.) o
r:u...v.:a

.

 

. r . ;
wwfiww
.. . (4' r l.
A»?

 

“an...”

nu»..-
mm“?

V . . 3mm”?
.2-.-

 

.. . w

.r . u!» 541.

m :mruwhr ,
A...

«ha»

.9»:
.10 0|! 1.

10..

 

gamma win.
. u. .3 .u
. . d 1..
. , 1..
. I ...:..
tum“? a!

.ILJM..- J1.
um
scans. .1 I1.

 

m .-,

.1.

4..
if”;
<::-,
Innnluli

v1.1

m6...
.. i .-

J

n‘nvflzflll
.l' xii." 741%»..FT
'I l

 

or.

.41ng
III.

                               

 

 

 

 

 

“WINNIHUIHHIU!llllllllllmll

293 015812153

LIBRARY

Mlchlgan State
University

 

 

 

 

This is to certify that the

dissertation entitled

Evaluation of oxidative stability of lipids by
fluorescence spectroscopy and characterization
of flavonoid antioxidant chemistry

presented by

Arti Arora

has been accepted towards fulfillment
of the requirements for

Ph . D. degree in Food Science

 

 

11/1

 

Major professor \(

Date 3/2 9/77

MS U is an Affirmative Action/Equal Opportunity Inuirun'on 042771

“.7 ‘-

-—v-v~ Y‘fl"

*‘f

7 ‘u ""

PLACE N RETURN BOX to remove this checkout from your rocord.
TO AVOID FINES return on or bdoro date duo.

ll DATE DUE DATE DUE DATE DUE
8 GMT

*ri— JUL 2919T

1le

MSU IcAn Affirmative Adlai/Ema! Opportunity Inctltwonm

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

EVALUATION OF OXIDATIVE STABILITY OF LIPIDS BY
FLUORESCENCE SPECTROSCOPY AND CHARACTERIZATION OF
FLAVONOID ANTIOXIDANT CHEMISTRY
By

Arti Arora

A DISSERTATION

Submitted to
Michigan State University
in partial fiJlfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Food Science and Human Nutrition

1997

ABSTRACT
EVALUATION OF OXIDATIVE STABILITY OF LIPIDS BY
FLUORESCENCE SPECTROSCOPY AND CHARACTERIZATION OF
FLAVONOID ANTIOXIDANT CHEMISTRY
By

Arti Arora

Fluorescence spectroscopy is a sensitive and rapid technique that can be exploited for
evaluation Of the oxidative stability of lipids. In this study, two assays based on quenching
of the fluorescence intensity of an extrinsic probe and an increase in its fluorescence
anisotropy due to the free radicals generated during lipid peroxidation were developed for
evaluation of antioxidant efficacy and for monitoring the progress of lipid peroxidation. The
assays were used to evaluate the efficacies of metal chelating and free radical scavenging
antioxidants.

By use of the assays, the effects of chelators on metal-ion-induced peroxidations were
found to be dependent on the type of metal ion used to initiate peroxidation and on the mole
ratio of chelator relative to lipid. F lavonoids and isoflavonoids were the class of compounds
selected to study prototypic, phenolic free radical scavenging antioxidants. These compounds
were more effective at suppressing metal-ion-induced peroxidations than peroxidations
induced by free radicals, an indication that metal chelation plays a larger role towards their
antioxidant mechanism than has previously been believed. Additionally, fluorescence
anisotropy measurements were conducted using a series of probes with the chromophore at
varying levels of penetration into the membrane. The results suggest that the flavonoids and

isoflavonoids partition preferentially into the hydrophobic interiors of membranes and

decrease membrane fluidity in this region. The reduced fluidity could limit accessibility of free
radical species generated during lipid peroxidation and stabilize the membrane, a Structural
mechanism of antioxidant action for these compounds.

Distinct structure-activity relationships were revealed for the antioxidant activities of
flavonoids and isoflavonoids. Presence Of electron-withdrawing groups enhanced activity
whereas electron-donating groups decreased activity, with the substitution pattern on the B-
ring of these compounds being an especially important determinant of activity. Among the
isoflavonoids, the biological metabolites of these compounds had similar or superior
antioxidant activities in comparison to the parent compounds, a finding that suggests that
these compounds may retain their activity under in vivo conditions. An oxidation product of
the reaction between the isoflavonoid genistein and peroxyl radicals was characterized and

could serve as a marker for genistein antioxidant chemistry.

To my parents

iv

ACKNOWLEDGMENTS

I would like to express my Sincere appreciation to my major professor Dr. Gale M.
Strasburg for his patience, support and encouragement throughout my studies at M.S.U. He
has been the best advisor, mentor and friend that a student could ever hope for, and I cannot
thank him enough for all the help and guidance he has provided me. He has been an excellent
role model and I hOpe tO live up to his standards of professional integrity.

I am grateful to all my committee members for their guidance and support. I would
like to acknowledge Dr. J. Ian Gray for all his support and enthusiasm for the project, for all
his helpfiil suggestions, for the innumerable letters of recommendations and especially, for his
constant good humor. I would also like to thank Dr. Muraleedharan G. Nair for providing
the flavonoid and isoflavonoid samples used in the research, for use of the equipment in his
laboratory, for help with data analyses and interpretation and for all the helpful meetings and
suggestions. I would like to thank Dr. Alfred Haug for the help he has provided and also for
all his challenging and thought-provoking questions that helped improve the quality of this
dissertation. I am also gratefiil to Dr. Robert Ofoli for serving on my committee.

I thank Dr. Russel Ramsewak for his assistance with the NMR analyses. 1 would also
like to acknowledge all the past and present colleagues in my laboratory. A special thanks

to my fiiends Stephanie Smith and Ivy Hsu for all their help and support during the past six

months, for their company during all the late nights in the department and for making the
dissertation completion process a lot more fiJn than it otherwise may have been. I am also
gratefiJl to all my other wonderful friends in the department- Zsuzsanna Balogh, Christine
Bergman, Mridvika, Alicia Orta-Ramirez, Vanee Chonhenchob, Scott Worthington, Jay
Chick, Arnie Sair and James Clarke, for giving me so many special memories of my days at
M.S.U. that I shall always treasure.

Finally, my heartfelt gratitude to my parents for all their love, support and

encouragement.

vi

TABLE OF CONTENTS

LISTOFFIGURES...................................-. ...............
LIST OF ABBREVIATIONS ..........................................
INTRODUCTION ...................................................

CHAPTER 1.
LITERATURE REVIEW ..............................................
1.1 Lipid Peroxidation .............................................
1.1.1 Mechanism of Lipid Peroxidation .............................
1.1.2 Role of Transition Metals ...................................
1.1.3 Role of Antioxidants .......................................
1.1.3.1 Free Radical Scavenging Antioxidants ..................
1.1.3.2 Metal Chelators ...................................
1.1.4 Methodology for Measuring Oxidative Stability of Lipids ...........
1.1.4.1 Need for a rapid and sensitive assay ....................
1. 1.4.2 Current Methodology ...............................
1.1.5 Model Systems for Measurement of Lipid Peroxidation .............
1.1.5.1 Biological Membranes as Substrates for Lipid Peroxidation . .
1.1.5.2 Liposomes as Models for Membrane Lipid Peroxidation .....
1.2 Fluorescence Spectroscopy .......................................
1.2.1 Advantages of Fluorescence Spectroscopy ......................
1.2.2 Basis of the Fluorescence Assay ..............................
1.2.3 Selection of Probe .........................................
1.2.3.1 Perturbation of membrane ...........................
1.2.3.2 High quantum yield ................................
1.2.3.3 Location of the extrinsic fluorescent probe ...............
1.2.3.4 Photodamage .....................................
1.2.3.5 Labeling the lipid assembly by the fluorescent probe ........
1.2.3.6 Probe concentration ................................
1.3 Flavonoids And Isoflavonoids .....................................
1.3.1 General Structure and Major Classifications .....................
1.3.2 Dietary Intake and Food Sources .............................

vii

xi

xiii

. S
. 5
. 5
. 6
. 7
. 8
10
12
12
l3
I6
16
17
18
19
19
20
20
21
21
23
23
24
24
24
27

1.3.2.1 Flavonoids ....................................... 27

1.3.2.2 Isoflavonoids ..................................... 28
1.3.3 Absorption and Metabolism ................................. 29
1.3.3.1 General Pathway .................................. 29
1.3.3.2 Flavonoids ....................................... 30
1.3.3.3 Isoflavonoids ..................................... 32
1.3.4 Biological Activities of Flavonoids and Isoflavonoids .............. 36
1.3.5 Antioxidant Activity of Flavonoids and Isoflavonoids .............. 38
1.3.5.1 Flavonoids ....................................... 38
1.3.5.2 Isoflavonoids ..................................... 41

CHAPTER 2
DEVELOPMENT AND VALIDATION OF FLUORESCENCE SPECTROSCOPIC
ASSAYS TO EVALUATE ANTIOXIDANT EFFICACY. APPLICATION TO METAL

CHELATORS ....................................................... 44
2.1 Abstract ..................................................... 44
2.2 Introduction .................................................. 45
2.3 Experimental Procedures ......................................... 48

2.3.1 Materials ................................................ 48
2.3.2 Preparation of Large Unilamellar Vesicles ....................... 49
2.3.3 Fluorescence Experiments ................................... 50
2.3.4 Preparation of SLPC Hydroperoxides .......................... 51
2.3.5 I-IPLC—Chemiluminescence and Conjugated Diene Determination ..... 52
2.4 Results ...................................................... 53
2.4.1 Characterization of the LUVS ................................ 53
2.4.2 DPH-PA Fluorescence Properties in the LUVS ................... 53
2.4.3 Fe(II)-Induced Peroxidation Monitored by Fluorescence
Intensity Decay ........................................... 58
2.4.4 F e(II)-Induced Peroxidation Monitored by Increase in
Fluorescence Anisotropy .................................... 60
2.4.5 F e(III)-Induced Peroxidation of LUVS ......................... 60
2.4.6 Cu(II)-Induced Peroxidation of LUVS .......................... 64
2.4.7 Effect of Chelator Concentrations ............................. 64
2.4.8 Validation of the F luorescence-Based Assays .................... 69
2.5 Discussion ................................................... 72

CHAPTER 3

STRUCTURE-ACTIVITY RELATIONSHIPS FOR ANTIOXIDANT ACTIVITIES OF A

SERIES OF FLAVONOIDS IN A LIPOSOMAL SYSTEM .................... 80
3.1 Abstract ..................................................... 80
3.2 Introduction .................................................. 81
3.3 Experimental Procedures ......................................... 86

3.3.1 Materials ................................................ 86

viii

3.3.2 Preparation of Large Unilamellar Vesicles ....................... 87

3.3.3 Antioxidant Evaluation of Flavonoids .......................... 87

3.4 Results ...................................................... 88

3.4.1 Effects of Flavonoids on F e(II)-Catalyzed Peroxidation in LUVS ..... 88

3.4.2 Effects of F lavonoids on Fe(III)—Catalyzed Peroxidation in LUVS ..... 91

3.4.3 Effects of F lavonoids on AAPH-Induced Peroxidation in LUVS ...... 94

3.5 Discussion .................................................. 101
CHAPTER 4

ANTIOXIDANT ACTIVITIES OF ISOFLAVONES AND THEIR BIOLOGICAL

METABOLITES IN A LIPOSOMAL SYSTEM ............................ 111

4.1 Abstract .................................................... 111

4.2 Introduction ................................................. 112

4.3 Experimental Procedures ........................................ 118

4.3.1 Materials ............................................... 118

4.3.2 Preparation of Large Unilamellar Vesicles ...................... 119

4.3.3 Antioxidant Evaluation of F lavonoids ......................... 119

4.4 Results ..................................................... 120

4.4.1 Effect of Isoflavonoids on Fe(II)-Catalyzed Peroxidation in LUVS . . . 120
4.4.2 Effects of Isoflavonoids on Fe(III)-Catalyzed Peroxidation in LUVS . . 126
4.4.3 Effects of Isoflavonoids on AAPH-Induced Peroxidation in LUVS . . . 126

4.5 Discussion .................................................. 133
CHAPTER 5

MODULATION OF LIPOSOMAL MEMBRANE FLUIDITY BY FLAVONOIDS AND

ISOFLAVONOIDS .................................................. 142

5.1 Abstract .................................................... 142

5.2 Introduction ................................................. 143

5.3 Experimental Procedures ........................................ 148

5.3.1 Materials ............................................... 148

5.3.2 Preparation of Large Unilamellar Vesicles ...................... 150

5.3.3 Fluorescence Anisotropy Measurements ....................... 151

5.4 Results ..................................................... 153

5.4.1 Effects of F lavonoids on Membrane F luidity .................... 153

5.4.2 Effects of Isoflavones on Membrane Fluidity .................... 156

5.4.3 Effects of Isoflavone Metabolites on Membrane F luidity ........... 159

5.4.4 Effects of Cholesterol and a-Tocopherol on Membrane F luidity ..... 162

5.5 Discussion .................................................. 162

ix

CHAPTER 6
OXIDATION PRODUCTS OF GENISTEIN FORIVIED BY REACTION WITH

ALKYLPEROXYL RADICALS ........................................ 170
6.1 Abstract .................................................... 170

6.2 Introduction ................................................. 171

6.3 Experimental Procedures ........................................ 174
6.3.1 Materials ............................................... 174

6.3.2 HPLC ................................................. 174

6.3.3 Spectroscopy ........................................... 175

6.3.4 Oxidation Studies with Genistein ............................ 175

6.4 Results ..................................................... 175
6.4.1 Formation of Genistein Oxidation Products ..................... 175

6.4.2 Chemical Structures of Oxidation Products ..................... 176

6.5 Discussion .................................................. 179
SUMMARY AND CONCLUSIONS ..................................... 181
FUTURE RESEARCH ............................................... 184
LIST OF REFERENCES ............................................. 186

Figure 1.1

Figure 1.2
Figure 1.3
Figure 1.4
Figure 2.1
Figure 2.2
Figure 2.3
Figure 2.4
Figure 2.5
Figure 2.6
Figure 2.7
Figure 2.8
Figure 2.9

Figure 2.10
Figure 2.11
Figure 2.12
Figure 2.13

Figure 3.1
Figure 3.2
Figure 3.3

Figure 3.4
Figure 3.5

Figure 3.6
Figure 3.7

Figure 3.8
Figure 3.9
Figure 4.1
Figure 4.2

Figure 4.3

LIST OF FIGURES

Structures of some extrinsic probes commonly used

for membrane studies ....................................... 22
Main classes of the flavonoid family ............................ 26
Metabolites of genistein detected in human urine .................. 34
Metabolites of daidzein detected in human urine ................... 35
Freeze-fracture electron micrographs of LUVS .................... 55
Fluorescence intensity and anisotropy of DPH-PA in LUVS .......... 57
Peroxidation of LUVS induced by F e(II) and chelators .............. 59
Peroxidation of LUVS induced by Fe(II) and chelators .............. 61
Peroxidation of LUVS induced by Fe(III) and chelators ............. 62
Peroxidation of LUVS induced by F e(III) and chelators ............. 63
Peroxidation of LUVS induced by varying ratios of NTA to F e(II) ..... 65
Peroxidation of LUVS induced by varying ratios of ADP to Fe(II) ..... 66
Peroxidation of LUVS induced by varying ratios of citrate to F e(II) . . . . 67
Peroxidation of LUVS induced by varying ratios of citrate to Fe(III) . . . 68
Peroxidation of LUVS induced by F e(II) and chelator ............... 70
Peroxidation of LUVS induced by Fe(III) and chelator .............. 71
Structure of 3-(p-(6-phenyl)-l,3,5-hexatrienyl)phenylpropionic acid . . . . 73
Structures of commonly-occurring flavonoids and other antioxidants . . . 84
Substitution pattens for the series of flavones examined ............. 85
Effects of commonly-occurring flavonoids, coumarin and TBHQ

on rates of F e(II)-induced peroxidation ......................... 90
Effects of series of flavones on rates of Fe(II)-induced peroxidation . . . . 93
Effects of commonly-occurring flavonoids, coumarin and TBHQ

on rates of Fe(III)-induced peroxidation ......................... 96
Effects of series of flavones on rates of Fe(III)-induced peroxidation . . . 98
Effects of commonly-occurring flavonoids, coumarin and TBHQ

on rates of AAPH-induced peroxidation ........................ 100
Effects of series of flavones on rates of AAPH-induced peroxidation . . 103
Structural criteria that enhance antioxidant activity of flavonoids ..... 110
Structures of the plant-derived isoflavonoids included in the study . . . . 115
Structures of the isoflavone metabolites examined for their

antioxidant activity ........................................ l 16

Effects of plant-derived isoflavones on rates Of Fe(II)-induced

xi

Figure 4.4
Figure 4.5
Figure 4.6
Figure 4.7
Figure 4.8
Figure 4.9
Figure 5.1
Figure 5.2
Figure 5.3
Figure 5.4
Figure 5.5
Figure 5.6
Figure 5.7
Figure 6.1
Figure 6.2

Figure 6.3

peroxidation ............................................. 122
Effects of isoflavone metabolites on rates of Fe(II)-induced

peroxidation ............................................. 125
Effects of plant-derived isoflavones on rates of Fe(III)-induced

peroxidation ............................................. 128
Effects of isoflavone metabolites on rates of Fe(III)-induced

peroxidation ............................................. 130
Effects of plant-derived isoflavones on rates of AAPH-induced

peroxidation ............................................. 132
Effects of isoflavone metabolites on rates of AAPH-induced

peroxidation ............................................. 135
Structural criteria that enhance the antioxidant activity of

isoflavonoids ............................................ 141
Structures of naringenin, hesperetin, rutin, a-tOCOpherol

and cholesterol ........................................... 146
Structures of isoflavonoids and their metabolites ................. 147
Structures of the fluorescent probes employed ................... 149
Effects of flavonoids on membrane fluidity ...................... 155
Effects of isoflavonoids on membrane fluidity .................... 158
Effects of isoflavonoid metabolites on membrane fluidity ........... 161
Effects of cholesterol and a-tocopherol on membrane fluidity ....... 164
Structures for genistein and 2,2'-azobis(isobutyronitrile) ............ 173
Reversed-phase HPLC analysis of products formed by

reaction between genistein and AIBN-derived peroxyl radicals ....... 177
Structure for Compound 1 .................................. 178

xii

AAPH
ADP
AIBN

l 6-AP
6-AS

1 2-AS
BHA
BHT
cis-PA
DMSO
DPH
DPH-PA
DPPC
EDTA
HEPES
HPLC
LUV
MDA
MLV
MOPS

NTA
O-Dma
PCOOH
PDA
SLPC
SUV
TBA
TBARS
TBHQ
UV

LIST OF ABBREVIATIONS

2,2'-Azobis(2-Amidinopropane) Dihydrochloride
Adenosine Diphosphate
2,2'-Azobis(isobutyronitrile)
16-(9-Anthroyloxy)Palmitic Acid
6-(9-Anthroyloxy)Stearic Acid
12-(9-Anthroyloxy)Stearic Acid

Butylated Hydroxyanisole

Butylated Hydroxytoluene

cis-Parinaric Acid

Dimethyl Sulfoxide

1 ,6-Diphenyl-1 , 3 , S-Hexatriene

3-(p-(6-Phenyl)-1 ,3 , 5-Hexatrienyl)Phenylpropionic Acid
Dipalmitoylphosphatidylcholine
Ethylenediaminetetraacetic Acid
4-[2-Hydroxyethyl]-1-Piperazine Ethanesulfonic Acid
High Performance Liquid Chromatography
Large Unilamellar Vesicle

Malondialdehyde

Multilamellar Vesicle

3-[N-Morpholino] Propanesulfonic Acid

Nuclear Magnetic Resonance

Nitrilotriacetic Acid

O-Desmethylangolensin

Phosphatidylcholine Hydroperoxide

Photodiode Array

1 -Stearoyl-2-Linoleoy1-sn-Glycero-3-Phosphocholine
Small Unilamellar Vesicle

2-Thiobarbituric Acid

2-Thiobarbituric Acid Reactive Substances
tert-Butylhydroquinone

Ultraviolet

xiii

INTRODUCTION

Lipid peroxidation is a very undesirable reaction in foods and is generally recognized
as the primary factor limiting the shelf-life of most processed food products (Loliger, 1992).
To mitigate the detrimental effects of lipid peroxidation on food quality, many highly effective
synthetic antioxidants have been developed. In recent years, however, considerable interest
in the use of plant derived antioxidants has developed, fueled in part by consumer concerns
about the safety of synthetic compounds in the food supply. This interest has been stirred
further by observations that many of these plant compounds, which are potent antioxidants,
also display antiatherogenic, antimutagenic, and anticarcinogenic activities (Duthie, 1991;
Kinsella et al., 1993). Hence, introduction of such compounds in food products may impart
health benefits in addition to stabilizing the product.

The screening of a large number of compounds or plant extracts for possible
antioxidant activity requires a rapid and sensitive assay for evaluation of oxidative stability
of lipids. The current methods for antioxidant evaluation, however, suffer from serious
limitations. Most of the assays are time-consuming, have limited sensitivity and specificity,
and require large amounts of test material. Ofien the methods used (such as high
temperatures to accelerate peroxidation) introduce artifacts into the sample which complicates

interpretation of the antioxidant efficacy. Additionally, the model systems employed in many

2
of these assays are not representative of the structural and functional characteristics of the
substrate in the food product.

The primary aim of this research was to develop a rapid and sensitive assay for
evaluation of the oxidative stability of lipids that employed a model system which addressed
many of the concerns about the experimental systems used by other researchers. The model
system selected for the study was a chemically, well-defined liposomal system that was
representative of biological membranes.

Fluorescence spectroscopy was chosen as the technique for use in development of an
assay for routine evaluation of the eflicacy of antioxidants due to its inherent advantages of
speed, simplicity and sensitivity. The overall hypothesis of this research was that fluorescence
spectroscopy could be utilized for the evaluation of the antioxidant potencies of metal
chelators and free radical scavengers, and that the technique had the required sensitivity to
distinguish between the antioxidant activities of structurally related compounds.

Using fluorescence, the susceptibility of a probe to peroxidative damage with
accompanying loss of its fluorescent character can be used to monitor the progress of lipid
peroxidation directly and sensitively. In addition, the anisotropy parameter of these probes
provides sensitive responses to changes in fluidity of the environment surrounding the probe.
This can be used to follow the decrease in membrane fluidity that accompanies a rise in
peroxidation.

Upon validation of the fluorescence spectroscopic assay using direct measures of lipid
peroxidation, the second objective of this research was to test the ability of the method to

discriminate between antioxidant activities of compounds within a structurally related class.

3

F lavonoids and isoflavonoids were the classes of compounds selected for evaluation in this
study because they display excellent potential for use as antioxidants. These compounds
exhibit a broad spectrum of positive pharmacological properties, including vasoprotective,
antiinflammatory, antiviral, and antitumor activities. Many of these pharmacological
properties are believed to be related to the antioxidant activity of these compounds (Rapta
etal., 1995).

Having established the ability of this method to distinguish differences in antioxidant
activities among these compounds, the third objective of the study was to define the
substitution patterns on the flavonoids and isoflavonoids that are necessary for a high
antioxidant activity. Establishment of a structure activity relationship for flavonoids and
isoflavonoids will allow for their selection as antioxidants based on structure alone, without
having to screen every possible compound for activity.

For successful use of any compound to enhance oxidative stability of lipids, it is
imperative that their mechanism of antioxidant action be completely understood. The
antioxidant mechanisms for flavonoids and isoflavonoids have not been clearly established.
The polyphenolic structure of flavonoids and isoflavonoids confers them with the ability to
scavenge free radicals and also to chelate transition metals, two possible mechanisms of
action. Additionally, these compounds may act as membrane stabilizers by decreasing fluidity
in the lipid bilayer. The next aim of this study was to investigate the mechanisms of
antioxidant action for flavonoids and isoflavonoids.

Prior to incorporating the pharmacologically relevant doses of these compounds in

foods, their oxidation products need to be characterized to ascertain the safety of these

4

breakdown products. Profiling the non-radical products formed during the antioxidant
reactions of a representative flavonoid or isoflavonoid would be the final aim of this study.

This dissertation was organized into a series of chapters. Each chapter covered a
specific aim of the study and was prepared as a manuscript with its specific abstract,
introduction, experimental procedures, results and discussion sections. Sections that were
common to the entire dissertation included the initial abstract, introduction and literature

review sections, and the final conclusion, fiiture research and list of references sections.

CHAPTER 1

LITERATURE REVIEW

1.1 LIPID PEROXIDATION

Lipid peroxidation, the oxidative degradation of polyunsaturated lipids, is well-
recognized as a major concern to food scientists since it results in the development of
undesirable ‘off-flavors’ and potentially toxic reaction products (Coupeland and McClements,
1996). In recent decades, its relevance to the field of biology and medicine has also received
increasing attention. Lipid peroxidation is a major contributor to membrane damage in cells,
disrupting their important structural and protective functions (Barclay and Vinquist, 1994).
It has also been implicated in a variety of pathological processes, such as inflammation,
muscular dystrophy, ischemia/ reperfusion and aging (Choe and Yu, 1995; Dix and Aikens,

1993; Yu et a1. 1992).

I. I. 1 Mechanism of Lipid Peroxidation
Lipid peroxidation proceeds by a classic free-radical chain mechanism involving three
discrete phases: (i). Initiation, the formation of a carbon-centered lipid radical (L') by

abstraction of a hydrogen atom from an unsaturated fatty acid (LH) in the presence of an

6

initiator (Equation 1); (ii). Propagation, fithher reactions of L' with molecular oxygen and
lipid to yield peroxyl radicals (LCD) and lipid hydroperoxides (LOOH) (Equations 2 and 3);
and, (iii). Termination, the formation of non-radical final products upon combination of two

radicals (Equation 4) (Gutteridge, 1988).

Initiation LH + Initiator —’ L' (1)
Propagation L' + 02 —* L00 (2)

L00" + LH —+ LOOH + L' (3)
Termination L' + L' —' LL (4)

I. 1.2 Role of Transition Metals

The direct reaction of LH with oxygen is spin-forbidden since the ground-state of
lipids is of singlet multiplicity whereas that of oxygen is of triplet multiplicity (Miller et al.,
1990). Lipid peroxidation must, therefore, occur via reactions that relieve this spin-restriction
between lipids and oxygen (Minotti and Aust, 1992). Transition metals such as iron and
copper are efficient catalysts of lipid peroxidation because their reactions with dioxygen are
not spin-restricted (Buettner and Jurkiewicz, 1996) and also because they can oscillate
between reduced and oxidized states with great ease (Minotti, 1993).

Of the transition metals, iron plays the most significant role in lipid peroxidation. Two
basic mechanisms have been established for understanding the role of iron in catalyzing lipid

peroxidation. According to the first mechanism, Fe(II) catalyzes the decomposition of trace

7

amounts of preformed LOOHs to form the highly reactive alkoxyl radicals (LO') which
abstract hydrogen from a neighboring allylic bond (Minotti, 1992) (Equations 5 and 6). This

is referred to as the “LOOH-dependent” lipid peroxidation (Minotti and Aust, 1987).

Fe(II) + LOOH —» Fe(III) + OH’ + LO (5)

LO- + LH —’ L-+ LOH (6)

According to the second mechanism, F e(II) reacts with oxygen to yield the superoxide
anion (02") and hydrogen peroxide (H202) (Equations 7 and 8). This is called the “LOOH-
independent” lipid peroxidation since the reaction of Fe(II) with H202 eventually liberates
hydroxyl radicals ('OH) that can abstract hydrogen from LH irrespective of any preformed

LOOH (Equations 9 and 10) (Minotti, 1992).

Fe(II) + O2 —v Fe(III) + 02" (7)
2 O," + 2 H‘ —’ H202 + 02 (3)
Fe(II) + H202 —» Fe(III) + OH’ + OH (9)
'OH + LH —> H20 + L- (10)

1. 1.3 Role of Antioxidants
The ubiquity of oxygen and transition metal catalysts in food products poses a serious

challenge to the food technologist to minimize the undesirable peroxidation of the lipid

8

constituents (Graf, 1994). An antioxidant is defined as “any substance that, when present at
low concentrations compared to those of an oxidizable substrate, significantly delays or
prevents oxidation of that substrate” (Halliwell, 1990).

Based on the mode of action, antioxidants can be primarily divided into three
categories: Type 1, Type II and Type III (Labuza, 1971). The Type I antioxidants, such as
tocopherols, work by scavenging free radicals and suppressing fi'ee radical oxidation by
molecular oxygen. The Type II antioxidants, including metal chelating agents like
ethylenediaminetetraacetic acid (EDTA),citric acid and various forms of ascorbic acid,
prevent the production of free radicals. By sequestering the metal ions catalysts, these
compounds suppress the generation of free radicals and reduce the overall rate of oxidation.
The Type III antioxidants include environmental factors such as the physical removal of

headspace oxygen through vacuum processing or modified atmosphere packaging.

1.1.3.] Free Radical Scavenging Antioxidants

Chain-breaking antioxidants (AH) can interfere with the lipid peroxidation chain
reaction by providing an easily-donatable hydrogen atom for abstraction by lipid radicals.
These reactions compete with the chain propagating reactions of lipid peroxidation
(Equations 11 and 12). The antioxidant-derived radical (A') may further react with another
peroxyl radical, dimerize to A1 or get converted back to AH by reaction with another

molecule (Equations 13, 14 and 15) (Gutteridge and Halliwell, 1990).

AH + LOO- —v LOOH + A- (11)
LO- + AH —* LOH + A- (12)
LOO + A- —» LOOA (13)
L0 + A' -’ LOA ( I 4)
A" + A' —' A2 (15)

Many of the food antioxidants, such as a-tocopherol, butylated hydroxyanisole
(BHA), butylated hydroxytoluene (BHT) and propyl gallate, act by a mechanism of free
radical scavenging. They reduce the primary radical by a one-electron reduction to a non-
radical chemical species, and in the process, get transformed into oxidized antioxidant radicals
themselves. The two basic conditions that must be satisfied for a molecule to be defined as
a chain-breaking antioxidant are: (i) when present in low concentrations, relative to the
substrate to be oxidized, it can delay, retard or prevent the free-radical mediated oxidation;
and (ii) the resulting radical formed after scavenging must be of such low reactivity that no
further reactions with lipids can occur (Halliwell, 1990).

These chain-breaking, free radical scavenging antioxidants are generally phenolic
compounds with one or more hindered phenolic hydroxyl groups. These phenolic compounds
are excellent hydrogen atom or electron donors; in addition, their radical intermediates are
relatively stable due to resonance delocalization and lack of suitable sites for attack by
molecular oxygen (Shahidi and Wanasundara, 1992). The efficacy of these compounds as

antioxidants is related to their chemical structure. In addition to modulating the radical

lO

scavenging activity of the hydroxyl group, the chemical structure of these antioxidants also
influences their physical properties such as volatility, lipid solubility and thermal stability

(Halliwell, 1994).

1.1.3.2 Metal Chelators

In certain cases, chelators fonn thermodynamically stable complexes that can remove
metals that are complexed to biomolecules, thereby protecting them from site-Specific
oxidation. Complexation of metals may also hinder their ability to participate in redox
reactions (Buettner and Jurkiewicz, 1996). However, the use of chelators cannot guarantee
the removal of metal ions as catalysts of lipid peroxidation.

Various chelators like adenosine diphosphate (ADP), EDTA, citrate and histidine have
been shown to have a prooxidant, antioxidant or no effect on the rates of metal-catalyzed lipid
peroxidation (Ahn etal. 1993; Tampo et al. 1994; Ursini et al. 1989; Yoshida et al. 1993).
The information available in the literature on the effects Of chelators on metal-catalyzed lipid
peroxidation is contradictory and not well understood due to the lack of standardized test
conditions used in these studies. This is because the results obtained with metal chelators are
greatly dependent on the source and preparation of tissues, cells, membranes or purified
lipids, the solvents used, the concentrations of metal complexes, the presence or absence of
oxygen source, and the methods used for measurements of these effects (Schaich, 1992).

Many studies have demonstrated the complexity imparted by chelators on the rates

of metal-catalyzed lipid peroxidation. Chelators are critical determinants of the catalytic

mode and effectiveness of metals (Schaich, 1992). Kinetics, mechanisms, and products of

ll

metal-catalyzed lipid peroxidation can be altered by chelation. The presence and nature of
chelators and the chelator to iron ratio is known to have a strong effect on iron autoxidation
(Minotti, 1993). Chelators alter the redox potential of metals and may also render steric
effects which may influence the rate or efficiency of coordination or binding to the target
molecule and/or hydroperoxide (Chevion, 1988).

The effects of chelators on rates of transition metal catalyzed lipid peroxidation are
dependent on the metal affinities and charges of the chelators, and also on their partitioning
properties between lipid and aqueous phases. The valence state of metal they stabilize, the
metal coordination sites they occupy, the type of electron transfer reactions they mediate and
the redox potential of their complexes all influence the rates of lipid peroxidation (Schaich,
1992).

While there is some information available on the effects of other chelators, most
Studies have dealt with EDTA and its effects on metal-catalyzed lipid peroxidation. EDTA
is known to remove the “free” or loosely bound complexed iron fi'om solution, thereby
making it unavailable for participation in peroxidation reactions. Additionally, it lowers the
Fe(II) / F e(III) redox potential and thus, limits the capability of iron to act as an oxidizing
agent. However, the lower redox potential makes EDTA-iron a better reductant, so EDTA-
Fe(II) chelates reduce lipid hydroperoxides faster than uncomplexed iron. In the presence of
reducing agents that can recycle the iron, EDTA complexation may even result in marked
acceleration of chain propagation and branching reactions (Tien et al., 1982). The net effect

ofEDTA results fi'om the balance between these actions in individual systems, and has led to

12

apparently contradictory reports and interpretations of effects of EDTA on lipid peroxidation
(Schaich, 1992).

The ratio of chelator to metal cation can also markedly affect the mechanism of
initiation of lipid peroxidation (Kanner et al., 1987). EDTA is known to accelerate lipid
peroxidation at chelator to metal ratios of less than one, whereas it inhibits peroxidation at
chelator to metal ratios of greater than one. This is because EDTA, in stoichiometric excess,
deactivates the iron ion by surrounding it with tightly bound ligands that cannot be replaced
with reagents such as hydroperoxides (Waters, 1971).

In addition to chelator to metal ratio, lipid configuration also alters the effects of the
chelator-metal complex on the rate of lipid peroxidation. In liposomes and microsomes,
EDTA appears to prevent iron penetration through the membrane and the formation of a site-
specific attack of 'OH radical towards the unsaturated fatty acids. This effect is eliminated
upon dispersion of the liposomes or microsomes with detergents (Kanner etal., 1986; Girotti
and Thomas, 1984). Surface charge on membranes and chelators also affects molecular
access and binding as well as the dynamics of electron transfer reactions through lipid phases,

thereby altering membrane peroxidation rates (Schaich, 1992).

1. 1. 4 Methodology for Measuring Oxidative Stability of Lipids

1.1.4.1 Need for a rapid and sensitive assay

Extremely potent synthetic antioxidants such as BHA, BHT, tert-butylhydroquinone

(TBHQ) and various esters of gallic acid have been developed and proven to be extremely

13

effective in increasing the oxidative stability of foods (Namiki, 1990). Though these synthetic
antioxidants have been very thoroughly tested for their toxicological behavior, some of them
are now coming into disfavor as new toxicological data impose some caution on their use
(Haigh, 1986). This has led to increased interest in the use of plant-derived compounds as
antioxidants to satisfy consumer concerns regarding synthetic compounds. In order to screen
a large number of compounds for possible antioxidant activity, it is necessary to have a

simple, sensitive and rapid assay for evaluating the oxidative stability of lipids.

1.1.4.2 Current Methodology for Evaluation of Oxidative Stability of Lipids

The current methods for antioxidant evaluation and for measurement of oxidative
stability of lipids suffer from serious limitations. Accelerated stability tests are most
commonly used to estimate the shelf-life of a food. To speed up the oxidation process in
these tests, several parameters such as temperature, metal catalysts and oxygen pressure are
manipulated. Heating is most commonly used to accelerate oxidation since the rate of
reaction increases exponentially with the absolute temperature (Ragnarrson and Labuza,
1977)

However, the higher the temperatures, the greater the number of limitations associated
with these tests. Results obtained with accelerated tests like the Rancimat (produced by
Metrohm Ltd, CH-9100, Herisau, Switzerland) and the Active Oxygen Method are not
representative of real-life storage conditions since the mechanism of lipid peroxidation
changes significantly at the elevated temperatures (loo-140°C and 98°C respectively) used

in these tests (Frankel, 1993).

14

The effects of temperature on multi-component food systems can be very complex.
The physical and chemical changes occurring at these higher temperatures could drastically
affect the chemical reactivity through their effect on the distribution of reactants, metal
binding properties, viscosity, metal prooxidant effectiveness, water activity and other
variables. Other limitations associated with these high temperatures are that volatile
antioxidants such as BHA and BHT undergo significant losses at the elevated temperatures,
decomposition of phenolic antioxidants in natural extracts occurs, and the rate of oxidation
becomes dependent on oxygen concentration because solubility of oxygen decreases at
elevated temperatures. Additionally, Maillard reaction products that possess proven
antioxidant activities may be formed at the higher temperatures (Ragnarsson and Labuza,
1977)

In the accelerated stability tests, the induction period or the onset of oxidation is
determined using standard methodology for measurement of lipid peroxidation. The 2-
thiobarbituric acid (TBA) test is the most frequently used method for assessing peroxidative
damage to lipids. This spectrophotometric test measures the amount of pink chromagen
formed by reaction of two molecules of TBA with one molecule of malondialdehyde (MDA),
one of the many decomposition products of peroxidation of polyunsaturated fatty acids. As
lipids are peroxidized, the amount of TBA-reactive substances (TBARS) increases in the
product, and as such, the measurement of TBARS provides a usefirl index of lipid
peroxidation (Gray and Monahan, 1992).

However, the interference by other aldehydes and a large number of biological

molecules (such as amino acids, bile pigments, and some carbohydrates) and the generation

15

ofMDA during the assay itself make the TBA test somewhat non-specific and unsuitable for
certain test substrates (Gray and Monahan, 1992). In addition, the formation of MDA occurs
so far downstream from the initial production of lipid hydroperoxides that it is difficult to
translate rates of MDA formation into direct rate constants for expressing antioxidant activity
of test compounds.

Peroxide value determination is often used for measurement of hydroperoxides, one
of the initial products formed upon peroxidation of lipids. Peroxides may be measured by a
variety of methods. The most commonly used methods utilize iodometric techniques Similar
to the 1993 Association of Official Analytical Chemists (AOAC) method, in which peroxide
value is reported as milliequivalents (meq) of iodine per kilogram of fat. Determination of
hydroperoxides may not be a useful measure of lipid peroxidation in foods during prolonged
storage, as breakdown to secondary products such as aldehydes may occur, leading to an
underestimation of the degree of oxidation (Melton, 1983).

Another method for assessing the extent of lipid peroxidation involves measurement
of conjugated dienes structures which absorb ultraviolet (UV) light in the range of 230-235
nm. Conjugated dienes reflect the initial damage to the unsaturated fatty acids upon exposure
to oxygen, and as such, are useful in assessing the early stages of the peroxidation process in
studies of pure lipids (Halliwell and Chirico, 1993). However, application of this simple
absorbance technique to more complex biological and food systems can lead to serious
misinterpretation problems because changes in absorbance in this region of the spectrum
cannot be uniquely attributed to conjugated dienes. It has been shown that human body fluids

contain a non-oxygen-containing isomer of linoleic acid, octadeca-9-cis-11-trans-dienoic acid,

16

that absorbs UV radiation at the conjugated diene wavelength (Dormandy and Wickens,
1987). Though this compound was initially proposed to result from hydrogen abstraction of
linoleic acid and reaction of the resulting carbon-centered radicals with protein, it was later
identified as a product of bacterial fatty acid metabolism (Jack et al., 1991). Since conjugated
diene products may also be present in animal diets, use of this technique can also lead to
erroneous results in attempts to measure lipid peroxidation in tissues (Holley and Slater,

1991).

I. 1. 5 Model Systems for Measurement of Lipid Peroxidation

The inconsistent reports in the literature on the effects of metals on rates of lipid
peroxidation in biological systems result fiom the lack of precisely controlled conditions and
defined composition of laboratory test tube conditions in the cells, tissues, membranes and

organelles used in these Studies (Schaich, 1992).

1.1.5.] Biological Membranes as Substrates for Lipid Peroxidation

For many food products, the membrane constituents of the cells comprising the food
are the primary sites of peroxidative damage. This is likely related to the large surface
exposure of membrane phospholipid molecules to the aqueous phase containing peroxidation-
initiating species, in contrast to neutral lipid molecules in fat droplets which have a much
lower surface to volume ratio. Sarcoplasmic reticulum microsomes (Monahan et al., 1994;
Dinis et al., 1993) and mammalian erythrocyte membranes (McKenna et al., 1991; Thomas

et al., 1990) are usefiJl models in membrane peroxidation studies. However, use of biological

l7

membranes as lipid substrates to evaluate antioxidant activity of compounds is complicated
by the presence of numerous endogenous prooxidative and antioxidative factors, including
transition metals, heme proteins, catalases, glutathione reductase, and superoxide dismutase

(Kanner etal., 1988; Kellogg and Fridovich, 1975; Girotti and Thomas, 1984).

1. 1. 5. 2 Liposomes as Models for Membrane Lipid Peroxidation

Liposomes - self-assembling colloidal particles in which a lipid bilayer encapsulates
a fraction of the surrounding aqueous medium (Lasic and Papahadjopoulos, 1995) -
constitute a simple and convenient system for studying membrane lipid peroxidation and its
inhibition by antioxidants without the ambiguities introduced by enzymes or scavengers that
may be present in more complex biological systems (Vigo-Pelfrey and Nguyen, 1991). Using
liposomes as the substrate, the oxidation process can be reproduced with different levels of
complexity.

On the basis of size and number of bilayers, liposomes are primarily categorized into
three types - multilamellar vesicles (MLVS), small unilamellar vesicles (SUVs) and large
unilamellar vesicles (LUVS). MLVS, which are typically greater than 400 nm in diameter,
consist of numerous concentric bilayers separated by narrow aqueous channels (Hope et al.,
1986). Liposomes comprising a single, bimolecular layer of lipids are divided into two classes
based on size: vesicles that range in diameter from 20 to 50 nm are considered to be SUVS,
whereas those with a diameter of 60 nm or greater are classified as LUVS (Chapman, 1984).

MLVS are formed spontaneously when dry phospholipid films swell in excess water or buffer,

I i

18

whereas some energy must be dissipated into the system in order to produce SUVS and
LUVS, since they possess higher free energies (Lasic, 1988).

Despite their ease of preparation, MLVS represent a poor choice as a model for lipid
peroxidation studies due to their heterogeneity in size, the relatively small proportion of lipid
that is exposed to the aqueous phase and the different degrees of accessibility from the
external aqueous medium into the internal bilayers (Hope et al., 1986). SUVS also
inadequately mimic the properties of natural biomembranes. In SUVS, a disproportionate
amount (approximately 70%) of the total lipid is located in the outer leaflet of the vesicle.
In addition, the small radius of curvature in SUVS imposes strains in packing of the lipid
molecules that are not associated with biological membranes (Chapman, 1984). LUVS offer
the combined advantages of MLVS and SUVS. Their single bilayer, uniform Size and large

radius of curvature makes LUVS an excellent model for lipid peroxidation studies.

1.2 FLUORESCENCE SPECTROSCOPY

Fluorescence spectroscopy is a method that can easily provide extensive information
regarding the structure and interactions of macromolecules that may only be obtainable by
other approaches with difficulty (Bentley et al., 1985). Despite the inherent simplicity and
reliability of this technique, the utility and potential of fluorescence methods has yet to be fully

recognized.

l9

1. 2. 1 Advantages of Fluorescence Spectroscopy

Fluorescence Spectroscopy offers several advantages for characterization of molecular
reactions and interactions. First, it is approximately 1000 times more sensitive than
spectrophotometric techniques, an attribute that becomes important when characterizing
mechanisms and identifying small amounts of unique reaction products. Second, fluorescent
compounds are exquisitely sensitive to their environment. Lipophilic and amphipathic
membrane probes are virtually non-fluorescent in the aqueous phase, but when embedded in
the hydrophobic environment of the lipid bilayer, they are highly fluorescent. This
environmental sensitivity allows characterization of phenomena such as changes in membrane
fluidity and cross-linking reactions that may occur during lipid peroxidation. Third,
fluorescence methods once established are relatively rapid; thus, a substantial amount of

information can be quickly obtained (Strasburg and Ludescher, 1995).

1. 2. 2 Basis of the Fluorescence Assay

The underlying principles which make fluorescence a useful method for peroxidation
studies are derived in part from the molecular characteristics of the fluorescent probes
selected for these studies. Typically, these probes possess a conjugated double bond structure
which gives rise to their fluorescent properties, as well as to a susceptibility to reaction with
free radical intermediates of lipid peroxidation produced by added initiators such as F e(II)
metal ions, resulting in the formation of non-fluorescent products. The rate of decay of
fluorescence intensity, therefore, reflects the rate of peroxidation. Addition of free-radical

scavenging antioxidants (which compete with the fluorescent probe for reaction), or of metal

20

chelators (which alter the reactivity of the metal ion initiator species), will alter the rate of
fluorescence intensity decay. The extent to which fluorescence decay is inhibited by the
added compound is an indicator of its antioxidant efficacy. These properties combine to make
fluorescence spectroscopy a usefirl tool for rapid evaluation of antioxidant candidates and for
identifying Structural characteristics which define the antioxidant activities of structurally

related molecules.

1. 2.3 Selection of Probe
Today, many well-characterized membrane probes are commercially available, and

selection of a probe must be made after careful consideration of the factors outlined below.

1. 2. 3. 1 Perturbation of membrane

The insertion of extrinsic probes into the membrane inevitably results in some
perturbation of the system (Lentz, 1993) which may, in tum, alter a membrane’s susceptibility
to peroxidation. Use of fluorescent reporter molecules, which are structurally similar in chain
length and bulk to the membrane lipids, can reduce the amount of structural perturbation of
the lipid bilayer. Further reductions in the levels of macroscopic perturbation can be achieved
by use of a low concentration of probe molecules relative to the lipid molecules. A ratio of
less than one probe molecule to one hundred lipid molecules is usually desirable in membrane

Studies.

21
1.2.3.2 High quantum yield
Minimization of fluorescent probe concentration results in reduction of fluorescent
intensity. To partially offset the reduced signal, the fluorophore of choice Should have a high
quantum yield; i.e., a high ratio of photons emitted to photons absorbed. Probes with high
quantum yield can be used at low concentrations, thereby minimizing the structural

perturbations to the membrane while maintaining adequate signal intensity.

1. 2. 3. 3 Location of the extrinsic fluorescent probe

The precise orientation and location of a fluorOphore in the lipid bilayer are also of
great importance for the interpretation of experimental observations, and hence need to be
clearly established. The partitioning of the fluorophore in the lipid bilayer can be predicted
from the geometry of the fluorophore as well as of the whole probe molecule (Borenstain and
Barenholz, 1993). In membranes, symmetric, non-polar molecules such as 1,6-diphenyl-
1,3,5-hexat1iene (DPH, Figure 1.1) difluse rapidly into the hydrophobic interior of the bilayer
with a lack of orientational specificity, thereby providing only bulk-averaged information
(Beck et al., 1990). By appropriate chemical modification, such probes can be localized in
specific regions of the membrane. The fluorescent probe 3-(p-(6-phenyl)-l,3,5-
hexatrienyl)phenylpropionic acid (DPH-PA, Figure 1.1), an anionic derivative of the parent
DPH molecule, has a more restricted orientation in the membrane. The polar substituent of
DPH-PA provides a surface anchor that ensures that the charged substituent on the probe is

localized at the lipid-water interface of the membrane, whereas the lipophilic tail of the probe

22

(a)

\\\I

(b)

CHZCHZCOOH
O \ \ \ C

(C)

(C H2)7C OOH
/ / / /

H3C

Figure 1.1 Structures of some extrinsic probes commonly used for membrane studies: (a),
1,6-diphenyl-l,3,5-hexat1iene (DPH); (b), 3-(p-(6-phenyl)1,3,5-hexatrienyl)phenylpropionic

acid (DPH-PA); and (c), cis-trans-trans-cis-9,11,13,15-octadecatetraenoic acid (cis-PA).

23

is embedded in the hydrophobic interior of the membrane, lying parallel to the acyl chains

(Beck et al., 1990).

1.2. 3. 4 Photodamage

Fluorescent molecules are, to varying degrees, subject to photooxidation during the
course of an experiment. Photochemically-induced damage to the fluorescent probe usually
results in loss of fluorescence which may confound interpretation of chemical peroxidation
reaction kinetics. For example, the fluorescent probe cis-trans-trans-cis—9,11,13,15-
octadecatetraenoic acid, commonly referred to as cis-parinaric acid (cis-PA, Figure 1.1), has
been used by various researchers to monitor lipid peroxidation in membranes because of its
close structural similarity to intrinsic membrane lipids (de Hingh et al., 1995; Dinis et al.,
1993; Kuypers et al., 1987; Laranjinha et al., 1992; McKenna et al., 1991; Van den Berg
et al., 1988). However, it is very photolabile and undergoes light-induced dimerization
(Morgan et al., 1980). This necessitates Special handling requirements for the probe, and the
use of excitation light of the lowest possible intensity for short periods of time (Lentz, 1993).
This disadvantage may preclude the use of cis-PA in experiments requiring long exposure of
sample to light. In such cases, one may select alternative probes like DPH and DPH-PA that

exhibit greater photostability.

1.2.3.5 Labeling the lipid assembly by the fluorescent probe
Another important point to consider during probe selection for the study of

membranes is the requirement for a high lipid to aqueous phase partition coefficient (2 l x

24

108). Probes that are hydrophobic (DPH) or amphipathic (DPH-PA and cis-PA) have very
low fluorescence in the aqueous phase and partition spontaneously into the lipid environment
of the membranes. This ensures that the level of fluorophore remaining in the aqueous phase
during the experiment is minimal (Borenstain and Barenholz, 1993), and that the observed

fluorescence is reporting molecular reactions occurring in the lipidic regions alone.

1. 2. 3. 6 Probe concentration

For unambiguous interpretation of results, the optimal concentration of fluorescent
probe and the ratio of probe molecules to that of lipid molecules needs to be determined. The
fluorescent intensity of the probe must be a linear function of its concentration; i.e., as
fluorescent molecules react with free radicals during peroxidation, the loss of fluorescence
intensity Should be directly proportional to the reaction progress. This linearity is generally
valid only at low concentrations. As membranes become concentrated with probes, the
fluorescent intensity dependence on the concentration of the probe is reduced and may even
decrease because of the inner filter effect or quenching of fluorescence by the probe molecules

themselves (Borenstain and Barenholz, 1993).

1.3 FLAVONOIDS AND ISOFLAVONOIDS

1. 3. 1 General Structure and Major Classifications
Flavonoids are ubiquitous, naturally occurring, low molecular weight aryl compounds

found in photosynthesizing cells, seeds, fruits and flowers (Das, 1994). Over 4,000

25

flavonoids have been identified from both higher and lower plants, and the list is constantly
expanding (Cook and Samman, 1996).

Flavonoids exhibit a wide diversity in structure. Based on a few backbone structures,
various hydroxylation, methoxylation, sulfation and/or glycosylation patterns exist (Bors et
al., 1990). The basic feature common to all flavonoids is the flavone nucleus made up of two
benzene rings (A and B) linked through a heterocyclic pyrane C ring. The position of the
benzenoid B ring divides the flavonoid class into the two subclasses of flavonoids (2-position)
and isoflavonoids (3-position) (Figure 1.2). The six-membered C ring is either a y-pyrone,
i.e., has a C2-C, double bond (flavonols and flavones) or its dihydro derivative (flavanols and
flavanones) (Havsteen, 1983).

These compounds are ofien hydroxylated at positions 3, 5, 7, 3', 4' and 5'. The
presence or absence of a hydroxyl group at position 3 firrther results in the generation of two
main subgroups for the flavonoids and isoflavonoids: 3-hydroxy flavonoids and isoflavonoids
(flavonols, flavanols, isoflavonols, isoflavanols) and 3-deoxy flavonoids and isoflavonoids
(flavones, flavanones, isoflavones, isoflavanones). These compounds occasionally occur
naturally in plants as their fi'ee aglycone forms; however, the most frequently encountered
forms are the glycoside derivatives. For the flavonoid glycosides, the glycosidic linkage is
normally located at positions 3 or 7 and the carbohydrate can be L-rhamnose, D-glucose,

glucorhamnose, galactose or arabinose (Brandi, 1992).

26

 

O
O
Flavone Isoflavone
0 O
OH
0 O
F lavanol

1F lavonol

O

Flavanone

Figure 1.2 Main classes of the flavonoid family.

OH

27

1.3.2 Dietary Intake and Food Sources

1.3.2.1 F lavonoids

Flavonoids are widespread among plant and plant products (i.e., fermented and
processed foods) (Formica and Regelson, 1995). Their daily consumption in the diet is
difficult to determine since intake is strongly dependent on feeding habits and, in this field,
exhaustive tables on food composition are not always available (Manach et al., 1996). Until
recently, data on human flavonoid intake were obtained from Kilhnau (1976) who estimated
the average intake of all dietary flavonoids in the Western diet to be approximately 1 g/day
(expressed as glycosides) of which about 170 mg (expressed as aglycones) consisted of
flavonols, flavanones and flavones. These values have been widely cited in the literature;
however, they were based mainly on food analysis techniques of doubtful accuracy (Hertog
et al., 1993a). Recently, the consumption of the flavonols quercetin, luteolin, kaempferol,
apigenin, and myricetin (and their glycoside derivatives) was documented in The Netherlands
using more advanced technologies (Hertog et al., 1993a; Hertog et al., 1993b; Hertog et al.,
1992). Based on these analyses, the average dietary intake of flavonoids in The Netherlands
was estimated to be approximately 23 mg/day (expressed as aglycones); with quercetin being
the major dietary flavonoid (16 mg/ day) (Hertog et al., 1993b).

Tea, coffee, cocoa, fiuit juices, red wines, beer and vinegar are some of the important
dietary sources of flavonoids, accounting for approximately 25-3 0% of total flavonoid intake
(Kilhnau, 1976). The flavonoid glycosides, particularly quercetin and kaempferol glycosides,

are found in the edible portions of a majority of food plants, e.g. citrus and other fruits;

28

berries, leafy vegetables; roots, tubers and bulbs; herbs and spices; legumes; cereal grains;

and tea and cocoa (Brown, 1980).

1. 3. 2. 2 Isoflavonoids

In marked contrast to the flavonoids, isoflavonoids have a very limited distribution in
the plant kingdom. Over 90% of the firlly characterized isoflavonoids (aglycones and
glycosides) are produced by species belonging to the large and taxonomically-advanced
subfamily Papilionoideae of the Leguminosae (Ingham, 1983). Even within the two other
smaller and more primitive subfamilies Caesalpinioideae and Mimosoideae of the
Leguminosae, only one or two plants have been reported to contain isoflavonoids (Dewick,
1988). As a consequence of their limited distribution, restricted primarily to tropical legumes,
isoflavonoids represent a very minor part of Western diets (Coward et al., 1993).

Soybeans are a rich dietary source of isoflavonoids. In soybeans and foods derived
from soy, isoflavones are found in concentrations ranging from 0.1 to 5 mg/g (Coward et al.,
1993). Genistein (4',5,7-trihydroxyisoflavone) and daidzein (4',7-dihydroxyisoflavone) are
the two most abundant isoflavones found in soy (Wang and Murphy, 1994a). These two
isoflavones occur principally as their 7-0-glucosides, genistin and daidzin (Wang and Murphy,

1994b)

29

1.3.3 Absorption and Metabolism

1. 3. 3. 1 General Pathway

Most of the flavonoids enter the diet as glycosides, with quercetrin and rutin
(glycosides of quercetin) being the most commonly consumed flavonoids (Bokkenheuser and
Winter, 1988). The hydrophilicity and relatively high molecular weight of the glycosides
preclude their absorption in the small intestine. Furthermore, these flavonoid fl-glycosides
are resistant to hydrolysis by the intestinal digestive enzymes and pass largely unaltered into
the large intestine (Brown, 1980). The microflora of the mammalian lower bowel produce
glycosidases capable of hydrolyzing the flavonoid glycosides to their constituent aglycones
and sugars (Hawksworth et al., 1971; Prizont et al., 1976; Scheline, 1968). In addition to
glycoside hydrolysis, the resident microflora can cleave the pyrone ring (ring C) to yield a
variety of phenolic acids, such as phenylpropionic and phenylacetic acid derivatives, and other
derivatives (Brown, 1980; Hackett, 1986). However, as glycosidase activity proceeds at a
faster rate than ring cleavage, the intact flavonoid aglycone can persist in the large intestine
with the clear potential for absorption (Formica and Regelson, 1995).

The major metabolic reactions occurring in the lower gut appear to be hydrolysis and
reductions. Liver is the site for most conjugation and oxidation reactions of these
compounds. A conjugation reaction with glucuronic acid or sulfate in the liver is perhaps the
most common final step in the metabolic pathways of intact flavonoids. The polar, water
soluble flavonoid glucuronides and sulfates appear to be readily excreted by mammals. These

polar conjugates are either excreted into the urine or into the duodenum with bile salts, where

30

they come into contact with the enzymes of intestinal micro-organisms. It appears likely that
in case of the flavonoids preferentially excreted in the bile, ring-fission products and flavonoid
aglycones are the metabolic products not only of unabsorbed orally administered flavonoids,
but also of their biliary metabolites. The probable fate of these biliary flavonoid glucuronides
is metabolic hydrolysis in the lumen of the intestine. The liberated flavonoids may undergo
firrther metabolism by intestinal micro-organisms or be reabsorbed from the intestine,
transferred via the hepatic portal vein to the liver and metabolized and re-excreted in bile
creating an enterohepatic circulation (Hackett, 1986).

Despite the potentially significant effects of these compounds in vivo, there is a
paucity of information about the absorption, metabolism and excretion of individual
flavonoids and isoflavonoids in humans. Most studies of individual flavonoid metabolism in
humans have used pharmacological doses of these compounds rather than the estimated
dietary intake levels of 23 to 170 mg/day; thereby, making it difficult to extrapolate results
of these studies to explain the absorption and metabolism of dietary flavonoids (Cook and

Samman, 1996).

1. 3. 3. 2 F lavonoids

The metabolism and absorption of quercetin, the major representative of the flavonol
subclass, has been examined in some studies. Gugler et al. (1975) investigated the
metabolism ofquercetin in six volunteers after the administration of single intravenous (100
mg) or oral (4 g) doses. Of the intravenous dose, approximately 7% was excreted in the urine

as a conjugated metabolite, and less than 1% was excreted unchanged. In contrast, after the

31

oral administration ofquercetin, no measurable plasma concentrations could be detected and
nor was any quercetin found in the urine, either unchanged or in a metabolized form. The
fecal recovery for quercetin after the oral dose was around 53%, indicating extensive
degradation by gut microflora. This led the researchers to conclude that oral administration
of flavonoids may be of questionable value. An earlier study with quercetin and rutin also
failed to detect any of the two compounds in urine after their oral administration (Clark and
MacKay, 1950).

Recent studies, however, appear to dispute these findings. In a human study with
ileostomy subjects, absorption of quercetin aglycone was found to be 24% (Hollman et al.,
1995). The absorption of quercetin glucosides from onions was found to higher, at 52%.
The data from this study suggest that humans absorb appreciable amounts of quercetin and
that the absorption of glycosides in the small intestine is possible.

A later study by the same group of researchers extended these findings. Hollman et
al. (1996) examined the time course of plasma quercetin concentration in two subjects after
ingestion of fiied onions containing quercetin glucosides equivalent to 64 mg of quercetin
aglycone. The mean peak plasma level of quercetin was 200 ng/ml and was reached 2.9 h
after ingestion of the onions, with an average half-life of absorption of 0.87 h. The half-life
of the elimination phase was 17 h, suggesting that repeated intake of quercetin glucosides
would lead to a build-up of the concentration in plasma. Quercetin was still detectable in the

plasma, at a concentration of 10 ng/ml, 48 h after ingestion of the onions.

32
1.3.3.3 Isoflavonoids

Current knowledge of isoflavonoid metabolism stems largely from sheep studies.
Since a massive outbreak of permanent and temporary infertility in sheep grazing on certain
cultivars of subterranean clover (T rifolium subterraneum L.) in Western Australia was
reported in the 19405 (Bennetts et al., 1946), a substantial amount of work on the metabolism
and estrogenic effects of isoflavones on sheep has been conducted. For ruminants, the major
metabolic transformation of isoflavones is performed by microorganisms in the rumen
(Nilsson et al., 1967). Biochanin A undergoes demethylation to genistein, which is converted
to p-ethyl phenol and other phenolic acids by ring cleavage (Lundh, 1995). In contrast,
fonnononetin, which is indirectly responsible for estrogenic disturbances in Sheep (Millington
et al., 1964), is primarily demethylated to daidzein and subsequently to equol via
hydrogenation (Lundh, 1995).

In spite of the purported significant role of isoflavonoids in human health, the
pathways of metabolism of daidzein and genistein, the two principal isoflavones found in
legumes and soy, remain unclear. These two isoflavones can be obtained from plant foods
as the free, unconjugated forms, referred to as daidzein and genistein, or through the
deconjugation of the glycosidic derivatives, daidzin and genistin. Additionally, they can be
synthesized in the intestinal tract from their plant precursors, forrnononetin and biochanin A,
respectively (Adlercreutz et al., 1991, Hutchins et al., 1995). In vitro anaerobic incubation
of isoflavones with human feces suggests that intestinal half-life of genistein and daidzein may
be as little as 3.3 h and 7.5 h respectively, indicating rapid degradation to other compounds

by the gut microflora (Xu et al., 1995).

33

Following ingestion of isoflavone-rich foods such as soy, the following compounds
have been described in human urine: the isoflavones daidzein, genistein and formononetin;
and their metabolites equol (7-hydroxy-3-(4’-hydroxyphenyl)-chroman), 0-
desmethylangolensin (O-Dma), dihydrodaidzein, dihydrogenistein, tetrahydrodaidzein, 6'-
hydroxy-O-Dma, and dehydro-O-Dma (Adlercreutz et al., 1986; Adlercreutz et al., 1991;
Axelson et al., 1982; Axelson et al., 1984; Bannwart et al., 1984; Joannou et al., 1995; Kelly
et al., 1993; Kelly et al., 1995).

Biochanin A and formononetin undergo demethylation by intestinal bacteria, giving
rise to genistein and daidzein respectively. Genistein is metabolized within the gut by ring
cleavage to yield the non-estrogenic, hormonally inert compound p-ethylphenol (Kelly et al.,
1993). Recently, dihydrogenistein and 6'-OH-0-Dma were also identified as catabolic
products of genistein metabolism in humans (Joannou et al., 1995) (Figure 1.3). Daidzein
undergoes intestinal microbial transformation in humans to yield the isoflavan equol (about
70%) and O-desmethylangolensin (5-20%) (Setchell and Adlercreutz, 1988). Equol does not
appear to be degraded firrther and most of the absorbed equol is conjugated in the liver with
glucuronic acid before being excreted in the urine (Axelson and Setchell, 1980; Axelson et
al., 1984). Other daidzein intermediates formed during this conversion are dihydrodaidzein,
4-hydroxy equol, 2-dehydro-0-Dma, and possibly dehydroequol (Joannou et al., 1995)

(Figure 1.4).

34

Oct: 58:: E 388% EBmEOm mo meg—onfioz m; oSwE

65.: ESE

O = Ewen—owe.“
EOHmEomEPEO -_»fioEo?O.>xo€»:-.o
:0
o :0 no

 

O 0: fl

52250

:0
O :0

35

Oct: 58:: 5 532% we mug—onfiofi «2 SEE

10 6:383:60
\
0 CI
52235
_>£os%.o-oazmo-m a
:0
0
536.3
:0
0
N10 l III
:0 O:
O O:
_o:um
IO
5223:0359:va
foo
O
O O:

n

:0
IO OI

:0

I0

.035 E02314

53282255

0

OI

OI

36

1 . 3. 4 Biological Activities of F lavonoids and Isoflavonoids

F lavonoids and isoflavonoids are known to display a bewildering array of
pharmacological and biochemical actions. Much of the recent interest in flavonoids and other
polyphenolic compounds was created by the anomaly of the “French Paradox”, the apparent
compatibility of a high fat diet with a low incidence of coronary atherosclerosis (Renaud and
De Lorgeril, 1992). It was suggested that the polyphenolic substances in red wine
(flavonoids, catechins, anthocyanins and soluble tannins) may provide protection against
coronary heart disease. In vitro (Frankel et al., 1993) and in vivo (Fuhrman et al., 1995)
studies confirmed the inhibition of oxidation of low density lipoproteins by the phenolic
substances in red wine.

In support of flavonoids exerting a protective effect in vivo were the results of a
Dutch epidemiological study (the Zutphen Elderly Study) demonstrating that heart disease
in elderly males was inversely correlated with their intake of flavonoids (Hertog et al., 1993).
The major sources of flavonols and flavones were tea (61%), onions (13%) and apples (10%).
Mortality from coronary heart disease was strongly and inversely correlated with flavonoid
intake in the Zutphen Elderly Study; a reduction in the mortality risk of more than 50% was
found. A recent cohort Study involving a 26-year follow-up of Finnish men and women found
a similar inverse association between intake of flavonoids and coronary mortality (Knekt et
al,1996)

In addition to their protective effect against heart disease, flavonoids are known to
exhibit a broad spectrum of positive pharmacological properties, including antiinflammatory

(Brasseur, 1989; Ferréndiz and Alcaraz, 1991), antiviral (Selway, 1986), antimetastatic

37

(Menon et al., 1995), anticarcinogenic (I-Iirose et al., 1994) and antitumor (Smith and Banks,
1986; So et al., 1996) activities. While some of these effects (i.e., antiproliferative activity)
may be attributed to the topoisomerase-II-dependent DNA cleavage induced by flavonoids
(Yamashida et al., 1990); others (such as vasoprotective and antiinflammatory) are believed
to be related to their antioxidant properties (Rapta et al., 1995).

Flavonoids have the ability to modify the activities of a host of enzyme systems
including protein kinase C and various other kinases, tyrosine kinase, aldose reductase,
myeloperoxidase, NADPH oxidase, xanthine oxidase, phospholipase A2, phospholipase C,
reverse transcriptases, omithine decarboxylase, salidase, several ATPases, nucleotide
phosphodiesterases, lipoxygenases, cyclooxygenase, cytochrome P450-dependent mixed
firnction oxidases, epoxide hydrolase, glutathione-S-transferase, and aromatase, amongst
others. Many of these enzyme systems are critically involved in immune function,
carcinogenesis, cellular transformation, and tumor growth and metastasis (Kandaswami and
Middleton, 1994).

Studies with mammalian cell systems demonstrate that flavonoids can modify the
firnction of mast cells, basophils, neutrophils, eosinoiphils, macrophages/monocytes, B and
T lymphocytes, platelets, nerve, smooth muscle and various cancer cells. This enables these
compounds to affect diverse and numerous physiological and pathological processes,
including secretion, platelet aggregation and adhesion to endothelial surfaces, mitogenesis,
cell motility and malignant cell proliferation, cancer metastasis and, the function or expression

of adhesion molecules in various mammalian cell types (Kandaswami and Middleton, 1993).

38

Isoflavonoids, being closely related in structure to estrogenic steroids, have also been
shown to bind to estrogen receptors and exert weak estrogenic activity in various in vivo and
in vitro assays (Martin et al., 1978; Shutt and Cox, 1972); in addition, significant estrogenic
effects in animals and in man have been observed (Setchell et al., 1987; Van Thiel et al.,
1991). Definite antiestrogenic effects have also been reported in vivo since high levels of
synthetic estrogens seem to be counteracted by administered isoflavonoids or their presence

in the diet (F olman and Pope, 1966; Folman and Pope, 1969).

1. 3. 5 Antioxidant Activity of F lavonoids and Isoflavonoids

1.3.5.1 Flavonoids

Flavonoids are second only to the tocopherols as the most common and the most
active antioxidant compounds naturally occurring in foods, possessing activity in both the
hydrophilic and lipophilic systems (Kirhnau, 1976). The antioxidant activity of these
compounds is conferred by their polyphenolic Structures. Polyphenols, depending on their
precise structure, can act as antioxidants either by virtue of their free radical scavenging
ability or their metal chelating activity (Salah et al., 1995). The relative contributions of these
two mechanisms are in dispute (van Acker et al., 1996a). It is widely believed that the
antioxidant ability of the flavonoids and isoflavonoids resides mainly in their ability to donate
hydrogen atoms, thereby scavenging the free radicals generated during lipid peroxidation.
Metal chelation has generally been regarded to play a minor role in the antioxidant activity

of these compounds, and so, has not been studied much by researchers in the area (Morel et

39

al., 1994). This is despite the early realization by researchers that the structures of these
compounds allowed them to form heavy metal complexes (Kuhnau, 1976). Two points of
attachment to the flavonoid were established: first, the o-diphenol (3',4'-dihydroxy-) grouping

in ring B, and secondly, the ketol structure in ring C of the flavonols:

 

 

 

 

Lack of at least one of these groups was found to reduce or delete the chelating
activity. However, the oxo group at C-4 could be reduced to a hydroxy group without loss
of metal-complexing ability. The ability of the flavonoid aglycone quercetin and of its
corresponding glycoside rutin to form stable, inert complexes with iron have also been
demonstrated earlier (Afanas’ev et al., 1989). In another study, the cytoprotective effects of
the flavonoids catechin, quercetin and diosmetin were investigated on iron-loaded hepatocyte
cultures, considering two parameters: the prevention of iron-increased lipid peroxidation and
the inhibition of intracellular enzyme release (Morel et al., 1993). The flavonoids tested were
found to be capable of removing of iron from the hepatocytes; and their iron-chelating ability

was correlated with their cytoprotective effect.

40

The redox chemistry of flavonoids is a predictor of their free radical scavenging
activity. The redox properties of flavonoids enable them to act as reducing agents, hydrogen-
donating antioxidants and Singlet oxygen quenchers (Rice-Evans and Miller, 1996). The
reduction potentials of flavonoid radicals are lower than those of the alkyl, peroxyl and
superoxide radicals. This enables the flavonoid radicals to inactivate the damaging oxyl
species generated during lipid peroxidation and prevent the deleterious consequences of their
reactions (Jovanovic et al., 1992; Wardman, 1989). While the general capability of
flavonoids to scavenge superoxide radicals has been demonstrated in numerous studies
(Cotelle et al., 1992; Hanasaki etal., 1994; Yuting etal., 1990; Zhou and Zheng, 1991),
the specific scavenging abilities proposed for hydroxyl (Husain et al. 1987), superoxide
(Robak and Gryglewski, 1988) and peroxyl radicals (Torel et al. 1986) are still subject to
debate (Bors et al. 1990). This is because of the non-specific techniques used to generate free
radicals in these studies coupled with the nonspecific assay methods employed for their
detection.

The structural requirements that are considered essential for effective radical
scavenging by the flavonoids are indicated below:

(i). The o-dihydroxy (catechol) structure in the B ring, a radical target site for all flavonoids
with a saturated 2,3-bond (Bors et al. 1990; Heilmann et al. 1995).

(ii). The presence of a hydroxyl group at position 3 on the C ring (Afanas’ev et al. 1989; Hu
et al. 1995; Mora et al. 1990; Ratty and Das, 1988). Flavonoid aglycones with a 3-OH
group such as fisetin, (+)-catechin, quercetin, myricetin, and morin are potent inhibitors of

lipid peroxidation when compared with those lacking a 3-OH substitution such as the flavones

41

diosmetin and apigenin and the flavanones hesperetin and naringenin (Cook and Samman,
1996)

(iii). The 2,3-double bond in conjugation with a 4—oxo function, which participate in electron
delocalization from the B ring (Bors et al. 1990).

(iv). The number of hydroxyl groups, a higher number providing maximal radical-scavenging
potential and the strongest radical absorption (Bors et al. 1990; Chen et al. 1996, Cotelle et
al. 1996; Roginsky et al. 1996).

(v). The patten of hydroxylation (Cholbi et al. 1991). Hydroxyl groups on positions C-5 and
C-7 of the A-ring (De Whalley et a1. 1990; Salvayre et al. 1988); C-3' and C-4' of the B-ring
(Salvayre et a1. 1988; Yuting et al. 1990); and position C-3 of the C ring appear to contribute
to inhibition of lipid peroxidation (Cholbi et al. 1991). Flavonols appear to require a C-2'
hydroxyl and the pyrogallol group (C-3', C-4', C-5') for antiperoxidative activity (Cook and
Samman, 1996).

Despite the wealth of data on the importance of flavonoids in conferring protection
against lipid peroxidation, the correlation between antioxidant activity and chemical structure
is far from clear. This can be attributed, in part, to the different methods of assessment,
varying substrate systems, and differential concentrations of active antioxidants used (Rice-

Evans et al. 1996).

1. 3. 5. 2 Isoflavonoids
In contrast to the numerous studies examining the abilities of flavonoids to inhibit lipid

peroxidation, the subgroup of isoflavonoids have remained relatively unexplored in terms of

42

their antioxidant activities. This is despite the numerous potent biological activities associated
with these compounds.

In a comparative study, the inhibition of in vitro microsomal lipid peroxidation
induced by a Fe(II)-ADP complex and NADPH by naturally occurring isoflavones and their
reduced derivatives (isoflavanones and isoflavans) was examined (Jha et al. 1985). While all
the isoflavonoids examined had an inhibitory activity, the isoflavanones were found to be
more effective than the parent isoflavones and the isoflavans were shown to be the most
potent inhibitors.

The other studies examining the antioxidant activities of isoflavonoids have mainly
focused on genistein. Genistein has been shown to strongly suppress tumor promoter-induced
H202 formation in both in vitro and in vivo conditions (Wei et al. 1993). A later study by the
same researchers (Wei et al. 1995) confirmed the antioxidant and antipromotional effects of
genistein. Dietary administration of genistein enhanced activities of antioxidant enzymes,
prolonged tumor latency and decreased tumor multiplicity in mice. The researchers
concluded that the potent inhibition of oxidant formation and proto-oncogene expression by
genistein suggested that the antioxidant and antiproliferative effects of genistein may, at least
in part, be responsible for its anticarcinogenic mechanism.

In another in vitro study, genistein was found to be an effective scavenger of H202
but was less effective against other peroxidative systems (Record et al. 1995). The
researchers could not find any evidence of iron chelation by genistein and concluded that in
vivo, genistein would be unlikely to be sufficiently potent a chelator to compete with

endogenous iron-binding ligands.

43

Another group of researchers found that dietary genistein enhanced the activities of

antioxidant enzymes in various organs of mice, and this may be the mechanism of its

chemopreventive actions (Cai and Wei, 1996).
None of the studies that were conducted attempted to elucidate the mechanism of

action by which these compounds act as antioxidants or to establish the structural criteria

necessary for a high activity.

CHAPTER 2

DEVELOPMENT AND VALIDATION OF FLUORESCENCE
SPECTROSCOPIC ASSAYS TO EVALUATE ANTIOXIDAN T

EFFICACY. APPLICATION TO METAL CHELATORS.I

2.1 ABSTRACT

Two fluorescence-based assays were developed for rapid evaluation of compounds
for antioxidant activity. These assays were based on the quenching of intensity of the
fluorescent probe and an increase in its fluorescence anisotropy due to the free radicals
generated during lipid peroxidation. A large unilamellar vesicle (LUV) system containing the
fluorescence probe diphenylhexatriene-propionic acid (DPH-PA) was used to study the
effects ofchelators on metal-ion-induced lipid peroxidation. In this paper, the actions of the
chelating agents ethylenediaminetetraacetic acid disodium salt (EDTA), nitrilotriacetic acid
trisodium salt (NTA), adenosine-5'-diphosphate disodium salt (ADP) and Sodium citrate on
Fe(H) and F e(III)-induced peroxidation were compared. The effects of chelators on metal-
ion-induced peroxidation were found to be dependent on the type of metal used to initiate

peroxidation and, in the case of citrate, also on the concentration of chelator used. EDTA

 

1Journal of American Oil Chemists’ Society. In Press.

44

45

strongly suppressed both F e(II) and Fe(III)- induced peroxidation in this system. NTA and
ADP inhibited Fe(III)-induced peroxidation but enhanced F e(II)-induced peroxidation at all
the concentrations tested. Citrate promoted both Fe(II) and Fe(III)-induced peroxidations
at lower chelator to metal ratios; however, at higher ratios, it inhibited both peroxidations.
The results of the two fluorescence-based assays agreed well with the quantitation of
conjugated dienes and hydroperoxides by HPLC. The combination of sensitivity, speed and
general utility associated with these methods suggest that they will be useful in rapid

screening of extracts and purified compounds for antioxidant activity.

2.2 INTRODUCTION

The peroxidation of lipids via uncontrolled free radical chain reactions is a critically
important reaction in physiological and toxicological processes in human health and disease
as well as in the Stability of food products during storage (Schaich, 1992). In biomembranes,
it leads to a disruption in structure and a loss of protective function (Barclay, 1993).

Membrane lipid peroxidation is greatly stimulated by the presence of transition metals,
partially through enhancement of initiation reactions as well as through metal ion catalysis of
lipid hydroperoxide decomposition reactions (Braughler et al. , 1987). Chelators can alter the
rate of metal-catalyzed peroxidation by steric effects, variations in redox potentials and
alterations in solubility properties of the metal (Aust et al., 1985). Several studies have been
conducted to determine the effects of metal chelators on lipid peroxidation (Dikalov et al.,

1993; Fujii et al., 1991; Miller et al., 1992; Spear and Aust, 1994; Tampo et al., 1994;

46

Yoshida et al., 1993); however, the effects are complicated and not fully understood
(Yoshida et al., 1993).

The choice of model system and the methods used in the study of lipid peroxidation
often introduce artifacts which complicate interpretations of the results. Results from many
commonly used model systems such as methyl linoleate micelles may have little relationship
to food products since they are not structurally representative of membrane phospholipids or
triglycerides (Frankel, 1993). SarCOplasmic reticulum microsomes (Dinis et al., 1993;
Monahan et al., 1994; Tien et al., 1982) and mammalian erythrocyte membranes (McKenna
et al., 1991; Thomas et al., 1990) are usefirl models in membrane peroxidation studies.
However, use of biological membranes as lipid substrates to evaluate antioxidant activity of
compounds is complicated by the presence of numerous endogenous prooxidative and
antioxidative factors, including transition metals, heme proteins, catalases, glutathione
reductase, and superoxide dismutase (Girotti and Thomas, 1984; Kanner et al., 1988; Kellogg
and Fridovich, 1975; Svingen et al., 1979). Although biological membranes may provide
useful information on oxidative damage or antioxidant status in an individual, the inherent
variability from preparation to preparation makes their use less attractive as systems for
evaluation of antioxidant activity of compounds or plant extracts.

As an alternative model to study lipid peroxidation, artificial membranes such as
liposomes offer clear advantages over biological membranes. In particular, large unilamellar
vesicles (LUVS) of a defined lipid composition constitute a simple and convenient system for

studying lipid peroxidation and its inhibition by antioxidants without the ambiguities

47

introduced by enzymes or scavengers that may be present in more complex biological systems
(Vigo-Pelfrey and Nguyen, 1991).

Use of a defined lipid substrate and structure coupled with fluorescence spectroscopy
offers the inherent advantages of speed, simplicity and sensitivity for probing membrane
structure. This methodology can be readily adapted to peroxidation studies on the mechanism
of action and efficacy of food antioxidants (Strasburg and Ludescher, 1995). The objective
of this study was to develop a sensitive, quantitative and rapid fluorescence-based assay for
measuring lipid peroxidation and to apply the assay to evaluate antioxidant efficacy. LUVS
containing the fluorescent probe 3-(p-(6-phenyl)-1,3,5-hexatrienyl)phenylpropionic acid
(DPH-PA) were used to study the modulation of F e(H), F e(HI) and Cu(II)-induced membrane
peroxidation by the chelators disodium EDTA, nitrilotriacetic acid trisodium salt (NT A),
adenosine-5'-diphosphate disodium salt (ADP) and sodium citrate. Results from these
experiments were correlated with data obtained by measurement of conjugated dienes and the
direct quantitation of hydroperoxides by HPLC-chemiluminescence to validate the

fluorescence-based assays.

48

2.3 EXPERIMENTAL PROCEDURES

2. 3.1 Materials

Synthetic 1-stearoyl-Z-linoleoyl-sn-glycero-3-phosphocholine (SLPC) of greater than
99% purity in chloroform solution was purchased from Avanti Polar Lipids (Alabaster, AL).
The purity of the lipids was confirmed by thin layer chromatography using two different
solvent systems (chloroform: methanol: water :: 65:25:4; chloroform: methanol: ammonium
hydroxide :: 652254). The lipids were stored in amber glass vials, layered with nitrogen,
sealed with teflon tape and stored at -20°C. The fluorescent probe 3-(p-(6-phenyl)—l,3,5-
hexatrienyl)phenylpropionic acid (DPH-PA) was purchased from Molecular Probes (Eugene,
OR). Adenosine—5'-diphosphate disodium salt (ADP), nitrilotriacetic acid trisodium salt
(NT A), 3-[N-morpholino] propanesulfonic acid (MOPS), potassium tetraborate, cytochrome
c from horse heart (prepared without using TCA), 5-amino-2,3-dihydro-1,4-phthalazinedione
(luminol), Chelex 100, methylene blue and xylenol orange were from Sigma Chemicals (St.
Louis, MO). 4-[2-hydroxyethyl]-l-piperazine ethanesulfonic acid (HEPES) was purchased
from Boehringer Mannheim (Indianapolis, IN). Ethylenediaminetetraacetic acid disodium salt
(EDTA), F eCl,'6I-I,O, F eCL' 4H 0 and sodium phosphate were from Mallinckrodt (Paris, KY),
citric acid monohydrate was from Fisher Scientific (Fair Lawn, NJ) and CuCl,'2H,O was from
J .T. Baker (Phillipsburg, NJ). All the glassware used in the study was acid-washed.
Contarninating transition metals were removed from the sodium citrate, sodium chloride and
buffer stock solutions by maintaining the solutions in Chelex 100 (5g/ 100 mL, w/v). The

solutions were sparged with nitrogen prior to use. The metal solutions were made up fi'esh

49

in nitrogen-sparged water immediately before use and stored on ice. The sodium citrate and

ADP stock solutions were adjusted to a pH of 7.0.

2. 3. 2 Preparation of Large Unilamellar Vesicles

LUVS were prepared immediately before use according to the procedure outlined by
MacDonald et al. (1991) with a few modifications. Briefly, the lipid and fluorescent probe
stocks dissolved in chloroform and N,N-dimethylformamide respectively, were dried under
vacuum, onto the wall of a round-bottomed flask, using a rotary evaporator. The mole ratio
of fluorescent probe to lipid was maintained at 1:350. The lipid film obtained by evaporation
was maintained under vacuum for at least an additional 0.5 h to remove any residual solvent,
hydrated at a 10 mM lipid and 28.6 11M probe concentration in 500 11L of a solution
containing NaCl (0.15 M), MOPS (pH 7.0) (0.01 M) and EDTA (0.1 mM) for 30 min at a
temperature at least 10°C higher than the transition temperature of the lipid (-16.2°C), and
fieeze-thawed ten-times in a solid carbon-dioxide/ethanol bath. The resulting multilamellar
vesicles were passed 29 times through 100 nm polycarbonate filters using a Liposofast
extruder apparatus (Avestin, Ottawa, Canada). An odd number of passages was used to
avoid any contamination of the sample by vesicles which may not have passed through the
filters. The LUVS were characterized by freeze-fracture studies using a scanning electron

microscope as described by MacDonald et al. (1991).

50
2. 3. 3 Fluorescence Experiments

For kinetic measurements, a 20 11L aliquot of the liposome suspension was diluted
to 2 mL in buffer containing 100 mM NaCl and 50 mM Tris-HEPES (pH 7.0) to achieve final
concentrations of 100 11M lipid and 0.286 pM DPH-PA. The suspension was pre—incubated
at room temperature for 5 min with continuous stirring, using a magnetic stirrer to
incorporate oxygen into the experimental system. Another 5 min incubation was done in the
cuvette for temperature equilibration. An aliquot of the stock solution of the chelator to be
tested was added to achieve the desired final concentration in the cuvette. Peroxidation was
initiated by the addition of 20 pL of a 100 11M stock metal ion solution to achieve a final
metal concentration of 1 11M. The control sample did not contain either added metal ions or
chelator. The reaction was monitored over 21 min with a reading being taken at 0 min, 1 min
and every 3 min thereafter.

Cuvette temperature was maintained at 22°C with a circulating water bath. The
cuvette holder was also fitted with a magnetic stirring mechanism to maintain a well dispersed
suspension of vesicles. Fluorescence experiments were conducted using a SLM Instruments,
Model 4800, spectrofluorometer (Urbana, IL) interfaced to a computer with data acquisition
hardware from On-Line Instrument Systems (Bogart, GA). The samples were excited with
polarized light at 384 nm (Slit width 2 nm), and vertical and horizontal components of the
sample fluorescence, In and 1,, emitted through optical filters (KV 418, Schott, Duryea, PA),

were detected in the T-format. The steady-state anisotropy, r,, was calculated as:

rs = (In ”Ill/(11+ 21.)

51

where II and Law the fluorescent intensities of the vertically ( n) and horizontally (.L) polarized
emission when the sample is excited with vertically polarized light (Lakowicz, 1983). Light
scattering by the vesicle suspension was determined by measuring intensity in the absence of
fluorescent probe using vertically polarized excitation light with the emission polarizer set to
55°. The ratio of this signal to that determined for membranes containing the probe gave the
fraction of signal resulting from light scattering. The light scattering fraction was negligible

relative to the fluorescence signal and no corrections were made to the anisotropy values.

2. 3. 4 Preparation of SLPC H ydroperoxides

SLPC hydroperoxides were prepared according to the procedure described by
Miyazawa et al. (1987) and modified by Zhang et al. (1995). Twenty five milligrams of lipid
dissolved in 25 mL of methanol containing 0.1 mM methylene blue were placed in a 250-mL
beaker cooled with ice water and photoirradiated with two 150 watt-lamps for 8 h. After
photooxidation, the methylene blue was removed from the solution by solid phase sequential
extraction using two Supelclean LC-Si 6-mL, 1 g columns (Supelco, Bellefonte, PA). The
resulting hydroperoxide-containing solutions were placed in amber glass vials (Avanti Polar
Lipids), layered with nitrogen, sealed with teflon tape and maintained at -80°C until use. The
concentration of hydroperoxides was determined using the xylenol orange assay developed

by Jiang et al. (1991).

52

2. 3. 5 HPLC-Chemiluminescence Assay and Conjugated Diene Determination

LUVS (2 mM) were suspended in 100 mM NaCl and 50 mM Tris-HEPES (pH 7.0).
Oxidation was initiated with or without addition of chelator to a final concentration of 400
11M, followed by addition of Fe(II) ions to a final concentration of 200 uM. The reaction
mixture of a total volume of 2.5 mL was maintained at 22°C throughout the experiment.
Aliquots of 500 uL were removed at 0 min, 1 min and at 30 min intervals thereafter, during
the 90 min experiment. Lipid was extracted from the LUV suspension by adding 3 volumes
of ice-cold chloroform:methanol (2:1, v/v). The suspension was vortexed for 1 min,
centrifirged at 750 x g for 10 min and the lower chloroform layer was collected. The
extraction procedure was repeated twice. The chloroform fractions were combined, dried
under nitrogen, immediately resuspended in 30 uL of methanol and analyzed by HPLC-
chemiluminescence using a normal phase Supelcosil LC-NH2 (5 pm, 250 x 4.6 mm) column
with a Supelguard LC-NH2 guard column (Supelco, Bellefonte, PA). The mobile phase,
consisting of a 90:10 (v/v) solution of methanol and 10 mM sodium phosphate monobasic
buffer, was adjusted to a final pH of 6.5 and was used at a flow rate of 1 mL/min with
continuous helium sparging. A Waters 991 photodiode array (PDA) detector (Milford, MA)
set at a range of200-240 nm was used to detect the shift from unconjugated diene absorption
at 205 nm to conjugated diene absorption at 234 nm.

The eluate passing through the PDA detector was mixed with a chemiluminescence
reagent at a post-column mixing tee. The luminescence reagent was prepared as described
by Miyazawa et al. (1987). It consisted of cytochrome c (10 ug/mL) and luminol (1 ug/mL)

dissolved in 50 mM nitrogen-sparged borate buffer (pH 10.0) containing 1% methanol to

53
improve mixing with the mobile phase. The reagent was prepared daily, placed in amber glass
containers, degassed and continuously sparged with helium. The flow rate of the reagent was
1 mL/min. The chemiluminescence generated by the presence of hydroperoxides was
detected by a Waters 474 fluorescence detector. The peak areas of the prepared standards
chromatographed under the same conditions as the samples were used to construct the
calibration curves. Separate calibration curves were generated for the quantitation of

conjugated dienes and the hydroperoxides.

2.4 RESULTS

2. 4. 1 Characterization of the LU Vs

Figures 2.1A and 2.1B consist of electron micrographs of replicas of LUVS that were
slarn—fi'ozen, fractured and shadowed. The micrograph in Figure 2. IA confirms that virtually
all the vesicles formed by extrusion contained single bilayers, a critical feature of our
experimental system. Figure 2.1B illustrates the relatively homogenous size distribution of

the vesicles. The average diameter of the LUVS was 80 nm.

2. 4.2 DPH-PA Fluorescence Properties in the LU Vs

Initial experiments were conducted to determine the optimal concentrations of DPH-
PA and the ratio of probe to phospholipid to be used in the LUVS. DPH-PA had negligible
fluorescence in the absence of phospholipids (Figure 2.2A). Fluorescence intensity increased

and reached saturation at 100 pM phospholipid as increasing amounts of LUVS were added

54

Figure 2.1. Freeze-fiacture electron micrographs of the LUVS at a 10 mM concentration in
a solution containing 0.15 M NaCl, 0.01 M MOPS (pH 7.0) and 0.1 mM EDTA. The lengths
of the bars represent 100 nm. These micrographs confirm the unilamellar nature of the

vesicles (A) and their relatively homogenous size distribution (B).

 

55

 

Figure 2.1 A

 

Figure 2.2 B

56

Figure 2.2. Fluorescence intensity (O) and anisotropy (I) of DPH-PA in LUVS as a firnction
of concentration. (A) 2 mL of buffer (100 mM NaCl/SO mM Tris-HEPES, pH 7.0) containing
0.286 11M DPH-PA and various concentrations of phospholipid and (B) 2 mL of buffer

containing 100 pM phospholipid and various concentrations of DPH-PA.

Intensity (A.U.)

Intensity (A.U.)

 

57

   

 

 

0.60 T 0.300
0.50 ~—
0.40 « — 0.250
0.30 «
0.20 ~ ~ 0.200
0.10 ~
0.00 . 1 r 0.150
0 50 100 150
Phospholipid Concentration (pM)
Figure 2.2A
2.0 --

 

0.0

 

l l I I
1 l I I

l
I

0.0 0.2 0.4 0.6 0.8 1.0

DPH-PA Concentration (p M)

Figure 2.23

Anisotropy

58

to a fixed concentration of DPH-PA. The anisotropy values were stable over the
phospholipid concentration range of 20-125 pM. When the phospholipid concentration was
held constant at 100 11M (Figure 2.2B), the fluorescence intensity increased linearly with the
concentration ofDPH-PA at lower concentrations. Above a concentration of 0.4 pM DPH-
PA, the increase in fluorescence became non-linear though saturation was not observed at the
probe concentrations tested. In all subsequent experiments, concentrations of 0.286 11M
DPH-PA and 100 uM phospholipid were used. Under these conditions, the fluorescence
intensity of DPH-PA was a linear function of its concentration in the membrane and the

lowest possible probe to lipid ratio was used, while maintaining a strong fluorescence signal.

2. 4.3 F e(II)-Induced Peroxidation Monitored by Fluorescence Intensity Decay

Figure 2.3 illustrates the effects of chelators on the rates of F e(II)-induced
peroxidation in the LUVS, studied at a 2:1 molar ratio of chelator to metal. The rate of
peroxidation was monitored by quenching of the fluorescence intensity of the probe by the
free radicals generated during lipid peroxidation. The control LUVs containing DPH-PA
showed very stable intensity values over the twenty-one minute time period, indicative of the
stability of the probe in the absence of initiators of peroxidation. When Fe(II) ions were
added to initiate peroxidation in the vesicles, the fluorescence intensity decreased to
approximately 30% of the original value by the end of the twenty-one minute assay. Addition
of EDTA followed by Fe(II) ions resulted in only a small decrease in fluorescence intensity
values of the probe, indicating that EDTA behaved as an antioxidant under these conditions.

However, the other three chelators studied - ADP, NTA and citrate - all acted as

59

 

1.20

   

 

1.00

 

 

 

 

  

01 i-CJ-l E>

 

 

 

o
2
0.30
.8.
2
3 060 ~
E .
0
z
‘5 0.40 F
0
r:
0.20 ~
0.00 ‘ ‘
0 3 6 9 12 15 10 21

Time (min)

Figure 2.3. Peroxidation of 100 pM SLPC LUVS induced by 1 pM Fe(II) and 2 uM
chelator. The rate of peroxidation was monitored by a decrease in fluorescence intensity as
a firnction of time, as described under Experimental Procedures. Relative fluorescence on the
y-axis represents the ratio of the fluorescence intensity after a given time of oxidation versus
the initial fluorescence intensity at time = 0 min. Values represent the mean :1: standard

deviation of triplicate measurements. A, Control; 0, Fe(II); 0, Fe(II)-EDTA; O, Fe(II).
ADP; A, Fe(II).NTA; 0, Fe(II)-Citrate.

6O

prooxidants, as evidenced by a more rapid drop in fluorescence intensity of the probe than
that observed with the Fe(II) ions alone. When the chelators were added to the LUVS in the

absence of metal ions, there was no change in fluorescence intensity values (data not shown).

2. 4. 4 F e(II)-Induced Peroxidation Monitored by Increase in Fluorescence Anisotropy

Fluorescence anisotropy was used to monitor the change in membrane fluidity that
accompanied peroxidation (Figure 2.4). Membrane peroxidation in the LUVS was
accompanied by an increase in the steady-state anisotropy parameter for DPH-PA. This
increase in anisotropy, indicating a reduction of DPH-PA mobility in the lipid bilayer and a
decrease in membrane fluidity, may be attributed to an increase in the molecular order of the
fatty acyl chains in the bilayer (Monahan et al., 1994).

The fluorescent anisotropy values for the control LUVS were stable over the course
of the assay, indicative of a membrane of unchanging fluidity. Similarly, there was no
Significant change in anisotropy in the LUVS in the presence of the EDTA-Fe(II) complex at
a 2:1 molar ratio. However, the chelators ADP, NTA and citrate, when complexed to Fe(II),
all caused a greater increase in anisotropy values than did the F e(II) ions alone, indicating a
substantial loss of membrane fluidity upon peroxidation. These results were consistent with

the time course of the loss of fluorescence intensity upon peroxidation observed in Figure 2.3.

2. 4. 5 F e(III)-Induced Peroxidation of L U Vs
Using either the decrease in fluorescence intensity (Figure 2.5) or the increase in

anisotropy (Figure 2.6) to monitor the progress of membrane peroxidation, Fe(III) ions were

61

 

 

 

 

 

0.24
“l
0.23 — T .:
i 1
T 1- ‘r
I 4 ’
E 0.21 r T "
.9. T T i "3
§ ..
2 0.20 — . . i
4
T J
'r .1.
r
019 — ./ r
-r gég4= 3
<fslr .. u a.
0.18 I l I I I
0 3 6 9 12 15 18 21

Time (min)

Figure 2.4. Peroxidation of 100 pM SLPC LUVS induced by 1 pM Fe(II) and 2 pM
chelator. The rate of peroxidation was followed by an increase in fluorescence anisotropy as
a function of time, as described under Experimental Procedures. Values represent the mean
:t standard deviation of triplicate measurements. A, Control; 0, Fe(II); 0, F e(II)-EDTA; O,
Fe(II)-ADP; A, Fe(II)-NTA; 0, Fe(II)-Citrate.

62

 

 

 

 

 

 

 

 

 

 

 

1.20
8 1.00 a a, ‘1;
c .1.-
8
a
3 I i: 3
o ..
a 0.00 i ..
LL
:1
3
1'5
3
a: 0.60

040 I I I I I I

0 3 6 9 12 15 18 21
Tlme(mln)

Figure 2.5. Peroxidation of 100 pM SLPC LUVS induced by 1 uM Fe(III) and 2 uM
chelator. The rate of peroxidation was monitored by a decrease in fluorescence intensity as
a function of time, as described under Experimental Procedures. Relative fluorescence on the
y-axis represents the ratio of the fluorescence intensity after a given time of oxidation versus
the initial intensity at time = 0 min. Values represent the mean :1: standard deviation of
triplicate measurements. A, Control; 0, Fe(III); 0, Fe(III)-EDTA; O, Fe(III)-ADP; A,
Fe(III)-NTA; 0, Fe(III)-Citrate.

63

 

0.20

 

 

 

 

 

 

>.
IL
9
‘6
.2 T
5 ES .1. .1. .1. .1.
l- : ’ T T T
018 O T T 9 ¢
-L O + 'l
A m m
r * I * ii}
0.17 I I I I
0 3 6 9 12 15 1B 21
Tlme (min)

Figure 2.6. Peroxidation of 100 uM SLPC LUVS induced by 1 11M Fe(III) and 2 pM
chelator. The rate of peroxidation was followed by an increase in fluorescence anisotropy as
a function of time, as described under Experimental Procedures. Values represent the mean
:1: standard deviation of triplicate measurements. A, Control; 0, Fe(III); 0, Fe(III)-EDTA;
o, Fe(III)-ADP; A, Fe(III)-NTA; 0, Fe(III)-Citrate.

64

less effective than Fe(II) ions in inducing peroxidation in the LUVS. Furthermore, the effects
of chelators on the rate of Fe(III)-induced lipid peroxidation were very different from the
F e(II)-induced peroxidation in the LUVS. Of all the chelators examined, it was found that
only the citrate-Fe(III) complex acted as a prooxidant. EDTA, ADP and NTA all acted as
antioxidants in the presence of F e(III) ions, as indicated by the reduction in fluorescence-
intensity decline (Figure 2.5) or by the smaller increase in anisotropy values (Figure 2.6)

compared to the results obtained with Fe(III) alone.

2. 4. 6 Cu(ID-Induced Peroxidation of LU Vs
Under the assay conditions employed in this study, neither the Cu(II) ions alone
(present in the upper, non-reductive state) nor the Cu(II) ions complexed with chelators

initiated peroxidation in the vesicles (data not shown).

2. 4. 7 Effect of Chelator Concentrations

For the chelators that exhibited a prooxidant effect in the presence of Fe (II) or Fe(III)
ions at the initial molar ratios of 2:1 tested, the effects of varying the concentration of
chelators relative to the metal ions was also examined. Figure 2.7 shOws the effects of a 1:1,
5:], 10:1 and 20:1 molar ratio of NTA relative to Fe(II) on rates of peroxidation. NTA
maintained its prooxidant effect at all the ratios tested. Similar results were obtained when
varying ratios of ADP to Fe(II) (Figure 2.8) were evaluated, with ADP enhancing the rates
of LUV peroxidation at all the concentrations tested. Citrate, on the other hand, reversed its

prooxidant effect at higher chelator concentrations in the presence of both F e(II) (Figure 2.9)

65

1.20

 

Relative

 

 

 

 

Trrne (min)

Figure 2.7. Peroxidation of the LUVS induced by varying molar ratios of NTA to Fe(II).
The concentration of metal was kept constant at 1 11M (O) and different concentrations of
NTA were used for final NTA to metal molar ratios of 1:1 (A), 5:1 (e), 10:1 (0) and 20:1
(O). Values represent the mean i standard deviation of triplicate measurements.

66

 

 

 

 

 

 

 

1.20
1.00
g 0.80 r
a
g 0....
O
3
2
g 0.40 ~—
0.20 - A D
"~ A
‘%= fl-k
0.00 1 1 1 1 r "'
0 3 6 9 12 15 18 21

Time (min)

Figure 2.8. Peroxidation of the LUVS induced by varying molar ratios of ADP to Fe(II).
The concentration of metal was kept constant at 1 pM (O) and different concentrations of

ADP were used for final ADP to metal molar ratios oflzl (A), 5:1(0), 10:1 (0) and 20:1 (0).
Values represent the mean i standard deviation of triplicate measurements.

67

 

1.20

/1-1

1.00

 

4+u4

 

 

 

 

_ ‘ T
0.80 \ -
2'. i
u 0.60 ” .
O J.
i
t: 0.40 ”‘
0.20 "
O 00 1 r L 1 r I
O 3 6 9 1 2 1 5 1 8 21

Time (min)

Figure 2.9. Peroxidation of the LUVS induced by varying molar ratios of citrate to Fe(II).
The concentration of metal was kept constant at 1 uM (0) and different concentrations of
citrate were used for final citrate to metal molar ratios of 1 :1 (A), 5:1 (e), 10:1 (0) and 20:1
(O). Values represent the mean i standard deviation of triplicate measurements.

68

 

 

 

 

 

 

 

1.20
1.00
. S" WTKM —————__~"_—_*_u .- a
i I E i ‘
0.80 F t O O Q
t l ,T, T
A .1. I I I G
g 060 l— T T .I %
II - __
.L .L B
o ..
a 1
9
a: 0.40 ‘—
0.20 r
000 1 1 r 1 1 1
0 3 6 9 12 15 18 21
Time (min)

Figure 2.10. Peroxidation of the LUVS induced by varying molar ratios of citrate to F e(III).
The concentration of metal was kept constant at 1 11M (O) and difi‘erent concentrations of
citrate were used for final citrate to metal molar ratios of 1 :1 (A), 5:1 (0), 10:1 (0) and 20:1
(O). Values represent the mean i standard deviation of triplicate measurements.

69

and Fe(III) ions (Figure 2.10). With Fe(II) ions, citrate did not exhibit any antioxidant
activity up to a chelator to metal molar ratio of 5:1. At chelator to Fe(III) molar ratios of
10:1 and 20:] though, citrate strongly suppressed peroxidation in the LUVS, as evidenced by
the small drop in intensity over the assay period. In the presence of Fe(III) ions, citrate

reversed its prooxidant effect at chelator to metal ratios of 5:1 and higher.

2. 4. 8 Validation of the F luorescence-Based Assays

Quantitation of conjugated diene and hydroperoxide formation as a firnction of time
was done by HPLC to establish the validity of the fluorescence intensity and anisotropy
measurements in the model system as assays for lipid peroxidation. The data in Figures 2.1 1
and 2.12 show the amounts of conjugated dienes and phosphatidylcholine hydroperoxides
(PCOOHs), respectively, that were formed during the Fe(II)-induced peroxidation of SLPC
vesicles over a 90 minute time period. The PCOOHS were specifically detected by
chemiluminescence as a single peak eluting at 6.5 min. By UV detection at 234 nm, an
unoxidized PC peak was seen at 3.7 min and an oxidized PC peak at 6.4 min. Formation of
conjugated dienes and PCOOHS were consistent with the peroxidation measurements by
fluorescence. At a chelator to metal molar ratio of 2: 1, citrate caused the greatest
acceleration of F e(II)-induced formation of conjugated dienes and PCOOHS in the LUVS,
followed by NTA and ADP. EDTA strongly suppressed the Fe(II)-induced formation of

conjugated dienes and PCOOHS.

7O

 

2.00

1.50

1.00

0.50

Conjugated Clone: (nrnoles)

 

 

 

 

 

0.00

 

Tlme (mln)

Figure 2.11. Peroxidation of 2 mM SLPC LUVS induced by 200 uM Fe(II) and 400 uM
chelator. The rate of peroxidation was monitored by the measurement of conjugated diene
absorption at 234 nm, as described under Experimental Procedures. Values represent the
mean i standard deviation of triplicate measurements. A, Control; 0, Fe(II); 0, Fe(II)-
EDTA; e, Fe(II)-ADP; A, Fe(II)-NTA; 0, Fe(II)-Citrate.

71

 

 

 

 

 

 

 

 

2.00 11%
a . D
o 1.50 ._
T:
E
5
8
g 1.00
x
2
o
a.
2
1:
£ 0.50
0.00

 

Tlrne (mln)

Figure 2.12. Peroxidation of 2 mM SLPC LUVS induced by 200 11M Fe(II) and 400 pM
chelator. The rate of peroxidation was monitored by quantitation of hydroperoxides
(PCOOHS) using HPLC-chemiluminescence, as described under Experimental Procedures.

Values represent the mean 3: standard deviation of triplicate measurements. A, Control; 0,

72

2.5 DISCUSSION

Fluorescence spectroscopy has previously been used successfirlly to monitor lipid
peroxidation in submitochondrial particles (de Hingh et al., 1995), low-density lipoproteins
(Laranjinha et al., 1992), microsomes (Dinis et al., 1994), human erythrocyte membranes
(McKenna et al., 1991; Van den Berg et al., 1988) and phospholipid vesicles (Kuypers et al. ,
1987). However, the probe used in these studies, cis—parinaric acid, requires special handling,
has a relatively low quantum yield in membranes and is subject to complicating photochemical
and oxidative reactions, including dimerizations (Lentz, 1993). DPH-PA overcomes these
disadvantages, and therefore, offers great potential as a probe for membrane studies.

DPH-PA (Figure 2.13) is an anionic derivative of the parent DPH molecule. In
contrast to DPH which is non-Specifically partitioned in various domains of the membrane
(Ivessa et al., 1988), the polar substituent of DPH-PA provides a surface anchor that ensures
that the polar head of the probe is localized at the lipid-water interface of the membrane, with
the hydrophobic tail portion lying parallel to the acyl chains of the membrane. In addition,
the probe is similar to long-chain free fatty acids both in its anionic carboxylic group and in
its approximate chain length, making it a useful probe for lipid peroxidation studies. The
probe also lacks the extrinsic bulk of some other membrane probes like the n—(9-anthroyloxy)
free fatty acid and pyrene derivatives, and can be expected to cause minimal disturbance to
the lipid bilayer (Trotter and Storch, 1989). In contrast to cis-parinaric acid, DPH-PA is
relatively photostable and requires little Special handling, it exhibits strong fluorescence
enhancement in a lipid environment and offers sensitive fluorescence anisotropy responses to

lipid ordering (Mateo et al., 1991).

73

CH,—CH,— C —OH

\\\

Figure 2.13. Structure of 3-(p-(6-phenyl)-l, 3, 5-hexatrienyl)phenylpropionic acid.

74

These favorable properties of the probe DPH-PA were utilized in developing
fluorescence intensity and anisotropy assays for monitoring lipid peroxidation in membranes.
The basis for the fluorescence intensity assay is the structure of the DPH-PA molecule, its
conjugated double bond system giving rise to its fluorescent properties, as well as to its
susceptibility to peroxidation. The probe reacts readily with a variety of free radicals to yield
non-fluorescent products. Degradation of the probe that accompanies membrane
peroxidation is indicated by a decrease in fluorescence, which can then be measured directly.
In contrast, the fluorescence anisotropy assay developed here follows the changes in
membrane structure that accompany peroxidation. Free radical reactions such as lipid
peroxidation have been implicated in decreased membrane fluidity (Yu et al., 1992; Choe et
al. , 1995). A decrease in membrane fluidity is indicated by an increase in the steady-state
fluorescence anisotropy parameter of the probe.

The two fluorescence-based assays developed in this study were used to explore the
effects of chelators on metal-ion—induced peroxidation. Of the three metal ions tested, Fe(II)
ions caused the greatest acceleration of peroxidation in the liposomes followed by Fe(III)
ions. By contrast, Cu(H) ions were unable to induce peroxidation under the conditions used
in this assay.

These difl‘erences in the abilities of the three metal ions to initiate peroxidation in the
LUVS are probably a result of the differences in accessibility of these oxidizing species to the
site of peroxidation in the membranes. Cu(II) ions have previously been shown to be less
effective at initiating lipid peroxidation in liposomes as compared to micelles (Maiorino et al.,

1995). These researchers found that liposomal peroxidation by Cu(II) occurred after an initial

75

lag phase during which peroxidation did not take place, suggesting that a physical constraint
had to be overcome before massive lipid peroxidation could occur. When the lipid
organization was shifted from bilayer to micellar dispersion, this lag phase was abolished.

The different coordination geometry of Cu(II) and Cu(I) may also affect their ability
to participate in redox reactions of lipid peroxidation. Cu (H) complexes have a square planar
geometry whereas Cu(I) complexes are mostly tetrahedral. On the other hand, both Fe(II)
and F e(III) complexes are almost always octahedral or distorted octahedral. It is thought that
the fact that the two valence States of copper usually have different coordination environments
may impose a kinetic hindrance on redox reactions involving copper (Miller et al., 1990).

Another factor that may explain the ability of iron and the inability of copper to initiate
peroxidation in our system may be the different mechanisms through which these compounds
catalyze lipid peroxidation. The presence of hydroperoxides is essential for a peroxidative
response to copper, suggesting that copper catalyzes the degradation of lipid hydroperoxides
and thus promotes the propagation phase of the peroxidation mechanism (Ding and Chan,
1984). Iron functions in both the initiation and propagation phases of lipid peroxidation
(Minotti, 1993). Since the lipid used in this study was of a very high purity and contained no
detectable amounts of hyroperoxides, it may explain why Cu(II) was unable to initiate
peroxidation in this liposomal system.

The two valence states of iron gave different rates of peroxidation and also showed
different effects in the presence of chelators. F e(II) ions caused greater acceleration of lipid
peroxidation than did the Fe(III) ions. A possible explanation for this difference in activity

for the two metal ions is that only Fe(II) ions are active in peroxide decomposition and

76

although the system is seemingly pure, it is probable that trace peroxides and trace F e(II) are
responsible for much of the small peroxidation seen with Fe(III) ions.

All the chelators examined in this study had a marked effect on the rates of metal-ion-
induced peroxidations. Depending on the concentration of the chelator and on the metal ion
used for peroxidation, these chelators either suppressed or promoted the rates of
peroxidation. EDTA was an effective antioxidant in this system, irrespective of the valence
State of iron used as the initiator of peroxidation. NTA and ADP, on the other hand, behaved
as antioxidants in the presence ofFe(III) ions, but were prooxidants in the presence of F e(II)
ions. They maintained their prooxidant effect even at the higher concentrations of chelators
tested. In contrast, citrate displayed a prooxidant effect in the presence of Fe(II) and F e(III)
at the lower concentrations of chelators tested. However, when the concentration of
chelators used was in high molar excess over that of the metal ions, the prooxidant efl‘ect of
citrate was reversed and it became an effective antioxidant.

This biphasic, paradoxical behavior of citrate may be attributed to the fact that at the
lower concentrations, citrate may be preferentially complexing the oxidized form of iron,
leaving traces of the lower valence form in solution to participate in the Haber-Weiss reaction,
reductively activating peroxides through formation of the hydroxide ion and the extremely
reactive alkoxyl or hydroxyl free radicals (Porter, 1993). This would also explain the
promotion of citrate to antioxidant status, at the higher concentrations, when all of the iron
would be chelated. Since one would expect citrate to be in much greater preponderance to
the metal in a real food or biological system, the antioxidant activity of citrate would be the

predominant one observed.

77

Another factor that may explain the different effects of the chelators studied on the
rates of metal-catalyzed lipid peroxidation may be their effect on the redox potential of the
metal ions. Transition metals have a range of accessible oxidation states that enables them
to transfer electrons. The redox potential for such a transfer is altered by chelation of the
metals (Kanner et al., 1987). The reduction potential of F e(III)/F e(II) (aqueous) is +110 mV
at pH 7.0. The reduction potential of the chelated Fe(III)EDTA/Fe(II)EDTA increases to
+120 mV, whereas those of the chelated F e(III)citrate/Fe(II)citrate and
Fe(III)ADP/Fe(II)ADP complexes decrease to +100 mV at neutral pH (Buettner, 1993).

In general, chelators in which oxygen atoms ligate the metal tend to preferentially bind
to the oxidized forms of iron or copper, thereby decreasing the redox potential of these
metals. In contrast, chelators in which nitrogen atoms primarily bind the metal favor the
reduced forms of iron and copper and tend to increase the redox potential of the metal (Miller
et al., 1990). This alteration of reduction potential influences the reactions in which iron can
participate, which in turn changes the yields of the different reactive oxygen species obtained
during iron autooxidation.

Alteration in the accessibility of the metal ions to the site of peroxidation after
complexation with the different chelators can also influence the ability of the metal to
participate in lipid peroxidation. Since metal ions are present in the aqueous phase and
catalysis of membrane peroxidation must occur at the phase interface or membrane surface,
any chelator which increases the binding of the metal to the membrane will increase the
catalytic effectiveness of the metal (Willson, 1982). Chelators like ADP are known to

increase fatty acid solubilization of the metals, thereby placing the metal at the site of

78

peroxidation and enhancing the rate of lipid peroxidation (Schaich and Borg, 1988). Other
factors like the charge on the chelator and the steric effects of the complex formed can also
influence the rate of peroxidation. The net effect on lipid peroxidation often results from a
complex competition between the individual effects (Schaich, 1992).

Numerous other studies have demonstrated the complexity imparted by chelators on
the reactivity of metals in lipid peroxidation. Chelators have variously been shown to have
a prooxidant, antioxidant or no effect on lipid peroxidation depending on the source and
preparation of tissues, cells, membranes or purified lipids, the solvents used, the
concentrations of metal complexes, the presence or absence of oxygen sources, and also, how
these effects are measured (Schaich, 1992).

This illustrates the need of working with a clean and simple model system that is free
from other complicating factors, in order to better elucidate the molecular bases of these
reactions. The SLPC lipid substrate used in this model system has a composition that is
representative of the phospholipids found in biological membranes, with a saturated fatty acid
at the sn-1 position, an unsaturated fatty acid at the sn-2, and a phosphate-containing polar
group at the sn-3 position (Stanley, 1991). The defined composition and high purity of this
lipid also offered other obvious advantages. Since the model system was free of
contaminating endogenous pro- or antioxidants associated with biologically derived materials
and extreme care was taken to keep the system free from oxygen and metal contaminants until
the time of initiation of oxidation, interpretations of the effects of chelators on membranal

metal-induced peroxidation were subject to fewer ambiguities.

79

The results of the fluorescence studies characterizing the effects of chelators on
F e(II)-induced peroxidation were compared with the rates of PCOOH and conjugated diene
formation, for validation of the fluorescence assays. The HPLC data showed that both the
hydroperoxides and conjugated dienes increased in concert with the rates of decrease in
fluorescence intensity and membrane fluidity. This was strong evidence to support our
hypothesis that the fluorescence changes observed upon the initiation of peroxidation stem
from the metal-ion-induced generation of free radicals.

The fluorescence assays are especially usefirl in monitoring the initial stages of lipid
peroxidation. The advantages of the fluorescence assays include extremely high sensitivity
(< 1 pmol of pure lipid is required to run an assay), very small concentration of probe
required relative to lipid (ratio of probe molecules to lipid molecules is 1:350) so that the
overall composition of the membrane is not significantly altered, and the low quantum yield
of DPH-PA in water, assuring that the measured fluorescence is emanating fiom within the
membrane and not from the surrounding aqueous environment. In addition, these assays offer
the speed of other accelerated tests, while avoiding the artifacts introduced by the high
temperature-abuse conditions of other accelerated tests.

The complexity and variability of food systems and biological materials precludes the
use of any single method to define antioxidant activity. However, the experimental system
developed here offers a promising alternative as a screening tool to evaluate antioxidant
efficacy of purified compounds or plant extracts on membrane lipids prior to long-tenn

studies in foods or in biological systems.

CHAPTER 3

STRUCTURE-ACTIVITY RELATIONSHIPS FOR ANTIOXIDANT

ACTIVITIES OF A SERIES OF FLAVONOIDS IN A LIPOSOMAL SYSTEM

3.1 ABSTRACT

Structurally diverse plant phenolics were examined for their abilities to inhibit lipid
peroxidation induced either by Fe(II) and Fe(III) metal ions or by azo-derived peroxyl radicals
in a liposomal membrane system. The antioxidant abilities of the flavonoids were compared
against coumarin and the widely used synthetic antioxidant tert-butylhydroquinone (TBHQ).
The antioxidant eflicacies of these compounds were evaluated using a previously developed
fluorescence spectroscopic assay based on their abilities to inhibit the fluorescence intensity
decay of an extrinsic probe 3—(p-(6-phenyl)-1,3,5-hexatrienyl)phenylpropionic acid (DPH-PA)
caused by the free radicals generated during lipid peroxidation. All the flavonoids tested
exhibited higher antioxidant eflicacies against metal-ion-induced peroxidations as opposed
to peroxyl-radical-induced peroxidation, indicating that metal chelation plays a larger role in
determining the antioxidant activities of these compounds than has previously been believed.
Distinct structure-activity relationships were also revealed for the antioxidant abilities of the

flavonoids. Presence of hydroxyl substituents on the flavonoid nucleus enhanced activity,

80

81

whereas substitution by methoxy groups diminished antioxidant activity. Substitution patterns
on the B-ring were especially important towards antioxidant potencies of the flavonoids. In
cases where the B-ring could not contribute to the antioxidant activities of flavonoids,
hydroxyl substituents in an catechol structure on the A-ring were able to compensate and

become a larger determinant of flavonoid antioxidant activity.

3.2 INTRODUCTION

Apart from the fat-soluble tocopherols, the most common and active antioxidant
compounds naturally occurring in foods are the flavonoids, possessing activity in both the
hydrophilic and lipophilic systems (Kuhnau, 1976). In recent years, there has been a renewed
interest in investigating the many positive pharmacological properties of flavonoids. Much
of this interest in the bioactivity of flavonoids has been spurred by the dietary anomaly
referred to as the “French Paradox”, the apparent compatibility of a high fat diet with a low
incidence of coronary atherosclerosis (Renaud and De Lorgeril, 1992). It was suggested that
the polyphenolic substances such as flavonoids in red wine may provide protection against
coronary heart disease. In vitro (Frankel et al., 1993; Pace-Asciak et al., 1995) and in vivo
(Fuhrman et al., 1995) studies confirmed the inhibition of oxidation of low density
lipoproteins by the phenolic substances in red wine.

The mechanism of this protective action of the flavonoids is a subject of considerable
debate. As polyphenolic compounds, flavonoids have the ability to act as antioxidants by a
free radical scavenging mechanism (Cotelle et al. , 1992; Hanasaki et al. , 1994; Heilman et al. ,

1995; Montinsinos et al., 1995) with the formation of less reactive flavonoid phenoxyl

82
radicals (Equations 1 and 2); on the other hand, through their known ability to chelate
transition metals (Afanas’ev et al., 1989; Morel et al., 1993; Laughton et al., 1991;
Thompson and Williams, 1976; van Acker et al., 1996a), these compounds may inactivate
iron ions through complexation, thereby suppressing the superoxide-driven F enton reaction
(Equations 3 and 4), which is currently believed to be the most important route to active

oxygen species (Afanas’ev et al., 1989).

R00 + Fl-OH _. ROOH + Fl-O' (1)
Ho- + Fl-OH -+ H20 + Fl-O‘ (2)
02-: + Fe(III) _. 02 + Fe(II) (3)
Fe(II) + H202 —+ Fe(III) + HO + HO’ (4)

There is much discussion regarding the relative contributions of these two mechanisms
(van Acker et al., 1996a). It is widely believed that the antioxidant ability of flavonoids
resides mainly in their ability to donate hydrogen atoms, thereby scavenging the fi'ee radicals
generated during lipid peroxidation. Despite the early realization by researchers that the
structures of these compounds allow them to form heavy metal complexes (Kuhnau, 1976),
metal chelation has generally been regarded to play a minor role in the antioxidant activity of
these compounds and so, has not been studied much by researchers in the area (Morel et al.,
1994).

In the present study, we assessed the abilities of a range of structurally diverse

flavonoids to inhibit lipid peroxidation induced by Fe(II) and Fe(III) ions. In addition, we

83

examined the abilities of these compounds to scavenge peroxyl radicals generated by the
aqueous-phase azo compound 2,2'-azobis(2-amidinopropane) dihydrochloride (AAPH).
Quercetin, the most abundant flavonol in fi'uits and vegetables (Hollman et al., 1995), was
one of the flavonoids examined in the study (Figure 3.1). Since flavonoids principally enter
the diet as glycosides (Hollman et al. , 1996), nrtin, a glycoside of quercetin, was also included
in the study as it represents the naturally occurring form of quercetin. The other commonly-
occuning, stnrcturally-related flavonoids investigated were luteolin, chrysin, naringenin and
hesperetin. In addition, a series of seven flavones with different hydroxyl and methoxy
substitution patterns were evaluated to better elucidate the structural criteria necessary for
a high antioxidant activity of these compounds (Figure 3.2). The antioxidant activities of the
flavonoids were compared against two other compounds: the structurally-related coumarin
which is sometimes referred to as a neoflavonoid and tert-butylhydroquinone (TBHQ), a
widely used food antioxidant .

In a preceding chapter (Chapter 2), we described the development of a fluorescence-
based assay utilizing the flourescent probe 3-(p-(6-phenyl)«1,3,5-hexatrienyl)phenylpropionic
acid (DPH-PA) for monitoring lipid peroxidation. The basis of the assay is provided by the
structure of DPH-PA, its conjugated double bond system giving rise to its fluorescent
properties as well as a susceptibility to peroxidation. Peroxidation of the probe molecule
results in a decrease in fluorescence, which can then be measured directly and sensitively.
Another aim of this study was to test the effectiveness of this new method for its ability to

discriminate between antioxidant activities of structurally related compounds.

84

   

R = OH Quercetin R, = H. R: = OH Narinsenjn
R = O-rutinose Rutin R1 = OH, R: = OCH; Hesperetrn
R = H Luteolin

 

Chrysin

OH

OH

TBHQ

Figure 3.1. Structures of the commonly-occurring flavonoids and the other antioxidants,
coumarin and TBHQ, included in this study.

85

 

 

 

 

 

 

 

 

 

 

 

 

 

Number
1
2
3 OH OH H OCH,
4 OCH, H H OCH,
5 H OH H OCH,
6 OCH, OCH, H OCH,
I 7 H OCH, H 0%

 

 

Figure 3.2. Substitution patterns for the series of flavones examined for their antioxidant

activity.

86

3.3 EXPERIMENTAL PROCEDURES

3. 3.1 Materials

Synthetic 1-stearoyl-2-linoleoyl-sn-glycero-3 -phosphocholine (SLPC) (Avanti Polar
Lipids, Alabaster, AL) was the lipid substrate used in the study. Its purity was confirmed by
silica gel TLC using two different solvent systems (chloroform : methanol : water :: 65 : 25
:4; chloroform : methanol : ammonium hydroxide :: 65 : 25 : 4). The lipid stock solutions,
dissolved in chloroform, were maintained at -20°C in amber glass vials that were layered with
nitrogen. The fluorescent probe 3-(p-(6-phenyl)-1,3,5-hexatrienyl)phenylpropionic acid
(DPH-PA) was obtained from Molecular Probes (Eugene, OR). Quercetin and rutin were
obtained from Sigma Chemical Co. (St. Louis, MO). The other flavonoids were purchased
from Indofine Chemical Company, Inc. (Somerville, NJ). 2,2'-Azobis(2-amidinopropane)
dihydrochloride (AAPH) was fi’orn Wako Chemical Company (Richmond, VA). F eC12‘4HZO
and FeCl,'6H,O of greater than 99% purity were obtained from Aldrich (Milwaukee, WI) and
stored at -20°C under anaerobic conditions. All other reagents were of the highest grade
available.

The DPH-PA and flavonoid stock solution were prepared in N,N-dimethylfonnamide
and dimethyl sulfoxide (DMSO) respectively. The iron and AAPH stock solutions were
prepared immediately before use in nitrogen-sparged, double-distilled, deionized water and
maintained on ice under anaerobic conditions. To remove metal contaminants, all the
glassware was acid-washed prior to use and the salt and buffer solutions maintained in a

chelating resin Chelex 100 (sodium form, mesh 50-100, Bio-Rad, Richmond, CA).

87

3. 3.2 Preparation of Large Unilamellar Vesicles

LUVS containing the probe DPH-PA were prepared fi'esh for each day of
experimentation using the extrusion procedure of MacDonald et al. (1991). Briefly, a mixture
containing 5 pmol of SLPC and 15 nmol of DPH-PA was dried under vacuum onto the wall
of a 15 mL round-bottomed flask. The resulting lipid film was hydrated in 500 11L of a
solution containing 0.15 M NaCl, 0.01 M MOPS (pH 7.0) and 0.1 mM EDTA. After 10
freeze-thaw cycles using a dry ice/ethanol bath, the suspension was passed 29 times through
two stacked polycarbonate filters (pore size 100 nm) using a Liposofast extruder apparatus

(Avestin, Ottawa, Canada).

3. 3.3 Antioxidant Evaluation of Flavonoids

The fluorescent intensity assay described in Arora and Strasburg (1997) was used to
evaluate the antioxidant eflicacy of the flavonoids. In the assay, the peroxidative degradation
of the probe, indicated by a decrease in its fluorescence, is used to monitor the sensitivity of
the membrane towards oxidative stress. A 20 pL aliquot of the LUV suspension was diluted
to 2 mL in a buffer containing 100 mM NaCl and 50 mM Tris-HEPES (pH 7 .0) to achieve
final concentrations of 100 pM lipid and 300 nM probe in the cuvette. After a 5 min pre-
incubation at room temperature with continuous stirring, the quartz cuvette containing the
suspension was transferred to a thennostatted cuvette holder (3 7°C) in the spectrofluorometer
and incubated for an additional 10 min for temperature equilibration. An aliquot of the

flavonoid stock solution, dissolved in DMSO, was added to yield a final concentration of 10

88

uM. The volume of added stock solution never exceeded 0.3 % of the total volume; under
these conditions, DMSO had no effect on the rate of peroxidation.

Following a 5 min incubation to allow partitioning of the flavonoid into the membrane,
peroxidation was initiated by addition of 20 pL of 1 mM stock FeCl2 or F eCl3 solutions or
40 pL of stock AAPH solution for final concentrations of 10 pM for FeCl2 and FeCl,, and 10
mM for the AAPH. The control samples did not contain any peroxidation initiator or
flavonoid to be tested. The decay in fluorescence intensity was monitored over 21 min, with
readings being taken at 0 min, 1 min and every 3 min thereafter. The fluorescent experiments
were conducted using a SLM Instruments, Model 4800, spectrofluorometer (U rbana, IL) with
data acquisition hardware and software from On-Line Instrument Systems (Bogart, GA). The
samples were excited with light of 384 nm, and the emitted light was passed through optical

filters (KV 418, Schott, Duryea, PA), prior to detection.

3.4 RESULTS

3. 4. 1 Effects of Flavonoids on Rates of F e(II)-Catalyzed Peroxidation in LU Vs

Figure 3.3 illustrates the inhibitory effects of the commonly-occurring flavonoids, the
neoflavonoid coumarin and the commercial food antioxidant TBHQ on rates of F e(II)-
induced lipid peroxidation. The antioxidant efiicacies of these compounds were evaluated by
their degrees of inhibition of the decay in fluorescence intensity of the extrinsic probe DPH-
PA. The abilities of the flavonoids to inhibit rates of Fe(II)-induced peroxidation varied

widely among the compounds tested, with the degree of inhibition depending primarily on the

89

Figure 3.3. Effects of the commonly-occurring flavonoids, coumarin and TBHQ on rates
of F e(II)«induced peroxidation in the LUVS at 37°C. Peroxidation was initiated in the LUVS
containing 100 pM lipid, 300 nM probe and 10 pM test compound suspended in 2 mL of
buffer (100 mM NaCl, 50 mM Tris-HEPES, pH 7.0) by addition of Fe(II) for a final
concentration of 10 11M. The rate of peroxidation was monitored by a decrease in
fluorescence intensity as a function of time. Relative fluorescence represents the fluorescence
intensity after a given time of oxidation over the fluorescence intensity at time = 0 min.

Values represent the means of triplicate measurements.

an 2.5.“.

 

90

 

 

 

 

3.5 we:
FN or mp mp a o n c
L L L L A _ _ 86
Em“. IT
£68300 II
n . . .. on:
Eat—Lo 101 - .,
53233: 1x1 : 9.:
:Eomctmz LT
5.093 I - .. .- 86
arm... I
:33“. IT .. 86
£5925 for w - d A m .. .
“lull-II :lrIrlllrlr -. I1”
111. r 8..
63:00 I

 

 

 

 

eoueoseronl :1 eArrelea

91

stnrcture of these compounds. The flavon-3 -ol quercetin, with five hydroxyl groups, was the
superior antioxidant under these conditions. Its corresponding glycoside, the flavone rutin,
was also extremely efi‘ective in this system. Both quercetin and rutin possessed antioxidant
activities that were comparable to the potent food antioxidant TBHQ. Luteolin, a flavone
with 4 hydroxyl groups, also demonstrated a high antioxidant activity under these conditions.
Next, in decreasing order of effectiveness, were the two flavanones naringenin and hesperetin.
Of the two, naringenin with three hydroxyl groups was a superior antioxidant to hesperetin
which contains a methoxy group in addition to the three hydroxyl groups. The flavone
chrysin, which lacked any hydroxyl substitution on the B ring, possessed minimal antioxidant
activity against Fe(II)«induced peroxidation. Coumarin, which lacks a B ring completely, also
exhibited negligible antioxidant activity.

A series of flavones with different substitution patterns were also tested to precisely
define the chemical features required for a high antioxidant activity of these compounds. In
the Fe(II)-catalyzed system, only 2 was effective as an antioxidant (Figure 3.4). 3 also
inhibited the F e(II)-induced LUV peroxidation, although to a much lesser degree as compared
to 2. The other five flavones examined in this study were quite ineffective under these
conditions. All these flavones lacked hydroxyl groups on the B-ring. In addition, with the

exception of 2, they all contained at least one methoxy group.

3. 4. 2 Effects of F lavonoids on Rates of F e(III)-Catalyzed Peroxidation in the LU Vs
Similar trends were observed for antioxidant activity of these compounds when Fe(III)

ions were used to catalyze lipid peroxidation. Again, among the commonly-occurring

92

Figure 3 .4. Effects of the series of seven flavones on rates of Fe(II)-induced peroxidation
in the LUVs at 37°C. Peroxidation was initiated in the LUVs containing 100 uM lipid, 300
nM probe and 10 uM flavone suspended in 2 mL of buffer (100 mM NaCl, 50 mM Tris-
HEPES, pH 7.0) by addition of Fe(II) for a final concentration of 10 uM. The rate of
peroxidation was monitored by a decrease in fluorescence intensity as a function of time.
Relative fluorescence represents the fluorescence intensity after a given time of oxidation over
the fluorescence intensity at time = 0 min. Values represent the means of triplicate
measurements. For a description of the chemical structures of the flavones, refer to Figure

3.2.

an 23E

 

 

93

 

 

 

 

:55 as:
«N ow mp up a o n o
econ II +1
I
N .. owd
0 I
m .. ovd
V +
..on
m lxl
N I? . r and
r . /
.9250 I o: P

 

 

 

eoueoseronl :1 01019193

94

flavonoids, the flavonol aglycone quercetin was the most efl‘ective antioxidant in the liposomal
system used in the study (Figure 3.5). Substitution of the 3-OH group of quercetin with the
disaccharide rutinose in rutin only led to a minimal drop in activity. Luteolin was also an
excellent inhibitor of Fe(III)-catalyzed LUV peroxidation. These three compounds had
superior antioxidant activities in comparison to TBHQ. Similarly, as was observed with the
Fe(II) ions, naringenin appeared to be slightly more effective as an antioxidant than
hesperetin, whereas chrysin and coumarin were very ineffective in this system.

The series of seven flavones tested were less effective at inhibiting Fe(III)-catalyzed
LUV peroxidation than the flavonoids shown above (Figure 3.6). 2 was the only flavone in
this series that possessed an appreciable amount of antioxidant activity. 3 possessed a
minimal amount of antioxith activity. The other five flavones studied did not inhibit the

Fe(III)-induced lipid peroxidation to any appreciable extent.

3. 4.3 Effects of Flavonoids on Rates of AAPH-Induced Peroxidation in the LU Vs
Next, as opposed to metal-catalyzed peroxidation, the aqueous-phase, free-radical-
generator AAPH was used to initiate peroxidation and the inhibitory effects of the flavonoids
was studied. All the flavonoids were less effective at inhibiting AAPH-induced peroxidation
than the metal-ion-induced peroxidations (Figure 3.7). TBHQ was superior to all the
flavonoids studied at scavenging AAPH-induced peroxyl radicals. Among the set of
commonly-occurring flavonoids, quercetin was more effective than the other compounds
tested. Rutin, luteolin, naringenin and hesperetin possessed only marginally higher antioxidant

activities than chrysin and coumarin.

95

Figure 3.5. Effects of the commonly-occurring flavonoids, coumarin and TBHQ on rates of
F e(III)-induced peroxidation in the LUVs at 37°C. Peroxidation was initiated in the LUVs
containing 100 uM lipid, 300 nM probe and 10 uM test compound suspended in 2 mL of
buffer (100 mM NaCl, 50 mM Tris-HEPES, pH 7.0) by addition of Fe(II) for a final
concentration of 10 uM. The rate of peroxidation was monitored by a decrease in
fluorescence intensity as a function of time. Relative fluorescence represents the fluorescence
intensity after a given time of oxidation over the fluorescence intensity at time = 0 min.

Values represent the means of triplicate measurements.

96

md 0.59“.

 

 

 

253+

:_._mE:oo .6:
:3ch I?
£5.38: le
EcooEEz LT
Gimp IT
5.023 Ion
.53. If.
£50.30 In:

.9250 +

 

 

 

 

 

 

 

 

 

 

 

 

 

 

AEEV 95...
pm 2 2. up a o n
u . F . _ . etc
and
n_ 1T
[i lT
ii if
.“lvi . ,. 1r OQ.O
«H UH. " u
all llT HO” ”O”
m ‘IT I / / w
u/lfl/
? if . i Ll ? r 8..

 

 

aouaoselonld aAgtelea

97

Figure 3.6. Effects of the series of seven flavones on rates of Fe(III)-induced peroxidation
in the LUVs at 37°C. Peroxidation was initiated in the LUVS containing 100 uM lipid, 300
nM probe and 10 uM flavone suspended in 2 mL of buffer (100 mM NaCl, 50 mM Tris-
HEPES, pH 7.0) by addition of Fe(II) for a final concentration of 10 uM. The rate of
peroxidation was monitored by a decrease in fluorescence intensity as a function of time.
Relative fluorescence represents the fluorescence intensity after a given time of oxidation over
the fluorescence intensity at time = 0 min. Values represent the means of triplicate
measurements. For a description of the chemical structures of the flavones, refer to Figure

3.2.

98

 

 

 

 

 

 

 

 

 

 

 

eoueoselom :1 eAgte|ea

MHHME’
T /
l I
//
J. //
l //
l //
//
a a 3.-

15 18 21

12

Time (min)

Figure 3.6

99

Figure 3.7. Effects of the commonly-occurring flavonoids, coumarin and TBHQ on rates of
AAPH-induced peroxidation in the LUVs at 37°C. Peroxidation was initiated in the LUVs
containing 100 uM lipid, 300 nM probe and 10 uM test compound suspended in 2 mL of
buffer (100 mM NaCl, 50 mM Tris-HEPES, pH 7.0) by addition of AAPH for a final
concentration of 10 mM. The rate of peroxidation was monitored by a decrease in
fluorescence intensity as a fimction of time. Relative fluorescence represents the fluorescence
intensity after a given time of oxidation over the fluorescence intensity at time = 0 min.

Values represent the means of triplicate measurements.

100

fin 8:9“.—

 

 

 

 

 

 

Eé we;
a 3 3 9 o o a o
F p p p p h h O‘eo
Im<< III
atmeaoo I9:
535 I?
.. omd
c3833.... IXI I,
Ecomémz LT -Illlllll... HUI .llll ,,
mi N l will/Ill- :l/f,
OImn—ullu L. L. l” w a/ r- omd
5.85.. +
:33”. In!
I i
£02030 lml o3

.0950 lel

 

 

 

 

GOUGOSSJOl'lL-J aime'ea

101

The series of seven flavones studied also demonstrated lower effectiveness against
AAPH-induced peroxidation (Figure 3.8). In this series, 2, 3 and 5 were moderately effective
at scavenging AAPH-induced peroxyl radicals. l, 4, 6 and 7 possessed minimal antioxidant

activities against AAPH-induced peroxidation.

3.5 DISCUSSION

In the present study, the protective effects of a range of structurally diverse flavonoids
against aqueous-phase generators of LUV peroxidation were assessed. The aim was to
compare the efl‘ectiveness of these compounds at inhibiting metal-ion-induced peroxidations
versus peroxidations induced by the water-soluble fiee radical generator AAPH. We also
sought to ascertain the relationship between chemical structure of the flavonoids and their
antioxidant activity. Another goal was to test the abilities of the previously developed
fluorescence assay to distinguish between antioxidant activities of structurally similar
compounds. The flavonoids tested in this study demonstrated vastly different inhibitory
effects on rates of lipid peroxidation in liposomes. In general, all the compounds examined
were more effective at inhibiting metal-ion-induced peroxidation than AAPH-induced
peroxidation.

Quercetin, a flavon-3-ol, was the most effective antioxidant in this system, indicating
the importance of the hydroxyl group at the C-3 position for a high antioxidant activity.
Replacement of the hydroxyl group at the 03 position of quercetin by the disaccharide
rutinose in rutin led to some drop in activity; however, the drop in activity was very small.

Interestingly, rutin was still more effective than the flavone luteolin which has the same

102

Figure 3.8. Effects of the series of seven flavones on rates of AAPH-induced peroxidation
in the LUVs at 37°C. Peroxidation was initiated in the LUVs containing 100 uM lipid, 300
nM probe and 10 pM flavone suspended in 2 mL of buffer (100 mM NaCl, 50 mM Tris-
HEPES, pH 7.0) by addition of AAPH for a final concentration of 10 mM. The rate of
peroxidation was monitored by a decrease in fluorescence intensity as a function of time.
Relative fluorescence represents the fluorescence intensity after a given time of oxidation over
the fluorescence intensity at time = 0 min. Values represent the means of triplicate

measurements. For a description of the chemical structures of the flavones, refer to Figure

3.2.

103

 

 

 

 

 

E a
8FN€OVIOCON§
liiiiili'l'
4L
lr
.8. 2::- 2

 

 

 

 

aaueosejonu among

15 18 21

12

Time (min)

Figure 3.8

104

substitution pattern as quercetin and rutin, with the only difference being the presence of a
hydrogen at the 03 position.

These small differences in antioxidant activity based on difl‘erences in substitution at
the C-3 position are probably a result of the planarity of the molecule. van Acker et al.
(1996b) determined that the torsion angle of ring B with the rest of the flavonoid molecule
was correlated with scavenging activity of the flavonoids. In quercetin, the 3-OH moiety
anchored the position of the ring B in the same plane as rings A and C via hydrogen-bonds,
resulting in a torsion angle of ring B with the rest of the molecule of close to 0° and thereby
yielding a completely planar molecule. This extra hydrogen bond formed as a result of the
presence of the 3-OH group had a large stabilizing effect on quercetin. The researchers
further demonstrated that removal of the 3-OH moiety from the flavonoids induced a torsion
angle of approximately 20° of ring B with the rest of the molecule and decreased the
conjugation. In mtin, the sugar moiety caused some loss of coplanarity of ring B with the rest
of the flavonoid. Similarly, in luteolin, replacement of the 3-OH with a hydrogen atom and
the subsequent loss of a hydrogen bond caused a slight twist of the ring.

Hendrickson et a1. (1994) investigated the electrochemical properties of quercetin,
rutin and luteolin and determined that their initial oxidation at a glassy carbon electrode was
a 2e’-2H+ process yielding an o—quinone species. The o-quinone further underwent
homogenous rearrangement by at least two processes: one zero-order and the other pseudo
first-order. The zero-order process was likely an intra-molecular rearrangement involving the
substituent at C-3. The catechol group on these compounds was oxidized to quinone which

further underwent intra-molecular rearrangement followed by keto-enol tautomerization.

105
This zero-order process was strongly dependent on the leaving group at C-3; with the poorer
the leaving group at this position, the slower the rearrangement kinetics of the
electrochemically generated intermediate. Quercetin with a proton attached to the oxygen
at C-3 exhibited the fastest zero-order kinetics. Rutin with a disaccharide attached to this
oxygen exhibited slower zero-order kinetics. Luteolin, being unsubstituted at 03, could not
react by this mechanism and exhibited the highest degree of reversibility.

As a group, quercetin, rutin and luteolin exhibited higher antioxidant activities than
the rest of the flavonoids studied. This provided evidence for the highly significant
contribution of the 3',4'-dihydroxy (catechol) substitution pattern on the B-ring towards the
antioxidant activity of these compounds. The presence of a o-dihydroxy (catechol) structure
in the B-ring confers a higher degree of stability to the flavonoid phenoxyl radicals by
participating in electron delocalization and is, therefore, an important determinant for
antioxidative potential (Bors er al. , 1990).

The flavanones naringenin and hesperetin had moderate antioxidant activities and were
not as effective as the flavones quercetin, rutin and luteolin. Presence of the 2,3-double bond
in conjugation with a 4-oxo group has previously been shown to be important for antioxidant
activities of these compounds (Bors et al., 1990; Rice-Evans er al., 1995; Rice-Evans and
Miller, 1996) and saturation of the 2,3-double bond is believed to cause a loss of antioxidative
potential. However, our study could not provide conclusive evidence for the requirement of
a 2,3-double bond. The lower antioxidant activities observed for the flavanones naringenin
and hesperetin as compared to the flavones quercetin, rutin and luteolin appeared to be related

more to the substitution pattern on the B-ring rather than to the presence of a 2,3-double

106

bond. Naringenin, with a hydroxyl group at 04', had a higher antioxidant activity as
compared to hesperetin, which carries a hydroxyl group at the 03' position and a methoxy
group at the C-4' position. This reiterated the importance of substitution patterns on the B-
ring where electron-withdrawing substituents like hydroxyl groups enhanced antioxidant
activity for the flavonoids whereas electron-donating groups like methoxy groups decreased
antioxidant activity. Presence of electron-donating or withdrawing groups at the aromatic
system has previously been shown to strongly influence the redox potential of phenols
(Steenken and Neta, 1982). This altered reduction potential of the flavonoids has an effect
on the lipid peroxidation reactions in which they can participate which in turn, alters their
ability to scavenge deleterious oxy radicals.

The flavone chrysin, which lacks hydroxyl substitution at the B-ring and coumarin,
which lacks a B-ring, demonstrated negligible activity in this system. This is flirther proof
that the antioxidant activity of these compounds is governed by the position and number of
hydroxyl groups on the B-ring. Hydroxyl groups on the A-ring did not appear to be
significant contributors to antioxidant activity in these compounds. These results are in line
with other studies that suggested that hydroxyl groups associated with the A-ring or the pyran
ring of flavonoids had poor reactivity towards peroxyl radicals and were not significant
contributors towards their antioxidant activity (Roginsky et al., 1996). van Acker et al.
(1996b) examined spin distributions on flavonoids and determined that, upon oxidation of the
flavonoid molecule, an event occurring mainly in the B-ring, almost all the spin remained in

the B-ring, even in the case of a completely conjugated molecule. This agreed with the

107

hypothesis that ring B mainly determined the antioxidative potential of these compounds and
that the rest of the flavonoid molecule only had a small influence.

The experiments conducted with the series of seven flavones extended these findings.
Interestingly, regardless of the initiator of peroxidation employed, 2 was the most effective
of this set of compounds. Though 2 lacked any hydroxyl substitution on the B-ring, it carried
two hydroxyl groups on the A-ring. Unlike chrysin, where the two hydroxyl groups on the
A-ring did not form a catechol structure and the compound did not display any significant
antioxidant activity, the hydroxyl groups on 2 were at the C-7 and C-8 positions, thereby
forming a catechol structure. This surprisingly high activity for 2 was quite unexpected and
suggested that, in cases where the B-ring could not contribute to the antioxidant activity of
the flavonoids, certain substituents on the A-ring may be able to compensate and become a
larger determinant of the antioxidant efficacy of these compounds.

Of the other flavones, 3 showed some antioxidant activity, whereas, none of the
other compounds studied inhibited the liposomal peroxidation to any considerable extent. All
these flavones shared one common feature - absence of any hydroxyl groups on the B-ring
and presence of at least one methoxy group on the B-ring. This, too, confirmed our earlier
observations that electron-withdrawing groups at the B-ring enhanced activity of these
compounds while electron-donating groups suppressed it. This is probably a consequence
of the altered spin-distribution on the molecule. Extensive spin delocalization would enhance
activity through an increased stabilization of the flavonoid phenoxyl radical (Mayouf and

Lemmetyinen, 1993).

108

Another striking observation of the study was that all the flavonoids investigated were
more effective at inhibiting the metal-ion-induced peroxidation than the peroxidation
catalyzed by AAPH-induced peroxyl radicals. This is despite the fact that flavonoids possess
lower reduction potentials than peroxyl radicals and can therefore inactivate these damaging
oxyl species and prevent the deleterious consequences of their reactions (Jovanovic et al. ,
1994). This suggests that metal chelation may play a larger role in determining the
antioxidant activity of these compounds than has previously been believed. The higher
antioxidative potentials of flavonoids against metal-ion-induced peroxidations are probably
a consequence of their combined metal chelating and free radical scavenging abilities. In the
AAPH-induced peroxidations, however, only the flu radical scavenging mechanism
contributes towards the inhibitory action of these compounds, thereby accounting for the
lower antioxidant potencies.

In an earlier study, the inhibitory effects of quercetin and rutin on Fe(II)-ion-
dependent and independent lipid peroxidation were examined to elucidate the chelating and
free radical scavenging activities of these compounds (Afanas’ev et al., 1989). Both
flavonoids were significantly more effective at inhibiting Fe(II)~dependent lipid peroxidation
due to their ability to chelate iron ions with the formation of inert complexes unable to initiate
peroxidation. Additionally, the iron complexesof flavonoids retained their free radical
scavenging activities. Morel et al. (1993) studied the antioxidant and iron-chelating actions
of flavonoids on iron-loaded rat hepatocyte cultures. They directly demonstrated that the

flavonoids possessed good chelating activities and were capable of removing iron already

109

inside the hepatocytes, thereby suggesting that the metal-chelating mechanism was a
significant contributor to the antioxidant mechanism of these compounds.

In our study, another possible explanation for the higher antioxidant activities
observed against metal-ion-dependent peroxidation versus AAPH-induced peroxidation may
be that the aqueous-phase, AAPH-derived-peroxyl radicals, being significantly bulkier than
the radicals generated by the metal ions, may not be able to localize into the same region of
the membrane as the flavonoids and hence may not be accessible to the flavonoids for
scavenging. Determination of the precise localization of flavonoids in the membrane bilayer
is the subject of a later chapter in the dissertation. The results from that study indicate that
the flavonoids preferentially partition into the hydrophobic core of the bilayer, a region that
may be different than that in which the AAPH-derived peroxyl radicals localize.

Figure 3.9 summarizes the structural criteria that enhanced the antioxidant activity of
flavonoids. Substitution patterns on the B-ring were found to be the most important
contributor to the antioxidant activity of flavonoids. Hydroxyl groups boosted the antioxidant
activity; whereas methoxy groups suppressed antioxidant activity. Presence of o-dihydroxy
substitution on the flavonoids substantially enhanced the antioxidant activity of these
compounds. A hydroxyl group at the C-3 position was also beneficial to the ability of
flavonoids to inhibit lipid peroxidation.

This study also confirms the sensitivity of the fluorescence-based assay previously
developed in our laboratory in evaluating plant-derived compounds for antioxidant activity.
The simple and rapid assay can be used to characterize structure-activity relationships within

a class of structurally related compounds.

110

 

Figure 3.9. Structural criteria that enhance the antioxidant activity of flavonoids.

CHAPTER 4

ANTIOXIDANT ACTIVITIES OF ISOFLAVONES AND

THEIR BIOLOGICAL METABOLITES IN A LIPOSOMAL SYSTEM

4.1 ABSTRACT

Genistein and daidzein, the two major soy isoflavones, principally occur in nature as
their glycosylated or methoxylated derivatives, which are cleaved in the large intestine to yield
the free aglycones and further metabolites. The objective of this study was to compare the
antioxidant activities of genistein and daidzein with their glycosylated and methoxylated
derivatives, and their human metabolites. The abilities of these compounds to inhibit lipid
peroxidation in a liposomal system were evaluated using fluorescence spectroscopy, and
structural criteria that enhance antioxidant activity were established. The peroxidation
initiators employed in the study were Fe(II) and Fe(III) metal ions, and aqueous-phase, azo-
derived peroxyl radicals. Both the parent isoflavonoids and their metabolites were more
effective at suppressing metal-ion-induced peroxidations than the peroxyl-radical-induced
peroxidation. Antioxidant activities for the isoflavone metabolites were comparable to or
superior to those for the parent compounds themselves. Equol and its 4-hydroxy and 5-

hydroxy derivatives were the most potent antioxidants in the study, suggesting that absence

111

112

of the 2,3 —double bond coupled with a loss of the 4-oxo group enhanced antioxidant activity.
Additionally, the number and position of hydroxyl groups were determining factors for
isoflavonoid antioxidant activity, with hydroxyl substitution being of utmost importance at
the C-4' position, of moderate importance at the 05 position and of little significance at the

C-7 position.

4.2 INTRODUCTION

Certain phytochemicals in fruits, vegetables and grains are now being recognized as
possessing possible cancer-preventive properties through inhibition of tumor initiation,
prevention of oxidative damage, or by affecting steroid hormones or prostaglandin
metabolism to suppress tumor promotion (Caragay, 1992). In recent years, there has been
a surge of interest in studying the beneficial physiological efl‘ects of one such class of
compounds, the isoflavonoids. Isoflavonoids have a very restricted distribution in the plant
kingdom and occur almost exclusively in legumes (Leguminosae family); with soybeans,
chickpeas, lentils and beans representing the major dietary sources (Ingham, 1983).

The principal isoflavonoids occurring in legumes are genistein and daidzein which are
present primarily as their glycoside conjugates (genistin and daidzin) or their respective 4'-
methoxy derivatives (biochanin A and formononetin) (Price and Fenwick, 1985). Upon
ingestion, the isoflavonoids are subjected to acidic and enzymatic hydrolysis and
demethylation by the gut microflora to yield the free aglycones, demethylated products and
further metabolites, with both the metabolites and the unferrnented parent aglycones being

liable for firrther absorption (Kelly et al., 1995).

113

The pathways for human metabolism of isoflavones have not been clearly established.
Biochanin A and formononetin are known to undergo demethylation by intestinal bacteria,
giving rise to genistein and daidzein respectively. Daidzein undergoes intestinal microbial
transformation in humans to yield the isoflavan equol (7-hydroxy-3 —(4'-hydroxyphenyl)-
chroman) (about 70%) and a ring-cleavage compound O-desmethylangolensin (O-Dma) (5-
20%) (Setchell and Adlercreutz, 1988). Other daidzein intermediates formed during this
conversion are dihydrodaidzein, 4-hydroxy equol, 2-dehydro—0-Dma, and possibly
dehydroequol (Joannou et al., 1995). Genistein is metabolized within the gut by ring cleavage
to yield the non-estrogenic, hormonally inert compound p-ethyl phenol (Kelly et al. , 1993).
Other compounds that have recently been identified as catabolic products of genistein
metabolism in humans include dihydrogenistein and 6’-OH-0-Dma (Joannou er al., 1995;
Kelly et al., 1993).

The reported presence of isoflavones of plant origin in human urine (Adlercreutz et
al., 1991; Adlercreutz et al., 1986; Axelson et al., 1984; Axelson et al., 1982; Bannwart et
al., 1984) has renewed intense scientific interest in the study of these compounds because of
their association with a broad range of biological activities which include antiestrogenic
(Cassidy et al., 1995), anticancer (Peterson and Barnes, 1996), antiinflammatory (Yamarnoto
et al., 1996), cardioprotective (Anthony et al., 1996) and enzyme-inhibitory effects
(Yamashita et al., 1990). Since it is believed that many of these protective effects of
isoflavonoids may be related to their antioxidant activities, there has been a surge of interest
in exploring the antioxidant activities of these compounds. However, most of the studies

have focused solely on the antioxidant activity of genistein (Cai and Wei, 1996; Record er al. ,

114

1995; Wei et al., 1993). As genistein is principally obtained from the diet as its 7-[3-
glucoside, genistin, it would be appropriate to include both the free aglycone and the
conjugated glycoside in studies on antioxidant evaluation.

In addition, in vitro anaerobic incubation of isoflavones with human feces has shown
that intestinal half-life of genistein and daidzein may be as little as 3.3 h and 7 .5 h respectively
(Xu et al., 1995), indicating rapid degradation to other compounds by gut microflora (Chang
and Nair, 1995). It would, therefore, also be logical to include these metabolites in studies
related to antioxidant activity of isoflavonoids. However, other than a recent report on the
antioxidant activities of daidzein metabolites equol and O-Dma (Hodgson er al., 1996), there
have been no studies examining the antioxidant efficacies of isoflavone metabolites.

The aims of this study were to do a comprehensive study comparing the antioxidant
activities of isoflavone aglycones, glycosides and their biological metabolites, and to establish
structural criteria that enhance the antioxidant activity of these compounds. The parent
isoflavones included in this study were the free aglycones genistein and daidzein, the methoxy
derivatives biochanin A and formononetin and the glycosides genistin and daidzin (Figure
4.1). Daidzein metabolites equol, 4-hydroxy equol and dihydrodaidzein, and genistein
metabolite dihydrogenistein were evaluated for their antioxidant abilities (Figure 4.2). 5-
hydroxy equol, a possible metabolite of genistein, has to date not been detected in human
urine. However, it represents the firlly reduced form of genistein (just as equol represents the
fully reduced form of daidzein) and hence, was examined in the study for better elucidation
of structure—activity relationships. Other known metabolites like p—ethylphenol and O-Dma,

where the isoflavone nucleus was no longer intact, were not included in the study.

115

 

 

Substituents

 

 

 

 

 

 

 

 

Isoflavonoid R-S R-7 R-4'
Genistein OH OH OH
Genistin OH O-glucose OH
Daidzein H
Daidzin H
Biochanin A OH
Formononetin H

 

 

 

 

 

 

Figure 4.1. The structures of the plant-derived isoflavonoids included in this study.

116

.3333 “52535 :05 com 35:58 «8:382: 28268“ 05 .«o 8.5835 .3» 23E

< 5.3.85 :0... x
.25 teaza mo . m EoSSEom = u a .85 Exam... : u a

 

/ 52:80 :9. a \
582.5 m .. x

 

/. 503.5235 :0» a
\ 5836233 z .. a

 

117

Another aim of this investigation was to better understand the mechanism of action
by which these compounds act as antioxidants. To accomplish that objective, the inhibitory
effects of these compounds against lipid peroxidation induced by metal ions versus
peroxidation induced by peroxyl radicals generated in the aqueous phase by thermal
decomposition of an azo compound were studied. For the sake of comparison, the well-
- known antioxidant tert-butylhydroquinone (TBHQ) was also included in the study.

Many different methods are available for evaluating the oxidative stability of lipids
(Gutteridge and Halliwell, 1990). Most of these approaches suffer fi'om serious
methodological drawbacks. In addition, they can be time-consuming, have limited sensitivity
and require large amounts of material. A recently developed assay in our laboratory (Chapter
2) overcomes many of these disadvantages. The approach is based on the susceptibility of
the extrinsic probe 3-(p-(6-phenyl)-1,3,5-hexatrienyl)phenylpropionic acid (DPH-PA) to
peroxidative damge with concomitant loss of its fluorescent character. The decay of DPH-
PA fluorescence is directly related to the extent of lipid peroxidation and can be followed
continually. This fluorescence assay was used in this study for evaluation of isoflavonoid
antioxidant efficacy. This study was also aimed at testing the ability of this assay to

discriminate between antioxidant activities of compounds that are closely related in structure.

118

4.3 EXPERIMENTAL PROCEDURES

4.3.] Materials

Synthetic 1-stearoyl-2-linoleoyl-sn-glycero-3 -phosphocholine (SLPC) was obtained
from Avanti Polar Lipids (Alabaster, AL). Lipid purity was confirmed by silica gel TLC using
two different solvent systems (chloroform : methanol : water :: 65 : 25 :4; chloroform :
methanol : ammonium hydroxide :: 65 : 25 : 4). The lipid stock solutions in chloroform were
maintained at -20°C in amber glass vials that were layered with nitrogen. The fluorescent
probe 3—(p-(6-phenyl)—1,3,5~hexatrienyl)phenylpropionic acid (DPH-PA) was obtained from
Molecular Probes (Eugene, OR). The isoflavone glycosides were purified from soy molasses,
and the aglycones and their metabolites were obtained by synthesis (Chang et al. , 1994;
Chang et al., 1995). 2,2'-Azobis(2-amidinopropane) dihydrochloride (AAPH) was obtained
fiom Wako Chemical Company (Richmond, VA). F eClz'4HZO and F eCl,'6HZO of greater
than 99% purity were obtained from Aldrich (Milwaukee, WI) and stored at -20°C under
anaerobic conditions. All other reagents were of the highest grade available.

The DPH-PA and isoflavonoid stock solution were prepared in N,N-
dirnethylformamide and dimethyl sulfoxide (DMSO) respectively. The iron and AAPH stock
solutions were prepared immediately before use in nitrogen-sparged, double-distilled,
deionized water and maintained on ice under anaerobic conditions. To remove metal
contaminants, all the glassware was acid-washed prior to use and the salt and buffer solutions
maintained in a chelating resin Chelex 100 (sodium form, mesh 50-100, Bio-Rad, Richmond,

CA).

119

4. 3. 2 Preparation of Large Unilamellar Vesicles

LUVs containing the probe DPH-PA were prepared fi'esh for each day of
experimentation using the extrusion procedure of MacDonald et al. (1991). Briefly, a mixture
containing 5 nmol of SLPC and 15 nmol of DPH-PA was dried under vacuum onto the wall
of a 15 mL round-bottomed flask. The resulting lipid film was hydrated in 500 uL of a
solution containing 0.15 M NaCl, 0.01 M MOPS (pH 7.0) and 0.1 mM EDTA. After 10
fi'eeze-thaw cycles using a dry ice/ethanol bath, the suspension was passed 29 times through
two stacked polycarbonate filters (pore size 100 nm) using a Liposofast extruder apparatus

(Avestin, Ottawa, Canada).

4. 3.3 Antioxidant Evaluation of F lavonoids

The fluorescent intensity assay described in Arora and Strasburg (1997) was used to
evaluate the antioxidant eflicacy of the isoflavonoids. A 20 uL aliquot of the LUV
suspension was diluted to 2 mL in a buffer containing 100 mM NaCl and 50 mM Tris-HEPES
(pH 7.0) to achieve final concentrations of 100 uM lipid and 300 nM probe in the cuvette.
Afier a 5 min pre-incubation at room temperature with continuous stirring, the suspension
was transferred to a thermostatted cuvette holder (37°C) in the spectrofluorometer and
incubated for an additional 10 min for temperature equilibration. An aliquot of the
isoflavonoid stock solution, dissolved in DMSO, was added to yield a final concentration of
10 uM. The volume of added stock solution never exceeded 0.3 % of the total volume; under

these conditions, DMSO had no effect on the rate of peroxidation.

120

Following a 5 min incubation to allow partitioning of the isoflavonoid into the
membrane, peroxidation was initiated by addition of 20 uL of 1 mM stock FeCl2 or FeCl3
solutions or 40 uL of stock AAPH solution for final concentrations of 10 uM for FeCl2 and
FeCl,, and 10 mM for the AAPH. The control samples did not contain any peroxidation
initiator or flavonoid to be tested. The decay in fluorescence intensity was monitored over
21 min, with readings being taken at 0 min, 1 min and every 3 min thereafier.

Fluorescence measurements were performed in a 2 mL thermostated quartz cuvette
at 37°C, and the fluorescent signal was followed with a SLM Instruments, Model 4800,
spectrofluorometer (Urbana, IL) with data acquisition hardware and software from On-Line
Instrument Systems (Bogart, GA). The excitation wavelength used was 384 nm, slit width
2 nm, and the emitted light was passed through optical filters (KV 418, Schott, Duryea, PA),

prior to detection.

4.4 RESULTS

4. 4. I Effect of Isoflavonoids on Rates of F e(II)-Catalyzed Peroxidation in LU Vs

The inhibitory effects of genistein, daidzein, and their methoxy and glycoside
derivatives on rates of Fe(II)-induced peroxidation are illustrated in Figure 4.3. The efficacies
of these compounds as antioxidants were evaluated as the degree of inhibition of the
fluorescence intensity of the probe DPH-PA. All the isoflavonoids examined in the study
inhibited this drop in intensity to some degree; however, the rate of inhibition varied widely

depending on the structures of these compounds. The isoflavone genistein, with hydroxyl

121

Figure 4.3. Effects of the plant-derived isoflavones on rates of Fe(II)-induced peroxidation
in the LUVs at 37°C. Peroxidation was initiated in the LUVs containing 100 uM lipid, 300
nM probe and 10 uM test compound suspended in 2 mL of buffer (100 mM NaCl, 50 mM
Tris-HEPES, pH 7.0) by addition of Fe(II) for a final concentration of 10 1.1M. The rate of
peroxidation was monitored by a decrease in fluorescence intensity as a function of time.
Relative fluorescence represents the fluorescence intensity after a given time of oxidation over
the fluorescence intensity at time = 0 min. Values represent the means of triplicate

measurements.

122

3. 2:9...

 

 

 

01m... Ill

camcococeou +

< 5:235 I¢I

ENEmo II

5323 lml

53:30 IXI

:8250 +

A=xyaurvu

.9550 +

 

 

 

 

3.595.?
a 8 2 a. a o m o
x x u u x n L. ed
:3
is
.foac
.f../../
I f l ../ +2.
ll: ,,
///.ufl
- y
3

 

eouaoseromd aaueleu

123

groups at 5, 7 and 4' positions exhibited the highest antioxidant activity among this group of
isoflavonoids. Blocking the hydroxyl group at the 07 position of genistein by a glucose in
genistin did not have any effect on the antioxidant activity, as evidenced by the identical decay
in fluorescence in the presence of either of the two compounds. Daidzein, which lacks the
C-5 hydroxyl of genistein, was less effective as an antioxidant. Again, the antioxidant activity
for daidzin, the 7-,B-glucoside of daidzein, was identical to the aglycone itself. However,
substitution of the C-4' hydroxyl of genistein and daidzein by a methoxy in biochanin A and
formononetin, respectively, substantially reduced their antioxidant activities. The antioxidant
activity of the commercial food antioxidant TBHQ against Fe(II)-induced peroxidation was
comparable to or surpassed that of the isoflavones examined in the study.

The isoflavone metabolites inhibited the rates of Fe(II)-induced peroxidation in the
LUVs at rates that were comparable to or superior than the parent compounds themselves
(Figure 4.4). Of the metabolites, the isoflavans (with a saturated 2,3 bond and no keto group
at the C-4 position) displayed the highest antioxidant activities. S-hydroxy equol, the fully
reduced derivative of genistein, was the most potent antioxidant under these conditions,
showing intensity values that were almost indistinguishable from the control experiment. 4-
hydroxy equol and equol were also very effective under these conditions. The antioxidant
activities displayed by equol and its 4-and S-hydroxy derivatives against Fe(II)-induced
peroxidation were even superior to the highly effective antioxidant TBHQ. The isoflavanones
(with a saturated 2,3 bond) dihydrogenistein and dihydrodaidzein were less effective as

antioxidants than the isoflavans.

124

Figure 4.4. Effects of the isoflavone metabolites on rates of Fe(II)-induced peroxidation in
the LUVs at 37°C. Peroxidation was initiated in the LUVs containing 100 uM lipid, 300 nM
probe and 10 uM test compound suspended in 2 mL of buffer (100 mM NaCl, 50 mM Tris-
HEPES, pH 7.0) by addition of Fe(II) for a final concentration of 10 uM. The rate of
peroxidation was monitored by a decrease in fluorescence intensity as a function of time.
Relative fluorescence represents the fluorescence intensity after a given time of oxidation over
the fluorescence intensity at time = 0 min. Values represent the means of triplicate

measurements.

 

we 8:9“...

Es: oEF
E 3 9 N?

 

125

 

07.9.11

5328 2.: at...

semamo 2: 5+

.035 303+

68m. $03.?

Baum lel

2an :1

6:80 Isl

 

 

 

~03
"(O

aoueosaronH enumeu

I’

ll

 

126

4. 4.2 Effects of Isoflavonoids on Rates of F e(III)-Catalyzed Peroxidation in the LU Vs

The trends for the antioxidant activities of the isoflavonoids remained essentially
unchanged when Fe(III) ions were used to initiate peroxidation instead of Fe(II) ions (Figure
4.5). Here too, the aglycone genistein and its corresponding 7-fl-glucoside, genistin, were
the most effective of the set of dietary isoflavanones examined. In fact, genistein and genistin
surpassed the inhibitory effects observed with TBHQ. Daidzein and its glycoside daidzin also
displayed very similar rates of inhibition. These inhibitory effects of daidzein and daidzin were
less than those seen for genistein and genistin. Methoxylation of the C-4' hydroxyl of
genistein and daidzein led to a dramatic drop in their antioxidant activities.

The isoflavone metabolites were very effective at inhibiting Fe(III)-induced liposomal
peroxidation (Figure 4.6). Again, 5-hydroxy equol, a possible genistein metabolite, and equol
and 4-hydroxy equol, daidzein metabolites, exhibited the highest antioxidant potencies. These
activities were greater than those observed with TBHQ. The dihydro derivatives of daidzein
and genistein were slightly less effective than TBHQ; however, they still showed substantial

rates of inhibition against Fe(III)-induced peroxidation.

4. 4.3 Effects of Isoflavonoids on Rates of AAPH-Induced Peroxidation in the LU Vs
Figure 4.7 illustrates the inhibitory effects of the parent isoflavones against
peroxidation induced by peroxyl radicals generated at a constant rate in the aqueous phase
by thermal decomposition of the azo compound AAPH. The isoflavones studied were less
effective at inhibiting the AAPH-induced peroxidation. TBHQ exhibited greater antioxidant

potencies against AAPH-induced peroxidation than the isoflavones. The antioxidant trends

127

Figure 4.5. Effects of the plant-derived isoflavones on rates of F e(III)-induced peroxidation
in the LUVs at 37°C. Peroxidation was initiated in the LUVs containing 100 uM lipid, 300
nM probe and 10 uM test compound suspended in 2 mL of bufl‘er (100 mM NaCl, 50 mM
Tris-HEPES, pH 7.0) by addition of Fe(II) for a final concentration of 10 uM. The rate of
peroxidation was monitored by a decrease in fluorescence intensity as a function of time.
Relative fluorescence represents the fluorescence intensity after a given time of oxidation over
the fluorescence intensity at time = 0 min. Values represent the means of triplicate

measurements.

128

 

3. 2:9".

 

 

 

 

 

 

 

E25 25...
pm 3 m r N? a o n
x l. L. . l. . “ oxo
01m... ITI
camcocoscou lcl
< Eco—.005 1.1 v 00.0
5330 I61
50350 101
.rl u Hlu/ .. and
E E0 . -
.«m. 0 111 .ll ll,” " -/-
I I -/
l N
523501 / M
scent-I 1 . o 19 T 1 8..

55:00 1.1

 

 

 

 

eouaasaronu eagreleu

129

Figure 4.6. Effects of the isoflavone metabolites on rates of Fe(III)-induced peroxidation in
the LUVs at 37°C. Peroxidation was initiated in the LUVs containing 100 uM lipid, 300 nM
probe and 10 M test compound suspended in 2 mL of bufi‘er (100 mM NaCl, 50 mM Tris-
HEPES, pH 7 .0) by addition of Fe(II) for a final concentration of 10 uM. The rate of
peroxidation was monitored by a decrease in fluorescence intensity as a firnction of time.
Relative fluorescence represents the fluorescence intensity after a given time of oxidation over
the fluorescence intensity at time = 0 min. Values represent the means of triplicate

measurements.

130

of 959....

 

 

 

 

.55. «at
a 8 2 N. a o m o
012.11
5828 3 611
.. on...
52380 is BLT
.85 foIImT
.85 £011 1 .. 85
Baum 1‘1 n n u -
n - r u, .
r u l .l/ . .
. H “/H/
253+ ? «I 61 lo ll . .
8.,

.9550 1.1

 

 

 

 

 

eoueaseronu engines

131

Figure 4.7. Efl‘ects of the plant-derived isoflavones on rates of AAPH-induced peroxidation
in the LUVs at 37°C . Peroxidation was initiated in the LUVs containing 100 [AM lipid, 300
nM probe and 10 11M test compound suspended in 2 mL of bufi‘er (100 mM NaCl, 50 mM
Tris-HEPES, pH 7.0) by addition of AAPH for a final concentration of 10 mM. The rate of
peroxidation was monitored by a decrease in fluorescence intensity as a function of time.
Relative fluorescence represents the fluorescence intensity afier a given time of oxidation over
the fluorescence intensity at time = 0 min. Values represent the means of triplicate

measurements.

132

3 28¢

 

 

 

 

 

€75 05C.
_.N a P m r N_. a o m c
I I1 .. . w .r 8...
01m... 1+1
£3an 141
< 5:286 11 m H A cod
£58 101 n
-/1 , - /1
camcococton. lml . //
.i I 1. I. . l. /, w/ .. 88
5.350 le
52286 IT
01 10.1 0
Ia<<+ 1.1 101 lfil 8..

.8500 1.1

 

 

 

 

aouaosaronlg aAIreIau

133

observed earlier against Fe(II) and Fe(III)-induced peroxidations were maintained here;
however, there were only marginal differences between the antioxidant activities of the
different isoflavones.

The trends remained unchanged with the metabolites of the parent isoflavones (Figure
4.8). Again, these compounds were less effective at inhibiting the AAPH-induced
peroxidations than the inhibitions observed earlier against Fe(II) and Fe(III) ions. Equol and
its 4- and S-hydroxy derivatives had a greater inhibitory effect than the dihydro derivatives
of daidzein and genistein. All the metabolites displayed lower antioxidant activities than

TBHQ.

4.5 DISCUSSION

In spite of the various reports linking many of the beneficial properties of
isoflavonoids to their antioxidant activities, no comprehensive studies have been conducted
examining the antioxidant efficacies of the dietary isoflavonoids versus their biological
metabolites. The aim of this work was to compare the antioxidant potencies of the naturally
occurring glycosidic and methoxylated forms of isoflavones, the free aglycones and their
biological metabolites. The goal was to establish structural features that enhance the
antioxidant activities of this class of compounds. The difl‘erent initiators of peroxidation used
to test the inhibitory actions of these compounds were the Fe(II) and Fe(III) metal ions and
peroxyl radicals generated from thermal decomposition of the water-soluble azo compound

AAPH. Another objective of the study was to test the sensitivity of the fluorescence

134

Figure 4.8. Effects of the isoflavone metabolites on rates of AAPH-induced peroxidation in
the LUVs at 37°C. Peroxidation was initiated in the LUVs containing 100 uM lipid, 300 nM
probe and 10 uM test compound suspended in 2 mL of buffer (100 mM NaCl, 50 mM Tris-
HEPES, pH 7.0) by addition of AAPH for a final concentration of 10 mM. The rate of
peroxidation was monitored by a decrease in fluorescence intensity as a function of time.
Relative fluorescence represents the fluorescence intensity afler a given time of oxidation over
the fluorescence intensity at time = 0 min. Values represent the means of triplicate

measurements.

13S

 

01m... IT.

smficow

$30le

5323 iv 5 LT

.85 £941.?

53% AIOYm IOI

Baum LT

In_<<lll

.8500 IOI

 

_.N

mi 0.59“.

.55. as:

3. 3 N_. a 0 n o

h _ _ h
u . q .

 

 

 

Ill-

qr-

l

 

 

 

 

0‘6

- cod

, and

co. _.

asueasaronu OAQIBIOH

136

spectroscopic assay previously developed in our laboratory in distinguishing between
antioxidant activities of structurally-related compounds.

The inhibitory potencies of isoflavonoids were dependent on the system used to
initiate peroxidation. All the compounds studied had a more pronounced inhibitory effect
against the metal-ion-induced peroxidations. The inhibitory effects of the isoflavonoids
decreased considerably when AAPH was used instead to initiate lipid peroxidation in the
vesicles. This was in spite of the fact that the reduction potential of isoflavonoids is lower
than that of peroxyl radicals, thereby allowing them to reduce these damaging oxy radicals
(Jovanovic er al. , 1994).

These results suggest that the free radical scavenging ability of these compounds only
partly accounts for their antioxidant capabilities. Because of their polyphenolic structures,
these compounds can donate hydrogen atoms to deleterious oxy radicals and form the less
reactive phenoxyl radicals in the process. The same structure also confers isoflavonoids with
an ability to chelate metal ions. The greater antioxidative potential of the isoflavonoids
against metal-ion-induced peroxidation observed in the study is probably a consequence of
their combined metal chelating and free radical scavenging abilities.

Metal chelation, though, cannot solely account for the greater antioxidant activities
of the isoflavonoids in the presence of metal ions. This is because the molar ratio of
isoflavonoid to metal ions was maintained at one to one in this study and most chelators of
similar structure complex iron with a 3:1 (chelator : iron) stoichiometry (Ryan and Petry,

1993).

137

In a study comparing antioxidant effectiveness of genistein against AAPH and
F e(II)ions, Record et a1. (1995) also observed that genistein afforded significant protection
against Fe(II)—induced peroxidation but was less effective against AAPH-induced
peroxidation Using the catechol-binding assay, these researchers were unable to demonstrate
any evidence of a substantial chelating ability for genistein. However, the assay used in the
study lacked sensitivity and was also limited by solubility properties of genistein.

Though no similar studies have been conducted with isoflavonoids, a study conducted
with a series of estrogens, compounds with structures similar to isoflavonoids, found that
inhibitory actions of these compounds were greater when Fe(II) ions were used as initiators
and much less pronounced against AAPH (Lacort et al., 1995). It was suggested that these
compounds may either be exerting their protective effects through chelation of metals or by
altering iron redox chemistry (Ruiz-Larrea et al., 1995).

In addition to being dependent on the peroxidation initiator used in the assay, the
antioxidant potencies for the isoflavonoids were also influenced by their chemical structures.
Although all the isoflavonoids examined in the study inhibited lipid peroxidation to a certain
extent, the degree of inhibition varied widely according to the structure of these compounds.
These structure-activity relationships were valid in all of the three peroxidation systems used
in the study.

Genistein, with hydroxyl groups at C-5, 7 and 4' positions, was an effective
antioxidant in the liposomal system. Daidzein, lacking the C-5 hydroxyl group of genistein,
was less effective as an antioxidant. This suggested that the hydroxyl at the 05 position

contributed towards the antioxidant activity of these compounds. In contrast, blocking the

138

C-7 hydroxyl ofgenistein or daidzein by a glucose in genistin or daidzin had no effect on the
antioxidant activities of the aglycones, thereby indicating that the hydroxyl at the C-7 position
had a negligible effect on the antioxidant activities of these compounds.

On the other hand, the contribution of the C-4' hydroxyl on the B-ring towards the
antioxidative potentials of the isoflavones was highly significant, as demonstrated by the
dramatic drop in the abilities of these compounds to inhibit lipid peroxidation when this 4'-
hydroxyl group of genistein or daidzein was substituted by a methoxy group. Biochanin A
and formononetin were very weak antioxidants in this in vitro system. In an in viva system,
however, since biochanin A and formononetin undergo fast demethylation to genistein and
daidzein, their antioxidant potencies may be vastly increased.

Comparing antioxidant potencies of genistein, daidzein and biochanin A, Wei et al.
(1995) also determined that the loss of the hydroxyl group at the C-4' position of isoflavones
totally diminished their inhibitory activity towards lipid peroxidation whereas replacement at
other positions had less of an influence. Genistein was found to be a more potent inhibitor
of lipid peroxidation than daidzein, and biochanin A was very ineffective under their assay
conditions.

An interesting observation of our study was that the metabolites of genistein and
daidzein exhibited antioxidant potencies that were comparable to or superior than the parent
compounds themselves. Reducing the isoflavone nucleus to yield the isoflavan structure
substantially enhanced the antioxidant activities of these compounds. This was evidenced by
the superior antioxidant actions of the derivatives equol, 4-hydroxy equol and S-hydroxy

equol as compared to the parent isoflavones. This indicated that the absence of the 2,3-

139

double bond in conjunction with a loss of the 4-oxo group enhanced antioxidant activities of
these compounds.

These results were surprising since it is widely believed that presence of a 2,3-double
bond coupled with a 4-oxo group increases the antioxidant activities for flavonoids due to the
greater stability conferred to the flavonoid phenoxyl radical by this increase in conjugation
(Bors et al.,l990). The only other study comparing the antioxidant activities of equol with
daidzein and genistein also determined that equol was a more potent antioxidant, as compared
to the parent isoflavones (Hodgson et al., 1996). Since the inhibitory actions in the study
were evaluated against Cu(H)-induced lipoprotein oxidation in serum, the authors suggested
that equol may be inhibiting oxidation via additional mechanisms such as by acting as co-
antioxidants. However, use of our well-defined system with no contaminating pro- or
antioxidants present, ensures that the activity observed is attributable to the test compound
alone. The higher antioxidant activity of equol may be a result of its non-planar structure that
confers equol with a greater flexibility for conformational changes, thereby enabling it to
penetrate into the interior of the membrane with greater ease than some of the other
isoflavonoids that are more rigid in structure.

With the dihydro metabolites of genistein and daidzein, however, no conclusive
differences could be seen with respect to the parent compounds. The antioxidant activities
observed for dihydrogenistein and dihydrodaidzein were close to those observed for genistein
and daidzein. Hence, our study could not demonstrate any significant effect of loss of the 2,3-

double bond alone on the antioxidant activities of these compounds.

140

In summary, the protective effects of isoflavonoids in a liposomal system were
dependent on the chemical structures of these compounds and the system used to initiate
peroxidation. The isoflavonoids were more potent inhibitors of metal-ion-induced
peroxidation than of the peroxidation induced by peroxyl radicals. For the structural criteria,
the number and position of hydroxyl groups was found to be an important determinant of
antioxidant activity (Figure 4. 9). Hydroxyl groups were found to be of critical importance
at the C-4' position, of moderate importance at the 05 position and of negligible importance
at the C-7 position. Loss of the 2,3-double bond coupled with the absence of the 4-oxo
group conferred the greatest antioxidant activities to these compounds.

As consumer interest in the use of plant-derived antioxidants in food products grows,
a systematic approach will be required to rapidly evaluate plant extracts and purified
compounds for antioxidant activity. The simplicity, sensitivity, and rapidity of the
fluorescence assay developed in our laboratory combine to make this an attractive screening
method to identify compounds or extracts for subsequent comprehensive evaluations in food
products. In addition, the results of this study indicate that this method has the ability to

discriminate between the antioxidant abilities of a class of structurally related compounds.

141

 

Figure 4.9. Structural criteria that enhance the antioxidant activity of isoflavonoids.

CHAPTER 5

MODULATION OF LIPOSOMAL MEMBRANE FLUIDITY BY

FLAVONOIDS AND ISOFLAVONOIDS

5.1 ABSTRACT

The polyphenolic structures of flavonoids and isoflavonoids confer them with the
ability to scavenge free radicals and to chelate transition metals, a basis for their potent
antioxidant abilities. Another contributory mechanism towards their antioxidant activities
could be their ability to stabilize membranes by decreasing membrane fluidity. In this study,
the effects of representative flavonoids, isoflavonoids and their metabolites on membrane
fluidity, and their preferential localization in the membrane were investigated using large
unilamellar vesicles (LUVs) as the membrane models. These results were compared with
cholesterol and a-tocopherol. Changes in fluorescence anisotropy values for the probes 3-(p-
(6-phenyl)-1,3,5-hexatrienyl)phenylpropionic acid (DPH-PA) and a series of n-(9-
anthroyloxy) fatty acids (11 = 6, 12, 16) upon addition of the test compounds were used to
monitor alterations in membrane fluidity at graded depths in lipid bilayer. The results of the
study suggested that the flavonoids and isoflavonoids, similar to cholesterol and a-tocopherol,

partitioned into the hydrophobic core of the membrane and caused a dramatic decrease in lipid

142

143

fluidity in this region of the membrane. Localization of flavonoids and isoflavonoids into the
membrane interiors and their restrictions on the fluidity of membrane components would limit

accessibility of free radicals and thus decrease the kinetics of free radical reactions.

5.2 INTRODUCTION

Free radical peroxidation of unsaturated lipids in biomembranes disrupts the various
important stmctural and protective functions associated with biomembranes, and various in
viva pathological events are implicated as a result of this oxidation (Barclay, 1993; Choe et
al., 1995; Niki et al., 1991). The deleterious consequences of membrane peroxidation have
stimulated numerous studies on the eflicacies and mechanisms of action of biologically
relevant antioxidants. A common structural criteria shared by the most potent membrane
antioxidants is the presence of an aromatic nucleus with at least one phenolic hydroxyl group
(Kagan et al., 1990). It is the presence of an easily donatable hydrogen atom coupled with
the stability of the resulting phenoxyl radical due to electron delocalization that contributes
to the effectiveness of these phenolic compounds as prototypic chain-breaking, fi'ee radical
scavenging antioxidants. Recent research, however, suggests that another factor contributing
to the effectiveness of certain phenolic compounds as antioxidants is their degree of
incorporation, uniformity of distribution and orientation in the membrane bilayer (Barclay et
al., 1990; Kaneko et al., 1994; Niki et al., 1985; Serbinova et al., 1991).

Flavonoids, a class of naturally-occurring benzo-y-pyrone compounds, are
increasingly gaining recognition for their ability to inhibit lipid peroxidation in biological

membranes (Chen et al., 1996; Ioku et al., 1995; Ramanathan and Das, 1992; Ratty and

144

Das, 1988; Terao et al., 1994). The mechanism of antioxidant action for these compounds
has not yet been fiilly elucidated and is still a matter of considerable debate. As a
consequence of their polyphenolic structure, these compounds are considered effective
scavengers of peroxyl (Ioku et al., 1995; Montesinos et al., 1995; Torel er al., 1986),
hydroxyl (Hanasaki et al., 1994; Rekka and Kourounakis, 1991; Sanz et al., 1994) and
superoxide (Cotelle et al., 1996; Cotelle et al., 1992 Hu et al., 1995) radicals. This fi'ee
radical scavenging mechanism has generally been ascribed as the basis of their antioxidant
activity. In addition, flavonoids are known to chelate metal ions (Afanas’ev et al., 1995;
Afanas’ev et al., 1989; Thompson and Williams, 1976), another possible contributory
mechanism towards their antioxidant action. However, the ability of flavonoids to alter
peroxidation kinetics by modification of the lipid packing order and fluidity of membranes has
not been investigated as a possible antioxidant mechanism.

In order to better understand the antioxidant activities of flavonoids in biomembranes,
it is important to determine the precise location of these compounds in membranes and to
examine their effects on membrane fluidity and packing order. a-Tocopherol, the major lipid-
soluble, chain—breaking antioxidant in biological membranes, is known to act as an antioxidant
partly through stabilization of biological membranes by restricting the molecular mobility of
their components (Kagan et al., 1990; Niki et al., 1985; Urano et al., 1990; Urano et al.,
1988). Similar observations have also been reported for the potent anticancer drug tamoxifen
and for cholesterol (Clarke et al., 1990; Wiseman et al. , 1993; Wiseman et al., 1990).

The aim of the present study was to investigate the effects of a range of structurally

representative flavonoids and isoflavonoids on the fluidity of liposomal membranes and to

145

ascertain their precise location in the lipid bilayer. Of the different types of liposomes
available, large unilamellar vesicles (LUVs) were selected as the model system in the study
since they are the most representative of real biological membranes. Representative flavonoid
and isoflavonoid samples exhibiting a broad multiplicity in structure were included in the
study. The flavanones naringenin and hesperetin and the flavone glycoside rutin were selected
to represent the flavonoids (Figure 5.1). For the subgroup of isoflavonoids, the isoflavone
aglycones genistein and biochanin A and the glucosides genistin and daidzin were examined
for their effects on membrane fluidity (Figure 5.2). As these compounds are rapidly degraded
to other compounds in the human body, it is meaningful to include their metabolites in studies
on biomembrane stabilization. Genistein metabolite dihydrogenistein and daidzein metabolites
dihydrodaidzein, equol and 4-hydroxy equol were included in the study. S-hydroxy equol,
a possible genistein metabolite that has not been detected in human urine to date, was also
examined for its effects on membrane fluidity (Figure 5.2). The effects of flavonoids and
isoflavonoids on membrane fluidity were compared against a-tocopherol and cholesterol
(Figure 5.1).

The localization of flavonoids and isoflavonoids in membranes and their efl‘ects on
membrane fluidity were studied by fluorescence anisotropy. Fluorescence anisotropy
measurement of fluorescent probes is one of the most widely used techniques in study of the
influence of different molecules on membrane biophysical properties (Jemiola-Rzeminska et
al., 1996). A series of n-(9-anthroyloxy) fatty acids with the anthroyloxy moiety covalently
linked at various positions along the alkyl chains were selected as the fluorescent probes in

this study. As the position of the fluorophore is fixed along the fatty acid acyl chain for these

146

   

H, R,= OH Naringenin Rutin
OH, lg = OCH3 Hesperetin

 

or-Tocopherol

OH.

I /
CH —CH,— CH,—-CH,— CH

CH, CH,

CH,

    

Cholesterol

Figure 5. 1. Structures of naringenin, hesperetin, rutin, a-tocopherol and cholesterol.

147

   

Rl = OH, R2 = OH Genistein Biochanin A
Rl = O-glu, R7 = OH Genistin
R1 = O-glu, R2 = H Daidzin

 

= H Equol
= OH 4-Hydroxy Equol
= H S-Hydroxy Equol

757070

H
H,
O

36:"?

 

H Dihydrodaidzein
OH Dihydrogenistein

R
R

Figure 5.2. Structures of the isoflavonoids and their metabolites included in the study.

148

probes, their location in the lipid bilayer is well defined (Garrison et al., 1994). Previous
studies have confirmed that the n-(9-anthroyloxy) fatty acids fit into the bilayer with the acyl
chains parallel to those of the phospholipids and with their anthroyloxyl moieties located at
a graded series of depths in the bilayer, as revealed by energy transfer, NMR and fluorescence
quenching studies (Mason, 1994; Tricerri et al., 1994).

In the study, the fluorescent probes employed for monitoring changes in membrane
fluidity at graded depths in the phospholipid bilayer were the 6—(9-anthroyloxy)stearic acid
(6-AS), 12—(9-anthroyloxy)stearic acid (12-AS), 16—(9-anthroyloxy)palmitic acid (l6-AP) and
3-(p—(6-phenylH,3,S-hexatrienyl)phenylpropionic acid (DPH-PA) (Figure 5.3). The anionic
charge of DPH-PA ensures that this probe labels the lipid-water interfacial region of the
membrane bilayer and reports on fluidity changes near the membrane surface; whereas the
anthroyloxyl moieties included in the study are linked to carbons 6, 12 and 16 of the base fatty
acids and report on fluidity changes at defined depths within the membrane bilayer. As each
of the four fluorescence probes labels a different region of the membrane, the information

obtained with them is complementary.

5.3 EXPERIMENTAL PROCEDURES

5. 3. 1 Materials
The fluorescent probes 3-(p—(6-phenyl)-l,3,5-hexatrienyl)phenylpropionic acid (DPH-
PA), l6-(9-anthroyloxy)palmitic acid ( l6-AP), 6-(9-anthroyloxy)stearic acid (6-AS) and 12-

(9-anthroyloxy)stearic acid (lZ-AS) were purchased fi'om Molecular Probes (Eugene, OR).

149

0
II

 

CH,(CH,)_CH(CH,), c — OH

m = 11 6-(9—anthroyloxy)stearic acid
= 5 12-(9-anthroyloxy)stearic acid

0 0
II
C —O—1(CH),,—-C —OH

16-(9-anthroyloxy)palmitic acid

0

cn;——crr,—~ c —orr

 

3-(p-(6-phenyl)-1,3,5—hexatrienyl)phenylpropionic acid

Figure 5.3. Structures of the fluorescent probes employed in the study.

150

The purity of the n-(9-anthroyloxy) fatty acids was confirmed by silica gel thin-layer
chromatography using ethanol/water (95 :5, v/v) as solvent followed by visualization with
iodine (Mason, 1994). Synthetic 1-stearoyl-2-linoleoyl-sn-glycero-3 -phosphocholine (SLPC)
was obtained from Avanti Polar Lipids (Alabaster, AL) and its purity checked by silica gel
thin-layer chromatography employing two different solvent systems (chloroform / methanol
/ water : 65/25/4 and chloroform / methanol / water : 65/25/4) followed by sulfuric acid
charring.

The isoflavone glycosides were purified from soy molasses, and the aglycones and
their metabolites were obtained by synthesis (Chang et al., 1994; Chang et al., 1995). Rutin
was purchased from Sigma Chemical Co. (St. Louis, MO) and the other flavonoids were from
Indofine Chemical Co., Inc. (Somerville, NJ). Cholesterol (99%) was fi'om Sigma Chemical
Co. (St. Louis, MO) and a-tocopherol (99%) was donated by Henkel Corporation (La
Grange, IL).

The lipid stock solutions, dissolved in chloroform, were maintained at -20°C in amber
glass vials that were layered with nitrogen. The fluorescent probe stock solutions were
prepared in N,N-dimethylfonnamide and the flavonoid and isoflavonoid stocks were prepared
in dirnethyl sulfoxide (DMSO). Cholesterol and a-tocopherol stock solutions were in ethanol

and methanol respectively. All the stock solutions were stored under nitrogen at -20°C.

5. 3.2 Preparation of Large Unilamellar Vesicles
LUVs containing the probe DPH-PA were prepared fresh for each day of

experimentation using the extrusion procedure of MacDonald et al. (1991). Briefly, a mixture

151

containing 10 nmol of SLPC and 40 nmol of the fluorescent probe to be used was dried under
vacuum onto the wall of a 15 mL round-bottomed flask to yield a probe to lipid molar ratio
of 1:250. At this mole ratio, the fluorescence intensity for the four probes was well within
their linear concentration ranges and no inner filter efi‘ects were displayed. The dried lipid and
probe mixture was dispersed in 1 mL of a solution containing 0.15 M NaCl, 0.01 M MOPS
(pH 7 .0) and 0.1 mM EDTA. After 10 freeze-thaw cycles using a dry ice/ethanol bath, the
suspension was passed 29 times through two stacked polycarbonate filters (pore size 100 nm)
using a Liposofast extruder apparatus (Avestin, Ottawa, Canada). During the preparation
steps, the lipid and probe were shielded from light and oxygen as much as possible. As the
fluorescent probes were present in the dried lipid films when they were hydrated, it can be

assumed they were symmetrically incorporated into both leaflets of the phospholipid bilayers.

5. 3.3 Fluorescence Anisotropy Measurements

Steady-state fluorescence emission anisotropy measurements were obtained with a
SLM Instruments, Model 4800, spectrofluorometer (Urbana, IL) equipped with a xenon light
source, a thermostated cuvette holder and rotatable polarizers. The samples were excited
with vertically polarized light at 384 nm (slit width 2 nm), and vertical and horizontal
components of the sample fluorescence, II and 1,, emitted through optical filters (KV 418,
Schott, Duryea, PA), were detected in the T-format. The steady-state anisotropy, r,, was

calculated as:

r, = (II — Ii) / (I. + 2Il)

152

where II and Il are the fluorescent intensities of the vertically ( u) and horizontally (4.) polarized
emission when the sample is excited with vertically polarized light (Lakowicz, 1983). The
contribution of scattered light to the fluorescent emission was determined using an unlabeled
reference solution of the same composition excited with vertically polarized light and
detection of the emitted light with the polarizer at 55°. The ratio of this signal to that
determined for membranes containing the probe gave the fiction of signal resulting fi'om light
scattering. At the lipid concentrations employed, the contribution of scattered light to the
total fluorescence emission was negligible and no corrections were made to the anisotropy
values.

For fluorescence anisotropy measurements, a 20 uL aliquot of the liposome
suspension was diluted to 2 mL in buffer containing 100 mM NaCl and 50 mM Tris-HEPES
(pH 7.0) to achieve final concentrations of 100 uM lipid and 400 nM fluorescent probe.
Cuvette temperature was maintained at 22°C with a circulating water bath. For each
compound to be tested, different volumes of the stock solutions were sequentially titrated into
the cuvette and changes in anisotropy values for each probe as a function of concentration of
the test compound were determined. During the titrations, there was a 5 min incubation
period following each addition of test sample to allow the compound to partition into the
membrane. Control experiments were conducted with addition of equivalent volumes of each

of the blank solvents and found to have no effect on the anisotropy values.

153

5.4 RESULTS

5. 4. 1 Effects of F lavonoids on Membrane F luidity

The effects of the flavonoids naringenin, hesperetin and rutin on membrane fluidity
along graded depths in the phospholipid bilayer are presented in Figure 5.4. Data are
reported as increases in the steady-state anisotropy parameters for the four probes used in the
study; an increase in the anisotropy parameter of a probe being indicative of a decrease in the
fluidity of the membrane. The two flavanones naringenin and hesperetin displayed similar
effects on the fluidity of the membrane. In both cases, only a marginal increase in the
anisotropy parameter for the probe DPH-PA was observed with increasing concentrations of
these flavanones. This suggested that naringenin and hesperetin were not influencing
membrane fluidity properties in the exterior region of the membrane. As the anthroyloxy
fluorophore was shifled along the acyl chain towards the interior of the membrane, the
influence of naringenin and hesperetin on membrane fluidity increased. The greatest
restriction of membrane mobility with naringenin and hesperetin was noted in the interior
regions of the membrane labeled by the probe 16-AP. For the flavone rutin, a linear increase
in anisotropy values of l6-AP as a fimction of rutin concentration was observed, also
demonstrating the strong influence of flavonoids on membrane fluidity in the interior of the
membrane. The influence of rutin on membrane fluidity diminished considerably as the

fluorophore position shifted to the exterior regions of the membrane.

154

Figure 5.4. Efi‘ects of flavonoids on membrane fluidity as studied by increases in anisotrOpy
values of the fluorescent probes DPH-PA (x), 6-AS (c), 12-AS (I) and l6-AS (A) at 22°C.
Measurements were carried out as described under Experimental Procedures. Values

represent the means of triplicate measurements.

155

 

 

Naringenin

888

$Anlaotropy Increase
8 8

 

 

 

 

a
00

10
o
0.0 0.2 0.4 0.6 0.8 1.0
Molar Ratio to SLPC
Hesperetin
60
g 50
g 40
§ so
’5‘; 20
as

 

0.0 0.2 0.4 0.6 0.8 1 .0
Molar Ratio to SLPC

 

 

 

Rutin

*Anlsotropylncraasa
068888838

 

Molar Ratio to SLPC

 

Figure 5.4

 

 

 

156

5. 4.2 Effects of Isoflavones on Membrane F luidily

Figure 5.5 illustrates the effects of the isoflavones genistein, genistin, daidzin and
biochanin A on membrane fluidity at different locations along the bilayer, as revealed by
changes in anisotropy values for the fluorescent probes. When increasing concentrations of
genistein were titrated into the liposomes, there were minimal changes in the anisotropy
values of the probes DPH-PA and 6-AS. This indicated that incorporation of genistein had
little effect on the dynamics of the external regions of the membrane bilayer. There was a
moderate increase in anisotropy values for 12-AS with increasing concentrations of genistein
suggesting that incorporation of genistein did contribute to a decrease in membrane fluidity
in this region of the membrane. The maximal effects of increasing genistein concentration on
anisotropy values were observed with the probe 16-AS. With 16-AS, there was almost a
linear increase in anisotropy values with increasing concentrations of genistein; evidence that
genistein was causing a strong membrane rigidifying effect in the interior of the bilayer.

Introduction of increasing amounts of genistin, the 7-fl-glucoside of genistein, also
resulted in a maximal reduction in fluidity in the membrane interior, as indicated by the
increase in anisotropy parameter of the probe 16-AP. A slight reduction in fluidity was
observed in the region of the membrane bilayer labeled by the probe 12-AS. On the other
hand, the anisotropy parameters for the probes 6-AS and DPH-PA showed a slight decrease
in value when increasing amounts of genistin were titrated into the liposomes.

The glycoside daidzin showed similar patterns to genistin. The incorporation of
daidzin into the membrane bilayer resulted in a linear increase in anisotropy values of 16-AP,

though the increase was somewhat smaller than that observed with genistin. Again, the

157

Figure 5.5. Effects of isoflavonoids on membrane fluidity as studied by increases in
anisotropy values of the fluorescent probes DPH-PA (x), 6—AS (e), 12-AS (I) and l6-AS (A)
at 22°C. Measurements were carried out as described under Experimental Procedures.

Values represent the means of triplicate measurements.

158

 

Genistein

    

 

 

ao
- 4o
2 f,
as
0.0 0.2 0.4 0.6 0.8 1.0
Molar Ratio to $ch
Genistin

$6 Anisotropy
Increase

 

0.0 0.2 0.4 0.6 0.8 1 .0

 

 

 

  

Molar Ratio to SLPC
Daidzin
30
g 3
ti ‘5
_. 1o
5 3. 5
* o
-5

 

0 0.2 0.4 0.6 0.8 1 .0
Molar Ratio to SLPC

.0

 

 

 

Biochanin A

 

0.0 0.2 0.4 0.6 0.8 1.0
Molar Ratio to SLPC

 

Figure 5.5

 

 

 

159

region probed by the probe 12-AS showed a negligible reduction in membrane fluidity,
whereas the external regions of the membrane probed by 6-AS and DPH-PA showed no effect
or a slight increase in fluidity.

Biochanin A, the 4'-methoxy derivative of genistein, demonstrated the strongest
influence on membrane fluidity. Incorporation of biochanin A into the liposomes led to strong
decreases in fluidity in the interior region of the bilayer labeled by the probes 16-AP and 12-
AS. In contrast, the region of the membrane probed by 6-AS was affected to a considerably
lower degree by the introduction of biochanin A into the bilayer. The superficial regions of

the membrane probed by DPH-PA demonstrated negligible changes in membrane fluidity.

5. 4.3 Effects of Isoflavone Metabolites on Membrane Fluidity

Similar trends were obtained with the isoflavone metabolites equol, 4-hydroxy equol,
dihydrodaidzein and dihydrogenistein, and with the possible metabolite 5-hydroxy equol
(Figure 5.6). In each of the cases, there was no perturbation or very little perturbation to the
superficial region of the membrane labeled by the probe DPH-PA. As the fluorophore was
moved from position 6 to position 16 along the fatty acid chain, the effects of these isoflavone
derivatives on membrane fluidity increased accordingly. Maximal reduction in fluidity was
observed in the interior regions of the phospholipid bilayer, as evidenced by the substantial

increase in anisotropy values for 16-AP.

160

Figure 5.6. Effects of isoflavonoid metabolites on membrane fluidity as studied by increases
in anisotropy values of the fluorescent probes DPH-PA (x), 6-AS (o), 12-AS (I) and 16-AS
(A) at 22°C. Measurements were carried out as described under Experimental Procedures.

Values represent the means of triplicate measurements.

161

 

 

 

Equol

5-(OH) Equol

 

'loAnlsotropy
Increase
03888883

 

 

 

 

 

 

 

 

 

 

 

 

96 Anisotropy
so

 

 

0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0
MolarRatioto SLPC Molar Ratlo to SLPC
5° 4-(OH) Equol Dihydrogenistein
40
>
if 3°
5 E 2°
:2 10
o
0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0
MolarRatioto SLPC MolarRatloto SLPC
Dihydrodaidzein

0.2 0.4 0.6 0.8 1.0
Molar Ratio to SLPC

 

 

Figure 5.6

 

162

5. 4.4 Effects of Cholesterol and a- Tocopherol on Membrane F luidity

Cholesterol and a-tocopherol appeared to localize exclusively in the interior regions
of the membrane (Figure 5.7). With either of these two compounds, there was a linear
increase in the anisotropy parameter of the probe 16-AP as a function of the concentration
of these two test compounds, indicative of the strong membrane stabilizing effect of
cholesterol and a-tocopherol in the hydrophobic core of the membrane. In strong contrast,
both of the compounds trivial effects on the fluidity of the membrane in the domains probed

by 12-AS, 6-AS and DPH-PA.

5.5 DISCUSSION

Membrane function is of vital importance to normal processes and can be influenced
by a wide range of factors (\Vrseman, 1996). One such factor is the modulation of membrane
function by dietary components such as vitamin E (Bisby, 1990; Morrissey et al., 1994),
vitamin D (Wiseman, 1993), vitamin C (Halliwell and Gutteridge, 1989; Thumham, 1994),
B—carotene (Halliwell and Gutteridge, 1989; Thumham, 1994), flavonoids (Chen et al., 1996;
Saija et al., 1995) and isoflavonoid-type phytoestrogens (Jha et al., 1985; Messina et al.,
1994) through alteration of membrane characteristics such as fluidity, stability, and
susceptibility to peroxidative damage.

The antioxidant activity of one such class of compounds, the flavonoids and
isoflavonoids, in membranes is well established. In contrast, there are no available data on
the precise location of these compounds in membranes and on their consequent effect on

membrane integrity. An understanding of the effects of flavonoids and isoflavonoids on

163

Figure 5.7. Effects of cholesterol and a-tocopherol on membrane fluidity as studied by
increases in anisotropy values of the fluorescent probes DPH-PA (x), 6-AS (c), 12-AS (I)
and 16-AS (A) at 22°C. Measurements were carried out as described under Experimental

Procedures. Values represent the means of triplicate measurements.

 

 

Cholesterol

 

0.0 0.2 0.4 0.6 0.8 1.0
Molar Ratio to SLPC

 

 

 

it Anisotropy Increase

or Tocopherol

8888

.a
CO

 

-10

0.0 0.2 0.8 1 .0

0.4 0.6
Molar Ratio to SLPC

 

Figure 5.7

 

 

165

biophysical properties of membranes may help to better elucidate their mechanism of action
as antioxidants. Other prototypic free radical scavenging antioxidants have been shown to
act as membrane stabilizers partly through a restriction of membrane fluidity. a-tocopherol
is generally believed to act as a chain-breaking antioxidant by donating a hydrogen atom from
the phenolic hydroxyl group present in the 6-hydroxychromane ring to the chain-propagating
lipid peroxyl and alkoxyl radical intermediates of lipid peroxidation, thereby producing the
stable tocopheroxyl radical and terminating the chain reaction (Liebler, 1993). In addition to
its chain-breaking antioxidant action, a structural membrane antioxidant action for a-
tocopherol mediated by the hydrocarbon chain through decreased membrane fluidity has also
been demonstrated (Kagan er al., 1990; Urano er al. , 1992; Urano et al., 1990; Urano et al.,
1988)

A similar mechanism of action has also been proposed for the anticancer drug
tamoxifen, which does not possess a labile hydrogen atom for free radical scavenging, and for
4-hydroxy tamoxifen and estrogens and vitamin D (Clarke et al., 1990; Wiseman et al.,
1993). Though the rest of these compounds possess a potentially donatable hydrogen atom,
time—course data indicate that they are not predominantly chain-breaking antioxidants (Lacort
et al., 1995; Ruiz-Larrea et al., 1994; Wiseman et al., 1990). For all these compounds, a
structural antioxidant action has been proposed based on their ability to mimic the membrane
stabilization against lipid peroxidation demonstrated by the structurally similar endogenous
membrane sterol cholesterol (Wiseman, 1994).

The purpose of the present study was to investigate the effects of flavonoids and

isoflavonoids on lipid fluidity in model membranes. The term “lipid fluidity” refers to the

166

relative motional freedom of the lipid molecules in the membrane bilayer (Jourd’heuil et al.,
1993). Probes such as DPH-PA possess a relatively high limiting hindered anisotropy, and
it is hypothesized that membrane order plays an important role in determining the motional
freedom of these probes. Hence, the use of such probes allows an estimation of the static
component of membrane fluidity (Schachter, 1984). In contrast, probes such as the
anthroyloxy fatty acid derivatives possess a relatively low value for hindered anisotropy, and
the anisotropy parameter for these probes reflects primarily their rotational movement, i.e.,
a dynamic component of membrane fluidity (Thulbom and Sawyer, 1978).

By the use of DPH-PA, 6-AS, 12-AS and 16-AP as the fluorescent probes in our
study, it was possible to determine fluidity of the membrane as a firnction of depth within the
bilayer. Our results suggest that all the flavonoids, isoflavonoids and their metabolites
examined in the study partitioned preferentially into the hydrophobic core of the model
membrane, where they modified the lipid packing order. These compounds decreased
membrane fluidity in a concentration dependent manner in the interior region of the membrane
labeled by the probe l6-AP. As the fluorophore position was moved from position 16 to
position 6 on the fatty acyl chain, these molecules appeared to have a smaller effect on the
lipid packing order. Similarly, the superficial regions of the membrane probed by DPH-PA
were not influenced by introduction of flavonoids and isoflavonoids into the bilayer.

The ability of flavonoids and isoflavonoids to influence membrane fluidity and their
preferential localization in the membrane were contrasted against the known membrane
stabilizers cholesterol and a-tocopherol. Like the flavonoids and isoflavonoids, cholesterol

and a-tocopherol had a strong, dose-dependent effect on the fluidity of the hydrophobic core

167

of the membrane, as revealed by the increase in anisotropy values for 16-AP. However, while
most of the flavonoids and isoflavonoids examined did exert some influence on lipid packing
order in other regions of the bilayer, cholesterol and a-tocopherol exclusively partitioned into
the internal regions of the membrane.

Other studies have indicated that, for a-tocopherol, the chromanol moiety fits into the
“space” formed by the polyunsaturated phospholipids of the membrane bilayer and the
isoprenoid side chain is embedded in the membrane interior to retain the molecule in the lipid
core (Urano et al., 1993; Urano er al., 1992; Urano er al., 1990). From our data, we can
conclude that a-tocopherol does not have much impact on the membrane mobility properties
in the exterior regions of the bilayer, but the isoprenoid side chain considerably restricts
mobility in the core of the lipid bilayer. Similarly, cholesterol is believed to act as a membrane
stabilizer via interactions between its hydrophobic rings and the saturated, monounsaturated
and polyunsaturated residues of phospholipid fatty acids (Wiseman, 1996; Wiseman, 1994),
and would thereby be expected to exert its strongest influence in the interior regions of the
membrane as was observed in this study.

Since the flavonoids and isoflavonoids mimicked the membrane stabilizing effects of
cholesterol and a-tocopherol, this mechanism may partly account for their antioxidant
activities, as has been previously confirmed for cholesterol and a-tocopherol. By imposing
a greater degree of structural order and rigidity to the membrane, the flavonoids and
isoflavonoids could reduce the mobility of free radicals in the lipid bilayer. Consequently,
the decreased membrane fluidity would result in inhibition of lipid peroxidation due to a slow-

down of fiee radical reactions.

168

A previous study indicated that the ability of flavonoids to interact with and penetrate
the lipid bilayer influenced their antioxidant capabilities in biomembranes (Saija et al. , 1995).
Comparing the flavonoids quercetin, rutin, hesperetin and naringenin, these researchers
demonstrated that the ability of these compounds to shift the main transition peak temperature
for dipalmitoylphosphatidylcholine (DPPC) liposomes to lower values was related to their
ability to inhibit peroxidation in rat cerebral membranes. The results obtained by these
researchers were in direct contrast to our study. These researchers observed that the
introduction of flavonoids into the liposomes caused a membrane fluidifying efi‘ect. The
different results for the two studies are probably a consequence of differences in the lipid
substrates employed. The authors of the previous study had used the highly saturated DPPC
as their lipid substrate, whereas we employed the more unsaturated SLPC that has a
composition that is truly representative of phospholipids in biological membranes, with a
saturated fatty acid at the sn-1 position, an unsaturated fatty acid at the sn-2, and a
phosphate-containing polar group at the sn-3 position (Stanley, 1991).

Another flavonoid, diosmetin, has previously been shown to exert a protective effect
against in vitro cell membrane damage and oxidative stress in cultured rat hepatocytes (Villa
er al., 1992). This protective effect of diosmetin towards cell membranes was attributed in
part to its ability to stabilize the stmcture of biological membranes by modifying fluidity of
the lipid bilayer.

In most of these studies, it has been assumed that flavonoids and isoflavonoids are
localized near the lipid-water interface of the membrane. The precise location of these

compounds in the membrane has not been examined in any of these studies and would be

169

crucial for fully understanding the role of flavonoids and isoflavonoids in stabilizing the
membrane, both through chemical and physical means. In contrast to the generally held belief,
our results suggest that despite the presence of polar substituents, flavonoids and
isoflavonoids partition preferentially into the hydrophobic core of the membrane, where they
exert a membrane stabilizing effect by modifying the lipid packing order.

On the surface, such a conclusion appears highly improbable since it would involve
the unfavorable localization of polar hydroxyl groups in a very hydrophobic environment.
However, this may not be the case as the final conformation of these compounds could be
influenced by various other factors such as the strength of the hydrophobic interactions
between these compounds and the phospholipids, and also their ability to form intrarnolecular
and intermolecular hydrogen bonds.

Summarizing, our results demonstrate that flavonoids and isoflavonoids partition
preferentially into the hydrophobic core of membranes. They stabilize the membrane through
a decrease in lipid fluidity, and this may be a contributory mechanism towards their known

ability to inhibit membrane peroxidation.

CHAPTER 6

OXIDATION PRODUCTS OF GENISTEIN FORMED BY

REACTION WITH ALKYLPEROXYL RADICALS

6.1 ABSTRACT

To facilitate a better understanding of genistein antioxidant chemistry, oxidation
studies were conducted with genistein and alkylperoxyl radicals generated by thermal
decomposition of an azo compound in a homogenous system. The oxidation products of
genistein were separated and analyzed by reversed-phase HPLC. Successive analyses of the
reaction mixture over a 24 h time period indicated a gradual disappearance of genistein, and
the appearance of two major oxidation products. By 1H-NMR and MS analyses, one of the
products was assigned the structure 2-dehydro-o-demethyl angolensol, a C-ring-cleavage
product of genistein. Although an unambiguous chemical structure could not be assigned to
the second oxidation product, lH-NMR analysis indicated that genistein was undergoing
Opening at the C-ring to yield this oxidation product. The structural identification of these
two products could provide biochemical markers of oxidation of genistein by peroxyl radicals

in biological systems.

170

171

6.2 INTRODUCTION

In recent years, the potential role of soybeans in cancer prevention has received a lot
of attention. Epidemiological data indicate that consumption of soybean-containing diets is
associated with the lower incidence of certain human cancers in Asian compared to Caucasian
populations (King and Locke, 1980; Locke and King, 1980; Setchell et al., 1984). It is
hypothesized that many of the anticancer properties of soy could be attributable to their
isoflavone content (Messina er al., 1994; Messina and Barnes, 1991; Messina and Messina,
1991)

Genistein, the most abundant soy isoflavone, possesses a wide range of biological
properties that could contribute to the possible anticancer abilities of soybeans. Some of
these properties include its potent inhibition of protein tyrosine kinases (Akiyama et al.,
1987), DNA topoisomerases I and II (Okura et al., 1988), and ribosomal S6 kinase (Linassier
et al., 1990). The antioxidant ability of genistein is also believed to be linked to its anticancer
properties. Genistein has been reported to inhibit tumor promoter-induced formations of
hydrogen peroxide in vivo and in vitro in mouse skin (Wei et al., 1993). Other studies have
demonstrated the antioxidant abilities of genistein in a coupled linoleic acid/B-carotene system
(Pratt and Birac, 1979), in a liposomal system against UVA and UVB or peroxyl-radical
induced peroxidation (Record et al., 1995), and in a microsomal system using
Fe(II)/ADP/NADPH (Jha et al., 1985).

Genistein is thought to act as a chain-breaking antioxidant through its demonstrated

ability to scavenge hydroxyl, superoxide and peroxyl radicals (Record et al. , 1995; Wei et al. ,

172
1995; Wei et al., 1993). Since kinetic, spin-trapping and product analysis studies indicate that
the principal chain carrier of lipid peroxidation is the peroxyl radical (Matsuo et al., 1989),
the ability of genistein to scavenge peroxyl radicals would be critical towards its biological
relevance as an antioxidant. The structure of genistein (Gen-H) confers it with three
potentially donatable hydrogen atoms for radical scavenging (Figure 6.1). Lipid peroxyl
radicals (ROO') can react with genistein via abstraction of a hydrogen radical to yield lipid

hydroperoxides (ROOH) and the resonance-stabilized genistein phenoxyl radical (Gen')

(Equation 1).

Gen-H + ROO' -t ROOH + Gen' (1)

The genistein phenoxyl radical may react further with other peroxyl radical to yield a variety

of non-radical products (Equation 2).

Gen. + ROO' -t non-radical products (2)

Further understanding of this chain-breaking antioxidant mechanism for genistein
requires a knowledge of the products formed by the reaction between genistein and peroxyl
radicals. Characterization of these products could provide useful markers for genistein
antioxidant chemistry. So far, however, no reports have been published on the structural

determination of genistein oxidation products.

173

 

Genistein

T”’ T“’
CH,—c -N =N -—c—CH_,

CN CN

2,2'-Azobis(isobutyronitrile)

Figure 6.1. Structures for genistein and 2,2'-azobis(isobutyronitrile).

174

The present investigation was undertaken to profile the products formed by reaction
of genistein with alkylperoxyl radicals obtained by thermal decomposition of the lipophilic azo
initiator 2,2'-azobis(isobutyronitrile) (Figure 6.1). Use of the azo initiator enabled us to

generate initiating radicals at a known and constant rate.

6.3 EXPERIMENTAL PROCEDURES

6. 3. 1 Materials
Genistein was obtained by synthesis (Chang et al. , 1994). 2,2'-azobis(isobutyronitrile)
(AIBN) was obtained from Wako Chemical Company (Richmond, VA). All the solvents used

in the study were HPLC grade.

6. 3.2 HPLC

Analytical reverse-phase HPLC experiments were performed on a Supelcosil LC-18
(5 pm, 250 x 4.6 mm) column with a Supelguard LC-18 guard column (Supelco, Bellefonte,
PA) eluted with methanol/water (50:50, v/v) at a flow rate of 1.4 mL/min with continuous
helium sparging. A Waters 991 photodiode array detector (Milford, MA) set at a range of
200-300 nm was used to detect the fonnation of oxidation products. Preparative HPLC was
conducted on and LC-20 Recycling Preparative HPLC (Japan Analytical Instruments, Japan)
with a C-18 reversed-phase column (Jaigel, S-343-15, 15 pm, 250 x 20 m) using a mobile

phase of methanoVwater (70:30, v/v) at a flow rate of 4 mL/min and UV detection at 260 nm.

175

6. 3.3 Spectroscopy
‘H-NMR spectra were recorded on a Varian VXR 300 spectrometer in CD3OD
solution at ambient temperature. Mass spectra were acquired on a JEOL HX-110 double

focusing mass spectrometer (JEOL, Tokyo, Japan).

6. 3.4 Oxidation Studies with Genistein

In experiments studying oxidative fate of genistein over time, 250 pg (925 nmol) of
genistein and 20.5 mg (125 nmol) of AIBN were dissolved in 1 mL of oxygen-saturated
acetonitrile and the mixture was heated in screw-capped vials at 50°C. At various time-
points, 50 uL aliquots of the reaction mixture were removed to micro-centrifuge tubes for
product analysis, and rapidly chilled by immersion in ice. The resulting AIBN crystals were
removed by filtration, the supernatant dried under nitrogen and redissolved in the mobile
phase for analytical HPLC analysis. Larger-scale oxidations contained 75 mg (278 nmol)
genistein and 6.15 g (37.5 mmol) AIBN and were purified by preparative HPLC for collection

of oxidation products for structural analysis.

6.4 RESULTS

6. 4. 1 Formation of Genistein Oxidation Products
Successive analyses of the genistein-AIBN oxidation products over time indicated a
gradual disappearance of the genistein peak eluting at 12 min (peak G), concomitant with the

appearance of other fractions. As indicated in the representative chromatograms in Figure

176

6.2, two major reaction products were observed by 8 h of oxidation. One of these two
products eluted around 19 min (peak B) on the C-18 column, indicating that it was more non-
polar than the reactants. The second product at a retention time of 3.5 min (peak A), eluted
very close to the solvent fi'ont which indicated that it had higher polarity than genistein.
Virtually all of the genistein was oxidized by the end of 20 h. Photo diode array detection
revealed that both genistein and its two major oxidation products detected by HPLC displayed

UV absorption maxima at 210 and 260 nm.

6. 4.2 Chemical Structures of Oxidation Products

lHNMR spectrum for compound 1 (CD3OD) b 8.20 (1H, s, H—l), 7.40 (2H, d, J =
9 Hz, H-2', H-6'), 6.40 (2H, d, J = 9 Hz, H-3', H-S'), 6.30 (1H, d, J = 1.5 Hz, H-6), 6.20 (1H,
d, J = 1.5 Hz, H-7) and 3.40 (3H, s, OCH3). A structure for compound 1 is given in Figure
6.3. It appears that ring C in genistein is cleaved with methoxylation of the 4'-hydroxyl of
genistein to yield 2-dehydro-o-demethyl angolensol. This indicates that, during oxidation
reactions with peroxyl radicals, genistein forms a monophenoxyl radical at the 4'-position.
During work up of the reaction mixture, the phenoxyl radical may have reacted with methanol
to yield the stable methoxy derivative 1.. The MS of Compound 1 did not give the molecular
ion at m/z 302. However, a peak was observed at m/z 286 with 60% intensity which would
result fi'om cyclization to form a C ring with the concomitant loss of an oxygen atom (shown
in the scheme in Figure 6.3). This is not unusual for compounds with this type of a structure.

It is well known that chalcones undergo rapid cyclization to form flavones during MS

177

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0.50 “‘- ~
.1
.4
0.40—
1
.1
0.30——
3 3
0.20....
1
0.10— W
o,oo_:._ _J \kg__ * __‘__V
1 F 1 1 j 1 1 1 _r—‘— ’— 1 '1 r r—’“
10.00 20.00
Minutes
0.20-:
2 G
0.15——
3 I A
0.10;
'1 B
0.05—1 J
o 00.: "A ——’-~——.
‘ 1 1 1 1 I 1 1 1 1 l r 1 r r 1 1 1 r 1 1 1 1
5 00 10.00 15.00 20.00
mm
-1
.1
0.60 «
.J
30.40-4
0.20—~
.4
J J “A
0.00 - . “A“ M_/W_,'/\W_//¥_fi_ ’__‘ . _
L“ ‘ 1"" "'“r""“‘ I“ '“"1‘“‘""l " '“' 1_‘ ”'1"“"1“‘—'1‘ “"" "‘—"'1" ”—‘1' "_"r “‘ ‘ 1'—““
10.00 20.00
Minutes

Figure 6.2. Reversed-phase HPLC analysis of products formed by reaction between genistein
and AIBN-derived peroxyl radicals during incubation at 50°C for 3 h (top), 8 h (middle) and
20 h (bottom).

 

ocrr,

Compound 1
2-dehydro-0-demethy1 angolensol
(m/z 302, not detected)

ocrr,

 

 

m/z 270
100 %

Figure 6.3. Structure for Compound 1.

179

analyses. The base peak seen at m/z 270 was identical to genistein. The rest of the
fragmentation pattern was similar to the MS fragmentation of genistein.

The second oxidation product was not collected in a sufficiently pure form to yield
unambiguous structural identification. Preliminary NMR data indicates that oxidation of
genistein resulted in the opening of ring C of genistein to yield this product, while rings A and
B appeared to remain intact. Additional NMR experiments are necessary to determine the

structure of this product.

6.5 DISCUSSION
Azo initiators such as AIBN decompose thermally to yield alkyl radicals (R') which,

in turn, react with molecular oxygen to form alkylperoxyl radicals (ROO') (Equations 3 and

4).

R-N=N-R —+ 2R' + N2 (3)

R: + 02 —» R00 (4)

By use of such initiators, the generation of fiee radicals occurs at a known and specific
rate, and at a specific site (Niki, 1990). Admittedly, such initiators are not representative of
biological conditions. However, use of these chemically well-defined compounds can serve
as kinetically reproducible oxidative challenges (Ham and Liebler, 1995).

The fate of genistein during reaction with peroxyl radicals has not been examined

previously. As peroxyl radicals are recognized as being the primary chain carriers of

180

membrane peroxidation, profiling the stable end-products formed during the oxidation
reactions of genistein with peroxyl radicals would lead to a better understanding of genistein
antioxidant chemistry. Similar studies examining the oxidation products of a-tocopherol have
greatly clarified the reaction pathways involved in cr-tocopherol antioxidant chemistry (Liebler
er al., 1996; Liebler and Burr, 1995; Liebler and Burr, 1992; Liebler er al., 1991; Winterle
et al., 1984; Yamauchi et al., 1990; Yamauchi et al., 1989).

The results from this study indicate that oxidation of genistein with azo derived
alkylperoxyl radicals results in the formation of two major reaction products. It appears that
ring C of genistein undergoes cleavage in both of these products, whereas rings A and B
remain intact. Further advanced level NMR experiments are needed to yield unequivocal

chemical structures for the genistein oxidation product detected by HPLC as peak A.

SUMMARY AND CONCLUSIONS

As consumer interest in the use of plant-derived antioxidants in food products is
increasing, a systematic approach for rapid evaluation of plant extracts and purified
compounds for antioxidant activity is required. This study was primarily aimed at utilizing
the technique of fluorescence spectroscopy for development of assays to monitor lipid
peroxidation and to evaluate the efficacy of antioxidants.

Two fluorescence spectroscopic assays were developed for following lipid
peroxidation in membranes. The assays involved incorporation of the fluorescent probe
diphenylhexatriene-propionic acid into unilamellar vesicles and monitoring changes in its
intensity and anisotropy parameters as a function of time. The fluorescence intensity assay
was based on the quenching of the intensity of the probe due to the free radicals generated
during lipid peroxidation, whereas the fluorescence anisotropy assay reflected the decrease
in membrane fluidity that occurred as a result of cross-linking reactions in lipid peroxidation.
Validation studies using conjugated diene and hydroperoxide measurements confirmed that
the fluorescence spectroscopic assays were accurately reflecting the progress of initial stages
of peroxidation.

These new assays were used to evaluate the effrcacies of metal chelating and free

radical scavenging antioxidants. With the metal chelators, markedly different effects on rates

181

182

of lipid peroxidation were observed depending on the type of metal ion used to initiate lipid
peroxidation and on the molar ratio of chelator to metal ions employed.

Flavonoids and isoflavonoids were selected as the representative class of prototypic,
phenolic, free radical scavenging compounds. The abilities of these compounds to inhibit
metal-ion-induced and free-radical-induced peroxidations were compared. Flavonoids and
isoflavonoids exhibited superior antioxidant activities against metal-ion-induced
peroxidations, suggesting that the metal chelation abilities of these compounds are a major
contributor to their antioxidative abilities.

The results from our study confirmed earlier findings that the chemical structure of
these compounds is a large determinant of their antioxidant activities. The B-ring substitution
pattern was especially important, with hydroxyl group substitutions increasing activity and
methoxy substitutions diminishing it. Studies with isoflavonoids revealed that in addition to
the number of hydroxyls, the position of substitution was also important. Comparing
antioxidant activities of plant-derived isoflavonoids with their biological metabolites, it was
demonstrated that the activities for the metabolites were similar to or higher than the parent
compounds. This suggests that these compounds may maintain their antioxidant activities
under in vivo conditions. Additionally, the results of our studies with flavonoids and
isoflavonoids indicated that the fluorescence assays developed were usefiil tools for
characterization of structure-activity relationships within a class of structurally related
compounds.

Experiments to determine the localization of flavonoids and isoflavonoids in

membranes revealed that these compounds partitioned preferentially into the hydrophobic

 

183

core of the membrane, the major site of most peroxidation reactions. The flavonoids and
isoflavonoids substantially reduced fluidity in the inner regions of the membrane. This would
limit mobility of free radical species in the lipid bilayer, a possible structural explanation of
their antioxidant activities in membranes.

An oxidation product of the reaction between genistein and alkylperoxyl radicals in
homogenous system was characterized. This reaction product could be used as a marker of

genistein antioxidant chemistry.

FUTURE RESEARCH

Some suggestions for future research are listed below:

1. Experiments with chelators demonstrated that their effects on metal-ion-induced
peroxidation vary depending on the type of metal ions used to initiate peroxidation.
Measuring the redox potential of the fiee and complexed metals would be important for a full

understanding of the eflects of metal chelators on rates of metal-catalyzed lipid peroxidation.

2. The studies conducted with flavonoids and isoflavonoids suggest that metal chelation is
a significant contributor towards the antioxidant activities of these compounds. Conducting
experiments that can specifically measure their metal chelating capacity and can determine the
sites of chelation on the flavonoids and isoflavonoids will help in firrther clarification of their

mechanisms of antioxidant action.

3. The flavonoids and isoflavonoids tested in this study show great potential for use as
antioxidants. However, the fluorescence assays used for evaluating the efficacy of these
compounds were only following the initial stages of oxidation. To ensure that these

compounds maintain their antioxidant activities at later stages of lipid peroxidation, they

184

185

should be tested using an assay that measures some secondary products of peroxidation such

as hexanal or malondialdehyde.

4. To conclusively demonstrate the ability of these compounds to prolong the shelf-life of
foods, their activity should be tested in representative food lipid systems such as emulsions,

bulk oils and phospholipid membranes.

5. An oxidation product of the reaction between genistein and peroxyl radicals in
homogenous system was characterized in our study, a first important step in clarifying
isoflavone antioxidant chemistry. Additional experiments need to be conducted in membranes
using lipid-derived free radicals as this system would be more representative of biological

conditions.

LIST OF REFERENCES

LIST OF REFERENCES

Adlercreutz, H., Honjo, H., Higashi, A., Fotsis, T., Hamalainen, E., Hasegawa, T., and
Okada, H. 1991. Urinary excretion of lignans and isoflavonoid phytoestrogens in
Japanese men and women consuming a traditional Japanese diet. Am. J. Clin. Nutr.
54:1093-1100.

Adlercreutz, H., Fotsis, T., Bannwart, C., Wahalat, K., Makela, T., Brunow, G., and Hase,
T. 1986. Determination of urinary lignans and phytoestrogen metabolites, potential

antiestrogens and anticarcinogens, in urine of women on various habitual diets. J.
Steroid Biochem. 25:791-797.

Afanas’ev, I.B., Dorozhko, AI, Brodskii, A.V., Kostyuk, VA, and Potapovitch, AL 1989.
Chelating and free radical scavenging mechansims of inhibitory action of rutin and
quercetin in lipid peroxidation. Biochem. Pharmacol. 38: 1763-1769.

Afanas’ev, I.B., Ostrachovitch, E.A., Abramova, NE, and Korkina, LG. 1995. Different
antioxidant activities of bioflavonoid rutin in normal and iron-overloading rats.
Biochem. Pharmacol. 50:727-635.

Ahn, D.U., Wolfe, RH, and Sim, JS. 1993 The efi‘ect of free and bound iron on lipid
peroxidation in turkey meat. Poultry Sci. 72:209-215.

Akiyama, T., Ishida, J., Nakagawa, S., Ogawara, H., Watanabe, S.-I., Itoh, N., Shibuya, M.,
Fukami, Y. 1987. Genistein, a specific inhibitor of tyrosine-specific kinases. J. Biol.
Chem. 262:5592-5595.

Anthony, M.S., Clarkson, T.B., Hughes, C.L. Jr., Morgan, T.M., and Burke, G.L. 1996.
Soybean isoflavones improve cardiovascular risk factors without affecting the
reproductive system of peripubertal rhesus monkeys. J. Nutr. 126243-50.

Arora, A., and Strasburg, GM. 1997. Development and validation of fluorescence
spectroscopic assays to evaluate antioxidant efficacy. Application to metal chelators.
J. Am. Oil Chem. Soc. In Press.

Aust, S.D., Morehouse, LA. and Thomas, CE. 1985. Role of metals in oxygen radical

186

187

reactions. .1. Free Radic. Biol. Med 123-25.

Axelson, M., Kirk, D.N., Farrant, R.D., Cooley, G., Lawson, A.M., and Setchell, K.D.R.
1982. The identification of the weak oestrogen equol (7-hydroxy-3-(4'-
hydroxyphenyl) chroman) in human urine. Biochem. J. 201 2353-3 57.

Axelson, M., Sjovall, J., Gustalfsson, BE, and Setchell, K.D.R. 1984. Soya - a dietary
source of the non-steroidal oestrogen equol in man and animals. J. Endocrinol.
102249-56.

Bannwart, C., Fotsis, T., Heikkinen, R., and Adlercreutz, H. 1984. Identification of the
isoflavonoid phytoestrogen daidzein in human urine. C lin. Chim. Acta 136: 165-172.

Barclay, L.R.C. 1993 Model biomembranes: Quantitative studies of peroxidation,
antioxidant action, partitioning, and oxidative stress. Can. J. Chem. 71:1-16.

Barclay, L.R.C., Baskin, K.A., Dakin, K.A., Locke, SJ, and Vinquist, MR. 1990. The
antioxidant activities of phenolic antioxidants in free radical peroxidation of
phospholipid membranes. Can. J. Chem. 68:2258-2269.

Barclay, L.R.C., and Vinquist, MR. 1994. Membrane peroxidation: inhibiting effects of
water soluble antioxidants on phospholipids of different charge types. Free Radic.
Biol. Med 16:779-788.

Beck, A., Heissler, D., and Duportail, G. 1990. 1-Palrnitoyl-2-[3-(diphenylhexatrienyl)
propanoyl]-sn-glycero-3 -phosphoethanolamine as a fluorescent membrane probe.
Synthesis and partitioning properties. Chem. Phys. Lipids 55: 13-24.

Bennetts, H.W., Underwood, BI, and Shier, FL. 1946. A specific breeding problem of
sheep on subterranean clover pastures in Western Australia. Aust. Vet. J. 2222-12.

Bentley, K.L., Thompson, L.K., Klebe, RI, and Horowitz, PM. 1985. Fluorescence
polarization: A general method for measuring ligand binding and membrane
microviscosity. Biotechniques 3:356-366.

Bisby, RH. 1990. Interactions of vitamin E with free radicals and membranes. Free Rad.
Res. Commun. 8 2299-3 06.

Bokkenheuser, V.D., and Winter, J. 1987. Hydrolysis of flavonoids by human intestinal
bacteria. Prog. Clin. Biol. Res. 280: 143-145.

Borenstain, V., and Barenholz, Y. 1993. Characterization of liposomes and other lipid
assemblies by multiprobe fluorescence polarization. Chem. Phys. Lipids 64: 1 17-127.

188

Bors, W., Heller, W., Michel, C., and Saran, M. 1990. Flavonoids as antioxidants:
Determination of radical-scavenging efficiencies. Meth. Enzymol. [86343-3 55.

Brandi, ML. 1992. Flavonoids: Biochemical effects and therapeutic applications. Bone
Mineral 19:S3-Sl4.

Brasseur, T. 1989. Anti-inflammatory properties of flavonoids. J. Pharm. Belg. 44:235-
241.

Braughler, J .M., Chase, R.L., and Pregenzer, J.F. 1987. Oxidation of ferrous ions during
peroxidation of lipid substrates. Biochim. Biophys. Acta 921:457-464.

Brown, JP. 1980. A review of the genetic effects of naturally occurring flavonoids,
anthroquinones, and related compounds. Mutation Res. 75 2243-277.

Buettner, GR 1993. The pecking order of fiee radicals and antioxidants: lipid peroxidation,
a-tocopherol, and ascorbate. Arch. Biochem. Biophys. 300:535-543.

Buettner, GR, and Jurkiewicz, BA. 1996 Catalytic metals, ascorbate and free radicals:
Combinations to avoid. Radiat. Res. [45:532-541.

Cai, Q., and Wei, H. 1996. Effect of dietary genistein on antioxidant enzyme activities in
SENCAR mice. Nutr. Cancer. 25 :1-7.

Caragay, AB. 1992. Cancer-preventive foods and ingredients. Food T echnol. 46:65-68.

Cassidy, A, Bingharn, S., and Setchell, K. 1995. Biological effects of isoflavones in young
women: importance of the chemical composition of soyabean products. Brit. J. Nut.
74:587-601.

Chang, Y.-C., and Nair, MG. 1995. Metabolism of daidzein and genistein by intestinal
bacteria. .1. Natural Prod 58:1892-1896.

Chang, Y.-C., Nair, MG, and Nitiss, IL 1995. Metabolites of daidzein and genistein and
their biological activities. .1. Natural Prod. 58: 1901-1905.

Chang, Y.-C., Nair, M.G., Santell, RC, and Helferich, W.G. 1994. Microwave-mediated
synthesis of anticarcinogenic isoflavones fiom soybeans. J. Agric. Food Chem.
42:1869-1871.

Chapman, D. 1984. Physicochemical properties of phospholipids and lipid-water systems.
In Liposome Technology, Gregoriadis, G. (ed.), p. 1-18, Vol. 1, CRC Press,Boca
Raton, FL.

189

Chen, Z.Y., Chan, PT, Ho, K.Y., Fung, K.P., and Wang, J. 1996. Antioxith activity of
natural flavonoids is governed by number and location of their aromatic hydroxyl
groups. Chem. Phys. Lipids 79:157-163.

Chevion, M. 1988. A site-specific mechanism for free radical induced biological damage:
The essential role of redox-active transition metals. Free Radic. Biol. Med 5:27-37.

Choe, M., Jackson, 0, and Yu, BR 1995. Lipid peroxidation contributes to age-related
membrane rigidity. Free Radio. Biol. Med 18:977-984.

Cholbi, MR, Paya, M., and Alcaraz, M.J. 1991. Inhibitory effects of phenolic compounds
on CCl4-induced microsomal lipid peroxidation. Experentia 47: 195-199.

Clark, W.G., and MacKay, EM. 1950. The absorption and excretion of rutin and related
flavonoid substances. .1. Am. Med Assoc. 143:1411-1415.

Clarke, R., van den berg, H.W., and Murphy, RF. 1990. Reduction of the membrane
fluidity of human breast cancer cells by tamoxifen and 17B-estradiol. J. Natl. Cancer
Inst. 82:1702-1705.

Cook, NC, and Samman, S. 1996. Flavonoids - chemistry, metabolism, cardioprotective
effects, and dietary sources. J. Nutr. Biochem. 7:66-76.

Cotelle, N., Bemier, J.-C., Catteau, J.-P., Pommery, J., Wallet, J.-C., and Gaydou, EM.
1996. Antioxidant properties of hydroxy-flavones. Free Radio. Biol. Med 20:35-
43.

Cotelle, N., Bemier, J.L., Henichart, J .P., Catteau, J .P., Gaydou, E., and Wallet, J .C. 1992.
Scavenger and antioxidant properties of ten synthetic flavones. Free Radio. Biol.
Med 13:211-219.

Coupeland, J .N., and McClements, DJ. 1996. Lipid oxidation in food emulsions. Trends
Food Sci T eohnol. 7:83-91.

Coward, L., Barnes, N.C., Setchell, K.D.R., and Barnes, S. 1993. Genistein, daidzein and
their B-glycoside conjugates: Antitumor isoflavones in soybean foods from American
and Japanese diets. .1. Agric. Food Chem. 41: 1961-1967.

Das, UK. 1994. Naturally occurring flavonoids: stmcture, chemistry, and high-performance
liquid chromatography methods for separation and characterization. Meth. Enzymol.
234:410-420.

de Hingh,Y.C.M., Meyer, J ., Fischer, J .C., Berger, R., Smeitink, J AM. and Op den Kamp,

190

J .A.F . 1995 Direct measurement of lipid peroxidation in submitochondrial particles.
Biochemistry 34: 12755-12760.

De Whalley, C.V., Rankin, S.M., Henlt, J.R.S., Jessup, W., and Leake, DS. 1990.
Flavonoids inhibit the oxidative modification of low density lipoproteins by
macrophages. Biochem. Pharmacol. 39: 1743-1750.

Dewick, PM. 1988. Isoflavonoids. In The Flavonoids, Harbom, J.P. (ed.), p. 125-210,
Chapman and Hall, New York, NY.

Dikalov, S., Alov, P. and Rangelova, D. 1993 Role of iron ion chelation by quinones in their
reduction, OH-radical generation and lipid peroxidation. Biochem. Biophys. Res.
Commun. [95:113-119.

Ding, A.-H., and Chan, PC. 1984 Singlet oxygen in copper-catalyzed lipid peroxidation in
erythrocyte membranes. Lipids 19:278-284.

Dinis, T.C.P., Almeida, L.M., and Madeira, V.M.C. 1993. Lipid peroxidation in
sarcoplasmic reticulum membranes: Effect on functional and biophysical properties.
Arch. Biochem. Biophys. 301 :256-264.

Dinis, T.C.P., Madeira, V.M.C. and Almeida, L.M. 1994 Action of phenolic derivatives
(acetaminophen, salicylate, and 5-aminosalicylate) as inhibitors of membrane lipid
peroxidation and as peroxyl radical scavengers. Arch. Biochem. Biophys. 315:161-
169.

Dix, TA, Aikens, J. 1993 Mechanisms of biological relevance of lipid peroxidation initiation.
Chem. Res. Toxicol. 6:2-13.

Dormandy, TL, and “fickens, D.G. 1987. The experimental and clinical pathology of diene
conjugation. Chem. Phys. Lipids 45:356-364.

Duthie, G.G. 1991. Antioxidant hypothesis of cardiovascular disease. Trends Food Sci.
T eohno]. 2:205-207.

Ferrandiz, ML, and Alcaraz, M.J. 1991. Anti-inflammatory activity and inhibition of
arachidonic acid metabolism bu flavonoids. Agents Action 32:283-288.

Folman, Y., and Pope, GS. 1969. Effect of northisterone acetate, dimethylstilbesterol,
genistein and coumesterol on uptake of [3H]oestradiol by uterus, vagina and skeletal
muscle of immature mice. J. Endocrinol. 44:213-218.

Folman, Y., and Pope, GS. 1966. The interaction in the immature mouse of potent

191

oestrogens with coumesterol, genistein and other uterovaginotrophic compounds of
low potency. J. Endocrinol. 34:215-225.

Formica, J .V., and Regelson, W. 1995. Review of the biology of quercetin and related
bioflavonoids. Food Chem. T oxioo]. 33:1061-1080.

Frankel, E. 1993. In search of better methods of evaluating natural antioxidants and
oxidative stability in food lipids. Trends Food Sci. T eohnol. 4:220-225.

Frankel, E.N., Kanner, 1, German, J.B., Parks, E., and Kinsella, J.E. 1993. Inhibition of
oxidation of low-density lipoprotein by phenolic substances in red wine. The Lancet
341:454-457.

Fuhrrnan, B., Lavy, A, and Aviram, M. 1995. Consumption of red wine with meals reduces
the susceptibility of human plasma and low-density lipoprotein to lipid peroxidation.
Am. J. Clin. Nutr. 61:549-554.

F ujii, T., Hiramoto, Y., Terao, J. and Fukuzawa, K. 1991 Site-specific mechanisms of
initiation by chelated iron and inhibition by a-tocopherol of lipid peroxide-dependent
lipid peroxidation in charged micelles. Arch. Biochem. Biophys. 284: 120-126.

Garrison, M.D., Doh, L.M., Potts, R.O., and Abraham, W. 1994. Fluorescence
spectroscopy of 9-anthroyloxy fatty acids in solvents. Chem. Phys. Lipids 70: 155-
162.

Girotti, AW., and Thomas, JP. 1984. Superoxide and hydrogen peroxide-dependent lipid
peroxidation in intact and triton-dispersed erythrocyte membranes. Biochem.
Biophys. Res. Commun. 118:474-480.

Graf, E. 1994 Copper (11) ascorbate: A novel food preservation system. J. Agric. Food
Chem. 42:161-1619.

Gray, J .I., and Monahan, F .J . 1992. Measurement of lipid oxidation in meat and meat
products. Trends Food Sci. Technol. 3:315-319.

Gugler, R., Leschik, M., and Dengler, H.J. 1975. Disposition of quercetin in man after
single oral and intravenous doses. Eur. J. Clin. Pharmacol. 9:229-234.

Gutteridge, J.M.C. 1988. Lipid peroxidation: some problems and concepts. In Oxygen
Radicals and Tissue Injury, B. Halliwell (ed.), p. 9-19, The Upjohn Company,
Bethesda, MD.

Gutteridge, J.M.C., and Halliwell, B. 1990. The measurement and mechanism of lipid

192

peroxidation in biological systems. TIBS 15:129-135.

Hackett, AM. 1986. The metabolism of flavonoid compounds in mammals. Prog. Clin.
Biol. Res. 213:177-194.

Haigh, R. 1986. Safety and necessity of antioxidants: EEC approach. Fd Chem. Toxic.
24:1031-1034.

Halliwell, B. 1994. Free radicals and antioxidants: A personal view. Nutr. Rev. 52:253-265.

Halliwell, B. 1990. How to characterize a biological antioxidant. Free Rad Res. Commun.
9: 1-32.

Halliwell, B., and Chirico, S. 1993. Lipid peroxidation: its mechanism, measurement, and
significance. Am. J. Clin. Nutr. 57:715S-725S.

Halliwell, B., and Gutteridge, J .M.C. 1989. In Free Radicals in Biologt and Medicine, 2'“I
edition, Clarendon Press, Oxford, UK.

Ham, A.-J.L., and Liebler, DC. 1995. Vitamin E oxidation in rat liver mitochondria.
Biochemistry 34 :5754-5761 .

Hanasaki, Y., Ogawa, S., and Fukui, S. 1994. The correlation between active oxygens
scavenging and antioxidative effects of flavonoids. Free Radio. Biol. Med 16:845-
850.

Havsteen, B. 1983. Flavonoids, a class of natural products of high pharmacological potency.
Biochem. Pharmacol. 32:1 141-1148.

Hawksworth, G., Drasar, BS, and Hill, M.J. 1971. Intestinal bacteria and the hydrolysis of
glycosidic bonds. .1. Med Microbial. 4 :451-459.

Heilmann, J ., Merfort, 1., and Weiss, M. 1995. Radical scavenger activity of different 3',4'-
dihydroxyflavonols and 1,5-dicaffeoquuinic acid studied by inhibition of
chemiluminescence. Planta Med 61:435-438.

Hendrickson, H.P., Kaufman, AD, and Lunte, CE. 1994. Electrochemistry of catechol-
containing flavonoids. J. Pharmao. Biomed Anal. 12:325-334.

Hertog, M.G.L., Hollman, P.C.H., and Katan, MB. 1992. Content of potentially
anticarcinogenic flavonoids of twenty eight vegetables and nine fiuits commonly
consumed in The Netherlands. J. Agric. Food Chem. 40:23 79-23 83.

193

Hertog, M.G.L., Hollman, P.C.H., Katan, MB, and Kromhout, D. 1993a. Intake of
potentially anticarcinogenic flavonoids and their determinants in The Netherlands.
Nutr. Cancer 20:21-29.

Hertog, M.G.L., Hollman, P.C.H., and van de Putte, B. 1993b. Content of potentially
anticarcinogenic flavonoids of teas infusions, wines and fiuit juices. J. Agric. Food
Chem. 41:1242-1246.

Hirose, M., Hoshiya, T., Akagi, K., Futakuchi, M., and Ito, N. 1994. Inhibition of mammary
gland carcinogenesis by green tea catechins and other naturally occurring antioxidants
in female Sprague Dawley rats pretreated with 7,12-dimethylbenz[a]nthrene. Cancer
Lett. 83:149-156.

Hodgson, J.M., Croft, K.D., Puddey, I.B., Mori, TA, and Beilin, L.J. 1996. Soybean
isoflavonoids and their metabolic products inhibit in vitro lipoprotein oxidation in
serum. J. Nutr. Biochem. 72664-669.

Holley, AE., and Slater, TE 1991. Measurement of lipid hydroperoxides in normal human
blood plasma using HPLC-chemiluminescence linked to a diode array detector for
measuring conjugated dienes. Free Rad Res. Commun. 15 :51-63.

Hollman, P.C.H., Hertog, M.G.L., and Katan, MB. 1996. Role of dietary flavonoids in
protection against cancer and coronary disease. Biochem. Soc. Trans. 24 2785-789.

Hollman, P.C.H., de Vries, J.H.M., van Leeuwen, S.D., Mengelers, M.J.B., and Katan, MB.
1995. Absorption of dietary quercetin glycosides and quercetin in healthy ileostomy
volunteers. Am. J. Clin. Nutr. 62:1276-1282.

Hope, M.J., Bally, M.B., Mayer, L.D., Janoff, AS, and Cullis, PR. 1986. Generation of
multilamellar and unilamellar phospholipid vesicles. Chem. Phys. Lipids 40289-107.

Horan, K.L., Lutzke, B.S., Cazers, A.R., McCall, J.M., and Epps, DE. 1994. Kinetic
evaluation of lipophilic inhibitors of lipid peroxidation in DLPC liposomes. Free
Radio. Bio]. Med 17:587-596.

Hu, J .P., Colomme, M., Lasure, A, De Bruyne, T., Pieters, L., Vlietinck, A, and Van den
Berghe, DA. 1995. Structure-activity relationship of flavonoids with superoxide
scavenging activity. Biol. Trace Element Res. 4 7:327-331.

Husain, S.R., Cillard, J., and Cillard, P. 1987. Hydroxyl radical scavenging activity of
flavonoids. Phytaohemistry 26 :2489-249 1 .

Hutchins, A.M., Slavin, J .L., and Lampe, J .W. 1995. Urinary isoflavonoid phytoestrogen

194

and lignan excretion after consumption of fermented and unfermented soy products.
.1. Am. Diet. Assoc. 95:545-551.

Ingham, J.L. 1983. Naturally occurring isoflavonoids (1855-1981). Fortschritte Chemie
Organisoher Naturstaffe 4321-265.

Ioku, K., Tsushida, T., Takei, Y., Nakatani, N., and Terao, J. 1995. Antioxidative activity
of quercetin and quercetin monoglucosides in solution and in phospholipid bilayers.
Biaohim. Biophys. Acta 1234:99-104.

Ivessa, E.N., Kalb, E., Paltauf, F. and Herrnetter, A. 1988. Diphenylhexatrienyl propanoyl
hydrazyl stachyose: A new oligosaccharide derivative of diphenylhexatriene.
Synthesis and fluorescence properties in artificial membranes. Chem Phys. Lipids
49:185-195.

Jack. C.I.A., Jackson, M.J., Ridgeway, E., and Hind, C.R.K. 1991. Octadeca-9,11-dienoic
acid - a measurement of free radical activity or a marker of infection in the lung?
Clin. Sci. 81: 17P-17P (abstr.).

Jemiola-Rzerninska, M., Kruk, J ., Skowronek, M., and Strzalka, K. 1996. Location of
ubiquinone homologues in liposome membranes studied by fluorescence anisotropy
of diphenyl-hexatriene and trimethylammonium-diphenyl-hexatriene. Chem. Phys.
Lipids 79:55-63.

Jha, H.C., Recklinghausen, G.V., and Zilliken, F. 1985. Inhibition of in vitro microsomal
lipid peroxidation by isoflavonoids. Biochem. Pharmacol. 34:1367-1369.

J iang, Z.-Y., Woollard, A.C.S. and Wolff, SM. 1991. Lipid hydroperoxide measurement by
oxidation of Fe2+ in the presence of xylenol orange. Comparison with the TBA assay
and an iodometric method. Lipids 26:853-856.

Joannou, G.E., Kelly, G.E., Reeder, A.Y., Waring, M., and Nelson, C. 1995. A urinary
profile study of dietary phytoestrogens. The identification and mode of metabolism
of new isoflavonoids. J. Steroid Biochem. Maleo. Biol. 54:167-184.

Jourd’heuil, D., Vaananen, P., and Meddings, J .B. 1993. Lipid peroxidation of the brush-
border membrane: Membrane physical properties and glucose transport. Am. J.
Physiol. 264261009-G1015.

Jovanovic, S.V., Jovanovic, I., and Josimovic, L. 1992. Electron-transfer reactions of alkyl
peroxyl radicals. J. Am. Chem. Soc. 114:9018-9022.

Jovanovic, S.V., Steenken, S., Tosic, M., Maryanovic, B., and Simic, MG. 1994.

195

F lavonoids as antioxidants. J. Am. Chem. Soc. 116:4846-4851.

Kagan, V.E., Serbinova, E.A., Bakalova, R.A., Stoytchev, Ts.S., Erin, A.N., Prilipko, LL,
and Evstigneeva, RP. 1990. Mechanisms of stabilization of biomembranes by alpha-
tocopherol. The role of the hydrocarbon chain in the inhibition of lipid peroxidation.
Biochem. Pharmacol. 40: 2403 -241 3.

Kandaswarni, C., and Middleton, E. 1994. Free radical scavenging and antioxidant activity
of plant flavonoids. Adv. Expta]. Med Biol. 366:35 1-3 76.

Kaneko, T., Kaji, K., and Matsuo, M. 1994. Protection of linoleic acid hydroperoxide-
induced cytotoxicity by phenolic antioxidants. Free Radic. Biol. Med 16:405-409.

Kanner, 1., German, J.B., and Kinsella, J .E. 1987. Initiation of lipid peroxidation in
biological systems. CR C. Crit. Rev. Food Sci. Nutr. 25:317-364.

Kanner, J ., Harel, S. and Hazan, B. 1986 Muscle membrane lipid oxidation by an iron redox
cycle system: Initiation by oxy radicals and site-specific mechanism. J. Agric. Food
Chem. 34:506-510.

Kanner, J ., Hazan, B., and Doll, L. 1988. Catalytic “fi'ee iron” in muscle foods. J. Agric.
Food Chem. 36:412-415.

Kellogg, E.W., III, and Fridovich, I. 1975. Superoxide, hydrogen peroxide, and singlet
oxygen in lipid peroxidation by a xanthine oxidase system. J. Biol. Chem. 25 0:88 12-
8817.

Kelly, G.E., Joannou, G.E., Reeder, AY., Nelson, C., and Waring, MA. 1995. The variable
metabolic response to dietary isoflavones in humans. Prao. Soc. Exp. Biol. Med
208140-43.

Kelly, G.E., Nelson, C., Waring, MA, Joannou, GE, and Reeder, A.Y. 1993. Metabolites
of dietary (soya) isoflavones in human urine. Clin. Chim. Acta 223:9-22.

King, H., and Locke, RB. 1980. Cancer mortality among Chinese in the United States. J.
Natl. Cancer Inst. 65:1141-1148.

Kinsella, J E Frankel, E., German, B., and Kanner, J. 1993. Possible mechanisms for the
protective role of antioxidants in wine and plant foods. Food T echnal. 4 7:85-89.

Knekt, P. 1996. Flavonoid intake and coronary mortality in Finaland: A cohort study. BMJ
312:478-481.

196

Kuhnau, J. 1976. The flavonoids. A class of semi-essential food components: Their role
in human nutrition. Wld Rev. Nutr. Diet. 24:117-191.

Kurechi, T., and Kunugi, A. 1983. Studies on the antioxidants XVII: Photooxidation

products of concomitantly used butylated hydroxyanisole and pr0pyl gallate. J. Am.
0i] Chem. Soc. 60:109-113.

Kuypers, F.A., Van den Berg, J .J .M., Schalkwijk, C., Roelofsen, B. and Op den Kamp,
J .AF . 1987 . Parinaric acid as a sensitive fluorescent probe for the determination of
lipid peroxidation, Biochem. Biophys. Acta 921 .2266-274.

Labuza, TR 1971 Kinetics of lipid oxidation in foods. CRC Crit. Rev. Food Techno]. 2:355-
405.

Lacort, M., Leal, AM., Mariana, L., Martin, C., Martinez, R., and Ruiz-Larrea, MB. 1995.
Protective efl‘ect of estrogens and catecholestrogens against peroxidative membrane
damage in vitro. Lipids 30:141-146.

Lakowicz, J .R. 1983. In Principles of Fluorescence Spectroscopy, Plenum Press, New
York, NY.

Laranjinha, J .A.N., Almeida, L.M. and Madeira, V.M.C. 1992 Lipid peroxidation and its
inhibition in low density lipoproteins: quenching of ois-parinaric acid fluorescence.
Arch. Biochem. Biophys. 29 7: 147-1 54.

Lasic, DD. 1988. The mechanism of vesicle formation. Biochem J. 256: 1-1 1.

Lasic, DD, and Papahadjopoulos, D. 1995. Liposomes revisited. Science 267: 1275-1276.

Laughton, M.J., Evans, P.J., Moroney, M.A., Hoult, IRS, and Halliwell, B. 1991.
Inhibition of mammalian 5-lipoxygenase and cyclo-oxygenase by flavonoids and

phenolic dietary additives. Biochem. Pharmacol. 42: 1673-1681.

Lentz, BR. 1993 Use of fluorescent probes to monitor molecular order and motions within
liposome bilayers, Chem. Phys. Lipids 64:99-116.

Liebler, DC. 1993. The role of metabolism in the antioxidant function of vitamin E. Crit.
Rev. T oxiool. 23:147-169.

Liebler, D.C., and Burr, J .A. 1995. Antioxidant stoichiometry and the oxidative fate of
vitamin E in peroxyl radical scavenging reactions. Lipids 30:789-793.

Liebler, D.C., and Burr, IA. 1992. Oxidation of vitamin E during iron-catalyzed lipid

197

peroxidation: Evidence for electron-transfer reactions of the tocopheroxyl radical.
Biochemistry 31 :8278-8284.

Liebler, D.C., Burr, J .A., Philips, L., and Ham, A.J.L. 1996. Gas chromatography-mass

spectrometry analysis of vitamin E and its oxidation products. Anal. Biochem.
236:27-34.

Liebler, D.C., Kaysen, KL, and Burr, JA 1991. Peroxyl radical trapping and
autooxidation reactions of a-tocopherol in lipid bilayers. Chem. Res. T oxicol. 4 :89-
93.

Linassier, C., Pierre, M., Le Peco, J.-B., and Pierre, J. 1990. Mechanisms of action in NIH-

3T3 cells of genistein, an inhibitor of EGF receptor tyrosine kinase activity. Biochem.
Pharmacol. 39:73 7-744.

Locke, F .B., and King, H. 1980. Cancer mortality risk among Japanese in the United States.
J. Natl. Cancer Inst. 65:1149-1156.

Loliger, J. 1992. The use of antioxidants in foods. In Free Radicals and Food Additives,
0.1. Aruoma and B. Halliwell (eds), Ch. 6, p. 121-150.

Lundh, T. 1995. Metabolism of estrogenic activity in domestic animals. P.S.E.B.M.
208:33-39.

MacDonald, RC, MacDonald, RI, Menco, B.P.M., Takeshita, K., Subbarao, N.K., and Hu,
L. 1991. Small-volume extrusion apparatus for preparation of large, unilamellar
vesicles. Biaohim. Biophys. Acta 1061:297-303.

Maiorino, M., Zamburlini, A, Roveri, A, and Ursini, F. 1995. Copper-induced lipid
peroxidation in liposomes, micelles, and LDL: which is the role of vitamin E. Free
Radic. Biol. Med 18:67-74.

Manach, C., Regerat, F ., Texier, 0., Agullo, G., Demigne, C., and Remesy, C. 1996.

Bioavailability, metabolism and physiological impact of 4-oxo flavonoids. Nutr. Res.
16:517-544. '

Marcollet, M., Bastide, P., and Tronche, P. 1969. Angioprotective effects of
anthocyanisides of Vaccinum myrttillus shown by the release of lactate dehydrogenase

(LDH) and their cardiac isoenzymes in rats subjected to swimming test. C.R. Soc.
Biol. 163:1786-1789.

Martin, P.M., Horwitz, K.B., Ruyan, D.S., McGuire, W.L. 1978. Phytoestrogen interaction

198

with estrogen receptors in human breast cancer cells. Endocrinology 103:1860-1867.

Mason, IT 1994. Properties of phosphatidylcholine bilayers as revealed by mixed-acyl
phospholipid fluorescent probes containing n-(9-anthroyloxy) fatty acids. Biaohim.
Biophys. Acta 1194:99-108.

Mateo, C.R., Lillo, M.P., Gonzalez-Rodriguez, J. and Acuna, AV. 1991 Molecular order
and fluidity of the plasma membrane of human platelets from time-resolved
fluorescence polarization. Eur. Biophys. J. 20:41-52.

Mayoufl A.M., and Lemmetyinen, H. 1993. Oxidation-reduction reaction of photoinduced
benzyl radicals in water: para substituent effect. J. Phatachem. Photabial. A:Chem.
73:205-211.

McKenna, R., Kedzy, F.J., and Epps, DE. 1991. Kinetic analysis of the free radical-induced
lipid peroxidation in human erythrocyte membranes: evaluation of potential

antioxidants using ois-parinaric acid to monitor peroxidation. Anal. Biochem.
196:443-450.

Melton, S.L. 1983 Methodology for following lipid oxidation in muscle foods. Food T eohnol
37:111-116.

Menon, L.G., Kutton, R., and Kutton, G. 1995. Inhibition of lung metastasis in mice

induced by B16F10 melanoma cells by polyphenolic compounds. Cancer Lett.
95:221-225.

Messina, M., and Barnes, S. 1991. The role of soy products in reducing risk of cancer. J.
Natl. Cancer Inst. 83:541-546.

Messina, M. and Messina, V. 1991. Increasing use of soyfoods and their potential role in
cancer prevention. J. Am. Diet. Assoc. 91:836-840.

Messina, M.J., Persky, V., Setchell, K.D.R., and Barnes, S. 1994. Soy intake and cancer
risk: A review of the in vitro and in vivo data. Nutr. Cancer 21:113-131.

Middleton, E., Jr. 1984. The flavonoids. Trends Pharmacol. Sci. 5:335-338.

Miller, D.M., Buettner, GR. and Aust, SD. 1990. Transition metals as catalysts of
“autoxidation” reactions, Free Radio. Biol. Med 8:95-108.

Miller, D.M., Spear, NH. and Aust, SD. 1992 Effects of defenioxarnine on iron-catalyzed
lipid peroxidation. Arch. Biochem. Biophys. 295:240-246.

199

Millington, A.J., Francis, CM, and McKeown, NR. 1964. Wether bioassay of annual
pasture legumes: II. The oestrogenic activity of nine strains of Trifolium
subterraneum L. Aust. J. Agric. Res. 15:527-536.

Minotti, G. 1992 The role of endogenous nonheme iron in microsomal redox reactions. Arch.
Biaohim. Biophys. 29 7: 189-198.

Minotti, G. 1993. Sources and role of iron in lipid peroxidation. Chem. Res. T oxicol.
6:134-146.

Minotti, G., and Aust, SD. 1987 The role of iron in initiation of lipid peroxidation. Chem.
Phys. Lipids 44: 191-208.

Minotti, G., and Aust, SD. 1992 Redox cycling of iron and lipid peroxidation. Lipids
27:219-226. '

Miyazawa, T., Yasuda, K., and Fujimoto, K. 1987. Cherniluminescence-high performance
liquid chromatography of phosphatidylcholine hydroperoxide. Anal. Letters 20:915-
925.

Monahan, F.J., Gray, J.I., Asghar, A, Haug, A, Strasburg, G.M., Buckley, DJ, and
Morrissey, PA. 1994. Influence of diet on lipid oxidation and membrane structure
in porcine muscle microsomes. J. Agric. Food Chem. 42:59-63.

Montensinos, M.C., Ubeda, A., Terencio, M.C., Paya, M., and Alcaraz, M.J. 1995.
Antioxidant profile of mono- and dihydroxylated flavone derivatives in fi'ee radical
generating systems. Z. Naturfarsch. 50c:552-560.

Mora, A, Paya, M., Rios, J .L., and Alcaraz, M.J. 1990. Structure-activity relationships of
polymethonylavenes and other flavonoids as inhibitors of non-enzymic lipid
peroxidation. Biochem. Pharmacol. 40:793-797.

Morel, 1., Lescoat, G., Cillard, P., and Cillard, J. 1994. Role of flavonoids and iron chelation
in antioxidant action. Meth. Enzymol. 234 :43 7-443.

Morel, 1., Lescoat, G., Cogrel, P., Sergent, 0., Pasdeloup, N., Brissot, P., Cillard, P., and
Cillard, J. 1993. Antioxidant and iron-chelating activities of the flavonoids catechin,
quercetin and diosmetin on iron-loaded rat hepatocyte cultures. Biochem.
Pharmacol. 45 : 13-19.

Morel, 1., Sergent, 0., Cogrel, P., Lescoat, G., Pasdeloup, N., Brissot, P., Cillard, P., and
Cillard, J. 1995. EPR study of antioxidant activity of the iron chelators proverdin
and hydroxypyrid-4-one in iron-loaded hepatocyte culture: Comparison with that of

200

desferrioxamine. Free Radic. Biol. Med 18:303-310.

Morgan, C.G., Hudson, 8., and Wolber, PK. 1980. Photochemical dimerization of parinaric
acid in lipid bilayers. Proo. Nat]. Acad Sci. USA 77:26-30.

Monissey, PA, Quinn, PB, and Sheehy, P.J.A 1994. Newer aspects of micronutrients in
chronic disease: Vitamin E. Proo. Nutr. Soc. 53:571-582.

Namiki, M. 1990. Antioxidants/antimutagens in food. CR C. Crit. Rev. Food Sci. Nutr.
29:273-300.

Niki, E. 1990. Free radical initiators as source of water- or lipid-soluble peroxyl radicals.
Meth. Enzymal. 186: 100-108.

Niki, E., Kawakami, A, Saito, M., Yamamoto, Y., Tsuchiya, J., and Kamiya, Y. 1985.
Effect of phytyl side chain of vitamin E on its antioxidant activity. J. Biol. Chem.
260:2191-2196.

Niki, E., Yamamoto, Y., Komuro, E., and Sato, K. 1991. Membrane damage due to lipid
oxidation. Am. J. Clin. Nutr. 53:2015-2055.

Nilsson, A, Hill, J.L., and Davies, H.L. 1967. An in vitro study of formononetin and
biochanin A metabolism in rumen fluid from sheep. Biaohim. Biophys. Acta 148292-
98.

Okura, A, Arakawa, H., Oka, H., Yoshinari. T., and Monden, Y. 1988. Effect ofgenistein
on topoisomerase activity and on the growth of [val 12]Ha-ras-transformed NIH 3T3
cells. Biochem. Biophys. Res. Commun. 15 7 2 183-189.

Pace-Asciak, C.R., Hahn, S., Diamandis, E.P., Soleas, G., and Goldberg, D.M. 1995. The
red wine phenolics trans-resveratrol and quercetin block human platelet aggregation

and eicosanoid synthesis: Implications for protection against coronary heart disease.
Clin. Chim. Acta 235 :207-219.

Peterson, G., and Barnes, S. 1996. Genistein inhibits both estrogen and grth factor -
stimulated proliferation of human breast cancer cells. Cell Growth Difl'er. 7: 1345-
1351.

Porter, W.L. 1993 Paradoxical behavior of antioxidants in food and biological systems.
T oxioal. Indust. I-Iealth 9:93-122.

Pratt, DE, and Birac, PM. 1979. Source of antioxidative activity of soybeans and soy
products. J. Food Sci. 44:1720-1722.

201

Price, K.R., and Fenwick, GR. 1985. Naturally occurring oestrogens in foods - A review.
Food Add Cantam. 2:73-106.

Prizont, R., Konigsberg, N., and Aminoff, D. 1976. Glycosidase activity in the rat cecal
content. Gastroenterolagy A 7 02A928 (abstr.).

Ragnarrson, J .O., and Labuza, TR 1977. Accelerated shelf-life testing for oxidative
rancidity in foods - a review. Fd Chem. 2:291-308.

Ramanathan, L., and Das, NP. 1992. Studies on the control of lipid oxidation in ground fish
by some polyphenolic natural products. J. Agric. Food Chem. 40:17-21.

Rapta, P., Misik, V., Stasko, A, and Vrabel, I. 1995. Redox intermediates of flavonoids and
cafi‘eic acid esters fiom propolis: An EPR spectroscopy and cyclic voltammetry study.
Free Radic. Biol. Med 18:901-908.

Ratty, AK, and Das, NP. 1988. Effects of flavonoids on nonenzymatic lipid peroxidation:
Structure-activity relationship. Biochem. Med Metab. Biol. 39:69-79.

Record, I.R., Dreosti, IE, and McInemey, J.K. 1995. The antioxidant activity of genistein
in vitro. Nutr. Biochem. 6 2481-485.

Rekka, E., and Kourounakis, RN. 1991. Effect of hydroxyethyl rutosides and related
compounds on lipid peroxidation and free radical scavenging activity. Some
structural aspects. J. Pharm. Pharmacol. 43:489-491.

Renaud, S., and De Lorgeril, M. 1992. Wine, alcohol, platelets, and the French paradox for
coronary heart disease. The Lancet 339: 1523-1526.

Rice-Evans, CA, and Miller, NJ. 1996. Antioxidant activities of flavonoids as bioactive
components of food. Biochem. Soc. Trans. 24 2790:795.

Rice-Evans, CA, Miller, N.J., Bolwell, P.G., Bramley, PM. and Pridham, J.B. 1995. The
relative antioxidant activities of plant-derived polyphenolic flavonoids. Free Rad Res.
22:375-3 83.

Rice-Evans, C.A, Miller, NJ, and Pagaaga, G. 1996. Structure-antioxidant activity
relationships of flavonoids and phenolic acids. Free Radio. Biol. Med 20:933-956.

Richter, C. 1987. Biophysical consequences of lipid peroxidation in membranes. Chem.
Phys. Lipids. 44: 175-189.

Robak, J. and Gryglewski, RI. 1988. Flavonoids are scavengers of superoxide anion.

202

Biochem. Pharmacol. 3 7: 83 7-841 .

Roginsky, V.A., Barsukova, T.K., Remosova, AA, and Bors, W. 1996. Moderate
antioxidative efficiencies of flavonoids during peroxidation of methyl linoleate in
homogenous and micellar solutions. J. Am Oil Chem. Soc. 73:777-786.

Ruiz-Larrea, M.B., Leal, A.M., Liza, M., Lacort, M., and de Groot, H. 1994. Antioxidant
effects of estradiol and 2-hydroxyestradiol on iron-induced lipid peroxidation in rat
liver microsomes. Steroids 59:383-388.

Ruiz-Larrea, M.B., Leal, A.M., Martin, C., Martinez, R., and Lacort, M. 1995. Effects of
estrogens on the redox chemistry of iron: A possible mechanism of the antioxidant
action of estrogens. Steroids 60:780-783.

Ryan, TR, and Petry, T.W. 1993. The effects of 21-aminosteroids on the redox status of
iron in solution. Arch. Biochem. Biophys. 300:699-704.

Saija, A, Scalese, M., Lanza, M, Marzullo, D., Bonino, F., and Castelli, F. 1995.
Flavonoids as antioxidant agents: Importance of their interaction with biomembranes.
Free Radio. Biol. Med 4 :481-486.

Salah, N., Miller, N.J., Paganga, G., Tijburg, L., Bolwell, GP, and Rice-Evans, C. 1995.
Polyphenolic flavanols as scavengers of aqueous phase radicals and as chain-breaking
antioxidants. Arch. Biochem. Biophys. 322:339-346.

Salvayre, R., Negre, A, Affany, A, Lenoble, M., and Douste-Blazy, L. 1988. Protective
effects of plant flavonoids, analogs and vitamin E against lipid peroxidation. In Plant
F lavonoids in Biology and Medicine 1]. Biochemical, Cellular and Medicinal
Properties, p. 313-316, Alan R. Liss, New York, NY.

Sanz, M.J., Ferrandiz, M.L., Cejudo, M., Terencio, M.C., Gil, B., Bustos, G., Ubeda, A,
Gunasegaran, R., and Alcaraz, M.J. 1994. Influence of a series of natural flavonoids
on free radical generating systems and oxidative stress. Xenobiotioa 24:689-699.

Schachter, D. 1984. F luidity and function of hepatocyte plasma membrane. Hepatology
4 : 140-1 5 1 .

Schaich, KM. 1992. Metals and lipid oxidation. Contemporary issues. Lipids 27:209-218.
Schaich, KM, and Borg, DC. 1988 Fenton reactions in lipid phases, Lipids 23:570-579.

Scheline, RR. 1968. The metabolism of drugs and other organic compounds by the
intestinal microflora. Acta Pharmacol. Toxicol. 26:332-342.

203

Selway, J.W.T. 1986. Antiviral activities of flavones and flavans. In Plant Flavonoids in
Biology and Medicine, V. Cody, E. Middleton, Jr., and J .B. Harbome (eds), p. 521-
536, Alan Liss, New York.

Serbinova, E., Kagan, V., Han, D., and packer, L. 1991. Free radical recycling and
intramembrane mobility in the antioxidant properties of alpha-tocopherol and alpha-
tocotrienol. Free Radic. Biol. Med 10:263-275.

Setchell, K.D.R., and Adlercreutz, H. 1988. In Role of Gut Flora in T axioity and Cancer,
Ch. 14, Academic Press, New York, NY.

Setchell, K.D.R., Borriello, S.P., Hulme, P., Kirk, D.N., and Axelson, M. 1984.
Nonsteroidal estrogens of dietary origin: Possible roles in hormone-dependent disease.
Am. J. Clin. Nutr. 40:569-578.

Setchell, K.D.R., Gosselin, 8.1, Welsh, M.B., Johnston, J.O., Ballistreri, W.F., Kramer,
L.W., Dresser, BL, and Tarr, M.J. 1987. Dietary estrogens - A probable cause of
infertility and liver disease in captive cheetahs. Gastroenteralagy. 93:225-233.

Shahidi, F., and Wanasundara, P.K.J.P.D. 1992. Phenolic antioxidants. CRC Crit. Rev. Food
Sci. Nutr. 32:67-103.

Shutt, DA, and Cox, RI. 1972. Steroid and phytoestrogen binding to sheep uterine
receptors in vitro. J. Endocrinol. 52:299-310.

Smith, DA, and Banks, W. 1986. Biosynthesis, elicitation and biological activity of
isoflavonoid phytoalexyns. Phytoohemistry 25 :979-995.

So, F .V., Guthrie, N., Chambers, AF ., Moussa, M., and Carrole, K.K. 1996. Inhibition of
human breast cancer cell proliferation and delay of mammary tumorigenesis by
flavonoids and citrus juices. Nutr. Cancer 26:167-181.

Spear, N., and Aust, SD. 1994. Thiol-mediated NTA-Fe(III) reduction and lipid
peroxidation. Arch. Biochem. Biophys. 312:198-202.

Stanley, D.W. 1991. Biological membrane deterioration and associated quality loss in food
tissues. C.R.C.Crit. Rev. Food Sci. Nutr. 30:487-553.

Steenken, S., and Neta, P. 1982. One-electron redox potentials of phenols. Hydroxy- and
aminophenols and related compounds of biological interest. J. Phys. Chem. 8613661-
3667.

204

Strasburg, G.M., and Ludescher, RD. 1995. Theory and application of fluorescence
spectroscopy in food research. Trends Food Sci. Techno]. 6:69-75.

Stratton, S.P., Schaefer, W.H., and Liebler, DC. 1993. Isolation and identification of singlet
oxygen oxidation products of B-carotene. Chem. Res. T oxioo]. 6:542-547.

Svingen, B.A., Buege, J .A, O'Neil, PO. and Aust, SD. 1979 The mechanism of NADPH-
dependent lipid peroxidation. The propagation of lipid peroxidation, J. Biol. Chem.
254:5892-5899.

Tampo, Y., Onodero, S. and Yonaha, M. 1994 Mechanism of the biphasic effect of
ethylenediarninetetraacetate on lipid peroxidation in iron-supported and reconstituted
enzymatic system. Free Radic. Biol. Med 17:27-34.

Terao, J ., Piskula, M., and Yao, Q. 1994. Protective effect of epicatechin gallate and
quercetin on lipid peroxidation in phospholipid bilayers. Arch. Biochem. Biophys.
308:278-284.

Thomas, J.P., Maiorino, M., Ursini, F., and Girotti, AW. 1990. Protective action of
phospholipid hydroperoxide glutathione peroxidase against membrane-damaging lipid
peroxidation. J. Biol. Chem. 265 :454-461.

Thompson, M., and Williams, C .R. 1976. Stability of flavonoid complexes of copper (II)
and flavonoid antioxidant activity. Anal. Chim. Acta 85:375-381.

Thulbom, KR, and Sawyer, W.H. 1978. Properties and the location of a set of fluorescent
probes sensitive to the fluidity gradient of the lipid bilayer. Biaohim. Biophys. Acta
511:125-140.

Thumham, D1. 1994. Newer aspects of micronutrients in chronic disease: B—carotene, Are

we misreading the signals in risk groups? Some analogies with vitamin C. Proo.
Nutr. Soc. 53:557-569.

Tien, M.T., Svingen, BA and Aust, SD. 1982 An investigation into the role of hydroxyl
radical in xanthine oxidase-dependent lipid peroxidation. Arch. Biochem. Biophys.
216:142-151.

Torel, J ., Cillard, J ., and Cillard, P. 1986. Antioxidant activity of flavonoids and reactivity
with peroxy radical. Phytochemistry 25 2383-385.

Tricerri, M.A., Garda, HA, and Brenner, RR. 1994. Lipid chain order and dynamics at
different bilayer depths in liposomes of several phosphatidylcholines studied by
differential polarized phase fluorescence. Chem. Phys. Lipids l 7 :61-72.

205

Trotter, PI, and Storch, J. 1989. 3-[p-(6-phenyl)-l,3,5-hexatrienyl]phenylpropionic acid
(PA-DPH: Characterization as a fluorescent membrane probe and binding to fatty
acid binding proteins. Biaohim. Biophys. Acta 982: 131-139.

Urano, S., Inomori, Y., Sugawara, T., Kato, Y., Kitahara, M., Hasegawa, Y., Matsuo, M.,
and Mukai, K. 1992. Vitamin E: Inhibition of retinol-induced hemolysis and
membrane-stabilizing behavior. .1. Biol. Chem. 26 7:18365-18370.

Urano, S., Kitahara, M., Kato, Y., Hasegawa, Y., and Matsuo, M. 1990. Membrane
stabilizing effect of vitamin E: Existence of a hydrogen bond between a-tocopherol
and phospholipids in bilayer liposomes. J. Nutr. Sci. Vitamina]. 36:513-519.

Urano, S., Matsuo, M., Sakanaka, T., Uemura, 1., Koyma, M., Kumadaki, I., and Fukuzawa,
K. 1993. Mobility and molecular orientation of vitamin E in liposomal membranes
as determined by l"F NMR and fluorescence polarization techniques. Arch. Biochem.
Biophys. 303210-14.

Urano, S., Yano, K. and Matsuo, M. 1988. Membrane-stabilizing effect of vitamin B: Effect
of a-tocopherol and its model compounds on fluidity of lecithin liposomes. Biochem.
Biophys. Res. Commun. 150:469-475.

Ursini, F ., Mairino, M., Hochstein, P., and Emster, L. 1989 Microsomal lipid peroxidation:
mechanisms of initiation. Free Radic. Biol. Med 6:31-36.

van Acker, S.AB.E., de Groot, M.J., van den Berg, D.-J., Tromp, M.N.J.L., den Kelder,
G.D.-O., van der Vijgh, W.J.F., and Bast, A. 1996b. A quantum chemical
explanation of the antioxidant activity of flavonoids. Chem. Res. Toxicol. 921305-
1312.

van Acker, S.AB.E., Van den Berg, D.-J., Tromp, M.N.J.L., Griffioen, D.H., Van
Bennekom, W.P., Van der Vijgh, W.J.F., and Bast, A. 1996a. Structural aspects of
antioxidant activity of flavonoids. Free Radio. Biol. Med 20:331-342.

Van den Berg, J.J.M., Kuypers, F.A, Oju, J.H., Chiu, D., Lubin, B., Roelefsen, B. and Op
den Kamp, J .AF. 1988 The use of cis-parinaric acid to determine lipid peroxidation
in human erythrocyte membranes. Comparison of normal and sickle erythrocyte
membranes, Biochem. Biophys. Acta 944:29-39.

Van Thiel, D.H., Galvaoteles, A, Monteiro, E., Rosenblum, E., and Gavaler, J .S. 1991. The
phytoestrogens present in de-ethanolized bourbon are biologically active - a
preliminary study in a postmenopausal woman. Alcohol Clin. Exp. Res. 15:822-823.

Vigo-Pelfiey, C., and Nguyen, N. 1991. Modulation of lipid peroxidation of liposomes and

206

micelles by antioxidants and chelating agents. In Membrane Lipid Oxidation, Vol.
II, Vigo-Pelfrey, C. (ed), p. 135-149, Ch. 7, CRC Press, Boca Raton, FL.

Villa, P., Cova, D., Francesco, L.D., Guaitani, A, Palladini, G., and Perego, R. 1992.
Protective effect of diosmetin on in vitro cell membrane damage and oxidative stress
in cultured cell hepatocytes. Toxicology 73 2 179-189.

Wang, H., and Murphy, P.A. 1994a. Isoflavone composition of American and Japanese
soybeans in Iowa: Effect of variety, crop year and location. J. Agric. Food chem.
42:1674-1681.

Wang, H., and Murphy, P.A 1994b. Isoflavone content in commercial soybean foods. J.
Agric. Food Chem. 42:1666-1673.

Wardman, P. 1989. Reduction potentials of one-electron couples involving free radicals in
aqueous solution. .1. Phy. Chem. Ref Data Ser. 18:1637-1755.

Waters, WA. 1971 The kinetics and mechanism of metal-catalyzed autoxidation. J. Am. Oil
Chem. Soc. 48:427-433.

Wei, H., Bowen, R., Cai, Q., Barnes, 8., and Wang, Y. 1995. Antioxidant and
antipromotional effects of the soybean isoflavone genistein. P. S. E.B.M. 208: 124-130.

Wei, H., Wei, L., Frankel, K., Bowen, R., and Barnes, S. 1993. Inhibition of tumor
promoter-induced hydrogen peroxide formation in vitro and in vivo by genistein.
Nutr. Cancer. 20:1-12.

Willson, R.L. 1982. Iron and hydroxyl free radicals in enzyme inactivation and cancer. In
Free Radicals, Lipid Peroxidation, and Cancer, McBrien, D.C.H and Slater, T.F .,
p. 275-304, Academic Press, New York, NY.

Winterle, J., Dulin, D., and Nfill, T. 1984. Products and stoichiometry of reaction of vitamin
E with alkylperoxy radicals. J. Org. Chem. 49:491-495.

Wiseman, H. 1996. Dietary influences on membrane firnction: Importance in protection
against oxidative damage and disease. J. Nutr. Biochem. 7:2-15.

Wiseman, H. 1994. Tamoxifen and estrogens as membrane antioxidants: Comparison with
cholesterol. Meth. Enzymol. 234.2590-602.

Wiseman, H. 1993. Vitamin D is a membrane antioxidant. Ability to inhibit iron-dependent
lipid peroxidation in liposomes compared to cholesterol, ergosterol and tamoxifen and
relevance to anticancer action. FEBS Lett. 326:285-288.

207

Wiseman, H., Laughton, M.J., Amstein, H.R.V., Cannon, M., and Halliwell, B. 1990. The
antioxidant action of tamoxifen and its metabolites . Inhibition of lipid peroxidation.
FEBS Lett. 263:192-194.

Wiseman, H., Quinn, P., and Halliwell, B. 1993. Tamoxifen and related compounds
decrease membrane fluidity in liposomes. FEBS Lett. 330253-56.

Xu, X., Harris, K.S., Wang, H.-J., Murphy, RA, and Hendrich, S. 1995. Bioavailability of
soybean isoflavones depends upon gut microflora in women. J. Nutr. 125 :23 07-23 1 S.

Yamamoto, S., Shirnizu, K., Oonishi, 1., Hasebe, K., Takamura, H., Inone, T., Muraoka, K.,
Tani, T., Hashimoto, T., and Yagi, M. 1996. Genistein suppresses cellular injury
following hepatic ischemia/reperfiision . Transplant. Proo. 28:1111-1115.

Yarnashita, Y., Kawada, S.Z., and Nakaro, H. 1990. Induction of mammalian topoisomerase
II dependent DNA cleavage by nonintercalative flavonoids, genistein and orobol.
Biochem. Pharmacol. 39:737-744.

Yamauchi, R, Matsui, T., Kato, K., and Ueno, Y. 1990. Reaction products of a-tocopherol
with methyl linoleate-peroxyl products. Lipids 25:152-158.

Yamauchi, R, Matsui, T., Satake, Y., Kato, K., and Ueno, Y. 1989. Reaction products of
a-tocopherol with a free radical initiator, 2,2'-azobis (2,4-dimethylvaleronitrile).
Lipids 24:204-209.

Yoshida, Y., Furuta, S. and Niki, E. 1993 Effects of metal chelating agents on the oxidation
of lipids induced by copper and iron. Biochim. Biophys. Acta 1210:81-88.

Yu, B.P., Suescun, EA. and Yang, S.Y. 1992 Effect of age-related lipid peroxidation on
membrane fluidity and phospholipase A2: Modulation by dietary restriction. Meoh.
Age. Dev. 65:17-33.

Yuting, C., Rongliang, Z., Zhongjian, J ., and Yong, J. Flavonoids as superoxide scavengers
and antioxidants. 1990. Free Radio. Biol. Med 9:19-21.

Zhang, J.-R., Cazers, AR, Lutzke, BS, and Hall, ED. 1995. HPLC-Chemiluminescence
and therrnospray LC/MS study of hydroperoxides generated fiom
phosphatidylcholine. Free Radic. Biol. Med 18:1-10.

Zhou, Y.-C. and Zheng, R-L. Phenolic compounds and an analog as superoxide anion
scavengers and antioxidants. Biochem. Pharmacol. 42:1177-1179.

"7111111111111111111