Physicochemical properties of liposomes in relation to uptake by alveolar macrophages


Material Information

Physicochemical properties of liposomes in relation to uptake by alveolar macrophages
Physical Description:
xvi, 215 leaves : ill. ; 29 cm.
Cacace, Janice L
Publication Date:


Subjects / Keywords:
Liposomes -- chemistry   ( mesh )
Liposomes -- metabolism   ( mesh )
Macrophages, Alveolar -- physiology   ( mesh )
Phospholipids   ( mesh )
Drug Carriers   ( mesh )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph. D.)--University of Florida, 1991.
Includes bibliographical references (leaves 189-214).
Statement of Responsibility:
by Janice L. Cacace.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
oclc - 50900412
System ID:

Full Text







Copyright 1991


Janice L. Cacace

For My Parents

Thomas and Diana Cacace


My Husband

John McDonnell

For Never Questioning

And Always Supporting


I would like to express my thanks and appreciation to

my advisor Dr. Hans Schreier for his guidance and friendship

as well as the means and support necessary to complete this

dissertation. I thank the members of my committee, Dr. John

Perrin, Dr. Stephen Schulman and Dr. Dinesh Shah, for their

insights into the physicochemical aspects of this project. I

am ever indebted to Dr. Ricardo Gonzalez-Rothi of the

Veterans Administration Medical Center and his staff, Leslie

Straub and Dave Soucy, for guidance on the cell culture

aspects of this project. Special thanks go to Dr. Berndt

Mueller and his staff at Christian Albrechts University for

guidance on the Malvern Zetasizer, to Alan Thein for

assistance on the Rank Brothers apparatus, to Dr. Richard

Prankard for his assistance with the Perkin Elmer DSC 7, and

to Dr. Ramesh Chander for help characterizing alveolar

lining material and for sharing his knowledge of

macrophages. These acknowledgements would not be complete

without thanking the remaining staff of the Department of

Pharmaceutics, as well as my fellow graduate students,

especially Gwen Saldajeno-DeLeon and Anup Zutshi. Special

thanks go to the Pharmaceutical Manufacturers Association

for the Advanced Predoctoral Fellowship in Pharmaceutics.



ACKNOWLEDGEMENTS .................................... iii

LIST OF TABLES...................................... vii

LIST OF FIGURES...................................... viii

GLOSSARY OF TERMS.................................... xiii

ABSTRACT........................ .... ................ xiv


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

Background Alveolar Macrophages................. 5
Reticuloendothelial System (RES)........... 5
Endocytosis............................... 6
Phagocytosis.............................. 9
Intracellular Fate of Ingested Substances. 11
Factors Affecting Phagocytosis............. 12
Properties of Phagocytic Cells............. 24
Alveolar Macrophages (AM)................. 25
Macrophage Culture........................ 27
Lung Surfactant........................... 29
Background Liposomes ........................... 33
Phospholipids............................. 33
Liposomes ................................. 40
Physicochemical Parameters of Liposomes........ 45
Thermotropic Phase Transitions............ 45
Effects of Water on Phospholipid
Phase Transitions................... 47
Differential Scanning Calorimetry.......... 51
Factors Affecting Thermotropic
Transitions of Phospholipids........ 55
Thermal Transitions in Phospholipid
Mixtures and Phase Diagrams.............. 58
Use of DSC in Liposome Formulation....... 63
Surface Charge of Liposomes............... 65
Factors Affecting Electrophoresis......... 69
Electrophoresis of Phospholipids......... 72
Interactions Between Liposomes and Cells....... 79
Selection of Marker............................ 84


Objective.................. ........ ............ 90

2 MATERIALS AND METHODS............................ 95

Chemicals and Reagents......................... 95
Liposome Preparation...................... 95
Cell Culture............................... 95
Experimental Procedures.......................... 97
Phospholipid Purity by TLC................ 97
Carboxyfluorescein Purification............ 98
Fluorescence.............................. 99
Liposome Preparation....................... 99
Liposome Separation from Unencapsulated CF. 100
Particle Size Measurement.................. 101
Differential Scanning Calorimetry........... 101
Zeta Potential............................ 101
Stability...................................... 103
Macrophage Culture............................. 104
Harvesting of Alveolar Macrophages.......... 104
AM Viability............................... 105
AM Phagocytic Activity..................... 105
Alveolar Lining Material (ALM)............. 106
Lung Homogenate (LH)....................... 107
Qualitative Uptake......................... 108
Quantitative Uptake........................ 108
In vitro Stability......................... 109
Statistical Methods ........................ 109

3 RESULTS AND DISCUSSION........................... 110

Physicochemical Parameters ...................... 110
Phospholipid Purity........................ 110
CF Fluorescence............................ 111
Liposome Size and Size Distribution
by Laser Light Scattering............. 115
Differential Scanning Calorimetry............... 117
Tc of HEPC:DPPG Liposomes.................. 117
T. of HEPC:DPPS Liposomes.................. 121
Tc of SPM ................................ 124
Effect of ALM on T ........................ 130
Effect of Ca2" on DSC Measurement.......... 132
Zeta Potential Measurements by Rank
Brothers Mark II............................ 134
Zeta Potential of HEPC:DPPG Liposomes...... 134
Zeta Potential of HEPC:DPPS
Liposomes .............................. 134
Zeta Potential Measurements by Malvern
Zetasizer.................................. 134
Zeta Potential of HEPC:DPPG Liposomes...... 134
Zeta Potential of HEPC:DPPS Liposomes...... 138
Discussion Zeta Potential of Negatively
Charged Phospholipids................. 143

Zeta Potential of HEPC:SPM Liposomes....... 146
Discussion Zeta Potential.................. 150
Comparison of Measurement Methods........... 153
Liposome Stability............................. 155
Stability of HEPC:DPPG Liposomes........... 155
Stability of HEPC:DPPS Liposomes........... 155
Liposome Stability Discussion.............. 162
Cell Culture.................................... 167
Viability and Functional Macrophage
Studies................................ 167
ALM Analysis ............................... 167
Liposome Uptake by Alveolar Macrophages.... 170
Uptake in ALM vs. DMEM ..................... 176
Discussion Uptake.......................... 181

4 SUMMARY AND CONCLUSIONS......................... 183


BIOGRAPHICAL SKETCH ..................................


Table page

1. Composition of Pulmonary Surfactant.............. 30

2. Phase Transitions of Anhydrous Phospholipids..... 46

3. Effect of Chain Length on To of DPPC ........... 49

4. Effect of Phospholipid Headgroup on T. .......... 49

5. Effect of pH on Tc of DPPG and DPPS............... 56

6. Effect of Vesicular Structure on T .............. 58

7. Isomorphous Mixtures of Phospholipids............ 62

8. Effect of Temperature on Dielectric Constant and
Viscosity of Water............................... 70

9. Effect of Sugars on Dielectric Constant and
Viscosity of Water at 25 C....................... 70

10. Dielectric Decrement Values of Salts............. 71

11. Cation Effect on Zeta Potential of
Phosphatidylglycerol............................ 77

12. Effect of Calcium and Temperature on Zeta
Potential of Phosphatidylserine................... 77

13. Effect of Ionic Strength on Zeta Potential
of Phosphatidylserine Liposomes................... 78

14. Composition of DMEM............................... 96

15. Thin Layer Chromatography of Phospholipids....... 110

16. Particle Sizing of HEPC:DPPG Liposomes........... 116

17. Particle Sizing of HEPC:DPPS Liposomes........... 116


Figure page

1. Structural representation of phospholipid (A)
and sphingolipid (B). The circled area (top left)
represents the hydrophilic headgroup with
possible X substituents (bottom box), and the
shaded area (right) represents the hydrophobic,
fatty acid portion of the molecule, with m and n
denoting the length of the hydrocarbon chain....... 34

2. Phospholipids with corresponding dynamic molecular
shapes and polymorphic phases; (A) micellar,
(B) bilayer, (C) interdigitated, (D) hexagonal..... 37

3. Schematic diagram of a multilamellar liposome,
an "onion skin" configuration with concentric lipid
bilayers separated by aqueous spaces, surrounding
an aqueous core.................................... 39

4. Preparation of multilamellar vesicles, and sizing
by extrusion through a polycarbonate membrane....... 43

5. Representative endothermic phase transition for
heating (A) and cooling (B) of hydrogenated
egg phosphatidylcholine........................... 53

6. Idealized phase diagram for a binary mixture of
phospholipids whose components are completely
miscible in the liquid and solid state. Above the
fluidus line, the phospholipid is in a liquid
state; below the solidus line, the phospholipid
is in a solid or "gel" state; in between the lines,
both states exist in equilibrium................... 61

7. The electric double layer, represented by a
spherical particle with a net negative surface
charge suspended in aqueous medium and surrounded
by a layer of opposite charge. Below the particle
is a diagram indicating the electrical potential as
a function of distance from the surface of the
particle. The potential at the plane of shear is
the zeta potential................................. 67

8. Structure of 5,6-carboxyfluorescein................ 86


9. Concentration versus relative fluorescence for
5,6-carboxyfluorescein at excitation X 492 nm and
emission X 524 nm. Note the decrease in
fluorescence due to self-quenching at
concentrations > 1 x 10-'M......................... 89

10. Phospholipid structures............................ 93

11. Calibration curve for 5,6-carboxyfluorescein in
PBS and Triton X-100 plotted as molar concentration
CF versus relative fluorescence intensity at
excitation X 492 nm and emission X524 rnm.......... 112

12. Calibration curve for 5,6-carboxyfluorescein in
PBS or PBS containing 1 X 10-IM HEPC or DPPS,
plotted as molar concentration CF versus
relative fluorescence intensity at excitation
X 492 nm and emission X 524 nm ................... 113

13. Effect of the molar concentration of HEPC on
the relative fluorescence intensity of CF at
excitation X 492 nm and emission X 524 nm.......... 114

14. Representative DSC endotherm for DPPG (left, solid
line), and HEPC (right, dashed line)............. 118

15. Phase diagram of HEPC:DPPG liposomes based on
onset temperature of heating (top) and
cooling (bottom).................................. 119

16. Transition temperature versus mol% DPPG for
HEPC:DPPG liposomes in PBS (time 0) or DMEM
(4 hours)......................................... 120

17. Representative DSC endotherm for DPPS (50 mM)
in 3 mM CF (left, solid line) and DPPS (25 mM)
in PBS (right, dashed line)........................ 122

18. Phase diagram of onset temperature of heating
(top), and cooling (bottom) for HEPC:DPPS
liposome series................................... 123

19. Transition temperatures of HEPC:DPPS liposomes
in PBS (filled) or DMEM (diagonal)................ 125

20. T of HEPC:DPPS liposomes encapsulating
PBS or CF (LIPCF)................................. 126

21. Representative thermogram for SPM (Left, solid
line), and HEPC (right, dashed line).............. 127

22. Phase diagram of HEPC:SPM liposomes represented
by temperature of onset (bottom) and
completion (top)................................... 129

23. Effect of incubation of liposomes in ALM for
4 hours on T, of HEPC, HEPC:DPPG (50:50),
and HEPC:DPPS (50:50)............................. 131

24. Zeta potential of HEPC:DPPG liposomes by Rank
Brothers Mark II Microelectrophoresis
Apparatus in PBS (bottom) and DMEM (top).......... 135

25. Zeta potential of HEPC:DPPS liposomes by Rank
Brothers Mark II Microelectrophoresis
Apparatus in PBS (bottom) and DMEM (top).......... 136

26. Zeta potential of HEPC:DPPG liposomes by Malvern
Zetasizer 3 in PBS (bottom) and DMEM (top)........ 137

27. Zeta potential of HEPC:DPPG liposomes by Malvern
Zetasizer 3 in DMEM (top) and DMEM without
FBS (bottom)..................................... 139

28. Zeta potential of HEPC:DPPG liposomes by Malvern
Zetasizer 3 in PBS (bottom), PBS + 5% FBS
(middle), and PBS + Ca" (200mg/L) + Mg2+
(200 mg/L) (top) .................................. 140

29. Zeta potential of HEPC:DPPS liposomes by Malvern
Zetasizer 3 in PBS (top) or 3mM CF (bottom)....... 141

30. Zeta potential of HEPC:DPPS liposomes by Malvern
Zetasizer 3 in PBS (bottom) or DMEM (top)......... 142

31. Zeta potential of HEPC:DPPS liposomes by Malvern
Zetasizer 3 in PBS (top) or 3 mM CF (bottom)....... 144

32. Zeta potential of LIPCF versus mol% charged
phospholipid..................................... 147

33. Zeta potential of HEPC:SPM liposomes in PBS
and DMEM.......................................... 148

34. Stability of HEPC:DPPG liposomes in PBS, DMEM,
ALM, and LH. A: HEPC; B: HEPC:DPPG (80:20);
C: HEPC:DPPG (50:50); D: HEPC:DPPG (20:80);
E: DPPG. ......................................... 156

35. Stability of HEPC:DPPG liposomes plotted as % CF
retained versus mol% DPPG in various media.
A: PBS; B: DMEM; C: ALM; D: LH.................... 158

36. Stability of HEPC:DPPG plotted as % CF remaining
versus time in vitro in DMEM (open circle) or in
cell culture (closed circle)..................... 159

37. Stability of HEPC:DPPS liposomes in PBS, DMEM,
ALM, and LH. (A) HEPC:DPPS (50:50);
(B) HEPC:DPPS (80:20)............................ 160

38. Stability of HEPC:DPPS liposomes plotted as % CF
retained versus mol% DPPS in various media.
A: PBS; B: DMEM; C: ALM; D: LH..................... 161

39. Stability of HEPC:DPPS liposomes in vitro in
DMEM (open circle) or in cell culture (closed
circle). (A) HEPC:DPPS (80:20);
(B) HEPC:DPPS (50:50)............................ 163

40. Stability of CF in lung homogenate at 370C
represented as % remaining versus time............ 166

41. Representative SDS-Polyacrylamide gel
electrophoresis of alveolar lining material,
with molecular weights of the protein standards
(left side) and calculated molecular weights
for the samples (right side) ...................... 168

42. Log R, of SDS-PAGE protein standards versus
molecular weight of protein, used for
determination of molecular weights of proteins
in ALM ........................................... 169

43. Qualitative uptake of liposomes by AM. Fluorescence
micrograph (left) with the corresponding phase
micrograph (right) (to locate cell position),
after incubation for 15 minutes (top) or 1 hour
(bottom). Note the distinct punctate vacuoles
which are more prominent in the latter time point. 171

44. Uptake of HEPC:DPPG liposomes by AM over 4 hours.
(A) HEPC; (B) HEPC:DPPG (80:20); (C) HEPC:DPPG
(63:35); (D) HEPC:DPPG (50:50); (E) HEPC:DPPG
(20:80)........................................... 173

45. Uptake of HEPC:DPPG liposomes by alveolar
macrophages plotted against mol% DPPG, after 1 and
4 hours incubation............................... 175

46. Uptake of HEPC:DPPS liposomes by alveolar
macrophages over 4 hours. HEPC (solid);
HEPC:DPPS (80:20) (diagonal); HEPC:DPPS
(65:35) (cross hatch).................. ........... 177

47. Uptake of HEPC:DPPS (80:20) (diagonal) versus
HEPC:DPPG (80:20) (solid) by AM over 4 hours...... 178

48. In vitro uptake of HEPC:DPPS (63:35 (diagonal)
versus HEPC:DPPG (65:35) (solid) by AM over
4 hours ........................................... 179

49. In vitro uptake of HEPC:DPPG (65:35) liposomes
by AM in ALM (diagonal) versus DMEM (solid)
over 4 hours..................................... 180


Abbreviation Definition

CF 5,6-carboxyfluorescein
DCP Dicetylphosphate
DLPG Dilaurylphosphatidylglycerol
DMPC Dimyristoylphosphatidylcholine
DMPE Dimyristoylphosphatidylethanolamine
DOPC Dioleoylphosphatidylcholine
DPPC Dipalmitoylphosphatidylcholine
DPPE Dipalmitoylphosphatidylethanolamine
DPPG Dipalmitoylphosphatidylglycerol
DPPS Dipalmitoylphosphatidylserine
DSC Differential scanning calorimetry
DSPC Distearoylphosphatidylcholine
DSPE Distearoylphosphatidylethanolamine
DSPS Distearoylphosphatidylserine
HEPC Hydrogenated egg phosphatidylcholine
LIPCF Extruded liposomes encapsulating CF
MLV Multilamellar vesicle
PA Phosphatidic acid
PC Phsophatidylcholine
PG Phosphatidylglycerol
PS Phosphatidylserine
PI Phosphatidylinositol
SPM Sphingomyelin
SUV Small unilamellar vesicle
Tc Transition temperature
UNEXLIP Unextruded liposomes encapsulating CF


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



Janice L. Cacace

December, 1991

Chairman: Hans Schreier
Major Department: Pharmaceutical Sciences

Treatment of intracellular infections is limited by the

inability of most therapeutic agents to cross the cell

membrane and interact with the infecting organism. One

option is to target the drug directly to the reservoir of

infection--the macrophage--by utilizing the natural capacity

of these cells to ingest particulate matter.

The studies presented here attempt to elucidate the

interactions between a colloidal dosage form, liposomes, and

macrophages of the lungs, i.e., alveolar macrophages. The

primary variable was the percentage of negatively charged

phospholipid in the liposomal membrane. Characterization of

the physicochemical properties of zeta potential, transition

temperature, and stability, as well as uptake by alveolar

macrophages, was assessed in vitro. Stability and uptake

studies were performed by fluorescence using the self-

quenching probe 5,6-carboxyfluorescein. The major

phospholipids used were hydrogenated egg phosphatidylcholine

(HEPC), dipalmitoylphosphatidylglycerol (DPPG) and

dipalmitoylphosphatidylserine (DPPS).

These liposomes retained their contents under in vitro

conditions simulating storage, cell culture, and the lung

lumen. Their ability to release encapsulated materials upon

enzyme contact was demonstrated by retention of < 20% of

encapsulated carboxyfluorescein after 4 hours incubation

with lung homogenate.

Several pitfalls were elucidated by the interaction of

liposomes with the culture medium--Dulbecco's modified Eagle

medium (DMEM). The presence of calcium ions in this medium

caused considerable charge neutralization, in some cases

liposome aggregation, and phospholipid phase separations.

Even though charge neutralization occurred, the uptake of

liposomes varied with DPPG and DPPS content. Maximum uptake

occurred with HEPC:DPPG (65:35 molar ratio). The uptake of

HEPC:DPPS liposomes was similar to HEPC alone; however,

these results may have been complicated by the physical

instability of DPPS containing liposomes.

When HEPC:DPPG (65:35) liposomes were incubated with

alveolar macrophages in alveolar lining material (ALM), the

amount taken up was comparable to the same study in DMEM.

However, the initial rate of uptake was faster in ALM.

Thus, it appears that hydrogenated phospholipid

vesicles provide a suitable carrier for substances targeted

to alveolar macrophages. The uptake process is dependent on

the type and amount of phospholipid present, with an

apparent preference for DPPG containing liposomes.


Liposomes have been widely used for the study of model

membranes and potential drug delivery systems for about 25

years (Bangham 1986). Interest in their use as drug carriers

is based on the potential to encapsulate a diverse number of

materials of biological interest (Gregoriadis 1976). In

addition, it is possible to alter such parameters as

liposome size, surface charge, bilayer fluidity and

stability to adapt the carrier to a wide range of experimen-

tal conditions.

Although considerable technological advances have been

made, liposome use in drug delivery has achieved limited

success. One of the major obstacles to attaining specific

goals was stated recently by Lopez-Berestein and Fidler

(1989): "Selective targeting of therapeutic agents to appro-

priate sites of action while avoiding the reticuloendo-

thelial system is still a challenging and unresolved issue";

therefore, the focus for drug delivery has switched to "new

approaches to selective targeting as well as elaborate

methods for prolonging drug availability and improving

targeting to intracellular sites." Thus, although the use of

liposomes may improve drug therapy, by far the major

rationale for their use involves targeting either to sites

of infection or away from sites of toxicity. This can be to

a specific organ or cell population.

The rapid uptake of liposomes by cells of the reticulo-

endothelial system (RES), mainly in the liver and spleen

(Gregoriadis and Ryman 1972a,b, Segal et al. 1974, Kimelberg

et al. 1979, Abra and Hunt 1981), makes these cells the

perfect target, since it enables the use of their natural

physiologic function of phagocytosis to improve drug


Since these cells can also be the host to a variety of

intracellular infections (e.g., mycobacteria, Brucella,

Listeria, and Salmonella sp.) (DeDuve et al. 1974), the

rationale for cell-directed delivery is even more


One of the factors inhibiting the treatment of these

infections in a traditional manner is the ability of the

causal agent to survive and proliferate intracellularly,

most often within the confines of the vacuolar system into

which it has been introduced through phagocytosis.

Therefore, if one wishes to reach intracellular bacteria

with an appropriate agent (e.g., antibiotic delivery [Lopez-

Berestein 1987, Swenson et al. 1988]), it must be capable of

entering the cell and of acting under the conditions

prevailing within the vacuoles occupied by the bacteria.

Liposomes are potentially useful for this purpose.

Their localization in cells harboring intracellular diseases

has been discussed (Fidler et al. 1980, Yatvin and Lelkes

1982), with the classic example being the treatment of

leishmaniasis, an intracellular parasitic disease localized

in Kupffer cells of the liver (Alving et al. 1978a,b, New et

al. 1978, New 1990).

Other potential sites for delivery to cells of the RES

are the spleen and the lungs. The lungs, because of their

daily interaction with airborne diseases, are prone to

infections and therefore a useful target organ for attempts

to localize drugs. It has been shown that intravenously

delivered liposomes 1 um or greater in diameter exhibit

improved localization in the lungs compared to smaller-sized

particles (Hunt et al. 1979, Fidler et al. 1980). This

effect has been attributed to trapping in the capillary beds

of the lungs (Hunt et al. 1979). It has also been shown that

positively charged (stearylamine) and negatively charged

(phosphatidylserine, phosphatidylglycerol and phosphatidic

acid) liposomes accumulate in lungs to a greater extent than

neutral liposomes of similar size (Jonah et al. 1975,

Kimelberg 1976, Kimelberg et al. 1979, Fidler et al. 1980,

Abra et al. 1984). In each case, however, the majority of

the liposome preparation still accumulated in the liver

rather than in the lungs.

Therefore, there has been increasing interest in the

delivery of liposomes directly into the pulmonary system

offering a non-invasive route of administration by pulmonary

instillation (Juliano and McCullough 1980) or aerosolization

(Debs et al. 1987, Mihalko et al. 1988) as a way to achieve

sustained localized delivery while specifically targeting

drugs to the airways. Several anatomical and physiological

properties of the lungs and airways render them an

attractive portal of entry for systemic drug delivery: a

large surface area (ca. 70 m2), a very thin (<0.5 pm)

diffusional barrier between the alveolar air space and its

surrounding capillary bed, a high blood flow and a

relatively low activity of enzymes when compared to the

liver (thus, a low first-pass effect). However, another

potential use for this approach is the treatment of

intracellular infections localized in the lungs, e.g., with

mycobacteria, histoplasma, and cryptococcus (Schroit et al.

1983, Lopez-Berestein 1987).

In this situation, the role of the RES in the clearance

of particles from the pulmonary system after aerosolization

is of concern. Thus, knowledge of the intimate interactions

between liposomes and cells of the pulmonary tract, as well

as between liposomes and other aspects of the pulmonary

environment (i.e., pulmonary surfactant), is important for

the development of dosage forms targeted to these areas.

To this end, we have studied the phagocytic properties

of cultured alveolar macrophages in an attempt to optimize

the delivery of liposomes to these cells, as well as to

correlate some physicochemical properties of liposomes that

might be important in this process.

Background Alveolar Macrophaqes

Reticuloendothelial System (RES)

While gazing at transparent starfish larvae, where

mobile cells could easily be viewed, Ilya Metchnikoff, in

the late 1800, conceived the idea of a host defense system

composed of amoeboid "wandering" cells which could ingest

solid particles and thus serve to defend an organism against

"noxious intruders" (Karnovsky 1981). This was termed

phagocytosis, after "cpayo~rTE" the Greek word for

phagocytee" or "devouring cell." This process became the

basis for a host defense system composed of specific cells

capable of detecting foreign material and clearing it from

the body and for the theory of "cellular immunity."

The concept of cellular immunity was further developed

by Aschoff (1924), who introduced the reticuloendothelial

system (RES) as a diffuse system composed of mesenchymally

derived fixed and wandering macrophages which could rapidly

ingest and accumulate foreign particulate matter. It is now

recognized that the cells responsible for this are

circulating blood monocytes and various tissue macrophages,

collectively termed the mononuclear phagocyte system (MPS).

The cells of the MPS are produced from stem cells in

the bone marrow where they undergo proliferation and are

delivered to the blood as monocytes (Bellanti and Kadvec

1985). After a period of maturation through a monoblast--

promonocyte--monocyte phase in the blood, the monocytes

migrate to their main site of action in various tissues

where they differentiate further into macrophages. These

cells include Kupffer cells of the liver and macrophages of

the spleen, bone marrow, and lungs.


One of the primary functions of macrophages is the

ingestion and destruction of foreign materials. They are

highly specialized to carry out this function by the process

of endocytosis (Silverstein et al. 1977, Pratten and Lloyd

1986), a general term which includes both phagocytosis

(ingestion of particles) and pinocytosis (uptake of

nonparticulates, i.e., fluid droplets). Both represent the

clearance of substances from the surrounding environment

through the formation of an intracellular vesicle and

subsequent delivery of entrapped materials to lysosomes.

These substances can be endogenous or exogenous, including

bacteria, viruses, damaged or effete cells, neoplastic

cells, macromolecules and colloidal materials including

particles with synthetic (latex), denatured (boiled yeast

cell walls, i.e., zymosan) and chemically modified surfaces

(aldehyde-treated red blood cells) (DeDuve et al. 1974).

However, macrophages are not attracted to viable undamaged

animal cells or encapsulated strains of pathogenic bacteria,

unless these are coated with opsonins such as antibodies or

complement (Poznansky and Juliano 1984).

The first encounter of host and foreign substance leads

to a stereotypical response causing mobilization of

phagocytic elements to the areas where the substance was

introduced. This may occur as an isolated event or as part

of the inflammatory response. Once mobilized, the phagocytic

cells mount an attack on their target by a process called

phagocytosis, a multiphasic act requiring the following

steps (Bellanti and Kadvec 1985):

(1) Recognition of material to be ingested. Circulating

monocytes are attracted to an area of injury by a number of

factors, some of which are derived from the complement

system secreted by T-lymphocytes. Here they may further

differentiate into macrophages and may be activated in a

variety of ways. Once activated, the cells display

heightened metabolic activity and enhanced function.

(2) Movement towards the material (chemotaxis). The

interaction of a particle with the plasma membrane of a

phagocytic cell results in the generation of signals that,

when relayed to the cell's interior, initiate the movement

of cytoplasm, the formation of membrane pseudopods, and the

remodeling and fusion of the membrane to form a phagocytic


(3) Adhesion and ingestion. The process of adhesion is

separate from that of ingestion (internalization) and does

not necessarily predestine a particle for ingestion.

Ingestion of an attached particle requires the sequential

interaction of membrane receptors on the phagocyte with

ligands distributed throughout the surface of the particle.

This is referred to as the "zipper" model due to the

requirement for a zipper-like interaction of cellular

receptors with particle-bound ligands. The zipper model

predicts that a particle with ligands (antibodies)

distributed over only one part of its surface will bind to

the macrophage plasma membrane but will not be ingested.

This in part depends on the activity of the macrophage. For

example, complement receptors lead to binding but not

ingestion for unstimulated unprimedd) macrophages, but

binding and ingestion occur for stimulated (inflammatory)

macrophages (Silverstein et al. 1978). In addition,

attachment has been shown to be independent of temperature

and energy expenditures, whereas ingestion is considered

highly temperature dependent and requires active cellular

metabolism (Rabinovitch 1969, Silverstein et al. 1977,

Pratten and Lloyd 1986).

(4) Intracellular digestion. Following ingestion, the

particle enters the cell in a vacuole composed of

internalized plasma membrane (phagosome), which in turn

fuses with a lysosome to form a phagolysosome where the cell

attempts to destroy (digest) the foreign particle using a

wide array of enzymes (DeDuve et al. 1974).


Phagocytosis has been described by Jacques (1970) as a

process during which a macrophage which appears to be moving

with a "flapping ruffle" at its leading edge, pushes out

towards the particle and quickly flows around it. This

process can be very rapid, with complete ingestion taking

place within minutes. In addition, in the course of 10 to 20

minutes, a phagocytic cell can internalize a quantity of

particles whose combined surface area is equivalent to 30%

to 50% of the area of the phagocyte's plasma membrane

(Silverstein et al. 1977, Mahoney et al. 1977). This can

proceed in either a specific or nonspecific manner (i.e.,

with or without the involvement of immunologically

determined ligands). Nonspecific interactions occur as a

result of physicochemical conditions rather than as a

consequence of specific epitope-antibody binding.

While in some systems phagocytosis can proceed in the

absence of protein, in general it is stimulated by its

presence in the medium. This may be related to the

adsorption of specific plasma proteins termed opsonins

(Howard and Wardlaw 1958, Saba 1970, Silverstein et al.

1978). These substances may be similar to natural

antibodies, thus different from specific immunoglobulins.

For instance, the presence of serum and divalent cations in

culture medium was found necessary for the ingestion of red

blood cells, but not for all particles (Wilkins and Bangham

1964, Carr 1973), and in the case of bacteria, specific

opsonic antibodies have been indicated. Aside from

increasing the rate of phagocytosis, it has been suggested

that opsonins in normal serum may be mediating in part the

recognition mechanism of the nonimmune animal. This may

indicate that specific receptors on macrophage membranes are

necessary to initiate phagocytosis.

Indeed, the macrophage surface has been shown to bear

more than 30 receptors, which respond to particle

opsonization by immunoglobulins (IgG), complement and other

proteins. There are two Fc portions of IgG involved,

however, the Fab fragments are devoid of such activity

(Ogmundsdottir and Weir 1980).

Complement is a series of sequentially reacting serum

proteins (Esser 1982) that may possibly be an amplifier of

phagocytosis. Complement receptor activity is governed by

the physiological state of the macrophage (Silverstein et

al. 1978). Complement receptors of unstimulated macrophages

promote binding but not ingestion of complement coated

particles; complement receptors of inflammatory macrophages

mediate binding and ingestion of complement coated


particles. There is evidence to suggest the C3b component of

complement plays a part (Lambris and Ross 1982).

Of the other proteins, there are receptors for

mannosyl/fucosyl-terminated glycoproteins (Stahl and Gordon

1982) as well as strong evidence for the involvement of

fibronectins in the opsonization process (Hsu and Juliano


There is also evidence to suggest that the stimulus to

phagocytize a particle is initiated by the particle. The

membrane response of the phagocyte to this stimulus is local

and is confined to the segment of the plasma membrane

adjacent to the particle initiating the stimulus. Membrane

components that mediate this "non-specific" uptake of

particles (e.g., polystyrene and latex particles) have been

proposed; however, they are purported to act independently

of Fc receptors and complement receptors (Hsu and Juliano


Intracellular Fate of Ingested Substances

Once ingested, the particle enters the cell in a

vacuole (phagosome) which in turn fuses with one or more

lysosomes to form a phagolysosome where the cell attempts to

destroy (digest) the foreign particle with a wide array of

lysosomal enzymes (DeDuve et al. 1974). In order to survive

in this vacuole, the ingested substance must be stable at pH

4-5 intracellularr lysosomal pH) and withstand attack by 40

or more digestive enzymes. The end product of digestion, the

residual body, is a large irregular electron dense mass

often containing laminated myelin whorls presumably

consisting of phospholipid, and sometimes ferritin from

breakdown of red blood cells.

At this point, the material may be totally digested,

may persist in the form of an indigestible residue, may fill

the cell, or, if toxic, may kill the cell. Completely

indigestible material such as carbon may persist unaltered,

while bacteria may survive and either kill the macrophage or

live in symbiosis with it (i.e., M. leprae [Allen et al.

1965] and Brucella [Karlsbad et al. 1964]). Cases where the

bacteria are dependent on an intracellular environment for

their survival and multiplication are referred to as

facultative intracellular infections. For instance, Allen et

al. (1965) showed by electron microscopy that bacilli from

murine leprosy within histiocytes of mice were neither

destroyed nor damaged by their host cells. In addition, even

in very advanced stages of infection, where huge numbers of

bacteria were present, the cytoplasmic components of the

host cells were still intact.

Thus, the final outcome of the encounter is dependent

on the properties of the substance, including size,

structure, chemical nature, and amount presented.

Factors Affecting Phaqocytosis

In vivo, several physiological processes influence the

vascular clearance and distribution of foreign substances.

These were comprehensively reviewed by Saba (1970) and

include species, blood flow, metabolic activity, status of

the cells, and particle dose. In general, species and blood

flow are not applicable to the in vitro study of

phagocytosis. However, the others, in conjunction with

certain physicochemical factors such as particle size,

surface charge, surface affinity including opsonization, and

membrane fluidity, influence the phagocytic response in vivo

and in vitro (Saba 1970, van Oss et al. 1975, Illum and

Davis 1982).

Metabolic activity. The energy sources for the cells of

the MPS are dependent upon the degree of cellular maturity,

the level of endocytic activity, and the environment.

Following formation of a vacuole, the primary metabolic

processes involved include the glycolytic pathway and the

hexose monophosphate shunt. In addition, the main energy

source for alveolar macrophages is the aerobic tricarboxylic

acid (TCA) pathway. Collectively, stimulation of these

pathways is termed respiratory burst and consists of an

increase in glycolysis, hexose monophosphate shunt, oxygen

consumption, and hydrogen peroxide and lactic acid

production. The increase in lactic acid is partly

responsible for the decreased pH in the phagosome.

Accompanying the respiratory burst is an increase in RNA and

phospholipid turnover, which is important for protein

synthesis and membrane formation. This is most prominent in

neutrophils and to a lesser extent in mononuclear

phagocytes. Both in vitro and in vivo phagocytosis are

energy dependent; however, particle adherence to the

phagocytic cell membrane is not (Oren et al. 1963).

In addition, it has been shown that particle uptake in

vitro can be inhibited by low temperature (4C) or by

inhibiting energy metabolism through the addition of NaF or

nitrophenol (Steinman et al. 1974, Pratten and Lloyd 1986).

Incubation at 4C causes an inhibition of uptake since

phagocytosis does not occur at these low temperatures, and

any uptake here would likely be due to binding.

Status of the reticuloendothelial cells (activated or

non-activated). The RES does not function solely as a

passive scavenger. It can be activated or depressed by a

variety of endogenous and exogenous factors. Activation can

occur naturally during bacterial infection, neoplastic

diseases, and diseases of autoimmunity, while reticuloendo-

thelial depression associated with circulatory failure may

be a crucial factor in the development and progression of a

disease process. Reticuloendothelial depression can also be

induced by the administration of a bacterial cell wall

component, such as muramyl dipeptide (Fidler 1986) or lipid-

containing emulsions (e.g., Intralipid") (Fischer et al.

1980). Illum and Davis (1984) have used this approach to

produce reticuloendothelial blockade resulting in a

redirection of colloids from the liver to other organs by

administering a placebo colloid or a macromolecular material

(e.g., dextran sulfate).

Particle dose. Small colloid doses removed primarily by

the liver cannot be used as an index of reticuloendothelial

activity. With increments in dose administered, there is an

associated decrease in clearance rate and a progressive

increment in relative extrahepatic colloid localization.

This may be due to progressive saturation of reticulo-

endothelial cell capacity or to a progressive saturation of

the available opsonin pool. With a large enough dose,

reticuloendothelial blockade can be induced. This condition

can persist for several days, in which no further uptake can

proceed (Saba and DiLuzio 1966, Gregoriadis and Ryman 1972a,

Gregoriadis and Neerunjum 1974, Kavet and Brain 1980). This

appears to be related to the total number and surface area

of particles administered rather than particle composition

(Abra and Hunt 1981).

Particle size. In vivo, large particles (>7 pm) will be

removed from the blood rapidly and efficiently by the

filtering propensity of the lung capillaries (Fidler et al.

1980, Davis and Illum 1986). Smaller particles that are not

removed by the lung will normally be removed by macrophages.

In relation to liposomes, the longest half life values

have been obtained with the smallest (<30 nm) liposomes

composed of solid phase neutral phospholipids. This may be

related to surface area, since on an equal weight basis of


phospholipid, small unilamellar vesicles have a much larger

total surface area and greater particle numbers compared

with those of larger multilamellar vesicles (Pidgeon and

Hunt 1981).

In vitro, the rate of particle uptake by peritoneal

macrophages has been shown to be linear over time and to

increase with particle size for polyvinylpyrrolidone

particles (Pratten and Lloyd 1986). Polyvinlypyrrolidone

particles of 100 or 1,100 nm diameter were phagocytized in a

rate-dependent manner over 2.5 hours, with the larger

particles being captured more rapidly. This effect was also

shown by Hsu and Juliano (1982) using large multilamellar

liposomes which were able to more efficiently deliver their

contents to mouse peritoneal macrophages than smaller

unilamellar vesicles, although a greater number of smaller

particles were captured.

Surface charge. Cell membranes contain large amounts of

surface carbohydrates, in particular surface sialic acid

residues which impart a negative charge to the surface

(Allen and Chonn 1987). When these sialic groups are

enzymatically removed from the cell surface, rapid uptake of

cells occurs by Kupffer cells of the liver. Likewise, when

sulfatides and gangliosides are used to impart surface

hydrophilicity and negative charge (by expression of sialic

acid groups), blood circulation times have been shown to

increase. However, the expression of a negative charge (by

incorporation of phosphatidylserine) may lead to enhanced

trapping in the lung vasculature (Fidler et al. 1980, Fidler


Juliano and Stamp (1978) found that liposomes of

negative charge due to the presence of PS were cleared more

rapidly from the circulation after intravenous

administration than positive and neutral liposomes,

regardless of size. However, positively charged and neutral

liposomes were cleared more rapidly if they were of larger

size. In addition, large positively and negatively charged

liposomes were cleared at a similar rate, which was more

rapid than for neutral particles.

In vitro, it has been consistently shown that

macrophages possess a negatively charged surface and that

interactions with positively charged particles lead to

enhanced endocytosis (Schwendener et al. 1984, Mutsaers and

Papadimitriou 1988). However, there is conflicting evidence

on the effects of the interactions of negatively charged

particles with macrophages. There were measurable

differences in the uptake rate in vitro for negatively

charged colloids, which also showed differences with respect

to rate of vascular clearance and relative hepatic

distribution in vivo. Although these particles had different

electrophoretic mobilities initially, after plasma

incubation, there was no measurable difference, presumably

due to the adsorption of plasma proteins (Wilkins and Meyers

1966, Wilkins 1967). A similar reversal of charge was seen

by Black and Gregoriadis (1976) where neutral and positive

liposomes acquired a negative charge on incubation with rat

plasma; however, the surface charge on negatively charged

liposomes was essentially unchanged. Thus, while it is

recognized that the surface of the entity (bacteria,

particles, etc.) being phagocytized will have significant

effects on its rate of removal, the relationship between the

charge in vitro and that acquired instantaneously in vivo

remains to be ascertained.

Opsonins. In addition to the uptake of soluble

proteins, macrophage surface receptors can also mediate

particle uptake. Thus coating a particle with proteins

capable of interacting with a macrophage (opsonization) can

enormously enhance the uptake of the particle. The surface

characteristics of a particle determine whether or not it

will be opsonized and by which component. As a consequence,

the mechanism of adhesion will be different depending on the

nature of the opsonic component and the particular receptor

mediated process. The effect of opsonization of liposomes

was demonstrated by Hsu and Juliano (1982), by coating

liposomes with IgG, thus promoting interaction with the

macrophage Fc receptors, and causing a 1000-fold increase in

the rate of phagocytosis. It was also demonstrated by

Sunamoto et al. (1984) that coating liposomes with

palmitoylamylopectin increased their uptake in alveolar

macrophages in vitro and their accumulation in lungs in

vivo. Similar enhancement would be expected with other types

of carriers.

The acquisition of an opsonic coat is many times an

unavoidable process, as seen with gelatin microspheres that

inevitably acquire a coating of fibronectin upon injection

into the circulation and are rapidly cleared by macrophages

(Saba et al. 1978). In addition, studies by Saba and DiLuzio

(1966) showed that in some cases diminished opsonin levels

could cause decreased particle uptake by macrophages.

Surface affinity. In some instances, phagocytosis will

occur regardless of opsonization. This may be due to the

affinity of the particle for the macrophage surface as in

the case of the lipophilic surface of triglyceride emulsions

and certain latex particles (Saba and DiLuzio 1966).

Illum and Davis (1984) reasoned that the opsonization

of particles followed by adhesion to macrophages would be

greatly modified if particles were given a hydrophilic

surface that provided a steric repulsive barrier since this

would create a high potential energy barrier and negate

short-range attractive forces, thus diminishing protein

adsorption and/or cell adhesion events. This could also

account for the increased blood circulation times seen by

Allen and Chonn (1987) when gangliosides and sulfatides were

used to impart surface hydrophilicity and negative charge

(by expression of sialic acid groups) to liposomal surfaces.

Several studies using this technique (Illum and Davis

1984, Illum et al. 1986a,b, 1987, Davis and Hansrani 1985)

showed colloid phagocytosis by mouse peritoneal macrophages

was altered in relation to surface hydrophobicity. Large

hydrophobic groups induced rapid and efficient (ca. 90%)

particle clearance, whereas large hydrophilic groups

resulted in much slower clearances. This is thought to be

due to the hydrophilic group inhibiting uptake of plasma

components (opsonization) and adhesion between the particles

and the macrophages in vivo, i.e., a steric repulsion

effect. This also led to a redistribution of particles into

other organs. For example, uncoated particles were taken up

by the liver, particles coated with Poloxamer 407 went

almost exclusively to the bone marrow, and particles coated

with Poloxamine 908 were retained in the vascular

compartment. In each case, the adsorbed layer thickness was

similar, therefore, chemical and physical differences must

have been contributing to the recognition of particle


Other polymers such as polyethylene glycols (PEG) were

found to decrease the uptake of liposomes by the RES. Blume

and Cevc (1990) were able to demonstrate approximately ten

times less uptake into cultured cells for PEG-coated

vesicles with the corresponding particle removal from blood

decreasing to only 15% of the rate characteristic for

uncoated liposomes. Whereas, Woodle et al. (1990) were able

to demonstrate approximately a three-fold decrease in the

percent of dose taken up by the RES, with approximately 20%

of the dose remaining in the bloodstream after 24 hours.

Likewise, Klibanov et al. (1990) demonstrated an increase in

the circulating half-life of liposomes in the bloodstream

from 0.5 hours to 5 hours by using PEG-conjugated

phospholipids. There are two reasons postulated for this

effect. The PEG coating may increase the hydrophilicity of

the liposome surface such that the nonspecific interaction

of liposomes with RES cells is decreased or it may

sterically prevent the coating of opsonins to the liposome,

resulting in decreased specific interactions with RES cells.

However, the exact reason for these effects is not known.

Bacterial hydrophobicity has also been related to

opsonization (Absalom 1988). Opsonization with IgG increases

the hydrophobicity of the bacterial surface, which in turn

increases phagocytic ingestion in vitro. However, with some

bacteria, i.e., Listeria monocvtogenes, which already have

maximum hydrophobicity, IgG opsonization does not increase

their hydrophobicity, but still noticeably increases


Membrane fluidity. Membrane fluidity is primarily a

concern when dealing with liposomes. The differences in the

interactions between cells and solid and fluid liposomes

were specifically addressed by Poste and Papahadjopoulos

(1978) and Margolis (1988).

The presence of cholesterol in the liposomal membrane

alters the distribution of liposomes in vivo (Senior et al.

1985). While a relationship between the uptake of liposomes

and bilayer fluidity can be postulated, it is not known

whether these effects are due to improved uptake of more

rigid particles, an affinity for cholesterol by certain

cells, or decreased opsonization. Moghimi and Patel (1987)

demonstrated that the uptake of liposomes by liver and

spleen cells in vitro was influenced by the incorporation of

cholesterol in the liposome membrane. These effects seem to

be somewhat cell specific and greatly influenced by the

presence of serum. In the absence of serum, both hepatic and

splenic cells took up cholesterol-free liposomes much more

than those containing cholesterol. The addition of serum

suppressed the uptake of the cholesterol-free liposomes by

both cell types. In hepatic cells, liposomes containing 20%

cholesterol were taken up twice as much as those containing

46% cholesterol in the absence of serum. In the presence of

serum, the uptake of 20% cholesterol liposomes was enhanced,

but that of 46% cholesterol liposomes was suppressed. In

contrast, in splenic cells 46% cholesterol liposomes were

taken up twice as much as 20% cholesterol liposomes. The

serum effects may be related to opsonins present in serum

specific for liver and spleen phagocytes, since these same

studies showed more serum proteins associated with 20%

versus 46% cholesterol liposomes by electrophoresis. It was

speculated that this effect was most likely due to the

decreased interaction with high density lipoproteins. This

may be due to increased membrane rigidity, thus increased

stability of the liposomes with higher cholesterol content.

This is consistent with later studies by Moghimi and Patel

(1987 and 1989) where liver and splenic cells showed less

affinity for liposomes composed of sphingomyelin and

saturated phospholipids, all of which would be in a solid

phase at the temperature of study compared to

phosphatidylcholine. However, in the presence of serum, the

uptake was decreased for the liver cells, but increased for

the splenic cells.

Thus, any of these factors could be used to alter the

uptake of particles by macrophages. In addition,

optimization of several factors, such as surface sialic acid

and bilayer viscosity properties, has been shown to have a

synergistic effect in increasing circulation times of

liposomes (Allen and Chonn 1987).

Most research efforts in this area have centered on the

mechanisms by which this process could be inhibited or

circumvented for the delivery of therapeutic agents in the

form of colloidal particles. However, as mentioned above, in

some cases it may be desirable to deliver the particle to

the cells or tissues where the cells are localized.

Properties of Phagocvtic Cells

Although phagocytic cells are functionally similar,

there is no reason to assume they are identical. Laskin et

al. (1988) found Kupffer cells and peritoneal macrophages to

be functionally and biochemically distinct, although both

displayed the morphologic and histochemical characteristics

of mononuclear phagocytes. While both phagocytize in a time

dependent manner, Kupffer cells were 2 to 3 times more

phagocytic. Upon stimulation, peritoneal macrophages release

greater amounts of superoxide anion and hydrogen peroxide.

In addition, differences in protein production by these

cells were observed.

Some other differences seen between macrophages include

the response to chemotactic (movement initiation) factors.

Both rat (Laskin et al. 1988) and mouse (Hashimoto et al.

1984) peritoneal macrophages responded to the chemotactic

factors complement C5 and phorbol ester tumor promoter.

Whereas, it had previously been seen by Ward (1968) that

alveolar macrophages are much more responsive to chemotactic

factors than peritoneal macrophages. In many cases, these

differences arise due to the tissue of origin and

specialization of the macrophage cell type. This is

particularly important when discussing the energy

requirements for phagocytosis.

The energy sources for the cells of the mononuclear

phagocyte system are dependent upon the degree of cellular

maturity, the level of endocytic activity, and the

environment. Following formation of a vacuole, the primary

metabolic processes involved are the glycolytic pathway and

the hexose monophosphate shunt. However, the main source of

energy for alveolar macrophages is the aerobic TCA pathway.

This is accompanied by a burst of respiration, which is much

greater in alveolar macrophages than in polymorphs or

peritoneal macrophages (Carr 1973). It has been shown by

Oren et al. (1963) that alveolar macrophages differ from

both polymorphonucleocytes and monocytes in that the resting

respiration of alveolar macrophages is much higher and

phagocytosis causes only a small increase in oxygen uptake

and glucose metabolism. Monocytes and polymorphonucleocytes

phagocytize efficiently without aerobic metabolism, but

require glycolysis for the process. Inhibition of glycolysis

in this case can block particle uptake. With alveolar

macrophages, interference with aerobic metabolism or

oxidative phosphorylation can depress particle uptake, which

is depressed even further if glycolysis is inhibited. In

addition, alveolar macrophage respiration can be stimulated

in culture by the presence of serum and glucose. This is

contrary to the depression of respiration seen in

polymorphonucleocytes in the presence of glucose.

Alveolar Macrophaqes (AM)

Of particular interest here are cells found in the

pulmonary tract. In pulmonary tissue, cells which resemble

macrophages elsewhere may be found in interstitial

connective tissue of the alveolar wall, forming part of the

lining epithelium of the alveolus, the great alveolar cells

of Sorokin, and free in the lumen of the alveolus (Carr

1973). These alveolar macrophages belong to the group of

free macrophages which are scattered diffusely throughout

the mammalian body that also includes macrophages of

connective tissue or histiocytes, macrophages of serosal

sacs, and macrophages of inflammatory exudates. Pulmonary AM

are derived from circulating monocytes originating from bone

marrow (van oud Alblas and van Furth 1979), as are

peritoneal macrophages and Kupffer cells. This influx of

monocytes is a steady state process, responsible for cell

renewal in the alveoli. They are considered end cells, as

they do not divide appreciably upon arrival (Myrvik and

Kohlweiss 1980). Most AM leave the alveoli by way of the

airways and are expelled via the mucociliary pathway, with a

mean turnover time around 27 days.

Particulates less than 10 pm can penetrate the airways

in varying degrees and those in the range of 1 to 2 um can

enter the respiratory space without difficulty. It is

probable that up to 50% of airborne particles in this size

range reach the alveoli. Particles that reach the terminal

airways and alveoli are destined to be phagocytosed by

alveolar macrophages (Green and Kass 1964, Green 1970).

Once phagocytized, normal macrophages are quite

efficient in killing many avirulent organisms as well as

some microorganisms of relatively low virulence. In

contrast, normal alveolar macrophages are totally

incompetent to handle highly virulent intracellular

parasites like Mvcobacterium tuberculosis and Francisella

tularensis (Myrvik and Kohlweiss 1980). Only immunologically

activated macrophages which are mobilized and activated by

the mechanisms of cell-mediated immunity can successfully

cope with these virulent organisms.

Macrophaqe Culture

Perhaps the best way to obtain a pure culture of

macrophages is by washing out a serosal cavity, usually the

peritoneum (Carr 1973). These macrophages will be either

resident or elicited, if they normally populate a given

anatomic site in an untreated animal, or have their

production stimulated (Jacques 1970), respectively. This

process may affect the surface of the macrophages. It has

been shown by Silva Filho et al. (1987) that resident,

elicited, and activated mouse peritoneal macrophages all

possess a negative charge, however, the charge is greatest

when the macrophages are activated.

Since cells of the monocyte-macrophage lineage are

capable of rapid firm adherence to solid surfaces, which

enables the enrichment of the macrophage population to

produce macrophage rich monolayers by adherence to plastic

or glass as first described by Mosier (1967) and later

reviewed by Pennline (1981). This procedure allows for the

separation of mixed-cell suspensions into adherent

(macrophage-rich) and non-adherent (lymphocyte rich)

populations. Other cells present either degenerate rapidly,

as do the polymorphonuclear cells, or fail to stick to the

glass and are washed off with the first change of medium, as

are the lymphocytes. The adherent fraction is heterogeneous

and has been shown to contain populations of cells that

differ in their ability to bind antigen, form rosettes with

antibody-coated erythrocytes, and kill tumor cells.

Functional and morphological differences have been described

between adherent populations isolated from sources

containing activated, elicited, or resident macrophages as

well as within a given population between poorly adherent

and strongly adherent cells.

The procurement of macrophages from the lung is

especially useful when a highly purified population of

macrophages is required to study morphological, metabolic

and functional parameters. When macrophages were recovered

from rabbit lungs, 90% to 95% of the total cells recovered

were macrophages with 85% to 95% cell viability (McGee and

Myrvik 1981). The same was found for AM of mice, rats and

guinea pigs (Oren et al. 1963, Kirkawa and Roneda 1974, van

oud Alblas and van Furth 1979). The lower respiratory tract

in a healthy individual is normally sterile. It has been


reported that >98% of the cells obtained via tracheal lavage

are pulmonary alveolar macrophages with 1% to 2% lymphocytes

and granulocytes as contaminants (Coggle and Tarling 1984).

Almost all macrophages isolated by lavage are viable;

morphological, cytochemical and functional studies can be

performed as usual for comparison to other mononuclear

phagocytes. In vitro, AM exhibit rapid spreading and high

mitotic rate, and their cytoplasm is notably pyrominiphilic

(Carr 1973). Pulmonary macrophages are positive for esterase

and negative for peroxidase, carry Fc receptors and show

avid phagocytosis and pinocytosis. However, mature AM in

lavage fluid rarely carry C3 receptors (van oud Alblas and

van Furth 1979).

Lung Surfactant

Mammalian lung stability requires the presence of a

tensioactive material at the air/water interface, able to

support very high surface pressures on dynamic compression,

to prevent alveolar collapse at the end of expiration

(Clements et al. 1961, Schwick et al. 1982, Keough 1985,

Dobbs 1989). Not only do these structures not collapse due

to surface tension, they also can undergo appreciable

expansion and contraction. The surface forces regulating the

mechanics and stability of lung alveoli are modulated by the

presence of amphipathic substances at the air-water

interface. This material is known as pulmonary surfactant


and has been well characterized, particularly in connection

with infant respiratory distress syndrome (Ivey et al. 1977,

Fujiwara et al. 1980, Haagsman and van Golde 1985, Jobe and

Ikegami 1987, Robertson and Lachmann 1988).

This surface-active material is a complex of lipids and

protein, as can be seen in Table 1. Dipalmitoylphosphatidyl-

choline (DPPC) accounts for -60% of the phospholipid

fraction, the other 40% being unsaturated phosphatidyl-

choline (PC), and about 5% neutral lipid which is mostly

cholesterol (Keough 1985).


Composition of Pulmonary Surfactant

Composition by Weight (%)

Phospholipids 85

Saturated phosphatidylcholine 60
Unsaturated phosphatidylcholine 20
Phosphatidylglycerol 8
Phosphatidylinositol 2
Phosphatidylethanolamine 5
Sphingomyelin 2
Others 3

Neutral lipids and cholesterol 5

Proteins 10

Contaminating serum proteins 8
Sp-A: 32 to 36,000 kDa -1
Lipophilic proteins: 6 to 18,000 kDa -1

from Jobe and Ikegami (1987)

When Chapman (1975) analyzed the lung lipids of 7

animal species, all contained considerable quantities of a

fully saturated palmitate lipid. Around 50% of these were

disaturated like DPPC and rigid at 370C, which he postulated

to be due to the aerobic environment in the lungs, and the

nature of these phospholipids to be less readily oxidized

than their unsaturated counterparts.

Recent studies of surfactant metabolism in vivo

indicate alveolar surfactant phospholipids are in a complex

dynamic equilibrium with the intracellular surfactant pool

of alveolar type II cells. Physiologically, type II

pneumocytes synthesize the phospholipid and concentrate it

in the form of lamellar bodies which are excreted in the

aqueous alveolar subphase by exocytosis where the

phospholipid becomes part of lung surfactant. The lamellar

bodies undergo transformation into other characteristic

structures such as tubular myelin, certainly responsible for

the very fast adsorption rate at the air/water interface of

the surfactant (Keough 1985). This surfactant is then

believed to be recycled by the alveolar macrophages. For

instance, in three-day-old rabbits, the majority of the

approximately 5 Vmole phosphatidylcholine pool seems to be

recycled back into type II cells and subsequently resecreted

with a turnover time around 10 hours (Jacobs et al. 1982).

Therefore, this system can be considered as a process of

synthesis, storage and secretion of surfactant components

(Goerke 1974).

In addition to phospholipids, pulmonary surfactant

contains a protein component composed of three specific

surfactant apoproteins: SP-A is a higher MW glycoprotein

(35,000), relatively hydrophilic and water soluble; and two

low MW apoproteins SP-B (8,000) and SP-C (5,000) which are

extremely hydrophobic (Hawgood 1989). Metabolic experiments

show the larger protein to be excreted into the alveolar

lumen at the same rate as the lipids of pulmonary


DPPC is considered the major component responsible for

pulmonary surfactant stability and surface activity (King

and Clements 1971). However, clinical data indicate that

phosphatidylglycerol (PG) also plays a key role (Hallman et

al. 1977). In addition, various studies have examined the

role of the protein components. King and MacBeth (1979)

found >71% binding of apoprotein to DPPC at room

temperature, which greatly enhanced the adsorption

(spreading) of DPPC at an air/water interface. The

apoprotein presence did not alter the capacity of DPPC to

lower the surface tension below 5 dynes/cm. It has also been

shown that synthetic phospholipid mixtures with SP-B and SP-

C alone or in combination adsorb rapidly to an air-liquid

surface and lower surface tension during prolonged dynamic

compression (Holm et al. 1990, Venkitaramam et al. 1991).

Immunoglobulin studies (LaForce 1976) indicate the

presence of both IgG and IgA in lavage supernatant, but


neither are found in the extracted lipid fraction. These may

specifically be involved in the opsonization of particles

and therefore important for the recognition and uptake of

particles by AM.

In addition, LaForce et al. (1973), LaForce (1976) and

Juers et al. (1976) have shown alveolar lining material

(ALM), the proposed pulmonary surfactant fraction extracted

from bronchoalveolar lavage fluid, to be important in the in

vitro bactericidal capacity of AM.

It was the similarities between the composition of

pulmonary surfactant and a potential drug delivery system

(liposomes) that prompted us to study them in association

with the phagocytic ability of AM.

Background Liposomes


Phospholipid molecules consist of a glycerol backbone

which is esterified with two fatty acid chains, and a

phosphate group esterified with an aminoalcohol, alcohol or

carbohydrate. The molecules are clearly amphiphilic with a

hydrophilic "head group" and a hydrophobic "tail". Many

early studies discussing the physicochemical properties of

phospholipids, primarily organized as a monolayer at an air

water interface were reviewed by Bangham (1968).

There are two types of phospholipids, -- glycolipids

and sphingolipids-- the protypes being phosphatidylcholine

and sphingomyelin, respectively (Figure 1). Although both


Water-soluble head

CH 2 -0-

2 CH

Water-insoluble tail

X= -CH2- CH -N-(CH ) OH OH

choline OH

I + inositol
sere -CH-CH CH OH
serine 2 2

+ OH
2 2 3 glycerol

Figure 1: Structural representation of phospholipid (A) and
sphingolipid (B). The circled area (top left)
represents the hydrophilic headgroup with possible
X substituents (bottom box), and the shaded area
(right) represents the hydrophobic, fatty acid
portion of the molecule, with m and n denoting the
length of the hydrocarbon chain.


Water-soluble head

Water-insoluble tail




Figure 1: (continued)

groups contain the characteristic phosphorylcholine head

group, they are distinctly different in their hydrophobic

tails. Whereas the phosphatidylcholines have two long chain

hydrocarbons of nearly equal length attached to carbon two

and three of the connecting glycerol backbone, sphingomyelin

has an acyl chain of 16 to 24 carbons attached to the second

carbon linked by an amide bond, and a paraffinic residue of

the sphingosine base, which contributes only 13 to 15 carbon

atoms to the nonpolar region. Synthetic derivatives of these

phospholipids can be made with varying degrees of saturation

and chain lengths.

Phospholipids belong to a class of lipid compounds

including surfactants, which are known to undergo lyotropic

mesomorphism, or they exist in various hydration states

called liquid crystals. According to the classification of

Small (1968), phospholipids belong to the Class II:

Insoluble Swelling Amphiphiles. They are insoluble in water,

but water can penetrate the hydrophilic phospholipid

headgroups. Due to their amphipathic nature, they can exist

in different types of liquid crystalline organization upon

hydration (Luzzati et al. 1968). The form occupied

uponhydration, depends on the physical structure of the

molecule, including micelles, bilayers and hexagonal arrays

(Figure 2) (Cullis and DeKrujff 1979). If the molecule has a

single hydrophobic tail (i.e. lysophosphatidylcholine) or

the headgroup is small in relation to the tails, it will

Molecular shape





stacked cylinder interdigitated bilayer


Phosphatidyl- /D
ethanolamine /

(pH < 4.0)

hexagonal ( )

Figure 2: Phospholipids with corresponding dynamic molecular
shapes and polymorphic phases; (A) micellar, (B)
bilayer, (C) interdigitated, (D) hexagonal.




assume the structure of a micelle (Figure 2A), if there are

two chains of similar length in balance with the headgroup

(i.e. PC) a lamellar structure will result (Figure 2B), if

the chains are of sufficiently different length (i.e.

sphingomyelin (SPM)) interdigitation of the bilayer (Figure

2C) can occur, and if the headgroup is large in relation to

the tails (i.e. phosphatidylethanolamine (PE)) a hexagonal

phase (Figure 2D) will form. In addition, bilayer to

hexagonal phase transitions can be induced by binding of

divalent cations to acidic phospholipids such as

phosphatidylserine (PS) (Harlos and Eibl 1980, reviewed by

Crowe and Crowe 1984).

When bilayer forming phospholipids swell upon contact

with water they spontaneously rearrange to form closed

concentric bilayers of phospholipid enclosing an aqueous

space (Figure 3) with the hydrophobic tails opposing each

other and the hydrophilic head groups facing the surrounding

aqueous bulk phase. This derives from the opposing

hydrophobic interactions of the hydrocarbon tails that force

lipids to assemble and hydrophilic tendencies of the

headgroups seeking to be solvated by water. For pure lipid

molecules, this leads to the formation of an extended

bilayer that must close and form a vesicle to prevent

exposure of the hydrocarbon chains to water at the edges of

the bilayer. Since the molecules are essentially insoluble

and do not form micelles, there is no equilibrium between

SAqueous space

Phospholipid headgroups

Lipophilic tails

Figure 3: Schematic diagram of a multilamellar liposome, an
"onion skin" configuration with concentric lipid
bilayers separated by aqueous spaces, surrounding
an aqueous core.

individually dissolved molecules, micellar aggregates and

bilayers (Lasic 1988).


The early characterizations of phospholipid vesicles

were done by Bangham and Horne in 1964, who observed that

purified phospholipid mixtures extracted from biological

membranes, spontaneously swell in aqueous salt solutions to

form liquid crystals. These lamellar structures were shown

to be able to entrap ions in water and release them at

variable rates, suggesting that each consisted of completely

closed phospholipid bilayers forming selective permeability

barriers (Bangham et al. 1965b). In 1968 these dilute

phospholipid dispersions were termed "liposomes" by Sessa

and Weissmann.

Bangham's group characterized the diffusion of water

(Bangham et al. 1967) and ions (Bangham et al. 1965 a,b)

across these multilamellar vesicles. They were found to be

osmotically sensitive, with permeability characteristics for

simple univalent cations, anions, and water, qualitatively

similar to those occurring across biological membranes; i.e.

freely permeable to water, Cl-, and I- but less so to F-,

NO3_, SO42- and HP042-. In particular, Cl- and I-, are by

several orders of magnitude more free to diffuse than simple

univalent cations and uncharged molecules. Likewise, cations

do not diffuse through positively charged multilamellar

vesicles whereas anions diffuse freely. Due to their osmotic


sensitivity, these bilayers are predicted to be impermeable

to small polar solutes (Bangham et al. 1972).

There are two basic types of lipid dispersions:

multilamellar vesicles (MLV) and small unilamellar vesicles

(SUV). MLV, first described by Bangham and Horne (1964),

consist of vesicles in which phospholipid is organized into

concentric bimolecular lamellae, each separated from its

neighbor by an interspersed water lamella. Each lamella is a

topologically closed surface, with a single liposome

composed of many concentric vesicles. A typical aqueous

dispersion of this type contains multilamellar liposomes

which vary in size in the micron range. The second type, SUV

or small unilamellar vesicles, described by Huang (1969),

are spherical vesicles, homogeneous in size, consisting of a

single continuous lipid bilayer enclosing a volume of

aqueous solution. These have been shown to be metastable

with time, undergoing fusion into larger unilamellar

vesicles, or into MLV (Suurkuursk et al. 1976, Lasic 1988).

Through the years, many methods have been established

to prepare liposomes (Szoka and Papahadjopoulos 1980, Martin

1990, New 1990). The original method described by Bangham et

al. (1965b) is also the simplest and involves the deposition

of a phospholipid film after evaporation of organic solvent,

phospholipid film reconstitution with an aqueous buffer,

then gentle shaking until the lipid is hydrated and has

formed MLV (Figure 4). Sonication or extrusion can be used

to obtain a smaller and/or more uniform preparation.

Sonication is difficult to standardize (Huang 1969),

whereas, pressure extrusion through polycarbonate filters

with a defined pore size is easy to employ and gives

reproducible size distribution (Olson et al. 1979).

Encapsulated water soluble compounds can be separated from

non-encapsulated by dialysis, gel filtration or

centrifugation (Olson et al. 1979).

The type of phospholipid used in liposome preparation

can influence the intracellular volume and hence, the

entrapped volume of MLV. Charge repulsion via incorporation

of a charged phospholipid such as phosphatidylglycerol in

the liposome membrane is a useful means for increasing the

encapsulated volume due to a wider spacing distance between

bilayers (Bangham 1968, Gregoriadis et al. 1977). This can

also improve the physical stability of the liposome due to

reduction of aggregation and fusion. In addition, lipophilic

drugs can be anchored into the liposomal lipid phase

(Gregoriadis 1973, Juliano and Stamp 1978).

The type of phospholipid used also has a great effect

on liposome stability. In particular, incorporation of 30 to

50 mol% cholesterol has been shown to stabilize bilayers of

unsaturated phospholipids (Ladbrooke et al. 1968). Certain

phospholipids are inherently unstable due to their

propensity to undergo oxidation (ie., unsaturated PC).

a o

Thin Lipid Film

Dispersion in Aqueous

Figure 4: Preparation of multilamellar vesicles, and sizing
by extrusion through a polycarbonate membrane.


Incorporation of an antioxidant or the use of fully

saturated phospholipids can inhibit this effect (Allen


In vivo, liposomes are susceptible to release of

encapsulated materials when they come in contact with plasma

components. This happens in a non-linear fashion presumably

due to a rearrangement of membrane lipids and adsorbed

proteins to form their most stable configuration (Hunt

1982). According to Damen et al. (1980, 1981), both high

density lipoprotein and non-lipoprotein components of human

and rat plasma cooperate in the destructive action of plasma

on phosphatidylcholine liposomes.

In general, the permeability of liposomes to entrapped

solutes has also been shown to increase when they interact

with various cell types (Szoka et al. 1979, van Renswoude et

al. 1979, van Renswoude and Hoekstra 1981, Hoekstra et al.

1981, Margolis et al. 1982).

Therefore, the success of liposomes as vehicles for the

targeted delivery of specific drugs depends on their

compatibility with the encapsulated drug, their stability

under physiological conditions and their ability to interact

with specific target sites. This will depend on the types

and amounts of phospholipids used for the preparation and

will require the characterization of a wide variety of

phospholipid mixtures in conjunction with biologically

relevant ions and molecules.

Physicochemical Parameters of Liposomes

Thermotropic Phase Transitions

In addition to lyotropic mesomorphism, phospholipids

also undergo thermotropic mesomorphism or temperature

induced phase changes. This is similar to the thermotropic

behavior of the soaps. When a soap is heated to a certain

temperature (sometimes several hundreds of degrees below the

final melting point), the hydrocarbon chain portion "melts",

the all planar trans configuration breaks up, and the chains

now contain gauche isomers (Chapman 1958). Phospholipids

exhibit this same property. The capillary melting points

range from 2000C for phosphatidylethanolamine to 2300C for

phosphatidylcholine. However, in addition to the capillary

melting point, a temperature dependent endothermic

transition occurs around 1200C and again at ca. 1350C (Byrne

and Chapman 1964, Chapman and Collin 1965) (Table 2). This

transition does not involve the change of state from solid

crystal to "normal" lipid usually implied by the term

"melting", but rather a shift from a crystalline gel to a

"liquid crystal". In other words, above the transition

temperature, the lipid chains are "melted and fluid",whereas

below the transition temperature the chains are organized in

a crystalline manner (Chapman and Salsbury 1966).

One of the first to characterize the gel to liquid

crystal transition (Tc) in phospholipids was Chapman (1968)

who related the temperature of this phase change to a


Phase Transitions of Anhydrous Phospholipids

Phospholipid Liquid Crystal ('C) Melting Point ('C)

DMPC 105 236-237

DPPC 90 234-237

DSPC 90 219-234

DSPS 120 159-160

DMPE 86 195-207

DPPE 135 175-185

DSPE -- 170-182

DSPC -- 190-198

C16 180-190 209-211
C18 209-210
C2 213-216
Natural from brain 165-175 205-207

from Dervichian (1964) and references therein.

dependence on phospholipid chain length and degree of

saturation. The phase transition was found to be primarily

concerned with the hydrocarbon chains since the space taken

up by glycerol and the polar group remained essentially

unchanged when the phase transition occurred (Chapman et al.

1967). The glycerol and polar groups retain a fairly regular

organization although the polar group does have considerable

mobility, but the fatty acid chains "melt" and acquire

considerably more mobility, with the methyl end of the chain

having greatest motion (Chapman and Salsbury 1966). Only one

main "melting" occurs even when there are two different

types of chain present in the phospholipid. Thus, near the

endothermic T,, a given phospholipid can be in a highly

mobile condition with its hydrocarbon chains flexing and


Temperatures at which phase transitions occur are

dependent upon the headgroup, the hydrocarbon chain length

and the degree and type of unsaturation present (Chapman

1968). For the same headgroup and extent of hydration,

lipids with more unsaturated, branched or bulky acyl chains

have lower transition temperatures than more saturated ones

(Ladbrooke et al. 1968). Thus, the transition temperature is

high for fully saturated long chain phospholipids, lower

when there is a trans double bond in one of the chains, and

lower still when there is a cis double bond present (Chapman

et al. 1966).

The temperature of the transition increases with

increasing phospholipid chain length (Chapman et al. 1967).

In phospholipids with the same acyl chain in both positions,

To increases ca. 14-170C with every 2 methylene unit

increase in chain length. With mixtures of phospholipids, T,

occurs over a wider temperature range compared to that of a

single phospholipid.

Effects of Water on Phospholipid Phase Transitions

Small amounts of water have profound effects upon the

mesomorphic behavior of phospholipids causing the appearance

of additional liquid crystalline forms between the first

transition temperature and the capillary melting point

(Chapman et al. 1967). As the amount of water increases, the

transition temperature is decreased until it reaches a

limiting value corresponding to the maximum uptake of bound

water by the phospholipid gel (20% or more water by weight)

(Ladbrooke and Chapman 1969). In this excess of water, the

hydrated phospholipid transforms from a gel phase to a

smectic mesophase or liquid crystal at temperatures above

this transition temperature (Ladbrooke and Chapman 1969,

Ladbrooke et al. 1968) called the gel to liquid crystalline


In the ordered gel state below a characteristic

temperature T,, the hydrocarbon chains are in an all-trans

configuration. On increasing temperature, an endothermic

phase transition occurs during which there is a trans-gauche

rotational isomerization along the chains resulting in a

lateral expansion and a decrease in the thickness of the

bilayer (Wilkinson and Nagale 1981).

This transition temperature occurs at a temperature

characteristic of both the nature (chain length and

saturation) of the hydrocarbon chains (Table 3) and the

polar head group (Table 4) of the molecule and can range

from approximately -20C to about 600C (refer to Szoka and

Papahadjopoulos 1980 for a list of phospholipids with charge

and To). If water is added to DPPC in the anhydrous state,


Effect of Chain Length on Tc of DPPC

Phospholipid Tc (OC) Reference
pH 7.0

1,2 (12:0) -1.1 Marsh 1990

1,2 (14:0) 23.5

1,2 (16:0) 41.4

1,2 (18:0) 55.1

1,2 (20:0) 66.0

1,2 (22:0) 75.0


Effect of Phospholipid Headgroup on Tc

Phospholipid Tc (OC) Reference
pH 7.0

DPPC 41.4 Wilkenson and
Nagele (1981)

DPPG 41.5 Findlay and
Barton (1978)

DPPS 54.0 Cevc et al.

to a degree approximating physiologic conditions, the T,

decreases sharply from 100C to ca. 420C. Thus < 420C, the

lipid is in a gel state, while > 42C it is in a fluid

state. On the other hand, unsaturated phosphatidylcholine in

the presence of water undergoes a transition around -200C.

Thus at biological temperatures, phospholipids with highly

unsaturated chains can be expected to be in a highly mobile

and fluid state.

Multilamellar dispersions of homogeneous phospholipids

exhibit two reversible phase transitions: the major chain

melting transition characterized by a sharp symmetric first-

order endothermic transition, and an additional broader

transition of lower enthalpy occurring 5* to 10C below the

major transition. The distance between the 2 transitions

decreases with increasing chain length. Structural changes

accompanying the pretransition are unclear. The

pretransition of DPPC has been associated with the motion of

the phospholipid polar headgroups by Ladbrooke and Chapman

(1969), the co-operative movement of the rigid acyl side

chains in a transition between crystal forms below their

melting temperature (Hinz and Sturtevant 1972a,b) and, as a

third possibility, tilting of the hydrocarbon chains before

melting (Chapman et al. 1974). Interestingly enough, the

pretransition is negated when mixtures of phospholipids are


Differential Scanning Calorimetry

Throughout the years, many methods have been used to

study the phase transition of phospholipids (Chapman et al.

1966). The first thermal analysis was done by differential

thermal analysis (Chapman and Collin 1965) where it was

shown that a marked endothermic reaction (heat absorption)

occurs at the gel to liquid crystal transition. However, T,

can most readily be measured in a calorimeter which gives

sensitive readings on heat absorption by a preparation.

A useful tool in the study of these thermally induced

phase transitions of lipid bilayers and biological membranes

is differential scanning calorimetry (DSC) (Ladbrooke and

Chapman 1969, Mabrey and Sturtevant 1978, Melchior and Steim


Differential scanning calorimetry (DSC) of

phospholipids is demonstrated by a reversible two-state

process on heating: A <---> B

with an equilibrium constant:


where AA and A. are the activities of A and B, respectively.

This varies with temperature according to the van't Hoff


(61nK) = H
( 6T ) p RT2

where HVH is the standard enthalpy change of the process.

Thus, if the scan rate is low enough so that the thermal and

chemical equilibria are maintained, the reaction must

proceed in an endothermic direction when the temperature is

raised, and exothermic peaks will be observed only if the

process is kinetically rather than thermodynamically limited

(Mabrey-Gaud 1981). Since the shift from gel to liquid

crystal is endothermic (i.e. relatively large quantities of

heat must be absorbed in order to break the bonds that hold

the molecules in the rigid gel structure), a sharp peak in

heat absorption at a given temperature signals the change of

state that in fact has occurred (Chapman 1975) (Figure 5).

The DSC apparatus consists of two cells: one for the

sample and the other for an inert reference material, which

can be heated at a programmed rate controlled to maintain

zero temperature difference between the cells. If the sample

is in solution or suspension, the reference material will be

the corresponding solvent. When a thermally initiated

process takes place in the sample cell the control system

responds by supplying either more or less heat to the sample

cell to hold its temperature equal to the reference cell.

Data output, of either excess heat or the corresponding

excess power, is presented as a function of temperature

(Mabrey and Sturtevant 1978).

The point of departure from the baseline, the onset

temperature as seen in Figure 5, is normally taken as the

temperature of transition. Since the peak maximum

corresponds to the temperature at which change is occurring

at the maximum rate and the point at which the curve returns

A: HEPC heating thermogram

3.75 Peak from, 49.53
toa 58.28
Onset- 51. 17 Peak- 52.88
J/gm.- 89. 41



35.00 40.00 45.00 50.00 55.00

Temperature (C)

Figure 5: Representative endothermic phase transition for
heating (A) and cooling (B) of hydrogenated egg

B: HEPC cooling thermogram

Paok from 43.C2
to& 51.97
Onsat- 49. 82
J/m. --72.49

49. 08

35.00 40.00 45.00 50.00 55.00

Temperature (C)

Figure 5: (continued)


to the baseline is influenced by instrumental factors,

neither of these points are useful in determining the

temperature range of the transition. In order to distinguish

between an isothermal transition and a change taking place

over a finite temperature range, the transition region can

be defined by the difference between onset temperature on

heating and cooling. For a pure lecithin this difference

should be less than one degree celsius, indicating a truly

reversible isothermal transition (Phillips et al. 1970).

Ladbrooke and Chapman (1969), showed the importance of

water on the mesomorphic behavior of phospholipids and

emphasized the need to maintain a constant amount of water

in samples for thermal analysis. However, when the lipid

contains 50% weight water or more, the phase transitions and

thermal spectrum are insensitive to variations in the actual

water content of the phospholipid.

Factors Affecting Thermotropic Transitions of Phospholipids

The thermogram can also be affected by pH, divalent

ions, and vesicular structure (packing constraints of the

acyl chains) (Szoka and Papahadjopoulos 1980). This is

particularly true for acidic phospholipids.

Effect of pH. The phase behavior of phosphatidylcholine

is little affected by pH between 3 and 13, corresponding

with the respective pKa's for the ionizable groups

(Kimelberg and Mayhew 1978). In contrast to

phosphatidylcholine, the phase behavior of phospholipids


with glycerol or serine headgroups is pH sensitive and thus

sensitive to the charge on the polar headgroup (Table 5).

This has been ascribed to a decrease in the repulsive forces

between adjacent negatively charged phosphate groups

(Jacobson and Papahadjopoulos 1975, Findlay and Barton 1978,

van Dijck et al. 1978).


Effect of pH on Tc of DPPG and DPPS














Cevc et al. (1980)
Watts et al. (1978)

van Dijck et al.
Jacobson and

Cevc et al. (1981)

Cevc et al. (1981)

Cevc et al. (1981)

Effect of ions. Generally, the

phospholipids is little affected by

transition of neutral

monovalent cations

including sodium and potassium even at concentrations of 1

M. The presence of 1 M magnesium increases the transition

temperature of DPPC slightly, while calcium concentrations


greater than 10 mM result in substantial increases in the

pretransition temperature, and a less marked increase in the

main transition, until the two transitions merge at calcium

concentrations > 250 mM (Chapman et al. 1974, Chapman et al

1977, Graddick et al. 1979). The T, of DPPC only shifts

slightly from 390 to 450C in the presence of calcium

concentrations from 10-' to 1 M (Simon et al. 1975).

However, it has been shown that Ca2+ concentrations

exceeding a 1:2 molar ratio of Ca2+ to DPPC can cause the

transition peak to split into two distinguishable components

(Ganesan et al. 1982). In addition, Castelli et al. (1990)

observed a shift to a higher melting point (60C) complex in

excess Ca2+. Thus, the presence of divalent cations seems to

have a greater effect on the pretransition, and little

effect on the main transition except at very high


The transition of glycerol and serine phospholipids is

sensitive to the presence of divalent cations. MacDonald et

al. (1976) and Verkleij et al. (1974) studied the effects of

Ca2 and Mg;2 on the thermotropic behavior of the negative

phospholipids, dipalmitoylphosphatidylserine (DPPS) and

dilaurylphosphatidylglycerol (DLPG), respectively. Small

increases in the ratio of divalent cation to lipid can lead

to small increases in T. for DLPG and DPPS until the ratio

exceeds 1:2, cation to phospholipid, when T, rather abruptly

increases to approximately 500C.

Vesicular structure. In addition to phase separations

with negatively charged phospholipids, the addition of

cations to DPPS or DPPG can induce a change in vesicular

structure from a bilayer to a hexagonal phase. This change

in state can also produce changes in T, as seen in Table 6.


Effect of Vesicular Structure on T,

Phospholipid Conditions T, (*C) Reference

DMPC pH 7.0 23

1M CaCl2 81 Harlos and Eibl
pH 4.6 1980

DPPS pH 7.0 54

>1.5M LiC1, 90 Seddon et al.
pH 7.5 1984

Thermal Transitions in Phospholipid Mixtures
and Phase Diagrams

The phase behavior of mixed phospholipid and water

systems was first reported by Ladbrooke and Chapman (1969)

and was studied in more detail by Phillips et al. (1970)

using calorimetric methods. In contrast to pure

phospholipids which exhibit sharp, highly cooperative phase

transitions, mixtures of phospholipids containing different

hydrocarbon chains melt over a much broader temperature

range. The shape of the transition as well as its position

is dependent on lipid composition, and a significant


asymmetry is apparent at compositions other than equimolar

(Mabrey and Sturtevant 1976).

The equilibrium between two component lipids in the

same two-dimensional bilayer plane at various temperatures

are frequently described in terms of a phase diagram. Such

diagrams be can constructed from the phase transition curves

of binary lipid mixtures obtained using high sensitivity

DSC. The phase diagram is constructed based on the onset

temperature of the cooling and heating curves (Chapman et

al. 1974) of a series of phase transition curves for the two

component mixtures at various molar ratios, generally

plotted as a function of the relative concentration of the

higher melting component. However, the onset and completion

temperatures can also give information on the miscibility of

phospholipid mixtures as described by Mabrey and Sturtevant


The simplest mixing behavior of binary lipid mixtures

in the bilayer is exhibited by the isomorphous system in

which the two component lipids are completely miscible over

the entire composition range in both the gel and liquid

crystalline phases (Mabrey and Sturtevant 1978, Melchior and

Steim 1979). Based on the onset of cooling and heating, the

phase diagram for this type of mixture gives rise to a

cigar-like enclosed region (Figure 6) with three

identifiable regions. The area above the upper curve (i.e.

the onset temperature of heating), represents the point

above which the phospholipid is in a more fluid state called

"liquidus"; the area below the lower curve (i.e., the onset

temperature of cooling) represents the point below which the

phospholipid is in a more solid or gel state called

soliduss"; and the enclosed area (region of calorimetric

peak) contains both phases in equilibrium. The phase

boundaries of such a system are defined by smooth and

continuous solidus and liquidus curves. The composition and

amounts of liquid and solid phases at any temperature can be

obtained directly from the phase diagram.

For binary mixtures of diacyl phospholipids, if the

component lipids are very similar in both structural and

packing properties, then one component lipid can be replaced

isomorphously by the other in the lamella in both the gel

and liquid crystalline states. For example, when the chains

of two components are similar (e.g., nCj1 and nC,1), co-

crystallization and ideal mixing of the component

phospholipids continuously in the bilayer plane occurs

(Chapman 1976); consequently, lipid dispersions of these

binary mixtures will display phase diagrams with a cigar

shape. In addition, saturated phospholipids of the same

headgroup, differing by 2 methylene units in acyl chains,

will exhibit complete miscibility in all proportions. These

isomorphous systems are often detected for different diacyl

phospholipids with the same polar headgroup and with a small

difference in acyl chain lengths as seen for the mixtures in


S60-- Liquid
Z) liquidus

J 55-

z 50- L + S solidus

z 45
45- Solid

0 10 20 30 40 50 60 70 80 90 100
Mol % Phospholipid

Figure 6: Idealized phase diagram for a binary mixture of
phospholipids whose components are completely
miscible in the liquid and solid state. Above the
fluidus line, the phospholipid is in a liquid
state; below the solidus line, the phospholipid is
in a solid or "gel" state; in between the lines,
both states exist in equilibrium.


Table 7 which produce single, broad, reasonably symmetrical


Isomorphous Mixtures of Phospholipid

Composition Reference

DMPC/DPPC Mabrey and Sturtevant (1976)
DMPC/POPC Curatolo et al. (1985)
DMPC/DPPC Shimshick and McConnell (1973)
DPPS/DPPC Shimshick and McConnell (1973)

However, a four carbon difference is considerably

removed from ideality, isothermal melting is not observed,

and lateral phase separation is possible. Lateral phase

separation is the separation in the plane of the bilayer, of

the solid and fluid regions. Above a certain temperature,

the system is a single liquid phase of uniformly mixed

lipids. As the temperature is lowered, the first crystalline

material appears. The solid region appears as patches which

can be visualized by freeze fracture electron microscopy

(Ververgaert et al. 1973). This is a reversible process that

occurs continuously as the temperature is decreased and

crystalline areas grow. However, at no point do pure A or B

separate; the growth of the solid phase always contains both


Monotectic behavior can be predicted when phospholipids

with widely different chain lengths or unsaturation are

mixed together. A six carbon difference is so far from ideal

that monotectic behavior is observed even though the

initiation temperature remains constant over most of the

concentration range (Mabrey and Sturtevant 1978, Cherry

1976). For example, T. of DOPC is hardly affected by the

presence of equimolar amounts of saturated phospholipids

with chains nC.4 nC22. To of saturated component (higher

m.p.) however, is broadened and occurs at decreased

temperature (Chapman 1976).

In some situations the presence of Ca2, can induce

phase separations in phospholipid mixtures. Findlay and

Barton (1978) found PG and PC exhibit virtually identical

thermograms and mixtures appeared to be completely miscible.

Broader transitions at intermediate transition temperatures

were observed for mixtures of varying chain lengths.

However, upon the addition of calcium, the complex

mesomorphism exhibited with PG was not seen in mixtures

containing > 10% PC and narrow transitions were seen as long

as the chain lengths were the same.

Use of DSC in Liposome Formulation

Differential scanning calorimetry has been shown to be

potentially useful in studying phospholipid interactions

with drugs and other macromolecules (Juliano and Stamp 1978)

especially if the drug is sufficiently hydrophobic and

creates an easily quantifiable change in Tc or transition

endotherm. These include cortisol esters (Fildes and Oliver

1978), morphine derivatives and antidepressants (Cater et


al. 1974, Knight and Shaw, 1979) and various small molecules

(Jain and Wu 1977).

Other compounds causing a thermogram change include

proteins (Papahadjopoulos et al. 1975a, Melchior and Steim

1979), sterols (Ladbrooke et al. 1968) and metal ions

(Hauser et al. 1969).

When phospholipids react with proteins, a lower

temperature reversible T. due to phospholipid and a higher

temperature transition due to protein are observed. This

implies the polar lipid-protein interaction is not extensive

(Melchior and Steim 1979).

It has been shown that some substances can stabilize

phospholipid membranes by eliminating the gel to liquid

crystalline phase transition. This includes the sterols:

cholesterol (Ladbrooke et al. 1968, Papahadjopoulos et al.

1973), a-tocopherol (Ortiz et al. 1987) and vitamin D3

(Castelli et al. 1990). Thus, DSC may be a very useful tool

for the characterization of liposome formulations.

Knowledge of T, is practically indispensable in

liposome formulation since the ability of phospholipids to

form liposomes increases markedly above the transition

temperature where dried phospholipid is hydrated easiest.

Only those phospholipids with T, in water near or below room

temperature spontaneously form bilayers. Fully saturated

phospholipids with high transition temperatures do not form

bilayers unless the temperature is raised above T,.


The permeability of vesicles to entrapped compounds is

relatively low below T,. However, a great deal of

instability and phase separation within the bilayer membrane

occurs during the phase transition, i.e., when the bilayer

structure changes from gel to "ripple" to liquid-

crystalline. Accordingly, the majority of encapsulated

material will be lost if temperature cycling through the Tc

occurs. The increase in ion permeability and penetration of

various molecules into the vesicle bilayer in the vicinity

of Tc has been an area of potential exploitation, i.e., to

achieve a measure of drug targeting by use of hyperthermia

(Yatvin et al. 1978).

Surface Charge of Liposomes

There is a net charge that arises at a surface in

contact with a polar medium that is governed by ionization

(pH), ion adsorption (surfaces which are already charged

e.g., by ionization, usually show a preferential tendency to

adsorb counter-ions) and ion dissolution. This charge can

arise due to the presence of negatively charged phospho-

lipids in a liposomal membrane or due to the degree of

ionization of the phospholipid. It can also be affected by

the presence of various counter ions in solution. A useful

tool used to study this phenomenon in association with

colloidal and cell surfaces is electrophoresis. The

principles of electrophoresis have been discussed by Haydon

(1964) and Shaw (1969).

Electrophoresis involves the motion of dissolved or

suspended material under the influence of an applied

electric field. It is one of four related electrokinetic

phenomena, the others being electro-osmosis, streaming

potential, and sedimentation potential. All four involve the

relative movement between the rigid and mobile parts of an

electric double layer (Figure 7).

The electric double layer arises due to the acquisition

of a surface electric charge when a substance is brought

into contact with a polar (i.e., aqueous) medium and the

mixing tendency of thermal motion. It is comprised of the

charged surface and a neutralizing excess of counter-ions

over co-ions distributed in a diffuse manner in the medium

(Figure 7). The situation is that of a double layer of

charge: one localized on the surface of the plane and the

other in a diffuse region extending into the solution. The

quantity which is a measure of both the degree of surface

charge and the distance the effect extends into the solution

is the "zeta potential". Ions or molecules close to the

particle's surface are not free to migrate in the liquid

until they pass the plane (or surface) of shear.

If an electric field is applied tangentially along the

charged surface a force is exerted on both plates of the

electric double layer. The charged surface (plus attached

material) tends to move in the appropriate direction, while

the ions in the mobile part of the double layer show a net

+ +



+ +
+ +
+ +

+ +

+ +




Figure 7: The electric double layer, represented by a
spherical particle with a net negative surface
charge suspended in aqueous medium and surrounded
by a layer of opposite charge. Below the particle
is a diagram indicating the electrical potential
as a function of distance from the surface of the
particle. The potential at the plane of shear is
the zeta potential.

migration in the opposite direction carrying solvent along

with them, thus causing its flow. Conversely, a potential

gradient is created if the charged surface and the diffuse

part of the double layer are made to move relative to one


Electrokinetic phenomena are only directly related to

the nature of the mobile part of the electric double layer

and may, therefore, only be interpreted in terms of zeta

potential or charge density at the surface of shear. The

electrophoretic charge or zeta potential does not represent

the actual charge on the surface of the particle, as each

will move in the electric field with an associated cloud of

counterions which are in equilibrium with those present in

the bulk aqueous phase. The actual mobility therefore will

depend on the charge existing at the plane of shear between

the particles and the solution.

There is a direct relationship between the zeta

potential and the electrophoretic mobility (the velocity of

the particle per unit electric field) of colloidal particles

according to the Smoluchowski equation.

pe= D(e/47r

where Ae = electrophoretic mobility (A/sec per volt/cm), D=

dielectric constant of medium, e= strength of applied

electric field (V/cm), n = viscosity of medium (poise), and

C = zeta potential in mV. For the derivation of this

equation one can refer to Overbeek and Wiersema (1967) and

Wiersma et al. (1966).

This equation can be simplified to:

Me = Cc/y

where e now equals permittivity. Permittivity is related to

the dielectric constant by (6x/evacu,)

For an aqueous medium at 250C, the dielectric constant

of water is 78.8, and the viscosity is .00899 poise, thus

the zeta potential is related to the electrophoretic

mobility by: C = 12.85 ge millivolts

with Ae expressed in micron/sec per volt/cm.

At 370C this becomes: C = 10.35 Me

Factors Affecting Electrophoresis

Particle size. The Smoluchowski equation is applicable

to particles in which the double layer is effectively flat

(i.e. large Ka, where Ka is the ratio of the radius of

particle curvature to double layer thickness), including

colloids and cells (Shaw 1969). Generally this includes all

particles > 10 nm. Accordingly, the electrophoretic mobility

of a non-conducting particle for which Ka is large at all

points on the surface should be independent of its size and

shape provided that the zeta potential is constant.

Phospholipid bilayers were previously shown to conform

to double layer theory, with the Smoluchowski equation being

useful for the conversion of the measured electrophoretic

mobility to the zeta potential (McLaughlin et al. 1981).

Viscosity. The viscosity of water decreases with

increasing temperature as seen in Table 8. In addition,

various other molecules such as sugars can profoundly affect

viscosity as seen in Table 9. This can affect the measured

pe, particularly at high temperatures.


Effect of Temperature on Dielectric Constant
and Viscosity of Water

Temperature 17 (cp)

20 1.002 80.1
25 0.8993 78.8
30 0.798 76.54
35 -- 75.04
40 0.653 73.15
50 0.547 69.9

from Robinson and Stokes (1955)


Effect of Sugars on Dielectric Constant
and Viscosity of Water at 250C

composition r e

10% sucrose 1.33 76.3
20% sucrose 1.905 --
10% glycerol 1.29 --
20% glycerol 1.73 --
10% mannitol 1.34 77.1

from Robinson and Stokes (1955)

Dielectric constant. The dielectric constant of water

decreases with increasing temperature as seen in Table 8

from 78.8 at 25c to 73.8 at 380C. It has been shown that at


dilute concentrations, the dielectric constant of water can

be related to the dielectric constant of a salt solution

based on:

Ess = CSW + C

where y = dielectric decrement. This is a crude estimate and

only holds true for salt solutions < 1M. The y for several

salts are listed in Table 10. The lowering of es. is more

than a single volume effect arising from the addition of

non-polar particles. Ions orient water molecules around them

thereby reducing their ability to orient in the applied

electric field, and so reduce the dielectric constant by a

local high field effect. Solutions of 1M electrolytes have

been shown to lower esw by a few percent.


Dielectric Decrement Values of Salts

Salt Y

Na* -8
K* -8
Mg2' -24
Ba2' -22
Cl1 -3

from Robinson and Stokes (1955)

Temperature. Dielectric constant and viscosity are both

temperature sensitive as seen in Tables 8 and 9. Thus, it

has been shown that increases in temperature can cause an

increase in electrophoretic mobilities by approximately 2%

for every 10C (Shaw 1969).

Electrophoresis of Phospholipids

The visibility of colloidal particles permits the

direct measurement of the electrophoretic mobility under the

ultra or light microscope. The principle result of

microelectrophoretic measurements is to determine the zeta

potential. One such microelectrophoretic apparatus using a

capillary tube was developed by Bangham and Dawson (1958).

It was found that there are two electrophoretic phenomena in

a closed glass tube containing an aqueous suspension of

particles when an electric field is placed across the tube.

1) motion of the particle with respect to the

surrounding medium due to the applied electric


2) motion of liquid relative to walls of tube

(electroosmotic effect).

There exists a position in the tube where the

electroosmotic effect equals zero, called the stationary

layer. The stationary layer for a cylindrical tube can be

calculated to be 0.293R from the wall of a tube of radius R.

In order to avoid electroosmotic effects, the objective of

the microscope must be focused exactly at this point:

This type of apparatus has been used for studying many

charge interactions of colloidal particles. Two areas where

electrophoretic data have been significant are blood

clotting and action of hydrolyzing enzymes on phospholipid

mesophases. Both reactions show strong correlations with

surface charge density.

Bangham and Dawson (1959) were the first to apply

electrophoresis measurements to phospholipid liquid crystals

during a series of studies on phospholipids and their

interaction with enzymes. They found zero electrophoretic

mobility for pure lecithin in 0.025M NaCl over a pH range of

3.0 to 13.0. This result is theoretically predictable

knowing the zwitterionic structure of the lecithin molecule.

They also found a direct correlation in electrophoretic

mobility vs. mol% negatively charged phospholipid

(dicetylphosphate (DCP)). However, in connection with

lecithinase activity, they found the zeta potential was not

linearly associated but rather commenced at approximately 30

mol% negative charge. This appeared to be a critical

concentration for activation since lecithin breakdown

rapidly reached a maximum and declined as further negative

charge was expressed. In contrast, Bangham and Dawson (1962)

found another enzyme, a-toxin, only lysed phospholipids if

they were positively charged. This effect also occurred at a

maximum concentration, above which enzyme activity


Bangham (1961) correlated the surface charge of

phospholipids to their coagulant action. Neutralization of

negative charge proportionately decreased the effectiveness

of Russel viper venom on blood clotting, causing increased

clotting times to be seen. Positively charged surfaces

clotted at a rate 2 to 6 times slower than negatively

charged surfaces and 1 to 2 times slower than neutral

surfaces, depending on concentration.

Bangham et al. (1965a) studied the interactions of egg

PC:DCP (9:1) liposomes with local anesthetics. They found

various anesthetics could essentially neutralize the charge

on egg phosphatidylcholine liposomes made negative by the

addition of 10% dicetylphosphate (from -20mV to zero). This

change in potential led to a decrease in the diffusion rate

of cations out of the vesicles. The effects of Ca2* and

local anesthetics was later correlated with their ability to

decrease the magnitude of negative surface potential, which

in turn resulted in a lower concentration of monovalent

cations in the vicinity of the membrane (McLaughlin et al.


It has been shown many times that the acidic

phospholipids PS, DCP, and PI bear one net negative charge

between pH 5 to 8 and thus the zeta potential is pH

sensitive (Hauser and Dawson 1967, Papahadjopoulos 1968,

Black and Gregoriadis 1976, Hauser et al. 1979). At any

given pH then, the zeta potential of liposomes is directly

proportional to the concentration of added negatively

charged phospholipid. Hauser and Dawson (1967) showed that

the progressive addition of PI, phosphatidic acid (PA) and

DCP to a PC bilayer produced a decrease in mobility which at

low concentrations (1 to 9 mol%) was directly proportional

to the mol% acidic phospholipid added. The magnitude of the

change was directly related to the total number of negative

charges available from the concentration of negatively

charged phospholipid, suggesting that the added anionic

phospholipids were uniformly distributed throughout the

vesicle. Hauser et al. (1979) provided additional evidence

that charged phospholipids randomly distribute between two

halves of a bilayer as well as throughout concentric layers

of multilamellar liposomes using electrophoresis and spin

labelled studies.

MacDonald and Bangham (1972) reviewed the

electrophoresis of phospholipid monolayers and bilayers and

confirmed that there was a good correlation between the

charge and the electrophoretic mobility by both methods. In

contrast, Szoka and Papahadjopoulos (1981) claimed that

mobility varies directly with the log of mol fraction of

charged species and that as little as 2 mol% charged

phospholipid could be detected in a PC/PS bilayer.

It has also been shown that the zeta potential of

negatively charged phospholipids is sensitive to the

presence of various cations. Bangham and Dawson (1959) noted

the effect of metal ions (ie., Ca" and Ba2", on lecithin

which had been made negatively charged by the addition of

dicetylphosphate. A plot of electrophoretic mobility vs.

molar concentration of metal ions showed an interaction at

the lecithin surface resulting in a positive mobility. The

magnitude of this potential was small and did not increase

with an increase in molarity >0.02M CaCl. Papahadjopoulos

(1968) showed the zeta potential of PS vesicles at neutral

pH was -55 mV, however, the presence of calcium increased

the mobility by neutralization of the negative charge.

This is consistent with studies of other negatively

charged substances such as DCP, stearylphosphate and

laurylphosphate where the zeta potential was little affected

by the presence of sodium, but significant charge

neutralization occurred in the presence of calcium which was

even more drastic in the presence of aluminum ions, in some

cases causing charge reversal (Adler 1973). In addition, the

positively charged amine group of cationic amines such as

stearylamine is capable of repelling Ca2 and Na* ions.

However, in contact with Al3*, the charge was not repelled

and an increase in zeta potential was observed. This was

probably due to the attraction of hydrated hydroxyl ions,

which are attached to the Al"3, to the positively charged

amine (Adler 1973).

The effect of divalent cations on zeta potential can be

seen in the work of Lau et al. (1981) who showed an increase

in the zeta potential of phosphatidylglycerol liposomes as

the concentration of Ca2* or Mg2 was increased (Table 11).

McLaughlin et al. (1981) showed PS vesicles behaved in a

similar manner in the presence of increasing concentration


Cation Effect on Zeta Potential

Concentration (M)


of Phosphatidylglycerol

Zeta Potential (mV)

ca Mg.

-55 -55
-40 -45
-17 -20
-2 -5

from Lau et al. (1981)

of calcium, as well as a change in-zeta potential dependent

on the T. of the phospholipid (Table 12). The largest

dependence on.temperature was in the absence of Ca2 as the

temperature decreased from 25 to 50C. Similar temperature

dependence had previously been seen for DPPS by McDonald et

al. (1976). The same trends in charge neutralization were

seen by Klein et al. (1987) with mixtures of PS and PC or



Effect of Calcium and Temperature on
Zeta Potential of Phosphatidylserine

Temp (C)

Ca" (M)

5 25

Zeta Potential (mV)

0 -50 -60 -60
0.5 -40 -42 -45
5.0 -23 -23 -22

from McLaughlin et al. (1981)


The electrophoresis of liposomes has also been shown to

be affected by ionic strength by Crommelin (1984), who

showed an increase in zeta potential as ionic strength

increases with increasing NaCl concentration (Table 13).


Effect of Ionic Strength on Zeta Potential
of Phosphatidylserine Liposomes

NaCl (mM) 6.6 150

Mol% PS Zeta Potential (mV)

6.6 -57 -12
13.3 -58 -20
20.0 -65 -34

from Crommelin (1984)

In addition, other substances have been shown to alter

the zeta potential of liposomes. Black and Gregoriadis

(1976) showed liposomes recovered from plasma have altered

electrophoretic mobilities, indicating possible adsorption

of plasma proteins. Neutral and positively charged liposomes

acquired a negative charge on interaction with rat plasma,

but the charge on already negative liposomes was not

affected. This process was reversible after washing except

for positively charged liposomes which still retained a

slight negative charge. They attributed this to the presence

of a2-macroglubulin found bound to the surface.

More recently, Krumbiegel et al. (1990) studied the

interactions of phosphatidylcholines with the

glycoaminoglycans -- dextran sulfate, chondroitan sulfate,


and heparin, by microelectrophoresis. All these experiments

were done in the presence of calcium which is biologically

relevant since calcium occurs in relatively high

concentrations in the extracellular space. The zeta

potential was found to vary with glycoaminoglycan, however,

the concentration of calcium, as well as the type of

phospholipid played a part in the interaction.

Interactions Between Liposomes and Cells

There are several types of interactions possible

between liposomes and cells, including adsorption on the

cell surface, phospholipid exchange, fusion, and endo-

cytosis. Experiments with cultured cells suggest that the

phospholipid composition is rather crucial in determining

the mode(s) of vesicle-cell interaction (Poste and

Papahadjopoulos 1978). These were recently reviewed by

Margolis (1988) with reference to solid (gel) and fluid

(liquid crystalline) state liposomes. There seem to be many

differences in these interactions, however, our primary

concern was with endocytic reactions with some possible

implications of adsorption since it has been shown that

attractive forces are involved in the initiation of

phagocytosis, as well as interactions between cells and

solid phase liposomes.

Adsorption on cell surface. Both fluid and solid phase

liposomes have been shown to readily adsorb onto cell

surfaces (Pagano and Takeichi 1977, Margolis and Neyfakh

1983, Margolis et al. 1984). Solid MLV bind to the cell

surface as revealed by electron microscopy (Martin and

McDonald 1976) and are capable of retaining entrapped marker

in this state (Margolis and Neyfakh 1983).

It has been consistently shown that vesicles made

negative with PS associate to a greater extent with cells

compared to neutral PC vesicles (Schroit and Fidler 1982,

Tanaka and Schroit 1983, Schroit et al. 1984, Rimle et al.

1984, Dijkstra et al. 1985). For example, as the amount of

PS decreases in PC/PS liposomes, the amount of PS bound to

macrophages during early incubation periods decreases (Chen

et al. 1977, Fraley et al. 1981), with 0.2-0.4% of the total

liposomes (PC:PS (10:1.5 W/W)) in culture tightly bound to

cells (Chen et al. 1977).

Rimle et al. (1984) observed a 3- to 4-fold increase in

binding of liposomes to RAW 264 macrophages with the

addition of 5% PS to the liposomal membrane, however, the

presence of calcium in the culture medium was necessary for

the effect. This may be due to the decreased charge as

measured by electrophoresis, which may lead to decreased

charge repulsion with the negatively charged surface of the


However, Stendahl and Tagesson (1977) found a greater

extent of liposome association with polymorphonuclear

leukocytes as the amount of DCP was increased in the

membrane from 1 to 10 mol% with no changes in reaction rate

when calcium or magnesium were omitted from the culture

medium. Little or no association was observed with neutral

or positive vesicles.

High concentrations of PG and cardiolipid found in some

bacteria (e.g., mycobacteria) appear to stimulate the

proliferation of those macrophages that phagocytize them

(Kates 1964). Yui and Yamazaki (1986) found that

phospholipid species such as serine, glycerol, ethanolamine

and cardiolipin, with growth stimulation activity were more

extensively bound to macrophages, whereas those without this

activity, PC and SPM had very little binding affinity.

Specifically, they found <1% PC compared to ca. 4% PS bound

to peritoneal macrophages after a 24 hour incubation period.

Endocvtosis. In vitro investigations of cellular uptake

show that liposomes are taken up by "coated pit" endocytosis

and their contents accumulate in a low pH intracellular

compartment (Raz et al. 1981, Straubinger et al. 1983).

Liposomes are endocytosed by coated pits in a manner similar

to that of a number of macromolecules with specific surface

bound receptors and also share their intracellular fate:

progressing from coated pits to coated vesicles, to uncoated

vesicles to large vesicles that correspond to endosomes and

lysosomes. Eventually, most liposomally delivered material

which remains in the cell can be found in the trans-Golgi

region, either in dense bodies (secondary lysosomes) or in

other vesicles.

The inclusion of negatively charged phospholipids

within the lipid bilayer of liposomes enhances their binding

to and phagocytosis by blood monocytes and tissue

macrophages. In particular, the inclusion of PS in the

phospholipid bilayer of small or large uni- or multi-

lamellar liposomes results in their enhanced binding to and

phagocytosis by all cells of the RES including mouse

peritoneal macrophages, mouse Kupffer cells, mouse or rat

alveolar macrophages, human alveolar macrophages, and human

peripheral blood monocytes (Fidler 1985).

Negatively charged liposomes are generally superior to

positively charged or neutral liposomes in the functional

delivery of macromolecules such as DNA and RNA to a variety

of animals (Fraley et al. 1981 and Schaefer-Ridder et al.

1982), and plant cells (Fraley et al. 1982). For example,

liposomes containing PS were phagocytosed up to 10 fold

faster than lipsomes of the same size and configuration that

contain positively charged stearylamine, or neutral PC.

Like PS, the addition of another negatively charged

phospholipid, PG, to liposomes also significantly enhances

liposome internalization by macrophages (Mehta et al. 1982).

Fraley et al. (1981) saw the amount of liposomes associated

with CV-1P, an established cell line of African green monkey

kidney cells, was greater after incubation with PS or PG



Schroit and Fidler (1982) found efficient uptake of PC

vesicles by rat or mouse AM was dependent on the presence of

PS (30 Mol%). In addition, they found higher phagocytosis of

DSPC liposomes which was approximately two times greater in

the presence of 30% PS. Although the majority of the cell

associated phospholipid was due to endocytosis (as indicated

by fluorescence), the presence of PS and also DSPC increased

binding. No significant fluorescence was observed at the

cell surface suggesting phagocytosis ensues very rapidly

after initial cell binding, supporting the theory that

lipids below Tc (like PC:PS) adsorb to cell surfaces whereas

lipids above Tc become cell associated by other pathways.

Interactions between cells and solid or fluid phase

liposomes may be somewhat cell specific. Solid SUV were

shown to remain on the surface of fibroblasts and epithelia

for at least 2 hours without endocytosis (Margolis and

Neyfakh 1983, Margolis et al. 1982). However, Poste et al.

(1976) reported association of solid SUV with 3T3 cells

which was sensitive to inhibitors like cytochalasin B and

was decreased dramatically upon incubation of the cells with

2-desoxy-D-glucose and sodium azide or at low temperature.

While there is a general tendency towards the theory

that negatively charged phospholipids increase binding, and

therefore, endocytosis of liposomes, there have also been

several reports of increased endocytosis with cationic

substances (Pagano and Weinstein 1978, Poste and Kirsch