2088-2091.

Plard, J. C., Elsoda, M., Alkhalaf, W., Rousseau. M.. Desmazeaud. M., Vassal, L. and

Gripon, J. C. (1986) Biotech. Lett. d(4), 241-246.

Poste, G. (1983) Biol. Cell. 47,19-38.

153
Poznansky, M. J. and Juliano, R. L. (1994) Am. Soc Pharmacol. Exp. Ther. 277-

336.

Pratt, W.B. (1977) Chemotherapy of Infection. Oxford University Press. New York.

Qualssi, N. A and Cabron, A. (1985) Ann. Inst. Pasteur Immun. L=, 169-185.

Rahman, Y. E. (1980) In: Liposomes in biological systems (Gregoriadis. G. and

Allison, A. C., eds. ) pp. 265-298, John Wiley and Sons, Chichester, England.

Rahman, Y. E., Cerny, E.A., Patel, K. R., Lau, E.H. and Wright, B.J. (1982) Life Sci.

31,2061-2063.

Rahman, A., Carmichael, D., Harris, M and Roh, J, K., (1986) Cancer Res. :M. 2295-

2299.

Rahman, A., Kessler, A., More, N., Sikic, B., Rowden, G., Wooley. G. and Schein.

P.S. (1980) Cancer Res. 40,1532-1536.

Ralston, E., Hjelmeland, L. M., Klauser. R.D., Weinstein, J. N and Blumenthal, R.

(1981) Biochim Biophys. Acta. 649,133-137.

Rao, L. S. (1984) In: Liposome Technology, Vol. L (Gregoriadis, G.. ed. ) pp. 247-257,

CRC Press, Boca Raton, Florida.

Readio, J. D and Bittman, R. (1982) Biochim. Biophys. Acta. fiýLI, 219-224.

Rogerson, A. (1986) Ph.D. Thesis, University of Strathclyde.

154
Rogerson, A., Cummings, J., Wilmott, N and Florence, A. T. (1988) in press.

Rossi, J. D and Wallace, B. A. (1983) J. Biol. Chem. 2M. 3327-3331.

Russell, F. E. (1980) Snake Venom Poisoning, J. B.Lippincott Company, Philidelphia,

Toronto.

Ryan, P. J., Davis, M. A., De Gaeta, L. R. Woda, L. R. and Melcior, D. L. (1984)

Radiology JR, 759-762.

Ryman, B. E. and Tyrrell, D. A. (1980) Essays Biochem. Lý, 49-98.

Scherphof, G., Roerdink, F., Waite, M. and Parks, J. (1978) Biochim. Biophys. Acta.

542,296-307.

Scherphof, G., Damen, J. and Hoekstra, D. (1981). In: Liposomes.- from physical

structure to therapeutic actions. (Knight, C. G.. ed) pp. 299-332. Elsevierl North

Holland, Amsterdam.

Scherphof, G., Damen, J and Wilschut. J (1984) In: Liposomes Technology, Vol. LLL

(Gregoriadis, G., ed.) pp. 205-224. CRC Press, Boca Ralon, Florida.

Scherphof, G. and Morselt, H. (1994) Biochem. J. 2ZL. 423-429.

Schieren, H. (1978) Biochim. Biophys. Acta. JU, 137-140-

Schneider, M. (1977) Battelle Memorial Institute US patent 224' 179.

155
 Seltzer, S.E., Davis, M. A., Adams, D.F., Shulkin, P.M., Landis, W.J. and Havron, A.

(1984) Invest. Radiol. 12,142-151.

Shaw, D. J. (1968) Electrophoresis, Acad. Press, London.

Shinozawa, S., Araki, Y and Oda, T (1981) Acla. Med. Okayama J-1,395-405.

Sjoholm, 1. and Edman, P. (1984) In: Microspheres and Drug Therapy.

Pharmaceutical, Immunological and Medical Aspects. (Davis, S.S., Illum, L., Mc. Vie,

J. G. and Tomlinson, E., eds.) pp. 245-262.

Slotboom, A. J., Verheij, H. M and de Haas, G.H (1982) In: Phospholipids

(Hawthorne, J. N and Ansell, G.B.,eds.) pp. 359-434. Elsevier, Amsterdam.

Spector, A. A and Fletcher, J. E. (1977) In: Albumin: structure, blosynthesis,

function. (Peters, T and Sjoholm, I., eds.) pp.51-60, Pergamon Press.

Strauss, G., Schurtenberger, P and Hausser, H (1986) Biochim. Biophys. Acla. M,

169-180.

Sweet, C. and Zull, J. E. (1970) Biochim. Biophys. Acla. M, 253-262.

Szoka, F. and Papahadjopoulos, D. (1978) Proc. Nall. Acad. Sci. USA Zj.

4194-4198.

Szoka, F. and Papahadjopoulos, D. (1980) Ann. Rev. Biophys. Bioeng. 2,467-508.

Tagesson, C., Stendahl, 0. and Magnusson, K. E. (1977) Stud. Biophys. 6j, 151-160.

156
Tall, A. R. and Small, D. M. (1977) Nature 2.61 163-164.

Tall, A. R and Lange, Y. (1978) Biochim. Biophys. Acta. ILI, 185-197.

Tanford, C. (1980) The Hydrophobic Effect. Formation of Micelles and Biological

Membranes. 2nd. Edition, Wiley- Interscience, New York.

Tomlinson, E., Burger, J. J., Schoonderwoerd, E.M. A. and McV! e, J. G. (1984) In:

Microspheres and Drug Therapy. Pharmaceutical, Immunological and Medical

Aspects, (Davis, S.S., Illum, L., McVie, J. G. and Tomlinson, E., eds. ) pp. 75-91

Towbin, H., Staehlin, T and Gordon, J. (1979) Proc. Nall. Acad. Scl. USA Zj, 4350-

4354.

Tyrrell, D. A., Ryman, B.E., Keeton, B.R. and Dubowitz, V. (1976) Brit. Afed. J. 2.

88-91.

Tyrrell, D. A., Healh, T. D., Colley, C.M. and Ryman, B.E. (1976) Biochim. Biophys.

Acta. 4J7,259-302-

Ugoretz, R.J. (1976) J. Am. Med. Assoc. (JAMA) M. 295-296.

Unlyal, S and Brash, J. L. (1982) Thromb. Haemostas. JZ, 285-290.

Van Berkel, T. J. C., Kar Kruijt, J., Harkes. L. Vagelkerke. J. S., Spanjer, 11 and

Kempen, H. J. M. (1996) In: Site-specific drug delivery, (Tomlinson. E and Davis,

S.S., ed.), pp. 49-68, John Wiley and Sons Ltd; Chichester.

Van Deenen, L. L. M. and de Haas, G. H. (1964) Ady. Lipid Res. 2,167-234.

157
Van Renswoude, J. and Hoekstra, D. (1980) Biochemistry n, 540-546.

Vanlerberghe, G., Ribier, A and Handiani-Vila, R.M (1978) Communicalion au

Colloque du CNRS sur la physicochemie des composes amphiphiles, Bordeaux.

Vroman, L., Adams, A. L.. Fisher, G.C and Munoz, P.C (1980) Blood 51,156-159.

Wallach, D. F. H., Surgenor, D. M., Soderberg, J. and Delano, E. (1959) Anal. Chem.,

11,456-461.

Wallach, D. F.H and Steck, T. L. (1963) Anal. Chem. a, 1035-1044.

Weast, R. C. (1984) Handbook of Chemistry and Physics pp. F-37. E-50.65th Edn.

CRC Press, Cleveland.

Weinstein, J. N., Yoshikami, S., Henkart, P., Blumenthal, R. and Hagins. IV.A. (1977)

Science 195.489-492.

Weinstein, J. N., Ralston, E., Leserman, L. D., Klausner, R.D., Dragsten, P.. Ilenkart,

P. and Blumenthal, R. (1984) In: Liposome Technology, Vol UL (Gregoriadis. G..

)
ed. pp. 183-204. CRC Press, Boca Raton, Florida.

Weinstein, J. N., Blumenthal, R. and Klausner, R. D. (1986) Meth. Enzymol. M,

657-668.

Weissmann, G., Sessa. G. and Weissman, S. (1966) Biochem. Pharmacol. Ll.

1537-1551.

158
Weissmann, G., Bloomgarden, D., Kaplan, R., Cohen, C., Hoffstein. S., Collins, T.,

Gottieb. A. and Nagle, D. (1975) Proc. Nall. Acad. Sci. USA Z2,88-92.

Weissmann, G., Cohen, C. and Hoffstein, S. (1977) Biochim. Biophys. Acta. JR,

375-385.

Whitmore, A., Brooks, L. G and Wheeler, K. P (1979) J. Pharm. Pharmacol. U, 277-

283.

Wililams, R. M and Chapman, D. (1970) In: Progress in the Chemistry of Fats and

Other Lipids, 2_1,1-10 (Holman, R. T., ed..) Pergamon Press, Oxford.

W11schut, J. C., Regis, J., Westenburg,H. and Scherphof, G. (1976) Biochim.

Biophys. Acta. 131,20-31.

Wi1schut, J. C., Regts, J., Westenberg,H. and Scherphof, G. (1978) Blochim.

Biophys. Acta. U-0,185-196.

Wilschut, J. and Papahadjopoulos, D. (1979) Nature ZU, 690-692.

Wilschut, J. C., Regts, J. and Scherphof, G. (1979) FEBS Left. 91,181-186.

Wilson, T., Papahadjopoulos, D. and Taber, R. (1977) Proc. Nall. Acad. Sci. USA

74,3471-3475.

Yatvin, M. B., Weinstein, J. N., Dennis, W.H. and Blumenthal, R. (1979) Science ZU,

1290-1293.

Yoshioka, S. (1984) Chem. Pharm. Bull. JI, 308-312.

159
Yoshioka, S., Ishibashi, M., Shibazaki, T and Uchiyama, M. (1984) J. Par. Sci.

Tech. Bf 6). 222-227.

Zborowski, J., Roerdink, F. and Scherphof, G. (1977) Blochim. Biophys. Acta. 497,

183-191.

160
 THERMAL TRANSITIONS

Cammund TemDerature for itel-to-Rel

crystal (lamtU.Qdlc,. Ia

DPPC (C16) 41

DMPC (C 12) 23

egg PC -15/-7
surfactant 1 (m. wt. 473) 41-43

surfactant 2 (m. wt. 972) 41-47

surfactant 3 (m.wt. 404) 40-42

161
Enzymes of Snake Venom. [Reproduced from Russell, F. E (1980)]

Ace tylcholinesterase: catalyses the hydrolysis of acetylcholine to choline and

acetic acid.

Arginine ester hydrolase: non-cholineste rase. The substate specificities are

directed to the hydrolysis of the ester or peptide linkage, to which an arginine

residue contributes the carboxyl group.

Collagenase: specific kind of proteinase that digests collagen.

Hyaluronidase: catalyses the cleavage of internal glycoside bonds of certain acid

mucopolysaccharides.

Lactate dehydrogenase: reversibly catalyses the conversion of lactic acid to

pyruvic acid.

L-aminoacid oxidase: catalysesthe oxidation of L-a-amino acids and a-hydroxy

acids and is the most active of the known amino acid oxidases. This gives the

venom it's yellow colour.

NAD-nucleotidase: catalyses the hydrolysis of the nicotinamide N-riboside linkage

of NAD, yielding nicotinamide and adenosine diphosphate riboside.

162
5'-nucleotidase: specifically hydrolyses phosphate monoesterase which links with a

5' position of DNA and RNA.

Phosphollpase A2: catalyses the hydrolysis of one of the fatty ester linkages in

diacyl phosphatides, forming lysophosphaticlesand releasing both saturated and

unsaturated fatty acids.

Phosphollpase B: hydrolyses lysophosphatides.

Phospholipase B and C: hydrolyses lysophosphatides.

Phosphomonoesterase: hydrolyses mono esters.

Phosphodiesterase: an orthophosphoric diester phosphohydrolase that releases 5-

mononucleotide from the polynucleotide chain and thus acts as an exonucleotidase.

Proteolytic enzymes: trypsin-like enzymes that digest tissue proteins and

peptides.

163
 A semi-logarithmic [Aut ()I
distances
i
aqaýnst molecular weig-tlt___Q_f
pro-t-eins

100000

p
i
R
0

N
w

10000

0248 10 12
DISTANCE OF BAND ('ýN ((--rii)
APPENDIX 3
Iý., ,

as explained in the teýxt

164
(o

OD
" 'I'

!_-
-

-ý
I-
I
 (9)

Z)
-. I
0
CIA
 0"

4D
ý

I !, -1 . 10 .II-.
liw
N

Qk-
 N

i&
. "0

Iwo
80

C4-
 113

Ile
I,