Professional Documents
Culture Documents
KÅRE LARSSON
Camurus Lipid Research Foundation, Lund, Sweden
PETER QUINN
King’s College London, London, UK
KIYOTAKA SATO
Hiroshima University, Japan
FREDRIK TIBERG
Camurus AB, Lund, Sweden
Woodhead Publishing India Private Limited, G-2, Vardaan House, 7/28 Ansari Road,
Daryaganj, New Delhi – 110002, India
www.woodheadpublishingindia.com
This book contains information obtained from authentic and highly regarded sources.
Reprinted material is quoted with permission, and sources are indicated. Reasonable
efforts have been made to publish reliable data and information, but the authors and the
publisher cannot assume responsibility for the validity of all materials. Neither the authors
nor the publisher, nor anyone else associated with this publication, shall be liable for any
loss, damage or liability directly or indirectly caused or alleged to be caused by this book.
Neither this book nor any part may be reproduced or transmitted in any form or by any
means, electronic or mechanical, including photocopying, microfilming and recording, or
by any information storage or retrieval system, without permission in writing from
Woodhead Publishing Limited.
The consent of Woodhead Publishing Limited does not extend to copying for general
distribution, for promotion, for creating new works, or for resale. Specific permission
must be obtained in writing from Woodhead Publishing Limited for such copying.
Over ten years have passed since the publication of Kåre Larsson’s Lipids –
Molecular Organization, Physical Functions and Technical Applications (LMO)
by The Oily Press, then based in Dundee, Scotland, and run by Dr William W.
Christie. The book was soon recognized as a major contribution to the literature
(“This is a book without comparison in the lipid literature...”, Stig E. Friberg in
the Journal of Dispersion Science and Technology, 1995, Vol.16, p.295 and
“...the content is excellent”, Philip W. Wertz in Chemistry and Physics of
Lipids, 1994, Vol.74, p.99). Dr Christie’s excellent choice of Kåre Larsson as
the author was also confirmed (“His expertise in describing the various states
of lipids is second to none”, Edward G. Perkins in INFORM, 1994, Vol.5,
p.1394 and “...written by an acknowledged world expert in his field”, Fred B.
Padley, Lipid Technology, 1994, Vol.6, p.102). Until the publication of LMO
there had been no single, concentrated source of so much information on the
subject: “The strength of this book – and it is enormous – is the fact that the
author has been able to compile in one volume information otherwise found
only in the most widely different kinds of scientific journals” (Friberg).
When LMO was published, Kåre Larsson was a Professor of Food Technol-
ogy in the Chemical Centre at Lund University, Sweden. He is now cofounder
of Camurus AB and Probi AB, and earlier of Biogram AB (which later became
Bioglan AB), and serves as Chairman of the Board of the Camurus Lipid
Research Foundation in Lund. Camurus is a provider of drug delivery systems
and works closely with pharmaceutical manufacturers. Kåre Larsson is a
Fellow of the Royal Swedish Academy of Science and the Academy of
Engineering Science and has authored more than 200 original papers and five
books covering areas of lipid biophysical chemistry, food science and
nutrition, and biomedicine. He is also the named inventor on several patents, of
which four have led to industrial products. In 2001 he won the Rhodia prize of
the European Colloid and Interface Society for his discovery of cubosomes and
hexosomes and explorative work on their applications.
When I asked Kåre Larsson to write a second edition, the comprehensive
coverage of LMO became an obstacle – the subject area had expanded to such
an extent that one author could not cover it alone. But we did not want to resort
to the usual edited book with each chapter written by a different author. The
solution was to invite three other well-known scientists in this field to act as
coauthors – and it is a tribute to Kåre’s reputation that all three agreed.
Peter Quinn, Professor of Biochemistry in the Department of Life Sciences
at King’s College London, UK, is renowned for his work on biological
v
vi FOREWORD
The ambition behind the new edition of this book is to provide an up-to-date
description of the diversity of lipid molecular arrangements in different
physical states, as a basis for the understanding of lipid functionality in
biological and technical systems. The first edition was published in 1994 with
Kåre Larsson as author, and when he was asked by the publisher to revisit the
text he realized that he could not cover this broad field alone. Three colleagues
joined forces with him, and the present edition has therefore in many aspects
been extended. In some cases the description is deeper with a more narrow
focus. For example, the chapter on the solid state in the earlier edition covered
all lipids, whereas in this new edition there is a very complete demonstration of
the crystal structures and crystallization properties of fatty acids and fats. These
general principles, however, can be applied to all lipids.
In biology, as well as in technical applications such as foods, we are dealing
with soft matter. Lipids form aqueous phases alone or in conjunction with
proteins and polysaccharides. The combination of short-range disorder and
long-range order into liquid-crystalline structures plays a crucial role. A
driving force is the dualistic properties of the molecules in relation to water.
Molecular regions avoiding water contact, in combination with regions
striving towards such contact and interaction, lead to self-assembly, and even
in the liquid state to the formation of organized structures on the colloidal level.
This new edition presents many new results, particularly on the structure and
functions of dispersions of liquid-crystalline phases forming nanostructures
and mesoporous systems.
With regard to the role of lipids in cellular and molecular biology, this book
focuses on biophysical aspects, and discussion of lipid biochemistry is limited
to a chapter on cell membranes.
Kåre Larsson, Lund, Sweden
Peter Quinn, London, UK
Kiyotaka Sato, Hiroshima, Japan
Fredrik Tiberg, Lund, Sweden
January 2006
vii
Contents
Foreword v
Preface vii
1 Basic concepts 1
A. Classification of lipids 1
1. Non-polar lipids
2. Polar lipids
B. Structure, physical properties and functional properties 5
C. Methods for structure characterization 6
References 8
8 Emulsions 175
A. Oil-in-water emulsions 175
1. Emulsification
2. Protein-stabilized emulsions
3. Lipid-stabilized emulsions
B. Water-in oil emulsions 180
C. Demulsification 181
References 181
12 Foods 251
A. Starch–lipid interaction 251
B. Milk-based lipid products and their analogues 254
1. Milk
2. Whipping of cream, butter and ice cream
3. Drying of milk
4. Hydrogenation of oils – hardened fats
5. Artificial milk/cream, margarine and low-fat spread products
C. Cocoa butter and chocolate 259
D. Vegetable oils, frying oils and emulsion products 260
1. Frying oils
2. Mayonnaise and dressing products
E. Lipids in bread and other baked products 261
1. Interaction of wheat flour with water
2. The baking process
3. Cakes and biscuits/cookies
References 264
Recommended further reading 264
Index 265
CHAPTER 1
Basic concepts
A. Classification of lipids
There is no strict definition of lipids. Extracts in organic solvents from
biological tissues have been termed lipids for a long time. A definition more
consistent with modern views, proposed by Christie (1987), says: “Lipids are
fatty acids and their derivatives, and substances related biosynthetically or
functionally to these compounds”. From a physical point of view it is natural to
divide lipids into two groups:
• polar lipids, which interact with water and form aqueous phases;
• non-polar lipids, which do not form aqueous phases.
These are the definitions we will use in this book. Interfaces, however, are a
special case, where even non-polar lipids can interact with water and form
monomolecular layers.
Simple fatty acids exhibit most of the structural states described in this book.
Oleic acid, for example, behaves like oil in relation to water at room tempera-
ture. If titrated into soap, however, a wide variety of aqueous phases are
obtained. This means that by lowering the pH, dissociated oleic acid – a polar
lipid – becomes undissociated and a non-polar lipid. In this respect fatty acids
represent a special case.
The structures of fatty acids in different states of order represent well the
structures in the solid state, described in Chapter 2, as well as in the disordered
liquid-crystalline phases, described in Chapter 3. The overall organization into
bilayers in the solid state is driven by the carboxyl groups, which associate the
molecules by hydrogen bonds into dimers or polymers. The hydrocarbon
chains are extended to different degrees. In solids, planar zigzag conformations
of the carbon–carbon bonds exist (all-trans), whereas in disordered states,
occurring in liquid crystals and in melts, gauche conformations form dynami-
cally along the chains. Recording of the trans/gauche ratio, for example by
Raman spectroscopy, has demonstrated the conformational changes that occur
during chain-melting transitions.
The raw materials used in lipid technology are fats and oils, with plant
extracts dominating. About 120 million tonnes of vegetable oils were produced
in 2001, the main types being soybean oil (29 million tonnes), palm oil (23
million tonnes) and rapeseed oil (14 million tonnes). The most important polar
lipids, phospholipids, are obtained as by-products during the industrial refining
of vegetable oils into pure triacylglycerols.
1
2 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
1. Non-polar lipids
they were viable even though cholesterol was replaced by another sterol
(Wechsler et al., 2003). An important mechanism induced by cholesterol in the
cell membranes is phase separation of the bilayer into cholesterol-rich domains
(lipid rafts and caveolae; see also Chapters 3 and 9) and cholesterol-poor
regions. Perhaps this segregation was still achieved in the knockout mice by the
alternative sterol.
2. Polar lipids
Monoacylglycerols or monoglycerides are the simplest type of polar lipids, and
are the dominant functional additives used in industrial food processing. They
are produced on a large scale to high purity by molecular distillation. Of the two
isomers, the 1-isomer dominates in the equilibrium mixtures formed by acyl
migration (about 90% 1-isomer and 10% 2-isomer).
Phospholipids or phosphoglycerolipids associate spontaneously into lipid
bilayers in water, which is the basic mechanism behind the formation of
biological membranes. They have acyl groups ester-bound in the 1- and 2-
positions of the glycerol backbone, and a polar group involving phosphate
in the 3-position. The most important members of this lipid class
are phosphatidylcholine (PC), phosphatidylethanolamine (PE) and
4 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
involving optical axis variations were analysed in smectic phases. It was shown
that the so-called focal–conic texture is due to curved arrangements of equidis-
tant and uniform layers, which are flexible (liquid-like). Lack of birefringence
means that a phase is cubic or a true liquid, whereas the birefringence texture
is quite characteristic of other liquid-crystalline phases described later in this
book.
X-ray diffraction is the most powerful method for structure determination in
the solid state. One problem is to prepare single crystals that are perfect enough
to allow resolution of the reflections from the different crystal planes. If this is
achieved, it is possible to determine the exact position in the unit cell of all
atoms other than hydrogen atoms. The unit cell is the smallest repetition unit
within a crystal.
Even when single-crystal data are not available, it is still possible to derive
the main features of the molecular packing from X-ray powder diffraction
curves, which can always be obtained. The reason is the arrangement of the
lipid molecules into layers, where the distance between hydrocarbon chains
can be derived from one region of the diffraction pattern (‘short-spacings’) and
the thickness of the layers can be obtained from another region (‘long-
spacings’). The solid-state behaviour of lipids also involves transitions between
different crystal forms (polymorphs).
The aqueous phases of lipids are liquid-crystalline or liquid, and small-angle
X-ray scattering/diffraction (SAXS) methods are used to characterize the
structures, particularly of the more ordered liquid crystals. As the repetition
distances in the hydrocarbon chain direction are usually very long, small
diffraction angles must be recorded. The detailed diffraction patterns used in
order to identify the structures of different phases is further discussed in
Chapter 3.
Neutron scattering methods are also applied to studies of lipid structures.
The replacement of hydrogen atoms by deuterium in lipids provides the
possibility of locating hydrogen atoms, a feature that is normally not possible
by X-ray diffraction.
The frequency of X-ray (or neutron) radiation is so high in relation to the
movement of the individual atoms that only a static structure is seen by
individual photons. Therefore the diffraction pattern reflects the time average
of the atomic position. In order to obtain information on the dynamic behaviour
of the molecules, various spectroscopic methods are used. The most powerful
technique is NMR (nuclear magnetic resonance) relaxation, and the NMR
method also provides structural information for the identification of liquid and
liquid-crystalline phases. From infrared and Raman spectra we can obtain
information on molecular conformation, for example whether the hydrocarbon
chains of lipid crystals have their zigzag planes parallel or not.
The enormous increase in X-ray intensity achieved by synchrotron sources
has recently made possible time-resolved X-ray diffraction analysis. This has
8 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
References
Bragg, Sir W. (1934) Nature 133, 445.
Christie, W.W. (1987) High-Performance Liquid Chromatography and Lipids, Pergamon
Books, Oxford, UK.
Friedel, G. (1922) Ann. Phys. 18, 2373.
Gunstone, F.D., Harwood, J. L. & Padley, F.B., eds (1986; second edition 1994) The Lipid
Handbook, Chapman & Hall, London, UK.
Wechsler, A., Braufman, A., Skafir, M., Heverin, M., Gottlieb, H., Damari, G., Gozlan, S.,
Kelner, I., Spivak, O., Moshkin, O., Friedman, E., Becker, Y., Shaliter, R., Einat, P.,
Faerman, A., Björkhem, I. & Feinstein, I. (2003) Science 302, 2087.
CHAPTER 2
Solid-state behaviour of polymorphic fats and
fatty acids
(a)
Physical state Function
crystal shape, melting,
liquid density, rheology
crystal
gel morphology,
liquid rheology
emulsion
stability, barrier,
o/w rheology, carrier
+H 2O &
emulsifier stability, rheology,
w/o texture
(b)
External conditions
Figure 2.1 (a) Relationships between the solid-state behaviour of fats and lipids in different physical
states, and lipid function. (b) Relationships between the microscopic and macroscopic features of fat
crystals.
crystallization properties of the principal fats and fatty acids are discussed in
this chapter as they relate to the physical states of bulk, emulsion, and gel. The
fundamental aspects of lipid crystal structures will be discussed first.
(a) (c)
L A L
G G
A B
B
TA–B Tm(B) Tm(A) Tm(B)
T T
(b) (d)
ΔG#S ΔG#m
A ΔG#c
G ΔG# G L
meta-
stable
stable
B B
Figure 2.2 Gibbs energy (G) values of two polymorphic forms, A and B, and liquid (L), as a function
of temperature (T). (a) Enantiotropic relation. (b) Activation of Gibbs free energy (ΔG#) for transforma-
tion from metastable to stable forms. (c) Monotropic relation. (d) Solid-state transformation from A to B
(dashed arrow) and melt-mediated transformation from A to B (solid arrows). Tm, melting temperature.
1. Fatty acids
This section discusses the polymorphic structures of the principal saturated and
unsaturated fatty acids. Fatty acids are the main hydrophobic moieties of lipids
present in biological tissues, and are also employed for multiple purposes in the
food, pharmaceutical, and polymer industries (Chow, 1992). Saturated fatty
acids exhibit a straight chain configuration, allowing for melting points higher
than those of unsaturated fatty acids with the same number of carbon atoms.
The inclusion of a double bond reduces the melting point of an unsaturated fatty
acid, depending on its conformation (cis or trans) and position in the carbon
14 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
nc Polymorphic forms
Even-numbered
12 A1(t), A-super(t), C(m)
14 A2(t), B(m), C(m)
16 A2(t), B(m), C(m), E(m)
18 A(t), B(m), C(m), E(m)
Odd-numbered
13 A′(t), C′(t)
15 A′(t), B′(t), C′(m)
17 A′(t), B′(t), C′(m)
c/2 c/2
c/2
(a)
a a
a
B form C form E form
(b)
monoclinic orthorhombic
Figure 2.3 (a) Molecular structures of three polymorphs of stearic acid: B, C, and E. In each case, the
projection of half of the unit cell (a, c/2) is shown. (b) Illustration of polytypism of monoclinic type and
orthorhombic type.
H T M
cs
O⊥ O′
bs
as
Figure 2.4 Typical subcell structures present in lipid crystals. Subcell parameters as, bs and cs are given
for O⊥ as an example.
16 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
Figure 2.5 Relationship between Gibbs energy (G) and temperature for five crystal forms of stearic
acid.
Table 2.2 Molecular properties of polymorphism in oleic acid, elaidic acid and erucic
acida
α S-C-T O′ + O -like
⏐⏐ ⊥
β2 Unclear ⏐⏐-type
β1 A: (174º,cis,173º) (T-C-T) T ⏐⏐
B: (175º,cis,175º) (T-C-T)
Elaidic acid LM (118º,trans,118º) (S-T-S′) O⊥
HM Unclear O + ⏐⏐-type
⊥
γ1 S-C-S O′ + O -like
⏐⏐ ⊥
α S-C-T O′ + O -like
⏐⏐ ⊥
α1 S-C-S T ⏐⏐
a
Data taken from Suzuki et al., 1985; Ueno et al., 1994; Kaneko et al., 1997b; Kaneko et al., 1998.
S-C-S′, skew-cis-skew′; S-C-S, skew-cis-skew; T-C-T, trans-cis-trans; S-C-T, skew-cis-trans. A and B
are two asymmetric units in a unit cell. LM, low-melting; HM, high-melting.
Figure 2.8 Relationship between Gibbs energy (G) and temperature for the γ, α, β2, and β1 forms of
oleic acid.
c/2 c c
a
(a)
c/2
a
(b) b a
(c) (d)
Figure 2.10 Crystal structures of polymorphic forms of the principal cis-monounsaturated fatty acids:
(a) γ form of erucic acid; (b) α form of erucic acid; (c) LM (low-melting) form (orthorhombic polytype)
of petroselinic acid; (d) LM form of elaidic acid.
segment consisting of the methyl group of molecule A and the carboxyl group
of molecule B, and subcell 2 is made of the methyl group of molecule B and the
carboxyl group of molecule A. The subcell parameters are summarized in
Table 2.3.
The fact that the interdigitated lamellar structure of β1 was only observed in
oleic acid can be explained as follows. Oleic acid (C18:1) has the same number
of carbon atoms (9) in its ω and Δ chains. For most other cis-monounsaturated
fatty acids, the ω and Δ chains are of different length. Therefore, a void would
form in the interdigitated structure either between the methyl ends or the
carboxyl terminal groups of adjacent molecules. A void between adjacent
carboxyl groups would result in an excessive energy barrier to formation of the
hydrogen bonds that stabilize the structure, and so this type of crystal would not
form.
24 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
as 4.09 nm 4.18 nm
bs 5.36 nm 5.54 nm
cs 2.54 nm 2.54 nm
α 81.4º 72.2º
β 106.5º 109.7º
γ 120.8º 123.5º
a
Data taken from Kaneko et al., 1997b.
Table 2.5 Polymorphic properties of linoleic acid, α-linolenic acid, and oleic acida
Linoleic acid
LT –51.3 (→ MT) 2.6 8.1
MT –35.4 (→ HT) 0.27 0.86
HT –7.2 (→ melt) 33.6 120.9
α-Linolenic acid
LT –60.2 (→ HT) 0.11 0.46
HT –13.0 (→ melt) 27.8 106.1
Oleic acid
γ –2.2 (→ α) 8.76 32.3
α 13.3 (→ melt) 39.6 138.4
a
Data taken from Suzuki et al., 1985; Ueno et al., 2000.
Ttr, transformation temperature. ΔHtr, enthalpy of transformation. ΔStr, entropy of transformation.
2. Diacylglycerols
Monoacylglycerols and diacylglycerols are polar acylglycerols. The crystal-
line and lyotropic liquid crystalline properties of monoacylglycerols were
recently reviewed by Krog (2001). Therefore, this section focuses on
diacylglycerols (DAGs) as a representative group of polar acylglycerols.
Interesting properties were observed in the crystal structures of three types of
saturated–unsaturated mixed-acid DAGs, 1-stearoyl-2-oleoyl-sn-glycerol (sn-
1,2-SODG) (Di & Small, 1993), 1-stearoyl-3-oleoyl-sn-glycerol (sn-1,3-SODG)
(Goto et al., 1995), and 1-stearoyl-2-linoleoyl-sn-glycerol (sn-1,2-SLiDG) (Di
& Small, 1995). Molecular structures of the diacylglycerols composed of one
saturated and one unsaturated fatty acyl chain make up the hydrophobic core of
many biological membranes. Therefore, the interactions of the acyl chains in
the membrane bilayer are of interest because the saturated and unsaturated
chains have marked difficulty in packing together in the crystalline state
(Small, 1984).
The stearoyl and oleoyl chains were stacked in separate leaflets of the double
chain-length structure in the stable form of sn-1,3-SODG (Tm=42.5ºC), in
which both leaflets were packed in the T⏐⏐ subcell (Figure 2.14a). The
molecules formed an extended V-shaped conformation, with the oleoyl and
stearoyl chains coming off the two ends of the glycerol group with an angle of
94º between their planes. The combination of the T⏐⏐ subcell and the trans-skew-
cis-skew-skew-trans (T-S-C-S-S-T) conformation at carbons 7–13 in the oleic
acid leaflet is of particular interest. Hydrogen bonding formed within the
lamellar interface between the oxygen atom of the glycerol sn-2 carbon and the
–C=O group of the oleoyl chain.
There are eight polymorphic forms of sn-1,2-SODG in the dry state: γ2, γ1, α,
SOLID-STATE BEHAVIOUR OF POLYMORPHIC FATS AND FATTY ACIDS 27
Figure 2.13 Crystal shapes of the HT forms of linoleic acid and α-linolenic acid.
a
b b
(a) (b) (c)
Figure 2.14 Crystal structures of: (a) sn-1,3-SODG; (b) sn-1,2-DPDG; (c) sn-1,2-SODG.
β4, β3, β2, β1, and β′. Interestingly, the most stable form of sn-1,2-SODG is β′
(Tm=25.7ºC), and the second most stable form is β1 (Tm=23.1ºC). This contrasts
with triacylglycerols (TAGs), whose most stable form is usually β with the T⏐⏐
subcell (to be discussed below). A structure model of the β′ form of sn-1,2-
SODG was constructed as depicted in Figure 2.14c, based on the crystal
structure of 1,2-dipalmitoyl-sn-glycerol (sn-1,2-DPDG) in Figure 2.14b
(Dorset & Pangborn, 1988). A hairpin conformer structure formed in the
monoclinic crystal of sn-1,2-DPDG since two palmitoyl chains were aligned in
the same leaflet with an angle of 92.9º with respect to the lamellar plane.
Hydrogen bonding formed between the oxygen atom of the glycerol sn-3
28 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
carbon and the –C=O group of the sn-1 palmitoyl chain. These structures are
basically the same as that of 1,2-dilauroyl-sn-glycerol. An interesting point is
that the palmitoyl chains in sn-1,2-DPDG are packed according to the O⊥
subcell. X-ray powder diffraction data indicate that the β′ form of sn-1,2-
SODG is stacked in a double chain-length structure. However, we cannot
determine whether a hairpin structure like that of sn-1,2-DPDG forms, or a
V-shaped structure like that of sn-1,3-SODG. Figure 2.14c postulates that a
hairpin structure might be formed by decreasing the steric hindrance between
the stearoyl and oleoyl chains that are packed in the same leaflet.
There are four polymorphs of sn-1,2-SLiDG in the dry state: α, sub-α1,
subα2, and β′. The two sub-α forms are metastable low-temperature polymorphs,
and the melting points of the α and β′ forms are 11.6ºC and 16.1ºC,
respectively. Hydrated sn-1,2-SLiDG possesses three forms: αw, sub-αw1, and
sub-αw2. The subcell packing is hexagonal for the α form and O⊥ for the β′ form.
sn-1,2-SLiDG packs much less efficiently than sn-1,2-DPDG, but appears to
pack more efficiently than sn-1,2-SODG.
3. Triacylglycerols
The most naturally abundant non-polar lipids are triacylglycerols (TAGs),
waxes, and cholesterol esters. TAGs are discussed in this section as a repre-
sentative group of non-polar lipids, since TAGs are the predominant constituents
of fats and oils. A TAG is a three-fold ester of glycerol and fatty acids. TAGs
can be divided into three classes with respect to their fatty acid composition.
TAGs with only one type of fatty acid are called monoacid TAGs; those with
two or three types of fatty acids are called diacid and triacid TAGs, both of
which are referred to in this section as mixed-acid TAGs. Diacid TAGs are
further divided into two types, symmetric and asymmetric. Chiral properties
are revealed in the asymmetric diacid TAGs, and also in triacid TAGs.
The physical properties of TAGs are determined by the types of their fatty
acid moieties, e.g. saturated and unsaturated chains, cis and trans double
bonds, short and long chains, chains with even and odd numbers of carbons,
and the positions of the esterified fatty acids on the glycerol carbon atoms.
Knowing the fatty acid compositions of TAGs enables us to modify the
molecular shapes of the fats through hydrogenation, interesterification,
fractionation, or genetic engineering, to obtain desirable physical properties for
fat-based products.
TAG molecules adopt the ideal conformation and arrangement in relation to
their neighbours in the crystalline state to optimize the intra- and intermolecu-
lar interactions and achieve efficient close-packing states. Such states are
characterized by a packing of long-chain molecules within a plane normal to
the chain axis, called lateral packing and defined by subcell, and also by a
stacking of lamellae with a specific lateral packing along the chain axis, called
SOLID-STATE BEHAVIOUR OF POLYMORPHIC FATS AND FATTY ACIDS 29
chain-length (Sato et al., 2001) and hexa chain-length (Kodali et al., 1989)
structures have been observed in TAGs with complicated fatty acid composi-
tions. We focus on the polymorphic structures of mixed-acid TAGs in this
section, since saturated monoacid TAG structures were summarized in a recent
article (van Langevelde et al., 1999a).
Figure 2.16 Melting points () and long-spacing values () of asymmetric diacid TAGs of the form
PPn (see text for details).
of two β forms of SOS (Arishima & Sato, 1989) was recently confirmed by a
powder X-ray diffraction study (Peschar et al., 2004). Figure 2.18b depicts the
melting behaviour of the polymorphic forms of 1,3-dipalmitoyl-2-sn-oleoyl-
glycerol (POP), 1,3-palmitoyl-stearoyl-2-oleoyl-rac-glycerol (POS) (Arishima
et al., 1991; Koyano et al., 1991), 1,3-stearoyl-arachidoyl-2-oleoyl-rac-glyc-
erol (AOS) (Arakawa et al., 1998), 1,3-diarachidoyl-2-oleoyl-glycerol (AOA),
and 1,3-dibehenoyl-2-sn-oleoyl-glycerol (BOB) (Wang et al., 1987), which is
basically the same as that of SOS. Their primary polymorphic transformation
properties are summarized below, mostly as referred to SOS, which was
studied with synchrotron radiation X-ray diffraction (Ueno et al., 1997),
vibrational spectroscopy (Yano et al., 1993, 1999) and solid-state nuclear
magnetic resonance (NMR) (Arishima et al., 1996) techniques.
The α forms of all the TAGs shown in Figure 2.18b revealed the same crystal
structure, having the double chain-length arrangement. This indicates that the
stearoyl and oleoyl moieties in the same leaflet are packed in accordance with
32 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
Figure 2.18 (a) Molecular model of polymorphism of SOS. (b) Melting points of polymorphic forms
of POP, POS, SOS, AOS, AOA, and BOB (: α, : γ, : β′, Δ: β2, : β1).
the H subcell, and that disordered, non-specific chain packing of the hexagonal
subcell may not cause steric hindrance between saturated and oleic acid
moieties. The α form transforms to the γ form of the triple chain-length
structure, in which the oleoyl and stearoyl leaflets are separated as a result of
chain sorting. The stearoyl leaflet is assumed to be parallel packed, and the
oleoyl leaflets retain a hexagonal subcell structure. The total chain of SOS in
the γ form is arranged normal to the lamella interface. The triple chain-length
structure in the β′ form is maintained in all the above TAGs except for POP, in
which the β′ form has a double chain-length structure. The stearoyl leaflet in the
β′ form is packed according to the O⊥ subcell, while the hexagonal subcell
structure is retained in the oleoyl leaflet. The SOS chains are inclined with
respect to the lamella plane. Finally, the β′ form transforms to two β forms,
SOLID-STATE BEHAVIOUR OF POLYMORPHIC FATS AND FATTY ACIDS 33
which reveal an inclined chain arrangement against the lamellae interface with
long-spacing values of 6.75 nm (β2) and 6.60 nm (β1) for SOS. The subcell
structure is T⏐⏐ for both the stearoyl and the oleoyl leaflets in the β1 form. It is
assumed that the subcell structures of the three leaflets of the β2 form are similar
to T⏐⏐, but the molecular conformation of the oleoyl leaflet may differ from that
of β1.
The differences in the conversion of the subcell structure between the
stearoyl and oleoyl leaflets of SOS were analysed by a polarized FTIR (Fourier
transform infrared) technique. Infrared CH2 scissoring and CH2 rocking re-
gions are good indicators of subcell packing (Yano et al., 1999). The bands of
oleoyl leaflets overlapped those of stearoyl leaflets for the usual hydrogenated
specimen, and thus partial deuteration was attempted so that the stearoyl chains
would be deuterated and the oleoyl chains would be hydrogenated to form
SDOHSD. The FTIR scissoring and rocking spectra of the stearoyl and oleoyl
leaflets of SDOHSD could thus be separated. A single band was observed for both
the SD chain and the OH chain of the α form, corresponding to the hexagonal
subcell. A sharp single band for the SD chain and a broad band for the OH chain
were observed in the γ form, suggesting that the stearoyl groups form parallel
packing and the oleoyl moiety packs in the hexagonal subcell. The scissoring
and rocking spectra of the SD chain in the β′ form exhibited two components,
indicating a O⊥ subcell. Parallel packing was indicated in the oleoyl leaflet in
the β′ form, since no splitting occurred. Both the stearoyl and oleoyl moieties
were packed in the T⏐⏐ subcell in the β1 form. Knowledge of the polymorphic
structures of SOS and other homologues is very important for the confectionery
industry, since the stable polymorphs of cocoa butter, which contains ~80%
POP, POS, and SOS, are thought to be identical to the two β forms of SOS.
Mykhaylyk & Hamley (2004) and Mykhaylyk et al. (2004) recently carried
out an analysis of the electron density profiles of the chain-length structures of
SOS polymorphs, using long-spacing X-ray diffraction patterns. Their findings
supported the structure models depicted in Figure 2.18a. A crystal structure
analysis of the β2 form of SOS performed using powder X-ray diffraction
methods confirmed the existence of two β forms in SOS (Peschar et al., 2004).
However, the kink model at the cis double bond depicted in Figure 2.18a was
not observed. It was revealed instead that the oleoyl and stearoyl chains were
both relatively straight.
The polymorphism of the series of saturated-oleic-saturated mixed-acid
TAGs summarized in Figure 2.18 is notable in cocoa butter (CB), which is the
most popular confectionery fat. CB consists of three major TAGs (POP, POS,
and SOS), plus other minor components (Timms, 2003). The three TAGs
determine the polymorphic nature of CB, which exhibits six polymorphs,
Form I through Form VI (Wille & Lutton, 1966). Form V is important for the
functionality of chocolate, and thus crystallization of CB in Form V and
preservation of this polymorph during long storage are prerequisites for quality
34 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
Table 2.6 Thermal data for SSS, SOS, SRS, and SLiS polymorphsa
Figure 2.19 Structure models of: (a) the β′ form of SRS; (b) the α and γ forms of SLiS.
The interactions among the linoleoyl chains at the sn-2 position in SLiS may
stabilize the γ form, prohibiting transformation into the more stable forms β′ or
β. The transformation from γ to β′ or β is associated with the chain inclination
with respect to the lamellar interface, which may be prohibited by the chain–
chain interactions between the linoleoyl moiety with two cis double bonds
(Figure 2.19 b). For this reason, the enthalpy and entropy values for melting of
the γ form of SLiS are much greater than those of SOS and SRS.
Atomic-level structure analyses have not been successful despite the impor-
tance of diacid TAGs that are abundantly present in natural fats, except for the
two cases discussed below. It is worth noting that the structure models cited
above are only hypothetical and are based on powder X-ray diffraction and
infrared absorption data.
Crystal structures of β′ form
Determination of the atomic-level crystal structure using single crystals and
X-ray diffraction can provide the most precise structural information, although
infrared absorption, Raman scattering, solid-state NMR spectroscopy, and
X-ray diffraction studies using polycrystalline samples or oriented samples
also provide valuable information.
36 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
Figure 2.20 (a) Structure of the β form of tricaproyl-glycerol (CCC). (b) Structure of the β form of sn-
1,2-dipalmitoyl-3-acetoyl-glycerol (PP2). (c), (d), (e) Possible structure models for the β′ form of
triacylglycerols.
The first atomic-level structure determination was performed for the β forms
of tricaproyl-glycerol (CCC) (Jensen & Mabis, 1963, 1966) and trilauroyl-
glycerol (LLL) (Larsson, 1964), both of which form a double chain-length
structure (Figure 2.20a). Goto et al. (1992) found a new crystal structure with
a triple chain-length structure composed of palmitoyl–acetoyl–palmitoyl leaf-
lets for the β form of the diacid TAG PP2 (sn-1,2-dipalmitoyl-3-acetoyl-glycerol)
(Figure 2.20b). Furthermore, atomic-level analyses of β′ crystals have been
reported for two saturated diacid TAGs, CLC (C10C12C10; 1,3-didecanoyl-
2-dodecanoyl-glycerol) (van Langevelde et al., 2000) and PP14 (1,2-palmitoyl-
3-myristoyl-sn-glycerol) (Sato et al., 2001). Surprisingly, the two β′ crystals
exhibited remarkably different molecular features, probably because CLC is a
symmetric-type diacid TAG and PP14 is an asymmetric-type diacid TAG.
Many researchers have argued about β′ structure, and proposed structure
models are summarized in Figures 2.20c–e (Hernqvist & Larsson, 1982;
Hernqvist, 1988; Birker et al., 1991; Birker & Blonk, 1993; van Langevelde et
al., 1999b; van de Streek et al., 1999). Distinctive differences among the three
models appear in the chain inclination and methyl-end stacking. Figure 2.20c
indicates that all the extended straight chains involving the glycerol groups are
synchronously inclined with respect to the lamella plane. In fact, this structure
was observed in CCC and LLL. In contrast, the aliphatic chains are alternately
inclined against the lamella plane at the methyl end (Figure 2.20d) or at the
glycerol group (Figure 2.20e).
The β forms of CCC and PP2 in Figures 2.20a and 2.20b did not reveal a flat
methyl-end stacking mode arrangement; a zigzag stacking mode comprising a
terrace and step was revealed instead. The angle made by the terrace plane and
the lamella plane, defined as θ, was 12º in CCC and 11º in PP2. The chain
SOLID-STATE BEHAVIOUR OF POLYMORPHIC FATS AND FATTY ACIDS 37
Figure 2.21 Crystal structures of: (a) the β′ form of CLC; (b) the β′2 form of PP14.
inclination structure depicted in Figure 2.20d was observed in the β′2 form of
PP14, whereas the β′ form of CLC displayed the structure depicted in
Figure 2.20e. Interestingly, the formation of each chain inclination is closely
related to the unique pattern of the methyl-end stacking mode.
CLC is a CnCn+2Cn-type TAG in which β′ is the most stable polymorph, as
described above. Melt-grown single crystals were analysed with a synchrotron
radiation X-ray beam to uncover an orthorhombic structure with a space group
of Iba2 with the unit cell lattice parameters and molecular structures depicted
in Figure 2.21a. The CLC molecules were packed in a chair conformation, in
which zigzag planes of the acyl chains were orthogonally packed and displayed
an O⊥ subcell structure.
The crystal structures of the β forms revealed in CCC and LLL and the β′
forms of CLC differ in the following three aspects, in addition to having
different subcell structures.
(1) The TAG molecules in the β form have an asymmetric tuning-fork
conformation, whereas the CLC molecules in the β′ form adopt a chair
conformation.
38 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
(2) The TAG molecules in the β form are all straight at the glycerol moieties,
whereas the CLC molecules in the β′ form are bent at the glycerol
moiety.
(3) The mode of methyl-end stacking is straight and flat.
Of the two stable β′ forms of PP14 (β′1 and β′2), a single crystal of β′2 was
crystallized from solution and its atomic-level structure was analysed
(Figure 2.21b). The primary structural properties of the β′2 form of PP14 can be
summarized as follows.
(1) A unit lamella revealed a quatro chain-length structure consisting of two
double-layer leaflets.
(2) The two double-layer leaflets were combined end-by-end in a unit
lamella, and the chain axes were alternately inclined against the lamellar
interface.
(3) The methyl-end stacking mode was largely different, a less stepped
structure at the outer plane (θ1 = 9.6º) and a very stepped structure at the
inner plane (θ2 = 38º).
(4) The two asymmetric units revealed different glycerol conformations,
trans for sn-1(P, palmitic) and sn-2(P) but gauche for sn-3(M, myristic)
in A, and trans for sn-2(P) and sn-3(M) but gauche for sn-1(P) in B
(Figure 2.22a).
(5) The outer methyl end consists of all the palmitic acid chains (–PPP–),
whereas the inner methyl end consists of palmitic–myristic–myristic
chains (–PMM–).
(6) Two asymmetric units, referred to as A and B, formed a hybrid-type
orthorhombic perpendicular subcell (HS3) (Figure 2.22 b), which was
different from the usual O⊥ type depicted in Figure 2.4 that was observed
for the β′ form of CLC.
The β′2 form of PP14 exhibited other unique properties of dynamic molecular
movements, expressed in the thermal movements of fatty acid chains of
different leaflets and the conformational stability of the glycerol group (Sato
et al., 2001), and the growth and equilibrium of the crystal morphology
(Hollander et al., 2003).
We briefly compared the two atomically analysed β′ forms. The CLC β′ form
displayed a bent glycerol conformation and straight methyl-end stacking,
whereas a straight glycerol conformation and stepped methyl-end stacking at
the inner interfaces of the four chain-length structures were revealed in the β′2
form of PP14. Arranging the lauric acid chain at the sn-2 position and the capric
acid chains at the sn-1 and sn-3 positions in the CLC β′ form may produce very
smooth methyl-end stacking. This would occur because the sn-2 chain, which
is longer than the sn-1 and sn-3 chains by two carbon atoms, does not form the
stepped methyl-end stacking mode that was revealed in the β′2 form of PP14,
SOLID-STATE BEHAVIOUR OF POLYMORPHIC FATS AND FATTY ACIDS 39
C1 6 C14
C14
O
O
as
O
O O 3
O 1
2 O 2
3
1 O O
O O
bs
O
C16
C16 HS3
A C16 B
Figure 2.22 (a) Asymmetric units, A and B, and (b) hybrid subcell of a PP14 β′2 crystal.
CCC, and PP2. However, the bent conformation at the glycerol group may be
accompanied by some excess lattice energy. We assume that the bent glycerol
conformation must have orthogonal-packed zigzag aliphatic chains, whereas
the parallel-packed chains in the β form do not require a bent glycerol
conformation. The uniqueness of the β′2 form of PP14 is revealed in the presence
of two asymmetric units in a unit cell, the chain-end stacking mode, the lateral
chain packing expressed in the HS3 subcell, the alternate chain inclination
against the lamella interface, and the straight glycerol conformation.
The fact that the β′ form is stabilized when a TAG contains different fatty
acid moieties (diacid TAGs), such as CLC and PP14, may indicate a general
tendency. Milk fat is a natural fat that is fairly stable in the β′ form; its stability
may be attributed to the presence of a high concentration of mixed-acid TAGs.
de Man (1999) indicated that the following factors are prerequisites for
stabilization of the β′ form, based on his observations: (1) fatty acid chain
length diversity, (2) TAG carbon number and diversity, (3) TAG structure, (4)
concentration of liquid oil and (5) temperature fluctuation. The first three
factors are of a molecular nature, which may be partially explained by
stabilization of the methyl-end stacking (as revealed in the β′2 form of PP14)
and/or bending at the glycerol group (as revealed in the β′ form of CLC).
Therefore, the following suggestions may conceivably explain a β′–β transfor-
mation, considering the β′ and β structures presently clarified.
The transformation of CLC from the β′ form causes conformational changes
in the glycerol groups, combined with a rotation of the half-leaflets between the
glycerol and methyl-end groups around the lamella plane normal, and conver-
sion of the subcell structure from O⊥ to T⏐⏐. The transformation of PP14 from the
β′2 form causes rotation of the double-layer leaflets I and II around the lamella
40 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
glycerol conformation
plane normal and conversion of the subcell structure from hybrid orthorhombic
to T⏐⏐. Therefore, stabilization of the β′ structure may be achieved by increasing
the activation energies necessary for all or some of the molecular movements
associated with the β′–β transformation described above. Clarification of these
processes is very important and further basic research is needed. Furthermore,
diversity in β′ structure should be elucidated; for example, atomic-level
understanding of the β′1 form of PP14 has not been achieved, although a
spectroscopic vibrational study revealed the presence of a O⊥ subcell (Yano et
al., 1997).
(2) Glycerol conformation among the glycerol groups determines the total
configuration (straight or bent) of the TAG molecule. Crystallographic
clarification of membrane lipid structures has revealed diverse varieties
of glycerol conformations of phospholipids and other polar lipids,
which are influenced by glycerol–glycerol interactions (Pascher et al.,
1981; Hauser et al., 1988; Pascher, 1996). This interaction is replaced by
carboxyl groups for fatty acids.
(3) Methyl-end stacking may play important roles in organizing the chain
inclination and chain-length structures. Methyl-end stacking in
combination with the glycerol conformation determines formation of
the triple chain-length structure.
These three factors interrelate in a complicated manner. For example,
optimum stabilization of the methyl-end stacking is related to the quatro chain-
length structure and the formation of HS3 subcell structures in the crystal
structure of the β′2 form of PP14. Stabilization of the stearic and oleic acid chains
causes a conversion from double to triple chain-length structures in SOS. The
same effect forms an interdigitated chain-length structure at the expense of
methyl-end stacking in the β1 form of oleic acid. It is very important to improve
our quantitative understanding of how the polymorphism of TAGs is controlled
by structural stabilization at the molecular level.
Figure 2.24 Typical three-phase diagrams of binary mixtures of two components, A and B: (a) solid
solution phase formation; (b) eutectic phase formation; (c) molecular compound (C) formation. S, solid;
L, liquid.
(Wille & Lutton, 1966; Rousset et al., 1998), and SSS–SSE (E; elaidoyl)
(Rossel, 1967; Timms, 1984) revealed eutectic and miscible properties in quite
complicated manners. However, this is not the case between PPP and LLL, in
which eutectic phases are formed in the three polymorphs (Takeuchi et al.,
2003). Mixtures of unsaturated TAGs and saturated TAGs exhibited eutectic
properties, but molecular compound formation occurred only in special combi-
nations of mixtures of saturated–unsaturated mixed-acid TAGs, as reviewed by
Sato et al. (1999). This section discusses new results regarding different types
of binary mixtures of TAGs.
Figure 2.25 Binary phase behaviour of (a) MMM–LLL and (b) PPP–LLL mixtures.
occurred separately for LLL concentrations between 50% and 90%, indicating
that phase separation occurred in the three polymorphic forms; (3) the LLL
fraction was dissolved in the PPP fraction for LLL concentrations below 50%.
Hence, the phase behaviours of the LLL–PPP mixtures were mainly governed
by PPP. An SSS–LLL mixture exhibited similar properties.
These results enable us to draw the following conclusions. The metastable
44 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
forms have miscible phases when the carbon numbers (Cn) of the fatty acids of
the component TAGs differ by two, but the most stable form has a eutectic
phase. Neither TAG is miscible in any polymorphic form when Cn differs by
four or six.
Figure 2.26 Phase diagrams of binary mixtures of (a) PPP–POP and (b) POP–OPO.
45
46 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
SSO OSO
Figure 2.27 Structure models of stable polymorphs of SOS, SSO, and OSO and the molecular
compound crystals of SOS–SSO and SOS–OSO.
(a) (b)
40 α SOS/SLiS
liquid
Temperature (°C)
30
γ
γ SOS/SLiS
20
α
0 20 40 60 80 100
SLiS concentration (%)
Figure 2.28 (a) Phase behaviour of the SOS–SLiS mixture. (b) Structure model of the co-crystal of
SOS and SLiS.
fν2γ3
ΔG# = –––––2 (2)
(Δμ)
ΔG#A ⎛ γA ⎞ 3 ⎛ ΔSfBΔTB ⎞ 2
–––– = ⎜ –– ⎟ ⎜ ––––––– ⎟ (4)
ΔG#B ⎝ γB ⎠ ⎝ ΔSfAΔTA ⎠
where the superscripts of γA and γB, etc. indicate the values of interfacial energy
and other factors for the A and B forms. The term [(ΔSfBΔTB)/(ΔSfAΔTA)]2 in
Equation 4 is always greater than 1 since metastable A forms have lower
melting points and smaller ΔSf values than the stable B form. This gives rise to
a significantly higher activation energy for nucleation for A compared with B
SOLID-STATE BEHAVIOUR OF POLYMORPHIC FATS AND FATTY ACIDS 49
Figure 2.29 Schematic illustration of activation energy for nucleation (ΔG#) of two crystal forms
(metastable A and stable B) with varying supercooling (ΔT).
if the interfacial energy values are the same for the two forms. However, the
ratio of the ΔG# values of A and B is determined by conflicting factors of
supercooling and interfacial energy if we reasonably suppose that the γ value of
the low-enthalpy form B is greater than that of the high-enthalpy form A.
Two extreme cases are worthy discussion topics: high supercooling with low
Tc, and low supercooling with high Tc. The ratio of (ΔSfBΔTB)/(ΔSfAΔTA)
approaches 1 in the former and form A is preferably nucleated since its
activation energy for nucleation becomes smaller than that of B. In contrast, the
ratio of (ΔSfBΔTB)/(ΔSfAΔTA) becomes smaller than 1 when supercooling is
decreased, resulting in a smaller ΔG# value and a greater nucleation rate for the
metastable A form than the stable B form. This argument is illustrated in
Figure 2.29.
Figure 2.30 provides an example of polymorph-dependent nucleation of a
TAG. This figure depicts the small-angle X-ray diffraction patterns of the β′
and β forms of LLL, which were monitored by a synchrotron radiation X-ray
beam at isothermal crystallization temperature (Tc) (Ueno et al., 2003b). A
small-angle X-ray diffraction line of 3.35 nm, corresponding to presence of the
β′ form, appeared earlier at Tc = 30ºC than that of the β form (3.25 nm). The
induction time, τ, was obtained by measuring the time interval after the
temperature of LLL reached Tc until the small-angle diffraction lines of the two
forms were detectable. The τ values thus obtained at Tc = 25.0ºC and 30.0ºC are
provided in Table 2.7, together with the values of Tf and ΔSf.
50 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
Figure 2.30 Synchrotron radiation small-angle X-ray diffraction (SR-XRD) of polymorphic crystalli-
zation of the β′ and β forms of trilaurin, taken during isothermal crystallization at Tc = 30ºC: (a) without
ultrasound stimulation; (b) with ultrasound stimulation. Details of first-occurring diffraction patterns are
inserted.
The τ value of the β′ form was always smaller than that of the β form. The
results in Table 2.7 indicate that ΔG# was always smaller for the β′ form than for
the β form during crystallization at 25.0ºC and 30.0ºC, since τ is inversely
proportional to the nucleation rate function (J–1). We can use the values of ΔT
and ΔSf in Table 2.7 to calculate that (ΔSfβΔTβ)/(ΔSfβ′ΔTβ′) is 3.2 at Tc = 25.0ºC
and 5.6 at Tc = 30.0ºC. The fact that (ΔSfβΔTβ)/(ΔSfβ′ΔTβ′) is greater than 1 at the
two crystallization temperatures indicates that the thermodynamic driving
force for nucleation is greater for the β form than for the β′ form. This contrasts
with the experimental result and suggests that the smaller interfacial energy for
the β′ form may exceed the thermodynamic driving force for nucleation, as
illustrated in Figure 2.29. A similar observation was made for POP, SOS
(Koyano et al., 1989) and PPP (Sato & Kuroda, 1987).
Table 2.7 Thermal data and crystallization properties of β′ and β polymorphs of trilaurin without and with ultrasound stimulationa
Tc = 30.0ºC Tc = 25.0ºC
Ultrasound stimulation
Ultrasound at high power and low frequencies can generally assist various
processes in food technology (Povey & Mason, 1998). It has been reported that
high-power ultrasound stimulation remarkably influences both nucleation and
crystal growth by creating additional fresh nucleation sites in the crystalliza-
tion medium (Eskin, 1994). Ultrasound stimulation for a short period at
supercooled conditions accelerated the crystallization of Form V in cocoa
butter (Baxter et al., 1995). We discuss here the effects of ultrasound stimula-
tion on the polymorphic crystallization of fats, which was recently studied for
triacylglycerols (Ueno et al., 2003a,b), cocoa butter (Higaki et al., 2001) and
palm oil (Patrick et al., 2004).
The crystallization behaviour of PPP and LLL in the presence of ultrasound
stimulation was investigated in situ using time-resolved SR-XRD measure-
ments. The results for LLL with ultrasound stimulation at an exposure time of
2 seconds with a power of 100 W are provided in Figure 2.30b and Table 2.7,
together with those without ultrasound stimulation. Figure 2.30b shows a
marked decrease in the induction times (increased nucleation rates) for crystal-
lization of both PPP and LLL, and indicates that crystallization of the most
stable β form was promoted rather than the β′ form as Tc increased (supercool-
ing decreased). However, ultrasound stimulation with a longer exposure time
delayed the crystallization rate due to the rise in sample temperature caused by
the absorption of ultrasound dissipation energy. Similar results were observed
for PPP. This conflicting behaviour indicates that there is an optimal condition
for the acceleration of crystallization by ultrasound stimulation. A pronounced
decrease in the induction time for nucleation results from the melting point shift
due to high-pressure pulses associated with collapsing bubbles. The mecha-
nism that clarifies the preferred nucleation of the β polymorph over the β′ form
may involve conflicting factors, such as the effects of pressure on the melting
temperature of the two polymorphs, which prevailed for the β′ form, and the
effects of rising temperature on stabilization of the crystal nuclei, which
prevailed for nucleation of the β form, but the details are unclear. Thus, details
of the crystallization processes of fats subjected to ultrasound stimulation are
open to interpretation.
SOLID-STATE BEHAVIOUR OF POLYMORPHIC FATS AND FATTY ACIDS 53
Figure 2.31 Time variation of the diffraction intensity of SR-XRD small-angle peaks of polymorphic
forms of cocoa butter. Open symbols, without shear; closed symbols, with shear.
Shear field
Shear is applied incidentally or purposely during the crystallization procedure
in the actual production processes of fat-based products. The effect of applying
a shear field on fat crystallization was first examined in cocoa butter using a
viscometer and DSC under varying shear rates (Feuge et al., 1962). The
polymorphic forms of crystallized cocoa butter (CB) converted from metastable
to more stable forms with an increased shearing rate, and the crystallization rate
was increased (Ziegleder, 1985). Stapley et al. (1999) examined the effect of
shear on the solidification of chocolate. Recent studies using SR-XRD have
demonstrated that transformations from metastable to more stable forms,
particularly to Form V, are accelerated by high shear stress (MacMillan et al.,
2002; Mazzanti et al., 2003, 2004).
Figure 2.31 depicts the time variation of relative intensities of X-ray diffrac-
tion peaks of CB crystals formed after cooling from 50ºC to 18ºC at a rate of
3ºC/minute. Form III appeared first without shear, after which Form IV crys-
tallized at the expense of Form III. In contrast, an accelerated transformation
from Form III to Form V was caused by applying shear stress at 1440 s–1,
without the occurrence of Form IV. The same result was observed with lower
shear rates, and the persistence time of Form III was reduced as the shear rate
was increased. Mazzanti et al. (2003) also observed that the CB crystals were
aligned and that a new phase was created in the shear field. Similar observa-
tions were obtained for milk fat and palm oil.
The effects of the shear field on fat crystallization are considered to involve
changes in the crystallization pathways by means of modification of the mass
54 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
transfer conditions to the crystal surface via the boundary layer, segregation of
individual crystallites or some hindering of cluster formation, improved heat
transfer, and increased crystallite collisions. However, further research is
needed to obtain deeper insight into the effects of shear field on fat crystalliza-
tion.
Figure 2.32 Three types of heterogeneous nucleation occurring in o/w emulsion droplets.
size (Dickinson et al., 1991), the type of emulsifier (McClements et al., 1993;
Palanuwech & Coupland, 2003), droplet–droplet interactions (Hindle et al.,
2000), polymorphism (Ueno et al., 2003a), the effects of additives (Katsuragi
et al., 2001), cooling rate (Lopez et al., 2002), and temperature variation
(Vanapalli et al., 2002). This section briefly discusses the basic properties of
the crystallization processes of fat crystals in o/w emulsion droplets.
(a)
1.2
1.0
0.8
Z V/Z V, m
0.6
0.4
0.2
0
0 0.2 0.4 0.6 0.8 1.0 1.2
r (µm)
(b)
1440
USV (m/sec)
1430
1420
1410
0 400 800 1200
t (min)
Figure 2.33 (a) Size distribution of palm oil-in-water emulsion droplets. Bars indicate the actual size
distribution. ZV and ZV, m are the size distribution function and its average value, respectively, and r is the
diameter of the emulsion droplets. (b) Change in ultrasound velocity (USV) with time during crystal
nucleation and growth. Closed circles are experimental data. Solid and dotted curves are calculated based
on poly-dispersed and mono-dispersed emulsions, respectively.
SOLID-STATE BEHAVIOUR OF POLYMORPHIC FATS AND FATTY ACIDS 57
motion and gravity in the emulsion state. Thermal movement and droplet–
droplet interactions cause various destabilization mechanisms, in which
crystallization due to droplet–droplet interactions plays an important role. The
already-crystallized droplets that collide with other droplets containing super-
cooled liquid cause inter-droplet nucleation, provided that the crystal phase
comes into contact with the supercooled liquid. Figure 2.32 depicts the three
types of heterogeneous nucleation.
Figure 2.34 (a) Molecular structure of polyglycerol fatty acid esters (PGFEs) with 10 glycerol groups.
R, fatty acid moiety. (b) Temperature variations of the ultrasonic velocity (USV) values of PMF–water
emulsions without (‘pure’) and with the addition of PGFEs with 3, 6, 9 or 12 esterified stearic acid
moieties. The additive concentration was 1% (w/w) with respect to the PMF.
Water phase
Interface
heterogeneous nucleation
fat
crystals
Volume
Oil phase heterogeneous nucleation
Template of
Tween20 additive
additive
Figure 2.35 Schematic illustration of two types of heterogeneous nucleation processes caused by
PGFE additives.
β A: sol
Tf
α
Tc
Time B: gel
Figure 2.38 Formation of a β-fat gel from a binary mixture of FHR-B and sal fat olein.
SOLID-STATE BEHAVIOUR OF POLYMORPHIC FATS AND FATTY ACIDS 63
Figure 2.39 Phase behaviour of binary mixtures of SFO and FHR-B (α, β′, and β forms).
(3) The phase behaviour of binary mixtures of SFO and FHR-B with α, β′,
and β polymorphs indicates that rapid cooling can result in the α form
and that melt mediation into the β form occurs when Tf is set between the
melting points of the α and β forms, as illustrated in Figure 2.39.
D. Conclusions
This chapter discussed the solid-state behaviour of fats and fatty acids, begin-
ning with the fundamental aspects of the structures and transformations of
crystalline materials and ending with crystallization in complex fluid
systems.
Future research areas that may interest scientists and technologists involved
with the main subjects of this chapter include structural analysis of fats with
complicated structures, such as asymmetric mixed-acid TAGs or binary and
ternary fat mixtures. Many naturally occurring and industrially valuable fats
are made of these TAGs and their mixtures, and many of their fine structures
have been unveiled. Clarification of the kinetic processes of transformations
and mixing processes is also very important and warrants further investigation,
possibly using advanced in situ observation techniques such as a synchrotron
radiation X-ray beam. Finally, dispersed systems, such as emulsions and gels,
that exhibit micro-sized and nano-sized structures will become increasingly
significant.
References
Aquilano, D. & Squaldino, G. (2001) Fundamental Aspects of Equilibrium and Crystalliza-
tion Kinetics. In: Crystallization Processes in Fats and Lipid Systems (Garti, N. & Sato,
K., eds), Marcel Dekker, New York, USA, pp.1–51.
Arakawa, H., Kasai, T., Okumura, Y. & Maruzeni, S. (1998) Polymorphism of SOA triglerol
prepared from sal fat. J. Jpn. Oil Chem. Soc. 47, 19–24.
Arishima, T. & Sato, K. (1989) Polymorphism of POP and SOS. III. Solvent crystallization
of β2 and β1 polymorphs. J. Am. Oil Chem. Soc. 66, 1614–1617.
Arishima, T., Sagi, N., Mori, H. & Sato, K. (1991) Polymorphism of POS. I. Occurrence and
polymorphic transformation. J Am. Oil Chem. Soc. 68, 710–715.
Arishima, T., Sugimoto, K., Kiwata, R., Mori, H. & Sato, K. (1996) 13C cross-polarization
and magic-angle spinning nuclear magnetic resonance of polymorphic forms of three
triacylglycerol. J Am. Oil Chem. Soc. 73, 1231–1236.
Bailey, A.E. (1950) Melting and solidification of pure compounds. In: Melting and Solidifi-
cation of Fats, Wiley, New York, USA, pp.117–180.
Barton, P.G. & Gunstone, F.D. (1975) Hydrocarbon chain packing and molecular motion in
phospholipids bilayers formed from unsaturated lecithins. J. Biol. Chem. 250, 4470–
4476.
Baxter, J.F.G., Morris, J. & Gaim-Marsoner, G. (1995) Process for retarding fat bloom in fat-
based confectionery masses. EU Patent Application 95306833 5.
Bernstein, J. (2002) Polymorphism in Molecular Crystals, Oxford University Press, Oxford,
UK.
Birker, P.J.M.W.L. & Blonk, J.C.G. (1993) Alkyl chain packing in a β′ triacylglycerol
measured by atomic force microscopy. J. Am. Oil Chem. Soc. 70, 319–321.
SOLID-STATE BEHAVIOUR OF POLYMORPHIC FATS AND FATTY ACIDS 65
Birker, P.J.M.W.L., de Jong, S., Roijers, E.C. & van Soest, T.C. (1991) Structural investiga-
tion β′ triacylglycerols: an X-ray diffraction and microscopic study of twinned β′ crystals.
J. Am. Oil Chem. Soc. 68, 895–906.
Boistelle, R. (1988) Fundamentals of nucleation and crystal growth. In: Crystallization and
Polymorphism of Fats and Fatty Acids (Garti, N. & Sato, K., eds), Marcel Dekker, New
York, USA, pp.189–226.
Boubekri, K., Yano, J., Ueno, S. & Sato, K. (1999) Polymorphic transformations in SRS (sn-
1,3-distearoyl-2-ricinoleyl glycerol). J. Am. Oil Chem. Soc. 76, 949–955.
Cardew, P.T. & Davey, R.J. (1985) The kinetics of solvent-mediated phase transformations.
Proc. R. Soc. London A398, 415–428.
Chow, C.K. (1992) Fatty Acids in Foods and Their Health Implications, Marcel Dekker, New
York, USA.
Chronakis, I.S. (1997) Structural-functional and water-holding studies of biopolymers in low
fat content spreads. Lebensm. Wissen.-Technol. 30, 36–44.
Clark, A.H. (1992) Gels and gelling. In: Physical Chemistry of Foods (Schwartzberg, H.G.
& Hartel, R.W., eds), Marcel Dekker, New York, USA, pp.263–305.
Clarke, C. (2004) The Science of Ice Cream, The Royal Society of Chemistry, Cambridge, UK.
Coupland, J.N. (2002) Crystallization in emulsions. Curr. Opin. Coll. Interface Sci. 7, 445–450.
de Jong, S, van Soest, T.C. & van Schaick, M.A. (1991) Crystal structures and melting points
of unsaturated triacylglycerols in the β phase. J. Am. Oil Chem. Soc. 68, 371–378.
de Man, J.M. (1999) Relationship among chemical, physical, and textural properties of fats.
In: Physical Properties of Fats, Oils, and Emulsifiers (Widlak, N., ed.), Marcel Dekker,
New York, USA, pp.79–95.
de Man, J.M. & de Man, L. (2002) Texture of fats. In: Physical Properties of Lipids
(Marangoni, A.G. & Narine, S.S., eds), Marcel Dekker, New York, USA, pp.191–217.
Di, L. & Small, D.M. (1993) Physical behavior of the mixed chain diacylglycerol, 1-stearoyl-
2-oleoyl-sn-glycerol: difficulties in chain packing produce marked polymorphism. J.
Lipid Res. 34, 1611–1623.
Di, L. & Small, D.M. (1995) Physical behavior of the hydrophobic core of membranes:
properties of 1-stearoyl-2-linoleoyl-sn-glycerol. Biochemistry 35, 16672–16677.
Dickinson, E. & McClements, D.J. (1996) Advances in Food Colloids, Blackie Academic &
Professional, London, UK.
Dickinson, E., McClements, D.J. & Povey, M.J.W. (1991) Ultrasonic investigation of the
particle size dependence of crystallization in n-hexadecane–water emulsions. J. Coll.
Interface Sci. 142, 103–110.
Dorset, D.L. & Pangborn, W.A. (1988) Polymorphic forms of 1, 2-dipalmitoyl-sn-glycerol:
a combined X-ray and electron diffraction study. Chem. Phys. Lipids 48, 19–28.
Engstrom, L. (1992) Triglyceride systems forming molecular compounds. J. Fat Sci.
Technol. 94, 173–181.
Eskin, G.I. (1994) Influence of cavitation treatment of melts on the processes of nucleation
and growth of crystals during solidification of ingots and castings from light alloys.
Ultrasonic Sonochem. 1, S59–S63.
Feuge, R.O., Landmann, W., Mitcham, D. & Lovegren, N.V. (1962) Tempering triglycerides
by mechanical working. J Am. Oil Chem. Soc. 39, 310–313.
Gallegos, C. & Franco, J.M. (1999) Rheology of food, cosmetics and pharmaceuticals. Curr.
Opin. Coll. Interface Sci. 4, 288–293.
Gennis, R.B. (1989) Biomembranes: Molecular Structure and Function, Springer-Verlag,
New York, USA.
Goff, H. (1997) Ice cream. In: Lipid Technologies and Applications (Gunstone, F.D. &
Padley, F.B., eds), Marcel Dekker, New York, USA, pp.355–368.
Goto, M. & Asada, E. (1978) The crystal structure of the B-form of stearic acid. Bull. Chem.
Soc. Japan 51, 2456–2459.
66 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
Goto, M., Kodali, D.R., Small, D.M., Honda, K., Kozawa, K. & Uchida, T. (1992) Single
crystal structure of a mixed-chain triacylglycerol: 1,2-dipalmitoyl-3-acetyl-sn-glycerol.
Proc. Natl Acad. Sci. USA 89, 8083–8086.
Goto, M., Honda, K., Di, L. & Small, D.M. (1995) Crystal structure of a mixed chain
diacylglycerols, 1-stearoyl-3-oleoyl-glycerol. J. Lipid Res. 36, 2185–2190.
Gunstone, F.D. & Padley, F.B., eds (1997) Lipid Technologies and Applications, Marcel
Dekker, New York, USA.
Hachiya, I., Koyano, T. & Sato, K. (1989a) Seeding effects on solidification behavior of
cocoa butter and dark chocolate. I. Kinetics of solidification. J. Am. Oil Chem. Soc. 66,
1757–1762.
Hachiya, I., Koyano, T. & Sato, K. (1989b) Seeding effects on solidification behavior of
cocoa butter and dark chocolate. II. Physical properties of dark chocolate. J. Am. Oil
Chem. Soc. 66, 1763–1770.
Hagemann, J.W., Tallent, W.H., Barve, J.A., Ismail, I.A. & Gunstone, F.D. (1975) Polymor-
phism in single-acid triglycerides of positional and geometric isomers of octadecanoic
acid. J. Am. Oil Chem. Soc. 52, 204–207.
Hauser, H., Pascher, I. & Sundell, S. (1988) Preferred conformation and dynamics of the
glycerol backbone in phospholipids. An NMR and X-ray single-crystal analysis.
Biochemistry 27, 9166–9174.
Heertje, I. & Leunis, M. (1997) Measurement of shape and size of fat crystals by electron
microscopy. Lebensm.-Wissenschaft Technol. 30, 141–146.
Hermansson, M. (1999) The fluidity of hydrocarbon regions in organo-gels, studied by NMR
– Basic translational and rotational diffusion measurements. Coll. Surf. A 154, 303–309.
Hernqvist, L. (1988) On the crystal structure of the β′1-form of triglycerides and mechanism
behind the β′1-β transition of fats. Fat Sci. Technol. 90, 451–454.
Hernqvist, L. & Larsson, K. (1982) On the crystal structure of the β′-forms of triglycerides
and structural changes at the phase transitions liq–α–β′–β. Fette Seifen Anstrichm. 84,
349–354.
Higaki, K., Ueno, S., Koyano, T. & Sato, K. (2001) Effects of ultrasonic irradiation on
crystallization behavior of tripalmitoylglycerol and cocoa butter. J. Am. Oil Chem. Soc.
78, 513–518.
Higaki, K., Sasakura, Y., Koyano, T., Hachiyal, I. & Sato, K. (2003) Physical analyses of gel-
like behavior of binary mixtures of high-melting and low-melting fats. J. Am. Oil Chem.
Soc. 80, 263–270.
Higaki, K., Koyano, T., Hachiya, I. & Sato, K. (2004a) In-situ optical observation of
microstructure of β-fat gel made of binary mixtures of high-melting and low-melting fats.
Food Res. Int. 37, 2–10.
Higaki, K., Koyano, T., Hachiya, I., Sato, K. & Suzuki, K. (2004b) Rheological properties
of β-fat gel made of binary mixtures of high-melting and low-melting fats. Food Res. Int.
37, 799–804.
Hindle, S., Povey, M.J.W. & Smith, K. (2000) Kinetics of crystallization in n-hexadecane and
cocoa butter oil-in-water emulsions accounting for droplet collision-mediated nuclea-
tion. J. Coll. Interface Sci. 232, 370–380.
Hiramatsu, H., Sato, T., Inoue, T., Suzuki, M. & Sato, K. (1990) Pressure effect on
transformation of cis-unsaturated fatty acid polymorphs. 2. Palmitoleic acid (cis-9-
hexadecenoic acid). Chem. Phys. Lipids 56, 59–63.
Hollander, F.F.A., Boerrigter, S.X.M., van de Streek, J., Bennema, P., Meekes, H., Yano, J.
& Sato, K. (2003) Comparing the morphology of β-n.n.n with β′-n.n+2.n and β′-n.n.n-
2 triacylglycerol crystals. J. Phys. Chem. B 107, 5680–5689.
Inaoka, K., Kobayashi, M., Okada, M. & Sato, K. (1988) Stability, occurrence and step
morphology of polymorphs and polytypes of stearic acid, II. Monolamella step morphol-
SOLID-STATE BEHAVIOUR OF POLYMORPHIC FATS AND FATTY ACIDS 67
Kodali, D.R., Atkinson, D. & Small, D.M. (1990a) Molecular packing in fats, influences of
acyl chain length and unsaturation. J. Disp. Sci. Technol. 10, 393–440.
Kodali, D.R., Atkinson, D. & Small, D.M. (1990b) Polymorphic behavior of 1,2-dipalmitoyl-
3-lauroyl (PP12)- and 3-myristoyl (PP14)-sn-glycerols. J. Lipid Res. 31, 1853–
1864.
Koyano, K., Hachiya, I., Arishima, T., Sato, K. & Sagi, N. (1989) Polymorphism of POP and
SOS. II. Kinetics of melt crystallization. J. Am. Oil Chem. Soc. 66, 675–679.
Koyano, T., Hachiya, I., Arishima, T., Sato, K. & Sagi, N. (1991) Polymorphism of POS. II.
Kinetics of melt crystallization. J. Am. Oil Chem. Soc. 68, 716–718.
Koyano, T., Hachiya, I. & Sato, K. (1992) Phase behavior of mixed systems of SOS and OSO.
J. Phys. Chem. 96, 10514–10520.
Krog, N. (2001) Crystallization properties and lyotropic phase behavior of food emulsifiers.
In: Crystallization Processes in Fats and Lipid Systems (Garti, N. & Sato, K., eds), Marcel
Dekker, New York, USA, pp.505–526.
Krog, N. & Larsson, K. (1992) Crystallization at interfaces in food emulsions. Fat Sci.
Technol. 94, 55–57.
Larsson, K. (1964) The crystal structure of the β-form of trilaurin. Ark. Kemi. 23, 1–15.
Larsson, K. (1966) Classification of glyceride crystal forms. Acta Chem. Scand. 20, 2255–
2260.
Larsson, K. (1972) Molecular arrangement in glycerides. Fette Seifen Anstrichm. 74, 136–
142.
Larsson, K. (1994) Solid state behavior. In: Lipids – Molecular Organization, Physical
Functions and Technical Applications, The Oily Press, Dundee, UK, pp.7–46.
Lopez, C., Bourgaux, C., Lesieur, P., Bernadou, S., Keller, G. & Ollivon, M. (2002) Thermal
properties and structure of milk fat: 3. Influence of cooling rate and droplet size on cream
crystallization. J. Coll. Interface Sci. 254, 64–78.
Lutton, E.S. (1950) Review of the polymorphism of saturated even glycerides. J. Am. Oil
Chem. Soc. 27, 276–281.
MacMillan, S.D., Roberts, K.J., Rossi, A., Wells, M.A., Polgreen, M.C. & Smith, I.H. (2002)
In situ small angle X-ray scattering (SAXS) studies of polymorphism with the associated
crystallization of cocoa butter fat using shearing conditions. Cryst. Growth Des. 2, 221–
226.
Malkin, T. (1954) The polymorphism of glycerides. Progr. Chem. Fats Other Lipids II, 1–
50.
Malta, V., Celotti, G., Zanetti, R. & Martelli, A.F. (1971) Crystal structure of the C form of
stearic acid. J. Chem. B, 548–553.
Marangoni, A.G. (2005) Fat Crystal Networks, Marcel Dekker, New York, USA.
Marangoni, A.G. & Hartel, R.W. (1998) Visualization and structural analysis of fat crystal
networks. Food Technol. 52, 46–51.
Mazzanti, G., Guthrie, S.E., Sirota, E.B., Marangoni, A.G. & Idziak, S.H.J. (2003) Orienta-
tion and phase transitions of fat crystals under shear. Cryst. Growth Des. 3, 721–725.
Mazzanti, G., Guthrie, S.E., Sirota, E.B., Marangoni, A.G. & Idziak, S.H.J. (2004) Novel
shear-induced phases in cocoa butter. Cryst. Growth Des. 4, 409–411.
McClements, D.J., Dickinson, E., Dungan, S.R. Kinsella, J.E., Ma, J.G. & Povey, M.J.W.
(1993) Effect of emulsions type on the crystallization kinetics of oil-in-water emulsions
containing a mixture of solid and liquid droplets. J. Coll. Interface Sci. 160, 293–
297.
Minato, A., Ueno, S., Yano, J., Wang, Z.H., Seto, H., Amemiya, Y. & Sato, K. (1996)
Synchrotron radiation X-ray diffraction study on phase behavior of PPP–POP binary
mixtures. J. Am. Oil Chem. Soc. 73, 1567–1572.
Minato, M., Ueno, S., Smith, K., Amemiya, Y. & Sato, K. (1997a) Thermodynamic and
SOLID-STATE BEHAVIOUR OF POLYMORPHIC FATS AND FATTY ACIDS 69
kinetic study on phase behavior of binary mixtures of POP and PPO forming molecular
compound systems. J. Phys. Chem B 101, 3498–3505.
Minato, A., Ueno, S., Yano, J., Smith, K., Seto, H., Amemiya, Y. & Sato, K. (1997b) Thermal
and structural properties of sn-1,3-dipalmitoyl-2-oleoylglycerol and sn-1,3-oleoyl-2-
palmitoylglycerol binary mixtures examined with synchrotron radiation X-ray diffraction.
J. Am. Oil Chem. Soc. 74, 1213–1220.
Minato, M., Yano, J., Ueno, S., Smith, K. & Sato, K. (1997c) FT-IR study on microscopic
structures and conformations of POP–PPO and POP–OPO molecular compounds. Chem.
Phys. Lipids 88, 63–71.
Mullin, J.W. (2001) Phase equilibria. In: Crystallization (4th edn), Butterworth-Heineman,
Oxford, UK, pp.135–180.
Mutaftschiev, B. (2001) Some specific cases of nucleation. In: The Atomistic Nature of
Crystal Growth, Springer, Berlin, Germany, pp.249–266.
Mykhaylyk, O.O. & Hamley, I.W. (2004) The packing of triacylglycerols from SXAS
measurements: Application to the structure of 1,3-distearoyl-2-oleoyl-sn-glycerol crys-
tal phases. J. Phys. Chem. B 108, 8069–8083.
Mykhaylyk, O.O., Casteletto, V., Hamley, I.W. & Povey, M.J.W. (2004) Structure and
transformation of low-temperature phases of 1,3-distearoyl-2-oleoyl-sn-glycerol. Eur. J.
Lipid Sci. Technol. 106, 319–324.
Palanuwech, J. & Coupland, J.N. (2003) Effect of surfactant type on the stability of oil-in-
water emulsions to dispersed phase crystallization. Coll. Surf. A 223, 251-262.
Pascher, I. (1996) The different conformations of the glycerol region of crystalline acylglycerols.
Curr. Opin. Struct. Biol. 6, 439–448.
Pascher, I., Sundell, S. & Hauser, H. (1981) Glycerol conformation and molecular packing
of membrane lipids: the crystal structure of 2,3-dilauroyl-D-glycerol. J. Mol. Biol. 153,
791–806.
Pascher, I., Lundmark, M., Nyholm, P-G. & Sundell, S. (1992) Crystal structures of
membrane lipids. Biochim. Biophys. Acta 1113, 339–373.
Patrick, M., Blindt, R. & Janssen, J. (2004) The effects of ultrasonic intensity on the crystal
structure of palm oil. Ultrason. Sonochem. 11, 251–255.
Peschar, R., Pop, M.M., de Ridder, D.J.A., van Mechelen, J.B., Driesse, R.A.J. & Schenk,
H. (2004) Crystal structure of 1,3-distearoyl-2-oleoylglycerol and cocoa butter in the
β(V) phase reveal the driving force behind the occurrence of fat bloom on chocolate. J.
Phys. Chem. B 108, 15450–15453.
Povey, M.J.W. (2001) Crystallization of oil-in-water emulsions. In: Crystallization Proc-
esses in Fats and Lipid Systems (Garti, N. & Sato, K., eds,) Marcel Dekker, New York,
USA, pp.251–288.
Povey, M.J.W. & Mason, T.J. (1998) Ultrasound in Food Processing, Blackie Academic &
Professional, London, UK.
Rossel, J.B. (1967) Phase diagram of triglyceride systems. In: Advances in Lipid Research
Vol. 5 (Paolettei, R. & Kritchevsky, D., eds), Adademic Press, New York, USA, pp.353–
408.
Rousseau, D. (2002) Fat crystal behaviour in food emulsions. In: Physical Properties of
Lipids (Marangoni, A. & Narine, S.S., eds) Marcel Dekker, New York, USA, pp.219–
264.
Rousset, P., Rappaz, M. & Minner, E. (1998) Polymorphism and solidification kinetics of the
binary system POS–SOS. J. Am. Oil Chem. Soc. 75, 857–864.
Sakamoto, M., Ohba, A., Kuriyama, J., Maruo, K., Ueno, S. & Sato,K. (2004) Influences of
fatty acid moiety and esterification of polyglycerol fatty acid esters on the crystallization
of palm mid fraction in oil-in-water emulsion. Coll. Surf. B 37, 27–33.
Sato, K. & Garti, N. (1988) Crystallization and polymorphic transformation: an introduction.
70 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
In: Crystallization and Polymorphism of Fats and Fatty Acids (Garti, N. & Sato, K., eds),
Marcel Dekker, New York, USA, pp.1–7.
Sato, K. & Kobayashi, M. (1991) Structure, stability and crystal growth of polymorphs and
polytypes of long chain alphatic compounds. In: Crystals – Growth, Properties and
Applications, Vol.13: Organic Crystals I. Characterization (Karl, N., ed.), Springer-
Verlag, Berlin, Germany, pp.65–108.
Sato, K. & Kuroda, T. (1987) Kinetics of melt crystallization and transformation of
tripalmitin polymorphs. J. Am. Oil Chem. Soc. 64, 124–127.
Sato, K. & Suzuki, M. (1986) Solvent crystallization of α, β and γ polymorphs of oleic acid.
J. Am. Oil Chem. Soc. 63,1356–1359.
Sato, K., Morishita, H. & Kobayashi, M. (1988) Stability, occurrence and step morphology
of polymorphs and polytypes of stearic acid, I. Stability and occurrence. J. Cryst. Growth
87, 236–242.
Sato, K., Arishima, T., Wang, Z.H., Ojima, K., Sagi, N. & Mori, H. (1989) Polymorphism
of POP and SOS. I. Occurrence and polymorphic transformation. J. Am. Oil Chem. Soc.
66, 664–674.
Sato, K,.Yoshimoto, N., Suzuki, M., Kobayashi, M. & Kaneko, F. (1990) Structure and
transformation in polymorphism of petroselinic acid (cis-ω-12-octadecenoic acid). J.
Phys. Chem. 94, 3180–3185.
Sato, K., Yano, J., Kawada, I. & Kawano, M. (1997) Polymorphic behavior of gondoic acid
and phase behavior of its binary mixtures. J. Am. Oil Chem. Soc. 74, 1153–1159.
Sato, K., Ueno, S. & Yano, J. (1999) Molecular interactions and kinetic properties of fats.
Prog. Lipid Res. 38, 91–116.
Sato, K., Goto, M., Yano, J., Honda, K., Kodali, D.R. & Small, D.M. (2001) Atomic
resolution structure analysis of β′ polymorph crystal of a triacylglycerol: 1,2-dipalmitoyl-
3-myristoyl-sn-glycerol. J. Lipid Res. 42, 338–345.
Shipley, G.G. (1986) X-ray crystallographic studies of aliphatic lipids. In: The Physical
Chemistry of Lipids, from Alkanes to Phospholipids (Handbook of Lipid Research Vol.
4) (Small, S.M., ed.), Plenum Press, New York, USA, pp.97–147.
Small, D.M. (1984) Lateral chain packing in lipids and membranes. J. Lipid Res. 25, 1490–
1500.
Small, D.M. (1986) Glycerides. In: The Physical Chemistry of Lipids, from Alkanes to
Phospholipids (Handbook of Lipid Research Vol. 4) (Small, S.M., ed.), Plenum Press,
New York, USA, pp.345–394.
Stapley, A.G.F., Tewkesbury, H. & Fryer, P.J. (1999) The effects of shear and temperature
history on the crystallization of chocolate. J. Am. Oil Chem. Soc. 76, 677–685.
Suzuki, M., Ogaki, T. & Sato, K. (1985) Crystallization and transformation mechanisms of
α, β and γ polymorphs of ultra-pure oleic acid. J. Am. Oil Chem. Soc. 62, 1600–
1604.
Takeuchi, M., Ueno, S., Yano, Y., Floter, E. & Sato, K. (2000) Polymorphic transformation
of 1,3-distearoyl-sn-linoleoyl-glycerol. J. Am. Oil Chem. Soc. 77, 1243–1249.
Takeuchi, M., Ueno, S., Floeter, E. & Sato, K. (2002a) Binary phase behavior of 1,3-
distearoyl-2-oleoyl-sn-glycerol (SOS) and 1,3-distearoyl-2-linoleoyl-sn-glycerol (SLS).
J. Am. Oil Chem. Soc. 79, 627–632.
Takeuchi, T., Ueno, S. & Sato, K. (2002b) Crystallization kinetics of polymorphic forms of
a molecular compound constructed by SOS (1,3-distearoyl-2-oleoyl-sn-glycerol) and
SSO (1,2-distearoyl-3-oleoyl-rac-glycerol). Food Res. Int. 35, 919–926.
Takeuchi, T., Ueno, S. & Sato, K. (2003) Synchrotron radiation SAXS/WAXS study of
polymorph-dependent phase behavior of binary mixtures of saturated monoacid
triacylglycerols. Cryst. Growth Des. 3, 369–374.
Timms, R.E. (1984) Phase behaviour of fats and their mixtures. Prog. Lipid Res. 23, 1–38.
SOLID-STATE BEHAVIOUR OF POLYMORPHIC FATS AND FATTY ACIDS 71
Timms, R.E. (2003) Confectionery Fats Handbook, The Oily Press, Bridgwater, UK.
Ueno, U., Suetake, T., Yano, J., Suzuki, M. & Sato, K. (1994) Structure and polymorphic
transformations in elaidic acid (trans-ω9-octadecenoic acid). Chem. Phys. Lipids 72, 27–
34.
Ueno, S., Minato, A., Seto, H., Amemiya, Y. & Sato, K. (1997) Synchrotron radiation X-ray
diffraction study of liquid crystal formation and polymorphic crystallization of SOS
sn-1,3-distearoyl-2-oleoyl glycerol. J. Phys. Chem. B 101, 6847–6854.
Ueno, S., Miyazaki, A., Yano, J., Furukawa, Y., Suzuki, M. & Sato, K. (2000) Polymorphism
of linoleic acid (cis-9, cis-12-octadecadienoic acid) and α-linolenic acid (cis-9, cis-12,
cis-15-octadecatrienoic acid). Chem. Phys. Lipids 107, 169–178.
Ueno, S., Hamada, Y. & Sato, K. (2003a) Controlling polymorphic crystallization of n-alkane
crystals in emulsion droplets through interfacial heterogeneous nucleation. Cryst. Growth
Des. 3, 935–939.
Ueno, U., Ristic, R.I., Higaki, K. & Sato, K. (2003b) In-situ studies of ultrasound stimulated
fat crystallization using synchrotron radiation. J. Phys. Chem. B 107, 4927–4935.
Vanapalli, S., Palanuwech, J. & Coupland, J.N. (2002) Influence of crystallization on the
stability of flocculated emulsions. J. Ag. Food Chem. 50, 5224–5228.
van de Streek, J., Verwer, P., de Gelder, R. & Hollander, F. (1999) Structural analogy between
β′ triacylglycerols and n-alkanes: toward the crystal structure of β′-2 p.p+2.p
triacylglycerols. J. Am. Oil Chem. Soc. 76, 1333–1341.
van Langevelde, A., van Malssen, K., Hollander, F., Pechar, R. & Schenk, H. (1999a)
Structure of mono-acid even-numbered β′-triacylglycerols. Acta Crystallogr. B 55, 114–
122.
van Langevelde, A., van Malssen, K., Sonneveld, E.D., Pechar, R. & Schenk, H. (1999b)
Crystal packing of a homologous series β′-stable triacylglycerols. J. Am. Oil Chem. Soc.
76, 603–609.
van Langevelde, A., van Malssen, K., Dressen, R., Goubits, K., Hollander, F., Peschar, R.,
Zwart, P. & Schenk, H. (2000) Structure of CnCn+2Cn-type (n=even) β′-triacylglycerols.
Acta Crystallogr. B 56, 1103–1111.
Walstra, P., Kloek, W. & van Vliet, T. (2001) Fat crystal networks. In: Crystallization
Processes in Fats and Lipid Systems (Garti, N. & Sato, K., eds), Marcel Dekker, New
York, USA, pp.289–328.
Wang, Z.H., Sato, K., Sagi, N., Izumi, T. & Mori, H. (1987) Polymorphism 1,3-disaturated-
acyl-2-oleoylglycerols: POP, SOS, AOA, BOB. J. Jpn. Oil Chem. Soc. 36, 671–
679.
Wille, R.L. & Lutton, E.S. (1966) Polymorphism of cocoa butter. J. Am. Oil Chem. Soc. 43,
491–496.
Yano, J., Ueno, S., Sato, K., Arishima, T., Sagi, N., Kaneko, F. & Kobayashi, M. (1993) FT-
IR study of polymorphic transformations in SOS, POP and POS. J. Phys. Chem. 97,
12967–12973.
Yano, J., Kaneko, F., Kobayashi, M., Kodali, D.R., Small, D.M. & Sato, K. (1997) Structural
analyses of triacylglycerol polymorphs with FT-IR techniques: II. β′1-form of 1,2-
dipalmitoyl-3-myristoyl-sn-glycerol. J. Phys. Chem. B 101, 8120–8128.
Yano, J., Sato, K., Kaneko, F., Small, D.M. & Kodali, D.R. (1999) Structural analyses of
polymorphic transitions of sn-1,3-distearoyl-2-oleoylglycerol (SOS) and sn-1,3-dioleoyl-
2-stearoylglycerol (OSO): assessment on steric hindrance of unsaturated and saturated
acyl chain interactions. J. Lipid Res. 40, 140–151.
Yoshimoto, N., Suzuki, M. & Sato, K. (1991) Polymorphic transformation in asclepic acid
(cis-ω-octadecenoic acid) Chem. Phys. Lipids 57, 67–73.
Ziegleder, G. (1985) Verbesserte kristallization von kakaobutter unter dem einfluss eines
scherge falles. Int. Z. Lebensm. Technol. Verfallensyechm. 36, 412–418.
CHAPTER 3
Liquid-crystalline lipid–water phases
A. Structures
Figure 3.1 Formation of a liquid-crystalline phase just before melting of a crystalline polar lipid. The
hydrocarbon chain axis and the polar head groups are indicated.
Figure 3.2 A liquid-crystalline phase formed from a polar lipid (with two hydrocarbon chains in the
molecule) by the addition of water.
2. Hexagonal phases
The basic structural studies mentioned above were done on soap–water sys-
tems. At that time it was known that soap molecules associate in water into
spherical micelles, which with increased concentration become elongated.
When the concentration of such rod-shaped aggregates is increased, it is natural
to expect that they become oriented and arranged in a hexagonal structure, as
shown in Figure 3.3. There are two simple geometric ways to organize parallel
cylinders or rods, either in a two-dimensional square lattice or in a hexagonal
lattice. The hexagonal lattice has the advantage of a higher packing density per
cross-sectional unit area and a better interaction between the rods. The hexago-
nal structure was also the one found experimentally in the first-studied
soap–water system, and it is now known to be a general structure in amphiphile–
water systems when the amphiphiles are surfactants containing hydrocarbon
chains. One type of surfactant lipid that forms micelles in water is the
lysophospholipids.
76 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
Figure 3.3 (Left) A cross-section of the HI structure formed by lipids associated into infinite cylindrical
micelles. (Right) The inverse type of structure (HII), with water cylinders arranged in a continuous
hydrocarbon matrix with interfaces formed by the polar head groups.
Most lipids have hydrocarbon chain regions that require much space. A rod
structure forming a water-continuous phase is therefore not possible, whereas
the inverse type of structure is ideal. Such a structure is also shown in
Figure 3.3. The open and inverse types of hexagonal structures are termed HI
and HII, respectively.
3. Cubic phases
The main features of the bicontinuous cubic lipid–water phases, which are now
generally accepted, are based on minimal surfaces. We will therefore start by
defining some of the concepts behind minimal surfaces.
During the 19th century, a popular topic in mathematics was the analysis of
the minimum surface area defined by borders of various shapes. Soap films
could then be used as experimental models, as a film spanning a particular
frame will minimize its surface area. An example is shown in Figure 3.4.
Figure 3.4 Illustration of a soap film that spans two parallel square-shaped frames.
LIQUID-CRYSTALLINE LIPID–WATER PHASES 77
1
G = ––––
R1R2
The mathematical condition for the minimum area of a surface (such as a soap
film) spanning a certain frame is that H = 0 everywhere along the surface. This
means that the surface is as convex as it is concave at all points within the
surface.
Repetition of a minimal surface unit in three dimensions results in an infinite
three-dimensional structure. Further, if we assume that a surface formed in this
way is space-filling, continuous, and free of self-intersections, it is called a
periodic minimal surface (PMS).
The lipid bilayer of bicontinuous cubic lipid–water phases form PMS.
Earlier theoretical work showed that there are three fundamental PMS struc-
tures. One was called the diamond surface (D-surface), one discovered by
Schwarz more than 100 years ago was called Schwarz’s primitive surface (P-
surface), and one discovered by Schoen (1970) was called the gyroid surface
(G-surface). As will be shown below, these three types of structures exist in
lipids. It had been proposed at an early stage that microemulsions and liquid-
crystalline phases of surfactants might form minimal-surface types of structures
(Scriven, 1976). An experimentally based general introduction to the minimal
surface concept in structure chemistry was reported by Andersson et al. (1988).
We will now describe in detail the step-wise evolution of the cubic PMS
concept of the bicontinuous structures of lipid–water phases. In this way we
can also demonstrate the methods available to determine liquid-crystalline
structures, as well as the limitations of these methods.
Many studies in the area have involved the use of 1-monoolein. However, a
complication of phase studies of monoacylglycerols is isomerization during the
study. Even if the samples initially consist of pure 1-monoolein, acyl migration
during thermal equilibration will result in an equilibrium mixture of about 90%
of the 1-isomer and 10% of the 2-isomer. Even a few hours is sufficient to result
78 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
Figure 3.5 Proposed minimal surface structure of a lipid–water phase (Larsson et al., 1980). Two P-
surface structure units along the x-axis are shown. The dotted lines on the unit to the right illustrate a
square frame, which together with frames in the other directions will span a soap film with this P-surface
structure.
in some acyl migration. It might therefore be better to use the name glycerol
monooleate (GMO), without stating the isomer purity.
A detailed structure of the cubic phase of anhydrous sodium myristate was
published by Luzzati & Spegt in 1967. Many observed X-ray diffraction lines
were consistent with space group Ia3d, and a structure consisting of a rod
arrangement of polar groups in a continuum of hydrocarbon chains was
proposed. An X-ray study of a cubic GMO–water phase was first proposed to
indicate space-filling polyhedrons with lipid bilayer faces arranged according
to space group Im3m (Larsson, 1972).
Then an NMR-diffusion study by Lindblom et al. (1979) of a series of cubic
GMO–water phases indicated that the structure was both water and lipid
continuous. As relations in X-ray data with the adjacent Lα phase indicated a
lipid bilayer-based structure unit, opening of the square faces of the earlier
proposed polyhedrons (space group Im3m) was a structure model that seemed
to fulfil both NMR and X-ray data. It was also realized that if the planar faces
were allowed to be curved as minimal surfaces, the proposed structure was
identical to Schwarz’s P-surface (Larsson et al., 1980). The structure is shown
in Figure 3.5. This represented a new structure model of cubic lipid–water
phases different from the prevailing model at that time, with the cubic lipid–
water phases proposed to be arranged as rod systems (Luzzati et al., 1968). The
proposed space group Im3m was not correct, however, as the indexing
involved two phases with related structures, a feature discussed below which
provides strong evidence for the minimal surface structures.
Later, Longley & Mcintosh (1983) analysed a cubic GMO–water phase in
equilibrium with excess water, and from somewhat better diffraction data they
could convincingly conclude that its space group was Pn3m. They also
proposed that the probable structure followed the diamond minimal surface,
LIQUID-CRYSTALLINE LIPID–WATER PHASES 79
Figure 3.6 Illustration of the cubic GMO–water structure with the bilayer centred on the D-surface.
The surface was calculated by the nodal surface approximation; see text for details.
80 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
Figure 3.7 Schematic illustration of the G-surface structure of cubic lipid–water phases.
Figure 3.8 The structure unit of the P-type of nodal surface illustrated by a net with a proposed standing
wave breathing vibration mode of the bilayer, indicated by transparent layers (with maximal amplitude
at the flat point). (After Andersson et al., 1997).
LIQUID-CRYSTALLINE LIPID–WATER PHASES 81
⎛ 1 1 ⎞
γ ⎜ –– + –– ⎟
⎝ R1 R2 ⎠
where γ is the interfacial tension.
82 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
1. Introduction
The phase rule (also called Gibb’s phase rule) defines the relation at equilib-
rium between the number of phases, P, the number of components, C, and the
number of degrees of freedom, F (e.g. temperature, pressure, composition), by
the relation:
P+F=C+2
Applications of the phase rule and its use in determining phase diagrams of
aqueous systems of lipids will be discussed here. Different liquid-crystalline
phases provide different functions, and the significance of knowing the
composition and temperature range where a particular phase exists is obvious.
The self-assembly of lipid molecules in water due to their dual properties in
relation to water (their amphiphilicity) was mentioned in Chapter 1. There are
a few lipid types which are so polar that they are water-soluble in the form of
micellar aggregates. Above a critical micellar concentration (cmc), the
monomers solved in water associate into aggregates which are usually
84 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
Figure 3.9 A bilayer description of the HII phase is shown to the left, as an alternative to the usual
monolayer model shown to the right.
k1H2
––– + k2G
2
where H and G are the average curvature and Gaussian curvature, respectively
(as defined in Section A.3 above), and k1 and k2 are elastic moduli constants.
The HII phase is usually described as water cylinders covered by lipid
monolayers in a hexagonal packing. Alternatively the HII phase can be regarded
as a bilayer phase with a honeycomb shape, i.e. bilayer units shaped as infinite
ribbons intersecting one another into trigonal corners. These two views of the
HII structures are illustrated in Figure 3.9. Both models might be fruitful to
consider when molecular packing is taken into account. The circular cross-
section of water channels in a hexagonal phase formed by a single lipid means
that there is a regular variation in molecular length in different hexagonal
directions, whereas there can be constant length in the other alternative shown
in Figure 3.9 (except at the intersections of bilayer units). Aqueous systems of
alkanes/phosphatidylcholine have been studied by Sjölund et al. (1988). They
demonstrated how alkanes favour the HII phase and reduce structural frustra-
tions by filling the trigonal void space between phosphatidylcholine monolayers
covering the water cylinders.
The effect of temperature on phase transitions is discussed in Section C
below, and the different phase diagrams shown in Section C will illustrate the
effect of water content.
We will now consider the transitions between the cubic phases. It seems
likely that the cubic transitions discussed below are martensitic in their
character; see Hyde et al. (1997).
86 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
Figure 3.10 Phase transition behaviour when water is added to the GMO–water G-phase at its
maximum swelling. The shaded region illustrates the D-phase successively formed.
88 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
Figure 3.11 Proposed interface structure between a G-phase or a D-phase bilayer and a water phase.
The curved double line illustrates the bilayer, and closing of the bilayer structure outwards towards water
by an Lα type of bilayer is indicated by the thick lines at the top.
in this connection that the surface structure depicted in Figure 3.11, with rod-
shaped water channels perpendicularly oriented to the surface, has hexagonal
symmetry when viewed along these channels. Both the G-phase and the
D-phase have closely related geometry and symmetry along this plane, which
is a strong argument for proposing that this plane corresponds to the interface
between the D-phase and the G-phase.
Figure 3.12 The structure unit of the proposed tetragonal minimal surface organization of the lipid
bilayer lining the alveolar surface, reproduced by permission (M. Larsson, 2002).
1. Monoacylglycerols
We will start with GMO, which has been the subject of numerous recent studies
of the structure of the bicontinuous cubic phases. The first complete phase
diagram is shown in Figure 3.13 (Hyde et al., 1984). Qui & Caffrey (2000) later
also described this phase diagram at temperatures below 20ºC, where there is
complicated behaviour involving metastable phases. Their diagram also shows
some minor deviations at higher temperatures from that shown in Figure 3.13.
The phase transition sequence obtained by heating illustrates the effect of
increased thermal mobility of the hydrocarbon chain; increasing proportions of
gauche conformations correspond to increased chain divergence. The packing
parameter values at the Lα→cubic and cubic→HII phase transitions can be
calculated from available data on the composition and on the corresponding
X-ray data. The effect of water content on the transitions will be discussed in
the last section of this chapter.
A further illustration of the effect of molecular geometry on the phase
behaviour is illustrated in Figure 3.14, where saturated 1-monoacylglycerols
from C8 to C20 are compared. It is obvious that the effect of chain divergence
will increase with chain length; therefore, the hexagonal HII phase grows at the
expense of cubic phases, which in turn grow at the expense of the lamellar Lα
phase.
In many technical applications, industrially distilled monoacylglycerol mix-
tures are used. A common type is obtained from fully hydrogenated lard, which
90 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
Figure 3.13 Phase diagram of the glycerol monooleate (GMO)/water system (after Hyde et al., 1984).
Equilibration of the phases was started from the 1-isomer. The two cubic phases are termed the G-phase
and the D-phase, after the corresponding minimal surface structures. The liquid L2 phase is discussed in
the next chapter. Two-phase regions are shaded.
has a fatty acid composition dominated by stearic acid. The phase diagram of
such a mixture behaves like that of a pure component (Krog & Larsson, 1968).
The transition temperatures are very close to those of 1-monopalmitin (Larsson,
1967). The presence of impurities in the form of charged lipid species results in
increased swelling.
2. Phospholipids
Phospholipids, especially phosphatidylcholines, have been the subject of
extensive studies concerning their aqueous interaction and phase properties,
because of their significance in biomembrane physiology. This is illustrated by
several reviews in this field. A special issue of the journal Chemistry and
Physics of Lipids has for example been devoted to phospholipid phase transi-
tions (Volume 57, 1991).
Phosphatidylcholines (PCs)
There is probably no natural lipid that has been studied as thoroughly as egg
yolk PC. The phase diagram is illustrated in Figure 3.15. At a very low water
content, the phase behaviour is very complex, and it is omitted here as it is
hardly relevant in any application. The lamellar Lα phase dominates the phase
diagram.
At low water content, however, the transitions Lα→ cubic→HII are seen,
LIQUID-CRYSTALLINE LIPID–WATER PHASES 91
Figure 3.14 The main features of the aqueous phase diagrams of the saturated monoacylglycerols from
C8 to C20: (a) C8:0–C12:0, (b) C14:0–C18:0, (c) C20:0.
Figure 3.15 Phase diagram of egg yolk PC. C denotes a cubic phase region and S denotes the solid state
(after Small, 1986).
Figure 3.17 Main features of the phase diagram of water/C6:0-C16:0-PC. Dotted regions contain two
phases (after Svensson et al., 1993).
Phosphatidylethanolamines (PEs)
Seddon et al. (1984) have reported the phase behaviour of the aqueous systems
of the dialkyl-PE didodecyl-PE, and of the diacyl-PE diarachidonoyl-PE.
Didodecyl-PE in excess water transforms into the Lα phase at about 40ºC, and
at about 100ºC the Lα phase transforms into a cubic phase. At higher tempera-
tures, above 120ºC, the HII phase is formed. The diacyl-PE also forms an Lα
phase in a very narrow temperature interval, and it is transformed directly into
the HII phase on heating. The binary system dipalmitoyl-PE–water is shown in
Figure 3.18. This diagram is rather similar to that of dipalmitoyl-PC–water
(Figure 3.16) discussed above.
There are also reports on the aqueous phase properties of unsaturated PEs
(e.g. see Lindblom & Rilfors, 1989).
When PE–water Lα phases are cooled to temperatures below the chain
crystallization temperature, crystals separate directly in the water phase
94 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
Figure 3.18 Phase diagram of the binary system dipalmitoyl-PE/water, drawn by combining data from
Cevc (1991) and Seddon et al. (1984). The unknown region at low water content is shaded.
Phosphatidylinositols (PIs)
The phase diagram of soybean PI–water is shown in Figure 3.19. The acyl
chains are dominated by C 16:0 and C18:2. The Lα phase dominates the
investigated region (room temperature to 70ºC) with a swelling limit of about
77% (w/w) water.
Figure 3.19 Phase diagram of the binary system consisting of the sodium salt of soybean PI/water
(after Söderberg, 1990).
lipids become non-ionic, they tend to form HII phases (Hope & Cullis, 1980);
divalent counter-ions like calcium (which contract the polar sheets laterally)
also tend to give HII phases (Harlos & Eibl, 1980; Farren et al., 1983).
In mixed phospholipid systems it is fruitful to consider the average molecu-
lar shape, and whether or not there is an average net charge. An illustrative
example is the mixed system DOPE (dioleoyl-phosphatidylethanolamine) plus
DOPC (dioleoyl-phosphatidylcholine) and milk sphingomyelin, which was
reported recently (Waninge et al., 2003). The phases observed in this system
are the same as those exhibited by each of the components, and their regions in
the phase diagram can be explained by the average molecular shape.
Lysophospholipids
Phospholipases A1 and A2 split off the acyl chains in the 1- and 2-positions
respectively. Phospholipase A2 is particularly important from a physiological
point of view as phospholipids in biological tissues often have arachidonic acid
in the 2-position, and this fatty acid can initiate a cascade of inflammatory
reactions. Lysophospholipids are water soluble in micellar form, and these
micelles can solubilize and destroy cell membranes. Phospholipase degrada-
tion of membrane phospholipids into fatty acids and the lyso compounds is the
reason why snake venom can cause haemolysis.
Arvidsson et al. (1985) have studied lyso-PCs of different chain length
(saturated and unsaturated chains), and found cubic phases in addition to the
expected HI and Lα phases.
96 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
Figure 3.20 Molecular arrangement in the different phases of cerebroside from bovine brain and in the
corresponding psychosine formed when the acyl chain is removed (after Abrahamsson et al., 1972).
3. Sphingolipids
Sphingomyelin (SM) is rather similar to PC from a chemical point of view (see
Chapter 1). The SM–water system is also quite similar to the PC–water system
with a corresponding chain melting point (dipalmitoyl-PC, for example, is
similar to milk SM in this respect). The Lα phase dominates, and the zwitterionic
character results in a maximum water layer thickness of about 20 Å. The
saturated and unsaturated members examined behave in a similar way except
for the difference in chain melting temperature (Shipley et al., 1985; Sripada et
al., 1987).
Cerebrosides are sphingolipids with a sugar group attached to the polar head
group. They occur in the brain and their aqueous phases have been studied by
Abrahamsson et al. (1972); see Figure 3. 20. When the acyl group is removed,
the detergent-like molecule that is formed is termed psychosine, the aqueous
phase behaviour of which is also shown in Figure 3.20.
When a hydroxyl group of the sugar in cerebrosides is exchanged for a
sulfate group, sulfatides are obtained. These form micellar solutions in water
(Abrahamsson et al., 1972) and behave in a similar way to another type of lipid
containing a sugar group: the gangliosides, which were described in Chapter 1.
Gangliosides are also soluble in micellar form (Curatolo et al., 1977).
LIQUID-CRYSTALLINE LIPID–WATER PHASES 97
4. Glyceroglucolipids/galactolipids
This type of lipid is particularly abundant in plant cell membranes, and is also
often found in microorganisms. Monogalactosyl-diacylglycerols (MGDG)
form HII phases with excess water, whereas digalactosyl-diacylglycerols
(DGDG) form Lα phases. This general behaviour was observed in MGDG and
DGDG from plant leaves (Shipley et al., 1973), from wheat endosperm
(Larsson & Puang-Ngern, 1979) and from mycoplasma (Wieslander et al.,
1978).
The thylakoid membranes of chloroplasts (where photosynthesis takes place)
contain MGDG and DGDG as major components. It is therefore most interest-
ing to note that in the vegetative state of chloroplasts, the lamellar stacks of flat
membrane sacs have been transformed into a cubic phase (Hyde et al., 1997).
In the mycoplasma membrane, a transition between the Lα phase and the HII
phase appears to occur when approximately equal proportions of MGDG and
DGDG are present (Lindblom & Rilfors, 1989).
5. Lipid–glycerol systems
There are also water-free lipid systems that behave like aqueous systems; these
include polar liquids, which allow hydrogen bonding and dipole–dipole inter-
actions like water. Glycerol is an example considered here. The binary system
GMO–glycerol exhibits the same phases as GMO–water, with similar exist-
ence ranges in the phase diagram (K. Larsson, unpublished data). The transition
temperatures are slightly higher, and the kinetics of the phase changes much
slower. Proteins such as lysozyme can be incorporated and they show similar
denaturation behaviour as in the corresponding aqueous phase. Phases with Lα
structure and L2 phases were prepared from galactolipids and phospholipids
extracted from oats. It was found that oxidation appeared to be almost totally
inhibited in such phases, swollen in glycerol. The reason might be that lipases
in this environment become inactive and lipoxidases require free fatty acids as
substrate. Possibilities for replacing water by an alternative non-toxic sub-
stance, such as glycerol, in technical applications of liquid-crystalline phases
may have been neglected. From a fundamental point of view also it is
remarkable that many different types of lipids, alone or in combination with
other types of biomolecules, show similar behaviour in glycerol as in an
aqueous environment.
Figure 3.21 Simplified model diagram of one-phase regions of aqueous systems of (left) non-ionic
lipids and (right) ionic or zwitterionic lipids. The limit of swelling is shifted to higher water concentra-
tions in ionic compared to non-ionic lipids, but can be reduced by counter-ions. The hatched regions
represent two phases in equilibrium.
The other two types of phase diagrams are illustrated in Figure 3.21. A
situation was chosen where the three types of phases – cubic, hexagonal and
lamellar – exist in the same system. Only the structures that exist above the
chain melting temperature are considered.
It is interesting to note the opposite tilt direction of the phase boundaries in
the two types of system (the reason why we have classified them in this way).
In non-ionic lipids there is a tendency for the packing parameter to change with
both increased water content and increased temperature, which may induce
Lα→cubic bicontinuous→HII transitions. The increased disorder with increased
water content is considered to be due to increased motion of the polar head
group in and out in relation to the end group surface. Such disorder would also
increase the gauche to trans ratio along the chains. This is confirmed by
diffraction data; the bilayer thickness decreases with increasing water content.
When the polar head groups are held together by ionic forces, the effect of
increased water content in the phase will be quite different. Let us consider a
lipid with one charged group. The lateral electrostatic repulsion between the
polar head groups will be reduced if the distance between them within the polar
sheet is increased. Such a change could take place with an increase in water
content, when water penetrates the polar sheets as well as forming a gap
between the bilayers. This would be equivalent to a change with water
concentration in the sequence HII→cubic→Lα. The effect of increased tem-
perature, however, would be the same as in non-ionic lipids, as discussed
above. This explains the difference in tilt direction of the phase boundaries in
Figure 3.21.
The limit of swelling is similar in zwitterionic and non-ionic lipids, corre-
sponding to a water layer thickness of about 20 Å. In the case of ionic lipids,
where the limit of swelling is increased (related to the electric double-layer),
LIQUID-CRYSTALLINE LIPID–WATER PHASES 99
1. Introduction
There are now numerous phase diagrams involving two or more lipids reported
in the literature. Here we will only consider some diagrams of general rel-
evance which can be helpful in order to understand the main features of
multicomponent lipid–water systems.
The binary phase diagrams discussed earlier in this chapter show the effects
of water content as well as temperature. If the number of components is
increased to three, and we want to show the effects of composition, we have to
select a particular temperature and illustrate the composition by a triangle. As
mentioned above, there is a tendency for mixtures within a certain lipid class
(such as soybean triacylglycerols) to behave like one component in relation to
the phase rule. This simplifies the determination of phase diagrams of lipids,
and is most important when attempting to apply basic knowledge from phase
diagrams to discussions of lipid functionality. We will start with examples with
applications in biology (gastrointestinal fat digestion) and technology (peroral
drug delivery).
Figure 3.22 Ternary phase diagram at 40ºC for soybean oil triacylglycerols (TG)/water/sunflower oil
monoacylglycerols (MG) (after Lindström et al., 1981). The two-phase region L2+water is dotted, and
the liquid-crystalline (LC) region consists of a lamellar liquid-crystalline phase, a cubic phase, and a
hexagonal HII phase.
Figure 3.23 Phase diagram for saponin/water/monoacylglycerols from sunflower oil (Barla et al.,
1979). Regions with saponin crystals are indicated by black, I shows the cubic region, and D indicates
the lamellar Lα phase region. Regions containing two or three phases are shaded. There is a continuous
transformation of the L2 phase into an L1 phase, and it is therefore not possible to strictly show the
corresponding phase boundaries. Position A in the liquid region represents an L1 phase, which towards
position B becomes an L2 type. The line B–C exhibits a remarkable similarity in extension to that of the
region of the lamellar phase (D), both having stoichiometric monacylglycerol:saponin ratios.
This behaviour can be explained by the phase diagram. There is one single
liquid phase from the water corner to the butanol corner, which might be
regarded as an L1 phase transforming into an L2 phase. Another interesting
feature is the transition from an Lα phase to an HI phase with increasing
solubilization of n-butanol, an effect of changes in the average molecular
shape. The effects of butyric acid were found to be similar to those of butanol.
This might be of physiological interest as this acid is produced by fermentation
in the colon, and by penetration provides nutrition to the epithelial cells.
The relations between liquid-crystalline phases and liquid lipid–water phases,
as reflected in phase diagrams, should be briefly mentioned here. We have seen
above that L2 phases often occur near Lα phases, and that their structures are
also related (with partial ‘melting’ and loss of crystallographic order). In a
102 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
Figure 3.24 Phase diagram for soybean phosphatidylcholine (SPC)/ethanol/water at room temperature
(after Söderberg, 1990).
Figure 3.25 Phase diagram for egg yolk PC/n-butanol/water (I. Söderberg, unpublished results). The
large empty region consists of one liquid phase, the dotted region is an L2 phase, and the black region is
an HI phase. Two-phase regions are shaded.
Figure 3.26 The ternary phase diagram for propylene glycol (PG)/GMO/water at room temperature
(after Engström et al., 1998). The cubic phase is termed Q here.
E. Cholesterol esters
Fatty acid esters of cholesterol are important from a physiological point of
view; they are found in particles that carry lipids in the circulation associated
into lipoproteins. Phase properties of lipid systems involving cholesterol esters
have been studied by Small & co-workers (e.g. see Small, 1986). The liquid-
crystalline phases are unique in their structure compared to other lipids. Before
melting they can form smectic, cholesteric, and blue phases. The cholesterol
skeletons can pack in layers with a helical twist when going from one layer to
the next. The structural changes are very sensitive to temperature variations,
and this can be observed as colour variations and used as a temperature
indicator in certain esters.
The blue phases are cubic, and they obtain their blue colour due to the size
of the structure repetition unit, with a unit cell length that interferes with the
wavelength of blue light. Three blue phases (BPI, BPII and BPIII) can form
successively on heating. The blue phases BPI and BPII have been shown to
have structures corresponding to the gyroid and diamond minimal surface
104 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
structures, respectively, and the lattice parameters are also in agreement with
those required by a Bonnet transformation (Hyde & Fogden, 1998).
References
Abrahamsson, S., Pascher, I., Larsson, K. & Karlsson, K-A. (1972) Chem. Phys. Lipids 8,
152.
Andersson, S., Hyde, S.T., Larsson, K. & Lidin, S. (1988) Chem. Rev. 88, 221.
Andersson, S., Larsson, K. & Jacob, M.Z. (1997) Kristallogr. 212, 5.
Andersson, S., Larsson, K., Larsson M. & Jacob, M. (1999) BIOMATHEMATICS –
Mathematics of Biostructures and Biodynamics, Elsevier, Amsterdam, the Netherlands.
Arvidsson, G., Brentel, I., Kahn, A., Lindblom, G. & Fontell, K. (1985) Eur. J. Biochem. 152,
753.
Barla, P., Larsson, K., Ljusberg-Wahren, H., Norin, T. & Roberts, K. (1979) J. Sci. Food
Agric. 30, 864.
Baumgart, T., Hess, S.T. & Webb, W.W. (2003) Nature 425, 821.
Cevc, G. (1991) Chem. Phys. Lipids 57, 293.
Curatolo, W., Small, D.M. & Shipley, G.G. (1977) Biochim. Biophys. Acta 168, 11.
de Neveu, D.M., Rand, R.P., Ginger, D. & Parsegian, A. (1976) Science 191, 399.
Deng, Y., Marko, M., Buttle, K.F., Leith, A.D., Mieczkowski, M. & Manella, C.A. (1999)
J. Struct. Biol. 127, 231.
Engström, S., Alfons, K., Rasmusson, M. & Ljusberg-Wahren, H. (1998) Progr. Colloid
Polymer Sci. 108, 93.
Farren, S.B., Hope, M.J. & Cullis, P.R. (1983) Biochem. Biophys. Res. Commun. 111, 675.
Harlos, K. & Eibl, H. (1980) Biochemistry 19, 895.
Hauser, H., Pascher, I. Pearson, R.H. & Sundell, S. (1981) Biochim. Biophys. Acta 650, 21.
Helfrich, W. (1973) Z. Naturforsch. A28, 693.
Hope, M.J. & Cullis, P.R. (1980) Biochem. Biophys. Res. Commun. 21, 846.
Hyde, S.T., Ericsson, B., Andersson, S. & Larsson, K. (1984) Z. Kristallogr. 168, 213.
Hyde, S.T., Andersson, S., Larsson, K., Blum, Z., Landh, T., Lidin, S. & Ninham, B. (1997)
The Language of Shape. The Role of Curvature in Condensed Matter Physics, Chemistry
and Biology, Elsevier, Amsterdam, the Netherlands.
Hyde, S.T. & Fogden, A. (1998) Progr. Colloid Polymer Sci. 108, 139.
Imberg, A. (2003) On Phase Behaviours in Lipid/Polymer/Solvent/Water Systems and their
Application for Formation of Lipid/Polymer Composite Particles. Thesis, University of
Uppsala, Sweden.
Israelachvili, J.N., Mitchell, D.J. & Ninham, B.W. (1977) Biochim. Biophys. Acta 470, 341.
Israelachvili, J.N. & Wennerström, H. (1990) Langmuir 6, 873.
Janiak, M.J., Small, D.M. & Shipley, G.G. (1976) Biochemistry 15, 4575.
Johansson, A.K., Linse, P., Picurell, L. & Engström, S. (2001) J. Phys. Chem. B105, 12157.
Kekicheff, P. & Cabane, B. (1988) Acta Crystallogr. B44, 395.
Krog, N. & Larsson, K. (1968) Chem. Phys. Lipids 2, 129.
Landau, E.M. & Rosenbusch, J.P. (1996) Proc. Natl Acad. Sci. USA 93, 14532.
Landh, T. (1996) Cubic Cell Membrane Architecture. Thesis, Lund University, Sweden.
Landh, T. (1994) J. Phys. Chem. 98, 8453.
Larsson, K. (1967) Z. Phys. Chem. (Frankfurt am Main) 56, 173.
Larsson, K. (1972) Chem. Phys. Lipids 9, 181.
Larsson, K. & Puang-Ngern, S. (1979) In: Advances in the Biochemistry and Physiology of
Plant Lipids (Appelquist, L-Å. & Liljenberg, C., eds), Elsevier, Amsterdam, the Nether-
lands, p.27.
LIQUID-CRYSTALLINE LIPID–WATER PHASES 105
Larsson, K., Fontell, K. & Krog, N. (1980) Chem. Phys. Lipids 27, 321.
Larsson, K. (1983) Nature 304, 664.
Larsson, K. (1989) J. Phys. Chem. 93, 7304.
Larsson, M. (2002) A Surface Phase Model of The Alveolar Lining: Ultrastructural Analysis
and In Vivo Applications. Thesis, Lund University, Sweden.
Larsson, M., Larsson, K., Nylander, T. & Wollmer, P. (2003) Eur. J. Biophys. 31, 633.
Larsson, K. & Tiberg, F. (2005) Curr. Opin. Colloid Interface Sci. 9, 365.
Lindblom, G. & Rilfors, L. (1989) Biochim. Biophys. Acta 988, 221.
Lindblom, G., Larsson, K., Johansson, L., Fontell, K. & Forsen, S. (1979) J. Am. Chem. Soc.
101, 5465.
Lindström, M., Ljusberg-Wahren, H., Larsson, K. & Borgström, B. (1981) Lipids 16, 749.
Longley, W. & McIntosh, T.J. (1983) Nature 303, 612.
Luzzati, V., Mustacchi, H., Skoulios, A. & Husson, F. (1960) Acta Crystallogr. 13, 660.
Luzzati, V. & Spegt, P.A. (1967) Nature 215, 710.
Luzzati, V., Tardieu, A., Gulik-Krzywicki, T., Rivas, E. & Reiss-Husson, F. (1968) Nature
220, 485.
Luzzati, V. (1997) Curr. Opin. Struct. Biol. 7, 661.
Mortensen, K., Pfeiffer, W., Sackmann, E. & Knoll, W. (1988) Biochim. Biophys. Acta 945,
221.
Nollert, P., Qui, H., Caffrey, M., Rosenbusch, J.P. & Landau, M. (2001) FEBS Lett. 504, 179.
Orädd, G., Lindblom, G., Fontell, K. & Ljusberg-Wahren, H. (1995) Biophys. J. 68, 1865.
Qui, H. & Caffrey, M. (2000) Biomaterials 21, 223.
Rappolt, M., Pabst, G., Rapp, G., Kriechbaum, M., Amenitsch, H., Krenn, C., Bernstorff, S.
& Laggner, P. (2000) Eur. Biophys. J. 29, 125.
Schoen, A.H. (1970) NASA Technical Report No. 05541.
Scriven, L.E. (1976) Nature 263, 123.
Seddon, J.M. (1990) Biochim. Biophys. Acta 1031, 1.
Seddon, J.M., Cevc, G., Kayl, R.D. & March, D. (1984) Biochemistry 23, 2634.
Shipley, G.G., Green, J.P. & Nichols, B.W. (1973) Biochim. Biophys. Acta 311, 531.
Shipley, G.G., Avecilla, L.S. & Small, D.M. (1985) Biochemistry 24, 2902.
Simons, K. & Ikonen, E. (2000) Science 290, 1721.
Sjölund M., Rilfors, L. & Lindblom, G. (1988) Biochemistry 28, 1323.
Small, D.M. (1986) Handbook of Lipid Research. 4. The Physical Chemistry of Lipids,
Plenum Press, New York, USA.
Snapp, E.L., Hegde, R.S., Francoline, M., Lombardo, F., Colombo, S., Pedrazzini, E.,
Bergese, N. & Lippincott-Schwartz, J. (2003) J. Cell. Biol. 163, 257.
Söderberg, I. (1990) Structural Properties of Monoglycerides, Phospholipids and Fats in
Aqueous Systems. Thesis, University of Lund, Sweden.
Sripada, P.K., Malik, P.R., Hamilton, J.A. & Shipley, G.G. (1987) J. Lipid Res. 28, 710.
Svensson, I., Adlercreutz, P., Mattiasson, B., Miezis, Y. & Larsson, K. (1993) Chem. Phys.
Lipids 66, 195.
Vargas, M., Mariani, P., Gulik, A. & Luzzati, V. (1992) J. Mol. Biol. 225, 137.
von Schnering, H.G. & Nesper, R. (1991) Z. Phys. B Condensed Matter 83, 407.
Waninge, R., Nylander, T., Paulsson, M. & Bergenståhl, B. (2003) Chem. Phys. Lipids 125,
59.
Wieslander, Å., Ulmius, J., Lindblom, G. & Fontell, K. (1978) Biochim. Biophys. Acta 512,
241.
CHAPTER 4
The liquid state
A. Liquid triacylglycerols
It is known that the melting of a fat into an oil may keep a ‘memory’ of the
crystal structure before melting, and in that way influence the polymorphic
behaviour. On the basis of X-ray scattering data from liquid triacylglycerols,
the persistence of bilayer regions after melting, as shown in Figure 4.1, has
been proposed. Thus a band in the X-ray scattering curve with an integrated
intensity value almost as high as that of the first-order diffraction line has been
observed. At nucleation, the organization within these associated regions
should therefore be expected to influence the structure of the crystallization
nuclei.
The freeze-fracture freeze-etching electron micrograph of a disordered lipid
melt is shown in Figure 4.2, and the corresponding texture of a triacylglycerol
melt is shown in Figure 4.3. Freeze-fracturing is an informative electron
microscopy method for structure studies of lipids. A sample fractured in frozen
condition will open up surfaces along planes where the molecular interaction
forces are weakest, which are the methyl end group planes/regions.
107
108 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
Figure 4.1 Proposed bilayer arrangement in the liquid state of triacylglycerols after melting (after
Larsson, 1972).
Figure 4.2 Electron micrograph of freeze-fractured freeze-etched liquid n-dodecanol (after Gulik-
Krzywicki & Larsson, 1984).
Figure 4.3 Electron micrograph of a freeze-fractured sample of milk fat before crystallization, kindly
provided by Wolfgang Buchheim, Kiel. The dimension of the terraces corresponds to a liquid
triacylglycerol bilayer according to X-ray scattering. The terrace height is about 50 Å and the shadow on
the micrograph is of the same order of magnitude.
0.3 μm
Figure 4.5 Freeze-fracture electron micrograph of a monoacylglycerol melt (after Gulik-Krzywicki &
Larsson, 1984).
This indicates that the L2 phase has a structure similar to that of the Lα phase,
with reduction of the infinite periodicity of the liquid-crystalline phase at the
transition into bilayer units, with a size of a few hundred Ångströms along the
bilayer and perpendicular to the bilayer plane. The bilayer thicknesses in both
THE LIQUID STATE 111
structures are about the same. The heating increases the disorder and gives a
tendency to divergence along the hydrocarbon chains. Therefore, after transi-
tion the polar regions with finite water layers are embedded into a continuous
hydrocarbon chain matrix.
The L2 phase is therefore considered to have a structure closely related to
that of the Lα phase: a melted Lα phase with change of long-range order to
‘medium’-range order. Studies by freeze-fracture freeze-etching electron
microscopy of the sunflower oil monoacylglycerols/water system have con-
firmed this structure (Gulik-Krzywicki & Larsson, 1984); see Figure 4.4.
Differently oriented stacks of lamellae can be seen, and their size is also in
agreement with the X-ray line broadening estimation.
When the water composition is reduced to zero, there is no discontinuity in
the scattering curves. Galactolipids in the liquid state show similar X-ray
scattering characteristics. An electron micrograph of a monoacylglycerol melt
is shown in Figure 4.5. This indicates that the structure in the melt of polar
lipids is the same as discussed here for L2 phases: a lamellar bilayer structure
with short-range disorder and medium-range periodicity.
D. Lipid microemulsions
The term microemulsion is used for thermodynamically stable liquid phases
formed by oil, water, a surfactant, and usually also a co-surfactant. The term
was introduced in the description of the transparent liquid phase in systems
based on mineral oils and water (Danielsson & Lindman, 1981). Lipid systems
are different, however, as the oil itself is amphiphilic. Still it seems adequate to
use this term for a liquid transparent phase consisting of water aggregates in a
triacylglycerol oil with a polar lipid acting as surfactant. An example showing
the existence range of such a phase is shown in Figure 3.22, which here is
termed L2. An electron micrograph of a corresponding freeze-fractured phase
is shown in Figure 4.6.
These inverse aqueous phases in lipid systems involving triacylglycerols are
liquid, thermodynamically stable and transparent. It seems logical from a
functional point of view to call them microemulsions, as an alternative to L2
phases. It also seems likely that by increasing the proportion of triacylglycerol
with respect to the polar lipid, the shape of the water aggregates will succes-
sively approach spheres.
E. The L3 phase
A third liquid phase beside the L1 and the L2 phases has been observed in non-
ionic surfactant–water systems (and in some ionic surfactant–water systems
containing salts), and it has been termed L3 (Anderson et al., 1989). It has since
also been observed in lipid systems, where it appears as an extension of a
112 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
References
Anderson, D., Wennerstöm, H. & Olsson, U.J. (1989) J. Phys. Chem. 93, 4243.
Danielsson, I. & Lindman, B. (1981) Colloids and Surfaces 3, 39.
Engström, S., Alfons, K., Rasmusson, M. & Ljusberg-Wahren, H. (1998) Progr. Colloid
Polym. Sci. 108, 93.
Gulik-Krzywicki, T. & Larsson, K. (1984) Chem. Phys. Lipids 35, 127.
Imberg, A. (2003) On Phase Behaviours in Lipid/Polymer/Solvent/Water Systems and their
Application for Formation of Lipid/Polymer Composite Particles. Thesis, University of
Uppsala, Sweden. [Comprehensive Summaries of Uppsala Dissertations from the
Faculty of Pharmacy, No. 304].
Larsson, K. (1972) Fette-Seifen-Anstrichm. 74, 136.
CHAPTER 5
Lipids at the air–water interface – monolayers
and multilayers in surface films, bubbles and
foams
Figure 5.1 A pressure–area (Π–A) isotherm of 1-monomyristin monolayer phases at 25ºC (after Krog
et al., 1985). The proposed structures of the two monolayer phases (form I at low pressure and form II
at high pressure) are also indicated.
Figure 5.2 Phase diagram (phase existence region in relation to temperature and monolayer pressure)
of 1-monomyristin monolayers (after Krog et al., 1985). The shaded area corresponds to collapse of the
monolayer.
116 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
relations with the aqueous phases in bulk are important, as so much more
information can be obtained from studies of three-dimensional phases. The
same two phases can also coexist in bilayers, such as cell membranes. If we
consider lipid rafts and caveolae (see Chapter 3), they correspond to two
coexisting monolayer phases, both with the liquid type of molecular conforma-
tion.
Figure 5.3 Tripalmitin pressure–area (Π–A) isotherm at 10ºC (after Burch et al., 1968). The transition
between the two solid phases, C1 and C2, is indicated by the arrow.
Figure 5.4 Illustration of competition between a monolayer spreading from a fat crystal and a
monolayer spreading from an oil droplet (where the monolayer conformation is first gaseous).
in the liquid state is dropped on the water surface of a surface balance, the
equilibrium pressure is reached within fractions of a second. If, on the other
hand a triacylglycerol crystal is deposited on the surface, it takes many minutes,
up to an hour, until a constant pressure (the ESP) is obtained.
Proteins are often used in food processing in order to stabilize emulsions and
foams. The ESP value of most proteins is about 20 mN/m, which is higher than
the ESP obtained from common fats and oils. This means that proteins will
squeeze out triacylglycerol molecules from air–water interfaces. The presence
of free fatty acids, however, may change this situation (see also next para-
graph).
Figure 5.5 Pressure–area (Π–A) isotherms of (a) palmitic acid and (b) stearic acid on a 0.001 M solution of hydrochloric acid (after Lundquist, 1978).
119
120 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
6. Mixed monolayers
Lipid monolayer phases show similar phase properties to lipid mixtures in
three-dimensional states. The phase rule can be applied (modified, as the
molecules are only free to move in two dimensions). If a binary lipid system is
studied by recording pressure–area isotherms at different proportions of the
two components, it is possible to directly determine if there is full mutual
molecular solubility as, if there is, the isotherms will have an intermediate
appearance to the isotherms of each of the two components studied alone. If the
mixed monolayer is expanded or compressed, an indication of interaction is
revealed by a change in the average molecular area in relation to the molecular
cross-sectional areas of the components at a particular pressure. If there is no
molecular solubility, on the other hand, the isotherm of the mixture will exhibit
the phase properties of each component (taking into consideration the fact that
a phase existing up to a higher pressure will squeeze out a phase with a lower
maximum pressure).
As mentioned above, the pioneering work on phase segregation in fluid
monolayer phases was reported by McConnel and co-workers (Subramaniam
& McConnel, 1987). By using a fluorescent probe they visualized the fluid–
fluid domains in mixtures of cholesterol and dipalmitoyl-phosphatidylcholine.
At a cholesterol concentration of 30mol%, they observed domain shape
fluctuations indicating the existence of a critical point of immiscibility. A
decade later the biological relevance of these observations was realized in
connection with studies of phase segregation in cell membranes (see Chap-
ter 6).
LIPIDS AT THE AIR–WATER INTERFACE 121
Figure 5.6 Formation of a triple chain-length structure of tripalmitin on a water surface (Larsson,
1973b).
Figure 5.7 The dynamic swelling behaviour of a lung surfactant extract, as seen under a microscope
(after Larsson et al., 2002). A network of branches at the air–water interface is formed in a time-scale of
minutes. The surface zone of the branches exhibits a uniform birefringence in polarized light.
1. Foam structure
The size of the gas cells and their shape can vary a lot in a given type of foam.
From a statistical study of foams, it has been reported that the average
polyhedral cell has 14 faces (William, 1968). This polyhedron with 6 square
and 8 hexagonal faces had already been described by Lord Kelvin as the
polyhedron with the smallest surface-to-volume ratio.
2. Foam stability
The following factors are the most important ones in stabilizing foams:
• the mechanical properties of the interfacial film;
• the viscosity of the water layer, which determines the drainage of foam
lamellae;
• the repulsive forces between the monolayers on each side of the
lamellae.
LIPIDS AT THE AIR–WATER INTERFACE 125
micellar lifetime had its maximum. This concentration agrees with the concen-
tration at which the spherical micelles start to become cylindrical.
If a foam is produced by gas injection from a capillary, the bubble will enter
the solution when the buoyancy force just exceeds the surface tension of the
contact circle between the bubble and the capillary. This demonstrates how the
bubble size decreases with surface tension. Above the critical micellar concen-
tration the monomer concentration is constant, but the bubble size has still been
found to vary (Oh et al., 1992). This could again be related to micellar stability.
The effects of salts on bubbles were discussed above. When a monolayer is
present on the surface, there will be electrostatic effects related to the ionic
character of the surfactant. Furthermore the critical micellar concentration is
reduced.
5. Destabilization of foams
There are many applications of foams, and there are also situations where
foaming must be avoided. Destabilizing mechanisms are therefore of interest.
The addition of salts will reduce the stability of foams based on ionic surfactants,
as explained above. Reduction of the viscosity of the surfactant monolayer is a
mechanism provided by a common foam destabilizer, 2-ethylhexanol. A
protein foam can be destabilized by a liquid fatty acid, such as oleic acid, and
this can be applied in order to reduce overly stable beer foams. The fatty acid
molecules will form separate domains along the surfaces of the foam lamellae,
where rupture will take place.
Another destabilizing mechanism has been described by Blute et al. (1994).
A series of organic electrolytes were studied, which supply counter-ions to a
foam stabilized by an ionic surfactant. Tetramethylammonium bromide and a
number of its homologues drastically reduced the stability compared to the
effect of sodium bromide. It was shown that the effect was related to the
increased area per surfactant molecule caused by the bulky shape of the
counter-ion. It was even seen that non-ionic surfactants were influenced,
indicating that these ions are amphiphilic and adsorb at the interface.
References
Ben-Jacob, E. & Levine, H. (2001) Nature 409, 985.
Blodgett, K.B. (1935) J. Am. Chem. Soc. 57, 1007.
Blute, I., Holmberg, K., Jansson, M., Oh, S.G. & Shah, D.O. (1994) J. Am. Oil Chem. Soc.
71, 41.
128 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
Burch, T., Larsson, K. & Lundquist, M. (1968) Chem. Phys. Lipids 2, 102.
Craig, V.S.J., Ninham, B.W. & Pashley, R.M. (1993) J. Phys. Chem. 97, 10192.
Dervichian, D.G. (1939) J. Chem. Soc. 7, 931.
Fahey, D.A. & Small, D.M. (1986) Biochemistry 25, 4468.
Fahey, D.A. & Small, D.M. (1988) Langmuir 4, 589.
Israelachvili, J.N. & Pashley, R.M. (1982) Nature 300, 341.
Krog, N., Riisom, T. & Larsson, K. (1985) Encyclopedia of Emulsion Technology, Marcel
Dekker, New York, USA, p.321.
Kuhn, H. & Möbius, D. (1971) Angew. Chem. 83, 672.
Langmuir, I. (1917) J. Am. Chem. Soc. 39, 1848.
Larsson, K. (1973a) Biochim. Biophys. Acta 318, 1.
Larsson, K. (1973b) In: Surface and Colloid Science 6 (E. Matijevic, ed.), J. Wiley & Sons,
New York, USA, p.261.
Larsson, K. & Krog, N. (1973) Chem. Phys. Lipids 10, 177.
Larsson, M., Haitsma, J., Lachmann, B., Larsson, K., Nylander, T. & Wollmer, P. (2002)
Clin. Physiol. Func. Imag. 22, 39.
Lösche, M., Duwwe, H. & Möwald, H. (1988) Colloid Interface Sci. 126, 432.
Lundquist, M. (1978) Progr. Chem. Fats Other Lipids 16, 101.
Moore, B.G., Knobler, C.M., Broseta, D. & Rondelez, J. (1987) Chem. Soc. Faraday Trans.
II 82, 1753.
Oh, S.G. & Shah, D.O. (1991) Langmuir 7, 1316.
Oh, S.G., Klein S.P. & Shah, D.O. (1992) AICAE J. 38, 149.
Okonogi, T.M. & McConnel, H.M. (2004) Biophys. J. 86, 880.
Richardson, G. (2004) Foam Formation and Starch Gelatinization with Alpha-Crystalline
Emulsifiers. Thesis, University of Lund, Sweden.
Stavans, J. & Glazier, J.A. (1989) Phys. Rev. Lett. A 37, 1318.
Subramaniam, S. & McConnel, H.M. (1987) J. Phys. Chem. 91, 1715.
von Tscharner, V. & McConnel, H. (1981) Biophys. J. 36, 409.
Weis, R.M. (1991) Chem. Phys. Lipids 57, 227.
William, R.E. (1968) Science 161, 276.
CHAPTER 6
Dispersions of lipid–water phases
What happens when two or more phases coexist in aqueous lipid systems? Can
the phases be dispersed in one another, and what are the flow, transport, and
stability characteristics of such dispersions? What are the responses to ambient
changes? These are questions of importance in many applications involving
lipid-based systems. Because the continuous phase typically is water we will
here concentrate on the properties of aqueous lipid dispersions of the lyotropic
lipid phases introduced in Chapter 3. Some related structures observed in
living systems are discussed towards the end of the chapter.
Temperature
Dilution +
shear force
Lα
Concentration
Lα
Liposomes
Figure 6.1 Schematic illustration of the Lα phase, phase diagram, and formation of liposomes in the
diluted two-phase region. The cryo-TEM image shows liposomes of dioleoylphosphatidylethanolamine/
Triton X-100. The small arrow indicates an ice crystal, the large arrow an inter-lamellar attachment. Scale
bar: 100 μm (Johnsson & Bergstrand, 2004).
Figure 6.2 Schematic illustration of the architecture of different types of liposomes: small and large
unilamellar vesicles (SUV and LUV), multivesicular liposomes (MVL), and multilamellar vesicles
(MLV).
20
10
0
0.01 0.1 1 10
Particle size (µm)
Figure 6.3 Particle size distributions of a homogenized and subsequently heat treated (125ºC) GMO–
Pluronic® F127 dispersion at a lipid:polymer ratio of 9:1. The size distribution increases with the total
amphiphile concentration (1%, 2%, 3%, 4% and 5%, w/w) used in the heat treatment step, and remains
unaltered by subsequent changes in concentration caused by, for instance, dilution (Barauskas et al.,
2005a).
micellar cubic (I2) particles. Cubosome® and Hexosome® are frequently used
trade names of cubic and hexagonal phase dispersions.
The size of non-lamellar particles can be controlled in the size range from
about 100 nm to several μm without changes in the inner morphology. One
example of size distributions obtainable for the GMO–Pluronic® F127 system
is shown in Figure 6.3.
By changing the phase structure the connectivity of water and lipid domains
can also be varied. This makes non-lamellar particles attractive for loading,
encapsulation and controlled release purposes in pharmaceutical and diagnos-
tic applications, as the phase structure can easily be adjusted to address a
specific technological need and fit a certain compound. By using two structur-
ing lipids with different spontaneous curvature, it is possible to change between
different particle phase structures simply by changing the ratios of the two
components. An example of such a system is GDO–DGMO, where GDO forms
a L2 phase and DGMO a lamellar phase. Combined in different ratios they form
a range of nanostructured phases including two cubic phases and the reversed
HII phase (Johnsson et al., 2005b). Some key properties of the different
reversed-phase dispersions are given below.
138 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
Figure 6.4 Cryo-TEM images of bicontinuous cubic phase (Q2) dispersions viewed in the [0,0,1]
direction.
Figure 6.5 Cryo-TEM images of reverse hexagonal phase (HII) dispersions showing a ‘top view’ of
well-ordered nanoparticles with reverse hexagonal ‘honeycomb’ structure.
DISPERSIONS OF LIPID–WATER PHASES 139
Figure 6.6 Cryo-TEM images at different magnifications of particles with sponge phase characteris-
tics.
1. Lamellar bodies
The special organelles for lipid storage and secretion have been called lamellar
bodies, and they can vary in size from 100 nm to 2400 nm (Schmitz & Muller,
1991). They occur for example in the gastrointestinal tract, in the peritoneum,
and in the mesodermal cell layer of joints. The lamellar bodies of two other
tissues will be considered here.
Assemblies of concentric lipid bilayers – like multilamellar liposomes – are
found in the lamellar bodies secreted from type II epithelial cells in the lung.
There are indications from cryo-TEM studies that the unit layer consists of two
lipid bilayers (M. Larsson, 2002), but still the main arrangement is similar to
that of liposomes. Epidermal lamellar bodies are involved in regeneration of
the stratum corneum layer of the skin. These organelles are different, however,
142 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
as the equidistant lipid bilayers are not concentrically arranged, but approxi-
mately planar. The stacks of membrane discs which form rods and cones in the
retina of the eye are another example of a parallel bilayer organization. These
two types of organization represent extremes of the Lα phase conformations in
closed particles/organelles.
Figure 6.7 Amoeba mitochondria showing transition into a cubic membrane conformation on starva-
tion (from Deng et al., 1999, with permission).
References
Albrecht, O., Gruler, H. & Sackmann, E. (1978) J. Phys. 39, 301–303.
Albrecht, O., Gruler, H. & Sackmann, E. (1981) J. Colloid Interface Sci. 79, 319–338.
Bangham, A.D. & Horne, R.W. (1964) J. Mol. Biol. 8, 660–668.
Barauskas, J., Johnsson, M., Joabsson, F. & Tiberg, F. (2005a) Langmuir 21, 2569–2577.
Barauskas, J., Johnsson, M. & Tiberg, F. (2005b) Nano Lett. 5, 1615–1619.
144 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
A. Introduction
Proteins are linear polymers of amino acids linked by peptide bonds between
the α-carboxyl group of one amino acid and the α-amino group of the next.
There are 20 protein amino acids and each protein is distinguished by the
primary sequence of these amino acids in the polymer. The factor that charac-
terizes the different amino acids is the side chain attached to the α-carbon atom.
Amino acids are categorized into several groups with common chemical
features. Some amino acids have charged functional groups such as carboxyl or
amino groups, while others are hydrophobic and contain branched aliphatic
groups or aromatic residues. The physical properties of the protein depend on
the proportion of each type of amino acid in the protein and the positions of the
amino acids in the polypeptide chain. These, in turn, dictate the higher-order
folding arrangements referred to as the three-dimensional secondary and
tertiary structure of the native protein.
The manner in which the protein folds depends largely on the local forces
experienced by the different side chains and the need for the polymer chain to
adopt a conformation with relatively high entropy. The forces at play are those
due to the solvent environment as well as the proximity of other amino acid
residues of the protein or of proteins with which it interacts. In the case of
certain membrane proteins the ‘solvent’ may include the lipid matrix of the
membrane, and interactions of this type may be required to fold the protein into
its native configuration. On the other hand there is evidence that interaction of
the proteins with membrane lipids is required to impose a bilayer conformation
on the surrounding membrane lipids, and is therefore an essential factor in
preserving the structure and properties of the membrane itself.
Non-ionic surfactants interact only weakly if at all with water-soluble
proteins but can solubilize intrinsic membrane proteins because of their
hydrophobic domains, which otherwise render them insoluble in aqueous
media. There is, however, strong interaction between ionic surfactants and all
proteins. One of the most studied of this type of interaction is between sodium
dodecyl sulphate and proteins. The non-specific cooperative binding of sodium
dodecyl sulphate to soluble proteins results in unfolding of the polypeptide
chain. After reduction of any disulphide bridges in proteins, sodium dodecyl
sulphate, above the critical micellar concentration, interacts with the polypep-
145
146 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
tide with a stoichiometry of 1.4 g detergent per gram of protein. This inter-
action imposes a regular helical structure on the polypeptide chain, which
becomes bent in the shape of a hairpin. The length of the resulting complex is
a function of the length of the polypeptide chain and, because of the predictable
conformation combined with a constant charge to mass ratio, differences in
hydrodynamic and electrophoresis properties can be exploited in separation
strategies for complex mixtures. The most notable system is the separation of
proteins by polyacrylamide gel electrophoresis in the presence of sodium
dodecyl sulphate (SDS-PAGE), which separates polypeptides on the basis of
size.
Where proteins interact with biological membranes at the aqueous–lipid
interface the charges of the acidic membrane lipids provide a particular
environment capable of interacting with basic amino acids of the protein. In the
case of proteins that are interpolated into the hydrophobic interior of the
membrane the environment is more conducive to the location of amino acids
with non-polar side chains. Clearly, both proteins and membrane lipids have
hydrophilic and hydrophobic groups, which interact to determine the structure
and conformation of the complex. This chapter will explore the nature of these
interactions and identify examples that illustrate the principles of how proteins
interact with lipids to modulate lipid structure and protein function.
Figure 7.1 Surface pressure–area isotherms of serum albumin monolayers at an air–water interface.
Monolayers were spread initially at high molecular density () and were subsequently expanded and
equilibrated at low surface density before recompression (). The protein molecules undergo a
conformational change to occupy a greater molecular area when unconstrained by pressure from
surrounding molecules in dense monolayers.
film is allowed to equilibrate at low surface density the isotherm changes so that
the molecules occupy almost twice the surface area per molecule at equivalent
surface pressures (open symbols). The measurement of surface radioactivity
confirms that there is the same number of protein molecules in the monolayer
so that the observed effect must be due to a change in conformation of the
protein molecules in the film. One explanation is that when the protein is spread
at high density the compact structure assumed by the protein in solution is
preserved by the tight packing of the molecules at the air–water interface.
Expansion of the surface area reduces these constraints, allowing the protein to
adopt a more expanded configuration consistent with the prevailing interfacial
forces associated with the surface tension of water. This illustrates the point
that protein folding (or unfolding) is dictated by the forces exerted by the
surrounding solvent molecules.
The structure of monomolecular films of rhodopsin purified from bovine
retinal rod outer segment disk membranes as a complex with octyl glucoside at
the air–water interface has been examined by infrared absorption and X-ray
reflectivity measurements (Lavoie et al., 2002). It was found that when the
148 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
proved to be invaluable for studies of the interaction between lipids and soluble
proteins; the lipid molecules at the interface are oriented and arranged in a well
understood way and their concentration can be readily altered by careful
expansion and compression of the film. A number of surface parameters can be
used to ascertain the extent and nature of the interaction with a soluble protein
in the bulk phase, including the consequences of interaction on the structure
and biological activity of the protein. Thus an increase in the surface pressure
of a film on adding protein is generally assumed to represent the penetration of
part at least of the space-occupying ‘volume’ of the protein into the film, albeit
this may not reach to the hydrophobic fatty acyl chains. Because of the various
factors that contribute to the surface pressure exerted by a monolayer (the
equation of state has kinetic, cohesive and electrostatic terms) it is clear that the
‘volume’ of protein penetrating into a particular film may not always be
proportional to the surface pressure increment. The ‘volume’ of protein pen-
etrating into the monolayer is related to the surface pressure increment by a
factor that depends on the force–area curve of the lipid film and any specific
interactions between the components. Measurement of the interfacial potential
of a monolayer also provides evidence of protein–lipid interactions, although
the interpretation of any changes is often problematic. This is because surface
potential is the sum of the vertical components of the intrinsic dipoles of the
lipid molecules, including the ionic dipoles of the head groups as well as the
oriented dipoles of water molecules hydrating the film. Any perturbation of the
aligned lipid molecules or ionic interactions or displacement of lipid molecules
by the adsorbing protein can affect these dipoles. Finally, with the use of
radioactively labelled proteins, direct measurement of the surface radioactivity
of the adsorbed protein is possible. Thus the total amount of protein adsorbed
per unit area of the surface phase can be derived but not the precise location or
orientation.
The monolayer method does have one possible disadvantage: unless the
penetrating parts of the proteins are exactly the same size and shape as the film
molecules, holes will tend to form in the unimolecular layer. To maintain the
planar structure the hole will fill with air unless some compensating collapse or
realignment of the fatty acyl chains can take place. Unless this realignment
occurs the adsorption energy may be changed somewhat by the energy required
to create the hole.
Penetration of protein into lipid monolayers can be observed when the initial
surface pressure of the monolayer is low, and increasing amounts of protein are
added to the aqueous subphase. This can be seen in Figure 7.2. [14C]-labelled
serum albumin was injected under a monolayer of egg phosphatidylethanolamine
spread at an initial surface pressure of 2 mN•m–1. The change in interfacial
junction potential shows a linear relationship but this parameter is the sum of
many vertically oriented dipoles and is not subject to a straightforward inter-
pretation. With increasing amounts of serum albumin the surface pressure
150 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
Figure 7.2 Changes in surface pressure (), radioactivity () and interfacial junction
potential () on injecting [14C]-labelled serum albumin underneath monolayers of
phosphatidylethanolamine initially spread at 2 mN•m–1.
Table 7.1 Percentage of protein added under lipid monolayers of 2 mN•m–1 adsorbed to
the film
Albumin Phosphatidylcholine 20 12
Phosphatidylethanolamine 14 7
Stearic acid 21 15
Cytochrome c Phosphatidylcholine 10 2
Phosphatidylethanolamine 26 19
Stearic acid 74 26
Table 7.2 Relationship between the space created in lipid monolayers and the area
occupied by protein at the interface at the first inflection point
surface pressure on the lipid molecules can be compared with the area occupied
by the penetrating protein assuming this is fully unfolded in the film. The
calculated area occupied by the protein at the first inflection point in the film
pressure showed reasonably close agreement with the area made ‘available’ by
compression of the lipid molecules (Table 7.2). This is consistent with the
penetration of folded protein molecules in their entirety into gaps in the
expanded films at low pressure and occupying spaces equivalent to that of the
unfolded protein at the air–water interface. At surface pressures greater than
the first inflection point the relationship completely breaks down and clearly
more protein is interacting with the film than could be accounted for by the
unfolded protein molecules entering the available space in the monolayer. This
means that only part of the protein enters the film, possibly the more hydropho-
bic domains of the polypeptide chain. It is likely that at low film pressures the
energy required for the protein molecules to enter the film and unfold at the
interface is comparatively low because of the greatly expanded nature of the
152 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
monolayer, but that at higher pressures the adsorption energy required for this
type of penetration is not available in the system. It is noteworthy that the
pressure of the first inflection point is less than the collapse pressure of the
protein monolayer at the air–water interface, whereas the second inflection
point is usually above this pressure.
The penetration of proteins containing the so-called START domain, a
conserved protein fold of a group of lipid-transfer and membrane-targeting
proteins, into lipid monolayers has been investigated (Angeletti et al., 2004).
The proteins themselves, such as StarD7, form stable monolayers by simple
adsorption from solution at the air–water interface. The maximum surface
pressure was found to be 18 mN•m–1 due to adsorption as a Gibbs monolayer,
from which a value of 28 kJ•mol–1 for the free energy of protein adsorption
could be derived. This value is similar to those of a variety of other amphipathic
proteins, consistent with the high tendency of this group of proteins to adsorb
at interfaces. Interestingly, no desorption of protein was observed after succes-
sive expansion and recompression cycles of the film up to the collapse pressure
of about 36 mN•m–1. However, a reversible rearrangement was observed in the
interfacial organization of the protein, which was evidenced by an abrupt
change in molecular packing area. The protein was found to penetrate into a
variety of phospholipid monolayers, with a preference for phosphatidylserine
over phosphatidylglycerol.
Figure 7.3 Changes in surface parameters on adding [14C]-labelled cytochrome c to the subphase
beneath phosphatidylcholine monolayers at various initial surface pressures.
Figure 7.4 Changes in surface parameters on adding [14C]-labelled cytochrome c to the subphase
beneath phosphatidic acid monolayers spread at various initial surface pressures on 10 mM NaCl (closed
symbols) and 1 M NaCl (open symbols).
subphase at intervals after injection of the protein into the subphase. The
adsorption of protein to the film was almost completely prevented when the
initial subphase was 1 M NaCl, consistent with the case presented in Fig-
ure 7.4. However, if adsorption was allowed to take place in 10 mM salt
solution and then solid NaCl was added to increase the concentration in the
subphase to 1 M, then only a proportion of the adsorbed protein was displaced.
The amount removed depended on the time of the interaction with the film,
reducing to a constant proportion of about 20% after equilibration had been
achieved. This result suggests that once electrostatic interaction has taken
place the adsorption is subsequently stabilized by non-ionic bonding. This
could be hydrogen bonding between the peptide chains and the polar head
groups of the phospholipid or, less likely, van der Waals interactions between
hydrophobic side chains of amino acids and the methylene groups of the
ethanolamine oriented in the interfacial region.
Figure 7.5 The effect of electrostatic charge of monolayers of phosphatidylethanolamine and [14C]-
labelled cytochrome c on (a) adsorption to films at surface pressures of 26 mN•m–1 and (b) penetration
into films spread on subphases of 1 M NaCl at a surface pressure of 10 mN•m–1.
156 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
statically facilitated penetration shows some specificity for the sign of the net
charges on the protein and lipid interfaces. The explanation for this differential
effect of the sign of the charge on the interface on penetration is obscure. It may
be expected that not all proteins would necessarily show the same charge-
assisted penetration characteristics, although the trend observed in the serum
albumin system has been observed in a wide variety of proteins. Furthermore,
it has been reported that a number of proteins bind only weakly to oil emulsions
if the stabilizer is a positively charged detergent, but relatively strongly if
negatively charged detergents are used.
and not dependent on the presence of unsaturated double bonds in the chain.
The molecular species containing the unsaturated acyl chain, however, does
occupy a larger surface area at equivalent surface pressure and this explains the
difference observed in the slopes of the initial pressure penetration curves. The
penetration of albumin into saturated and polyunsaturated molecular species of
phosphatidylcholines is found to produce similar results.
A systematic study of the effect of the chain length of phosphatidylcholines
on the penetration of serum albumin showed that the cutoff point for dimyristoyl
and dipalmitoyl molecular species was 20 mN•m–1 but increased to 35 and
41 mN•m–1 for the distearoyl and dibehenoyl molecular species, respectively.
These differences are likely to be due to the marked differences in the
properties of the respective monolayers. Thus distearoyl and dibehenoyl
monolayers are condensed at 20ºC, while dipalmitoyl- and dimyristoyl-
phosphatidylcholine monolayers are liquid-expanded. This again means that
the pressure limit of protein penetration cannot be related to the gaps between
the phospholipid molecules at the interface, since dimyristoyl-
phosphatidylcholine, which has the most space ‘available’, is not penetrated at
pressures which allow penetration into distearoyl- or dibehenoyl-
phosphatidylcholines.
It is not clear what additional forces assist protein penetration into the
longer-chain molecular species of phosphatidylcholines at higher pressures.
Direct hydrophobic interactions would involve extensive penetration of un-
folded domains of the protein because the hydrophobic side chains of individual
INTERACTION OF LIPIDS WITH PROTEINS AND POLYPEPTIDES 159
amino acid residues of the polypeptide chain are not long enough to penetrate
between the acyl chains of the phospholipid without some degree of unfolding.
enzyme activity. This suggests that while the substrate is presented in the form
of a bilayer the enzyme is sensitive to the local instability that assists orienta-
tion of the substrate about the active site of the enzyme. The interesting feature
of all reaction mixtures, however, was that the proportion of the substrate
hydrolysed during the lag period was invariably 0.10% of the total substrate
present. The explanation for this observation is that the creation of local rafts of
bilayer enriched in the diacylglycerol product of the reaction results in aggre-
gation of the vesicular substrate which, in turn, is responsible for the acceleration
in the rate of hydrolysis.
An indication of the effect of reaction products on the physical properties of
the substrate has been obtained from studies of the effect of phosphatidic acid
in monolayers of phosphatidylcholine subjected to hydrolysis by
phospholipase D from Streptomyces chromofuscus (El Kirat et al., 2002). The
enzyme is a member of a superfamily that includes endonucleases, helicases,
lipid synthetases and enzymes able to catalyse the formation or hydrolysis of
phosphodiester bonds. In its reaction against a substrate of phosphatidylcholine
a phosphatidyl enzyme intermediate is formed which is subjected to a nucle-
ophilic substitution by a molecule of water to release phosphatidic acid. The
effect of phosphatidic acid in monolayers of substrate at the air–water interface
on the surface elasticity modulus (Ks) and the lag period before accelerated
hydrolysis is observed is shown in Figure 7.7. The surface elasticity modulus is
derived from the surface pressure–area isotherm from the relationship:
⎛ ∂π ⎞
Ks = – A ⎜ –– ⎟
⎝ ∂A⎠
where A is the molecular area at the corresponding surface pressure, π. The
greater the value of Ks for a monolayer the less it is subject to deformation. The
correlation between Ks and lag time shown in Figure 7.7 suggests that deforma-
tion of the substrate–water interface is an essential step in orienting the enzyme
about the phospholipid substrate, and that the product is instrumental in
modulating this process.
A number of studies have examined the effects of substrate presentation and
the molecular species preferences of secretory phospholipase A2. This enzyme
hydrolyses the fatty acid esterified to the sn-2 position of the glycerol backbone
of diacylglycerophosphatides. The activities of the different subclasses of this
enzyme are known to be modulated by proteins and peptides such as melittin
and phospholipase A2-activating protein that are believed to act by modifying
the manner of presentation of the substrate (Koumanov et al., 2003). The
effects of such proteins in activating the enzymes differ depending on whether
the substrate is in the form of a dispersion of pure lipid or is present in a
biological membrane. These differences highlight the role of regulatory peptides
in the biological function of these phospholipases. The presence of non
INTERACTION OF LIPIDS WITH PROTEINS AND POLYPEPTIDES 161
Figure 7.7 The value of the surface elasticity modulus (Ks) at 30 mN•m–1 and the enzyme activity lag
time (determined as the time preceding a 5% decrease in monolayer area) plotted against the proportion
of phosphatidic acid mixed with the phosphatidylcholine substrate. Data taken from El Kirat et al., 2002.
Figure 7.8 A cartoon depiction of the fluid-mosaic model (after Singer & Nicolson, 1972).
Figure 7.10 Configurations of (a) fatty acid and (b) isoprenyl anchors of membrane proteins (P).
also a distinction between GPI-anchored proteins that are cycled through the
endosome compartment via clathrin-coated vesicles (such as prion protein) and
those that are not (including thy-1). There is accumulating evidence that the
lipid composition of membranes and the creation of lipid microdomains plays
a key role in the sorting of different GPI-anchored proteins during both
endocytosis from the cell surface and repackaging in the secretory process
(Mayor & Riezman, 2004).
The lipid anchors tethering proteins to the cytoplasmic surfaces of cell
membranes are either long-chain fatty acids or prenyl lipids. Cytosolic proteins
like v-Src associate with the plasma membrane via a single fatty acid, typically
myristate or palmitate, attached to an N-terminal glycine residue of the protein.
Other cytosolic proteins are anchored to the plasma membrane by prenylation
of one or two cysteine residues at or adjacent to the C-terminus of the protein;
examples of this group include Ras and Rab proteins. The structures of these
lipid anchors comprised of farnesyl (C15) and geranylgeranyl (C20) groups are
shown in Figure 7.10. The possession of a lipid anchor is not in itself sufficient
to permanently anchor the protein to the membrane; additional sites of interac-
tion with the lipids are also present. These additional attachment sites may be
a polybasic patch, exemplified by the six consecutive lysine residues in Ras that
bind to the head groups of acidic phospholipids. Alternatively, additional
palmitoylation via labile thioester linkages may strengthen the association of
the protein with the cytoplasmic membrane surface.
The lipidation of proteins appears to be a dynamic process, and palmitoylation
is not invariably accomplished with palmitic acid. Thus other fatty acids such
as myristic, stearic, oleic, linoleic and arachidonic acids have also been
INTERACTION OF LIPIDS WITH PROTEINS AND POLYPEPTIDES 167
F. Lipoproteins
Proteins can interact with lipids to form water-soluble complexes. Such
complexes are exploited in the mobilization of complex lipids within living
organisms. There are two main types of complex: complexes formed between
monomeric proteins and lipids, and large lipoprotein complexes.
1. Albumin
Free fatty acids bind to a variety of proteins during mobilization in the body.
They are transported in the bloodstream as a complex with plasma albumin.
Plasma albumin is a conformerically flexible protein that can adopt multiple
conformations of approximately equal energy to accommodate the binding of
ligands. The protein has at least five fatty acid binding sites, three of which are
of significantly higher affinity than the remainder. The mechanism of binding
of the fatty acids and other lipophilic drugs to the protein has been investigated
by nuclear magnetic resonance (NMR) methods (Lucas et al., 2004). The
average residence lifetime of a long-chain fatty acid bound to a high-affinity
site was found to be greater than 66 ms, whereas short-chain fatty acids like
octanoic acid have residence times of only a few milliseconds. The lifetimes are
relatively short because the fatty acids exchange readily between different
binding sites on the protein. Dissociation of the fatty acid from the protein, on
the other hand, takes place on a timescale of seconds.
The dissociation of fatty acids from plasma albumin is the rate-limiting step
in the delivery of fatty acids to target cells. Their dissociation from the complex
at the site of entry into cells is assisted by the presence in the plasma membrane
of proteins with a high affinity for fatty acids (McArthur et al., 1999). One such
protein, membrane fatty acid-binding protein (FABPpm), binds tightly to free
fatty acids and prevents destabilization of the membrane due to the presence of
the free fatty acid in the structure. The entry of the free fatty acid into the outer
168 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
2. Serum lipoproteins
Lipoprotein complexes circulate in the mammalian bloodstream to distribute a
cargo of lipids from their site of synthesis, usually the liver, to the peripheral
tissues. There has been an extensive research effort to characterize lipoproteins
because of their association with heart disease and atherosclerosis. The lipids,
mainly triacylglycerols, cholesterol and cholesterol esters, occupy a central
core surrounded by a shell of polar lipids and proteins. The proteins act to
stabilize the lipid droplet and provide recognition sites for targeting the
complex to the appropriate site of delivery. The particles are usually spherical
in shape and are classified according to their buoyant density. The characteris-
tics of the different classes of human lipoproteins are presented in Table 7.3. As
expected, the density increases with increasing protein to lipid ratio. The
protein components are synthesized in the liver and small intestine. While
many of the ten major apoproteins found in lipoproteins are common to more
than one class of lipoprotein, their combination is distinct in each of them.
Chylomicrons are the least dense of the lipoproteins and are responsible for
packaging fats, cholesterol and other lipids taken up from the diet in the
bloodstream and conducting them about the body. Because they consist almost
entirely of triacylglycerols they have a buoyant density of <0.95 g•cm–3. The
major protein is apolipoprotein B-48 (apoB-48), which has a molecular mass of
INTERACTION OF LIPIDS WITH PROTEINS AND POLYPEPTIDES 169
240 kDa and forms an amphipathic shell around the spherical fat globule, in
which the interior surface of the protein is hydrophobic and the exterior is
hydrophilic. Triacylglycerols and cholesterol are also synthesized in the liver,
and amounts in excess of requirements are packaged into very-low-density
lipoproteins and exported to the peripheral tissues. The proteins stabilizing
very-low-density lipoproteins include apoB-100 and apoE. ApoB-100 is an
extremely large protein comprised of more than 4500 amino acid residues and
has a molecular mass of 513 kDa. ApoB-48 is formed from the first 48% of
apoB-100 and arises from the post-transcriptional editing of apoB-100 mRNA
in the intestine. The relationship between apoB-100 and apoB-48 has been the
subject of considerable interest (Brodsky et al., 2004).
The intermediate-density lipoproteins result from depletion of the
triacylglycerol content of very-low-density lipoproteins by the action of lipases
associated with capillary surfaces and their consequent enrichment in choles-
terol esters. These intermediate-density lipoproteins may be taken up by the
liver and further processed or converted into low-density lipoproteins by
hydrolysis of more triacylglycerol. Low-density lipoprotein is the major carrier
of cholesterol and consists of a core of about 1500 cholesterol molecules
esterified mainly to linoleate. The non-polar lipid is stabilized by a monolayer
of phospholipid and apoB-100. High-density lipoprotein is also involved in
cholesterol transport but the source of cholesterol is that scavenged from
apoptosing and dying cells and membranes undergoing metabolic turnover.
The cholesterol is esterified to a long-chain fatty acid in a reaction catalysed by
an acyltransferase intrinsic to the high-density lipoprotein. The cholesterol
esters are rapidly transferred to very-low-density or low-density lipoproteins
by a specific transfer protein or targeted to the liver in their high-density
lipoprotein vector.
ApoA, apoC and apoE are referred to as exchangeable apolipoproteins and
they are responsible for regulating the traffic of lipids into and out of cells by
acting as cofactors for plasma enzymes and ligands for cell-surface receptors.
The exchangeable apolipoproteins share the same genomic heritage and there-
fore possess structural similarities (Saito et al., 2004). One particular feature is
170 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
nonpolar surface
positively-charged
residues Class A
negatively-charged
residues
Class Y
Class G
Figure 7.11 Helix wheel plots of the three classes of α-helical segments found in apoA-1, apoA-IV and
apoE apolipoproteins (adapted from Segrest et al., 1992).
a primary sequence arranged in α-helical motif in which the basic residues are
located near the hydrophilic–hydrophobic interface and acidic residues are
clustered at the centre of the polar face. The helical segments are often
interrupted by proline residues. This so-called Class A motif, which is exempli-
fied by apoA-1 lipoprotein, is illustrated together with Class Y and Class G
motifs in Figure 7.11. ApoE isoforms have a predominance of Class G helical
motifs and the Class Y helical motif dominates the secondary structure of
apoA-IV. The N-terminal amphipathic helices of apoE are bundled into 4
antiparallel strands forming an elongated globular structure with the hydropho-
bic faces oriented into the interior. The C-terminal adopts a coiled-coil helix
structure that is more exposed to the aqueous phase. In the absence of lipid,
apoA-1 adopts a two-domain configuration similar to apoE: an N-terminal
helical bundle extending into the central part of the primary structure and a
INTERACTION OF LIPIDS WITH PROTEINS AND POLYPEPTIDES 171
C-terminal helical domain that is less well organized. Both apoA-1 and apoE
aggregated into oligomers in the absence of lipid via hydrophobic interactions
between residues located at the C-termini of the respective molecules.
The self-association in aqueous media via the C-terminal domains of
apoA-1 and apoE is consistent with the role of the C-termini in the interaction
of these proteins with lipids. When apoE binds to lipid it has been suggested
that the initial binding takes place at the C-terminal, which then induces the
4-helix bundle of the N-terminus to reorganize so that the hydrophilic faces of
the helices open out and become available for binding to the lipid. Similar
reorganizations are believed to occur when apoA-1 interacts with lipid. Thus,
an initial binding takes place at the C-terminal domain, which is arranged in an
elongated hairpin structure. Following this there is a conformational change
involving residues 1–43 which serves to unmask a hydrophobic domain in
residues 44–65 of the protein.
The driving force for the formation of complexes between the apolipoproteins
and lipids appears to be a favourable change in enthalpy. A conformational
transition from random coil to α-helix on binding of apoA-1 to lipid, for
example, is associated with an enthalpy change of the order of –5 kJ•mol–1
per α-helical segment. This represents a total enthalpy change of about
–11 kJ•mol–1 and is additional to the enthalpy change accompanying the
interaction of apoA-1 with lipid, which is about –40 kJ•mol–1. The change in
conformation of the protein therefore contributes significantly to the binding
affinity between the protein and the lipid.
Two conformers of high-density lipoprotein have been characterized, one
discoid in shape and the other spherical. The discoid particles are comprised of
a phospholipid bilayer disk with two molecules of apoA-1 encircling the edges
where the acyl chains are exposed. The size is limited by the length of the
apoA-1 molecules, which are arranged in a belt of α-helices stacked one on the
other in an antiparallel orientation. A similar discoid particle has also been
described for apoE but in this structure 4 molecules of the protein are arranged
at the periphery of the disk. The spherical form of high-density lipoprotein
varies in diameter and has more neutral lipid than the discoid form. The
amphipathic helices of the apolipoproteins are believed to be interpolated
between the phospholipid molecules rather than at the periphery. The extent of
interaction of the protein with the lipid is greater as the proportion of protein in
the particle decreases.
presented in Figure 7.12. The most remarkable feature of the phase diagram is
that a relatively large proportion of the water-soluble protein can be incorpo-
rated to form a lipid–protein–water phase without any ionic interactions of the
type illustrated in monomolecular films described above. It was also found that
the protein occupied the aqueous phase of the complex in its native configura-
tion. As discussed below, this discovery was to have considerable implications
for the crystallization of both soluble and membrane proteins. The incorp-
oration of protein results in an expansion of the cubic phase lattice formed by
monoolein–water. The cubic domain of the phase diagram shown in
Figure 7.12 contains all three fundamental cubic structures observed in lipid–
water systems of the type found in monoolein–water, namely the gyroid surface
CG, the diamond surface CD, and the primitive surface CP, in order of increasing
protein:water ratio.
Complex cubic phases are formed with ternary lipid–water systems like
monoolein mixed in proportions of two parts protein solution or dispersion
with three parts of lipid. When such mixtures are treated with precipitants such
as non-ionic detergents or salts the protein begins to crystallize within hours of
incubation at 20ºC. The method can be used to grow crystals of soluble as well
as membrane proteins and other organic and inorganic molecules.
The precise process of crystallization from these tertiary lipid phases has
been examined in some detail (Misquitta et al., 2004). Precipitants like Na+/K+
phosphate salts, for example, provoke a reduction in water activity which
favours protein–protein interactions. Three-dimensional structures are created
INTERACTION OF LIPIDS WITH PROTEINS AND POLYPEPTIDES 173
References
Angeletti, S., Maggio, B. & Genti-Raimondi, S. (2004) Biochem. Biophys. Res. Commun.
314, 181–185.
Brezesinski, G. & Mohwald, H. (2003) Adv. Colloid Interface Sci. 100–102, 563–584.
Brockman, H.L. (2000) Biochimie 82, 987–995.
Brodsky, J.L., Gusarova, V. & Fisher, E.A. (2004) Trends Cardiovasc. Med. 14, 127–132.
El Kirat, K., Besson, F., Prigent, A.F., Chauvet, J.P. & Roux, B. (2002) J. Biol. Chem. 277,
21231–21236.
Ericsson, B., Larsson, K. & Fontell, K. (1983) Biochim. Biophys. Acta 729, 23–27.
Koumanov, K.S., Momchilova, A.B., Quinn, P.J. & Wolf, C. (2002) Biochem. J. 363, 45–51.
Koumanov, K., Momchilova, A. & Wolf, C. (2003) Cell Biol. Int. 27, 871–877.
Lahdo, R. & De La Fourniere-Bessueille, L. (2004) Biochem. J. 382, 987–994.
Lavoie, H., Desbat, B., Vaknin, D. & Salesse, C. (2002) Biochemistry 41, 13424–13434.
Lucas, L.H., Price, K.E. & Larive, C.K. (2004) J. Am. Chem. Soc. 126, 14258–14266.
Mayor, S. & Riezman, H. (2004) Nat. Rev. Mol. Cell Biol. 5, 110–120.
McArthur, M.J., Atshaves, B.P., Frolov, A., Foxworth, W.D., Kier, A.B. & Schroeder, F.
(1999) J. Lipid Res. 40, 1371–1383.
Misquitta, Y., Cherezov, V., Havas, F., Patterson, S., Mohan, J.M., Wells, A.J., Hart, D.J. &
Caffrey, M. (2004) J. Struct. Biol. 148, 169–175.
Moffett, S., Brown, D.A. & Linder, M.E. (2000) J. Biol. Chem. 275, 2191–2198.
Moffett, S., Rousseau, G., Lagace, M. & Bouvier, M. (2001) J. Neurochem. 76, 269–279.
Overath, P. & Engstler, M. (2004) Mol. Microbiol. 53, 735–744.
Quinn, P.J. & Dawson, R.M. (1970) Biochem. J. 116, 671–680.
Ruiz-Arguello, M.B., Goni, F.M. & Alonso, A. (1998) Biochemistry 37, 11621–11628.
Saito, H., Lund-Katz, S. & Phillips, M.C. (2004) Prog. Lipid Res. 43, 350–380.
Segrest, J.P., Jones, M.K., De Loof, H., Brouillette, C.G., Venkatachalapathi, Y.V. &
Anantharamaiah, G.M. (1992) J. Lipid Res. 33, 141–166.
Singer, S.J. & Nicolson, G.L. (1972) Science 175, 720–731.
Soulages, J.L., van Antwerpen, R. & Wells, M.A. (1996) Biochemistry 35, 5191–5198.
Tanner, M.J. & Boxer, D.H. (1972) Biochem. J. 129, 333–347.
Zidovetzki, R. & Lester, D.S. (1992) Biochim. Biophys. Acta 1134, 261–272.
CHAPTER 8
Emulsions
A. Oil-in-water emulsions
1. Emulsification
Dispersing one liquid phase in another requires energy input, which is directly
proportional to the increase in interfacial area and to the interfacial tension.
This energy can be provided by a stirrer, for example an ‘Ultra-Turrax’. The
higher the shear rate, the smaller are the droplets formed. Various mechanisms
behind droplet formation have been described by Walstra (1973). A droplet in
laminar flow will burst when the velocity gradient exceeds a critical value. The
viscosity of the dispersed phase in relation to the continuous phase is therefore
important. Turbulent flow is more efficient when the continuous phase has low
viscosity, for example.
If different equipment is compared, the particle size is reduced when an
Ultra-Turrax process is replaced by ultrasonication, and further reduced when
a valve homogenizer is used (the mixture is forced through a valve, resulting in
a pressure gradient). A log–log plot of the droplet diameter versus energy input
shows a linear relation. Pressure fluctuations result in cavity formation, which
is a major mechanism in the valve homogenizer and during ultrasonication. At
very high viscosity a colloid mill is the best equipment for emulsification.
175
176 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
Figure 8.1 Schematic illustration of a common globular protein, like β-lactoglobulin (left), and a
protein with a flexible peptide chain, like β-casein (right) at the oil–water interface.
2. Protein-stabilized emulsions
This is a common emulsion type in foods. The protein used will orient and
change its conformation at the oil–water interface in order to reduce the
interfacial energy. In the early literature it was usually assumed that the protein
molecules were unfolded. Studies by ellipsometry and atomic-force microscopy
have since modified this picture (e.g. see Wahlgren & Arnebrant, 1991). Thus
the molecules often tend to be globular and compact. There are two major
arrangements, as indicated in Figure 8.1.
Proteins with globular tertiary structure, like whey proteins, adsorb with a
conformation close to the native state. Proteins lacking a well-defined tertiary
structure, on the other hand, such as casein molecules, are more unfolded, with
the hydrophilic segments extending out into the water phase. The driving force
is the charges along the peptide chain, usually with a dominance of negative
groups. This provides an efficient steric stabilization, and the caseins are very
powerful emulsifiers. It takes several minutes for a protein to reach a plateau
value of interfacial tension, which means that after emulsification, various
relaxation phenomena take place. Ageing phenomena in interfacial protein
films have been followed by Dickinson (1992). In interfacial films of
β-lactoglobulin, an increase in film viscosity was seen corresponding to
disulphide bridge formation.
The main factor behind the high stability of emulsions based on proteins is
the irreversible character of the adsorption of the protein molecules at the
interface. Thus the peptide chain is attached at several positions and desorption
would require that all these positions are detached at the same time, which is
very unlikely.
EMULSIONS 177
Figure 8.2 Structure of an emulsion stabilized by a PC monolayer and a bilayer phase at the oil–water
interface.
3. Lipid-stabilized emulsions
In most cases when polar lipids are used to stabilize emulsions, they form a
separate phase at the oil–water interface. We might say that they form
‘interphases’. If a liposomal dispersion is formed by phosphatidylcholine (PC)
in water, the liposomes can encapsulate an introduced oil phase by mechanical
opening of the bilayer structure. The first emulsion used for parenteral nutri-
tion, ‘Intralipid’, prepared from egg yolk PC and soybean oil, consists of such
a lamellar liquid-crystalline phase at the oil–water interface, as shown in
Figure 8.2 (see also Chapter 11).
Friberg and co-workers introduced phase equilibria into discussions of lipid
emulsion stability (Friberg et al., 1969), which has been a fruitful approach.
When there is equilibrium between oil, water and the Lα phase, a stable
emulsion with a structure like that shown in Figure 8.2 can be obtained.
Such emulsions can be further stabilized if the Lα phase is transformed into
a gel phase (Larsson, 1978). Such an interface can be prepared by cooling
during homogenization through the temperature range at which the bilayers
will crystallize (Figure 8.3).
Technical emulsification processes used for foods and for pharmaceutical
products often involve homogenization under cooling. This results in forma-
tion of either a gel phase or surface-active crystals, which form a solid film at
the interface (Krog & Larsson, 1992). These leaf-shaped crystals expose a
polar head surface towards water and a methyl end group surface towards oil.
This phenomenon is neglected in the literature and will therefore be described
in detail here.
The interfacial tension at the sunflower oil–water interface at different
178 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
Figure 8.3 Effect of temperature on the oil–water interfacial tension with monostearin at different
concentrations solved in the oil (after Krog & Larsson, 1992).
Figure 8.4 Critical temperatures (Tγ) at which the interfacial tension starts to decline rapidly versus the
concentration of monomyristin (glycerol monomyristate; GMM), monostearin (GMS), and monobehenin
(GMB) (after Krog & Larsson, 1992).
EMULSIONS 179
Figure 8.5 Binary phase diagram for monomyristin–sunflower oil (after Krog & Larsson, 1992).
Figure 8.6 Interfacial tension versus temperature of: (A) 0.2% (w/w) glycerol monostearate
(monostearin) in sunflower oil towards distilled water; (B) pure sunflower oil towards 0.01% milk
proteins in water; (C) 0.2% glycerol monostearate in sunflower oil towards 0.01% milk proteins in water.
The situation in (C) is desired in ice cream (after Krog & Larsson, 1992).
Chapter 12). Milk proteins can give overly stable emulsions, but the addition of
some monoacylglycerols will solve the problem by squeezing out the protein
molecules from the interface, with the effect shown in Figure 8.6.
C. Demulsification
Flocculation and coalescence limit the lifetime of emulsions. Sometimes it is
important to break down an emulsion, and the mechanisms involved are
therefore important to understand. Flocculation is reversible, but if the inter-
facial film becomes thinner, film rupture and coalescence take place.
Addition of electrolytes will reduce the electrostatic repulsion and therefore
favour demulsification. Increasing the temperature also favours demulsification,
by decreasing the viscosity of the film between the particles. Physical
demulsification involves separation methods like separation, filtration and
flotation. More efficient is chemical demulsification, for example the use of
organic solvents or surfactants that destroy the interfacial film structure.
References
Busch, G. & Neuwald, F. (1973) In: Proceedings of the VII Congress IFSCC, Gesellschaft
Deutsches Kosmetik-Chemiker, Hamburg, Germany, p.171.
Dickinson, E.I. (1992) In: Emulsions – A Fundamental and Practical Approach (Sjöblom, J.,
ed.), Kluwer Academic Publishers, London, UK, p.23.
Friberg, S.E., Mandell, L. & Larsson, M. (1969) J. Colloid Interface Sci. 29, 155.
Hoppe, U. & Larsson, K. (1981) J. Dispersion Sci. Technol. 2, 433.
Krog, N. & Larsson, K. (1992) Fat Sci. Technol. 94, 55.
Larsson, K. (1978) Prog. Chem. Fats Other Lipids 16, 163.
McMahon, A.J.M. In: Emulsions – A Fundamental and Practical Approach (Sjöblom, J.,
ed.), Kluwer Academic Publishers, London, UK, p.135.
Pashley, R.M. (2002) J. Phys. Chem. B 107, 2724.
Wahlgren, M. & Arnebrant, T. (1991) Trends Biochem. 9, 209.
Walstra, P. (1973) Chem. Eng. Sci. 48, 333.
CHAPTER 9
Lipids of biological membranes
A. Introduction
All biological membranes contain a highly complex assortment of polar lipids.
While there are relatively few major lipid classes there is a whole spectrum of
molecular species within each class, which differ in the type, length and
number of unsaturated residues representing their hydrophobic component.
There is a general consensus, based on observations that have been obtained by
a variety of biophysical methods, that the lipid constituents of biological
membranes are arranged in a bilayer configuration in which the polar groups
are located in contact with the aqueous medium on the outside and the
hydrocarbon substituents are oriented towards the interior to form a hydro-
phobic domain that excludes water. It has been argued on the basis of the
hydrophilic to hydrophobic balance within the molecules that membrane lipids
have a relatively low critical micellar concentration, and that a discrete
distribution of domains within the molecule is responsible for creating a stable
bilayer structure.
Although the bilayer arrangement appears to be the dominant phase of many
lipids in biological membranes (Sachs et al., 2003) this is not necessarily the
phase preferred by individual molecular species of lipid isolated from particu-
lar biological membranes. If polar lipids, for example, extracted from biological
membranes are dispersed in aqueous media a mixture of phases is seen at
temperatures relevant to the growth of the organism from which the lipids were
extracted. If the constituent lipids are separated into molecular species and then
dispersed in aqueous media they are found to form one of a number of well-
characterized phases. The structures include bilayer phases in which the
hydrocarbon components are arranged in a crystal lattice, a gel phase or a liquid
crystal configuration. In addition, a number of non-bilayer phases, such as
hexagonal-II, cubic phases, etc. (Yang et al., 2003), are also found at tempera-
tures approximating the growth temperature of the organisms from which the
lipid was extracted. It is argued that lipids which tend to form non-bilayer
phases are constrained into a bilayer arrangement by their interaction with
other components of the membrane.
The composition of the molecular species, even of membranes that perform
very simple functions, is highly complex, and individual molecular species of
lipid may number more than 100 distinct chemical entities. In general, indi-
vidual molecular species of lipids that will form any one of these particular
phases are likely to be constituents of most biological membranes. Of particular
183
184 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
monitoring’ (SRM) mode with a maximum selectivity and sensitivity for each
of the lipid classes. Due to the experimental simplicity of the method, with no
pre-purification of the lipid extract required, the major advantages are (1) that
a definitive assessment of the fatty acid composition, which hitherto could not
be obtained from the overall carbon number and double bond content, can be
obtained, (2) minor molecular species among a few dominant species can be
identified, and (3) accurate quantification can be achieved, which is not
possible if the sample has to be pre-purified through multiple separation steps
with varying yield.
Nano-electrospray ionization tandem mass spectrometry (ESI–MS) brings
analysis of membrane lipids down to the picomole level. The soft ionization
procedure applied in ESI–MS is appropriate for small volumes of
chloroform:methanol extracts from less than 106 CHO cells (in practice down
to 103). The resolution of the molecular species of the distinct lipid classes and
for the plasmalogen analogues is obtained with negative ion mode [M–H]– for
anionic lipids and positive mode [M+H]+ for choline-containing phospholipids.
Ethanolamine phosphatides are responsive in both modes. When operated in
the single-stage MS, ESI mass spectra of lipid extracts show almost exclusively
molecular ion species without fragmentation. The assignments of the molecu-
lar ions can be obtained by recording the product ion spectra in Q2 after
selection in Q1 and activation in the collision cell. The polar head group
detected in Q2 can be lost as a charged or a neutral fragment. The proportion of
ion product (head group lost as a charged product) increases with the collision
energy, whereas neutral loss is favoured at voltages less than 30 V. In a
precursor ion scan the second mass analyser is set to transmit only ions from the
parent ion selected by Q1 (for instance Q2 is set to detect species with
m/z = 184 Da, corresponding to H + plus the phosphorylcholine of
phosphatidylcholine and sphingomyelin), whereas in the neutral loss scan the
analyser Q2 is scanned in a synchronized fashion with Q1 but with an offset to
lower m/z values equivalent to the neutral loss of interest (for instance, –183 Da
to monitor phosphatidylcholine).
Using the higher mass resolution of tandem Fourier transform ion cyclotron
resonance mass spectrometry (Fridriksson et al., 1999), over 90 different
phospholipid species were resolved in detergent-resistant membranes (DRM)
prepared from mast cells. Coupling between the high-affinity receptor for
immunoglobulin E (IgE), located on the external surface of the cell, and
tyrosine kinase, Lyn of the Src family, found exclusively on the cytoplasmic
side of the membrane, associated with the DRM, is accompanied by a shift to
polyunsaturated molecular species of choline- and amino-phosphatides found
in this sub-fraction. Taking into account the critical role of fatty acids in the
association of phospholipids with cholesterol, a transition of the liquid-ordered
to liquid-disordered fluid phase was inferred from this compositional shift (Ge
et al., 1999).
188 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
Table 9.1 Molar proportions of the major lipids of human, sheep and goat red blood cell
ghost membranes and detergent-resistant membranes (DRM) derived from them
Table 9.2 Composition of fatty acids amide linked to the sphingosine of sphingomyelin
isolated from human and goat erythrocyte ghosts and corresponding detergent-resistant
membrane (DRM) fractions
Fatty acid Human ghosts Human DRM Goat ghosts Goat DRM
Analysis of the proteins in such fractions shows that they are predominantly
attached to the membrane by lipid anchors. Such proteins include
glycosylphosphatidylinositol (GPI)-anchored proteins originating from the
outer surface of the plasma membrane and diacylated cytoplasmic proteins
such as certain Src family kinases originating from the inner surface of the
membrane (Bagnat et al., 2000; Moffett et al., 2000). The lipid composition of
rafts, like the protein constituents, differs from the composition of the parent
membrane, providing evidence that fractionation depends on the solubility of
the lipid in detergent solution.
Differential solubility in non-ionic detergents such as Triton X-100, CHAPS
and Lubrol has been exploited to define trafficking pathways of membrane
proteins in cells (Mairhofer et al., 2002; Sakyo & Kitagawa, 2002; Slimane et
al., 2003). The implication of such studies is that membrane lipid domains are
primarily responsible for directing the traffic.
Preparations of DRM fractions insoluble in 1% Nonidet P-40 have been used
to examine signal transduction pathways in lymphocytes (Zubiaur et al., 2002).
CD38 signal transduction in T-lymphocytes was said to be due to formation
of supramolecular signalling complexes of CD3 molecules bearing immuno-
receptor tyrosine-based activation motifs together with protein tyrosine kinases.
In other studies insoluble fractions in 1% Brij 58 at 4ºC were used to demon-
strate that T-cell receptors are associated into cell-surface domains upon
receptor stimulation, providing a mechanism for coupling receptor location
with downstream signalling cascades (Montixi et al., 1998).
A B
Figure 9.1 Freeze-fracture electron micrographs of: (A) chloroplast thylakoid membrane; (B) total
polar lipid extract of chloroplast thylakoid membrane dispersed in the same medium as in (A).
Figure 9.2 Electron micrographs of freeze-fracture replicas of chloroplasts showing the effect of
temperature on particle distribution in thylakoid membranes of Arabidopsis mutants defective in the
synthesis of unsaturated membrane lipids. (A) Chloroplasts from wild-type Arabidopsis and (B)
chloroplasts from Arabidopsis with the fad6 mutation, both thermally quenched from 20ºC. (C,D)
chloroplasts from wild-type Arabidopsis equilibrated and thermally quenched from 4ºC. The scale bar
corresponds to 100 nm.
Figure 9.3 Genetic manipulation of membrane lipid unsaturation of Nicotiana tabacum and Arabidopsis
and the consequence on exposure of the transgenic plants to chilling conditions.
This section will review the properties of some of these hydrolases and factors
that are responsible for their regulation.
1. Phospholipases A2 (PLA2)
Hydrolases that attack the acyl ester bonds linking fatty acids to the sn-1 and sn-
2 positions of the glycerol backbone of phospholipid molecules are categorized
by their positional specificity as A1 and A2, respectively. A number of early
studies showed evidence of phospholipase A activity hydrolysing fatty acyl
residues at the C-2 position of the glycerol moiety of phosphoacylglycerols
(PLA2) by enzymes that were bound to cell membranes and acting on endog-
enous substrates. The enzyme activity was shown to change the level of cAMP
in fat cells (de Cingolani et al., 1972) and to be influenced by the insulin:glucagon
ratio (Polonovski et al., 1974; Wolf et al., 1977). The possibility that the
enzyme retro-regulates membrane fluidity was also implied in these early
reports (Petkova et al., 1987). The notion that these enzymes are involved in
signal transduction by mediation of metabolic turnover of membrane
phospholipids has been less favoured by the discovery of cytosolic species of
PLA2 (cPLA2), whose action appears more appropriate to this function (Clark
et al., 1991). It was also realized that phospholipids of the membrane matrix
were relatively resistant to enzyme hydrolysis. A number of mechanisms are
responsible for the protection of phospholipids against PLA2 attack. These
include:
(1) Limitation of the penetration of the enzyme into the substrate due to the
tight packing of the lipid molecules.
(2) Restricted access of the enzyme to its preferred substrate. For instance,
exogenous secretory type II PLA2 outside cells cannot gain access to
phosphatidylethanolamine substrate located in the inner leaflet of plasma
membranes.
(3) Dilution of susceptible substrates within non-substrate membrane lipids
such as sphingolipids and cholesterol.
These limitations serve to restrict enzyme activity to a relatively small propor-
tion of the lipids forming the membrane lipid matrix, leading to the conclusion
that the hydrolysis of such a small number of molecules would be unlikely to
result in any physiological consequences.
Detection of phospholipid turnover resulting from the action of endogenous
PLA2 acting on the membrane lipid matrix requires pre-labelling of
phospholipids by biosynthetic incorporation of radiolabelled precursors, either
32
P- or 14C-labelled fatty acids, administered beforehand to the organism.
Furthermore, the release of lysoderivatives and fatty acids can only be related
to PLA2 under particular experimental conditions in which subsequent enzyme
action on the products (lysophospholipases, transacylase, oxygenation of
LIPIDS OF BIOLOGICAL MEMBRANES 203
PUFA, etc.) are negligible. These conditions are not easily achieved and assays
of enzyme activity against exogenous substrate, i.e. dispersions of radiolabelled
or fluorescent analogues of phospholipids, are assumed to reflect the mem-
brane-bound enzyme activities. Despite such problems it has been shown that
activities against exogenous and endogenous substrates vary according to the
fluidity of the membrane matrix (Momchilova et al., 1985, 1986).
The demonstration that cPLA2 is the particular phospholipase involved in the
release of arachidonic acid from the sn-2 position of phosphatidylcholine under
conditions where the cascade of reactions leading to eicosanoid biosynthesis is
triggered has been fundamental to understanding the role of PLA2 in this
process (Clark et al., 1991). In this work it was shown that cloning and
expression of a cDNA encoding a high molecular mass (85.2 kDa) cPLA2,
assigned to type IV PLA2, has no detectable sequence homology with the non-
pancreatic secreted forms of PLA2 (type II) described previously. Whereas
type II PLA2 represented non-specific activities observed earlier, cPLA2 selec-
tively cleaved arachidonic acid from microsomes of intact biological
membranes. It was demonstrated that cPLA2 translocated to membrane vesicles
in response to physiologically relevant changes in free calcium concentration.
By contrast, secretory type II PLA2 is known to be calcium-dependent with an
optimum concentration greater than 1 mM consistent with its preferential
extracellular activity. Moreover, an amino-terminal 140 amino acid fragment
was identified in cPLA2 which translocated to natural membrane vesicles in a
Ca2+-dependent fashion. Maximal activation required phosphorylation of Ser-
505 via the MAP kinase (mitogen-activated protein kinase) pathway (Lin et al.,
1993). Treatment of cells with agents that stimulated the release of arachidonic
acid caused increased serine phosphorylation and activation. The site of cPLA2
phosphorylation by MAP kinase, Ser-505, was identical to the major site of
cPLA2 phosphorylation observed in phorbol ester-treated cells. Replacement
of Ser-505 with Ala resulted in a mutant cPLA2 that caused little or no enhanced
agonist-stimulated arachidonate release from intact cells.
The delineation of the two functionally distinct domains of cPLA 2 was
approached experimentally by Nalefski et al. (1994). Isolation and sequence
analysis of cPLA2 cDNA clones from four different species revealed several
highly conserved regions. The N-terminal conserved region is homologous to
that of a number of other Ca2+-dependent lipid-binding proteins. The first 178
residues of cPLA2, containing the homologous Ca2+-dependent lipid-binding
(CaLB) motif, and another recombinant protein containing the cPLA2(1–178)
fragment placed at the C- terminus of the maltose-binding protein (MBP-
CaLB) associated with membranes in a Ca2+-dependent manner. Both cPLA2
and MBP-CaLB also bind to synthetic liposomes at physiological Ca2+ concen-
trations, demonstrating that accessory proteins are not required for this process.
In contrast, ΔC2, a truncated cPLA2 lacking the CaLB domain, failed to
associate with membranes and failed to hydrolyse liposomal substrates but
204 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
2. Acyltransferases
The activity of acyltransferase able to reverse the action of PLA2 was initially
LIPIDS OF BIOLOGICAL MEMBRANES 205
3. Acylation/deacylation cycle
Few details have been published on the relative activities of PLA2 and
acyltransferase in relation to the homeostatic regulation of membrane lipids.
Regulation of phosphorylation/dephosphorylation was suggested to be a key
factor from studies of the incorporation of 14C-labelled palmitoyl CoA into
membrane phospholipids via the deacylation/acylation cycle conducted in rat
LIPIDS OF BIOLOGICAL MEMBRANES 207
liver microsomes (Hutson & Higgins, 1987). This activity was reversibly
inactivated by treatment of the microsomes with 105 000 g supernatant in
either the presence or absence of ATP and MgCl2. These observations suggest
that the acylation cycle is controlled by a mechanism involving phosphoryla-
tion/dephosphorylation. Because the pool of lysolecithin in the membranes
was not altered by conditions that increase incorporation of palmitoyl CoA into
phospholipid, it is likely that the site of regulation of deacylation/acylation is
the acyltransferase reaction rather than the phospholipase. This conclusion was
reached on the basis of studies of resting tissue; however, results obtained on
activation of granulocytes provided a different perspective (Tou, 1981). The
effect of phorbol myristate acetate, known to reproduce the stimulated oxidative
activities characteristic of phagocytosis, was examined on the metabolism of
the fatty acyl groups of granulocyte phospholipids and compared with that of
phagocytic stimuli. Phorbol myristate acetate was found to stimulate the
labelling of phosphatidylethanolamine, phosphatidylcholine and
phosphatidylinositol by 1-[14C]-palmitic acid but not by U-[14C]-glycerol.
Challenge of the cells with starch granules, by contrast, selectively increased
the labelling of phosphatidylinositol by both radioactive tracers. Labelled
palmitic acid was found at both the sn-1 and sn-2 positions of phospholipids;
more radioactivity was recovered from the 2-position. The radioactivity at both
positions was enhanced in stimulated cells. These data suggest that phorbol
myristate acetate increased palmitic acid incorporation into glycero-
phospholipids by increasing the acylation of the lysoderivatives, and that starch
granules enhanced the formation of phosphatidylinositol via both de novo
synthesis and acylation of the lysoderivative. Both phorbol myristate acetate
and starch granules selectively augmented the incorporation of 1-[14C]-
arachidonic acid into phosphatidylinositol, which exhibited the highest
specific radioactivity among the phospholipids in control and in stimulated
cells. The significance of the increased incorporation of arachidonic acid into
phosphatidylinositol is thought to reflect the degradation that produces the
eicosanoid precursor.
The high specificity of the recipient lysoderivative has been examined in
platelets, another activable circulating cell (McKean & Silver, 1985). The
transfer of unsaturated fatty acyl groups to 1-alkyl-sn-glycero-3-phosphocholine
is many-fold slower than to 1-acyl-sn-glycero-3-phosphocholine. The CoA
esters of linoleate and arachidonate, two unsaturated fatty acyl groups com-
monly found in platelet phospholipids, are the preferred fatty acyl group
donors. In macrophages, three mechanisms of reacylation participating in the
remodelling of membrane phospholipids were examined. Intact alveolar
macrophages were found to acylate alkyl- and acyl-lysophospholipids with a
high selectivity for arachidonate. A specific mechanism appears responsible
for the incorporation of arachidonate into lysophospholipids in intact cells,
since the kinetic pattern for the formation of the C20:4 species was different from
208 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
that of all other species. This specificity was investigated in more detail by
examining the enzymatic acylation of 1-alkyl-2-lyso-sn-glycero-3-
phosphocholine by membrane fractions; in the absence of CoA, ATP, and
Mg2+, this lysophospholipid was still re-acylated with a high preference for
arachidonate that was independent of added free fatty acids. The addition of
CoA alone increased the rate of acylation of 1-alkyl-2-lyso-sn-glycero-3-
phosphocholine, mainly due to an increase in the formation of species other
than those containing arachidonate. When CoA, ATP, and Mg2+ were present,
the macrophage membranes catalysed the acylation of 1-alkyl-2-lyso-sn-
glycero-3-phosphocholine without preference for arachidonate. The acylation
of alkyl- and acyl-lysophospholipids by rabbit alveolar macrophages appears
to take place by three distinct mechanisms: a CoA-independent transacylation,
a CoA-dependent transacylation (reverse reaction catalysed by acyl-CoA
acyltransferase), and an acyl-CoA-dependent acylation. The CoA-independent
transacylation reaction is unique in that it is specific for arachidonate and
accounts for the selective acylation of alkyl- and acyl-lysophospholipids by
arachidonate in membrane preparations of alveolar macrophages. This reaction
appears to be extremely important in the remodelling of phospholipid molecu-
lar species and the mobilization of arachidonate into ether-linked lipids. The
transfer of arachidonate to 1-alkyl-2-lyso-sn-glycero-3-phosphocholine is also
of importance in the termination step for platelet activating factor (1-alkyl-2-
acetyl-sn-glycero-3-phosphocholine; PAF) activity, whereby 1-alkyl-2-
arachidonoyl-sn-glycerol-3-phosphocholine (a stored precursor of both PAF
and arachidonic acid metabolites) is restored.
In rabbit alveolar macrophages the transacylation system was shown to
exhibit a complex selectivity according to distinct donor and acceptor and CoA
dependency (Sugiura et al., 1987). Microsomes were found to acylate 1-[3H]-
alkyl-glycero-3-phosphocholine (1-alkyl-GPC; lyso-PAF) in the absence of
any cofactors, indicating the presence of transacylation activity. The
transacylation activity was comparable to the activity of acyl-CoA:1-alkyl-
GPC acyltransferase. The fatty acyl moieties introduced into 1-alkyl-GPC from
membrane lipids by microsomes were mainly C20:4n-6. [14C]-labelled C20:4n-6,
C20:5n-3, C22:4n-6, and C22:6n-3 were transferred efficiently from diacyl-GPC to
1-alkyl-GPC in a CoA-independent manner. The transfer rates for C16:0, C18:0,
and C18:1 from diacyl-GPC to 1-alkyl-GPC were relatively low in the presence
and absence of CoA. On the other hand, the transfer of C20:4 from diacyl-
glycero-3-phosphoethanolamine (GPE) or diacyl-glycero-3-phosphoinositol
(GPI) to 1-alkyl-GPC or 1-acyl-GPC was markedly increased by the addition of
CoA. This observation confirmed that acylation can be a specific and active
pathway for polyunsaturated fatty acids cleaved from the sn-2 position of
phospholipids serving as substrate during cell activation. In the activated
human neutrophil the circulation of arachidonate between alkyl and alkenyl
derivatives participates in the generation of the lysoderivative precursor of
LIPIDS OF BIOLOGICAL MEMBRANES 209
PAF (Sugiura et al., 1987). These studies indicate that lyso-PAF is formed by
the transfer of arachidonate from 1-O-alkyl-2-arachidonoyl-GPC to the alkenyl-
lyso-GPE by a CoA-independent transacylase reaction.
Mass measurements revealed a rapid loss of arachidonate from 1-radyl-2-
acyl-GPE and a concomitant increase in alkenyl-lyso-GPE upon stimulation of
the neutrophils by addition of ionophore A23187. The circulation of polyun-
saturated fatty acids between alkyl-PC and lyso-PAF via alkenyl-PE was
demonstrated in a variety of tissues (Blank et al., 1995; Nixon et al., 1996).
Microsomal membranes from six different rat tissues (spleen, lung, kidney,
brain, testis, and liver) were found to possess CoA-independent transacylase
activity that could both acylate lyso-PAF (1-hexadecyl-2-lyso-sn-glycero-3-
phosphocholine) and then deacylate the 1-hexadecyl-2-acyl-sn-glycero-
3-phosphocholine product via the transacylation of added exogenous 1-alk-1′-
enyl-2-lyso-sn-glycero-3-phosphoethanolamine. Analysis of molecular species
of 1-hexadecyl-2-acyl-sn-glycero-3-phosphocholine before and after addition
of 1-alk-1′-enyl-2-lyso-sn-glycero-3-phosphoethanolamine as the acyl
acceptor demonstrated a high selectivity for polyunsaturated fatty acids (>3
double bonds/acyl group) in both the acylation and deacylation processes that
occurred in testicular microsomal membranes. The transfer of acyl groups by
the transacylase appeared to be equally effective for either arachidonic or
docosapentaenoic (n-6) fatty acids, whereas linoleic and oleic acids were not
transferred. The results indicate the PAF-related transacylase is widely
distributed among tissues and, although highly selective for polyunsaturated
acyl groups, does not discriminate selectively among the polyunsaturates.
A very particular role was ascribed to the deacylation/reacylation
of lysoderivatives on the inner monolayer of red blood cell membrane. This
was the maintenance of the highly asymmetric distribution of
phosphatidylethanolamine (PE) in ruminant membrane and the relatively low
content of phosphatidylcholine (PC) (Florin-Christensen et al., 2001). Rumi-
nant erythrocytes are remarkable for their choline-phospholipid anomalies,
namely low or absent PC along with high sphingomyelin levels. Another
anomaly in bovine erythrocytes affects aminophospholipids: PE shows almost
absolute asymmetry, with only 2% of the total present in the outer leaflet.
Furthermore, PLA2, an enzyme located on the external surface of the erythro-
cytes, shows 3-fold greater activity against PC than against PE. Because
acylation of PE is by far the most important biosynthetic event in this cell
following deacylation by PLA2 to generate lyso compounds, the selective
reacylation of lyso-PE on the inner side can account for the asymmetry of PE
distribution, and the departure of lyso-PC extracted by serum albumin for the
low content of PC in bovine erythrocyte membranes.
210 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
4. Phospholipases C
5. Phospholipases D
Phospholipase D (PLD) hydrolyses phospholipids, yielding phosphatidic acid
and water-soluble bases such as choline (Singer et al., 1997). The role of
phosphatidic acid generation in membranes and its role in signalling remains
conjectural. The membrane-bound product does not occur in appreciable
proportions although locally it may be generated in amounts that may have
consequential effects on membrane structure and stability. Efforts to implicate
phospholipase D in signalling have therefore concentrated on how and where
the enzyme is activated on a subcellular level.
Two genes coding for phospholipase D activity have been identified in
mammals: PLD1 and PLD2. Both genes have been cloned and overexpressed
in different cell lines (Frohman et al., 1999). The enzyme PLD1 is located in
association with intracellular membranes and is known to be active in living
cells. PLD2, by contrast, is associated with plasma membranes but has low
resting activity that can be activated by a variety of factors including protein
kinase C, family members of ADP-ribosylation factor and Rho, and the lipid
PIP2 (Singer et al., 1997; Exton, 2002).
Targeting of PLD1 to membranes may involve palmitoylation of cysteine
residues at a domain near the C-terminus of the protein (Sugars et al., 2002).
The importance of the domain in the vicinity of the palmitoylated cysteinyl
residues suggested that not only did palmitoylation target the enzyme to the
membrane but there was specificity in the membrane site of interaction.
How phosphatidic acid may act as a signalling device at the sites of enzyme
activity has been inferred from evidence that the phospholipid binds to specific
proteins. Candidates for phosphatidic acid binding include Raf-1, a serine/
threonine kinase (Rizzo et al., 2000), cAMP-specific phosphodiesterase (Grange
212 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
et al., 2000), the mammalian target of rapamycin, mTOR (Fang et al., 2001),
protein phosphatase-1 (Jones & Hannun, 2002) and Src homology 2-domain
containing protein tyrosine phosphatase (Frank et al., 1999). Recently, a solid-
phase adsorption system has been described to identify traffic-related
phosphatidic acid binding proteins and proteins that may be implicated in PLD-
dependent pathways, and a number of specific proteins have been characterized
(Ktistakis et al., 2003).
6. Sphingomyelinases
Sphingomyelinases are enzymes that cleave the phosphorylcholine moiety
from sphingomyelin to yield lipophilic ceramide. The significance of
sphingomyelinases has been recognized by the discovery that sphingomyelin is
a substrate progenitor for initiation of signal transduction pathways (Kolesnick,
1991) in which ceramide acts as a second messenger in functions like cell
differentiation, proliferation and apoptosis (Venkataraman & Futerman, 2000;
Ohanian & Ohanian, 2001; Hannun & Obeid, 2002). Five different categories
of sphingomyelinase have been characterized on the basis of their optimal
conditions for catalytic activity (Samet & Barenholz, 1999), and a further
sphingomyelinase has been identified in bacteria (Goni & Alonso, 2002).
Evidence from model membrane studies has indicated that the action of
sphingomyelinase in generating ceramide from sphingomyelin can induce
lamellar to non-lamellar phase transitions leading to membrane fusion
(Kinnunen & Holopainen, 2000) and to lateral phase separations of ceramide-
enriched domains which exist in gel phase in fluid bilayers of phospholipid
(Huang et al., 1996). Ceramide, but not dihydroceramide, is also able to induce
the formation of pores in phospholipid bilayers, a property that has been
attributed to the extensive hydrogen-bonding capacity of ceramide (Siskind et
al., 2002).
There is evidence that the action of neutral sphingomyelinase causes domain
formation of phases enriched in ceramide and a consequent clustering of cell-
surface receptors. One example is the clustering of L-selectin in lymphocytes
(Junge et al., 1999). The action of a Zn-dependent acid sphingomyelinase in
response to interleukin-1β treatment of human fibroblasts has been found to be
associated with depletion of sphingomyelin and a corresponding increase of
ceramide in the caveolae compartment (Liu & Anderson, 1995). The interleukin-
1β treatment was also shown to be associated with decreased platelet-derived
growth factor-induced thymidine uptake by these cells, suggesting that pertur-
bation of transmembrane signalling processes had occurred. Similar studies of
the action of nerve growth factor signalling have emphasized the importance of
intact caveolae for sphingomyelinase action (Dobrowsky, 2000).
Ceramide generation and the creation of rafts has been shown to be essential
for optimal Fas signalling and induction of apoptosis in both B- and T-
LIPIDS OF BIOLOGICAL MEMBRANES 213
lymphocytes (Kirschnek et al., 2000). On the basis of these studies a model has
been proposed for the action of ceramide in signalling processes associated
with apoptosis (Cremesti et al., 2002). Essentially, the engagement of Fas
triggers translocation of acid sphingomyelinase to the plasma membrane,
where it acts on its substrate segregated into sphingomyelin–cholesterol rafts.
The formation of ceramide induces coalescence of the rafts into large domains
in which oligomerization of downstream effectors such as FADD/MORT-1
and pro-caspase-8 can take place, leading to the Fas death signal.
E. Conclusions
The lipid matrix of biological membranes consists of a complex assortment of
polar lipids arranged in a bilayer configuration. The matrix usually consists of
a mixture of many different molecular species within each lipid class present.
The arrangement of lipids in the bilayer is asymmetric. Lateral asymmetry and
the creation of domains with defined physical properties is produced by
specific interactions between certain lipids such as the complex formed be-
tween cholesterol and the more saturated molecular species of
choline-containing phospholipids. Transbilayer asymmetry is generated by
ATP-dependent aminophospholipid translocases that sequester the
aminophospholipids on the cytoplasmic membrane leaflet.
The precise molecular species of lipid and the proportions in which they are
present differ from one membrane to another but the identity is preserved
within relatively narrow limits. The biochemical mechanisms responsible for
maintaining membrane homeostasis are not fully understood. Hydrolytic en-
zymes specific for particular ester bonds in complex lipids are reasonably well
characterized and appear to be involved in regulating the metabolic turnover of
the lipids. The regulation of such enzymes has been shown to represent the
molecular mechanism of a range of membrane functions, including transmem-
brane signal transduction.
References
Ahmed, S.N., Brown, D.A. & London, E. (1997) Biochemistry 36, 10944.
Anderson, R.G. & Jacobson, K. (2002) Science 296, 1821.
Ariizumi, T., Kishitani, S., Inatsugi, R., Nishida, I., Murata, N. & Toriyama, K. (2002) Plant
Cell Physiol. 43, 751.
Artus, N.N., Uemura, M., Steponkus, P.L., Gilmour, S.J., Lin, C. & Thomashow, M.F. (1996)
Proc. Natl Acad. Sci. USA 93, 13404.
Bagnat, M., Keranen, S., Shevchenko, A. & Simons, K. (2000) Proc. Natl Acad. Sci. USA 97,
3254.
Blank, M.L., Smith, Z.L., Fitzgerald, V. & Snyder, F. (1995) Biochim. Biophys. Acta 1254,
295.
Braccia, A., Villani, M., Immerdal, L., Niels-Christiansen, L.L., Nystrom, B.T., Hansen,
G.H. & Danielsen, E.M. (2003) J. Biol. Chem. 278, 15679.
214 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
Illenberger, D., Walliser, C., Strobel, J., Gutman, O., Niv, H., Gaidzik, V., Kloog, Y.,
Gierschik, P. & Henis, Y.I. (2003b) J. Biol. Chem. 278, 8645.
Ishizaki-Nishizawa, O., Fujii, T., Azuma, M., Sekiguchi, K., Murata, N., Ohtani, T. & Toguri,
T. (1996) Nature Biotechnol. 14, 1003.
Janes, P.W., Ley, S.C., Magee, A.I. & Kabouridis, P.S. (2000) Semin. Immunol. 12, 23.
Jones, J.A. & Hannun, Y.A. (2002) J. Biol. Chem. 277, 15530.
Junge, S., Brenner, B., Lepple-Wienhues, A., Nilius, B., Lang, F., Linderkamp, O. & Gulbins,
E. (1999) Cell Signal. 11, 301.
Karlsson, O.P., Rytomaa, M., Dahlqvist, A., Kinnunen, P.K. & Wieslander, A. (1996)
Biochemistry 35, 10094.
Karlsson, O.P., Dahlqvist, A., Vikstrom, S. & Wieslander, A. (1997) J. Biol. Chem. 272, 929.
Katsikas, H. & Wolf, C. (1995) Biochim. Biophys. Acta 1258, 95.
Kelley, G.G., Reks, S.E., Ondrako, J.M. & Smrcka, A.V. (2001) EMBO J. 20, 743.
Kenworthy, A. (2002) Trends Biochem. Sci. 27, 435.
Kerkhoff, C., Trumbach, B., Gehring, L., Habben, K., Schmitz, G. & Kaever, V. (2000) Eur.
J. Biochem. 267, 6339.
Kinnunen, P.K. & Holopainen, J.M. (2000) Biosci. Rep. 20, 465.
Kirkpatrick, F.H., Gordesky, S.E. & Marinetti, G.V. (1974) Biochim. Biophys. Acta 345, 154.
Kirschnek, S., Paris, F., Weller, M., Grassme, H., Ferlinz, K., Riehle, A., Fuks, Z., Kolesnick,
R. & Gulbins, E. (2000) J. Biol. Chem. 275, 27316.
Klapisz, E., Masliah, J., Bereziat, G., Wolf, C. & Koumanov, K.S. (2000) J. Lipid Res. 41,
1680.
Klein, W., Weber, M.H. & Marahiel, M.A. (1999) J. Bacteriol. 181, 5341.
Kodama, H., Hamada, T., Horiguchi, G., Nishimura, M. & Iba, K. (1994) Plant Physiol. 105,
601.
Kodama, H., Horiguchi, G., Nishiuchi, T., Nishimura, M. & Iba, K. (1995) Plant Physiol. 107,
1177.
Kolesnick, R.N. (1991) Prog. Lipid Res. 30, 1.
Koumanov, K.S., Quinn, P.J., Bereziat, G. & Wolf, C. (1998) Biochem. J. 336, 625.
Koumanov, K.S., Tessier, C., Momchilova, A.B., Rainteau, D., Wolf, C. & Quinn, P.J. (2005)
Archiv. Biochem. Biophys. 432, 150.
Ktistakis, N.T., Delon, C., Manifava, M., Wood, E., Ganley, I. & Sugars, J.M. (2003)
Biochem. Soc. Trans. 31, 94.
Kunst, L., Browse, J. & Somerville, C. (1989a) Plant Physiol. 91, 401.
Kunst, L., Browse, J. & Somerville, C. (1989b) Plant Physiol. 90, 943.
Lehel, C., Los, D., Wada, H., Gyorgyei, J., Horvath, I., Kovacs, E., Murata, N. & Vigh, L.
(1993) J. Biol. Chem. 268, 1799.
Li, X.M., Momsen, M.M., Smaby, J.M., Brockman, H.L. & Brown, R.E. (2001) Biochemistry
40, 5954.
Lichtenbergova, L., Yoon, E.T. & Cho, W. (1998) Biochemistry 37, 14128.
Lin, L.L., Wartmann, M., Lin, A.Y., Knopf, J.L., Seth, A. & Davis, R.J. (1993) Cell 72, 269.
Lindblom, G., Brentel, I., Sjolund, M., Wikander, G. & Wieslander, A. (1986) Biochemistry
25, 7502.
Liu, P. & Anderson, R.G. (1995) J. Biol. Chem. 270, 27179.
London, E. (2002) Curr. Opin. Struct. Biol. 12, 480.
Lopez, I., Mak, E.C., Ding, J., Hamm, H.E. & Lomasney, J.W. (2001) J. Biol. Chem. 276,
2758.
Magargal, W.W., Dickinson, E.S. & Slakey, L.L. (1978) J. Biol. Chem. 253, 8311.
Mairhofer, M., Steiner, M., Mosgoeller, W., Prohaska, R. & Salzer, U. (2002) Blood 100, 897.
McKean, M.L. & Silver, M.J. (1985) Biochem. J. 225, 723.
McLaughlin, J.N., Thulin, C.D., Bray, S.M., Martin, M.M., Elton, T.S. & Willardson, B.M.
(2002) J. Biol. Chem. 277, 34885.
216 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
Miguel, M., Jas, D., Dooner, H. & Browse, J. (1993) Proc. Natl Acad. Sci. USA 90, 6208.
Moffett, S., Brown, D.A. & Linder, M.E. (2000) J. Biol. Chem. 275, 2191.
Momchilova, A., Petkova, D., Mechev, I., Dimitrov, G. & Koumanov, K. (1985) Int. J.
Biochem. 17, 787.
Momchilova, A., Petkova, D. & Koumanov, K. (1986) Int. J. Biochem. 18, 659.
Momchilova, A., Markovska, T. & Pankov, R. (1999) Biochem. Mol. Biol. Int. 47, 555.
Montixi, C., Langlet, C., Bernard, A.M., Thimonier, J., Dubois, C., Wurbel, M.A., Chauvin,
J.P., Pierres, M. & He, H.T. (1998) EMBO J. 17, 5334.
Murata, N. & Wada, H. (1995) Biochem. J. 308(1), 1.
Murata, N., Ishizaka-Nishizawa, O., Higashi, S., Hayashi, H., Tasaka, Y. & Nishida, I. (1992)
Nature 356, 710.
Nalefski, E.A., Sultzman, L.A., Martin, D.M., Kriz, R.W., Towler, P.S., Knopf, J.L. & Clark,
J.D. (1994) J. Biol. Chem. 269, 18239.
Nishida, I., Swinhoe, R., Slabas, A.R. & Murata, N. (1996) Plant Mol. Biol. 31, 205.
Nixon, A.B., Greene, D.G. & Wykle, R.L. (1996) Biochim. Biophys. Acta 1300, 187.
Ohanian, J. & Ohanian, V. (2001) Cell Mol. Life Sci. 58, 2053.
Okuyama, H., Saito, M., Joshi, V.C., Gunsberg, S. & Wakil, S.J. (1979) J. Biol. Chem. 254,
12281.
Oquist, G., Campbell, D., Clarke, A.K and Gustafsson, P. (1995) Photosynthesis Res. 46, 151.
Palta, J.P., Whitaker, B.D. & Weiss, L.S. (1993) Plant Physiol. 103, 793.
Parton, R.G., Molero, J.C., Floetenmeyer, M., Green, K.M. & James, D.E. (2002) J. Biol.
Chem. 277, 46769.
Petkova, D.H., Momchilova-Pankova, A.B. & Koumanov, K.S. (1987) Biochimie 69, 1251.
Philip, F., Guo, Y. & Scarlata, S. (2002) FEBS Lett. 531, 28.
Polonovski, J., Bard, D., Colard, O. & Bereziat, G. (1974) Comptes Rendues Séances Soc.
Biol. Fil. 168, 1211.
Pugh, E.L. & Kates, M. (1984) Lipids 19, 48.
Quinn, P.J. (1998) In: Aqueous Phase Organometallic Catalysis: Concepts and Applications
(Cornils, B.A.H. and Herrmann, W.A., eds), Wiley-VCH, Weinheim, Germany, p. 418.
Quinn, P.J. (2002) Subcell. Biochem. 36, 39.
Rebecchi, M.J. & Pentyala, S.N. (2000) Physiol. Rev. 80, 1291.
Rhee, S.G. (2001) Annu. Rev. Biochem. 70, 281.
Ritov, V.B., Banni, S., Yalowich, J.C., Day, B.W., Claycamp, H.G., Corongiu, F.P. & Kagan,
V.E. (1996) Biochim. Biophys. Acta 1283, 127.
Rizzo, M.A., Shome, K., Watkins, S.C. & Romero, G. (2000) J. Biol. Chem. 275, 23911.
Roberts, W.L., Santikarn, S., Reinhold, V.N. & Rosenberry, T.L. (1988) J. Biol. Chem. 263,
18776.
Roper, K., Corbeil, D. & Huttner, W.B. (2000) Nature Cell Biol. 2, 582.
Roughan, P.G. (1985) Plant Physiol. 77, 740.
Rouquette-Jazdanian, A.K., Pelassy, C., Breittmayer, J.P., Cousin, J.L. & Aussel, C. (2002)
Biochem. J. 363, 645.
Rutter, A.J., Thomas, K.L., Herbert, D., Henderson, R.J., Lloyd, D. & Harwood, J.L. (2002)
Biochem. J. 368, 57.
Sachs, J.N., Petrache, H.I. & Woolf, T.B. (2003) Chem. Phys. Lipids 126, 211.
Sakyo, T. & Kitagawa, T. (2002) Biochim. Biophys. Acta 1567, 165.
Samet, D. & Barenholz, Y. (1999) Chem. Phys. Lipids 102, 65.
Sato, N. & Murata, N. (1980) Biochim. Biophys. Acta 619, 353.
Scheideler, M.A. & Bell, R.M. (1986) J. Biol. Chem. 261, 10990.
Schroeder, R.J., Ahmed, S.N., Zhu, Y., London, E. & Brown, D.A. (1998) J. Biol. Chem. 273,
1150.
Shvedova, A.A., Tyurina, Y.Y., Tyurin, V.A., Kikuchi, Y., Kagan, V.E. & Quinn, P.J. (2001)
Biosci. Rep. 21, 33.
LIPIDS OF BIOLOGICAL MEMBRANES 217
Shvedova, A.A., Tyurina, J.Y., Kawai, K., Tyurin, V.A., Kommineni, C., Castranova, V.,
Fabisiak, J.P. & Kagan, V.E. (2002) J. Invest. Dermatol. 118, 1008.
Singer, W.D., Brown, H.A. & Sternweis, P.C. (1997) Annu. Rev. Biochem. 66, 475.
Siskind, L.J., Kolesnick, R.N. & Colombini, M. (2002) J. Biol. Chem. 277, 26796.
Slimane, T.A., Trugnan, G., Van, I.S.C. & Hoekstra, D. (2003) Mol. Biol. Cell 14, 611.
Song, C., Satoh, T., Edamatsu, H., Wu, D., Tadano, M., Gao, X. & Kataoka, T. (2002)
Oncogene 21, 8105.
Sonoki, S. & Ikezawa, H. (1976) J. Biochem. (Tokyo) 80, 1233.
Sugars, J.M., Cellek, S., Manifava, M., Coadwell, J. & Ktistakis, N.T. (2002) J. Biol. Chem.
277, 29152.
Sugiura, T., Masuzawa, Y., Nakagawa, Y. & Waku, K. (1987) J. Biol. Chem. 262, 1199.
Swinnen, J.V., Van Veldhoven, P.P., Timmermans, L., De Schrijver, E., Brusselmans, K.,
Vanderhoydonc, F., Van de Sande, T., Heemers, H., Heyns, W. & Verhoeven, G. (2003)
Biochem. Biophys. Res. Commun. 302, 898.
Tasaka, Y., Gombos, Z., Nishiyama, Y., Mohanty, P., Ohba, T., Ohki, K. & Murata, N. (1996)
EMBO J. 15, 6416.
Thomas, K., Rutter, A., Suller, M., Harwood, J. & Lloyd, D. (1998) FEBS Lett. 425, 171.
Tou, J.S. (1981) Biochim. Biophys. Acta 665, 491.
Tsvetkova, N.M., Brain, A.P. & Quinn, P.J. (1994) Biochim. Biophys. Acta 1192, 263.
Tsvetkova, N.M., Apsotovova, E.L., Brain, A.P.R., Williams, W.P. & Quinn, P.J. (1995)
Biochim. Biophys. Acta 1228, 201.
Tyurina, Y.Y., Tyurin, V.A., Carta, G., Quinn, P.J., Schor, N.F. & Kagan, V.E. (1997) Arch.
Biochem. Biophys. 344, 413.
Uemura, M. & Steponkus, P.L. (1997) Plant Physiol. 114, 1493.
Venkataraman, K. & Futerman, A.H. (2000) Trends Cell Biol. 10, 408.
Vieu, C., Terce, F., Chevy, F., Rolland, C., Barbaras, R., Chap, H., Wolf, C., Perret, B. &
Collet, X. (2002) J. Lipid Res. 43, 510.
Vigh, L., Horvath, I. & Thompson, G.A., Jr (1988) Biochim. Biophys. Acta 937, 42.
Wada, H., Gombos, Z. & Murata, N. (1990) Nature 347, 200.
Wang, D., Feng, J., Wen, R., Marine, J.C., Sangster, M.Y., Parganas, E., Hoffmeyer, A.,
Jackson, C.W., Cleveland, J.L., Murray, P.J. & Ihle, J.N. (2000) Immunity 13, 25.
Wang, T.Y. & Silvius, J.R. (2001) Biophys. J. 81, 2762.
Webb, M.S. & Steponkus, P.L. (1993) Plant Physiol. 101, 955.
Wilde, J.I. & Watson, S.P. (2001) Cell Signal. 13, 691.
Wolf, C., Colard-Torquebiau, O., Bereziat, G. & Polonovski, J. (1997) Biochimie 59, 115.
Wolter, F.P., Schmidt, R. & Heinz, E. (1992) EMBO J. 11, 4685.
Wu, J. & Browse, J. (1995) Plant Cell 7, 17.
Xu, X. & London, E. (2000) Biochemistry 39, 843.
Xu, X., Bittman, R., Duportail, G., Heissler, D., Vilcheze, C. & London, E. (2001) J. Biol.
Chem. 276, 33540.
Yang, L., Ding, L. & Huang, H.W. (2003) Biochemistry 42, 6631.
Zubiaur, M., Fernandez, O., Ferrero, E., Salmeron, J., Malissen, B., Malavasi, F. & Sancho,
J. (2002) J. Biol. Chem. 277, 13.
CHAPTER 10
Lipid barriers at the environment–body
interface
A. Introduction
Living organisms exist in a polar air/aqueous environment and the interior of
cells is dominated by a hydrophilic milieu. The interface between the environ-
ment and the interior of cells is created by the plasma membrane. The barrier to
the permeation of solutes in and out of the cell is represented by the hydropho-
bic domain of the lipid bilayer matrix of this membrane. Multicellular organisms
have evolved specialized barrier structures to supplement the function of the
plasma membrane, protect the underlying tissues and organs, and assist them in
their functions. Like the membranes of individual cells, the barrier to the
passage of water and solutes is a hydrophobic domain created by polar lipids
together with specialized proteins which serve to assist in the organization and
maintenance of the lipid structures. While the outer surface skin is dry the inner
surfaces of the body are wet. The internal surfaces consist of hydrophilic
proteoglycans and mucins which form an aqueous, gel-like phase, the mucus.
A surface film of polar lipids is often a component of the mucous zone, with a
significant influence on the properties of the surface. The flow and regenera-
tion of the serbium, representing the hydrophobic surface of the skin, and the
hydrophilic mucus, covering the inner surfaces, both serve to clean and protect
the underlying tissues from harmful microorganisms.
Below the mucous layer lies the endothelial plasma membrane, which is
replenished from within the cell. The transport of polar lipids into the lumen via
the mucous layer is responsible for conferring an amphiphilic character to
mucous surfaces, the turnover of which has often been neglected. The presence
of polar lipids in surface layers produces physical properties aiding tissue
function. The condensed monolayer lining the alveoli of the lung is essential to
prevent collapse of the lung. By contrast, saliva and the oral mucosa do not
have a lipid monolayer covering the surface, and the physical properties
reflected in the surface tension differ markedly from those of lung surfactant
(Sefton et al., 1992). The reduction in surface tension at the air–water interface
by components in the saliva is about 30 mN/m, which is significantly lower
than would be expected if the polar lipids were oriented at the interface.
This chapter will examine the composition and structure of barriers created
in the gastrointestinal tract, in skin, and in the surfactant lining the alveoli of the
lung. The mechanism of regulation and secretion of lipid barriers will also be
considered.
219
220 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
Figure 10.1 Ceramides of human stratum corneum. The particular features of CER1–CER9 are
indicated by the following keys: A, α-hydroxy fatty acid; EO, linoleic acid bound to ω-hydroxy fatty
acid; H, 6-hydroxysphingosine; N, non-hydroxy fatty acid; P, phytosphingosine; S, sphingosine (adapted
from de Jager et al., 2004).
1. Saliva
Normal saliva has relatively low amounts of polar lipids, which are present in
a concentration of about 10 mg/100 ml. The classes of lipids that are found
include sphingolipids, glyceroglucolipids and glycerophospholipids. The ma-
jor secretions from the parotid and submaxilliary glands are, however, rich in
proteins which have strongly surface-active properties and help to form a
protective interfacial film over the entire oral cavity.
Lipids of saliva are an important secretion from the salivary glands and they
are required to maintain the integrity of both soft and hard oral tissues. Elevated
levels of lipids in saliva are associated with a high incidence of dental caries
and the development of plaque, calculus and resulting periodontal disease. The
lipids of saliva have been shown to affect the penetration of hydrophobic
substances across the oral mucosa, alter the interaction of salivary proteins with
calcium, and influence glycosyltransferase activity associated with cariogenic
bacteria. Furthermore the phospholipids of saliva interact heterotypically with
proteins and glycoproteins to form a dynamic continuum that dictates many
physiochemical and biological properties of saliva (Slomiany et al., 1988).
These properties include viscoelasticity, lubricative properties, proteolytic
susceptibility and the formation of protective coatings over the oral mucosa and
tooth enamel (Slomiany et al., 1989).
The factors that regulate the secretion of lipids by the salivary glands are
unclear, although secretion in general from the glands is effected mainly
through neural, paracrine and endocrine systems. Stimulation via the
parasympathetic nerve results mainly in an increase in salivary fluid secretion,
while stimulation via sympathetic nerves affects the secretion of macromolecular
224 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
Mucus forms a lining along the entire intestinal wall. The small intestine is
particularly efficient at digesting food and absorbing the nutrients thereby
released. Water absorption is a major function of the colon. Common features
of this region of the mucous layer are the presence of microorganisms and the
location of important components of the immune system. T-cell lymphocytes in
the epithelium cells control reactions against microorganisms and food compo-
nents, and this region acts as the mammalian equivalent of the bursa in birds,
where B-cells are differentiated.
The low pH of the stomach provides an effective protection against poten-
tially pathogenic microorganisms. When food enters the duodenum the pH is
normalized and antimicrobial protection is afforded by the resident microbial
flora which inhabits the mucous layer (Strompfova et al., 2004). Lactobacilli
are known to inhibit the growth of potentially pathogenic bacteria (Johansson
et al., 1993) and there is evidence that they may stimulate an immune response
(Fuller, 1991). For this reason they are sometimes described as probiotics.
Their packing density increases progressively down the alimentary tract from
about 106 to 108 colony-forming units per millilitre of mucus from one end to
the other. Colonization with different Lactobacillus strains has been followed
in vivo and a corresponding decrease of Gram-negative bacteria was observed
(Johansson et al., 1993). Helicobacter pylori are among the bacteria that are
associated with gastrointestinal inflammatory diseases and ulcer formation
(Suerbaum & Michetti, 2002). Their mode of action is to neutralize the low pH
of gastric secretions by extracellular urease activity. The bacteria move through
the mucous layer and adhere to glycerolipid receptors in the gastric wall with
the aid of flagella (Lingwood et al., 1989). It is interesting to note that
colonization of the intestinal wall by lactobacilli appears to be host specific, so
the particular strain that colonizes the mucosa of one animal species will not
normally be found in the epithelium of another.
The surface area of the intestine is manyfold greater than the surface area of
the skin. To put the metabolic demands on the maintenance of the intestinal
epithelium into some perspective, the cell membranes are renewed every two
days. In consequence the transfer of the lipids of the cell membranes into the
mucous layer is considerable and the presence of these lipids imparts a non-
wettable character to the surface. It is possible that this hydrophobic character
is due to accumulation of phospholipid within a phase-separated domain in the
mucous layer, as suggested above for the gastric mucosa.
The lipid spectrum in the intestinal mucosa is highly complex. In addition to
the lipids of the constituent membranes of the epithelial cells, bile containing
phospholipids, cholesterol and bile acids, as well as food and the degradation
products produced by pancreatic enzymes, are present. The surface of the brush
border of the small intestine is covered by a diffuse layer of carbohydrate
LIPID BARRIERS AT THE ENVIRONMENT–BODY INTERFACE 227
that one function that these phospholipids may perform is to repel water and
lubricate the tissue (Ziegler et al., 1989). Prolonged dialysis has been shown to
result in a decrease in the phospholipid content, and the high risk of adhesions
following dialysis has been attributed to this decrease, which is believed to be
related to loss of phospholipids from the surface of mesothelial cells (Ar’Rajab
et al., 1991). This is consistent with experimental evidence showing that
intraperitoneal administration of phosphatidylcholine could prevent tissue
adhesions in rats.
E. Lung surfactant
The alveoli of the lung are lined with a surfactant that is essential for normal
pulmonary function. The surfactant consists of a few specialized proteins and
polar lipids which are synthesized and secreted by type II pneumocytes and
which then adsorb to the air–water interface in the alveoli (Piknova et al.,
2002). The process of synthesis and secretion of pulmonary surfactant onto the
alveolar surface is illustrated schematically in Figure 10.2.
The surfactant is required to function immediately at birth, when the air–
water interface expands dramatically as the amniotic fluid is expelled, and to
remain in place thereafter during cycles of compression and expansion in the
respiratory cycle. The remarkable feature of the surfactant is that it is able to
form a film that reduces surface tension at the alveolar air–liquid interface
(Goerke, 1998). In this way it stabilizes the small air spaces of the lung during
exhalation when compression of the structure reaches surface pressures well
above that where films would be expected to collapse. This property of
pulmonary surfactant is believed to reside in the lipid constituents rather than
the proteins that are found associated with the surfactant.
In addition to its primary role in preserving the structure of the functioning
lung, pulmonary surfactant represents the interface between the environment
and the underlying body tissues. The surfactant is known to contain agents that
participate in the innate immune response to microbial infection and to act in
other aspects of immune and inflammatory processes within the lung (Crouch
& Wright, 2001; Vayrynen et al., 2004).
Type I
pneumocyte
Air space
Air-water
interface
Adsorption Type II
pneumocyte
Air Lamellar
body
Fluid
Figure 10.2 Schematic of pulmonary surfactant in an alveolus of the lung. Type II pneumocytes
synthesize the surfactant constituents and package them into the concentric bilayers of lamellar bodies.
After secretion, the bilayers unravel and adsorb to the air–water interface (see inset). During exhalation,
the decreasing alveolar surface area compresses the interfacial film. The arrangement of the surfactant
lipid underlying the surface monolayer has not yet been clearly established (adapted from Piknova et al.,
2002).
Lipid % by weight
Dipalmitoylphosphatidylcholine 34.3
Unsaturated phosphatidylcholinesa 33.6
1-palmitoyl-2-myristoylphosphatidylcholine 7.1
Phosphatidylglycerol 9.6
Choline plasmalogen 6.8
Phosphatidylethanolamine 2.4
Phosphatidylinositol 1.0
Lyso-bis-phosphatidic acid 0.9
Sphingomyelin 1.6
Cholesterol 2.6
a
Mainly 1-palmitoyl-2-palmitoleoyl and 1,2-palmitoleoyl molecular species (adapted from Veldhuizen,
2004).
ultrastructural study to identify the location of fatty acid synthase and CTP-
phosphocholine cytidylyltransferase within type II cells showed that the
glycogen granules were populated by lipid-synthesizing enzymes as well as
surfactant proteins and surfactant storage organelles (Ridsdale & Post, 2004).
It was suggested that this location would ensure that the supply of substrate for
phospholipid biosynthesis would not be rate limiting at a time of maximum
surfactant production.
The surfactant produced within type II pneumocytes is accumulated in the
cytoplasm in the form of lamellar bodies. These structures are lysosome-
derived organelles and are characterized by a low luminal pH and a bounding
membrane containing lysosome-associated membrane proteins. Ultra-
structural studies have shown that multivesicular structures containing sur-
factant components fuse to form so-called composite bodies and ultimately
mature into lamellar bodies (Stahlman et al., 2000). During this process the
complete complement of surfactant components is assembled from their re-
spective sites of synthesis or post-translational modification in the Golgi
apparatus. The formation of lamellar bodies is dependent on the expression of
the surfactant-associated protein SP-B (Foster et al., 2003).
Lamellar bodies are secreted via the apical membrane of type II cells. The
factors that control secretion are conjectural because conflicting evidence has
been obtained from studies performed in vivo and those carried out in vitro.
Nevertheless, the process of secretion of lamellar bodies from type II
pneumocytes appears to be similar to that of other cell types in that it is
regulated by secretagogues (Rooney, 2001) and targeted to the plasma mem-
brane by syntaxin 2 and SNAP-23 (Abonyo et al., 2004).
232 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
Figure 10.3 An electron micrograph of tubular myelin structures from large aggregates of lung
surfactant isolated from rat lung lavage. Scale bar: 500 nm.
(A)
Figure 10.4 (A) Synchrotron X-ray diffraction pattern recorded from large aggregates of rat lung
surfactant isolated from lavage preparations. (B) Relationship between the d-spacing (S) and scattering
intensity of the sample shown in (A). The inset shows that the diffraction maxima fit a straight-line linear
regression giving the expression y = 33.5x – 0.0926.
234 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
References
Abonyo, B.O., Gou, D., Wang, P., Narasaraju, T., Wang, Z. & Liu, L. (2004) Biochemistry
43, 3499.
Araneo, B.A., Cebra, J.J., Beuth, J., Fuller, R., Heidt, P.J., Midvedt, T., Nord, C.E.,
Nieuwenhuis, P., Manson, W.L., Pulverer, G., Rusch, V.C., Tanaka, R., van der Waaij,
D., Walker, R.I. & Wells, C.L. (1996) Zentralbl. Bakteriol. 283, 431.
Ar’Rajab, A., Ahren, B., Rozga, J. & Bengmark, S. (1991) J. Surg. Res. 50, 212.
Ballard, P.L., Merrill, J.D., Godinez, R.I., Godinez, M.H., Truog, W.E. & Ballard, R.A.,
(2003) Am. J. Respir. Crit. Care Med. 168, 1123.
Batenburg, J.J. & Haagsman, H.P. (1998) Prog. Lipid Res. 37, 235.
Bernhard, W., Pynn, C.J., Jaworski, A., Rau, G.A., Hohlfeld, J.M., Freihorst, J., Poets, C.F.,
Stoll, D. & Postle, A.D. (2004) Am. J. Respir. Crit. Care Med. 170, 54.
LIPID BARRIERS AT THE ENVIRONMENT–BODY INTERFACE 235
Bouwstra, J.A., Gooris, G.S., van der Spek, J.A. & Bras, W. (1991) J. Invest. Dermatol. 97,
1005.
Bouwstra, J.A., Gooris, G.S., Bras, W. & Downing, D.T. (1995) J. Lipid Res. 36, 685.
Bouwstra, J.A., Gooris, G.S., Dubbelaar, F.E., Weerheim, A.M. & Ponec, M. (1998) J.
Investig. Dermatol. Symp. Proc. 3, 69.
Bouwstra, J.A., Gooris, G.S., Dubbelaar, F.E. & Ponec, M. (2002) J. Invest. Dermatol. 118,
606.
Bukholm, G., Tannaes, T., Nedenskov, P., Esbensen, Y., Grav, H.J., Hovig, T., Ariansen, S.
& Guldvog, I. (1997) Scand. J. Gastroenterol. 32, 445.
Cao, J., Hawkins, E., Brozinick, J., Liu, X., Zhang, H., Burn, P. & Shi, Y. (2004) J. Biol.
Chem. 279, 18878.
Chen, H., Mendelsohn, R., Rerek, M.E. & Moore, D.J. (2000) Biochim. Biophys. Acta 1468,
293.
Chen, Q., Sternby, B. & Nilsson, A. (1989) Biochim. Biophys. Acta 1004, 372.
Cockshutt, A.M., Weitz, J. & Possmayer, F. (1990) Biochemistry 29, 8424.
Cook, D.I., Van Lennep, E.W., Roberts, M.L. & Young, J.A. (1994) In: Physiology of the
Gastrointestinal Tract (Johnson, L.R., ed.), Raven Press, New York, USA, p.1061.
Corfield, A.P., Wagner, S.A., Clamp, J.R., Kriaris, M.S. & Hoskins, L.C. (1992) Infect.
Immun. 60, 3971.
Crouch, E. & Wright, J.R. (2001) Annu. Rev. Physiol. 63, 521.
de Jager, M.W., Gooris, G.S., Dolbnya, I.P., Bras, W., Ponec, M. & Bouwstra, J.A. (2004)
J. Lipid Res. 45, 923.
Demirbilek, S., Gurses, I., Sezgin, N., Karaman, A. & Gurbuz, N. (2004) J. Pediatr. Surg. 39,
57.
Duane, W.C., McHale, A.P. & Sievert, C.E. (1986) Am. J. Physiol. 250, G275.
Elias, P.M. (1983) J. Invest. Dermatol. 80(Suppl.), 44S.
Fabia, R., Ar’Rajab, A., Willen, R., Andersson, R., Ahren, B., Larsson, K. & Bengmark, S.
(1992) Digestion 53, 35.
Fenske, D.B., Thewalt, J.L., Bloom, M. & Kitson, N. (1994) Biophys. J. 67, 1562.
Foster, C.D., Zhang, P.X., Gonzales, L.W. & Guttentag, S.H. (2003) Am. J. Respir. Cell Mol.
Biol. 29, 259.
Fuller, R. (1991) Gut 32, 439.
Goerke, J. (1998) Biochim. Biophys. Acta 1408, 79.
Goggin, P.M., Marrero, J.M., Spychal, R.T., Jackson, P.A., Corbishley, C.M. & Northfield,
T.C. (1992) Gastroenterology 103, 1486.
Haitsma, J.J., Papadakos, P.J. & Lachmann, B. (2004) Curr. Opin. Crit. Care 10, 18.
Hills, B.A. (1990) Med. J. Aust. 153, 76.
Holm, B.A., Venkitaraman, A.R., Enhorning, G. & Notter, R.H. (1990) Chem. Phys. Lipids
52, 243.
Holmskov, U., Thiel, S. & Jensenius, J.C. (2003) Annu. Rev. Immunol. 21, 547.
Jensen, J.M., Folster-Holst, R., Baranowsky, A., Schunck, M., Winoto-Morbach, S., Neumann,
C., Schutze, S. & Proksch, E. (2004) J. Invest. Dermatol. 122, 1423.
Johansson, M.L., Molin, G., Jeppsson, B., Nobaek, S., Ahrne, S. & Bengmark, S. (1993) Appl.
Environ. Microbiol. 59, 15.
Kitson, N., Thewalt, J., Lafleur, M. & Bloom, M. (1994) Biochemistry 33, 6707.
Kuroki, Y. & Voelker, D.R. (1994) J. Biol. Chem. 269, 25943.
Larsson, M., Larsson, K., Andersson, S., Kakhar, J., Nylander, T., Ninham, B. & Wollmer,
P. (1999) J. Dispers. Sci. Technol. 20, 1.
Larsson, M., Larsson, K., Nylander, T. & Wollmer, P. (2003a) Eur. Biophys. J. 31, 633.
Larsson, M., van Iwaarden, J.F., Haitsma, J.J., Lachmann, B. & Wollmer, P. (2003b) Clin.
Physiol. Funct. Imaging 23, 199.
236 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
Lingwood, C.A., Law, H., Pellizzari, A., Sherman, P. & Drumm, B. (1989) Lancet 2, 238.
Lupton, J.R. (2004) J. Nutr. 134, 479.
McIntosh, T.J., Stewart, M.E. & Downing, D.T. (1996) Biochemistry 35, 3649.
McIntosh, T.J. (2003) Biophys. J. 85, 1675.
McIntosh, T.J. (2004) Chem. Phys. Lipids 130, 83.
Montagne, L., Piel, C. & Lalles, J.P. (2004) Nutr. Rev. 62, 105.
Moore, D.J., Rerek, M.E. & Mendelsohn, R. (1997) Biochem. Biophys. Res. Commun. 231,
797.
Moore, D.J. & Rerek, M.E. (2000) Acta Derm. Venereol. Suppl. (Stockh.) 208, 16.
Moreau, H., Bernadac, A., Tretout, N., Gargouri, Y., Ferrato, F. & Verger, R. (1990) Eur. J.
Cell Biol. 51, 165.
Morrow, M.R., Stewart, J., Taneva, S., Dico, A. & Keough, K.M. (2004) Eur. Biophys. J. 33,
285.
Nag, K., Munro, J.G., Hearn, S.A., Rasmusson, J., Petersen, N.O. & Possmayer, F. (1999)
J. Struct. Biol. 126, 1.
Neubert, R., Rettig, W., Wartewig, S., Wegener, M. & Wienhold, A. (1997) Chem. Phys.
Lipids 89, 3.
Ohta, N. & Hatta, I. (2002) Chem. Phys. Lipids 115, 93.
Orgeig, S. & Daniels, C.B. (2001) Comp. Biochem. Physiol. A Mol. Integr. Physiol. 129, 75.
Osada, M., Gniadecka, M. & Wulf, H.C. (2004) Exp. Dermatol. 13, 391.
Percot, A. & Lafleur, M. (2001) Biophys. J. 81, 2144.
Piknova, B., Schram, V. & Hall, S.B. (2002) Curr. Opin. Struct. Biol. 12, 487.
Ponec, M., Weerheim, A., Lankhorst, P. & Wertz, P. (2003) J. Invest. Dermatol. 120, 581.
Rau, G.A., Vieten, G., Haitsma, J.J., Freihorst, J., Poets, C., Ure, B.M. & Bernhard, W. (2004)
Am. J. Respir. Cell Mol. Biol. 30, 694.
Raudenkolb, S., Hubner, W., Rettig, W., Wartewig, S. & Neubert, R.H. (2003) Chem. Phys.
Lipids 123, 9.
Ridsdale, R. & Post, M. (2004) Am. J. Physiol. Lung Cell Mol. Physiol. 287, L743.
Robson, K.J., Stewart, M.E., Michelsen, S., Lazo, N.D. & Downing, D.T. (1994) J. Lipid Res.
35, 2060.
Rodriguez Capote, K., McCormack, F.X. & Possmayer, F. (2003) J. Biol. Chem. 278, 20461.
Rodriguez-Capote, K., Nag, K., Schurch, S. & Possmayer, F. (2001) Am. J. Physiol. Lung
Cell Mol. Physiol. 281, L231.
Rooney, S.A. (2001) Comp. Biochem. Physiol. A Mol. Integr. Physiol. 129, 233.
Schram, V., Anyan, W.R. & Hall, S.B. (2003) Biochim. Biophys. Acta 1616, 165.
Schram, V. & Hall, S.B. (2004) Biophys. J. 86, 3734.
Schultz, M.J. & Kesecioglu, J. (2004) Curr. Drug Targets 5, 445.
Sefton, J., Arnebrant, T. & Glantz, P.O. (1992) Acta Odontol. Scand. 50, 221.
Slomiany, B.L., Murty, V.L., Sarosiek, J., Piotrowski, J. & Slomiany, A. (1988) Biochem.
Biophys. Res. Commun. 151, 1046.
Slomiany, B.L., Murty, V.L., Mandel, I.D., Zalesna, G. & Slomiany, A. (1989) Arch. Oral
Biol. 34, 229.
Slomiany, B.L., Sengupta, S., Fekete, Z., Murty, V.L. & Slomiany, A. (1991) Gen.
Pharmacol. 22, 969.
Slomiany, B.L. & Slomiany, A. (2004) Biochem. Biophys. Res. Commun. 318, 247.
Stahlman, M.T., Gray, M.P., Falconieri, M.W., Whitsett, J.A. & Weaver, T.E. (2000) Lab.
Invest. 80, 395.
Stewart, M.E. & Downing, D.T. (1999) J. Lipid Res. 40, 1434.
Strompfova, V., Laukova, A. & Ouwehand, A.C. (2004) Folia Microbiol. (Praha) 49, 203.
Suerbaum, S. & Michetti, P. (2002) N. Engl. J. Med. 347, 1175.
Takagi, Y., Nakagawa, H., Matsuo, N., Nomura, T., Takizawa, M. & Imokawa, G. (2004)
J. Invest. Dermatol. 122, 722.
LIPID BARRIERS AT THE ENVIRONMENT–BODY INTERFACE 237
Uchida, Y., Hara, M., Nishio, H., Sidransky, E., Inoue, S., Otsuka, F., Suzuki, A., Elias, P.M.,
Holleran, W.M. & Hamanaka, S. (2000) J. Lipid Res. 41, 2071.
Vayrynen, O., Glumoff, V. & Hallman, M. (2004) Pediatr. Res. 55, 55.
Veldhuizen, R. & Possmayer, F. (2004) In: Subcellular Biochemistry. Membrane Dynamics
and Domains (Quinn, P.J., ed.), Springer, London, UK, p.359.
Venkitaraman, A.R., Hall, S.B. & Notter, R.H. (1990) Chem. Phys. Lipids 53, 157.
Walters, R.W., Jenq, R.R. & Hall, S.B. (2000) Biophys. J. 78, 257.
Wegener, M., Neubert, R., Rettig, W. & Wartewig, S. (1997) Chem. Phys. Lipids 88, 73.
Wertz, P.W., Miethke, M.C., Long, S.A., Strauss, J.S. & Downing, D.T. (1985) J. Invest.
Dermatol. 84, 410.
Wu, H., Kuzmenko, A., Wan, S., Schaffer, L., Weiss, A., Fisher, J.H., Kim, K.S. &
McCormack, F.X. (2003) J. Clin. Invest. 111, 1589.
Ziegler, C., Torchia, M., Grahame, G.R. & Ferguson, I.A. (1989) Perit. Dial. Int. 9, 47.
CHAPTER 11
Drug delivery
Most drugs are solids, which can be taken orally. When drugs are formulated
into pills from powders, lipids are used as solid lubricants. Lipids can also play
a direct role in improving drug efficacy, safety, and patient convenience,
thereby making it possible to introduce new and improved therapies. This has
stimulated the creative design and study of lipid self-assembly systems for
delivery through different portals of the body.
Drug substances are frequently amphiphilic and sometimes they have low
aqueous solubility. Lipid self-assembly structures present a wide spectrum of
different drug delivery systems in which amphiphilic and lipophilic drugs can
be solubilized. Lipid–water phases also have the obvious advantage of being
able to encapsulate water-soluble drugs by solubilization in aqueous compart-
ments. In this way sustained release can be achieved, as well as protection
against digestive enzymes. A further important aspect where lipid systems can
have a direct influence is permeability through biomembranes, such as the
surface lining of the gastrointestinal tract. Many lipid systems are colloidal in
nature, for which there are well-known pathways in the form of lymphoid
follicles, Peyer’s patches of the GALT (gut-associated lymphoid tissue), and
normal enterocytes. Our natural ‘delivery system’ operates partly by lipids in
synergistic interplay with various proteins and enzymes.
A wide variety of lipid self-assembly structures with specific structural and
functional attributes can be prepared by exploiting the self-assembly properties
unique to different lipids. These structures may therefore vary with regards to
both lipid composition and morphology, and include well-known colloidal
structures like micelles, emulsions, different types of liposomes, and more
sophisticated self-assembly structures such as the reverse liquid crystals and
their colloidal dispersions. The word lipid is generously used in this text and
also includes other surface-active agents (surfactants).
The spectrum of properties and functions represented by lipid self-assembly
objects appears almost unlimited, but is in practice restricted by cost, safety and
regulatory demands. In reality there is only a limited number of approved lipids
and surfactants with prior use in pharmaceutical products. Therefore, introduc-
tions of new excipients should greatly stimulate the design and use of novel
drug delivery systems based on self-assembled lipids, and lead to new and
improved therapies in the future.
Many drugs on the market and in different development stages have
suboptimal properties which potentially could be improved by the use of a
suitable delivery system; see Table 11.1.
239
240 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
There are also a large number of promising candidates which are currently
unavailable because a suitable means of delivery is lacking. As more water-
insoluble drugs are discovered through the use of modern screening techniques,
the need for effective and flexible drug delivery systems will increase in both
preclinical and clinical development stages. The large number of emerging
biotech drugs also face delivery problems that potentially can be solved by
using lipid-based drug delivery systems. This text focuses on colloidal self-
assembly structures, as they are particularly useful in many practical
applications. For controlled release purposes, however, bulk-phase structures
are often preferred due to their improved retention ability.
also used for solubility management in, for instance, Taxotere® (docetaxel;
Aventis).
As early as 1869 it was reported that fat could be given subcutaneously to
malnourished patients (Menzel & Perco, 1869). Stabilized emulsion droplets
have been used as vehicles for parenteral nutrition for decades. The emulsion
is injected intravenously, and the particles must be smaller than ~1 μm in size
to avoid emboli (blockage) of fine capillaries. In the circulation these particles
adsorb lipoproteins and become similar to chylomicrons (Zilversmit, 1971).
Lipophilic drugs such as diazepam can be solubilized in such oil droplets and
delivered this way (von Dardel et al., 1976). The most well-known parenteral
emulsion product on the market is probably Diprivan® (propofol; AstraZeneca),
which is presented as a ready-made o/w emulsion based on soybean oil and egg
lecithin as stabilizer. Water-in-oil emulsions have been used for intramuscular
injection in order to form depots for the controlled release of anti-cancer drugs
(Hashida et al., 1977). As early as 30 years ago, multiple emulsions (w/o/w)
were described for the delivery of hydrophilic anti-cancer drugs such as
methotrexate (Elson et al., 1970).
Drugs for transdermal administration can be dissolved in the oil or water
phases of an emulsion. The o/w emulsions have better cosmetic properties,
whereas the w/o emulsions often provide better uptake. The reason is the so-
called occlusion effect: as the stratum corneum layer becomes more hydrated,
its barrier function is reduced. Emulsions are the most commonly used carriers
for topical drug delivery, and are used in numerous marketed products.
3. Liposomes
Drugs encapsulated in liposomes may circulate in the bloodstream for extended
periods compared to the same drugs in a non-liposomal form, thereby extend-
ing the treatment effect and simplifying the dosing regimen for physicians and
patients. In some cases, liposomal drugs have shown improved accumulation at
the tumour or infection site, thus delivering higher concentrations of the drug
to the disease target. The most important application of liposomes is in
infectious disease and cancer therapy (Lasic, 1993).
244 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
these sites (Ishida et al., 2002). The first drug using the PEG-stabilized
liposome approach is Doxil®, an anti-cancer drug for the treatment of refractory
ovarian cancer and AIDS-related Kaposi’s sarcoma. In retrospect it is some-
what surprising that the well-known steric stabilization approach was not
explored at an earlier stage in the pharmaceutical development of liposome
products.
Aside from passive targeting, liposome structures can be equipped with
‘homing devices’ such as surface-tethered antibodies. However, such struc-
tures become increasingly complex. It is noteworthy that even ‘simple’ things
are difficult in pharmaceutical development, and that such complex systems
will require substantial potential rewards in terms of therapeutic index and
market potential to even be considered for drug development.
Utilization of liposomes in administration routes other than intravenous is so
far limited because of the complexity of liposome formulations and the
existence of better and simpler alternatives. However, there are a number of
liposomal drug delivery systems on the market intended for the subcutaneous
or intramuscular routes. One example of a depot technology based on liposomes
is DepoFoam® from SkyePharma (Mantripragada, 2002). This technology
exploits foam structures based on liposome aggregates that are 10–30 μm in
size, and has been shown to provide a sustained release of some drugs for up to
1 month.
The technology is used in the marketed product Depocyt® (cytarabine;
SkyePharma) for the treatment of neoplastic meningitis by intrathecal delivery,
and is amenable to clinical product development. The formulation, which is
administered once every 2 weeks, has overcome the limitations of current
therapies for the disease; these require 2–3 doses per week by means of lumbar
puncture, and the rapid clearance of the conventional formulations results in
inadequate distribution. The controlled-release formulation has significantly
improved the pharmacokinetic profile of the active drug substance.
Another current interesting clinical development concerns transdermal
liposome technology (Cevc & Blume, 1992). Such systems are in clinical
development, e.g. Transfersomes® with ketoprofen developed by IDEA, claim-
ing a significantly improved bioavailability.
phase I trial of its use for nasal immunization with diphtheria vaccine has
recently been performed (Haile et al., 2004).
An example of a colloidal self-assembly structure not belonging to the above
mentioned classes of phase structures is the ‘ribbon-like’ amphotericin B
Abelcet™ formulation. This is another example that clearly illustrates the kind
of benefits that can be achieved by incorporating drugs into novel lipid self-
assembly objects.
B. Liquid crystals
Drugs often have an optimum therapeutic concentration range, above which
they are toxic and below which they are essentially ineffective. In formulations
where there is no controlled release function, the in vivo concentration of the
drug can frequently build to a level which is supra-optimum and then rapidly
decrease to a sub-optimum level. The objective of a sustained release formula-
tion is to deliver the drug within the therapeutic concentration range for a
prolonged period of time. One way to obtain sustained release is to incorporate
the drug in a porous matrix, such as those characteristic of liquid crystalline
materials. It has been recognized for several decades that lyotropic liquid
crystalline phases formed by lipids and surfactants can provide suitable matri-
ces for the sustained release of different kinds of drugs (Engström et al., 1992).
Developments regarding the use of liquid crystalline phases in drug delivery
applications are summarized in two recent reviews (Drummond & Fong, 1999;
Shah et al., 2001).
The release profile of a drug from the matrix is partly determined by drug-related
factors such as the diffusion constant, solubility, and partition coefficient, and
partly by factors related to the delivery system, such as porosity, tortuosity, and
pore connectivity. For liquid crystalline systems the release rate can be regulated
by tuning of the phase structure. It is for instance intuitive that the release of an
aqueous-soluble drug will be much slower in a reverse micellar cubic phase (I2),
where aqueous domains are separated by lipid bilayers, than in a bicontinuous
cubic phase (Q2), in which water channels form a continuous network.
In cases where the encapsulation is effective, drug release becomes deter-
mined by the rate of biodegradation of the liquid crystalline matrix, which in
the case of lipids often involves lipases.
The first marketed product using the special features of liquid crystals for
controlled release purposes is Elyzol® dental gel (Colgate Oral Pharmaceuti-
cals). The formulation is based on a mixture of monoacylglycerols and
triacylglycerols presented in the form of a water-free suspension that trans-
forms to a controlled release matrix of reversed hexagonal phase when injected
into the periodontal pocket (Norling et al., 1992). The phase transition is
triggered by the absorption of water, which also results in the bioadhesive
properties of the lipid matrix.
248 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
C. Summary outlook
The past decades have witnessed increased momentum in the research and
development of lipid self-assembly objects as delivery vehicles for various
types of therapeutic agents. These can facilitate the solubilization, stabilization,
controlled release, and transport of a wide range of drug substances that would
otherwise pose problems. The majority of lipid-based drug delivery products
are micellar, microemulsion, and emulsion pre-concentrates or stable disper-
sions, used primarily to address solubility issues. Several liposomal products
have emerged on the market in recent years which typically improve the
therapeutic index (decrease toxicity) by decreasing local exposure to free drug
and facilitating more favourable biodistribution profiles. Non-lamellar particle
structures appear from many aspects ideal for a range of drug delivery applica-
tions. As key challenges related to manufacturing, stability, and reproducibility
have now finally been overcome, these systems provide new opportunities for
optimizing drug product performance. Non-lamellar lipid-based liquid crystal-
line nanoparticles (LCNP) are currently being implemented as a promising
approach in a number of products under development. The need for simple
controlled release technologies enabling, for instance, less frequent injections
is another issue where lipid-based systems can add substantial value, in
DRUG DELIVERY 249
References
Allen, T.M. & Hansen, C. (1991) Biochim. Biophys. Acta 1068, 133–141.
Cevc, G. & Blume, G. (1992) Biochim. Biophys. Acta 1104, 226–232.
Chung, H., Kim, J., Um, J.Y., Kwon, I.C. & Jeong, S.Y. (2002) Diabetologia 45, 448–451.
Drummond, C.J. & Fong, C. (1999) Curr. Opin. Colloid Interface Sci. 4, 449–456.
Elson, L.A., Mitchley, B.C.V., Collings, A.J. & Schneider, R. (1970) Eur. J. Clin. Biochem.
15, 87.
Engström, S., Larsson, K. & Lindman, B. (1992) US Patent US 5 753 259.
Glantz, P. & Larsson, K. (1982) J. Dispersion Sci. Technol. 3, 573.
Glantz, P., Larsson, K. & Nyquist, G. (1976) Odont. Rev. 27, 265.
Haile, M. et al. (2004) Vaccine 22, 1498–1508.
Hashida, M., Takahashi, Y., Muranishi, S. & Sezaki, H. (1977) J. Pharmacokin. Biopharm.
5, 225.
Ishida, T., Harashima, H. & Kiwada, H. (2002) Biosci. Rep. 22, 197–224.
Klibanov, A.L., Maruyama, K., Torchilin, V.P. & Huang, L. (1990) FEBS Lett. 268, 235–237.
Kovarik, J.M., Mueller, E.A., van Bree, J.B., Tetzloff, W. & Kutz, K. (1994) J. Pharm. Sci.
83(3), 444–446.
Larsson, K. (1968) In: Surface Active Lipids in Foods, Society of Chemical Industry, London,
UK, p.8.
Larsson, K., Björnberg, A. & Hellgren, L. (1975) Opusc. Med. 4, 162.
Lasic, D.D. (1993) Liposomes: From Physics to Applications, Elsevier Science Publishers,
New York, USA.
Lasic, D.D., Martin, F.J., Gabizon, A., Huang, S.K. & Papahadjopoulos, D. (1991) Biochim.
Biophys. Acta 1070, 187–192.
Mantripragada, S. (2002) Prog. Lipid Res. 41, 392–406.
Menzel, A. & Perco, H. (1869) Wien. Med. Wschr. 19, 517.
Mueller, E.A. et al. (1994a) Pharmaceutical 11, 151–155.
Mueller, E.A. et al. (1994b) Pharmaceutical 11, 301–304.
Norling, T. et al. (1992) J. Clin. Periodontol. 19, 687–692.
Shah, J.C., Sadhale, Y. & Chilukuri, D.M. (2001) Adv. Drug Delivery Rev. 47, 229–250.
Strickley, R.G. (2004) Pharm. Res. 21, 201–230.
Thuresson, K., Tiberg, F., Johansson, M., Harwigsson, I., Joabsson, F. & Johnsson, J. (2005)
Patent Application PCT GB05/002217.
von Dardel, O., Mebius, C. & Mossberg, T. (1976) Acta Anaesthesiologica Scand. 20(3),
221–224.
Zilversmit, D.B. (1971) J. Biol. Chem. 249, 2645.
CHAPTER 12
Foods
We will start by considering a typical food product: sausage. The structure and
biophysical properties of the muscle cell and the fat cell, the biological tissues
from which sausage is prepared, have been characterized in detail. In the
transformation of these raw materials into a food, the biological structure is
destroyed, membrane-bound enzymes are released, and small molecules (re-
lated to flavour) are redistributed between different compartments. We are,
then, far from a molecular understanding of the roles of different components
in the product. Only on the colloidal level can the new structure be described.
It is an emulsion of fat and a dispersion of fat cells and connective tissue in a
continuous aqueous gel phase, which is formed by the homogenized muscle
proteins.
This example is given as an illustration of the knowledge gap between
biology and degraded biological tissues, which are recombined into food
products. It also demonstrates that the fundamental description of foods on the
colloidal level requires knowledge of the structure of interfaces and the forces
between colloidal structural elements. Lipids are usually the most surface-
active components of a food system, and they are therefore of great importance
in food processing.
As most of the applications of lipid functionality discussed below are closely
linked to the fundamental properties described in Chapters 1–8 with references,
only a list of recommended reading will be given at the end of this chapter.
A. Starch–lipid interaction
Starch consists of two glucose polymers: amylose, which is linear with α-(1,4)
links; and amylopectin, which contains the same linear segments in a branched
structure. The branches are attached by an α-(1,6) glucoside bond, and the
structure is indicated in Figure 12.1. Amylopectin is considered to be one of the
largest molecules in nature. A common ratio of amylose to amylopectin is 1:3.
Cereals, tubers, and many other plants store energy as starch granules. These
are concentric structures with alternating semi-crystalline layers and amor-
phous layers – a remarkable structure, particularly as amylopectin molecules
mainly form the ordered layers, whereas most of the amylose is localized in the
disordered layers.
One of the most important reactions in food processing is starch gelatiniza-
tion. Less than 1% of potato starch in water, for example, is enough to give a
251
252 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
Figure 12.1 Illustration of the amylopectin branch structure and the location of the three types of
chains (A, B, and C) (after Eliasson et al., 1987).
Figure 12.2 Proposed structure of the amylose–monoacylglycerol complex. The amylose backbone is
shown with the monoacylglycerol acyl chain indicated by the hydrogen atoms (after Carlsson et al.,
1979).
carbon chain in an extended conformation. The polar head group is too large to
be included.
Amylopectin molecules do not interact as strongly with lipids as amylose.
The outer A-chains (see Figure 12.1), however, can also form a similar
inclusion complex. The effects of lipids on starch in foods can be explained by
these two mechanisms.
The presence of monoacylglycerol molecules at the surface of starch gran-
ules, when gelatinization starts, will result in precipitation of amylose–lipid
complex on the surface. This will hinder the diffusion of amylose molecules out
from the granule. The gel volume will be smaller as water transport inwards is
also reduced. Furthermore the gelatinization temperature may shift upwards
due to the barrier effect of the complex on the surface.
One application of amylose–lipid complex formation is to reduce the sticki-
ness of industrial potato powder products and pasta products. Amylose
molecules in the water phase outside the swollen granules will result in a sticky
consistency.
An important application of the amylopectin complexing reaction with lipids
254 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
1. Milk
Milk secreted from the mammary glands is a complex biological fluid contain-
ing growth factors, antibodies and enzymes with significance far beyond
nutrition. Here we will only focus on bovine milk as a colloidal system,
involving the transformations into various foods. Enzymes are important in
these changes. The proteolytic enzymes, for example, are responsible for the
degradation of caseins to form the cheese matrix, and lipase activity contributes
to the development of flavour.
The milk fat globules, which amount to about 4% (w/w) of milk, have a broad
size distribution with a predominating diameter range of 1–10 μm. The milk fat
globule membrane is illustrated in Figure 12.3.
When the fat globule is secreted, it brings the cell membrane from the gland
as a coat with an important stabilizing function. A protein-rich layer dominated
by xanthine oxidase is located between the globule surface and the membrane
bilayer. The triacylglycerols consist of ‘ordinary’ fatty acids (mainly oleic acid
and palmitic acid) provided via the blood, and short-chain fatty acids (such as
butyric acid) formed in the mammary gland. The short acyl chains are located
FOODS 255
Figure 12.3 Schematic illustration of the milk fat globule membrane. Phospholipid (with polar head
groups denoted by open circles) and cholesterol molecules (with polar groups shown by filled circles) at
the globule surface and in the outside bilayer are indicated. The aqueous protein layer is indicated with
PL.
in the sn-3 position, and the corresponding crystals tend to form triple chain-
length structures. Crystallization of milk fat is extremely complicated, and
even after prolonged storage all three types of crystal forms (α, β′ and β) can
coexist (see Chapter 2).
Milk proteins are dominated by caseins, consisting of α-, β- and κ-casein
molecules. These molecules are extremely amphiphilic, and are associated into
micelles in the milk. When milk is homogenized in order to obtain a uniform
globule size and avoid creaming, with a resultant tenfold increase in the surface
area between globules and milk serum, caseins will act as emulsifier and fill
empty space at the globule surface.
Consumer milk is standardized with regard to fat content, and is therefore
reconstituted after separation of the cream from the milk. It then has to be
pasteurized or sterilized, at about 80ºC or 125ºC, respectively.
Whipped cream
A typical composition after centrifugation into whipping cream is a fat content
of 30% (w/w). Whipping of cream involves the introduction of so many air
cells that we approach a foam structure. The air cells are stabilized by a two-
dimensional matrix of aggregated fat globules at the air–water interface.
Therefore a critical amount of fat globules is needed. Furthermore, fat crystals
must be present in the fat globules in order to hinder coalescence.
The probable structures at the three phase interfaces are illustrated in
Figure 12.4. When air cells are introduced, the fat globules will attach to their
256 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
Figure 12.4 Illustration of a probable interfacial structure at the initial stage of the whipping of cream.
surface in order to reduce the surface energy. Polar lipids from the surface will
spread along the interface, and the resulting partly ‘naked’ fat globules will
tend to adsorb at the interface with the exposed hydrophobic regions turned
towards air. Aggregation of globules at the interface will also be driven by the
exposed triacylglycerol core.
Butter
If the whipping is continued after the foam-like structure has been formed, the
new air cells will remove more of the globule coating, and aggregation of the
globules in three dimensions will start. This will result in a phase inversion into
a w/o emulsion. The emulsion oil continuum coexisting with an outside
buttermilk phase will spontaneously take up about 20% (w/w) of water. The
proposed interfacial extraction of the milk fat globule membrane on whipping
is confirmed by the presence of the membrane lipids in buttermilk.
The fat crystal network is an important quality factor of the butter, its
spreading properties, etc. Therefore, certain tempering regimes have been
developed for optimal crystal growth during the whipping process, termed
churning. Usually the temperature of the cream is first lowered to about 8ºC to
give numerous crystal nuclei. It is then raised to about 19ºC in order to favour
growth of the stable polymorphic forms over that of the unstable ones. The
temperature is then lowered to about 10ºC in order to obtain more crystals, and
churning is started.
An alternative to churning is to concentrate the cream by centrifugation to
the fat composition of butter. The concentration of fat globules is then above
that corresponding to close-packed spheres, which means that the globules
have to adopt a polyhedral shape. This emulsion is therefore very fragile, and
a minor mechanical disturbance is enough to cause phase inversion.
The advantage with this method is that the phase inversion can be initiated
when the desired composition of the butter is reached. Legislation in most
countries requires an exact composition of butter, which by traditional churn-
ing is not straightforward to achieve. Furthermore, the production of buttermilk
as a by-product is avoided. On the other hand, the presence of fat globule
membranes in the fat phase complicates the formation of a crystal network.
FOODS 257
Ice cream
Traditional ice cream is prepared from whipped cream, and represents an
interesting system with features in common with those of butter. About 10%
(w/w) of sugar and flavourings, like vanilla, are dissolved in the cream before
whipping. Whipping (or the actual agitation) has to be performed under
cooling. The crucial step is that the water phase must be transformed into ice
just before phase inversion of the fat globules takes place. The structure of the
fat globules on the verge of coalescence is then immobilized by the ice
crystals.
The typical sensory characteristics of ice cream are partly caused by the
cooling sensation that accompanies the melting process and partly by the
mouth-feel involved in coalescence of the fat globules during melting of the fat
and the release of flavour compounds.
Ice cream should have air cells to about half its volume. The ice crystals must
be less than about 30 μm in diameter in order to avoid a ‘sandy’ mouth-feel.
Hydrocolloids such as cellulose derivatives are often added in order to reduce
crystal growth during storage. Even at the low temperature at which ice cream
is stored, all the water is not frozen but can contribute to crystal growth.
Ice cream can also be prepared from a vegetable oil mix, which is first
emulsified into an artificial cream. Then the same processing conditions as
described above have to be fulfilled, including the presence of fat crystals. Milk
proteins are usually used for emulsification. A complication is that caseins are
such efficient emulsifiers that coalescence of the fat globules in the mouth is
inhibited. This problem can be solved by the addition of a small amount of
monoacylglycerols to the oil. These will displace enough protein molecules for
the coalescence process to take place. The mechanism behind this was de-
scribed in Chapter 8.
3. Drying of milk
It is often practical to store milk as a dry powder. When the surrounding water
is replaced by air, there is a tendency for the globule surface layer to rupture.
The addition of polar lipids and carbohydrates has proven to be an efficient way
of stabilizing the surface structure during drying so that the emulsion is
reconstituted on subsequent exposure to water. The polar lipid should supply
an Lα phase which can emulsify any oil phase that separates during the drying
process. The carbohydrate will accumulate at the globule surface during
drying. The polar surface is therefore stabilized and can persist even when
water is removed. Disaccharides and maltodextrins are usually used.
The same mechanism is used by certain microorganisms as protection
against drying damage. Trehalose is known to be secreted during the drying of
fungi, for example, and coats their surface.
258 LIPIDS: STRUCTURE, PHYSICAL PROPERTIES AND FUNCTIONALITY
to the stable β form may take place during storage. A fat composition favouring
stability of the β′ form is therefore required. Retardation of the phase transition
is favoured by:
• crystals containing fatty acids with varying chain lengths;
• co-crystallization of a lipid with a strong tendency to form stable crystals
with orthorhombic chain packing.
When margarines are used for frying, it is important to avoid spattering due
to the presence of large water droplets. Special emulsifiers with anti-spattering
effects are therefore used.
Margarines for use as spreads on bread are usually prepared from mixtures
of butter fat and vegetable fat in order to provide the taste of butter and good
spreading properties at refrigerator temperature. A typical spread has 75% milk
fat and 25% soybean oil or rapeseed oil as the fat phase, with a water phase
similar to that of butter. Margarines with a lower fat content have increased on
the market. Such spread products are also a convenient form for the incorpora-
tion of special nutrition-enhanced lipids, as so-called functional foods. One
example is a spread product containing plant sterols able to reduce blood
cholesterol levels.
In order to incorporate 40–60% (w/w) of a water phase, special emulsifiers
have to be used. The water droplets in an ordinary margarine are about 5 μm in
diameter. Larger water droplets are obtained when the water content is in-
creased, and then there is a risk of microorganism growth. Distilled
monoacylglycerol preparations are therefore used to reduce water droplet size.
Furthermore there is a need to increase the viscosity of the water phase; gelatin
is often used for this purpose.
1. Frying oils
Foods can be cooked by heating in water or in oil, and the difference in surface
temperature and chemical environment results in quite different food charac-
teristics, as we all know. A problem with the repeated re-use of frying oils is the
oxidative changes that occur, and the risk of formation of toxic degradation
products. Saturated oils will of course have better stability, but they have
obvious disadvantages from a nutritional point of view.
FOODS 261
Quality control of frying oils is based on measurements of the free fatty acids
content, the foaming properties of the oil, its viscosity, or its colour. An
interesting aspect of frying oil functionality is the presence of polar lipids,
which will influence heat diffusion at the surface and thus the heat transfer
efficiency. It has been shown that the amount of frying oil needed can be
reduced by half if phospholipids are added to the oil. A disadvantage, however,
is the risk of dissociation of the polar group, with the release of amines with
undesirable aroma effects.
Figure 12.5 Phase diagram of wheat lipids plus water (after Eliasson & Larsson, 1993). The different
phases are here termed: LC-L, denoting the lamellar liquid-crystalline phase; LC-H, the inverse
hexagonal phase; and L, a liquid phase without contact with the water or lipid corners, and therefore
probably an L3 phase.
hydrophobic, and their amphiphilic properties are the basis for their behaviour
in bread-making.
The lipid fraction also interacts with water, and is in fact a part of the gluten
gel (about 10%). The functionality of the lipids in baking will be summarized
here. In order to understand the effects of lipids it is convenient to start from a
phase diagram, as shown in Figure 12.5. It was constructed as a ternary system,
although there are thousands of different lipid species present. On chromato-
graphic fractionation of the total lipids, there is one type of fraction that does
not form aqueous phases. These non-polar lipids form one natural corner of the
phase diagram. The remaining fractions, dominated by galactolipids and
phospholipids, interact with water. They are therefore, according to our defini-
tion in Chapter 1, polar lipids, and together they are used here as the second
lipid corner of the phase diagram.
The dotted line on Figure 12.5 corresponds to the composition of the lipids
in wheat flour. If we follow this line, we will reach the water content where the
Lα phase, non-polar oil and a liquid phase (probably L3) are in equilibrium. The
partition of water between the gluten gel and the embedded lipid phases is not
known, but there is evidence indicating that the phases in this region contribute
significantly to the bread-making process.
FOODS 263
Figure 12.6 Illustration of the lamellae between gas pores in bread dough during expansion in the oven
(after Eliasson & Larsson, 1993). Lipid monolayers spread and stabilize the surface on drying and
contraction of the gluten gel phase.
References
Carlsson, T., Larsson, K., Dinh-Nguyen, N. & Krog, N. (1979) Starch 31, 222.
Eliasson, A-C., Larsson, K., Andersson, S., Hyde, S.T., Nesper, R. & Von Schnering, H-G.
(1987) Starch 39, 147.