Natural Food Colorants

Natural Food Colorants
JOIN US ON THE INTERNET VIA WWW, GOPHER, FTP OR EMAIL:

WWW: http://www.thomson.com
GOPHER: gopher.thomson.com
A service of lOOP
FTP: ftp.thomson.com
EMAIL: findit@kiosk.thomson.com
Natural Food Colorants
Second edition
Edited by
G.A.F. HENDRY
Honorary Senior Lecturer
NERC Unit of Comparative Plant Ecology
University of Sheffield
and
l.D. HOUGHTON
Lecturer
Department of Biochemistry and Molecular Biology
University of Leeds
SPRINGER-SCIENCE+BUSINESS MEDIA, B.V.

First edition 1992
Second edition 1996
Reprinted 1997
© 1996 Springer SciencetBusiness Media Dordrecht
Originally published by Chapman & Hall in 1996
Softcover reprint ofthe hardcover 2nd edition 1996
'JYpeset in 10/12 TImes by Photoprint, Torquay, Devon
ISBN 978-1-4613-5900-5 ISBN 978-1-4615-2155-6 (eBook)
DOI 10.1007/978-1-4615-2155-6
Apart from any fair dealing for the purposes of research or private study,
or criticism or review, as permitted under the UK Copyright Designs and
Patents Act, 1988, this publication may not be reproduced, stored, or
transmitted, in any form or by any means, without the prior permission in
writing of the publishers, or in the case of reprographic reproduction only
in accordance with the terms of the licences issued by the Copyright
Licensing Agency in the UK, or in accordance with the terms of licences
issued by the appropriate Reproduction Rights Organization outside the
UK. Enquiries conceming reproduction outside the terms stated here
should be sent to the publishers at the London address printed on this
page.
The publisher makes no representation, express or implied, with regard
to the accuracy of the information contained in this book and cannot
accept any legal responsibility or liability for any errors or omissions that
maybemade.
A catalogue record for this book is available from the British Library
Library of Congress Catalog Card Number: 95-78251
Printed on acid-tree text paper, manufactured in accordance with

ANSIINlSO Z39, 48-1992 (Performance of Paper)
Preface to the second edition
In this second edition of Natural Food Colorants two new chapters have
been added and we have taken the opportunity to revise all the other
chapters. Each of the original authors have brought up to date their
individual contributions, involving in several cases an expansion to the text
by the addition of new material. The new chapters are on the role of
biotechnology in food colorant production and on safety in natural
colorants, two areas which have undergone considerable change and
development in the past five years. We have also persuaded the publishers
to indulge in a display of colours by including illustrations of the majority
of pigments of importance to the food industry. Finally we have rearranged
the order of the chapters to reflect a more logical sequence. We hope this
new edition will be greeted as enthusiastically as the first.
It remains for us, as editors, to thank our contributors for undertaking
the revisions with such thoroughness and to thank Blackie A&P for their
support and considerable patience.
G.A.F.R.
J.D.R.
Contributors
Dr G .. Brittori Department of Biochemistry, University of

Liverpool, PO Box 147, Liverpool L69 3BX,
UK
Professor F.J. Francis Department of Food Science, College of Food
and Natural Resources, University of Massa-
chusetts, Amherst, MA 01003, USA
Dr G.A.F. Hendry NERC Unit of Comparative Plant Ecology,
Department of Animal and Plant Sciences,
University of Sheffield, Sheffield S10 2TN, UK
Mr B.S. Henry Phytone Limited, Third Avenue, Centrum
100, Burton-on-Trent, Staffordshire DE14
2WD, UK
Dr J.D. Houghton Department of Biochemistry and Molecular
Biology, University of Leeds, Leeds LS2 9JT,
UK
Professor R.L. Jackman Department of Food Science, Ontario Agricul-
tural College, University of Guelph, Guelph,
Ontario N1G 2W1, Canada
Dr M.C. O'Caliaghan Mary O'Callaghan & Associates, Castledowns,
Boycestown, Carrigaline, Co. Cork, Ireland
Professor J.L. Smith Ault Foods Ltd., 75A Bathurst Street, London,
Ontario N6B 1N8, Canada
Contents
1 Natural pigments in biology 1

G.A.F. HENDRY
1.1 Summary 1
1.2 Introduction 1
1.3 Definitions 1
1.4 Colour and colour description 2
1.5 Colour perception 4
1.6 Colour and chemical name 6
1.7 Electronic structure of pigments 7
1.8 Classification of biological pigments 9
1.9 Classification based on structural affinities 9
1.9.1 Tetrapyrroles 9
1.9.2 Tetraterpenoids 12
1.9.3 O-Heterocyclic compounds 12
1.9.4 Quinoids 15
1.9.5 N-Heterocyclic compounds (other than tetrapyrroles) 17
1.9.6 Metalloproteins 23
1.9.7 Miscellaneous 25
1.10 Colourless compounds in perspective 25
1.11 Pigment classification by natural distribution 26
1.11.1 Plants including algae 27
1.11.2 Vertebrates 28
1.11.3 Invertebrates 28
1.11.4 Fungi 29
1.11.5 Lichens 33
1.11.6 Bacteria 37
1.12 Biological systems as sources of pigments for commercial
exploitation 38
Acknowledgements 39
References 39
2 Natural food colours 40

B.S. HENRY
2.1 Summary 40
2.2 The role of colour in food 40
2.3 Classification of food colours 41
2.3.1 Synthetic colours 42
2.3.2 Nature-identical colours 42
2.3.3 Natural colours 42
2.4 Legislation 42
2.5 Factors affecting colour choice 44
2.6 Factors affecting colour application forms 45
2.6.1 Solubility 45
2.6.2 Physical form 45
2.6.3 pH 46
2.6.4 Microbiological quality 46
2.6.5 Other ingredients 46
viii CONTENTS
2.7 Performance and consumption of natural colours 46

2.8 Annatto 47
2.8.1 Source 47
2.8.2 Extracts and their application forms 49
2.8.3 Legislation 50
2.8.4 Factors affecting stability 50
2.8.5 Applications 51
2.9 Anthocyanins 53
2.9.1 Source 53
2.10 Beetroot 59
2.10.1 Source 59
2.11 Cochineal and carmine 64
2.11.1 Source 64
2.12 Curcumin 69
2.12.1 Source 69
2.12.3 Application forms 71
2.13 Other colours 74
2.13.1 Chlorophyll 74
2.13.2 Copper complexes of chlorophylls and chlorophyllins 75
2.13.3 Carotenoids 76
2.13.4 Caramel 77
2.13.5 Carbon black 78
2.14 Conclusions 78
References 79
3 Biotechnology in natural food colours:

The role of bioprocessing 80
M.C. O'CALLAGHAN
3.1 Summary 80
3.2 Introduction 80
3.3 Definitions 82
3.4 Plant biotechnology 83
3.4.1 Genetic engineering 83
3.4.2 Plant tissue culture 83
3.5 Fermentation 90
3.5.1 Fermentation development 91
3.5.2 Solid state (surface) fermentation 92
3.5.3 Algal fermentation 93
3.5.4 Submerged fermentation 97
3.6 Bioprocessing 98
3.6.1 Bioprocessing of j3-carotene 99
3.6.2 Bioprocessing of annatto 101
CONTENTS ix
3.6.3 Bioprocessing of astaxanthin 101

3.6.4 Bioprocessing of marigolds 102
3.6.5 Bioprocessing of betanin 103
3.7 Future perspectives 103
3.7.1 Biotechnology and carotenoid production 105
3.7.2 The influence of legislation 106
3.7.3 Commercial developments 107
Acknowledgements 108
References 108
4 Safety of food colorants 112

F.J. FRANCIS
4.1 Introduction 112
4.2 Toxicity testing 112
4.3 Acceptable daily intake (ADI) 114
4.4 Regulatory approaches 116
4.4.1 The European Union (EU) 116
4.4.2 The USA 117
4.4.3 The Delaney Clause 117
4.5 Safety of individual colorants 119
4.5.1 Beneficial aspects 119
4.5.2 Anthocyanins 120
4.5.3 Carotenoids 120
4.5.4 Canthaxanthin 121
4.5.5 Apocarotenol 121
4.5.6 Lycopene 122
4.5.7 Annatto 122
4.5.8 Paprika 122
4.5.9 Algae 122
4.5.10 Turmeric and curcumin 123
4.5.11 Cochineal and carmine 123
4.5.12 Lac 124
4.5.13 Caramel 124
4.5.14 Monascus 124
4.5.15 Iridoids 125
4.5.16 Betalains 125
4.5.17 Cocoa 125
4.5.18 Chlorophyll 125
4.5.19 Titanium dioxide 125
4.5.20 Carbon black 126
References 126
5 Chlorophylls and chlorophyll derivatives 131

G.A.F. HENDRY
5.1 Summary 131
5.2.1 Nomenclature 133
5.3 Chlorophylls under natural conditions 135
5.3.1 Function 135
5.3.2 Structure of chlorophylls and chlorophyll-protein
complexes 136
5.3.3 Biosynthesis of chlorophyll a and other natural porphyrin
pigments 138
5.3.4 Types of naturally-occurring chlorophylls 140
5.3.5 Turnover and degradation of chlorophylls in biology 141
5.3.6 Natural, and some unnatural, chlorophyll derivatives 142
X CONTENTS
5.4 Chlorophyll derivatives as food colorants 146

5.4.1 Sources of chlorophylls for food colorants 146
5.4.2 Alternatives to chlorophyll a 147
5.4.3 Extraction, isolation and derivatization 149
5.4.4 Structure of the derivatives 150
5.4.5 Stability and instability 151
5.4.6 Economics of chlorophyll derivatives 152
5.5 Future prospects 153
References 155
6 Haems and bilins 157

J.D. HOUGHTON
6.1 Summary 157
6.2 Haems 157
6.2.1 Introduction 157
6.2.2 Haems in nature 158
6.2.3 Free haems 165
6.3 Phycobilins 170
6.3.1 Introduction 170
6.3.2 Phycobilins in nature 173
6.3.3 Free phycobilins 182
References 193
7 Carotenoids 197
G. BRI1TON
7.1 Summary 197
7.2.1 General publications 197
7.2.2 Structure and nomenclature 198
7.3 Distribution, natural functions and actions 200
7.3.1 Distribution 200
7.3.2 Natural functions 202
7.3.3 Actions of carotenoids in human health 202
7.4 Biosynthesis 203
7.4.1 Regulation of biosynthesis 208
7.5 Absorption, transport and metabolism 209
7.6 Natural and synthetic carotenoids as colorants 211
7.6.1 Natural carotenoids and extracts 211
7.6.2 Synthetic carotenoids 212
7.7 General properties and stability 213
7.8 General procedures for carotenoid work 213
7.9 Extraction and purification 215
7.10 High-performance liquid chromatography (HPLC) 217
7.10.1 Normal-phase (adsorption) HPLC 217
7.10.2 Reversed-phase HPLC 218
7.10.3 HPLC of apocarotenoids 219
7.10.4 Resolution of optical isomers 219
7.11 UV-visible light absorption spectroscopy 219
7.11.1 Position of the absorption maxima 220
7.11.2 Spectral fine structure 221
7.11.3 Geometrical isomers 221
7.12 Quantitative determination 221
7.12.1 Spectrophotometry 221
7.12.2 Quantitative determination by HPLC 224
CONTENTS XI
7.13 Other physicochemical methods 225

7.13.1 Mass spectrometry (MS) 225
7.13.2 Nuclear magnetic resonance (NMR) spectroscopy 226
7.13.3 Circular dichroism (CD) and optical rotatory dispersion
(ORD) 226
7.13.4 Infra-red and Raman spectroscopy 227
7.14 Commercial uses and applications 227
7.14.1 Application forms and formulations 227
7.14.2 Uses 228
7.15.1 Other carotenoids 230
7.15.2 Carotenoids as health-promoting substances 230
7.15.3 Carotenoproteins 230
7.15.4 Biotechnology 231
Appendix: Structures of the individual carotenoids mentioned in the text 233
References 240
8 Anthocyanins and betalains 244

R.L. JACKMAN and I.L. SMITH
8.1 Summary 244
8.3 Functions 246
8.4 Anthocyanins 246
8.4.1 Structure 246
8.4.3 Biosynthesis 251
8.4.4 Factors influencing anthocyanin colour and stability 257
8.4.5 Extraction and purification 270
8.4.6 Current and potential sources and uses 278
8.5 Betalains 280
8.5.1 Structure 280
8.5.3 Biosynthesis 284
8.5.4 Colour and structural stability 286
8.5.5 Extraction and analysis 290
8.5.6 Current and potential sources and uses 294
8.6 Conclusions 296
References 296
9 Less common natural colorants 310

F.J. FRANCIS
9.1 Acylated ~-ring substituted anthocyanins 310
9.2 Annatto 315
9.3 Saffron 317
9.4 Gardenia pigments 318
9.5 Cochineal and related pigments 321
9.6 Turmeric 325
9.7 Carthamin 327
9.8 Monascus 328
9.9 Miscellaneous colorants 332
References 335
Index 343
1 Natural pigments in biology
G.A.F. HENDRY
1.1 Summary
This introductory chapter provides a working definition of 'natural'

pigments. The greater part of the chapter consists of a survey of pigmented
compounds found in biology. Two systems of classification are adopted,
one based on structural affinities, the second based on the natural
occurrence of the pigment in biology. The latter system provides a resume
of almost all but the rarest pigments in vertebrates and invertebrates,
plants including algae, fungi, lichens and bacteria. Where appropriate, the
classifications are cross-referenced to the fuller treatment of certain
pigments in the succeeding chapters. An outline is given of the way colours
or hues are defined throughout the book, and of the areas of confusion
surrounding colour description. The variation in colour perception by
human eyes is also described. Chemical nomenclature and certain physical
attributes common to pigmented compounds are provided, including a
brief synopsis of the structural features needed to make an otherwise
colourless molecule pigmented.
1.2 Introduction
The immediate purpose of this chapter is to introduce the major and at

least the most widespread of the minor pigments occurring naturally in
plants, fungi, lichens, animals and bacteria. Whether or not these have
previously been exploited as natural food colorants is not of immediate
consideration. Neither is the availability of the pigment for commercial
exploitation, particularly in view of the recent and promised advances in
gene transfer and biotechnology. Theoretically, almost all reasonably
stable biological pigments can now be considered for biotechnological
production and, in future, may become available to the food and other
industries as natural colorants (see chapter 3).
1.3 Definitions
By definition, a natural pigment in biological systems is one that is

synthesized and accumulated in, or excreted from, living cells. In addition,
G. A. F. Hendry et al. (eds.), Natural Food Colorants

© Springer Science+Business Media Dordrecht 1996
2 NATURAL FOOD COLORANTS
certain pigments, such as oxidized phenols and the more simple phenolic
derivatives such as the anti-coagulant coumarins, may be formed by the
dying cell. What is not a natural pigment in biology is less easy to define.
For the purposes of this chapter, inorganic pigments based on, for
example, iron or titanium are not considered natural if only because they
do not appear to playa significant or even minor role in biology; nor are
the several organic-iron complexes formed synthetically from naturally
occurring organic complexes. That 'work-horse' of the food colorist's
palette, caramel, is also not considered in depth in this chapter since it is
not a pigment of living cells, even if its constituent parts (sugars and amino
acids and amides) are.
This working definition and the delimitation of natural pigments inevit-
ably highlight certain 'grey' areas, and so a little pragmatism has been
employed to keep within the spirit of the above definitions while remaining
relevant to the particular interests of the food colorant user. For example,
green chlorophylls are truly natural compounds: the grey-green phaeophy-
tin pigments extracted from green plants are natural in that they may be
formed from chlorophyll during the natural senescence of the dying leaf
and, incidentally, during industrial extraction of chlorophylls from leaves.
The bright green Cu-chlorophyllins (see chapter 5) are not found in nature
but represent man's close attempt to restore the original colour of true
chlorophyll to these phaeophytins. All of these green pigments are
considered here as 'natural'. Indigo, a well known plant dye, is present in
living tissues but as the colourless glycoside. Only on extraction, hydrolysis
and oxidation does this pigment acquire its blue colour. For this reason, an
expanded definition of a natural colorant as a 'pigment formed in living or
dead cells of plants, animals, fungi or micro-organisms, including organic
compounds isola~ed from cells and structurally modified to alter stability,
solubility or colour intensity' is offered. The definition must also allow for
pigments whose synthesis is engineered in genetically transformed organ-
isms. For pragmatic reasons, structural colours arising from diffraction,
interference or Tyndall scattering of light are not discussed in depth;
nevertheless, many playa significant role in the coloration of animals.
1.4 Colour and colour description
Throughout this book and most of the literature colours are described by
the wavelength (>..) of the maximum absorption (Amax) in the visible part of
the electromagnetic spectrum, expressed in nanometres (nm). White light
is seen as the simultaneous incidence of the full range of the visible
spectrum (wavelengths between 380 and 730 nm approximately) at the
same relative intensities. Any object that lessens the intensity of one part of
the spectrum of white light by absorption (as in a monochromatic filter)
NATURAL PIGMENTS IN BIOLOGY 3
will bring about the perception of colour in the residual transmitted light,
as well as a reduction in irradiance (measured in Watts per m2 , Einsteins or
moles of photons) or more simply a darkening. As more of the spectrum is
filtered out the perception of colour will be increasingly monochromatic
(shades of grey) darkening ultimately to black.
To those familiar with quantifying and qualifying light, a compound with
an absorption maximum of, for example, 600 nm in the yellow part of the
spectrum, will be thought of as deep blue in perceived colour. One
absorbing light at say 470 nm, the blue-green part of the spectrum, will be
predictably a yellow, leaving the unfamiliar reader thoroughly confused.
The reason for the confusion is readily explained: light, or more specific-
ally visible radiation, when perceived simultaneously over the complete
range of the spectrum as in unfiltered sunlight, is seen as white light by the
human eye. When separated by passing the light through a prism, the
refracted beams of light are displayed as the sequence of colours familiar in
a rainbow. The human eye 'senses' six hues:
1. Red at around 700 nm
2. Orange at 625 nm
3. Yellow at about 600 nm
4. Green at 525 nm
5. Blue at around 450 nm
6. Violet at and below 400 nm
Hence the casual laboratory description of a compound as having a 'strong
red absorbance' instead of the more precise 'absorbance maximum at
700 nm.' Such a compound would probably appear to most human eyes to
have a deep blue or green-blue colour! The confusion arises because the
pigment in solution with say a strong red absorbance when placed in front
of a beam of light, would allow only the passage, that is transmission, of the
remaining non-red part of the spectrum-the broad area yellow-green-
blue-violet. In white light, such a compound would appear to be blue (or
green-blue). For a workable translation of description, based on the
perceived colour into absorption maxima, see Table 1.1. The first column
lists the colour perceived by the normal (see below) human eye when
viewing a solid or liquid pigment under white light. The second column
shows that part of the spectrum absorbed by that pigment and the third
column the approximate wavelength of the light absorbed. Thus a pigment
such as in a mauve-coloured grape skin, under white light, will absorb the
yellow-green part of the spectrum around say 520-530 nm. Yellow-green
light having been absorbed by the pigments in the grape skin, the
remaining, unabsorbed, part of the light spectrum will be seen by the eye
as a mixture of colours-purple, blue, orange and red-that is the hue
mauve.
Table 1.1 Human perception of colours according to the wavelength of light absorbed by a
pigment
Colour perceived Colour absorbed Wavelength of maximum

(from white light) absorption (nm)
Blue-green Red 660--700

Blue Orange 6()()....Q30
Violet Yellow 585-600
Mauve-red Yellow-green 560--580
Orange Green 520--540
Yellow Green-blue 480--500
Yellow-green Blue 440--460
Green-yellow Violet >430
In the succeeding chapters data will be presented with the absorption

maxima of several hundred different natural pigments. By reference to
Table 1.1 the portion of the spectrum of white light absorbed by the
pigment can be predicted as well as the colour of the compound as seen by
the human eye. Table 1.1 can also be used in the reverse order to predict
the part of the spectrum that any coloured compound might absorb when
viewed under white light. Thus from the previous example, grape juice
anthocyanins that appear to be mauve under white light are likely to
absorb light at around 510 to 540 nm, that is the green part of the
spectrum. There is then little point in observing the colour of a treasured
glass of wine under the soft green lights sometimes favoured by restaur-
ateurs for the wine will appear as black as ink!
1.5 Colour perception
Numerous definitions of colour are available, the most useful one in this
context being that part of the electromagnetic spectrum visible to the
human eye and generally regarded as lying between 380 and 730 nm. The
key element in the special context of food colorants, however, is the hue
perceived by the eye of the normal, healthy and average adult. Such an eye
is less well-defined.
One problem with defining pigments by their apparent or perceived hue
is that one man's mauve is another man's purple or red (and here male
gender is meant). Precisely where red moves to the description mauve is
less a matter of definition but more a subjective decision dependent on
variations in colour reception and processing that exist in the eyes of
different humans-particularly males. While the human eye can detect up
to six major colours and many more intermediate blends, just three
predominate-red, yellow and blue. These 'primary' colours (nearly)
coincide with the three types of eye pigments.
The retina at the back of the human eye is made up of some 3 million
colour-perceiving cone cells and about 100 million rod cells involved in
providing monochromatic vision under low light. Each cell contains one of
several types of visual pigment, which are structurally based on the ~
carotene and vitamin A derivative 11,cis-retinol and bound to a particular
class of proteins, the opsins. 11 ,cis-retinol itself, without modification,
absorbs light maximally at around 500 nm, coinciding conveniently for
Earth dwellers with the spectral maximum of the Sun. The cone-cell
pigment proteins are sensitive to different ranges of predominantly blue,
green and red light. Colour in this sense is the sum of excitation of red-
sensitive, green-sensitive and blue-sensitive cones. The difference between
the 'red' and 'green' protein lies in just 15 out of 348 amino-acid residues,
and there is evidence from studies of higher mammals that the evolution of
the IS-residue difference may have occurred relatively recently in humans.
Perhaps because of this recent evolution, considerable variation appears to
persist within the human genetic make-up in these 15 or so residues. Small
variations between two individuals in the DNA coding for the protein can
lead to quite distinct differences in distinguishing the colour perceived by
the different viewers. Disagreements over the matching or separation of
colours lie particularly at the junctions of what are described as blue/green,
red/orange and mauve/violet hues. These differences are not classed in
normal parlance as colour blindness since what is green, blue or blue/green
is a matter of unspoken consensus. The degree of variation in colour
perception and its prevalence is unknown but is believed to be consider-
able.
Clinically defined 'colour blindness' is, in most cases, the result of an X
chromosome-linked recessive mutation occurring in approximately 1 in 12
males. The incidence in females is about 1 in 170. As a brief description of
what is a complex and fascinating subject in its own right, the following is
offered as an introductory guide. An individual classes as 'colour blind' and
lacking say the 575 nm red-light cones will 'see' only green and blue, or if
lacking the 540 nm green-light cones will 'see' green as red, the latter
condition being the more common. Other individuals carry cone cells with
a maximum absorbance lying between green and red, resembling the cone
cells of the less recently evolved apes. In most cases of colour-blindness,
the individual is well able to distinguish the colour between blue and yellow
but distinctions between green and red are possible often only as differ-
ences in brightness. In the most common forms of colour blindness, a red
and green mixture of light is seen as either too 'red' or too 'green' by the
standards prevailing among the non colour-blind majority. Given the
immense effort to achieve food colorants that are perceived by the food
chemist as a reproducibly distinct green or red hue, it is worth noting that
the endeavour will never be appreciated by one in twelve male consumers!
1.6 Colour and chemical name
One way of pinpointing coloured compounds from a list of organic

chemicals is through the name of the substance. There has been a long
tradition particularly among German chemists of the last century of giving
coloured compounds a name derived from a Greek or Latin pigment.
These names have persisted in the trivial (and familiar) names of many
chemicals. A short list is given in Table 1.2. More comprehensive
treatments of the subject are provided by Stearn [1] and Flood [2].
Caution is needed, however, in using these names as indicators of
particular colours. For example, porphyra of the Greek aristocratic toga
gives rise in translation to purple; such a toga seen today would be
described by most observers as crimson. Similarly, not all melanins are
black as implied by the Greek melas. Given these limitations, however, the
terminology is a fast way to recognize a coloured chemical.
Table 1.2 Chemical names implying colour"
Chemical prefix or suf- Classical derivation Implied colour Chemical example

fix or root
anil- ai-nil (Arab) Deep blue Aniline

arg- argentum (L) Silver Arginine
aur- aurum (L) Gold Aureomycin
-azur azul (OSp) Sky blue Aplysioazurin
chloro- chloros (Gk) Yellow-green Chlorophyll
chrom- or -chrome chroma (Gk) Coloured Chromotropic acid
chryso- chrysos (Gk) Gold Chrysin
citr- citrus (L) Lemon-coloured Citruline
-cyanin kyanos (Gk) Blue Anthocyanin
erythro- erythros (Gk) Red Phycoerythrin
-flavin f1avus (L) Pale yellow Riboflavin
fulv- fulvus (L) Red-yellow Fulvine
fusc- fuscus (L) Brown Fuscin
haem- or heme- haima (Gk) Blood red Haematin
indigo- indikon (Gk) Deep blue Indigo
leuco- leukos (L) Colourless/white Leucoptrein
lute- luteus (Gk) Yellow Lutein
mela- melas (Gk) Black Melanin
phaeo- phaios (Gk) Dark-coloured Phaeophytin
porph(o)- porphyros (Gk) Purple Porphyrin
purpur- purpura (L) Purple Purpurin
-pyrrole pyrros (Gk) Fiery-red Pyrrole
rhodo- rhodon (Gk) Rose-coloured Rhodophyllin
rub- ruber (L) Red Bilirubin
ruf- rufus (L) Red-brown Anthrarufin
sepia- sepia (L) Brown-pink Sepiapterin
stilb- stilbein (Gk) Glitter Stilbene
verd- or virid- viridis (L) Bright green Biliverdin
viol- viola (L) Purple-blue Aplysioviolin
xanth- xanthos (Gk) Yellow Xanthophyll
"Arab = Arabic; Gk = Ancient Greek; L = Latin; OSp = Old Spanish.

1. 7 Electronic structure of pigments
Whether or not a biological molecule is coloured is determined by its

structure, particular electronic, the size of the molecule, solubility and
elemental composition. Fortunately most natural pigments have several
common features that immediately distinguish them from the larger
number of colourless compounds found in biological material. All biologi-
cal material is composed of a limited range of elements within the periodic
table. Within this narrow selection, pigmented organic compounds gener-
ally possess three features:
1. Almost all biological molecules are composed of no more than 17
elements within the periodic table. Of these, just four elements, H, C,
Nand 0 predominate.
2. Most pigmented compounds, particularly those other than pale yellow,
contain either N or 0, often both.
3. Most are relatively large molecules. Among the more common pig-
mented compounds, molecular weights range from about 200 (anthra-
quinones), 300 (anthocyanidins), 400 (betalaines), 500 (carotenoids) to
800 (chlorophylls).
Much greater molecular weights are of course found in pigment polymers
such as melanins.
With this limited range of attributes, it is possible to restrict a discussion
on the electronic structure of biological pigments to bare essentials (see
Brown [3] for a more full explanation). The electrons in the outer shells
over the surface of biological molecules, coloured or not, are concentrated
in distinct orbits. Each shell holds a finite number of electrons arranged to
occupy the least energy-demanding orbits possible. Such electrons are
described as being in a resting state. When irradiated with light of sufficient
energy, the electrons in the outermost orbitals oscillate in resonance with
light, and are raised from the resting state to a higher excited state. After
the excitation, as the electrons decay (relax) to the original resting state so
the energy is released in a less energetic form, most commonly as heat. As
biological molecules are composed of generally no more than four, perhaps
five out of seventeen elements, almost all of the electronic orbitals
participating in colour are from just three orbitals designated d, p and n.
On excitation, one or more electrons within these orbitals may be raised
respectively to d* and p* orbitals.
Transitions from d to d*, which are characteristic of saturated hydro-
carbons (with no double bonds), are the most energy demanding, requiring
more energy than that provided by the visible part of the spectrum and so
appear colourless. Transitions from p to p* as in unsaturated hydro-
carbons (which contain double bonds) are somewhat less demanding
energetically but are usually still beyond the energy levels provided by the
Table 1.3 Range of electro-magnetic spectrum absorbed by molecules and causing

excitation of electrons from lower (resting state) to higher orbital state
Light causing Electronic Molecular form Expected colour

excitation transition
Visible n~ p* Unsaturated hydrocarbon Red, blue or green

with N or 0
Violet to blue n~ d* Saturated hydrocarbon Pale yellow
with N or 0
Near UV p~ p* Unsaturated hydrocarbon Probably colourless
FarUV d~ d* Saturated hydrocarbon Colourless
visible part of the spectrum, beyond the ultra-violet. The larger molecules
at least may appear as pale yellow, a characteristic of many vegetable oils.
The third transition n to d*, characteristic of saturated hydrocarbons
particularly with nitrogen (N) or oxygen (0) substitutions may also just
appear in the visible part of the spectrum as pale yellow compounds, as in
many fungal pigments. However, the least energy demanding are the n to
p* transitions found in compounds that are essentially unsaturated hydro-
carbons with N or 0 substitutions (Table 1.3) and it is this group that
provides the greatest range of pigmented organic molecules.
The simple picture presented for alkenes needs to be improved,
however, to understand why biological molecules appear to be coloured
while most do not. Several important structural modifications cause the
electrons of the modified molecule to absorb and to become excited by
light at wavelengths considerably longer (and less energetic) than would be
predicted from the unmodified alkene. This shift to less energetic wave-
lengths is called a bathochromic shift.
Six types of structural modification are associated with a bathochromic
shift as follows:
1. An increase in chain length
2. Branching of the carbon chain
3. Arrangement in a single ring structure (cyclization)
4. Addition of Nor 0, particularly with 3
5. Bonding of two or more rings
6. Addition of certain transition metals
Ring structures, particularly in relatively large molecules, permit sharing
or delocalization of particular electrons orbiting over the face of the
molecule. Such de localized electrons are more readily raised to an excited
state on irradiance with visible light and contribute to most of the non-
yellow pigments in biology. The outstanding biological example of one
class of compounds with all of these modifications (that is a large number
of conjugated double bonds, branched structures, addition of Nand 0,
linked rings and addition of a transition metal) is the iron porphyrins such
as the red haems.
1.8 Classification of biological pigments
It will be apparent that pigmented compounds in biology will tend to be

large and complex molecules. Given the long time-scale over which these
complex structures have evolved, it is no surprise to find that they are also
highly varied. Among just one group, the anthocyanins, over two hundred
and fifty structurally and spectrally distinct pigments have been recorded so
far [4]. The classification of anything approaching the majority of pig-
mented compounds in biology into a single workable synopsis will
inevitably result in grouping together compounds that have little or no
biosynthetic, functional or structural or even colour relationship. Several
bold attempts have been made by other workers and the account that
follows is in no way the only means of classifying pigments. Good
alternatives can be found in the works of Britton [5], Fox [6] and Fox and
Vevers [7], and also, in more detail, in Needham [8].
Two methods of classification are adopted here, which are linked
through cross references and through the index. The first is based on
structural affinities. This has the advantage of brevity but the disadvantage
that it groups together compounds whose biosynthetic origins are quite
unrelated. The second is based on the natural occurrence of pigments in
biology. For convenience, the groupings used here cover the pigments of
plants including algae, fungi, lichens, higher animals (vertebrates) and
lower animals (invertebrates) and bacteria. Throughout, reference will be
made to chapters in this book that specifically cover either the pigment
itself or at least a group of structurally related pigments.
1.9 Classification based on structural affinities
Almost all biological pigments can be reduced to no more than six

major structural classes: the tetrapyrroles, tetra-terpenoids, quinones, 0-
heterocyclic, N-heterocyclic and metallo-proteins. Given the wealth of
variation thrown up by 3000 million years of evolution there will inevitably
also be a group of miscellaneous pigments that defy simple classification. In
Table 1.4 this latter category has been limited by excluding pigments that
are rare or restricted in occurrence.
1.9.1 Tetrapyrroles
This is a relatively small group of pigments that contribute the greatest
range of colours as well as being the most abundant globally. All are based
Table 1.4 Naturally occurring pigments in biology
Group Alternative or Major example Predominant See

familiar name colour chapter
Tetrapyrroles Porphyrins and Chlorophylls Green 5

porphyrin Haems (hemes) Red 6
derivatives Bilins (Bile Blue-green- 6
pigments) yellow-red
Tetraterpenoids Carotenoids Carotenes Yellow-red 7
Xanthophylls Yellow 7
O-Heterocyclic Flavonoids Anthocyanins Blue-red 8,9
compounds Flavonols Yellow-white
Flavones White-cream
Anthochlors Yellow
Quinones Phenolic Naphthaquinones Red-blue-green 2,9
compounds Anthraquinones Red-purple 2,9
Allomelanins Yellow-brown
Tannins Brown to red
N- Heterocyclic Indigoids and Betalaines Yellow-red 8,9
compounds indole Eumelanins Black-brown
derivatives Phaeomelanins Brown
Indigo Blue-pink
Substituted Pterins White-yellow 2
pyrimidines Purines Opaque white
Flavins Yellow
Phenoxazines Yellow-red
Phenazines Yellow-purple
Metalloproteins Cu-proteins Blue-green
Haemerythrin Red
Haemovanadins Green
Adenochrome Purple-red
Miscellaneous Lipofuscins Brown-grey 9
Fungal pigments Various but
commonly yellow
on the same structure, the tetrapyrrole either in its linear form (Figure 1.1)
and known as bilins or bile pigments or in the cyclic form (Figure 1.2)
where it constitutes the chromophore of haem proteins (chelated with iron)
or chlorophyll proteins (chelated with magnesium). In total, the number of
naturally occurring cyclic tetrapyrroles (haems, chlorophylls and their
precursors) that have been recorded is about 28, with a further 6 or so
linear tetrapyrroles (see Hendry and Jones [9] for a full list).
The proteins to which all cyclic and most of the linear pigments are
linked alter the solubility properties of the tetrapyrrole and under natural
1)
Figure 1.1 A linear tetrapyrrole.

Figure 1.2 A cyclic tetrapyrrole, porphyrin.
conditions probably determine the stability of the chromophore. In living,

healthy cells, free haems and chlorophylls that have been detached from
their proteins, probably do not exist or at least do not accumulate to
detectable concentrations. The linear bilins are also bound either to a
protein (as in phycobilins of certain algae) or to sugars or as amorphous
polymers when formed as breakdown products of, typically, haemoglobin.
The colour of the particular tetrapyrrole is determined largely by the
structure and substitutions to the tetrapyrrole molecule itself and relatively
little by the metal. The colour range of this group of pigments is
considerable and is summed up in Table 1.5.
Evolutionary evidence suggests that the tetrapyrroles must have been
among one of the earliest of biological pigments functioning in photosyn-
thesis (the various chlorophylls and phycobilins), electron transfer (the
haem-based cytochromes), oxygen transport (haem-based haemoglobin
and myoglobin) and in protection from reactive forms of oxygen (perox-
idases). As all of these functions are of primary importance in cellular
metabolism, tetrapyrroles are found in all biological organisms-probably
without exception.
In living cells the tetrapyrroles are relatively stable with half-lives of
many weeks. On isolation and particularly where ingested by animals, the
tetrapyrroles are rapidly degraded. Chlorophylls ingested with the rest of
the photosynthetic apparatus retain the potential to generate electrons on
exposure to light. In transparent animals, this could lead to the formation
Table 1.5 Colour range of naturally occurring tetrapyrroles
Colour Tetrapyrrole Occurrence
Blue Phycocyanin Blue-green algae

Blue-green Chlorophyll a Plants
Green Chlorophyll b Plants
Yellow-green Phaeophytin Senescing plants
Yellow-orange Phycoerythrin Red algae
Orange Bilirubin Vertebrates
Red Haem All organisms
Figure 1.3 A tetraterpenoid, lycopene.
of oxygen radicals and at best indigestion! There are examples of animals

that retain ingested chlorophyll for an appreciable length of time (certain
hydra, sea anemones, corals and sea slugs) but in most cases the
chlorophyll is rapidly and perhaps selectively degraded.
1.9.2 Tetraterpenoids
More familiar as carotenoids, this group of pigments is also present in
abundance, globally, both in plants and in animals. In total over 600
carotenoids have been described making this one of the larger groups of
natural pigments.
Most carotenoids are yellow to yellow-orange. Distinctly red and
indistinctly yellow-green examples can also be found. The structures of all
carotenoids are based on a common 40-carbon (40C) linear terpene
(Figure 1.3). In plants, carotenoids are principally associated with chi oro-
phylls in the photosynthetic organelle, the chloroplast or its degenerate
successor, the chromoplast. At this point, carotenoids enter the animal
food chains by ingestion with chlorophylls. In many animals, the caroten-
oids are selectively assimilated for a wide range of purposes (see chapter
7), while in almost all animals the chlorophylls are excreted in a partially if
not wholly degraded form. This perhaps emphasizes the potentially
positive value of the carotenoids to animals in contrast to the possible
photo-toxicity of chlorophylls should they be absorbed. Carotenoid-
protein complexes are common in animals but in photosynthetic organisms
(plants and bacteria) carotenoids are more usually present dissolved in
lipid membranes. The more widespread carotenoids in nature are summar-
ized in Table 1.6.
1.9.3 a-Heterocyclic compounds

An enormous and somewhat diverse group, these compounds are more
familiar as flavonoids and related structures. They are pre-eminently
compounds of terrestrial (flowering) plants and appear to have evolved
along with flower petals and pollinator attractants. A total of some 4100
flavonoids have been described so far in structural detail (see Harborne
[4]), and almost one quarter within the past 10 years. Of the several classes
of flavonoids, most are at best pale yellow in appearance, the majority
being visible only under ultraviolet (UV) light. To that extent, most
Table 1.6 The more widespread of the naturally

occurring carotenoids in biology (unless otherwise
indicated all are to be found in land plants)
a-Carotene
J3-Carotene (particularly widespread)
Lutein
Violaxanthin
Neoxanthin
J3-Cryptoxanthin
Antheraxanthin
Fucoxanthin (brown algae)
Lycopene (and several related derivatives)
Astaxanthin (fish, crustacea)
Canthaxanthin (beetles)
flavonoids are not pigments in the narrow sense of this book. The most
truly colourful of the flavonoids are the anthocyanins, the pigments
contributing to much of the colour of petals and ripening fruits of higher
plants and to senescing leaves of trees in autumn.
Structurally, the flavonoids are based on a common tri-cyclic phenyl
benzopyran (Figure 1.4) consisting oftwo benzene rings (designated A and
B) joined with an oxygen-containing pyran group. Further modifications to
the heterocyclic ring determine the class (and indeed colour) of the various
flavonoids. In terms of visible-light pigments, much the most important
group are the anthocyanidins (Figure 1.5). In biological systems, antho-
cyanidins are probably always bound to one or more sugars, most usually
I3-D-glucose, I3-D-galactose and a-D-rhamnose with one, often more,
hydroxyl groups on the benzene rings, these latter substitutions dictating
the colour of the anthocyanidin. Although some 200 anthocyanins (the
flavonoid plus sugar) have been documented (Harborne [4]), only 17
anthocyanidins (the aglycone structure) are involved. Omitting the uncom-
mon or rare, Table 1.7 lists the six most widespread anthocyanidins with
their structural attributes. The trivial names reflect, in most cases, the plant
from which the pigment was first isolated and disguise the very widespread
presence of the six throughout the flowering plants.
In addition, interactions of the ring B with certain metals, usually
aluminium or iron, and interactions elsewhere in the molecule with
Figure 1.4 Basic structure of a flavonoid.

R'
HO
OH
Figure 1.5 Basic structure of an anthocyanidin (where RI and R2 are typically H, OH,
OCH3)'
Table 1.7 The most widely occurring anthocyanidins
Name Colour Amax in acidic

methanol
Cyanidin Blue-red 535

Delphinidin Purple-blue 546
Malvidin Purple 542
PeIargonidin Scarlet-red 520
Peonidin Blue-red 532
Petunidin Purple-blue 543
magnesium confer a blueing to the pigment. Interactions between antho-

cyanins and between ftavonoids extend the range of colours. Flavonoid
polymers are common. The reddish-brown pigment theaftavin (Figure 1.6)
in the beverage tea is one example.
OH
HO
HO
OH
Figure 1.6 A theaflavin.
1.9.4 Quinoids
These phenolic compounds vary from the simple monomeric dihydroxy-
phenol derivatives such as l,4-benzoquinone (Figure 1.7), to the reduced
dihydroquinone, to the dimer such as l,4-naphthaquinone (Figure 1.8) to
the trimer 9,1O-anthraquinone (Figure 1.9) on to the more complex
polymeric forms as in hypericin (Figure 1.10).
o
o
o
Figure 1.7 l,4-Benzoquinone.
Oc) o
Figure 1.8 l,4-Naphthaquinone.
Figure 1.9 9,1O-Anthraquinone.
HO
HO
Figure 1.10 Hypericin.

The simpler quinones may function as electron and proton carriers in

primary metabolism and are to be found, if only in trace amounts, in all
actively respiring organisms. In terms of contributing to natural pigmenta-
tion, observable coloured quinones are widely distributed in plants,
particularly trees where they contribute to the pigmentation of the
heartwood and therefore have considerable interest to the timber and
furniture trades.
Most quinones and their derivatives are probably bitter to the taste and
relatively toxic. Perhaps for that reason they figure in the literature as anti-
herbivore defences, that is chemical defences formed in plants against
challenges by would-be herbivores. There are problems with proving that
this is the natural function of this class of pigment if only because it is
difficult to show that animals avoid plants rich in say naphthaquinones
solely because of the taste of the quinone. Certainly the plant genus
Hypericum, which contains the toxin hypericin, has few natural enemies
and is a notorious poisonous plant to livestock. As an introduced weed
species in Australia it is a serious problem.
Quinoids, as a group, probably make little contribution to the colours of
the visible parts of plants. They do, however, contribute to some of the
duller colours of certain fungi and lichens (typically yellow, orange or
brown) but also to the brilliant red, purple and blues of sea urchins,
crinoids (sea lilies), coccid insects and several species of visibly coloured
aphids.
In general, the simpler quinones, such as benzoquinones, are colourless
except in high concentration when they may have a pink hue. The
naphthaquinones and anthraquinones are frequently present in plants with
complex additions or substitutions. These substitutions provide colours
ranging from almost black or deep purple to deep wine-reds to orange and
through to yellow. Further colour changes are readily made in vitro by
simple additions of hydroxyl groups under basic conditions.
Historically, certain quinone pigments have been the mainstay of the
textile dyestuff trade. Quinoid-rich plants were formerly cultivated solely
for their pigment. The deep-red pigment from the roots of Rubia species,
such as madder (R. tinctorum) is a mixture of anthraquinones, while the
yellow-red-brown naphthaquinones of Lawsonia alba produce the cos-
metic dye henna. Two related anthraquinones have long been established
as food colorants, the dyestuff kermes from the coccid insect Kermococcus
ilicis and cochineal (carminic acid) from another insect Dactylopius coccus
(see chapter 9).
A final group of quinoid pigments are the allomelanins, usually dark
brown to black pigments of fungi and some plants, which are chemically
unrelated to the true or eumelanins of animals. The precise biosynthetic
origin of most of the allomelanins is unknown but they appear, structurally
at least, to be related to polymers of simple phenols and their quinones.
1.9.5 N-Heterocyclic compounds (other than tetrapyrroles)

From a range of largely unrelated N-heterocyclic compounds found in
biology there are two groups that show at least some structural affinities as
well as being pigmented; namely (i) indigoid and indole derivatives and
(ii) substituted pyrimidines.
1.9.5.1 Indigoid and indole derivatives The blue textile dye indigo
(Figure 1.11) was until relatively recently an economically important
product of the semi-tropical plant Indigo/era tinctoria (and near relatives)
and the European woad plant !salis tinctoria. Indigo constitutes one of the
oldest colorants known to man. The intact living plants, however, contain
not indigo but colourless glycosides such as 3-hydroxyindole. Only when
extracted, hydrolysed, oxidized and dimerized is the blue dye, indigo,
formed. Another indigoid, the ancient dye Tyrian purple, was derived by
photochemical oxidation and extraction of a colourless indigoid from
certain Mediterranean molluscs to yield 6,6' -dibromo indigo (Figure 1.12).
This illustrates the important point that several pigments considered to be
'natural' are actually colourless in the living cell. Only on isolation from the
host organism, and following relatively complex chemical modifications,
do they provide dyes or colorants of commerce. It is often their long
history of use that endows them with the accolade 'natural'. For all their
artificiality, these indigo dyes are vigorously defended as 'natural dyes' by
the wool dyers and weavers in contrast to the identical indigo synthesized
by Baeyer in 1878 and for many years an important product formed from
distillation of coal tar (formed from fossilized land plants) and later from
oil (formed from degraded marine plants). The boundaries between
'natural' and 'artificial' come no closer than in indigo chemistry.
Structurally related to indigo through the common indole group, an
important group of plant pigments are the betalains. Unlike indigo, the
betalains are true pigments in the living plant that give rise to the distinct
o:XJo H 0
Figure 1.11 Indigo.
o H
~
N -..;;:Br
~I I~
Br
H 0
Figure 1.12 Tyrian purple, dibromoindigo.

deep-red pigmentation of beetroot and the crimson of Amaranthus

flowers. As they are considered in detail in chapter 8, only an outline will
be provided here. The betalains consist of two sub-groups, the betacyanins
(Figure 1.13) and betaxanthins (Figure 1.14). Being water-soluble and
produced in abundance by, for example, the beet plant Beta vulgaris, the
betacyanins are an established food colorant. Of interest to those studying
the origin and evolution of higher plants, the betalains are something of a
mystery. They are found in just one sub-group of extant flowering plants,
the Centrospermae. The living Centrospermae have either lost, or perhaps
never acquired, the genes for the synthesis of the other great family of
plant pigments, the flavonoid anthocyanins. Genes for the synthesis of
betalains appear again in a few otherwise unrelated fungi, most colourfully
perhaps in fly agaric Amanita muscaria as violet and yellow pigments.
The betalains are, however, structurally and biosynthetically related to
the widespread group of animal pigments the melanins, more correctly the
eumelanins. In mammals, including man, the melanins are the major
pigments familiar as the colorant of black and brown skin and hair.
Although the natural function of eumelanins is probably the protection
from ultraviolet light (UV-B radiation 280 to 320 nm) and, in association
with the pineal gland in light perception, interest in these pigments has
arisen largely because they may serve as indicators of malignant mela-
nomas in man. In mammals, melanins are frequently linked to metals (Fe,
H0XQ<
HO~
I ~ ",H
N "" COOH
HOOC~H H COOH
Figure 1.13 A betacyanin, betanidin.
n."",H
HOOC'X~~I
'"
COOH
H ~ COOH
Figure 1.14 A betaxanthin, indicaxanthin.

Figure 1.15 Eumelanin, part of a linked series of indole-5,6-quinone.
eu and Zn) and attached to proteins isolated in discrete packages called

melanosomes, which are present in light-receptive areas ofthe body. Other
related but simpler melanins have quite different functions; in the cuttle
fish Sepia officinalis, for example, melanins are released as a defensive
ink.
For many years the structure of the eumelanins was debated, the starting
material for most analyses being an amorphous, insoluble and frequently
impure protein-bound isolate. The simplified 'textbook' structure was that
of a long-chain polymer of units of indole-5,6-quinone (Figure 1.15). In
practice, the true structure of a eumelanin molecule will probably remain
elusive given numerous and perhaps random variants in points of attach-
ment and the presence of branches and cross-linkages. Whatever the
eumelanins lack in structural consistency, they are all virtually insoluble in
water and most organic solvents and, except under fairly drastic chemical
conditions, they are highly stable. Eumelanins are essentially black in
perceived colour.
An unrelated class of pigments, the phaeomelanins, also contribute to
mammalian colours as light brown, red, yellow or blond colours of hair.
Similar pigments contribute to some of the colorants present in feathers of
birds. At least the red phaeomelanins reveal an indole origin within the
complex structures of these sulphur-containing pigments. Unlike the
eumelanins, the phaeomelanins can be dissolved in dilute alkali but in
other respects they are stable and relatively inert. An even less well
defined group of pigments, the phytomelanins, are found in seed coats,
particularly of the Compositae family. Despite having been in the literature
for 50 years or more, little has been published either on the occurrence of
phytomelanin outside the Compositae or on its structure or chemical
composition. The pigment appears to be polymeric, highly insoluble,
resistant to alkali and acid and in high concentration is black in colour. Its
natural function is unknown. Black-coated seeds, however, can be recov-
ered after many years of burial in the soil and which will have resisted
weathering, soil acidity and microbial attack. This suggests a remarkable
degree of stability. (The name phytomelanin should not be confused with a
flavonoid phytomelin, otherwise known as quercitin 3-rutinoside which is

yellow.)
1.9.5.2 Substituted pyrimidines Components of this diverse group show

little biosynthetic relationship, their grouping here is entirely due to their
superficial structural similarity and to some important physical characteris-
tics in common.
The major groups are the purines (Figure 1.16), and the pterins (Figure
1.17) each with the 6C pyrimidine ring substitute with a 5C imidazole or 6C
pyrazine ring respectively, bearing in all four nitrogens. In addition, three
minor groups are also conveniently considered here, structurally related to
pteridine by a further 6C substitution to form flavin (note spelling) Figure
1.18) or without the nitrogens on the first ring to form the colourful
phenazines (Figure 1.19) or with an O-substitution for a N-group and
called phenoxazine (Figure 1.20). Given this somewhat tenuous link
between these diverse structures, they possess several common features.
While many are not at all common in biology, each group contains several
individually important pigments.
o=:~
N H
Figure 1.16 Basic structure of a purine.
Figure 1.17 Basic structure of a pterin.
Figure 1.18 Isoalloxazine, basic structure of flavin.
Figure 1.19 Basic structure of phenazine.

Figure 1.20 Basic structure of phenoxazine.
A significant group of pigments in the animal world is formed from

purines, particularly guanine, xanthine and uric acid. In the form of micro-
crystals or granules they are the chemical basis for several structural
colours, notably white, semi-transparent cream colours and silver. They
are particularly common in the scales of fish.
More widespread perhaps are the pterins, contributing to the white,
cream, yellow and reds of many insect groups. For example, the intricate
patterns of colour in the wings of butterflies and moths are often formed
from pterins including the mustard-yellow colours of leucopterin (Figure
1.21). The bright yellow of many wasps is due to xanthopterin (Figure
1.22) while other yellows, oranges and reds of crustaceans, fish, amphi-
bians and reptiles are due partly to several other pterins. Most of the
pterins are soluble in dilute acid or alkali and in the dark are relatively
stable. Unfortunately most are also photo-labile when isolated from their
biological carrier.
Another widespread group of substituted pyrimidines are the flavins,
and one in particular, riboflavin, is an important component of redox
proteins throughout biological systems, particularly in respiratory
enzymes. Its more familiar name is vitamin B2 . Despite its widespread
occurrence, riboflavin seems rarely to function as a visible pigment in
biology. Exceptionally, a small number of yellow-coloured marine inverte-
brates owe their pigmentation to riboflavin, and flavin-pigmented bacteria
Figure 1.21 Leucopterin.
Figure 1.22 Xanthopterin.

are occasionally found in nature, although more recently in deliberately

genetically engineered vitamin B2 -forming microbes. The free riboflavin is
highly soluble in water, appears bright yellow in concentrated solution and
has an absorbance spectrum similar to the otherwise unrelated caroten-
oids.
The phenazines have, superficially, a structural resemblance to the
flavins. Reported so far only from certain bacteria, most commonly species
of Pseudomonas and Streptomyces, the phenazines are often brightly
coloured-mostly yellow but exceptions include the deep-blue pyocyanine
(Figure 1.23), and violet-blue iodinin (Figure 1.24). Occasionally, these
and other phenazines are found in mammalian (including human) tissues
associated with, for example, infected wounds, presumably from a bacter-
ial invasion. Many of the phenazines are (or can be readily converted to
become) water-soluble and appear to be relatively stable at least under
laboratory conditions. Functionally, the phenazines may be formed as
chemical protectors (antibiotics) against other bacteria. They have import-
ant applications as redox mediators in biochemistry and some are available
commercially.
The final group of substituted pyrimidines, the phenoxazines, are
structurally similar to the phenazines but with an oxygen (0) substituted
for a nitrogen (N). They constitute a group of invertebrate pigments, the
ommochromes, and enjoy a welter of names. Among the more common
ommochromes is the yellow xanthommatin (Figure 1.25), most of these
pigments being yellow, golden-yellow or yellow-brown, although darker
IN~
((
~ +N N
I
CH,
Figure 1.23 Pyocyanine.
YI~~
~NN
b-
OH
Figure 1.24 Iodinin.

COOH
H,N-CH
&~
~
¢H
C=O N
I
H,C
;s
0
:
O!
IN
"
OH
~ 0
Figure 1.25 Xanthommatin.
browns are common too. Occasional variants give rise to red-browns and
mauves. Many ommochromes may serve to screen out stray light in the
eyes of some invertebrates. They are also widespread in higher invert~
brates where they may function in camouflage. Phenoxazines are also
reported from microbes, for example some Streptomyces spp. produce a
pink-red phenoxazine, actinomycin, with antibiotic properties. Soluble in
acids and in acidified alcohols, many phenoxazines undergo a colour shift
often acquiring a red hue. In alkali, they are more yellow and not
infrequently fairly unstable. It seems likely that in vivo the ommochromes
are stable in the reduced form.
1.9.6 Metalloproteins
An otherwise unrelated group of pigments, the metalloproteins form an

important group of proteins in biological systems. One group, the metallo-
porphyrin haems and chlorophylls, has already been described above. A
list of the more widespread (and a few that are uncommon) metallopro-
teins are given in Table 1.8. The list is incomplete, however. Many
proteins, on isolation, appear to be associated with one or more metals.
Such associations may be artifacts of extraction. Other protein-metal
complexes almost certainly exist in vivo but have perhaps rarely remained
intact on isolation or purification. Many other metalloproteins are well
known but are not shown in Table 1.8 because they are not present in
sufficient quantities to form a recognizable pigment. The cobalt-containing
porphyrin-derivative chromophore of vitamin B12 that is naturally present
in certain bacteria in trace amounts is therefore excluded from the list, as
are the many haem-containing enzymes such as the cytochromes and
peroxidases, the large number of other iron and iron-sulphur containing
proteins, many of which have been purified and concentrated to the point
where the protein shows a clear colour. Cytochrome c for example, in the
reduced state, is bright scarlet. The Fe-S ferredoxin is yellow. Neither is
present in any organism in such concentrations that they make any
distinguishable contribution to the colour of the cell. It has to be conceded,
Table 1.8 The more widespread metalloporphyrins with minor examples of interest
Metal Chelator Chromophore Chromophore Colour

co-factor plus protein
Fe Protoporphyrin Haems (of several Haemoglobin Red

types) Myoglobin Red
Chlorocruorin Green
Complexed directly to Oxyhaemerythrin Brick-red to
protein scarlet
Ferritin Red
Haemosiderin Brick-red
Mg Protoporphyrin Chlorophylls (of Chlorophyll- Green
several types) proteins (several
types)
Cu Complexed directly to Oxyhaemocyanin Blue
protein Ceruloplasmin Blue
Erythrocuprein Blue-green
Oxyplastocyanin Blue
Uroporphyrin Turacin Purple
V Complexed directly to Haemovanadin Apple-green
protein
Zn Protoporphyrin Pink-brown
however, that through biotechnology it might be possible to synthesize

many of the metalloproteins in substantial amounts; therefore these should
be considered as potential colorants of the future. However, to keep Table
1.8 to a reasonable size and within the bounds of the natural and present-
day availability of the proteins, only those metalloproteins that function
naturally as pigments or are present in quantities high enough to be visibly
pigmented are included.
Iron-proteins are widespread in biology; they are usually pigmented and
inevitably come in a wide variety of forms. In one important group, the
iron is bound to its own non-protein chromophore, a porphyrin (see Figure
1.2), which, in turn, is associated in various ways with different proteins.
This group includes the familiar haem (or heme) proteins that are, in the
reduced form at least, usually bright red. One exceptional haem-protein is
the oxygen-carrying pigment chlorocruorin (also known as chlorohaemo-
globin, Spirographis haem, or haem s protein), the pigment in the green
'blood' of four families of polychaete worms. However, in most other
major examples of the iron-porphyrin proteins, where there is any distinct
colour, then it is red to purple.
Haems, in whatever form, are pigments composed of or derived from
protoporphyrin chelates or iron (see chapter 6 for fuller descriptions). The
word 'haem' therefore has a precise meaning. Unfortunately this was not
always so. The use of the name haem for several red (and indeed blue)
non-porphyrin pigments stems from the 19th century and reflects unfortun-
ate attempts to name and classify chemicals by their characteristic spectral

qualities. The legacy is a number of metalloproteins with haem as part of
the name but which are not haems in the modern meaning of the word (Fe-
protoporphyrin or its Fe-derivatives). One of these is the misnamed haem-
erythrin, a simple iron-protein, again functioning in oxygen transport,
which although red in colour has no structural resemblance to haem
proper. Present in a few otherwise unrelated marine invertebrates, the iron
is directly bound to amino acids of the protein. Copper (Cu) metallo-
proteins, generally blue to blue-green, are present in a number of inver-
tebrates. Exceptionally a Cu-containing porphyrin, turacin, is a bronze-
purple colorant in one family of tropical birds. A vanadium-containing
protein, haemovanidin, is apple-green in colour. Zn-porphyrins are wide-
spread in the shells of birds' eggs including those of the domestic hen.
1.9.7 Miscellaneous
There are many pigments throughout biology, particularly within the fungi
and invertebrates, that are often of very limited distribution. No single
comprehensive available work attempts to cover this area. However,
reviews from time to time do give some exposure to these often obscure
pigments. Among the more useful sources both for a perspective and for
reports of the more unusual structures the annual (or more frequent)
reviews in Progress in the Chemistry of Organic Natural Products (Fort-
schritte der Chemie Organischer Naturstoffe) and Natural Product Reports
and the journals Liebigs Annalen der Chemie, Lloydia and Phytochemistry.
The pharmacological literature is also helpful.
1.10 Colourless compounds in perspective
Most biological molecules are not coloured, that is, they do not absorb
light in the visible part of the spectrum. As a class, carbohydrates ranging
from simple sugars to polymers of glycogen, starch and cellulose, are
colourless (they appear white). Most naturally occurring amino acids are
colourless as are many proteins (which tend to have a strong absorbance
between 250 and 300 nm). Lipids, including oils and waxes, are at best pale
yellow but generally colourless. Nucleic acids are also without colour. All
of these widespread organic compounds function directly or indirectly in
the primary metabolism of all organisms but such functions do not include
the absorption of light.
Coloured compounds are then much less common than colourless ones
and, in most cases, have quite specialized functions. As a consequence,
their distribution is often restricted. For example, pigments in higher
animals are frequently involved in camouflage (as in amphibians) or in

mating attraction (as in male birds). Pigments in plants and in the fruiting
bodies of some fungi may function as animal attractants in seed or spore
dispersal, other pigments in lichens and algae may be synthesized in
response to extreme desiccation although their natural function thereafter
may not be understood. All of these functions, however vital to the
particular organism, generally reflect the highly specialized life led by that
organism. Only in the cases of chlorophylls (chapter 5), haems and bilins
(chapter 6), and some carotenoids (chapter 7) is there a more widespread
occurrence reflecting directly the involvement of these few pigments in
fundamental process of oxygen and electron transport, photosynthesis and
as antioxidants.
1.11 Pigment classification by natural distribution
The second approach to summarizing the pigments found in biological

systems is to classify them according to the type of organisms (animal,
plant or bacteria) in which they occur. In this way, it is possible to pinpoint
those pigments that are confined to one particular type, family or species.
For example, many lichen pigments are, in terms of abundance, relatively
uncommon. In a classification based on structure, they would inevitably be
grouped as a 'hotchpotch' of miscellanea. By considering lichens separately
as a source of pigments, attention is drawn to one otherwise obscure class
of pigments known as depsides. These depsides and their derivatives form
the basis of the long-practised art of textile dyeing. In more recent decades,
depside derivatives such as litmus have formed the basis for several
chemical indicators and cytological stains.
Despite the considerable advances in biotechnology and gene transfer
across the whole field of biology, the gene products of certain types of
organism lend themselves to exploitation more readily than others. In the
context of food colorants, bacteria, single-celled and simple fungi together
with single-celled algae and perhaps the simpler zooplankton are the most
likely sources of new pigments because they have the potential of being
exploited using existing culture techniques. The pigments of higher
organisms, animal, plant and fungal, may be less accessible to exploitation
because of the structural complexity of the pigment-bearing tissue or
because the pigment is formed only at critical points of development within
a complex life cycle. For example, pigments that function as attractants in
sexual reproduction may be formed only after completion of other aspects
of the life cycle. They may then be less amenable to exploitation through
genetic manipulation. Pigments that are present in low concentrations for
much of the life cycle of the organism and which function in, for example,
stress tolerance may perhaps be more readily exploited.
1.11.1 Plants including algae

While the colour of vegetation is visible from orbiting satellites and
contributes to the overall colour of the planet, the range of pigments
present in plants is surprisingly small. The predominant colours in land
plants result from two types of chlorophyll, no more than four or five
carotenoids with a seasonal contribution from three or so flavonoids. The
oceans yield four common chlorophylls, perhaps six or seven widespread
carotenoids and two forms of phycobilins. The contribution from other
plant pigments, including the betalains, melanins, anthraquinones, naph-
thaquinones, the less-common carotenes, xanthophylls, and the several
thousand flavonoids is relatively insignificant when considered on a global
basis. This fact helps explain why a few plant pigments are produced, or
at least can be produced, commercially in many countries; chlorophyll,
~-carotene and grape anthocyanins being good examples. Other pigments,
such as annatto (extracted from the fruits of one species of tree Bixa
orellana), saffron (from the anthers of Crocus sativa), and turmeric (an
extract from roots of Curcuma longa) , are the products of one or two
countries only. Table 1.9 shows the major classes of pigments present in
plants, including algae, with a short list of the most commonly occurring
forms. More detailed treatments of the occurrence of these pigments is
given in chapters 5 to 9. The purpose of Table 1.9 is to highlight the most
globally abundant plant pigments.
Several thousand plant pigments are excluded from Table 1.9 simply
because they are uncommon (many carotenoids), or of limited natural
Table 1.9 The most abundant pigments in plants including algae
Pigment Common examples Natural occurrence and abundance
Chlorophylls a All photosynthetic eukaryotes

b All land plants, many algae
c,d Brown and other algae
Phycobilins Phycocyanin Blue-green and other algae
Phycoerythrin Red and other algae
Carotenoids Lutein Most abundant xanthophyll. Most
photosynthetic organisms
~-Carotene Most abundant carotene. Most photosynthetic
organisms
Violaxanthins Common in higher plants
Neoxanthins Common in higher plants
Fucoxanthin Brown and other algae
Anthocyanidins Cyanidin The most common anthocyanidin widely
distributed in higher plants
Pelargonidin }
Oelphinidin Common in higher plants
Betalains Betacyanin Widely distributed but confined to one order

of plants
distribution (e.g. turmeric pigments), or natural production occurs only

under specific environmental conditions (e.g. certain anthocyanins of
ripening fruits of senescing leaves) or simply because in overall quantity
the pigment is small relative to other pigments of the type (e.g. bacterio-
chlorophylls) .
1.11 .2 Vertebrates
Fox [6] and Fox and Vevers [7] have provided excellent summaries of the
distribution of pigments among all the major taxa of animals. The account
provided by Needham [8] is more comprehensive. Table 1.10 is then a
digest of a much more extensive consideration but at least emphasizes the
relatively limited distribution and range of pigments in the vertebrates. All
vertebrates (with a few rare exceptions) contain haemoglobin and myo-
globin, which may also contribute to the pigmentation of the animal.
Of the vertebrates, the more colourful classes are the birds, amphibians,
the bony fish and some reptiles. Mammals are generally remarkably dull
and considering that they constitute a substantial element of the human
diet, they provide relatively little to the intake of natural pigments. More
natural pigments are consumed through a diet of fish.
Table 1.10 The major pigments of the vertebrates
Class Pigment
Mammals Mainly melanins

Birds (including their eggs) Melanins
Carotenoids
Tetrapyrroles
Reptiles and amphibians and Melanins
bony fish (teleosts) Carotenoids
Pterins
Riboflavin
Cartilaginous fish Melanins
(elasmobranchs)
1.11.3 Invertebrates
The distribution of pigments in the invertebrates is considerably greater
than in the vertebrates and rivals the land plants in terms of variety. Rather
than give an extensive table covering the many invertebrate classes, Table
1.11 lists only the major invertebrate groupings that are significantly
pigmented. Needham [8] gives a more comprehensive account.
In the context of human diets, the crustacea and molluscs provide
perhaps the most diverse source of pigments including several unusual
carotenoids, bilins, ommochromes and, in many instances, pigments that
have not been characterized. It is for this reason that no serious attempt
Table 1.11 The major pigments of the invertebrates and protozoa
Phylum or major class Pigment
Echinoderms (star fish, sea urchins, Generally highly pigmented, mainly carotenoids,
crinoids) quinones and melanins
Molluscs (snails, bivalves, squids) Melanins, porphyrins (particularly in shells),
ommochromes, many with haemocyanin as a
visible pigmentation. The group also contain
numerous and so far unidentified pigments and
others that are rare but of potential economic
interest
Insects Often highly coloured and containing many
carotenoids, flavonoids, wide range of melanins,
ommochromes, pterins, flavins and occasional
bilins and unusual polycyclic quinones
Malacostraca (centipedes, millipedes) Melanins and carotenoids predominate but many
examples of ommochromes and rare pigments.
Haemocyanin may contribute to pigmentation
Crustacea (crabs, lobsters, shrimps) Highly pigmented with carotenoids, bilins,
melanins, pterins, flavins and some with
haemocyanin as a visible pigment
Arachnida (spiders, scorpions and Varied pigments, including unusual carotenoids,
alJies) pterins, qui nones (especially spiders), some with
haemocyanin and pterins, many with so far
unidentified pigments
Annelids (segmented worms and Varied in. pigmentation , often due to bilins, some
leeches) with haemoglobin or chlorohaemoglobin
Cnidaria (jellyfish, anemones and Often with unusual carotenoids, some with bilins,
alJies) the group contains numerous poorly described
pigments
Porifera (sponges) Mainly carotenoids
Protozoa (plankton) Some with photosynthetic pigment (chlorophylls
and carotenoids)
can be made to establish a comprehensive list of 'natural' pigments

appearing in the average human diet, particularly one that is based, at least
in part, on an intake of sea food. Natural food colorant regulations that are
based on the dozen or so pigments widely consumed in a diet of
mammalian meat, fish, vegetable and fruits (from widely grown sources)
must inevitably ignore the fact that significant proportions of the popula-
tion ingest natural pigments of which we have little knowledge. The most
ill-defined foods, from a description of their pigments, will be among the
shrimps, crabs, krill, lobsters, crayfish and their relatives. On this point
alone, most of us have no idea what we are eating! Apart from a couple of
insects and molluscs (see above), the invertebrates have contributed little
to commercialized colorants.
1.11 .4 Fungi
Fungi, particularly the fruiting bodies, are often brightly pigmented and
even where the colours are dull, exposure of the fungal sap to air (as occurs
in wounding) often yields pigmented oxidation products. The number of

different pigments so far described from all classes of fungi including slime
moulds probably exceeds 1000. Gill and Steglich [10] give structural
descriptions of over 500 in their review. Steglich [11] provides a broad
summary of groups of pigments in the larger fungi while Shibata et al. [12]
give detailed coverage to pigments from other orders.
The summary that follows here is a much reduced digest, highlighting
the major classes of pigment at the expense of the minor or obscure. For
the greater part, the review by Gill and Steglich [10] should be referred to
for fuller details of structure and natural occurrence.
Perhaps the most remarkable feature of fungal pigments is their
extraordinary diversity. If flavonoids are excluded, then fungi must rival
plants in the variety of pigments they contain. This becomes even more
significant when one considers the types of pigments not found in fungi.
The chlorophylls, abundant in plants including algae, are entirely absent
from fungi. Carotenoids are absent from the major families of the familiar
toadstools (with a few exceptions) and confined to just four orders of the
Phragmobasidiomycetidae and one order of Ascomycetes (Discomycetes).
Flavonoids have not been reported (reliably) from fungi. Instead, the
fungal kingdom excels in other areas by accumulating pigmented com-
pounds that are either unreported from other biological groups or if
present then in only trace amounts. An example of the latter is riboflavin
(vitamin B 2), which is present in most organisms where it functions in
electron transport but in concentrations that do not contribute to any
visible pigmentation. In the fungal genera Russula and LyophyUum,
riboflavin and its close derivatives are accumulated to such high concentra-
tions in some species as to contribute a deep yellow colour to the organism.
Almost the only pigments common to fungi and to plants or animals are the
betalains, melanins, a relatively small number of carotenoids and certain
anthraquinones. Many of the pigments found in fungi (or lichens) have not
been recorded in any other biological organisms. Fungi, particularly the
more simple single-celled organisms amenable to large-scale culture, offer
the greatest potential for exploration in the field of natural pigments.
As with plants, fungi as a group have a well-develped terpenoid
metabolism; however, remarkably little of this results in the accumulation
of pigmented compounds. This presumably reflects the limited role of
pigments in fungal biology. Sesquiterpenoids are widespread in many
orders of higher plants as aromatic and volatile oils but rarely in
concentrations that would impart visible pigmentation. Most unusually
then, in one fungal genus Lactarius sesquiterpenoids accumulate in the
latex of the fruiting body to yield a range of red, orange, green and blue
pigments. Tetraterpenoid carotenoids are present in all higher plants and
widespread in the animal and bacterial kingdoms. They have, however, a
limited distribution in fungi and are generally confined to j3-carotene with
smaller amounts of a- and 'Y-carotene and lycopene. Xanthophylls appear

to be rare in fungi (Britton [5]). In the higher fungi (which include the
familiar toadstools and bracket fungi) (class Basidiomycetes), carotenoids
appear to be largely confined to three orders, the jelly fungi Tremellales
and Dacrymycetales and to the stinkhom Phallales where odour and
pigmentation function as insect attractants in spore distribution. Unexpect-
edly, at least on systematic grounds, carotenoids appear again in the
familiar chantarelles (Cantharellus and related spp.). In the edible North
American C. cinnabarinus the dominant orange-red pigment has been
reported to be canthaxanthin (13,I3-carotene 4,4'-dione) (although this has
been disputed). If true, this would at least justify the claim that this
pigment, familiar to the food colorant industry, is a 'natural' colorant.
Older (and often still cited) reports of carotenoids in other groups of higher
fungi, including the Agaricales, are to be treated with reservation (Gill and
Steglich [10]) although, as with much else in biology, there are always the
surprising exceptions.
Among the Ascomycetes, carotenoids are also of limited distribution and
again largely confined to l3-carotene and its close derivatives. Carotene
pigmentation is characteristic of the orange-peel fungi (genus Aleuria) , the
related genera Morchella (morels) and Helvella (false morel). Other
carotene-containing groups of discomycetes include the Helotiales. A more
extensive list of fungal species from which carotenoids have been reported
is provided by Gill and Steglich [10]. The red yeast Phaffia rhodozyma is
the subject of biotechnological exploitation as a source of astaxanthin (see
chapter 3).
A biosynthetically unrelated group of malonate derivatives, the ketides,
are found predominantly in the Ascomycetes and notably in the genera
Daldinia, Bulgaria, Hypoxylon, Chlorosplenium or in the Basidiomycete
order Polyporales. The simplest, the tetraketides, contribute to the yellow
pigments of Albatrellus and Phlebia spp. and to the red-brown pigments of
Gloeophyllum spp.
The dime ric oosporeine (Figure 1.26) that is widespread among the
ascomycetes and deuteromycetes, illustrates the quinoid structure of
many, but not all, of these and more elaborate ketides. Naphthalene and
naphthaquinone structures (also known as pentaketides) are widespread as
dark-red, brown or black pigments. 1,8-Dihydroxynaphthalene (colour-
less) on dimerization forms the brown and black perlene quinone of the
fruiting bodies of Daldinia and Bulgaria spp. Several of these naphtha-
quinones are identical to those isolated from higher plants.
Other pentaketides are purple to violet and more elaborated forms
(hexaketides) contribute to the pigmentation of Penicillium and the
otherwise unrelated Hypoxylon spp., a feature emphasizing the impos-
sibility of relating the distribution of many of these pigments to any
recognized systematic arrangement. As many of these naphthalene-
HO~O
°OH
I I I I
H,C 0 CH,
o H H 0
Figure 1.26 Oosporeine.
naphthaquinone derivatives absorb strongly in the ultraviolet part of the

spectrum and particularly below 300 nm, it is possible that they playa part
in both higher plants and fungi as UV-B-screening pigments. Another
group of pigments with no established function, the octaketide anthra-
quinones, feature large in the Agaric genus Dermocybe but are also
present in several otherwise unrelated higher fungi as yellow, orange, pink,
red, purple and brown pigments. Many are also related anthraquinones of
higher plants. Ironically, anthraquinones (and naphthaquinones) of trees
have been shown, in isolation, to have anti-fungal properties, which does
nothing to explain the presence of similar anthraquinones in many fungi of
woodlands. The dimeric anthraquinone xylindeine is the beautiful
turquoise-green pigment of dead wood infested with Chlorosplenium
aeruginosum. Other dimeric forms may have anti herbivore potential; one
pink-red anthraquinone hypericin (Figure 1.10) present in the fungal genus
Dermocybe is the probable toxic component of the chemically well-
defended plant genus Hypericum. Among the often brilliantly pigmented
slime moulds, of which the bright yellow Metatrichia vesparium is an
example, octaketides are important yellow, orange and red colorants.
As with one group of higher plants, the Centrospermae, so several
genera of higher fungi produce deep-red betalain pigments similar to the
red pigments of beet. Unlike beet, most of these fungal betalains are not
produced in large quantities although there are exceptions. The commer-
cial mushroom Agaricus bisporus and others show on wounding, albeit
faintly sometimes, a colour change from pale pink to grey-black, recording
the oxidation of the betalain precursor dihydroxyphenylalanine (L-dopa) to
melanin. The familiar hallucinogenic toadstool Amanita muscaria and
others of the genus produce a range of purple, red, orange and yellow
betalain pigments (Figure 1.13) derived from L-dopa. Final elaboration of
some of these involves condensation of betalamic acid with unusual amino
acids to form a group of yellow musca-flavin, orange musca-aurin and
purple musca-purpurin pigments, several seemingly unique to Amanita and
Hygrocybe. Some of these yellow pigments are structurally related to
bet ax an thins (Figure 1.14) of higher plants and, unlike the carotenoids, are
water soluble.
One pre-eminently fungal pigment group is the terphenylquinones of
which pulvinic acid and its modified forms are examples (Figure 1.27). In
Figure 1.27 Pulvinic acid derivative, leprarinic acid.
their many variants, these aryl pyruvic acid derivatives produce intense
colours across the complete spectrum, although sometimes unrevealed in
the intact (uninjured) fungus. The blue pigment of boletes (Boletales), for
example, appears rapidly on damage due to oxidation of variegatic
(tetrahydropulvinic) acid. In the hydroxylated forms, pulvinic acids are
responsible for the yellow and reds of most boletes. In other groups,
terphenylquinones produce striking pigments particularly greens, violets
and dark reds and, in isolation, will readily form other colours when
titrated against alkali. In the dime ric form, the terphenylquinones con-
tribute to the deep-chocolate colour of boletes.
Another group of fungal phenol derivatives, more properly phenyl-
propanoid and cinnamic acid derivatives, share some similarities, at
least in biosynthesis, to higher plants. The wood-rotting Fomes fomentarius
produces brown-red purpurogallin (Figure 1.28) derivatives which form
striking blood-red pigments on treatment with alkali. Simpler hydroxyben-
zoic acid derivatives, often with terpene side-chains, contribute to the
yellows and reds of several fungi including Suillus, Polyporus, Gomphi-
dius, Chroogomphus and Coprinus spp. Many appear to have antibiotic
properties.
1.11.5 Lichens
No single group of organisms has such an ancient and extensive use as dyes
as the lichens. These fungal-algal and fungal-cyanobacterial associations
provide pigments found nowhere else in biology and this may explain the
'-':
Qf(
~ I h
OH
OH
HO 0 OH
Figure 1.28 Purpurogallin.

relatively limited number of comprehensive (and readable) accounts on the

subject. Ely et al. [13] have provided a major review of recent discoveries,
Richardson [14], Hale [15] and Hawksworth and Hill [16] have published
brief summaries.
As a class, many lichens are brightly pigmented. Some of these pigments
may act as sunlight filters that protect the photosynthetic algal component;
others appear to have antibiotic properties at least in vitro. For the
majority, however, the functions of these pigments are unknown. Neither
the light filtration nor antibiotic properties have promoted a significant
amount of research. Inevitably, more is known about the pigments that
have (or have had) a place in human economy, the lichen textile dyes,
although these probably constitute a relatively minor part of the total
complement of pigments in the group. Once of considerable commercial
interest, formation of lichen textile dyes depends largely on the structural
modification of compounds, many of which are colourless in their native
state. Interest in these dyes waned considerably in the latter half of the
19th century, although their production has continued in northern Scot-
land, Norway and elsewhere in home-based natural dyeing. Unfortunately,
their limited distribution and very slow rates of growth make their
collection neither commercially worthwhile nor desirable from a conserva-
tion viewpoint. Small quantities of lichen dyes have also been produced
for many years as reagents for biological stains and as pH indicators.
Perhaps the most important group of lichen dyes has been those derived
from the numerous depsides and depsidones characteristic of many
lichens. The basis of depside dye formation is now well understood as
shown:
l
Depsides
decomposition products ~ orcinol (orcin)
gyrophoric acid (mild alkaline hydrolysis) (brown)
evernic acid ~ (1.1)
lecanoric acid orcinol + NH4 0H + O 2 ~ orcein
(purple)
Orcein itself (Figure 1.29) is a mixture of oxy- and amino-phenoxazone or

phenoxazine, the individual components varying from brown-red to blue-
violet. Insoluble in water, most components are soluble in dilute acid or
alkali. Some eight different pigmented components of orcein have been
characterized with several other partially identified. One group of orcein-
based dyes from Rocella tinctoria, Ochrolechia tartarea and other lichens
gives rise to the purple to red-mauve pigments known as orchil, or to the
Scottish variants corkir and cudbear. Related species have provided the
HO
N;:xCH·
R,
'" "
o ~ 2
Figure 1.29 Orcein (orcinol).
@
HO 0 OH
~
II"
h
CH,O CH,
o
Figure 1.30 Parietin.
source material for the pH indicator litmus. Litmus itself has also been
used for colouring beverages (Hawksworth and Hill [16]).
Probably more widespread than depsides are the phenolic pigments of
lichens, although as a group they are less well-documented. Some 40 or so
anthraquinones, ranging from yellow to red, have been described in some
detail and are exemplified by the bright orange to orange-red colours found
in the genus Xanthoria. The structure of one, parietin, typifies the group
(Figure 1.30). Naphthaquinones contribute more to the red to blood-red
colours found, for example, in the apothecia of Cladonia (rhodocladonic
acid) and Haematomma spp. (haemoventosin). The structurally more
varied terphenylquinones provide deep-red to purple colours of which
polyporic acid (Figure 1.31) is one of a small number of pigments in this
class. Biosynthetically related to the terphenylquinones, pUlvinic acid and
its derivatives represent a widespread class which provides a range of
yellow to orange pigments. Some of the pulvinic acid derivatives may have
a role as anti-herbivore defence chemicals. One, vulpinic acid from
Letharia vulpina, appears to be one of the ingredients in a concoction
designed to poison wolves in Lapland. Absorption if not digestion needs to
be aided by the addition of ground glass in the bait (Hawksworth and Hill
[16])! The more elaborate chromones, of which siphulin is an example,
have a structural resemblance to flavonoids, including the anthocyanins of
higher plants. As lichen pigments they are generally pale yellow.
Bright and dull yellow or yellow-orange pigments are particularly
widespread among lichens. Apart from the phenyl derivatives discussed
above lichen xanthones, norlichexanthone and lichexanthone (Figure
1.32), are relatively widespread pale yellow to yellow pigments. Chlorin-
Figure 1.31 Polyporic acid.
&-
CH, 0 OH
11-"":::
~ h
CH,O 0 OCH,
Figure 1.32 Lichexanthone.
HO -j::ft;(
COCH' 0
~ ~ :-- 0
CH - CH, COCH,
, OH OH
Figure 1.33 Usnic acid.
ated examples have also been recorded, although these are not apparently
common. Some have larvicidal properties. Even more widespread is the
class of pale yellow pigments based on usnic acid (Figure 1.33) and its
several dibenzofuran derivatives. Relatively little is known of the function
of this large class of pigments. In some species, usnic acid may function as a
sunlight filter under conditions of high irradiance and desiccation but its
antibiotic properties have given rise to a substantial literature. All lichens
contain carotenoids (as well as chlorophylls) if only because part of the
dual organism is photosynthetic. However, several xanthophylls have been
found synthesized within the fungal component. In addition, a few of the
algal Xanthophyta and Phaeophyta form lichen so giving rise to marine
xanthophylls (and chlorophyll c) in land plants.
Most reviews of lichen pigments lack a perspective either by concentrat-
ing on one class of pigment to the exclusion of others (e.g. dyes) or tend to
be chemotaxonomic 'block-busters' with no indication of the obscurity or
commonness of each group of pigments. Chemotaxonomy rarely demands
quantification. The following summary is an attempt to redress this state.
Chemically, the lichens are outstandingly rich in compounds derived from
relatively simple phenols including quinones. It is doubtful if there are
many genera that do not accumulate phenol derivatives to concentrations

where they impart visible coloration to the organism. Many of these
compounds, perhaps the majority, are unique to lichens. The most
widespread groups probably include the anthraquinones, rivalling higher
plants in diversity, pulvinic and usnic acid derivatives and perhaps the
xanthones. Despite their apparent abundance throughout the lichens,
many are poorly described (particularly structurally). From reviews such as
that of Ely et al. [13] it is highly likely that many more of these compounds,
particularly the terphenylquinoids, await description. Just as widespread
are the carotenoids common to many algae (and higher plants). The
carotenoids (trans)formed within the fungal partner are probably less
common. At the other end of the scale of abundance, the depsides,
particularly those of commercial interest, are not at all widespread, being
confined to a few species. The abundant literature covering lichen depsides
can give a false view of these uncommon compounds.
1.11.6 Bacteria
In general, bacteria contain many pigments that are similar, if not
identical, to those of more complex organisms-particularly plants. Bac-
terial chlorophylls differ from plant chlorophylls in the reduction of one
double bond. Although bacteria also possess their own complement of
distinct carotenoids, these pigments are structurally and biosynthetically
closely related to the carotenoids of plants and animals. Many bacteria,
both photosynthetic and non-photosynthetic, also accumulate the more
familiar carotenoids such as ~- and 'V-carotene.
In general, bacteria appear to be relatively poorly endowed with
pigmented quinoids, with melanins and (almost) not at all with ftavonoids.
They do, however, make up for this in the production of some strikingly
brilliant pigments apparently unique to bacteria and often to just one
genus. One group of pigments apparently confined to bacteria are the
phenazines based on a dibenzopyrazine skeleton. Among these often
intensely coloured compounds, are the purple iodinin from species of
Chromobacterium and the dark blue (in acid solution) pyocyanine (Figure
1.23) isolated from Pseudomonas aeruginosa. Many of the several dozen
phenazines so far described (see Ingram and Blackwood [17]) have
potential commercial interest particularly as antibiotics.
Several other intensely coloured compounds have been isolated from
certain bacteria and which have little resemblance to pigments in other
biological systems. Indigoidine or 'bacterial indigo', a dimeric pyridine
structurally unrelated to the indigo of plants, is found in Pseudomonas
indigo/era. The highly pigmented genus Chromobacterium has also yielded
(in acidic solution) the dark-red antibiotic prodigiosin with a most
uncommon structure, a trimeric pyrrole. The same genus also produces
dimeric indoles such as the purple violacein, although this has, at least,
some resemblance to the indole derivatives of higher plants. Little is
known of the function and properties of these unusual pigments.
1.12 Biological systems as sources of pigments for commercial

exploitation
Increasingly, with the improvements in fermentation and other bio-

technological techniques, bacteria, single-celled fungi and protozoa
(including photosynthetic plankton) offer considerable scope for the
commercial production of many pigments. The more future prospects
using gene transformation techniques will allow, in theory, the biological
production of almost any stable natural pigment. The limits, however, are
not in the technologies, despite some major problems, but in our ignorance
of what is present in the biological world. At its simplest, despite centuries
of interest in natural pigments, we have had a very myopic view of the
coloured natural pigments.
The chemistry, biochemistry and chemosystematics of higher plant
pigments are relatively well-developed, often to considerable detail. In
contrast, those of algae, particularly single-celled ones, are not. For
example, of the five new chlorophylls described in the past 20 years, all are
algal. The likelihood is that the next two or three chlorophylls will also be
from algae (S.W. Jeffrey [18]). Yet algae lend themselves to biotechno-
logical exploitation. Some progress is being made with a few organisms of
which Dunaliella is one example (see later chapters). In contrast, bacterial
pigments appear to be relatively well-documented and several pigments
(including pigmented antibiotics) are already produced industrially from
bacteria.
Much less is known of animal pigments, perhaps because of their
diversity particularly among the invertebrates. Despite this plethora,
animals have never contributed more than perhaps a dozen pigments of
commercial value and it is not clear why this has occurred. The mounds of
mussel shells from archaeological digs of dye works in the Mediterranean
suggest that our forefathers had a different attitude to the value of animal
byproducts.
Perhaps the richest source of pigment 'pickings' will be among the fungal
kingdom. Many of these, including the myxomycetes (slime moulds), are
highly pigmented. A significant proportion of fungal (including lichen)
pigments are unique to the group and their attributes barely explored.
Other than as textile and indicator dyes, few fungal pigments have been
successfully exploited for their colorant properties for any length of time.
Yet the several hundreds of fungal pigments so far described are among the
most colourful of all biological systems.
Acknowledgements
Several colleagues have provided me with advice on their particular areas

of expertise and I readily acknowledge, with thanks, the valuable discus-
sions with George Britton (animal carotenoids), Rod Cooke and Tony
Lyon (fungal pigments), David Hill, Oliver Gilbert and David Lewis
(lichen pigments) and Owen Jones and Stan Brown (tetrapyrroles). I am
also grateful to Richard Winton for correcting my now rusty Latin and
Greek.
References
1. Steam, W.T. Botanical Latin. David & Charles, Newton Abbot (1973).
2. Flood, W.E. The Origin of Chemical Names. Chemical Society, London (1963).
3. Brown, S.B. Introduction to Spectroscopy for Biochemists. Academic Press, London
(1980).
4. Harborne, J.B. The Flavonoids, Advances in Research Since 1980. Chapman & Hall,
London (1988).
5. Britton, G. The Biochemistry of Natural Pigments. Cambridge University Press, Cam-
bridge (1983).
6. Fox, D. Biochromy. University of California Press, Berkeley (1979).
7. Fox, H.M. and Vevers, G. The Nature of Animal Colours. Sidgwick & Jackson, London
(1960).
8. Needham, A.E. The Significance of Zoochromes. Springer-Verlag, Berlin (1974).
9. Hendry, G.A.F. and Jones, O.T.G. Journal of Medical Genetics 17 (1980) 1-14.
10. Gill, M. and Steglich, W. Progress in the Chemistry of Organic Natural Products 51 (1987)
1-317.
11. Steglich, W. in Pigments in Plants (F.-C. Czygan Ed.) 2nd edn., Fischer, Stuttgart (1980)
pp. 393-412.
12. Shibata. S., Naume, S. and Udagawa. S. List of Fungal Products. University of Tokyo
Press, Tokyo (1964).
13. Ely, J.A., Whitton, A.A. and Sargent, M.V. Progress in the Chemistry of Organic
Natural Products 45 (1984) 103-233.
14. Richardson, D.H.S. The Vanishing Lichens. David & Charles, Newton Abbot (1975).
15. Hale, M.E. The Biology of Lichens. Edward Arnold, London (1983).
16. Hawksworth, D.L. and Hill, D.J. The Lichen-Forming Fungi. Blackie, Glasgow (1984).
17. Ingram, J.M. and Blackwood, A.C. Advance in Applied Microbiology 13 (1970) 267-282.
18. Jeffrey, S.W. Personal communication.
2 Natural food colours
B.S. HENRY
2.1 Summary
Natural colours have always formed part of man's normal diet and have,
therefore, been safely consumed for countless generations. The desirability
of retaining the natural colour of food is self-evident, but often the
demands of industry are such that additional colour is required. Contrary
to many reports, natural sources can provide a comprehensive range of
attractive colours for use in the modern food industry. In particular, five
natural colours-annatto, anthocyan ins , beetroot, turmeric and carmine
-are widely used in everyday foodstuffs. The factors affecting the stability
of these and other permitted natural colours and their commercial
applications are fully discussed.
2.2 The role of colour in food
The first characteristic of food that is noticed is its colour and this
predetermines our expectation of both flavour and quality. Food quality is
first judged on the basis of colour and we avoid wilting vegetables, bruised
fruit, rotten meat and overcooked food.
Numerous tests have demonstrated how important colour is to our
appreciation of food. When foods are coloured so that the colour and
flavour are matched, for example yellow to lemon, green to lime, the
flavour is correctly identified on most occasions. However, if the flavour
does not correspond to the colour then it is unlikely to be identified
correctly [1].
Colour level also affects the apparent level of sweetness; in one study
Johnson et al. [2] showed that sweetness appeared to increase between 2
and 12% with increasing colour of a strawberry flavour drink. The colour
of a food will therefore influence not only the perception of flavour, but
also that of sweetness and quality. It is also important not to underestimate
aesthetic value. The best food with a perfect balance of nutrients is useless
if it is not consumed. Consequently, food needs to be attractive. Domestic
cooking has traditionally attempted to enhance or preserve food colour.
Pies are glazed with beaten eggs, and lemon juice is used to prevent

NATURAL FOOD COLOURS 41
browning of fruit. Recently, new varieties of peppers of different colours,

yellow and purple, have become available.
Likewise, there is a need for processed foods to be visually appealing.
Colour may be introduced in several ways. The raw materials-the fruit,
vegetables, meat, eggs-have their own intrinsic colour, the processing
conditions may generate colour or a colour may be added. Some food
products have little or no inherent colour and rely on added colour for their
visual appeal.
However, although the colour of fruits and vegetables can vary during
the season and processing can cause colour loss, food manufacturers need
to ensure uniformity of product appearance from week to week-a factor
that does not concern the domestic cook where a slight variation in recipe
or cooking time is usual. For the manufacturer, colour consistency is seen
as visual proof of absolute consistency of the manufacturing process.
Colours may be added to foods for several reasons, which may be
summarized as follows [3]:
1. To reinforce colours already present in food but less intense than the
consumer would expect
2. To ensure uniformity of colour in food from batch to batch
3. To restore the original appearance of food whose colour has been
affected by processing
4. To give colour to certain foods such as sugar confectionery, ice lollies
and soft drinks, which would otherwise be virtually colourless
The continued use of colour in food is acknowledged by the Food
Additives and Contaminants Committee (FACC) who concluded that 'if
consumers are to continue to have an adequate and varied diet, attractively
presented, the responsible use of colouring matter, the safety-in-use of
which has been fully evaluated, still has a valid part to play in the food
industry' [3].
2.3 Classification of food colours
It is only in the last 100 years or so, following the discovery of the first
synthetic dye by Sir William Perkin in 1856 and the subsequent develop-
ment of the dyestuffs industry, that synthetic colours have been added to
food. For centuries prior to this, natural products in the form of spices,
berries and herbs were used to enhance the colour and flavour of food.
During this century, the use of synthetic colour has steadily increased at
the expense of these products of natural origin, due principally to their
ready availability and lower relative price. In the last 20 years following the
delisting of several synthetic colours, notably that of amaranth in the USA
in 1976 and that of all synthetic colours by Norway also in 1976, there has
been an increase in the use of colours derived from natural sources.

Generally three types of organic food colours are recognized in the
literature: synthetic colours, nature-identical colours and natural colours.
2.3.1 Synthetic colours

These are colorants that do not occur in nature and are produced by
chemical synthesis (e.g. sunset yellow, carmoisine and tartrazine).
2.3.2 Nature-identical colours

These are colorants that are manufactured by chemical synthesis so as to be
identical chemically to colorants found in nature, for example ~-carotene, 1
riboflavin and canthaxanthin.
2.3.3 Natural colours

These are organic colorants that are derived from natural edible sources
using recognized food preparation methods, for example curcumin (from
turmeric), bixin (from annatto seeds) and anthocyanins (from red fruits).
This description of a natural colour would exclude caramels manufac-
tured using ammonia and its salts and also copper chlorophyllins, since
both of these products involve chemical modification during processing
using methods not normally associated with food preparation.
2.4 Legislation
Natural colours are widely permitted throughout the world. However,

there is no universally accepted definition of this term and many countries
exclude from their list of permitted colours those substances that have both
a flavouring and a colouring effect. Thus spices are generally not regarded
as colours. Sweden, for example, states that 'turmeric, paprika, saffron
and sandalwood shall not be considered to be colours but primarily spices,
providing that none of their flavouring components have been removed'
[5]. Italian legislation states that 'natural substances having a secondary
colouring effect, such as paprika, turmeric, saffron and sandalwood, are
not classed as colours but must be declared as ingredients in the normal
way' [6]. Similar comments are included in the food legislation of The
Netherlands, Switzerland and Norway.
1 j3-Carotene is fundamental as a precursor of vitamin A and also acts as an antioxidant in the

diet as a result of its ability to scavenge singlet oxygen [4].
Table 2.1 Natural colours (and colours of natural origin) listed by the EU
Number Colour
Eloo Curcumin
ElOl Riboflavin, riboflavin-5 ' -phosphate
E120 Cochineal, carminic acid, carmines
El40 Chlorophylls and chlorophyllins
E141 Copper complexes of chlorophylls and chlorophyll ins
E150a Plain caramel
E153 Vegetable carbon
El60a Mixed carotenes and J3-carotene
El60b Annatto, bixin, norbixin
El60c Paprika extract, capsanthin, capsorubin
El60d Lycopene
El60e J3-Apo-8' -carotenal (C30)
E161b Lutein
E161g Canthaxanthin
E162 Beetroot red, betanin
E163 Anthocyanins
The European Union (EU) permits a wide range of colours, some of

which are of natural origin and these are listed in Table 2.1. It is, however,
important to note that lycopene is not widely available commercially and
four of the colours are only available commercially as nature-identical
products, namely E101 riboflavin, riboflavin-S'-phosphate, E160e l3-apo-
8'-carotenal and E161g canthaxanthin. I3-Carotene is available in both
natural extract and nature identical forms, the latter of which is more
widely used.
Council Directive 94/36IEC specifies conditions of use for permitted
colours [7] and is fully implemented with effect from 20th June 1996. The
directive stipulates the purposes for which individually approved colours
may be used and in many cases it also limits the levels at which they may be
applied. Generally the directive is less restrictive in its treatment of natural
colours many of which can be used at quantum satis, a flexibility that is not
granted to any of the azo-dyes.
The USA has a different set of 'natural' colours and those currently
permitted by the Food and Drug Administration (FDA) are listed in Table
2.2; these do not require certification (therefore do not have FD & C
numbers) and are permanently listed.
One of the advantages of using natural colours is that they are generally
more widely permitted in foodstuffs than synthetic colours. It should be
remembered that colour usage may be controlled in three distinct ways as
follows:
1. National legislation lists those colours that may be used in foods
2. Colour use within a country is limited by the type of food that may be
coloured
Table 2.2 Natural colours (and colours

of natural origin) listed by the FDA for
food and beverage use
Annatto extract
~-Apo-8' -carotenal"
~-Carotene"
Beet powder
Canthaxanthin"
Caramel
Carrot oil
Cochineal, carmine
Cottonseed flour, toasted
Fruit and vegetable juices
Grape colour extract
Grape skin extract
Paprika and paprika oleoresin
Riboflavin"
Saffron
Turmeric and turmeric oleoresin
" Nature-identical forms only.
3. The maximum quantity of colour that can be added to a food may also
be specified
Thus for example, beetroot extract is permitted in Sweden. However, its
use is limited to specified food products only, such as sugar confectionery,
flour confectionery and edible ices. There is a maximum dose level limit of
20 mglkg as betanin in the first two categories of food and 50 mglkg in
edible ices. 20 mglkg equates to 0.4% of beetroot extract containing 0.5%
betanin, which is the standard strength for such an extract. Its use,
however, is not currently permitted in dry mix desserts or milk shakes,
although now that Sweden is a member of the ED its food legislation will
need to change and beetroot extract will become a quantum satis colour.
It is therefore essential to consult the legislation relating to the particular
food product before using any colour. It is obviously insufficient just to
check the simplified list of permitted colours since this relates solely to one
aspect of colour regulation.
2.S Factors affecting colour choice
When deciding which natural colour to use in a specific application it

should be noted that there are several factors to influence that choice as
follows:
1. Colour shade required. It is often necessary to blend two or more
colours together in order to achieve the desired shade.
2. Legislation of the countries in which the food is to be sold.
3. Physical form required. Liquid natural colours are generally more cost
effective than powder forms.
4. Composition of the foodstuff-in particular whether it is an aqueous
system or whether there is a significant level of oil or fat present. The
presence of tannins and proteins may limit the use of some colours such
as anthocyanins. It should also be determined whether a cloudy or
crystal clear colour is required.
5. Processing conditions-particularly the temperatures used and the
times for which these temperatures are held.
6. pH. The pH of a food will often determine the suitability of a particular
colour for a given application. The stability or colour shade of most
natural colours are affected by pH.
7. Packaging. This will determine the amount of oxygen and light that
reaches the product and hence the suitability of such colours as
carotenoids and curcumin.
8. Required shelf-life and storage conditions.
2.6 Factors affecting colour application forms
Once these factors have been defined it will enable the most suitable
natural colour to be selected by reference to the relevant sections of this
chapter. It should also be remembered that natural colours are a diverse
group of colorants with widely differing solubility and stability characteris-
tics. Consequently each colorant is available in several different appli-
cation forms, each formulated to ensure that the colour is compatible with
a particular food system. A product application form is a formulation that
enables a specific food additive to be easily and efficiently incorporated
into a manufactured food product. It may, for example, be a very fine
spray-dried powder of low pigment content for mass coloration. Or it may
be a more complex emulsion incorporating an oil-soluble colour dissolved
in citrus oils, which is subsequently emulsified into an aqueous phase
containing emulsifiers and stabilizers. There are, therefore, several factors
relating to these application forms that should be considered by the food
technologist.
2.6.1 Solubility
Anthocyanins and beetroot are water soluble. Curcumin, chlorophylls and
xanthophylls are oil soluble. Some blends of curcumin and annatto may be
both oil and water miscible.
2.6.2 Physical form

Colours are available in the form of liquids, powders, pastes and suspen-
sions. When using suspensions it is important to realize that colour shade
will change if the pigment becomes dissolved during processing. An

increase in temperature is often sufficient to dissolve a suspended pigment
and so effect a colour change; this particularly applies to ~-carotene and
annatto suspensions. The viscosity of liquids and pastes is temperature
dependent and their ease of dispersibility into a foodstuff will decrease
with decreasing temperatures.
2.6.3 pH
Most liquid, water-soluble natural colours are manufactured with their pH

close to that of their maximum stability. Thus extracts containing norbixin
are alkaline, whereas anthocyanins are acid. Addition of such products to
an unbuffered solution is likely to alter the pH of that solution.
2.6.4 Microbiological quality
Oil-soluble pigments tend to have low moisture contents and are therefore
not generally susceptible to microbiological spoilage. Anthocyanin and
beetroot extract contain significant levels of water and sugars and thus
steps must be taken to ensure good microbiological quality.
2.6.5 Other ingredients
Some oil-soluble colours such as curcumin and ~-carotene need the

addition of gums, stabilizers or emulsifiers in order to render them water
miscible. It is important that these ingredients are compatible with the food
system to which the colour is being added.
2.7 Performance and consumption of natural colours
As has been stated, natural colours are a very diverse group of compounds
and it is therefore extremely difficult to make general comments about
their nature and performance. However, the literature contains many such
statements that, over the year, have led to general misunderstandings.
One of those frequently repeated statements says that natural colours
have a lower tinctorial strength than synthetic colours and thus require
higher levels of addition. However, in reality the reverse is generally true.
Beetroot, ~-carotene, bixin and curcumin, for example, are all intense
colours and their use in food generally results in a decrease in colour dose-
level. It is interesting to compare the absorptivities of some natural colours
with azo-dyes of a similar shade.
Table 2.3 Relative absorptivities of some natural and synthetic dyes
Natural dyes Synthetic dyes

Dye El%
1 em Wavelength Dye El%
1 em Wavelength
(nm) (nm)
Norbixin 2870 482 Sunset yellow 551 480

Curcumin 1607 420 Tartrazine 527 426
Betanin 1120 535 Amaranth 438 523
Table 2.3 clearly illustrates that these natural colours are considerably
more intense than some commonly used synthetic colours-a fact that is
barely mentioned in the literature.
When natural colours are added to food it is also quite common for the
dose level to be adjusted so that a pastel, more 'natural' appearance is
achieved. This is particularly noticeable when considering the colour
strength of yoghurts and fruit squashes where colour inclusion rates have
been reduced following the introduction of natural or nature-identical
colour. Thus colour inclusion levels in food are tending to reduce on
account of both the strength of many natural colours and the move to more
pastel shades.
Another view is that natural colours are only available in small amounts
and that excessive areas of land would be required to cultivate commercial
quantities. In reality, the colour supply industry has responded quickly to
any increase in demand, although the quantities of most natural colours
required by industry are small in relation to those produced by nature. The
consumption of natural colours as an integral part of the diet is far in excess
of the quantities added as food colour. Kuhnau [8] stated that the normal
daily intake of anthocyanins in food was around 215 mg in summer and
180 mg during winter. Assuming an annual consumption of 70 g per person
this equates to approximately 4000 tonnes of anthocyanin being consumed
each year in the UK whereas the quantity of anthocyanins added as colour
to food in the UK is certainly less than 10 tonnes per year. A similar
comparison can also be made for carotenoids, chlorophylls and beetroot.
2.8 Annatto
2.B.1 Source
Annatto (Figure 2.1) is the seed of the tropical bush Bixa orellana. The
major colour present is cis-bixin, the monomethyl ester of the diapocarot-
enoic acid norbixin, which is found as a resinous coating surrounding the
seed itself. Also present, as minor constituents, are trans-bixin and cis-
norbixin. The annatto bush is native to Central and South America where
(a)
..~Q04
•v
i
:
~
A:I
c 0.0
400 500 100 700
Wanlength nm
(b) 0·'
-~
:
~ 0'4
COOCH, •v
i
:
...
~
A:I
C 0.0
500 700
Wavelength nm
Figure 2.1 Annatto (class carotenoid; synonyms rocou, achiote, CI natural orange 4):
structural and physical characteristics of the constituents. (a) Norbixin chemical formula:
C24H2S04' Molecular weight: 380.5. Visible spectrum is of a 0.0003% solution in NaOH
solution. Colour shade: yellow-orange to orange. Solubility: water-soluble. Absorptivity of
norbixin El% = 2870 at 482 nm in 0.1 N NaOH. (b) Bixin chemical formula: C25H3004'
Molecular weight: 394.5. Visible spectrum is of a 0.0003% solution in chloroform. Colour
shade: yellow to orange-yellow. Solubility: oil-soluble. Absorptivity of bixin El% = 2870 at
502 nm in chloroform.
its seeds are used as a spice in traditional cooking. In Brazil, substantial

quantities of processed annatto seeds are sold in retail outlets often
blended with other ingredients for addition to soups and meat dishes
similar to the use of paprika seasonings in Europe.
Although annatto seed is harvested in many tropical countries including
Bolivia, Ecuador, Jamaica, the Dominican Republic, East Africa, India
and the Philippines, it is Peru and Brazil that are the dominant sources of
supply. The principle export form of annatto is the seed, although to
increase export values, several seed suppliers now also manufacture
extracts. It is likely that around 7000 tonnes of annatto seed are used
annually as a food colour worldwide and, assuming an average colour
content of 2%, this equates to 140 tonnes of bixin. The main market for
annatto is the USA and Western Europe, although there is also consider-
able inter-trade between the Central and South American suppliers.
2.8.2 Extracts and their application forms

Bixin, the principal colouring compound, is oil soluble, although only
sparingly so, and occurs in cis and trans forms. Alkaline hydrolysis of bixin
yields the salts of the corresponding cis and trans dicarboxylic acid
norbixin, which are water soluble. On heating, cis-bixin is converted to the
more stable, more soluble trans-bixin. Further heating of bixin leads to the
formation of degradation compounds (Figure 2.2). Note that the heating of
cis-norbixin does not yield trans-norbixin.
Annatto seeds contain cis-bixin with smaller quantities of trans-bixin and
cis-norbixin. Annatto extracts contain varying proportions of colouring
compounds depending on the extraction process used and the processing
temperature.
Carotenoids, because of the highly conjugated system of double bonds,
are particularly intense colours. Dose levels are thus correspondingly low
and often in the range 5 ppm to 10 ppm.
2.8.2.1 Oil-soluble extracts. Extracts of annatto seed produced using

hot vegetable oil as the extractant contain trans-bixin as the major colour
component and usually have a dye content of only 0.2 to 0.3%. Higher
concentrations are not possible because of the poor solubility of bixin in
oil. Although of low tinctorial strength, such products are ideally suited for
use in foods where there is a substantial quantity of oil present, such as
dairy spreads, salad dressings and extruded snack foods.
2.8.2.2 Suspensions in oil. A more cost effective way of incorporating

bixin is to suspend the undissolved pigment in vegetable oil. By this
method products with 4% bixin are manufactured, both cis- and trans-bixin
being present. However, it is important to remember that the colour shade
obtained depends upon the amount of bixin that dissolves in the oil phase
which, in turn, depends upon dose level and the temperature attained
during processing. A suspension of bixin is orange in colour whereas a
solution of bixin in vegetable oil is yellow and so, as the end-product
cis-bixin heat heat

_ _ _ _ _ _........ trans-bixin _ _ _ _ _ _........ degradation compounds
1 hydroly," 1 hydroly'"
cis-norbixin trans-norbixin
Figure 2.2 The interrelationship of different annatto colouring compounds.

temperature is increased, the bixin dissolves and the colour becomes more
yellow.
2.8.2.3 Water-soluble extracts. Norbixin, the cis form in particular, is

very water soluble and solutions with more than 5% dye can be achieved.
It is usual commercial practice to manufacture a range of alkaline solutions
containing 0.5 to 4.0% norbixin by alkaline hydrolysis of annatto. Extrac-
tion of norbixin from the seeds is not commercially viable because of its
low level of occurrence. In the UK, an aqueous solution containing 0.5%
norbixin is often termed 'single strength' annatto and 1.0% norbixin
'double strength'. However, in the USA, 1.4% norbixin is termed 'single
strength' and 2.8% 'double strength'. All such products are often termed
'cheese colour' and use dilute potassium hydroxide as the diluent.
Norbixin may also be spray dried using a carrier such as acacia,
maltodextrin or modified starch, to give a water-soluble powder colour.
Norbixin concentrations range from approximately 1 to 14% with 1%,
3.5%, 7% and 14% being the most commonly manufactured products.
Due to their large surface area, such products are particularly prone to
oxidation.
2.8.2.4 Process colours. It is also possible to blend together bixin and

norbixin extracts by the use of suitable carriers such as propylene glycol
and polysorbates to produce a colour that can be added to water- or oil-
based foods. Bixin content is usually in the range 1 to 2%. Other extracts,
notably those of turmeric and paprika, may also be blended with such
products to allow a wider range of colour shades to be obtained. These
blends find application in flour confectionery and dairy products where
some oil or fat is present.
2.8.3 Legislation
Annatto extracts are permitted virtually world-wide including the EU,
Scandinavia and North America. Some countries impose limitations on the
levels of use of annatto extracts including those in the EU where maximum
levels are specified for all approved applications [7].
2.8.4 Factors affecting stability
2.8.4.1 pH. Norbixin will precipitate from an acid solution as the free
acid. Thus a cheese colour should not be used in water ice, acid sugar
confectionery or soft drinks. Bixin, however, is not affected by pH, and
special application forms, often based on bixin, can be used in acidic
products.
2.8.4.2 Cations. Divalent cations (particularly calcium) will form salts

with norbixin. Since calcium norbixinate has very poor water solubility,
norbixin extracts are incompatible with products containing high calcium
levels. It is therefore not recommended to make dilutions of cheese colour
using hard water. Particular application grades of nor bixin have to be used
where high levels of salts may be present such as in the processing of cured
fish.
2.8.4.3 Heat and light. When bound to proteins or starch, norbixin is

stable to heat and light. Nevertheless, stability is significantly reduced
when norbixin is present in dilute aqueous systems.
Bixin and norbixin are reasonably stable to heat, although degradation
of bixin can occur at temperatures over lOO°C leading initially to a more
lemon yellow shade and subsequently to loss of colour. Exposure to light
will result in a gradual loss of colour.
Aqueous annatto extracts should not be allowed to freeze otherwise
norbixin may be thrown out of solution.
2.8.4.4 Oxygen. Because of the conjugated double-bond structure, all

carotenoids are susceptible to oxidation. The addition of ascorbic acid as
an oxygen scavenger is beneficial.
2.8.4.5 Sulphur dioxide (S02)' Colour intensity will reduce in the

presence of sulphur dioxide thus alternative preservative systems are
recommended when using annatto as a colour.
2.8.5 Applications
Since annatto extracts are available in so many different application
forms-more so than any other natural colour-virtually any yellow to
orange food product may be successfully coloured with annatto. The most
difficult task is to establish which application form is best suited to a given
food system. The simplest answer to the problem is to consult an annatto
colour manufacturer.
2.8.5.1 Dairy products. Dairy products are the single most important
application for annatto extracts. Cheese, particularly hard cheese such as
Red Leicester and Cheshire, has traditionally been coloured with a
solution of norbixin. Norbixin binds to the milk protein and forms a stable
colour, which is not lost during the separation of the whey. For Cheddar
cheese production, dose levels of norbixin are usually in the range 0.75 to
1.25 ppm in the milk. Similar norbixin extracts are also used to colour
other dairy products, notably vanilla-flavour ice cream where a blend of
norbixin and curcumin (extracted from turmeric) is commonly used. A
typical dose level would be 10 ppm norbixin together with 15 ppm of

curcumin. Bixin is used to colour dairy spreads at a dose in the range 1 to
5 ppm.
2.B.5.2 Flour confectionery. Norbixin is ideally suited for the colouring

of such products since it binds to the flour and forms a stable colour that
will not leach. Dose levels are approximately 4 to 8 ppm for sponge cake.
Other related applications include breadcrumbs, ice-cream wafers, biscuits
and snack foods. For flour confectionery containing a significant level of
fat, then a solution of bixin in oil or a process annatto colour can be used.
2.B.5.3 Fish. Annatto extracts, in the form of special application grades

of norbixin, are widely used in the coloration of smoked herrings and
mackerel. Prior to smoking, the prepared fish are dipped in a brine to
which has been added norbixin at 200 to 300 ppm. The norbixin binds well
to the protein such that no significant amount of colour is lost during the
subsequent processing and the final cooking by the consumer.
2.B.5.4 Sugar confectionery. Norbixin is suitable for use in a variety of

sugar confectionery products, although special application forms need to
be used in acidic products where a crystal-clear colour is required, such as
high boilings.
2.B.5.5 Soft drinks. Norbixin extracts that are acid and light stable,
therefore suitable for use in soft drinks have been developed. Dose levels
for norbixin usually lie within the range 1 to 10 ppm in the ready-to-drink
product. This is one of the most demanding applications for any colour.
2.B.5.6 Meat products. Annatto extracts find application in the color-

ation of edible casings used in the production of frankfurter sausages and
similar products.
They are also suitable for use in combination with carmine in the
manufacture of chicken dishes. A variety of colour shades can be obtained
to enable different ethnic styles to be reproduced.
2.B.5.7 Snack foods. Annatto extracts are used in many snack food
products. The colour is incorporated in several different ways depending
on the nature of the food. Norbixin can be added to the cereal starch prior
to extrusion or bixin may be added to the oil slurry that is used to flavour
the snack after extrusion.
2.B.5.B Dry mixes. Norbixin in powder form is ideally suited for the
coloration of both savoury and sweet dry mixes. For powders that are acid
on reconstitution, special application forms are required.
Plate 1 The traditional staple diet of fruits Plate 2 Blue grapes are the main commercial
and vegetables provides man with a rich source of anthocyanin colours. Anthocyanins
variety of foods with an interesting variety of E163 are the mainly red pigments that are
colours. These colours provide vital informa- responsible for the colours of many edible
tion about the safety, freshness and quality fruits and berries. They also contribute to the
of these foods and even today man continues colours of some vegetables and flowers.
to use colour in making such subjective Together with carotenoids, they provide us
assessments. However, the aesthetic value of with the attractive autumnal colours.
colour should not be overlooked. Eating
satisfies more than just a nutritional need. It
is frequently a social occasion and
its enjoyment is critically dependent upon
appearance.
Plate 3 The majority of anthocyanins providing red and purple shades are normally extracted
from the raw material using alcohol or acidified water. (a) Three solutions of grape skin
anthocyanins demonstrate the change of intensity and shade with changing pH, pH 2.7 (left),
4.5 (middle) and 6.5 (right) . (b) Anthocyanin reactivity to gelatine is dependent upon the
monomeric/polymeric balance, varying from predominantly monomeric (top) to precipitated
polymeric (bottom).
Plate 4 Beetroot Red E162 is obtained by aqueous ex-
traction of red table beet (Beta vulgaris), consisting of
betanin and vulgaxanthin. Freely soluble in water, it pro-
vides an intense red colour for convenient use. It has
limited stability to heat, light and sulphur dioxide; shelf
life in finished products may be partially determined
by the water activity of the system in which it is used.
(b )
Plate 5 Carmine has a long history of use as a food colour. It is obtained from the Coccus cacti
insect principally by a process of aqueous extraction. Preparation of the carmine pigment is
achieved by the formation of the aluminium lake of carminic acid. Carmine provides a bright
cherry red shade. It is chemically a very stable colour and is unaffected by oxygen, light,
sulphur dioxide, heat and water activity. It may precipitate under low pH conditions and
concentrated blends of carmine with strongly acidic ingredients such as flavours should be
avoided. The effect of low pH conditions is shown for (a) a carmine solution at pH 3.0; and
(b) carmine in aqueous solution at pH 7.0. (c) Three solutions show the influence of pH on
the shade of carminic acid solutions at pH 3.0 (left), 5.5 (middle) and 6.5 (right).
Plate 6 The seeds of the annatto bush (Bixa orellana)
have long been recognised and used in some cultures to
provide both colour and flavour to the diet and are the
commercial source of annatto colours.
(b) .-~-
I (c)
.
Plate 7 Annatto seeds provide two pigments, bixin which is oil soluble and norbixin which is
soluble in water. Both pigments are carotenoids and may therefore be adversely affected by
light and oxygen. In extreme cases it is helpful to protect sensitive products using ascorbic acid
(Vitamin C). Norbixin is sensitive to sulphur dioxide at levels in excess of 100 ppm, whereas
hard water or low pH conditions can lead to pigment precipitation unless specially formulated
products are used. These pigments are heat stable and provide an orange hue. They are
frequently offered as blends with other pigments, especially curcumin, to ensure that precise
yellow/orange shades are achieved. Traditionally the main use of norbixin has been in cheese
coloration, but it is used in a much wider variety of applications including breadcrumbs, sugar
confectionery, flour confectionery, dairy products, soft drinks and ice cream. (a) Precipitation
of annatto solutions when used in hard water conditions. (b) Effect of pH 4.0 conditions on
unprotected annatto solution. (c) Clear solution of protected product that withstands low pH
or hard water conditions.
50 50
Plate 8 (a) Curcumin ElOO is the principle pigment

of turmeric, a spice that is obtained from the
rhizomes of Curcuma Zanga. As the spice turmeric,
curcumin has been a component of the diet for
many years. Obtained by extraction from the plant
to produce an oleoresin which is then purified,
curcumin provides a bright, strong yellow shade in
solution. It is an oil soluble pigment that is available
in convenient water dispersible forms that are used
in a wide range of foods. (b) Curcumin is extremely
heat stable and may be generally used in products
throughout the acid pH range. The influence of pH
on colour shade is shown at pH 3.0 (left) and 8.5
(right).
(a) (b)
Plate 9 Chlorophyll E140 is the green pigment that is found in all plants that are capable of
photosynthesis and it has thus always been a component of the diet. The pigment is obtained
by solvent extraction from three main sources, grass, alfalfa and nettles. It is an oil soluble
pigment that provides an olive-green colour. Copper chlorophyllin E141 is obtained by the
coppering of chlorophyll following its alkaline hydrolysis. The improved stability and
brightness of copper chlorophyllin leads to its much wider use as a food colour. The influence
of pH on the stability of copper chlorophyllin solutions is shown for (a) protected product at
pH 4.0; and (b) unprotected product at pH 4.0.
2.9 Anthocyanins
2.9.1 Source
Anthocyanins (Figure 2.3) are those water-soluble compounds responsible
for the red to blue colour of a wide range of fruits and vegetables. The list
of sources is very long and includes grapes, redcurrants and blackcurrants,
raspberries, strawberries, apples, cherries, red cabbages and aubergines
[9]. Chemically, they are glycosides of ftavylium or 2-phenylbenzopyrylium
salts and are most commonly based on six anthocyanidins: pelargonidin,
cyanidin, peonidin, delphinidin, petunidin and malvidin. The sugar moiety
present is most commonly one of the following: glucose, galactose,
rhamnose, arabinose.
In addition, the sugar may be acylated with a phenolic or aliphatic acid.
Some 300 anthocyan ins are found in nature, some fruits (e.g. strawberry)
containing just one or two while others such as 'Concord' grapes have at
least 15 present.
It is possible to extract colour from any of these raw materials but for
economic reasons, grape skins, a byproduct of the wine industry, are the
usual source.
Grapes are the single most abundant fruit harvested in the world from
which an annual consumption of about 10 000 tonnes of anthocyanins has
been estimated [10]. Grapes with highly pigmented skins such as Ancel-
lotta, Lambrusco, Alicante and Salamina have sufficient colour remaining
after wine production to justify colour extraction [11]. It is estimated that
o·a
~
;§ 0·4
•I!!
OH j
Co· 0 ~'--'--L.-L.-'&""";-"
400 700
Wavelength nm
Figure 2.3 Anthocyanins (class anthocyanin: synonyms grape skin extract, grape colour
extract, enocianina): structural and physical characteristics of malvidin-3-glucoside. Chemical
formula: C23H2S012. Molecular weight: 529. Visible spectrum is of grape skin extract as a
0.002% aqueous solution at pH 3.0. Colours shade: red at pH 2, blue-red at pH 4. Solubility:
water-soluble. Absorptivity of grape skin extract: El% = 500 at pH 1.5 at Am"x> closest to
520 nm.
some 10 000 tonnes of grape skins are extracted annually in Europe,

yielding approximately 50 tonnes of anthocyanins. Extraction is carried out
using a dilute aqueous solution of an acid, usually sulphurous acid, and
yields a product containing sugars, acids, salts and pigments all derived
from the skins. It is normal to concentrate this extract to 20 to 30° Brix, at
which strength the anthocyanin content is usually in the range 0.5 to 1%. It
is technically possible to concentrate further but the cost of doing so is not
usually justified.
2.9.1.1 Grape extracts and their colour strength. It is important to

remember that natural food colours, as used by the food industry, are
commonly in the form of a standardized dilute solution of a particular
colour or blend of colours. Thus when calculating pure dye inclusion levels,
it is essential to know the dye content of the colour. It is usual for the
specification for a natural colour to include a measurement of colour
strength based upon the absorbance of a solution of the colour measured
spectrophotometrically. This is often expressed as the calculated absorb-
ance of a 1% w/v solution in a 1 cm cell. However, this is often not
translated into a dye content.
With anthocyanin extracts, the most common method is to calculate
either the absorbance per gram of colour or the absorbance of a 1% w/v
solution of the colour in a pH 3.0 citric/citrate buffer.
2.9.1.1.1 Grape skin extracts. Most liquid anthocyanin extracts de-

rived from grape skins have a colour strength (absorbance/g) of between
150 and 300 (Table 2.4), which equates to 0.5% and l.0% anthocyanin
colour respectively.
Since grapes contain from at least four to over sixteen separate
anthocyanins, depending on the variety, it is not possible to calculate the
exact colour content by a simple analytical technique. To do so you would
need to know the exact percentage anthocyanin composition of the extract.
However, by measuring colour strength at pH 1.0 when the anthocyan ins
are essentially monomeric and by taking an average value for the E} ~m of
Table 2.4 Colour strengths of grape skin extracts
Grape skin E:o/~m Absorbance Approximate %

extract per gram anthocyanin
Standard 1.5 150 0.5

Double strength 3.0 300 1.0
Concentrate 6.0 600 2.0
Powder 12.0 1200 4.0
anthocyanins of 500 at 520 nm then a calculation [12] can be made, giving

the results shown in Table 2.4.
Such extracts can be oven- or spray-dried, using malto-dextrin as the
carrier if necessary, to yield a water-soluble powder. Such a product
usually contains 4% anthocyanin, although more dilute grades are pro-
duced for specific applications.
Thus there are essentially only two application forms of grape antho-
cyanin extract-a water-soluble liquid and a water-soluble powder. Never-
theless, different varieties of grapes are extracted and extraction
techniques vary; this in turn produces extracts with different properties.
These differences will affect colour shade and stability such that those
extracts with a higher level of non-pigmented phenolic compounds are
likely to be bluer in shade. However, too high a level may well induce a
haze or precipitation if used in a soft drink.
2.9.1.1.2 Grape colour extracts. It is also possible to extract antho-

cyanins from the lees derived from grape juice production. Some grape
juices, particularly those produced from Concord grapes (Vitis labrusca),
have high levels of anthocyan ins and tartrates.
On storage, some of the tartrates precipitate bringing down with them a
proportion of the colour. Aqueous extraction of this precipitate (the lees)
will yield an anthocyanin, which, after ion-exchange to convert the
insoluble tartrates into soluble tartaric acid, is suitable for use as a food
colour. Since this type of extract consists of monomeric pigments and a low
level of phenols, it tends to be redder in shade than Vitis vinifera extracts.
In the USA, they are termed 'grape colour extracts' and are classified
separately from 'grape skin extracts'.
2.9.1.2 Other sources. There are also several food products that have
significant levels of anthocyanin present. These include the concentrated
juice of blackcurrant, elderberry, cranberry and cherry, the aqueous
extract of roselle (the calyces of hibiscus) and of red cabbage. These
products contain all the colouring and flavouring principles of the raw
materials, no attempt having been made to concentrate the colour at the
expense of the flavour. Such products are therefore food ingredients and
not colour extracts.
Colour, however, is commercially extracted from red cabbage and this
flavour-free extract, although considerably more expensive than grape
colour, is used in the food industry because of its impressive stability to
heat and light [13].
2.9.2 Legislation
Anthocyanins, particularly those extracted from grapes, are widely permit-
ted. In the USA, grape colour extracts (from V. labrusca) are permitted in
non-beverage foods, whereas grape skin extracts (from V. vinifera) are

permitted in beverages only. In the EU the proposed specification for
anthocyan ins states that they should be obtained by extraction with water,
methanol or ethanol from vegetables or edible fruits. Maximum limits on
levels of colour addition are imposed by many countries; however in the
EU anthocyanins may be used in all approved applications at quantum satis
with the sole exception of fruit-flavoured breakfast cereals where a
maximum level of 200 ppm pure pigment is imposed.
2.9.3.1 pH. Anthocyanins act as pH indicators. They gradually change

from red through blue-red, purple, blue and green to yellow as the pH
increases from pH 1 through 4,6,8,12, to 13 respectively. From a practical
point of view, anthocyanins are only used in acidic products where the pH
is 4 or below. Not only does colour shade change with pH but colour
intensity is also pH dependent being greatest at pH 1.0 and decreasing
rapidly as pH rises. As seen from the plot of absorbance against pH (Figure
2.4), the colour intensity at pH 3.5 is half that at pH 2.0. Also, the gradient
is steepest in the range pH 2.0 to 3.5. Since this is the pH range in which
colour intensity is usually measured, it is necessary to take great care when
preparing the relevant buffer.
90
8 70
j
o
~
11
tC
~ 50
30
o 2 3 4 5
pH
Figure 2.4 Variation of absorbance of grape skin extract with pH.

2.9.3.2 Cations. Some cations, particularly di- and trivalent metal ions,
will cause a bathochromic shift of wavelength of maximum absorption.
This is seen as a distinct 'blueing' of the colour and may result eventually in
precipitation of the pigment. Iron, mild steel and copper surfaces should be
avoided and tinplate containers should be lacquered.
2.9.3.3 Heat and light. Stability to heat is good and is adequate for
processes such as jam and sugar boiling and fruit canning. Acylation of the
sugar moiety will increase stability to both heat and light. Colours from red
cabbage contain a significant quantity of mono- and di-acylated anthocyan-
ins and are particularly stable.
2.9.3.4 Oxygen. Anthocyanins will slowly oxidize when in aqueous

solution. Ascorbic acid does not improve stability under such conditions.
2.9.3.5 Sulphur dioxide (S02). Sulphur dioxide reacts with anthocyan-

ins to form a colourless addition product. This reaction is reversible and on
heating S02 is released to yield the anthocyanin and so restore the
colour-a process that is demonstrated whenever suI phi ted fruit is boiled
to make a preserve.
Sulphur dioxide should not be used as a preservative in products
containing monomeric anthocyan ins as the colour source. In such products
a mixed benzoate/sorbate system is preferred. Alternatively where a
natural red colour is required for a soft drink preserved with sulphur
dioxide it is essential to select an anthocyanin extract which has been
processed so as to ensure maximum S02 stability.
2.9.3.6 Proteins. Some grape extracts will react with proteins (such as
gelatin) to form a haze or even a precipitate. This reaction appears to be
caused by non-pigmented phenolic compounds present in the extract
rather than the anthocyanins themselves, since purified pigments are
compatible with gelatin.
2.9.3.7 Enzymes. Enzyme treatment of fruit juices can cause loss of

anthocyan ins [14] which may, in part, be due to the presence of glucosidase
in the enzyme preparation.
2.9.4 Applications
2.9.4.1 Soft drinks. The principal use of anthocyanin colour is in soft

drinks. Clear drinks with a pH below 3.4 and not containing S02 as a
preservative present an ideal application. Grape extracts, high in oligo-
meric or polymeric colours, have the advantage that they are more stable
in the presence of SOz than those with monomeric colours since the
position of attack of the sulphite ion is blocked. It is a wise precaution
when evaluating natural colours and particularly anthocyanins to let
the food, to which the colour has been added, stand for 24 h before
appraising the colour. This will allow the pigments time to equilibrate and,
in the case of grape skin extract, it is possible to see an increase in colour
during this period as anthocyanin is released from its sulphite derivative.
Dose levels of around 30 to 40 ppm anthocyanin in a ready-to-drink
beverage are usually sufficient to give a deep-red colour. This is a relatively
low level bearing in mind that blackcurrants contain 2000 to 4000 ppm of
anthocyanins.
Storage at elevated temperature (25°C or above) or exposure to sunlight
will cause a significant loss of colour.
Anthocyanins are not usually suitable for use in a cloudy beverage. The
presence of the cloud causes a very noticeable 'blueing' of the colour
because of adsorption onto the cloudifier and the 'thin layer' effect.
2.9.4.2 Fruit preserves. Anthocyanins are also used in fruit prepara-

tions, jams and preserves. The nature and quality of the fruit is important-
-fresh or frozen being preferable to sulphited or canned fruit. Canned
fruit, in particular, can be rather brown in shade and this is hard to mask
using anthocyanins since they themselves absorb in the brown region (420
to 440 nm). Dose levels in such applications vary widely depending on the
natural anthocyanin content of the fruit and the degree of brownness but
are typically in the range 20 to 60 ppm.
2.9.4.3 Sugar confectionery. Acid sugar confectionery, particularly high

boilings, and pectin jellies are an ideal application for anthocyanins where
a variety of red shades can be obtained.
As mentioned previously, some anthocyanin extracts, particularly those
derived from grape, are incompatible with gelatin and thus care must be
taken to select the correct application form to ensure good clarity of the
end product. When a concentrated grape anthocyanin colour is added to a
solution of gelatin, a haze or precipitate is likely to occur. The higher the
concentration of the extract, the more severe the problem. Thus it is
always a wise precaution to dilute the colour prior to use and to establish its
gelatin compatibility before carrying out production trials.
2.9.4.4 Dairy products. It is uncommon to colour dairy products with

anthocyanins since their pH is such that a violet to grey colour would be
achieved. In addition, the presence of the suspended fat particles increases
the visual blueness of the colour. Acid dairy products such as yoghurt can,
however, be successfully coloured although the shade achieved is distinctly
purple. Thus black-cherry flavoured yoghurt is an intense colour because
of the presence of anthocyanins derived from either the grape skin extract
added as colour or the cherry juice present in the formulation.
2.9.4.5 Frozen products. Ice cream is not usually coloured with antho-
cyanins because its pH is too high, beetroot red being the preferred red
colour. However, although water ice with a pH of around 3.0 would seem
an ideal application, when frozen it is distinctly bluer than the red solution
before freezing. This is probably due to internal reflections creating a 'thin-
layer' effect similar to that seen at the meniscus of a glass of young red
wine, which appears bluer than the bulk of the wine.
2.9.4.6 Dry mixes. A variety of acid dessert mixers and drink powders
can be successfully coloured with spray-dried anthocyanin extracts.
2.9.4.7 Other applications. It is also technically possible to colour

alcoholic drinks and products containing vinegar with anthocyanins,
although commercial applications are limited. Red wine would usually be
used in the former example and very few vinegar-based products require
the addition of a red colour. The prime requirement for the successful use
of anthocyanin is a low pH and absence of a cloud; under such conditions
attractive red shades are obtained.
2.10 Beetroot
2.10.1 Source
The red beetroot has been cultivated for many hundreds of years in all
temperate climates. The pigments present are collectively known as
betalains and can be divided into two classes, the red betacyanins and
yellow betaxanthins; both are very water soluble. The betalains have a
limited distribution in the plant world and it would appear that betalains
and anthocyanins are mutually exclusive. Plants producing betalains do not
contain anthocyanins.
Most varieties of beetroot contain the red betacyanin, betanin (Figure
2.5) as the predominant colouring compound and this represents 75 to 90%
of the total colour present. Vulgaxanthin I and II are the principal yellow
betaxanthins.
Beetroot is an excellent source of colour and some varieties contain up
to 200 mg/IOO g fresh weight of betacyanins representing up to 2% of the
soluble solids. Very large quantities, over 200 000 tonnes, of beetroot are
grown annually in Western Europe, most of which is either consumed as
the boiled vegetable or as a conserve [15]. Of this quantity some 20000
tonnes are converted into juice and colour. As a proportion of this is
0·8
.GIUCOse-ox;cr
HO""
1-"-:
h
+ COOH ....
·e:::) 0·.
HOOC~~ H
j
i
.a
c 0.0 I.....:-'~~":""""""~:-'--:'
.00 500 100 70.
Wayelength nm
Figure 2.5 Beetroot (class betalain; synonyms beetroot red, beet red, betanin): structural and
physical characteristics of betanin. Chemical formula: C24H26N2013. Molecular weight:
550.5. Visible spectrum is of beet red as a 0.00075% aqueous solution at pH 5.0. Colour
shade: red to blue-red. Solubility: water-soluble. Absorptivity of betanin: EI% = 1120 in
pH 5 buffer at Am.x> closest to 530 nm.
exported, the quantity of beetroot colour consumed as a food additive is

small compared with that consumed as a constituent of the vegetable.
Another way of looking at this is that when eating 100 g of beetroot the
quantity of betanin consumed is 200 mg, whereas when eating 100 g of
strawberry yoghurt containing beetroot as the added colour the quantity of
betanin consumed is only 0.5 mg.
Beetroots are processed into juice in a manner very similar to that used for
fruit-juice production, using either pressing or diffusion techniques. The
juice is then centrifuged, pasteurized and concentrated to yield a viscous
liquid concentrate with approximately 70% sugar and 0.5% betanin. This
product, the concentrated beetroot juice, is widely used as a food
ingredient. It is possible to produce a colour extract (i.e. to increase the
colour content and so reduce the flavour) by fermenting some of the sugar
to alcohol and removing the alcohol during concentration. The benefits
gained are limited since, for many applications, the juice is a satisfactory
ingredient.
The juice can be spray-dried to a powder although maltodextrin has to
be added as a carrier since the high sucrose content precludes direct drying
of the juice.
Betanin is a particularly intense colour and is stronger than many
synthetic colours (Table 2.5). Thus dose levels for betanin in, for example,
yoghurt are very low at around 5 ppm and for strawberry ice cream around
20 ppm.
Table 2.5 Comparative colour strengths of

betanin and some equivalent synthetic
colours
Colour El%
1 em A max (nm)
Betanin 1120 537

Amaranth 438 523
Carmoisine 545 515
Ponceau 4R 431 505
Table 2.6 Calculation of the betanin content
Commercial beetroot True % betanin

concentrate
'1% betanin' liquid 5.60 0.5

'0.66% betanin' powder 3.60 0.33
It should be mentioned that concentrated beetroot juice is usually

offered for sale with a specification that includes a betanin content. The
figure quoted is usually 1% betanin.
When spray dried to give a powder, the betanin content is lower since
the amount of maltodextrin required is greater than the quantity of water
removed during drying. Betanin figures quoted for beetroot juice powder
are usually in the range 0.4 to 0.7%
The betanin content is calculated by measuring the spectrophotometric
absorption at 535 nm in a 1 cm cuvette and using 550 as the absorptivity of
betanin. However, as mentioned previously, the absorptivity is really 1120
[16] and thus the true betanin content of a standard juice concentrate is
0.5% (Table 2.6). This is an important point to remember when calculating
dye content.
2.10.3 Legislation
Beetroot-juice concentrate is universally permitted as a food ingredient.
Beetroot red is widely permitted in Europe and North America, although
it is not listed by a number of countries such as Egypt, Iceland, India and
Thailand. The EU has proposed [16] a specification for beetroot red of
0.4% betanin minimum (both liquid and powder forms), which would
distinguish the purified colour from the vegetable juice. The specification
prepared by the Joint Food and Agriculture Organization/World Health
Organization Expert Committee on Food Additives (JECFA) [17] is very
similar, with a minimum limit of 1.0% for liquids and 4% for powders and
both calculations use 1120 as the Er/~m for betanin.

When considering the stability of beetroot colour, it is betanin stability that
is being measured. Thus although other pigments are present the measure-
ments of colour strength made are at the wavelength of maximal absorb-
ance of betanin, about 535 nm. Vulgaxanthin is less stable than betanin
and is also present at relatively low levels and absorbs maximally at
477 nm.
2.10.4.1 pH. Beetroot colour shows greatest stability at pH 4.5. At pH

7.0 and above the betanin degrades more rapidly and thus is not
recommended for alkaline applications. Colour shade is not significantly
affected by pH change in the range 3.0 to 7.0. In very acidic conditions, the
shade becomes more blue-violet as the red anionic form is converted to the
violet cation. In alkaline conditions. The colour rapidly becomes yellow-
brown due to the loss of betanin.
2.10.4.2 Heat. Beetroot pigments are susceptible to heat degradation

and this is the most important factor that determines their use as food
colorants. The rate of loss of colour at different temperatures is also
dependent on several factors including pH and water activity.
At high sugar levels, beetroot pigments will withstand pasteurization but
not retorting. When pasteurizing at 60% solids, a colour loss of around 5%
would be expected assuming rapid cooling after the heat process. Gener-
ally, a high temperature for a short period of time is acceptable but it is
most important to cool the product immediately after processing. Often it
is not the heat process itself that causes colour degradation but the heat
input during a slow (unforced) cooling cycle. It is interesting to note that
betanin loss during heating is partially reversible since some betanin is
regenerated from its two degradation products betalmic acid and
cyclodopa-5-0-glycoside. Thus when evaluating beetroot in a food system
it is important to allow the end product to fully equilibrate before
evaluating the final colour shade and strength. This important point relates
to many other natural colours-particularly anthocyanins.
2.10.4.3 Oxygen. Betanin is susceptible to oxidation and loss of colour

may be noticeable in some long-life dairy products. Oxidation is most rapid
in products with high water activity. Some antioxidants such as ascorbic
acid have been shown to be beneficial though ~-hydroxyanisole (BHA)
and ~-hydroxytoluene (BHT) are ineffective.
2.10.4.4 Light. Light does cause degradation of beetroot pigments;

however other factors have a more limiting effect on product applications.
2.10.4.5 Water activity. It is important to remember that water activity

(in this case the amount of free water available for the dissolution of a
colour) significantly affects stability. Beetroot-juice powder stored in dry
conditions is very stable even in the presence of oxygen. Carotenoid
powders under similar conditions will oxidize more rapidly. In aqueous
solution, the higher the solids content, the more stable the colour.
2.10.4.6 Cations. Some di- and trivalent metal ions, particularly iron
and copper, will accelerate betanin oxidation. Sequestration of metal ions
will improve colour stability.
2.10.4.7 Sulphur dioxide (S02). S02 will completely de colorize beet-

root pigments. Thus if a preservative is required in a foodstuff containing
beetroot it should be either benzoate or sorbate.
2.10.5 Applications
The susceptibility of betanin to heat, oxygen and high water activity
restricts its use as a food colorant. Its applications are therefore confined to
products that receive limited heat processing, have either a low water
activity or a short shelf-life and contain no S02. Beetroot pigments are thus
particularly suited to use in powder mixes and most dairy and frozen
products.
2.10.5.1 Ice cream. This is the most important application of beetroot

colour. Most pink ice cream in Germany and the UK contains betanin
either from the use of beetroot juice or beetroot colour. Betanin levels are
usually in the range 15 to 25 ppm (0.3 to 0.5% beetroot juice).
The exact colour shade obtained using a beetroot juice will depend upon
the variety of beetroot used and the extraction and processing conditions
employed. The differences will be evident in the degree of blueness of the
colour. When using a 'blue' beetroot juice it may be necessary to add a
yellow or orange colour to achieve an acceptable strawberry colour. A
water-soluble annatto extract, at a dose level equivalent to 10 ppm
norbixin, would be sufficient.
2.10.5.2 Yoghurt. Again, beetroot is the preferred colouring, and

annatto may also be added to achieve the correct shade. Since the colour
associated with flavoured yoghurt is paler than that of ice cream, the
betanin dose level is lower-usually in the range 4 to 8 ppm. The
microbiological quality of a beetroot colour used in a set yoghurt is
particularly important. Care must be taken to ensure that yeast counts are
extremely low-preferably less than WIg. Beetroot juice is particularly
susceptible to yeast and mould growth because of its high sugar content
and lack of added preservative. Good quality is dependent upon a pro-

duction process designed to ensure any yeasts and moulds are destroyed,
carried out under hygienic conditions, a high (above 66%) sugar content,
storage at below 5°C and careful handling at all times to avoid contamina-
tion.
2.10.5.3 Dry mixes. Beetroot-juice powders are ideal for such appli-
cations because of their excellent solubility characteristics and good
stability. They are commercially used in both instant desserts and soup
mixes. Again, colour shade may be too blue for strawberry and tomato
varieties, and thus a yellow or orange colour may need to be included in
the formulation.
2.10.5.4 Sugar confectionery. The fourth of the commercially important

uses of beetroot extracts is in the coloration of some sugar confectionery
products, principally those products based on sugar pastes having no added
acid. Fondants, sugar strands, sugar coatings and 'cream' fillings are all
suitable applications.
2.10.5.5 Other applications. Beetroot is used to colour some snack

foods. The colour is applied in stripes onto an extruded maize snack
immediately after extrusion, giving the snack the appearance of a strip of
bacon.
Beetroot may also be used in those comminuted meat products with a
low moisture content and which do not contain S02 such as salami-style
sausages. However, the coloration of meat products is not permitted in
several countries. Anthocyanins are also unsuitable because of the pH,
thus carmine is commonly used as a colorant for meat products. The above
comments about end-product applications are all applicable to both
beetroot juices and the further processed beetroot colours. The distinctive
flavour and low pigment strength of the juice may, on occasion, limit its
application. However, since betanin is a very intense pigment its presence
at only 0.5% in a juice concentrate is still sufficient to provide an attractive
colour when the juice is used at relatively low levels.
2.11 Cochineal and carmine
2.11.1 Source
Carmine is the word used to describe the aluminium chelate of carminic
acid. Carminic acid (Figure 2.6) is the colour extracted from the dried
female coccid insect Dactylopius coccus costa (Coccus cacti L.). The word
(a) o·s
OH 0 CH, ·c~
::) 0·4
Glucose@",=COOH
I I 8
HO ~ OH i.a
.5
h
OH 0 .a
io( 00
400 500 700
Wavelength nm
(b)
0·8
..c
:'!:
..
::) 04
u
C
to
..0
.CI
.a
io( 0.0
700
Wavelength nm
Figure 2.6 Cochineal and carmine (class anthroquinone): structural and physical characteris-
tics of carminic acid. Chemical formula: ~2H20013. Molecular weight: 492.4. Visible
spectrum (a) is of carminic acid as a 0.004% solution in pH 7 buffer. Colour shade: orange at
pH 3, red at pH 5.5, purple at pH 7. Solubility: water-soluble. Absorptivity of carminic acid:
El% = 174.7 in 2 N HCI at Amax closest to 494 nm. Visible spectrum (b) is of carmine as a
0.0025% solution in pH 7 buffer.
cochineal is used to describe both the dried insects themselves and also the
colour derived from them.
Coccid insects of many species have been used for thousands of years as
a source of red colour. Each insect is associated with a specific host plant
and each is the source of a particular colour such as Armenian red, kermes,
Polish cochineal, lac dye and American cochineal. The Spanish conquest of
Central and South America brought the latter product to Europe and now
American cochineal is the only cochineal of commercial importance,
although Lac dye (from Laccifera lacca) is available in the Far East.
The current principal source of cochineal insects is Peru, although these
insects are also available in the Canary Islands where they live on various
species of cacti (principally Nopalea cochenillifera). About 300 tonnes of
dried cochineal are produced annually, all of which is used to produce
colour, principally carmine. A significant proportion of the carmine
produced is actually used in the cosmetics industry.
The extraction of colour from the cochineal takes place principally in

Peru, France, UK, USA and Japan.
2.11.2.1 Carminic acid. Carminic acid is very water soluble and its
colour shade in solution is pH dependent. It is orange in acid solution and
violet when alkaline, with a rapid colour change through red as the pH
increases from 5.0 to 7.0. Its colour intensity is relatively low, having an
Enm of only 175; thus the number of commercial applications is limited.
2.11.2.2 Carmine. The ability of carminic acid to complex with metals,

such as aluminium, is used in the manufacture of the more intense pigment
known as carmine. Carmine is a chelate of carminic acid with aluminium
and calcium and is precipitated out of solution by the addition of acid at the
final stage of production. It is soluble in alkaline solutions but insoluble in
acids. Its colour is essentially independent of pH, being red at pH 4.0 and
changing to blue-red at pH 10.0. The intensity of carmine is almost twice
that of carminic acid and it is thus a more cost-effective colour.
2.11.2.3 Commercially available forms. Carminic acid is usually avail-

able as an aqueous solution with a dye content of below 5% and from this
spray-dried powders can be prepared.
Carmine is manufactured as a sparingly water-soluble powder with a
carminic acid content in the range 40 to 60%. This product is used by both
the food and cosmetic industries. Two widely accepted specifications exist
for carmine. One is detailed in the British Pharmaceutical Codex and
equates to approximately 60% carminic acid. This measures carmine
content by its absorbance in dilute ammonia solution. The other is detailed
in the Food Chemicals Codex, 2nd Edition, which specifies a minimum of
50% carminic acid; and also in the more recent Food Chemicals Codex III,
where a revised assay procedure gives a more accurate figure of minimum
42% carminic acid for the same absorbance value. These both measure
carminic acid content following treatment with boiling 2 N hydrochloric
acid. It is still the 50% figure that is widely quoted in the trade.
Due to its insolubility in acid solutions, carmine is usually produced as an
alkaline solution with a carminic acid content in the range 2 to 7%.
Traditionally, ammonia solutions have been used as the solvent since
carmine is particularly soluble in ammonia solutions. However, because of
the unpleasant nature of ammonia, other more acceptable diluents such as
potassium hydroxide are now in use. It is possible to spray-dry these
solutions using mal to dextrin as the carrier to give very water-soluble
powder colours with carminic acid contents in the range 3.5 to 7%.
2.11.3 Legislation
Cochineal and carmine are widely permitted in Europe and North
America. Within the EU they may be used in a wide variety of processed
foodstuffs at specified dose-level maxima varying from 50 to 500 ppm (pure
pigment) in soups and sauces respectively.
Finland and Sweden have both recently extended the use of carmine to a
wider range of foods including sugar confectionery where the maximum
level of use is 200 ppm in Sweden [18] and 100 ppm in Finland.

The usual product used by the food industry is an alkaline solution of
carmine and the following description relates to this solution.
2.11.4.1 pH. Colour shade is fairly constant with changing pH. How-
ever carmine will precipitate out of solution when the product pH is below
3.5. The exact point of precipitation depends on a number of factors
including viscosity, water content and pH.
2.11.4.2 Heat, light and oxygen. Carmine is very stable to heat and light
and is resistant to oxidation.
2.11.4.3 Sulphur dioxide (S02)' S02 does not bleach carmine at levels
usually found in foodstuffs.
2.11.4.4 Cations. Cations will affect colour shade, generally increasing

the blueness of the colour.
Carmine is thus a very stable colour and ideally suited to a variety of

applications.
2.11.5 Applications
The only significant technical limitation on the use of carmine is low pH.
However, carmine is less cost-effective in use than both beetroot and
anthocyanins, principally because it is less intense and hence more needs to
be added in order to obtain the same visual effect. Based on 1990 prices,
beetroot is approximately half the cost in use of anthocyanin extracts and
one third that of carmine when considering the dose levels required to
obtain the same visual colour strength in food. In the past, supplies and
prices of carmine have been adversely affected by a shortage of raw
material. However, during the period 1990-1994 there was a steady
increase in cochineal availability resulting in a gradual reduction in price
and subsequent increase in demand. This position, though, changed in
1994, when a reduced crop volume caused an increase in price which still
persists in 1995.
Historically, carmine was widely used as a textile dye but it has been
totally replaced by synthetic dyes during the past 100 years, principally
because of their lower cost and ready availability. As previously men-
tioned, carmine is used in cosmetics where the insoluble pigment finds
application because of its consistency of colour and its stability.
Other important applications include the coloration of certain alcoholic
aperitifs and use within the food industry where the solubilized colour is
widely used.
2.11.5.1 Meat products. Being somewhat blue-red in shade and stable

in the presence of S02 it is ideally suited for use in sausages and other
comminuted meat products, where inclusion levels are usually in the range
10 to 25 ppm as carminic acid. With the increasing popularity of further
processed poultry products, carmine finds application in ethnic chicken
dishes where a blend of yoghurt and spices coloured with carmine is used as
a marinade for chicken portions. In order to obtain a variety of different
shades other colours, often annatto, are used in combination. Being heat
stable, the colour is retained during the cooking process.
2.11.5.2 Jams and preserves. For this application, beetroot is usually

insufficiently heat stable and anthocyan ins may not be totally effective,
either because of the length of the heat process or the underlying
brownness of the preserve. Under these conditions, carmine is used to give
a bright-red colour with good stability.
2.11.5.3 Gelatin desserts. It is possible to use beetroot for gelatin

products that are designed to have a short shelf-life. Nevertheless, carmine
is the preferred colour for products stored at ambient temperatures for an
extended period, since under these conditions beetroot may oxidize.
Anthocyanin extracts, particularly those derived from grapes, are often
incompatible with gelatin.
2.11.5.4 Flour confectionery. Carmine, being heat stable, is ideally

suited for use in baked goods. At a level equivalent to 40 ppm carminic
acid, carmine will produce a pink-coloured sponge. At a similar dose level,
carmine is also used to produce a pink icing for the decoration of cakes and
biscuits.
2.11.5.5 Dairy products. Beetroot is suitable for most dairy appli-

cations; however, carmine finds application in long-life flavoured milk
because of its resistance to oxidation during storage. For strawberry
flavour varieties a yellow colour would also be added in order to achieve

the correct shade.
2.11.5.6 Soft drinks. Carminic acid is particularly suited to the color-

ation of soft drinks where a crystal clear orange-red colour is required. It is
also interesting to note that after the addition of a clouding agent the
colour shade is still attractive, unlike the effect of cloud on anthocyanin
extracts.
2.12 Curcumin
2.12.1 Source
Curcumin (Figure 2.7) is the principal colour present in the rhizome of the
turmeric plant (Curcuma Zanga). Turmeric has been used as a spice for
many thousands of years and is today still one of the principal ingredients
of curry powder.
Turmeric is cultivated in many tropical countries including India, China,
Pakistan, Haiti and Peru and is usually marketed as the dried rhizome,
which is subsequently milled to a fine powder. This imparts both flavour
and colour to a food product. Several types of turmeric are recognized but
the most important are Madras, Alleppy and West Indian.
Ground turmeric powder is insoluble in water and imparts colour either
by dispersion throughout the food or by dissolution of the curcumin into
Wavelength nm
Figure 2.7 Curcumin (class phenalone: synonyms turmeric yellow, diferoylmethane, natural
yellow 3): structural and physical characteristics of curcumin. Chemical formula: C21H2006.
Molecular weight: 368.4. Visible spectrum is of curcumin as a 0.0005% solution in acetone.
Colour shade: lemon yellow at pH 3, orange at pH 10. Solubility: oil-soluble. Absorptivity of
curcumin: El% = 1607 at 426 nm in ethanol. Curcumin: R, = R2 = OCH 3 . Demethoxycur-
cumin: R, = OCH 3 , R2 = H. Bis-demethoxycurcumin: R, = R2 = H.
vegetable oil. Curcumin, the major pigment present, is accompanied by

smaller quantities of related compounds, and all are insoluble in water.
India is the largest producer with annual quantities usually in the range
of 250 000 to 300 000 tonnes. Most of this is consumed in the ground form
as a spice and only a small amount, 1000 to 1500 tonnes, is converted into
extract.
The quantity of curcumin colour produced annually is very small-pro-
bably of the order of 30 tonnes-when compared with the quantity of
curcumin consumed as a spice. Taking only the Indian production of
250 000 tonnes of turmeric and assuming an average curcumin content of
3% gives an annual consumption of 7500 tonnes of curcumin.

There are three principal types of turmeric extract, namely essential oil of
turmeric, turmeric oleoresin and curcumin.
2.12.2.1 Essential oi/of turmeric. This is the oil obtained by the steam
distillation of turmeric powder and contains all the volatile flavour
components of the spice and none of the colour. It is present in the
turmeric at levels of 3 to 5%. There is only a small commercial demand for
this product.
2.12.2.2 Turmeric oleoresin. This is the most commonly produced

extract and contains the flavour compounds and colour in the same relative
proportion as that present in the spice. It is obtained by solvent extraction
of the ground turmeric, a process identical to that used in the production of
other spice oleoresins.
Spice oleoresins have several advantages over ground spices, principally
their excellent microbiological quality, standardized organoleptic proper-
ties and freedom from contaminants; for these reasons the use of
oleoresins has increased steadily during the past 30 years. Turmeric
oleoresins usually contain 37 to 55% curcumin [19].
2.12.2.3 Curcumin. This is the pure colouring principle and contains

very little of the flavour components of turmeric. It is produced by
crystallization from the oleoresin and has a purity level of around 95%,
which is the standard commercially available quality. The proposed ED
specification for curcumin [16] states that the dye content must be not less
than 90% when measured spectrophotometrically at 426 nm in ethanol.
It is important to note that the distinction between these three products
lies in the ratio of colour to flavour. A spice oleoresin contains the total
sapid, odorous and related characterizing principles normally associated
with the spice. Thus the ratio of flavour components to curcumin in ground
spice and oleoresin is the same. A colour, however, results when an

attempt is made to reduce the flavour of the product and increase the
relative concentration of colour such that the ratio of flavour to colour is
altered in favour of the colour. Thus, in the case of turmeric, the ratio of
curcumin to volatile oil is of the order 50:50, usually lying within the range
40:60 to 60:40. However, curcumin made to the proposed EU specification
has a ratio of 99:1 whereas the spice oleoresin ratio lies within the same
range as that of the ground spice. Perhaps this technique could be applied
to other extracts to distinguish between the total extract and the colour
additive.
2.12.3 Application forms

Pure 95% curcumin is not an ideal product for direct use by the food
industry since it is insoluble in water and has poor solubility in other
solvents. Thus it is usual for curcumin to be converted into a convenient
application form. In many countries, this is achieved by dissolving the
curcumin in a mixture of food-grade solvent and permitted emulsifier. In
this form, the product contains 4 to 10% curcumin and is easily miscible in
water. Polysorbate 80 is the favoured emulsifer/diluent for such products
since it is an ideal carrier for curcumin.
Other forms are also commercially available, including suspensions of
curcumin in vegetable oil and dispersions onto starch but these are not so
commonly used.
2.12.4 Legislation
Turmeric oleoresin is permitted universally as a spice oleoresin. The
Swedish legislation, for example, specifically states that 'turmeric, paprika,
saffron and sandalwood shall not be considered to be colours but primarily
spices, providing that none of their ftavouriI.g components have been
removed' [5]. It is not a permitted colour in the EU.
Curcumin is specifically permitted as a colour in the EU; however many
countries simply list turmeric without a specification for its colour strength.

All the following comments relate to solubilized curcumin dissolved in an
aqueous medium.
2.12.5.1 pH. Curcumin gives a lemon yellow colour in acidic media

with a distinct green shade. As the pH increases, so the green shade
becomes less distinct. In alkaline conditions (above pH 9.0) the colour
becomes distinctly orange.
2.12.5.2 Heat. Curcumin is essentially stable to heat and will withstand

baking.
2.12.5.3 Light. Curcumin is sensitive to light and this factor is the one
that usually limits its application in foods. Curcumin as the suspended
pigment is more stable than the solubilized colour.
2.12.5.4 Cations. In general, cations will tend to induce a more orange-

brown shade.
2.12.5.5 Sulphur dioxide (S02)' S02 reduces the colour intensity of

solubilized curcumin, particularly when present at over 100 ppm.
2.12.6 Applications
Curcumin is an intense colour with a very bright yellow appearance even at

low doses. It is noticeable that the colour easily becomes saturated and at
levels over 20 ppm it is hard to perceive a small increase in colour-dose
level. Thus when using curcumin it is important to determine the minimum
dose level required to give the desired colour. Dose levels are often very
low usually in the range 5 to 20 ppm. Its colour shade is very similar to that
of tartrazine and its use has increased in the USA, following the
requirement to specifically identify the presence of tartrazine.
For many products where an egg yellow colour is required, its shade is
too green and another more orange colour has to be used in combination.
It is usual for annatto to perform this function and numerous blends are
available commercially, particularly for the dairy and flour confectionery
industries.
2.12.6.1 Dairy industry. This is one of the principal applications of

curcumin. Vanilla ice cream is often coloured with a combination of
curcumin and nor bixin and usually contains about 20 ppm curcumin
together with 12 ppm norbixin. In yoghurt, 5 ppm curcumin will give an
acceptable lemon yellow colour. Curcumin colours are generally fairly
viscous products and thus care must be taken to ensure that they are fully
mixed into the milk. A premix made using a high-speed mixer is often a
preferred method to ensure complete homogeneity.
Dairy spreads can successfully use either vegetable oil extracts or the
standard water-miscible form, the curcumin, however, will always migrate
to the oil phase. The only dairy products application where curcumin may
not be suitable is in the coloration of long-life flavoured milks that are
packed in transparent containers. Here, light stability is the limiting factor.
2.12.6.2 Flour confectionery. It is traditional to use a yellow colour in

cakes and biscuits. The required colour is achieved using a blend of
curcumin and annatto similar to that used in ice cream but at a lower dose
level-usually 10 to 15 ppm curcumin together with 5 to 10 ppm norbixin.
Any cut surfaces should be protected from prolonged exposure to light by
suitable packaging [20].
2.12.6.3 Sugar confectionery. Curcumin at 20 ppm will impart a deep,

bright yellow colour to high boilings. It is recommended to use more dilute
propylene glycol based curcumin colours for wrapped confectionery where
polysorbate may not be compatible. Generally, curcumin is acceptable in
all sugar confectionery products not excessively exposed to light.
It is quite common, especially in the sugar confectionery industry, to
prepare stock solutions of colour such that these can be easily measured
out and mixed into the sugar mass. However, curcumin is not particularly
soluble in a concentrated water-based stock solution and it is likely to
crystallize out if left overnight. If stock solutions have to be made, this
should be discussed with the colour supplier so that the solubility limits can
be ascertained.
2.12.6.4 Frozen products. Sorbets and water ice can be successfully

coloured with 5 to 15 ppm curcumin and their low pH does not pose any
problems. Nevertheless, it should be remembered that curcumin solutions
should not, themselves, be frozen since the curcumin may precipitate out
of solution.
2.12.6.5 Dry mixes. Spray-dried curcumin colour using gum acacia as

the carrier is ideal for use in dessert mixes and instant puddings. Such
products usually contain up to 8% curcumin. When using powders based
on gum acacia it is necessary to use warm water when reconstituting the
product in order to fully solubilize the gum and so release the colour.
Alternatively, dispersed starch based colours with 1% curcumin or less
may be used. The lower the curcumin content the less susceptible the
colour to specking. When a concentrated colour is present at a low level in
a dry mix and water is added, each individual particle of colour slowly
dissolves creating a temporary speck of deep colour that is very noticeable.
However, if a powder colour of low tinctorial strength is used at a relatively
high level in a dry mix, and if that colour is bound to the starch, then
coloration is by mass effect and there are consequently no specks of bright
colour when the water is added. It should also be pointed out that powders
using a spray-dried 5% curcumin would be white in appearance with a few
yellow specks visible whereas the second product, using a weaker dispersed
colour would appear an even yellow colour.
2.12.6.6 Savoury products. If turmeric flavour is an acceptable part of a

seasoning then turmeric oleoresin can be used as a spice to give both colour
and flavour to savoury products. This would be true of some spiced chicken
dishes or soups. There are many applications of turmeric oleoresin but the
use of such products is outside the scope of this chapter. However, it
should be remembered that turmeric oleoresin is widely used in the food
industry and thus curcumin is present in many savoury foods.
2.13 Other colours
The five natural colours previously discussed are those most widely used by
the food industry in Europe and North America. They are available in
substantial quantities and at an affordable price. While carmine is more
expensive in use than beetroot, this disadvantage is offset by its excellent
stability. Several other natural colours are permitted but for reasons of
price, availability or colour performance these are not so widely used.
2.13.1 Chlorophyll
Chlorophyll is a vitally important pigment in nature and is present in all
plants capable of photosynthesis. In the food industry, much effort is put
into retaining the chlorophyll naturally present in the green vegetables (see
chapter 5). However, the addition of chlorophyll as a colour to foodstuffs is
very limited, principally because of its poor stability.
Chlorophyll is an oil-soluble colour that can be extracted from a range of
green leaves, but usually grass, nettles or alfalfa is used. Chlorophyll
degrades easily, particularly in acidic conditions, losing its magnesium ion
to yield phaeophytin, which is yellow-brown in colour. Chlorophyll colours
tend to be rather dull in appearance and of an olive green-brown colour.
This obviously limits their food applications. Chlorophyll extracts can be
standardized using vegetable oil for oil-soluble products or blended with a
food solvent or permitted emulsifier to give a water-miscible form. An
extract of chlorophyll will contain approximately 10% chlorophyll together
with other colouring compounds, principally lutein and carotenes as well as
fats, waxes and phospholipids. Raw material quality and the extraction
techniques used will affect the ratio of chlorophyll to carotenoids and
hence the colour shade of the extract.
Water-miscible forms can be used in sugar confectionery, flavoured
yoghurts and ice cream if the colour shade obtained is acceptable.
However, most chlorophyll extracted is used in cosmetics and toiletries,
with only very small quantities in foods. In the USA chlorophyll is not a
permitted food ingredient, whilst in the UK consumption of chlorophyll as
an added colour in food is probably less than 400 kg per year.
2.13.2 Copper complexes of chlorophylls and chlorophyllins

The replacement of the central magnesium ion with copper produces a
more stable complex with greater tinctorial strength. The subsequent
removal of the phytol chain by hydrolysis with dilute alkali renders a water-
soluble product termed copper chlorophyllin. This particular complex, in
the form of its sodium or potassium salt, is the most widely used green
colour of natural origin and it is usually manufactured from either grass or
alfalfa. One of the purification steps involves the precipitation of the
chlorophyllin, which ensures the elimination of the yellow carotenoids.
Copper chlorophyllin is permitted widely as a food colour throughout
Europe, although its use in the USA is limited to dentifrice. Copper
chlorophylls and copper chlorophyllins are chemically modified natural
extracts and cannot truly be termed natural.
2.13.2.1 Application forms. Copper chlorophyll extracts are oil-soluble

viscous pastes often standardized with vegetable oil. Dye contents are of
the order of 5 to 10%. There is no internationally recognized method of
chlorophyll content estimation though White et al. [21] have reported a
method for the determination of chlorophylls and copper chlorophylls,
which, although difficult mathematically, seems to give reproducible
results.
In contrast, salts of copper chlorophyllin are available in both liquid and
powder forms. There are many grades of powder product available with
dye contents varying from 10 to 100%. Colour content is determined by a
widely used method involving measuring the absorption at 405 nm in a pH
7.5 buffer and calculating the dye content using 565 as the absorptivity of
sodium copper chlorophyllin. However, it is now realized that the true
absorbancy is actually higher than 565 but no new figure has been agreed.
The powders will dissolve in water to give an alkaline solution but special
application forms are required if the powder needs to dissolve in acid
solution.
The other widely used forms are alkaline solutions containing up to 10%
chlorophyllin, the strength of solution being related to the intended use of
the colour. In addition, some special application forms that are both
soluble and stable in acid conditions are available.
2.13.2.2 Principal factors affecting stability. So far as pH is concerned,

all chlorophylls are most stable under alkaline conditions. In dilute acids,
chlorophylls will hydrolyse and lose colour rapidly and copper chlorophyl-
lins will precipitate. In addition, copper chlorophyllins are relatively heat
stable but will lose colour on exposure to light.
2.13.2.3 Applications. Green is not a widely used colour in the food

industry and hence applications are limited principally to lime flavour sugar
confectionery and pistachio flavour ice cream. Dose levels for copper
chlorophyllin lie between 30 and 50 ppm for clear sugar confectionery
products and 50 to 100 ppm for ice cream. In most foodstuffs, copper
chlorophyllin gives a mint green (blue-green) shade, therefore it is
necessary to add a yellow colour to achieve the required shade of yellow-
green. The addition of an orange colour would tend to give a brown cast to
the colour therefore a yellow, preferably a green-yellow, is necessary.
Curcumin meets this requirement perfectly and is often used in combi-
nation with copper chlorophyllin at a dose level approximately half that of
the copper chlorophyllin.
Other applications include cucumber relishes, dessert mixes and some
cheese either to give a green vein or at very low levels (less than 1 ppm), to
whiten a soft cheese.
2.13.3 Carotenoids
Carotenoids are very widely distributed in nature and it has been estimated
that nature produces some 3.5 tonnes of carotenoids every second [22].
Over 400 different carotenoids have been identified and many of these are
present in our diet (see chapter 7). Lutein is found in all green leaves and
l3-carotene is an essential source of vitamin A. Several of these carotenoids
can be extracted and used as colorants. Bixin has already been described
but lutein, l3-carotene, paprika extracts and crocin are also available
commercially.
2.13.3.1 Lutein. Lutein is an oil-soluble colour either produced as a by-

product of chlorophyll extraction or extracted from marigolds. Its principal
use is in the fortification of the xanthophyll content of poultry feed and its
food applications are very limited. Annatto and turmeric are the most cost-
effective source of natural yellow-orange colours and lutein only finds
application where the light sensitivity of turmeric limits its use. Lutein is
used commercially in some lemon-flavoured cloudy soft drinks, selected
sugar confectionery products and emulsified salad dressings. Consumption
in Europe as an added colorant to foodstuffs, is estimated at less than
1000 kg per year.
A concentrated extract contains 5 to 12% lutein and this extract can be
solubilized in citrus or vegetable oils for use in the above mentioned
products.
2.13.3.2 I3-Carotene. I3-Carotene is an oil-soluble colour that can be

extracted from a number of sources including algae, carrots and palm oil.
However, a large proportion of the l3-carotene used as a colorant in
foodstuffs is the nature-identical form produced by synthesis. Naturally
extracted l3-carotene is significantly more expensive in use and is mainly
used in the pharmaceutical/health food industry as a dietary supplement.
Special application forms of nature-identical l3-carotene have found ex-

tensive use in the soft drink and dairy industries.
Natural l3-carotene is available as 20 to 30% suspensions, solutions in
vegetable oil and as emulsified, water-miscible forms. Spray-dried powders
can also be manufactured. Consumption of the natural form in Europe is
increasing as supplies become more readily available and is estimated to be
more than 2000 kg per annum of pure pigment.
2.13.3.3 Paprika. The total extract of paprika (Capsicum annuum) ,

containing all the flavour and colour components present in the spice, is
widely used as a spice oleoresin. Several carotenids, principally capsanthin,
capsorubin and l3-carotene, are responsible for its orange-red colour. It
also has a distinctive flavour, which is responsible for the characteristic
taste of such products as goulash and chorizo. Its use as a colour is very
much limited by its flavour and there is no significant production of the
purified flavour-free colour.
The extract is available as oil-soluble solutions standardized with
vegetable oil to a colour strength that is quoted in multiples of 10 000 with
40 000, 60 000, 80 000 and 100 000 being the most commonly available. A
100 000 colour-strength extract equates to approximately 10% caroten-
oids.
Water-miscible forms are usually prepared by incorporating polysorb-
ates during production, although emulsions can also be formed by using
gum acacia. Of the 400 tonnes of spice extract manufactured annually in
Europe, most is used as a spice in meat products, soups and sauces. Very
little is used as a colour in sweet products because of the flavour carry-over,
although occasional use in sugar confectionery and gelatin desserts is
encountered.
2.13.3.4 Crocin. Crocin is the water-soluble yellow carotenoid found in

saffron and in gardenia fruits. It is responsible for the yellow colour of
paella and saffron rice. It is not currently permitted by the EU as a food
colour but it finds extensive use in the Far East. The use of saffron as a
source of crocin is limited by its distinctive flavour and high cost.
2.13.4 Caramel
Caramel colour is, in volume terms, the most widely used colour in the
food industry. It is totally derived from sugar and most is modified
chemically by the use of ammonia, ammonium salts and sulphites. The
sugary, aromatic product obtained from heating sugar solutions without
the addition of any other substance is caramel syrup, sometimes called
burnt-sugar syrup. This material falls outside the scope of the colour
directives. Creme caramel is a good example of a natural caramel product.
Caramels are very stable colours but usually carry an electrical charge
and thus may cause precipitation in the presence of oppositely charged
molecules.
2.13.5 Carbon black

This is derived from vegetable material, usually peat, by complete
combustion to the insoluble carbon. It is widely used in Europe in sugar
confectionery but is not a permitted colour in the USA. The powder colour
has a very fine particle size, usually less than 5 /J-, and is consequently very
difficult to handle. It is therefore usual for carbon black to be sold as a
viscous paste where the carbon is suspended in glucose syrup. Carbon
black is a very stable pigment.
2.14 Conclusions
Colour is essential to the full enjoyment of our food. Our diet contains a
wide range of colours that are naturally present in the food we eat.
Chlorophylls, carotenoids and anthocyanins are consumed virtually every
day.
In order to ensure that our processed food is attractive, it is often
necessary, for reasons stated previously, to enhance their appearance by
the addition of colour. It is common practice, especially in Europe, to use
natural colours for this purpose and five product groups in particular-an-
natto, anthocyanins, beetroot, cochineal and curcumin---dominate the
natural colour industry. These five groups represent over 90% of the
market for natural colour extracts. It may seem that the choice of natural
colour is limited to just a handful of products. However, it should be
remembered that each colour is available in numerous forms and it is the
choice of the correct application form that requires careful consideration.
In the future, it is unlikely that new colours will be added to the
permitted lists since the cost of the necessary toxicological testing is
extremely high. Future development will be aimed towards the production
of more stable forms of the currently permitted colours. If you consider
that natural colours are quite stable to direct sunlight for many days when
present in the plant itself, the researcher has the comfort of knowing that
colour stability is possible.
As the number of permitted artificial colours gradually reduces and as
consumers express a preference for products of a natural origin, so the
range of natural colours available to the food industry has increased. This,
in tum, has led to an increased awareness of their many attributes and how
best to exploit these qualities to the advantage of both the food manufac-
turer and the consumer.
References
1. DuBose, C.N., Cardello, A.V. and Maller, O.J. Food Science 4S (1980) 1393--1399.
2. Johnson, J., Dzendolet, E. and Clydesdale, F.M. J. Food Protect. 46 (1983) 21-25.
3. Food Advisory Committee, FdACIREP/4, HMSO, London (1987).
4. Ames, B.N. Science 221 (1983) 1256.
5. Swedish Food Regulations, Food Additives (1985) SLV FS 1985:1.
6. Italian Official Gazette 28 (1968) 159.
7. Official Journal of the Commission of the European Communities, L23737, 10.9.94.
8. Kuhnau, J. World Rev. Nutr. Diet 24 (1976) 117.
9. Timberlake, C.F. and Bridle, P. 'Anthocyanins' in Developments in Food Colours,
Vol. 1 (J. Walford ed.), Applied Science Publishers, London (1980) pp. 115-149.
10. Timberlake, C.F. Food Chern. S (1980) 69-80.
11. Timberlake, C.F. and Henry, B.S. 'Anthocyanins as natural food colorants' in Plant
Flavonoids in Biology and Medicine Ill, A.R. Liss Publishers, New York (1988) pp. 107-
121.
12. Knuthsen, P. Lebensm.-Unters. Forsch. 184 (1987) 204-209.
13. La Bell, F. Food Processing (USA), April (1990) 69-72.
14. Jiang, J., Paterson, A. and Piggot, J.R. Int. J. of Food Science Technol. 2S (1990) 59&-
600.
15. Verniers, C. NATCOL Quart. Inform. Bull. No.3 (1987) 4-10.
16. European Commission Document III/5218/94-Rev. 4, April 1995.
17. Food and Agriculture Organization, Food and Nutrition Paper 3111 (1990).
18. Swedish Food Regulations, Food Additives (1989), SLV FS 1989. 31.
19. Food and Agriculture Organization, Food and Nutrition Paper 49 (1984).
20. Knewstubb, C.J. and Henry, B.S. Food Technology International (Europe) (1988)
179-186.
21. White, R.c., Jones, I.D., Gibbs, E. and Butler, L.S. J. Agricul. Food Chern. 2S (1977)
143.
22. Weedon, B.C.L., in Carotenoids (0. Isler ed.), Birkliuser Verlag, Basel (1971).
3 Biotechnology in natural food colours:
The role of bioprocessing
M.C. O'CALLAGHAN
3.1 Summary
There has been much interest over the past decade in the use of
biotechnology for the production of food additives. The demand for
naturally obtained food additives would not in itself guarantee the
employment of biotechnology in their production. It would have to
compete with traditional agricultural sources of supply, which often have
the advantages of being cheap and well established. The factor which could
tip the balance in favour of biotechnology is concern in the food industry
over the reliability of supply of agricultural sources of colours which can be
subject to the vagaries of political instability and climate. This has caused
the industry to look more favourably on biotechnology as a source of
supply, which because of the high costs implicit in the necessary research
and development might not otherwise seem so attractive. A working
definition of biotechnology in the context of natural colours is provided.
There are three main areas of biotechnology which it is hoped will produce
the desired colours: plant cell tissue culture, microbial fermentation and
enzyme and gene manipulation technology, all of which have been found
to have advantages and disadvantages. The uncertain policies of regulatory
agencies towards biotechnology are a major stumbling block. In spite of
these difficulties the literature highlights a few successes.
3.2 Introduction
Biotechnology is of special interest to the colours industry because of:

• the consumer's demand for higher quality and naturalness
• the need to develop more efficient processes with fewer environmental
problems
• the fact that it may enable the mass production of important colouring
compounds at relatively low cost [1]
Traditional breeding techniques and mutagenesis have been used to
achieve increased yields in plants and microorganisms. However since it is

BIOTECHNOLOGY IN NATURAL FOOD COLOURS 81
difficult to select for specific molecules, the extent to which breeding or

mutagenesis may be used to enhance the level of anyone compound is
limited and now that the previous difficulties associated with transferring
genes from one organism to another are rapidly fading, this is not the
method of choice.
Conceptually, the production of colour compounds and other natural
products using biotechnology seems straightforward. The strategy that
seems most obvious is to amplify the rate-limiting steps of appropriate
biosynthetic pathways. In addition to amplification, it may be possibJe to
eliminate points of negative regulation within the pathways or competing
metabolic pathways. In reality, targeting and amplifying the enzymes
intrinsic to the specialised metabolic pathways of natural product forma-
tion is no small task. A detailed biochemical knowledge of colour
biogenesis pathways is essential for pinpointing the molecular targets for
cloning and, in many cases, our knowledge of these metabolic pathways
remains obscure.
The pharmaceutical industry has used biotechnology to build many new
profitable products and processes on the foundation of basic research
which it has funded over the last 45 years. Such a commitment to basic
research has been lacking in the food industry. The challenge to the colours
industry, then, lies in its willingness and ability to conduct and fund
research to further understand the molecular basis of food colour and
appearance. This effort is a prerequisite to the successful application of
modern biotechnology's tools to food colour systems.
Biotechnology plays a role, and a large one, in most types of naturally
derived colours, but chemical (hydrogenation, preservation) and physical
(separation, heat treatment, drying, blending, preservation) technologies
dominate. It is likely that the role of biotechnology will increase, although
not necessarily in a revolutionary way. Arguably, biotechnology will have
the greatest impact on the colours industry through changes in raw
materials, that is to say plant products. The main targets are crop
protection which will increase yields and bring down costs, and crop
modification which allows the properties of plant materials to be adapted
to the demands of the consumer and the colours processor. Application of
new biotechnology to the agricultural aspects of naturally derived colours
is developing rapidly. Most countries are involved and The Netherlands
has large seed companies, which are in the forefront of biotechnology in
this area.
There has been a lot of acquisition and merger activity in the colours
industry since the mid-1980s displaying a clear trend towards international-
isation and concentration. Biotechnology has been cited as one reason for
these mergers. The second reason is that non-traditional colour companies
are becoming involved in the production of colours, coming from a base in
fermentation of chemicals. Examples are Gist-brocades and Nestle. They
continue with their successful efforts to improve the yield of their key
products while at the same time they are trying to find new uses for their
workhorse microorganisms by introducing foreign genes, which they
sometimes acquire from small R&D boutiques. The organisms experi-
mented with or used most widely in the colours industry are moulds, yeasts
and algae. Colours papers and patents in these active research areas
account for the majority of all recent colours biotechnology publications.
3.3 Definitions
What is biotechnology? In short it is the synthesis or transformation of

product by microorganisms, plant or animal cells or enzymes from living
cells, within the framework of technical processes in industrial production.
Biotechnology can be broadly divided into traditional and modern
biotechnologies. Traditional or classical biotechnology encompasses con-
tinuous processing, fermentation and enzymatic processing. Genetic en-
gineering and plant biotechnology are considered the new or modern
biotechnologies. Recombinant DNA, hybridoma technology, protoplast
fusion, cell tissue culture have all contributed to the new elan in
biotechnology. However, the new biotechnology has led to an upsurge in
classical biotechnology and the colours industry was among those indus-
tries that were quick to increase efforts in this area.
Fermentation or microbial transformation is the primary tool of biotech-
nology. In this case living microorganisms are used to exploit their enzyme
systems to perform a special chemical conversion to produce a specific
metabolite, e.g. a colour pigment. Fermentation development activities
must optimise both cell growth and the expression of the desired metabol-
ite. Biotechnological processes based on microorganisms are attractive
because they duplicate their biomass so quickly (between 20 minutes and
2 hours) and thus produce cost-effective metabolites.
Plants, however, grow much more slowly (several hundreds times
slower) but they are often able to produce more complicated metabolites
than microorganisms. Plant biotechnology applications relevant to colours
can be divided into two categories: (i) improvement of plant strains that
produce a lot of the desired secondary metabolite; and (ii) the production
of secondary metabolites by cells or organs in vitro, plant tissue culture and
the newer hairy root culture technology. A plant tissue culture is a way of
growing plant cells outside the plant organism. The culture cells can, when
stressed or placed in the correct environment, produce high value bio-
chemicals like flavours and pigments-so-called secondary metabolites. A
hairy root culture is a plant tissue culture consisting of root cells infested
with Agrobacterium rhizogenes. The 'hairy root' disease results in a
proliferation of fast growing roots from the infection site. The roots can be
separated from the plant and after removal of the bacteria, the root cells
will develop into a culture which grows much faster than the normal
untransformed root culture [2].
3.4 Plant biotechnology
An inherent problem of using plants as reliable sources of colours/pigments

involves seasonal and geographic variations [3]. Climatic, political and pest
influences can lead to considerable fluctuations in supply and price.
Biotechnological processes are considered attractive because of the poten-
tial to produce these compounds directly, compounds which otherwise
would have to be synthesised through many (often costly) chemical steps
[4]. Furthermore the possibility of avoiding use of pesticides/herbicides and
reducing the amount of fertiliser could be an argument for developing
industrial scale production of plant pigments in reactors.
3.4.1 Genetic engineering

Genetic engineering of plants has been focused mainly on improved
properties for agriculture including: pest resistance, virus infection resist-
ance, shelf-life, colour and texture. It is a long and expensive way to
modify a plant into a more attractive variety, and it needs significant sales
to justify the investment. Therefore, application of genetic engineering for
pigment production is still an embryonic technology. On top of that there is
the disadvantage of regulatory approval needed for the use of genetically
modified strains in production.
3.4.2 Plant tissue culture

Plant tissue culture is the rapid propagation of selected plant cells
containing expensive metabolites which are in high demand. Plant tissue
(leaf, stem, root) is selected and brought into a medium with selected
nutrients and plant hormones to induce 'callus tissue'. The isolated callus is
further grown as separate cells in a liquid culture medium. When enough
material is available one can start a 'suspension culture' on a large scale [5].
By adding precursor, the cells can be used to perform bio-conversions or
directly as producers of a colour pigment. Although much research is being
done with plant tissue culture in colours there is still only one large scale-
up, notably the commercial production of the dye Shikonin by Mitsui
Petrochemical Industries, Japan [6].
There have been many attempts to produce anthocyanins from carrot
cell cultures. Dougall [7 a] examined the production of anthocyanin from
cell suspension cultures of wild carrot and found that the pigment
accumulation ability of the cells changed reversibly during culture. Kin-

nersley and Dougall [7b] showed that anthocyanin accumulation depended
on the size of the carrot cell aggregates; small aggregates were higher
yielding than large, and it was discovered that if the cultures were screened
for the small aggregates, high yielding cell lines could be firmly established.
Ozeki et al. [8] also examined the induction of anthocyanin synthesis in a
carrot suspension culture and found that synthesis could be triggered by
transferring the cells from a medium containing 2,4-D to one without.
Hinderer et al. [9] found that the presence of exogenous gibberellic acid
blocks anthocyanin synthesis in carrot cell cultures, and Ozeki et al. [10]
discovered that anthocyanin yields in carrot cell cultures decreased at high
inoculum densities, but that this problem could be offset by the addition of
sucrose. Attempts have also been made to produce anthocyanin pigments
from tissue cultures of plants other than carrot. These include that of
Yamakawa et al. [11] who grew grape callus tissues which were capable of
producing anthocyanin in the dark in both static and suspension cultures,
despite limited success in establishing stable cell lines. Also Zryd et al. [12]
examined the biosynthesis of the pigment, betanin, in red beet cell
suspension cultures, in order to try to understand how the biosynthesis
might be switched on and off in these cells.
Attempts have been made in Japan to develop a micropropagation
method for shoot regeneration from callus in tissue culture media of
saffron corms [13-16]. Work has been on-going in Brazil and South Africa
on the in vitro tissue culture of annatto (Bixa orellana). Organogenic (both
direct and indirect) and embryogenic approaches have been explored for
the successful micropropagation of annatto [17, 18].
Plant tissue culture can be used to:
• produce a metabolite (colour pigment) from the cells
• do a bio-transformation
• utilise the whole biomass for further processing
The principal colour pigments (plant metabolites), the plant species of
origin and the organ and type of culture used in the main areas of research
to date are outlined in Table 3.1 [2].
It has been found, however, that many secondary metabolites are
produced in the root cells or accumulated in the root. Many of these
metabolites can only be produced in suspension cultures if root cells are
developed in the culture [19].
3.4.2.1 Development of plant tissue culture. When the first plant tissue
cultures were initiated in the 1930s in order to study the differentiation of
cells in the plants, root tips were cultured in a medium consisting of
nutrients, sugar and vitamins [20]. The root tips underwent elongation and
branching and became the first model system used for studying the
Table 3.1 Examples of secondary metabolites produced in vitro in plant cell cultures
Plant metabolite Plant species Organ Type of culture
Betalain Beta vulgaris Root Root culture

Anthocyanin Vitis vinifera L. Fruits and leaves Cell suspension
Daucus carota Callus (A. reptans)
Ajuga reptans
Shikonin Lithospermum Root Cell suspension
erythrorhizon
Capsaicin Capsicum fructescens Fruits Cell suspension
Bixin Bixa orellana Seed coat Callus
Crocin Crocus sativus L. Ovaries Callus
behaviour of plant cells and the production of secondary metabolites in the

plant. However, the root cultures were slow in growing and in the 1950s
they were displaced by the rapidly growing so-called plant callus and plant
tissue suspension cultures as described above.
In the late 1970s and early 1980s the introduction of special media and
conditions stressing the plant cells led to stimulation of the accumulation of
the secondary metabolites. Moreover, more control of the process was
obtained by introducing special techniques like:
• two stage fermentation
• cell cloning
• elicitation
• immobilised plant cells
These techniques are described briefly.
3.4.2.1.1 Two-stage fermentation. Secondary metabolites are pro-

duced in highest quantity when growth of the cell mass is stopped
(stationary phase). This has led to two-stage fermentations where biomass
production is the goal for the first stage. When sufficient biomass has been
obtained in the broth, secondary metabolite production is increased by
adding or applying the necessary trigger material, which may not be the
same materials necessary for optimum growth of biomass [21].
3.4.2.1.2 Cell cloning. A clone is a group of organisms or plants

produced non-sexually from one ancestor and the tissue culture itself is a
clone. The target of cloning is to generate cells which produce more of the
required metabolite than an equilibrium mixture of the cells. Cells are
selected from the starting clone and grown individually to develop
subclones, which are normally identical to the original clone. Sometimes,
however, the clone does not consist of totally identical cells since
differences may occur from spontaneous mutation or other changes.
Subclones are chemically tested or visually inspected for production of
desired metabolites and yield. High yielding strains can be selected by
virtue of the analytical results. Cells producing colour pigments are
normally selected visually. Eichenbeger (1951) visually selected carrot
callus with high levels of J3-carotene and his strain was maintained for more
than 10 years [20]. Strains of Lithospermum erythrorhizon, which produces
shikonin, have been selected by cloning methods and these strains are now
used to produce shikonin commercially in Japan. Clones tend to revert to
an equilibrium mixture which produces lower amounts of pigment. This
problem is common during scale-up to large batch fermentations. More
work must be done to develop methods of maintaining the high yielding
cells in the culture especially in industrial scale production.
3.4.2.1.3 Elicitation. Elicitation is a technique of growing the plant

culture together with a microbial culture in order to promote production of
secondary metabolites. Normally plant cultures are grown under sterile
conditions. However, it has been found that many plants produce valuable
chemical compounds in order to defend the plant against microbial attack.
Production of these compounds can be induced by adding microorganisms
to the cell culture. So far no description of elicitors in connection with
colour production has been reported.
3.4.2.1.4 1mmobilised plant cells. Immobilisation means entrapping

plant cells in some kind of gel (or algal beads) or combination of gels which
are allowed to polymerise around them. This technique has been used for
immobilisation of microbial cells for years but has only recently been used
for plant cells. Immobilisation can also be achieved by use of blocks of
polyurethane foam or hollow fibre membranes in which the cells are
distributed.
The advantage of immobilisation is that it is possible to produce a culture
with high density in which the metabolites are synthesised and then
released to the medium [5). Yeast and microbial cultures are easier to work
with because the microbe and the yeast itself are the 'original organism'. A
plant cell culture is a much more complex collection of cells from the plant.
In a multicellular plant, cells are differentiated and each cell has its own
function. A collection of cultured plant cells cannot therefore be compared
with a microbial cell culture. Problems associated with scaling up in large
reactors are many and various. Immobilisation techniques are interesting
and promising ways to make cell cultures more stable and make it possible
to make continuous production of the secondary metabolites also in large
scale. Table 3.2 summarises some of the advantages obtained by immobil-
isation techniques.
Table 3.2 Some advantages of plant cell immobilisation (after Fontanel and Taboo 1987;
Knoll 1986)
Advantage Influence on cell culture
Protection of cells by matrix Simplification of scale up

Protection from physical damage
Maintenance of stable and active Elimination of non-productive growth
biocatalyst phase
Reduced risk of declining productivity from
a selected cell line
Continuous process possible Continuous removal of metabolic inhibitors
Improved process control and product
recovery
Increased productivity Increased cell densities
Elimination of growth phase
Table 3.3 Secondary metabolites produced or transformed

by root cultures
Compound Plant source Use
Betalains Beta vulgaris Food colour

Shikonin Lithospermum
erythrorhizon Anti-bacterial
Dye
Anthocyanins Daucus carota Food colour
3.4.2.2 Limitations of cell suspension cultures. The current interest is in

root cultures, because of the limitations in cell and tissue cultures. As
previously shown it is necessary to develop two stage cultures in order to
accumulate significant quantities of a secondary metabolite. Changes in
culture performance are common in suspension cultures and mean that the
cell suspensions must be carefully controlled and screened regularly for
their desirable characteristics. Because of these difficulties in cultures of
'undifferentiated cells' interest in organ cultures has resurfaced in recent
years. The detailed metabolism of the root is not very well understood and
work is in progress to establish hairy root cultures of different plant genera
and to determine the biosynthetic pathways of the secondary metabolites.
Table 3.3 shows the colour compounds produced by normal or hairy root
cultures [19].
Most of the currently reported work on tissue cultures and root cultures
for pigments concentrates on anthocyanin, annatto and red beet. Annatto
and anthocyanins are produced by Bixa orellana and various plants and
flowers respectively and most of the reported work is performed in callus
or cell cultures. Red beet can be produced by hairy root cultures. Recently
work in Korea on the production of anthocyan ins by Daucus carota hairy

root cell culture in vitro has also been reported. The use of fungal elicitors
and specific aeration techniques for scale-up in air-lift bioreactors show
promise [22, 23]. In this review anthocyanin pigments are discussed as an
example of a cell suspension culture and red beet pigments as an example
of a hairy root culture. It was judged most interesting to review only those
articles describing work done on cultures where the explant is an edible
plant such as grape, carrots or red beets. Authorisation to use pigments
produced from plant cells is likely to proceed more easily if the cell culture
originates from an edible plant.
3.4.2.3 Anthocyanin cell suspension cultures. Anthocyanins have been

produced in pilot scale suspension of grapeskin or in a mutated carrot cell
line. A plant cell culture of grapeskin was established in 1978 in Toulouse,
France [24] with pulp fragments from young fruits as the explant.
Suspension cultures are now maintained by weekly transfers to the growth
medium from selected calli. The obvious advantage of working with
pigments is that high producing cells can be selected visually. In the
suspension cultures reported here approximately 50% of the culture
consisted of pigmented cells [20].
Several reports describe attempts to optimise pigment production once
the cell culture has been established [25-29]. 20 I to 150 I fermentor tanks
are used and it seems as if sucrose in combination with other types of sugar
is used to promote anthocyanin formation after reaching a stationary phase
of the culture. Just growing the cells does not promote anthocyanin
growth. Enhancement was obtained by stressing the plant cells with excess
amounts of sucrose in the medium. It took 6 to 8 days to establish adequate
amounts of pigmented cells in the medium. Maximum concentration was
reached after 11 days of culturing.
Until now the literature has reported only optimisation of the cell growth
phase and the production of metabolites and the relative cost of different
nutrients. The carbon source is the most expensive part of the medium.
Combinations of different sugars have been tested for efficiency and at the
same time low cost. Galactose in combination with sucrose seems to be the
best combination [28].
No attempts have so far been reported on the extractions from the cells,
and there is no price of a full process including extraction and concen-
tration of the broth. A problem associated with the cell cultures is that cells
often return to a state where they produce none or little of the secondary
metabolite. Furthermore the anthocyanin pigments in the cell culture are
often difficult to extract since the pigments are formed inside the cells in
vacuoles. Work is being done to find a way to extract the pigments without
damaging the cells. No description of a more commercial production of
anthocyanins is available. It shows that there is still a long way to go before
commercial production is viable--or even before a price of such a product

can be estimated.
3.4.2.4 Hairy root cultures of red beet pigments. Some work done at
Osaka University in Japan on red beet hairy root cultures is interesting
because it shows that the red beet pigments can be released from the cells
without damaging the cells and that the cell culture can continue producing
pigment after exchange of the medium with fresh medium [30, 31]. Parts of
red beet plants, which were grown freely for 2 months were used for
preparation of the hairy root culture. The culture was infested with
Agrobacterium rhizogenes to promote formation of the hairy roots. The
culture of hairy roots was grown in the dark, either in flasks placed in a
rotary shaker or in a jar fermentor.
It was discovered spectrophotometrically that the pigments produced
were betanin and vulgaxanthin. Colour hue of the pigments was similar to
the hue of pigments obtained from a normal red beet plant. Two smaller
peaks were found in the HPLC chromatogram-these still need to be
identified. Furthermore it was discovered that betanin and vulgaxanthin
could be released into the broth when the culture was not being shaken.
The reason for this was later established to be the drop in the concentration
of dissolved oxygen during the period of the shaking. Shaking the culture
was stopped for a period of time every 10 days, the pigmented broth was
replaced by fresh medium and the culture was maintained for more than
30 days in total. This finding from the Osaka laboratory working on
cultures of red beet hairy root in bioreactors, led to the recovery of pigment
released from the cells by repeated treatment of oxygen starvation [32].
The pigments were not contaminated with extraction solvents or other
additives. In addition it was found that the concentration of betanin and
vulgaxanthin in the hairy root cells was at the same level as in the normal
red beet. More than 30 different secondary metabolites have been reported
in cell suspension cultures in quantities equal to or even higher than in the
corresponding plant. With the hairy root techniques more metabolites are
likely to follow in the near future. Of these secondary metabolites
anthocyanin and red beet are the food pigments being investigated. Other
pigments such as blue anthocyanin precursors and less well known plant
pigments have also been produced [14, 15, 33]. Some other interesting cell
lines are also being investigated as alternatives for tissue culturing of
anthocyanins and betacyanin [34].
3.4.2.5 Economic factors. Interest in using plant cell or tissue culture

did not appear until the success in 1983 of plant culture production of the
dye and antibacterial agent, shikonin [35]. Today only the pigment
shikonin is produced commercially in a two-stage fermentation in Japan.
Shikonin is used as a colorant in lipsticks. The reason for its success is that
this pigment is relatively easily released from the surface of the cells and
does not create major problems for extraction. Furthermore the price of
shikonin is high (£3000/kg; US$2050/lb )-a fact that makes the process
attractive.
It is stated in the literature [4] that the lowest price at which production
by suspension cultures is viable is £300/kg (US$20S/lb) of pure pigment.
This means that red beet pigment and anthocyanin pigments are unlikely to
be produced in the near future. A pigment like saffron has a value of
£5000/kg (US$341O/lb). Since this is one of the few water-soluble yellow
food colorant carotenoids allowed in Europe it could be an interesting
opportunity for cell culturing [13].
3.4.2.6 Disadvantages/technical successes and problems. There are dis-

advantages to plant tissue culture:
• the production of cells is very slow
• the metabolic pathways of plant cells are not known
• large (expensive) reactors are needed
• engineering of plant cells is not as easy as microbial R-DNA technology
The technique has been beset with problems, such as the need for special
culture conditions to induce secondary metabolite production, low yields
and genetic instability of established cell lines leading to loss of production
[3]. Generally however the low yield obtained via plant tissue culture has
not made it an economical source of colours [36].
It has also been predicted that plant tissue culture will only be suited to
reactions unique to plants, because of competition from microbial fer-
mentation, which is cheaper because microbes grow faster than plant cells
and thus need less sterile conditions. Microbes do not have large and rigid
cell walls and so are less susceptible to damage in fermentors [37]. Plant
tissue culture will prevail in conditions specific to plants because of the
difficulty of transferring the complex DNA responsible for secondary
metabolite production into microbes [38]. Since colours are often complex
compounds produced as secondary metabolites in plants, tissue cultures
derived from plants would appear to be a good source of these compounds.
3.5 Fermentation
Microbial fermentation appears to have had a more immediate impact on

the colours industry. Colorants from microbial species offer considerable
advantages since they can be produced in any quantity and are not subject
to the vagaries of nature. Genetic manipulation also appears to have had
more success in microbial fermentation than in plant tissue culture. In
Table 3.4 Examples of colour pigments produced by microbial transformation
Colour pigment Non-microbial source Microbial source
~-Carotene Carrots, palm oil, synthesis Dunaliella salina

Blakeslea trispora
Astaxanthin Crustacea, birds' feathers, Phaffia rhodozyma
synthesis, Adonis sp. Haematococcus sp.
flowers
Lutein Alfalfa, com, marigold Spongiococcum excentricum
flowers, maize, green plants Chlorella pyrenoidosa
Zeaxanthin Alfalfa, com, marigold Flavobacterium sp.
flowers, maize, green plants
Canthaxanthin Crustacea, birds' feathers, Cantharellus cinnabarinus
synthesis Brevibacterium KY -4313
Monascus red pigments Monascus purpureus
Phycocyanin Spirulina sp.
research concerning the production of food colours by fermentation

processes, three main areas can be identified:
• surface (solid state) fermentations
• algal fermentations
• submerged fermentations
Natural extracts, particularly of fruits, flowers and leaves, have been used
for many years as sources of carotenoids as colorants. Despite the
availability of a variety of natural and synthetic carotenoids, there is
currently an increasing general interest in biotechnology and the commer-
cial production of carotenoids by microbial systems is being evaluated.
Specifically, algal production of l3-carotene has been examined in some
detail. A red yeast, Phaffia rhodozyma, has been explored as a source of
astaxanthin for fish feed (salmonoid) and has enjoyed some limited
success. Biotechnology has also been used to improve traditional fermenta-
tion processes e.g. Monascus, traditionally grown on rice in the Orient to
produce a red food colorant.
The principal colour pigments, the non-microbial sources and the
microbial species used in the main areas of research to date are outlined in
Table 3.4 [39].
3.5.1 Fermentation development

Dramatic advances have occurred in fermentation technology during the
past 40 years. Techniques available for exploitation of microbial systems
have become increasingly specialised. Recombinant DNA technology has
expanded both the potential product list and the selection of microbial
systems available to produce those products. The successful transfer of

basic research from laboratory curiosity to commercial success requires
technology development in the areas of strain development, fermentation
development and engineering design. Fermentation development activities
must optimise both cell growth and the expression of the desired metabol-
ites. The development of a commercial fermentation process, regardless of
product, requires R&D in those same areas of strain development,
fermentation development and process engineering.
3.5.1.1 Strain development. Without doubt, advances in microbial

genetics/molecular biology have had the greatest impact on the biotech-
nology revolution in general and on strain development in particular. The
aim of strain development work is to identify 'new' or improved strains for
production of the desired product. These activities begin with the initial
laboratory screening and should continue throughout the life of the
product.
3.5.1.2 Fermentation development and process engineering. Fermenta-

tion development work is usually initiated soon after strain development
efforts have begun. This work combines biological and engineering
expertise to identify the optimum growth/production media, process
variables (temperature, pressure, pH, aeration) and fermentation mode
(batch, fed-batch, continuous) for a particular process. Knowledge of
when during the microbial growth cycle the product is elaborated must be
established first. For the sake of simplicity, fermentation products can be
considered to be primary or secondary metabolites. Primary metabolites
are products that are obligate products of cell growth. Conversely,
fermentation products whose enhanced production is not directly asso-
ciated with cell growth are considered secondary metabolites. Therefore in
tandem with the media development work, consideration must be given to
the most efficient fermentation mode in which the process should operate.
The use of continuous culture systems to increase productivity (mass of
product/volume of medium/time) is often suggested but seldom commer-
cialised. Following optimisation of the biology, the engineering group must
scale-up the process so that an acceptable commercial process is achieved.
3.5.2 Solid state (surface) fermentation

The primary example of solid state or surface fermentation for the
production of colour pigments has been work carried out on improving the
production of red pigment from the traditional oriental Monascus fer-
mentation process. The genus Monascus includes several fungal species
which will grow on various solid substrates, especially steamed rice.
Monascus species have been used for a long time as a meat colorant,
disinfectant and as a natural food colorant. They are currently being

produced commercially in Japan, Taiwan and China.
Monascus has been the focus of much work concerned with improving
general culture conditions and substrates, which has led to over 50 patents
being registered in Japan, the USA and France [40, 41].
Publications include the research of Wong et al. [42] who found that
maximum pigment production from Monascus purpureus was obtained in
the presence of high levels of ammonium nitrate with low levels of glucose.
Lin et al. [43] derived a mutant of Monascus kaoliang by mutagenesis,
which when grown on a solid culture of steamed bread produced a deep red
pigment with a hundredfold greater yield than the wild type.
For many years, Monascus species were grown on solid cereal substrates
and the whole mixture was ground and used as a food additive. It became
obvious that the fungus could be cultured in liquid media or semi-solid
state, so several studies were made to optimise the production of pigment
in a variety of culture modes. The six well known Monascus pigments are
produced mainly intracellularly and are insoluble in water. Thus many
patents have focused on the extraction and solubilisation of the pigments.
More recent work has focused on culturing Monascus spp. in glutamate as
a nitrogen source to shift the location of pigment from predominantly
intracellular to mainly extracellular in order to produce a new type of
water-soluble red pigment [44, 45].
3.5.3 Algal fermentation
The major developments in recent years have occurred with micro algae
[46], especially Dunaliella species and, more recently, Haematococcus
species. Microalgal production combines photosynthesis with the charac-
teristics of microbial fermentations: high growth rates and hydraulic
production systems. Microalgal production R&D was initiated 40 years ago
in the USA, with human food production as the goal. The first commercial
operations were achieved in Japan, where the single-cell alga Chlorella was
produced and sold as a health food. The filamentous blue-green alga
Spirulina, a traditional food in several countries, is also being produced,
in Mexico, the Far East and the USA. In Japan, phycocyanin, a blue
pigment, is extracted from Spirulina and used as a food colouring agent.
Microalgal biotechnology has become a well established branch of indus-
trial microbiology.
3.5.3.1 r>-Carotene. Most natural r>-carotene is produced through the

mass cultivation of the micro alga Dunaliella salina. Another species
D. bardawil is now thought to be one and the same as D. salina. Following
reports of the occurrence of natural blooms of D. salina around the world
Table 3.5 Major producers of Dunaliella carotene
Company name Location Dunaliella sp.
Microbio Resources (TrouwIBP) Death Valley, California D. salina

Cyanotech Keahole Pt., Hawaii D. salina
Salt Research Institute Tianjin, China D. salina
Betatene Whyalla, S. Australia D. salina
Western Biotechnology Hutt Lagoon, W. Australia D. salina
Nature Beta Technology/Koor Foods Eilat, Israel D. bardawil
(35% shareholding by Nikkon Sohansha)
in the early part of this century (summarised in [47]), research was carried
out on the growth requirements, cultivation and biochemistry of the alga.
Much of this was done by groups in Israel [48--51] and Australia [52-54].
The production of j3-carotene from Dunaliella salina was examined by
Borowitzka et al. [55], who reported the discovery that salt concentration
was the major factor determining the level of algal growth in the pilot plant
since a high salt concentration inhibited the growth of a predatory
protozoan. They also discovered that high light intensity assists D. salina
and inhibits the growth of its non-carotenogenic algal rivals. The discovery
of these conditions greatly assisted the commercial exploitation of the alga.
Dunaliella spp. belong to the Chlorophyceae, with the cupshaped chloro-
plast in the lower half of the oval-to-spherical cell. The cell does not have a
cellulosic cell wall but only a cell membrane, thus allowing the cell to
expand rapidly for osmotic control. Longer term osmotic control is
achieved by balancing the external solutes with internal glycerol as is well
documented in the references given above. Two apical flagellae propel it
through the water and the cell is also phototactic.
Under conditions of stress (high light, high temperature, high salt,
mineral stress) these green algae become orange because they produce
large amounts of 'secondary carotenoids' outside the chloroplast, in the
cell wall or in oil droplets; Dunaliella species accumulate j3-carotene up to
10% of the dry weight. The algae are now being cultured on a large scale in
vast open ponds in countries with a suitable climate, especially Australia,
Israel and the USA (Table 3.5) [56]. Culture methods have evolved along
two different lines: extensive and intensive. Extensive algaculture involves
harvesting natural halotolerant algae from saline water. Intensive algacul-
ture involves the growth of the microalgae in bioreactors.
Since no additional energy input is needed and the extreme salt
concentrations used (up to 4 molar with Dunaliella) reduce contamination
with other organisms, production of carotene-rich dried algae or oil-based
extracts is cheap and such preparations are competitive. The purification of
j3-carotene from the extracts, however, increases the cost substantially.
A number of commercial ventures have started in the last 10 years with

the aim of cultivating the alga to make l3-carotene products. Commercial
operations have usually been established in arid areas with high light
intensities and a comparatively low number of cloudy days. The commer-
cial systems for cultivation of the alga differ considerably, from the
intensive or raceway system with mechanical agitation, through larger
structured ponds to the extensive open-lake system [57]. The latter system is
the one currently operated by Betatene Ltd. at Whyalla in South Australia.
This operation, believed to be the largest area of man-made algal
cultivation in the world, covers over 300 hectares which is divided into
three major cultivation lakes [58]. A requirement of the extensive system is
cost-effective harvesting to handle the relatively large culture volumes.
The extraction plant at the lakeside is designed to handle up to one million
litres per hour (220000 gallons per hour) in four parallel modules using a
proprietary purely physical harvesting system.
The advantages of extensive algaculture include lower capital costs and
potentially lower production costs. Extensive culture does not require
energy for mixing the cultures; it needs lesser amounts of nutrients
including CO 2 , because of lower biological productivity than intensive
culture. One disadvantage is the necessity to process the large volume of
water because of the lower carotene content per unit volume. It is very
difficult to control the physical, chemical or biological parameters of the
cultures in natural bodies of water. Chemical nutrients or pollutants and
biological contaminants, such as bacteria and predatory protozoa, can also
be difficult to control.
Intensive culture has advantages in that the high-rate monoculturing
permits the control of the physical, chemical, and biological parameters;
salinity and medium nutrient concentrations can be maintained at optimum
levels. Moreover, selected strains of Dunaliella can be introduced and
maintained in pure monoculture. Populations of deleterious bacteria and
protozoa can be more easily detected and quickly eliminated in bioreac-
tors. One disadvantage of intensive culture is that capital costs can be
higher and the system can require sophisticated personnel and equipment.
3.5.3.2 Astaxanthin. Industrial interest is now gradually shifting away

from the yellow carotenoids such as l3-carotene and lutein towards the
considerably more valuable orange-red ketocarotenoids, for which at
present no cheap commercially exploitable plant or animal sources exist.
The green microalgae, Haematococcus pluvialis is known as a potent
producer of astaxanthin such that accumulation of astaxanthin under
conditions of nitrogen deficiency can be as high as 2% of its dry weight.
The genus Haematococcus has an interesting developmental cycle; under
unfavourable conditions (such as nitrogen deprivation) cysts germinate and
simultaneously cells begin the massive accumulation of astaxanthin. Fully
mature cysts contain up to 5% by weight astaxanthin, predominantly in the

form of monoesters of fatty acids [59]. In 1987, Microbio Resources Inc.
developed a process for the commercial production of natural astaxanthin
from Haematococcus pluvialis [60].
The cultivation is in open ponds i.e. production bioreactors of 500 000
litres (110000 gallons), 4500 m2 (5382 sq. yards), in which astaxanthin
production consistently proceeds, while encystment is occasionally difficult
to control. As with any individual microalgal scale-up the two important
biological problems were contamination of the ponds with foreign algae
and with protozoal predators. Methods for harvesting Haematococcus
were listed as proprietary. After harvesting, the tough cell wall has to be
broken to release the free astaxanthin and its esters, but without damaging
the molecules. The method used was patented and involved sonication.
Microbio Resources did not disclose any production costs other than to
report comparative production costs for the production of both Haemato-
coccus and Dunaliella.
In the culture, astaxanthin remains entirely intracellular so the final yield
depends on the amount of biomass produced. Biomass concentration is
usually low because the cells are grown autotrophically for economy (i.e.
they receive no complex nutrients from the medium). However, for some
species initial heterotrophic conditions reportedly accelerate the process of
secondary carotenoid formation, presumably by causing the nitrogen
supply to be exhausted sooner. Even then, the time course of astaxanthin
accumulation is slow and remains unfavourable from a practical point of
view [39). So despite the fact that Haematococcus has a high concentration
of astaxanthin, the industrial application is limited by the lengthy auto-
trophic cultivation in open freshwater ponds and the requirement for
disruption of the cell wall to liberate the carotenoid.
More recent research work in this area was from the Fermentation
Technology Department in Hiroshima University, where research is
concentrating on the effects of light intensity, light quality and illumination
cycles on astaxanthin formation [61]. Further work on the growth and
astaxanthin formation of Haematococcus pluvialis in heterotrophic and
mixotrophic conditions supports the theory that in H. pluvialis, unlike in
most green algae, both photosynthesis and oxidative metabolism of acetate
can occur simultaneously for mixotrophic growth [62].
3.5.3.3 Zeaxanthin. There is a renewed interest in zeaxanthin for use in

poultry pigmentation as a competitor for xanthophyll. Among microbes,
free zeaxanthin has been detected in marine ftavobacteria. Although
information is scarce, the commercial production of zeaxanthin from a
Flavobacterium sp. appears to be approaching the stage of commercialisa-
tion [63). A mutant strain No. ATCC 21.588 or ATCC 21.081 reportedly
produces 335 mg of zeaxanthin/I. A number of patents on zeaxanthin
production have been assigned to different companies including Nestle

[64].
3.5.3.4 Canthaxanthin. Some bacteria have been considered for can-

thaxanthin production by fermentation, for example strains of Brevibacter-
ium and Rhodococcus maris, but commercial production has not yet been
established. Brevibacterium KY-4313, originally isolated from a petroleum
field in Japan, has been the subject of fermentative overproduction studies,
including mutant preparation, activation of carotenogenesis and growth
promotion for canthaxanthin [65]. However, although a substantial im-
provement in canthaxanthin yield was achieved compared with the optimal
result of Tanaka et al. [66], the pigment levels remain insufficient to make
Brevibacterium KY-4313 a realistic candidate for the production of
'natural' canthaxanthin. Unfortunately, Rhodococcus maris is inferior to
Brevibacterium KY -4313 in terms of canthaxanthin yield, reproducibility of
growth and pigmentation in hydrocarbon media. Attempts to grow
Rhodococcus maris efficiently in the various non-hydrocarbon media
previously developed for Brevibacterium were also less successful.
3.5.4 Submerged fermentation

As mentioned previously, natural sources of astaxanthin include crustacea,
algae, plant and yeast. The discovery in 1976 that astaxanthin also occurs in
the red yeast Phaffia rhodozyma as a secondary metabolite, caused
considerable excitement [67, 68]. Since then several biotechnology com-
panies active in the pigment business have devoted considerable research
and development efforts to this organism [69-76].
P. rhodozyma has desirable properties as a biological source of pigment,
including rapid heterotrophic metabolism and production of high cell
densities in fermentors, but its content of astaxanthin in wild strains is only
200-600 J.Lg/g yeast (0.02 to 0.06%). The growing market for astaxanthin,
together with the trend towards using natural sources of feed nutrients,
make the production of astaxanthin by P. rhodozyma attractive. The usual
issues have been addressed; the optimisation of the astaxanthin yield, the
search for cheaper media and the extraction of the astaxanthin from the
cells.
3.5.4.1 Optimisation of astaxanthin yield. Astaxanthin production is

growth-associated, and the astaxanthin concentration gradually decreases
after depletion of the carbon source. The biomass concentration decreases
rapidly during the stationary phase with a concomitant increase in the
intracellular content of astaxanthin. Initial efforts concentrated on the
important fermentation parameters such as pH, aeration rate, vitamin
requirements and particularly the nature and source of the carbon [77, 78].
Since the yield improvement from media optimisation is insufficient,

attention recently turned to the preparation of astaxanthin-overproducing
mutants [79-81]. Eventually, mutant strains containing 2000--2500 f.Lg of
astaxanthin/g yeast have been prepared through successive mutation steps
using the laboratory mutagens antimycin A and nitrosoguanidine. How-
ever, the increased astaxanthin content of the mutants tends to be at a cost,
slower growth and lower yield of cell densities; two drawbacks which could
possibly outweigh the gain in cellular pigment content. High producers are
often unstable and further strain development is required [82].
3.5.4.2 Reduction of media costs. Apart from the insufficient yield,

another factor negatively affecting future commercialisation of Phaffia
rhodozyma is the relatively high costs of the media used (yeast, nitrogen
base and sugars). Certain cheap food-processing waste products (e.g.
alfalfa residual juice) effectively support yeast propagation [83], but also
suppress the astaxanthin formation [84]. The inhibiting factor has been
attributed to the presence of saponins.
3.5.4.3 Extraction of astaxanthin. Astaxanthin is absorbed from Phaffia

rhodozyma and deposited in fish or chicken egg yolk only when the yeast
cells are ruptured prior to their incorporation in the feed [85]. A number of
alternatives have been suggested including preliminary autolysis of the
cells in distilled water or a citrate buffer [86] and digestion of the cell walls
by enzymes secreted by Bacillus circulans [85]. An experimental two-step
process involving the addition of Bacillus after completion of the growth of
Phaffia required heat treatment to inactivate the yeast cells and pH
adjustment of the medium. It was found to be more convenient to grow the
two organisms in mixed culture [87, 88]. This has the additional advantage
that the cell-free culture liquid can be re-used, because it still supports,
after fortification with some nutrients, the growth of Phaffia and also
retains the lytic activity necessary to modify its cell walls. A scheme
including mixed culture and recycling of culture filtrate has been worked
out to satisfy the needs of an economic large-scale process [88]. Unfortun-
ately, mixed culture partly suppresses the astaxanthin production.
From the extensive work on Phaffia rhodozyma which is still ongoing
[89], a picture emerges of a promising microorganism that has yet to be
engineered further to make it really commercially attractive.
3.6 Bioprocessing
Processing is the key for making biotechnology inventions work. It is the

strength of established fermentation companies in this area which enables
them to turn laboratory developments into viable products. Of course,
traditional colorant companies have similar strength in their traditional

areas, but it takes a great deal of effort to graft biotechnological techniques
onto their technologies. Arguably, while it cannot be strictly labelled
'biotechnology' there is a great deal of bioprocessing ongoing in the area of
traditional colours technology. This includes plant improvement tech-
niques to boost pigment yield, the application of innovative extraction
technologies to yield new sources of traditional pigments, and the use of
the 'biotechnologies' to improve the quality of pigments from traditional
sources.
During the last decade, carotenoids have emerged as possible anti car-
cinogens and as antioxidants protecting against free radical damage. As a
result, the demand for carotenoids is increasing and new methods are being
developed for their production. Commercial carotenoids are obtained by
synthesis, natural products and fermentation (Table 3.4). Traditionally,
carotenoids such as [3-carotene have been produced by organic synthesis
and very efficient methods have been developed to produce them in multi-
tonne quantities. This traditional approach has changed in the past decade
and methods have been improved or newly developed for production of
carotenes and xanthophylls from natural sources [90]. These include mass
culturing of algae, fermentation and extraction from natural sources.
3.6.1 Bioprocessing of [3-carotene

Production of palm oil carotene is increasing rapidly in Southeast Asia,
especially in Malaysia and Indonesia. Crude palm oil contains 500 to
1000 ppm of carotenoid and many researchers and enterprises have tried to
recover the carotenes by means of solvent extraction, absorption, frac-
tionation, saponification-solvent extraction etc. since the 1940s. No one
succeeded in commercialising a process, and all attempts ceased when
Hoffmann-La Roche established the synthetic route for [3-carotene in
1954. Lion Corporation of Japan, which had been working on recovering
carotene from palm oil since the 1970s, developed a unique process [91].
The process used is illustrated in Figure 3.1. Initially, crude palm oil is
converted into methyl esters by transesterification with methanol. This
reaction proceeds at low temperature and results in little, if any, loss of
carotene. In the second step, methyl esters are mixed with an organic
solvent, and then the mixture is separated into a carotene-rich layer and a
decolorised methyl ester layer. It was based on an interesting phenomenon
initially observed accidentally by a researcher.
This decolorised methyl ester can be used for soap, detergents, fabric
softeners, toiletries, cosmetics and chemicals. Carotene is further concen-
trated from the carotene-rich layer by standard methodology. Finally, it is
purified by an innovative commercial scale HPLC with hexane. Palm oil
carotene (after HPLC treatment) contains more than 95% carotenoids of
.....--T---;==~-- Methanol
E,,1raction
Detergent
'---....;;.;T;.;.;.;.-;:=~--Hexane
Softener
Purification
(Chromatography) Soap
Toiletries
Cosmetics
Figure 3.1 Lion procedure for producing palm oil carotene [91].
Table 3.6 Carotenoid composition of various sources
Carotene source a-Carotene ~-Carotene Other

carotenoids
Carrot carotene 30-50% 40-60% 10-15%

Synthetic 100%
Dunaliella carotene 6-10% 84-90%
Palm oil carotene 30-35% 60-65% 5-10%
which 60-70% is ~-carotene, 30-35% is a-carotene plus a small amount of

'Y- and A-carotenes. In this respect it is quite similar to carrot carotene
composition (Table 3.6). This distinguishing feature of palm oil is that it
contains a significant amount of a-carotene.
It has been generally accepted that 'carotene' equates to ~-carotene,
because ~-carotene has occupied almost all markets for carotene. How-
ever, new and natural sources of carotene containing a considerable level
of a-carotene (palm oil) revealed some interesting physiological effects. It

is expected that this will significantly expand the carotene market.
3.6.2 Bioprocessing of annatto

Traditionally annatto is cultivated from the tropical shrub Bixa orellana
which grows rapidly in Brazil, Peru, Central America, Kenya and India.
High transportation costs for untreated seeds and a variable low percent-
age (2 to 4%) of useful pigment extracted from the seeds, has focused
attention on the application of plant improvement techniques, novel
pigment extraction methods as well as plant tissue culture to annatto
cultivation and processing.
The main purpose of researching in vitro culture is to perfect a new
method of vegetative propagation. The potential advantages would be
reduction in the general heterogeneity found in the progeny produced by
the conventional method (e.g. seed sowing), and an increase in the
productivity of the pigment bixin compared to the mother plant. Consider-
able increase in productivity is therefore expected by using homogeneous
plants with high individual qualities. Direct organogenesis, indirect
organogenesis and somatic embryogenesis are the three processes gener-
ally applied to achieve in vitro micropropagation. As outlined earlier in this
chapter, work is ongoing in this area in South Africa and Brazil utilising the
organogenic and embryogenic approach [17, 18].
Classical propagation studies are also ongoing at nursery level and small
commercial scale in Brazil. These involved 'slipping' i.e. taking roots from
annatto plants with root systems having the best growing characteristics
and transposing them onto the stems of annatto plants with the highest
yields of bixin. The outcome of this work over a period of 6 years has been
the gradual increase in bixin concentrations in indigenous Brazilian
annatto shrubs from an average of 1.8% to an average of 2.8% and the
development of a cultivated annatto industry. Traditionally annatto was
simply used as a seasoning in Brazil and high yielding shrubs characteristic
of an organised annatto industry were not available. Shrubs of plants are
possible up to 4% bixin and typical concentrations in Peru are of the order
of 3 to 4%, while Kenyan levels vary from 2 to 3.5% bixin. There is still
some way to go in optimising the classical propagation route.
3.6.3 Bioprocessing of astaxanthin

Salmon farming increased substantially in the 1980s, which created a large
market for astaxanthin, the principal pigment of salmon and there has been
considerable interest within the aquaculture industry in using natural
sources of astaxanthin. Although widely distributed throughout the plant
and animal kingdom, only four natural sources of astaxanthin are realistic:
• krill and crustacea waste

• algae (Haematococcus pluvialis)
• flowers of the plant genus Adonis
• yeast (Phaffia rhodozyma)
Each pigment source has its limitations and none can currently compete
economically with the synthetic variant. Yeast provides a most promising
source due to the high intrinsic content of astaxanthin, its ease of culture
and its lack of waste byproducts. Algae, such as Haematococcus pluvialis,
are known to be a rich source of astaxanthin and have been extensively
studied and cultured under laboratory conditions for many years. These
algae are however, difficult to culture yielding relatively low biomass
levels.
Crustacean wastes have relatively low contents of astaxanthin and high
levels of moisture, ash and chitin. Several researchers have evaluated
carotenoid-containing crustacea or crustacean processing wastes as addit-
ives to sal monoid diets to improve pigmentation [92]. However, crustacean
shells are very low in carotenoid content (10-50 mg/kg) and high in
minerals which, without extensive processing to improve their dietary
quality, restricts their inclusion in sal monoid diets. Low levels of protein
and other nutrients further reduce their potential.
Astaxanthin has been identified in higher plants, such as in the flowers
of Adonis spp. Within the genus Adonis, two species A. annua and A.
aestivalis have been cited in the literature as containing astaxanthin within
their red flowers. Two botanically similar species, having significantly
larger flowers have been recorded growing wild in the Middle East, these
being A. aleppica and A. palaestina. By estimating flower surface area and
assuming that pigment concentration is similar, suggests that the above two
wild species may contain the greatest amount of pigment per flower.
Work on astaxanthin from plant sources has been carried out by
Unilever Research Laboratory, Colworth House and a patent has been
registered for seed variety modification [93]. There have been no reported
full scale field trials to confirm the commercial viability of the plant route.
The astaxanthin in the Adonis sp. mainly occurs as esters with little or no
free astaxanthin. Further mild processing would be required to produce the
required free pigment. Also, astaxanthin in some micro-algae is esterified
with fatty acids, and these forms of the pigments are thought to be more
poorly absorbed in the gut than the unesterified astaxanthin [60].
3.6.4 Bioprocessing of marigolds

The application of new biotechnology to the agricultural aspects of
naturally derived colours is developing rapidly. Work is ongoing in the area
of seed modification of Tagetes erecta to produce stronger hybrids of
mother plants in order to get higher colour (lutein) yielding varieties.
Furthermore, seed modification is now taking place to produce varieties

which are more suited to wetter climatic conditions and hence broaden the
growing season and the geographical location for marigold cultivation.
Similar work is being carried out on Capsicum spp. with regard to
producing varieties both high in colour (capsanthin) and low in flavour.
3.6.5 Bioprocessing of betanin

The production of red beet concentrate (Beta vulgaris) and powder has
been studied for many years and the many patents concerning it show
considerable interest and progress [94-97]. Attempts to increase the
pigment content have been made by fermenting out the sugar, whose
normally high level prevents further concentration of the juice. Fermenta-
tion work has been carried out in France, the former Czechoslovakia and
the USA. However, such a pure concentrate could no longer be claimed as
red beet juice concentrate.
Yeasts which have been used include Candida utilis [98] and Saccharo-
myces cerevisiae [99-101]. These yeasts can ferment high levels of sucrose
but do not have the side effects (such as oxidation, enzymatic activity and
damaging metabolites) which would adversely affect the pigment content.
Fermentations where a combination of enzyme from Aspergillus niger and
yeast (Saccharomyces oviformis) were used resulted in both the removal of
sugars and a reduction in the percentage vulgaxanthin (a yellow pigment)
[102]. This has the added effect of producing a redder red beet concentrate
with a single absorbance peak, a quality which is often required for the
Japanese market. The application of fermentation in red beet concentrate
production is an example of biotechnology to improve the quality of the
pigment from a traditional colour source.
3.7 Future perspectives
The availability of food colorants historically has depended on three

sources: specialised plant and animal sources, byproducts, and chemical
synthesis. Today, two other sources could be added, fermentation and
tissue/cell culture possibly with DNA transfer biotechnology. Specialised
plant and animal sources have been known for many centuries. Examples
are annatto, safflower, saffron, and turmeric from plants and cochineal
from insects. Byproduct sources are anthocyanins from wine grape skins
and carotenoids from many sources such as carrots and palm oil. Chemical
synthesis has provided the carotenoids, canthaxanthin, l3-carotene, 13-8-
apo-carotenal, and ethyl 13-8-apo-carotenoic acid.
The cell and tissue-culture concepts have been of much research interest,
possibly because of the pharmaceutical and fermentation industries.
Several reports have appeared on the tissue-culture applications of plants

for colorant applications, possibly due to the background of information
available from plant propagation and plant culture developments in these
areas. Plant products have been used in the pharmaceutical, cosmetic and
food industries for many years. The choice of supply by chemical synthesis
or plant biosynthesis depends on many factors, but three are predominant:
• Some chemicals are difficult to synthesise chemically and possibly could
be obtained more easily by biosynthesis.
• Others contain too many components.
• The ability to select strains or conditions that allow much greater
concentrations of the desired compounds to be produced.
Cell culturing will no doubt become a commercial process for production
of important secondary metabolites including colour pigments from plants
in the future. Today, there is still a need for basic understanding of the
metabolism, especially in the root, and all of the techniques are not ready
for commercialisation of plant cell culture production.
The demand for astaxanthin is expected to increase markedly as
aquacultural processes are developed for farming of prawns, shrimp,
lobsters and various fishes. The most promising biological sources at
present are the alga Haematococcus spp. and the yeast Phaffia rhodozyma.
Currently the market for astaxanthin is dominated by the synthetic
product, but it is possible that biological sources will begin to compete in
the late 1990s. There are serious limitations for biological production by
algae and yeasts. Economic production of astaxanthin by algae will depend
on developments in cultivation to increase the rate of growth and prevent
contamination in the sweet water conditions, and development of methods
to rupture the cells while maintaining carotenoid stability. Astaxanthin
production by P. rhodozyma depends on the isolation of carotenoid high
producing strains that will produce stable astaxanthin in industrial ferments
and perhaps utilisation of sugar waste streams as inexpensive growth
substrates. Isolation of P. rhodozyma strains that grow at temperatures
> 26°C would also be beneficial.
The culturing of Haematococcus is less favourable, since the time-course
of astaxanthin accumulation is too long and the yields of astaxanthin are
much lower, but the high cost of synthesising optically active astaxanthin
makes the feasibility of production by Haematococcus worth exploring. In
the longer term, the genetic transfer or introduction of the ability to
produce the more expensive commercially important carotenoids (e.g.
astaxanthin or zeaxanthin with the correct sterochemistry) into Dunaliella
is an attractive prospect. More research is also necessary in fundamental
aspects of carotenogenesis, a research area in biotechnology that at present
deserves considerable attention owing to the rapid growth in aquaculture.
Very recent molecular genetic work has achieved the introduction of the
genes for carotenoid biosynthesis from carotenogenic bacteria into the

normally non-carotenogenic Escherichia coli. High concentrations of ~
carotene and, potentially, of other carotenoids can be obtained in this way,
and rapid developments in this area can be expected.
3.7.1 Biotechnology and carotenoid production

While some success has been achieved in the production of natural
carotene, the production of natural carotenoids is still limited by low
producing organisms, cultural inefficiency and extraction/purification. To
overcome these difficulties, a number of biochemical- and biotechnology-
based methods are being applied to increase carotenoid production using
whole organisms and isolated carotenogenic enzymes.
3.7.1.1 Chemical regulation of carotenoid biosynthesis. Chemical reg-

ulation which provides a means to alter the production of a specific
carotenoid by fermentation or in cell-free systems has been used for this
purpose. The advent of site-directed mutagenesis and genetic engineering
has, however, provided an alternative to the use of chemicals for the
increased production of carotenoids.
3.7.1.2 Genetic manipulation and engineering. The first studies on

carotene mutants led to new insights into the carotenoid biosynthesis
which, in turn, have resulted in genetic manipulation of carotenogenic
encoding genes. Other studies have focused on expression of specific
enzymes involved in carotenogenesis. A number of commercial companies
are also exploring genetic manipulation of carotenoid biosynthesis [56].
For example, a research group in a major oil company has isolated the
genes for geranylgeraniol pyrophosphate synthase, phytoene synthase and
phytoene dehydrogenase and expressed the genes in E. coli, Agrobacter-
ium spp. and yeast to produce phytoene and lycopene. Two companies
producing natural carotene from D. salina have applied strain selection
and mutagenesis to yield algae producing up to 8% ~-carotene (on a dry
weight basis). A high astaxanthin-producing strain of a yeast has been
developed by a US genetic engineering company and will be used to
produce this carotenoid on a commercial basis [90].
3.7.1.3 Enzyme isolation. Current scientific evidence suggests that

there are three to four enzymes involved in the latter part of the synthesis
of carotenoids to ~-carotene and that these are membrane-bound [103].
For this reason, most biosynthetic studies on carotenoids have utilised cell-
free systems. Other workers have also found that carotenogenic enzymes
must be stabilised once removed from their membrane environment. More
efficient methods are also being developed for the isolation and purifica-
tion of carotenoid biosynthetic enzymes and binding proteins. The isola-

tion of carotenoid biosynthetic enzymes and/or the cloning of these
enzymes opens up the potential for mass production of carotenoids using
immobilised enzymes and bioreactors. A number of commercial compan-
ies are already investigating the immobilisation of carotenogenic enzymes
for continuous production of carotenoids. Such methods could overcome
the limitations of mass culturing and provide large quantities of inexpens-
ive natural carotenoids [90].
Besides having one of the widest distributions of any class of pigments in

nature, carotenoids also appear to have the highest diversity of function:
from precursor to vitamin A to phototropism and photoprotection. Given
this historical context, it is not surprising that we continue to find new
functions and applications for carotenoids. The application of carotenoids
as food and cosmetic colorants provided their first commercial market.
Their application in animal feeds extended this use and now represents one
of the largest markets for carotenoids. The potential use of carotenoids as
anti carcinogens and antioxidants with protection against free radical
damage is potentially their largest market. Biotechnology is playing a
central role in carotenoid chemistry and production. Genetically en-
gineered carotenoid producing microbes and carotenogenic enzymes,
immobilised enzyme systems and bioreactors, and new methods to stimu-
late and control biosynthesis are actively under development. These new
production methods will meet the demands for natural carotenoids as
further insights are gained into their functions and applications.
3.7.2 The influence of legislation

Finally a few words about the regulatory scene for natural colours
produced from biotechnological techniques. Assuming that technical
problems can be overcome and economic biotechnological production of
colours established, what legislative hurdles are likely to arise? The
uncertain policies of regulatory agencies towards biotechnology is a major
stumbling block [104]. As it is, many microorganisms have not been
approved for food use. Other problems lie with the slowness of the
regulatory machinery with the tendency to examine each product on a
case-by-case basis under different legislation in each country [105].
There is optimism in the belief that regulatory agencies will accept
biotech no logically derived products as being natural. The US legal defini-
tion of 'natural' includes products modified by cells or enzymes. However,
at present most regulatory authorities do not accept 'nature identical'
products as being natural. If biotechnological products were subsequently
deemed by regulatory agencies to be only nature identical this could have
an adverse effect on their marketability. Furthermore, the most important
threat to the use of plant pigments produced by cell culturing and

fermentation will be the legislation. The authorities are likely to prohibit
any pigment produced by cell cultures for use in the food industry unless it
can be shown that it is safe for human consumption. This will require
toxicological studies, which are expensive and time-consuming. Broadly
speaking the intent behind the Novel Foods Directive (which is now
tortuously proceeding through the EU machinery) is to make safety
clearance for biotechnologically derived products mandatory.
3.7.3 Commercial developments

The outlook for the commercial application of biotechnologically derived
colours in the food industry appears to be more limited. Of the natural
colours, the carotene family is considered to be the most technically sound,
other groups such as the anthocyanins and betalaines having shown
technical performance problems, e.g. colour fastness and stability. In line
with this the only commercial developments appear to have been with
carotenes and astaxanthin. In spite of some of these success stories, the
colours industry as a whole often shows a very conservative attitude to the
new technologies. This is because of their traditional technology approach
and adoption of a 'wait and see' attitude [106].
The potential role of biotechnology in the colours industry has come
under scrutiny. There are problems with the slowness in responding to the
opportunities offered by biotechnology, even though they exist in the food
industry with respect to the use of fermentation processes, plant tissue
culture, enzymes and gene manipulation. On examination of the appli-
cation of biotechnology in the colours industry, one comes to the con-
clusion that many of the possibilities for biotechnology with respect to
colours would fail on economic grounds. Problems highlighted included
failure by technologies, such as plant tissue culture, to produce good
product yields coupled with high capital and processing costs making the
products astronomically expensive, and hence not competitive with the
traditional natural sources.
Many obstacles still remain. A new process must first solve all of its,
often numerous, technical problems, produce a product that is competitive
with other sources of supply, be acceptable to the consumer, pass
legislative requirements and convince an often conservative colours indus-
try of its utility. This means that in spite of the many optimistic reports
suggesting success, the eventual outcome of these techniques remains very
uncertain.
Unless some radical advances occur in plant tissue culture technology,
the technologies that would appear to be the future leaders are microbial
fermentation and enzyme technology, areas which will be more acceptable
to the industry anyway because of their traditional use.
Acknowledgements
Several colleagues have provided me with advice on their particular areas

of expertise and I readily acknowledge, with thanks, the contributions
from Charlotte Didriksen, Chr. Hansen (plant biotechnology); Lance
Schlipalius, Betatene (algal J3-carotene); Dr Fons Peters and Kate Maume,
Quest International (astaxanthin); Marech Wallach, RSSL (microalgae);
Noel Sexton, Aine Curtin and Kieran Dwane, Quest International (helpful
discussions and proof reading). Finally, I gratefully appreciate the expert-
ise of Catherine Side, Side Hellon Marketing with the correction of the
manuscript, a task above and beyond the call of friendship.
References
1. Bell, E.R.J. and White, E.B. Int. Ind. Biotechnol. 9(3) (1989) 20-26.
2. Didriksen, C. Novel colouring agents. The Colour Conference, Leatherhead FRA
(1994) pp. 81-91.
3. Kirsop, B. Chem & Ind. April (1981) 218.
4. Kempler, G.M. Adv. in App. Microbiol. 29 (1983) 29.
5. I1ker, R. In-vitro pigment production: An alternative to color synthesis, Food Technol.
41(4) (1987) 70-72.
6. Hahlbrock, K. Biotechnology: Potentials & Limitations (S. Silver ed.) (1986) p. 241.
7a. Dougall, D.K. et al. Planta 149 (1980) 292.
7b. Kinnersley, A.M. et al. Planta 149 (1980) 200.
8. Ozeki, Y. and Komamine, A. Plant Cell Tissue Organ Culture 5 (1985) 45-53.
9. Hinderer, W. et al. Planta 160 (1984) 544.
10. Ozeki, Y. et al. Plant Tissue Culture 1982 Proc., 5th Inti. Congo Plant Tissue and Cell
Culture (Fujuwara, A. ed.) (1982) 355.
11. Yamakawa, T. et al. Proc. 5th Inti. Congo Plant Tissue and Cell Culture (Fujiwara, A.
ed.) (1982) p. 273.
12. Zryd, J.P. et al. Proc. 5th IntI. Congo Plant Tissue and Cell Culture (Fujiwara, A. ed.)
(1982) p. 387.
13. Igarashi, Y. and Yuasa, M. Effects of ammonium and total nitrogen content in culture
medium on shoot generation from calli in Saffron (Crocussativus L.), Plant Tissue
Culture Letters 11(1) (1994) 61-64.
14. I1ahi, I. et al. J. Plant Physiol. 128 (1987) 227-232.
15. Isa, T. and Ogasawara, T. Japan. J. Breed. 40 (1988) 153-158.
16. Vasil, V. and Vasil, I.K. Cell Culture and Somatic Cell Genetics of Plants, Academic
Press, Orlando (1984) pp. 36-42.
17. Lins, A.C.R. The tissue culture research on Urucum (Bixa orellana), Proceedings of the
1st International Symposium on Annatto, Sao Paulo (1991) pp. 193-195.
18. Lemos, O.F. et al. Propagation of Urucum in vitro, Proceedings of the 1st International
Symposium on Annatto, Sao Paulo (1991) p. 320.
19. Signs, M.W. and Flores, H.E. The biosynthetic potential of plant roots, Bio Essays
12(1) (1990) 7-13.
20. Vasil, I.K. Cell Culture and Somatic Cell Genetics of Plants, Academic Press (1987).
21. Knorr, D. Plant tissue culture for the production of naturally derived food ingredients,
Food Technol. (1990) 71-79.
22. Yoon, H.Y. et al. Anthocyanin production by Daucus carota root cell culture in vitro,
HortScience 27(6) (1992) 206.
23. Kim, C.R. et al. Enhanced anthocyanin production in hairy root culture of Daucus
carota by fungal elicitation, HortScience 27(6) (1992) 206.
24. Cormier, F. et al. Effects of sucrose concentration on the accumulation of anthocyanins
in grape (Vitis vinifera) cell suspension, Can. J. Bot. 68 (1990) 1822-1825.
25. Chi Bao Do and Cormier, F. Effects of high ammonium concentrations on growth and
anthocyanin formation in grape (Vitis vinifera L.) cell suspension cultured in a
production medium, Plant Cell, Tissue and Organ Culture 27 (1991) 169-174.
26. Callebaut, A. et al. Production of anthocyanins by cell cultures of Ajuga reptans, Med.
Fac. Landbouw. Rijksuniv. Gent. 53(4a) (1988) 1713--15.
27. Callebaut, A. et al. Anthocyanins in cell cultures of Ajuga reptans, Phytochem., 29(7)
(1990) 2153--2158.
28. Nagarajan, Raj. P. et al. Optimization of anthocyanin yield in a mutated carrot cell line
(Daucus carota) and its implications in large scale production, J. Ferm. & Bioeng., 68(2)
(1989) 102-106.
29. Chi Bao Do and Cormier, F. Accumulation of anthocyanins enhanced by a high osmotic
potential in grape (Vilis vinifera L.) cell suspensions, Plant Cell Reports 9 (1990) 143--146.
30. Masahito, Taya et al. Production and release of pigments by culture of transformed hairy
root of red beet, J. Ferment. & Bioeng., 73(1) (1992) 31-36.
31. Masahito, T. et al. Influence of medium constituents on enchancement of pigment
production by batch culture of red beet hairy roots, J. Ferment. & Bioeng. 77(2) (1994)
215-217.
32. Masahiro Kino-Oka et al. Culture of red beet hairy root in bioreactor and recovery of
pigment released from the cells by repeated treatment of oxygen starvation, J. Chem.
Eng. of Japan 25(5) (1992) 490-495.
33. Nakajima, H. et al. Pigment synthesis by immobilised cultured cells of Lavandula vera
and characterisation of a component of the pigments, Agri. Bioi. Chem. 54(1) (1990)
53--59.
34. Bianco-Colomas, J. and Hugues, M. Establishment and characterization of a betacyanin
producing cell line of Amaranthus tricolor: inductive effects of light and cytokinin,
J. Plant Physiol. 136 734-739.
35. Stafford, A. Enz. Microb. Technol. 8, (1986) 578.
36. Collinge, M. Trends Biotechnol. 4(12) (1986) 299.
37. Whitaker, R.J. et al. Biogeneration of Aromas, American Chemical Society, Wash-
ington DC (1987) p. 347.
38. Tudge, C. New Scientist, January (1984) 25.
39. Nellis, H.J. and de Leenheer, A.P. A review: Microbial sources of carotenoid pigments
used in foods and feeds. J. Appl. Bacteriol. 70 (1991) 181-191.
40. Francis, F.J. Food Technology-Special Report Colour Additives 41 (1987) 62-{)8.
41. Wong, H.C. Antiobiotic and pigment production by Monascus purpureus, Ph.D Thesis,
University of Georgia, Athens, USA (1982).
42. Wong, H.C. et al. Mycologia 73 (1981) 649.
43. Lin, T.F. et al. App & Envir. Micro. 43(3) (1982) 671.
44. Lin, T.F. and Demain, A.L. Effect of nutrition of Monascus sp. on formation of red
pigments, Appl. Microbiol. Biotechnol. 36 (1991) 70-75.
45. Lin, T.F. et al. Formation of water-soluble Monascus red pigments by biological and
semi-synthetic processes, J. Ind. Microbiol. 9 (1992) 173--179.
46. Ben-Amotz, A. and Avron, M. Ann. Rev. Microbiol. 37 (1983) 95-119.
47. Butcher, R.W. An Introductory Account of the Small Algae of British Coastal Waters.
Fisheries Investigations Secretariat, Ministry of Agriculture, Fisheries & Food, UK
(1959).
48. Ben Amotz, A. and Avron, M. Glycerol, f3-carotene and dry algal meal production by
commercial cultivation of Dunaliella. In Algae Biomass & Use (G. Shelef and c.J.
Soeder eds), Elsevier, Amsterdam (1980) pp. 603-{)1O.
49. Ben Amotz, A. and Avron, M. Glycerol and f3-carotene metabolism in the halotolerant
alga Dunaliella: A model system for biosolar energy conversion. Trends Biochem. & Sci.
November (1981) 297.
50. Ginzburg, B.Z. and Ginzburg, M. Studies of the comparative physiology of the genus
Dunaliella (Chlorophyta, Volvocales). 1. Response of growth to NaCl concentration.
Br. Phyc. J. 20 (1985) 277.
51. Ginzburg, M. and Ginzburg, B.Z. Influence of age of culture and light intensity on
solute concentrations in two Dunaliella strains. J. Exp. Botany 36 (1985) 701-712.
52. Borowitzka, L.J. and Brown, A.D. The salt relations of marine and halophilic species of
the unicellular green alga, Dunaliella. Arch. Microbiol. 96 (1974) 37-52.
53. Borowitzka, L.J., Kessly, D.S. and Brown, A.D. The salt relations of Dunaliella. Arch.
Microbiol. 113 (1977) 131-8.
54. Brown A.D. and Borowitzka, L.J. Halotolerance of Dunaliella. In Biochemistry and
Physiology of Protozoa, 2nd edn, Vol. 1 (M. Levandowsky and S. Hunter eds),
Academic Press, New York (1979) pp. 139-90.
55. Borowitzka, L.J. Nova Hedwigia 83 (1986) 224.
56. Taylor, R.F. Carotenoids: Products, Applications and Markets; Decision Resources
Inc., Burlington, MA (1990).
57. Curtain, c.c., West, S.M. and Schlipalius, L.E. Manufacture of f3-carotene from the
salt lake alga Dunaliella salina; the scientific and technical background. Australian J.
Biotec. (1987) 51.
58. Schlipalius, L.E. The extensive commercial cultivation of Dunaliella salina. Bioresource
Technol. 38 (1991) 241-243.
59. Renstrom, B. and Liaaen-Jensen, S. Fatty acid composition of some esterified
carotenols. Compo Biochem. Physiol., B. Compo Biochem. 69 (1981) 625-627.
60. Bubrick, P. Production of Astaxanthin from Haematococcus. Bioresource Technol. 38
(1991) 237-239.
61. Kobayashi, M. et al. Effects of light intensity, light quality, and illumination cycle on
astaxanthin formation in a green algae, Haematococcus pluvialis. J. Ferment. Bioeng.
74(1) (1992) 61-63.
62. Kobayashi, M. et al. Growth and astaxanthin formation of Haematococcus pluvialis in
heterotrophic and mixotrophic conditions. J. Ferment. Bioeng. 74(1) (1992) 17-20.
63. Dasek, J., Sheperd, D. and Traelnes, K.R. Swiss Patent No. 537456 (1973).
64. Ninet, L. and Renaut, J. Carotenoids. In Microbial Technology, 2nd edn (eds H.J.
Peppler and D. Perleman), Academic Press, New York (1979) pp. 529-544.
65. Nelis, H.J. and De Leenheer, A.P. Reinvestigation of Brevibacterium sp. strain KY-
4313 as a source of canthaxanthin. Appl. & Env. Microbiol. 55 (1989) 2505-2510.
66. Tanaka, A. et al. Studies on the formation of vitamins and their function in hydrocarbon
fermentation. (IX) Production of carotenoids from hydrocarbons by Brevibacterium. J.
Ferment. Techno/. 49 (1971) 778-791.
67. Andrewes, A.G. et al. Carotenoids of Phaffia rhodozyma, a red-pigmented fermenting
yeast. Phytochemistry 15 (1976) 1003-1007.
68. Miller, M.W. et al. Phaffia, a new yeast genus in the Deuteromycotina (Blastomycetes).
Int. J. Syst. Bact. 26(2) (1976) 286-291.
69. Danisco Bioteknologi PCT Patent No. 88 08 025 (1988).
70. Igene Biotechnology, PCT Patent No. 90 01552 (1990).
71. Phillips Petroleum, European Patent No. 454024 (1991).
72. Quest International, European Patent No. 474 347 (1991).
73. Kyowa Hakko Kogyo, European Patent No. 543 023 (1992).
74. Villadsen, PCT Patent No. 92 22 648 (1992).
75. Gist Brocades, European Patent No. 590707 (1992).
76. Gist Brocades, European Patent No. 551 676 (1993).
77. Johnson, E.A. and Lewis, M.J. Astaxanthin formation by the yeast Phaffia rhodozyma.
J. Gen. Microbiol. 115 (1979) 173-183.
78. Haard, H.F. The production of carotenoid pigments from mineral oil by bacteria. J.
Bacteriol. 48 (1988) 219-231.
79. An, G.-H. et al. Isolation of Phaffia rhodozyma mutants with increased astaxanthin
content. J. Appl. Env. Microbiol. 55 (1989) 116-124.
80. An, G.-H. et al. Isolation and characterization of carotenoid hyperproducing mutants of
yeast by flow cytometry and cell sorting. Biotechnology 9 (1991).
81. An, G.-H. and Johnson, E.A. Influence of light on growth and pigmentation of the
yeast Phaffia rhodozyma, Antonie van Leeuwenhoek. J. Microbiol. Serol. 57 (1990) 191-
203.
82. Johnson, E.A. and An. G.-H. Astaxanthin from microbial sources. Crit. Rev. Biotech-
nolo 11 (1991) 297-326.
83. Okagbue, R.N. and Lewis, M.J. Use of alfalfa juice as a substratefor propagation of the
red yeast Phaffia rhodozyma. Appl. Microbiol. Biotechnol. 20 (1984) 33-39.
84. Okagbue, R.N. and Lewis, M.J. Inhibition of the red yeast Phaffia rhodozyma by
saponin. Appl. Microbiol. Biotechnol. 20 (1984) 278-280.
85. Johnson, E.A. et al. Simple method for the isolation of astaxanthin from the
Basidiomycetous yeast Phaffia rhodozyma. Appl. Env. Microbiol. 35 (1978) 1155-1159.
86. Okagbue, R.N. and Lewis, M.J. Autolysis of the red yeast Phaffia rhodozyma: a
potential tool to facilitate extraction of astaxanthin. Biotechnol. Letts. 6 (1984) 247-250.
87. Okagbue, R.N. and Lewis, M.J. Mixed culture of Bacillus circulans WL-12 and Phaffia
rhodozyma on different carbon sources: yeast-wall lytic enzyme production and
extractability of astaxanthin. Biotechnol. Letts. 5 (1983) 731-736.
88. Okagbue, R.N. and Lewis, M.J. Influence of mixed culture conditions on yeast-wall
hydrolytic activity of Bacillus circulans WL-12 and on extractability of astaxanthin from
the yeast Phaffia rhodozyma. J. Appl. Bacterial. 59 (1985) 243-255.
89. Meyer, P.S. and Du Preez, J.e. Effect of culture conditions on astaxanthin production
by a mutant of Phaffia rhodozyma in batch and chemostat culture. Appl. Microbial.
Biotechnol. 40 (1994) 780-785.
90. Taylor, R.F. Production of natural carotenoids. J. Amer. Chem. Soc., Chap. 4 (1992)
27-37.
91. Murakoshi, M. and Iwasaki, R. Palm oil yields carotene for world markets. Inform
(Oleochemicals) 3(2) (1992) 210-217.
92. Torrissen, O.J. et al. Pigmentation of salmonids-carotenoid deposition and metabolism
in salmonids. Crit. Rev. Aquatic Sci. 1 (1989) 209.
93. Unilever Research Laboratory, Scottish Patent Filing No. NCIMB 40309 (1990)
Colworth EDI.
94. Haber, G.J. et al. US Patent No.4 132 793 (1977).
95. Oragvelidze, N.!. et al. USSR Patent No. 565049 (1977).
96. Yoshotomi, T. et al. Japanese Patent No. 78 119 929 (1978).
97. Cherukuri, S.R. et al. US Patent No.4 371 549 (1981).
98. Adams, J.P. et al. Production of a betacyanine concentrate by fermentation of red beet
juice with Candida uti/is. J. Food Sci. 41 (1976) 78.
99. Verniers, C. French Patent No.2 399 467 (1977).
100. Drdak, M. and Nascakova, M. Czechoslovakian Patent No. 225 529 (1985).
101. Drdak, M. et al. Red beet pigment composition: effects of fermentation by different
strains of Saccharomyces cerevisiae. J. Food Sci. 57(4) (1992) 935-936.
102. Pourrat, H. et al. Purification of red beetroot dye by fermentation. Biotechnol. Lett. 5(6)
(1983) 381.
103. Britton, G. Pure & Appl. Chem. 63 (1991) 101-108.
104. McCormick, R.D. Prepared Foods 155(1) (1986) 131.
105. Peyvatin, J.L. Biofutur 50 (1986) 19.
106. Van Brunt, J. Biotechnology 3(6) (1985) 525.
4 Safety of food colorants
F.J. FRANCIS
4.1 Introduction
Safety of food colorants has been a very controversial topic in recent

decades but nearly all of the criticism has been directed towards the
synthetic colorants. The natural colorants have been singularly free of
criticism possibly due to the belief that most are derived from well known
food sources that have been consumed for many years. This concept is
reinforced by the belief that Man has somehow become conditioned to
tolerate certain compounds after a long history of use. This belief is
certainly naive because 5000 years is too short a time to create significant
genetic changes. It is true, however, that history has allowed us to identify
acute hazards and subsequently eliminate them from the diet. However,
long-term or even lifetime exposures are another story. Regardless,
natural colorants have a good history of safety.
4.2 Toxicity testing
Regulation of safety of food colorants is similar in principle around the

world, but individual countries differ in both protocol and interpretation.
In general, safety testing for all compounds depends on animal testing.
Animals, typically rodents, are fed increasing dosages of the chemical
under test and the effects are determined. A simplified diagram of the
relationship between dosage and effect is shown in Figure 4.1 [1]. In part
(a), the increasing dosage, within reason, has no effect and examples
would be sugar and starch. In part (b), every increase in dosage creates a
response. The classical example of this situation is radiation and few, if
any, compounds fall into this category. In part (c), no response is seen until
the body's defense mechanisms are overwhelmed and we see a response.
Nearly all chemicals fall into this category. In part (d), we see a harmful
effect at low dosage, followed by a beneficial optimum, followed by an
increase in response. Oxygen, selenium and vitamins A and D fall into this
category.
In Figure 4.2 [1, 2], a simplified animal experiment for chemicals in the
above category (c) is shown. First the total adverse response (TAR) is
determined. This is important because a high TAR response is indicative of
SAFETY OF FOOD COLORANTS 113
(a) (b) (c) (d)
IO[2J[2JO Dose Dose Dose Dose
Figure 4.1 Four diagrammatic responses to a toxic chemical [1). (a) No effect. (b) No
threshold. (c) Threshold. (d) Low dose beneficial threshold.
problems with the experiment. Then the background level of response is

determined followed by the dosage level at which no observed effect level
(NOEL) is seen. This is also called the NOAEL (no observed adverse
effect level). The NOEL is divided by 100 to obtain the acceptable daily
intake (ADI). The 100-fold safety factor comes from a factor of 10 to
change animal data to humans and another 10 to account for the variability
of humans. This procedure works well for many chemicals. However, if
there are non-reversible immunologic or reproductive effects, the safety
factor may be 1000. With carcinogens, the safety factor may be as high as
5000. Carcinogens create special problems because the time span for
humans may be over a generation or more. In any event, the uncertainties
lie in the determination of the NOEL, the extrapolation of high doses to
low doses, and the interpretation for humans. There are a number of
statistical models to extrapolate from high doses to low doses and they
could vary in estimates of risk by a factor of 200 000. One extreme example
is dioxin in which seven methods of calculating risks produced a 14-fold
difference. When we add the uncertainties because of the interpretation
for humans the degree of uncertainty rises even more, This makes it
possible for organizations to interpret the same data for a 'worst case' or a
'best case' scenario according to their own agenda. They would both be
statistically correct. The only way to reduce the level of uncertainty is to
understand the mechanism of action which may enable the researcher to
choose a more appropriate mathematical model to extrapolate from high
to low doses rather than using a linear model. It may also allow for the
choice of a more appropriate animal model. For example, the carcinogenic
effect of saccharin in rats is mediated through a protein in rat urine which
does not occur in humans. Rat data on saccharin are clearly unsuited for
humans.
The worst case scenario for the classification of substances as carcino-
gens has been criticized in a number of ways.
1. The use of the maximum tolerated dose (MTD) has been accused of
causing tumors because the injured tissue may have greater cell division
which by itself promotes more mutations and some mutations may
become carcinogenic.
2. Many experiments use the B6C3Fl strain of mice which is very sensitive
to tumor production.
3. Many classifications do not include the degree of potency for carcino-
genesis.
All of these concerns introduce uncertainty.
The extrapolation of effect from high to low doses has been criticized
because a linear extrapolation is usually used whereas in many cases the
curve is non-linear. Figure 4.3 shows a simplified case of a linear and a non-
linear extrapolation. One statistical calculation which seems to be gaining
favor is the virtually safe dose (VSD) possibly because the interpretation
(but not the calculation) is easy to understand. In Figure 4.3, at the same
acceptable risk level, the VSD is more than doubled. This again points out
the necessity of understanding the shape of the curve in order to choose the
most appropriate statistical calculation.
4.3 Acceptable daily intake (ADI)
The above concerns primarily deal with carcinogenicity but most safety
evaluations of food colorants do not involve carcinogenicity. Consequently
the most useful index of safety is the ADI (Figure 4.2). The ADI is usually
expressed as mg of the test substance per kg of body weight per day.
-
r
Typical
"0
c Safety
::>
0
Factor
Q)
<II
0,
)
oX
C (J
0
a.
<II
Q)
'"
.0
Q)
a: >
0
.0
<{
iii Background
~
Dose - 0.01 1.0 100 10000

I I
ADI NOAEL
Figure 4.2 A simplified dose response experiment with an animal model [1, 2].
ADI = acceptable daily dose. NOAEL = no observed adverse effect level. NOAEL and
NOEL are interchangeable.
Q)
</)
c
o
a.
</)
Q)
a:
Acceptable 1--_01<----,....
Risk Level
o VSD VSD Dose

(Linear (Nonlinear
Model) Model)
Figure 4.3 The calculation of the virtually safe dose (VSD) with a low-dose linear (-) and a
low-dose non-linear (-) mathematical model [1,2].
Expressing the ADI based on body weight minimizes the size difference
between animals. Expressing the dose as mg per day minimizes the
differences in life expectancy between animals. Also the ADI takes into
account contributions from all aspects of diet. The ADI is used by the Food
& Drug Administration (FDA) in the USA and worldwide by the World
Health Organization (WHO). In the USA, the Environmental Protection
Agency (EPA) chose to replace the lOO-fold safety factor by an 'uncer-
tainty factor'. Also because the ADI implies an inference of acceptability,
the EPA prefers the term 'reference toxicity dose' (RID), which implies a
non-scientific value judgment. Regardless, in most cases the ADI and the
RID are identical.
The ADI values are usually determined from chronic toxicity studies,
not acute toxicity experiments. The chief parameter in acute studies is the
LDso which is defined as the dosage which results in the death of 50% of
the experimental animals. The LDso may be modified to include a time
factor e.g. the LDso (100 days) is the dosage which results in the death of
50% of the animals within 100 days. For example, the LDso (100 days) for
sucrose for male albino rats is between 80 and 95 g/kg/day. LDso values can
be calculated for almost anything. For example, the LDso value for starch
with rats is 168 g/kg at which point gastric rupture occurs followed by death
in a few hours [3]. The LDso is a useful value for judging acute toxicity but
it is not of much help for long term chronic toxicity studies.
In the USA, the ADI or RID usually refers to a hypothetical average
person who weighs 70 kg [4]. For example, synthetic ~-carotene has an
ADI of 5 mg/kg/day. The ADI for the average person IS 5 X 70 =

350 mg/day.
4.4 Regulatory approaches
4.4.1 The European Union (EU)

The principles of food law are similar throughout the world but the
specifics vary considerably between countries. The EU has three main
principles:
1. Protection of the health of the consumer
2. Prevention of fraud
3. Removal of non-tariff barriers to intra-community trade
The third point has been the subject of most efforts since the intent to bring
about uniformity in the laws that affect the functioning of the Common
Market has required 'harmonization' of the laws between the various
countries. The need for harmonization was recognized with the awareness
that each country has its own national customs, tastes and social systems.
Harmonization was effected under Article 100 of the Treaty of Rome,
which established two types of EU Directives. The Colours Directive is an
example of the horizontal type and its provisions apply across the board.
The vertical Directives apply only to specific types of foods. Directives are
drawn up by the European Commission which may also take into account
the recommendations of the Codex Alimentarius Commission, the Joint
FAOIWHO Expert Committee on Food Additives (JECFA) and the EU
Scientific Committee for Food (SCF). Regardless of the source of the
Directive, it only has standing when approved by individual countries.
The EU Colours Directive lists colorants deemed to be suitable for food
use and specifies the limits of impurities. The SCF recommended colorants
for which an ADI could not be established for use in foods. The SCF made
an exception for colorants normally present in foods, and in situations
where consumption would not greatly exceed that normally present in
foods. All other colorants would require testing for safety prior to
acceptance. This lent support to the 'natural' vs. 'synthetic' categorization
of colorants. Since each country can modify the Directive according to its
own agenda, not surprisingly, there are some differences. Italy is probably
the most restrictive of the EU countries whereas Norway is probably the
most restrictive of the non-EU countries. Norway does not allow any
synthetic colorants and allows natural colorants only in amounts deemed to
be technologically necessary even though there are no specified limits.
Each colorant is identified by a specific E number. For example,
anthocyanins are listed as E163. Parker [5] published a list of colorants and
their current status in the EU and the UK.
4.4.2 The USA

The development of food laws in the USA followed those of Europe and
are governed by the Food and Drug Act. The Act has been modified
several times but the 1938 principles generally are still in effect. The 1938
Act established the terms F D & C Colors and stated that colorants,
presumably held to higher specifications for purity, would be allowed in
foods, drugs, and cosmetics. There is also a D & C classification for drugs
and cosmetics. Colorants used in industry for other applications could still
be referred to by their common name. The 1960 Color Additive Amend-
ment to the Act specified two groups of colorants: certified color additives,
and color additives exempt from certification. The first group comprises
seven synthetic F D & C color additives which are certified to comply with
the purity specifications required by the FDA. The second group has 26
colorants which do not require that each batch be analysed and certified.
Many of the colorants that would be called natural in other countries are
on the list of 26. The FDA does not accept the concept of 'natural' vs.
'synthetic' and requires that compounds in each group be subjected to the
same standards of safety. In practice, the safety studies required for some
of the natural additives were considerably less than that required for
the synthetics. This is understandable if one accepts the concept of the
Decision Tree approach suggested by the Food Safety Council [2]. The
Nutrition Labelling and Education Act of 1990 mandated that certified
color additives be specifically declared by their individual names, but the
requirements for exempt colorants were not changed. Exempt colorants
can still be declared generically as 'artificial color', 'color' or other generic
or specific term. However the term 'natural' is prohibited because it may
lead the consumer to believe that the color is derived from the food itself.
There is no such thing as a 'natural' color additive in the USA. Hallagan
et al. [6] published a detailed account of the status of the non-certified
colorants in the USA.
Testing for safety in the USA is described in detail in the FDA
publication Toxicological Principles for the Safety Assessment of Direct
Food and Color Additives Used in Food. The 1993 draft is commonly
known as 'Redbook-2'. Dr H.B.S. Conacher, with the Health Protection
Branch in Canada, commented that Redbook-2 has a great potential to be
the focal point for international harmonization of the safety assessment of
chemicals used in the food industry [7].
4.4.3 The Delaney Clause

The US situation is complicated by the so called 'Delaney Clause' which
was included in the 1958 Amendment to FDA law. Briefly the clause states
that no additive shall be added to food
'if it is found to induce cancer when ingested by man or animal, or if it is

found, after tests which are appropriate for the evaluation of safety of
food additives, to induce cancer in man or animals'.
Unfortunately Congressman Delaney did not appreciate the ingenuity of
modern chemists to push back the limits of sensitivity (Table 4.1). Today,
chemists can find almost anything in anything. The ability to analyze for
infinitely small amounts has far outstripped the ability of the toxicologists
to interpret the results. The end result is an unworkable regulatory clause.
Scientific opinion in the USA is almost unanimous in declaring that the
Delaney Clause is hopelessly obsolete and should be declared invalid.
Opposition to its abandonment is coming from some environmental groups
who fear that abolition of the Delaney Clause will result in a symbolic
weakening of the food safety laws. Also it is difficult to get some
Congressmen to agree to allow 'a little bit of cancer' in our food.
Regardless, the Delaney Clause is likely to be replaced in the near future
with some kind of de minimis, which comes from de minimis curat lex.
Freely translated, this means that the law does not concern itself with
trifles. The FDA in 1958 was not unduly concerned about the Delaney
Clause because they considered that food safety problems could be
handled under the existing food safety laws. They underestimated the
advances in analytical chemistry, but they have been really creative in
attempts to circumvent the Delaney Clause. These involve the Grand-
father Clause, the Generally Recognized as Safe (GRAS) list, the DES
proviso, the Sensitivity of Method (SaM) approach, the Constituents
Policy, and the De Minimis concept. Regardless, replacement of the
Delaney Clause is likely because nearly all current legislation includes
some form of a risk/benefit concept. Of more importance in the immediate
Table 4.1 The receding zero: sensitivity of analysis
Year Level of detection Approach

(moles)
1900 I x 10-3 Wet chemistry

1950 1 x 10--6 Gas chromatography (GC)
1960 1 x 10 9 High performance liquid chromatography (HPLC)
1970 1 x 10- 12 GC with mass spectrometry (MS)
1980 1 x 10- 15 GC-MS + clean-up
1990 1 x 10 lX HPLC with clean-up
1992 1 x 10 21 High performance capillary electrophoresis (HPCE)
1993 I x 10- 21 Laser excitation and fluorescence
1994 6 x 10-26 Confocal fluorescence microscopy, diffraction-limited laser
(one molecule) excitation and a high-efficiency avalanche diode
future is the recent action of the US Congress to change the Dietary

Supplement Health and Education Act to remove dietary supplements
from the food additive category thereby exempting them from the Delaney
Clause. Quite possibly we will see an increasing number of preparations
described as food supplements.
4.5 Safety of individual colorants
4.5.1 Beneficial aspects

Safety of food components is a two-way street since many compounds
exhibit a beneficial effect on human health. Riboflavin is well known for its
vitamin effects as well as its colorant potential. j3-Carotene, in addition to
its pro-vitamin A effects, is well established by a number of studies for its
beneficial effect on lung cancer [8]. But one well done study in Finland
suggested that j3-carotene actually increased the susceptibility of smokers
to lung cancer [9, 10]. Carotenoids are helpful in coronary heart disease
[11, 12]. Lycopene is beneficial in digestive tract cancers [13] and tomatoes
are claimed to be anti carcinogenic [14]. Lutein and zeaxanthin and, to a
lesser extent, j3-carotene, protect against macular degeneration in the eyes
[15]. Carotenoids, particularly astaxanthin and canthaxanthin, are well
known for their coloration effects in salmon and other seafood [16].
The f1avonoids have been the subject of a large amount of physiological
research. Cody et al. [17] edited a 591 page book entitled The Plant
Flavonoids in Biology and Medicine. The book covers a wide range of
physiological activity from antiviral activity, proliferation of human lym-
phocytes, anti-hepatoxic effects, anti-inflammatory effects to a number of
beneficial effects on a series of enzyme and hormone systems. Current
research is centered on the antioxidant activity of f1avonoids in chronic
diseases such as atherosclerosis [18]. Bioflavonoids in citrus extracts are
believed to make ascorbic acid more available, particularly to slow down
the rate of galactose-induced cataracts [19]. Anthocyanin extracts are
being promoted in Europe for a variety of physiological benefits including
cardiovascular diseases, atherosclerosis and cancer [20]. They are believed
to act through their antioxidant effect on lipids. Anthocyanins are also
claimed to lower cholesterol level and to have platelet anti-aggregating
activity, anti-thrombotic activity, beneficial ophthalmologic effects,
beneficial capillary permeability, increased wound healing, etc. [21].
Turmeric has been used for a variety of illnesses because of its reported
therapeutic properties [22]. It is reported to be an antioxidant and to be
capable of lowering both LDL and VLDL and raising HDL cholesterol
levels in blood [23]. Five reports have claimed that turmeric is anti-
carcinogenic [24-29].
Many of the benefits claimed above are complicated because the

epidemiological studies make it difficult to establish which compounds are
creating the effect. There are over 4000 known flavonoid compounds [30],
including the red anthocyanins, and over 600 carotenoids. Many food
colorants are extracts and it is economically prohibitive to purify the
components. Hence, many of the natural colorants on the market are
complex chemical mixtures. Colorants containing anthocyanins certainly
contain an array of yellow flavonoids. Extracts from beets will also contain
yellow betaxanthins. Colorants containing mainly l3-carotene will contain a
number of other carotenoids. The impurities in synthetic 'nature-alike' 13-
carotene will be different from those obtained from plant sources. This is
really not important from a food colorant point of view but it makes the
determination of physiological effects very difficult. This is also the reason
that safety tests should be carried out on the actual colorant extract in
addition to the pure compound.
4.5.2 Anthocyanins
Colorants containing anthocyanins, usually made from the skins of wine
grapes, are well known worldwide. The anthocyanins themselves are not
genotoxic as shown by a number of studies [23, 31-33]. The oral toxicity of
mixed anthocyanins was greater than 20 g/kg/day for rats [34, 35]. Dogs fed
a diet containing 15% grape color extract from Concord grapes for 13
weeks showed no significant toxic changes [36]. A similar experiment for
90 days also showed no effect [6]. Another experiment with multigenera-
tion studies on rats fed 7.5% and 15% powder from Concord grapes
showed no effects on reproduction [37]. Since grapes are the largest fruit
crop grown around the world with a long history of human consumption,
the lack of toxic effects is not surprising.
4.5.3 Carotenoids
I3-Carotene (CI 75130), in both the synthetic form and in natural extracts,
is well known as a colorant for a variety of foods. The acute oral toxicity of
l3-carotene is very low. The oral LDso in dogs is greater than 1000 mg/kg
[38] and a single intramuscular injection of 1000 mg/kg in rats had no
significant effect [39]. Lifetime dietary administration in both rats and mice
showed no carcinogenicity with a NOEL of 100 mg/kg/day for rats and
1000 mg/kg/day for mice [40]. No teratogenic or reproductive toxicity was
shown when four generations of rats were fed up to 100 mg/kg/day [41,42].
No cytogenic or teratogenic effects were observed in the offspring of
rabbits given up to 400 mg/kg/day by stomach tube [43]. With rats fed
1000 mg/kg/day, or the same dose by intubation, there was no embryo-
toxicity, teratogenicity or reproductive effects. Apparently rodents can
tolerate large quantities of J3-carotene, but extrapolation to humans is not

simple because they metabolize J3-carotene in different ways. Humans
absorb carotene unchanged while rats do not. With humans, the absorbed
J3-carotene is largely converted to vitamin A, esterified and transported in
the lymph. In the rat, J3-carotene is largely converted to non-saponifiable
substances [44, 45]. Regardless, there is ample evidence that J3-carotene in
reasonable quantities is harmless to humans. There is, however, evidence
that humans with a high intake of J3-carotene may develop hypercarot-
enemia leading to an orange skin coloration [46]. The JECFA established
an ADI of 0--5 mg/kg/day for the sum of carotenoids used as colorants.
4.5.4 Canthaxanthin
Canthaxanthin is a red carotenoid available in synthetic form as a colorant

for a wide variety of foods. The Hoffman-LaRoche Co. sponsored an
extensive series of safety studies which were submitted to JECF A [47]. The
oral toxicity is very low. The oral LDso for mice is greater than 10 g/kg.
Lifetime dietary studies in rats revealed liver changes at 250 mg/kg/day.
Increased blood cholesterol levels were seen in mice fed 1000 mg/kg/day
for 90 weeks. No effect was seen in dogs fed 500 mg/kg/day for 15 weeks.
Canthaxanthin was not carcinogenic at feeding levels of 1000 mg/kg/day
for 104 weeks for rats or 98 weeks for mice. It may even be anti-
carcinogenic [48]. No reproductive or teratogenic effects were seen when
three generations of rats were fed up to 1000 mg/kg/day. Deposition of
canthaxanthin crystals, producing retinal impairment, occurred in the eyes
of humans, cats and rabbits but not in rats, mice and dogs, upon ingestion
of large amounts of canthaxanthin. In humans, the visual impairment was
reversible in a few months [49]. Effects of high doses are available because
canthaxanthin is used as a tanning aid. One young woman died after taking
35 mg/day for four months [50]. Other adverse effects reported were
hepatitis and urticaria [51]. The JECFA was unable to assign an ADI for
cantha:xanthin because of the problem with retinal crystal deposition in
humans.
4.5.5 Apocarotenol
Apocarotenol is the third of the nature-alike carotenoids introduced by

Hoffman-LaRoche. Bagdon et al. [38, 52] investigated the safety and
metabolism of 8' -apo-J3-carotenol in dogs and found that daily ingestion of
1000 mg for 14 weeks had no effect, other than an increased level of
vitamin A in the kidneys. 8' -Apo-J3-carotenoic acid was found as a
metabolite.
4.5.6 Lycopene
Lycopene is being promoted as a food colorant [53] but no safety data are
available. This may be influenced by the observation that any possible use
as a colorant would be far less than that consumed as a food.
4.5.7 Annatto
Annatto (CI 75120) is a very old colorant. It is believed to be non-geno-
toxic by a series of studies [23, 54, 55]. The acute oral toxicity is very low.
The oral LDso in rats is greater than 50 g oil-soluble extract/kg and greater
than 35 g/kg for the water soluble extract [56]. Lifetime toxicity studies
revealed that annatto did not produce toxic effects in rats or mice when
fed 26 mg/kg/day for rats or one drop of 10% annatto in soy oil per
day. Painting the skin with 0.05 ml of a 50% solution of annatto in oil in
mice or subcutaneous injection of 0.1 % annatto in mice produced no toxic
effects. The authors concluded that annatto is not carcinogenic [57]. Rats
fed up to 500 mg/kg/day for three generations showed no adverse effects
on reproduction [56]. The JECFA established an ADI for bixin of
0-0.65 mg/kg/day.
4.5.8 Paprika
Paprika acts both as a colorant and a spice. It is available as a dry powder
and as an oleoresin extract. It is not considered to be genotoxic [54, 58,59].
The acute oral toxicity is very low since the LDso for mice is 11 g/kg
[60,61]. Information on lifetime toxicity, carcinogenicity and reproductive
toxicology is not available. The JECFA did not allocate an ADI because
they considered that the levels of paprika and its oleoresins in foods were
self-limiting.
4.5.9 Algae
The JECFA considered extracts from algae in 1990, namely Dunaliella
bardawi/, D. salina and D. kona only to be told in 1993 that the species
were identical [62]. The species used commercially is D. salina. Seven short
term toxicity studies on algal powder submitted to WHO [62] reported no
problems. Five similar studies on oil extracts indicated no problems [62].
Jenson et al. [63] compared the absorption of ~-carotene in humans
ingesting synthetic all-trans ~-carotene and extracts from D. salina contain-
ing equal proportions of trans ~-carotene and cis ~-carotene. The cis ~
carotene was absorbed to a lesser extent indicating a selective absorption
of trans ~-carotene. Furuhashi [62] suggested a NOEL of 2.5 g for
D. bardawil. The JECFA did not establish either a NOEL or an ADI for
powdered extracts of Dunaliella because of the variation in products and

insufficient data [62]. There are no systematic toxicological studies on the
oil extracts.
There is little toxicological data available for extracts of carrots, alfalfa,
corn oil, palm oil, etc. The JECFA had no objection to their use as food
colorants provided that the level of use did not exceed that normally
present in vegetables [62]. Some food supplements which conceivably have
a colorant function are well established in the human diet. For example,
Spirulina products were found to be non-toxic for hamsters [34].
4.5.10 Turmeric and curcumin

Turmeric is available in two forms, a dry powder and an oleoresin.
Turmeric has been the subject of a number of safety studies because its
consumption in Europe [64], India and the Middle East is very high. In
India, consumption was estimated at up to 3.8 g/day [65]. Both the powder
and the oleoresin were concluded to be non-genotoxic [66-79] and possibly
anti-mutagenic [80]. The acute oral toxicity is very low. The oral LDso of
turmeric oleoresin in mice and rats is greater than 10 g/kg [81] and the oral
LDso for curcumin in mice is greater than 2 g [82]. Rats fed turmeric for 52
weeks at 500 mg/kg/day [65] or 60 mg/kg/day of oleoresin for 60 weeks
showed no significant toxicity [29]. Monkeys fed 500 mg/kg/day for 9
months [65] or dogs fed 750 mg/kg/day of turmeric [29] showed no adverse
effects. There is no evidence of carcinogenicity [83], reproductive toxicity
or teratogenicity in rats and mice [84]. The JECFA did not allocate an ADI
for turmeric because it is considered to be a food. A temporary ADI of
0-0.1 mg/kg was established in 1990 by the JECFA for turmeric oleoresin
pending additional studies. An ADI of 0-0.1 mg/kg was established for
curcumin in 1990.
4.5.11 Cochineal and carmine

Cochineal extract and carmine (CI 75470) from the insect Coccus cacti are
not genotoxic [20, 83, 85-92]. Carmine was not carcinogenic when fed up
to 500/mg/kg/day for 19 weeks following in utero exposure [93]. Rats fed
carmine up to 100/mg/kg/day for 90 days showed only a reduced growth
rate at the higher levels [94]. There were no significant adverse effects on
reproduction in rats fed up to 100 mg/kg/day by intubation in a single
generation study or up to 500 mg/kg/day in a three generation
reproduction/teratology study [60, 95]. Some fetal malformations (2 in 85
treated mice) were observed in mice receiving a single subcutaneous
injection of 150/mg/kg on the eighth day of pregnancy [96]. JECFA
assigned a combined ADJ of 0-5 mg/kg/day for cochineal and carmine.
Carmine should not be confused with indigocarmine or carmOisme.

Indigocarmine, also know as indigotine, E132 or F D & C Blue No.2 is a
synthetic colorant in the indigoid class. Carmoisine, E122, CI 14720 is also
a synthetic colorant in the azo class.
4.5.12 Lac
Lac is a red colorant obtained from the insects Laccifera lacca and Kerria
lacca. It is the sodium salt of laccaic acid and is closely related to cochineal.
Chakravarti [97] reported that high concentrations of lac in the human diet
could possibly cause cell membrane damage in the liver. No effect was
found in hepatic enzymes [98-101], or on DNA by the Ames and Chinese
hamster lung cell tests [102]. Lac is permitted in India at an ADI of
2 mg/kg/day. The authors concluded that lac was safe at the permitted
level.
4.5.13 Caramel
Caramel has recently been submitted to a series of safety studies. Eleven
papers were published in 1992 in the same volume of Food Chemical
Toxicology. Caramels can be divided into four groups depending on the
process and the components but they are relatively similar. The history and
safety [103], characterization [104, 105], specification [106] and antioxidat-
ive and antimutagenic [107] properties were described in 1992.
Caramel is not genotoxic [108-110]. The subchronic oral toxicity of
caramel is very low. No significant toxicity was shown in rats consuming
16 g/kg/day in drinking water for 13 weeks [111] or in 20 g/kg/day for 13
weeks [112]. In two-year studies, caramel was not toxic or carcinogenic and
the NOEL was 10 g/kg/day for rats and mice [113]. No effects were noted
in humans who ingested 200 mg/kg/day for 7 days [114]. The JECFA
assigned an ADI of 0--200 mg/kg/day to caramel.
4.5.14 Monascus
Monascus molds produce yellow, red and purple pigments. The yellow
pigment was found to have no mutation effect with a series of microorgan-
isms [115]. The LDso for mice was 132 mg/20 g. Monascus extracts were
positive for Salmonella mutation and negative for chromatid exchange and
alkali elution. The authors concluded the preparation of rice fermented by
Monascus purpurous (Angkak) was safe for human consumption [116]. No
toxicity was observed in rodents [117] and in fertile chicken eggs [118].
Monascus colorants have a long history of consumption in Asia but no ADI
is available.
4.5.15 1ridoids
The geniposides from the fruits of the Gardenia plant were found to have
some hepatoxicity due to the aglycone genipin produced on hydrolysis of
the geniposides [119]. The toxicology of the yellow, red, blue and green
pigments from Gardenia was studied [120] and they were found to be safe
for human consumption.
4.5.16 Betalains
The betalain pigments from red beets were tested on rodents with four
conditions: (i) 50 mg/kg pure betanin; (ii) 50 mg/kg degraded betanin; (iii)
50 mg betanin after fermentation to remove the sugar; and (iv) 2000 ppm
betanin in the diet. No carcinogenic effects were observed [121]. Rats fed
beet extracts at 50 mg/kg betanin in the diet showed no effect and the
authors concluded the red beet extract was safe as a food colorant [41].
4.5.17 Cocoa
Tarka [122] wrote an extensive review of the toxicology of cocoa and the
methylxanthenes, theobromine, caffeine and theophyllin. The review
involved biological and behavioral effects, metabolism, species variation,
pregnancy, lactation, reproduction, teratology, mutagenicity, lactation,
reproduction, teratology, mutagenicity, fibrocystic breast diseases and
drug and dietary factors. No implications for human use as a beverage were
included. Cocoa extracts are mainly used as a food ingredient but they are
effective colorants.
4.5.18 Chlorophyll
The JECFA classified chlorophyll under List A1 which means that the
colorant has been ful1y cleared and not limited toxicological1y since its use
in accordance with 'good manufacturing practice' does not represent a
hazard to health. The only natural colorants on the A1 list are four
carotenoids, chlorophyll, caramel and riboflavin [123]. Chlorophyll is
usually used in the form of the copper complexes of chlorophyll.
4.5.19 Titanium dioxide

Food grade titanium dioxide (CI 77891) is a white powder prepared
synthetically, even though it occurs in nature, in order to avoid contamin-
ants. Titanium dioxide is not genotoxic [14, 25, 124-129]. The oral toxicity
is very low. The LD50 values are greater than 25 g/kg/day for rats [130] and
10 g for mice [131]. Long-term feeding studies on dogs, cats, rabbits,
guinea pigs and mice showed no evidence of carcinogenicity [132, 133].

No obvious signs of toxicity were evident with dogs fed 900 mg/kg/day, cats
fed 1500 mg/kg/day, rabbits fed 1500 mg/kg/day and guinea pigs fed
800 mg/kg/day all for 390 days [132]. The NOEL values were reported to
be 3.75 g/kg/day for female mice, 7.5 g for male mice and 2.5 g for rats
[133]. No significant toxicity was observed in humans with a single oral
dose of 450 g [134]. When orally administered to rats, 92% of Ti0 2
appeared in the feces after 13 days [135]. Apparently titanium dioxide is
both poorly absorbed [136] and non-toxic. Technologically, it is a very
effective stable whitener for foods.
The JECFA has not established an ADI for titanium dioxide and its use
is regulated by the GMP. In the USA, it is allowed in amounts up to 15 by
weight of the food.
4.5.20 Carbon black

Carbon black is a large-volume industrial commodity but food grade usage
is much smaller. It is used as a food colorant worldwide usually in the form
of a paste in liquid glucose because the fine powder by itself is very difficult
to handle. There is little safety data available and in the USA in the 1970s
when the GRAS list was being reviewed, toxicological data were requested
in view of the theoretical possibility of contamination with heterocyclic
amines. Apparently the cost of obtaining the requested data was greater
than the entire annual sales of food grade carbon black so the tests were
never done. Carbon black is not permitted as a food colorant in the USA.
Carbon black is technologically an effective stable colorant. The JECFA
did not set an ADI anticipating that its use would be self-limiting under the
GMP.
References
I. Francis, F.l. Food Safety: The Interpretation of Risk. Council for Agricultural Science
and Technology, Ames, Iowa (1992) 21 pp.
2. Anon, A Proposed Food Safety Evaluation Process. Food Safety Council (1982) p. 142.
3. Boyd, E.M. Toxicity of Pure Foods. CRC Press, Boca Raton, FL (1973).
4. Rodricks, 1. V. Calculated Risks. The Toxicity and Human Health Risks of Chemicals in
Our Environment. Cambridge University Press (1992).
5. Parker, L.E. Chapter 1 in Developments in Food Colours-2. Applied Science, London
(1984) I.
6. Hallagan, 1.B .. Allen. D.C and Borzellaca, 1.F. Food Chern. Toxicol. (in press).
7. Anon, Food Chern. News 35 (1994) 28.
8. Dartiques, 1.F., Dabis, F., Gros. A., Moise, G., Bois, R., Salamon, 1.M., Dilhuydy, M.
and Courty, G. European 1. Epidem. 6 (1990) 261.
9. Heinonon, O.P. and Albanes, D. New England 1. Med. 330 (1994) 1080.
10. Hennekins. CH., Buring, 1.E. and Peto, R. New England 1. Med. 330 (1994) 1080.
11. Abbey. M., Nestel, P.l. and Baghurst, P.A. Amer. 1. Clin. Nut. 58 (1993) 525.
12. Morris, D.L., Kritchevsky, S.B. and Davis, CE. 1. Amer. Med. Ass. 272 (1994) 1439.
13. Franceschi, S., Bidoli, E., Vechia, C.L., Talamini, R., D'Avanzo, B. and Negri, E.
Internat.l. Cancer 59 (1994) 181.
14. Dipaolo, J.A. and Castro, B.C. Cancer Res. 39 (1979) 1008.
15. Seddon, J.M., Agani, U.A. and Sperduto, RD. 1. Amer. Med. Assoc. 272 (1994) 1413.
16. Meyers, S.P. Proc. First Intern. Symp. on Natural Colorants, Amherst, MA (1993).
17. Cody, V., Middleton, E. and Harborne, J.B. Plant Flavonoids in Biology and Medicine
Pub!. Alan R. Liss, New York (1986) 581 pp.
18. Hertog, M.G.L., Feskins, E.J.M., Hollman, P.C.H., Katan, M.B. and Kronhaut, D.
Lancet 342 (1993) 1007.
19. Vinson, J.A., Courey, J.M. and Maro, N.P. Nutrition Res. 12 (1992) 915.
20. Brown, J.P. and Brown, R.J. Mutation Research 66 (1979) 223.
21. Minati, E. and Cole, R. Proc. First Intern. Symp. on Natural Colorants, Amherst, MA
(1993).
22. Ravindranath, V. and Chandrasekara, N. Toxicology 16 (1981) 259.
23. Haveland-Smith, R.B. Mutation Research 91 (1981) 285.
24. Huang, M.T., Smart, R.e., Wong, e.Q. and Conney, A.H. Cancer Res. 48 (1988) 5941.
25. Ivett, J.L., Brown, B.M., Rodgers, e., Anderson, B.E., Resnick, M.A. and Zeiger, E.
Molecular Mutagenesis 14 (1989) 165.
26. Kuttan, R., Bhanumathy, P., Nimala, K. and George, M.e. Cancer Letters 29 (1985)
197.
27. Nagabhushan, M. and Bhide, S.V. Food Chem. Toxicol. 22 (1987) 545.
28. Sondamini, K.K. and Kuttan, R.I. Ethnopharmacol. 27 (1989) 227.
29. Truhaut, R. Proc. Intern. Federation Pharm. (1978) 8.
30. Harborne, J.B. The Flavonoids. Advances in Research since 1980. Chapman & Hall
(1988) 621 pp.
31. Brown, J.P. and Dietrich, P.S. Mutation Research 66 (1979) 223.
32. MacGregor, J.T. and Jurd, L. Mutation Res. 54 (1978) 297.
33. Viola, M. and Nosotti, A. Bolletino Chimico-Farmaceutico 117 (1978) 402.
34. Chamorro, G., Salazar, M., Pages, N. and Dedict-Cofoa, B. de. Belgian 1. Food Chem.
and Biotech. 42 (1987) 188.
35. Pourrat, H., Bastide, P., Dorier, P., Pourrat, M.A. and Tranche, P. Chimie Therapeu-
tique 2 (1967) 33.
36. Becci, P.J., Hess, F.G., Gallo, M.A., Johnson, W.D. and Babish, J.G. Food Chem.
Toxico!. 21 (1983) 75.
37. Becci, P.J., Hess, F.G., Babish, J.G., Gallo, M.A. and Voss, K.A. Food Chem.
Toxicol. 21 (1983) 79.
38. Bagdon, R.E., Zbinden, G. and Studer, A. Toxicol. Appl. Pharmacol. 2 (1960) 225.
39. Zbinden, G. and Studer, A. Zeits. fur Lebens. und Forschung 108 (1958) 113.
40. Coulson, J. Chap. 7 in Developments in Food Colours - J (J. Walford, ed.), Applied
Science, London (1980).
41. Comforti-Froes, N., Varella-Garcia, M. and Silva, A.A. Alimentos e Nutriciio 4 (1992)
33.
42. COQper, R.M. Food and Drug Law. Food and Drug Law Inst., Washington D.e. (1991)
482.
43. Heywood, R, Palmer, A.K., Gregson, RL. and Himmler, H. Toxicology 36 (1985) 91.
44. Huang, H.S. and Goodman, D.S. 1. Bioi. Chem. 240 (1965) 2839.
45. Krause, R.F. and Sanders, P.L. Proc. Soc. Exp. Bioi. & Med. 95 (1957) 549.
46. Thomsen, K., Schmidt, H. and Fischer, A. Dermatologica 159 (1979) 82.
47. Hoffman-La Roche. Canthaxanthin. Unpublished report submitted to WHO. Cited in
JECFA 31st Report 1988 and 35th Report 1990.
48. Schwartz, J. and Shklar, G. Nutrition Cancer 11 (1988) 35.
49. Weber, U., Goerz, G., Kern, W. and Michaelis, L. Concepts in Toxicol. 4 (1987) 105.
50. Bluhm, R., Branch, R, Johnston, P. and Stein, R.I. Amer. Med. Assoc. 264 (1990)
1141.
51. Herbert, V. Amer. 1. Clin. Nut. 53 (1991) 573.
52. Bagdon, R.E., Impellizzeri, e. and Osadca, M. Toxicol. Appl. Pharmacal. 4 (1962) 444.
53. Nir, Z., Hartel, D. and Raveh, Y. Proc. First Inter. Symp. on Natural Colorants,
Amherst, MA (1993).
54. Ishidate, M., Sofuni, K., Yoshiyawa, K., Hayashi, M., Nohmi, M. and Matsuoka, A.
Food Chern. Toxicol. 22 (1984) 263.
55. Murthy, M.S., Mutation Res. 64 (1979) 1.
56. van Esch, G.J., van Genderen, H. and Vink, H.H. Zeits. Leben. Unter. Forsch. 111
(1959) 93.
57. Englebreth-Holm, J. and Iverson, S. Acta Path. et Microbiol. Scandinavia 37 (1955) 483.
58. Goodpasture, C.E. and Arrighi, F.E. Food and Cosrnetic Toxicol. 14 (1987) 9.
59. Ueno, S., Ayamoda, N., Kubota, K., Kurosawa, K. and Ishizaki, M. I. lap. Soc. Food
Sci. & Technol. 30 (1983) 172.
60. Grant, D. and Gaunt, I.F. Food Chern. Toxicol. 2S (1987) 903.
61. Noda, T., Shimiez, M., Yamada, A., Morta, S. and Ishibashi, T. Seikatsu Eisei 28
(1984) SO.
62. World Health Organization. Food Additive Series, No. 32 (1993).
63. Jensen, C.O., Howes, T.W., Spiller, G.A., Pattison, T.S., Whittam, J.H. and Scala, J.
Nutr. Rep. Int. 3S (1987) 413.
64. Perotti, A.G. Industrie Alirnentari 14 (1975) 66.
65. Bhavanishanker, T.N., Murthy, K.N. and Murthy, K.S. I. Food Sci. & Technol. 23
(1986) 287.
66. Abraham, S.K. and Kesavan, P.C. Mutation Research S3 (1978) 142.
67. Abraham, S.K. and Kesavan, P.C. Mutation Research 136 (1984) 85.
68. Abraham, S.K. and Kesavan, P.C. Mutation Research 143 (1985) 219.
69. Giri, A.K., Das, S.K., Taiukder, G. and Sharma, A. Cytobios 62 (1990) 111.
70. Hachiya, N. Akita I. Medicine 14 (1987) 269.
71. Ishidata, M., Sofuni, K. Yoshikawa, K., Hayashi, M., Nohmi, M. and Matsuoka, A.
Food Chern. Toxicol. 22 (1984) 623.
72. Jensen, N.J. Mutation Res. lOS (1982) 393.
73. Kawachi, T., Komatsu, T., Kada, T., Ishidate, M., Sasaki, M., Sugiyama, T. and
Tazima, Y. Applied Methods in Oncology 3 (19SO) 252.
74. Luck, H. and Rickeri, E. Food Chern. Toxicol. 30 (1992) 759.
75. Mortelmans, K., Haworth, S., Lawlor, T., Speck, W., Tainor, B. and Zeiger, E.
Environ. Mutag. 8 (1986) 1.
76. Sankaranarayanan, N. and Murthy, M.S. Mutation Res. 67 (1979) 309.
77. Seetharamaiah, G.S. and Chandrasahhara, N. I. Food Sci. & Technol. India 30 (1983)
249.
78. Tazima, Y. Appl. Methods in Oncology 3 (19SO) 252.
79. Vijalaxmi, V.Y. Mutation Res. 12 (1992) 915.
SO. Polasa, K., Sesikaren, B., Krishna, T.P. and Krishnaswamy, K. Food Chern. Toxicol.29
(1991) 699.
81. Egg Mason Research Inst. Report No. MRI-NPT 11-83-22. (1983).
82. Srimal, R.C. and Dhawan, B.N. I. Pharmacy and Pharmacol. 2S (1973) 447.
83. NTP Tech. Report No. 427 (1993).
84. WHO. Unpublished report from the Central Food Technology Research Inst. India.
Submitted by Chr. Hansen's Lab. (1980).
85. Brown, J.P. Mutation Research 7S (1980) 243.
86. Haveland-Smith, R.B. and Coombes, R.D. Food and Cosrnet. Toxicol. 18 (19SO) 215.
87. Ishidate, M. and Yoshikawa, K. Arch. Toxicol. Suppl. 4 (19SO) 41.
88. Kada, T., Tutikawa, K. and Sadiae, Y. Mutation Res. 16 (1972) 165.
89. Kornbrust, D. and Barfknecht, T. Envir. Mutag. 7 (1985) 101.
90. Loprieno, G., Boncristieni, G. and Loprieno, N. Food Chern. Toxicol. 30 (1992) 759.
91. Mori, H., Iwata, H., Tonaka, T., Morichita, Y., Mori, Y., Koyama, T. and Okumura,
A. Food Chern. Toxicol. 29 (1988) 585.
92. Sasaki, M., Sugimura, K., Yoshida, M. and Abe, S. Senshokutai 20 (1980) 574.
93. Ford, G.P., Gopel, T., Grant, D., Gaunt, I.F., Evans, G.J. and Butler, W.H. Food
Chern. Toxicol. 2S (1987) 897.
94. Battelle. Unpubl. report from the Battelle Memorial Inst. submitted to WHO.
95. Grant, D., Gaunt, I.F. and Carpenini, F.M.B. Food Chern. Toxicol. 2S (1987) 913.
96. Shulter, G. Naunyn-Sehrniedeberg's Arch. Pharmacal. 270 (1971) 316.
97. Chakravarti, I. Int. Congo Food Sci. & Teehnol. Abs. (1978) 290.
98. Chakravarti, I., Sreeder, R. and Roy, B.R. Indian J. Nut. & Dietetics 17 (1980)
377.
99. Ghosh, A. and Chakravarti, I. Indian J. Nut. & Dietet. 19 (1982) 310.
100. Ghosh, A. and Chakravarti, I. Indian J. Nut. & Dietet. 22 (1985) 147.
101. Ghosh, A. and Chakravarti, I. J. Food Sci. & Teehnol. India 25 (1988) 94.
102. Banerjee, T.S., Bhawmik, G., Yu, C.L., Swminathan, B., Giri, A.K., Shrivastava, S.
and Bhattachargee, S.B. Food Chern. Toxieol. 22 (1984) 677.
103. Chappel, C.1. and Howell, J.C. Food Chern. Toxieo/. 30 (1992) 351.
104. Licht, B.H., Shaw, K., Smith, e., Mendoza, M., Orra, J. and Myers, D.V. Food Chern.
Toxieol. 30 (1992) 565.
105. Myers, D.V. and Howell, J.e. Food Chern. Toxieol. 30 (1992) 359.
106. Licht, B.R, Shaw, K., Smith, e., Mendoza, M., Orr, J. and Myers, D.V. Food Chern.
Toxieol. 30 (1992) 375.
107. Houben, G.F., Abma, P.M.H., Berg, H. vanden, Dokkum, W. van, Loveran, H. van,
Pennicks, A.H., Seinen, W., Spanhack, S., Vos, J.G. and Ockhuizen, T. Food Chern.
Toxieol. 30 (1992) 749.
108. Adams, K., Allen, J.K., Brookes, P.C., Jones, E. and Proudlock, R.J. Food Chern.
Toxieol. 30 (1992) 397.
109. Allen, J.A., Brookes, P.C., Jones, E., Adams, K. and Reichold, M. Food Chern.
Toxicol. 30 (1992) 389.
110. Brusik, D.J., Jaganath, D.R., Galloway, S.M. and Nestman, E.R. Food Chern.
Toxieol. 30 (1992) 403.
111. MacKenzie, K.M., Carter, J.L., Petsell, S.R.W., Chappel, C.I., Emerson, J.L. and
Stanley, J. Food Chern. Toxieol. 30 (1992) 411.
112. MacKenzie, K.M., Boysen, B.G., Field, W.E., Petsell, S.R.W., Chapell, C.I.,
Emerson, J.L. and Stanley, J. Food Chern. Toxieol. 30 (1992) 417.
113. McKenzie, K.M., Boysen, B.G., Field, W.E., Petsell, S.R.W., Chappell, e.1.,
Emerson, J.L. and Stanley, J. Food Chern. Toxieol. 30 (1992) 431.
114. Hong-Sik, C. and Chang-Long, L. J. Korean Soc. Food Sci. and Nut. 22 (1993) 246.
115. Kim, H.S., Chang, U., Son, e.H., Bae, J.C. and Yu, J.K. Korean J. App. Micro. &
Engin. 9 (1981) 117.
116. Jahn, A. Mitt. der Bund. fur Fleisch, Kulrnbaeh 114 (1991) 379.
117. Fink-Gummels, J. and Leistner, L. Fleisehwirtschaft 69 (1989) 115.
118. Youngsmith, B. and Tabloka, W. Kasetsart J. 19 (1985) 45.
119. Yammono, T., Tsugimoto, Y., Noda, T., Shimizu, M., Ohmori, M., Morta, S. and
Yamada, S. Food Chern. Toxieol. 28 (1990) 515.
120. Yoshizumi, S., Okuyoma, H. and Toyoma, R. Shouhuhin Kogyo 23 (1980) 41.
121. Schwartz, S.J., Elbe, J.R, von, Pariza, M.W., Goldsworthy, T. and Pitat, Re. Food
Chern. Toxieol. 2l( (1983) 531.
122. Tarka, S.M. CRC Critical Reviews in Toxicology 9 (1982) 275.
123. Huschke, G. Chap. 8 in Natural Colours For Food and Other Uses. Edit. J.N. Counsel.
Applied Science, London (1979).
124. Dunkel, V.C., Zeiger, E., Brusick, D., McCoy, E., McGregor, D., Mortelmans, K.,
Rosenkrantz, H.S. and Simmon, V.F. Environ. Mutag. 7 (1985) 1.
125. Kanetmatsu, N., Hara, M. and Kada, T. Mutation Res. 77 (1980) 109.
126. Mikalsen, S.O., Rivadel, E. and Sanner, T. Carcinogenesis 9 (1988) 891.
127. Poole, A., Brown, R.e. and Rood, A.P. Brit. J. Exper. Path. 67 (1986) 289.
128. Tennant, R.W., Margolin, B.H., Shelby, M.D., Zeiger, E., Haseman, J.K., Spalding,
J., Caspary, W., Resnick, M., Staciewicz, S., Anderson, B. and Minor, R. Science 236
(1987) 933.
129. Zeiger, E., Anderson, B., Haworth, S., Lawior, T. and Mortelmans, K. Envir. and
Molecular Mutag. 11 (1988) 1.
130. Trochinowitz, RJ. J. Appl. Toxicol. 8 (1988) 383.
131. Roy, D. and Saha, J. The Eastern Pharmacist 24 (1981) 125.
132. Lehman, K.B. and Herget, L. Chemiker-Zeitung 51 (1927) 793.

133. National Cancer Inst. Tech. Report No. 97 (1979).
134. Heaton, N. The Chemical Trade 1. and Chern. Engineer 85 (1929) 439.
135. Lloyd, L.E., Rutherford, B.E. and Crampton, E.W. 1. Nut. 56 (1955) 265.
136. Fournier, P. C.R. Acad. Science (Paris) 231 (1950) 1343.
5 Chlorophylls and chlorophyll derivatives
G.A.F. HENDRY
5.1 Summary
Chlorophyll derivatives form a significant and growing element in the

range of natural pigments used as food colorants. An account is provided
of the methods of production and of the products formed under industrial
conditions. Consideration is given to the formation of chlorophyll under
biological conditions, together with a comprehensive summary of the
structure and occurrence of all known natural chlorophylls. Brief details
are provided of the function and natural environment of chlorophylls in
biological systems, with information on their relative stability in vivo and in
vitro, together with a description of the known major products of natural
and artifical degradation. The economic value of chlorophyll derivatives is
described, together with future prospects for exploitation.
5.2 Introduction
The chlorophylls are a family of naturally-occurring pigments present in

photosynthetic plants, including algae, and in some photosynthetic bac-
teria. As an integral part of vegetable foodstuffs, chlorophylls have formed
a component of the natural diet of animals, including humans, from
earliest times. Throughout recorded history, the presence, or absence, of
chlorophylls has been used as a sensitive indicator of the health and
ripeness of vegetables and fruit and the freshness of harvested produce. As
a guide to ripening, the presence of chlorophyll may be valid but to
freshness it can be misleading. Once plants are harvested the chlorophylls
are invariably degraded, in some species within a few hours, in others over
several weeks. Prolonged storage, particularly if associated with several
forms of deliberate preservation, usually completes the destruction of
chlorophyll to leave, often at best, a grey-brown colour indicative or
suggestive of loss of freshness. Further processing, particularly heating,
may result in the breakdown of any remaining chlorophylls. The wish to
modify food to restore the colour to that of the freshly-harvested crop is
understandable and has, over the years, given rise to the practice of

deliberately adding back natural, inorganic and later organic synthesized

pigments.
Chlorophylls are almost the only natural green plant pigment and
certainly the only ones in super-abundance. Unfortunately, the inherent
instability of chlorophylls in isolation has been a drawback to their more
widespread application as food colorants. The instability and degradation
of chlorophylls is a natural process in senescing leaves and ripening fruits
[1]. Whatever the evolutionary significance of this natural degeneration on
death, were chlorophylls not to be rapidly broken down in plants, the
billions of tonnes of chlorophylls produced each year on land and in the
oceans would accumulate with unthinkable consequences.
Given the long history of animal (and human) consumption of chloro-
phylls and of the many natural breakdown products of chlorophyll, there is
little evidence to suggest that this family of pigments is in any way harmful
following digestion, at least to the healthy herbivore. The few recorded
adverse effects of ingested chlorophylls arise in diseased animals [2] or
those with genetic defects such as albinoism [3]. The practice of improving
on nature's bounty, particularly the natural pigmentation of vegetables and
fruits, has evolved with the sophistication of the market place in the area of
preserved and processed foods. In part, this has been achieved by adding
back naturally-occurring plant pigments, notably the yellow-red carot-
enoids, the subject of several recent reviews [4-7]. Similarly the use of the
water-soluble red-purple anthocyanins and betalains is a growing practice
[8, 9]. Chlorophylls as food colorants have had less-detailed attention,
although short reviews on the structure and reactivity of chlorophylls have
been given in recent years [to].
The aim of this chapter is to provide a concise but up-to-date synopsis of
the principal chemical, biochemical and biological characteristics of the
chlorophylls, with some consideration given to the stability and breakdown
of chlorophyll derivatives, both in natural systems and as a food colorant.
An outline description is given of the preparation of the known chlorophyll
derivatives for industrial purposes, together with a discussion of the likely
future prospects for the commercial exploitation of the these pigments.
Finally, the opportunity is taken to introduce the revised nomenclature
for porphyrins (and chlorophylls) approved by the IUPAC-IUB Joint
Commission on Biochemical Nomenclature [11]. There may be merit in
reconsidering the current trade terminology surrounding chlorophyll-
derived food colorants, if only to bring these into line with current
scientific trends towards simplicity with precision in chemical nomen-
clature. Within the narrow context of chlorophylls as food colorants, the
important name changes are few. To those involved in research into the
chemistry and biochemistry of the porphyrins, the changes in nomen-
clature have involved the demise of many cherished, if rather quirky,
friends but are being increasingly adopted in the current literature.
CHLOROPHYLLS AND CHLOROPHYLL DERIVATIVES 133
(a) (b)
Figure 5.1 The numbering system of chlorophyll a showing: (a) the recently abandoned
Fischer system; and (b) the recommended IUPAC-IUB system.
5.2.1 Nomenclature
Before considering the natural and unnatural processes to which chloro-

phylls are exposed, it is necessary to introduce the changes in porphyrin
nomenclature made by the IUPAC-IUB Joint Commission [11].
These changes make clear the relationship between the chlorophyll
derivatives found in nature and those formed industrially. In the narrow
context of this chapter, the most notable change is the abandoning of the
porphyrin ring-numbering system (1 to 8 and a to 8) originated by Fischer
and colleagues [12]. In its place, a complete carbon numbering system
(from 1 to 24) has been recommended. The two systems are shown in
Figure 5.1.
Instead of a mixed system of Roman numerals for the outer carbons of
the pyrroles and Greek letters for the methene bridge carbons, the new
system numbers each and every carbon in sequence. The pyrrole rings
previously numbered I to IV are now lettered A to D, the fifth or
cyclopentanone ring being lettered E. Side-groups take their number in
sequence from the originating alkyl group. Thus the two carbons of ring E,
previously called positions 9 and to, are now 13 1 and 13 2 .
Chlorophyll a and many of its derivatives have long been known as
chlorins, which, by definition, are dihydroporphyrins. In biology, the
reduction occurs on ring D, thus chlorophyll a is a 17 ,18-dihydroporphyrin.
All of the naturally-occurring chlorophylls have a fifth or cyclopentanone
ring E (Fischer's ring V) and are based on the structure known as
phaeophytin. Thus, again, chlorophyll a would be based on 17,18-
dihydrophaeophytin. All natural chlorophylls are magnesium (Mg) coord-
ination complexes and the structures are defined, by earlier IUPAC rules,
by adding the ending 'ato', followed by the name of the metal and its
oxidation number. Thus chlorophyll a is now 17,18-dihydrophaeophytinato
Table 5.1 The nomenclature of chlorophylls and derivatives as recommended by the

IUPAC-IUB Joint Commission on nomenclature [11]
Retained trivial names Based on
Chlorophyll 17, 18-Dihydrophaeophytinato Mg(II)

Chlorophyll ide 17,18-Dihydrophaeophorbideato Mg(II)
Phaeophytin 17, 18-Dehydrophaeophytin
Phaeophorbide 17, 18-Dihydrophaeophorbide
Abandoned names Recommended names
Protochlorophyllide Didehydrophaeophorbideato Mg(II)

Phyllin e6
Chlorin e6
Chlorin e4 Didehydrorhodochlorin with a specific
Purpurin 7 description of the substitute at C13 and 15
Purpurin 5
Rhodin equivalents of the above
Chlorophyllin
Mg(II) and if the Mg were replaced by copper (Cu), the pure compound
would be 17,18-dihydrophaeophytinato Cu(II). These names would come
into play principally in naming unusual porphyrins and, in this context, the
derivatives of chlorophyll. Fortunately, a substantial body of long-
established trivial names (as they are called) have been retained, while
some of the less-common derived porphyrins, often sequentially and
chronologically named at the bench in Fischer's laboratory 70 years ago,
have gone. Among these abandoned names are the various chlorins,
rhodins and erythrins. Examples of retained and discarded trivial names, in
the special context of this review, are shown in Table 5.1.
The term chlorophyllin, widely used in the food colorant industry, is not
an accepted name. Indeed the food colorant 'chlorophyllin' covers a range
of compounds identical to, or structurally related to, the porphyrins
previously known as chlorin e6, isochlorin e4, probably their 3-HO
derivatives, purpurins 5 and 7 and their corresponding rhodins. Presum-
ably the survival of the term 'chlorophyllin' in trade descriptions reflects its
desirable similarity to the word chlorophyll. Whether onomatopoeia or
malaprop, the acceptable (and indeed more accurate) name for several,
although perhaps not all, of the various industrial derivatives of chloro-
phyll would now be rhodochlorin. Individual chlorophyll derivatives might
be, for example, 15-methylrhodochlorin or 13-ethylester rhodochlorin. If
the unspecific term rhodochlorin were to be adopted as the name for water-
soluble 'chlorophyll(in)s', it would at least have the merit of greater
chemical accuracy and meaningfulness, particularly to new generations of
porphyrin chemists.
5.3 Chlorophylls under natural conditions
5.3.1 Function
To make clear the difference between chlorophylls in their natural and
their artificial (man-made) environments, it is useful to consider the
function of these pigments in nature.
In the intact plant, more particularly in subcellular organelles known as
chloroplasts, different groups of chlorophylls are associated structurally
with particular proteins or polypeptides, for the most part buried in the
hydrophobic environment of the chloroplast lipid membranes or thyl-
akoids. The greater part of the individual chlorophyll molecules are so
arranged, physically, to trap or harvest incoming particles of light in the
form of photons or visible-light energy. At full sunlight, groups of
chlorophyll molecules will capture about 40 to 45 photons per second. This
trapped energy causes each of the electron-rich chlorophyll molecules to
wobble or resonate, passing the resonance on to neighbouring chlorophyll
molecules, ultimately concentrating the energy into a smaller number of
specialized reaction centres that contain an epimer of chlorophyll a and
phaeophytin.
In these activated reaction centres, the now intensely concentrated
energy flips an electron to a higher energetic state before ejecting it.
Through an elaborate circuit of quinones and other electron carriers, the
electron, with associated protons, yields chemical reducing power. Almost
simultaneously, the electron imbalance in the reaction centres are restored
by further electrons formed by the splitting of water
(5.1)
which in turn also yields protons used to drive the formation of storable
chemical reducing power, together with oxygen, which is released into the
atmosphere. The chemical reducing power of nicotinamide adenine di-
nucleotide phosphate (NADPH) and adenosine triphosphate (ATP) are
then consumed in the reduction of CO 2 (with NO) or SO~-) to give rise to
simple carbohydrates, phosphorylated fructose and glucose, and sucrose
together with amino acids used largely in protein synthesis. The whole
process, photosynthesis, sustains life on earth by capturing energy from an
external source, the sun.
The role of the chlorophylls is central to the process of photosynthesis,
although other pigments, notably the xanthophylls, certain carotenes and
bile pigments, play an auxiliary but essential role. Photosynthesis itself is at
least 2600 million years old and, in all but its fine details, has probably not
altered significantly in that time. Considering the powerful selective forces
of evolution this is a remarkable longevity and says much about its
efficiency. Perhaps solely for this reason, the structure and biosynthesis of
the chlorophyll molecule has been highly conserved since earliest times.
The relatively modest differences between one type of chlorophyll and
another do not alter the basic function of the pigment in photosynthesis.
5.3.2 Structure of chlorophylls and chlorophyll-protein complexes

The chlorophyll molecules of higher plants consist of a reduced porphyrin
chromophore with a colourless 20C terpenoid, usually phytyl, side-group
(Figure 5.2). The chlorophyll molecule itself is relatively large; the
porphyrin head being about 1.5 x 1.5 nm in size and the phytol chain
being overall about 2 nm in length. The 10 double bonds and the extensive
ring structure allow electrons to become de localized (see chapter 1) over
three of the rings, giving numerous oxidation-reduction states that are
important for the transfer of energy from one excited chlorophyll molecule
to its neighbour. The bonding of Mg (rather than any other metal) is
critical to photosynthesis, the electronic structure of Mg enabling it to
change the electron distribution of the porphyrin ring to produce powerful
excited states. The side-groups, particularly the vinyl and methyl groups at
positions 3 and 7, dictate the predominant wavelengths of light that will be
absorbed (and so alter the perceived colour of the pigment). Substitutions
such as a formyl group at these two positions, as in chlorophylls band dare
sufficient to increase or decrease the blue and red light-absorption patterns
to give rise to a more yellow-green or more blue-green coloured pigment,
respectively.
Similar colour changes can be obtained by the oxidation of carbons 17
and 18 (so forming didehydroporphyrin) and disruption and oxidation of
the cyclopentanone ring, although these are biologically destructive
changes. Reduction of ring B to yield the tetrahydrochlorins (or bacterio-
chlorophylls) drastically alters the area of de localized electrons by remov-
ing a carbon double bond. The effect is to shift the red absorption peak
into the infra-red spectrum. Such pigments are no longer green but, to our
eyes, grey-pink.
In the living cell, within the discrete chloroplasts, the chlorophylls are
complexed with, but not covalently bonded to, one of a series of
polypeptides. Here, the whole chlorophyll-polypeptide complex is ordered
in particular locations in the chloroplast membranes or thylakoids, from
which water is excluded.
The chlorophyll-polypeptide (or protein) complexes are closely asso-
ciated with several types of carotenes or xanthophylls (depending on the
species of organism) together with a family of hydrophobic molecules the
tocopherols (in the human diet better known as vitamin E). These are
present at a concentration of about one or two molecules for every ten or
so chlorophylls. The carotenoids act to quench excess chlorophyll excita-
tion energy and also act as scavengers (or antioxidants) of highly reactive
Other arrangements in
Vn
\
-------;:711
I
chlorophylls ll., k and l1.
Porphyrin ring
----,~--- Cyclopentanone ring
Bond disrupted in alkali

Position subject to oxidation
Figure 5.2 The structure of chlorophyll a. The two positive charges on the central magnesium
(Mg2+) ion are balanced by two negative charges shared randomly among the four pyrrole-
nitrogens. The arrangements of the ten double bonds within the ring system may also
vary. Vn = CHCH2 ; Me = CH 3 ; Et = CH2 CH 3 . The delocalized system of electrons is
shown as ----.
oxygen radicals and singlet oxygen. The tocopherols (or tocophenyl-

quinones) also scavenge singlet oxygen and act as a terminal trap for
various damaging radicals and lipid peroxides. The tocopherols, in
essence, prevent the unsaturated-lipid environment of the chlorophylls
from becoming peroxidized or rancid. The oxygen radicals or high-energy
forms of oxygen, are an inevitable byproduct of photosynthesis arising, in
part, from the photo-splitting of water. They have the potential to
irreversibly oxidize and to destroy (decolorize) the chlorophylls in a
process known as photo-bleaching or photo-oxidation. In the intact healthy
cell, photo-oxidation is strictly controlled and damage limited to diseased,
environmentally stressed or ageing plants [13].
In the chloroplast thylakoid membranes, the greater part of chlorophyll
(at least in higher plants) is associated with a group of proteins that make
up two types of light-harvesting complexes (LHCs), designated LHCI and
LHCII [14]. The latter, on isolation, resolves into about seven polypep-
tides, three of which are known as LHCII. This complex alone holds more
than half of all the chlorophylls and about one-third of the total protein in
higher plant thylakoids. The remainder of the chlorophylls are associated
with some seven other complexes whose characteristics are less well
known. Within the major complex, the LHCII, each of the three poly-
peptides is associated with about 10 chlorophyll molecules, approximately
6 chlorophyll a and 4 chlorophyll b, together with 1 to 2 xanthophylls (the

precise number varies in different species). The chlorophyll a molecules
appear to surround the densely packed chlorophyll b. The three polypep-
tides, each with their complement of chlorophylls and xanthophylls, form a
single unit folded in a structurally constant way from one side of the lipid
membrane across to the other. Those parts of the polypeptide buried
within the hydrophobic environment of the lipids appear to have a helical
structure and to these the chlorophyll molecules are linked. The precise
form of the ligand bonding between the chlorophylls and the amino acids
of the helices is not known with certainty. The hydrophobic phytyl tail
probably dictates the precise orientation of each chlorophyll molecule in
the lipid membrane.
This brief outline should be sufficient to indicate that the environment in
which chlorophyll functions, during photosynthesis, is elaborate, highly
protected and comparatively stable. In this environment, the chlorophyll
molecule is relatively immune from photo-bleaching and from oxidation. It
remains green. It should be no surprise then that when such a photo-
reactive molecule as chlorophyll is extracted from the plant and isolated
from its associated radical and singlet oxygen scavengers (antioxidants),
the pigment becomes unstable and is rapidly destroyed in air.
5.3.3 Biosynthesis of chlorophyll a and other natural porphyrin

pigments
An outline of the biosynthetic pathway of the porphyrins and chlorophyll is
provided in Figure 5.3. Fuller accounts are available elsewhere [15, 16].
Among the various steps in the pathway, it is worth noting that the first
six, from 5-aminolaevulinic acid to protoporphyrin, are common to all
living organisms, bacteria, fungi, animals and plants, if only because all
such organisms contain haem proteins. Plants and some photosynthetic
bacteria contain chlorophyll in addition. Protoporphyrin is the branch
point of the haem and chlorophyll pathways, and may be chelated to iron
to form protohaem-the prosthetic group of many haem proteins. Most of
these haems are a deep-red colour and include the familiar red chromo-
phore of haemoglobin and myoglobin of blood and muscle. In plants,
protoporphyrin is also chelated to magnesium before undergoing a series
of elaborations including cyclization of the side-groups at positions 13 and
15 to form the cyclopentanone ring E. In the presence of light, the
porphyrin is reduced (at positions 17 and 18) to form the class of
compounds known as chlorins. Final elaboration takes place in the form of
esterification with the acyclic diterpenoid phytol (geranylgeraniol or
farnesol in some chlorophylls). At this or possibly at an earlier step, the
chlorophyll molecule undergoes its final insertion into the polypeptide-
lipid environment of the thylakoid membranes.
Comment
5-Amino laevulinic acid Universal precursor of all
!
Porphoblilinogen
biological prophyrins
A mono-pyrrole
J.
Uroporp ynnogen III
CU
Turacin In feather of certain birds
Co
Cobalamin Prosthetic group vitamin B 12
Fe
Sirohaem Primitive haem protein
Coproporphyrinogen III
1
Protoporphyrinogen IX
Zn
• Z-bilin Bird egg-shells
~
Protoporphyrin
(Protoporphyrin IX) Prosthetic group of haemo-
Fe globin, myoglobin, cytochrome
Protohaem
!
b. peroxiases, catalase etc.
Mg
Haem a Cytochrome a
!
Haemc Cytochrome c,f
Protoporphyrinato
methylester Mg II
1
Fonnation of ring E Involves several steps
D'd<h ,1,,",,""00"',"'"," Mg II Accumulates in dark-grown
j
(Protochlorophyllide) plants
2", + ligh<
Phaeophorbideato Mg II Does not accumulate

(Chlorophyllide)
1Phy<o]
Phytyl group is a 20T
Chlorophy 11 or terpenoid
Phaeophytinato Mg II
Figure 5.3 The porphyrin biosynthetic pathway showing some of the adaptations that have
arisen during the course of evolution (adapted from [15]).
Regulation of the biosynthetic pathways is complex and involves internal

control by simple inorganic molecules, by porphyrins themselves as well as
the external stimulus, light. In addition, many steps in the pathway operate
in close association with the synthesis of otherwise unrelated molecules.
The formation of chlorophyll a is coordinated to the synthesis of several
thylakoid polypeptides, to the tocopherols, phytol and carotenes.
With such a long-conserved, ancient, biosynthetic pathway it is not
surprising that evolution has thrown up the occasional variation on an
otherwise constant theme [15]. For example, where cobalt (Co) rather than
iron (Fe) or magnesium (Mg) is chelated to a particular porphyrin, after
modification the end-product is cobalamin, the prosthetic group of vitamin
B12, synthesized by certain anaerobic intestinal microbes. Protoporphyrin
itself and zinc-protoporphyrin constitute the pink-brown and blue pig-
ments on many bird egg shells. More bizarre perhaps is the existence of a
natural copper porphyrin confined, however, to the purple-brown flight
feathers of the tropical Musophagidae family of birds. At least humans are
not the only ones to have exploited copper porphyrins! Numerous other
unusual porphyrins exist, including a vanadium-substituted haem in brown
algae [17]. The green pigment in certain water beetles, although not a
chlorophyll derivative, is a related pigment, biliverdin IX [18], more likely
to be a haem derivative (see chapter 6).
Nevertheless, the main products of the porphyrin biosynthetic pathway
are a limited range of pigments whose function is fundamental to such
processes as oxygen transport (haemoglobin), electron transfer (cyto-
chromes), oxygen radical destruction (through haem-peroxidases) to pho-
tosynthesis (the chlorophylls). The end-products are all highly coloured,
ranging from the avian purples and blues, to the reds and mauves of haem
pigments and the greens and blue-greens of the chlorophylls. It is no
wonder that poets [19] and scholars [20] alike should have marvelled at this
paint box of natural colours.
5.3.4 Types of naturally-occurring chlorophylls

Several different types of chlorophylls have been described and it is likely
that more will follow with further studies of certain less well-known algal
groups. In addition, a further family of related pigments is present in
certain photosynthetic bacteria. These bacteriochlorophylls are 7,8,17,18-
tetra hydroporphyrins. In Table 5.2 the distribution of all recorded
chlorophylls and bacteriochlorophylls is given and in Table 5.3 the
structures and spectral characteristics are provided. Further more-detailed
information is available [22-28].
All land plants, from mosses through to flowering plants, contain just
chlorophylls a and b (see Table 5.2), as does one group of photosynthetic
bacteria, the Prochlorophyta. (There has been much speculation that the
Table 5.2 Distribution of naturally-occurring chlorophylls and bacteriochlorophyllsa
Chlorophyll/ Organism
bacteriochlorophyll
Chlorophyll:
a All oxygen-evolving photosynthetic organisms including all higher
plants, all algae and the photosynthetic bacteria Cyanophyta and
Prochlorophyta
b High plants, the algal Chlorophyta and Euglenophyta and the
bacterial Prochlorophyta
c Algal Phaeophyta (brown algae), Pyrrophyta (dinoflagellates),
Bacillariophyta (diatoms), Chrysophyta, Prasinophyta and
Cryptophyta
d In some Rhodophyta (red algae) and Chrysophyta
e Reported from algal Xanthophyta
Bacteriochlorophyll:
a and b Purple bacteria including Chromatiaceae and Rhodospirillaceae
c, d and e Green and brown sulphur bacteria including Chlorobiaceae and
Chloroflexaceae
a Nomenclature follows ref. 21.
genus Prochloron may be directly related to the ancestral organism that

evolved to form the chloroplasts of higher plants.) The remaining chloro-
phylls c, d and e have been found only in algae, including the brown and
red seaweeds, as well as the single-celled marine algae making up the
phytoplankton of the oceans. The chlorophylls are also found in certain
photosynthetic bacteria, including the widespread blue-green Cyanophyta;
other bacteria possess tetrahydroporphyrins or bacteriochlorophylls.
Structurally (see Table 5.3), the variations in the ten known chlorophylls
and bacteriochlorophylls are relatively modest. The most common varia-
tions are at positions 3, 7 and 8 and when compared to chlorophyll a, many
of these variations involve an oxidation. Alternatives to the phytol group at
position 17 are found in some photosynthetic bacteria and algal groups.
The overall impression, however, is that much the greater part of the
original Mg-protoporphyrin structure has been conserved throughout all
known photosynthetic organisms.
5.3.5 Turnover and degradation of chlorophylls in biology

For all its relative stability in un diseased tissue chlorophyll is subject to
natural turnover, involving simultaneous destruction and synthesis, prob-
ably throughout the life of the organism. The rates of degradation are
considered to be fastest at the initial greening-up phase of the seedling [29]
and at the end of the life-span during leaf senescence [1] and fruit ripening
[30]. The latter processes, involving a net loss of chlorophyll, are
accompanied by profound physical changes to the immediate environment
of the pigments. These changes will include alteration to the fluidity of the
Table 5.3 Structures of all major chlorophylls and bacteriochlorophylls
Pigment Position (see Figure 5.1)
3 7 8 17 18 20
Dihydroporphyrins:
Chlorophyll a Vn Me Et PhyH MeH
Chlorophyll b Vn CHO Et Phy H MeH
Chlorophyll c CHO Me Et Phy H MeH
Tetrahydroporphyrins:
Bacteriochlorophyll a COCH 3 MeH Et Phy H MeH
Bacteriochlorophyll b COCH 3 MeH CHCH 3 H Phy H MeH
Bacteriochlorophyll c HO-Et MeH CH 2CH 3 H Farn H MeH Me or Et
Bacteriochlorophyll d HO-Et MeH CH 2CH 3 H Farn H MeH
Bacteriochlorophyll e HO-Et CHOH CH 2CH 3 H Farn H MeH Me or Et
Porphyrins:
Chlorophyll Cl Vn Me Et Acr Me
Chlorophyll C2 Vn Me Vn Acr Me
(i) Positions: 2, 12, 18-Me throughout, except B'chl e 12 position Et; 13, 15-CP throughout;
other unspecified positions unsubstituted.
(ii) Abbreviations: Me, -CH3;Et, -CH2 CH 3 ; Vn, -CHCH 2 ; Acr, -CHCHCOOH; Cp,
cyclopentanone; Phy, phytol; Farn, farnesol.
(iii) Original data derived and modified from ref. [1].
thylakoid membrane and peroxidation of membrane-bound thylakoid

proteins [31], almost certainly as a result of increased attacks from various
forms of activated oxygen. Just as the assembly of chlorophyll into the
photosynthetic pigment is elaborate, so is its disassembly. Unfortunately,
much less is known about the order in which disassembly takes place and
even less about the precise mechanisms by which chlorophyll is turned
from a stable green pigment to a colourless compound, often over a few
hours or days. The phase of destruction is usually much shorter than the
period of net synthesis. For example, in the northern temperate areas,
deciduous trees develop green leaves over a period of several days, often
weeks, and may bear fully green leaves for up to 5 months, during which
time chlorophyll turnover is probably extremely slow. But in early October
most, if not all, of that chlorophyll is rapidly destroyed with a half-life of as
little as 3 to 4 days [1].
The nature of the colourless end-products were until recently unknown.
In aquatic systems, and particularly under anoxic conditions, the destruc-
tion of chlorophyll is considerably slower, sometimes with identifiable
products, often phaeophytin, being formed in substantial amounts.
5.3.6 Natural, and some unnatural, chlorophyll derivatives

Despite the rapid destruction of the chlorophylls in senescence and on fruit
ripening, naturally-occurring breakdown products (as opposed to artifacts)
can sometimes be detected under ce~tain conditions. For example, grasses
when cut for hay, rapidly form the blue-grey colour of phaeophytin.
Grazing of chlorophyll-containing algae, in the oceans, gives rise to
phaeophorbides, detected mainly in the faecal pellets of the grazers. Some
of the phaeophorbides survive to sink to the ocean floor, where they may in
time contribute to the most long-lived of chlorophyll derivatives, the geo-
or petroporphyrins of oils and coal deposits [32]. In addition, numerous
in vitro experiments have documented the conditions required to bring
about particular modifications to the original chlorophyll molecule. Piecing
all this evidence together, it is possible to draw up a somewhat simplified
flowchart (see Figure 5.4) of the routes by which chlorophylls in natural
and man-made conditions become altered and ultimately destroyed.
The action of weak acids on chlorophylls in vitro will readily bring about
the loss of Mg to form phaeophytins. In vivo there may be an enzyme
responsible for this demetallation [33]. Under natural and industrial
conditions, in some plant species but not others, the action of the lipid
protein chlorophyllase readily cleaves the phytyl ester to yield chlorophyl-
lide (dihydrophaeophorbidato Mg(II».
Acidification would give rise to the demetallated phaeophorbide, the
starting material for most of the chlorophyll food colorants. In biology,
however, leaf senescence is associated with characteristic changes in
optical properties of the chlorophylls, including reflectance, and which
appear to signal the onset of chlorophyll destruction [34]. The quest over
the past few years has been to identify the nature of these transient
pigments and their ultimate fate. One approach has been to induce
senescence in green algae in culture by, for example, nitrogen starvation.
Under such conditions, algal cultures of Chiarella excrete red pigments
simultaneously with the disappearance of chlorophyll. The major product
appears to be a bile pigment (phycobilin) cleaved between rings A and B
with the release of the methane carbon [35]. But the situation in land plants
must be different if only because these plants do not excrete waste
products. Considerable progress in studying chlorophyll in plants has been
made in recent years in Matile's laboratory [36, 37]. Unlike the algal
system, cells of grasses maintain large vacuoles. Chlorophyll breakdown
products are converted to relatively water soluble forms and are actively
transported from the chloroplast to the vacuole. The process of chlorophyll
degradation is initiated by the action of chlorophyllase, at least in the
grasses studied so far. This initial solubilization step is followed by
oxidative cleavage of the bridge between rings A and B but without loss of
the methine carbon. In isolation and in the presence of acid the otherwise
colourless linear pyrrole can be detected as rust-coloured pigments.
Presumably these pigments are further degraded in the course of senes-
cence.
Much of the global production of chlorophylls is ultimately grazed and
pass through one, or more, animals' guts. In the oceans, grazing leads to
Chlorophyll or
Chlorophyllase PhaeoporphinatO~Mg
II -:---aCe!'d
ak unknown~~::~c~
Phytol Mg reactions
*Phaeophorbideato Mg II *PhaeOPhytin!
or Chlorophyllide *Various
Chlorophyll
Mg Phytol Strong ring E
Weak acid isomers
j
acid
*Phaeophoribide
[Cu-free oil SOI::~:loroP~hYll

~:: *Low molecular
weight
solvation colourless
effects compounds
Heat *Phaeophorbideato Cull

Heat
Anoxia [oil-soluble
Time *Phytoporphyrin Cu-chlorophyll]
H,:l~"g, ri"gE~
(Phylloerythrin)
Didehydro
rhodochlorins
(chlorin e6 '
*Decarboxy-
purpurins and others)
deoxophyto-
I j
porphyrin Phyllophorbide
(Dioxophyllo- (Pyrochlorin)
erythroaetio-
porphyrin)
and other Gut Na-didehydro
Aetioporphyrins digestion rhodochlorinato
[Water-soluble Cu-free
chlorophyll (in)]
jCu+~id
Low molecular
NiorV=O weight colourless
compounds
*The geo- or Na-didehydro

petro- rhodochlorinato Cu II
prophyrins of [Water-soluble Cu
oils and coal chlorophyll (in)]
Figure 5.4 Natural and unnatural pathways of the degradation of chlorophyll. Fischer
nomenclature in parentheses, industrial chlorophyll-derivatives in brackets. 'Products
formed naturally during plant senescence or natural decomposition.
the formation of phaeophorbide, although again the major product

appears to be a colourless unknown (Figure 5.4). On land, among those
strict herbivores whose diet is composed of the green parts of plants, much
of the ingested chlorophyll is excreted as the decarboxylated phaeophor-
bide derivative phytoporphyrin. Having survived grazing and photo-
destruction, some remaining chlorophylls will be deposited in oxygen-
depleted conditions, such as the ocean floor or on a water-saturated peat
bed. Undisturbed over several tens of millions of years, given the right
physical and chemical environment, including prolonged heating, even-
tually those chlorophylls may emerge with considerable modifications, to
contribute to the make-up of lignite, coal, shale, bitumen and oils.
Characteristically, the porphyrin skeleton, having long lost its Mg, is
chelated with nickel (Ni) or, under certain conditions, with a vanadyl
group (V=O). Much is known of such compounds [32] because these
contaminants can add to the costs of cracking during petroleum produc-
tion, as well as having a diagnostic significance for the exploration for new
reserves of fossil fuels.
To all these many products of degradation, the. food, cosmetic and
pharmaceutical industries have added a few more, particularly during the
early stages of isolation and concentration of chlorophyll derivatives.
Figure 5.4 also provides a highly condensed account of some of these
events (end-products' names in square brackets). Phaeophorbide itself, in
the appropriate solvent, may be marketed as an oil-soluble Mg-free
pigment. Acidification in the presence of copper (Cu) acetate or other salts
gives the oil-soluble Cu phaeophorbide (dihydrophaeophorbidato Cu(II)).
The preparation of the water-soluble derivatives involves several steps,
which aim to achieve a controlled cleavage of the cyclopentanone ring.
Among the many products are a class of porphyrins that may best be
described, generally, as di(de)hydrorhodochlorins. This mixture on acidifi-
cation ultimately yields the soluble sodium (or potassium) salt or copper
rhodochlorins (the so-called copper chlorophyllins).
In porphyrin chemistry, few degradative steps, however well controlled,
yield only one product. Part of the reason for the multiplicity of products
may be due to the ease with which chlorophyll molecules in solution form
aggregates that can be both complex and highly variable. Such randomized
aggregates will have different side-groups exposed to (or hidden from) the
external effects of solvents and will respond, chemically, in different and
seemingly random ways. An analogous situation arises in the random
formation of lignins from lignin radicals where any two molecules are
highly likely to be structurally different from each other. Perhaps, as a
consequence, it is commonly observed in the laboratory that the recovery
of derivatives from an original concentration of chlorophyll is rarely
complete. Much, often a substantial part, of the original chlorophyll
disappears (spectrally) to form unknown colourless products. This alone
seems to be one of the few consistent patterns in studies of the degradative

pathway of the chlorophylls.
Table 5.4 Total net annual global pro-

duction of chlorophylls·
Environment Tonnes x 108
Terrestrial 2.92
Aquatic 8.63
Total 11.55
• From ref. [1].
5.4 Chlorophyll derivatives as food colorants
5.4.1 Sources of chlorophylls for food colorants

To date, most, if not all, chlorophylls entering the food-colorant market
are derived from land plants which, in recent decades, have included
lucerne, or alfalfa (Medicago sativa), nettles (Urtica dioica) or several high-
yield pasture grasses. One notable exception has been the use in Japan of
the green faeces of the silk-worm where the larvae feed on mulberry leaves
and excrete 13 2HO-chlorophyll, among other pigments [38]. Convenience
of harvesting, ease of pigment extraction, yield of useful derivatives of
chlorophyll and the value of waste byproducts as cattle feed, dictate the
choice of raw material. However, there is no evidence that the current
selection of species necessarily gives the highest yields. An additional
problem, in the temperate regions, is that the supply of raw material is
restricted, often being confined to a few weeks in the year. Outside the
summer period, the availability of suitable raw material is limited. In
contrast, by far the greater synthesis of the chlorophylls takes place in the
oceans, mainly in single-celled phytoplankton and, in the North Atlantic,
over much of the year. One candidate among the flowering plants would be
one of the water hyacinths such as Eichhornia crassipes. These sub-tropical
free floating aquatic plants under field conditions can outstrip most other
species in productivity with yields of up to 500 tonnes per hectare in just 4
months. Originally confined to tropical waters, water hyacinths have been
introduced as a benign ornamental plant to Southern Europe and North
America where they have escaped to become a serious weed of waterways.
In the author's experience, aquatic plants tend to have low activity of
chlorophyllase with the potential to yield un degraded chlorophylls follow-
ing even prolonged storage.
Estimates of the global production of chlorophyll are shown in Table
5.4. Some 75% of the total annual production of chlorophyll, globally,
takes place in the aquatic (and mainly marine) environment. Interestingly,
the estimated figure of approximately 11 x 108 tonnes of chlorophyll

production compares quite well with an independent estimate of 1 x 108
tonnes given for carotenoids [39]. In plants, the ratio of chlorophylls to
carotenoids by weight (not molarity) is about 5:1, giving a value for the
global natural production of carotenoids as about 2 x 108 tonnes annually.
Much the greater part, probably well in excess of 99%, of both the
chlorophylls and carotenoids is degraded within days of the death of the
organism. The traces that remain to be recovered in geological deposits are
an exceedingly small part of the whole [32]. Nevertheless, in terms of
pigment production in living plants, there is a super-abundance of both
chlorophylls and carotenoids. The supplies, whether terrestrial or aquatic
in origin, are self-renewing and, in terms of their economic value, have
been barely exploited.
5.4.2 Alternatives to chlorophyll a

Much the greater part of the industrially produced chlorophyll colorants
are derived from chlorophyll a, the principal (occasionally the only) form
of chlorophyll in all oxygen-evolving photosynthetic organisms. Some 20 to
30% of chlorophyll colorants may also be derived from chlorophyll b. In
all, some ten chlorophylls or bacteriochlorophylls have been described (see
Table 5.3) from biological systems. Each pigment has its own distinct
colour and these are listed, together with the spectral characteristics, in
Table 5.5.
Given adequate access to a supply of algae and with suitable isolation
techniques, it should be possible to prepare chlorophylls with spectral
characteristics ranging from yellow-green to blue-green. Derivatives of
these chlorophylls would be likely to produce orange or, under drastic
chemical conditions, even red colours.
Two other photosynthetic pigments exist, orange and blue coloured,
which are structurally related to the chlorophylls. Among the algae, the
Rhodophyta and Cryptophyta, together with the photosynthetic bacteria,
the Cyanophyta, produce large quantities of highly coloured phycobilins
(see chapter 6). These linear tetrapyrroles are derived not from chloro-
phylls but from protohaem [40], and are related to the mammalian bile
pigment bilirubin. In the living cell, these phycobilins are highly stable and
covalently bonded to certain proteins. Cleavage from these proteins yields
two classes of pigments, the orange-red erythrobilins and the blue
cyanobilins, of which there are several variants [41]. They have immediate
interest as they are water-soluble and relatively stable. Very large concen-
trations of these pigments are formed in algal cells; 20% or more of the dry
weight may consist of the phycobilin plus its protein.
Finally, for the sake of completion, there exists a third class of linear
tetrapyrroles, which are present in all plants including algae. This blue
.......
~
Table 5.5 Spectral characteristics of the naturally-occurring chlorophylls and bacteriochlorophylls
Spectra
Z
Soret Red >
C!
Pigment Solvent Colour Xmax EmM Xmax EmM References ~
t""
Chlorophyll a 80% acetone Blue-green 431.2 95.8 663.2 86.3 25 0
'"
Chlorophyll b 80% acetone Green 459.0 135.4 646.8 49.2 25 0
Chlorophyll CI 90% acetone Yellow-green 443.2 318.0 630.6 44.8 28 ~
(")
Chlorophyll C2 90% acetone Yellow-green 443.8 374.0 630.9 40.4 28 0
Chlorophyll d Ether Blue-green 445.0 97.8 686.0 117.8 22 t""
0
Bacteriochlorophyll a Methanol Grey-pink 365.0 53.9 772.0 42.0 26 ~
Bacteriochlorophyll b Acetone Brown-pink 368.0 795.0 22
Bacteriochlorophyll C Acetone Green 428.0 660.0 27 ~
Bacteriochlorophyll d Acetone Green 424.0 654.0 27
Bacteriochlorophyll e Acetone Green 456.0 646.0 27
pigment, called phytochrome, has a highly specific and dominant func-

tion as a light receptor, controlling many critical aspects of plant meta-
bolism and development. Its concentration in plants is exceedingly low
and for this reason it is unlikely to be of commercial interest.
5.4.3 Extraction, isolation and derivatization

Chlorophylls are usually extracted, not from freshly-harvested plants, but
from bulk-harvested material previously sun-dried or subjected to artifi-
cial, rapid, heat-usually at closely controlled temperatures. The subse-
quent yield of pigment from the dried plants, pellets or powder varies with
the stage of development of the plant species itself, the duration and
temperature of the drying process and the length of storage of the
dehydrated raw material. The drying process itself leads to the formation
of several degraded (euphemistically called 'altered') chlorophylls. Heat-
ing for 5 min at 70°C was found to produce at least six derivatives with
substantial amounts of colourless degradation products [36]. Different
products were formed at different pH and were dependent on the
availability of O 2 . Most of the identified products show alterations to the
13 1_ and 13 2-position side-groups on the cyclopentanone ring, indicative of
oxidative attack.
Primary extraction, under strict control, using an aqueous solvent such
as acetone or a chlorinated hydrocarbon, is followed by washing, concen-
tration and solvent recovery. The degradative effects of organic solvents
have been described by many authors [e.g. 34, 36, 42]. For example, plant
species rich in the enzyme chlorophyllase readily form chlorophyllides,
particularly during extraction with aqueous acetone. Other products,
similar to the 'altered' chlorophylls of heat treatment, are also formed
during contact with certain water-organic solvent mixtures in the presence
of air. For this reason, the ratio of solvent to plant material is critical
during industrial extraction. Overall, the yield of chlorophyll (and green
derivatives) may be no better than 20%, although much higher and lower
values are quoted [43]. During these early steps, other plant products,
including resins, waxes and fats released by the solvent extraction process,
are removed by differential fractionation.
The partially purified extracts containing phaeophytin, and other
degraded chlorophylls are subjected to further processing depending on
the desired end-product. Oil-soluble colorants can be secured after further
washing against water-immiscible solvents and are marketed as the metal-
free phaeophytin, or further acidified in the presence of copper salts to
form oil-soluble copper phaeophytin. The possibility of producing authen-
tic Mg-phaeophytin (that is, chlorophyll) exists, although the extraordinary
care needed to produce this makes the product comparatively costly and
creates further problems in maintaining its stability.
Water-soluble derivatives are prepared by saponification of the crude

mixtures of degraded chlorophylls and phaeophytins allowing the hydro-
phobic phytyl group to be displaced by sodium (Na) (or potassium (K».
The acidic groups at 13 1 and 13 2 of the cyclopentanone ring will also be
converted to the Na (or K) salt. After further fractionation and washing to
remove much of the lipophilic contamination, the chlorin (or porphyrin)
satt may be marketed as the metal-free grey-green water-soluble pigment.
Conversion to a stable green colorant is achieved by acidification in the
presence of copper salts. Outlines of the process have been provided [43],
although most producers of chlorophyU-based colorants will have pro-
cedures that vary in detail and in sequence from those described. While the
logic of the design of the procedures is clear, the actual process of
extraction of chlorophyll derivatives has been described as more of an art.
5.4.4 Structure of the derivatives

Chlorophyll derivatives, as food colorants, are usually sold as 'chloro-
phylls' or 'chlorophyllins' with or without copper, soluble in water or
insoluble, in various strengths (concentrations) and diluents. Several
products are marketed with a subsidiary description based on the word
'phaeophytin' , others use the base-word 'chlorophyllin'. The term chloro-
phyllin is broadly understood to refer to an unspecified product of alkaline
hydrolysis of chlorophylls [44] and may be further equated with chloro-
phyHide, also know as dihydrophaeophorbidato Mg(II) [10], or the term
'phytochlorin' [43]. The implication is certainly that chlorophyllin is one
substance. However, the number of products formed both during the
drying of the raw material, in the initial extractions and on saponification is
potentially large [45] and inevitably the term 'chlorophyllin' has only a
generic meaning. This is even more true of the word 'chlorophyll' in the
sense used in the food trade.
To summarize a complex series of reactions, the products marketed as
chlorophyUins are water-soluble salts of derivatives of phaeophorbide a
and b or aUomerized forms of Mg-free chlorophyl\s a and b, characterized
by disruption of the cyclopentanone ring with the presence of nil, one or
two carboxyl groups at positions 13 1 and 132 with oxidation of ring D and
possible reduction of the vinyl group at position 3. The products marketed
as oil-soluble chlorophyUs are perhaps less varied and, under carefully
controlled conditions, may be based substantially on phaeophytin a and b
and their epimers and isomers with more severe re-arrangements around
positions 13 1 and 132 • In so far as any name could accurately describe this
varied family of pigments, the most appropriate, and perhaps most
accurate term would be one based either on phaeophytin (the oil-soluble
pigments) or rhodochlorin (the water-soluble form) (see Figure 5.5).
Further precision, where necessary, could be obtained by additional
CH 2
II
CH
\
Figure 5.5 The structure of a rhodochlorin where R1 and R2 are simple derivatives or
substitutions to the positions 13 1 and 132 of the disrupted cyclo~entanone ring. R3 would be
either 2H+ or a metal ion such as Cu +.
qualifying terms such as disodium didehydrorhodochlorin-13,15-diacetic

acid.
5.4.5 Stability and instability

The origin of most of the industrial chlorophyll derivatives lies in the
demand for a green pigment, of natural origin, closely related to authentic
chlorophyll, but by necessity restructured to form a pigment that is stable
in isolation. This stability has been achieved by restoring not magnesium
but copper to the metal-free pigment. The Cu-substituted derivatives are
relatively unresponsive to light, due to the electronic structure of the Cu
ion. Neither are they readily decomposed by mineral acids [43]. In
contrast, the isolated and purified Mg-chelated derivatives are unstable,
particularly in the presence of even dilute acids, increasingly so in the light.
Even the metal-free derivatives such as phaeophytin and phaeophorbide
with an intact cyclopentanone ring have a propensity to become oxidized,
particularly on exposure to light. However desirable (from the point of
view of food colorant regulations) it may be to provide authentic Mg-
chelated 'natural' chlorophylls, these pigments are inherently unstable and
acquire short-term stability only at pHs of more than 7.0, under anoxic
conditions in the absence of light. Degradation to the level of phaeophytins
is to be expected, with the presence of oxygen actually hastening the
decomposition.
The biological explanation for the apparent instability of the Mg-
chelated porphyrins is that, in the excited state, these metallorporphyrins
are strong reducing agents and thereafter readily oxidized. Mg (and Zn)
porphyrins are, essentially, porphyrin di-anions and so lose an electron
following excitation by light as in photosynthesis (the biological reason) or
in photo-bleaching (the artifactual phenomenon). Following excitation,

the resulting oxidized chlorophyll (Chl+) undergoes a series of electronic
rearrangements that further add to excitation, or instability. If the electron
deficiency is not rapidly balanced, as it is in the photosynthesis, the
chlorophyll anion species will combine with oxygen. The precise details of
the order (if there is order) in the subsequent oxidative reactions are
unknown; the products degenerate to colourless compounds, usually
rapidly. However, a variety of intermediates of chlorophyll oxidative
breakdown have been detected from time to time [1, 10, 34-37,52], several
of which suggest that the first site of oxidative attack is the cyclopentanone
ring. Certain transition metals, such as Cu or Fe, are readily chelated to
porphyrins but the resulting complexes (copper porphyrins and proto-
haem) are largely inert photochemically, that is they do not readily lose
electrons on exposure to light energy. The apparent instability of Mg
porphyrins in vitro is a direct consequence of their functional photoreact-
ivity in vivo.
The stability of some porphyrins is legendary, however. Proof of this lies
in the occurrence of nickel and vanadyl porphyrins, derived from chloro-
phylls a and b, in ancient oil deposits [32]. Mg coordinate petro-porphyrins
have, however, never been reported. Nevertheless, from time to time
claims are made for the discovery of 'fresh' green plants exposed during
archaeological investigations. Some of these plant remains have been
preserved under the most exotic conditions, in one case buried in
waterlogged conditions (near anoxia) under thousands of tonnes of chalk
(thus no light and pH > 7.0) at the heart of the neolithic structure Silbury
Hill, Wiltshire, UK [1]. None of these archaeological finds, to the author's
knowledge, has even been shown to contain the original unaltered
chlorophyll in more than what may be barely detectable traces. Much the
greater part of such long-preserved pigments consist of the metal-free grey-
coloured phaeophytin a. Indeed, the fresh greenness, jubilantly reported
on exposure of archaeological plant remains, frequently turns to a dull-
brown colour within minutes, strongly indicative of oxidation of green-
coloured phenols.
5.4.6 Economics of chlorophyll derivatives

Estimates of the worth of the world production of food colorants vary. A
value of $150 to $200m in 1985 has been given [46], increasing at a rate of
about 10% annually. Other estimates [47] suggest that 10 to 20% of the
market might be accounted for by natural colorants giving a conservative
value for these of say $60m to $80m (at 1994 prices). This value covers
carotenoids (much the most important of the 'natural' colorants), antho-
cyanins, betalains, as well as chlorophylls. Informed estimates (see
Acknowledgements) place the UK production of chlorophylls as worth
$4.0m to $5.5m annually with the suggestion that this might represent up to
one-third of the world production. Given these estimates, this places a
world value on the sales of chlorophyll-derivatives at around $15m in 1994,
a figure that includes both food colorants as well as for pharmaceutical and
cosmetic usage.
From trade estimates, the major demand, in volume, is for the water-
soluble chlorophyll derivatives with specialized interest in the more costly
oil-soluble product. Much the greater part, perhaps 75% or more, of
industrially prepared chlorophyll-derivatives are used in the growing
demand for natural colorants in food and drinks, including dairy products,
edible oils, soups, chewing gum and sugar confectionery. Some 20% of the
industrial production, at least in Europe, is for the more volatile fashion-
led cosmetic and toiletry market, including herbal bath foams and gels,
shampoos, hair conditioners and soaps. A small but traditionally-based
demand exists in the pharmaceutical trade for chlorophyll-derivatives as
deodorants, mouthwashes and surgical dressings.
To put all this into perspective, if the estimate of $15m is accepted for
the worth of the chlorophyll colorants, how does this compare with one
other aspect of the economics of chlorophyll? Once a year, in early
October, the destruction of chlorophyll in the leaves of the trees of the
north-east United States of America creates a major tourist industry.
Estimates for just one state, New Hampshire, put the value of fall tourism
in 1990 as $350m! The chlorophyll colorant industry has some catching up
to do.
5.5 Future prospects
There are three areas in which recent scientific and technological advances
could have a significant effect both on production costs and on stimulating
demand for chlorophylls as food colorants.
In recent years, the advances in molecular biology have resulted in the
identification, isolation and engineering of particular genes that have
properties open for exploitation. Considerable progress has been made in
very recent years in the area of plant genetic transformations. In the
context of chlorophyll degradation, a gene that controls one of the early
steps in disassembly and degradation of chlorophyll, in the intact plant, has
been identified [48]. This sid-gene (senescence inducing determinant) has
been partially characterized in the grass Festuca pratensis of which a
mutant (BF993) lacks the ability to express the gene. The consequence is
that the mutant remains green on death, that is, it retains its chlorophyll
following the natural senescence of the plant. While the wild-type turns
yellow during senescence (following destruction of the chlorophyll), the
mutant retains its deep green colour. Although this may have interesting
commercial possibilities for perhaps sports grounds in summer conditions,

it may also have an application in the extraction of chlorophyll as a
colorant. One immediate attraction would be that a supply of green plant
material could be stored following the summer harvest but then used as the
raw material for extraction throughout the rest of the year. A more long-
term benefit might arise if the sid-gene from F. pratensis were to be
isolated, copied in vitro, and inactivated by perhaps inverting part of the
genetic sequence before transferring the now unexpressed or nonsense
gene into new plant species such as nettles or high-yielding fodder grasses.
Genetic lines from the seeds of these transformed plants would be selected
to provide 'evergreen' plant material. The benefit would come in providing
a year-round supply of desirable species of green plants and eliminating the
need to rely on dehydrated plant tissues with their attendant slow oxidative
losses of pigment during prolonged storage.
The second prospect lies in stabilizing the original Mg-chlorophyll (the
authentic natural product). The prospect of achieving this might involve
biochemically encapsulating the chlorophyll molecule in an environment
designed to quench the destructive effects of activated forms of oxygen
(radicals and singlet oxygen). The environment might, for example, retain
one of the natural chloroplast LHC polypeptides, a complement of
xanthophylls and tocopherols (vitamin E). Such a partial reconstitution of
the environment of the photosynthetic membranes has already been
achieved using photosynthetic bacteria [49]. The aim would be to provide a
product that was stable in air, acid and light and which was in its entirety
biologically 'natural'. Some encouragement for this prospect will come
from the common laboratory observation that isolated thylakoid mem-
branes remain remarkably green, in the light and air, for many days after
the complete destruction of chlorophylls in organic solvents.
The third prospect would be in the exploitation of different chlorophylls,
other than chlorophyll a, particularly where there is a possibility of
improved stability. There are, for example, several reports [50, 51] of the
unusual relative stability of chlorophyll c in biological systems and even in
zooplankton faeces. Indeed the indications from the literature [1, 52] are
that the order of stability of the chlorophylls in living and senescing systems
is:
chlorophyll c > b > a (5.2)
It is perhaps ironic that chlorophyll a, the least stable of the three
chlorophylls, has been the only one deliberately selected for exploitation.
Chlorophyll c is a discarded (indeed combusted) product of the alginate
industry, based on brown seaweeds. An alternative source of chlorophyll c
is single-celled phytoplankton (see Table 5.2), which may be adaptable to
large-scale continuous culture [53]. Chlorophyll c, unlike other natural
chorophylls, is not esterified at position 17. Without further modification,
this pigment is readily soluble in methanol and other relatively polar

solvents [26]. As to its suitability as a natural food colorant, chlorophyll
c-containing brown seaweeds have long formed part of the traditional diet
of Asiatic and some maritime European cultures.
The production and use of chlorophylls as food colorants has become
increasingly important in recent years. The challenging opportunities for
new production methods and new products may help to promote the wider
use of this colourful group of pigments. The challenges will bear on
chemists and biochemists and may provoke them to answer John Donne's
perspective lines:
'Why grass is green or why our blood is red
Are mysteries which none have reached unto.'
John Donne, 1612 [19]
Acknowledgements
The author wishes to acknowledge the considerable contribution made by

Philipe Matile and his group in the area of chlorophyll degradation in
recent years and to thank Dick Patrick, Fred Fost and Dr K. He\liwell for
discussions on the industrial aspects ofchlorophylls as food colorants.
References
1. Hendry, G.A., Houghton, J.D. and Brown, S.B. New Phytol. 107 (1987) 255.
2. Rimmington, C. and Quinn, J.r. Ondersteepoort J. Vet. Sci. Anim. Ind. 10 (1934) 421.
3. Lohrey, E., Tapper, B. and Hove, E.L. Brit. J. Nutr. 31 (1974) 159.
4. Counsell, J.N. 'Some synthetic carotenoids as food colours' in Developments in Food
Colours-l (J. Walford, ed.), Applied Science Publishers, London (1980) pp. 151-188.
5. Coulson, J. 'Miscellaneous naturally occurring colouring materials for foodstuffs' in
Developments in Food Colours-l (J. Walford, ed.), Applied Science Publishers,
London (1980) pp. 189-218.
6. Counsell, J.N. Natural Colours for Food and Other Uses, Applied Science Publishers,
London (1981).
7. Rayner, P.B. 'Colours' in Food Additive Users Handbook (1. Smith ed.), Blackie,
Glasgow (1991) pp. 89-113.
8. Taylor, A.J. 'Natural colours in food' in Developments in Food Colours-2 (J. Walford,
ed.), Elsevier Applied Science Publishers, London (1984) pp. 159-206.
10. Scheer, H. (ed.) Chlorophylls. CRC Press, Boca Raton, Florida (1991).
11. Merritt, E. and Loening, K.L. Eur. J. Biochem. 108 (1980) 1.
12. Fischer, H. and Orth, H. Die Chemie des Pyrrols-2, Akademische Verlag, Leipzig
(1937).
13. Price, A. and Hendry, G.A.F. Plant Cell Environ. 14 (1991) 477.
14. Thornber, J.P., Peter, G.F. and Nechustai, R. Physiol. Plant 71 (1987) 236.
15. Hendry, G.A.F. and Jones, O.T.G. J. Med. Genet. 17 (1980) 1.
16. Jones, O.T.G. 'Chlorophyll biosynthesis' in The Porphyrins VI (D. Dolphin, ed.),
17. Vi Iter, H. Phytochem. 23 (1984) 1387.
18. Kayser, H. and Dettner, K. Compo Biochem. Physiol. 778 (1984) 639.
19. Donne, J. On the Second Anniversary (1612).
20. Brown, S.B. Univ. Leeds Review 30 (1987) 1.
21. Holmes, S. Outline of Plant Classification, Longmans, London (1983).
22. Svec, W. 'The isolation, preparation, characterization and estimation of the chlorophylls
and bacteriochlorophylls' in The Porphyrins VC (D. Dolphin, ed.), Academic Press, New
York (1979) pp. 341-397.
23. Smith, K.M. 'Chemistry of chlorophylls: Synthesis' in The Chlorophylls (H. Scheer, ed.)
CRC Press, Boca Raton (1991) pp. 115-143.
24. Falk, J.E. Porphyrins and Metallorporphyrins, Elsevier, Amsterdam (1964).
25. Lichtenthaler, H.K. Methods in Enzymology 148 (1987) 340.
26. Jones, O.T.G. 'Chlorophylls and related pigments' in Data for Biochemical Research,
2nd edn. (R.M.e. Dawson, D.e. Elliott, W.H. Elliott and K.M. Jones, eds), Clarendon
Press, Oxford (1969) pp. 318-320.
27. Gloe, A., Pfennig, N., Brockman, H. and Trowtzsch, W. Arch. Microbiol. 102 (1975)
103.
28. Jeffrey, S.W. Biochim. Biophys. Acta 279 (1986) 2735.
30. Goldschmidt, E.E. 'Pigment changes associated with fruit maturation and their control'
in Senescence in Plants (K.V. Thimann, ed.), CRC Press, Boca Raton, Florida (1980)
pp.207-217.
31. Thompson, J.C., Legge, R.L. and Barber, R.F. New Phytol. 105 (1987) 317.
32. Baker, E.W. and Palmer, S.E. 'Geochemistry of porphyrins', in The Porphyrins fA
(D. Dolphin, ed.), Academic Press, New York (1978) pp. 485-551.
33. Ziegler, R., Dresken, M., Herl, B., Menth, M., Schneider, H. and Zange, P. Deutsch.
Bot. Gess. Meeting (Freiburg) (1982) Abstract 472.
34. Gitelson, A. and Merzlyak, M.N. J. Pl. Physiol. 143 (1994) 286.
35. Engel, N., Jenny, T.A., Mooser, V. and Gossauer, A. FEBS Letters 293 (1991) 131.
36. Matile, Ph., Schellenberg, M. and Peisker, Ch. Planta 187 (1992) 230.
37. Krautier, B., Jaun, B., Bortlik, K., Schellenberg, M. and Matile, Ph. Angew. Chem. 103
(1991) 1354.
38. Nakatani, Y., Orisson, G. and Beck, J-P. Chem. Pharmacal. Bull. 29 (1981) 2261.
39. Kearsley, N.W. and Katsabaxakis, K.Z. J. Food Technol. 15 (1980) 501.
40. Brown, S.B., Holroyd, J.A., Troxler, R.F. and Offner, G.D. Biochem. J. 194 (1981) 137.
41. Rudiger, W. 'Plant biliproteins' in Pigments in Plants (F.-e. Czygan, ed.), Gustav Fischer
Verlag, Stuttgart (1980) pp. 314-351.
42. Schoch, S. and Brown, J. J. Plant Physiol. 126 (1987) 483.
43. Kephart, J.e. Ecan. Bot. 9 (1955) 3.
44. Strain, H.H. and Svec, W.A. 'Extraction, separation, estimation and isolation of the
chlorophylls' in The Chlorophylls (L.P. Vcrnon and A.R. Sceley, eds), Academic Press,
New York (1966) pp. 21-66.
45. Strell, M., Kalojanoff, A. and Zuther, F. Arzneimittel-Forsch. 6 (1956) 8.
46. I1ker, R. Food Technol. 41 (1987) 70.
47. Blenford, D. Food 7 (1985) 19.
48. Thomas, H. Theor. Appl. Genet. 73 (1987) 551.
49. Hunter, e.N., Van Grondelle, R., Holmes, N.G. and Jones, O.T.G. Biochim. Biophys.
Acta 548 (1979) 458.
50. Jeffrey, S.W. and Allen, M.B. J. Gen. Microbiol. 36 (1964) 277.
51. Jeffrey, S.W. Marine Bioi. 26 (1974) 101.
52. Brown, S.B., Houghton, J.D. and Hendry, G.A.F. 'Chlorophyll degradation' in The
Chlorophylls (H. Scheer, cd.), CRC Press, Boca Raton, Florida (1991).
53. Dixon, G.K. and Syrett, P.J. New Physiol. 109 (1988) 289.
6 Haems and bilins
J.D. HOUGHTON
6.1 Summary
Haems and bilins provide some of the most visually apparent pigments in
nature, from blood-red to the azure blue of marine algae. Their colours are
associated with the natural world, with health and well-being, and have
been used for centuries to reinforce these associations in traditional food
and cosmetic preparations.
Man's interest in these pigments has also led to an accumulation of
knowledge of their chemical and biochemical properties, and a growing
research interest in the emerging field of molecular biology. Today, these
have an important role to play in the modern use of natural colours, yet
little of the scientific knowledge available concerning these pigments has
been made accessible to the commercial or industrial user. Nomenclature
of the natural or nature-alike pigments that are available is still inaccurate
and confusing, and sometimes bears little relationship to scientifically
accepted terms. In addition, much ignorance exists as to the true form of
these pigments in nature, and the justification for their modification in
commercial use.
It is hoped here to bring together the available information concerning
haems and bilins, in a form that is accessible to the industrialist and
research scientist alike, and to create a basis for interaction, which must
surely offer future benefits to all
6.2 Haems
6.2.1 Introduction
The term 'haem' is generally used to describe an iron chelate of the cyclic
tetrapyrrole protoporphyrin IX (Figure 6.1). Haems are ubiquitous in
nature and perform a central role in cellular energy transduction and
metabolism in all known species. They are represented in this role by a
great variety of apoprotein conjugates. All of these haemoprotein com-
plexes are pink or red in colour by virtue of their common prosthetic
group, but the structure most intensively studied is undoubtedly haemo-

(a) (b) (c)

CH,
I
h -CH, h -CH,
\ \
CH, <;:H,
CH, CH,
CO,H CO,H
Figure 6.1 The structures of protoporphyrin IX (b) and haem (c) with that of their common
porphyrin nucleus (a). The numbering of the porphyrin ring is according to the IUB-IUPAC
recommendations (1].
globin. It is the aim of this section to characterize the known physical and
biochemical properties of the haemoproteins with particular emphasis on
haemoglobin, and assess their present use and future potential as color-
ants.
6.2.1.1 Structure and nomenclature. The structure of haem is shown in

Figure 6.1, together with the correct IUB-IUPAC system of numbering of
the constituent carbon atoms [1]. The central iron atom of haem is
coordinated to the four pyrrole nitrogens of protoporphyrin IX to form an
almost planar metalloporphyrin ring.
The fifth and sixth coordination positions of the iron atom are perpendic-
ular to the ring and may be occupied by several ligands, including a donor
group from the surrounding apoprotein. Several trivial names persist,
which describe the valency of the central iron atom and/or the ligands to
which it is bound. Thus a haemin is a ferric (Fe III) haem in which a
halogen or other anion is coordinated. The term haematin is also used to
denote the ferric state of the central iron atom, and may be qualified to
indicate the nature of the ligand bound, for example, alkaline haematin.
This nomenclature is commonly used to describe the valency and ligand
states of haems in vitro, and is summarized in Table 6.1. In nature, most
haemoproteins perform their physiological functions in the ferrous (Fe II)
state, but may also exist in the Fe III form, and their names are sometimes
prefixed to denote this fact, for example ferrihaemoglobin/ferrohaemo-
globin. Other terms will be defined, where appropriate, within the body of
the chapter.
6.2.2 Haems in nature
6.2.2.1 Function and occurrence. Haem forms the prosthetic group of a

large family of proteins involved in several diverse functions throughout
HAFMS AND BILINS 159
Table 6.1 Nomenclature of some haems and haem derivatives
Name Synonyms Valency 5th ligand
Haem Ferroprotohaem IX Fe II H 2O
Protohaem
Ferroprotoporphyrin IX
Iron protoporphyrin IX
Reduced haematin
Haematin Ferriprotohaem IX Fe III OH
Hydroxyferriprotoporphyrin
Alkaline haematin
Haemin Proto haemin IX Fe III CI
Chlorohaemin
Chloroferriprotoporphyrin
the animal and plant kingdoms. It is most visually prominent as the red
blood pigment haemoglobin, which functions as an oxygen carrier, and
also in muscle as myoglobin where it performs the same role. Haem is also
central to the energy transducing mechanisms of the cytochromes, respons-
ible for the use of energy from photosynthesis and respiration, and to
several enzymes such as catalase and peroxidase, which use haem in their
structures. The importance of haem in these vital processes is manifest and
their role is so central to life processes that, along with the chlorophylls,
they have been referred to as 'the pigments of life' [3].
The variety of functions that the haemoproteins as a whole perform are
so wide as to be outside the scope of this chapter. It seems appropriate here
to describe the function of haemoglobin as an example of one particular
haemoprotein, and as an introduction to the structural and biochemical
studies on haemoglobin described later.
The evolution of oxygen-carrying systems was a major step in the
development of aerobic life on earth in overcoming the limitations to life of
the low solubility of oxygen in water. The presence of haemoglobin in
blood increases its oxygen carrying capacity by a factor of fifty. Haemo-
globin' also facilitates the removal of waste carbon dioxide from the blood.
The binding of both of these adducts is a function of the haem moiety of
the protein, and is achieved by a ligand bond to the sixth ligand position of
the central iron atom. The binding of these ligands is stabilized by the
three-dimensional structure of the surrounding apoprotein, in the presence
of which the degree of binding is controlled over the physiological range of
oxygen concentrations. This control is modified by the fact that haemo-
globin exists as a non-covalently attached tetramer of four protein
subunits. This tetrameric configuration allows cooperation of oxygen
binding between subunits, and produces a sigmoidal curve of binding at
increasing oxygen concentrations. By these means, haemoglobin is able to
respond extremely sensitively to small changes in the concentration of
available oxygen, and perform its role both as an acceptor and as a donor
of oxygen in the different physiological compartments of the body [4].
The functions of the other haemoproteins are as various as the
apoproteins to which they are bound. All, however, are involved in the
binding either of oxygen or of reducing power in the form of electrons.
Haemoproteins such as haemoglobin and the respiratory cytochromes act
mainly as carriers of oxidation or reduction potential, while others such as
the P-450 cytochromes, themselves use this potential to transform meta-
bolic substrates. The subject of diversity and similarity of structure and
function within the family of haemoproteins is one of great current
interest, and it seems likely that as our knowledge of these proteins
increases so will our understanding of the common themes in their make-
up.
6.2.2.2 Biosynthesis. Haems are produced from the biosynthesis of

protoporphyrin IX, in common with chlorophylls and bilins (see chapter 5
and section 6.3.2.2). In animals and most bacteria, the first committed step
in porphyrin synthesis is the condensation of the amino-acid glycine with a
metabolic precursor succinyl coenzyme A, to form 5-aminolaevulinic acid
(ALA). This reaction, carried out by the enzyme ALA synthase was
characterized as long ago as 1958, and has become known as the classic
Shemin pathway [5]. The next step in the biosynthesis of porphyrins is the
formation of the monopyrrole porphobilinogen from two molecules of
ALA by the enzyme porphobilinogen synthase. This step is followed by the
cyclization of four molecules of porphobilinogen to produce the cyclic
tetrapyrrole Uroporphyrinogen III. This intermediate contains six more
hydrogen atoms than are found in porphyrins, the presence of which
prevents conjugation of the electrons of the macrocyclic ring, rendering the
porphyrinogens colourless.
Consecutive alterations to the constituents at the periphery of the
porphyrinogen structure lead to the formation of protoporphyrinogen IX,
which is then oxidized to produce the coloured protoporphyrin. This is a
common precursor of both the haems and chlorophylls, and the pathway of
conversion of ALA to form protoporphyrin IX is remarkably constant in
the organisms in which it has been studied so far. A single enzyme,
responsible for the chelation of iron into the centre of the porphyrin
nucleus, results in the formation of free haem [6]. (This pathway is
summarized in Figure 6.2).
The free haem chromophore may then be specifically attached to its
apoprotein in one of two possible ways [7]. Firstly several haemoproteins
exist in which haem is covalently bound to the surrounding apoprotein to
produce a highly stable pigment-protein complex. Among these are the
c-type cytochromes, relatively ubiquitous in muscle tissue. In these
structures, the two propionic acid substituents at positions 13 and 17 of the
HAEMS AND BILINS 161
<;:O,H <;:O,H <;:O,H
..
<;:H, (a) <;:H, (b) CH, <;:O,H
<;:H, + ~H, <;:H, tH, CH,
~ \ I
0
<;:H, CO
~~Co.A CO,H 9 H,
Succinyl NH, N ...... CH NH
coenzyme A Glycine
H "
ALA
porphobilinogen
<;:O,H
CH,
tH,
\
uroporphyrinogen coproporphyrinogen
CHCH,
\
~ -CH,
\
CH,
tH,
CO,H
protoporphyrinogen protoporphyrin IX
Figure 6.2 The biosynthesis of protoporphyrin IX from glycine and succinyl coenzyme A.
The enzymatic steps shown are carried out by the following enzymes: (a) ALA synthetase,
(b) ALA dehydratase, (c) uroporphyrinogen synthetase and cosynthetase, (d) uroporphyrin-
ogen decarboxylase, (e) coproporphyrinogen oxidative decarboxylase, (f) protoporphyrin-
ogen oxidase.
haem chromophore are esterified to amino-acid residues of the surround-

ing apoprotein, to form a stable covalent attachment. The remaining
apoprotein structure is arranged around the chromophore to maximize
hydrophobic and hydrophilic interactions between the two. This arrange-
ment also presents a nitrogen atom from a histidine amino acid of the
apoprotein in the correct orientation to form a fifth ligand to the central

iron atom of the haem, further stabilizing the complex.
In a second type of interaction found in b-type cytochromes and
haemoproteins such as haemoglobin and myoglobin, the chromophore
protein complex is held together entirely by the non-covalent interactions
described above, that is by hydrophobic and hydrophilic forces, and the
presence of a ligand bond from a histidine nitrogen to the fifth ligand
position of the iron atom in haem. This structure, although stable in
nature, is considerably less resistant to chemical change in vitro than that of
the c-type cytochromes, and this aspect of their chemistry will be discussed
in section 6.2.3.1.
6.2.2.3 Metabolism. The metabolism of haem in mammalian tissues is a

well-documented and active research area [8, 9]. In human beings and
other vertebrates, endogenous haem is degraded by oxidative cleavage of
the porphyrin ring to yield the linear tetrapyrroles biliverdin IX a and
bilirubin IX a, in a process that releases carbon monoxide and iron (Figure
6.3). Ring cleavage of haem is a common step in the synthesis of both
animal and plant bile pigments (see section 6.3.2.2), and is catalysed by the
enzyme haem oxygenase. The product of this reaction is the green pigment
biliverdin IX a, a transient intermediate in the pathway of breakdown in
mammals, and relatively unstable as a free compound. The transformation
of this pigment by reduction to form the yellow bilirubin IX a is achieved
by the enzyme biliverdin reductase, and is followed by conjugation to
sugars to yield stable excretion products. The potential of animal bile
pigments themselves as colorants and for medicinal purposes is worth
mentioning here, and will be discussed further in section 6.2.3.3.
(a) (b)
P, Pr
haem bifiverdin bilirubin
Figure 6.3 The degradation of haem to the bile pigments biliverdin and bilirubin. The
reactions shown are carried out by the following enzymes: (a) haem oxygenase, (b) biliverdin
reductase. V = CHCH 2 , Me = CH3 , Pr = CH2 CH2CH2 0H.
It should also be stated that the metabolism of ingested haem, as

opposed to endogenous pigments, is most unlikely to occur via this
pathway. It is probable that most ingested pigment chromophores-
whether from haem, chlorophyll or bilins-pass through the human gut
with few chemical changes and little or no absorption of nutrient. This has
been proved extensively in the case of chlorophyll [10] except where
degradation by light is possible, for example in the digestive tracts of
transparent marine organisms close to the water surface. The metal ion
component of porphyrins may, however, be used--<:hlorophyll is reported
as being a major source of magnesium in cows and haem is well known as a
valuable source of 'available' iron and, in this respect, must be entirely
beneficial.
The protein component of the pigment-protein complexes will be
digested in the normal way following ingestion and will itself have a
nutritive value.
6.2.2.4 Physical properties. These can be summarized by their localiza-

tion, absorbance and stability.
6.2.2.4.1 Localization. As has already been alluded to, the distribu-

tion of haemoproteins both among species, and within a given organism is
virtually comprehensive. Almost no single tissue within an animal is
without its complement of haemoproteins, some soluble, some membrane-
bound but all essential to the function of the organism. Some tissues,
however, will obviously offer richer sources of haem pigments than others,
and by far the greatest source of haem in any mammalian tissue must be
that of haemoglobin. Comprising about 14 g in every 100 ml of blood, and
readily available from abattoirs, it is tempting to overlook the possibility of
extraction from other sources. Firstly, certain tissues such as spleen and
liver have a very high haemoglobin content themselves, and methods for
extraction from these tissues are sometimes more convenient than those
for whole blood (see section 6.2.3.1). Secondly, the possibility exists of
isolating different haem compounds from tissues, which may have advant-
ages in terms of colour or stability, over haemoglobin itself. The b-type
cytochromes, which are membrane bound and present generally in small
concentrations in the tissues in which they are found, are unlikely to
provide practical sources of haem in quantity. Also, in terms of stability
and absorbance, they are rather similar in their properties to haemoglobin.
The soluble c-type cytochromes may, however, offer certain advantages,
and can be found in relatively high concentrations in specialized tissues.
Thus, for example, heart muscle which is particularly rich in cytochrome c,
could provide a source of highly stable, soluble haem of specific absorb-
ance characteristics rather different from those of haemoglobin itself.
6.2.2.4.2 Absorbance. The absorbance properties of all porphyrins

are generated by the conjugated system of double and single bonds that are
characteristic of the molecules. Thus the aromatic porphyrin nucleus of
haems and chlorophylls produces strongly coloured pigments with high
extinction coefficients for their main absorption bands. In addition to the
absorption bands in the visible region of the spectrum, there is also an
intense band around 400 nm called the 'Soret' band, which is a characteris-
tic of the macrocyclic ring conjugation. This band is lost, for example,
following cleavage of the porphyrin, to yield linear tetrapyrroles or bilins
[11].
The absorption spectra of the porphyrins vary with the nature of the
porphyrin itself, and with the various substituents upon it. There is a strong
theoretical basis to describe these changes, but the subject has been
admirably described in qualitative terms elsewhere [11] and readers are
directed to this work for a fuller explanation. There are significant
differences between the absorption spectrum of say haemoglobin in blood
with that of its chromophore haem in organic solution, and consideration
of absorption properties in vitro will be given in section 6.2.3.1.
The absorbance characteristics of haems in vivo depend on the nature of
the apoproteins to which they are attached, the means by which this
attachment is achieved, and the presence of additional ligands binding to
the central iron atom. Thus haemoglobin has a characteristic absorption
spectrum similar to that of myoglobin, but relatively dissimilar to that of
the c-type cytochromes [4]. In addition, the spectrum of haemoglobin even
in vivo is dependent on the presence of oxygen (or carbon dioxide)
occupying the sixth ligand position of the central iron atom. Thus partial
pressures of oxygen sufficient to oxygenate all of the haemoglobin present
will result in a shift of the absorbance to form two discrete maxima, one at
higher and one at lower wavelength to that of the single peak of
deoxyhaemoglobin (Figure 6.4). Under normal circumstances in vivo, a
proportion of the haemoglobin would be present in both forms, and the
observed spectrum would be a rather broad peak representing an average
of the two states. It is normal practice for the purpose of quantification to
measure the absorbance spectra of haems in the reduced form, or as a
difference spectrum of the reduced minus the oxidized form. The absorb-
ance characteristics of the b- and c-type cytochromes obtained by these
methods are given in Table 6.2 for comparison with those of haemoglobin.
6.2.2.4.3 Stability. The stability of haems in vivo is of some interest, as

in mammals there appear to be two separate pools of haem, subject to
turnover at different rates. The first pool has a life-span of between 1 and
3 h and is thought to consist of haems present in cytochromes and the
haem-containing enzymes. The second pool represents haemoglobin
haem, and is remarkable in its long life-span of 60 days for rats and 120
1.25
1.0
,,
,
,
I
,,
0.5 \
,
\
~~50~--~5~070----~55~O~----6~O-O~~~
Wavelength (nm)
Figure 6.4 The visible absorption spectra of haemoglobin as oxyhaemoglobin (..•• ) and
deoxyhaemoglobin (---).
Table 6.2 Absorbance characteristics of globins and cytochromes
Name Spectrum Am." Am;n (nm) EmM
Haemoglobin reduced 555 12.5

oxidized 535 14
575 15
Myoglobin as haemoglobin as haemoglobin
Cytochrome b reduced--oxidized 559,570 20
Cytochrome c reduced--oxidized 550,540 20
Cytochrome a reduced--oxidized 605,630 16
days for humans [9], which illustrates the great natural stability of
biomolecules in vivo. The task for the biochemist must be to equal this
achievement in the laboratory.
6.2.3 Free haems
6.2.3.1 Physical properties. These are divided into (i) extraction and
purification, (ii) absorbance, (iii) stability, and (iv) quantitative analysis.
6.2.3.1.1 Extraction and purification. It should be mentioned here

that the absorbance properties of haems are largely maintained during
processes such as freeze-drying of tissues and of blood itself, without the

necessity for extraction or purification procedures. As many of these
sources are themselves foodstuffs, it may sometimes be appropriate to use
the colorants directly from the crude product. The disadvantages of this
approach are usually associated with problems of stability, and the
associated changes in spectral properties. The advantages of increased
stability and the reproduction of particular spectral characteristics often
justify the use of extraction and purification, and these processes will now
be described.
Extraction of haems from proteins such as haemoglobin, to which they
are not covalently attached, is relatively easily achieved by the use of
acidified organic solvents [11]. Several biological materials such as blood
and tissue samples may be treated this way, and with a choice of solvents
appropriate to the proposed use. Ether-acetic acid or ethyl acetate-acetic
acid are the most commonly used solvent mixtures, with ether having the
advantage of being easier to remove by evaporation. Hel-acetone mix-
tures provide a more vigorous extraction medium and this method is used
widely [13-16] for the separation of haem and globin. Acidified methyl-
ethylketone is another solvent suitable for extraction of haems and, unlike
acetone, is immiscible with water. This allows direct extraction of haem
into the organic phase leaving the un denatured protein behind in the
aqueous medium [17]. Preparation of haem from blood has been known
since the nineteenth century and is classically achieved by direct addition of
blood to salt-saturated glacial acetic acid at 100°C to product precipitation
of crystalline haemin [18, 19]. For bulk preparations, use of acetone/acetic
acid with 2% strontium chloride is recommended [20, 21] with yields of up
to 80% being reported by this method.
Isolation of c-type cytochromes and other haems covalently bound to
their proteins is best achieved as the whole pigment protein complex.
Extraction of the protein in aqueous medium is possible, followed by
separation by ion-exchange chromatography [22] to yield a relatively pure
solution of cytochrome. Separation of haem from its covalently attached
apoprotein in c-type cytochromes is rather difficult, and has only satisfac-
torily been achieved by the use of silver sulphate in hot acetic acid [23].
This is a reflection of the natural stability of cytochrome c in vitro, and in
the preparation of colorants is actually advantageous.
In terms of the preparation of material for food use, the suitability of any
of the methods so far described may be limited. However, the existence of
a logical rationale for the extraction of haem from tissues, and the large
body of scientific reference available must surely augur well for such
techniques being developed in the near future.
6.2.3.1.2 Absorbance. We have already seen that the binding of

oxygen to haemoglobin in vivo produces a shift in the observed absorption
Table 6.3 Effects of ligand substitutes on the absorbance characteristics of haems
Name 5th and 6th ligands Solvent hmax EmM
Haem H 2 0, H 2 O pH 7.0 to 12.0 buffer 570-580 5.5--6.5

Haematin OH, H 2 O pH 10 buffer 600 4.5
0.1 M NaOH 610 4.6
385 58.4
Haemin Cl, H 2O Ether 638
Acetic acid 635
Glacial acetic acid 510 90
Ethanol 400
Cyan met-haemoglobin CN, H 2O buffer 540 44
Protohaemochrome Pyridine, pyridine 0.2 M NaOH 557 34.4
25% pyridine 525 17.5
Cytochrome b- Pyridine, pyridine 0.2 M NaOH 557 34.4
haemoglobin-myoglobin- 25% pyridine 525 17.5
haemochromes
Cytochrome c- Pyridine, pyridine 0.2 M NaOH 550 29.1
haemochrome 25% pyridine 522 18.6
spectrum of the pigment. The binding of a number of alternative ligands is

also possible in vitro. Molecules such as carbon monoxide, nitrous oxide,
hydroxide and cyanide are all capable of forming ligand bonds to the
central iron atom in place of oxygen. The overall effect of their presence on
the spectrum of haem is dependent on the electronic nature of the ligand
itself. Ligands capable of donating electrons to the delocalized system of
the porphyrin ring will, in general, produce a shift to shorter wavelengths
in the absorbance maximum of the a peak. Conversely, addition of
electron withdrawing ligands should produce a shift to longer wavelengths
[11 ].
The advantages of using such ligands to bind to haemoglobin are several.
Firstly, irreversible binding of such a ligand may lead to increased stability
of haemoglobin itself, and secondly it leads to the formation of a single
absorbing form of the parent molecule. Thus the rather broad spectrum of
haemoglobin described in vivo is transformed to a single absorbing peak at
540 nm by the binding of cyanide. Other ligands known to have similar
effects--carbon monoxide, nitrous oxide and hydroxide groups have been
characterized in the laboratory, and have some commercial potential.
Thirdly it may be possible to modify the wavelength of absorption by the
use of different ligands to produce a group of pigments of varying
absorption characteristics for different purposes. The haemin-anion com-
plexes could lend themselves to this use as they are relatively easily
prepared [24] and complexation with several different anions is possible.
Table 6.3 summarizes the effects of several known ligand substitutions on
the absorbance characteristics of haems.
The absorption properties of free haems without their associated

apoproteins are well documented [2] and that of free protohaem is
included in Table 6.3 for comparison. The spectra of free haems in aqueous
solution are known to be dependent on several factors that affect their
aggregation [25]. These factors include the concentration of the haem
itself, the pH, ionic strength and temperature of the solution and the
presence in it of several cations. In the case of protoferrihaem, over 95% of
the molecules may exist as dimers in aqueous solution, and at pHs greater
than 11.0 even higher aggregates may form. The dimeric structure has been
shown to be joined by an oxygen molecule bridging the two iron atoms of
the haems, and results in a pronounced decrease in the intensity of the
Soret band. There is also an increase in the visible region of the spectrum
of a component at around 360 nm. Although this may have relatively little
effect on the perceived colour of haems in aqueous solution, it is
nevertheless important to remember when using absorption spectra to
characterize or quantify haems; indeed, this is one of the reasons why
absorption of free haems is usually used only in conjunction with other
methods for this purpose.
6.2.3.1.3 Stability. The relative stability of b- and c-type haemo-

proteins has already been discussed with reference to their extraction and
purification procedures. The effect of ligand binding in increasing the
stability of haems has also been mentioned. It is a fact of nature that
haemoglobin, whose function it is to transport oxygen, is itself susceptible
to oxidative breakdown. Paradoxically, oxygen itself protects the haem
from oxidative attack, which appears to occur only in that fraction of
molecules that at anyone moment do not have bound oxygen [26]. For this
reason, the use of ligands that are capable of binding irreversibly to the
haem molecule, protects it from oxidation, as would the binding of oxygen
itself were it not in dynamic equilibrium with its unbound form.
This principle underlies the relative lack of stability in crude blood or
tissue extracts, where in addition the presence of endogenous proteolytic
enzymes may lead to the breakdown of the haemoprotein itself. Progress-
ive breakdown of apoprotein is probably the single most important cause
of loss of spectral characteristics of ageing blood or tissue preparations.
This problem may be overcome either by purification of the haemoprotein
or by simple extraction of the haem moiety itself from the sample, by the
methods already described. Several adaptations of these methods have
been made for stabilization of haems for commercial use, and these are
discussed in section 6.2.3.2.
6.2.3.1.4 Quantitative analysis. The property of haems in producing

single sharp absorption bands on binding of certain ligands, is used to
effect in their quantitative assay. The extinction coefficients given in Table
6.3 for the various haem-ligand complexes may thus be used for quantitat-
ive analysis. One ligand complex has, however, proved particularly useful
for qualitative and quantitative analysis of haems from a variety of sources.
That is the use of their dipyridine complexes [2] known as a dipyridine
ferrohaemochrome or pyridine haemochromogen. Sodium dithionite
(NaZSZ04) is used to reduce the haem to its ferro (Fe II) form, while
sodium hydroxide is required to disrupt the interaction of the protein with
the fifth ligand position of iron in haemoproteins. The dipyridine complex
formed under these circumstances results in a reproducible spectrum
characteristic of the individual haem compounds. Table 6.3 also gives the
absorption maxima and extinction coefficients of several examples of
these.
As the pyridine haemochromogen method for quantification of haems is
simple, cheap and suitable for haems from almost any source, it is rarely
necessary to use more modern techniques for quantification. Such tech-
niques do, however, exist and have been used mainly in the structural
determination and analysis of natural products on a milligram scale. High-
performance liquid chromatography (HPLC) techniques have been devel-
oped for the separation of porphyrins both as the naturally occurring free
acids and as their methyl ester derivatives [27]. Despite the central
importance of these techniques to clinical and biochemical studies of
porphyrins, it seems likely that, in situations where microscale analysis is
not required, the older (and less expensive) techniques will prevail.
6.2.3.2 Commercial exploitation. Some of the physical properties and

chemical characteristics, capable of being used for the commercial ex-
ploitation of haems, have already been alluded to. It seems pertinent here
to review the processes and patents already in existence as commercial
propositions and the scientific basis of their use.
Several patents covering the use of haem compounds exist [28], and may
be considered to fall into three main classes: (i) those dealing with
preparations of unmodified whole blood or tissue products; (ii) several
concerned with stabilization by binding of different ligands to crude
samples; and (iii) a few dealing with the more complicted question of
purification. Some processes under patent for the preparation of apo-
protein from haemoprotein complexes have been reported to be suitable
also for recovery of haem but this is often not the case, for example
following decolorization with peroxide or active oxygen, which causes
oxidative cleavage of the haem macrocycle.
Patents covering the use of whole blood or tissue slurries are generally
rather limited in application, although both can be spray-dried for direct
use in foodstuffs. Concentration of haem from blood or red blood cells is
possible by a simple process of dehydration using alcohols. Several haem
derivatives are described including methods for ozonolysis, and formation
of carbon monoxide, nitrous oxide, hydroxy and carboxy adducts, all of

which serve to stabilize the haem content of the source without extraction
from it.
Simple methods for extraction have also been covered, e.g. release of
haem from haemoglobin by acid or alkali treatment, enzyme hydrolysis or
heating. Solvent extraction is rarely used, presumbly because of the
expense and the need to remove solvents prior to addition to food. One
process that uses imidazole to promote separation of haem and globin is
presumably similar to the action of pyridine in occupying the fifth and sixth
ligand position of the haem iron and thus disrupting its association with the
apoprotein.
It is generally only possible to discern from a patent whether the product
intended for use is in the form of a haemoprotein or free haem derivative,
by examination of the processes described within it. While this may be an
interesting academic pastime, it seems reasonable to expect that prospec-
tive users of the products should have available very clear information on
this, and on the nature of the bound ligands. Accurate dissemination of
such information would surely do much to allay current fears of 'artificial'
derivatives of these natural pigments.
Nature's own haem derivatives-the bile pigments bilirubin and bili-
verdin-are not without their commercial uses. Obtained from gall stones
and bezoar (hair balls), bilirubin has been used for centuries in Chinese
medicine and is reputed to have aphrodisiac properties. Demand for these
products in modern China now greatly outstrips the availability from
natural sources, and the opportunity for commercialization of a 'nature
alike' pigment is now open.
6.3 Phycobilins
6.3.1 Introduction
Phycobilins are deeply coloured, fluorescent, water-soluble pigment pro-
tein complexes. They are characteristic proteins of blue-green, red and
cryptomonad algae, and represent major biochemical constituents of the
organisms in which they are found. The phycobilins can be classified on
the basis of their spectral characteristics into three major groups: phyco-
erythrins (PEs), phycocyanins (PCs) and allophycocyanins (APCs). The
phycoerythrins are red in colour with a bright orange fluorescence, while
the phycocyanins and allophycocyanins are blue and fluoresce red. The
range of colours found within these groups of pigments is large, dependent
on the source of the biliprotein and the medium in which it is isolated.
These pigments thus provide a variety of natural colours, distributed
ubiquitously within the algal kingdom, and with the potential for future
HAEMS AND BrLINS 171
2 3 12 13 17 18
(a)
1 -:-,.A 19
N
5 10
COOH COOH
I I
CH, CH, CH, CH,
I I I I
CH H,C CH, H,C CH, H,C CH,
(b)
N N
H H
COOti COOH
I I
CH, CH, CH, CH,
I I I II
:H li·
H,C CH H,C CH, H,C. CH, H,C CH,
(e)
~ ~.1=5. )=\-:-,.
Od....N~N))~N~N~O
H H H H H
Figure 6.5 The structures of: (b) phycocyanobilin, (c) phycoerythrobilin and (a) their
common bilin nucleus. The numbering of the bilin structure is according to the IUB-IUPAC
recommendations [1].
commercial development. This section describes the known chemical and

biochemical properties of the biliproteins and aims to assess their future
commercial and industrial potential.
6.3.1.1 Structure and nomenclature. All biliproteins are based on a

structure consisting of a linear tetrapyrrole or bilin, covalently bound to an
apoprotein. The structure of these algal bilins is very similar to that of
mammalian bile pigments, and their nomenclature follows the same
pattern, as shown in Figure 6.5a [1]. The chromophore of the blue
phycocyanin and allophycocyanins is the same in both groups of pigments,
and is called phycocyanobilin. Its structure is shown in Figure 6.5b. The
chromophore of the red phycoerythrins differs from phycocyanobilin in
two respects: phycoerythrobilin has two extra atoms of hydrogen, forming
a 15,24-dihydrobilin, and also has an ethyl rather than a vinyl group at
position 18 of the bilin structure (Figure 6.5c). The chromophores released
by methanol hydrolysis of the pure biliproteins are spectrally distinct to
those produced by Hel cleavage of the proteins [29]. Evidence from
nuclear magnetic resonance and mass spectroscopic studies of the pigments
obtained by methanolysis led to the structures described in Figure 6.5 [30],
The phycocyanobilin chromophore released by methanolysis is identical to
that excreted by Cyanidium caldarium following exogenous administration
of ALA [31]. It is also known that bilin cleaved by methanol can be re-
incorporated into phycocyanin in C. caldarium, providing good evidence
that its structure represents that of the true biliprotein chromophore [31].
Some similar evidence exists in the case of phycoerythrobilin. When red
algae are eaten by the marine gastropod mollusc Aplysia their phyco-
erythrin chromophores are released by digestion and excreted as free-bile
pigments. The structure of this pigment is reported to be identical to that of
phycoerythrobilin obtained by methanol hydrolysis [33].
The simplest interpretation of the known data is that the two structures
shown represent the only prosthetic groups of the algal biliproteins-the
wide spectral differences observed among the biliproteins being a result of
their different protein environments. Thus phycocyanobilin is the only
chromophore found in C-phycocyanin, cryptomonad phycocyanin and
allophycocyanin, and only phycoerythrobilin is obtained from C-, R-, B-
and cryptomonad phycoerythrins. R-Phycocyanin remains the exception to
this rule and releases both phycoerythrobilin and phycocyanobilin on
methanol cleavage [34, 35]. Other chromophores have been reported from
time to time by several authors [29, 36-40], whether these pigments arise as
artifacts of extraction, or whether they are present in vivo is not yet clear.
They do not, however, represent major cleavage products of the common
biliproteins and for that reason they will not be considered further here.
Historically, individual phycocyanins and phycoerythrins have been
classified according to their algal origin and absorption characteristics (see
Bennet and Siegelman [30] for a review). Thus phycocyanins and phyco-
erythrins isolated from blue-green algae (Cyanophyta), which both have
single prominent visible absorption maxima, were given the prefix 'C-',
while phycobilins from red algae (Rhodophyta), which have two or three
such maxima were prefixed by 'R-' [41]. Later a 'new' red algal phyco-
erythrin was found with two major visible absorption maxima and was
given the prefix 'B-', as the original source was a red alga of the order
Bangiales. A final addition to this system of classification came with the
discovery of phycobilins in cryptomonad algae (Cryptophyta), which are
unicellular flagellate algae of indefinite taxonomic position. Cryptomonad
phycocyanin and phycoerythrin have been classified separately due to the
wide range of absorption maxima of the phycobilins found in the various
representatives of this family. Allophycocyanin (APC) is rather simpler to
classify, as there are no spectral differences between allophycocyanins
isolated from red or blue-green algae, and it has not yet been characterized
in cryptomonad species.
This nomenclature is still fairly widely used, although it is cumbersome,
and its usefulness is limited by several exceptions to the taxonomic rules.
The eight classes of phycobilins described by this system of classification
are listed in Table 6.4, along with their characteristic absorption and
fluorescence maxima.
Table 6.4 Classification of phycobilins and their characteristic absorption and

fluorescence maxima
Phycobiliprotein Absorption >-max (nm) Fluorescence

emission >-max (nm)
C-Phycoerythrin 280 308 380 565 577

R -Phycoerythrin 278 308 370 497 538 578
278 308 370 497 556
B-Phycoerythrin 278 307 370 497 545 578
278 307 370 497 563
Cryptomonad phycoerythrin 275 310 370 544 NR
275 310370556 580
275 310 370 568 NR
C-Phycocyanin 280360620 654
280360615 647
R-Phycocyanin 275 355 522 610 637
275 355 533 615 565637
Cryptomonad phycocyanin 270 350 583 625 643 600
270 350 588 625 630 NR
270350588615625 637
Allophycocyanin 280 350 598 629 650 663
NR = not recorded.
The great diversity of absorption maxima observed for the biliproteins in

general is a function of the different apoproteins to which the chromo-
phores are attached, and their physico-chemical environment within the
organism. The attachment of the bilin chromophore to its apoprotein is
particularly stable and is achieved by means of one or two covalent
thioether bonds. The first of these is between a cysteine amino-acid residue
of the apoprotein and an unusual ethylidene group at position 3 of the
bilin. A second covalent linkage is also possible, between another cysteine
residue and the vinyl group at position 18 of the bilin chromophore [43].
The particular properties that this covalent attachment confers on the
stability of the biliproteins will be discussed in section 6.3.3.1.
The native forms of many algal bilins consist of several bilin chromo-
phores attached to one of two apoproteins, to produce distinct biliprotein
subunits designated a and /3. These subunits are present in the organism in
different aggregated forms, as part of the architecture of the light
harvesting apparatus of photosynthesis. This aspect of bilin structure will
now be discussed.
6.3.2 Phycobilins in nature
6.3.2.1 Function and occurrence. Phycocyanin and phycoerythrin func-

tion as major light-harvesting pigments of the photosynthetic algae. In
these organisms, phycobilins exist alongside chlorophyll and act in concert

with it to collect and transform light energy, by means of photosynthesis,
into the chemical energy of the cell. The broad absorption maxima of
phycobilins in their native forms allow algae to trap light energy at
wavelengths over a large proportion of the spectrum of sunlight. The
absorption spectra of phycocyanin and phycoerythrin effectively fill the
optical gap left by the absorbance of chlorophyll.
Table 6.5 shows the occurrence of phycocyanins and phycoerythrins
among the three algal classes in which they occur. It illustrates both the
variation between species, and also the possibility of identifying species
that predominantly synthesize individual pigments.
6.3.2.2 Biosynthesis. Plant bilins are synthesized by ring cleavage of

haem to form a linear tetrapyrrole, in exact analogy to the formation of
bile pigments in mammalian systems following haem breakdown [47]. The
biosynthesis of chlorophyll (as discussed in chapter 5) occurs in parallel to
that of bile pigments, so that in organisms using both chlorophyll and
biliproteins for photosynthesis, a single branched pathway is responsible
for the biosynthesis of both pigments. This pathway is outlined in Figure
6.6.
As has been described previously for mammalian haem synthesis, the
biosynthetic pathway of haem in plants has the substituted amino acid 5-
aminolaevulinate (ALA) as its first committed intermediate. However, in
contrast to the mammalian pathway plants, algae and several photosyn-
thetic bacteria are now known to produce ALA from the intact 5-carbon
skeleton of glutamate, rather than by the condensation of glycine and
succinyl co-enzyme A [48]. This process is presently the subject of intense
interest, as the formation of ALA is likely to be a key control point in the
biosynthesis of both chlorophyll and the bilins in these photosynthetic
organisms. In particular, it appears that light control of pigment synthesis
is likely to be exerted at this point, although the mechanisms by which this
occurs are not yet known.
The next phase in the biosynthesis of phycobilins is the formation of
haem. The process by which this occurs is essentially identical to that
known for the biosynthesis of haem in mammalian tissues (see section
6.2.2.2), and will not be described further here. It is, in fact, remarkable
that the process of tetrapyrrole biosynthesis is so strictly conserved over
the range of organisms in which it occurs.
The formation of phycocyanobilin and phycoerythrobilin following the
ring cleavage of haem, is believed to occur through the intermediates
biliverdin IX a and phytochromobilin [32] (see Figure 6.6) with the two
algal bilin chromophores being produced by the action of one of two
possible reductions of the phytochromobilin precursor. This pathway is
particularly attractive as it provides a universal means by which the major
Table 6.5 Distribution of biliproteins in some algae of the classes Cyanophycae,
Rhodophycae and Cryptophycae [44-46]
Species containing Species containing both Species containng

predominantly phycoerythrin and phycocyanin, with no
phycoerythrin, with trace phycocyanin phycoerythrin
amounts of phycocyanin
Cyanophyceae Cyanophyceae Cyanophyceae

(contain C-PE) (contain C-PE & C-PC) (contain C-PC)
Hydrocoleum spp. Aphanocapsa spp. Anabaena cylindrica
Oscillatoria rubescens Aphanothece sacrum Anabaena variabilis
Phormidium autumnale Calothrix membranacea Anabaena 64JJ
Phormidium ectocarpii Dermocarpa violacea Anabaenopsis spp.
Phormidium fragile Fremyella diplosiphon Anacystis nidulans
Phormidium persicinum Glaucosphaera vacuolata (=Synechococcus 6301)
Phormidium uncinatum Schizothrix alpicola Arthrospira maxima
Pseudanabeana spp. Tolypothrix distorta tenuis Aphanizomenon fios-aquae
Calothrix scopulorum
Rhodophyceae Rhodophyceae Chlorogloea fritschii
(contain B-PE) (contain B-PE & R-PC) Lyngbya lagerheimii
Rhodelia violacea Porphyridium aerugineum Mastigocladus laminosus
maculata Porphyridium cruentum Nostoc punctiforme
Porphyridium marinum
Cryptophyceae Oscillatoria amoena
(contain cryptomonad PE)
(=purpureum) Oscillatoria animalis
Cryptomonas ovata Cryptophyceae Oscillatoria aghardii
maculata Unknown Oscillatoria chalybea
Hemiseimis rufescens Oscillatoria formosa
Plagioseimis prolonga Oscillatoria minima
Rhodomonas lens Oscillatoria subbrevis
Oscillatoria tenuis
Phormidium faveolarum
Phormidium luridum
Plectonema boryanum
Plectonema calothricoides
Spirulina platensis
Rhodophyceae
(contain C-PC)
Cyanidum caldarium
Cryptophyceae
(contain cryptomonad PC)
Chroomonas spp.
Cryptomonas cyanomagna
Hemiselmis virescens
algal bilins may be synthesized, it also incorporates the chromophore of

phytochrome, the higher plant bile pigment responsible for light percep-
tion, as an intermediate. Indeed, it is proposed that the general pathway
outlined here for the synthesis of algal bilins exists throughout the plant
kingdom for the biosynthesis of phytochrome.
The final step in the biosynthesis of all bile pigments is likely to be the
attachment of the bilin chromophore to its apoprotein [32]. Although little
is known about the enzymology of this process, it is clear that in most
.......
-...I
0\
~ chlorophylls
Me/ I
I Pr
~-V Mg. protoporphyrin IX
z
CO,H CH 2 NH 2
(b)
~
CHNH, (a) CO ~
¢H, t""
CH, ~ --..
CH, C;:H2 c:l
CO,H C0 2 H o
t:I
glutamate ALA (j
phycoerythrobilin o
S
~
(d) I
~ biliverdin ~ phytochromobilin ~
phycocyanobilin
haem
"
Figure 6.6 The branched biosynthetic pathway for the formation of haem, bilins and chlorophylls from glutamate. The enzymes responsible for the
biosynthetic steps shown are: (a) glutamic acid tRNA ligase, glutamyl tRNA dehydrogenase, and glutamate I-semialdehyde aminotransferase, (b) as
described in Figure 6.2, (c) ferrochelatase, (d) as described in Figure 6.3. V = CHCH2 , Me = CH 3 , Pr = CH2 CH2 C0 2 H.
HAEMS AND BIUNS 177
organisms strict mechanisms exist for the coordinate regulation of apo-

protein and chromo ph ore synthesis. As a result, the accumulation of free
chromophore within an organism occurs only when the normal metabolic
state of the organism is altered. This will be discussed in section 6.3.2.3.
The converse situation, in which free apoprotein is produced without
concomitant synthesis of the appropriate chromophore, is not known to
occur under any conditions. It would appear from this that formation of the
chromophore is not limiting in biliprotein synthesis [49]. This is supported
by results recently obtained using cloned genes for phycocyanin apo-
proteins. The genes for the a and 13 subunits of phycocyanin from the
blue-green alga Synechococcus spp. PCC7002 were cloned [50] and re-
introduced into Synechococcus at higher multiplicity. This resulted in over
expression of the apoproteins and the accumulation of corresponding
greater amounts of phycocyanin [51]. The results suggest that cells are able
to synthesize sufficient free bilin to accommodate the increased demand for
pigment, and this opens up the possibility of genetic manipulation of algal
strains for increased pigment production.
In addition to the major plant and algal bilins already described, there
have, over several years, been reports of unusual bilin-like pigments from
other sources. A blue bile-pigment based on the parent structure of
uroporphyrin has been described in two bacteria-Clostridium tetano-
morphum and Propionibacterium shermanii [52]. This pigment has visible
absorption maxima at 369 and 644 nm. However, such pigments are not
major products of the organisms concerned and will not be considered
further here.
6.3.2.3 Metabolism. This is divided into: (i) chromatic adaptation;

(ii) sun to shade adaptation; (iii) effectors of bilin synthesis; and (iv) pig-
ment mutants.
6.3.2.3.1 Chromatic adaptation. One of the characteristic properties

of many algal species is their ability to produce different proportions of
individual light absorbing pigments in response to the different qualities
of light in which the organism is growing. This photoregulated response of
pigment synthesis is known as chromatic adaptation. Typically red light at
wavelengths above 600 nm promotes the production of phycocyanin and
allophycocyanin but not phycoerythrin, while green light of 500 to 600 nm
promotes formation of phycoerythrin alone [53]. In nature, this allows the
algae to make the most efficient use of light at the available wavelengths,
with the minimum expenditure of metabolic energy in synthesizing pig-
ments.
This characteristic has been known for some time [44], and has been
analysed in detail in several different species. In the filamentous blue-green
alga Tolypothrix tenuis, for example, the production of bilins is known to

be specifically driven towards phycoerythrin by light of 541 nm wavelength
and towards phycocyanin by illumination at 641 nm [54]. It is thus possible
to maximize production of individual bile pigments by using the organisms'
natural mechanisms for adapting to light quality. It is of particular interest
that the genetic basis of photoregulation is beginning to be unravelled [55]
as this may offer further possibilities for the future manipulation of
pigment synthesis in algal species.
Two other metabolic influences on the overall level of bilins are
important. The accumulation of any metabolite is dependent on its net rate
of biosynthesis and degradation within the cell. Although specific degrada-
tion of phycocyanin is known to occur in several species in response to
nitrogen starvation [56], so far as is known adaptation to changes in the
wavelength and intensity of light does not involve specific degradative
processes [57] and decreases in individual bilin concentrations within the
cell are usually considered to be a result of dilution following cell division.
This is consistent with studies on exponentially growing cells, which report
the remarkable stability of phycobilins in vivo [57, 58].
In many algal species when cells are grown under conditions in which
nitrogen supply is non-limiting, phycobiliproteins comprise a significant
proportion of the total cellular protein. moving such cells to conditions of
limiting nitrogen results in the specific degradation of these pigments. This
has been shown to occur in several cyanobacterial species [55, 59]. A
similar effect is also known following depletion of sulphur compounds from
the growth medium of cyanobacteria [60], although this seems surprising as
C-phycocyanin has a rather low ratio of sulphur to nitrogen compared with
many proteins. These characteristics may, however, be used to increase
pigment synthesis under certain conditions, as it has been observed that
administration of sulphate following a period of starvation leads to a rapid
and preferential increase in C-phycocyanin content prior to resumed cell
growth.
6.3.2.3.2 Sun to shade adaptation. Just as algae are able to adapt to

the quality of light available for photosynthesis, synthesis of phycobili-
proteins is also affected by light intensity. Increasing the light intensity
causes an overall decrease in phycobiliprotein concentration, as part of an
adaptation mechanism in which photosynthesis is maintained at its maxi-
mum rate using the minimum pigment necessary for light harvesting and
energy transduction. Reductions of up to three-fold in phycobiliproteins
were reported in the unicellular rhodophyte Porphyridium cruentum [61]
in response to increasing light intensity. Similar reductions were observed
in the blue-green alga Anacystis nidulans [62] in which it was concluded
that light intensity was one of the primary factors affecting pigment
synthesis as a whole [63].
6.3.2.3.3 Effectors of bilin synthesis. Addition of several substrates

and intermediates of tetrapyrrole metabolism is known to induce bilin
chromophore synthesis [32]. Addition of aminolaevulinic acid, as a
committed intermediate in bilin synthesis, is found not to increase overall
levels of biliprotein, but instead causes the excretion of the free bilin
chromophore. The addition of excess glutamate would be expected to have
a similar although less marked effect as it is a substrate for several
metabolic processes other than porphyrin synthesis. It appears that the
coordinate control of chromo ph ore and apoprotein synthesis prevents
increased biliprotein formation in response to increased levels of bilin
alone; this suggests that efforts to specifically increase biliprotein content
would be more logically aimed at promoting apoprotein formation, as the
rate of bilin biosynthesis may respond to the availability of apobiliprotein
[49].
6.3.2.3.4 Pigment mutants. One means by which pigment synthesis is

controlled very strictly, if artificially, is by the use of certain mutations in
which biosynthesis of an individual pigment is blocked. In the case of the
unicellular red alga Cyanidium caldarium, several such mutants exist [64].
Like the wild type cells, these mutants are capable of growing hetero-
trophically in the presence of glucose in the dark. On transfer of dark
grown cells to the light, they synthesize pigments characteristic of their
specific mutation as summarized in Table 6.6. The existence of these
mutants demonstrates the possibility of using the independent functions of
the branched biosynthetic pathway to produce individual pigments.
6.3.2.4 Physical properties. Cellular localization, absorbance and fluor-

escence, and stability of the phycobilins in nature will be discussed here.
6.3.2.4.1 Cellular localization. In red and blue-green algae phyco-

bilins participate in the overall process of photosynthesis in the form of
large multimeric aggregates called phycobilisomes. These structures are
present in the chloroplast of the algae, and appear as discrete morpho-
logical features on the outer face of the thylakoid membrane of the
chloroplast, when viewed under an electron microscope. The size, compo-
sition, structure and function of the phycobilisomes from several organisms
have now been studied [65~7], and much is known about the mechanisms
by which light energy is absorbed and transferred within their structure. In
cryptomonad algae phycobilisomes are absent, instead, the biliproteins are
aggregated within the intrathylakoid spaces of the chloroplasts. This has a
similar effect of producing closely packed pigments, and thereby enhancing
their energy transfer efficiency.
The phycobilisomes of red and blue-green algae vary in their composi-
tion according to several factors. These include the organism from which
PE PC AP PC PE
ROD CORE ROD
Figure 6.7 Schematic representation of a phycobilisome. The core structure consists of

allophycocyanin (AP) associated with specific linker polypeptides. The rods contain phyco-
cyanin (PC) and phycoerythrin (PE) hexamers, and their linker polypeptides.
they are isolated and the metabolic and spectral conditions in which the
organism was grown. A model structure for a phycobilisome has been
proposed; it allows for the variation in pigments found in different
organisms and accounts for the known characteristics of energy transfer
between pigments within the photosynthetic apparatus (Figure 6.7).
The basic phycobilisome building block is a monomer containing two
dissimilar apoprotein subunits a and 13. The a subunit contains one
chromophore per apoprotein, and the 13 subunit has two [68]. The model
proposed consists of a core composed primarily of allophycocyanin,
present as trimers of af3 subunits. As allophycocyanin is thought to be
present in trace amounts in all species [42], it is probable that this core is a
universal structure on which all phycobilisomes are built. The core is sur-
rounded by several rod structures composed of hexamers of phycocyanin
and phycoerythrin af3 subunits. The whole structure is stabilized by the
presence of associated linker peptides, which also serve to modulate
the absorption characteristics of the pigments within the phycobilisome.
The model illustrated is specifically proposed for the alga Synechocystis
6701 [69], but it is possible to adapt it for organisms containing different
proportions of phycocyanin and phycoerythrin by substituting the pigments
in their different proportions in the rod segments of the structure.
The overlapping absorption and emission spectra produced by the
biliproteins in the environment of the phycobilisome, result in a one-way
flow of light energy from pigment to pigment in the following order:
phycoerythrin ~ phycocyanin ~ allophycocyanin ~ chlorophyll
The overall efficiency of this process is over 90% [70], making the
phycobilisome one of the most efficient structures for transducing light
energy in nature.
(a) (b)
0.8
,
,,"
~ "
> 0.6
N ,
~ ,,
,
,,,
.~
c:
,, ,
~ 0.4
. ,,, ,,,
fl
c: I
u
'"
l'!
0
, ,
, \ 0.2
'- ,
:J
u:: I
700 750 400
Wavelength (nm)
Figure 6.8 Fluorescence emission (a) and absorbance (b) spectra of intact phycobilisomes
from Fremyella diplosiphon grown under differing light conditions. Cells grown under red
light (----) contain phycocyanin and allophycocyanin (>'max 620 nm and 650 nm respectively),
while cells grown under green light (--) also contain phycoerythrin (>'max 565 nm).
6.3.2.4.2 Absorption and fluorescence. The spectrum of an intact

phycobilisome is a composite representation of its component chromo-
phores and the individual associations and energy coupling of the overall
structure. Representative absorption and fluorescence spectra of phyco-
cyanin and phycoerythrin contained within the intact phycobilisomes of
Fremyella diplosiphon [65] are shown in Figure 6.8. The diversity of
spectra obtained from the photosynthetic apparatus of an organism is a
reflection of the different environmental factors affecting pigment accumu-
lation, as described in section 6.3.2.3. Definitive spectra of individual
biliproteins are only obtained following isolation from in vivo structures,
and these are described in section 6.3.3.1.
6.3.2.4.3 Stability. The stability of phycobilins in vivo has already

been mentioned, with regard to the degradation and turnover of pigment
during growth under different metabolic conditions. Among the character-
istics of the algae that use phycobilins as light-harvesting pigments, is the
ability of phycobilisomes to function under a variety of environmental
stresses. Thus algae that are classified as thermophilic (high temperature
tolerant), acidophilic (acid pH tolerant), halophilic (high salt tolerant) or
psychrophilic (low temperature tolerant) might be expected to show
different characteristics with regard to the stability of their pigments.
Published data on the phycocyan ins of several blue-green and red algae
[71] indicate that resistance to denaturation varies from one species to
another. Using denaturation by urea as model system, greatest stability

to denaturation was shown in three thermophilic organisms tested-
Mastigocladus laminosus, I-30-m and NZ-DB2-m, and the acid-tolerant
unicellular red alga Cyanidium caldarium. The stability of phycocyanins
from the halophilic organism Coccochloris elabens was also reported as
being enhanced compared with those of organisms growing at room
temperature and lower. The authors suggest that the increased stability of
the different phycocyanins is produced directly by differences in the
primary amino-acid structure of the proteins.
6.3.3 Free phycobilins
6.3.3.1 Physical properties. Extraction and purification, absorbance

and fluorescence, stability and quantitative analysis will be considered here.
6.3.3.1.1 Extraction and purification. Phycobilins have been

extracted from algae and purified in a range of forms from intact
phycobilisomes to the protein-free chromophore. These forms, by defini-
tion, have very different properties of absorption, fluorescence and
chemical stability, thus it is useful to define the purification procedures and
the nature of the bilin products before discussing their individual character-
istics. Before doing this it may be noted that phycobilins may be obtained
without purification by simple freeze-drying of algal cells. This produces a
brightly coloured powder retaining many of the absorption characteristics
of the phycobilins in their native form and without loss of stability of the
intact structure. The spectrum of phycocyanin and chlorophyll in freeze-
dried cells of Cyanidium caldarium that have been resuspended in aqueous
solution is indistinguishable from that of whole cells in vivo (J.D.
Houghton, unpublished reSUlt).
The disadvantages of this system are obvious in that the colours obtained
are mixtures of the total cellular pigment contents-usually bilins, chloro-
phyll and carotenoids. One possible solution to this could be by use of the
pigment mutant GGB (see Table 6.6). This mutant is capable of growing
heterotrophically in the dark using glucose as a carbon source at rates
comparable with those of wild-type cells. On transfer to light, the mutant
synthesizes phycocyanin without any detectable synthesis of chlorophyll a,
and the cells acquire a pale-blue coloration. The amount of phycocyanin
accumulated by the mutant is over 80% of that achieved by wild-type cells
[72].
Traditionally the problem of purity has been approached by successive
separation of phycobilins from the associated cellular pigments and
proteins. This process is aided by the fact that phycobilins are very readily
Table 6.6 Pigmentation of Cyanidium caldarium

mutants transferred to light following heterotrophic
growth [45, 72]a
Cell type Ch la PC APC
Wild type +(100) +(100) +

III-D-2 +(145) +(140) +
III-C +(55) -(0)
GGB -(0) +(83) +
GGB-Y -(0) -(0)
a Numbers in parentheses represent percentage of pig-

ment produced in mutant compared with wild-type
cells.
Table 6.7 Extinction coefficients and stable assembly forms of algal biliproteins [23, 80]
Stable Amax Ex 103 1/ Solvent

assembly mol!cm!
form chromophore
Allophycocyanin (a~h 662.5 32.2 8M urea

C-Phycocyanin (a~h (a~)6 662.5 35.5 8M urea
R-Phycocyanin (a~h 662 36.2 8M urea
Cryptomonad phycocyanin ?
Phycocyanobilin 670 37.9 5% HCllMeOH
Phycocyanobilin 600 18.9 Neutral CHCI 3
C-Phycoerythrin (a~)6 550 43.5 8M urea
B-Phycoerythrin (a~ky 550 42.9 8M urea
R-Phycoerythrin (a~ky 567 8M urea
Cryptomonad phycoerythrin a~
Phycoerythrobilin 556 HCllMeOH
Phycoerythrobilin 576 Acid CHCl 3
soluble in water. Following cell breakage it is therefore possible to separate

the phycobilins from the lipid soluble chlorophylls and carotenoids by high
speed centrifugation of the homogenate. This process yields brightly
coloured solutions of soluble protein, of which 40% may be phycobilin. In
this form, the phycobilins are probably present as aggregates, derived from
partial breakdown of the phycobilisome structure (Table 6.7).
Further purification of the supernatant by ammonium sulphate precipi-
tation and ion-exchange chromatography yields almost pure phycobilin in
the form of linked al3 subunits [73]. Separation of the subunits themselves
is possible by denaturation with urea [74], while purification of the protein-
free chromophore may be achieved by refiuxing in boiling methanol for
16 h [75]. Each of these purification steps is associated with a loss of overall
yield and would require justification in terms of the separation from

contaminating coloured proteins or a requirement for absolute purity.
An alternative approach to that of purifying phycocyanobilin from the
native biliprotein, in situations where absolute purity is required, would be
to use the ability of certain organisms to synthesize free chromophore.
Thus in Cyanidium caldarium addition of 5-aminolaevulinic acid to the
growth medium causes excretion of phycocyanobilin as an almost pure
product. Harvesting of the whole cells yields a clear-blue solution from
which phycocyanobilin may be separated from any contaminating porphyr-
ins by simple chloroform extraction of the supernatant [76].
6.3.3.1.2 Absorption andftuorescence. The visible spectral properties

and fluorescence of free biliproteins are dependent on their attached
apoproteins and the stable isolation state that is characte~stic of the
organism in which they are found. The spectral characteristics of phyco-
cyanins and phycoerythrins from three classes of algae are shown in Figure
6.9. The spectra of the individual biliproteins themselves are dependent on
the conformational state of the apoproteins, for example, with the
transition from aggregate (hexamer) to trimer to monomer [77]. Fluores-
cence is even more sensitive to dissociation of pigment aggregates as, for
example, in the observed shift of fluorescence absorption following
dissociation of intact phycobilisomes [45] as shown in Figure 6.10.
Complete unfolding or denaturation of the apoprotein causes a substan-
tial decrease in the visible light absorbance [77] and completely abolishes
the visible fluorescence [78]. The spectral shifts associated with these
changes are considerable and are shown for phycocyanin from Porphyr-
idium luridum in Figure 6.10.
More subtle effects are also known. In dilute solution, the extinction
coefficient decreases at the absorption maxima, and absorbance at a given
concentration varies with ionic strength and pH [79-81]. Thus, factors
affecting aggregation also affect absorbance, and it is important when
making comparisons between different pigments to ensure that the
physical conditions in which they are measured are comparable. Table 6.7
is a summary of the stable aggregation states and known extinction
coefficients for phycobilins, and gives for comparison the known values for
free phycocyanobilin and phycoerythrobilin.
6.3.3.1.3 Stability. The stability of phycobilins in vitro is a notional

concept, as the stable state of the biliprotein varies from simple a{3
monomers to distinct hexameric aggregates. In the case of C-phycoerythrin
and C-phycocyanin, this aggregation behaviour appears to vary according
to the species of origin [46], and is dependent on concentration, pH, ionic
strength and the presence of denaturants and detergents. In contrast,
(a)
0.8 C - phycobilins
0.6
'\
: I
/\
I
: I I
, I
~ I
r. /
0.4 I'
,:
1
I
I
I
300 400
(b)
R - phycobilins
0.8
!~
\/
I
I
I
, I
,
I
I
I
/ -'
500 700
(c)
cryptomonad phycobilins
0.6
jIl
·c
:::>
g 0.4
-e'"
5l
~ 0.2
300
Wavelength (nm)
Figure 6.9 Absorbance spectra of isolated biliproteins from three classes of algae, taken in
pH 7 phosphate buffer. The spectra show phycoerythrins ( .... ), phycocyanins (----), and
allophycocyanins (--) from representative species of the (a) cyanophyta e.g. C-phycobilins;
(b) rhodophyta e.g. R-phycobilins; and (c) cryptophyta e.g. cryptomonad phycobilins.
native forms of R- and B-phycoerythrins exist as stable a~ hexamers, with

some evidence for irreversible dissociation to form a~ monomers [82]. R-
phycocyanin appears to exist in two well-defined aggregation states as a~
trimers and hexamers. The transition between the two forms is regulated
by pH, with values of 6.5 or above favouring the trimer and values below
(a) (b)
0.8
.~
c:
::>
0.6 CD
0
1ij
-e0
..:'"
0.4 £J
0.2
Wavelength (nm)
Figure 6.10 Fluorescence emission (a) and absorbance (b) spectra of phycobilins from two
species of Porphyridium. The spectra show the shifts occurring following dissociation of
aggregates (--) to form partially (----) and totally ( .... ) dissociated products and, finally,
on denaturation of the subunits (- . - . - .).
6.5 the hexamer. C-Phycocyanin follows a similar aggregation pattern, but

tends to dissociate at low protein concentrations to form a~ monomers [83]
as shown in Table 6.7.
Phycocyanin has been extensively studied in terms. of aggregation
phenomena in particular and physico-chemical properties in general, and a
review on this subject has appeared recently [77].
Phycoerythrins as a class are more stable towards denaturation than
phycocyanins. All phycocyanins are known to be denatured in 4M urea
while phycoerythrins are not [83]. Among the phycoerythrins, the Rand B
types show increased stability over C and Cryptomonad pigments, denatur-
ing only slowly in 8M urea, and retaining absorption and fluorescence
maxima at temperatures up to 70°C [83]. These differences in conforma-
tional stability have been suggested to arise from the presence of covalent
disulphide bonds between cysteine residues of the separate protein
subunits [83]. Thus R- and C-phycocyanins have only enough cysteine
residues to provide for covalent attachment of their chromophores, while
R- and B-phycoerythrins contain sufficient residues to allow intra-subunit
bonding to take place. Dissociation of biliproteins into subunits using
mercurial or sulphydryl reagents may be mediated through interaction with
sulphydryl groups on the proteins [84-86], while the action of denaturants,
such as urea and guanidine hydrochloride, and detergents may be assumed
to be through hydrophobic interactions with the proteins. It is interesting
to note that renaturation of biliproteins can be facilitated by simple
removal of denaturants [87], for example, by dialysis. Although this
reflects an inherent stability of the pigment protein complex of biliproteins,

there is an implication that renaturation is not complete as only partial
restoration of spectral properties is observed [88].
The stability of the free chromophores of phycocyanobilin and phyco-
eyrthrobilin is rather less well studied. It is thought that the free phycobilin
structures are stabilized by intermolecular hydrogen bonds to form a semi-
cycle structure [89, 90]. This structure is soluble in acid, methanol and
chloroform but is both acid-labile and subject to oxidative degradation.
6.3.3.1.4 Quantitative analysis. In many instances, the most conven-

ient method for quantitative analysis of biliproteins is by use of their UV-
visible absorption spectra and known extinction coefficients. The accuracy
of this method is, however, dependent on a knowledge of the physical state
of the phycobilin with regard to the presence of apoprotein and the state of
aggregation.
For absolute accuracy, it may sometimes be desirable to use modern
methods of high-pressure liquid chromatography (HPLC) for separation
and quantification of bilins. The subject of bile pigment separation of
HPLC has been reviewed recently by McKavanagh and Billing [91].
Although this specially deals with bile pigments derived from the break-
down of haem in mammalian tissues, the principles of separation are
identical to those employed in dealing with phycobilins, and the methods
may be adapted accordingly.
In general, it is more convenient to quantify bile pigments by HPLC as
protein-free chromophores either in the free-acid form or with the
propionic acid substituents esterified to form the corresponding bilin
dimethyl ester.
Separation of unesterified bile pigments is achieved by reverse-phase
HPLC using a hydrophilic column support and a hydrophobic mobile
phase. Esterified bilins have been separated by both normal and reverse-
phase systems with success, and this method of sample preparation would
seem to confer several advantages in terms of stability and ease of
handling. A method for the HPLC separation of algal bilin methyl esters
has also recently been published [92].
One advantage of using HPLC as a tool for the quantification of bile
pigments is that, in addition to information on the amount of pigment
present, the HPLC provides details of the nature and quantity of
contaminating pigments, and may in the future prove to be of considerable
use as a tool in quality control of pigment production.
6.3.3.2 Commercial exploitation. The commercial exploitation of bili-

proteins is a rapidly advancing field. The range of intensity of colours
available, their relative abundance in the organisms in which they are
found, and the inherent stability of their pigment-protein complexes, make
them ideal candidates for future development. Current interest is roughly

divided into biotechnology and applications in industry.
6.3.3.2.1 Biotechnology. Although there is a long tradition in certain

cultures of harvesting algae for food use, the single most promising area for
further advances lies in the use of modern biotechnological methods for
large-scale culture and extraction of micro algae [93-95].
This field of technolgoy has advanced considerably in recent years
following pioneering work on the extraction of J3-carotene from the
halophilic green alga Dunaliella salina. Other microalgae are now begin-
ning to be exploited for mass cultivation and harvesting of fine chemicals,
the enhanced value of which justifies the technology required for optimum
algal growth. To date, the algae most commonly cultured among the
Rhodophyta and Cyanophyta are respectively from the genera Porphyr-
idium and Spirulina. It is fair to say that a great many other species of algae
could prove suitable for mass culture given the basic knowledge of their
physiology and biochemistry and the financial incentive to develop appro-
priate technology. This financial incentive is becoming increasingly more
apparent in the use of microalgae for production of vitamins, pigments,
polysaccharides and pharmaceutical compounds [96].
Current technology for mass culture is commonly based on open pond
plants in hot areas where algae with halophilic or alkali-tolerant properties
can grow in near monoculture all year round. This system represents
relatively low cost technology and uses natural sunlight to promote
photosynthetic growth. Its use is also limited to those organisms whose
growth conditions preclude major contamination by indigenous species,
and this has proved to be a major limiting factor. Several other methods for
large-scale culture are available and, although their use represents
increased expense in equipment and running costs, the production of high-
value metabolites from individual algae could well offset these. Closed
systems using polyethylene tubes to incubate algae under natural sunlight
represent an advance on techniques for photosynthetic growth [97], as
have several variations on this theme such as the use of covered troughs
and vertical glass columns. Such enclosed photobioreactors will shortly be
available for industrial systems using micro algae [93] and should consider-
ably enhance the number of species for which it is feasible to consider
large-scale culture.
An alternative approach is to use already existing fermentation technol-
ogy to grow those algal species capable of heterotrophic growth using an
external carbon source. Again, the investment in equipment and substrates
may be justified by production of novel high-value products. It is in this
area in particular that the use of genetically engineered algae might be
envisaged, as the economic rewards of unique product synthesis would be
great.
So far as pigment synthesis from microalgae is concerned, the tech-

nology existing for the growth and extraction of the blue-green alga
Spirulina platensis is typical and may be used as a general example. The
natural occurrence of this alga is in saltwater ponds and inland lakes with
saline and alkaline water; as such, it is capable of colonizing extreme
environments that are unsuitable for many other organisms. In the wild, S.
platensis grows at 27°C at pHs from 7.2 to 9.0 and at salt concentrations of
over 30 gil, under which conditions it is found almost as a monoculture
[98]. The high amino-acid content of Spirulina and digestibility to humans,
has led to its being used as a food supplement from earliest times among
indigenous tribes in Africa and Mexico where Spirulina is found. Its use is
fashionable now as a health food, and the extraction of C-phycocyanin
from Spirulina is a recent addition to the commercial potential of the
organism. Mass cultivation is generally in covered ponds or closed
reactors, and in such conditions of temperature, light and nutritional status
as are suitable for maximum cell growth, pigment synthesis etc. [98, 99].
Harvesting may be achieved by use of a vibrating screen or filtration and
the cells are then sun- or spray-dried.
The situation for the unicellular red alga Porphyridium cruentum is
similar [100]. It is generally grown in artificial seawater at temperatures of
21°C and with an added nitrogen source such as potassium nitrate (KN0 3 ).
The alga can be grown in ponds or closed tubular reactors and, in addition
has been immobilized on glass beads for use as a biocatalyst in the
production of excreted polysaccharides [93, 100, 101]. Other products of
interest from this alga include the dietary supplement arachidonic acid and
R-phycoerythrin.
The processes described for the culture of Spirulina and Porphyridium
are similar to those required for any number of algal species, given the
biochemical knowledge to optimize their culture conditions and the
economic incentive for doing so. It is thus reasonable to expect that, should
the future demand for pigments justify it, a range of algal organisms might
be employed to produce not only the C-phycocyanin and R-phycoerythrin
currently available but also the many other spectral variants of the
pigments occurring in nature.
The economic potential of any individual alga is a combination of the
input in its cultivation and the value of its products. In this respect, the
concept of obtaining several products from a single organism must appear
attractive. In the case of Porphyridium this possibility has been assessed
and costed [100, 102], with the conclusion that the gross value of a number
of products would repay any additional cost of their individual separation.
The removal of pigments from algal biomass may, in fact, be a requisite for
its proposed use, for example as animal feed, and in this case the pigment
itself becomes a byproduct of the feed production process, further
enhancing the commercial potential of the process.
The potential of several other microalgae remains to be investigated. It

seems highiy likely that among the organisms listed in Table 6.5, are some
whose economic possibilities are such that efficient production of indivi-
dual pigments would be possible, and the time is now ripe to capitalize on
this.
6.3.3.2.2 Applications in industry. Although in its infancy, the de-

velopment of bilins as food colorants has already received much interest
[28,103-106]. Several patents already exist for the extraction, stabilization
purifications and use of phycocyanin from Spirulina spp. and Aphanothece
nidulans [28]. Niohan Siber Hegner Ltd. (Tokyo) and Dainippan Ink and
Chemicals Inc. (Tokyo) both market phycocyanin as a food colorant. It has
been used in chewing gums, and is suggested as an additive in frozen
confections, soft drinks, dairy products, sweets and ice-creams [103, 107].
No patents have been filed to date on the use of phycoerythrin as a food
colorant, but the potential for the development of a natural red pigment is
thought to be great because of concern about the safety of the alternatives
currently available. It is likely that this will be an area of rapid develop-
ment in the immediate future.
In addition to the use of phycobilins in the food industry, several other
high-value uses have recently emerged. The fluorescent properties of
phycobilins are being employed as novel tracers in biochemical research
(see Glazer and Stryer [108] for review). Biliproteins are conjugated with
molecules that confer biological specificity to produce a new class of
fluorescent probes-the phycofluors. The high absorbance coefficients and
intense fluorescence emission of phycobiliproteins make them highly
sensitive fluorescent labels. These properties are almost unchanged on
conjugation with other molecules through acylation of specific amino-acid
residues of the protein.
The phycobiliproteins may be conjugated to specific linker molecules,
allowing formation of several derivatives [109]. One of the most widely
used methods involves formation of a derivative with a monoclonal
antibody for use in fluorescent immunoassay [110]. Phycobiliprotein
conjugates have also been used for fluorescence-activated cell sorting and
fluorescence microscopy and are promising tools in several new areas of
research [108]. R-Phycoerythrin conjugates are already commercially
available from the Sigma Chemical Co. (Poole), and Vector Laboratories
(Peterborough), and represent an important high-value outlet for phyco-
erythrin production from algae.
6.3.3.3 Comparison with synthetic colours. Several comparisons have

been made of the colours of the phycobiliproteins with the synthetic red
and blue pigment available commercially Phycocyanin has been reported
1.0 (3) 1.0 (b)
,,/...", !:."
~
, '
,," ,.
c:
i:
::J /
g / \
.e 0.5 /
/
' 0.5 I :
,, ,'\
o
'" "
,\ .:
/
\
J:l "
< \, ,
/
\
...
,,
\.
, ,, :' \
OL----"'---'--"'-----::-' o '.""- .--~ \
300 400 500 600 700 300 400 SOO 600 700
1.0 (C) 1.0 (d)
,I:,,
,r""\, ,, ,,
/
:, ',,
0.5 /
,: ~,,
/ \, ;, :'
, ,
~\ ,, } ~
,, / ,,
.... ,
/
'- ,
\
"
..... ....
o "'.... ,.' '. .. ,,'
,/ \"
OL-~_~~~~.
300 400 SOO 600 700 300 400 500 600 700
1.0 \(e) 1.0 (f)

,
,'\ r,
,/, ",, " /'
,f \,
,, ,,: ',,,
;\
I '
, , 0.5
~
"\ ,,
'\ I
, ,, ,, ,,
,,' / \"
/
-~,/
~
o \...... ",---",,' ..
300 400 500 600 700 300 400 500 600 700
Wavelength (nm)
Figure 6.11 The absorbance spectra of several synthetic blue and red food colorants:
(a) ponceau 4R (E124); (b) erythrosine (E127); (c) red 2G; (d) patent blue V (E131);
(e) indigo carmine (E132); and (f) brilliant blue FCF.
as being 'between blue colour no. 1 (brilliant blue) and blue colour no. 2
(indigo carmine)' [107]. However, as already discussed, the exact spectral
properties of all the phycobiliproteins are dependent on their algal source,
state of purification and physicochemical environment. These spectra have
been presented throughout the chapter in relation to each particular state
of the pigment. It seems pertinent here to present for accurate comparison
with the phycocyanins and phycoerythrins the spectra of several synthetic
blue and red colours [111] (Figure 6.11, Table 6.8).
Table 6.8 Absorbance characteristics of some synthetic red and blue food colorants
Name ~max (nm) E1% EmM
Ponceau 4R (E124) 505 431 71

Erythrosine (E127) 526 1154
Red2G 528 620 122
Patent blue V (E131) 635 2000 172
Indigo carmine (E132) 610 489 105
Brilliant blue (FCF) 629 1637 207
6.3.3.4 Future prospects. In the first edition of this book the future of
natural and nature-alike pigments seemed assured. The public awareness
of such pigments was growing and the acceptance of natural food colorants
as 'healthy' additives seemed certain. In this atmosphere it seemed likely
that haems and bilins might play a role in expanding the spectrum of
natural colorants available for use in foods. Advances in the areas of algal
aquaculture, and DNA manipulation of pigment biosynthesis seemed set
to increase the yields and commercial viability of such products.
In the last few years a slow but certain undermining of these basic
assumptions has begun. The recent MAFF review on the average daily
intake for food colours highlighted for the first time the possibility of
natural colorants being consumed at levels exceeding the proven safe limits
[112]. This applied in particular to annatto, which as a result will no longer
be legal in the UK for use in sugar confectionery, or other applications
where the majority of consumers are children. It is a very real fear that
such restrictions may ultimately limit the produces that food manufacturers
will be able to sell.
Present concerns regarding dietary intake above the approved Average
Daily Intake (ADI) are founded in the assumption that the current ADI
values are scientifically meaningful. In the case of natural food colorants
this is often far from the case and little research has been carried out on
their metabolism in humans. There is an urgent need for modern metabolic
studies to be conducted on a range of natural colorants to establish
meaningful ADI values for each.
A second factor affecting the possible introduction of new natural food
colorants is the need to bring UK Food Regulations in line with those of
other EU countries. The new EU colours directive, which aims to allow
manufacturers to formulate their proqucts for free distribution in all EU
states, seem likely to restrict some current practices [113].
At present permitted colours in the UK may, with a few exceptions, be
freely used in any given food product. The EU Directive will specify in
which foods each colour may be used and give maximum levels for their
use. Any new colorant would have to be approved for each individual
product in which it was used under this directive, and it is likely that this
will prove a significant barrier to the introduction of new colorants.
The EU Directive has also highlighted a second area which requires

clarification. The present UK definition of an additive does not include
natural foodstuffs which also have a colouring effect. Thus a manufacturer
who wished to include a meat or a seaweed extract in a product could do so
providing the ingredient was of nutritive value as well as providing colour.
The new directive will prevent this type of additive being used unless its
colouring effect is secondary to its nutritive value.
Despite these legislative problems there is still hope that the very real
values of this class of natural colorants may still be realized. Prominent
amongst these is the fact that seaweeds have recently been approved in
France for human consumption, both as vegetables and condiments [114].
This approval covers some twelve algal species which interestingly include
a broad spectrum of natural colours from the red Chondrus crispus through
browns and greens to the blue Spirulina species. A number of seaweed
based food products are already available in France, some of them
produced by large multinational companies. Many such products represent
the modern counterparts of traditional national foods, for example from
Japan and the former USSR. They are particularly valued in Western diets
as seaweeds contain high levels of so called 'soluble' dietary fibre, which is
regarded as beneficial due to a correlation with the lowering of blood
cholesterol levels.
At a time when Western eating habits are becoming increasingly
international, it would seem wise for us to maintain an open mind on the
introduction of novel food ingredients into our diets. Where those
ingredients may provide both nutritive benefit and useful colour we may
still have much to learn from other cultures.
References
1. Merrit, J.E. and Loening, K.L. Eur. I. Biochem. 108 (1980) 1-30.
2. Dawson, R.M.C., Elliot, D.C., Elliott, W.H. and Jones, K.M. Data for Biochemical
Re,5earch 3rd edn. Clarendon Press, Oxford (1986) pp. 213-231.
3. Battersby, A.R., Fookes, C.J.R., Matchan, G.W.F. and McDonald, E. Nature 285
(1980) 17-21.
4. Stryer, L. Biochemistry, W.H. Freeman & Co., San Francisco (1975).
5. Kikuchi, G., Kumar, A., Talmage, P. and Shemin, D. I. BioI. Chern. 233 (1958) 1214-
1219.
6. Bogorad, L. The Porphyrins Vol. 6 (D. Dolphin, ed.), Academic Press Inc., London
(1979) pp. 125-172.
7. Lemberg, R. and Barrett, J. Cytochromes, Academic Press, London (1973).
8. Brown, S.B. and Troxler, R.F. in Bilirubin Vol. 2 (K.P.M. Heirwegh and S.B. Brown,
eds), CRC Press, Boca Raton, Florida (1982) pp. 11-38.
9. Schmid, R. and McDonagh, A.F. in The Porphyrins Vol. 6 (D. Dolphin, ed.),
10. Hendry, G.A.F., Houghton, J.D. and Brown, S.B. New Phytol. 107 (1987) 255-302.
11. Smith, K.M. (ed.) Porphyrins and Meta/loporphyrins (1975), Elsevier Scientific,
Amsterdam.
12. Hendry, G.A.F., Houghton, J.D. and Jones, O.T.G. Biochern. J. 194 (1981) 743-751.
13. Anson, M.L. and Mirsky, A.E. J. Gen. Physioi. 13 (1930) 469-472.
14. Drabkin, D.L. J. Bioi. Chern. 158 (1945) 721-722.
15. Lewis, U.J. J. Bioi. Chern. 206 (1954) 109.
16. Sumner, J.B. and Dounce, A.L. J. Bioi. Chern. 127 (1938) 439-447.
17. Teale, F.W.J. Biochirn. Biophys. Acta 35 (1959) 543.
18. Shalfejeff, M. Chern. Ber. 18 (1885) 232-233.
19. Fischer, H. Org. Syn. 3 (1955) 442-445.
20. Labbe, R.F. and Nishida, G. Biochirn. Biophys. Acta 26 (1957) 437.
21. Chu, T.C. and Chu, E.J. J. Bioi. Chern. 212 (1955) 1-7.
22. Dixon, H.B.F. and Thompson, C.M. Biochern. J. 107 (1968) 427-431.
23. Morrison, M. and Stotz, E. J. Bioi. Chern. 213 (1955) 373-378.
24. Hamsik, A. and Hofman, M. Hoppe-Seylers, Z. Physiol. Chern. 305 (1956) 143-144.
25. Brown, S.B., Hatzikonstantinou, H. and Herries, D.G. Int. J. Biochern. 12 (1980)
701-707.
26. Perutz, M.F. Trends in Biochern. Sci. 14 (1989) 42-44.
27. Rossi, E. and Curnow, D.H. in HPLC of Small Molecules: A Practical Approach (C.K.
Lim, ed.), IRL Press, Oxford (1986) pp. 260-304.
28. Francis, F.J. Handbook of Food Colorant Patents, Food and Nutrition Press Inc.,
Westport, Connecticut (1986).
29. O'Carra, P., Murphy, R.F. and Killilea, S.D. Biochern. J. 187 (1980) 303-309.
30. Bennet, A. and Siegelman, H.W. in The Porphyrins, Vol. 6 (D. Dolphin, ed.),
31. Troxler, R.F. and Bogorad, L. Plant Physiol. 41 (1966) 491-499.
32. Brown, S.B. Houghton, J.D. and Vernon, D.I. J. Photochern. Photobiol. B. Biology 5
(1990) 3-23.
33. Rudiger, W., O'Carra, P. and O'hEocha, C. Nature 215 (1967) 1477-1478.
34. O'Carra, P. and O'hEocha, C. Phytochern. 5 (1966) 993-997.
35. Chapman, D.J., Cole, W.J. and Siegelman, H.W. Biochern. J. 105 (1967) 903-905.
36. Chapman, D.J., Cole, W.J. and Siegelman, H.W. Am. J. Bot. 55 (1968) 315-316.
37. Bryant, D.A., Glazer, A.N. and Eiserling, F.A. Arch, Microbiol. 110 (1976) 61-75.
38. Morschel, E. and Wehrmeyer, W. Arch. Microbiol. 113 (1977) 83-89.
39. MacCoIl, R., Berns, D.S. and Gibbons, o. Arch. Biochern. Biophys. 177 (1976) 265-
275.
40. Glazer, A.N. and Cohen-Bazire, G. Arch. Microbiol. 104 (1975) 29-32.
41. Svedberg, T. and Katsuari. I.T., J. Am Chern. Soc. 51 (1929) 3573-3583.
42. Chapman, D.J. in The Biology of Blue Green Algae (N.G. Carr and B.A. Whitton, eds),
Univ. California Press, Berkeley (1973) pp. 162-185.
43. Lagarias, J.C., Klotz, A.V., Dallas, J.L., Glazer, A.N., Bishop, J.E., O'Connell, J.F.
and Rapoport, H. J. Bioi. Chern. 263 (1988) 12977-12985.
44. Hattori, A. and Fujita, Y. J. Biochern. 46 (1959) 521-524.
45. Gantt, E. and Lipschultz, C.A. Biochirn. Biophys. Acta 292 (1973) 858-861.
46. Bennett, A. and Bogorad, L. Biochemistry 10 (1971) 3625-3634.
47. Brown, S.B., Houghton, J.D. and Wilks, A. in Biosynthesis of Heme and Chlorophylls
(H.A. Dailey, ed.), McGraw-Hill, New York (1990) pp. 543-575.
48. Kannangara, C.G., Gough, S.P., Bruyant, P., Hoober, J.K., Kahn, A. and von
Wettstein, D. Trends in Biochern. Sci. 13 (1988) 139-143.
49. Beale, S.1. and Weinstein, J.D. in Biosynthesis of Heme and Chlorophylls (H.A. Dailey,
ed.), McGraw-Hill, New York (1990) pp. 287-391.
50. de Lorimer, R., Bryant, D.A., Porter, R.D., Liu, W-Y., Jay, E. and Stevens, S.E. Jr.
Proc. Natl. Acad. Sci. (USA) 81 (1984) 7946-7950.
51. de Lorimer, R., Wang, Y-J. and Yeh, M-L. Proc. Third Ann. Penn. State Syrnp., Plant
Physiol. (1988) pp. 332-326.
52. Brumm, P.J., Fried, J. and Friedmann, H.C. Proc. Natl. Acad. Sci. (USA) 80 (1983)
3943-3947.
53. Bogorad, L. Ann. Rev. Plant Physiol. 26 (1975) 369-401.
54. Fujita, Y. and Hattori, A. Plant and Cell Physiol. 1 (1960) 293-303.
55. Tandeau de Marsac, N., Mazel, D., Damerval, T., Guglielmi, G., Capuano, V. and
Houmard, J. Photosyn. Res. 18 (1988) 99-132.
56. Wood, N.B. and Haselkorn. in Limited Proteolysis in Microorganisms (G.N. Cohen and
H. Holzer, eds), US DHEW Publication no (NIH) Bethesda, MD (1979) pp. 79-1591.
57. Bennet, A. and Bogorad, L. 1. Cell BioI. 58 (1973) 419-435.
58. Tandeau de Marsac, N. 1. Bacteriol. 130 (1977) 82-91.
59. Canto de Loura, I., Dubacq, J.P. and Thomas, J.C. Plant Physiol. 83 (1987) 838-843.
60. Schmidt, A., Erdle, I. and Kost, H-P. Z. Naturforsch. 37c (1982) 870-876.
61. Brody, M. and Emerson, R. Am. 1. Bot. 46 (1959) 433-440.
62. Myers, J. and Kratz, W.A. 1. Gen. Physiol. 39 (1955) 11-22.
63. Jones, L.W. and Myers, J. 1. Phycol. 1 (1965) 7-14.
64. Nichols, K.E. and Bogorad, L. Botan. Gaz. 124 (1962) 85-93.
65. Gantt, E. Int. Rev. Cytol. 66 (1980) 45-80.
66. Gantt, E. Ann. Rev. Plant Physiol. 32 (1981) 327-347.
67. Grossman, A.R., Lemaux, P.G. and Conley, P.B. Photochem. Photobiol. 44 (1986)
827-837.
68. Brown, A.S., Offner, G.D., Ehhart, M.M. and Troxler, R.F. 1. Bioi. Chem. 254 (1979)
7803-7811.
69. Glazer, A.N. Ann. Rev. Biochem. Biophys. 14 (1985) 47-77.
70. Porter, G., Tredwell, C.J., Searle, G.F.W. and Barber, J. Biochim. Biophys. Acta. 501
(1978) 532-545.
71. Chen, C-H. and Berns, D.S. Biophys. Chem. 8 (1978) 203-213.
72. Troxler, RF. in Chemistry and Physiology of Bile Pigments (P.D. Berk and M.1. Berlin,
eds), Dept. Health Education and Welfare, US, pp. 431-454.
73. Brown, A.S. and Troxler, R.F. Biochem. 1. 163 (1977) 571-581.
74. Glazer, A.N. and Fang, S. 1. Bioi. Chem. 248 (1973) 659-662.
75. Troxler, RF., Kelly, P. and Brown, S.B. Biochem. 1. 172 (1978) 569-576.
76. Brown, S.B., Holroyd, J.A., Troxler, R.F. and Offner, G.D. Biochem. 1. 194 (1981)
137-147.
77. Berns, D.S. and MacColl, R. Chem. Rev. 89 (1989) 807-825.
78. Rudiger, W. in Pigments in Plants (F-C., Czygan, ed.), Akademie-Verlag, Berlin (1981)
pp. 314-351.
79. MacColl, R, Lee, J.J., Berns, D.S. Biochem. 1. 122 (1971) 421-426.
SO. Neufeld, G.J. and Riggs, A.F. Biochim. Biophys. Acta 181 (1969) 234-243.
81. Davis, L.C., Radke, G.A., Guikema, J.A. 1. Liq. Chromatogr. 9 (1986) 1277-1295.
82. van der Velde, H.H. Biochim. Biophys. Acta 303 (1973) 246-257.
83. O'Carra, P. and O'hEocha, C., in Chemistry and Biochemistry of Plant Pigments 2nd
edn. Vol. 2 (T.W. Goodwin, ed.), Academic Press, London (1976) pp. 328-376.
84. Fujimori, E. and Pecci, J. Biochemistry 5 (1966) 3500-3508.
85. Fujimori, E. and Pecci, J. Biochim. Biophys. Acta 207 (1970) 259-261.
86. Pecci, J. and Fujimori, E. Biochim. Biophys. Acta 154 (1968) 332-341.
87. Murphy, R.F. and O'Carra, P. Biochim. Biophys. Acta 214 (1970) 371-373.
88. Boucher, L.J., Crespi, H.L. and Katz, J.J. Biochemistry 5 (1966) 3796-3802.
89. Sheldrick, W.S. 1. Chem. Soc. Perkin Trans. 2 (1976) 1457-1461.
90. Falk, H. and Grubmayr, K. Angew. Chem. Int. Ed. 16 (1977) 470-471.
91. McKavanagh, S. and Bi11ing, B.H. in HPLC of Small Molecules: A Practical Approach
(C.K. Lim, ed.), IRL Press, Oxford (1986) 305-323.
92. Dawson, L., Houghton, J.D., Vernon, DJ. and Brown, S.B. Mol. Asp. Med. 11 (1989)
133-134.
93. Parkinson, G., Shota, U., Hunter, D. and Sandler, N. Chem. Eng. 94 (1987) 19-23.
94. Borowitzka, M.A. and Borowitzka, L.J. Micro-algal Biotechnology, Cambridge Univer-
sity Press, Cambridge (1988).
95. Cohen, Z. in CRC Handbook of Microalgal Mass Culture (A. Richmond, ed.), CRC
Press, Boca Raton (1986) pp. 421-454.
96. Borowitzka, M.A. in Microalgal Biotechnology (M.A. Borowitzka, and L.J.
Borowitzka, eds), Cambridge University Press (1988) pp. 173-175.
97. Lee, Y.K. Trends in Biotechnol. 4 (1986) 186-189.
98. Richmond, A. in Microalgal Biotechnology (M.A. Borowitzka and L.J. Borowitzka,

eds), Cambridge University Press, Cambridge (1988) pp. 85-97.
99. Tomasseli, L., Giovannetti, L., Sacchi, A. and Bocci, F. in Algal Biotechnology (T.
Stadler et al., eds), Elsevier Applied Science, London (1988) pp. 305-314.
100. Vonshak, A. in Microalgal Biotechnology (M.A. Borowitzka and L.J. Borowitzka, eds),
Cambridge University Press, Cambridge (1988) pp. 130-133.
101. Robinson, P.K., Mak, A.L. and Trevan, M.D. Process Biochem. 21 (1986) 122-127.
102. Thepenier, c., Chaumont, D. and Gudin, C. in Algal Biotechnology (T. Stadler et al.,
eds), Elsevier Applied Science, London (1988) pp. 413-421.
103. Francis, F.J. Food Technol. April (1987) 62--68.
104. Taylor, A.J. in Developments in Food Colours-2 (J. Walford, ed.), Elsevier Applied
Science, London (1984) pp. 159-206.
105. Wadds, G. in Developments in Food Colours-2 (1. Walford, ed.), Elsevier Applied
Science, London (1984) pp. 23-74.
106. Cohen, Z. in CRC Handbook of Microalgal Mass Culture (A. Richmond, ed.), CRC
Press Inc., Boca Raton, Florida (1986) 421-454.
107. Jacobson, G.K. and Jolly, S.O. in Biotechnology Vol. 7b (R-J. Rehm and G. Reed,
eds), VCR Publishers, Weinheim (1989).
108. Glazer, A.N. and Stryer, L. Trends in Biochem. Sci. 9 (1984) 423-427.
109. Vernon, T.O., Glazer, A.N. and Stryer, L. 1. Cell. BioI. 93 (1983) 981-986.
110. Kronick, M.N. 1. lmmunol. Meth. 92 (1986) 1-13.
111. Walford, J. (ed.) Developments in Food Colours-2, Elsevier Applied Science, London
(1984).
112. Perkin, J. Food Processing 1 (1994) 28-30.
113. Law, M. Food Processing 2 (1995) 38.
114. Mabeau, S. and F1eurence, J. Trends in Food Science and Technology 4 (1993) 103-107.
7 Carotenoids
G. BRITTON
7.1 Summary
The carotenoids are very widespread natural pigments in plants and

animals, so they provide the natural yellow, orange or red colours of many
foods as well as being used extensively as non-toxic natural or nature-
identical colorants. This chapter discusses all aspects of carotenoids,
especially in relation to their presence or use in food. Topics discussed
include the distribution of carotenoids, particularly in plants and animals,
the natural functions and commercial uses and applications of carotenoids,
the pathways and regulation of carotenoid biosynthesis, the absorption and
transport of carotenoids and their metabolism, especially into vitamin A.
Examples are given of the use of natural carotenoid extracts (e.g. annatto)
and synthetic carotenoids (e.g. (3-carotene, canthaxanthin, 8' -apo-(3-
caroten-8' -al) as colorants in food, feed, cosmetics and medical and health
products. The properties and stability of carotenoids in the free form and
in oil-based and water-dispersible formulations are assessed. All practical
aspects of the handling, isolation and analysis of carotenoids are outlined,
with emphasis on their purification and identification by chromatography,
including HPLC, and their characterization and quantitative determination
by physicochemical methods such as UV-visible and NMR spectroscopy
and mass spectrometry. Future prospects for the production and use of
carotenoids as colorants are assessed, and substantial advances in the
biological and biotechnological production of carotenoids by micro algae or
bacteria are predicted.
7.2 Introduction
7.2.1 General publications

In a short chapter such as this, only an outline of the main features of
carotenoids and their applications can be given. Other publications are
available which give a more extensive coverage of the carotenoid field.
Although it was published more than 20 years ago, the classic monograph
edited by Isler [1] remains an essential reference book containing a vast
store of information about the chemistry, biochemistry, properties, func-

tions and applications of carotenoids. Complementary to this is a book
edited by Bauernfeind [2] which deals with the industrial and technological
aspects of carotenoids in relation to their applications as colorants and
vitamin A precursors. A new series of books on carotenoids is designed to
survey more recent developments, with emphasis on practical guidance.
The two parts of the first volume in this series [3, 4] give a comprehensive
survey of methods of isolation, analysis and spectroscopy of carotenoids. A
Key to Carotenoids has been produced [5] which lists the structures and
semi-systematic names of all natural carotenoids known up to 1986 and
gives extensive references to original publications of their isolation,
spectroscopic properties and synthesis. An appendix to this Key [6] lists
new carotenoids identified since 1986.
In addition to the above, the published proceedings of International
Symposia on Carotenoids, held every three years since 1966, provide a
regular update of progress in all areas of carotenoid science [7-16].
7.2.2 Structure and nomenclature

Carotenoid hydrocarbons collectively are called carotenes. Derivatives that
contain oxygen functions (most commonly hydroxy, keto, epoxy, methoxy
or carboxylic acid groups) are called xanthophylls. Some carotenoids are
acyclic (e.g. lycopene) but more common are those that contain a six-
membered (or occasionally five-membered) ring at one end or both ends of
the molecule. The structure and numbering of the parent acyclic and
dicyclic carotenoids, lycopene and j3-carotene, are illustrated in Figure 7.1.
Traditionally, natural carotenoids have been given trivial names that are
usually derived from the name of the biological source from which they
17 19 20
It
16
19' 17'
Lycopene
~
:r
2'
3
4 18
B-Carotene
Figure '_I Structures and numbering scheme for the parent acyclic and dicyclic carotenoids,
Jycopene and ~-carotene.
CAROTENOIDS 199
C,! end-group types
.'
/(
16
17 17
18
Figure 7.2 Structures of the seven carotenoid end-group types.
were first isolated. In recent years, however, a semi-systematic nomen-

clature that conveys structural information has been devised. According to
this scheme, the carotenoid molecule is considered in two halves. Each
individual compound is then named as a derivative of the parent carotene,
as specified by the Greek letters that describe its two end groups, with
conventional prefixes and suffixes being used to indicate changes in
hydrogenation level and the presence of substituent groups. The seven end
groups that have been recognized are illustrated in Figure 7.2.
Compounds from which an end group has been removed from the
normal C40 structure are known as apocarotenoids; some of these are
important food colorants. In keeping with current practice, trivial names
will be used in this article for common carotenoids, but the semi-systematic
names of these are given in Table 7.1 (see also Appendix for the structures
of individual carotenoids mentioned in the text).
Because of the extensive double-bond system in the molecule, any carot-
enoid can, theoretically, exist in many geometrical isomeric forms (Z-E
isomers). Most carotenoids occur naturally in the all-E (all-trans) form, but
Z-isomers (cis-isomers) are frequently present in small amounts, and are
easily formed as artifacts from the all-E isomer.
A comprehensive description of carotenoid structures, stereochemistry
and nomenclature, including the current rules for applying the IUPAC
Table 7.1 Semi-systematic names of carotenoids mentioned in the text (stereochemistry is

not specified)
Trivial name Semi-systematic name
Actinioerythrol 3,3' -dihydroxy-2,2' -dinor-~,~-carotene-4,4' -dione

Antheraxanthin 5 ,6-epoxy-5 ,6-dihydro-~,~-carotene-3,3' -diol
Astaxanthin 3,3' -dihydroxy-~,~-carotene-4,4' -dione
Bixin methyl hydrogen 9' -Z-6,6' -diapocarotene-6,6' -dioate
Canthaxanthin ~,~-carotene-4,4' -dione
Capsanthin 3,3' -dihydroxy-~,K-caroten-6' -one
Capsorubin 3,3' -dihydroxY-K,K-carotene-6,6' -dione
a-Carotene ~ ,E-carotene
~-Carotene ~, ~-carotene
"V-Carotene ~,IjI-carotene
I)-Carotene E,IjI-carotene
E-Carotene E,E-carotene
,-Carotene 7,8,7' ,8' -tetrahydro-IjI,IjI-carotene
Citranaxanthin 5' ,6' -dihydro-5' -apo-18' -nor-~-caroten-6' -one
Croce tin 8,8' -diapocarotene-8,8' -dioic acid
a-Cryptoxanthin ~,E-caroten-3-ol
~-Cryptoxanthin ~,~-caroten-3-ol
Isocryptoxanthin ~, ~-caroten-4-ol
Isozeaxanthin ~,~-carotene-4,4' -diol
Lactucaxanthin E,E-carotene-3,3' -diol
Lutein ~,E-carotene-3,3' -diol
Lutein-5,6-epoxide 5 ,6-epoxy-5 ,6-dihydro-~ ,E-carotene-3,3' -diol
Lycopene IjI,IjI-carotene
Neoxanthin 5' ,6' -epoxy-6,7-didehydro-5,6,5' ,6' -tetrahydro-~,~-carotene-3,5,3' -triol
Neurosporene 7 ,8-dihydro-IjI ,IjI-carotene
Norbixin 6,6' -diapocarotene-6,6' -dioic acid
Phytoene 7,8,11,12,7' ,8' ,11' ,12' -octahydro-IjI,IjI-carotene
Phytofluene 7,8,11,12,7' ,8' -hexahydro-IjI,IjI-carotene
Violaxanthin 5,6:5' ,6' -diepoxy-5,6,5' ,6' -tetrahydro-~,~-carotene-3,3' -diol
Violerythrin 2,2' -dinor-~, ~-carotene- 3,4 ,3' ,4' -tetrone
a-Zeacarotene 7' ,8' -dihydro-E,IjI-carotene
~-Zeacarotene 7' ,8' -dihydro-~,IjI-carotene
Zeaxanthin ~,~-carotene-3 ,3' -diol
semi-systematic nomenclature, is given by Weedon and Moss [17], and an

older, though largely still acceptable, version of the rules is included in
Isler [1].
7.3 Distribution, natural functions and actions
7.3.1 Distribution
Carotenoids are often considered to be plant pigments alone but they also
occur widely in bacteria, fungi, algae and in animals, especially birds, fish
and invertebrates. Two volumes of Goodwin [18, 19] provide extensive
details of the distribution of carotenoids in plants and animals, respect-
ively.
CAROTENOIDS 201
7.3.1.1 Plants. In plants, the carotenoids occur universally in the

chloroplasts of green tissues, but their colour is masked by the chloro-
phylls. The leaves of virtually all species contain the same main carot-
enoids, that is ~-carotene (usually 25 to 30% of the total), lutein (around
45%), violaxanthin (15%) and neoxanthin (15%). Small amounts of (X-
carotene, (X- and ~-cryptoxanthin, zeaxanthin, antheraxanthin and lutein-
5,6-epoxide are also frequently present, and lactucaxanthin is a major
xanthophyll in a few species, notably lettuce (Lactuca sativa).
There is considerable variation in the total carotenoid content in leaves
of different species and varieties; dark green leaves such as spinach contain
the largest amounts of carotenoids. The carotenoid composition of leaves
of most species are similar quantitatively, although some changes are seen
in plants under stress.
Carotenoids are also widely distributed in non-photosynthetic tissues of
plants and are responsible for the yellow, orange and red colours of many
flowers and fruit. They are usually located in chromoplasts, and the
xanthophylls are frequently present as complex mixtures of fatty acyl
esters.
Several distinctive carotenoid patterns have been recognized in fruit:
• the normal collection of chloroplast carotenoids
• large amounts of lycopene and its hydroxy-derivatives (e.g. tomato)
• large amounts of ~-carotene and its hydroxy-derivatives (e.g. peach)
• collections of 5,6- or 5,8-epoxycarotenoids (e.g. carambola)
• large amounts of apocarotenoids (e.g. Citrus spp.)
• some specific or unusual carotenoids (e.g. capsanthin in red pepper,
Capsicum annuum)
Although carotenoids are not widely found in roots, there are some well-
known examples, for example carrot and sweet potato, where large
amounts of carotenoids, mainly carotenes, are present. Similarly, some
seeds are coloured by carotenoids, for example zeaxanthin in maize. One
particular example, Bixa orellana (annatto) contains very large amounts of
the apocarotenoid bixin in its seed coat (up to 10% dry weight).
Most yellow flowers are coloured by carotenoids; here xanthophyll
epoxides are common. Natural extracts of some flowers, such as marigold
(Tagetes erecta) contain large amounts of lutein and are used for color-
ation.
7.3.1.2 Animals. In animals, although carotenoids are found in some

birds and fish, they are particularly associated with invertebrate animals. In
birds, carotenoids colour some yellow or red feathers but they are also
important for skin colour in chickens and especially in egg yolk. In fish,
important examples are the flesh of salmon and trout (astaxanthin and
canthaxanthin). In many marine invertebrate animals (e.g. shrimps, crabs
and lobsters), astaxanthin and related carotenoids may be present in large

amounts, often as carotenoprotein complexes, which are green, purple or
blue in the living animal but are denatured to reveal the red carotenoid
colour when the animal is cooked.
7.3.2 Natural functions
The carotenoids owe their distinctive properties and functions to the

presence of a long chromophore of conjugated double bonds, which gives
them their light-absorbing (colour) properties and renders the molecules
extremely susceptible to oxidative degradation.
In flowers, fruits and many animals, the function of carotenoids is simply
to provide colour. In green plant tissues, they play important roles in
photosynthesis [20-23]. All aspects of carotenoids in photosynthesis are
surveyed in a new monograph [24]. The carotenoids are located, in the
chloroplast thylakoid membranes, in the pigment-protein complexes of
photosystems 1 and 2, where they serve as accessory light-harvesting
pigments (mainly the xanthophylls). They are particularly important as
protectants against photo-oxidation by singlet oxygen, 102 . When more
light energy is absorbed by the light-harvesting chlorophylls than can be
used to drive photosynthesis, some excited chlorophyll molecules will
undergo intersystem crossing to give the slightly lower energy but longer-
lived triplet excited state, 3Chl. The energy of the 3Chl can be transferred
to oxygen to give singlet oxygen, 102 , which is a highly reactive species that
can rapidly cause damage to lipids, membranes and tissues. Carotenoids,
particularly j3-carotene in the chloroplast pigment-protein complexes, give
protection by quenching the energy of 3Chl and thus preventing the
formation of 102 , and they can also quench the energy of 102 if any should
be formed. Carotenoids can similarly protect non-photosynthetic organ-
isms or tissues against photo-oxidation by 102 since this can also be
produced in the presence of other suitable photosensitizers.
7.3.3 Actions of carotenoids in human health
The nutritional importance of j3-carotene and other carotenoids containing

an un substituted j3-ring as provitamin A has been known for many years.
These carotenoids are cleaved enzymically in the intestine to yield retinal
and are the main source of vitamin A for most people [25]. j3-Carotene has
also been used successfully for the treatment of erythropoietic proto-
porphyria, a condition in which the patients are extremely sensitive to light
because of an abnormality in haem synthesis, which leads to the accumu-
lation of free porphyrins in the skin. The free porphyrin molecules act as
CAROTENOIDS 203
photosensitizers, in the same way as chlorophyll in plants, and cause the

production, in the skin, of 102 which causes tissue damage and inflamma-
tion. ~-Carotene can afford substantial protection against this and alleviate
the unpleasant symptoms [26, 27].
In recent years there have been many reports to suggest that ~-carotene
may also be important in affording protection against cancer, heart disease
and other conditions, and that it may have anti-ageing benefits [28-31].
These reports indicate that there may be a physiological role for intact
carotenoids, independent of their conversion into vitamin A. In 1981, a
now classic paper by Peto et al. [28] discussed the possibility that dietary ~
carotene could afford protection against some forms of cancer. Several
studies since then have concluded that individuals whose dietary intake and
serum levels of ~-carotene are low have higher risk of developing some
cancers, for example of the lung, skin and bladder. Similar correlations
have been found for low carotene levels and increased risk of coronary
heart disease. Experiments with animals as model systems have strongly
suggested that the administration of carotene can give substantial protec-
tion against the development and proliferation of induced cancers.
It is usually considered that the beneficial protective actions of ~
carotene are due to its antioxidant properties. Under appropriate con-
ditions, ~-carotene can function as an effective antioxidant [32], not only
against 102 but also against lipid peroxidation and the highly destructive
hydroxyl radical HO' that is implicated in many diseases. It is by no means
sure, however, that carotenoids can act as antioxidants at the low
concentrations found in vivo.
Evidence is now being obtained which suggests that ~-carotene may
have direct stimulatory effects on the immunoresponse system [33, 34].
These effects, which mayor may not involve an antioxidant action, could
form the basis of a protective action against cancer and also against HIV
and AIDS [35].
7.4 Biosynthesis
Carotenoids are biosynthesized in higher plants, algae, fungi and bacteria.

Animals cannot biosynthesize them de novo, although many are able to
metabolize and modify structurally some carotenoids that they ingest.
Many review articles on carotenoid biosynthesis are available, and
should be consulted for details of the pathways, reactions, enzymes and
regulation [36-43].
Rapid progress is being made in studies of the molecular genetics of
carotenoid biosynthesis, especially in bacteria but also now in higher plants
[40, 44]. The clusters of genes responsible for carotenoid biosynthesis in
the phototrophic bacteria Rhodobacter capsulatus and Rhodobacter
e'UD"r--
Mevalonic acid (MVA)
!
Isopentenyl diphosphate
Dimethylallyl
diphosphate
(DMADP)CS
(GD'l--- m'
Geranyl diphosphate Sterols
Tr'""
Monoterpenes ••- - - -
e,"
Sesquiterpenes ••- - - Farnesyl diphosphate - - -.t. Squalene

CI5(FDP)
---- r
!--IDP
Diterpenes Geranylgeranyl diphosphate Phytoene
I
C20(GGDP)
Phytol
I
m
,
Long-chain polyprenols
Ubiquinone
Chlorophyll Plastoquinone CAROTENOIDS
Figure 7.3 The isoprenoid biosynthetic pathway.
sphaeroides have been isolated and the individual genes identified and
sequenced [45, 46]. A cluster of genes responsible for the biosynthesis of
zeaxanthin and its glucoside in the non-phototrophic bacterium Erwinia
has also been isolated and the genes cloned in Escherichia coli [47]. The
carotenoid biosynthesis genes of higher plants are not clustered, but some
have been isolated, notably from the tomato [48].
The carotenoids are isoprenoid compounds and are biosynthesized from
acetyl coenzyme-A via mevalonic acid as a branch of the great isoprenoid
or terpenoid pathway (Figure 7.3). The main stages of carotenoid bio-
synthesis are outlined in Figure 7.4. The early stages of the pathway are
common to the biosynthesis of all isoprenoid compounds. The first step
that is specific to carotenoids is the formation of the first C40 hydrocarbon,
phytoene, from two molecules of geranylgeranyl diphosphate (GGDP) via
prephytoene diphosphate (PPDP). In plants, the phytoene that is formed
appears to be the 15Z isomer, although the 15E form is made by some
bacteria directly. The colourless phytoene (with three conjugated double
bonds) then undergoes a series of de saturation reactions (Figure 7.5), each
of which introduces a new double bond and increases the chromophore by
two conjugated double bonds (c.d.b.). The intermediates are phytofluene
(5 c.d.b.), t-carotene (7 c.d.b.), neurosporene (9 c.d.b.) and the normal
final product, lycopene (11 c.d.b.).
CAROTENOIDS 205
1
MVA
E"" ...... ----------
1
GGDP
Phytoene
CormatIOn
-----------
Phytoene
D"" .."'oo---------- 1
1
Lyeopene
c,,';,.;;oo ----------
f3-Carotene
""''''''''°"'''---------1
Xanthophyll.
Figure 7.4 Summary of the main stages of carotenoid biosynthesis.
Lycopene can then undergo cyclization at one end or both ends of the
molecule to give the monocyclic 'Y-carotene and 8-carotene and the dicyclic
~-carotene, a-carotene and E-carotene (Figure 7.6). Similar cyclization
of the more desaturated end group of neurosporene gives ~- and a-
zeacarotenes. The cyclizations occur by the mechanism shown in Figure
7.7, in which alternative losses ofH+ from C-6, C-18 or C-4 give rise to the
~-, 'Y- or e-rings, respectively.
Oxygen functions are normally introduced as the final steps in the bio-
synthesis. Thus the common 3-hydroxy-xanthophylls lutein and zeaxanthin
are formed by the stereospecific hydroxylation of the corresponding
carotenes (a- and ~-carotene) by mixed-function oxidase enzymes (Figure
7.8).
Pathways have been proposed for the formation of most of the structural
variations that are found in carotenoids, for example, of several different
end groups in plant carotenoids from a 5,6-epoxy-~-ring (Figure 7.9). In
most cases, however, there is little or no biochemical evidence to support
these proposals.
The apocarotenoids, including the apo-~-carotenals, crocetin and bixin,
that occur in plants are believed to be biosynthesized by excision of
fragments from the ends of a C40 carotenoid molecule. Again, there is no
direct biochemical evidence to support these reasonable suggestions.
Phytoene
2H1
Phytofluene
r 1
r-Carotene
or
7, 8, 11, 12-Tetrahydro-",', ",-carotene
Neurosporene
Lycopene
Figure 7.5 The sequence of desaturation reactions involved in carotenoid biosynthesis.

'Y - Carotene (:J- Carotene
I
Lycopene i
a- Carotene :2
~
E- Carotene
0- Carotene
Figure 7.6 The biosynthesis of monocycIic and dicycIic carotenes from lycopene. s
/.\
~..
{3 -ring
ec--. €-ring
Cc'"
r-ring
Figure 7.7 Mechanisms for the alternative formation of ~-, "Y- and E-rings from the acyclic
precursor, Iycopene.
H
)
~~P _-..c;..)_ H~
', p
,,.:::; ,
Ii HO
Figure 7.S Stereospecific hydroxylation of carotenes.
7.4.1 Regulation of biosynthesis

In plants, carotenoid biosynthesis occurs universally as an essential part of
the construction of chloroplasts, but it is also particularly associated with
chromoplast development as fruit ripen or flowers open. Thus, in Capsi-
cum or tomato fruit, the carotenoid content increases several fold during
ripening.
CAROTENOIDS 209
~ ~~, ~'
~"O I / 0
~...
M'
/ " I ~ L
~
/
~ '. ~... ~c~·-·
--OH '.
HO~~~~ HoM-oH
Figure 7.9 Scheme for the formation of different end groups of plant carotenoids.
Carotenoid contents and compositions in plants are affected by environ-

mental and nutritional conditions, and can be altered drastically by
treatment with a range of chemicals, especially amines. The genetics of
carotenoid biosynthesis in tomatoes has been studied intensively and many
strains with different carotenoid compositions have been produced.
7.5 Absorption, transport and metabolism
There is very wide variation in the ability of different animal species and
even of different individuals to absorb and metabolize carotenoids [19, 49].
Humans are generally good and indiscriminate absorbers of all carotenoids
that are offered in the diet. Many other mammals, for example the cat,
rat, mouse and sheep, are very poor absorbers and metabolizers. Cattle
efficiently absorb l3-carotene but not xanthophylls; l3-carotene colours the
fat and may accumulate in high concentration in the corpus luteum, and it
is considered to be implicated in maintaining fertility in cattle. Although
most work on the absorption and metabolism of carotenoids has been
performed on rats and mice, neither these nor, with the possible exception
of other primates, other mammals provide a reliable model system that is
applicable to humans.
In birds, l3-carotene is absorbed poorly but xanthophylls are absorbed in
large amounts and transported to the egg yolk. They may also colour skin
(in chickens) and feathers; indeed, flamingos need a constant supply of

oxocarotenoids to maintain their characteristic pink colour.
Although they cannot synthesize carotenoids, birds, fish and inverte-
brate animals, can, in many cases, introduce structural modifications into
the carotenoids they obtain from the diet. Many oxidative and reductive
metabolic pathways have been proposed for the transformations of
carotenoids in these organisms, the most important processes perhaps
being those that introduce oxygen functions into the j3-ring to produce
astaxanthin. In almost all cases, the proposed pathways are based only on
the structures of the compounds present and direct conversions have not
been demonstrated. None of the enzymes involved has been isolated. In
comm~rcially important examples such as salmon, these transformations
do not occur; satisfactory pink-fleshed salmon are only obtained when
astaxanthin itself is supplied in the diet [50].
Dietary carotenoids are absorbed in the gut along with other lipids. A
most important process, following absorption, is the conversion of 13-
carotene and other suitable molecules, that is those having one unsubsti-
tuted j3-ring, into vitamin A (retinol) [25]. This conversion occurs mainly in
the small intestine, although there are reports of its occurrence also in
other tissues, for example liver and kidney. The accepted mechanism of
conversion is central cleavage, which requires a 15,15' -dioxygenase
enzyme and can provide two molecules of retinal from one molecule of 13-
carotene. In addition to central cleavage, it is likely that excentric cleavage
by oxidation of other double bonds also occurs to give apocarotenals of
different chain lengths (C 22 to C30), which undergo further oxidation to
give retinal (one molecule from each molecule of j3-carotene). The retinal
that is produced by either of these processes is then reduced enzymically to
retinol, which is transported by retinol-binding protein to its required sites
of action or to the liver for storage as acyl esters. The formation of vitamin
A from the parent carotenoid is regulated so that excessive amounts of the
highly toxic vitamin are not produced.
Carotenoids that are absorbed intact are also transported on blood
lipoproteins, and can be deposited in tissues [27, 51]. j3-Carotene and other
carotenes are associated mainly with the low-density lipoprotein, the
various xanthophyl1s in the human diet are transported mainly on the high-
density lipoprotein fraction [52]. When excessive amounts of carotenoids
are supplied they may be deposited in many tissues and can accumulate in
concentrations sufficient to cause yellow-orange coloration of the skin,
especially that of the hands and feet. Although this normally happens only
in individuals who ingest a large amount of pure carotenoid, caroteno-
dermia is also sometimes seen in people who eat large amounts of, for
example, carrots and oranges which have a high carotenoid content.
Coloration by carotenoids is generally harmless and reversible, although
there are cases where the intake of large amounts of canthaxanthin
CAROTENOIDS 211
(200 mg/day) in the form of oral 'sun-tanning' capsules not only imparts
the sought-after bronzed (although somewhat orange) appearance of the
skin, but has resulted in the appearance of crystals of canthaxanthin in the
eye. There is, fortunately, no risk of this happening with the much smaller
amounts (microgram quantities) of canthaxanthin that are ingested as food
colorants. The absorption of ~-carotene and other carotenoids from plant
food sources is described as 'poor to fair', while that of pure ~-carotene is
rather better. Absorption is facilitated by the presence of fats and agents
such as lecithin, which aid emulsification. Virtually nothing is known about
the degradative metabolism of carotenoids in humans.
Carotenoids can persist in the human body for a considerable time. ~
Carotene has a half-life of a few days [53] but canthaxanthin, if large
amounts have been taken, can persist for 6 to 12 months. The ~-carotene
status and general dietary carotenoid profile of an individual can be
monitored rapidly by HPLC analysis of a small sample of blood or serum.
7.6 Natural and synthetic carotenoids as colorants
Carotenoids and carotenoid-containing extracts have been widely used for

centuries as colorants in food. The coloration may be introduced by direct
addition to the foodstuff or indirectly by feeding to animals (e.g. chickens
and fish), which thus become suitably coloured ready for use as food. All
aspects of the use of carotenoids as colorants are covered in detail in
Bauernfeind [2].
7.6.1 Natural carotenoids and extracts

The natural extracts first used were those of annatto, carrot, palm oil,
saffron, tomato and paprika. The powdered, dried plant materials and
extracts of them have been used for many years. These are not, however,
pure carotenoids or simply mixtures of carotenoids but generally contain
large amounts and large numbers of other, mainly unidentified substances.
7.6.1.1 Paprika. One of the oldest and most important such extracts is
that of paprika (Capsicum annuum), which is produced as a dry powder or
an oil extract or oleoresin. These products provide hot and spicy flavour as
well as colour, and their use is thus limited to suitably savoury products.
The main carotenoids present are capsanthin and capsorubin (largely as
acyl esters) but there are many other minor components.
7.6.1.2 Annatto. The use of annatto also has a long history. Annatto is
obtained as extracts of the red-brown resinous coating of the seeds of Bixa
orellana, a tree that grows abundantly in the tropics. The major pigment
present is the apocarotenoid bixin (9-cis) and this methyl ester is the main
component in oil-based preparations. Hydrolysis (saponification) liberates
the dicarboxylic acid, norbixin, salts of which provide the water-soluble
annatto preparations. The seedcoat contains a high concentration of bixin
together with small amounts of many unidentified but probably related
compounds. The biosynthesis of bixin has not been elucidated but is
presumed to follow the normal C40 pathway to lycopene (normal C4Q
phytoene, phytofluene and ~-carotenes are present). Many preparations of
annatto are available with different hues (usually pinkish) and are used to
colour a wide range of food products.
7.6.1.3 Saffron. One of the earliest additives, saffron is the powdered

dried flowers of Crocus sativus. The main constituent is crocin, the bis-
gentiobioside (and other glycosides) of the diapocarotenedioic acid,
crocetin. Saffron is used particularly to impart a pure yellow colour to rice
and other foods. It is also used as a spice.
7.6.1.4 Tomato. Extracts of tomato (Lycopersicon esculentum) have a

high lycopene content (80 to 90% of total carotenoid). Their use as
colorants is limited because of the strong tomato flavour.
7.6.1.5 Xanthophyll pastes. These usually consist of concentrated

extracts of leaves (e.g. nettles, alfalfa and broccoli), and are green rather
than yellow unless saponification has been used. Many 'xanthophyll' pastes
contain as much as 30% carotene. Other xanthophyll extracts are obtained
from flowers. Marigold extracts (Tagetes erecta) , or the dried, crushed
flowers themselves, are very rich in lutein and are used to pigment egg-yolk
and chickens.
7.6.1.6 Palm oil. The red oil obtained from the fruit of the tropical
palm tree Elaeis guineensis contains 500 mg of carotenoid, mainly 13-
carotene, per kilogram of oil.
7.6.2 Synthetic carotenoids

Following the successful commercial synthesis of vitamin A, a similar
procedure was used for the production of l3-carotene, which was intro-
duced onto the market in 1954. Several companies now manufacture
l3-carotene by processes based on Grignard reactions, enol ether condens-
ations, ethynylation, Wittig and sulphone reactions, the total annual
output now exceeding 500 tonnes. Application of the same synthetic
methods subsequently led to the commercial production of 8' -apo-l3-
caroten-8' -aI, 8' -apo-l3-caroten-8' -oic acid ethyl or methyl ester and
CAROTENOIDS 213
citranaxanthin, which are also widely used. More recently canthaxanthin

has entered production, being made originally by oxidation of l3-carotene,
later by direct synthesis. Commercial production of the natural optical
isomers of zeaxanthin (3R, 3' R) and astaxanthin (3S, 3' S) is now under
development [54].
The commercial synthetic products are crystalline and of high purity and
they are identical in all respects to the natural substances. They are
marketed as crystals, as micronized suspensions in oil and as water-
dispersible formulations.
7.7 General properties and stability
Carotenoids crystallize in a variety of forms, and the crystals vary in colour

from orange-red to violet and almost black, depending on their shape and
size. The melting points are high, usually in the range 130 to 220°C. Even
the crystals are very sensitive to oxidative decomposition when exposed to
air, and they must be stored in an inert atmosphere or under vacuum. The
pure carotenoids are greatly stabilized by suspension or solution in
vegetable oil, especially in the presence of antioxidants such as (l-
tocopherol. Peroxidation of unsaturated lipids in the oil can, however,
cause the rapid oxidative destruction of carotenoids. Carotenoids in situ
are also very susceptible to enzymic degradation by lipoxygenases when
tissues are disrupted.
Carotenoid crystals tend to have only poor solubility. They are insoluble
in water, only slightly soluble in vegetable oils and very soluble only in
chlorinated solvents (e.g. chloroform and dichloromethane). The crystals
are generally slow to dissolve, although rates of solution increase with
heating.
7.8 General procedures for carotenoid work
Recommended methods for working with carotenoids are described in

several articles [55-58]. The new Carotenoids books give details and critical
evaluation of many methods, and present some worked examples of
recommended procedures [3].
Carotenoids are rather unstable substances and precautions have to be
taken when handling them to avoid degradation and the formation of
artifacts [59]. The conjugated polyene chromophore of carotenoids renders
them sensitive to oxygen, light and heat. Other structural features are
easily modified by acid or alkali. Stringent precautions must be observed,
therefore, if losses of material or unwanted structural changes are to be
avoided. Speed of manipulation is important. All procedures that inevit-
ably introduce risks of oxidation, isomerization, and so on, must be carried

out as rapidly as possible. The sensitivity of modem analytical methods has
greatly increased the possibility of detecting artifacts produced during
extraction and purification. The strong colour of carotenoids can mask the
presence of substantial amounts of colourless contaminants that are not
detected during the normal spectrophotometric assay of carotenoids.
The need to protect carotenoids against oxidation cannot be over-
emphasized. The presence of traces of oxygen in stored samples (even at
deep-freeze temperatures) and of peroxides in solvents (especially diethyl
ether) or of any oxidizing agents even in extracts containing carotenoids
can rapidly lead to bleaching, or to the formation of artifacts such as
epoxides or apocarotenals. The presence of unsaturated lipids and metal
ions can greatly enhance the oxidative breakdown, especially if lipoxyge-
nase enzymes are also present. Carotenoids, or extracts containing them,
should always be stored in the complete absence of oxygen, either in vacuo
or in an inert atmosphere (Ar or N2 ).
In commercial, water-dispersible formulations and in the blue caroteno-
proteins, carotenoid molecules are generally stabilized by protein. When
plant tissues are disrupted, however, even the carotenoids that are located
in the pigment-protein complexes of the thylakoids can undergo rapid
photochemical or enzymic oxidation; ~-carotene is particularly susceptible.
Changes in the geometrical isomeric composition of a carotenoid can
easily occur during isolation and purification. To minimize such changes,
exposure of carotenoids to heat or light, especially direct sunlight, should
be avoided if possible. In the presence of chlorophyll or other potential
sensitizers, photoisomerization, presumably via the carotenoid triplet
state, occurs very rapidly and appreciable amounts of artifactual geo-
metrical isomers can be produced. Carotenoids should be shielded from
light during chromatography. To avoid excessive heating, solvents with low
boiling points should be used whenever possible, since these can subse-
quently be removed at low temperature.
Almost all carotenoids are susceptible to decomposition, dehydration or
isomerization if subjected to acid conditions. Carotenoid 5,6-epoxides
undergo particularly facile isomerization to the corresponding furanoid
5,8-epoxides; many plant tissues are sufficiently acid to bring about this
isomerization during extraction, but it can usually be prevented by the
inclusion of a neutralizing agent such as sodium bicarbonate (NaHC03 )
during extraction. Acidic adsorbents, especially silica gel and silicic acid,
can cause isomerization during chromatography and should be avoided, as
should acidic solvents, notably chloroform, which usually contains traces of
hydrochloric acid (HCI). Acidic reagents and strong acids should not be
used in rooms where carotenoids are being handled.
Most carotenoids are stable to alkali, so saponification is used routinely
to hydrolyse carotenoid esters or to destroy contaminating chlorophylls or
CAROTENOIDS 215
oils. Some carotenoids, however, notably those containing the 3-hydroxy-

4-oxo-~-ring as in astaxanthin, are altered by treatment with even weak
alkali. Saponification must be avoided if it is suspected that any such
compounds may be present or, of course, if carotenoid esters are to be
isolated.
7.9 Extraction and purification
Carotenoids should be extracted from tissues as rapidly as possible, in

order to minimize oxidative or enzymic degradation. Extraction normally
employs a water-miscible organic solvent, usually acetone, methanol or
ethanol; chloroform-methanol is not recommended for carotenoids
because of the possible presence of HCl. Extraction from dried tissue is
usually more efficient if the tissue is first treated with a little water.
Efficient extraction from plant material usually requires mechanical tissue
disruption. On the laboratory scale, the usual procedure involves direct
homogenization of the tissue with the solvent in a suitable electric blender.
For small samples, grinding the tissue in acetone with clean sand (pestle
and mortar) may be used, but quantitative extraction by this means is not
easy to achieve. On a commercial scale, prolonged steeping in solvent,
with suitable agitation, is often used. Although cold solvent should be used
if possible, extraction may sometimes be so much more efficient with hot
solvent that the advantages of rapid extraction outweigh the possible
deleterious effects of heat. After filtration, the solid debris is re-extracted
with fresh solvent until no more colour is recovered (usually twice or three
times). For large-scale commercial preparations, the solvent extracts are
concentrated directly. In analytical work, the carotenoid-containing lipid
extract is transferred to ether, washed two to three times with water and
evaporated to dryness. The extract may then be saponified, if required, in 6
to 10% ethanolic potassium hydroxide (KOH) at room temperature, in the
dark, under Nz, overnight. All traces of acetone must be removed before
saponification, to avoid the possibility of producing artifacts by facile aldol
condensation between apocarotenals and acetone, or introducing contam-
inants by the polymerization of acetone. Details of recommended pro-
cedures are given elsewhere [59].
Column chromatography [60], thin-layer chromatography (TLC) [61]
and high-performance liquid chromatography (HPLC) [62] are widely used
with carotenoids, but GLC is not suitable because of their instability. A
rapid preliminary examination by TLC gives an indication of the number
and variety of carotenoids present and allows the best separation and
purification procedure for the extract to be determined. Column chroma-
tography is used mainly for large-scale purifications or, in analytical work,
to separate an extract into fractions that contain groups of compounds of
similar polarity. TLC may then be used to isolate and purify the individual
carotenoids. A general strategy that will usually yield a pure carotenoid is
to use column chromatography on alumina followed by TLC successively
on silica gel, MgO-kieselguhr G and silica again, but with a different
solvent. Samples for analysis by mass spectrometry (MS) and nuclear
magnetic resonance (NMR) need further treatment to bring them to the
rigorous state of purity required. HPLC is now the method of choice for
routine qualitative and quantitative analysis of carotenoid compositions. It
should be noted, however, that identification by HPLC, even combined
with UV-visible absorption spectra, does not constitute rigorous character-
ization.
Alumina and silica are widely used for column chromatography and
TLC. Other materials, particularly basic inorganic substances such as
magnesium oxide (MgO) and calcium hydroxide (Ca(OH)z), are also
extremely valuable for carotenoid purification. Separation on alumina and
silica depends upon polarity; the carotenes are very weakly adsorbed, the
xanthophylls more strongly, depending on the functional groups present.
In contrast, separation on MgO, Ca(OH)z and so on is determined by the
number and arrangement of double bonds in the molecule. Carotenoids
with the most extensive conjugated polyene system are most strongly
adsorbed on MgO (e.g. lycopene > neurosporene > ,-carotene). Acyclic
carotenoids are much more strongly adsorbed than cyclic ones that have
the same number of double bonds (e.g. lycopene > ,),-carotene > 13-
carotene), and l3-ring compounds are more strongly held than the corres-
ponding e-ring isomers (e.g. l3-carotene > a-carotene; zeaxanthin >
lutein). Carotenoid-5,6-epoxides are much less strongly adsorbed than the
corresponding 5,8-epoxides. Other polar groups, even hydroxy groups, are
generally not a strong influence; lycopene is much more strongly adsorbed
than the diol zeaxanthin. Other basic adsorbents, especially Ca(OH)z and
zinc carbonate (ZnC03 ) are useful for separating geometrical isomers that
may otherwise be difficult or impossible to resolve.
The instability of some carotenoids can cause problems even with the
commonly used adsorbents. Thus, unless neutralized, silica gel or silicic
acid can be sufficiently acidic to cause isomerization of 5,6-epoxycarot-
enoids. On alumina, some carotenoids, especially those containing a 3-
hydroxy-4-oxo-l3-ring (e.g. astaxanthin) undergo irreversible oxidation to
the corresponding 2,3-didehydro (diosphenol) derivatives, which are vir-
tually impossible to elute. Indiscriminate use of activated alumina (grade 0
or I) can cause geometrical isomerization. If used without dilution with a
filter aid (celite) or binder (kieselguhr G), MgO is sufficiently basic to
cause polymerization of acetone in the solvent, and to catalyse aldol
condensation between acetone and carotenoid aldehydes.
The usual adsorbent for column chromatography is alumina (deactivated
to grade III by addition of 6% water), although silica can also be used. The
CAROTENOIDS 217
Table 7.2 Solvents suitable for the elution of different groups of carotenoids in column
chromatography on alumina, grade III
Solvent system Carotenoids separated
Light petroleum (petrol) (or diethyl ether Carotenes

(ether)):petrol (1:99)
Ether:petrol (5:95) Carotene epoxides
Ether:petrol (10:90) Xanthophyll esters
Ether:petrol (20:80) Monooxocarotenoids
Ether:petrol (1: 1) Dioxocarotenoids, Monohydroxy carotenoids
Ether (or ethanol:ether) (5:95) Dihydroxy-more polar xanthophylls
Ethanol:ether (1:4) Carotenoid glycosides
column is eluted with solvents of increasing polarity. Suitable solvents and

the carotenoid groups eluted by each are shown in Table 7.2.
For TLC on silica, ether-petrol or acetone-petrol mixtures are very
suitable for carotenoids, and the solvent composition that was used to elute
the fraction from an alumina column usually gives effective separation.
Other mixtures, for example of methanol-toluene, propan-2-01-petrol or
ethyl acetate--carbon tetrachloride may b·~ used for the second TLC on
silica. The most effective solvents for TLC on MgO-kieselguhr G are
mixtures of petrol with acetone and toluene, or both. The adsorptive
strength of MgO is somewhat variable, but the following solvent mixtures
are approximately correct: 4% acetone in petrol for (X- and j3-carotene, 20
to 25% acetone in petrol for lutein and zeaxanthin and acetone-toluene-
petrol (1:1:4) for lycopene. Elution from MgO is best achieved with
acetone, plus some toluene and ethanol for strongly adsorbed carotenoids.
For further details and discussion of the purification of carotenoids, the
general references [55-{)2) should again be consulted.
7.10 High-performance liquid chromatography (HPLC)
HPLC is the method of choice for both qualitative and quantitative

analysis of carotenoids. The carotenoids are readily detected by their UV-
visible light absorption, and the procedure is extremely powerful when a
photodiode array detector is used to permit simultaneous detection and
monitoring at any chosen wavelengths and also continuous determination
and memorizing of absorption spectra during the chromatography. More
detailed treatment of HPLC of carotenoids is given by Riiedi [63], Pfander
and Riesen [62] and by Goodwin and Britton [64].
7.10.1 Normal-phase (adsorption) HPLC

Gradient elution is usually employed for complex extracts that contain
carotenoids of widely different polarities. Impressive resolution of carot-
21S NATURAL FOOD COLORANTS
enediols and more polar xanthophylls can be achieved with the following
programme.
• ~ min: isocratic, 10% solvent A, 90% solvent B
• 8-20 min: linear gradient 10-25% solvent A
• 20-30 min: isocratic 40% solvent A
• 30-40 min: isocratic 70% solvent A
where solvent A is hexane-propan-2-01 (S:2) and solvent B is hexane, and
the flow rate is 2 mllmin.
This procedure will separate efficiently mixtures of geometrical isomers
of the main xanthophylls such as zeaxanthin, lutein, violaxanthin and
neoxanthin. The use of propan-2-01 is not so satisfactory, however, for
separating less polar carotenoids. For this, mixtures of hexane with ethyl
acetate are recommended, but long pre-equilibration is needed to remove
residual traces of more polar solvents, especially alcohols.
Bonded nitrile phase columns (e.g. Spherisorb S5-CN) [63] permit the
very efficient resolution of mixtures of similar carotenoids, including
structural isomers (lutein and zeaxanthin), geometrical isomers, 5,6- and
5,S-epoxides and even the SR and SS epimers of the 5,S-epoxides. The best
results are achieved by isocratic separation of fractions (e.g. 'dihydroxy-
fraction') that have been obtained from natural extracts by classical column
chromatography or nc. Combinations of two solvent mixtures are used;
solvent A is of constant composition, namely hexane containing 0.1%
N-ethyl-diisopropylamine, whereas solvent B comprises dichloromethane
containing a variable proportion of methanol. Solvent compositions
recommended for separating the carotenoids of various polarity groups
have been tabulated by Riiedi [63].
7.10.2 Reversed-phase HPLC

Reversed-phase partition chromatography is now most widely used for the
routine analysis of carotenoids in natural extracts [62]. There is virtually no
risk of decomposition or structural modification, even of the most unstable
carotenoids. The stationary phases usually used are those with CIs-bonded
chains (ODS) and the resolution is somewhat dependent on the extent of
end-capping and carbon loading of the stationary phase. The order of
elution from reversed-phase columns is not exactly the opposite of that
from normal-phase columns, for example lutein is eluted before zeaxanthin
in both methods. As well as the overall polarity of the compounds
(presence of polar substituent groups) other important factors influence
separations. The strongest association with the CIS stationary phase is that
of an unsubstituted carotenoid end group, so that a diol with both hydroxy
groups in one end group will be substantially more strongly retained than
one with one hydroxyl in each end group. Acyclic carotenoids generally
CAROTENOIDS 219
have longer retention times than cyclic ones containing the same functional
groups, and carotenoids with a greater degree of saturation are usually
more strongly retained.
Non-aqueous solvent systems (acetonitrile-dichloromethane-methanol)
have been widely explored for reversed-phase HPLC of carotenoids [65].
In the author's laboratory, a simple gradient procedure is used as a routine
method for screening carotenoid compositions [64]. This procedure
(adapted from Wright and Shearer [66]), employs a linear gradient of ethyl
acetate (0 to 100%) in acetonitrile-water (9:1, containing 0.1% triethyl-
amine) at a flow rate of 1 ml/min over 25 min, and gives good resolution
both of polar xanthophylls and of non-polar carotenes, carotene epoxides
and xanthophyll acyl esters in the same run. The gradient is easily
modified, if required, to improve resolution of selected parts of the
chromatogram. A natural esterified carotenoid will usually contain a
mixture of molecular species containing different esterifying fatty acids.
The presence of esters is therefore usually revealed by a pattern of unusual
peaks in the chromatogram in the vicinity of f3-carotene. The absorption
spectra of xanthophylls are not altered by esterification so the carotenoid
components can tentatively be identified.
7.10.3 HPLC of apocarotenoids

The chromatographic behaviour of apocarotenoids on a reversed-phase
column can cause some confusion. The retention times are short; apo-f3-
carotenals, for example, are found in the same area of the chromatogram
as the very polar xanthophyll, neoxanthin. The retention times increase
with increasing chain length, which is a more important determining
feature than the nature of the functional groups present.
7.10.4 Resolution of optical isomers

Several methods are available for the resolution of optical isomers of
natural carotenoids, either direct on optically active columns or after
formation of diastereoisomeric derivatives.
7.11 UV-visible light absorption spectroscopy
UV-visible light absorption spectroscopy is the basic spectroscopic method

used as a first identification criterion for carotenoids. The principles of the
method, and the relationships between spectrum and structure of carot-
enoids have been discussed in detail [67]. Both the position of the
absorption maxima (Amax) and the shape or fine structure of the spectrum
are characteristic of the chromophore of the carotenoid molecule. It is
stressed, however, that the spectrum provides information only about the
light-absorbing chromophore and not about other structural features of the
molecule, such as the presence of functional groups. Thus, for example,
hydroxycarotenoids generally have spectra identical to that of the parent
hydrocarbon.
7.11.1 Position of the absorption maxima
Spectra of apolar carotenoids are usually determined in petrol or hexane,

those of the more polar xanthophylls in ethanol. The absorption spectra of
most carotenoids exhibit three maxima. Values of Amax are markedly
dependent on solvent. The values recorded in petrol, hexane and ethanol
are almost identical, but values recorded in acetone are greater by around
4 nm, those recorded in chloroform or benzene greater by 10 to 12 nm,
and those in carbon disulphide greater by as much as 35 to 40 nm. When
spectra are determined on-line during HPLC, it must be remembered that
the values for Amax in the eluting solvent frequently do not correspond to
the published values recorded in a pure solvent.
Also, with gradient HPLC, the solvent composition is changing through-
out the chromatography, so compounds which in published tables are given
the same maximum values may not give the same maximum values during
chromatography.
The Amax values of the main carotenoids are given in Table 7.3.
Comprehensive tables giving Amax values for a wider range of carotenoids
in a variety of solvents are available in other articles [56, 57, 67, 68].
In any given solvent, Amax values increase as the length of the chromo-
phore increases. Non-conjugated double bonds, for example the C-4,5
double bond of the e-ring, do not contribute to the chromophore.
Extension of the conjugated double bond system into a ring (the C-5,6
double bond of the ~-ring) does extend the chromophore but, because the
ring double bond is not coplanar with the main polyene chain, the Amax
occur at shorter wavelengths than those of the acyclic carotenoid with the
same number of conjugated double bonds. Thus, although they are all
conjugated undecaenes, the acyclic, monocyclic and dicyclic compounds
lycopene, "),-carotene and ~-carotene have Amax at 444, 470, 502 nm, at 437,
462,494 nm and at 425, 450,478 nm, respectively.
Carbonyl groups, in conjugation with the polyene system, also extend
the chromophore. A ring keto-group increases Amax by 10 to 20 nm, so that
echinenone and canthaxanthin have Amax at 461 and 478 nm, respectively.
Other substituents, such as hydroxy and methoxy groups do not affect
the chromophore. All such substituted carotenoids therefore have Amax
virtually identical to those of the parent hydrocarbon with the same
chromophore. For example, ~-carotene and its hydroxy-derivatives ~-
CAROTENOIDS 221
cryptoxanthin, zeaxanthin, isocryptoxanthin, and isozeaxanthin, all have

virtually identical spectra with Amax at 425, 450, 478 nm.
7.11.2 Spectral fine structure

The overall shape or fine structure of the spectrum (equivalent to
persistence) is also diagnostic and generally reflects the degree of planarity
that the chromophore can achieve. Thus the spectra of acyclic compounds
are characterized by sharp maxima and minima. The degree of fine
structure decreases a little when the chromophore exceeds nine double
bonds. Cyclic carotenoids in which conjugation does not extend into the
rings have simple linear polyene chromophores.
When conjugation extends into a l3-ring, steric strain causes the ring to
adopt a conformation in which the ring double bond is not coplanar with
the 'IT-electron system of the polyene chain. This effect is even more
pronounced in carotenoids that have carbonyl groups in conjugation with
the polyene chain. Spectral fine structure therefore decreases in the order
lycopene > 'V-carotene > l3-carotene > canthaxanthin (Figure 7.10).
Canthaxanthin shows only a single, rounded, almost symmetrical absorp-
tion peak (in ethanol), but a slight degree of fine structure remains if the
spectrum is determined in a non-polar solvent such as light petroleum.
7.11.3 Geometrical isomers

For Z-isomers, the Amax are generally 1 to 5 nm lower, the spectral fine
structure is decreased and a new absorption peak, usually referred to as the
'cis-peak' appears at a characteristic wavelength in the UV region
142 ± 2 nm below the longest wavelength peak in the main visible
absorption region. The intensity of the 'cis-peak' is greatest when the Z-
double bond is located at or near the centre of the chromophore.
7.12 Quantitative determination
7.12.1 Spectrophotometry
Spectrophotometric analysis is normally used for the quantitative determi-
nation of carotenoids. The amount of carotenoid present is calculated from
the equation
x = Ay/(Ar~m x 100) (7.1)
where x is the mass of carotenoid (g), y the volume of solution (ml), A the
measured absorbance, and Ar~m is the specific absorption coefficient, that
is, the absorbance of a solution of 1 g of that carotenoid in 100 ml of
Table 7.3 Light absorption maxima and specific absorption coefficients (Ar/~) of some
carotenoids
Carotenoid ~m"x (nm) Solvent" A1%

I em
b
Actinioerythrol 470 496 529 P

480 508 538 A
518 C
Antheraxanthin 422 445 472 P
422 444 472 E
430 456 484 C
8' -Apo-f3-caroten-8' -al 457 P 2640
463 E
477 C
8' -Apo-f3-caroten-8' -oic acid ethyl or 445 470 P 2500
methyl ester
Astaxanthin 468 P
478 E
480 A
485 C
Bixin 432 456 490 P 4200
433 470 502 C
Canthaxanthin 466 P 2200
474 E
482 C
Capsanthin 450 475 505 P
460 483 518 B 2072
Capsorubin 445 479 510 P
460 489 523 B 2200
a-Carotene 422 444 473 P 2800
423 444 473 E
424 448 476 A
433 457 484 C
f3-Carotene 425 449 476 P 2592
450 476 E 2620
(429) 452 478 A
435 461 485 C 2396
')I-Carotene 437 462494 P 3100
440 460489 E
439 461 491 A
446 475 509 C
8-Carotene 431 456 489 P 3290
440 470 503 C
£-Carotene 416 440470 P 3120
417 440 470 E
~-Carotene 378 400 425 P
380 400 425 H 2555
377 399 425 E
Citranaxanthin 463 495 H
463 495 P 2145
475 E
Crocetin 400 422 450 P 4320
401 423 447 E
413 435 462 C
CAROTENOIDS 223
Table 7.3 continued
Carotenoid ~max (nm) Solventa A1%

1 em b
a-Cryptoxanthin 421 445 475 H 2636

434 456 485 C
I3-Cryptoxanthin 425 449 476 P 2386
428 450 478 E
Isocryptoxanthin 425 448 475 P
425 450 477 E
438 464 489 C
Isozeaxanthin 427 450 475 P 2400
430 451 478 E
435 463 489 C
Lactucaxanthin 438 468 P
419 440470 E
427 450 479 C
Lutein 421 445 474 P
422 445 474 E 2550
435 458 485 C
Lutein 5,6-epoxide 420 443 472 P
420 442 471 E
433 453 483 C
Lycopene 444 470502 P 3450
446 472 503 E
448 474 505 A
458 484 518 C
Neoxanthin 416 438 467 P
415 439 467 E 2243
423 448 476 C
Neurosporene 414 439 467 P
416 440 470 H 2918
416 440 469 E
424 451 480 C
Norbixin 442 474 509 C
Phytoene 276 286 297 P 1250
276 286 297 H 915
Phytofluene 331 348367 P 1350
331 347366 H 1577
Violaxanthin 416 440 465 P
419 440 470 E 2250
426 449 478 C
Violerythrin 566 A
a·Zeacarotene 398 421 449 H 1850
I3-Zeacarotene 406 428 454 P 2520
406 427454 H 1940
405 428 455 E
Zeaxanthin 424 449 476 P 2348
428 450478 E 2540
430 452 479 A 2340
433 462 493 C
a Solvents: P = light petroleum; A = acetone; C = chloroform; E = ethanol; H = hexane;

B = benzene.
bThe A:o/~m values given are for the wavelength printed in italics.
Canthaxanthin
y-Carolene
350 400 450 500 550
Wavelength (nm)
Figure 7.10 Spectral characteristics of common carotenoids.
solution. A~o~ values for some common carotenoids are included in Table
7.3. More extensive tables of AI~m values have been published elsewhere
[56, 57, 67, 68]. An arbitrary value of 2500 is often taken when no
experimentally determined value has been reported, for an unknown
compound, or to give an estimate of the total carotenoid content of an
extract. It is difficult to achieve complete solution, especially of crystalline
carotenoids even in favourable solvents, so carotenoid contents are easily
underestimated. Also there is doubt about the accuracy of some of the
published Ar~m values.
7.12.2 Quantitative determination by HPLC
HPLC provides the most sensitive, accurate and reproducible method for
quantitative analysis of carotenoids, particularly when the instrumentation
includes automatic integration facilities for measuring peak areas. The
relative amounts of each component in the chromatogram can be deter-
mined, provided the peak area can be calculated for each component at its
Amax by use of a multi-wavelength detector. For the estimation of absolute
amounts or concentrations, calibration is necessary and this can be
achieved by means of a calibration graph or by use of an internal standard.
Worked examples of the quantitative analysis of isomers of J3-carotene
and astaxanthin in commercial formulations, food and feed products have
been presented [69, 70].
CAROTENOIDS 225
7.13 Other physicochemical methods
7.13.1 Mass spectrometry (MS)

Although chemical ionization has been used, most MS work on carot-
enoids has employed ionization by electron impact. Carotenoids have very
low volatility, and samples are usually inserted by means of a direct probe,
heated to 200 to 220°C. Almost all carotenoids give good molecular ions.
In addition, many fragmentations that are diagnostic of particular struc-
tural features have been identified. These have been discussed in detail in
several articles [71-77].
7.13.1.1 Fragmentations of the polyene chain. It is characteristic of all

carotenoids that they undergo reactions in which the polyene chain is
folded and portions of it are then excised. The most intensively studied are
the losses of toluene (92 mass units) and m-xylene (106 mass units) but
similar losses of 79 and 158 mass units are also frequently seen. The
abundance ratio of the [M-92]t and [M-106r fragment ions can give a good
indication of the carbon skeleton of a carotenoid; for example, for acyclic,
monocyclic and dicyclic carotenoids the [M-92r:[M-106r abundance
ratios are in the ranges 0.02 to 0.3:1, 0.6 to 1.0:1, and 2 to 10:1,
respectively. Carotenoids with a higher degree of saturation, for example
phytoene, ,-carotene, show an analogous loss of 94 mass units.
7.13.1.2 Functional groups and end groups. The presence of functional

groups in a carotenoid is indicated by characteristic fragmentations, for
example hydroxy groups, especially if in an allylic position, give rise to
strong losses of water (18 m.u.). Acetylation and trimethylsilylation
increase the molecular mass by 42 and 73 mass units, respectively, for each
hydroxy group reacting. Those single bonds in the carotenoid molecule
that are in a position allylic to the main polyene chain and also to an
isolated double bond in an end group undergo particularly facile cleavage
(e.g. C-3,4 of lycopene [M-106]+ , C-7,8 of ,-carotene [M-137]+ and C-
11,12 of phytoene [M-69]+). It is usually possible to detect the presence of
an e-ring in a carotenoid by losses of 56 mass units (by a retro-Diels-Alder
fragmentation to lose C-1, C-2 and the C-1 geminal methyl groups) and of
123 mass units due to the cleavage of the C-6,7 bond (138 mass units for a
3-hydroxy-e-ring) .
Carotenoid epoxides are readily identified. Both 5,6- and 5,8-epoxides
show strong loss of 80 mass units (two successive losses for bis-epoxides)
and also abundant ions at mlz 165 and 205 (181 and 221 if a 3-hydroxy-
group is also present in that end group).
7.13.2 Nuclear magnetic resonance (NMR) spectroscopy

Englert [78-80] has discussed the general features of the assignment of
NMR spectra of carotenoids and also described the application in the
carotenoid field of some of the sophisticated 2-dimensional NMR methods
that are now available. He has also provided an extremely valuable
collection of figures and tabulated data which illustrate the IH and 13C
assignment for more than one hundred end groups found in natural and
synthetic carotenoids, together with additional data that facilitate the
assignment of geometrical configuration. A brief summary of recent
progress in the application of NMR to plant carotenoids, together with
some tabulated data, is also available [64].
7.13.2.1 End group assignments. The IH and 13C chemical shifts for a
particular carotenoid end group are essentially the same for any carotenoid
that contains that end group. The methyl group signals are usually
diagnostic.
7.13.2.2 Olefinic protons and geometrical configuration. In a high-field

IH-NMR or 13C-NMR spectrum, for example at 400 MHz, the signals in
the olefinic region are well resolved and can be assigned, their coupling
relationships identified and the geometrical configuration of the carotenoid
determined.
7.13.3 Circular dichroism (CD) and optical rotatory dispersion (ORD)

An extensive article by Bartlett et al. [81] provided ORD data for a wide
range of carotenoid half molecules, and formulated an empirical additivity
rule by which ORD curves for carotenoids could be predicted. Since then,
most work has used CD rather than ORD. Some recent reviews and papers
have described general features such as the origin of CD in carotenoids,
factors that affect the CD, and empirical rules for the interpretation of CD
data [82-86].
In optically active carotenoids, the asymmetric centres are located in the
end groups and determine the preferred conformation of the end group.
The chiral end groups impose chirality on the main polyene chain and
determine the preferred angle of twist about the C-6,7 bond which, in turn,
determines the chirality of the twist imposed on the polyene chain. CD is
therefore observed in the electronic transitions of the main polyene chain.
Carotenoid CD spectra are strongly dependent on factors such as
temperature, concentration (e.g. molecular aggregation) and especially
geometrical configuration. Great care must therefore be taken to ensure
geometrical isomeric purity of a carotenoid sample, and also in interpreting
CD data.
CAROTENOIDS 227
7.13.4 Infra-red and Raman spectroscopy

Infra-red spectroscopy does not playa major role in carotenoid analysis,
although it can assist in the identification of carbonyl groups and special
structural features such as acetylene and allene groups [74, 87].
Resonance Raman spectroscopy has made great advances in recent years
[88, 89]. No extensive systematic study of carotenoids has yet been
reported, but carotenoids can be detected in situ by resonance Raman
spectroscopy of intact tissues.
7.14 Commercial uses and applications
Only a brief outline can be given here. For a comprehensive review, the
monograph by Bauernfeind [2] is recommended. Shorter, but useful
accounts are given by Bauernfeind et al. [90] and by KUiui [91]. Natural
carotenoids and carotenoid-containing extracts have been used for many
years to colour foods. Carotenoids are also added to animal feeds. With
the introduction of pure, synthetic, nature-identical carotenoids onto the
market and the move away from unnatural synthetic compounds as
colorants, the use of carotenoids as colour additives has increased enor-
mously. The use of carotenoids and carotenoid-containing extracts and
preparations as health-promoting products is also increasing rapidly and
shows great potential for further exploitation.
7.14.1 Application forms and formulations
7.14.1.1 Natural extracts. Coloration by natural carotenoids does not

normally use the purified pigments. Traditionally, dried and powdered or
homogenized plant material was used direct, a practice that is still
followed. Now, however, the usual preparations are obtained by extraction
with a suitable solvent and removal of the solvent to give a concentrated
crude extract. Alternatively, the extract may be prepared with, or
reconstituted in, a suitable vegetable oil.
Supercritical-ftuid extraction (SFE) with supercritical CO2 , with or
without small amounts of an organic modifier such as methanol, shows
promise in some applications for the extraction of carotenoids of low
polarity, such as lycopene from tomato waste material.
Annatto is the most widely used natural carotenoid extract, especially in
dairy and bakery products, and confectionery. Many annatto preparations
of widely differing quality are available. Oil-soluble preparations contain
the natural bixin, whereas water-soluble forms consist mainly of solutions
of norbixin as its potassium salt, as obtained from bixin by saponification.
Annatto preparations usually have good stability, but their coloration
properties are somewhat pH-sensitive, as would be expected for carboxylic

acids.
7.14.1.2 Synthetic carotenoids. The use of crystalline, pure, synthetic

carotenoids depends upon the preparation of suitable application forms or
formulations that will provide the desired hue and uniform, stable
coloration. The dry crystals are rarely used direct because of their poor
solubility properties. The usual formulations may be oil-based or water-
dispersible. The main oil-dispersible forms consist of solutions or suspen-
sions of micronized carotenoid crystals in a vegetable oil. These prepara-
tions are stable and can be stored for long periods, especially if an
antioxidant is included.
Pure crystalline carotenoids can be converted into water-dispersible or
colloidal preparations. Simple emulsions and colloidal dispersions are
generally not very satisfactory because only very low concentrations of
carotenoid can be achieved and colour stability is not good. The most
successful preparations use emulsions of supersaturated oil solutions or
solutions in suitable easily removed organic solvents, and are marketed in
the form of beadlets containing surface-active dispersing agents, stabilizing
proteins and antioxidants. Products currently on the market contain up to
10% of the carotenoid and dissolve readily in water to give slightly cloudy
dispersions. I3-Carotene and ethyl 8'-apo-l3-caroten-8'-oate give yellow to
orange colours (depending on concentration), whereas 8' -apo-l3-caroten-
8'-al gives yellow to red and canthaxanthin orange-red.
7.14.2 Uses
7.14.2.1 Direct use in food coloration. The main use of carotenoids is

for the direct coloration of food [51]. Natural extracts are used, but the
major market is for synthetic carotenoids in either oil-based or water-
dispersible formulation. I3-Carotene and 8'-apo-l3-caroten-8'-al are the
main carotenoids used in oil suspensions and solutions, and can provide
yellow-orange and orange-red colours, respectively, depending on con-
centration. The solubility of canthaxanthin in triglyceride oils is too low for
practical application.
Although lycopene shows good coloration properties, its poor solubility
and stability present serious difficulties for its practical application.
The oil-based preparatons, especially of l3-carotene, are widely used for
colouring butter, margarine, cheese, cooking fats, industrial egg products,
bakery products, pasta, salad dressings, dairy product substitutes, pop-
corn, potato products and many others.
Water-dispersible forms of l3-carotene, 8'-apo-l3-caroten-8'-al and can-
thaxanthin, and the water-soluble (norbixin) forms of annatto, are used
CAROTENOIDS 229
extensively for colouring soft drinks (especially 'orange juice'), ice-cream,

desserts, sweets, soups and meat products.
7.14.2.2 Animal feeds. Feeds for cattle, birds and fish will be con-
sidered.
7.14.2.2.1 Cattle. Natural pasture is very rich in J3-carotene and

provides sufficient of this compound to fulfil the vitamin A requirement of
the animals and also to give a desirable yellow colour to the fat and rich
cream colour to cream and butter. Artificial diets must be supplemented
with J3-carotene to ensure adequate vitamin A levels and to maintain the
required yellow colour of dairy products and fat [2].
7.14.2.2.2 Birds. Many ornamental birds owe their exotic yellow and
red colours to the presence of carotenoids in the feathers. Adequate
dietary supplies of carotenoid therefore need to be maintained in captive
birds. A well-known example is the flamingo, which needs to be provided
with substantial quantities of the oxocarotenoids that the wild birds obtain
from their diet of crustaceans, otherwise the characteristic deep pink
colour is lost. Of far greater commercial importance is the poultry industry.
Chickens absorb and accumulate xanthophylls rather than carotenes and
need substantial supplies of xanthophylls to ensure the required golden-
yellow colour of the egg yolk and also the yellow skin colour demanded in
many countries. The apocarotenoids and canthaxanthin can be used for
yolk coloration, but the most natural colours of yolk and skin are achieved
with zeaxanthin. Marigolds, as a source of lutein, are included in the diet in
some countries [50, 92-95].
7.14.2.2.3 Fish. With the recent rapid development of farming

methods for salmon and trout, it is essential that the products should have
the same desired pink flesh coloration as the wild fish. This is now attained
by the inclusion of astaxanthin (salmon) or canthaxanthin (trout) in the
diet for several weeks before harvesting [50, 94-97].
7.14.2.3 Medical and health products. The beneficial role of J3-carotene

as provitamin A has long been known. J3-Carotene is also now used
successfully to alleviate the symptoms of light-sensitivity diseases,
especially erythropoietic protoporphyria, that are characterized by
extreme irritation of the skin when exposed to strong light, because of the
formation of singlet oxygen sensitized by the free porphyrins that accumu-
late in this condition. J3-Carotene, administered at a high level (around
180 mg/day), is deposited in the skin and then quenches the triplet state of
the sensitizer and prevents formation of singlet oxygen [27].
The increasing reports that j3-carotene may be a useful antioxidant that

can afford protection against cancer and other diseases [28, 29] have led to
the appearance of a great number of j3-carotene preparations on the health
market, including crystalline carotene, extracts of natural sources such as
carrots and algae, carotene drinks and even dried algal cells (Dunaliella).
There seems little doubt that the demand for j3-carotene and perhaps other
carotenoids as beneficial health products will increase.
Carotenoids are also used simply for coloration purposes in health
products such as pills, capsules and suppositories [98].
7.15.1 Other carotenoids

Only a very small number of the 600 or so known naturally-occurring
carotenoids have been used in food coloration. Some of the other, longer
chromophore carotenoids that absorb at longer wavelengths may be worth
investigating for the possibility of extending the range of colours or hues
available further into the red and purple. Interesting examples include
actinioerythrol (cherry-red, from the sea anemone Actinia equina) and its
oxidized derivative violerythrin (purple-blue).
7.15.2 Carotenoids as health-promoting substances

The increasing use of carotenoids as food additives and in health products
can be predicted. There is currently intense research activity to try to prove
a beneficial role for carotenoids in health. Much more work is needed,
however, to elucidate their mechanism(s) of action. The focus of attention
may switch from simple consideration of an antioxidant role to invest-
igations of more specific interactions with cell surfaces and membranes, in
relation to effects on cell communication, immune response, etc.
The consensus at present is that it is desirable to maintain or increase
intake of carotenoids. It is also desirable that other, non-provitamin A,
carotenoids be studied, since these may have effects similar to or greater
than those of j3-carotene and may be equally beneficial dietary components
or supplements. Although toxicity studies have revealed no adverse effects
of j3-carotene, the effects of prolonged administration of large amounts of
other carotenoids are not known.
7.15.3 Carotenoproteins
Free carotenoids are yellow, orange or red in colour but, in many marine
invertebrate animals, they (especially astaxanthin) occur as caroteno-
CAROTENOIDS 231
protein complexes, which have a range of colours including green, blue and
purple [99-103]. The best-known example of a carotenoprotein is (X-
crustacyanin, the blue astaxanthin-protein of the carapace of the lobster
(Homarus gammarus) , which has Amax 630 nm in contrast to the 480 nm of
free astaxanthin. This 320 kDa protein has been studied intensively. Its
protein subunit structure is known and the primary sequences of the
apoprotein subunits have been determined. The main interactions between
protein and carotenoid occur via the ring oxygen functions of astaxanthin,
particularly the C-4 and C-4' oxo-groups, but the central part of the
chromophore is also highly perturbed. Details of the three-dimensional
structure of the protein and of the interactions between the protein and
carotenoid molecules are likely to be elucidated in the near future.
The carotenoprotein complexes are water soluble and very stable (in
some cases the colour is stable for years in air at room temperature) and
they have, therefore, aroused interest as possible colorants. In addition,
when the interactions with protein that lead to the large spectral shift and
colour change are understood, it may prove possible to devise formulations
that mimic or reproduce these interactions and would therefore provide
stable, water-soluble, blue, green or purple colorants.
7.15.4 Biotechnology
As outlined above, natural extracts, particularly of fruits, flowers and
leaves, have been used for many years as sources of carotenoids as
colorants. With increasing general interest in biotechnology, the commer-
cial production of carotenoids by microbial systems is being evaluated [104,
105]. The first sources to be studied were the carotenogenic moulds,
Blakeslea trispora and Phycomyces blakesleeanus. High-yielding mutants
and optimized culture conditions have been developed and large-scale
production of \3-carotene could certainly be achieved. It will be difficult,
however, for this product to compete commercially with synthetic \3-
carotene. A red yeast, Phaffia rhodozyma, has been explored as a source
of astaxanthin for fish feed (salmon) and has enjoyed some limited success,
but it is not an ideal organism to use because the astaxanthin it produces is
the wrong optical isomer [(3R, 3' R)], and extraction of it is difficult.
The major developments in recent years have occurred with micro algae
[106], especially Dunaliella species (D. salina and D. bardawil) and, more
recently, Haematococcus species. Under conditions of stress (high light,
high temperature, high salt, mineral stress) these green algae become
orange because they produce large amounts of 'secondary carotenoids'
outside the chloroplast, e.g. in the cell wall or in oil droplets. Dunaliella
species accumulate \3-carotene (up to 10% dry weight). With Haemato-
coccus, yields greater than 1-2% of astaxanthin and related compounds,
mainly as esters, have been achieved. Dunaliella is now being cultured on a
large scale in vast open ponds in countries with a suitable climate,

especially Israel and Australia. As no energy input is needed (except
sunlight) and the extreme NaCl concentrations used (up to 4M) prevent
contamination with other organisms, production of carotene-rich dried
algae or oil-based extracts is cheap and such preparations are competitive.
The purification of ~-carotene from the extracts, however, increases the
costs substantially. The culturing of Haematococcus is less favourable.
Haematococcus lacks the salt tolerance of Dunaliella and therefore cannot
be produced in open-pond culture. In spite of the expense of culturing in
bioreactors, the high cost of synthesizing optically active astaxanthin
makes the feasibility of production by Haematococcus worth exploring. In
the longer term, the genetic transfer or introduction of the ability to
produce the more expensive commercially important carotenoids (e.g.
astaxanthin or zeaxanthin with the correct stereochemistry) into Dunaliella
is an attractive prospect. Carotenoid production by the algae is, however, a
slow process.
Some bacteria have been considered for carotenoid production by
fermentation, for example strains of Brevibacterium (to yield canthaxan-
thin) and Flavobacterium (to yield zeaxanthin), but commercial production
has not been established. Very recent molecular genetic work has achieved
the introduction of the genes for carotenoid biosynthesis from the carot-
enogenic bacterium into the normally non-carotenogenic Escherichia coli
Erwinia (bacteria) and Saccharomyces cerevisiae (yeast) [105]. High
concentrations of ~-carotene and, potentially, of other carotenoids can be
obtained in this way, and rapid developments in this area can be expected.
CAROTENOIDS 233
Appendix: Structures of the individual carotenoids mentioned in the text
HO
OH
o acti nioeryth rol
08
80
antheraxanthin
CHO
8' -apo-B-caroten-8' -al
COOR
K"-apo-B-caroten-tI'-oic acid ester
80
astaxanthin
HOOC "
bixin
canthaxanthin
~OH
HO capsanthin
o
capsorubin
a-carotene
CAROTENOIDS 235
B-carotene
y-carotene
6-carotene
€-carotene
<:-carotene
citranaxanthin
COOH
HOOC "
crocetin
HO
a-cryptoxanthin
HO
B-cryptoxanthin
isocryptoxanthin
OH
OH
OH isozeaxanthin
CAROTENOIDS
237
.. ' OH
lactucaxa nthin
... OH
HO
lutein
.. OH
'
HO
lu tei n-5,6-epo xide
Iycopene
OH
~.'
M;0 H neoxanth in
HO
neurosporene
COOH
HOOC
nor-bixin
phytoene
pbytofluene
retinol
CAROTENOIDS 239
08
80
violaxanthin
o
o
violerythrin
a-zeacarotene
B-zeacarotene
08
80
zeaxanthin
References
1. Isler, O. Carotenoids, Birkhauser, Basel and Stuttgart (1971).

2. Bauernfeind, J.e. Carotenoids as Colorants and Vitamin A Precursors. Academic Press,
New York (1981).
3. Britton, G., Liaaen-Jensen, S. and Pfander, H. Carotenoids, Vol. lA: Isolation and
Analysis. Birkhiiuser, Basel, Boston and Berlin (1995).
4. Britton, G., Liaaen-Jensen, S. and Pfander, H. Carotenoids, Vol. IB: Spectroscopy.
Birkhauser, Basel, Boston and Berlin (1995)
5. Straub, O. Key to Carotenoids, 2nd edn. (H. Pfander, ed.), Birkhauser, Basel and
Boston (1987).
6. Kull, D. and Pfander, H. in Carotenoids, Vol. IA: Isolation and Analysis (G. Britton,
S. Liaaen-Jensen and H. Pfander, eds), Birkhauser, Basel, Boston and Berlin (1995)
pp. 295-317.
7. Pure Appl. Chern. 14 (1967) 215-278.
8. Pure Appl. Chern. 20 (1969) 365-553.
9. Pure Appl. Chern. 35 (1973) 1-130.
10. Pure Appl. Chern. 47 (1976) 97-243.
11. Pure Appl. Chern. 51 (1979) 435-675, 857-886.
12. Britton, G. and Goodwin, T.W. Carotenoid Chemistry and Biochemistry. IUPAC-
Pergamon, Oxford (1982).
13. Pure Appl. Chern. 57 (1985) 639-821.
14. Krinsky, N.I., Mathews-Roth, M.M. and Taylor, R.F. Carotenoids: Chemistry and
Biology. Plenum Press, New York (1990).
15. Pure Appl. Chem. 63 (1991) 1-176.
16. Pure Appl. Chern. 66 (1994) 931-1076.
17. Weedon, B.C.L. and Moss, G.P. in Carotenoids, Vol. IA: Isolation and Analysis
(G. Britton, S. Liaaen-Jensen and H. Pfander, eds), Birkhauser, Basel, Boston and
Berlin (1995) pp. 27-70.
18. Goodwin, T.W. The Biochemistry of the Carotenoids, Vol. 1, Plants. Chapman & Hall,
London (1980).
19. Goodwin, T.W. The Biochemistry of the Carotenoids, Vol. 2, Animals. Chapman &
Hall, London (1984).
20. Cogdell, R.J. Pure Appl. Chern. 57 (1985) 723--728.
21. Cogdell, R.J. in Plant Pigments (T.W. Goodwin, ed.), Academic Press, London (1988)
pp. 183--230.
22. Siefermann-Harms, D. Biochim. Biophys. Acta 811 (1985) 325-355.
23. Siefermann-Harms, D. Physiol. Plantarum 69 (1987) 561-568.
24. Young, A.J. and Britton, G. Carotenoids in Photosynthesis. Chapman & Hall, London
(1993).
25. Olson, J.A. in Vitamin A Deficiency and its Control (J.C. Bauernfeind, ed.), Academic
Press, Orlando and London (1986) pp. 19-67.
26. Mathews-Roth, M.M. in Carotenoid Chemistry and Biochemistry (G. Britton and T.W.
Goodwin, eds), Pergamon, Oxford (1982) pp. 297-307.
27. Mathews-Roth, M.M. in Carotenoids as Colorants and Vitamin A Precursors (J.C.
Bauernfeind, ed.), Academic Press, New York (1981) pp. 755-785.
28. Peto, R., Doll, R., Buckley, J.D. and Sporn, M.B. Nature 290 (1981) 201-208.
29. Santamaria, L. and Bianchi, A. Preventive Medicine 18 (1989) 603--623.
30. Krinsky, N.I. in Carotenoids: Chemistry and Biology (N.I. Krinsky, M.M. Mathews-
Roth and R.F. Taylor, eds), Plenum Press, New York (1990) pp. 279-291.
31. Krinsky, N.I. Pure Appl. Chern. 66 (1994) 1003--1010.
32. Burton, G.W. and Ingold, K.U. Science 224 (1984) 569-573.
33. Bendich, A. in Carotenoids: Chemistry and Biology (N.I. Krinsky, M.M. Mathews-Roth
and R.F. Taylor, eds), Plenum Press, New York (1990) pp. 323--335.
34. Bendich, A. 1. Nutr. 119 (1989) 112-115.
35. Bianchi-Santamaria, A., Fedeli, S. and Santamaria, L. Med. Oncol. Tumor Pharma-
cother. 9 (1992) 151-153.
CAROTENOIDS 241
36. Britton, G. in Chemistry and Biochemistry of Plant Pigments 2nd edn. (T.W. Goodwin,
ed.), Academic Press, London (1976) pp. 262-327.
37. Britton, G. in Regulation of Chloroplast Differentiation (G. Akoyunoglou and
H. Senger, eds), A.R. Liss, New York (1986) pp. 125-134.
38. Britton, G. in Plant Pigments (T.W. Goodwin, ed.), Academic Press, London (1988)
pp. 133--182.
39. Britton, G. in Carotenoids: Chemistry and Biology (N.1. Krinsky, M.M. Mathews-Roth
and R.F. Taylor, eds), Plenum Press, New York (1990) pp. 167-184.
40. Britton, G. Pure Appl. Chern. 63 (1991) 101-108.
41. Bramley, P.M. Adv. Lipid Res. 21 (1985) 243--279.
42. Bramley, P.M. and Mackenzie, A. Curro Topics Cell Regulat. 29 (1988) 291-343.
43. Britton, G. in Carotenoids in Photosynthesis (A.J. Young and G. Britton, eds) ,
Chapman & Hall, London (1993) pp. 96-126.
44. Sandmann, G. Eur. J. Biochem. 223 (1994) 7-24.
45. Armstrong, G.A., Alberti, M., Leach, F. and Hearst, J.E. Mol. Gen. Genet. 216 (1989)
254-268.
46. Coomber, S.A., Chaudhri, M., Connor, A., Britton, G. and Hunter, C.N. Molec.
Microbiol. 4 (1990) 977-989.
47. Misawa, N., Nakagawa, M., Kobayashi, K., Yamano, S., Izawa, Y., Nakamura, K. and
Harashima, K. J. Bacteriol. 172 (1990) 6704-6712.
48. Bird, C.R., Ray, J.A., Fletcher, J.D., Boniwell, J.M., Bird, A.S., Teulieres, C., Blain,
I., Bramley, P.M. and Schuch, W. Biotechnol. 9 (1991) 635-639.
49. Goodwin, T.W. Ann. Rev. Nutrition 6 (1986) 273--297.
50. Schiedt, K. in Carotenoids: Chemistry and Biology (N.1. Krinsky, M.M. Mathews-Roth
and R.F. Taylor, eds), Plenum Press, New York (1990) pp. 247-268.
51. Kliiui, H. and Bauernfeind, J.C. in Carotenoids as Colorants and Vitamin A Precursors
(J.C. Bauernfeind, ed.), Academic Press, New York (1981) pp. 48-317.
52. Krinsky, N.I., Cornwell, D.G. and Ondey, J.L. Arch. Biochem. Biophys. 73 (1958)
233--246.
53. Dimitrov, N.V. and Ullrey, D.E. in Carotenoids: Chemistry and Biology (N.1. Krinsky,
M.M. Mathews-Roth and R.F. Taylor, eds), Plenum Press, New York (1990) pp. 269-
277.
54. Bernhard, K. in Carotenoids: Chemistry and Biology (N.1. Krinsky, M.M. Mathews-
Roth and R.F. Taylor, eds), Plenum Press, New York (1990) pp. 337-363.
55. Britton, G. and Goodwin, T.W. Methods Enzymol. 18e (1971) 654-701.
56. Davies, B.H. in Chemistry and Biochemistry of Plant Pigments 2nd edn. (T.W.
Goodwin, ed.), Vol. 2, Academic Press, London (1976) pp. 38-165.
57. Britton, G. Methods Enzymol. 111 (1985) 113--149.
58. Britton, G. in Methods in Plant Biochemistry, Vol. 7 Terpenoids (D.V. Banthorpe and
B.V. Charlwood, eds), Academic Press, London (1991) pp. 473--518.
59. Schiedt, K. and Liaaen-Jensen, S. in Carotenoids, Vol. lA: Isolation and Analysis
(G. Britton, S. Liaaen-Jensen and H. Pfander, eds), Birkhiiuser, Basel, Boston and
Berlin (1995) pp. 81-108.
60. Bernhard, K. in Carotenoids, Vol. lA: Isolation and Analysis (G. Britton, S. Liaaen-
Jensen and H. Pfander, eds), Birkhiiuser, Basel, Boston and Berlin (1995) pp. 117-130.
61. Schiedt, K. in Carotenoids, Vol. lA: Isolation and Analysis (G. Britton, S. Liaaen-
Jensen and H. Pfander, eds), Birkhiiuser, Basel, Boston and Berlin (1995) pp. 131-144.
62. Pfander, H. and Riesen, R. in Carotenoids, Vol. lA: Isolation and Analysis (G. Britton,
S. Liaaen-Jensen and H. Pfander, eds), Birkhiiuser, Basel, Boston and Berlin (1995)
pp. 145-190.
63. Riiedi, P. Pure Appl. Chern. 57 (1985) 793--800.
64. Goodwin, T.W. and Britton, G. in Plant Pigments (T.W. Goodwin, ed.), Academic
Press, London and San Diego (1988) pp. 61-132.
65. Nelis, H.J.C.F. and De Leenheer, A.P. Anal. Chern. 55 (1983) 27G-275.
66. Wright, S.W. and Shearer, J.D. J. Chromatogr. 294 (1984) 281-295.
67. Britton, G. in Carotenoids, Vol. 1B: Spectroscopy (G. Britton, S. Liaaen-Jensen and H.
Pfander, eds), Birkhiiuser, Basel, Boston and Berlin (1995) pp. 13--62.
68. De Ritter, E. in Carotenoids as Colorants and Vitamin A Precursors (J.e. Bauernfeind,

ed.), Academic Press, New York (1981) pp. 883-923.
69. Schierle, J., Hardi, W., Faccin, N., Biihler, I. and Schiiepp, W. in Carotenoids, Vol.IA:
Isolation and Analysis (G. Britton, S. Liaaen-Jensen and H. Pfander, eds), Birkhiiuser,
Basel, Boston and Berlin (1995) pp. 265-272.
70. Schiiepp, W. and Schierle, J. in Carotenoids, Vol. IA: Isolation and Analysis (G.
Britton, S. Liaaen-Jensen and H. Pfander, eds), Birkhauser, Basel, Boston and Berlin
(1995) pp. 273-276.
71. Vetter, W., Englert, G., Rigassi, N. and Schwieter, U. in Carotenoids (0. Isler, ed.),
Birkhauser, Basel and Stuttgart (1971) pp. 189-266.
72. Johannes, B., Brzezinka, H. and Budzikiewicz, H. Org. Mass Spectrom. I) (1974) 1095-
1113.
73. Johannes, B., Brzezinka, H. and Budzikiewicz, H. Z. Naturforsch. 348 (1979) 300-305.
74. Moss, G.P. and Weedon, B.C.L. in Chemistry and Biochemistry of Plant Pigments, 2nd
edn. (T.W. Goodwin, ed.), Vol. 1, Academic Press, London (1976) pp. 149-224.
75. Enzell, e.R., Wahlberg, I. and Ryhage, R. Mass Spectrom. Rev. 3 (1984) 395-438.
76. Enzell, e.R. and Wahlberg, I. Biochem. Appl. Mass Spectrom. (Suppl. 1) (1980) 407-
438.
77. Enzell, e.R. and Back, S. in Carotenoids, Vol. IB: Spectroscopy (G. Britton, S. Liaaen-
Jensen and H. Pfander, eds), Birkhauser, Basel, Boston and Berlin (1995) pp. 261-
320.
78. Englert, G. in Carotenoid Chemistry and Biochemistry (G. Britton and T.W. Goodwin,
eds), Pergamon, Oxford (1982) pp. 107-134.
79. Englert, G. Pure Appl Chem. 57 (1985) 801-821.
80. Englert, G. in Carotenoids, Vol. IB: Spectroscopy (G. Britton, S. Liaaen-Jensen and H.
Pfander, eds), Birkhiiuser, Basel, Boston and Berlin (1995) pp. 147-260.
81. Bartlett, L., Klyne, W., Mose, W.P., Scopes, P.M., Galasko, G., Mallams, A.K.,
Weedon, B.C.L., Szabolcs, J. and T6th, G. J. Chem. Soc. C (1969) 2527-2544.
82. Noack, K. and Thomson, A.J. Helv. Chim. Acta 62 (1979) 1902-1921.
83. Sturzenegger, V., Buchecker, R. and Wagniere, G. Helv. Chim. Acta 53 (1980) 1074-
1092.
84. Buchecker, R., Marti, U. and Eugster, C.H. Helv. Chim. Acta 65 (1982) 896-912.
85. Noack, K. in Carotenoid Chemistry and Biochemistry (G. Britton and T.W. Goodwin,
86. Buchecker, R. and Noack, K. in Carotenoids, Vol. IB: Spectroscopy (G. Britton,
pp.63-116.
87. Bernhard, K. and Grosjean, M. in Carotenoids, Vol. IB: Spectroscopy (G. Britton,
pp. 117-134.
88. Merlin, J.e. Pure Appl. Chem. 57 (1985) 785-792.
89. Koyama, Y. in Carotenoids, Vol. IB: Spectroscopy (G. Britton, S. Liaaen-Jensen and
H. Pfander, eds), Birkhauser, Basel, Boston and Berlin (1995) pp. 135-146.
90. Bauernfeind, J.C., Brubacher, G.B., Klaui, H.M. and Marusich, W.L. in Carotenoids
(0. Isler, ed.), Birkhauser, Basel (1971) pp. 743-770.
91. Klaui, H. in Carotenoid Chemistry and Biochemistry (G. Britton and T.W. Goodwin,
92. Brush, A.H. in Carotenoids as Colorants and Vitamin A Precursors (J.C. Bauernfeind,
93. Marusich, W.L. and Bauernfeind, J.e. in Carotenoids as Colorants and Vitamin A
Precursors (J.C. Bauernfeind, ed.), Academic Press, New York (1981) pp. 320-462.
94. Schiedt, K., Leuenberger, F.J., Vecchi, M. and Glinz, E. Pure Appl. Chem. 57 (1985)
685-692.
95. Schiedt, K., Bischof, S. and Glinz, E. Pure Appl. Chem. 63 (1991) 89-100.
96. Simpson, K.L., Katayama, T. and Chichester, C.O. in Carotenoids as Colorants and
Vitamin A Precursors (J.C. Bauernfeind, ed.), Academic Press, New York (1981)
pp. 463-538.
97. Meyers, S.P. Pure Appl. Chern. 66 (1994) 1069-1076.
CAROTENOIDS 243
98. Miinzel, K. in Carotenoids as Colorants and Vitamin A Precursors (J.e. Bauernfeind,

99. Britton, G., Armitt, G.M., Lau, S.Y.M., Patel, A.K. and Shone, C.e. in Carotenoid
Chemistry and Biochemistry (G. Britton and T.W. Goodwin, eds), Pergamon, Oxford
(1982) pp. 237-251.
100. Findlay, J.B.e., Pappin, D.J.e., Brett, M. and Zagalsky, P.F. in Carotenoids:
Chemistry and Biology (N.I. Krinsky, M.M. Mathews-Roth and R.F. Taylor, eds),
Plenum Press, New York (1990) pp. 75-104.
101. Zagalsky, P.F. Pure Appl. Chem. 47 (1976) 103-120.
102. Zagalsky, P.F. Oceanis 9 (1983) 73-90.
103. Zagalsky, P.F., Eliopoulos, E.E. and Findlay, J.B.C. Compo Biochem. Physiol. 97B
(1990) 1-18.
104. Nonomura, A.M. in Carotenoids: Chemistry and Biology (N.!. Krinsky, M.M.
Mathews-Roth and R.F. Taylor, eds), Plenum Press, New York (1990) pp. 365-375.
105. Ausich, R. Pure Appl. Chem. 66 (1994) 1057-1062.
106. Ben-Amotz, A. and Avron, M. Ann. Rev. Microbiol. 37 (1983) 95-119.
8 Anthocyanins and betalains
R.L. JACKMAN and J.L. SMITH
8.1 Summary
Anthocyanins occur widely in plants, being responsible for their blue,

purple, violet, magenta, red and orange coloration; while betalains,
consisting of red-violet betacyanins and yellow-orange betaxanthins, occur
exclusively in families of the order Caryophyllales. The occurrence of these
two classes of pigments is mutually exclusive. Their stability is markedly
influenced by environmental and processing factors such as pH, temper-
ature, O 2 , enzymes and condensation reactions. Due to their inherent
instability these pigments, and betalains in particular, have not been widely
used as food colorants. However, structural variants of the anthocyanins,
i.e. polyacylated and copigmented species, have demonstrated remarkable
stability and have great potential for use as stable natural colorants.
Anthocyanin and betalain preparations can be used to colour a variety of
foods and pharmaceuticals with compatible physicochemical properties,
yielding highly coloured and high quality products. Their application could
be enhanced, however, with new sources and stable structural variants,
modification of current processes and foods, and technological advances
(e.g. industry-scale extractions/purifications, microbial purifications, bio-
technology) that would make purer and more stable preparations avail-
able. In this chapter the chemistry and biochemistry of anthocyanins and
betalains-their structure and distribution, functions, biosynthesis, factors
influencing their stability, methods of extraction and analysis, and current
and potential sources and uses-are reviewed.
8.2 Introduction
Anthocyanins are among the best known of the natural pigments, being
responsible for the blue, purple, violet, magenta, red and orange colour of
a majority of plant species and their products [1, 2]. Like the more
widespread anthocyanins, the betalains, a class of water-soluble pigments
consisting of the red-violet betacyanins and the yellow-orange betaxan-
thins, occur abundantly and uniquely in flowers, fruits and leaves of plants
that contain them [3, 4]. Betalains are also responsible for the coloration of

ANTHOCYANINS AND BETALAINS 245
some fungal species, e.g. toadstools [5,6], often accumulate in the stalk of
the plants, and in the red beet they are found in high concentration in the
root. Interestingly, these biogenetically and structurally distinct pigments,
the anthocyanins and betalains, have not been detected together in the
same species or even in different species of the same family: their
occurrence is mutually exclusive [7].
The striking colours associated with anthocyanins and betalains often
serve as the basis for identification and consumer acceptance of foods
containing these pigments. However, despite their obvious aesthetic
appeal and familiarity, the use of anthocyanins and betalains as colorants
in the food industry has been somewhat limited. They possess inherent
instability as a function of pH, temperature and light, low tinctorial
strength, low yields, in some cases they are associated with other
properties such as flavour, and they are generally difficult and/or expensive
to extract and purify from their natural sources. However, concern over
the toxicological safety of synthetic colorants and the banning of several
red dyes [8, 9] has encouraged the development and application of
anthocyanins, betalains and other natural colorants as food ingredients.
Anthocyanin powder obtained by physical means from fruit or vegetable
materials, e.g. mainly from the skins of grapes and other by-products of
wine manufacture and from red cabbage, and beetroot powder are
permitted food colour additives in most countries of the world [10, 11].
Food colorants extracted from 'natural' sources are exempt from the
rigorous and costly toxicological testing that synthetic dyes must undergo
prior to their clearance as safe food ingredients; natural colorants are
generally considered safe due to their presence in edible plant material.
Reports that have provided toxicological data on anthocyanins and
betalains [12-15] support the long-held view that these pigments pose no
threat to human health. Moreover, these pigments have demonstrated
therapeutic or medicinal properties, including antioxidant or antilipoper-
oxidant activity [16--19], antiinflammatory activity [20], anticonvulsant
activity [21], the ability to reduce serum cholesterol [22, 23] and serum
lipid levels [24], and to act as chemoprotectors in certain cancer therapies
[25]. Further investigation of these properties, with respect to structural
dictates and mechanisms of action, is required.
A knowledge of the properties of natural colorants and the factors that
influence them, including their interaction with other food components, is
germane to their successful application in a wide variety of food products.
This chapter will review the chemistry and biochemistry of anthocyanins
and betalains: their structure and distribution, in vivo functions, biosyn-
thesis, factors that influence their stability, methods of extraction and
analysis, and current and potential sources and uses. In addition to books
edited by Mazza and Miniati [2] and Markakis [26], several comprehen-
sive reviews have been recently published concerning the anthocyanins
[9, 27-32]. Betalains and related topics have also recently been reviewed
[3, 4, 33, 34].
8.3 Functions
The mutually exclusive occurrence of anthocyan ins and betalains, and

similarity in developmental and tissue distributions, suggests that their
functions in plants that produce them are essentially identical [35]. Both
pigments play a definitive role as pollination factors, floral and fruit
coloration serving to attract insect, bird and animal vectors, thereby aiding
in plant reproduction. However, the function of anthocyanins and beta-
lains in vegetative tissues (e.g. leaves, stems, roots) is not well defined.
Their accumulation at wound or injury sites in plants that synthesize them
normally suggests that they may function as phytoalexins, whereby they aid
in defence against viral and/or microbial infection. These particular
functions are of technological interest, with respect to the use of these pig-
ments in health beverages and/or as therapeutic agents. Both anthocyanins
and betalains have been shown to partake in biological oxidations, enzyme
inhibition, viral resistance and inhibition of microbial growth and respira-
tion [36-40]. Their accumulation as a result of various environmental
stresses such as chilling, drought, high salinity or nutritional deficiency may
be a type of wound response. The suggestions by Stafford [35] that
anthocyanins and betalains may function also as photoreceptors and/or as
filters needs to be more closely examined. When associated with proteins,
these pigments could function as green light photoreceptors: only small
amounts of pigment would be required to fulfil this function. Also, since
their major absorption peaks occur in the 465-560 nm range, when present
in large concentrations these natural pigments might act as green light
filters and thereby counteract the repression of plant growth that green
light has been reported to induce. Since these pigments also strongly
absorb UV light, they are suggested to playa protective role against the
adverse effects of UV irradiation [41a,b]. In plants that produce them,
betalains may also serve as nitrogen reservoirs [42]; the presence of
endogenous betalain-specific decolorizing enzymes [43, 44] is consistent
with this tenet.
8.4 Anthocyanins
8.4.1 Structure
Anthocyanins possess the characteristic C6 C3 C6 carbon skeleton of other
natural flavonoids, and the same biosynthetic origin [45, 46]. Yet, they
differ from other natural ftavonoids by strongly absorbing visible light. The
range of colours associated with the anthocyanins results from their ability
to form resonance structures, from distinct and varied substitution of the
parent C6 C3 C6 nucleus, and various environmental factors.
The anthocyanins are glycosides of eighteen different naturally occurring
anthocyanidins, these being polyhydroxy and polymethoxy derivatives of
2-phenylbenzopyrylium (ftavylium) salts (Table 8.1). An additional antho-
cyanidin, ricciniodin A, and its dimer linked at the 3' and 5' position and
referred to as ricciniodin B, have recently been isolated from the cell walls
of Ricciocarpus natans [47]. The red monomeric form is hydroxylated at
positions 6, 7,3' and 4' and possesses an ether linkage between the 3 and 2'
positions (Figure 8.1). Ricciniodin A constitutes the first reported naturally
Table 8.1 Structure of naturally occurring anthocyanidins
Anion"
Rs
Anthocyanidin Substitution pattern (R) Colour
3 S 6 7 3' S'
Pelargonidin (Pg) OH OH H OH H H Orange

Cyanidin (Cy) OH OH H OH OH H Orange-red
Delphinidin (Dp) OH OH H OH OH OH Blue-red
Peonidil'l (Pn) OH OH H OH OMe H Orange-red
Petunidin (Pt) OH OH H OH OMe OH Blue-red
Malvidin (Mv) OH OH H OH OMe OMe Blue-red
Apigeninidin (Ap) H OH H OH H H Orange
Luteolinidin (Lt) H OH H OH OH H Orange
Tricetinidin (Tr) H OH H OH OH OH Red
Aurantinidin (Au) OH OH OH OH H H Orange
6-Hydroxy-Cy (60HCy) OH OH OH OH OH H Red
6-Hydroxy-Dp (60HDp) OH OH OH OH OH OH Blue-red
Rosinidin (Rs) OH OH H OMe OMe H Red
Hirsutidin (Hs) OH OH H OMe OMe OMe Blue-red
S-Methyl-Cy (SMCy) OH OMe H OH OH H Orange-red
Pu1chellidin (PI) OH OMe H OH OH OH Blue-red
Europinidin (Eu) OH OMe H OH OMe OH Blue-red
Capensinidin (Cp) OH OMe H OH OMe OMe Blue-red
occurring anthocyanidin in which the Band C rings are connected; the

effect of the ether linkage on pigment stability has yet to be determined.
The occurrence of ricciniodins A and B in other liverworts [47] suggests
that these pigments may be more widespread than is currently realized.
With the exception of the 3-deoxy forms, which are relatively stable [48],
and ricciniodin A, anthocyanidins are rarely found in their free form in
plant tissues [45]. A free hydroxyl group in positions C-6, -8, -3' and,
especially, -3 destabilize flavylium chromophores, while hydroxylation of
positions C-5, -7, -2' (rare in natural anthocyanins) and -4' have a
stabilizing effect [49, 50]. Hydroxyl substitution at the C-7 position
generally has a greater stabilizing effect than at C-5. Without exception,
the 3-hydroxyl group is glycosylated, this conferring stability and aqueous
solubility to the anthocyanin molecule [51]. Loss of the 3-g1ycosyl moiety is
accompanied by rapid decomposition of the aglycone (i.e. anthocyanidin)
with irreversible loss of colour [52]. If a second site is glycosylated it is most
often at the C-5 hydroxyl when present. Glycosylation of 7-, 2'-, 3'-, 4'-
and/or 5'-hydroxyl groups may also occur, although glycosylation at both
the 3'-OH and 4'-OH is sterically hindered [53], and thereby contribute
further to pigment stabilization by blocking the reactive hydroxyl groups,
reducing the electrophilic character of the chromophore and, as a con-
sequence, reducing its susceptibility to nucleophilic attack, e.g. by water.
Anthocyanins have been classified into eighteen groups based on the
position of attachment and number of sugar residues [54]; the 3-
monosides, 3-biosides, 3,5-diglycosides and 3,7-diglycosides occur most
often [1, 2, 45]. The most common glycosyl group is glucose, although a
number of different monosaccharides (e.g. rhamnose, galactose, xylose,
arabinose or fructose), disaccharides (mainly rutinose, sambubiose or
sophorose; lathyrose, gentiobiose or laminariobiose occur less often) or
trisaccharides (homo- or mixed glycans) may also serve as a glycosyl
moiety. The nature of the sugar residue(s) appears to have a greater
influence on anthocyanin stability than the nature of the aglycone,
OH
HO
HO
Figure 8.1 Structure of the anthocyanidin riccionidin A.

although further investigation in this regard is required: anthocyanin

stability decreases for glycosyl moieties in the order glucose> galactose>
arabinose [55, 56].
The sugar residues attached to anthocyanins are often acylated with
aromatic acids such as p-coumaric, caffeic, ferulic, sinapic, gallic or p-
hydroxybenzoic acids, and/or with aliphatic acids such as malonic, acetic,
malic, succinic or oxalic acids. The latter acyl groups are particularly labile
in dilute hydrochloric acid, the traditional solvent used for extraction of
anthocyanins, and are more difficult to identify from spectra than the
aromatic acids [9]. In recent years, the use of weaker acids such as
perchloric, formic or acetic acids for anthocyanin recovery, and of more
selective and sensitive methods for structural elucidation, has revealed that
the aliphatic acids serve as acyl groups more often than was once thought.
Acyl substituents are usually bound to the C-3 sugar [57], esterified to the
6-0H or less frequently to the 4-0H group of the sugar [58, 59].
Anthocyanins containing two or more acyl groups have been reported,
including several with rather complex acylation patterns, e.g. cyanodelphin
[60], rubrocinerarin [61, 62], the ternatins [63-67] and lobelinins [68], and
the major anthocyanin of Tradescantia pallida [69]. Acylation has a
stabilizing influence on anthocyanins via intramolecular copigmentation,
i.e. through sandwich-type stacking of the acyl group(s) with the pyrylium
ring of the anthocyanin molecule (Figure 8.2) [27, 31, 53, 70, 71]. The
effect is more apparent for aromatic acyl groups. Deacylated pigments fade
immediately after their dissolution in neutral or slightly acidic media,
similar to the behaviour of non-acylated anthocyan ins [72]. Monoacylated
anthocyan ins do not generally display the colour stability of di-, tri- or
polyacylated anthocyan ins , indicating that at least two constituent acyl
groups are required for good colour stability/retention in neutral or slightly
acidic media. Based on limited data, the nature of the acyl moiety does
seem to have some influence on anthocyanin stability, e.g. anthocyanins
acylated with p-coumaric acid are less stable than those acylated with
caffeic acid [73, 74].
Methyoxylation of anthocyanidins and their glycosides also occurs, most
frequently at the C-3' and -5' positions but also at position C-7 and -5. It is
no coincidence that natural anthocyanins have not been reported in which
glycosylation or methoxylation occurs at all of the C-3, -5, -7 and -4'
positions. A free hydroxyl group at any of the C-5, -7 or -4' positions is
essential for formation of a quinonoidal (anhydro) base structure [53].
Anthocyanins normally exist in this coloured form in plant tissues, where
they are localized in cell vacuoles in weakly acidic or neutral aqueous
solution (e.g. pH 2.5-7.5) [31, 70, 75-77].
Under acidic conditions the colour of non- and monoacylated antho-
cyanins is determined largely by substitution in the B-ring of the aglycone
(Table 8.1). Increased hydroxyl substitution yields a bluer colour, while
Intramolecular stacking
(Sandwich-type)
Acylation
Intermolecular stacking
(Chiral-type)
Copigmentation Self-association
[
anthocyanidin copigment acyl group sugar
e.g. flavone
Figure 8.2 Schematic presentation of mechanisms of anthocyanin stabilization.
methoxylation causes the chromophores to become more red [2]. Thus,

aqueous extracts containing primarily pelargonidin and/or cyanidin glyco-
sides appear orange-red, those with peonidin glycosides are deep red, and
those containing glycosides of delphinidin, petunidin and/or malvidin
exhibit bluish-red coloration. The nature of the glycosyl group has no
apparent influence in anthocyanin coloration, although increased glycosyl
substitution and/or substitution specifically at the C-7 hydroxyl does
generally lead to increased colour intensity [9]. Glycosylation of B-ring
hydroxyl groups also generally increases perceived redness. Acylation of

glycosyl groups may have a slight blueing effect.
8.4.2 Distribution
A thorough survey of the distribution of anthocyanins in nature has

revealed that they are particularly characteristic of the angiosperms or
flowering plants, and are more prominent in fruits and berries than in other
plant parts [2, 32, 78, 79]. The pigmentation of plants and plant parts is
rarely due to a single anthocyanin. This is evident in Table 8.2, which lists
the individual anthocyanins in a number of food plants.
The major food sources of anthocyanins belong to the families Vitaceae
(grape) and Rosaceae (cherry, plum, raspberry, strawberry, blackberry,
apple, peach, etc.). Other families containing anthocyanin pigmented food
plants include the Solanaceae (tamarillo, aubergine), Saxifragaceae (red
and black currants), Ericaceae (blueberry, cranberry) and Cruciferae (red
cabbage). Most anthocyanin-containing food plants exhibit colours that are
characteristic of their constituent aglycone types; however, it should be
emphasized that many of the food plants listed in Table 8.2 also contain
pigments other than anthocyanins (e.g. chalcones, aurones, carotenoids,
chlorophylls), which influence the actual perceived colour (hue, tint,
intensity) [146]. The most commonly occurring anthocyanins are based on
cyanidin, followed by pelargonidin, peonidin and delphinidin, then by
petunidin and malvidin [2, 32, 78, 79). 3-Glycosides occur approximately
two and a half times as often as 3 ,5-diglycosides , the most ubiquitous
anthocyanin being cyanidin-3-g1ycoside.
8.4.3 Biosynthesis
The biosynthesis of anthocyanins follows a pathway common to other

flavonoids in plants, a pathway that is now well defined at both the
biochemical and genetic levels through the use of genetic mutants, cell and
tissue cultures and substrate feeding studies [40, 46,147-152]. Biosynthesis
is controlled by a number of regulatory and structural genes [153, 154] that
appear to be functionally conserved among plant species [155). Mutants at
most steps of the pathway have been identified and, at least in maize,
snapdragon and petunia, many of the regulatory and structural genes have
been cloned [154]. The structural genes, those encoding specific enzymes
of the biosynthetic pathway, appear to be expressed coordinately and in a
tissue-specific fashion. Control of expression of the structural genes
appears to occur at the level of transcription factors [156, 157]. Various
Table 8.2 Anthocyanins in selected food plants a
Botanical name Common name Anthocyanins present References

(organ)
Alium cepa Red onions Cy 3-glucoside and 3- 80-82

laminariobioside, nonacylated
and acylated with malonoyl
ester; Cy 3-galactoside and 3-
diglucoside; Pn 3-glucoside
Brassica oleracea Red cabbage (leaf) Cy 3-sophoroside-5-glucoside 83-86
acylated with malonoyl and
mono- and di-p-coumaroyl,
-feruloyl and -sinapoyl esters
Citrus sinensis Blood orange Cy and Dp 3-glucosides 87
Cyphomandra betacea Tamarillo Cy, Pg and Dp 3-glucosides and 88
3-rutinosides
Dioscorea alata Purple yam (tuber) Cy 3-glucoside, 3-diglucoside, 89-92
3-rhamnoside, 3-gentiobioside
acylated with sinapoyl ester,
and 3-glucoside acylated with
feruloyl ester; Pn 3-
gentiobioside acylated with
sinapoyl ester
Ficus spp. Fig (skin) Cy 3-glucoside, 3-rutinoside 93,94
and 3,5-diglucoside; Pg 3-
rutinoside
Fragaria spp. Strawberry Pg and Cy 3-glucosides 95-97
Glycine maxima Soybean (seedcoat) Cy and Dp 3-glucosides 98
Ipomoea batatas Purple sweet potato Cy and Pn 3-sophoroside-5- 99,100
glucosides acylated with
feruloyl and caffeoyl esters
Litchi chinensis Lychee fruit (skin) Cy 3-rutinoside and 3- 101
glucoside; Mv 3-acetylglucoside
Malus pumila Apple (skin) Cy 3-galactoside, 3-glucoside, 102-104
3-xyloside, and 3- and 7-
arabinosides; free and acylated
Mangifera indica Mango Pn 3-galactoside 105
Passiflora edulis Passion fruit Pg 3-diglucoside; Dp 3-
glucoside 106
Phaseolus vulgaris Kidney bean Pg, Cy and Dp 3-glucosides and 107
(seedcoat) 3,5-diglucosides; Pt and Mv 3-
glucosides
Prunus avium Sweet cherry Cy and Pn 3-glucosides and 3- 108-110
rutinosides
Prunus cerasus Tart cherry Cy 3-glucoside, 3-rutinoside, 3- 109-113
sophoroside, 3-
glucosylrutinoside and 3-
xylosylrutinoside; Pn 3-
glucoside, 3-rutinoside and 3-
galactoside
Prunus domestica Plum Cy and Pn 3-glucosides and 3- 97,110,
rutinosides 114
Punica granatum Pomegranate Cy, Dp and Pg 3-glucosides and 115
3,5-diglucosides
Table 8.2 continued
Botanical name Common name Anthocyanins present References

(organ)
Raphanus sativus Red radish (root) Pg and Cy 3-sophoroside-5- 57, 116,

glucosides acylated with p- 117
coumaroyl, feruloyl and
caffeoyl esters
Rheum rhaponticum Rhubarb (stem) Cy 3-glucoside and 3-rutinoside 117-119
Ribes nigrum Blackcurrant Cy and Dp 3-glucosides, 3- 120, 121
diglucosides and 3-rutinosides
Ribes rub rum Redcurrant Cy 3-glucoside, 3-rutinoside, 3- 110, 122,
sambubioside, 3-sophoroside, 3- 123
glucosylrutinoside and 3-
xylosylrutinoside
Rubus fruticosus Blackberry Cy 3-glucoside and 3- 110, 124,
rutinoside; free and acylated 125
Mv 3-biosides
Rubus ideaus Red raspberry Cy and Pg 3-glucosides, 3- 109, 126--
diglucosides, 3-rutinosides, 128
glucosylrutinosides, 3-
sophorosides, 3-sambubiosides,
3,5-diglucosides and 3-
rutinoside-5-glucosides
Solanum melongena Aubergine Dp 3-galactoside, 3-rutinoside, 129, 130
3 ,5-diglucoside, 3-rutinoside-5-
glucoside and 3-
glucosylrhamnoside-5-glucoside
acylated with p-coumaroyl ester
Solanum nigrum Huckleberry Pt and Mv 3-rutinoside-5- 131,132
glucoside; free and acylated
with mono- and di-p-coumaroyl
esters
Synsepalum dulcificum Miracle fruit (skin) Cy and Dp 3-galactosides, 3- 133
glucosides and 3-arabinosides
Vaccinium Lowbush blueberry Cy, Dp, Pn, Pt and Mv 3- 134
angustifolium glucosides, 3-galactosides and
3-arabinosides
Vaccinium Highbush blueberry Cy, Dp, Pn, Pt and Mv 3- 135, 136
corymbosum glucosides and 3-galactosides;
Dp, Pn, Pt and Mv 3-
arabinosides
Vaccinium Common cranberry Cy and Pn 3-galactosides, 3- 137-140
macrocarpon arabi nos ides and 3-glucosides
Vaccinium oxycoccus Small cranberry Pn and Cy 3-glucosides, 3- 141
galactosides and 3-arabinosides;
Dp, Pt and Mv 3-glucosides
Vitis spp. Grape Cy, Pn, Dp, Pt and Mv mono- 142
and diglucosides; free and
acylated
Zea mays Purple corn Cy, Pg and Pn 3-glucosides, 143-145
and Cy 3-galactoside; free and
acylated
aAbbreviations: Cy = cyanidin; Pg = pelargonidin; Pn = peonidin; Dp = delphinidin; Pt =

petunidin; Mv = malvidin.
patterns of expression have been observed in different plants, suggesting

that the control of biosynthesis by phytohormones such as gibberellic and
abscisic acids may vary among plant species [158-161]. Anthocyanin
accumulation is also influenced by environmental factors such as light,
temperature, plant nutrition, mechanical damage and pathogenic attack,
light being the most important of these. In general, anthocyanin accumu-
lation arising from prolonged exposure to red and far-red light is mediated
by phytochrome, while responses to blue and UV light are mediated by
cryptochrome (i.e. UV-Alblue light-photoreceptor) and/or a UV-B-
photoreceptor [162]. The identity of the UV photoreceptors is still
uncertain, but there is some evidence to suggest that it may be or involve
riboflavin [163, 164]. The signal transduction of the phytochrome system is
apparently different from and independent of that of cryptochrome [165]
and the UV-B-photoreceptor [166]. UV-light is often required in addition
to active phytochrome for enzyme induction.
Anthocyanin biosynthesis proceeds according to the scheme in Figure
8.3. Initial steps of the pathway lead to formation of a CIs-chalcone
intermediate synthesized in the plant from condensation of a molecule of
activated cinnamic acid (i.e. coenzyme A-ester of mainly p-coumaric acid)
with three molecules of malonyl-CoA, catalysed by chalcone synthase
(CHS). The free activated cinnamic acid is derived from L-phenylalanine by
general phenylpropanoid metabolism [147, 148, 150]; malonyl-CoA is
formed from the acetyl-CoA carboxylase reaction [167]. Cyclization of the
yellow coloured chalcone by chalcone isomerase (CHI) leads to formation
of colourless isomeric flavanones which themselves are converted to
dihydroflavonols (flavan 3-01s) by flavanone 3-hydroxylase (F3H) [46,147].
F3H is a 2-oxoglutarate-dependent dioxygenase requiring Fe2 +, oxygen
and ascorbate as cofactors [168]. Dihydroflavonols serve as the immediate
precursors to anthocyanidins and, specifically dihydrokaempferol, act
as direct substrates for specific B-ring hydroxylases (e.g. flavonoid 3'-
hydroxylase, flavonoid 3' ,5'-hydroxylase) [150, 169] and methyltransfer-
ases [148, 152] that lead to variation in B-ring substitution. Substitution in
the B-ring may also occur at the cinnamic acid level; CHS from several
plant sources has exhibited in vitro activity towards caffeoyl-CoA and
feruloyl-CoA esters [150]. However, the coenzyme A-ester of p-coumaric
acid appears to be the main physiological substrate for the enzyme. CHS
and the B-ring hydroxy lases are cytochrome P-450-like monooxygenases
that may be associated with microsomal membranes, and require NADPH
and O 2 as cofactors [152, 170]. Intermediates in the conversion of
dihydroflavonols to anthocyanidins are leucoanthocyanidins (flavan-3,4-
diols), formed via oxidative reduction of the 4-keto group and catalysed by
a dihydroflavonol reductase (DFR) with NADPH serving as a cofactor
[148, 171]. A similar enzyme, flavanone 4-reductase, catalyses the reduc-
tion of flavanones to flavan-4-0Is, precursors of the unusual 3-
IPOOtooyn-1
co, + H,O
co,
CHO
......... ~
HCOH
I
t-o® I 3AcelylCoA
"
CH,
HCOH
tH,O® 3CO'-i
~/
I ShIdmI< Acid Pathway I 3 MaIonyf-CoA
0- 0
COOH Co-SCoA
6;~6~ - C4H .4CL
OH OH
L-Phenytolanlne CInnamiC AcId
p-Coumar1cAcid p-Coumar1cAcid
OH
HOWI O : OH CHI
HO
CHS
~
I --
OH 0
Chalcone
Flavanone (yelloW)
(colo!1ess)
F3H 1 OH
HO
HO
OH 0
-OFR
+H,
OH OH
FlaVan 3-d leucocyanldln
(colo!1ess) (colorless)
FOH l-H,O
r OH
Anthocyanin
(coloUred)
~J +--
FGT
r OH
Anlhocyanldln
(coloUred)
Figure 8.3 The anthocyanin biosynthesis pathway. Abbreviations: PAL, phenyl ammonia
lyase; C4H, cinnamate 4-hydroxylase; 4CL, 4-coumaroyl-coenzyme A ligase; CHS, chalcone
synthase; CHI, chalcone isomerase; F3H, flavanone 3-hydroxylase; DFR, dihydroflavonol4-
reductase; FDH, flavan-3,4-diol dehydroxylase; FGT, flavonoid glycosyltransferase.
Gly = glycosyl group.
deoxyanthocyanidins (Table 8.1) [172, 173]. Subsequent dehydration of

leucoanthocyanidin into the corresponding anthocyanidin involves a puta-
tive dehydratase [46, 147, 171]. Since the naturally occurring anthocyan-
idins are unstable and somewhat insoluble in aqueous media [174], under
physiological conditions they are thought to be immediately glycosylated.
Glycosylation proceeds in a stepwise manner to higher glycosylated forms
of the anthocyanins, beginning with attachment of sugars to the 3-hydroxyl
residue catalysed by UDP-glycosyl:flavonoid 3-0-glycosyltransferase. Uri-
dine diphosphate (UDP)-sugars serve as glycosyl donors for all of the
glycosyltransferases. Glycosylation is presumed to be one of the latter
steps in anthocyanin biosynthesis. Acylation generally occurs subsequent
to glycosylation, catalysed by acyltransferases with aliphatic- or aromatic-
CoA esters serving as acyl donor substrates [148, 152, 175-177]. Evidence
for the enzymatic formation of hydroxycinnamic acid (RCA)-acylated
anthocyanins with RCA-glucose esters serving as acyl donors has also been
provided [178].
Anthocyanins do not accumulate in all cells of plants that synthesize
them. Rather, they are typically localized in flower and fruit tissues, and in
the epidermal and hypodermal cell layers of leaves and stems [46].
Anthocyanin biosynthesis is mediated via a membrane-associated enzyme
complex [179-181]. The first enzyme, phenylalanine ammonia lyase
(PAL), is located on the lumen face of the endoplasmic reticulum (ER)
where it has access to a pool of phenylalanine. Transmembrane cinnamate
4-hydroxylase channels its product to the cytoplasmic face of the ER, likely
in close proximity to the vacuolar membrane or tonoplast, where further
enzymes of the flavonoid pathway are located. Their products, dihydro-
flavonoids, are further transformed to anthocyanidins by membrane-
associated hydroxylases and cytosolic methyltransferases [182], and
membrane-associated oxido-reductases. Subsequent glycosylation and
acylation of anthocyanidins by glycosyltransferases and acyltransferases,
respectively, may be a requisite for translocation across the tonoplast into
the central or cell vacuole [183-185] where they accumulate in highly
pigmented spherical structures called anthocyanoplasts. These structures
are apparently not restricted to the cell vacuole, having been observed also
free in the cytoplasm [186]. They occur widely in both dicotyledons and
monocotyledons [46] as high density, strongly osmiophilic globules that
lack internal structure [187]. The self-association of anthocyanins in the
formation of anthocyanoplasts may serve to protect individual pigment
molecules from degradative reactions to which they are particularly prone.
The enhanced production of anthocyanins and of vacuolar anthocyan-
oplasts in plants or cell/tissue cultures as a result of low nitrate and high
sugar concentrations [188, 189] suggests that a nitrate-sensitive tonoplast
ATPase may be involved in the translocation and accumulation of
anthocyanins in cell vacuoles [190].
8.4.4 Factors influencing anthocyanin colour and stability
As with most natural colorants, anthocyanins suffer from inherent instabil-

ity; they are most stable under acidic conditions. They may degrade by any
of a number of possible mechanisms to soluble colourless and/or brown
coloured and insoluble products. Degradation may occur during
extraction/purification and normal food processing and storage. A know-
ledge of the factors governing anthocyanin stability, and of putative
degradation mechanisms, is vital to the efficient extraction/purification of
anthocyanins and to their use as food colorants. Such knowledge could also
lead to more judicious selection of pigment sources and development of more
highly coloured food products. The major factors influencing anthocyanin
stability are pH, temperature and the presence of oxygen and light, but
enzymatic degradation and interactions with food components (e.g. ascorbic
acid, metal ions, sugars, copigments) are no less important. The interdepen-
dence of these various factors precludes their evaluation in strict isolation.
8.4.4.1 pH. The presence of the oxonium ion adjacent to the C-2
position in anthocyanins is responsible for their characteristic amphoteric
nature. Thus, non- and monoacylated anthocyanins behave somewhat like
pH indicators, existing as either an acid or a base depending on pH. The
structural transformations of anthocyanins as a function of pH are funda-
mental to their colour and stability. Anthocyanin-containing solutions
generally display their most intense red coloration at acid pH (e.g. < pH
3). With increasing pH, anthocyanin-containing extracts normally fade to
the point where they may appear colourless before finally changing to
purple or blue at pH > 6. At any given pH, an equilibrium exists between
four main anthocyanin/aglycone structures (Figure 8.4): the blue quinon-
oidal (anhydro) bases (collectively denoted by A), the red flavylium cation
(AH+) and the colourless carbinol (pseudo)bases or hemiacetals (collec-
tively denoted by B; the equilibrium concentration of B4 is usually
considered to be negligible) and chalcones (C E and C z , collectively
denoted by C). At neutral or slightly acidic pH the anthocyanins exist
predominantly in their non-coloured forms. The colour of anthocyanin-
containing solutions and the relative concentrations of each of the coloured
(AH+, A) and colourless (B, C) species at equilibrium (Figure 8.5) are
dependent on the values of the equilibrium constants controlling the acid-
base or proton transfer (K a '), hydration (K{,) and ring-chain tautomeric
(KT ) reactions, where:
K; = ([A]/[AH+])aH+ (8.1)
K': = «[B] + [C])/[AH+])aH+ (8.2)

KT = [C]/[B] (8.3)
o
ccB°
~
"
0GIy
~
A·7
I
~
0GIy
I
4 •
w9
. ~
~I ....:
0GIy
~.
....:....:
0GIy
0
*00 ~
OGIy
~
A,
0GIy
4
*A ~ I . .-: :
0GIy
A,.
0GIy
~~
+H,O/-H"Y
~
t OGIy
AH'
1 ~
~+H'OI-H"
OH
OH
HO
0GIy
B,
-11
B.
OIl
OH
very slow
•
GIyO OH 4
0GIy
OH C,
C,
Figure 8.4 Structural transformations of anthocyanins (i.e. cyanidin 3,5 diglycoside): flavy-
lium cation (AH+); carbinol (pseudo)bases or hemiacetals (B 4 , B2); chalcones (C E , Cz );
neutral quinonoidal bases (A 4" A 7); ionized quinonoidal bases (A 4" A 7-). Gly = glycosyl
moiety.
100
c: B
0
-
;:;
IV
~
cQ):
(.)
c:
0 50
0
Q)
>
;:;
IV
Cii
0:::
0~
c
A
0
4 5 6
pH
Mixture of I Mixture of
Flavylium ion flavyllum cation Neutral qulnonoidal : neutral and
predominates and quinonoidal bases predominate I ionized
quinonoidal
bases
I bases
Figure 8.5 pH-distribution of anthocyanin species and predominant chromophores at

equilibrium for cyanidin 3,5-diglycoside at 25°C: flavylium cation (AH+); carbinol bases (B);
chalcones (C); quinonoidal base(s) (A). pKn' = 2.23; pK h ' = 3.38; p(K h ' KT ) = 3.03
(redrawn from Brouillard [53] and Mazza and Brouillard [30] with permission).
and aH + is the hydronium ion or proton activity (pH = -log aH +) [53, 70,
191, 192]. The expression for K h ' is often simplified, where the overall
conversion of AH+ into the colourless forms (hemiacetal and cha\cones)
are considered together in the notation B. In highly acidic media (i.e.
< pH 2) the red flavylium cation (AH+) is essentially the only anthocyanin
species present (Figure 8.5). With an increase in pH the flavylium cation
yields the quinonoidal base (A) through rapid loss of a proton. For the
naturally occurring anthocyanins so far investigated, pKa' values (pKa' =
-log K a') generally range from 3.36 to 4.85 [53, 70, 191]. The hydroxyl
groups at C-4', -7 and -5 (if present) have similar ionization constants;
therefore, the quinonoidal species normally exists as a mixture (e.g. A 4 "
A7; Figure 8.4). Anthocyanins substituted with more than one free
hydroxyl group become further deprotonated at pH 6.0 to 8.0 to yield a
mixture of resonance-stabilized quinonoidal anions (e.g. A 4,-, A 7-).
Unlike f1avylium salts that are unsubstituted at positions C-3 and -5 (i.e.
synthetic salts), for which hydration reactions occur more readily with
quinonoidal base than cationic forms [193-195], equilibrium between the
quinonoidal bases and hemiacetal of naturally occurring anthocyan ins
occurs exclusively via the f1avylium cation [192, 196]. The concentration of
the quinonoidal bases (A), which become apparent above pH 4, generally
remains low due to their thermodynamic instability relative to the hemi-
acetal (B) [197]. On standing, at pH values of about 3.0 to 6.0, nucleophilic
addition of water to the C-2 position of the f1avylium cation slowly yields
the colourless hemiacetal (B). Equilibration of this species with the open
cis-chalcone (C E ) is extremely rapid, i.e. in the range of 1 s or less.
However, the cis-trans equilibrium of C E with the corresponding (Z)-form
occurs very slowly [192]. The hydration-<iehydration equilibrium is usually
characterized by pKh ' values (pK h ' = -log Kh ') of 2-3 [53]. Any process
that reduces the efficiency of the hydration reaction, that is, induces an
apparent decrease of Kh ', should enhance the stability of the coloured
anthocyanin species. According to Brouillard [70], a 102-103-fold reduc-
tion in the value of K h ' is sufficient for such stabilization, whereby the
neutral and/or ionized quinonoidal bases are both the kinetic and thermo-
dynamic products of equilibrium reactions. In vivo, copigmentation (see
section 8.4.4.8) strongly favours the anthocyanin chromophores by reduc-
ing K h ' values, and thereby confers colour to plant products within a pH
range in which the anthocyanins generally display poor colour properties
(e.g. pH 3.0 to 7.0). Anthocyanins may have greater application as food
colorants if polyacylated forms or copigmentation were exploited by food
manufacturers. Both means of anthocyanin stabilization have received
considerable attention in recent years [31, 198].
8.4.4.2 Temperature. As with most chemical reactions the stability of

anthocyanins and the rate of their degradation is markedly influenced by
temperature. In general, structural features that lead to increased pH-
stability (i.e. methoxylation, glycosylation, acylation) also lead to
increased thermal stability [199-201]. Anthocyanin degradation is virtually
pH-independent under anaerobic conditions [202-204]. However, in the
presence of oxygen increased methoxyl, glycosyl and/or acyl substitution
generally leads to an increase in the pH at which maximum thermal
stability occurs [127, 199, 205-207]. Deprotonation of the f1avylium cation
by solvent (AH+ ~ A ~ A -) is exothermic, whereas cation hydration
(AH+ ~ B) and pyrylium ring opening (B ~ C E ) are both endothermic and
associated with positive entropy changes [208]. Thus, unless structural
features or other factors are successful in reversing the equilibrium towards
formation of the f1avylium cation or quinonoidal base, formation of chalcone
is favoured by increasing temperature (viz. during storage and processing) at
the expense of quinonoidal, f1avylium and hemiacetal species.
Coumarin glycoside has been identified as a common product of thermal

degradation of anthocyanidin 3,5-diglycosides; however, anthocyanidin 3-
glycosides do not form coumarin derivatives [209]. While deglycosylation
may not be required for anthocyan ins to undergo thermally-induced
degradation [210], it occurs readily at temperatures approaching 100°C and
higher [202, 203, 211] and is facilitated by the activity of certain enzymes
(see section 8.4.4.5). The kinetics of deglycosylation vary with the number
and nature of glycosyl constituents, thereby influencing the rate and,
possibly, the mechanism of anthocyanin thermal degradation. Once
deglycosylation has occurred, all anthocyanidins undergo a similar thermal
degradation pattern with chalcone serving as an intermediate product [212,
213]. Subsequent breakdown of the chalcone leads to carboxylic acids
(i.e. substituted benzoic acids) originating from the B-ring and carboxy-
aldehydes (i.e. 2,4,6-trihydroxybenzaldehyde) from the A-ring of the
parent anthocyanin [212-214]. a-Diketone, protochuic acid, quercetin and
phloroglucinol have also been identified as thermal degradation products
of anthocyanidin 3-glycosides [202, 212-215]. All of these products may
partake in further reaction to form melanoidins, rather ill-defined brown-
coloured complexes that may be evident as a precipitate in fruit juices and
other anthocyanin-containing beverages [204, 210, 215-218].
With few exceptions anthocyanin degradation follows first-order kin-
etics. The mechanism of anthocyanin degradation appears to be temper-
ature dependent. At storage temperatures (e.g. < 40°C) activation energy
(Ea) and z values of around 70 kJ mol-1 and 25°C, respectively, have been
reported; while at processing temperatures (e.g. > 70°C) Ea values of 95 to
113 kJ mol- 1 and z values of around 28°C have been reported [216, 219-
222]. (The z value is the temperature change required to change the
thermal destruction time by a factor of ten; it reflects the temperature-
dependence of thermal degradation.) It would appear, then, that antho-
cyanins have relatively greater stability at higher temperatures, despite
that such conditions favour the formation of the colourless anthocyanin
species, particularly the chalcones. The protective effect of various
constituents in a given system, and of condensation reactions, cannot be
overlooked. The concentration of polymeric pigments has been shown to
increase with temperature and storage time, and to contribute to the
coloration of juices and red wines [223-226]. Several authors have
recommended high-temperature short-time processing to achieve maxi-
mum pigment retention in anthocyanin-containing food products [210,
221, 224, 227]. The short times involved may be sufficient to prevent
significant degradation of anthocyanins and/or transformation to colourless
species.
B.4.4.3 Oxygen and hydrogen peroxide. Oxygen may cause degradation

of anthocyanins by a direct oxidation mechanism and/or by indirect
OY
OH Degradatton
I
~y
OxIdized an1hocyanln - products
h OH
Enzymattc Nonenzymattc
~AO:A __ 00
Figure 8.6 Coupled oxidation mechanism leading to decoloration and degradation of

anthocyanins (from Peng and Markakis [228] with permission).
oxidation whereby oxidized constituents of the medium react with the

anthocyanins (Figure 8.6). Ascorbic acid and oxygen have been shown to
act synergistically in anthocyanin degradation [229]. Maximum pigment
losses occur under conditions most favourable to ascorbic acid oxidation,
that is, at high levels or concentrations of both oxygen and ascorbic acid
[55, 230]. Ascorbic acid-induced destruction of anthocyanins results from
indirect oxidation by hydrogen peroxide (HzO z), formed during aerobic
oxidation of the ascorbic acid. Significant bleaching of anthocyanin-
containing juices occurs when HzO z is added [231]. Nucleophilic attack at
the C-2 position of anthocyanins by HzO z cleaves the pyrylium ring to yield
a chalcone, which subsequently breaks down to various colourless esters
and coumarin derivatives [59, 232]. In acid solution, HzO z was shown to
oxidize acylated 3,5-diglycosides directly to acylated O-benzylphenylacetic
acid esters [233]. The oxidation products may partake in further degrada-
tion and/or polymerization reactions [210, 230, 231]. Precipitate and haze
development in fruit juices may result from direct oxidation of hemiacetal
or chalcone species [204].
8.4.4.4 Light. Anthocyanins are generally unstable when exposed to

UV or visible light or other sources of ionizing radiation [201, 212, 214,
234, 235]. Anthocyanins substituted at the C-5 hydroxyl group, which
are known to fluoresce [236], are more susceptible to photochemical
decomposition than those unsubstituted at this position [51, 201, 237]. Co-
pigmentation can accelerate or retard the decomposition depending on the
nature of the co-pigment [238]. Light causes an increase in the rates at
which anthocyan ins undergo thermal degradation, via formation of a
flavylium cation excited state [212]. Photo-oxidation of anthocyanins yields
the same products as thermal degradation, whereby a C-4-hydroxy adduct
serves as an intermediate, which itself undergoes hydrolysis at position C-2

and consequent ring opening to yield chalcone. The chalcone then readily
undergoes further degradation to the usual anthocyanin degradation
products, e.g. 2,4,6-trihydroxybenzaldehyde and substituted benzoic acids
[212-214].
8.4.4.5 Enzymes. A number of enzymes, endogenous in many plant

tissues, have been implicated in anthocyanin degradation and concomitant
loss of colour. These enzymes have been generally termed 'anthocyanases',
but based on their activity two distinct groups of enzymes have been
identified.
1. Glycosidases, which hydrolyse the glycosidic bond(s) of anthocyanins to

yield free sugar and aglycone, the instability of the latter chromophores
resulting in their spontaneous degradation via the colourless chalcone
[213,239-241];
2. Polyphenoloxidases (PPO) which act on anthocyanins in the presence of
o-diphenols via a coupled oxidation mechanism (Figure 8.6) [228, 242-
248]. In addition to glycosidases and PPO, Grommeck and Markakis
[249] have reported a perioxidase-catalysed anthocyanin degradation.
PPO are virtually ubiquitous in the plant kingdom, catalysing the

oxidative transformation of catechol and other o-dihydroxyphenols to 0-
quinones, which may subsequently either react with each other, with
amino acids or proteins, and/or with other phenolic compounds including
anthocyanins, to yield brown coloured higher molecular weight polymers;
or oxidize compounds of lower oxidation-reduction potential [250].
Anthocyanins are poor substrates for PPO [246, 247]. The quinonoidal
base may be more susceptible to oxidative degradation by PPO than the
flavylium cation [245]; the rate of decolorization mediated by PPO may
also be dependent on the substitution pattern in the B-ring and extent of
glycosylation [251].
Various commercial enzyme preparations, usually obtained from fungal
sources, have been shown to contain glycosidases [213,239,241,252] and/
or PPO [253]. In most cases, secondary activity in commercial enzyme
preparations is desirable, and the costs of purification often preclude total
removal of such activity. Fungal 'anthocyanase' preparations have been
used to remove excess anthocyanin from blackberry jams and jellies that
were too dark and unattractive [254]; and similar preparations have been
suggested for use in the manufacture of white wines from mature red
grapes [239]. However, anthocyanase activity, whether endogenous or
exogenous, can be problematic if maximum pigment retention is desired.
A preliminary steam blanch prior to subsequent processing and/or storage
[255] and packing in high concentrations (e.g. > 20%) of sugar/syrups
[256] have proven effective in destroying and inhibiting, respectively,

endogenous anthocyanase activity in fruits. Glucose, gluconic acid and
glucono-5-lactone are competitive inhibitors of glycosidases [257]. PPO
activity in various anthocyanin-containing fruit extracts has been effect-
ively inhibited by S02 [258], bisulfite, dithiothreitol, phenyl-hydrazine and
cysteine [248, 259] and gallotannin [251]. The stabilizing effect of gallotan-
nin is also attributed to copigmentation, whereby excess gallotannin
complexes with anthocyanin via hydrophobic interactions. Ascorbic acid
also has a protective effect on anthocyanin degradation by PPO, by being
preferentially oxidized by the o-quinone formed from enzymatic oxidation
of the mediating phenol. Anthocyanins and their associated colour are not
destroyed as long as ascorbic acid is present in the system [245].
8.4.4.6 Nucleophilic and electrophilic agents. In acidic media (e.g.

< pH 3) anthocyanins occur mainly as the ftavylium cation, which itself
exists in as many as six resonance species, the positive charge being
delocalized over the entire heterocyclic structure [260]. Since the highest
partial positive charge occurs at the C-2 and -4 positions [261], non- or
monoacylated anthocyanins are particularly prone to nucleophilic attack at
these positions, generally yielding colourless species. The reaction of ftav-
ylium salts with nucleophiles occurs readily when the C-5 position is un-
substituted [50, 51]. Nucleophilic attack by water occurs readily at the C-2
position, yielding the colourless hemiacetal (Figure 8.4). Prevention of
the hydration reaction is fundamental to stabilization of anthocyanins. The
C-4 position has been considered less favourable for attack by water [53]
and, thus, at equilibrium the C-4 adduct occurs only as a minor species.
The C-4 position is more susceptible to attack by amino acids and carbon
nucleophiles such as catechin, phenol, phloroglucinol, 4-hydroxycoumarin
and dimedone, to yield 4-substituted ftav-2-enes [261-266]. These com-
pounds are highly reactive and may undergo further changes depending on
the nature of the 4-substituent.
The decolorizing action of S02, an antiseptic agent used extensively in
the wine industry, also results from formation of a colourless C-4 adduct
[267-270]. This reaction is reversible; however, acidification to about pH 1
is required for restoration of red colour. The equilibrium constants and
rates of reaction for formation of anthocyanin-bisulfite complexes have
been published elsewhere [268, 269]. The large values of these constants
are consistent with observations that only small amounts of free sulfur
dioxide are required to decolorize significant quantities of anthocyan ins.
Kinetically, anthocyanin-bisulfite complexes are quite stable; the bisulfite
moiety presumably deactivates the C-3 glycosidic bond, thereby preventing
its hydrolysis and consequent degradation reactions [202, 211]. Steric
factors also contribute to the stability of anthocyanin-bisulfite complexes.
Noteworthy, is that phenyl, methoxyphenyl, carboxyl and, especially,
methyl substitution at the C-4 position markedly stabilizes the flavylium

chromophore: these C-4 substituted anthocyan ins are virtually resistant to
nucleophilic attack [48,191-195,271]. Such C-4 substitution may therefore
provide a practical means to stabilize anthocyanins for use as food
colorants. The only natural C-4 substituted anthocyanin known is purpur-
inidin fructo-glucoside extracted from willow bark [272].
Anthocyanins have been shown to react with flavan-3-0Is, such as
catechin, in the presence of acetaldehyde to yield a covalently linked
product(s) of enhanced colour [273-275]. The reaction, which is acceler-
ated at reduced pH, has been suggested to occur via positions 6 and/or 8 of
both the flavan-3-01 and anthocyanin. Acetaldehyde alone, and other
aldehydes, readily cause fading of crude anthocyanin extracts through
electrophilic attack at these positions. Glycosyl substitution at C-5 reduces
the nucleophilic character of the C-6 and C-8 positions of the flavonoid
nucleus; thus, anthocyanidin 3,5-diglycosides are less prone to electrophilic
attack than 3-glycosides [273].
B.4.4.7 Sugars and their degradation products. The use of high sugar
concentrations (i.e. >20%) or syrups to preserve fruits and fruit products
has an overall protective effect on anthocyanin chromophores [256],
presumably by lowering water activity (a w )' Reduced aw is associated with
a reduced rate of anthocyanin degradation [217]: the hydration of antho-
cyanin chromophores to colourless species becomes less favourable as
water becomes limiting. Indeed, dried anthocyanin powders (a w ::; 0.3) are
relatively stable at room temperature for several years when held in
hermetically sealed containers [276-278].
Above a threshold level (e.g. 100 ppm), sugars and their degradation
products accelerate the degradation of anthocyanins [207]. Fructose,
arabinose, lactose and sorbose have greater degradative effects on antho-
cyanins than glucose, sucrose or maltose [205, 206, 279]. The rate of
anthocyanin degradation is associated with the rate at which the sugar itself
is degraded to furfural-type compounds [210, 280]. These compounds,
furfural (formed mainly from aldo-pentoses) and t-hydroxymethylfurfural
(HMF; formed from keto-hexoses), derive from the Maillard reaction
[281] or from oxidation of ascorbic acid [282, 283], polyuronic acids [55] or
anthocyan ins themselves. These degradation products readily condense
and/or react with anthocyanins, possibly via electrophilic attack, ultimately
leading to formation of colourless or complex brown coloured compounds.
The advanced sugar degradation products, levulinic and formic acids,
which are readily formed from HMF and furfural in the acidic environs of
most anthocyanin-containing fruits and their juices, do not react nearly as
rapidly with anthocyanins as their parent furyl aldehydes [206, 210, 279,
280]. Anthocyanin degradation in the presence of furfural and HMF is
directly temperature dependent and more pronounced in natural systems,
e.g. fruit juice. Oxygen enhances the degradative effects of all sugars and
sugar derivatives.
8.4.4.8 Co-pigmentation. Harborne [57, 284] suggested that all antho-

cyanins may be ionically bound in the cell vacuole to aliphatic organic acids
such as malonic, malic or citric acid. Such interactions could provide a
mechanism of colour stabilization in vivo. Physical absorption of the
flavylium cation and/or neutral or anionic quinonoidal bases onto a suitable
surface could also stabilize anthocyanin chromophores by taking them out
of the bulk solution, thereby preventing colour loss caused by hydration
reactions [70]. It is possible that the stability of anthocyan ins associated
with pectin arises from such interaction [285, 286]. Similarly, the high
stability of anthocyanin extracts from the dried flowers of Clitorea ternatia
used to colour Malaysian rice-cakes was attributed to adsorption of the
chromophores onto the starch of the glutinous rice [287].
In aqueous media anthocyanins are known to form weak complexes with
numerous compounds such as proteins, tannins, other flavonoids, organic
acids, nucleic acids, alkaloids, polysaccharides and metal ions through
what is referred to as intermolecular co-pigmentation [27, 30, 31, 53, 70,
288]. The most effective copigments have a planar 'IT-electron-rich moiety
and are colourless, but when complexed with anthocyanins they augment
the colour and stability of the chromophore. Thus, the flavonols quercitin
and rutin, the aurone aureusidin and C-glycosyl flavones such as swertisin
are effective copigments (Table 8.3). A double bond in the 2,3-position of
the C-ring is generally required of flavonoids for efficient co-pigmentation
[289]. C- and O-glycosylated flavones are also effective copigments [51], as
well as flavonolsulfonic acids [238]. Colour augmentation, that is, a
bathochromic shift or increase in the wavelength of maximum absorption
in the visible range, results from a reduction in the local charge-distribution
or polarity of the anthocyanin upon complexation with copigment [197].
Both the flavylium cation and quinonoidal base may partake in co-
pigmentation [27, 290, 291]. Co-pigmentation involving both anthocyanin
species results in a bathochromic shift in the visible wavelength of
maximum absorption, from red to a stable blue or purple colour; however,
increases in maximum absorption or tinctorial strength may occur only
when complex formation involves the quinonoidal base [290]. Coloured
anthocyanin species (flavylium cation and quinonoidal bases) are almost
planar with a strongly de localized system of 'IT-electrons [260] that facil-
itates good molecular contact with a copigment possessing similar struc-
tural features.
The extent of the bathochromic shift and increase in maximum absorp-
tion (Table 8.3) are empirical and qualitative parameters, and provide no
quantitative information regarding the co-pigmentation reaction itself. The
stability of a given complex is reflected in the value of the equilibrium
ANTHOCY ANINS AND BETALAINS 267
Table 8.3 Co-pigmentation of cyanidin 3,5-diglucoside (2 x 10-3 M) at pH 3.32
Copigments (6 x 10-3 M) Amax <1.Amax % Absorbance

(nm) (nm) increase at
Amax (nm)
None 508
Aurone
Aureusidin a 540 32 327
Alkaloids
Caffeine 513 5 18
Brucine 512 4 122
Amino acids
Alanine 508 0 5
Arginine 508 0 20
Aspartic acid 508 0 3
Glutamic acid 508 0 6
Glycine 508 0 9
Histidine 508 0 19
Proline 508 0 25
Benzoic acids
Benzoic acid 509 1 18
o-Hydroxybenzoic acid 509 1 9
p-Hydroxybenzoic acid 510 2 19
Protocatechuic acid 510 2 23
Coumarin
Esculin 514 6 66
Cinnamic acids
m-Hydroxycinnamic acid 513 5 44
p-Hydroxycinnamic acid 513 5 32
Caffeic acid 515 7 56
Ferulic acid 517 9 60
Sinapic acid 519 11 117
Chlorogenic acid 513 5 75
Dihydrochaicone
Phloridzin 517 9 101
Flavan 3-ols
( + )-Catechin 514 6 78
Flavone
Apigenin 7-glucoside a 517 9 68
C-Glycosyl flavones
8-C-Glucosylapigenin (vitexin) 517 9 238
6-C-Glucosylapigenin (isovitexin) 537 29 241
6-C-Glucosylgenkwanin (swertisin) 541 33 467
Flavonones
Hesperidin 521 13 119
Naringin 518 10 97
Flavonols
Kaempferol 3-glucoside 530 22 239
Kaempferol 3-robinobioside-7-rhamnoside (robinin) 524 16 185
Quercetin 3-glucoside (isoquercitrin) 527 19 188
Quercetin 3-rhamnoside (quercitrin) 527 19 217
Quercetin 3-galactoside (hyperin) 531 23 282
Quercetin 3-rutinoside (rutin) 528 20 228
Quercetin 7-glucoside (quercimeritrin) 518 10 173
7-0-Methylquercetin-3-rhamnoside (xanthorhamnin) 530 22 215
aFormed a slight precipitate. (From Asen et al. [289] by permission of the authors and
Pergamon Press.)
binding constant, K, for the complexation reaction [292]. Thus, for co-
pigmentation involving the flavylium cation:
(8.4)
where [AH(CP)n +] represents the concentration of copigment-flavylium
cation complex; [AH+]o and [CP]o the equilibrium concentrations of
flavylium cation and copigment, respectively; and n is the stoichiometric
constant, that is, the number of copigment molecules linked to the
flavylium cation in the complex. At a wavelength close to Amax for ab-
sorption of the free flavylium cation AH+, the equilibrium absorbance is
given by:
(8.5)
where EAH + is the molar absorption coefficient of AH+ and l is the optical
pathlength. In the presence of copigment, the absorbance of a given
anthocyanin-containing solution is expressed according to:
A = EAH+ [AH+]ol + EAH(CP)n+ + [AH(CP)n+]ol (8.6)
where EAH(CP)n + is the molar absorption coefficient of AH+ in the
copigment complex. The magnitude of the co-pigmentation effect under a
given set of experimental conditions is then generally described by the
relation:
(A - Ao)/Ao = Kr[CP]on (8.7)
The value of r (= EAH(CP) +/E AH +) is dependent on the wavelength of
absorbance measurement and is often estimated from the ratio A/Ao,
where A is the absorbance associated with an entirely complexed flavylium
cation, that is, the absorbance of a strongly acidic anthocyanin-containing
solution to which excess copigment is added. Plots of In{(A - Ao)/Ao}
versus In{[CP]o} are generally linear with slope n and intercept In(Kr). Such
plots have proven useful in quantifying the co-pigmentation reaction;
values of K and n for a number of anthocyanin-copigment complexes have
been reported [197, 291-294]. The magnitude ofthe co-pigmentation effect
has been shown to be influenced by pH, copigment and anthocyanin
structure, as well as the molar ratio of copigment to anthocyanin.
Quantitation of the co-pigmentation effect using equation (8.7) has
confirmed the hypothesis of Goto and coworkers [27, 31, 295] that complex
formation is mediated largely by hydrophobic forces [292, 293]. In the
absence of water, the co-pigmentation effect does not exist: the copigment
competes with water for association with the anthocyanin. Thus, stability is
rendered through displacement of the hydration-dehydration equilibrium
(Figure 8.4) towards the chromophores; for example, with quercitin acting
as copigment the hydration constant K h ' for cyanin is reduced from 10-2 to
7 X 10-4 M [53]. Association of anthocyanin with copigment occurs via a
vertical stacking process similar to that which occurs in intramolecular co-

pigmentation (Figure 8.2) [27, 31, 53, 70, 188, 197,291,295]. The chiral
stacking provides an efficient means of protecting the anthocyanin chromo-
phore against nucleophilic attack and, specifically, from hydration and
associated loss of colour. Intermolecular co-pigmentation would appear to
be less efficient in the stabilization of anthocyanin chromophores than
intramolecular co-pigmentation, that is, acylation; yet, at the molecular
level, the characteristics of both phenomena are similar. Hydrogen
bonding is likely involved in some co-pigmentation complexes, especially
when the anthocyanin and/or copigment contain sugar moieties [296];
however, hydrogen bonding is not likely to contribute much to complex
stabilization if only because water itself serves as an excellent hydrogen
bond donor and acceptor. The co-pigmentation effect is greater for
anthocyanin 3,5-diglycosides than for 3-monoglycosides [197] and
enhanced, also, with increased acyl and, especially, methyl substitution
[294]. These structural effects on co-pigmentation reflect differences in the
thermodynamics of the hydration process [197]. Variation in magnitude of
the co-pigmentation effect with pH also reflects the displacement of the
hydration equilibrium towards coloured anthocyanin species [293, 294].
Co pigments have been shown to associate with colourless forms of
anthocyan ins [197] and with cyclodextrins [291, 297, 298], to result in what
has been referred to as 'anti-co-pigmentation'. Such associations may have
a significant influence on the magnitude of the co-pigmentation effect via
competition of colourless anthocyanin species with the flavylium cation for
copigment, complex formation with colourless species being favoured
as the molar ratio of copigment/pigment increases. Generally, co-
pigmentation is more pronounced as anthocyanin concentration and the
ratio of copigment to anthocyanin increase [30, 293, 294, 299-301].
However, there appears to be an optimum molar ratio of pigment to
copigment at which colour intensity and stability are maximized, and
beyond which anti-co-pigmentation has a significant effect.
When pigment concentrations become relatively high, anthocyanins
themselves may act as copigments and participate in self-association
reactions [298, 302]. Schewffeldt and Hrazdina [301] noted that competi-
tion can take place between co-pigmentation and self-association re-
actions. They demonstrated that with low concentrations of anthocyanin
from Vitis spp., colour augmentation by co-pigmentation with rutin was
intensified, but decreased with corresponding increases in anthocyanin
concentration. The net result of competing co-pigmentation and self-
association reactions is that colour may increase more than proportionally
to pigment concentration. 4'-Hydroxyl and 5-glycosyl groups appear to be
essential structural elements for self-association [303, 304]. Urea and
dimethylsulfoxide disrupt self-associated aggregates, while sodium salts
and magnesium chloride promote self-association [303-304]. Figueiredo
and Pina [305] have alternatively suggested that ionic salts render colour
enhancement through formation of an ion-pair between the charged
flavylium cation and the anion, which displaces the equilibrium towards the
flavylium cation. Self-association constants for the flavylium cation and
quinonoidal base forms of the common anthocyanins have been reported
by Hoshino [306, 307]. Generally, the strength of mutual interaction
increases with substitution of hydroxyl or methoxyl groups in the B-ring,
and is greatest for the anionic quinonoidal base forms relative to self-
association of neutral quinonoidal bases or of flavylium cations. Self-
association occurs via chiral stacking (Figure 8.2) with the configuration of
a left-handed screw [27].
The co-pigmentation effect, both intra- and intermolecular, has been
deemed primarily responsible for the coloration of flower and fruit tissues,
fruit juices and red wines [292] since anthocyanins alone are known to be
virtually colourless at the pH of these products. That juices obtained from
enzyme-treated fruit mashes are more highly coloured than nonenzyme-
treated press-juices can be attributed to decompartmentalization of various
cellular constituents, including flavonoids, alkaloids, amino acids and
nucleosides, that may participate in co-pigmentation with anthocyanins to
varying degrees. Flavonoids, as copigments, are always found in conjunc-
tion with anthocyanins, likely because of their similar biosynthetic path-
ways. Polymeric flavonoids and anthocyanins have been shown to play an
important role in the coloration of grapes and red wines [223, 225, 226,
308]. The constituent tannins (condensed flavonoids) have a protective
effect on anthocyanins. During aging of red wines, monomeric anthocya-
nins are progressively and irreversibly replaced by polymeric pigments
through self-association reactions. Such polymeric material is less pH-
sensitive and relatively resistant to decoloration by S02, ascorbic acid and
light [142, 289, 309].
8.4.5 Extraction and analysis

The choice of a method to extract anthocyan ins from any of their various
sources depends largely on the purpose of the extraction, and also on the
nature of the constituent anthocyanins. A knowledge of the factors that
influence anthocyanin structure and stability is vital. If the extracted
pigments are to be subsequently analysed, qualitatively or quantitatively, it
is desirable that a method be chosen that retains the pigments as close to
their natural state (i.e. in vivo) as possible. On the other hand, if the
extracted pigments are to be used as a colorant/food ingredient then
maximum pigment yield, tinctorial strength and stability are of greater
concern. It is also important that the extraction and cleanup procedures
not be too complex, time-consuming or costly. Methods of anthocyanin
extraction and purification, and of their subsequent analysis, have been the
subject of several comprehensive reviews [27, 29, 32, 310-312], to which

the reader is referred.
8.4.5.1 Extraction. Anthocyanins are not stable in neutral or alkaline

solution. Thus, extraction procedures have generally involved the use of
acidic solvents which disrupt plant cell membranes and simultaneously
dissolve the water-soluble pigments. The traditional and most common
means of anthocyanin extraction involves maceration or soaking of plant
material in a low boiling point alcohol containing a small amount of
mineral acid (e.g. ::; 1% HCl) [311]. Methanol is most often used;
however; due to its toxicity, acidified ethanol may be preferred with food
sources, though it is a less effective extractant and more difficult to
concentrate due to its higher boiling point [310]. Acidification with HCl
serves to maintain a low pH, thereby providing a favourable medium for
the formation of flavylium chloride salts from simple anthocyanins. The
use of mineral acids such as HCl may, however, alter the native form of
complex pigments by breaking associations with metals, copigments, and
so on [313-315]. The loss of labile acyl and sugar residues may also occur
during subsequent concentration [202]. Thus, to obtain anthocyanins
closer to their natural state several neutral solvents have been suggested
for initial pigment extraction, including 60% methanol, n-butanol, ethy-
lene glycol, propylene glycol, cold acetone, acetone/methanol/water mix-
tures and, simply, boiling water [202, 276, 287, 316]. In addition, weaker
organic acids-mainly formic acid, but also acetic, citric and tartaric
acids--or stronger, more volatile acids such as trifluoroacetic acid have
become popular for use in extraction solvents, especially for extraction of
complex, polyacylated anthocyanins [277, 313-315, 317]. Extraction of
anthocyanins with ethanolic or aqueous S02 or bisulfite solutions has led to
increased pigment yields and to concentrates of higher purity, colour
intensity and pigment stability than those obtained via hot water extraction
[234, 318-320]. The initial decanted or filtered anthocyanin extracts are
usually quite dilute and require concentration prior to subsequent purifica-
tion or use as a food colorant. Depending on the means of extraction,
decreasing the ratio of extraction solvent to plant material, as in conti-
nuous counter-current extraction. [321], could avoid the need for a
concentration step. To minimize pigment degradation, concentration is
best carried out in vacuo at temperatures below 30°C [310]. The acidic
aqueous anthocyanin concentrates thus obtained can be used as such,
frozen [276], freeze-dried [234, 278, 322, 323] or spray-dried [277].
8.4.5.2 Purification. Purification of anthocyanin-containing extracts is

often necessary, since none of the solvent systems commonly used for
extraction is specific for the anthocyanins: there may be considerable
quantities of extraneous material (e.g. other polyphenolic substances,
pectin) that could influence the stability and/or analysis of these pigments.
Co-extracted flavonoids, which react similarly with the common reagents
used for phenolic analysis, are commonly removed via chromatographic
techniques employing insoluble polyvinylpyrrolidone (PVP) [324-326],
polyamide [327], combination polyamide-PVP [59, 328, 329]. Sephadex G-
25 [101] or LH-20 [101, 330, 331], octadecylsilane [101, 331, 332], weak
anion exchange (e.g. Amberlite CG-50) [111, 333, 334], polyethylene
glycol dimethacrylate [335] and cellulose-type resins [120, 127,240,336].
Droplet counter-current chromatography (DCCC) using n-butanol-glacial
acetic acid-water (BAW) as the solvent system [337-339], preparative
thin-layer chromatography (PTLC) [340--343], solvent-solvent extraction
with n-butanol [108], and precipitation with basic lead acetate [l08] or via
the spherical agglomeration technique employing a benzene-methanol-
aqueous hydrochloric acid solvent mixture [344] have also been used to
separate and/or purify anthocyanins from their crude or concentrated
extracts. If appreciable quantities of lipid, chlorophyll or unwanted
polyphenols are suspected to be present in anthocyanin-containing
extracts, these materials may be removed by washing with petroleum
ether, ethyl ether, diethyl ether or ethyl acetate [310, 311, 345]. This may
improve purification but is not always necessary.
Purification of anthocyanins for analytical purposes was traditionally
carried out using paper or thin-layer chromatographic techniques [45, 310,
345]. However, more rapid and efficient separation of complex mixtures is
achieved with reversed-phase high-performance liquid chromatography
(HPLC) [225, 226, 330, 332, 338, 339, 346-350]. The technique is non-
destructive and, therefore, separated peaks are readily collected for
subsequent analyses. With appropriate selection of eluent, column type
and length, flow rate and temperature, HPLC can be used to resolve and
quantitate microgram quantities of anthocyanin without the need for
preliminary purification of extracts [351-355]. A major application of
HPLC is, therefore, the simultaneous qualitative and quantitative analysis
of anthocyanins [356].
8.4.5.3 Qualitative analysis. Current analytical approaches that allow

for unambiguous identification of individual anthocyanins are based on
proven in vitro techniques. Well developed chromatographic and spectro-
scopic methods have been applied for rapid and accurate identification of
anthocyanins. However, absolute characterization of anthocyanins cannot
generally be established by chromatographic or spectral methods alone.
Structural characterization generally involves identification of the aglycone,
sugar moieties and acyl groups, if present, and the position(s) of attach-
ment of the sugar and acyl groups. Such is currently facilitated by a
combination of different spectroscopic techniques, including infra-red (IR)
[357], Raman resonance [358, 359], ultraviolet/visible (UVMS) [45,312,
360], circular dichroism (CD) [31, 295, 303, 304, 306, 307, 361], nuclear
magnetic resonance (NMR) [31, 92, 306, 307, 347, 361-372] and mass
spectrometry (MS) [347, 366, 369, 372-376]. The complete structural
characterization of anthocyanins is possible using current one- and two-
dimensional (lD and 2D) NMR techniques; however, relatively large
quantities of purified material are required for resolution of proton signals
associated with the glycosyl moieties [369] and positions C-6 and C-8 of the
flavylium nucleus [376]. When only small amounts of material are avail-
able, MS techniques are the preferred methods for analysis of anthocyanin
structure ,[312]. These techniques are often interfaced with liquid chro-
matography or HPLC to facilitate pigment purification, and employ such
ionization methods as fast-atom bombardment (FAB) [31, 366, 369],
electrospray [374, 375] and thermospray [377]. While the sophisticated
NMR and MS techniques have become indispensable for the structural
characterization of complex, polyacylated anthocyanins, the equipment is
costly and not always readily available. Thus, the classical hydrolysis
techniques for structural elucidation of anthocyanins will continue to be
important; they can be quickly and efficiently carried out when interfaced
with HPLC and gas-liquid chromatography (GLC) [378, 379].
Much information regarding the identity of a given anthocyanin and its
aglycone can be derived simply by observing its colour in aqueous solution
or on paper. The colour generally changes from orange-red to bluish-red as
the structure is modified through pelargonidin, cyanidin, peonidin, mal-
vidin, petunidin and delphinidin [310]. In addition, under UV-light
anthocyanins substituted at the C-5 position usually fluoresce [236]. The
spectral characteristics of most of the known anthocyanidins and many of
the anthocyanins in 0.1% HCI-methanol have been published by Har-
borne [45, 312, 360]. In acid solution anthocyanins and their aglycones
exhibit two characteristic absorption maxima; one in the visible region
between 465 and 550 nm and a smaller one in the UV region at about
275 nm. The wavelength of maximum absorption can be used to tentatively
identify the aglycone; however, confirmation by other methods is required.
A bathochromic shift of 15-35 nm from the visible maximum upon
addition of 5% alcoholic aluminum chloride (AICI 3 ) to pigments in 0.1 %
HCI-methanol indicates the presence of a free o-dihydroxyl group in the
aglycone, a structural characteristic of cyanidin, delphinidin and petunidin
[380]. Glycosylation of anthocyanidins at the C-3 position generally results
in a bathochromic shift in the wavelength of maximum absorption, while
glycosyl substitution at C-5 produces a shoulder on the absorption curve at
440 nm. The ratio of absorbance at the UV maximum to that at the visible
maximum, and the ratio of absorbance at 440 nm to that at the visible
maximum provide valuable information regarding the extent and position
of glycosyl substitution [50, 360]. Acylated anthocyanins exhibit an
additional weak absorption maximum in the 310 to 335 nm region, the
actual maximum indicating the type of aromatic acylation involved [59,

310]. The complex acylated anthocyanins exhibit yet an additional absorp-
tion maximum at 560---600 and/or 600---640 nm at pH > 4.0 [381].
The traditional means of anthocyanin characterization generally involves
(controlled) acid, alkali, enzyme and/or peroxide hydrolysis and subse-
quent chromatography of hydrolysis products to facilitate their identifica-
tion [45, 310, 382]. Analytical HPLC has become the standard
chromatographic procedure, especially since the advent of the photo diode
array detector [351, 352, 356, 383-385]. General anthocyanin structure-
retention relationships have been reviewed by Strack and Wray [32, 312).
The chromatographic mobility (R f value) of anthocyanins on paper [45,
310, 382] or cellulose strips (i.e. TLC) [126, 202, 343, 386] provide an
alternative useful parameter for identification purposes. Pigment char-
acterization by all chromatographic methods requires comparison with
authentic anthocyanin, aglycone and/or sugar standards. Although antho-
cyanins and their aglycones can be purchased from various suppliers, they
often require purification prior to their use as reference pigments.
Acid hydrolysis of anthocyan ins generally yields aglycones and sugars,
the sugars being cleaved from the pigment in a more or less random fashion
[120,310]. The course of hydrolysis can be followed, and the glycosylation
pattern determined, by the production of intermediate pigments [387].
Alkaline hydrolysis, while it too may be used to determine the nature of
the aglycone, is more commonly employed for acyl group determination.
Under nitrogen alkaline hydrolysis specifically removes acyl groups,
generally leaving the glycoside intact [120]. The removal of oxygen is
necessary since anthocyanins possessing o-dihydroxy groups are unstable
in alkali [310]. Enzymatic hydrolysis procedures generally involve the use
of f3-glycosidases of fungal origin [239, 257]. The activity of f3-glycosidase is
generally not influenced by the nature of the aglycone [239]. The glycosyl
moieties of simpler glycosides (e.g. 3-0-f3-glycosides) are rapidly hydro-
lysed, thereby permitting relatively easy identification and differentiation
from more complex anthocyanins: since ex-glycosides are hydrolysed more
slowly, if at all, by f3-glucosidase, acylated pigments are virtually unaffec-
ted during enzymatic hydrolyses. Oxidation of anthocyan ins with peroxide
liberates the C-3 sugar [388]. Subsequent permanganate treatment releases
sugars attached to the readily ruptured heterocyclic ring, and ozonolysis
releases sugars attached to enolic and phenolic hydroxyl groups [388].
Peroxidation of anthocyanidin 3,5-diglycosides under neutral conditions
yields coumarin derivatives, while peroxidation of anthocyanidin 3-
glycosides does not [209, 232].
8.4.5.4 Quantitative analysis. Quantitative methods may be con-

veniently divided into three groups depending on the needs of the analysis
[310, 389]:
1. Determination of total anthocyanin concentration in systems containing

little or no interfering substances,
2. Determination of total anthocyanin concentration in systems that
contain interfering material,
3. Quantitation of individual pigments.
In fresh plant extracts/juices there are generally few interfering com-
pounds that absorb energy in the region of maximum absorption of
anthocyanins (465-550 nm). Total anthocyanin concentration may thus be
determined by measuring the absorbance of the sample (diluted with
acidified alcohol) at the appropriate wavelength. It is important that Beer's
Law be obeyed in the concentration range under study; self-association
and co-pigmentation effects cause deviations in Beer's Law and can lead to
inaccurate estimates of total anthocyanin concentration [390]. Consider-
able sample dilution may be necessary to counteract these effects. Adams
[202] suggested that absorbance measurements be carried out at a single
pH as low as possible to ensure maximum colour development and low
response to changes in pH which may occur (Figure 8.7). Comparative
measurements of total anthocyanin may be made at pH 1.0 for samples
containing predominantly 3-glycosides [202] and at pH less than 1.0 for
samples containing mainly 3,5-diglycosides [268].
For samples containing a mixture of anthocyanins, absorbance measure-
ments at a single low pH value are proportional to their total concentration
[392]. Estimates of the concentrations of individual anthocyanins in the
mixture can only be obtained with prior knowledge of the proportions of
individual pigments therein. Absolute concentrations may then be estim-
ated by the use of weighted average absorptivities and absorbance
measurement at weighted average wavelengths [323]. More accurate
estimates are obtained with samples in which a single pigment is present or
predominates. Molar extinction coefficients (absorptivities) of a number of
anthocyanins have been reported [137, 310, 393]. It is important that the
same solvent and pH be used for absorbance measurement as were used
for determination of molar absorptivities [53, 137, 323, 382]. The single pH
method is subject to interference by and, therefore, cannot be used in the
presence of brown coloured degradation products, for example, from sugar
or pigment breakdown. Unpurified pigment extracts may also contain
compounds that at low pH absorb in the range of maximum absorption of
the anthocyanins [202]. Differential or subtractive absorption methods
[389] may be used to determine the total anthocyanin concentration in
those samples that contain interfering substances.
The substractive method involves measurement of sample absorbance at
the visible maximum, followed by bleaching and remeasurement to give a
blank reading. Subtraction gives the absorbance due to anthocyanin, which
can be converted to anthocyanin concentration with reference to a
~max 468
0.7
0.6
Q)
() 0.5
c
C'il
..0
~
0
(J) 0.4 ~ max 373
..0
<:
0.3
0.2
0.1
350 400 450 500

Wavelength (nm)
Figure 8.7 Effect of pH on the visible spectrum of 3'-methoxy-4',7-dihydroxyftavylium
chloride (5.2 x 10-3 gil) in aqueous solution after standing for 2 h (redrawn from Jurd (391)
with permission). pH: 1,0.63; 2, 2.30; 3, 2.65; 4,2.94; 5, 3.11; 6,3.40; 7,3.67; 8, 3.81; 9, 4.10;
10,4.36.
calibration curve prepared using a standard pigment [202]' The most

common bleaching agents have been sodium sulfite [389, 394, 395] and
hydrogen peroxide [396]. The subtractive method has come under criticism
since the bleaching agents used may cause a decrease in absorbance of
some interfering substances, resulting in erroneously high values of total
pigment concentration [310].
The differential method relies on the structural transformations of
anthocyanin chromophores as a function of pH (Figures 8.5 and 8.7). It is
also based on the observation that the spectral characteristics of interfering
brown (degradation) products are generally not altered with changes in pH
[223]. The difference in absorbance at two pH values and same wavelength
(e.g. the anthocyanin visible maximum or at 550-600 nm for polyacylated

pigments [381]) is assumed to provide a measure of anthocyanin concen-
tration since absorbance due to interfering substances cancels in the
subtraction. A number of different pH values have been used in the
differential method, but measurement at pH 1.0 and 4.5 is most common
[397]. Noteworthy, is that an insufficient time for equilibration between the
different anthocyanin species upon pH adjustment can yield erroneous
results.
An advantage of the differential method is that data can be used to
calculate an anthocyanin degradation index (DI), defined as the ratio of
total anthocyanin determined at a single pH to that determined by the
differential method [310]. The DI is a useful measure of browning products
that absorb at the visible maxima of anthocyanins. However, it cannot be
used with such products as aged red wines which contain condensed/
polymeric anthocyan ins that are relatively pH insensitive and absorb
strongly at the visible maxima of anthocyanins [223]. They cannot be
regarded as anthocyanin degradation products since they contribute to the
total anthocyanin concentration and colour of the product. The DI has
been used as an indicator of anthocyanin colour and overall quality
deterioration in a number of berry fruit products [55, 398-401]. The simple
ratio of absorbance of red (e.g. anthocyanin) pigments measured in the 465
to 550 nm region to that of brown (e.g. degradation) products measured in
the 400 to 440 nm region also provides an index of anthocyanin or, more
accurately, colour deterioration [389, 402, 403].
Measurements of anthocyanin concentration rarely correlate with the
actual colour of a given product or sample; since the pH of analysis and
endogenous pH of the sample usually differ, the distribution and concen-
tration of anthocyanin chromophores also differ. Wrolstad [389] reviewed
methods for determining colour density, polymeric colour, percentage
contribution by tannin and monomeric anthocyanin colour, calculated
from only a few absorbance readings at the native sample pH. A two point
index-the difference in absorbance at the visible maximum and at
420 nm-has been shown to correlate well with visual appearance [404].
However, for those with available equipment, tristimulus measurements
have been shown to correlate much better with human visual responses
[310, 403-405].
The quantitative analysis of individual anthocyan ins requires their prior
separation/purification from a mixture, usually carried out by chromatog-
raphy. Spectral measurements are usually of limited use since the absorp-
tion spectra of most anthocyanins are so similar [360]; yet, as previously
mentioned, molar absorptivities have been reported and used for quantita-
tion of a number of anthocyanins. Quantitation of individual anthocyanins
has traditionally involved their separation by paper or thin layer chroma-
tography, followed by photometric determination of pigments in situ by
reflectance or transmittance densitometry [55, 111, 139, 406]. Reversed-

phase HPLC is currently the method of choice for quantitation of
anthocyanins [32, 312, 356]. The speed, resolving power and ability to
quantitatively separate pigments has revolutionized the qualitative and
quantitative analysis of anthocyanins.
8.4.6 Current and potential sources and uses

Several synthetic flavylium salts have been suggested as potential food
colour additives [48, 271,191, 193, 195,407]. The production of synthetic
anthocyanins identical to their natural counterparts (i.e. nature-identical)
is also possible [51], although yields via chemical synthesis are generally
poor [408]. These latter products provide a source of food colour additives
with consistent quality/purity and known properties. However, all syn-
thetic flavylia, including the nature-identicals, currently require extensive
toxicological testing prior to clearance as safe food additives. Public
concern over the safety of food additives has restricted the sources of
commercial anthocyanin preparations to food plants naturally containing
these pigments; this is not likely to change in the very near future.
The potential of various food plants as commercial sources of antho-
cyanins is limited by availability of raw materials and overall economic
considerations. Numerous plants and plant parts have been suggested as
potential commercial sources (reviewed by Francis [9]), including cran-
berry [277, 334, 409, 410], blueberry [321], red cabbage [321, 322], roselle
(Hibiscus sabdariffa) calyces [276, 316, 411], miracle fruit [133, 412],
berries of Vibernum dentatum [413], bilberries (Vaccinium myrtillus) [414],
duhat [Sysyium cumini Linn.) [415], black olives [416], flowers of Clitorea
ternatia [63, 290] and Ipomea tricolor [417,418], and leaves of cherry-plum
[419,420], Perilla ocimoides (i.e. shiso) [233] and Tradescantia pallida [69,
198, 379]. The anthocyanins of interest from most of these sources are di-,
tri- or polyacylated, and many exist in more complex, supra-molecular
forms [27, 31, 421, 422]. They, therefore, display good stability in aqueous
systems at slightly acid or neutral pH; yet, their yields are often poor and
source materials are not generally readily available in quantities that may
be necessary for commercial extraction. Grapes provide the greatest
quantity of anthocyan ins for use as food colorants: production of grapes
constitutes about a quarter of the annual fruit crop worldwide [9, 239].
Both a spray-dried powder and a concentrated solution containing grape
anthocyan ins have been marketed for years from Italy under the trade
name of Enocolor (formerly Enocianina) [423].
Economically, the best potential commercial sources of anthocyanins are
those from which the pigment is a byproduct of manufacture of other
value-added products [424]. The extraction of grape anthocyan ins into
juice or wine is an inefficient process and large quantities of extractable
pigment are available in grape pomaces and waste material [9]. Antho-
cyanin extracts have been produced from these sources [207, 234, 277, 317,
318] and used to successfully colour a wide range of commercial products
[425]. Anthocyanin extracts have also been prepared from cranberry
presscake [277, 334, 409]. Approximately 40% of the anthocyan ins from
cranberries remains in the presscake following juice extraction [404]. The
tropical red berry known as miracle fruit contains a taste modifier,
Miraculin, that has potential as a sweetener; anthocyanin extracts have
been obtained as a by-product of taste modifier production [133, 411]. The
production of expensive food or other plant crops for their pigment content
alone is not economically feasible. Yet, the availability of highly pigmented
inexpensive crops such as red cabbage [321, 322] and bilberries [413] makes
the use of these crops as potential sources of commercial anthocyanin
preparations viable.
Anthocyanin preparations from natural sources may vary considerably
in quality [426] since they may contain partially degraded and/or con-
densed pigments, tannins, copigments and various impurities which are
co extracted with the anthocyanins. Stable, relatively pure anthocyanin
preparations have been produced in relatively large quantities (e.g. up to
about 5% on a dry basis) by suspension cultures of cells of Populus spp.
[427,428], Daucus carota [429, 430], Vitis spp. [431] and Euphorbia millii
[432]. Anthocyanin production by cell cultures from numerous other
sources has been reported; studies have been directed not only towards
improving productivity for large scale pigment production, but also
towards identifying those factors that influence anthocyanin accumulation.
One of the greatest limitations to commercial production of anthocyanins
via cell or tissue culture is the requirement for light. Only a few species of
plant cell culture have been reported to produce anthocyanins in the dark,
but levels of production have been generally low. The exception is a highly
productive cell line of Aralia cordata, obtained by continuous cell-
aggregate cloning, that has been reported to yield anthocyanins in
concentrations as high as 17% on a dry weight basis [433-435]. While an
effective scaled-up process has been documented [433], costs of production
must be competitive with conventional anthocyanin recovery processes.
Costs could be reduced considerably should an inexpensive alternative to
refined sugars be made available and found to be suitable, e.g. milk whey
[436]. Time may yet see the commercial production of anthocyanin
preparations via application of biotechnologies.
Natural anthocyanin preparations used as food colorants suffer the same
fate as endogenous pigments; their addition to food systems may encour-
age reactions with endogenous constituents that can lead to their stabiliza-
tion or destabilization, and thereby influence product quality. With careful
formulation/selection of certain ingredients, choice of appropriate stages
during formulation and processing at which the colorant and/or other
ingredients are added, and control of processing and storage conditions, a

wide range of high quality products can be attractively coloured with
anthocyanins. Counsel et al. [436] have reviewed numerous food appli-
cations for anthocyan ins and other natural pigments, including chewing
gums, hard candies, fruit chews, dry beverage powders, cream fillings,
icings and fruit coloration. Anthocyanins may also be successfully used in
high acid foods such as soft drinks, jams and jellies, and in red wines where
they contribute to the overall ageing process [408]. For those food products
for which endogenous pigments are to serve as the only or major source of
coloration, procurement of intensely coloured raw material is necessary to
ensure that if some pigment is destroyed during processing/storage, a
sufficient portion remains to impart the characteristic product colour [389].
Few practical stabilizing agents have been found for the anthocyan ins
[424]. Of nineteen different additives, only thiourea, propyl gallate, and
quercetin demonstrated a stabilizing effect on colour retention in straw-
berry juice and buffered pigment solutions [210]. Cysteine, which may act
as a reducing agent and/or inhibitor of PPO, was shown to inhibit
anthocyanin degradation in Concord grape juice below a temperature of
75°C [244]. Ascorbic acid can render a stabilizing effect on one hand, by
being preferentially oxidized by PPO, but a destabilizing effect on the
other, by indirectly oxidizing the anthocyanins via its oxidation products
[51]. Tartaric acid and glutathione have been shown to have a protective
effect on anthocyan ins by acting as mildly acidic and antioxidant agents,
respectively [437]. The addition of phenolic compounds, such as rutin and
caffeic acid, also markedly stabilized the colour of blood orange fruit juice,
presumably by means of intermolecular co-pigmentation [437]. Indeed, in
view of the factors that influence pigment stability, the most practical
means of stabilizing the colour of anthocyanins within the bounds of
maintaining their 'natural' status is via complex formation (e.g. surface
adsorption), intermolecular co-pigmentation and/or condensation re-
actions (e.g. with proteins, tannins or other polyphenols).
8.5 Betalains
B.5.1 Structure
The structure of the betalain chromophore may be described as a
protonated 1,7-diazaheptamethin system (Figure 8.8); betaxanthins and
betacyanins are distinguished by substitution on the dihydropyridine
moiety by specific Rand R' groups. The yellow betaxanthins are character-
ized by having Rand R' groups that do not extend the conjugation of the
1,7-diazaheptamethin system. However, if conjugation is extended, where-
by Rand R' comprise a substituted aromatic ring (i.e. cyclodopa), the
H
COOH HOOC.... COOH
Figure 8.8 Structural transformations of the betalain chromophore.
chromophore is red and characterized as a betacyanin [3, 4]. All beta-

cyanins are glycosylated [3] and derive mainly from the aglycones betanidin
or isobetanidin (the C-15 epimer of betanidin). 2-Decarboxybetanidin has
also been reported, as a minor pigment in Carpobrotus ancinaciformus
[438], and a glycoside of neobetanidin, 14,15-dehydrobetanidin, has also
been identified as a natural constituent of some plants [439, 440]. The most
common betacyanin is betanin, the 5-0-J3-glucoside of betanidin (Table
8.4). The most common glycosyl moiety is glucose; sophorose and
rhamnose occur much less frequently. Acylation of pigments may also
occur whereby the acyl group is attached via an ester linkage to the sugar
moiety. The most common acyl groups include sulfuric, malonic,
3-hydroxy-3-methylglutaric, citric, p-coumaric, ferulic, caffeic and sinapic
acids. The structures of many acylated betacyanins have yet to be fully
characterized [4].
In betaxanthins the cyclodopa unit of betacyanins is displaced by either
an amine or amino acid, suggesting the potential for existence of over
200 different betaxanthin pigments. For example, the R' group of
vulgaxanthin-I from Beta vulgaris is a residue derived from glutamic acid;
that of indicaxanthin from Opuntia fiscus-indica joins with the R group to
give a proline moiety (Table 8.5). Other R' groups of the betaxanthins
include glutamine, methionine sulfoxide, tyramine, DOPA and 5-hydroxy-
norvaline. The R group of betaxanthins is usually a hydrogen. Like the
betacyanins, the structures of many of the betaxanthins have yet to be fully
elucidated [4, 441, 442].
8.5.2 Distribution
Unlike the more widely distributed anthocyanins, betalains have been
detected only in red-violet, orange and yellow-pigmented botanical species
belonging to closely related families of the order Caryophyllales [3, 4, 33].
Table 8.4 Structures of some known betacyanins
N
00
N
H
COOH
~
Cl
Compound Substitution pattern Botanical source Reference ~
t""
Rs R6 ~
o
Betanin j3-glucose H Beta vulgaris 375,376 (":l
Amaranthin 2' -O-(j3-glucoronic acid)-j3-glucose H Amaranthus tricolor 377,378 "o
Celosianin-I 2' -0-[ O-(p-coumaroyl)-j3-glucuronic acidj-j3-glucose H Celosia cristata L. 380 S
Chenopodium rub rum 498
Celosianin-II 2' -0-[ O-(trans-feruloyl)-j3-glucuronic acidj-j3-glucose H Chenopodium rub rum 509
~
Iresinin-I 2' -O-(j3-glucuronic acid)-6' -O-(3-hydroxy-3- H lresine herbstii 380 ~
methylglutaryl)-j3-glucose
Prebetanin 6' -0-(S03H)-j3-glucose H Beta vulgaris 379
Phyllocactin 6' -0-( malonyl )-j3-glucose H Phyllocactus hybridus 380
Rivinianin 3' -0-(S03H)-j3-glucose H Rivinia humilis 510
Lampranthin-I 6' -O-(p-coumaroyl )-j3-glucose H Lampranthus spp. 381
Lampranthin-II 6' -O-feruloyl-j3-glucose H Lampranthus spp. 381
Bougainvillein-r-I j3-sophorose H Bougainvillea glabra var. 382
sanderiana
Gomphrenin-I H j3-glucose Gomphrena globosa 383
Gomphrenin-II H 6' -O-(p-coumaroyI)-j3-glucose Gomphrena globosa 497
Basella rubra 511
Gomphrenin-III H 6' -O-(trans-feruloyl)-j3-glucose Gomphrena globosa 497
Basella rubra 511
Gomphrenin-V H 6' -O-(trans-feruloyl)-j3-glucose Gomphrena globosa 383
Table 8.5 Structures of some known betaxanthins
R'O~I ~ +.'
.H
R,O N eOOH
Compound
H~~COOH H
I
Substitution pattern Botanical source Reference
R R'
Indicaxanthin Rand R' together form proline Opuntia ficus-indica 388

Portulacaxanthin-I Rand R' together form Portulaca grandiflora 389
hydroxyproline
Portulacaxanthin-II H tyrosine Portulaca grandiflora 492
Portulacaxanthin-III H glycine Portulaca grandiflora 492
Vulgaxanthin-I H glutamine Beta vulgaris 390
Vulgaxanthin-II H glutamic acid Beta vulgaris 390
Miraxanthin-I H methionine sulphoxide M irabilis jalapa 391
Miraxanthin-II H aspartic acid Mirabilis jalapa 391
Miraxanthin-III H tyramine M irabilis jalapa 391
Miraxanthin-V H dopamine M irabilis jalapa 391
Dopaxanthin H L-DOPA Glottiphyllum longum 392
Humilixanthin H hydroxynorvaline Rivinia humilis 393
Muscaaurin-I H ibotenic acid Amanita muscaria 394
Muscaaurin-II H stizolobic acid Amanita muscaria 394
Muscaaurin-VII H histidine Amanita muscaria 394
Many of these species have developed the ability to carry out C4 -

photosynthesis and typically contain unusual sieve-element plastids and
relatively large concentrations of ferulic acid in their cell walls [3].
Although remaining a point of phylotaxonomic contention, ten betalain-
producing families have been identified: Aizoaceae, Amaranthaceae, BaseJla-
ceae, Cactaceae, Chenopodaceae, Didiereaceae, Holophytaceae, Nyctagina-
ceae, Phytolaccaceae (including Stegnospermaceae) and Portulacaceae.
Betalains of fungal origin have also been found. A violet betacyanin,
muscapurpurin, and seven yellow betaxanthins, muscaaurins-I to -VII,
have been isolated from the poisonous mushroom Amanita muscaria (fly
agaric; Order Agaricinales) [5, 6]. The unusual amino acid stizolobic acid
and a cyclized form of this amino acid are found in muscaaurin-II and
muscapurpurin, respectively. Some of the muscaaurins have identical
structures to those of known betaxanthins produced by members of the
Caryophyllales, i.e. indicaxanthin, vulgaxanthin-I and -II, miraxanthin-II1
[4-6]. The yellow pigment muscaflavin, possessing a dihydropyridine

structure characteristic of the betalains, has also been found in pigment
extracts of Amanita muscaria [443, 444]. Muscaflavin also occurs naturally
in Hygrocybe toadstools accompanied by pigments that are muscaflavin
analogs of the betaxanthins, commonly referred to as hygroaurins [6, 445].
Betalains, as with the anthocyanins, accumulate in cell vacuoles of the
flowers, fruits and leaves of the plants that synthesize them, mainly in
epidermal and/or subepidermal tissues. Moreover, betalains often accumu-
late in plant stalks and are found in high concentrations in underground
parts of the common red table beet [5]. Of the numerous natural sources of
the betalains (e.g. red beet, cacti, cockscomb, pokeberry), the red beet
and prickly pear are the only food products containing this class of
pigments [33, 446]. The major betalain in beets, betanin, accounts for 75 to
95% of the total betacyanin content; the remainder is comprised of
isobetanin, prebetanin and isoprebetanin. These latter two pigments are
sulphate monoesters of betanin and isobetanin, respectively. The major
yellow pigments in the red beet are vulgaxanthin-I and -II.
8.5.3 Biosynthesis
The biosynthesis of betalains arises from arogenate of the shikimate
pathway, and its conversion to tyrosine via arogenate dehydrogenase [447].
The dihydropyridine moiety present in all betalains is synthesized in vivo
from two molecules of L-5,6-dihydroxyphenylalanine (L-DOPA), one of
which must first undergo 4,5-extradiol oxidative cleavage and recyclization
[448] through seco-DOPA [449, 450] to betalamic acid (Figure 8.9).
Alternatively, in toadstools L-DOPA undergoes 2,3-extradiol cleavage to
form muscaflavin [6]. Betalamic acid constitutes an essential structural
feature of all natural betalains, but accumulates as a natural constituent
only in those plants that produce betaxanthins [451-453]. The conversion
of L-DOPA to betalamic acid is catalysed by a dioxygenase [454, 455]. Its
subsequent condensation with cyclodopa or a derivative of cyclodopa leads
to betacyanidins; condensation with other amines or amino acids results in
betaxanthins. Glycosylation of betacyanidins occurs late in the biosynthetic
pathway [456], via uridine-5-diphosphate (UDP) linked sugars as mediated
by nucleoside-diphosphate-sugar-dependent glycosyl transferases [457].
Alternatively glycosylation of cyclodopa may occur prior to its condensa-
tion with betalamic acid [458, 459]. Acylation takes place subsequent to
glycosylation, and most often involves hydroxycinnamic acid transferases
with 1-0-hydroxycinnamic acid-acylglycosides as the acyl donors [460-
462].
The use of betalain-producing cell cultures has facilitated the elucidation
of enzymes involved in betalin biosynthesis and the influence of various
environmental factors such as light, temperature, precursor availability,
~
I
HO '"
.H
. COOH
Tyrosine
HO~.H
H O V H,NA COOH
_____> HO
WH
HO '" ~ COOH
H
L-DOPA
5-Cyclodopo
1
[ HO~'"
I, J ·.H
HO~
Ho~~Acoo .H
"0 A ~;: DO
I / eoo:;9-c~-_OG-;~_c:-on_atlo_n---.. Betacyanlns
H I Betanldln
HOOC" N COOH
I
H
5-Betalamlc acid
Amino acid or amine

other than cyclodopa 1
.1
Betaxanthlns
Figure 8.9 Biosynthesis of betalains.
presence of cytokinins and auxins, and availability of nutrients. A key

regulatory step in the biosynthesis and accumulation of betalains appears
to be the hydroxylation of tyrosine [463] which may be under auxin and/or
cytokinin control [464-466]. The ratio of auxin to cytokinin is an important
factor in establishing and stabilizing in vitro cell culture lines of specific
coloured phenotypes [467], and may be important in regulating the type
and accumulation of betalains in vivo. In contrast to anthocyanin accumu-
lation in both plant and cell culture systems, the accumulation of betalain
pigments is greatest during growth or cell division [468]. Light can serve as
a powerful stimulant of betalain biosynthesis, although its effect appears to
be species-dependent; while light is an absolute requirement for pigment
synthesis in some plants (e.g. Amaranthus and Phytalacca spp.), it is not
for others (e.g. Beta vulgaris). UV- and red-light-induced betalain synthe-
sis appears to be mediated by phytochrome via gene activation [469].
Cytokinins have also been shown to promote betalain biosynthesis,
presumably also via gene activation, even in the dark [470, 471]. The
mechanisms of light- and cytokinin-induction, while synergistic, appear to

be independent of one another [472].
8.5.4 Colour and structural stability

The stability of betalains is influenced by numerous confounding factors
(e.g. pH, temperature, oxygen, water activity, light) that have limited the
use of these pigments as food colorants. The majority of work to date has
focused on the influence of these various factors on betanin. Comparat-
ively fewer studies have been reported on the effects of these factors on the
betaxanthins. To date, no systematic work has been carried out to
investigate the effects of acylation and/or betalain substitution on the
structural stability of these pigments.
8.5.4.1 pH. Generally, the red colour of betanin solutions remains

unchanged from pH 3.0 to 7.0, exhibiting maximum absorption at
537-538 nm [446]. Below pH 3.0 the colour changes to violet as the
absorption maximum is shifted to 534 to 536 nm and its intensity decreases;
a slight increase in absorbance at 570-640 nm is also observed. Above pH
7.0 the colour of betanin solutions also becomes bluer, due to a bathochro-
mic shift in the wavelength of maximum absorption. The greatest blueing
effect occurs at pH 9.0, with maximum absorption at 543 to 544 nm. Above
pH 10.0 a decrease in intensity at the absorption maximum of 540-550 nm
is accompanied by an increase in absorption at 400-460 nm due to
liberation of betalamic acid, which is yellow; pigmentation thus changes
from blue to yellow as a result of alkaline hydrolysis of betanin to
betalamic acid and cycIodopa-5-0-glycoside [473] (Figure 8.10).
8.5.4.2 Temperature. The thermolability of the betalains is probably the

most restrictive factor in their widespread use as food colorants. The
thermal degradation of both betanin and vulgaxanthin-I have been shown
to follow first-order kinetics over a pH range 3.0 to 7.0 under aerobic
conditions [446, 474-477], but to deviate from first-order kinetics under
anaerobic conditions [478]. The activation energy (Ea) for betanin deg-
radation in the presence of oxygen is approximately 84 kJ mol-I at pH 3.0-
7.0 [476, 477,479,480]. Under similar conditions, vulgaxanthin-I has an Ea
of around 67 kJ mol-I [476], and is therefore more sensitive than betanin to
thermal degradation. The thermal stability of the betalain pigments is
greatest between pH 5.0 and 6.0 in the presence of oxygen [474, 476, 481]
and between pH 4.0 and 5.0 in its absence [474, 477]. The rate of
degradation of betanin and vulgaxanthin-I is more rapid in model systems,
suggesting a protective effect conferred by constituents in the natural
system (e.g. by polyphenols, antioxidants). In both natural and model
systems betanin may be regenerated via Schiff's base condensation
\0=
G~WH
00 ~
WZJlCOOH
I
H
lsobetonln
1~~~
GuO
WH
W ~t~
HO::-- 6+
'COOH ~~OH GIuO
• ::-- I + "H +
HO N COOH
H I
~
I
HOOC COOH H
I
H Cyciodopa-5-O-glucoside Betolamlc Acid
Betonin
l~
6
GluO~
~~H
COOH
HO::-- + CO,
H I
H COOH
I
H
Decorboxylated Betonln
Figure S.lO Mechanisms of betanin degradation,
of betalamic acid with cyclodopa-S-O-glycoside (Figure 8.10) [474, 481-

483]. Betanin may also yield isobetanin and/or decarboxylated betanin,
their formation being favoured upon heating at pH 3.0-4.0. These latter
compounds do not differ significantly from betanin in their absorption/
colour properties and they yield similar thermal and pH degradation
products (i.e. betalamic acid derivatives and cyclodopa-S-O-glycoside)
[473].
The primary steps in betanin degradation, as mediated by temperature
and/or pH, involve nucleophilic attack (e.g. by water) at the C-ll position
[484], i.e. at the carbon atom adjacent to the quaternary amino nitrogen
(Figure 8.10). As already alluded to, this reaction is reversible, the
regeneration of betanin being greatest in the pH range in which the
pigment is most stable [477, 482, 48S]. While maximum stability of
betalamic acid occurs at pH 6.0-8.0 [486], that of cyclodopa-S-O-glycoside
occurs at pH 3.0 [477]. The Ea for regeneration of pigment ranges from 2.S
to 14.7 kJ mol-I at pH 3.0 to S.O [483]. Thus, provided that betalamic acid
and cyclodopa-S-O-glycoside are present, regeneration of betanin is
favoured, especially at lower temperatures. However, betalamic acid is
heat sensitive; it may undergo aldol condensation or participate in Maillard
reactions thereby making it unavailable for the regeneration reaction [477].
At temperatures approaching 130a C under alkaline conditions betalamic
acid yields formic acid and 4-methylpyridine-2,6-dicarboxylic acid [487].
The glycosidic moiety of cyclodopa-S-O-glycoside may also be cleaved at
high temperatures. It is also very susceptible to oxidation reactions,
initiating polymerization to melanin-type compounds [477]. Thus, as
temperature increases, particularly in the presence of oxygen (viz. during
storage or processing), irreversible betanin degradation is promoted. The
quantity of betanin degradation during and regeneration after thermal
treatment is dependent on, among other things, the presence of nucleo-
philes, temperature, pH, and also initial betanin concentration; as initial
betanin concentration increases so does colour stability [483].
8.5.4.3 Light. Like the anthocyanins, betalains are susceptible to degra-

dation by various sources of radiation. The photochemical destruction of
these pigments involves molecular oxygen and follows first-order kinetics
[480]. Photooxidation of red beet pigments is also pH-dependent, greater
degradation occurring at pH 3.0 than at pH S.O [488]. Exposure of betanin
solutions to UV/visible light increases the rate of pigment degradation by
as much as lS% [446]. Absorbed light in the UV or visible range
presumably acts to excite 1T electrons of the pigment chromophore to a
more energetic state (i.e. 1T*). This causes a higher reactivity or lowered
activation energy for the molecule. The result is a decrease in the
dependence of degradation rate on temperature [480]. Gamma irradiation
also enhances the rate of betanin degradation; doses exceeding 100 krad
result in total loss of colour [489].
8.5.4.4 Oxygen. Involvement of molecular oxygen in pigment degrada-

tion is not restricted to photooxidation; it has been implicated in thermal
degradation as well. Pigment losses are reduced through nitrogen purging
of purified betanin solutions [474] and through application of glucose
oxidase (an oxygen scavenger) to red beet concentrates [490]. Metal ions,
common food constituents or contaminants from processing equipment,
may function as pro-oxidants. Metal cations (e.g. iron, copper, tin, alu-
minium) accelerate the rate of betanin degradation [491, 492]. As elec-
tron donors or acceptors, depending on their oxidation state, metal ions
may stabilize the electrophilic centre of betalains resulting in rearrange-
ment of associated bonds, destruction of the chromophore and concomi-
tant loss of colour [491]. No protective effect is provided by ascorbic acid
or a-tocopherol, commonly added to food systems as antioxidants;
extremely elevated concentrations of ascorbic acid (i.e. 1000 ppm) decrease

pigment shelf-life since it then acts as a pro-oxidant. High concentrations of
citric acid and EDTA (e.g. 10 000 ppm) have been found to markedly
improve betanin stability, possibly by partially neutralizing the electrophilic
centre (i.e. they may sequester betanin) and/or by sequestering metal ions
that otherwise act as pro-oxidants [491, 492]. The brunt of evidence suggests
that betanin oxidation does not occur via a free radical mechanism.
8.5.4.5 Water activity (a w ). The effect of aw on betanin stability has also

been documented [475, 493, 494]. First-order rate constants for betanin
degradation have generally varied exponentially with respect to aw , a four-
fold increase in pigment stability being observed as aw decreases from 1.0
to 0.37. Reduced aw may increase betalain stability by reducing the
nucleophile (i.e. water) concentration, by limiting reactant mobility and/or
by decreasing oxygen solubility. Using various water/alcohol model
systems and correcting for the appearance of degradation products, Simon
et al. [494] noted that there was a maximum in the dependence of the rate
constant on aw . Their results were explained in terms of the propensity of
the electrophilic centre of betalains to nucleophilic attack. Increasing
amounts of water in alcohol increased the rate constant since the water was
better able to transfer protons than alcohol. However, for a certain a w or
water/alcohol ratio, the contribution of the proton transfer rate to the
overall rate constant approached a maximum. Further increases in aw were
associated with a decrease in concentration of the stronger nucleophile
(i.e. alcohol), and the rate constant thus decreased. Their results [494]
demonstrate the important role of solvolysis in betanin degradation.
8.5.4.6 Enzymes. As with other natural pigments, specific decolorizing

enzymes also influence betalain stability. Decolorizing activity has gener-
ally been found concentrated where most pigment also accumulates, that
is, in subcellular tissue separated from the epidermal layers [495, 496].
However, while betalains are localized in cell vacuoles, the majority of
decolorizing activity (i.e. 70--76%) occurs in leucoplasts; additional activity
is found mainly associated with microsomal fractions [42]. This different
intracellular compartmentation of enzyme and substrate may be regarded
as an adaptation of betalain-producing plants to function in pigment
storage. Much of the decolorizing activity becomes strongly associated with
cell wall material during extraction, requiring high salt concentrations
(e.g. 2: 1 M NaCl) for its recovery [43,497]. The presence of decolorizing
activity is of consequence during extraction/purification, and in subsequent
use of the extract as a food colorant, since pigment yields may be
significantly reduced [496, 497]. Inactivation, for example through mild
thermal treatment, may be required to prevent or reduce enzyme catalysed
fading of pigment extracts.
The enzymatic conversion of betalains to colourless species is associated

with the consumption of molecular oxygen, but does not involve hydrogen
peroxide [42]; the enzyme has thus been referred to as betalain oxidase.
An enzyme with similar properties has been isolated from several sources
[43, 495, 496, 498]. It decolorizes both betacyanins and betaxanthins, and
exhibits optimum activity at pH 3.4-3.5 and 38-40°C. Its activity is
inhibited by NH4 N0 3 , KN0 3 , KCI and NaCl. Further inhibition of the
enzyme by sodium azide, sodium diethyldithiocarbamate and thiourea has
led to the suggestion that its activity may be dependent on the presence of a
metal ion (i.e. iron) in the active site and on disulfide linkages [495].
Enzymatic reaction products are reportedly identical to those initially
appearing as a consequence of thermal treatment or alkaline hydrolysis
(Figure 8.10) [498]; however, in the acidic conditions of the reaction,
betalamic acid is unstable and is readily converted to 2-hydroxy-3-
hydrobetalamic acid [42]. Three additional betalain oxidases were partially
purified from tissues of Phytolacca americana, with pH optima between 5.0
and 5.5 [43]. These particular enzymes were suggested to have a different
function in vivo from that with a pH optimum at pH 3.5, since they were
not inhibited by various salts. Further work is required to more fully
characterize the various betalain oxidases, to develop effective strategies
by which their activity can be controlled.
8.5.5 Extraction and analysis
8.5.5.1 Extraction and purification. The availability of commercial

preparations of betalains is currently restricted by legislation in most
countries to beet juice concentrates produced by concentrating juice under
vacuum to 60-65% total solids, or beet powders produced by spray- or
freeze-drying the concentrate [10, 11]. Beet juices have traditionally been
prepared by hydraulic press operations in which raw beets are blanched,
chopped, pressed and the expressed juice filtered. Less than 50% recovery
of betalains can be expected from most conventional press operations
unless macerating enzymes are used to facilitate pressing. However, up to
90% recovery of betalains from beet roots has been reported with use of a
continuous diffusion apparatus [499, 500]. Since commercial preparations
are not pure they vary in colour, depending on the proportions of
betacyanin and betaxanthin extracted, and often possess beet-like odour
and taste due to the presence of geosmin. Purification of the betalains is
possible on both small and large scale, although the methods used
generally give poor yields.
Separation and purification of betalains from crude plant extracts is
usually necessary if qualitative and/or quantitative analyses are to be
carried out. A number of different approaches have been reviewed [3, 4].
Preliminary separation of betalains in a variety of crude aqueous plant

extracts has been accomplished by non ionic absorption on strongly acid
resins (e.g. Dowex SOW) followed by polyamide column chromatography,
using increasing concentrations of methanol in aqueous citric acid as the
developing solvent [501, 502]. Citric acid is removed from resolved
betalains by resin treatment. Sequential chromatography on series of
Sephadex ion exchangers has also provided good separation of betalains
[5, 443, 459, 503]. Adams and von Elbe [504] reported that gel filtration
column chromatography on Sephadex or polyacrylamide (Bio-Gel P-6)
gels could be used to rapidly and efficiently separate the betalains from
beet juice. Numerous other chromatographic and electrophoretic pro-
cedures have been used over the years to separate/purify betalains from a
variety of sources. However, most of these procedures are time-
consuming. Since their initial application for analysis of the betalains [505],
HPLC methods have become widely adopted as a fast and efficient means
of betalain separation and analysis [440, 441, 506-511].
8.5.5.2 Qualitative analysis. Structural characterization and identifica-

tion of betalains generally involves direct comparison of spectroscopic,
chromatographic and electrophoretic properties with authentic standards,
before and after controlled hydrolysis [4, 33, 512]. Since the betalains
differ from anthocyanins in being more sensitive to acid hydrolysis, in their
colours with changes in pH, and in their chromatographic and electrophor-
etic properties, they may be easily differentiated by simple colour tests
[3, 311]. Since the occurrence of these two pigment types is mutually
exclusive, colour tests can usually be performed using crude plant extracts
[311 ].
Betacyanins and betaxanthins possess similar chromatographic and
electrophoretic properties; thus, similar techniques are used for their
isolation/purification. However, polyamide chromatography is only
moderately successful for the separation of betaxanthins. Preparative
paper electrophoresis and chromatography using alternate supports are
commonly used for separation of individual betaxanthins from co-
occurring betacyanins and other plant constituents [512]. HPLC is also
becoming an invaluable means by which to separate and analyse the
betalains; the number of betaxanthin structures was found to be consider-
ably greater using HPLC than was known from more traditional separation
techniques [441]. The spectral properties and RP-ion pair HPLC retention
characteristics of a number of natural and semisynthetic betaxanthins were
recently reported [442].
Tentative identification of betalains can be deduced from their chroma-
tographic or electrophoretic behaviour [512]. For example, on polyamide
or cellulose-based resins:
1. The retention of betacyanins decreases with increased glycosyl substitu-

tion.
2. The retention of 6-glycosides exceeds that of the 5-glycosides.
3. Iso-derivatives are retained slightly longer than their corresponding
C-15 epimers.
4. Acyl groups, aromatic more than aliphatic, increase the retention of
pigments.
Each pigment possesses characteristic electrophoretic mobility between
0.30 and 1.78, relative to betanin, usually determined at both pH 4.5 and
2.4 [502, 513].
Corroborative data may be provided by analysis of absorption spectra.
All betacyanins exhibit visible maxima between 534 and 552 nm [502],
while all betaxanthins have absorption maxima in the range 474 to 486 nm
[513, 514]. Acylated betalains generally exhibit a second absorption
maximum in the UV region of 260 to 320 nm, where the absorption of non-
acylated pigments is weak [512, 515]. As with the anthocyanins, the
number of acyl residues in betalains can be estimated from the ratio of
absorption at the visible maximum to that at the UV maximum [512].
Further valuable structural information may be obtained from lH-NMR
spectra [439, 441, 515]. Indeed, the current sensitivity of NMR spectrom-
eters permits complete structural characterization of betalain pigments,
even when sample size is small [4]. Corroborative molecular weight data
have been obtained by FAB MS and/or GC/MS [439, 441]. The former
technique has the advantage that the betalain pigments do not require
derivatization beforehand. The IR spectra of betalains have been found to
provide little useful information [515].
Betacyanins occur naturally as C-15 epimers, the betanidin derivatives
generally being present in greater quantities than their corresponding
stereoisomers [512]. Strong acid hydrolysis of betanin gives, along with
glucose, a mixture of both aglycones [516], while milder controlled acid
hydrolysis [517], or ~-glucosidase hydrolysis [518], yields only betanidin.
Acylation often renders the sugar moiety resistant to enzymatic cleavage.
Acid hydrolysis of isobetanidin derivatives yields only the isobetanidin
stereoisomers [501]. It is thus possible to establish whether two betacyanins
of unknown structure are C-15 epimers; if this is the case, subsequent
analyses can be conveniently carried out using a mixture of both pigments
[512]. In addition to acid and enzyme treatments, oxidation using alkaline
hydrogen peroxide can liberate constituent glycosyl moieties [519]. The
cleaved sugar group(s) may be identified by conventional paper chroma-
tography or HPLC; the aglycone and its degradation products may be
identified by conventional methods, for example by comparison with
authentic standards using paper chromatography, electrophoresis or HPLC.
The position(s) of glycosyl attachment may be determined by treatment
with excess diazomethane that converts the aglycone into an O-methyl

neobetacyanidin trimethylester glycoside in which the originally free
phenolic hydroxyl group of the aglycone has been methylated and thus
labelled for identification [512]. In acidic media, the neobetacyanidin
exhibits a bathochromic shift of about 100 nm relative to the parent
glycoside [520, 521]. Diazomethane methylation similarly converts beta-
xanthins into neo-pigments which exhibit absorption maxima at 340 to
360 nm and show a bathochromic shift of 80-100 nm in acid solution [522].
Alkaline fusion of the neo-pigments yields 4-methylpyridine-2,6-
dicarboxylic acid [3, 523].
The identity and position of sugar groups, as well as of acyl or sulfate
groups linked to the glycosyl moieties, are now most often determined by
NMR techniques. Traditionally, acyl and/or sulfate residues are removed
by cold alkaline hydrolysis under nitrogen (i.e. oxygen must be absent),
isolated and identified by chromatographic techniques [515]. Their position
of attachment has been determined by:
1. Mild periodate oxidation, subsequent borohydride reduction, mild acid
hydrolysis, a second borohydride reaction, then chromatographic iden-
tification of the resulting polyols; or
2. Pigment methylation with Mel/AgO in dimethylformamide and subse-
quent identification of the products released after acid hydrolysis of the
permethylated compound [512, 515].
8.5.5.3 Quantitative analysis. In the absence of interfering substances

(e.g. fresh extracts or juices, purified samples), the quantitation of
betalains can be accomplished using spectrophotometry, whereby absorb-
ance at the visible maximum is expressed in terms of concentration by use
of appropriate absorptivities [449, 502, 516, 524-526]. Nilsson [527]
developed a spectrophotometric method that directly determines beta-
cyanin and betaxanthin pigments in beets without their initial separation.
The method is based on the observation that while vulgaxanthin-I absorbs
only at 476-478 nm, betanin exhibits a maximum at 535-540 nm but also
absorbs at the visible maximum of vulgaxanthin-1. Calculation of the ratio
of absorbance at 538 nm to that at 477 nm as a function of concentration
(i.e. dilution) leads to determination of the measurable concentration of
vulgaxanthin-1. Absorbance of the sample at 538 nm is due only to
betanin, permitting its direct determination. Results are expressed in terms
of total betacyanin and total betaxanthin concentrations since minor
constituent betalains also contribute to measured absorbances.
If interfering substances are present in the sample, for example due to
betalain degradation, the betalain pigments must first be purified. The
procedures generally used, usually paper or column chromatography [504]
or electrophoresis [528], not only purify but also separate pigments,
thereby allowing for quantitation of total and individual betalains simultan-

eously. Due to its speed and high resolving power, and the fact that
preliminary purification is not always necessary, HPLC is the method of
choice for quantitative analysis of individual and total betalains [437, 480,
491,507,510,511].
8.5.6 Current and potential sources and uses

The use of betalain pigments as food colorants dates back to at least the
turn of the century when juice from pokeberries, which contain betanin,
was added to wine to impart a more desirable red colour [529]. With
passing of legislation against the adulteration of foods, this practice was
eventually prohibited. Current legislation restricts betalain colorants to
concentrates or powders obtained from aqueous extracts of red beets (Beta
vulgaris). Because of the relatively high pigment concentrations therein,
red beets are the only plant food source from which extraction of betalains
for commercial use as a colorant is economically feasible. Commercial beet
colorants typically contain 0.4-1.0% pigment (expressed as betanin), 80%
sugar, 8% ash and 10% protein [10]. Their colour can vary considerably,
depending on the content of yellow betaxanthins; thus their colour varies
with beet variety, beet root quality and age at harvest, and method of
pigment extraction.
Improvements in the pigment content of table beets have been reported
through traditional breeding procedures [530]. Selection of high pigment
yielding populations was attributed to high realized heritability for total
pigment; selection may have increased the frequency of a dominant allele
of the multiple allelic series at the R locus that is associated with pigment
concentration [531]. Increased pigment content of commercial beet color-
ants could also be achieved if legislative restrictions on the use of purified
pigments were lifted. Purification of beet extracts can be carried out via
industrial-scale chromatography. Although yields have not generally been
good to date, purification of crude extracts by ion-exchange and gel-
filtration chromatography has been proven successful on a laboratory scale
[532]. Further developments could make chromatographic systems very
useful for large-scale production of purified betalain colorants. Purification
of red beet extracts has also been carried out by fermentation using
Aspergillus niger [533, 534], Candida utilis [535, 536] and Saccharomyces
oviformis [511], resulting in products with improved stability and no off-
flavours, (i.e. geosmin is removed in the fermentation process). The
practical application of microbial purification for large-scale preparation of
commercial betalain colorants is uncertain.
Similarly, the accumulation of betalains in beetroot and other cell
suspension and tissue cultures has been reported, but the practical appli-
cation of this or other biotechnologies to the production of highly purified
betalain preparations for use as food colorants has still yet to be

demonstrated [537]. In vitro cell suspension or tissue cultures can be more
efficient and yield greater levels of pigment than whole plant cultivation.
The in vitro production system has several advantages over conventional
beet cropping and extraction for commercial betalain production [538]:
1. High culture growth rates allow fermentations to be carried out in the

time required for a single beetroot crop to be raised.
2. The compound responsible for the characteristic beetroot odour (i.e.
geosmin) is not formed.
3. Growth conditions can be established that eliminate the requirement
for secondary fermentations that may otherwise be necessary for
purification.
4. One can be selective with respect to the nature of the betalain
pigment(s) produced.
Yet, conventional red beet extract production is still less expensive than
comparable biotechnological processes, or than chemical syntheses that
tend to be plagued by low product yields. Further increases in pigment
yields are deemed necessary for biotechnological processes to become
economically viable [538]. Alternatively, world market demands for these
natural colorants in more highly purified forms than are currently available
from conventional processes could render in vitro processes more viable in
the future.
Commercial and purified betalain preparations have been shown to be
effective colorants of foods with compatible chemical and/or physical
properties. Microbiologically purified beetroot preparations have demon-
strated suitable stability for use in various pharmaceutical forms [539-541].
Beetroot colorants may be used alone to produce hues resembling
raspberry or cherry, or in combination with other colorants such as annatto
to obtain strawberry shades [10, 436]. Considering the factors that
influence pigment stability beet root preparations may be successfully used
to colour products with short shelf-lives that are marketed in a dry state,
that are packaged to reduce exposure to light, oxygen and humidity, and/or
that do not receive high or prolonged heat treatment. If thermal processing
is required pigment degradation can be minimized by adding the colorant
after the heat treatment or as near the end of the heating cycle as possible.
As reviewed by Counsell et al. [436] and von Elbe [542], beetroot colorant
can be effectively used to colour hard candies and fruit chews, dairy
products such as yoghurt and ice cream [543], salad dressing, frostings,
cake mixes, gelatin desserts, starch-based puddings [544], meat substitutes
[446], poultry meat sausages [545, 546], gravy mixes, soft drinks and
powdered drink mixes [547].
8.6 Conclusions
The successful use of anthocyanins and betalains (especially betacyanins)

to colour a variety of foods and pharmaceuticals has been clearly
demonstrated. Although these pigments may not replace the more stable
synthetic red dyes, they do provide suitable natural alternatives for the
production of highly coloured and high-quality products. Their successful
application is obviously limited to products with compatible physicochem-
ical properties. However, if current legislation restricting the use of more
highly purified extracts and sources other than grapes (and grape by-
products) and beets were lifted, wider application of anthocyanins and
betalains as food colorants could be realized.
At present, the anthocyanins would appear to be more stable than the
betacyanins. However, systematic research to investigate the stabilizing
influence of acylation, various condensation reactions and/or surface
adsorption has not been carried out on the betalains. Polyacylated
anthocyanins, and those which partake in condensation reactions (e.g. self-
association and co-pigmentation) and/or are adsorbed onto surfaces,
demonstrate marked stability and offer great promise as stable colorants.
With further technological advances, for example in the development of
industry-scale chromatographic purifications and/or separations and appli-
cation of biotechnology (e.g. cell suspension or tissue culture) for large-
scale pigment production, discovery of new pigment sources and structural
variants with enhanced stability, and modification of foods to stabilize
constituent anthocyanins and betalains, the potential of these natural
pigments for wider application as colorants would seem to be unlimited. A
greater knowledge of the structural features necessary for enhanced
stability and colour properties is required, however. Further investigations
into the physicochemistry of anthocyanins and betalains should lead to
increased application of these natural colorants in food products.
Acknowledgements
The authors gratefully appreciate the help of Cathy Young and Allen Mao
in the preparation of the chapter for the first edition of this volume, and the
Department of Food Science, University of Guelph, for financial assist-
ance.
References
1. Timberlake. C.F. and Bridle. P. in Anthocyanins as Food Colors (P. Markakis. ed.).
Academic Press, New York (1982) Chapter 5.
2. Mazza, G. and Miniati, E. Anthocyanins in Fruits, Vegetables and Grains, CRC Press,
Boca Raton (1993).
3. Mabry, T.J. in Secondary Plant Products (E.A. Bell and B.V. Charlwood, eds),
Springer-Verlag, Berlin (1980) Chapter 11.
4. Steglich, W. and Strack, D. in The Alkaloids: Chemistry and Pharmacology, Vol. 39
(A. Brossi, ed.), Academic Press, New York (1990) Chapter 1.
5. Musso, H. Tetrahedron 35 (1979) 2843-2853.
6. Gill, M. and Steglich, W. Prog. Chern. Org. Nat. Prod. 51 (1987) 1-286.
7. Kimler, L., Mears, J., Mabry, T.J. and RosIer, H. Taxon. 19 (1970) 875-878.
8. Meggos, H.N. Food Technol. 38 (1984) 70-74.
9. Francis F.J. CRC Crit. Rev. Food Sci. Nutr. 28 (1989) 273-314.
10. Marmion, D.M. Handbook of u.s. Colorants for Foods, Drugs and Cosmetics. 2nd edn.
John Wiley & Sons, New York (1984) Chapter 8.
11. Collins, P. and Timberlake, C. Inti. Food Ingred. 6 (1993) 32-38.
12. Singleton, V.L. and Esau, P.L. Phenolic Substances in Grapes and Wines and Their
Significance, Academic Press, New York (1969).
13. Horowitz, R.M. and Gentili, B. 1. Agric. Food Chern. 17 (1969) 696-700.
14. Haveland-Smith, R.B. Mut. Res. 91 (1981) 285-290.
15. von Elbe, J.H. and Schwartz, S.J. Arch. Toxicol. 49 (1981) 93-98.
16. Constantion, L., Albasini, A., Rastelli, G. and Benvenuti, S. Planta Med. 58 (1992)
342-344.
17. Frankel, E.N. Kanner, J., German, J.B., Parks, E. and Kinsella, J.E. Lancet 341 (1993)
454-457.
18. Tamura, H. and Yamagami, A. 1. Agric. Food Chern. 42 (1994) 1612-1615.
19. Tsuda, T., Ohshima, K., Kawakishi, S. and Osawa, T. 1. Agric. Food Chern. 42 (1994)
248-251.
20. Vlaskovska, M., Drenska, D. and Ovcharov, R. Probl. Vutr. Med. 18 (1990) 3-19.
21. Drenska, D., Bantutova, I. and Ovcharov, R. Farmatsiya (Sofia) 39 (1989) 33-40.
22. Igarashi, K., Abe, S. and Satoh, J. Agric. Bioi. Chern. 54 (1990) 171-175.
23. Pourrat, A. Lejeune, B., Grand, A., Bastide, P. and Bastide, J. Med. et Nut. 23 (1987)
166-17.
24. Igarashi, K. and Inagaki, K. Agric. Bioi. Chern. 55 (1991) 285-287.
25. Karaivanova, M.H., Drenska, D. and Ovcharov, R. Comptes Rendus de I'Academie
Bulgare des Sciences 42 (1989) 119-21.
26. Markakis, P. (ed.) Anthocyanins as Food Colors. Academic Press, New York (1982).
27. Goto, T. Progr. Chern. Org. Nat. Prod. 52 (1987) 113-158.
28. Jackman, R.L., Yada, R.Y., Tung, M.A. and Speers, R.A. 1. Food Biochem. 11 (1987)
201-247.
29. Jackman, R.L., Yada, R.Y. and Tung, M.A. 1. Food Biochem. 11 (1987) 279-308.
30. Mazza, G. and Brouillard, R. Food Chern. 25 (1991) 17-33.
31. Goto, T. and Kondo, T. Angew Chern. Int. Ed. Engl. 30 (1991) 17-33.
32. Strack, D. and Wray, V. in The Flavonoids: Advances in Research Since 1986 (J.B.
Harborne, ed.), Chapman & Hall, London (1993) Chapter 1.
33. Piattelli, M. in The Biochemistry of Plants. Vol. 7. Secondary Plant Products (P.K.
Stumpf and E.E. Conn, eds), Academic Press, New York (1981) Chapter 18.
34. Liebisch, H.W. in Biochemistry of Alkaloids (K. Mothes, H.R. Schutte and M.
Luckner, eds), VEB Deutscher Verlag der Wissenschaften, Berlin (1985) Chapter 14.
35. Stafford, H.A. Plant Sci. 101 (1994) 91-98.
36. McClure, J.W. in The Flavonoids (J.B. Harborne, T.J. Mabry and H. Mabry, eds),
Chapman & Hall, London (1975) Chapter 18.
37. Sosnova, V. Bioi. Plant. 12 (1970) 424-427.
38. Kimler, L.M. Abstr. Bot. Soc. Am. 70th Annu. Meet. (1975) 36.
39. Stenlid, G. Phytochem. 15 (1976) 661-663.
40. Hrazdina, G. in The Flavonoids: Advances in Research (J.B. Harborne and T.J. Mabry,
eds), Chapman & Hall, London (1982) Chapter 3.
41a. Takahashi, A., Takeda, K. and Ohnishi, T. Plant Cell Physiol. 32 (1991) 541-547.
41b. Koes, R.E., Quattracchio, F. and Mol, J.N.M. BioEssays 16 (1994) 123-132.
42. Mabry, T.J. in Comparative Phytochemistry (T. Swain, ed.), Academic Press, London
(1966) Chapter 14.
43. Zakharova, N.S., Petrova, T.A. and Bokuchava, M.A. Soviet Plant Physiol. 36 (1989)
273-277.
44. Kumon, K., Sasaki, J., Sejimi, M., Takeuchi, Y. and Hayashi, Y. Plant Cell Physiol. 31
(1990) 233-240.
45. Harborne. J.B. in The Comparative Biochemistry of the F1avonoids, Academic Press,
New York (1967) Chapter 1.
46. Grisebach, H. in Anthocyanins as Food Colors (P. Markakis, ed.), Academic Press,
47. Kunz, S., Burkhardt, G. and Becker, H. Phytochem. 35 (1994) 233--235.
48. Sweeny, J.G. and Iacobucci, G.A. 1. Agric. Food Chem. 31 (1983) 531-533.
49. Arnie, D. and Trinajstic, N. 1. Chem. Soc. Perkin Trans. 2 (1991) 891-895.
50. Timberlake, C.F. and Bridle, P. 1. Sci. Food Agric. 18 (1967) 473--478.
51. Iacobucci, G.A. and Sweeny, J.G. Tetrahedron 39 (1983) 3005-3038.
52. Jurd, L. Adv. Food Res., Supp. 3 (1972) 123-142.
53. Brouillard, R. in Anthocyanins as Food Colors (P. Markakis, ed.), Academic Press,
54. Harborne, J.B. in Chemical Plant Taxonomy (T. Swain, ed.), Academic Press, New
York (1963) Chapter 13.
55. Starr, M.S. and Francis, F.J. Food Technol. 22 (1968) 1293--1295.
56. Martinelli, E.M., Scilingo, A. and Pifferi, G. Anal. Chim. Acta 259 (1992) 109-113.
57. Harborne, J.B. Phytochem. 3 (1964) 151-160.
58. Gueffroy, D.E., Kepner, R.E. and Webb, A.D. Phytochem. 10 (1971) 813-819.
59. Hrazdina, G. and Franzese, A.J. Phytochem. 13 (1974) 225-229.
60. Kondo, T., Suzuki, Yoshida, K., Oki, K., Veda, M., Isobe, M. and Goto, T.
Tetrahedron Lett. 32 (1991) 6375-6378.
61. Terahara, N., Kenjiro, T. and Honda, T. Z. Naturforsch. 48 (1993) 430-435.
62. Yoshitama, K., Kaneshige, M., Ishikura, N., Araki, F., Yahara, S. and Abe, K. 1. Plant
Res. 107 (1994) 209--214.
63. Kondo, T., Veda, M. and Goto, T. Tetrahedron 46 (1990) 4749--4756.
64. Terahara, N., Saito, N., Honda, T., Toki, K. and Osajima, Y. Tetrahedron Lett. 30
(1989) 5305-5308.
(1990) 2921-2924.
66. Terahara, N., Saito, N., Honda, T., Toki, K. and Osajima, Y. Heterocycles 31 (1990)
1773--1776.
67. Terahara, N., Saito, N., Honda, T., Toki, K. and Osajima, Y. Phytochem. 29 (1990)
949-953.
68. Kondo, T., Yamashiki, J., Kawahori, K. and Goto, T. Tetrahedron Lett. 30 (1989)
6055-6058.
69. Shi, Z., Daun, H. and Francis, F.J. 1. Food Sci. 58 (1993) 1068--1069.
70. Brouillard, R. Phytochem. 22 (1983) 1311-1323.
71. Brouillard, R. Phytochem. 20 (1981) 143--145.
72. Yoshitama, K. Bot. Mag. Tokyo 91 (1978) 207-212.
73. Dangles, 0., Saito, N. and Brouillard, R. Phytochem. 34 (1993) 119--124.
74. Yoshitama, K. and Hayashi, K. Bot. Mag. Tokyo 87 (1974) 33--40.
75. Lu, T.S., Saito, N., Yokoi, M., Shigihara, A. and Honda, T. Phytochem. 30 (1992) 289-
295.
76. Stewart, R.N., Norris, K.H. and Asen, S. Phytochem. 14 (1975) 937-942.
77. Asen, S. Acta Hortic. 63 (1976) 217-223.
78. Timberlake, C.F. and Bridle, P. in The Flavonoids (J.B. Harborne, T.J. Mabry and H.
Mabry, eds), Chapman & Hall, London (1975) Chapter 5.
79. Harborne, J.B. and Grayer, R.J. in The F1avonoids: Advances in Research Since 1980
(J.B. Harborne, ed.), Chapman & Hall, London (1988) Chapter 1.
80. Fuleki, T. 1. Food Sci. 36 (1971) 101-104.
81. Du, C.T., Wang, P.L. and Francis, F.J. 1. Food Sci. 39 (1974) 1265-1266.
82. Terahara, N., Yamaguchi, M. and Honda, T. Biosci. Biotech. Biochern. 58 (1994) 1324-
1325.
83. Tanchev, S.S. and Timberlake, C.F. Phytochern. 8 (1969) 1825-1827.
84. Hrazdina, G., Iredale, H. and Mattick, L.R. Phytochern. 16 (1977) 297-299.
85. Idaka, E., Suzuki, K., Yamakita, H., Ogawa, T., Kondo, T. and Goto, T. Chern. Lett.
(1987) 145-148.
86. Jdaka, E., Yamakita, H., Ogawa, T., Kondo, T., Yamamoto, M. and Goto, T. Chern.
Lett. (1987) 1213-1216.
87. Chandler, B.V. Nature 182 (1958) 933.
88. Wrolstad, R.E. and Heatherbell, D.A. l. Sci. Food Agric. 35 (1974) 1221-1228.
89. Rasper, V. and Coursey, D.C. Experientia 23 (1967) 611--612.
90. Imbert, M.P. and Seaforth, C. Experientia 24 (1968) 445-447.
91. Shoyama, Y., Nishioka, I., Kerath, W., Uemoto, S., Fujieda, K. and Okuba, H.
Phytochern. 29 (1990) 2999-3001.
92. Yoshida, K., Kondo, T., Kameda, K., Kawakishi, S., Lubag, A.J.M., Mendoza,
E.M.T. and Goto, T. Tetrahedron, Lett. 32 (1991) 5575-5578.
93. Ishikura, N. and Sugahara, K. Bot. Mag. Tokyo 92 (1979) 57--61.
94. Puech, A.A., Rebeiz, C.A., Catlin, P.B. and Crane, J.C. l. Food Sci. 40 (1975) 775-
779.
95. Co, H. and Markakis, P. l. Food Sci. 33 (1968) 281-283.
96. Wrolstad, R.E., Hildrum, K.1. and Amos, J.F. l. Chrornatogr. 50 (1970) 311-318.
97. Ishikura, N. Bot. Mag. Tokyo 88 (1975) 41-45.
98. Yoshikura, K. and Hamaguchi, Y. l. lap. Soc. Food Nutr. 22 (1969) 367-370.
99. Odake, T., Terahara, N., Saito, N., Toki, K. and Honda, T. Phytochern. 31 (1992)
2127-2130.
100. Shi, Z., Bassa, LA., Gabriel, S.L. and Francis, F.J. l. Food Sci. 57 (1992) 755-757, 770.
101. Lee, H.S. and Wicker, L. l. Food Sci. 56 (1991) 466-468, 483.
102. Sun, B.H. and Francis, F.J. l. Food Sci. 32 (1967) 647--649.
103. Timberlake, C.F. and Bridle, P. l. Sci. Food Agric. 22 (1971) 509-513.
104. Mazza, G. and Velioglu, Y.S. Food Chern. 43 (1992) 113-117.
105. Proctor, J.T. and Creasy, L.L. Phytochern. 8 (1969) 2108.
106. Pruthi, J.S., Susheela, R. and Lal, G. l. Food Sci. 26 (1961) 385-388.
107. Yoshikura, K. and Hamaguchi, Y. l. lap. Soc. Food Nutr. 24 (1971) 275-278.
108. Lynn, D.Y. and Luh, B.S. l. Food Sci. 29 (1964) 735-743.
109. Li, K. and Wagenknecht, A.C. Nature 182 (1958) 657.
110. Harborne, J.B. and Hall, E. Phytochern. 3 (1964) 453-463.
111. Dekazos, E.D. l. Food Sci. 35 (1970) 237-241.
112. Shrikhande, A.J. and Francis, F.J. 1. Food Sci. 38 (1973) 649--651.
113. Chandra, H., Nair, M.G. and Iezzoni, A. l. Agric. Food Chern. 40 (1992) 967-969.
114. Tanchev, S.S. and Vasilev, V.N. Gradinar. Lozar. Nauka 10 (1973) 23-28.
115. Du, c.T., Wang, P.L. and Francis, F.J. l. Food Sci. 40 (1975) 417-418.
116. Ishikura, N. and Hayashi, K. Bot. Mag. Tokyo 78 (1965) 91-96.
117. Fuleki, T. l. Food Sci. 34 (1969) 365-369.
118. Gallop, R.A. Variety, Composition and Colour in Canned Fruits, Particularly Rhubarb.
Fruit & Veg. Canning Res. Assoc., Chipping Campden, Glos. Sci. Bull. No.5 (1965).
119. Wrolstad, R.E. and Heatherbell, D.A. 1. Food Sci. 33 (1968) 592-594.
120. Chandler, B.V. and Harper, K.A. Austral. l. Chern. IS (1962) 114-120.
121. Casoli, U., Cultrera, R. and Gherardi, S. Ind. Conserve 42 (1967) 255.
122. Nybom, N. Frukt. Baer. (1970) 106-118.
123. Oeydin, J. Hortic. Res. 14 (1974) 1-7.
124. Barritt, B.H. and Torre, L.c. l. Chrornatogr. 75 (1973) 151-155.
125. Jennings, D.L. and Carmichael, E. New Phytol. 84 (1980) 505-513.
126. Nybom, N. l. Chrornatogr. 38 (1968) 382-387.
127. Daravingas, G. and Cain, R.F. l. Food Sci. 31 (1966) 927-936.
128. Daravingas, G. and Cain, R.F. l. Food Sci. 33 (1968) 138-142.
129. Watanabe, S., Sakamura, S. and Obata, Y. Agric. Bioi. Chern. 30 (1966) 420-422.
130. Tanchev, S.S., Ruskov, P.J. and Timberlake, C.F. Phytochern. 9 (1970) 1681-1682.
131. Saito, N., Hotta, R., Imai, K. and Hayashi, K. Proc. lap. Acad. 41 (1965) 593-598.
132. Francis, F.J. and Harborne, J.B. 1. Food Sci. 31 (1966) 524--528.
133. Buckmire, R.E. and Francis, F.J. 1. Food Sci. 41 (1976) 1363-1365.
134. Francis, F.J., Harborne, J.B. and Barker, W.G. 1. Food Sci. 31 (1966) 583-587.
135. Ballinger, W.E., Maness, E.P. and Kushman, L.J. 1. Am. Soc. Hort. Sci. 95 (1970) 283-
285.
136. Ballinger, W.E., Maness, E.P., Galletta, G.J. and Kushman, L.J. 1. Am. Soc. Hort.
Sci. 97 (1972) 381-384.
137. Zapsalis, C. and Francis, F.J. 1. Food Sci. 30 (1965) 396-399.
138. Fuleki, T. and Francis, F.J. Phytochem. 6 (1967) 1705-1708.
139. Fuleki, T. and Francis, F.J. 1. Food Sci. 33 (1968) 471-478.
140. Hicks, K.B., Sondey, S.M., Hargrave, D., Sapers, G.M. and Bilyk, A. LC Mag. 3
(1985) 981-982, 984.
141. Andersen, 0.M. 1. Food Sci. 54 (1989) 383-384, 387.
142. Ribereau-Gayon, P. in Anthocyanins as Food Colors (P. Markakis, ed.), Academic
Press, New York (1982) Chapter 8.
143. Harborne, J.B. and Gavazzi, G. Phytochem. 8 (1969) 999-1001.
144. Lawanson, A.O. and Osude, B.A. Z. PJianzenphysiol. 67 (1972) 460-463.
145. Nakatani, N., Fukuda, H. and Fuwa, H. Agric. BioI. Chem. 43 (1979) 389-391.
146. Lancaster, J.E., Grant, J.E., Lister, C.E. and Taylor, M.C. 1. Am. Soc. Hort. Sci. 119
(1994) 63-69.
147. Grisebach, H. in The Biochemistry of Plant Phenolics (C.F. van Sumere and P.J. Lea,
eds), Clarendon Press, Oxford (1985) Chapter 10.
148. Heller, W. and Forkmann, G. in The Flavonoids: Advances in Research Since 1980
(J.B. Harborne, ed.), Chapman & Hall, London (1988) Chapter 11.
149. Mol, J. Stuitje, A., Gerats, A., van der Krol, A. and Jorgensen, R. TIBTECH7 (1989)
148-153.
150. Forkmann, G. Plant Breeding 106 (1991) 1-26.
151. Holton, T.A. and Tanaka, Y. TlBTECH 12 (1994) 4042.
152. Stafford, H.A. Flavonoid Metabolism, CRC Press, Boca Raton (1990).
153. Dooner, H.K., Robbins, T.P. and Jorgensen, R.A. Annu. Rev. Genet. 25 (1991) 173-
199.
154. van der Meer, I.M., Stuitje, A.R. and Mol, J.N.M. in Control of Plant Gene Expression
(D.P.S. Verma, ed.), CRC Press, Boca Raton (1993).
155. Quattracchio, F., Wing, J.F., Leppen, H.T.C., Mol, J.N.M. and Koes, R.E. Plant Cell
5 (1993) 1497-1512.
156. Ludwig, S.R., Habera, L.F., Dellaporta, S.L. and Wessler, S.R. Proc. Natl. Acad. Sci.
USA 86 (1989) 7092-7096.
157. Radicella, J.P., Turks, D. and Chandler, V.L. Plant Mol. BioI. 17 (1991) 127-130.
158. Jain, V.K. and Guruprusad, K.N. Physiol. Plant. 75 (1989) 233-236.
159. Weiss, D., van Blokland, R., Kooter, J.M., Mol, J.N.M. and van Tunen, A.J. Plant
Physiol. 98 (1992) 191-197.
160. Hattori, T., Vasil, V., Rosenkrans, L., Hannah, L.C., McCarty, D.R. and Vasil, I.K.
Genes Dev. 6 (1992) 609-618.
161. Ilan, A., Zanewich, K.P., Rood, S.B. and Dooner, D.K. Physiol. Plant. 92 (1994) 47-
52.
162. Mancinelli, A.L., Rossi, F. and Moroni, A. Plant Physiol. 96 (1991) 1079-1085.
163. Jain, V.K. and Guruprusad, K.N. 1. Exp. Bot. 41 (1990) 53-58.
164. Khare, M. and Guruprusad, K.N. Plant Sci. 91 (1993) 1-5.
165. Galbiati, M., Chiusi, A., Peteriongo, P., Mancinelli, A. and Gavazzi, G. Maydica 39
(1994) 89-95.
166. Shichijo, C., Hamada, T., Hiraoka, M., Johnson, C.B. and Hashimoto, T. Planta 191
(1993) 283-245.
167. Egin-Buyler, B., Loyal, R. and Ebel, J. Arch. Biochem. Biophys. 203 (1980) 90-100.
168. Forkmann, G. and Stotz, G. Z. Naturforsch. 36c (1981) 411-416.
169. Stotz, G. and Forkmann, G. Z. Naturforsch. 37c (1982) 19-23.
170. Menting, J.G.T., Scopes, R.K. and Stevensen, T.W. Plant Physiol. 106 (1994) 633-642.
171. Heller, W., Britsch, L., Forkmann, G. and Grisebach, H. Planta 163 (1985) 191-196.
172. Stich, K. and Forkmann, G. Phytochem. 27 (1988) 785-789.
173. Stich, K. and Forkmann, G. Z. Naturforsch. 43c (1988) 311-314.

174. Timberlake, C.F. and Bridle, P. Nature 212 (1966) 158-159.
175. Teusch, M., Forkmann, G. and Syffert, W. Phytochem. 26 (1987) 991-994.
176. Ino,1. and Yamaguchi, M.-A. Phytochem. 33 (1993) 1415-1417.
177. Ino, 1., Nishiyama, H. and Yamaguchi, M.-A. Phytochem. 33 (1993) 1425-1426.
178. Glassgen, W.E. and Seitz, H.U. Planta 186 (1992) 582-585.
179. Hrazdina, G., Zobel, A.M. and Hoch, H.c. Proc. Natl. Acad. Sci. USA 84 (1988) 8966-
8970.
180. Hrazdina, G. and Wagner, G.J. in The Biochemistry of Plant Phenolics (C.F. van
Sumere and P.J. Lea, eds), Clarendon Press, Oxford (1985) Chapter 7.
181. Hrazdina, G. and Wagner, G.J. Arch. Biochem. Biophys. 237 (1985) 88-100.
182. Jonsson, L.M., Donker-Koopman, W.E., Uitslager, P. and Schram, A.W. Plant
Physiol. 72 (1983) 287-290.
183. Joanny, K. and Reddy, A.R. Maydica 37 (1992) 165-167.
184. Hopp, W. and Seitz, H.V. Planta 170 (1987) 74-85.
185. Seitz, H.U. and Hinderer, W. in Cell Culture and Somatic Cell Genetics of Plants, Vol.5
(F. Constabel and 1.K. Vasil, eds), Academic Press, New York (1988) Chapter 3.
186. Nozzolillo, C. and Ishikura, N. Plant Cell Rep. 7 (1988) 389-392.
187. Nozue, M., Kubo, H., Nishimura, M., Katou, A., Hattori, C., Usuda, N., Nagata, T.
and Yasuda, H. Plant Cell Physiol. 34 (1993) 803-808.
188. Cormier, F., Crevier, H. and Do, C.B. Can. 1. Bot. 68 (1990) 1822-1825.
189. Do, C.B. and Cormier, F. Plant Cell Rep. 9 (1991) 500-504.
190. Hirasuna, T.J., Shuler, M.L., Lackney, V.K. and Spanswick, R.M. Plant Sci. 78 (1991)
107-120.
191. Mazza, G. and Brouillard, R.I. Agric. Food Chem. 35 (1987) 422-426.
192. Brouillard, R. and Lang, J. Can. 1. Chem. 68 (1990) 755-761.
193. Baranac, J., Amic, D. and Vukadinovic, V.I. Agric. Food Chem. 38 (1990) 932-936.
194. McClelland, R.A. and McGall, G.H. 1. Org. Chem. 47 (1982) 3730-3736.
195. Amic. D., Baranac, J. and Vukadinovic, V.I. Agric. Food. Chem. 38 (1990) 936-940.
196. Brouillard, R. and Dubois, J.E. 1. Am. Chem. Soc. 99 (1977) 1359-1364.
197. Dangles, O. and Elhajji, N. Helv. Chim. Acta 77 (1994) 1595-1610.
198. Francis, F.J. Trends Food Sci. Technol. 3 (1992) 27-30.
199. Hrazdina, G., Borzell, A.J. and Robinson, W.B. Am. 1. Enol. Vitic. 21 (1970) 201-204.
200. Robinson, W.D., Weiers, L.D., Bertino, J.J. and Mattick, L.R. Am. 1. Enol. Vitic. 17
(1966) 178-183.
201. Van Buren, J.P., Bertino, J.J. and Robinson, W.B. Am. f. Enol. Vitic. 19 (1968) 147-
154.
202. Adams, J.B. Changes in the Polyphenols of Red Fruits During Processing-The Kinetics
and Mechanism of Anthocyanin Degradation. Campden Food Pres. Res. Assoc.,
Chipping Campden, Glos. Tech. Bull. P. 2. (1972).
203. Adams, J.B. f. Sci. Food Agric. 24 (1973) 747-762.
204. Lukton, A., Chichester, C.O. and MacKinney, G. Food Techno!. 10 (1956) 427-432.
205. Meschter, E.E. f. Agric. Food Chem. 1 (1953) 574-579.
206. Tinsley, 1.1. and Bockian, A.H. Food Res. 25 (1960) 161-173.
207. Calvi, J.P. and Francis, F.J. 1. Food Sci. 43 (1978) 1448-1456.
208. Brouillard, R. and Delaporte, B. 1. Am. Chem. Soc. 99 (1977) 8461-8468.
209. Hrazdina, G. Phytochem. 10 (1971) 1125-1130.
210. Markakis, P., Livingston, G.E. and Fellers, C.R. Food Res. 22 (1957) 117-129.
211. Adams, J.B. Food Manuf. (1973) (Feb.) 19-20,41.
212. Furtado, P., Figueiredo, P., Chaves des Neves, C. and Pina, F. f. Photochem.
Photobiol. A: Chem. 75 (1993) 113-118.
213. Piffaut, B., Kader, F., Girardin, M. and Metche, M. Food Chem. 50 (1994) 115-120.
214. Macarone, E., Ferrigno, V., Longo, L.L. and Rapisarda, P. Annal. Chim. 77 (1987)
499-508.
215. Tanchev, S.S. and Ioncheva, N. Nahrung 20 (1976) 889-893.
216. Markakis, P. in Anthocyanins as Food Colors (P. Markakis, ed.), Academic Press, New
217. Erlandson, J.A. and Wrolstad, R.E. f. Food Sci. 37 (1~·72) 592-595.
218. Abers, J.E. and Wrolstad, R.E. J. Food Sci. 44 (1979) 75-78, 81.
219. Tanchev, S. Proc. 6th Int. Congr. Food Sci. Techno!. (1983) 96.
220. Mishkin, M. and Saguy, I. Z. Lebensrn. u. Forsch. 175 (1982) 410-412.
221. Havlikova, L. and Mikova, K. Z. Lebensrn. u. Forsch. 181 (1985) 427-432.
222. Cemeroglu, B., Velioglu, S. and Isik, S. J. Food Sci. 59 (1994) 1216-1218.
223. Somers, T.C. Phytochern. 10 (1971) 2175-2186.
224. Adams, J.B. and Ongley, M.J. J. Food Technol. 8 (1973) 139-145.
225. McCloskey, L.P. and Yengoyan, L.S. Am. J. Eno!. Vitic. 32 (1981) 257-261.
226. Bakker, J. and Timberlake, C.F. J. Sci. Food Agric. 37 (1986) 288-292.
227. Rommel, A., Wrolstad, R.E. and Heatherbell, D.A. J. Food Sci. 57 (1992) 385-391,
410.
228. Peng, c.Y. and Markakis, P. Nature 199 (1963) 597-598.
229. Keith, E.S. and Powers, J.J. J. Agric. Food Chern. 13 (1965) 577-579.
230. Sondheimer, E. and Kertesz, A.I. Food Res. 18 (1953) 475-479.
231. Sondheimer, E. and Kertesz, A.I. Food Res. 18 (1952) 288-297.
232. Hrazdina, G. Phytochern. 9 (1970) 1647-1652.
233. Yoshida, K., Kondo, T., Kameda, K. and Goto, T. Agric. Bioi. Chern. 54 (1990) 1745-
1751.
234. Palamidis, N. and Markakis, P. J. Food Sci. 40 (1975) 1047-1049.
235. Markakis, P., Livingston, G.E. and Fagerson, J.S. Food Res. 24 (1959) 520-528.
236. Pellegrino, F., Sekular, P. and Alfano, R.R. Photobiochern. Photobiophys. 2 (1981)
15-20.
237. Santhanam, M., Hautala, R.R., Sweeny, J.G. and Iacobucci, G.A. Photochern.
Photobiol. 38 (1984) 477-480.
238. Sweeny, J.G., Wilkinson, M.M and Iacobucci, G.A. J. Agric. Food Chern. 29 (1981)
563-567.
239. Huang, H.T. J. Agric. Food Chern. 3 (1955) 141-146.
240. Forsyth, W.G.c. and Quesnel, V.C. Biochern. J. 65 (1957) 177-179.
241. Blom, H. Food Chern. 12 (1983) 197-204.
242. van Buren, J.P., Scheiner, D.M. and Wagenknecht, A.C. Nature 185 (1960) 165-166.
243. Wagenknecht, A.C., Scheiner, D.M. and van Buren, J.P. Food Technol. 14 (1960) 47-49.
244. Skalski, C. and Sistrunk, W.A. J. Food Sci. 38 (1973) 1060-1062.
245. Pifferi, P.G. and Cultrera, R. J. Food Sci. 39 (1974) 786-791.
246. Wesche-Ebeling, P. and Montgomery, M.W. J. Food Sci. 55 (1990) 731-734, 745.
247. Oszmianski, J. and Lee, C.Y. J. Agric. Food Chern. 39 (1991) 1050-1052.
248. Siddiq, M., Arnold, J.F., Sinha, N.K. and Cash, J.N. J. Food Process. Pres. 18 (1994)
75-84.
249. Grommeck, R. and Markakis, P. J. Food Sci. 29 (1964) 53-57.
250. Matthew, A.G. and Parpia, H.A.B. Adv. Food Res. 19 (1971) 75-145.
251. Sakabura, H. and Ichinose, H. Agric. Chern. Soc. Jap. 56 (1982) 517-524.
252. Jiang, J., Paterson, A. and Piggott, J.R. Inti. J. Food Sci. Technol. 25 (1990) 596-600.
253. Tanchev, S., Vladimirov, G. and Ioncheva, N. Nauchni Trudove, Vissh Inst. Khranit.
Vkusova Prornyshl. 16 (1969) 77-82.
254. Yang, H.Y. and Steele, W.F. Food Technol. 12 (1958) 517-519.
255. Siegel, A., Markakis, P. and Bedford, c.L. J. Food Sci. 36 (1971) 962-963.
256. Wrolstad, R.E., Skrede, G., Lea, P. and Enersen, G. J. Food Sci. 55 (1990) 1064-1065,
1072.
257. Blom, H. and Thomassen, M.S. Food Chern. 17 (1985) 157-168.
258. Goodman, L.P. and Markakis, P. J. Food Sci. 30 (1965) 135-137.
259. Cash, J.N., Sistrunk, W.A. and Stotte, C.A. J. Food Sci. 41 (1976) 1398-1402.
260. Ingold, C.K. Structure and Mechanism in Organic Chemistry. Cornell University Press,
Ithica (1969).
261. Bendz, G., Martensson, O. and Nilsson, E. Ark. Kerni 27 (1967) 65-77.
262. Shriner, R.L. and Sutton, R. J. Am. Chern. Soc. 85 (1963) 3989-3991.
263. Jurd, L. Tetrahedron 21 (1965) 3707-3714.
264. Jurd, L. Tetrahedron 23 (1967) 1057-1064.
265. Jurd, L. J. Heterocycl. Chern. 18 (19981) 429-430.
266. Jurd, L. and Waiss, A.C. Jr. Tetrahedron 21 (1965) 1471-1483.
267. Jurd, L. J. Food Sci. 29 (1964) 16-19.

268. Timberlake, C.F. and Bridle, P. J. Sci. Food Agric. 18 (1967) 479-485.
269. Brouillard, R. and El Hage Chahine, J.M. Bull. Liaison-Groupe Polyphenols 9 (1980)
77-78.
270. Brouillard, R. and El Hage Chahine, J.M. Am. Chem. Soc. 102 (1980) 5375-5378.
271. Timberlake, C.F. and Bridle, P. Chem. Ind. Oct. (1968) 1489.
272. Bridle, P., Scott, K.G. and Timberlake, C.F. Phytochem. 12 (1973) 1103-1106.
273. Timberlake, C.F. and Bridle, P. J. Sci. Food Agric. 28 (1977) 539-544.
274. Bakker, J., Picinelli, A. and Bridle, P. Vitis 32 (1993) 111-118.
275. Garcia-Viguera, C., Bridle, P. and Bakker, J. Vitis 33 (1994) 37-40.
276. Esselen, W.B. and Sammy, G.M. Food Prod. Dev. 9(8) (1975) 37-38, 40.
277. Main, J.H., Clydesdale, F.M. and Francis, F.J. J. Food Sci. 43 (1978) 1693-1694, 1697.
278. Bronnum-Hansen, K. and Flink, J.M. J. Food Technol. 20 (1985) 725-733.
279. MacKinney, G., Lukton, A. and Chichester, C.O. Food Technol. 9 (1955) 324-326.
280. Debicki-Pospisil, J., Lovric, T., Trinajstic, N. and Sabljic, A. J. Food Sci. 48 (1983)
411-416.
281. Hodge, J.E. J. Agric. Food Chem. 1 (1953) 928-943.
282. Kurata, T. and Sakurai, Y. Agric. Bioi. Chem. 31 (1967) 170-176.
283. Sloan, J.L., Bills, D.D. and Libbey, L.M. J. Agric. Food Chem. 17 (1969) 1370-1372.
284. Harborne, J.B. Recent Adv. Phytochem. 12 (1979) 457-474.
285. Asen, S., Stewart, R.N., Norris, K.H. and Massie, D.R. Phytochem. 9 (1970) 619-627.
286. Bayer, F., Fink, A., Nether, K. and Wegmann, K. Angew Chem. Inter. Ed. Engl. 5
(1966) 791-798.
287. Lowry, J.B. and Chew, L. Econ. Bot. 28 (1974) 61-62.
288. Osawa, Y. in Anthocyanins as Food Colors (P. Markakis, ed.), Academic Press, New
289. Asen, S., Stewart, R.N. and Norris, K.H. Phytochem. 11 (1972) 1139-1144.
290. Hoshino, T., Matsumoto, U. and Goto, T. Phytochem. 19 (1980) 663-667.
291. Mistry, T.V., Cai, Y., Lilley, T.H. and Haslam, E. J. Chem. Soc., Perkin Trans. 2
(1991) 1287-1296.
292. Brouillard, R., Mazza, G., Saad, Z., Albrecht-Gray, A.M. and Cheminat, A. J. Am.
Chem. Soc. 111 (1989) 2604-2610.
293. Mazza, G. and Brouillard, R. Phytochem. 29 (1990) 1097-1102.
294. Davies, A.J. and Mazza, G. J. Agric. Food Chem. 41 (1993) 716-720.
295. Goto, T., Hoshino, T. and Takase, S. Tetrahedron Lett. 20 (1979) 2905-2908.
296. Wigand, M.C., Dangles, O. and Brouillard, R. Phytochem. 31 (1992) 4317-4324.
297. Dangles, 0., Wigand, M.C. and Brouillard, R. Phytochem. 31 (1992) 3811-3812.
298. Dangles, O. and Brouillard, R. J. Chem. Soc., Perkin Trans. 2 (1992) 247-257.
301. Schewffeldt, P. and Hrazdina, G. J. Food Sci. 43 (1978) 517-520.
302. Nakayama, T.O.M. and Powers, J.J. Adv. Food Res., Supp. 3 (1972) 193-199.
303. Hoshino, T., Matsumoto, U. and Goto, T. Phytochem. 20 (1981) 1971-1976.
304. Hoshino, T., Matsumoto, U., Harada, N. and Goto, T. Tetrahedron Lett. 22 (1981)
3621-3624.
305. Figueiredo, P. and Pina, F. J. Chem. Soc., Perkin Trans. 2 (1994) 775-778.
306. Hoshino, T. Phytochem. 30 (1991) 2049-2055.
308. Singleton, V.L. and Trousdale, E.K. Am. J. Enol. Vitic. 43 (1992) 63-70.
309. Jurd, L. and Asen, S. Phytochem. 5 (1966) 1263-1271.
310. Francis, F.J. in Anthocyanins as Food Colors (P. Markakis, ed.), Academic Press, New
311. Harborne, J.B. Phytochemical Methods. A Guide to Modern Techniques of Plant
Analysis, 2nd edn. Chapman & Hall, London (1984).
312. Strack, D. and Wray, V. in Methods in Plant Biochemistry, VoU (J.B. Harborne, ed.),
Academic Press, London (1989).
313. Anderson, D.W., Julian, E.A., Kepner, R.E. and Webb, A.D. Phytochem. 9 (1970)
1569-1578.
314. Moore, A.B., Francis, F.J. and Clydesdale, F.M. 1. Food Protect. 45 (1982) 738-743.
315. Moore, A.B., Francis, F.J. and Jason, M.E. J. Food Protect. 45 (1982) 590-593.
316. Esselen, W.B. and Sammy, G.M. Food Prod. Dev. 7 (1973) 80,82,86.
317. Philip, T. J. Food Sci. 39 (1974) 859.
318. Langston, M.S.K. US Patent 4 500 556 (1985).
319. Pouget, M.P., Lejeune, B., Vennat, B. and Pourrat, A. Lebensm.-Wiss. u. Technol.23
(1990) 103-105.
320. Dallas, C. and Laureano, O. Vitis 33 (1994) 41-47.
321. Weisenborn, D., Glowacki, A., Hettiarchchy, N. and Zander, L. Trans. ASAE 36
(1993) 1411-1416.
322. Shewfelt, R.L. and Ahmed. E.M. Food Prod. Dev. 11(4) (1977) 52, 57-59, 62.
323. Shewfelt, R.L. and Ahmed, E.M. J. Food Sci. 43 (1978) 435-438.
324. Hrazdina, G. J. Agric. Food Chem. 18 (1970) 243-245.
325. Wrolstad, R.E. and Struthers, B.J. J. Chromatogr. 55 (1971) 405-408.
326. van Teeling, e.G., Cansfield, P.E. and Gallop, R.A. J. Chromatogr. Sci. 9 (1971) 505-
509.
327. Strack, D. and Mansell, R.L. J. Chromatogr. 109 (1975) 325-331.
328. Hrazdina, G. Lebensm.-Wiss. u. Technol. 8 (1975) 111-113.
329. Manley, C.H. and Shubiak, P. Can. Inst. Food Sci. Technol. 8 (1975) 35-39.
330. Davies, A.J. and Mazza, G. J. Agric. Food Chem. 40 (1992) 1341-1345.
331. Kim, J.H., Nonaka, G.I., Fujieda, K. and Uemoto, S. Phytochem. 28 (1989) 1503.
332. Oszmianski, J. and Sapis, J.e. J. Food Sci. 53 (1988) 1241-1242.
333. Fuleki, T. and Francis, F.J. J. Food Sci. 33 (1968) 266-274.
334. Chiriboga, C. and Francis, F.J. J. Amer. Soc. Hort. Sci. 95 (1970) 233-236.
335. Goldstein, G. J. Chromatogr. 129 (1976) 466-468.
336. Spagna, G. and Pifferi, P.G. Food Chem. 44 (1992) 185-188.
337. Francis, G.W. and Andersen, 0.M. J. Chromatogr. 283 (1984) 445-448.
338. Andersen, 0.M. J. Food Sci. 50 (1985) 1230-1232.
339. Andersen, 0.M. J. Food Sci. 52 (1987) 665--666, 680.
340. Asen, S. J. Chromatogr. 18 (1965) 602--603.
341. Quarmby, C. J. Chromatogr. 34 (1968) 52-58.
342. Barritt, B.H. and Torre, L.e. J. Chromatogr. 75 (1973) 151-155.
343. Andersen, 0.M. and Francis, G.W. J. Chromatogr. 318 (1985) 450-455.
344. Lasagabaster, A., Martin, e. and Goni, M.M. J. Chem. Tech. Biotech. 60 (1994) 397-
403.
345. Francis, F.J. HortSci. 2 (1967) 170-171.
346. Wulf, L.W. and Nagel, e.W. Am. J. Enol. Vitic. 29 (1978) 42-49.
347. Bakker, J. and Timberlake, e.F. J. Sci. Food Agric. 36 (1985) 1315-1324.
348. Sapers, G.M., Hicks, K.B., Burgher, A.M., Hargrave, D.L., Son dey, S.M. and Bilyk,
A. J. Am. Soc. Hort. Sci. 111 (1986) 945-950.
349. Mazza, G. J. Food Sci. 51 (1986) 1260-1264.
350. Baj, A., Bombardelli, E., Gabetta, B. and Martinelli, E.M. J. Chromatogr. 279 (1983)
265-372.
351. Hong, V. and Wrolstad, R.E. J. Agric. Food Chem. 38 (1990) 698-708.
352. Hong, V. and Wrolstad, R.E. J. Agric. Food Chem. 38 (1990) 708-715.
353. Drdak, M., DauCik, P., Simko, P., Karovicova, J. and Rajniakova, A. Nahrung 36
(1992) 411-413.
354. Drdak, M., DauCik, P., Karovicova, J., Simko, P. Rajniakova, A. and Morova, E.
Nahrung 36 (1992) 411-413.
355. Lamuela-Raventos, R.M. and Waterhouse, A.L. Am. J. Enol. Vitic. 45 (1994) 1-5.
356. Lee, H.S. and Hong, V.J. J. Chromatogr. 624 (1992) 221-234.
357. Ribereau-Gayon, P.R. and Josien, M.L. Bull. Soc. Chim. Fr. (1960) 934-937.
358. Statoua, A., Merlin, J.e., Del Haye, M. and Brouillard, R. Raman Spectrosc., Proc.
Int. Conf 8th (1982) 629--630.
359. Merlin, J.-e., Statoua, A., Cornard, J.-P., Saidi-Idrissi, M. and Brouillard, R.
Phytochem. 35 (1993) 227-232.
360. Harborne, J.B. Biochem. J. 70 (1958) 22-28.
362. Hoshino, T., Matsumoto, U., Goto, T. and Harada, N. Tetrahedron Lett. 23 (1982)
433-436.
363. Goto, T., Takase, S. and Kondo, T. Tetrahederon Lett. 19 (1978) 2413-2416.
364. Goto, T., Kondo, T., Imagawa, H. and Miura, I. Tetrahedron Lett. 22 (1981) 3213-
3216.
365. Goto, T., Kondo, T., Tamura, H., Imagawa, H., Iiono, A. and Takeda, T. Tetrahedron
Lett. 23 (1982) 3695-3698.
366. Timberlake,.C.F., Bakker, J., Bridle, P., Loeffler, R.S.T., Saito, N. and Self, R. Bull.
Liaison-groupe Polyphenols 13 (1986) 96-101.
367. Andersen, 0.M., Aksnes, D.W., Nerdal, W. and Johansen, O. Phytochem. Anal. 2
(1991) 175-183.
368. Nerdal, W., Pedersen, A.T. and Andersen, 0.M. Acta Chem. Scand. 46 (1992) 872-
876.
369. Tamura, H., Hayashi, Y., Sugisawa, H. and Kondo, T. Phytochem. Anal. 5 (1994) 190-
196.
370. Love, G.D., Snape, e.E., Jarvis, M.C. and Morrison, I.M. Phytochem. 35 (1994) 489-
491.
371. Comus, G., Wyler, H. and Lauterwein, J. Phytochem. 20 (1981) 1461-1462.
372. Takura, H., Kumaoka, Y. and Sugisawa, H. Agric. Bioi. Chem. 53 (1989) 1969-1970.
373. Saito, N., Timberlake, C.F., Tucknott, O.G. and Lewis, I.A.S. Phytochem. 22 (1983)
1007-1009.
374. Glassgen, W.E., Seitz, H.U. and Metzger, J.W. Bioi. Mass Spectrom. 21 (1992) 271-
277.
375. Glassgen, W.E., Wray, Y., Strack, D., Metzger, J.W. and Seitz, H.U. Phytochem. 31
(1992) 1593-1601.
376. Pedersen, A.T., Andersen, 0.M., Aksnes, D.W. and Nerdal, W. Magn. Reson. Chem.
31 (1993) 972-976.
377. Wolfender, J.-L. Maillard, M. and Hostettmann, K. Phytochem. Anal. 5 (1994) 153-
182.
378. Gao, L. and Mazza, G. J. Agric. Food Chem. 42 (1994) 118-125.
379. Baublis, A.J. and Berber-Jimenez, M.D. J. Agric. Food Chem. 43 (1995) 640-646.
380. Geissman, T.A., Jorgenson, E.C. and Harborne, J.B. Chem. Ind. 1953) (Dec.) 1389.
381. Lin, M., Shi, Z. and Francis, F.J. J. Food Sci. 57 (1992) 766-767.
382. Harborne, J.B. J. Chromatogr. 1 (1958) 473-488.
383. Werner, D.J., Maness, E.P. and Ballinger, W.E. HortSci. 24 (1989) 488-489.
384. Hebrero, E., Santos-Buelga, e. and Rivas-Gonzalo, J.C. Am. J. Enol. Vitic. 39 (1988)
227-233.
385. Hebrero, E., Garcia-Rodriguez, C., Santos-Buelga, e. and Rivas-Gonzalo, J.e. Am. J.
Enol. Vitic. 40 (1989) 283-291.
386. Lees, D.H. and Francis, F.J. J. Food Sci. 36 (1971) 1056-1060.
387. BIom. H. Rev. Food Sci. Technol. 7 (1983) 587-589.
388. Chandler, B.Y. and Harper, K.A. Aust. J. Chem. 14 (1961) 586-595.
389. Wrolstad, R.E. Color and Pigment Analyses in Fruit Products. Bull. 624. Oregon Agric.
Exp. St., Corvallis (1976).
390. Staples, L.C. and Francis. F.J. Food Technol. 22 (1968) 611--614.
391. Jurd, L. J. Org. Chem. 28 (1963) 987-991.
392. Ponting, J.D., Sanshuck, W.D. and Brekke, J.E. Food Res. 25 (1960) 471-478.
393. Niketic-Aleksic, G.K. and Hrazdina, G. Lebensm.-Wiss. u. Techno!. 5 (1972) 163-165.
394. Dickinson, D. and Gawler, J.H. J. Sci. Food Agric. 7 (1956) 699-705.
395. Somers, T.e. and Evans, M.E. J. Sci. Food Agric. 25 (1974) 1369-1379.
396. Swain, T. and Hillis, W.E. J. Sci. Food Agric. 10 (1959) 63--68.
397. Fuleki, T. and Francis, F.J. J. Food Sci. 33 (1968) 78-83.
398. Tanchev, S. Z. Lebensm. U. Forsch. 150 (1972) 28-30.
399. Tanchev. S. Nahrung 18 (1974) 303-308.
400. Lees, D.H. and Francis, F.J. HortSci. 7 (1972) 83-84.
401. Godek, S. Przemysl Fermentacyjny i Owocowo-Warzywyny 25(5/6) (1981) 27-30.
402. Speers, R.A., Tung, M.A. and Jackman, R.L. Can. Inst. Food Sci. Technol. J. 20
(1987) 15-18.
403. Spayd, S.E. and Morris, J.R. 1. Food Sci. 46 (1981) 414--418.
404. Staples, L.c. and Francis, F.J. Food Technol. 22 (1968) 611-614.
405. van Buren, J.P., Hrazdina, G. and Robinson, W.B. J. Food Sci. 39 (1974) 325-328.
406. Torre, L.c. and Barritt, B.H. J. Food Sci. 42 (1977) 488-490.
407. Jurd, L. Food Technol. 18 (1964) 559-561.
408. Hrazdina, G. Z. Lebensm.-Wiss. u. Technol. 14 (1981) 283-286.
409. Chiriboga, C. and Francis, F.J. J. Food Sci. 38 (1973) 464--467.
410. Clydesdale, F.M., Main, J.H. and Francis, F.J. J. Food Protect. 42 (1979) 196-201.
411. Clydesdale, F.M., Main, J.H. and Francis, F.J. J. Food Protect. 42 (1979) 204-207.
412. Buckmire, R.E. and Francis, F.J. J. Food Sci. 43 (1978) 908-911.
413. Francis, F.J. Food Technol. 29 (1975) 52, 54.
414. Anon. Food Chem. 5 (1975) 69-80.
415. Martinez, S.B. and Del Valle, M.J. UP Home Econ. J. 9 (1981) 7-10.
416. Coudonis, M., Katsaboxakis, K. and Papanicolaou, D. Rev. Food Sci. Technol. 7 (1983)
567-572.
417. Asen, S., Stewart, R.N. and Norris, K.H. Phytochem. 16 (1977) 1118-1119.
418. Asen, S., Stewart, R.N. and Norris, K.H. US Patent No.4 172 902 (1979).
419. Baker, c.H., Johnston, M.R. and Barber, W.D. Food Prod. Dev. 8 (1974) (3) 83-84,
86-87.
420. Baker, C.H., Johnston, M.R. and Barber, W.D. Food Prod. Dev. 8 (1974) (4) 65-70.
421. Kondo, T., Yoshida, K., Nakagawa, A., Kawai, T., Tamura, H. and Goto, T. Nature
358 (1992) 515-518.
422. Kondo, T., Veda, M., Tamura, H., Yoshida, K., Isobe, M. and Goto, T. Angew.
Chem. Int. Ed. Engl. 33 (1994) 978-979.
423. Anon, Enocolor. The Importance of the Natural Colour. Reggiana Antociani, Industria
Coloranti Naturali, Italy (1981).
424. Taylor, A.J. in Developments in Food Colours-2 (J. Walford, ed.), Applied Science
Pub., London (1984) Chapter 5.
425. Clydesdale, F.M., Main, J.H., Francis, F.J. and Damon, R.A. Jr. J. Food Sci. 43 (1978)
1687-1692, 1697.
426. Lancrenon, X. Process Biochem. 16 (1978) (10) 16.
427. Matsumoto, T., Nishida, K., Noguchi, M. and Tamaki, E. Agric. Bioi. Chem. 37 (1973)
561-567.
428. Tholakalabavi, A., Zwiazek, J.J. and Thorpe, T.A. In Vitro Cell. Dev. Bioi. 30P (1994)
164-170.
429. Dougal, D.K. and Weyrauch, K.W. Biotech. Bioeng. 22 (1980) 337-352.
430. Nagarajan, R.P., Keshavarz, E. and Gerson, D.F. J. Ferment. Bioeng. 68 (1989) 102-
106.
431. Yamakawa, T., Kato, S., Ishida, K., Kodama, T. and Minoda, Y. Agric. Bioi. Chem. 47
(1983) 2185-2192.
432. Yamamoto, Y., Knoshita, Y., Watanabe, S. and Yamada, Y. Agric. Bioi. Chem. 53
(1989) 417-423.
433. Kobayashi, Y., Akita, M., Sakamoto, K., Liu, H., Shigeoka, T., Koyano, T.,
Kawamura, M. and Furuya, T. Appl. Microbiol. Biotech. 40 (1993) 215-218.
434. Sakamoto, K., Iida, K., Sawamura, K., Hajiro, K., Asada, Y., Yoshikawa, T. and
Furuya, T. Phytochem. 33 (1993) 357-360.
435. Sakamoto, K., Iida, K., Sawamura, K., Hajiro, K., Asada, Y., Yoshikawa, T. and
Furuya, T. Plant Cell Tiss. Org. Cult. 36 (1994) 21-26.
436. Counsell, J.N., Jeffries, G.S. and Knewstubb, c.J. in Natural Colors for Food and
Other Uses (J.N. Counsell, ed.), Applied Science Pub., New York (1981) Chapter 7.
437. Maccarone, E., Maccarone, A. and Rapisarda, P. J. Food Sci. 50 (1985) 901-904.
438. Piattelli, M. and Impellizzeri, G. Phytochem. 9 (1970) 2553-2556.
439. Alard, D., Wray, V., Grotjahn, L., Reznik, H. and Strack, D. Phytochem. 24 (1985)
2383-2385.
440. Strack, D., Engel, V. and Wray, V. Phytochem. 26 (1987) 2399-2400.
441. Strack, D., Schmitt, D., Reznik, H., Boland, W., Grotjahn, L. and Wray, V.
Phytochem. 26 (1987) 2285-2287.
442. Trezzini, G.F. and Zryd, 1.-P. Phytochern. 30 (1991) 1901-1903.

443. Dopp, H. and Musso, H. Naturwissenschaften 60 (1973) 477--478.
444. Dopp, H. and Musso, H. Chern. Ber. 106 (1973) 3473-3482.
445. von Ardenne, R., Dopp, H., Musso, H. and Steglich, W. Z. Naturforsch. 29C (1974)
637-{)39.
446. von Elbe, 1.H. and Maing, I.-v. Cereal Sci. Today 18 (1973) 263-264.
447. lensen, R.A. in The Shikirnic Acid Pathway (E.E. Conn, ed.), Plenum, New York
(1986) Chapter 3.
448. ImpeJlizzeri, G. and PiatteJli, M. Phytochern. 11 (1972) 2499-2502.
449. Chang, c., Kimler, L. and Mabry, T.J. Phytochern. 13 (1974) 2771-2775.
450. Terradas, F. and Wyler, H. Phytochern. 30 (1991) 3251-3253.
451. Kimler, L., Larson, R.A., Messenger, L., Moore, J.B. and Mabry, T.l. 1. Chern. Soc.,
Chern. Cornrnun. (1971) 1329--1330.
452. Mabry, T.l., Kimler, L and Larson, R.A. Hoppe-Seyler's Z. Physiol. Chern. (1972)
3532.
453. Reznick, H. Z. Pjlanzenphysiol. 87 (1978) 95-102.
454. Girod, P.-A. and Zryd, J.-P. Phytochern. 30 (1991) 169-174.
455. Terradas, F. and Wyler, H. Helv. Chirn. Acta 74 (1991) 124-140.
456. Sciuto, S, Oriente, G. and PiatteJli, M. Phytochern. 11 (1972) 2259--2262.
457. Heuer, S. and Strack, D. Planta 186 (1992) 626-{)28.
458. Sciuto, S., Oriente, G., PiatteJli, M., ImpeJlizzeri, G. and Amico, V. Phytochern. 13
(1974) 947-951.
459. Wyler, H., Meuer, U., Bauer, 1. and Stravs-MombeJli, L. Helv. Chirn. Acta 67 (1984)
1348.
460. Bockern, M. and Strack, D. Planta 174 (1988) 101-105.
461. Bockern, M., Wray, V. and Strack, D. Planta 184 (1991) 261-270.
462. Bockern, M., Heuer, S. and Strack, D. Bot. Acta 105 (1992) 146--151.
463. Hirano, H. and Komamine, A. Physiol. Plant. 90 (1994) 239-245.
464. Sakuta, M., Hirano, H. and Komamine, A. Physiol. Plant. 83 (1991) 154-158.
465. Hirano, H., Sakuta, M. and Komamine, A. Z. Naturforsch. 47C (1992) 705-710.
466. Sweet, S.S. and Guruprasad, K.N. Indian 1. Exp. Bioi. 31 (1993) 168-172.
467. Girod, P.-A. and Zryd, 1.-P. Plant Cell Tissue Organ Cui. 25 (1991) 13-16.
468. Hirose, H., Yamakawa, T., Kodama, T. and Komamine, A. Plant Cell Physiol. 31
(1990) 267-271.
469. Obrenovic, S. Ann. Bot. 66 (1990) 245-248.
470. PiattelJi, M., Giudici de Nicola, M. and Castogiovanni, V. Phytochern. 10 (1971) 289-
293.
471. Bianco-Colomas, 1. and Hugues, M. 1. Plant Physiol. 136 (1990) 734-739.
472. Bianco-Colomas, 1., Lenoel, c.P. and Bulard, C. Plant Growth Regulators 7 (1988) 19-
27.
473. Schwartz, S.l. and von Elbe, 1.H. Z. Lebensrn. u. Forsch. 176 (1983) 448--453.
474. von Elbe, 1.H., Maing, I.-V. and Amundsen, C.H. 1. Food Sci. 39 (1974) 334-337.
475. Pasch, 1.H. and von Elbe, J.H. 1. Food Sci. 40 (1975) 1145-1146.
476. Saguy, 1.1. 1. Food Sci. 44 (1979) 1554-1555.
477. Huang, A.S. and von Elbe, 1.H. 1. Food Sci. 52 (1987) 1689-1693.
478. Attoe, E.L. and von Elbe, 1.H. 1. Agric. Food Chern. 30 (1982) 708-712.
479. Saguy, 1., Kopelman, 1.1. and Mizrahi, S. 1. Agric. Food Chern. 26 (1978) 360-362.
480. Attoe, E.L. and von Elbe, 1.H. 1. Food Sci. 46 (1981) 1934-1937.
481. Savolainen, K. and Kuusi, T. Z. Lehensrn. u. Forsch. 166 (1978) 19-22.
482. von Elbe, 1.H., Schwartz, S.l. and Hildenbrand, B.E. 1. Food Sci. 46 (1981) 1713-1715.
483. Huang, A.S. and von Elbe, 1.H. 1. Food Sci. 50 (1985) 1115-1120, 1129.
484. Altamirano, R.C., Drdak, M., Simon, P., Rajniakova, A., Karovicova, 1. and PrecJik,
L. Food Chern. 46 (1993) 73-75.
485. Bilyk, A. and Howard, M. 1. Agric. Food Chern. 30 (1982) 906--908.
486. Attoe. E.L. and von Elbe, 1.H. Z. Lebensrn. u. Forsch. 179 (1984) 232-236.
487. Wyler, H. and Dreiding, A.S. Helv. Chirn. Acta 45 (1962) 638-{)40.
488. Sapers, G.M. and Hornstein, 1.S. J. Food Sci. 44 (1979) 1245-1248.
489. Aurstad, K. and Dahle, H.K. Z. Lebensrn. u. Forsch. 151 (1972) 171-174.
490. Mikova, K. and Kyzlink, V. Sbornik Vysokeskoly Chemicko-Technologicke v. Praze, E.

58 (1985) 9-15.
491. Pasch, J.H. and von Elbe, J.H. f. Food Sci. 44 (1979) 72-74.
492. Czapski, J. Z. Lebensm. u. Forsch. 191 (1990) 275-278.
493. Drd:ik, M., Vallov:i, M., Greif, G., Simko, P. and Kusy, P. Z. Lebensm. u. Forsch. 190
(1990) 121-122.
494. Simon, P., Drd:ik, M. and Altamirano, R.C. Food Chem. 46 (1993) 155-158.
495. Soboleva, G.A., UI'yanova, M.S., Zakharova, N.S. and Bokuchava, M.A Biokhimiya
41 (1976) 968-973.
496. Shih, e.e. and Wiley, R.C. f. Food Sci. 47 (1981) 164--166, 172.
497. Lashley, D. and Wiley, R.C. f. Food Sci. 44 (1979) 1568-1569.
498. Elliott, D.C., Schultz, e.G. and Cassar, R.A Phytochem. 22 (1983) 383-387.
499. Wiley, R.e. and Lee, Y.-N. f. Food Sci. 43 (1978) 1056-1058.
500. Wiley, R.e., Lee, Y.-N., Saladini, J.J., Wyss, R.C. and Topalain, H.H. f. Food Sci. 44
(1979) 208-212.
501. Piattelli, M. and Minale, L. Phytochem. 3 (1964) 307-311.
502. Piattelli, M. and Minale, L. Phytochem. 3 (1964) 547-557.
503. D6pp, H. and Musso, H. Z Naturforsch. 29C (1974) 640-642.
504. Adams, J.P. and von Elbe, J.H. f. Food Sci. 42 (1977) 410-414.
505. Vincent, K.R. and Scholz, R.G. f. Agric. Food Chem. 26 (1978) 812-816.
506. Singer, J.W. and von Elbe, J.H. f. Food Sci. 45 (1980) 489-491.
507. Schwartz, S.J. and von Elbe, J.H. f. Agric. Food Chem. 28 (1980) 540-543.
508. Strack, D. and Reznik, H. Z. Pjianzenphysiol. 94 (1979) 163-167.
509. Strack, D., Engel, D. and Reznik, H. Z. Pjianzenphysiol. 101 (1981) 215-222.
510. Poumit, A., Lejeune, B., Grand, A. and Pourrat, H. f. Food Sci. 53 (1988) 294--295.
511. Drd:ik, M., Vallova, M., Daucik, P. and Greif, G. Z. Lebensm. u. Forsch. 188 (1989)
547-550.
512. Piattelli, M. in Chemistry and Biochemistry of Plant Pigments, 2nd edn. Vol. 1. (T.W.
Goodwin, ed.), Academic Press, New York (1976) Chapter 11.
513. Mabry, T.J. and Dreiding, A.S. Recent Adv. Phytochem. 1 (1968) 145-160.
514. Piattelli, M., Minale, L. and Nicholaus, R.A. Rend. Accad. Sci. Fis. Mat., Naples 32
(1965) 55-56.
515. Minale, L., Piattelli, M., de Stefano, S. and Nicolaus, R.A. Phytochem. 5 (1966) 1037-
1052.
516. Wyler, H. and Dreiding, AS. Helv. Chim. Acta 42 (1959) 1699-1702.
517. Schmidt, O.Th., Becher, P. and Hubnerr, M. Chem. Ber. 93 (1960) 1296-1304.
518. Piattelli, M., Minale, L. and Prota, G. Tetrahedron 20 (1964) 2325-2329.
519. Piattelli, M. and Minale, L. Ann. Chim. (Rome) 56 (1966) 1060-1064.
520. Mabry, T.J., Wyler, H., Sassy, G., Mercier, M., Parikh, I. and Dreiding, A.S. Helv.
Chim. Acta 45 (1962) 640--647.
521. Mabry, T.J., Wyler, H., Parikh, I. and Dreiding, A.S. Tetrahedron 23 (1967) 3111-
3127.
522. Piattelli, M., Minale, L. and Nicolaus, R.A. Phytochem. 4 (1965) 817-823.
523. Piattelli, M., Minale, L. and Prota, G. Ann. Chim. (Rome) 54 (1964) 955-962.
524. Wilcox, M.E., Wyler, H., Mabry, T.J. and Dreiding, A.S. Helv. Chim. Acta 48 (1965)
252-258.
525. Piattelli, M., Minale, L. and Prota, G. Phytochem. 4 (1965) 121-125.
526. Piattelli, M., Giudici de Nicola, M. and Castrogiovanni, V. Phytochem. 8 (1969)
731-736.
527. Nilsson, T. Lantbrukshogskolans Annaler 36 (1970) 179-219.
528. von Elbe, J.H., Sy, S.H., Maing, I.-Y. and Gabelman, W.H. f. Food Sci. 37 (1972)
932-934.
529. Wyler, H. and Dreiding, AS. Helv. Chim. Acta 44 (1961) 249-257.
530. Wolyn, D.J. and Gabelman, W.H. f. Amer. Soc. Hort. Sci. 115 (1990) 165-169.
531. Wolyn, D.J. and Gabelman, W.H. f. Hered. 80 (1989) 33-38.
532. Wilkins, e.K. Inter. f. Food Sci. Technol. 22 (1987) 571-573.
533. Pourrat, A., Lejeune, B., Jean, D. and Pourrat, H. Ann. Pharm. Fr. 38 (1980) 261-266.
534. Pourrat, H., Lejeune, B., Regerat, F. and Pourrat, A. Biotech. Lett. 5 (1983) 381-384.
535. Adams, J.P., von Elbe, J.H. and Amundson, C.H. I. Food Sci. 41 (1976) 78-81.
536. Avila, A., Garcia-Hernandez, F. and Santos, E. Dev. Ind. Microbiol. 24 (1983) 553-
561.
537. Bell, E.R.J. and White, E.B. Inter. Ind. Biotech. 9(3) (1989) 20-26.
538. Leathers, R.R., Davin, C. and Zryd, J.-P. In Vitro Cell. Dev. Bioi. 18P (1992) 39-45.
539. Lejeune, B., Pouget, M.P. and Pourrat, A. Labo-Pharma 334 (1983) 638-643.
540. Lejeune, B., Pourrat, A. and Pouget, M.P. Ann. Pharm. Fr. 44 (1986) 461.
541. Lejeune, B., Grand, A. and Pourrat, A. S. T.P. Pharma 3 (1987) 400.
542. von Elbe, J.H. in Current Aspects of Food Colorants (T.E. Furia, ed.), CRC Press,
Cleveland (1977) Chapter 3.
543. Pasch, J.H., von Elbe, J.H. and Sell, R.J. I. Milk Food Technol. 38 (1975) 25-28.
544. Altamirano, R.C., Drdak, M., Simon, P., Smelik, A. and Simko, P. I. Sci. Food Agric.
58 (1992) 595-596.
545. von Elbe, J.H., Klement, J.T., Amundson, C.H., Cassens, R.G. and Lindsay, R.C. I.
Food Sci. 39 (1974) 128-132.
546. Dhillon, A.S. and Mauer, A.J. Poultry Sci. 54 (1975) 1272-1277.
547. Kopelman, 1.1. and Saguy, I. I. Food Proc. Pres. 1 (1977) 217-224.
9 Less common natural colorants
F.J. FRANCIS
9.1 Acylated J3-ring substituted anthocyanins
For many years anthocyanins were believed to be sugar-substituted only in

the A-ring. This concept was surprising since J3-ring substitution was well
known in the closely related yellow flavonoid group. However, the first 13-
ring sugar-substituted anthocyanin was isolated by Yoshitama [1] from
Lobelia (Figure 9.1). It was delphinidin 3-rutinoside-5,3'-5'-triglucoside
acylated with caffeic and p-coumaric acids. Since then, similar pigments
have been found in Tradescantia, Bromeliaceae, Commelinaceae, Com-
positae, Gentianaceae, Leguminoseae, Liliaceae and Lobeliaceae [2]. This
group of pigments usually has a complex structure with several sugar and
acyl groups. For example, ternatin (see Figure 9.1) from Clitoria ternatea is
delphinidin 3,3' ,5'-triglucoside with four more sugar units, four molecules
of p-coumaric acid and one molecule of malonic acid [3]. With a molecular
weight of approximately 23tO, it has replaced the anthocyanin from
Heavenly Blue morning glories (Ipomoea tricolor) [4] as the largest
monomeric anthocyanin. Figure 9.1 presents the structure of some selected
highly acylated anthocyanins; however, there are some disagreements in
structure in the literature. For example, Stirton and Harborne [5] reported
that zebrinin had three caffeic acid units whereas Idaka et al. [61 reported
the presence of four units. Idaka et al. also reported that setcreasin from
Setcreasea purpurea contained cyanidin-3,7 ,3' -triglucoside with four ferulic
acid units. Shi et al. [7] reported that Tradescantia pallida anthocyanin
(TPA) from T. paUida (a synonym of S. purpurea) contained the same
basic structure except with three units of ferulic acid and one extra glucose
molecule. Shi et al. [8] later reported that the molecule also contained
caffeic acid. Baublis and Berber-Jimenez reported that TPA contained 3
molecules of glucose, one molecule of arabinose and 4 units of ferulic acid
(Figure 9.1(a» [9]. The number of complex acylated anthocyanins is likely
to increase in view of more sophisticated methods of structure determin-
ation. The methods of structure determination that. work so well on the
simpler anthocyanins [to] are not adequate to determine the structure of
the highly acylated anthocyanins. The methods of choice today are high-
performance liquid chromatography (HPLC), for clean separation, with
concurrent three-dimensional photodiode array detection in order to yield
good spectra. The number, identity and points, of attachment of the acyl
LESS COMMON NATURAL COLORANTS 311
(c) (d)
6
lJ-CAFF
6<fLU-<AFF~LU
'CAFF-GL
GLU-
I
CAFF
Figure 9.1 Formulae of selected acylated anthocyanins: (a) Pigment from Tradescantia
pallida; (b) HBA from morning glories; (c) Monardein; (d) Cinerarin; (e) Gentiodelphin;
(f) Platyconin; (g) Tematin 0; and (h) Zebrinin. Abbreviations Glu and Rha refer to the
sugars glucose and rhamnose. The abbreviations Caff, Coum, Fer and Mal refer to the acyl
acids caffeic, coumaric, ferulic and malonic acid. The superscript refers to the carbon linkage
of the sugar acid (e.g. 6). HBA = Heavenly Blue Anthocyanin.
molecules can be determined by mass spectroscopy of the intact molecule

and selected products of hydrolysis. The newer physical methods of
analysis such as fast atom-bombardment mass spectroscopy (FAB-MS),
COSY (H-H correlated spectroscopy), proton and 13C NMR spectroscopy

(nuclear magnetic resonance) were described by Brouillard [11].
The acylated anthocyanins are of interest as food colorants because the
acylation usually increases the stability of the pigment in acid media [12,
13] but not always [14a,b, 15]. This concept was tested by Bassa and
Francis [16a] for the stability of the normal acylated pigments [16b] from
sweet potatoes (Ipomoea batatas) in a model beverage. The deacylated
pigments were much less stable. The normal pigments were more stable
than cyanidin-3-glucoside from blackberries and a commercial sample of
oenocyanin from grapes. Asen et al. [12] reported that an anthocyanin
from Heavenly Blue morning glories (HBA) pigment had six glucose
molecules and three caffeic acid units. Teh and Francis [17] confirmed the
high stability of the HBA pigment in a model beverage. The anthocyanins
with a high degree of acylation, together with sugar substitution in the ~
ring, have been reported to be very stable. Brouillard [18] attributed the
exceptional stability of the anthocyanins from Zebrina pendula (see Figure
9.1) to the ability of the acyl groups to layer above and below the
anthocyanidin group, thereby inhibiting the formation of a pseudobase or a
chalcone. Brouillard was referring to the stability of the pigment to change
colour with pH and not the stability in a model beverage. However, the
same mechanism seems to operate in beverages. Teh and Francis [17]
reported that, in a model beverage, in decreasing order of stability, were
the pigments from lpomeoa tricolor, Zebrina pendula, blackberries and
oenocyanin. Terahara et al. [3] recently reported that ternatin I was even
more stable than HBA from morning glories. Dangles [19] studied the
kinetics of the interactions between the chromophore and the cinnamic
acid residues in terms of intramolecular complexes in a series of pigments
in Pharbitis nil. These involved pelargonidin and none to 3 caffeyl residues
on the same molecule. The kinetic and thermodynamic parameters of the
hydration reaction of the ftavylium chromophore which may involve
folding of the cinnamic residues allows one to predict the degree of
intramolecular complexation.
The complex acylated anthocyanins may also have another feature that
may make them attractive as food colorants. Some of them have an extra
absorption band at 560 to 600 nm and/or 600 to 640 nm, at weakly acid or
neutral pH values. This means that these anthocyanins will be highly
coloured at pH values above 4.0, whereas conventional anthocyanins are
nearly colourless. The extra bands, according to Saito [20] are due to the
tendency of the anthocyanins to form quinoidal base structures at neutral
or weakly acidic solutions. Two forms may be present, the 7-keto structure
or the 4'-keto form. The 7-keto form predominates in the conventional
anthocyan ins and only the normal 520 to 560 nm band appears. If the 7-
OH is blocked by methylation or glycosylation, then the 4'-form appears;
this shows an extra band at the higher wavelength. Sugar substitution in the
pH 2·0 - - -
pH 5 ' 5 - -
/
/
/
----- l
400 500 600 700
WAVELENGTH (nm)
Figure 9.2 Absorption spectra of the major anthocyanin in Tradescantia pal/ida. - --,
pH 2.0; - , pH 5.5.
B-ring together with complex acylation favours the existence of the 4' -keto
form even without substitution of the 7 hydroxyl. For example, the Zebrina
anthocyan ins (see Figure 9.2) show an extra absorption band at 587 nm.
The anthocyanins from Tradescantia pallida show an extra band at 583 nm
at pH 5.5 (Figure 9.2). Platyconin and cinerarin show the extra band.
Gentiodelphin [21] shows only a shoulder at pH 5.5 and 560 nm, because
the 7 position is free and only one acyl sugar group is in the 3' position. The
HBA anthocyanin does not show the extra bands because all the acyl
sugars are on the 3 position.
Figure 9.3 shows typical stability data for the anthocyanins from T.
pallida at pH 3.5 and 5.5 [22]. Colour data for a beverage are appropriately
expressed using Hunter tristimulus data. The data for Hunter L versus
storage time for T. pallida are relatively stable at both pH values. The
values for Theta 1 are redder than those for cyanidin-3-glucoside (Cn-3-G)
or oenocyanin. The pigment retention data show that the T. pallida
pigments are more stable particularly at higher pH values. This is not
surprising since conventional anthocyanins are known to be unstable and
nearly colourless at pH values above 4.0.
The presence of an absorption band at 583 nm and pH 5.5 and its
absence at pH 2.0 suggests that the difference in absorption at the two pH
values would provide a simple analytical method. Min et al. [23] proposed
that a single absorption method at pH 5.5 and 583 nm be used for fresh
tissue and the difference in absorption at pH 5.5 and 2.0 be used for stored
I Theta is the angle that a line joining a point in Hunter space makes with the origin. In this
work, a lower Theta value indicates a redder sample.
pH 5.5 pH 3.5
90
90
'"'":J 80
~ 80
...l
... 70
~
I:
:J 70 60
:r:
50
10 20 30 40 50 60 0 10 20 30 40 50 60
80 70
'"'":J 60
~ 60 50
~ 40
t2...'" 40
30
~ 20
I: 20
:J
:r: 10
10 20 30 40 50 60 10 20 30 40 50 60
3
c
'"
C 2
o
U
c
'"
~ o+-~~~,-~,-~,-~,-~
0: 0 10 20 30 40 50 60 o 10 20 30 40 50 60
Time(Weeks) TIme(Weeks)
Figure 9.3 Typical stability data for the anthocyanins from T. pallida ([!]), blackberries (Cn-3-
G) (+), and oenocyanin (0) from grapes. The Hunter L value is a measure of lightness with
white = 100 and black = O. Theta is a function of hue. In these examples, which are all in the
+a, +b Hunter quadrant, a theta = 0 represents a bluish-red colour whereas 90 indicates a
yellow colour.
samples. The Afi'::m 583 nm values were 111 and 97 respectively. The
anthocyanin content of T. pallida, using this method was reported to be
120 mg/l00 g [23].
Several sources of acylated pigments have been patented as potential
food colorants. These include morning glories [24, 25], red cabbage [26-
36], Gibasis geniculata [37], Zebrina purpusii [38] and sweet potatoes [39].
The patent literature for all food colorants from 1969 to 1984 was surveyed
(a)
R
(b)
(c)
(d)
R
Figure 9.4 Carotenoid pigments of annatto «a) bixin, R = CH 3 ; and norbixin, R = H;

(b) trans-bixin; and (c) C 17 yellow pigment) and saffron «d) crocin, R = gentiobiose; and
crocetin, R = H).
by Francis [40]. Murai and Wilkins [41] and LaBell [42] described the
recent introduction of a stable red colorant (trade name San Red RC) from
red cabbages.
9.2 Annatto
Annatto is a yellow-orange carotenoid preparation obtained from the

seeds of the plant Bixa orellana, which are grown primarily in Central and
South America. The pigments in annatto are a mixture of bixin, the mono-
methyl ester of a dicarboxylic carotenoid compound, and norbixin, the
dicarboxylic derivative of the same carotenoid as in bixin (Figure 9.4).
Both bixin and norbixin normally occur in the cis form and a small
percentage of both is changed to the more stable trans form upon heating.
A yellow degradation product, termed C17 yellow pigment, with the
formula shown in Figure 9.4, is also produced on heating. The cis forms are
redder than the trans forms or the yellow C17 compound, thus a series of
yellow to red colours are available.
A colorant can be extracted from the seeds by treatment with water.
Rozinkov et al. [43] patented a continuous countercurrent extraction with
water while submitting the seed to intensive mechanical friction. Bixin, on
treatment with alkalis, is hydrolysed to the water-soluble dibasic acid form
norbixin. Normal commercial procedure is to soften the seeds with steam
and extract the pigments with propylene glycol contammg potassium

hydroxide. An oil-soluble form can be produced by treating steam-
softened seeds with a solvent such as ethanol, a chlorinated hydrocarbon or
a vegetable oil. The pigments are primarily in the outer coats of the seed
and can also be removed mechanically. Guimares et al. [44] published a
method to isolate pigment from seeds using agitation in a spouted bed
holding 50 kg of seeds. The dust from the agitation contained over 20% of
pigment. A dry powder is useful for colouring dry and instant foodstuffs.
The pigments in annatto can be separated and analysed using HPLC
methods. Rouseff [45] published a method using a Zorbax ODX column
with a 58/42 mixture of waterltetrahydrofuran. Earlier methods used paper
chromatography or TLC plates [46] but usually required separate analyses
for annatto and turmeric. The two colorants are often found together
because addition of turmeric produces a yellower shade and also lends
stability [47]. The HPLC methods allow analyses for both annatto and
turmeric in the same sample.
The pigments of annatto are unstable to oxidation, as are all carotenoid
compounds, but they are relatively more stable than the other carotenoid
colorants used in food. Najar et al. [48] reported the effects of light, air,
antioxidants and pro-oxidants on the stability of annatto extracts. Light
proved to be the most destructive and 10% ascorbyl palmitate was an
effective antioxidant. Heltiarachchy and Muffett [49] patented the use of in-
organic polyvalent cations together with hydrocolloids containing carboxyl
groups to produce stabilized colorant complexes with annatto. Ford and
Mellor [50] patented a process to use demineralized water or demineral-
ized glucose syrup together with 0.5% ascorbic acid to prevent fading of
annatto in a beverage. Berset and Marty [51] reported the use of annatto in
extrusion cooking. They were able to demonstrate good thermal stability at
between 175 and 185°C. Ford and Draisey [52] patented the use of annatto
together with cloudifier preparation containing beeswax and a natural gum
to minimize fading of annatto in fruit squashes. Ikawa and Kagemoto [53]
reported that a stable colorant could be made from bixin by addition of
ethanol, sucrose acetate hexaisobutyrate, coconut oil and gum arabic.
Mori et al. [54] reported stabilization of bixin with vanillin eugenol or
vitamin E. Harrison [55] patented the production of a yellOW and blue twin
curd cheese by adding annatto to half the batch. Both batches were
remelted, coagulated, mixed and ripened. The cheese had an attractive
mottled appearance. Collins [56] published a detailed account of the early
history of annatto together with the cultivation, harvesting, processing,
extraction degradation and chemical interactions. He also described the
production of water-soluble and oil-soluble preparations and their appli-
cations to foodstuffs.
Annatto has been used in foods, especially dairy products, for many
years. Technically it is a good colorant and the existence of both water-
soluble and oil-soluble forms has provided flexibility. The worldwide

movement towards natural colorants in the past two decades has led to a
great increase in the use of annatto.
9.3 Saffron
Pigments similar to those in annatto are found in saffron. They are

crocetin, a dicarboxylic carotenoid (Figure 9.4) together with its digentio-
bioside ester, crocin. Gentiobiose is a diglucoside with a beta 1-{) linkage.
The sugar portion makes the molecule water-soluble. It is one of the few
carotenoids found in nature that is freely water-soluble and this is one
reason for its application in the food and pharmaceutical industries.
Saffron is obtained from the stigmas of the flowers of Crocus sativus. The
same pigments also occur in C. albifloris, C. luteus, Cedrela toona,
Nyctanthes arbor-tristis, Verbascum phlomoides and the cape jasmine
(Gardenia jasminoides). Until recently, Crocus sativus was the only
commercial source of crocin and crocetin and, since it requires about
64 000 stigmas to produce one pound (454 g) of colorant, the price is
obviously very high (> US$500/Ib). Also production is limited due to the
cost of labour. Saffron, of course, provides both colour and flavour and is
considered to be a gourmet item.
The principal substance responsible for the bitter taste of saffron is
picrocrocin [57]. Picrocrocin hydrolyses to safranal (2,6,6-trimethyl-3-
cyclohexadiene-carboxaldehyde) and glucose. Safranal is formed in saffron
extracts during processing and is the main component responsible for the
odour. Alonso et al. [57] studied the auto-oxidation of saffron at 40°C and
75% relative humidity using a spectrophotometric analysis. Picrocrocin,
safranal and crocin absorb at 257, 297 and 327 nm respectively [58]. The
loss of colour followed first-order kinetics and the loss of bitterness
followed second-order kinetics. Solinas and Cichelle [59] were able to
analyse for the colour (crocin and four similar compounds) and flavour
(picrocrocin and safranal) using an HPLC approach. This method was
effective in judging the quality and authenticity of saffron extracts.
Klaui et al. [60] patented the synthesis of di-n-dodecyl, di-n-decyl, and
di-n-tetradecyl esters of crocetin. These compounds, together with alpha-
tocopherol and ascorbyl palmitate provided a stable yellow colorant for
ice-cream and beverages. The same concept was extended to the synthesis
of several other esters [61] but the preferred compounds are listed above.
Hashimoto [62] reported that croce tin , norbixin, ~-carotene, antho-
cyanins, Iycopene, curcumin, chlorophyll and iridoid preparations could be
stabilized by cyclodextrins. Cyclodextrins are cyclic non-reducing oJigo-
saccharides produced by the action of an enzyme, cyclomaltodextrin
glucanotransferase on starch. The ex, ~ and 'Y forms contain 6, 7 and 8
glucose units respectively. For example, a-cyclodextrin improved the

resistance to light of crocetin by a factor of 8. It also prevented the colour
shift to longer wavelength with change to acidic pH.
Saffron extracts contain other carotenoids such as l3-carotene and
zeaxanthin, as well as crocin and crocetin. They also contain, as might be
expected, a range of flavonoid compounds. Garrido et al. [63] reported
derivatives of myricetin, quercetin, kempferol, delphinidin and petunidin.
Saffron is a very old food colorant and spice. Its pure-yellow colour and
odour have made it an attractive ingredient in specialty foods such as
risotto, curry, bakery products, sausages, beverages and so on. Its current
price and availability will limit it to specialty gourmet foods.
9.4 Gardenia pigments
The technological success of the pigments in annatto and saffron combined

with their relatively high price led to searches for other plant sources with
the same pigments. It soon became obvious that the same pigments (but
not the flavour) could be obtained in much greater quantities and at a
lower price from the fruits of the cape jasmine.
The fruits of gardenia contain three major groups of pigments, crocins,
iridoid pigments and flavonoids. The carotenoid crocin and related
compounds develop in the fruit of G. jasminoides [64] during the eighth to
twenty-third weeks of growth but the iridoid pigments develop 1 to 6 weeks
after flowering. The structure of the iridoid pigments gardenoside, geni-
poside, shanzhiside, gardoside, methyldeacetylasperuloside, genipingen-
tiobioside, beniposidic acid, acetylgeniposide and scandoside methylester
are shown in Figures 9.5 and 9.6 [65]. The third group comprises a series of
flavonoid compounds as illustrated by the five flavonoids isolated by
Gunatilaka et al. [66,67] (Figure 9.7) from G. fosbergii. This is a different
species from G. jasminoides but flavonoid pigments tend to be similar in
closely related species. The flavonoid pigments are yellow but their
contribution to the overall colour of gardenia preparations is unclear.
Gardenia fruits and other portions of the plant have been a part of the
folklore of Oriental culture for many years [68]. There appears to have
been a revival of interest in recent years in terms of biosynthesis [69-74],
physiological effects [65, 68, 75-77], chemical composition [64, 78--82],
analysis [83, 84], and antimicrobial activity [85, 86].
The production of food colorants from gardenia is apparently being
investigated at the present time but the mechanism is unclear. Most patent
procedures involve extraction of the fruit with water, treatment with
enzymes having l3-glycosidic activity or proteolytic activity (bromelain) and
reaction with primary amines from either amino acids or proteins such as
those in soy. Manipulation of the reaction conditions such as time, pH,
(a)
(c)
Figure 9.5 Iridoid pigments of gardenia. (a) geniposide; (b) shanzhiside; (c) gardoside;
(d) gardenoside. Gl = glucose.
(a) (b)
~OCH3
~entiobi05e
H OOCH3
(c) (d)
Figure 9.6 Additional iridoid pigments of gardenia. (a) geniposidic acid; (b) genipingentio·
bioside; (c) scandoside methylester; (d) acetylgeniposide; (e) methyldeacetylasperuloside.
temperature, oxygen content, degree of polymerization and conjugation of

the primary amino reaction products, and so on, enables a series of
colorants to be produced that vary from yellow to green, red, violet and
blue. Several of the patents involve culturing preparations of gardenia with
microorganisms such as Bacillus subtilis [87], Aspergillus japonicus or a
~4
Flavonoid compounds
CPO. R1 ~ R3 ~ ~ R6
1 H H H OMe Ov1e OMe
2 Ov1e H H OH OMe OH
3 H H OMe OMe H OH
4 Ov1e H H OMeOMe OH
5 OMeOMe H H OH H
Figure 9.7 Flavonoid compounds of gardenia.
Rhizopus [88]. Depending on the condition, preparations with the colour

described above can be produced. The overall colour, of course, would be
a combination of the yellow-orange contributed by the carotenoids and the
colours produced by the iridoid pigments and their derivatives. Appar-
ently, seven preparations involving four greens, two blues and one red
have been commercialized in Japan. Flow sheets for their production
together with their physical properties such as solubility in various
solvents, heat and light stability, pH effect, hygroscopicity, reaction with
metals, antioxidants, and so on, were published by Yoshizumi et af. [65].
Nineteen patents on gardenia colorants were published up to 1984 [40].
More recent patents involved hydrolysis of the iridoid glycoside geniposide
by p-glucosidase to produce genipin. The genipin is then reacted with
taurine to produce a stable blue colorant [89, 90]. Tomara et af. [91]
patented a process involving gardenia extracts treated with acylase for
hydrolysis of geniposides, and then incubated with a microorganism. After
enzyme inactivation with heat, the colorant could be recovered with
ethanol. Fujikawa et af. [92] reported the reactions of genipin, formed by
hydrolysis of the geniposides, with the amino acids glycine, alanine,
leucine, phenylalanine and tyrosine. The resultant 'brilliant sky-blue'
pigment was stable for 2 weeks at 40°C in 40% ethanol. Fujikawa et al. [93]
also reported the continuous production of a blue pigment using B. subtilis
cells immobilized on a calcium alginate gel. This micro-organism had
l3-glucosidase activity and assimilated genipin as a carbon source. The
colorants from gardenia have been described in several reviews (e.g.
Toyama and Moritome [94]).
Colorant preparations from gardenia were suggested by Yoshizumi [75]
for use with candies, sweets, ices, noodles, imitation crab (Kamabago),
fish eggs, glazed chestnuts, boiled beans, dried fish substitute (furikaki) hot
cakes, liquor, etc. Mitsuta et al. [29] patented a green colour for beverages
using colorants from gardenia. In view of the wide range of available
colours and their apparent excellent stability, colorants from gardenia
appear to have good potential.
9.5 Cochineal and related pigments
Cochineal has been suggested as 'the best of all natural pigments' [95]. It
has been an article of commerce for many years in the form of extracts
from the bodies of the coccid insects from the families Cocco idea and
Aphidoidea. Several preparations are known under various names [86].
Armenian Red is obtained from the insect Porphyrophera hameli, which
grows on the roots and lower stems of several grasses in Azerbaijan and
Armenia. Kermes is obtained from the Old World insects Kermes ilicis or
Kermococcus vermilis, which grows above ground on several species of
oak, particularly Quercus coccifera-the 'Kermes Oak'. Polish cochineal is
obtained from Margarodes polonicus or Porphyrophera pomonica found
on the roots of Scleranthus perennis, a grass found in Central and Eastern
Europe. Lac is obtained from Laccifera lacca, which is found on the trees
Schleichera oleosa, Ziziphus mauritiana and Butea monosperma found in
India and Malaysia. The lac insects are better known as sources of shellac.
American cochineal is obtained from insects of the family Dactylopiidae
primarily Dactylopius coccus Costa found as parasites on the aerial parts of
cactus, Opuntis and Nopalea spp.-especially N. cochenillifera. There are
80 000 to 100 000 insects in each kilogram of raw, dried cochineal as
exported from Central and South America. World production today is a
small fraction of that available in the sixteenth to nineteenth centuries. For
example, the Canary Islands alone produced 3000 tons in 1875, but the
availability of the synthetic food colorants in the twentieth century and the
high cost of cochineal reduced their usage. Cochineal has been manufac-
tured in India, Persia, Europe, Africa, Mexico and the Canary Islands, but
the only source today is Peru [96].
Cochineal is extracted from the bodies of female insects just prior to egg-
laying time, and they may contain as much as 22% of their dry weight as
pigment; Historically the pigment was extracted with hot water and the
colorants were known as simple 'extracts of cochineal' [97]. Recently,
proteinase enzyme treatment and further purification has provided color-
ants known as the 'carmines of cochineal'. The word carmine may be used
as a general term for this family of anthraquinone pigments but is more
usually considered to be·the aluminium or magnesium lake of carminic acid
on aluminium hydroxide, and contains about 50% carminic acid. These
preparations. are likely to be the commercial source of cochineal for the
future since synthesis is unlikely at this time in view of the expense of
toxicity testing. Even assuming that the synthetic pigments were 'nature-
alike' ,it is likely that considerable testing for toxicity would be required.
Schul [96] discussed the production, chemistry, applications and particu-
larly the analytical problems for quality control of cochineal from Peru. He
proposed that the price of carmine should be based on its colour strength
and not carminic acid content and provided details of the analytical
methods. Valdes-Martinez et al. [98] reported details of the chemistry and
kinetics of degradation under a number of storage conditions.
The structure for carminic acid, one of the major colorants in cochineal,
is shown in Figure 9.8. The pigment in Kermes is kermesic acid, the
aglycone of carminic acid; It also occurs as an isomer, ceroalbolinic acid.
The lac colorants are a complex mixture of laccaic acids (Figure 9.9) and
the closely related· compounds erythrolaccin and deoxyerythrolaccin
(Figure 9.8) .. Solutions of carminic acid at pH 4 show a pale-yellow colour
and actually have little intrinsic colour below pH 7, but they do complex
with metals to produce brilliant red hues. Complexes with tin and
aluminium produce the most. desirable colours and most commercial
preparations involve complexes with aluminium. Floyd [85]. has shown that
a range of colours 'strawberry' to near 'blackcurrant' can be produced by
adjusting the ratio of carmine acid to aluminium.
Carmine can be used in powder form to colour a variety of foods. It can
be used in a solution of ammonia to colour foodstuffs, such as baked
products, confectionery, syrups, jams and preserves. Food with low pH
values may cause precipitation of the colorant but this may be an
advantage.
Alkannet (Figure 9.13 below) is a related pigment extracted with. alcohol
from the roots of Alkanna tinctoria Tausch and Anchusa tihctoria Lorn
from sQuthern· Europe, The red, pigment is only slightly soluble in· water
but readily soluble in organic solvents. It has. been used to colour
confectionery, ice-cream and wines [99].
Cochineal and related compounds have had a real resurgence ofinterest
as food colorants in view of the movement towards the 'naturals'. This has
led to more research interest-particularly when its safety was. questioned
(a)
H H
(c)
H H
(e)
Figure 9.8 Anthraquinone pigments of cochineal and kermes. (a) Carminic acid;
(b) erythrolaccin; (c) kermesic acid; (d) deoxyerthrolaccin; (e) ceroalbolinic acid.
(a)
OOH
(b)
Figure 9.9 Anthraquinone pigments in lac. (a) Laccaic acids. A: R = CH 2-NH-COCH 3 ;

B: R = CH20H; C: R = CH(NH 2)COOH; E: R = CHr NH 2 • (b) Laccaic acid D.
[100, 101), and a series of exhaustive toxicological studies were initiated

[102-107]. The consensus by an FAOIWHO Expert Committee on Food
Additives was that carmine, at doses employed as food colorants, had no
adverse effect. The renewed interest led to reports on methods of analysis

such as the spectrophotometric study of carmine acid in solution and its
application to analysis [108]. Recent interest in the quinone area may have
been stimulated by the extensive research on microbiological growth and
physiology related to the production of antibiotics, cell culture and
generaly biotechnology. Several new compounds have been reported.
Richardson et al. [109] reported a new anthraquinone (1,6-dihydroxy-4-
methoxy-9,1O-anthraquinone) from Xenorhabdus luminescens. It occurred
together with a hydroxystilbene antibiotic. Mukherjee et al. [110] reported
a new anthraquinone pigment from Cassia mimosoides. The yellow
pigment was found in the leaves and stems of the senna plant. Sakuma et
al. [111] reported two new napthopyrone pigments from the crinoid
Comanthus parvicirrus. The two new pigments occurred together with
comaparvin, 6-methoxycomaparvin, and 6-methoxy-comaparvin-S-O-Me-
ether. Ali et al. [112] reported the presence of deoxyerythrolaccin and a
new anthraquinone in Gladiolus segetum. Ikenaga et al. [113] studied the
growth and napthoquinone production in Lithospermum erythrorhizan.
They were interested in seasonal changes in order to optimize the content
of shikonen derivatives for use in oriental medicine. Nedentsov et al. [114]
studied the biosynthesis of the napthaquinone pigments in the fungus
Fusarium decemcellulase. They found more than ten pigments including
fusarubin, javanicin, nonjavanicin, anhydrojavanicin, anhydrofusarubin,
movarubin and bostrycoidin. Suemitsu et al. [115] reported the formation
of modified anthraquinones in an extensive study of the physiology and
metabolic products of Alternaria porri. They named the novel bioanthra-
quinones, alterporriols. In addition, Billen et al. [116] described the
stereochemistry of the pre-anthraquinones and the dimeric anthraquinone
pigments in a book chapter. Furthermore, Pari sot et al. [117] reviewed the
physiology and genetic control of the napthaquinone pigments produced
by the filamentous fungus Fusarium solani.
The patent literature from 1969 to 1984 contains twelve patents on
napthaquinones and anthraquinones [40]. They are involved mainly with
purification by alkali [118], alcohols [119], stabilization with metallic salts
[120, 121], sugars [122] and alum [123], and derivatization [124-126]. Two
more recent patents involved the application of quinone pigments to
cosmetics [127] and the stabilization of lac food colorants [128]. Lac
pigments in ethanol, propylene glycol, glycerol or sorbitol could be
stabilized by the addition of citric, malic, tartaric, acetic, butyric, lactic,
hydrochloric or sulphuric acid. Another patent [129] reported that carmine
and its lakes could be stabilized with carboxylic acid derivatives and
phosphates. Baldwin [130] patented the use of carmine to colour collagen
sausage casings.
The anthraquinone and napthaquinone pigments have a good reputation
as food colorants primarily because of their stability and high tinctorial
power [95]. They were the chromophore of choice in the 'linked polymer'
colorants pioneered by the Dynapol Corporation [40]. Unfortunately this
concept is no longer being commercialized but the concept is interesting
and may have application for other food additives. Some of the new
pigments described in the literature may have potential as food colorants.
9.6 Turmeric
Turmeric, also called Curcuma, is a fluorescent yellow-coloured extract

from the rhizomes of several species of the curcuma plant. Curcuma longa
L. is the usual commercial source. Three pigments occur in every species,
curcumin, demethoxycurcumin, and bisdemethoxycurcumin (Figure 9.10)
together with compounds that contribute to the flavour. These include
turmerone, cineol, zingeroni, and phellandrene. A dark-coloured turmeric
oleoresin containing both pigments and flavour compounds can be
prepared by extracting the dried rhizomes of C. longa with ethanol and
removing the solvent under vacuum. Turmeric colorants of varying degrees
of purity can be obtained from the oleoresin. Fukazawa [131] patented a
process to produce turmeric from C. domestica (syn. C. longa) by
extracting the rhizomes with ether, evaporating the ether, and dissolving
the residue in vegetable oil. Verghese and Joy [132] reported an extraction
process using ethyl acetate which produced a 2.77% yield. Govindarajan
[133] estimated that 160000 tons were produced annually. The large
production has stimulated research on optimization of yield. For example,
Tonnesen et al. [134] studied the variations in turmeric from Nepal. They
reported that the major pigment was bisdemethoxycurcumin as opposed to
sources from the other major producing areas in which the major pigment
was curcumin. Sampathee et al. [135] studied the relationship between
curing methods and the quality of the final product. Iyer [136] reported a
High Performance Thin Layer Chromatography (HPTLC) method for
quick and accurate profiling of the three major pigments for quality control
applications.
Turmeric oleoresins and colorants are unstable to light and alkaline
conditions. Thus several patents are involved with adjustment to acid
conditions. Leshnik [137, 138] prepared a stable preparation by spray-
drying with citric acid, waxy maize starch, Polysorbate 60, and sodium
citrate. The encapsulated product was stable for 16 weeks at 35°C.
Schrantz [139] patented a process using gelatin and citric acid. Obata et al.
[140] prevented colour loss by addition of gentisic acid, gallic acid, tannic
acid polyphosphate and citric acid.
Curcumin is insoluble in water but water-soluble complexes can be made
by complexing with metals. Maing and Miller [141, 142] employed
stannous chloride and zinc chloride to produce an intensely orange
Figure 9.10 Pigments of turmeric. Curcumin: RI = R z = OCH 3 . Demethoxycurcumin: RI =

H; R z = OCH3 . Bisdemethoxycurcumin: RI = R z = H.
colorant using a ratio of 4 to 5 moles of curcumin to each mole of metal.

Curcumin colorants can also be prepared by absorbing the pigment on
finely divided cellulose [143, 144,]. Goldscher [145] patented the addition
of glycine to reduce the bitter taste.
Tonneson and Karlsen [146] reported an HPLC method of analysis that
would separate the three components of turmeric. Separation and indivi-
dual estimation is important because the extinction values differ consider-
ably as well as the stability. They commented that the well known boric
acid test is sometimes inconclusive because of interference with other
extracted components. With the HPLC method and a UV detector, they
could detect greater than 20 ng of pigment. With a fluorescence detector
using 420 exciting and 470 emission nm they could detect greater than 2 ng.
Rouseff [45] reported an HPLC method that would separate and estimate
the three pigments in turmeric and the two in annatto in one analysis.
Turmeric and annatto are often found in combination, since apparently the
mixture is more stable [47]. C. domestica also contains 2-hydroxy methyl
anthraquinone [147].
Shah and Netrawadi [148] reported that turmeric extracts did not show
mutagenic activity even after activation with mammalian liver microsomal
extracts. Nagabhushan et al. [149] reported that turmeric extracts and all
three pigments reduced the nitrosation of methylurea by sodium nitrite and
thus reduced the mutagenicity of Salmonella typhimurium. EI Gazzar and
Marth [150] reported that Listeria monocytogenes could not survive in
dairy starter cultures when annatto or turmeric colorants were added.
These encouraging antibacterial and antimutagenic responses may encour-
age even wider use of turmeric in foods. Turmeric is already widely used in
the food industry in canned products, pickles, soups, mixes and confection-
ery. Turmeric is the usual yellow colorant in mustard and pickled
cauliflower.
(b)
(c)
H H
H
Figure 9.11 Pigments of carthamin: (a) carthamin; (b) safflor yellow A; and (c) safflor
yellow B.
9.7 Carthamin
Carthamin, sometimes called carthemone, is a yellow to red preparation

extracted from the flowers of Carthamus tinctorius. It contains three
chalcones [151, 152], the red carthamin, safftor yellow A, safftor yellow B
(Figure 9.11) and several precursors. Fresh yellow flower petals contain
precarthamin, which oxidizes to form the red colorant carthamin. Cartha-
min, upon treatment with dilute hydrochloric acid, yields two isomers-red
carthamin and yellow isocarthamin. Upon acid hydrolysis, carthamin
yields 'glucose and two aglycones, the flavanones carthamidin and isocarth-
amidin.
The formation of cart ham in involves an oxidizing or enzymatic step.
Saito et al. [153] and Saito and Fukushima [154], reported the conditions
for enzymatic synthesis and oxidation [155, 156]. They reported that
exogenous substances such as alcohols, amino acids, amines, carboxylic
acids and ethers produced a bathochromic shift in colour, whereas esters
and some fatty acids (except formic and acetic) caused a hypsochromic
shift.
The earlier patents on carthamin involved simple aqueous extraction of
the petals [157, 158] or crushing the petals prior to extraction to allow
oxidation [159]. Both a red and a yellow colorant could be obtained for use
with pineapple juice, and yoghurts [160]. Fukushima et al. [161] published
a method for the purification of the three pigments by adsorption on
polystyrene resins. Purification and stabilization has been greatly enhanced
by absorption of carthamin on cellulose. Apparently cellulose has great
affinity for carthamin and this is known as the 'Saito Effect' [154]. 'The
effect is so strong that the carthamin may be retained for more than 1000
years without appreciable change in the red coloration' [162]. No storage
data were presented to substantiate this claim. Precarthamins and safflor
yellow A+B were not absorbed on a CM Sephadex C-25 column, but
carthamin is strongly absorbed and could be eluted with dilute formic or
acetic acid or ammonia. Apparently the mechanism of absorption on
cellulose powder and cellulose ion exchanges is different. Saito [163]
reported that the site of absorption was the primary alcohol hydroxyl on
the glucose unit of the macromolecules. He arrived at this conclusion by
studying absorption with a series of synthetic glycosyl polymers. The
cellulose absorption method was improved and adapted to large-scale
production using a methanolic extract and a BD-cellulose column in acidic
media [164]. Elution with 3.3 M acetic acid in acetone and successive
chromatography also yielded pure safflor yellow B. Another method
involved the use of a cellulose derivative, diethylaminoethylcellulose. It
would absorb precarthamin, carthamin, safflor A and B and successive
chromatography yielded the pure pigments. The use of cellulose, or
cellulose and chitin, to provide a colorant that was dispersible in water or
oil was patented by the Sanyo-Kokusaku Pulp Co. [165, 166]. Three
patents were issued for tissue-culture methods for production of carthamin
and related pigments. Yomo et al. [167, 168] used flower bud cells on
Murashiga-Skoog agar followed by liquid culture with cellulose. Addition
of chitosan raised the production of carthamin from 5 to 50 mg/l. Daimon
et al. [169] patented an improved method involving liquid culture and 4%
powdered cellulose. The tissue-culture methods may involve production of
novel pigments as well as the expected group as reported by Saito et al.
[170]. The production of pigments of C. tinctorius by tissue culture was
recently viewed by Wakayama [171].
Carthamin is the only chalcone-type pigment suggested for colouring
foods. Little information is available on its stability in formulations but it
appears to be very promising. Its yellow to red-orange gives it some
flexibility. Its applications were reviewed by Onishi [172].
9.8 Monascus
The genus Monascus includes several fungal species that will grow on
various solid substrates, especially steamed rice. The combination of
(8) (b)
(c)
Figure 9.12 Pigments of monascus. (a) Monascin: R = CSH l1 ; ankaflavin: R = C7 H 1S .

(b) Rubropunctatin: R = CSH l1 ; monascorubin: R = ~HlS. (c) Rubropunctamine: R =
CSH l1 ; monascorubramine: R = ~HlS.
fungus and substrate has been known for many centuries in south and east
oriental countries. It was first mentioned in 1590 in a treatise on Chinese
medicine [81]. Traditionally in the Orient, Monascus species were grown
on rice and the whole mass eaten as such, or dried, powdered and
incorporated into food as desired. They were used to colour wine, bean
curd, and also as a general food colorant and a medicine [173]. They are
currently being produced commercially in Japan, Taiwan and China.
The structure of the monascus pigments is shown in Figure 9.12.
Monascin and ankaftavin are yellow, rubropunctatin and monascorubin are
red, whereas rubropunctamine and monascorubramine are purple [174,
175]. The yellow and red pigments are considered to be normal secondary
metabolites of the fungal growth whereas the two purple pigments may be
produced by chemical or enzymatic modification of the yellow and red
pigments. The red pigments readily react with compounds containing
amino groups via a ring opening and a Schiff rearrangement to form water-
soluble compounds (Figure 9.13). The monascus pigments have been
reacted with amino sugars, polyamino acids, amino alcohols, ethanol,
chitin amines or hexamines [176-178]; proteins, peptides and amino acids
[179, 180]; bovine serum albumin [181]; casein [182, 183]; gluten [184];
selected amino acids [181]; RNA [185]; nucleic acids [180]; sugar amino-
acid browning-reaction products [186]; aminoacetic acid and aminobenzoic
acid [82]. Sweeney et al. [187] prepared the N-glucosyl derivatives of
rubropunctamine and monascorubramine and investigated the protective
effect of these compounds by 1,4,6-trihydroxynaphthalene. Obviously,
several monascus derivatives are possible, and these apparently show
greater water solubility, thermostability and photostability.
(a) (c)
M'~' ~
HO i
HO bH OH 0
N
Pr
'-~ N'R'
Wi"'
R1
(b) (d)
H R
H H
Figure 9.13 Miscellaneous pigments: (a) rubrolone suggested by Iacobucci and Sweeney
[197]; (b) is deoxyanthocyanidin also suggested by Iacobucci and Sweeney [198]; (c) is a red
pigment suggested by Moll and FaIT [167, 168] where R represents an aliphatic radical and R 1
represents a radical of a compound of the formula HN-R which represents an amino sugar, a
polymer of an amino sugar or an amino alcohol; and (d) is alkannet.
For many years, Monascus species were grown on solid cereal substrates
and the whole mixture was ground and used as a food additive. It became
obvious that the fungus could be cultured in liquid media or semi-solid
state, so several studies were published to optimize the production of
pigment in a variety of culture modes. Han [188] listed 125 references to
growth studies. He screened thirteen strains for pigment production in
submerged flask culture and chose Monascus purpureus ATCC 16365 to be
most appropriate for solid-state fermentation [189, 190]. The optimum
conditions for pigment production were polished long rice at 50% moisture
and pH 5.0 to 6.0 at 30°C and RH-97-100. Supplementation with zinc
increased the red pigment content by a factor of 2.5 and the yellow
pigments by 2.0. Optimization of pigment production by other strains and
other modes will obviously differ [40]. Much recent research has been
devoted to studies on general culture conditions and substrates. Lin and
Iizuka [173] studied rice meal, wheat bran, wheat meal, bread meal, corn
meal and several others. Kim et al. [191] studied the use of molasses and
corn. Shepherd and Carels [192] developed a two-stage process that was
optimized first for growth and then for pigment production. Others studied
the growth processes to optimize the type and amount of pigment. For
example, growth of M. anka on bread produced only yellow pigments
[193], whereas culture on a solid medium gave superpigment production
[194, 195]. Lin [196] and Su et al. [197] also studied strain selection for
pigment production. Recent research has included higher pigment produc-
tion by selection of mutants induced by UV light or N-methyl-N-nitro-
nitrosoguanidine [198] or produced by protoplast fusion [199].
Monascus species produce other compounds such as ethanol, enzymes,

co-enzymes, monascolins, which modify fat metabolism and inhibit choles-
terol biosynthesis [200, 201], antibiotics, antihypertensives and flocculants.
For use as food colorants, the above compounds have to be removed or
preferably not produced in the process. The antibiotic content, in particu-
lar, would be undesirable in a food colorant. Wong and Bau [202] first
reported the antibacterial activity of M. purpureus. Wong and Koehler [81]
found that the increase in antibiotic content was usually associated with the
increase in pigment content but that acetate in the culture media inhibited
the formation of the antibiotic and enhanced the pigment production. They
suggested that this could be one approach to production of colorant
preparations free of antibiotic activity. Wong and Koehler [81] also
reported that cultures of A TCC 16365 had no antibiotic activity. This was
confirmed by Han [188]. Monascus species also produce ethanol, which, as
an alternate metabolic pathway, could decrease the production of pig-
ments. Apparently, the conditions that produce large quantities of ethanol
decrease pigment production; however, ethanol production in small
quantities apparently stimulates pigment production [188].
The monascus pigments are readily soluble in ethanol and only slightly
soluble in water [203]. Kumasaki et al. [204] reported that monascorubrin
was soluble in ether, methanol, ethanol, benzene, chloroform, acetic acid
and acetone but insoluble in water and petroleum ether. Ethanolic
solutions of monascorubrin are orange at pH 3.0 to 4.0, red at 5.0 to 6.0
and purplish red at 7.0 to 9.0. Su and Huang [203] reported that the
pigment from M. anka faded under prolonged exposure to strong light and
the pigment in 70% ethanol was more photostable than solutions in water.
They also reported that the pigment in 70% ethanol was very heat stable at
temperatures up to lOO°C. It also showed excellent heat stability at neutral
or alkaline conditions. Gan [205] used UV and visible spectroscopy and
NMR to study the stability of the yellow monascus pigments. They showed
good stability at lOO°C and much less at 120°e. There is little detailed data
on stability of monascus pigments in food formulations.
Colorants from microbial species offer considerable advantages since
they can be produced in any quantity and are not subject to the vagaries of
nature. The range of colours from yellow to purplish red combined with
their stability in neutral media is an advantage. Sweeney et al. [187]
suggested that they could be used in many applications such as processed
meats, marine products, jam, ice-cream and tomato ketchup. Their
stability in ethanol is a real advantage for colouring alcoholic beverages.
Hiroi [206] reported their application to sake and koji. Suzuki [207]
described their application to soy sauce and kamaboko. There is consider-
able commercial interest in the monascus group as shown by thirty-eight
patents issued in recent years [40].
9.9 Miscellaneous colorants
Several colorants from unusual sources have appeared in the patent

literature. For example, Iacobucci and Sweeney [208] reported the isola-
tion and properties of rubrolone (Figure 9.13), a red pigment produced by
the bacteria Streptomyces echinoruber. This colorant can be stabilized from
exposure to sunlight by addition of quercetin-5-sulphonate-5-sulphonate
and is suggested for use in dry beverage mixes for solution in water. Moll
and Farr [167, 168] reported a series of red colorants (Figure 9.13)
produced by reacting monascus with chitosan or monascorubin with the
hydrogen chloride salt of glucosamine. The corresponding water-soluble
compounds are substituted furoisoquinoline derivatives. A whole series of
compounds can be formed as described previously. Iacobucci and Sweeney
[209] reported the synthesis of a series of 3-deoxyanthocyanidins by
reduction of acetylated flavanones with sodium borohydride to the corres-
ponding flavan followed by oxidation with chloranil or bromanil. Yeowell
and Swearingen [210] described the preparations of a series of colorants
synthesized from benzyl alkyl compounds condensed with guanidine to
give benzylpyrimidine derivatives. The colorants are suggested for food,
textiles and medicines. Watanabe [211] reported the isolation of a red
pigment from the mould Penicillium purpurogenum. The isolation of
berberin, an isoquinoline type pigment, was reported by a Japanese
company Kakko Honsha [212]. Berberin can be extracted from the bark of
Phellodendron amurense, the roots of Coptis japonic or the xylem of
Berberis thunbergii. It also occurs in the leaves of P. amurense. The yellow
pigment is stable to heat and light. Koda [213] patented a potential blue
food colorant from the callus tissue of the plant Clerodendrum trichoto-
mum. Several potential colorants have also been reported in the general
literature. Hamburger et al. [214] reported a novel purple pigment from
the tree Dalbergia candenatensis, which occurred together with a mixture
of orange and red pigments from the heartwood. Achenbach [215]
described a new class of pigments of the flexirubin type in the bacteria
Flexibacter elegans. Sato and Hasegawa [216] reported the blue pigment
cyanohermidin in the roots of the plant Mercurialis leiocarpa. It resembled
allogachrome, a dyestuff used for silk and cotton in ancient Japan. Drewes
et al. [217], studying the medicinal plants of South Africa, reported four
new orange chalcone-type pigments in Brackenridgea zanguebarica. De
Rosa and De Stefano [218] reported a new guaiane sesquiterpene from the
fungus Lactarius sanguiftuus. Ueda et al. [219] studied the formation of a
purple pigment from cereals by fermentation. Jaegers et al. [220] clarified
the chemical structure of leucomelone, and protoleucomelone in the
fungus Boletopsis leucomelaena and the acetylated derivatives of cycloleu-
comelone, cyclovariegatin and atromentin in Anthracophylum species. The
orange pigments in the bacteria Streptomyces spp. C-72 were studied by

Rachev et al. [221] and the blue pigments in the bacteria Streptomyces
lavendula were described by Mikama et al. [222]. Oreshina et al. [223]
studied the biosynthesis of the pigments in strains of the mould Actinomy-
cetes (Streptoverticillium). Piskunkova et al. [224] and Torapova et al. [225]
reported the biosynthesis of the pigments in the fungus Hypomyces
rosellus. Brucker et al. [226] reported a broad spectrum of metabolites
including the pigments bikaverins and carotenoids in the mould Fusarium
moniliforme. In the mould Aspergillus oryzae, Manonmani and Shreek-
toniah [227] reported on orange red anthraquinone and Dahiya [228]
reported vismiaquinone and a flavone. Gill et al. [229] reported a series of
red and purple anthraquinones from the fungus Dermocybe austroveneta;
these included austrocorticin, austrocorticinic acid, austrocorticone, and a
series of related pigments [230]. They further reported the dermocanarins,
which are unique macrocylic lactones from Dermocybe canarea [231].
Suemitsu et al. [115,232,233] reported on HPLC analysis for macrosporin,
alters olano I A and alterporrioles A, B+C in the mould Alternaria porri.
These are anthraquinone-type pigments. Kondrateva and Moon [234]
reported a red pigment in the bacteria Arthrobacter. Etoh et al. [235]
reported the structures of the rhodonocardins produced by a Nocardia
species of bacteria. Gessner et al. [236] reported the synthesis of flaxin and
perlylyrine found in the j3-carboline group of pigments in sake and soy
sauce. Ito patented the use of flavins [127] and quinones [237] for use in
cosmetics. Sato et al. described a blue pigment in the mould Streptomyces
[238]. Anderson and Chung [239] reported the conversion of versiconal
acetate to versiconal and versicolorin C in extracts from the mould Asper-
gillus parasiticus. Nakamura and Homa [240] described prostaglandin-like
precursors of a red pigment formed during the autoxidation of methyl
arachidonate. Ozawa et al. [241] reported the isolation and identification of
a novel yellow j3-carboline pigment in salted radish roots (Raphanus sativus
L.). Yang et al. [242] patented the extraction method of a yellow pigment
from the roots of Glycyrrhiza. Cai et al. [243] patented the extraction for an
edible polyphenol colorant from bean seedcoats (Phaseolus spp.). Both a
brownish red and a cherry-red colorant could be produced. Aarnio and
Agathos [244] reported that variants of the mould Tolypocladium inflatum
produced white, red, orange and brown colours. This organism is used in
production of the immunosuppressive agent cyclosporin and apparently
the level of pigment is correlated to the level of cyclosporin produced. It
may be possible to prepare a colorant as a by-product of cyclosporin
production. Many of the above pigments were found in microorganisms
and may reflect the interest in metabolites that may have physiological
effects. Nevertheless, some of them may have potential as colorants, but
they would require toxicological clearance.
The availability of food colorants historically has depended on three

sources: specialized plant and animal sources, by-products, and chemical
synthesis. Today, a fourth should be added, that of tissue or cell culture
possibly with DNA transfer biotechnology.
Specialized plant and animal sources have been known for many
centuries. Examples are annatto, safflower, saffron, and turmeric from
plants and cochineal from insects. By-product sources are oenocyanins
from wine grape skins and carotenoids from many sources such as carrots
and palm oil. Chemical synthesis has provided the carotenoids, cantha-
xanthin, j3-carotene, j3-8-apo-carotenal, and ethyl j3-8-apo-carotenoate.
Another example would be the now discontinued linked polymer concept
with anthraquinone chromophores. The cell and tissue-culture concepts
have been of much research interest lately, possibly because of the
pharmaceutical and fermentation industries. The extensive literature on
the optimization of growth of the fungus Monascus for production of the
monascin colorants is one example. Nochida et al. [245] patented the
manufacture of the colorant canthaxanthin from the Rhodococcus species
of bacteria. Kumar and Lonsane [246] described the production of a novel
red pigment that is associated with the production of gibberellic acid from
the fungus Gibberella fujikuroi. The new pigment is not the normal indole
alkaloids, bikaverin, norbikaverin or o-dimethylanhydrofusarubin that are
ordinarily associated with in vivo growth of Gibberella. Sasaki et al. [247]
patented the production of pigments from the bacteria Serratia SSP-1
(Enterobacter). Ishiguro et al. [248] patented the production of a heat and
light stable blue-purple pigment from cultures of the bacterium lanthino-
bacterium lividium. Cioppa et al. [249] described the production of melanin
by a tyrosinase gene cloned into E. coli bacteria. Melanin is of considerable
interest as a general brown colorant, particularly in the pharmaceutical
industry for skin 'tanning' formulations.
Several reports have appeared on the tissue-culture applications of
plants for colorant and flavour applications, possibly due to the back-
ground of information available from plant propagation and plant-culture
developments. Nozue et al. [39] patented the production of the stable
acylated anthocyanins from the callus tissue of red sweet potatoes (Ipo-
moea batatas). Koda [250] patented the production of the acylated
anthocyanins from red cabbage (Brassica) by tissue culture. Other patents
involving anthocyanins are: callus culture of the beefsteak plant (Perilla)
by Ota [251J and Ota et al. [252, 253], a red plant (Basella rubra) by
Fukuzaki et al. [254], the red sepals of Hibiscus sabdariffa by Oohashi et al.
[255], and the evening primrose Oenothera [256]. Li and Zhu [257]
reported the formation of anthocyanins in panax ginseng. Two reports
involved the production of leucoanthocyanins by fermentation from wheat
germ [258] and barley [259]. Kargi and Friedel [260] reported the indole
alkaloids and related compounds produced in cell culture of the plant
Catharanthus roseus. Yomo et al. [167, 168] described the production of
carthamin and Saito et al. [261] reported a novel red pigment from cell
suspension cultures of Carthamus tinctorius. Three reports involved the
production ofcrocin from the stigmas of Crocus sativus [262-264]. Saffron
colorants are an obvious choice for cell culture because of their high price.
Kilby and Hunter [265] described the production of betanin from beetroot
(Beta vulgaris) by cell culture. The production of useful chemicals by
biotechnology has been reviewed by Trividi [266], Ilker [267], Hosono
[268], Whitaker and Evans [269], Fontanel and Tabata [270], Bomar and
Knopfel [271], Knorr et al. [272] and in Chapter 3 of this volume.
Plant products have been used in the pharmaceutical, cosmetic and food
industries for many years. The choice of supply by chemical synthesis or
plant biosynthesis depends on many factors but three are predominant.
First, some chemicals are difficult to synthesize chemically and possibly
could be obtained more easily by biosynthesis. One example in the
colorant area is the ternitins. These compounds, which are reported [3] to
be the most stable anthocyanins to date, have a complex acylated structure
that might be difficult to synthesize. Second, others contain too many
components. For example, jasmine with over 300 components and a
limited market is a poor candidate for chemical synthesis. Similarly, coffee
flavour has over 1000 components and may be a good candidate for tissue
culture. Grapes are reported to have as many as 30 anthocyanins [15],
whereas blueberries are reported to have 15 [273]. However, the number
of anthocyanins in a given plant is unlikely to be a deterrent for chemical
synthesis because not all components would be necessary for a good
colorant. A third important aspect is the ability to select strains or
conditions that allow much greater concentrations of the desired com-
pounds to be produced. For example, Fontanel and Tabata [270] reported
a WOO-fold increase in the production of berberine by Thalictrum minus
when tissue culture was used as opposed to the whole plant. Ilker [267]
reported that shikonen from Uthospermum roots showed an 845-fold
increase. With increases like this, cell culture approaches may be very
attractive. Some of the colorants may be good candidates for this emerging
area of biotechnology.
References
1. Yoshitama, K. Phytochem. 16 (1977) 1857.

2. Francis, F.J. Crit. Rev. Food Sci. Nutr. 28 (1989) 273.
(1990) 2921.
4. Goto, T. Proc. 5th Asian Symp. Med. Plants, Spices, Seoul, Korea (1984) p. 593.
5. Stirton, J.Z. and Harborne, J.B. Biochem. Syst. Ecol. 8 (1980) 285.
6. Idaka, E., Ogawa, T. and Kondo, T. Agricol. Bioi. Chern. 51 (1987) 2215.
7. Shi, Z., Lin, M. and Francis, F.J. J. Food Sci. 57 (1992) 761.
8. Shi, Z., Daun, H. and Francis, F.J. J. Food Sci. 58 (1993) t068.
9. Baublis, A.J. and Berber-Jimenez, M.D. J. Agric. Food Chern. 43 (1995) 640.
to. Francis, F.J. Analysis of Anthocyanins, in Anthocyanins as Food Colours (P. Markakis,
ed.), Academic Press, New York (1982) p. 181.
11. Brouillard, R. Flavonoids and Flower Culture. in The Flavonoids: Advances in Research
Since 1980. Ch. 16 (J.B. Harborne, ed.), Chapman & Hall, London (1988) p. 525.
12. Asen, S., Stewart, R.N. and Norris, K.H. Phytochem. 16 (1977) 1118.
13. Hoshimo, T., Matsumoto, U. and Goto, T. Phytochem. 19 (1980) 663.
14a. Francis, F.J. and Harborne, J.B. J. Food Sci. 31 (1966) 524.
14b. Francis, F.J., Draetta, I., Baldini, V. and Iaderoze, M. J. Amer. Soc. Hortic. Sci. 107
(1982) 789.
15. Singleton, V.L. and Esau, P. Phenolic Substances in Grapes and Wine and Their
Significance, Academic Press, New York (1969).
16a. Bassa, LA. and Francis, F.J. J. Food Sci. 52 (1987) 1753.
16b. Shi, Z., Bassa, LA., Gabriel, S.L. and Francis, F.J. J. Food Sci. 57 (1992) 755.
17. Teh, L.S. and Francis, F.J. J. Food Sci. 53 (1988) 1580.
18. Brouillard, R. Phytochem. 20 (1981) 143.
19. Dangles, O. Proceedings. First International Symposium on Natural Colorants. Amherst,
MA (1993) Hereld Organization, 200 Leeder Drive, Hamden, CT.
20. Saito, N. Personal communication (1990).
21. Goto, T., Imagawa, H., Kondo, T. and Miura, I. Heterocycles 17 (1982) 355.
22. Shi, Z., Lin, M. and Francis, F.J. J. Food Sci. 57 (1992) 766.
23. Min, L., Shi, Z. and Francis, F.J. J. Food Sci. 57 (1992) 768.
24. Asen, S., Stewart, R.N. and Norris, K.H. US Patent No.4 172 902 (1979).
25. K.K. Seiwa Kazuku Co. Japan Patent No. 82058871 (1982).
26. Sanei Chemical Co. Japan Patent No. 809 025 460 (1980).
27. Sanei Chemical Co. Japan Patent No. 59223756 (1984).
28. Idaka, H. Japan Patent No. 87209 173 (1987).
29. Idaka, H. Japan Patent No. 63 113 078 (1988).
30. Idaka, H. Japan Patent No. 63 278 971 (1988).
31. Mitsuta, H., Wada, S. and Nishino, Y. Japan Patent No. 86224979 (1986).
32. Obata, M., Koda, T. and Yasuda, A. Japan Patent No. 8703765 (1987).
33. Obata, M., Koda, T. and Yasuda, A. Japan Patent No. 8703775 (1987).
34. Yasuda, A., Koda, T. and Obata, M. Japan Patent No. 61282032 (1986).
35. Katake, K. and Yasuda, A. Japan Patent No. 8703739 (1987).
36. Katake, K. and Yasuda, A. Japan Patent No. 8703741 (1987).
37. Idaka, H. Japan Patent No. 89224389 (1988).
38. Idaka, H. Japan Patent No. 63 ItO 259 (1988).
39. Nozue, M., Kikuma, M., Miyamoto, Y., Fukuzaki, E., Motsumura, T. and Hashimoto,
Y. Japan Patent No. 88233993 (1988).
40. Francis, F.J. Handbook of Food Colorant Patents, Food and Nutrition Press, Westport,
CT (1986).
41. Murai, K. and Wilkins, D. Food Technol. 44 (1990) 131.
42. LaBell, F. Food Processing 51 (1990) 69.
43. Rozinkov, P.F., Gavilov, F.M. and Klochkova, N.V. USSR Patent No. 343987 (1972).
44. Guimares, I.S., Barbosa, A.L.S. and Nassariani, G. Rev. Bras. Eng. Quin 12 (1989) 22.
45. Rouseff, R.L. J. Food Sci. 53 (1988) 1823.
46. Reith, J.F. and Gielan, J.W. J. Food Sci. 36 (1971) 861.
47. Todd, P.H. US Patent No.3 162538 (1964).
48. Najar, S.V., Bobbio, F.O. and Bobbio, P.A. Food Chern. 29 (1988) 283.
49. Heltiarachchy, N.S. and Muffett, D.J. European Patent No. 200043 (1986).
50. Ford, M.A. and Mellor, C. GB Patent No.2 190922 (1987).
51. Berset, C. and Marty, C. Lebensmittel-Wissenschaft und -Technologie 19 (1986) 126.
52. Ford, M.A. and Draisey, A.H. GB Patent No.2 188823 (1987).
53. Ikawa, Y. and Kagemoto, A. Japan Patent No. 75 t04231 (1975).
54. Mori, K., Tonari, K. and Nishiura, Y. Japan Patent No. 75 113 533 (1975).
55. Harrison, R.M. GB Patent No.2 186 175 (1987).
56. Collins, P. Proceedings. First International Symposium on Natural Colorants. Amherst,
MA (1993).
57. Alonso, G.L., Varon, R., Gomez, R., Navarro, F. and Salinas, M.R. J. Food Sci. 55
(1990) 595.
58. Corradi, C. and Micheli, G. Boll. Chim. Farm. 118 (1979) 553.
59. Solinas, M. and Cichelle, A. Industries Alimentari 27 (1988) 634.
60. Klaui, H., Manz, U., Rigassi, N., Ryser, G. and Schweiter, U. German Patent No.
2053381 (1971).
61. Klaui, H., Manz, N.R., Gottleib, R. and Schweiter, R. US Patent No. 39 223 68 (1975).
62. Hashimoto, H. Proceedings. First International Symposium on Natural Colorants.
Amherst, MA (1993).
63. Garrido, J.L., Diez de Bethancourt, C. and Revilla, E. Anales de Bromatologia 39
(1987) 69.
64. Umetani, Y., Fukui, H. and Tobata, M. Yakugoku Zarshi 100 (1980) 920.
65. Yoshizumi, S., Okuyoma, H. and Toyoma, R. Shouhuhin Kogyo 23 (1980) 41.
66. Gunatilaka, A.A.L. and Sirimanni, S.R. J. Chern. Res. Synop. 7 (1979) 21.
67. Gunatilaka, A.A.L., Sirimone, S.R., Sothersworan, S. and Sreyani, H.T.B. Phyto-
chern. 21 (1982) 805.
68. Ishara, N. and Arichi, S. Wakanayaku Shinpojumu 14 (1981) 45.
69. Uesato, S., Ueda, S., Kobayashi, K. and Inouye, H. Proc. Int. Congo Plant Tissue Cell
Cult. 5th edn. (A. Fujiwara, ed.) (1982).
70. Uesato, S., Ueda, S. and Kobayashi, K. Chern. Pharm. Bull. 31 (1983) 4185.
71. Uesato, S., Uedo, S., Kobayashi, K., Miyauchi, M. and Inouye, H. Tetrahedron Lett. 25
(1984) 573.
72. Terao, T., Ohashi, H. and Mizukomi, H. Plant Sci. Lett. 33 (1984) 47.
73. Mizukomi, H., Terao, T., Miura, H. and Ohaski, H. Phytochem. 22 (1983) 679.
74. Tobata, M., Umetani, Y., Shima, K. and Tanaka, S. Plant Cell Tissue Organ Culture 33
(1984) 3.
75. Yoshizumi, S., Okuyoma, H. and Toyoma, R. Fureguransum Janaru 8 (1980) 104.
76. Aburoda, M. Gendai Toyo lzaku. 4 (1982) 42.
77. Kuwano, S. Gendai Toyo Izaku. 4 (1983) 55.
78. Joshi, K.C., Singh, P. and Pardosoni, R.T. J. Ind. Chern. Soc. 56 (1979) 327.
79. Yan, Y. Yaowa Fenxi Sazhi 4 (1984) 227.
80. Inouye, H. Gendai Toyo Izaku. 4 (1983) 48.
81. Wong, H.C. and Koehler, P.E. J. Food Sci. 46 (1981) 956.
82. Wong, H.C. and Koehler, P.E. J. Food Sci. 48 (1983) 1200.
83. Hagikara, K., Yamauchi, K. and Kuwano, S. Mukogawa Joshi Daigaku Kiyo.
Yakugaku Hen. 29 (1981) 2.
84. Noda, N., Yomada, S., Hayakawa, J. and Uno, K. J. Hygiene. Chern. 29 (1983) 7.
85. Floyd, A.G. Food Chern. 5 (1980) 19.
86. Ishiguro, K., Yamaki, M., Takagi, S. J. Nat. Prod. 46 (1983) 532.
87. Komura, K., Okayama, H., Toyoma, R. Sawada, Y. and Ichinose, S. Japan Patent No.
76 006 230 (1976).
88. Hasegawa, K. Japan Patent No. 79 152026 (1979).
89. Koga, K., Fujikawa, S. and Fukui, Y. European Patent No. 251 0639 (1988).
90. Koga, K., Fujikawa, S. and Fukui, Y. US Patent No.4 878 921 (1989).
91. Tomara, I., Komai, T. and Hotae, T. Japan Patent No. 61 296070 (1986).
92. Fujikawa, S., Fukui, Y., Koga, K. and Kumada, J. J. Ferment. Technol. 65 (1987) 419.
93. Fujikawa, S., Nakamura, S., Koga, K. and Kumada, J. J. Ferment. Techno/. 65 (1987)
711.
94. Toyama, R. and Moritome, N. New Food Indust. 30 (1988) l3.
95. Riboh, M. Food Eng. 49 (1977) 66.
96. Schul. I.J. Proceedings. First International Symposium on Natural Colorants. Amherst,
MA (1993).
97. Donkin, R.A. Trans. Amer. Philosoph. Soc. 67 (1977) Part 5.
98. Valdes-Martinez, S.E., Garcia-Casanova, A., Zaragoza-Santamaria, L., Reyes-Quiroz,

R., Sanchez Ortega, P. and Valle-Vega, P. Proceedings. First International Symposium
on Natural Colorants. Amherst, MA (1993).
99. Coulson, J. Developments in Food Colours (1. Walford, ed.), Applied Science, London
(1980) p. 189.
100. Gutteridge, J.M.C. and Quinlin, G.J. Food Add. and Contam. 3 (1986) 289.
101. Yamada, J., Tomita, Y. and Fukude, K. Agricol. Bioi. Chern. 52 (1988) 2893.
102. Ford, G.P., Gopal, T., Grant, D., Gaunt, 1.F., Evans, J.G. and Butler, W.H. Food and
Chern. Toxicol. 25 (1987) 897.
103. Ford, G.P., Stevenson, B.1. and Evans, J.G. Food and Chern. Toxicol. 2S (1987) 919.
104. Grant, D. and Gaunt, 1.F. Food and Chern. Toxicol. 25 (1987) 903.
105. Grant, D., Gaunt, 1.F. and Carpanini, F.M.B. Food and Chern. Toxicol. 25 (1987) 913.
106. Phillips, J.C., Bex, c., Walters, D.G. and Gaunt, 1.F. Food and Chern. Toxicol. 25
(1987) 927.
107. Ghosh, A. and Chakravarti, 1. J. Food Sci. Tech., India 25 (1988) 94.
108. Schwing-Weill, M.J. Analusis 14 (1986) 290.
109. Richardson, W.H., Schmidt, T.M. and Nealson, K.H. Appl. Environ. Microbiol. 54
(1988) 1602.
110. Mukherjee, K.S., Bhattachargee, P., Mukherjee, R.K. and Ghosh, P.K. J. Indian
Chern. Soc. 63 (1986) 619.
111. Sakuma, Y., Tanaka, J. and Hiza, T. Aust. J. Chern. 40 (1987) 1613.
112. Ali, A.A., Abdalla, O.M. and Steglick, W. Phytochem. 28 (1988) 281.
113. Ikenaga, T., Kibuta, S., Mimura, K. and Ohashi, H. Shoyakugaku Zasshi 40 (1986) 397.
114. Nedentsov, A.G., Baskunov, B.P. and Akimento, V.K. Biokhimiya (Moscow) 53
(1988) 413.
115. Suemitsu, R., Horiuchi, K., Ohnishi, K. and Yanagowase, S. J. Chromat. 454 (1988)
406.
116. Billen, G., Karl, U., Scholl, T. and Stroech, K.D. Chap. 3 in Natural Products
Chemistry (A. Rahmon and P.W. Le Quesne, eds), Springer Verlag, Berlin (1988)
p.305.
117. Parisot, D., Devys, M. and Barbier, M. Microbios 64 (1990) 31.
118. Kasumov, M.A., Amirova, G.S., Alekperov, U.K. and Musaev, V.R. USSR Patent No.
988845 (1983).
119. Tokano, M. and Konishi, T. Japan Patent No. 76 124 120 (1976).
120. Sanei Kaguku Kogyo Co. Japan Patent No.4 838 154 (1973).
121. Daisho Co., Ltd. Japan Patent No. 59 120075 (1984).
122. Shiaki, Y. Japan Patent No. 71 002 386 (1971).
123. Sanei Chemical Co. Japan Patent No. 81 139561 (1981).
124. Eguchi, A., Suzuki, Y., Oonuki, T., Hirata, T., Tosaka, O. and Yokokawa, Y. Japan
Patent No. 76 050 938 (1976).
125. Eguchi, A., Suzuki, Y., Onuma, T., Yokokawa, Y., Hirata, T. and Tosaka, O. Japan
Patent No. 52084220 (1977).
126. Rikio, G. and Takanoshi, M. Japan Patent No. 77 108061 (1977).
127. Ito, K. Japan Patent No. 61 225257 (1986).
128. Asahi, Y. and Domae, A. Japan Patent No. 62 151 467 (1987).
129. Crompton and Knowles Co. Japan Patent No. 62 131 071 (1987).
130. Baldwin, W.L. US Patent No.4 769 246 (1988).
131. Fukazawa, T. Japan Patent No. 70024 672 (1970).
132. Verghese, J. and Joy, M.T. Flavour Fragrance J. 4 (1989) 31.
133. Govindarajan, V.S. Crit. Rev. in Food Science and Nutrition, CRC Press, Boca Raton,
Florida (1980) 199.
134. Tonneson, H.H., Karlsen, J.K., Adhikary, S.R. and Pandey, R. Zeitschrift fur
Lebensmitteluntersuchung und Forschung 189 (1989) 116.
135. Sampathee, S.R., Krishnamurthy, N., Sowbhagya, H.B. and Shankaranarayana, M.L.
J. Food Sci. Technol. (India) 25 (1988) 152.
136. Iyer, V.S. Proceedings. First International Symposium on Natural Colorants. Amherst,
MA (1993).
137. Leshnik, R.R. US Patent No.4 307 117 (1981).
138. Leshnik, RR European Patent No. 37 204 (1982).

139. Schrantz, J.L. US Patent No.4 368 208 (1983).
140. Obata, S., Ohaski, S. and Hanatoku, M. Japan Patent No. 73014941 (1973).
141. Maing, Y. and Miller, I. European Patent No. 25637 (1981).
142. Maing, Y. and Miller, I. US Patent No.4 263 333 (1981).
143. Shinagawa, K., Nagai, S. and Kokagoshi, T. Japan Patent No. 79 163866 (1979).
144. Sanyo-Kokusaku Pulp Co. Japan Patent No. 56019220 (1981).
145. Goldscher, K. US Patent No.4 163 803 (1979).
146. Tonneson, H.H. and Karlsen, J.K. J. Chromat. 259 (1983) 367.
147. Ogbeide, O.N., Eduaveguavoen, 0.1. and Parvez, M. Pakistan J. Sci. 37 (1985) 15.
148. Shah, RG. and Netrawadi, M.S. Bull. Envir. Contam. Toxicol. 40 (1988) 350.
149. Nagabhushan, M., Nair, U.J., Amonkar, A.J., D'Souza, A.V. and Bhide, S.V. Mut.
Res. 202 (1988) 163.
150. El Gazzar, F.E. and Marth, E.H. Lebensmittel-Wissenschaft und Technologie 22 (1989)
406.
151. Takahashi, Y., Saito, K., Yanagiya, M., Ikura, M., Hikichi, K., Matsumoto, T. and
Wada, M. Tetrahed. Lett. 25 (1984) 2471.
152. An, X., Li, Y., Chen, J., Li, F., Fang, S. and Chen, Y. Zhongcaoyao 21 (199O) 188.
153. Saito, K., Takahashi, Y. and Wada, M. Biochem. Biophys. Acta 756 (1983) 217.
154. Saito, K. and Fukushima, A. Acta Soc. Bot. Pol. 55 (1986) 639.
155. Saito, K. and Fukushima, A. Food Chem. 26 (1987) 125.
156. Saito, K. and Takahashi, Y. Acta Soc. Bot. Pol. 54 (1985) 231.
157. Tokai University, Japan Patent No. 16346 (1970).
158. Kokusaku Pulp Co. Japan Patent No. 57025576 (1982).
159. Wada, M. Japan Patent No. 70 016 346 (1970).
160. Umeda, A. and Inoue, K. Japan Patent No. 70 040 266 (1970).
161. Fukushima, A., Hase, H. and Saito, K. Acta Soc. Bot. Pol. 56 (1987) 485.
163. Saito, K. Food Chem. 36 (1990) 243.
165. Sanyo-Kokusaku Pulp Co. Japan Patent No. 80 102 659 (1980).
166. Sanyo-Kokusaku Pulp Co. Japan Patent No. 80 102 661 (1980).
167. Yomo, H., Miyano, S. and Sekino, Y. Japan Patent No. 63253386 (1987).
168. Yomo, H., Miyano, S. and Sekino, Y. Japan Patent No. 62257390 (1987).
169. Daimon, E., Wakayama, Y., Sekino, Y. and Yomo, H. Japan Patent No. 0120092
(1989).
170. Saito, Y. and Hasegawa, M. Kobunkazai no Kagagu 31 (1988) 18.
171. Wakayama, S. New Food Industries 30 (1988) 7.
172. Onishi, K. New Food Industries 29 (1987) 22.
173. Lin, C.F. and Iizuka, H. Appl. Envir. Microbial. 43 (1982) 671.
174. Fielding, B.C., Holker, J.S.E., Jones, D.F., Powel, A.D.G., Richmond, K.W.,
Robertson, A. and Whalley, W.B. J. Chem. Soc. (1961) 4579.
175. Manchard, P.S. and Whalley, W.B. Phytochem. 12 (1973) 2531.
176. Moll, H.R. and Farr, D.H. US Patent No.3 993 789 (1976).
177. Moll, H.R. and Farr, D.H. German Patent No. 606433 (1978).
178. Moll, H.R. and Farr, D.H. Switzerland Patent No. 606 433 (1978).
179. Yamaguchi, Y., Ito, H., Watanabe, S., Yoshida, T. and Kumatsu, A. Japan Patent No.
3765906 (1973).
180. Kawabato, S. and Sato, K. Japan Patent No. 76099519 (1976).
181. Broder, C.U. and Koehler, P.E. J. Food Sci. 45 (1980) 567.
182. Nakagawa, N., Watenabe, S. and Kobayoshi, J. Japan Patent No. 76091937 (1976).
183. Nakagawa, N., Watenabe, S. and Kobayoshi, J. Japan Patent No. 76091939 (1976).
184. Nakagawa, N., Watenabe, S. and Kobayoshi, J. Japan Patent No. 76091 938 (1976).
185. Tsunenaga, T. and Onoe, A. Japan Patent No. 75036519 (1975).
186. Yoshimura, M., Yamanaka, S. and Hirose, Y. Japan Patent No. 76 130428 (1976).
187. Sweeney, J.G., Estrada-Valdez, M.C., Iacobucci, G.A., Sato, H. and Sakamura, S. J.
Agric. Food Chem. 29 (1981) 1189.
188. Han, O.H. Optimization of Monascus Pigment Production in Solid-State Fermentation.

PhD Thesis. Univ. Massachusetts, Amherst (1990).
189. Han, O.H. and Mudgett, R.E. Inst. Food Technologists Ann. Mtg. Abs (1991) 156.
190. Han, O.H. and Mudgett, R.E. Biotechnol. Prog. 8 (1992) 5.
191. Kim, C., Rhee, S. and Kim, I. Korean J. Food Sci. and Technol. 9 (1977) 277.
192. Shepherd, D. and Carels, M.S.C. Swiss Patent No. 606 421 (978).
193. Tadao, H., Suzuki, T., Tsukioka, M. and Takahashi, T. Japan Patent No. 81006 263
(1981).
194. Hiroi, T., Shima, T., Suzuki, T., Tsukioka, M. and Takahashi, T. Japan Patent No.
75 025 766 (1975).
195. Hiroi, T., Shima, T., Suzuki, T. and Tsukioka, M. Agric. BioI. Chem. 43 (1979) 1975.
196. Lin, C.F. J. Ferm. Tech. 53 (1973) 407.
197. Su, Y.C., Dhen, W.L. and Lee, Y.H. Res. Rep. Coli. Agri. Nat. Taiwan Univ. 14 (1973)
41.
198. Ryu, B.H., Lee, B.H., Park, B.G., Kim, H.S., Kim, D.S. and Lee, J.H. Korean J.
Food Sci. Technol. 21 (1989) 31.
199. Ryu, B.H., Lee, B.H., Park, B.G., Kim, H.S., Kim, D.S. and Roh, M.H. Korean J.
Food Sci. Technol. 21 (1989) 37.
200. Fink-Gremmels, J. and Leistner, L. Mitteilungsblatt der Bundesanstalt fur Fleischfor-
schung, Kulmbach 101 (1988) 8073.
201. Fink-Gremmels, J. and Leistner, L. Fleischwirtschaft 69 (1989) 115.
202. Wong, H.C. and Bau, Y.S. Wert. Plant Physiol. 60 (1977) 578.
203. Su, Y.c. and Huang, J.H. Proc. Nat. Sci. Council 4 (1980) 201.
204. Kumasaki, S., Nakanishi, K., Nishikawa, E. and Ohashi, M. Tetrahed. 18 (1962) 1171.
205. Gan, C. Kexueban 4 (1988) 63.
206. Hiroi, T. New Food Indust. 30 (1988) 1.
207. Suzuki, H. New Food Indust. 30 (1988) 20.
208. Iacobucci, G.A. and Sweeney, J.G. US Patent No.4 285 985 (1981).
209. Iacobucci, G.A. and Sweeney, J.G. British Patent No.1 552402 (1979).
210. Yeowell, D.A. and Swearingen, R.A. German Patent No.2 725 992 (1977).
211. Watanabe, H. Japan Patent No. 74093587 (1974).
212. Kakko Honsha, K.K. Japan Patent No. 80 155056 (1980).
213. Koda, T. Japan Patent No. 62 83 892 (1987).
214. Hamburger, M., Cordell, G.A., Ruangrungsi, N. and Tantivatana, P. J. Org. Chem. 53
(1988) 4161.
215. Achenbach, H. Prog. Chem. Org. Nat. Prod. 52 (1988) 73.
216. Sato, Y. and Hasegawa, M. Kobunkagai no Kagagu 31 (1986) 18.
217. Drewes, S.E., Hudson, N.A., Bates, R.B., Linz, G.S. J. Chem. Soc. (Perkin Trans.) 12
(1987) 2809.
218. De Rosa, S. and De Stefano, S. Phytochem. 26 (1987) 2007.
219. Ueda, S., Koba, Y. and Ohba, R. Hakko Kogaku Kaishi 65 (1987) 237.
220. Jaegers, E., Hillen-Maske, E., Schmidt, H., Steglich, W. and Horan, E. Z. Naturforsch
B: Chem. Soc. 42 (1987) 1354.
221. Rachev, R., Sakanekova, M., Darakchieva, M., Tsbetkova, R., Gesheva, R., Puntsag,
T. and Tsetseg, B. Acta Microbiol. Bulg. 21 (1987) 57.
222. Mikama, Y., Yazawa, K., Yokoyama, K., Takahashi, K. and Arai, T. Symp. BioI.
Hung. 32 (1986) 297.
223. Oreshina, M.G., Penzikova, G.A. and Bartoshevich, Y. Antibiot. Khimioter 33 (1988)
323.
224. Piskunkova, N.F., Maksimov, V.N., Toropova, E.G., Egorov, N.S. and Mardamshina,
A.D. Bioi. Nauki (Moscow) 2 (1988) 84.
225. Torapova, E.G., Mardomshina, A.D. and Piskunkova, N.F. Antibiot. Khimioter 33
(1988) 96.
226. Brucker, B., Blechsmidt, D. and Schubert, B. Zentralblattfur Mikrobiologie 144 (1989)
3.
227. Mananmani, H.K. and Shreektoniah, K.R. J. Food Sci. Tech. India 21 (1984) 195.
228. Dahiya, J.S. Ind. J. Microbiol. 27 (1987) 12.
229. Gill, M., Gimenez, A. and McKenzie, R.W. J. Nat. Prod. 51 (1988) 1255.
230. Gill, M. and Gimenez, A. J. Chern. Soc. (Perkin Trans.) 4 (1990) 1159.
231. Gill, M. and Gimenez, A. Tetrahed. Lett. 31 (1990) 3503.
232. Suemitsu, R., Ohnishi, K., Yanagawase, S., Yanamoto, K. and Yamada, Y. Phyto-
chern. 28 (1989) 1621.
233. Suemitsu, R., Sakurai, Y., Nakachi, K., Miyoshi, 1., Kubota, M. and Ohnishi, K. Agric.
Bioi. Chern. 53 (1989) 1301.
234. Kondrateva, L.M. and Moon, T.H. Microbiologiya 55 (1987) 1010.
235. Etoh, H., Iguchi, M., Nagasawa, T., Tani, Y., Yanada, H. and Fukomi, H. Agric. Bioi.
Chern. 51 (1987) 1819.
236. Gessner, W.P., Brossi, A., Bembenek, M.E. and Abell, C.W. Archiv. der Pharrnazie
321 (1988) 95.
237. Ito, K. Japan Patent No. 61 225 259 (1986).
238. Sato, K., Nihira, T., Sakuda, S., Yanagimoto, M. and Yamada, Y. J. Ferment.
Bioengin. 68 (1989) 170.
240. Nakamura, T. and Homa, Y. J. Agric. Food Chern. 36 (1988) 15.
241. Ozawa, Y., Kawakishi, S., Uda, Y. and Maeda, Y. Agric. BioI. Chern. 54 (1990) 1241.
242. Yang, B., Chang, X. and Pan, X. Chinese Patent No.1 033 459 (1989).
243. Cai, C., Dong, H., Dong, X. and Dong, T. Chinese Patent No.1 033459 (1989).
244. Aarnio, T. and Agathos, S. Appl. Microb. Biotechnol. 33 (1990) 435.
245. Nochida, K., Uejima, T., Ochiai, K. and Kawamoto, 1. Japan Patent No. 02138996
(1990).
246. Kumar, P.K.R. and Lonsane, P.K. Appl. Microbiol. Biotech. 28 (1988) 537.
247. Sasaki, M., Bandai, S. and Kamiyama, H. Japan Patent No. 63255293 (1988).
248. Ishiguro, Y., Morita, K. and Ito, Y. Japan Patent No. 63245666 (1988).
249. Cioppa, G., Garger, S.J., Sverlow, G.G., Turpen, T.H. and Grill, L.K. Biotechnol. 8
(1990) 634.
250. Koda, T. Japan Patent No. 62 181 796 (1987).
251. Ota, S. Gekkan Fudo Kernikaree 4 (1988) 55.
252. Ota, S., Tamaoki, Y. and Niwada, S. Japan Patent No. 63 156865 (1988).
253. Ota, S., Tamaoki, Y., Niwada, S. and Kawachi, M. Japan Patent No. 64 002 593 (1989).
254. Fukuzaki, E., Miyamoto, Y., Kikuma, M., Matsumura, T. and Hashimoto, Y. Japan
Patent No. 63 109786 (1988).
255. Oohashi, Y., Mizokami, H., Tomita, K_, Hiraoka, N. and Fugimoto, K. Japan Patent
No. 01 030 594 (1989).
256. Yamada, o. and Fugita, T. Japan Patent No. 62210 992 (1987).
257. Li, S. and Zhu, W. Zhiwu Xuebao 32 (1990) 103.
258. Ueda, S. Japan Patent No. 63230774 (1988).
259. Ohba, R., Koba, Y. and Ueda, S. Hakko Kogoku Kaishi 65 (1987) 507.
260. Kargi, F. and Freidel, 1. Biotech. Lett. 10 (1988) 409.
261. Saito, K., Daimon, E., Kusaka, K1., Wakayama, S. and Sekino, Y. Z. Naturforsch C.
Biosci. 43 (1988) 862.
262. Himeno, H. and Sano, K. Agric. and Bioi. Chern. 51 (1987) 2395.
263. Fakhrai, F. and Evans, P.K. J. Exper. Bot. 41 (1990) 47.
264. Sarama, K.S., Maesato, K., Hara, T. and Sonoda, Y. J. Exper. Bot. 41 (1990) 745.
265. Kilby, N.J. and Hunter, C.S. Appl. Microbiol. Biotechnol. 33 (1990) 448.
266. Trividi, N. Biotechnology of Food Processing (S. Harlander and T.P. LaBuza, eds),
Noyes Data Corp., Park Ridge, New Jersey (1986) pp. 15-132.
267. I1ker, R. Food Technol. 41 (1987) 70.
268. Hosono, T. Up-to-date Food Proc. 22 (1987) 674.
269. Whitaker, R.J. and Evans, D.A. Cherntech. 17 (1987) 93.
270. Fontanel, A. and Tabata, M. Nestle Research News (1987) 93.
271. Bomar, M.T. and Knopfel, S.A. Biosynthesis of Food Colorants by Microorganisf.1s
(T. Kutsch, ed.), in Research in the Service of Nutrition: A Tribute. German Federation
Research Inst. for Nutrition, Prague (1989) p. 253.
272. Knorr, D., Beaumont, M.D., Castor, C.S., Dornenburg, H., Gross, B., Pandya, Y. and
Romagnoli, L.G. Food Technol. 44 (1990) 71.
273. Francis, F.J., Harborne, J.B. and Barker, W.G. J. Food Sci. 31 (1966) 583.
Index
absorbance spectra 48, 53, 60, 65, 69, applications 51,57,59,63,64,72,

164-166,173,181,185,186,191, 228-230
219-224, 276, 313 Aralia cordata 279
acceptable daily intake 114-116, 192 Armenian red 65, 321
Actinidia equina 230 Arthrobacter 333
actinioerythrol 200, 222, 230, 233 Arthrospira maxima 175
Actinomycetes 333 Ascomycetes 30, 31
ADI see acceptable daily intake Aspergillus spp. 103, 294, 319, 333
Adonis spp. 102 astaxanthin 91, 95, 97, 101, 200, 210,
Agaricales 31 233
Agaricus bisporus 32 aubergine 53, 253
Agrobacterium rhyzogenes 82, 89 aurantinidin 248
Ajuga reptans 85 aurones 267
Albatrellus spp. 31
alcoholic drinks see drinks Bacillariophyta 141
alfalfa 146 Bacillus spp. 98, 319
algae 11, 27, 76, 82, 93-98, 122, bacteria 37, 46, 87, 200
170-190,200 bacteriochlorophy\ls 141, 148
Alkanna tinctoria Tausch. 322 bakery products see confectionery
Allium cepa 252 Basella rubra 282, 334
Basidiomycetes 31
allomelanin 16
beetroot 18,59-64,67,281-285,335
allophycocyanin 177-185
benzoic acids 267
Alternaria porri 324, 333
benzoquinone 15
Amanita muscaria 18, 32, 283, 284
Berberis thumbergii 332
amaranth 47,61 betacyan(id)in 18, 59-64
amaranthin 282 betalains 17,27,59-64,85-87,125,
Amaranthus spp. 18, 282, 285 280-296
amino acids 267 betan(id)in 18,47,84, 103
amphibians 28 Beta vulgaris 18, 281-285, 335
Anabaena spp. 175 betaxanthin 18, 59-64
Anabaenopsis spp. 175 beverages see drinks
Anacystis nidulans 175, 177, 178 bilberries 278
animals 200,209 bile pigments see bilins
annatto 47-52, 63, 72, 84, 101, 197-239, bilins 170-193
315 bilirubin 162
antheraxanthin 200, 232, 233 biliverdin 162
anthocyanin(id)ins 14, 27, 53-59, 67, bioprocessing 98-103
85-88, 120, 244-296 biotechnology 80-108, 188,231
Anthracophylum 332 birds 28, 200, 229
anthraquinones 15, 321-325 Bixa orellana 27,47,84,85, 101,201,
Aphanizomenon ftos-aquae 175 212, 315
Aphanocapsa spp. 175 bixin 47-52, 72, 85, 197-239
Aphanothece spp. 175, 190 blackberry 253, 314 see also Rubus spp.
Aphidoidea 321 blackcurrant 53, 55, 253 see also Ribes
apigeninidin 248 spp.
Aplysia spp. 172 Blakeslea trispora 91, 231
apo-carotenoids 121, 199 blood orange 252
apples 53, 252 blueberry 253,278 see also Vaccinium
application forms 45,49, 60, 66, 70, 227 spp.
344 INDEX
Boletales 33 Chlorophyta 141

Boletopsis leucomelaena 332 Chlorosplenium spp. 31,32
Bougainvillea glabra 282 cholesterol 119
bougainvillein 282 Chondrus crispus 193
Brackenridgea zanguebarica 332 Chromatiaceae 141
Brassica spp. 252, 334 Chromobacterium spp. 37
Brevibacterium spp. 91,97,232 chromones 34
brilliant blue FCF 191-192 Chroogomphus spp. 33
Bromeliacceae 310 Chroomonas spp. 141
Bulgaria spp. 31 Chrysophyta 141
Butea monosperma 321 cinnamic acids 267
butter see dairy products cinerarin 311
cis-bixin see bixin
cis-norbixin see bixin
cake mix see confectionery citraxanthin 200, 222, 233
Calothrix spp. 175 Citrus sinensis 252
cancer 118, 230 Clerodendrum trichotomum 332
Candida utilis 103, 294 Clitoria tematia 266, 278, 310
candies see confectionery cloning 85
Cantharellus spp. 31,91 Clostridium tetanomorphum 177
canthaxanthin 91,97, 121, 197,200, Coccochloris elabens 182
222, 229, 233 Coccoidea spp. 321
capensinidin 248 Coccus cacti L. 64
capsaicin 85 cochineal ~9, 123, 320
capsaxanthin 200, 222, 233 cocoa 125
Capsicum spp. 77, 85, 103, 211 colour blindness 5
capsorubin 77, 200, 222, 233 Comanthus parvicirrus 324
caramel 2, 77, 124 Commelinaceae 310
carbon black 78, 125 Compositae 19,310
carcinogens 113 confectionery 52, 58, 64, 68, 73, 76,
carmine ~9, 123 227,322
carminic acid ~9 co-pigmentation 267
carmoisine 61 copper chlorophylls 75
carotenes 12,27,76-77,86,91,99, 100, Coprinus spp. 33
105, 120, 197-239 Coptis japonica 332
carotenodermia 210 com 253
carotenoids 76, 197-239 coronary heart disease 119
Carpobrotus ancinacriformis 281 cosmetics 65
carrot 76, 83, 85-88, 100 coumarin 267
carthamin 327-328 cranberry 55,253,278
Carthamus spp. 327, 335 crocetin 200,222,236
Cassia mimosoides 324 crocin 77, 85
cattle 209 Crocus spp. 27,85,212,317,335
Cedrela spp. 317 Cryptomonas spp. 175
cell culture 88-90 Cryptophyta 141
Celosia cristata 282 cryptoxanthin 200,223,236
celosianin 282 Curcuma spp. 27,69,325
Centrospermae 18, 32 curcumin 47,69-74,76
chalcone 258 cyanidin 53, 248
Chenopodium spp. 282 Cyanidium caldarium 171, 175, 179, 184
cherry 53, 55, 252 Cyanophyta 141
Chlorella spp. 91, 93 cyanobacteria see Cyanophyta
chlorins see chlorophylls Cyphomandra betacea 252
Chlorobiaceae 141 cytochromes 165
Chloroflexaceae 141
Chlorogloea fritschii 175
chlorophyllide 134 Dacrymycetales 31
chlorophyllin 134 Dactylopiidae 321
chlorophylls 27,74, 125, 131-155 Dactylopius coccus costa 16,64, 321
INDEX 345
dairy products 51,58,60,63,68,72,76, flavonoids 13, 320

77,227,228 flavonol 267
Dalbergia candenatensis 332 Flexibacter elegans 332
Daldinia spp. 31 fluorescence 173, 181
Daucus carota 76, 83, 85-88, 100, 279 Fornes fomentarius 33
delphinidin 53 Food and Drug Administration 43, 44,
demethoxycurcumin 69 115-118
depsides 34 Fragaria spp. 252
Dermocarpa violacea 175 Fremyella diplosiphon 175, 181
Dermocybe spp. 32, 333 frozen products 59, 73
desserts 68 fungi 28, 200
diferoylmethane 69 Fusarium spp. 324, 333
dihydrochalcone 267
Dioscorea alata 252
Discomycetes see Ascomycetes Gardenia spp. 77, 125,317
dopaxanthin 283 genetic engineering 83, 105
drin~ ~,~,M,~,~,77,n1 Gentianaceae 310
duhat 278 gentiodelphin 311
Dunaliella spp. 38,77,93,94, 104, 122, Gibasis geniculata 314
188, 230, 231 Gibberella spp. 334
dry mixes see mixes Gladiolus segetum 324
Glaucosphaera vacuolata 175
Glottiphyllum longum 283
eggs 228 Glycine maxima 252
Eichhornia crassipes 146 Glycyrrhiza spp. 333
Elaeis guineensis 212 Gomphrena globosa 282
elderberry 55 gomphrenin 282
Enterobacter spp. 334 grape(s) 53-59, 253, 314
E numbers 43 grape extracts 53--59
Erwinia spp. 204, 232 grape skin 53--59
europinidin 248
erythrosine 191-192
Escherichia coli 105, 204, 232 haem 157-170
EU see European Union haematin 159, 167
European Union 43,50,61,67,71,77, Haematoccus spp. 91, 93, 95, 96, 102,
116, 192 104
Euglenophyta 141 Haematomma spp. 35
eumelanin 19 haemin 159, 167
Euphorbia milli 279 health 202, 229
Evening Primrose 334 Heliotales 31
extinction coefficients 47, 48, 53, 54, 60, Helvella spp. 31
61,65,69,148,165,167,183,192, heme see haem
222-223 Hemiselmis spp. 175,278,334
extraction 49,60, 66, 70, 149, 165, 182, Hibiscus spp. 55
215,257,286 high-performance liquid
chromatography 89, 118, 169, 187,211,
217-219,272,278,291,310,316
FDA see Food and Drug Administration hirscutidin 248
fermentation 85, 90-98 Homarus gammarus 231
Festuca pratensis 153--154 HPLC see high-performance liquid
Ficus spp. 252 chromatography
fig 252 huckleberry 253 see also Vaccinium
fish 28, 52, 200, 229 spp.
flamingos 210 Hydrocoleum spp. 175
flavanone 267 Hygrocybe spp. 32, 284
flavans 267 hypericin 15
flavin 20 Hypericum 32
Flavobacterium spp. 91, 96, 232 Hypomyces spp. 333
flavone 267 Hypoxylon spp. 31
346 INDEX
ices see dairy products Lyngbya lagerheimii l75

indicaxanthin 18, 283 Lyophyl/um spp. 30
indigo 2,17
indigo carmine 191-192 MAFF see Ministry of Agriculture,
Indigofera tinctoria 17 Fisheries and Food
iodinin 22 Malus pumila 252
invertebrates 28, 200 malv(in)idin 53, 248
Ipomoea spp. 252, 278, 310, 312, 334 mammals 28
lresine herbstii 282 Mangifera indica 252
iridoids 125, 319 mango 252
iresinin 282 margarine see dairy products
lsatis tinctoria 17 Margarodes polonicus 321
isocryptoxanthin 200, 223, 236 marigold 76, 102, 229
isozeaxanthin 200, 223, 236 mass spectrometry 225, 311
Mastigocladus laminosus 175, 182
jams see preserves maximum tolerated dose 113
lanhinobacterium lividium 334 meat 52, 69, 77
JECFA see Joint Expert Committee on Medicago sativa 146
Food Additives Mercurialis leiocarpa 332
Joint Expert Committee on Food metalloproteins 23
Additives 61, 116 Metatrichia vesparium 32
Ministry of Agriculture, Fisheries and
Kermococcus spp. 16, 321 Food 192
Kerria lacca 124 Mirabilis jalapa 283
kidney bean 252 miracle fruit 253, 278
Krill 102 miraxanthin 283
mixes 52, 59, 64, 73, 76
lac 124 monardein 311
Laccifera lacca 65, 124, 321 monascus 91,328-331,334
Lactarius spp. 30, 332 Monascus spp. 92, 104
Lactuca sativa 201 Morchella spp. 31
lactucaxanthin 200, 223, 237 morning glory - heavenly blue 314
lampranthin 282 moulds 82
Lampranthus spp. 282 MID see maximum tolerated dose
Lawsonia alba 16 musca-aurin 283
Musophagidae spp. 140
LDso see lethal dose (50)
legislation 42,50,55,61,67,71, 106, myoglobin 165
116-Jl7, 192
lethal dose (50) 115 naphthaquinone 15
Leguminosae 310 natural colours 42, 44, 46
leprarinic acid 33 nature identical colours 42
Letharia vulpina 35 neoxanthin 200, 223, 237
lettuce 201 nettles see Urtica dioica
leucopterin 21 neurosporene 200,206, 208, 223,238
lichens 33 N-heterocyclic compounds 17
Iichexanthone 36 NOAEL see No Observed Adverse
Liliaceae 310 Effect
Litchi chinensis 252 Nocardia spp. 333
Lithospermum spp. 85-87,324,335 No Observed Adverse Effect 114-116
litmus 35 Nopalea spp. 65, 321
Lobelia spp. 310 norbixin see bixin
lobster 231 Nostoc punctiforme 175
lucerne 146 nuclear magnetic resonance
luteins 76,91,200,223,229,237 spectroscopy 226
luteolinidin 248 Nyctanthes arbor-tristis 317
Iychee 252
Iycopene 119, 122, 198, ZOO, 206, 223, 237 Ochrolechia tartarea 34
Lycopersicum esculentum 212 Oenothera spp. 334
INDEX 347
O-heterocyclic compounds 12-24 Plectonema spp. 175

oil of turmeric 70 plum 252
oosporeine 32 Polyporales 31
Opuntia spp. 281,283,321 polyporic acid 36
orcein 35 Polyporus spp. 33
orcin( 01) 35 ponceau 4R 61, 191-192
Oscillatoria spp. 175 PopUlus spp. 279
oxygen 51,57,62,67,169,210,261, Porphyridium spp. 175, 179, 184, 188
288 porphyrins 138-140
Porphyrophera spp. 321
Portulaca grandiflora 283
palm oil 76, 100, 212 portulacaxanthin 283
pap.ri~a 77, 122 poultry feed 76
panetm 35 Prasinophyta 141
Passiflora edulis 252 prebetanin 282
passion fruit 252 preserves 58, 68
pasta 228 Prochloron spp. 141
patent blue V 191-192 Prochlorophyta 141
patents 311-335 Propionibacterium shermanii 177
peach 251 protohaem see haem
pelargonidin 53, 248 protoporphyrin (IX) 158
Penicillium spp. 31,332 protozoa 28
peonidin 53, 248 Prunus spp. 252
Perilla spp. 278, 334 Pseudomonas spp. 22,37
petunidin 53, 248 pterins 20
pH 46,50,62,66,67,71,75,184,257, pu\Chellidin 248
270,271, 276, 286 Punica granatum 252
phaeomelanins 19 purification 149, 165, 182, 215, 270, 290
phaeophorbide 134, 144 purines 20
Phaeophyta (brown algae) 141 purpurogallin 33
phaeophytin 74, 134, 144 pyocyanine 22
Phaffia rhodozyma 31,91,97,98, 102, Py"ophyta 141
103,231
Phallales 31
Pharbites nil 312 Quercus coccifera 321
pharmaceuticals 81,317 quinoids see qui nones
Phaseolus spp. 252, 333 quinones 15
Phellodendron amurense 332
phenazines 20 radish 252
phenoxazines 21 raspberry 53,253
Phlebia spp. 31 red 2G 191-192
Phormidium spp. 175 red cabbage 53, 55, 252, 278, 314
photo-oxidation see oxygen redcurrant 53, 253
photosynthesis 135 red onion 252
Phragmobasidiomycetidae 30 reptiles 28
phycobilins 27 retinal see vitamin A
phycocyanin 91, 173 Rhaphanus sativus 253, 333
phycoerythrin 173 Rheum rhaponticum 253
Phycomyces blakesleeanus 231 Rhizopus spp. 320
phyllocactin 282 Rhodella violacea maculata 175
Phyllocactus hybridus 282 Rhodobacter spp. 203
phytoene 200, 206, 223, 238 Rhodococcus spp. 97,334
phytofluene 200, 206, 223, 238 Rhodomonas lens 175
Phytolaccaca spp. 285 Rhodophyta (red algae) 141
phytomelin 20 Rhodospirillacae 141
phytoplankton see algae rhubarb 253
Plagioselmis prolonga 175 Ribes spp. 253
plant tissue culture 82-90 riboflavin 119
platyconin 311 Ricciocarpus natans 247
348 INDEX
riccionidin A 248 tetraterpenoids 12

Rivina humilis 282 Thalictrum minus 335
rivinianidin 282 theaflavin 14
RaceUa tinctaria 34 thin layer chromatography 215, 274
rocou see annatto tissue culture 83
root cultures 89 titanium dioxide 125
roselle see Hibiscus spp. TLC see thin layer chromatography
rosinidin 248 tocopherols see vitamin E
Rubia tinctorium 16 Tolypacladium infiatum 333
rubrolone 332 Tolypothrix distorta tenuis 175
Rubus spp. 253 tomato 212
Russula spp. 30 toxicity 112-114
Tradescantia paUida 249,278, 313, 314
Saccharomyces spp. 103, 232, 294 transbixin see bixin
safety 112-126 Tremallales 31
saffron 77, 84,212,317 tricetinidin 248
San Red RC 315 turmeric 69-74, 119-123, 325-327
sauces 77 tyrian purple 17
Schizothrix alpicola 175
Schleichera oleosa 321 Urtica dioica 146
Scleranthus perennis 321 usnic acid 36
secondary metabolites 82-90
Sepia officinalis 19
Serratia spp. 334 Vaccinium spp. 253,278
Setcreasea purpurea 310 see also Verbascum phlomoides 317
Tradescantia paUida vertebrates 28
shikonin 83-87 Viburnum dentatum 278
snacks 52 violaxanthin 200, 223, 239
soft drinks see drinks violerythrin 200, 223, 230, 239
Solanum spp. 253 vitamin A 119
soups 77 vitamin B12 139, 140
soy bean 252 vitamin E 136, 137, 140, 154
Spirulina spp. 91,93, 123, 188, 193 Vitis spp. 55,85,251,253,279
Spongicoccum excentricum 91 vulgaxanthin 59, 283
stability 51,56,62,67,71, 151, 164,
168, 181, 184,213,257, 286
water activity 63, 289
strawberry 53,252
Streptomyces spp. 22, 332, 333
Streptoverticillium spp. 333 xanthommatin 21
Suillus spp. 33 xanthophylls 212
sulphur dioxide (S02) 51, 57, 63, 67, 72 Xanthophyta 36, 141
sunset yellow 47 xanthopterin 21
sweet cherry 252 Xanthorrhia spp. 35
sweet potato 252, 314 Xenorhabdus luminescens 324
sweets see confectionery
Synechococcus 6301 see Anacystis
nidulans yams 252
Synechosysytis 6701 180 yeast see Saccharomyces spp.
Synsepalum dulcificum 253 yoghurts see dairy products
Sysyium cumini 278
synthetic colours 42, 190, 212 Zea mays 253
zeacarotene 200, 223, 239
Tagetes erecta 102,201,212 zeaxanthin 91,96,200,223,229,239
tamarillo 252 Zebrina spp. 312, 314
tartrazine 47, 72 see also Tradescantia pallida
tematin 311 zebrinin 311
tetrapyrroles 9-12 Ziziphus mauritiana 321

Natural Food Colorants - G. A. F. Hendry (Auth.), G. A. F. Hendry, J. D. Houghton

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Natural Food Colorants - G. A. F. Hendry (Auth.), G. A. F. Hendry, J. D. Houghton

Uploaded by

Copyright:

Available Formats

JOIN US ON THE INTERNET VIA WWW, GOPHER, FTP OR EMAIL:

SPRINGER-SCIENCE+BUSINESS MEDIA, B.V.

Printed on acid-tree text paper, manufactured in accordance with

Dr G .. Brittori Department of Biochemistry, University of

1 Natural pigments in biology 1

2 Natural food colours 40

2.7 Performance and consumption of natural colours 46

3 Biotechnology in natural food colours:

3.6.3 Bioprocessing of astaxanthin 101

4 Safety of food colorants 112

5 Chlorophylls and chlorophyll derivatives 131

5.4 Chlorophyll derivatives as food colorants 146

6 Haems and bilins 157

7.13 Other physicochemical methods 225

8 Anthocyanins and betalains 244

9 Less common natural colorants 310

This introductory chapter provides a working definition of 'natural'

The immediate purpose of this chapter is to introduce the major and at

By definition, a natural pigment in biological systems is one that is

G. A. F. Hendry et al. (eds.), Natural Food Colorants

1.4 Colour and colour description

Colour perceived Colour absorbed Wavelength of maximum

Blue-green Red 660--700

In the succeeding chapters data will be presented with the absorption

1.5 Colour perception

1.6 Colour and chemical name

One way of pinpointing coloured compounds from a list of organic

Table 1.2 Chemical names implying colour"

Chemical prefix or suf- Classical derivation Implied colour Chemical example

anil- ai-nil (Arab) Deep blue Aniline

"Arab = Arabic; Gk = Ancient Greek; L = Latin; OSp = Old Spanish.

1. 7 Electronic structure of pigments

Whether or not a biological molecule is coloured is determined by its

Table 1.3 Range of electro-magnetic spectrum absorbed by molecules and causing

Light causing Electronic Molecular form Expected colour

Visible n~ p* Unsaturated hydrocarbon Red, blue or green

1.8 Classification of biological pigments

It will be apparent that pigmented compounds in biology will tend to be

1.9 Classification based on structural affinities

Almost all biological pigments can be reduced to no more than six

Table 1.4 Naturally occurring pigments in biology

Group Alternative or Major example Predominant See

Tetrapyrroles Porphyrins and Chlorophylls Green 5

Figure 1.1 A linear tetrapyrrole.

Figure 1.2 A cyclic tetrapyrrole, porphyrin.

conditions probably determine the stability of the chromophore. In living,

Table 1.5 Colour range of naturally occurring tetrapyrroles

Colour Tetrapyrrole Occurrence

Blue Phycocyanin Blue-green algae

Figure 1.3 A tetraterpenoid, lycopene.

of oxygen radicals and at best indigestion! There are examples of animals

1.9.3 a-Heterocyclic compounds

Table 1.6 The more widespread of the naturally

Figure 1.4 Basic structure of a flavonoid.

Table 1.7 The most widely occurring anthocyanidins

Name Colour Amax in acidic

Cyanidin Blue-red 535

magnesium confer a blueing to the pigment. Interactions between antho-

Figure 1.9 9,1O-Anthraquinone.

Figure 1.10 Hypericin.

The simpler quinones may function as electron and proton carriers in

1.9.5 N-Heterocyclic compounds (other than tetrapyrroles)