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 FTP: ftp.thomson.com

A service of lOOP

EMAIL: [email protected]

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 Massachusetts, 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 Agricultural 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 2.9.2 Legislation 55 2.9.3 Factors affecting stability 56 2.9.4 Applications 57

2.10 Beetroot 59 2.10.1 Source 59 2.10.2 Extracts and their application forms 60 2.10.3 Legislation 61 2.10.4 Factors affecting stability 62 2.10.5 Applications 63

2.11 Cochineal and carmine 64 2.11.1 Source 64 2.11.2 Extracts and their application forms 66 2.11.3 Legislation 67 2.11.4 Factors affecting stability 67 2.11.5 Applications 67

2.12 Curcumin 69 2.12.1 Source 69 2.12.2 Extracts and their application forms 70 2.12.3 Application forms 71 2.12.4 Legislation 71 2.12.5 Factors affecting stability 71 2.12.6 Applications 72

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

3.6.3 Bioprocessing of astaxanthin 3.6.4 Bioprocessing of marigolds 3.6.5 Bioprocessing of betanin

3.7 Future perspectives 3.7.1 Biotechnology and carotenoid production 3.7.2 The influence of legislation 3.7.3 Commercial developments

Acknowledgements References

4 Safety of food colorants F.J. FRANCIS

4.1 Introduction 4.2 Toxicity testing 4.3 Acceptable daily intake (ADI) 4.4 Regulatory approaches

4.4.1 The European Union (EU) 4.4.2 The USA 4.4.3 The Delaney Clause

4.5 Safety of individual colorants 4.5.1 4.5.2 4.5.3 4.5.4 4.5.5 4.5.6 4.5.7 4.5.8 4.5.9 4.5.10 4.5.11 4.5.12 4.5.13 4.5.14 4.5.15 4.5.16 4.5.17 4.5.18 4.5.19 4.5.20

References

Beneficial aspects Anthocyanins Carotenoids Canthaxanthin Apocarotenol Lycopene Annatto Paprika Algae Turmeric and curcumin Cochineal and carmine Lac Caramel Monascus Iridoids Betalains Cocoa Chlorophyll Titanium dioxide Carbon black

5 Chlorophylls and chlorophyll derivatives G.A.F. HENDRY

ix

101 102 103 103 105 106 107 108 108

112

112 112 114 116 116 117 117 119 119 120 120 121 121 122 122 122 122 123 123 124 124 124 125 125 125 125 125 126 126

131

5.1 Summary 131 5.2 Introduction 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 Acknowledgements 155 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 Introduction 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 Future prospects 230 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 R.L. JACKMAN and I.L. SMITH 8.1 Summary 8.2 Introduction 8.3 Functions 8.4 Anthocyanins

8.4.1 Structure 8.4.2 Distribution 8.4.3 Biosynthesis 8.4.4 Factors influencing anthocyanin colour and stability 8.4.5 Extraction and purification 8.4.6 Current and potential sources and uses

8.5 Betalains 8.5.1 Structure 8.5.2 Distribution 8.5.3 Biosynthesis 8.5.4 Colour and structural stability 8.5.5 Extraction and analysis 8.5.6 Current and potential sources and uses

8.6 Conclusions Acknowledgements References

9 Less common natural colorants F.J. FRANCIS 9.1 Acylated ~-ring substituted anthocyanins 9.2 Annatto 9.3 Saffron 9.4 Gardenia pigments 9.5 Cochineal and related pigments 9.6 Turmeric 9.7 Carthamin 9.8 Monascus 9.9 Miscellaneous colorants 9.10 Future prospects References

Index

244

244 244 246 246 246 251 251 257 270 278 280 280 281 284 286 290 294 296 296 296

310

310 315 317 318 321 325 327 328 332 334 335

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 inevitably 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 phaeophytin 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 organisms. 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 specifically 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-greenblue-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

Blue-green Blue Violet Mauve-red Orange Yellow Yellow-green Green-yellow

Colour absorbed (from white light)

Red Orange Yellow Yellow-green Green Green-blue Blue Violet

Wavelength of maximum absorption (nm)

660--700 6()()....Q30 585-600 560--580 520--540 480--500 440--460 >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 restaurateurs 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 redsensitive, 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 considerable.

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 differences 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 biological material is composed of a limited range of elements within the periodic table. Within this narrow selection, pigmented organic compounds generally 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 pigmented compounds, molecular weights range from about 200 (anthraquinones), 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 hydrocarbons (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 hydrocarbons (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 hydrocarbons 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 wavelengths 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 nonyellow 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 pigmented 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 photosynthesis (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 (peroxidases). 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 orophylls 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 carotenoids 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. Carotenoidprotein 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 summarized 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

NATURAL PIGMENTS IN BIOLOGY

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)

13

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, anthocyanidins 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 uncommon 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 anthocyanins 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 dihydroxyphenol 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.

o

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 pigmentation, 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 antiherbivore 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 cosmetic 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 melanomas in man. In mammals, melanins are frequently linked to metals (Fe,

H0XQ< ~ I ",H

HO ~ 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 attachment 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 recovered 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 characteristics 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 microcrystals 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, amphibians 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 invertebrates 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 carotenoids.

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 bacterial 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 important 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

OH b-Figure 1.24 Iodinin.

NATURAL PIGMENTS IN BIOLOGY

COOH H,N-CH

¢H,;s! HCOOH &C=O N : IN

~ I " ~ 0 ~ 0

Figure 1.25 Xanthommatin.

23

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 metalloporphyrin haems and chlorophylls, has already been described above. A list of the more widespread (and a few that are uncommon) metalloproteins 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 presentday 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 chlorohaemoglobin, 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 (Feprotoporphyrin or its Fe-derivatives). One of these is the misnamed haemerythrin, 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) metalloproteins, generally blue to blue-green, are present in a number of invertebrates. Exceptionally a Cu-containing porphyrin, turacin, is a bronzepurple colorant in one family of tropical birds. A vanadium-containing protein, haemovanidin, is apple-green in colour. Zn-porphyrins are widespread 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 (Fortschritte 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, naphthaquinones, 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

Chlorophylls

Phycobilins

Carotenoids

Anthocyanidins

Betalains

Common examples

a b c,d

Phycocyanin Phycoerythrin

Lutein

~-Carotene

Violaxanthins Neoxanthins Fucoxanthin

Cyanidin

Pelargonidin } Oelphinidin

Betacyanin

Natural occurrence and abundance

All photosynthetic eukaryotes All land plants, many algae Brown and other algae

Blue-green and other algae Red and other algae

Most abundant xanthophyll. Most photosynthetic organisms

Most abundant carotene. Most photosynthetic organisms

Common in higher plants Common in higher plants Brown and other algae

The most common anthocyanidin widely distributed in higher plants

Common in higher plants

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. bacteriochlorophylls) .

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 myoglobin, 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

Mammals

Birds (including their eggs)

Reptiles and amphibians and bony fish (teleosts)

Cartilaginous fish (elasmobranchs)

1.11.3 Invertebrates

Pigment

Mainly melanins

Melanins Carotenoids Tetrapyrroles

Melanins Carotenoids Pterins Riboflavin

Melanins

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

Echinoderms (star fish, sea urchins, crinoids)

Molluscs (snails, bivalves, squids)

Insects

Malacostraca (centipedes, millipedes)

Crustacea (crabs, lobsters, shrimps)

Arachnida (spiders, scorpions and alJies)

Annelids (segmented worms and leeches)

Cnidaria (jellyfish, anemones and alJies)

Porifera (sponges) Protozoa (plankton)

Pigment

Generally highly pigmented, mainly carotenoids, quinones and melanins

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

Often highly coloured and containing many carotenoids, flavonoids, wide range of melanins, ommochromes, pterins, flavins and occasional bilins and unusual polycyclic quinones

Melanins and carotenoids predominate but many examples of ommochromes and rare pigments. Haemocyanin may contribute to pigmentation

Highly pigmented with carotenoids, bilins, melanins, pterins, flavins and some with haemocyanin as a visible pigment

Varied pigments, including unusual carotenoids, pterins, qui nones (especially spiders), some with haemocyanin and pterins, many with so far unidentified pigments

Varied in. pigmentation , often due to bilins, some with haemoglobin or chlorohaemoglobin

Often with unusual carotenoids, some with bilins, the group contains numerous poorly described pigments

Mainly carotenoids 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 population 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 compounds that are either unreported from other biological groups or if present then in only trace amounts. An example of the latter is riboflavin (vitamin B2), 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 concentrations 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. Unexpectedly, 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 (colourless) on dimerization forms the brown and black perlene quinone of the fruiting bodies of Daldinia and Bulgaria spp. Several of these naphthaquinones 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 impossibility 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 anthraquinones, 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 welldefended 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 commercial 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 contribute to the deep-chocolate colour of boletes.

Another group of fungal phenol derivatives, more properly phenylpropanoid 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 hydroxybenzoic acid derivatives, often with terpene side-chains, contribute to the yellows and reds of several fungi including Suillus, Polyporus, Gomphidius, 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('-': OH

~ I h 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 Scotland, 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 conservation 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:

Depsides

gyrophoric acid l evernic acid ~

lecanoric acid

decomposition products ~ orcinol (orcin) (mild alkaline hydrolysis) (brown)

orcinol + NH40H + O2 ~ orcein (purple)

(1.1)

Orcein itself (Figure 1.29) is a mixture of oxy- and amino-phenoxazone or phenoxazine, the individual components varying from brown-red to blueviolet. 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 orceinbased 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.

-j::ft;(COCH' 0

HO ~ ~ :-- 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 concentrating 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. Bacterial 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 biotechnological 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 biotechnological 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 discussions 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 development 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 manufactured 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].

NATURAL FOOD COLOURS

Table 2.1 Natural colours (and colours of natural origin) listed by the EU

Number

Eloo ElOl E120 El40 E141 E150a E153 El60a El60b El60c El60d El60e E161b E161g E162 E163

Colour

Curcumin Riboflavin, riboflavin-5 ' -phosphate Cochineal, carminic acid, carmines Chlorophylls and chlorophyllins Copper complexes of chlorophylls and chlorophyll ins Plain caramel Vegetable carbon Mixed carotenes and J3-carotene Annatto, bixin, norbixin Paprika extract, capsanthin, capsorubin Lycopene J3-Apo-8' -carotenal (C30) Lutein Canthaxanthin Beetroot red, betanin Anthocyanins

43

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 characteristics. Consequently each colorant is available in several different application 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 suspensions. 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 doselevel. 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 diapocarotenoic acid norbixin, which is found as a resinous coating surrounding the seed itself. Also present, as minor constituents, are trans-bixin and cisnorbixin. 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·'

-~ COOCH,

: ~ 0'4 • 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 considerable 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

heat heat ______ ........ trans-bixin ______ ........ degradation compounds

1 hydroly," 1 hydroly'"

cis-norbixin trans-norbixin

cis-bixin

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. Extraction 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 oilbased 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 coloration 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 and vegetables provides man with a rich variety of foods with an interesting variety of colours. These colours provide vital information about the safety, freshness and quality of these foods and even today man continues to use colour in making such subjective assessments. However, the aesthetic value of 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 2 Blue grapes are the main commercial source of anthocyanin colours. Anthocyanins E163 are the mainly red pigments that are responsible for the colours of many edible fruits and berries. They also contribute to the colours of some vegetables and flowers. Together with carotenoids, they provide us

with the attractive autumnal colours.

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).

(b )

Plate 4 Beetroot Red E162 is obtained by aqueous extraction of red table beet (Beta vulgaris), consisting of betanin and vulgaxanthin. Freely soluble in water, it provides 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.

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).

(b)

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.

.-~-

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.

(a)

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).

(b)

50 50

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 Ancellotta, 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 absorbance 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 derived 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 extract per gram

Standard 1.5 150 Double strength 3.0 300 Concentrate 6.0 600 Powder 12.0 1200

Approximate % anthocyanin

0.5 1.0 2.0 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 produced for specific applications.

Thus there are essentially only two application forms of grape anthocyanin extract-a water-soluble liquid and a water-soluble powder. Nevertheless, 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 anthocyanins 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 permitted. 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 anthocyanins 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 anthocyanins 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 oligomeric 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 preparations, 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 anthocyanins 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 'thinlayer' 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


.GIUCOse-ox;cr 1-"-: + COOH

HO"" h

HOOC~~ H

.. .. ·e

0·8

:::) 0·.

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 concentrate

'1% betanin' liquid '0.66% betanin' powder

5.60 3.60

True % betanin

0.5 0.33

61

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 measurements of colour strength made are at the wavelength of maximal absorbance 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 yellowbrown 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. Generally, 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 beetroot 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 production 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 contamination.

2.10.5.3 Dry mixes. Beetroot-juice powders are ideal for such applications 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)

(b)


OH 0 CH,

Glucose@",=COOH I I

HO ~ h OH

OH 0

o·s

~ ·c ::) 0·4

8 i .a 5 .. .a io( 00

0·8

.. :'!: c ::) 04 .. u C to .CI

0 .. .a io( 0.0

65

400 500 700

Wavelength nm

700 Wavelength nm

Figure 2.6 Cochineal and carmine (class anthroquinone): structural and physical characteristics 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 available 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. However 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 mentioned, 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 applications; 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 coloration 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 = OCH3 . Demethoxycur-

cumin: R, = OCH3 , 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-probably 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 properties 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 orangebrown 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 watersoluble 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 chlorophyllins 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 yellowgreen. 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 combination 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 byproduct 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 costeffective 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 extensive 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% carotenoids.

Water-miscible forms are usually prepared by incorporating polysorbates 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-annatto, 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 manufacturer 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 formation 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 internationalisation 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 experimented 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 continuous processing, fermentation and enzymatic processing. Genetic engineering 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 industries that were quick to increase efforts in this area.

Fermentation or microbial transformation is the primary tool of biotechnology. 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 metabolite. 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 biochemicals 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 potential 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 resistance, 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 scaleup, 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. Kinnersley 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 produced 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 immobilisation techniques.


Table 3.2 Some advantages of plant cell immobilisation (after Fontanel and Taboo 1987; Knoll 1986)

Advantage

Protection of cells by matrix

Maintenance of stable and active biocatalyst

Continuous process possible

Increased productivity

Influence on cell culture

Simplification of scale up Protection from physical damage

Elimination of non-productive growth phase

Reduced risk of declining productivity from a selected cell line

Continuous removal of metabolic inhibitors Improved process control and product

recovery

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 concentration 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 disadvantages 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 fermentation, 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

BIOTECHNOLOGY IN NATURAL FOOD COLOURS

Table 3.4 Examples of colour pigments produced by microbial transformation

Colour pigment

~-Carotene

Astaxanthin

Lutein

Zeaxanthin

Canthaxanthin

Monascus red pigments

Phycocyanin

Non-microbial source

Carrots, palm oil, synthesis

Crustacea, birds' feathers, synthesis, Adonis sp. flowers

Microbial source

Dunaliella salina Blakeslea trispora

Phaffia rhodozyma Haematococcus sp.

Alfalfa, com, marigold Spongiococcum excentricum flowers, maize, green plants Chlorella pyrenoidosa

Alfalfa, com, marigold Flavobacterium sp. flowers, maize, green plants

Crustacea, birds' feathers, Cantharellus cinnabarinus synthesis Brevibacterium KY -4313

Monascus purpureus

Spirulina sp.

91

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 commercial 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 fermentation 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 metabolites. 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 biotechnology 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. Fermentation 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 associated 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 commercialised. 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 fermentation 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 characteristics 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 industrial 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

Microbio Resources (TrouwIBP)

Cyanotech

Salt Research Institute

Betatene

Western Biotechnology

Nature Beta Technology/Koor Foods (35% shareholding by Nikkon Sohansha)

Location Dunaliella sp.

Death Valley, California D. salina

Keahole Pt., Hawaii D. salina

Tianjin, China D. salina

Whyalla, S. Australia D. salina

Hutt Lagoon, W. Australia D. salina

Eilat, Israel D. bardawil

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 chloroplast 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 algaculture 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 commercial 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 CO2 , 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 bioreactors. 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 Haematococcus 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 autotrophic 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 commercialisation [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 canthaxanthin production by fermentation, for example strains of Brevibacterium 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 improvement 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 companies 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. However, 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]. Unfortunately, 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 techniques 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 carcinogens 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 multitonne 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, fractionation, 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 concentrated 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

'---....;;.;T;.;.;.;.-;:=~--Hexane

Purification (Chromatography)

Detergent

Softener

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. However, 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 percentage (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. Considerable 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 generally 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 additives 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. Fermentation 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 Saccharomyces 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 biotechnologybased 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 regulation 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, Agrobacterium 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 cellfree 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 isolation 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 companies 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 inexpensive 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 engineered carotenoid producing microbes and carotenogenic enzymes, immobilised enzyme systems and bioreactors, and new methods to stimulate 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 definition 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 application of biotechnology in the colours industry, one comes to the conclusion 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 industry 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 expertise 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. Biotechnolo 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 carcinogens 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 carcinogenesis.

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 nonlinear 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.

-Typical

"0 Safety r c ::> Factor 0

Q) 0, <II oX C (J 0 '" ) a. .0 <II Q) 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:

SAFETY OF FOOD COLORANTS

Acceptable 1--_01<----, .... Risk Level

o VSD VSD Dose (Linear (Nonlinear Model) Model)

115

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 'uncertainty 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 Amendment 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 Grandfather 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 lymphocytes, 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 anticarcinogenic [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 multigeneration 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 embryotoxicity, 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 hypercarotenemia 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 anticarcinogenic [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-genotoxic 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 containing 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 antioxidative 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 microorganisms [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 contaminants. 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 Ti02

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 JECF A 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 bacteria. 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 chlorophylls 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 carotenoids, 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 chlorophyllderived food colorants, if only to bring these into line with current scientific trends towards simplicity with precision in chemical nomenclature. 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 nomenclature 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 chlorophylls 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 131 and 132 .

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) coordination 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

Chlorophyll Chlorophyll ide Phaeophytin Phaeophorbide

Abandoned names

Protochlorophyllide Phyllin e6 Chlorin e6 Chlorin e4 Purpurin 7 Purpurin 5 Rhodin equivalents of the above Chlorophyllin

Based on

17, 18-Dihydrophaeophytinato Mg(II) 17,18-Dihydrophaeophorbideato Mg(II) 17, 18-Dehydrophaeophytin 17, 18-Dihydrophaeophorbide

Recommended names

Didehydrophaeophorbideato Mg(II)

Didehydrorhodochlorin with a specific description of the substitute at C13 and 15

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 longestablished 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. Presumably 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 chlorophyll 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 watersoluble '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 thylakoids. 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 dinucleotide phosphate (NADPH) and adenosine triphosphate (ATP) are then consumed in the reduction of CO2 (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 bacteriochlorophylls) drastically alters the area of de localized electrons by removing 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 associated 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 excitation energy and also act as scavengers (or antioxidants) of highly reactive

CHLOROPHYLLS AND CHLOROPHYLL DERIVATIVES

Porphyrin ring

Vn -------;:71 \ I 1

Other arrangements in

chlorophylls ll., k and l1.

----,~--- Cyclopentanone ring

Bond disrupted in alkali

Position subject to oxidation

137

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 pyrrolenitrogens. The arrangements of the ten double bonds within the ring system may also vary. Vn = CHCH2 ; Me = CH3 ; Et = CH2CH3 . The delocalized system of electrons is

shown as ----.

oxygen radicals and singlet oxygen. The tocopherols (or tocophenylquinones) 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 polypeptides, 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 polypeptides 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 polypeptides, 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 photoreactive 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 chromophore 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 polypeptidelipid environment of the thylakoid membranes.


5-Amino laevulinic acid

! Porphoblilinogen

J. Uroporp ynnogen III

Coproporphyrinogen III

1 Protoporphyrinogen IX

~ Protoporphyrin (Protoporphyrin IX)

Mg

Protoporphyrinato methylester Mg II

1 Fonnation of ring E

CU

Co

Fe

Zn

Fe

D'd<h ,1,,",,""00"',"'"," Mg II (Protochlorophyllide)

j 2", + ligh<

Phaeophorbideato Mg II (Chlorophyllide)

1 Phy<o]

Chlorophy 11 or Phaeophytinato Mg II

Turacin

Cobalamin

Sirohaem

• Z-bilin

Protohaem

! Haem a

! Haemc

Comment

Universal precursor of all biological prophyrins

A mono-pyrrole

In feather of certain birds

Prosthetic group vitamin B 12

Primitive haem protein

Bird egg-shells

Prosthetic group of haemo-globin, myoglobin, cytochrome b. peroxiases, catalase etc.

Cytochrome a

Cytochrome c,f

Involves several steps

Accumulates in dark-grown plants

Does not accumulate

Phytyl group is a 20T terpenoid

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 pigments 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 (cytochromes), oxygen radical destruction (through haem-peroxidases) to photosynthesis (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/ bacteriochlorophyll

Organism

141

Chlorophyll: a All oxygen-evolving photosynthetic organisms including all higher

plants, all algae and the photosynthetic bacteria Cyanophyta and Prochlorophyta

b

c

d e

Bacteriochlorophyll: a and b c, d and e

High plants, the algal Chlorophyta and Euglenophyta and the bacterial Prochlorophyta

Algal Phaeophyta (brown algae), Pyrrophyta (dinoflagellates), Bacillariophyta (diatoms), Chrysophyta, Prasinophyta and Cryptophyta

In some Rhodophyta (red algae) and Chrysophyta Reported from algal Xanthophyta

Purple bacteria including Chromatiaceae and Rhodospirillaceae 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 chlorophylls 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 variations 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, probably 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 COCH3 MeH Et Phy H MeH Bacteriochlorophyll b COCH3 MeH CHCH3 H Phy H MeH Bacteriochlorophyll c HO-Et MeH CH2CH3 H Farn H MeH Me or Et Bacteriochlorophyll d HO-Et MeH CH2CH3 H Farn H MeH Bacteriochlorophyll e HO-Et CHOH CH2CH3 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, -CH2CH3 ; Vn, -CHCH2 ; 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 destruction 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 geoor 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 chlorophyllide (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 senescence.

Much of the global production of chlorophylls is ultimately grazed and pass through one, or more, animals' guts. In the oceans, grazing leads to

144

Heat Anoxia Time

*Decarboxydeoxophytoporphyrin (Dioxophylloerythroaetioporphyrin) and other Aetioporphyrins

NiorV=O

*The geo- or petroprophyrins of oils and coal

NATURAL FOOD COLORANTS

Chlorophyll or

Chlorophyllase PhaeoporphinatO~Mg II -:---aCe!'dak unknown ~~::~c~ reactions Phytol Mg

*Phaeophorbideato Mg II *PhaeOPhytin! or Chlorophyllide *Various

Weak acid

Mg Phytol

*Phaeophoribide

Strong acid

Chlorophyll ring E isomers

solvation

j *Low molecular weight colourless

[Cu-free oil SOI::~:loroP~hYll ~::

effects compounds

Heat *Phaeophorbideato Cull

*Phytoporphyrin (Phylloerythrin)

[oil-soluble Cu-chlorophyll]

H,:l~"g, ri"gE~

Phyllophorbide (Pyrochlorin)

I Gut digestion

Low molecular weight colourless compounds

Didehydro rhodochlorins (chlorin e6 '

purpurins and others)

j Na-didehydro rhodochlorinato [Water-soluble Cu-free chlorophyll (in)]

j Cu+~id Na-didehydro rhodochlorinato Cu II [Water-soluble Cu 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 phaeophorbide derivative phytoporphyrin. Having survived grazing and photodestruction, some remaining chlorophylls will be deposited in oxygendepleted 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, eventually 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 production, 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 acidification 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 production of chlorophylls·

Environment

Terrestrial Aquatic Total

• From ref. [1].

Tonnes x 108

2.92 8.63

11.55

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 highyield 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 132HO-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 following 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 chlorophylls 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 concentrations 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

....... ~

Tab

le 5

.5 S

pect

ral

char

acte

rist

ics

of t

he n

atur

ally

-occ

urri

ng c

hlor

ophy

lls

and

bact

erio

chlo

roph

ylls

Spe

ctra

Z

S

oret

R

ed

> C!

Pig

men

t S

olve

nt

Col

our

Xm

ax

Em

M

Xm

ax

Em

M

Ref

eren

ces

~ t""

Chl

orop

hyll

a

80%

ace

tone

B

lue-

gree

n 43

1.2

95.8

66

3.2

86.3

25

'" 0

Chl

orop

hyll

b

80%

ace

tone

G

reen

45

9.0

135.

4 64

6.8

49.2

25

0

Chl

orop

hyll

CI

90%

ace

tone

Y

ello

w-g

reen

44

3.2

318.

0 63

0.6

44.8

28

~

Chl

orop

hyll

C2

90%

ace

tone

Y

ello

w-g

reen

44

3.8

374.

0 63

0.9

40.4

28

("

) 0 C

hlor

ophy

ll d

E

ther

B

lue-

gree

n 44

5.0

97.8

68

6.0

117.

8 22

t""

0

Bac

teri

ochl

orop

hyll

a

Met

hano

l G

rey-

pink

36

5.0

53.9

77

2.0

42.0

26

~

Bac

teri

ochl

orop

hyll

b

Ace

tone

B

row

n-pi

nk

368.

0 79

5.0

22

~ B

acte

rioc

hlor

ophy

ll C

A

ceto

ne

Gre

en

428.

0 66

0.0

27

Bac

teri

ochl

orop

hyll

d

Ace

tone

G

reen

42

4.0

654.

0 27

B

acte

rioc

hlor

ophy

ll e

A

ceto

ne

Gre

en

456.

0 64

6.0

27


pigment, called phytochrome, has a highly specific and dominant function as a light receptor, controlling many critical aspects of plant metabolism 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 artificial, rapid, heat-usually at closely controlled temperatures. The subsequent 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. Heating 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 O2 . Most of the identified products show alterations to the 131_ and 132-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, concentration 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 metalfree phaeophytin, or further acidified in the presence of copper salts to form oil-soluble copper phaeophytin. The possibility of producing authentic 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 hydrophobic phytyl group to be displaced by sodium (Na) (or potassium (K». The acidic groups at 131 and 132 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 procedures 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 'chlorophylls' 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 chlorophyllin is broadly understood to refer to an unspecified product of alkaline hydrolysis of chlorophylls [44] and may be further equated with chlorophyHide, 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 131 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 \

151

Figure 5.5 The structure of a rhodochlorin where R1 and R2 are simple derivatives or substitutions to the positions 131 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 Mgchelated '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 Mgchelated 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 protohaem) 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 photoreactivity in vivo.

The stability of some porphyrins is legendary, however. Proof of this lies in the occurrence of nickel and vanadyl porphyrins, derived from chlorophylls 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 greycoloured phaeophytin a. Indeed, the fresh greenness, jubilantly reported on exposure of archaeological plant remains, frequently turns to a dullbrown colour within minutes, strongly indicative of oxidation of greencoloured 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), anthocyanins, 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 watersoluble 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 fashionled 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 longterm 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 membranes 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.),

Academic Press, New York (1979) pp. 179-232. 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 complexes are pink or red in colour by virtue of their common prosthetic group, but the structure most intensively studied is undoubtedly haemo-



158

(a)


(b)

h -CH,

\ CH, CH, CO,H

(c) CH, I

h -CH,

\ <;:H, CH, 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 colorants.

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 perpendicular 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/ferrohaemoglobin. 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, responsible 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. Haemoglobin' 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 haemoglobin 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 metabolic 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 makeup.

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

<;:O,H <;:H, <;:H, +

~~Co.A Succinyl

coenzyme A

~H, <;:H, CO,H

Glycine

uroporphyrinogen

protoporphyrinogen

HAEMS AND BILINS

<;:O,H (a) <;:H, (b)

~ <;:H, CO 9H, NH,

ALA

<;:O,H CH, <;:O,H tH, CH, .. \ I

0 N ...... CH NH H "

porphobilinogen

<;:O,H CH, tH, \

coproporphyrinogen

CHCH, \

~ -CH, \ CH, tH, CO,H

protoporphyrin IX

161

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) uroporphyrinogen decarboxylase, (e) coproporphyrinogen oxidative decarboxylase, (f) protoporphyrin-

ogen oxidase.

haem chromophore are esterified to amino-acid residues of the surrounding 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 arrangement 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 = CHCH2, Me = CH3 , Pr = CH2CH2CH20H.

HAEMS AND BILINS 163

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 chromophoreswhether 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 localization, absorbance and stability.

6.2.2.4.1 Localization. As has already been alluded to, the distribution 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 membranebound 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 advantages 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 absorbance 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 characteristic 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 absorbance 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 \ ,

\

HAEMS AND BILINS

~~50~--~5~070----~55~O~----6~O-O~~~ Wavelength (nm)

165

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 mixtures provide a more vigorous extraction medium and this method is used widely [13-16] for the separation of haem and globin. Acidified methylethylketone 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 satisfactorily 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 H20, H2O pH 7.0 to 12.0 buffer 570-580 5.5--6.5

Haematin OH, H2O 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 complexes 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 haemoproteins 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. Progressive 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 quantitative 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 techniques do, however, exist and have been used mainly in the structural determination and analysis of natural products on a milligram scale. Highperformance liquid chromatography (HPLC) techniques have been developed 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 exploitation 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 apoprotein 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 prospective 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 biliverdin-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 protein 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: phycoerythrins (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, I

CH, I

CH, I

CH, I

CH H,C CH, H,C CH, H,C CH, (b)

N N H H

COOti COOH I I

(e)

CH, CH, CH, CH, I I I II

H,C CH H,C CH, H,C. CH, H,C CH, :H ~ li· ~.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 reincorporated 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 phycoerythrin 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-, Band 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 phycoerythrins 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 phycoerythrin 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.

HAEMS AND BILINS

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.

173

The great diversity of absorption maxima observed for the biliproteins in general is a function of the different apoproteins to which the chromophores 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 chromophores 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 function 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 photosynthetic 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 predominantly phycoerythrin, with trace amounts of phycocyanin

Cyanophyceae (contain C-PE) Hydrocoleum spp. Oscillatoria rubescens Phormidium autumnale Phormidium ectocarpii Phormidium fragile Phormidium persicinum Phormidium uncinatum Pseudanabeana spp.

Rhodophyceae (contain B-PE) Rhodelia violacea maculata

Cryptophyceae (contain cryptomonad PE) Cryptomonas ovata maculata

Hemiseimis rufescens Plagioseimis prolonga Rhodomonas lens

Species containing both phycoerythrin and phycocyanin

Cyanophyceae (contain C-PE & C-PC) Aphanocapsa spp. Aphanothece sacrum Calothrix membranacea Dermocarpa violacea Fremyella diplosiphon Glaucosphaera vacuolata Schizothrix alpicola Tolypothrix distorta tenuis

Rhodophyceae (contain B-PE & R-PC)

Porphyridium aerugineum Porphyridium cruentum Porphyridium marinum ( = purpureum)

Cryptophyceae Unknown

Species containng phycocyanin, with no phycoerythrin

Cyanophyceae (contain C-PC) Anabaena cylindrica Anabaena variabilis Anabaena 64JJ Anabaenopsis spp. Anacystis nidulans

(=Synechococcus 6301) Arthrospira maxima Aphanizomenon fios-aquae Calothrix scopulorum Chlorogloea fritschii Lyngbya lagerheimii Mastigocladus laminosus Nostoc punctiforme Oscillatoria amoena Oscillatoria animalis Oscillatoria aghardii Oscillatoria chalybea Oscillatoria formosa Oscillatoria minima 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 perception, 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

CO

,H

CH

NH

, (a

) C

H, ~

CH

, C

O,H

glut

amat

e

CH

2NH

2

CO

¢H

, C;

:H2

C0

2H

ALA

(b)

--..

~

chlo

roph

ylls

Me/

I I

Pr

~-V

Mg.

pro

topo

rphy

rin I

X

phyc

oery

thro

bilin

(d)

I ~

biliv

erdi

n ~

phyt

ochr

omob

ilin

" ph

ycoc

yano

bilin

haem

Fig

ure

6.6

The

bra

nche

d bi

osyn

thet

ic p

athw

ay f

or t

he f

orm

atio

n of

hae

m,

bilin

s an

d ch

loro

phyl

ls f

rom

glu

tam

ate.

The

enz

ymes

res

pons

ible

for

the

bi

osyn

thet

ic s

teps

sho

wn

are:

(a)

glu

tam

ic a

cid

tRN

A l

igas

e, g

luta

myl

tR

NA

deh

ydro

gena

se,

and

glut

amat

e I-

sem

iald

ehyd

e am

inot

rans

fera

se,

(b)

as

desc

ribe

d in

Fig

ure

6.2,

(c)

fer

roch

elat

ase,

(d)

as

desc

ribe

d in

Fig

ure

6.3.

V =

C

HC

H2

, M

e =

C

H3

, P

r =

C

H2C

H2C

02H

.

.......

-...I

0\ z ~ ~ t"" c:l o t:I

(j o S ~ ~

HAEMS AND BIUNS 177

organisms strict mechanisms exist for the coordinate regulation of apoprotein 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 apoproteins. The genes for the a and 13 subunits of phycocyanin from the blue-green alga Synechococcus spp. PCC7002 were cloned [50] and reintroduced 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 tetanomorphum 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) pigment 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 pigments.

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 degradation 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 phycobiliproteins 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 maximum 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 heterotrophically 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 fluorescence, and stability of the phycobilins in nature will be discussed here.

6.3.2.4.1 Cellular localization. In red and blue-green algae phycobilins 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 morphological features on the outer face of the thylakoid membrane of the chloroplast, when viewed under an electron microscope. The size, composition, 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 composition 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 surrounded 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 , , , I c: , , ..

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 chromophores and the individual associations and energy coupling of the overall structure. Representative absorption and fluorescence spectra of phycocyanin 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 accumulation, 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 characteristics 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 testedMastigocladus 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 definition, 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 characteristics. 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 freezedried 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, chlorophyll 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

HAEMS AND BILINS

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 pigment produced in mutant compared with wild-type cells.

183

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 CHCI3

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 CHCl3

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 precipitation 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 proteinfree 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 porphyrins 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 phycocyanins 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]. Fluorescence 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 substantial 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 Porphyridium 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)

(b)

(c)

jIl ·c :::>

0.8

0.6

0.4

0.8

0.6

g 0.4

'" -e 5l ~ 0.2

HAEMS AND BILINS

C - phycobilins

300 400

R - phycobilins

/

cryptomonad phycobilins

300

'\ /\ : I I : I I , I ~ I

/ r. I' ,:

1 I

I

! ~ \/

I

I I

I I ,

I I ,

I

-' 500

Wavelength (nm)

185

700

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]. Rphycocyanin 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 -e 0

'" 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, denaturing only slowly in 8M urea, and retaining absorption and fluorescence maxima at temperatures up to 70°C [83]. These differences in conformational 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 phycoeyrthrobilin is rather less well studied. It is thought that the free phycobilin structures are stabilized by intermolecular hydrogen bonds to form a semicycle 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 convenient method for quantitative analysis of biliproteins is by use of their UVvisible 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 breakdown 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 reversephase 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 biliproteins 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 beginning 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 Porphyridium 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 appropriate 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 highvalue 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 considerably enhance the number of species for which it is feasible to consider large-scale culture.

An alternative approach is to use already existing fermentation technology 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 technology 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 (KN03).

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 individual 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 development 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 development 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 phycoerythrin 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

HAEMS AND BILINS

~ c: ::J

g

1.0 (3)

/

/ ... " , , , , ' /

/ \

.e 0.5 o ,\

/ ' .: \ '" " J:l "

< \ , , , , ... /

/ , \ , , \.

OL----"'---'--"'-----::-' 300 400 500 600 700

1.0 (C)

~\

'-.... , " \ .....

r""\ , , : ' , , , / \ , , , , , ,

OL-~_~~~~. 300 400 SOO 600 700

1.0 \(e) ,

,'\

, " / , , , : ' , , , , , , , , ,

'\ I ~

" , \ ,,' ~ /

o \ ...... ",---",,'

, , , ,

300 400 500 600 700

1.0 (b)

." !: " , , , . i:

0.5 I : , ' , \ ,

:' \ o '.""- .--~ \ 300 400 SOO 600 700

1.0 (d)

0.5

I: , , / , , , , , : ~ / , , , ; : , ' , ,

} ~ / ,

/ , .... ,/ \

o "' .... ,.' '. .. ,,' " 300 400 500 600 700

1.0 (f)

0.5

300 400

r, " /'

f \ , , I '

; \ , , , , / \" ~,/ - ..

500 600 700

Wavelength (nm)

191

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.),

Academic Press, New York (1979) pp. 257-292. 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.),

Academic Press, New York (1979) pp. 493-520. 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 natureidentical 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, functions 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 sixmembered (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

R

C,! end-group types

.'

/(

16

17 17

18

x

Figure 7.2 Structures of the seven carotenoid end-group types.

were first isolated. In recent years, however, a semi-systematic nomenclature 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 carotenoid 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

Actinioerythrol Antheraxanthin Astaxanthin Bixin Canthaxanthin Capsanthin Capsorubin a-Carotene ~-Carotene "V-Carotene I)-Carotene E-Carotene ,-Carotene Citranaxanthin Croce tin a-Cryptoxanthin ~-Cryptoxanthin Isocryptoxanthin Isozeaxanthin Lactucaxanthin Lutein Lutein-5,6-epoxide Lycopene Neoxanthin Neurosporene Norbixin Phytoene Phytofluene Violaxanthin Violerythrin a-Zeacarotene ~-Zeacarotene Zeaxanthin

Semi-systematic name

3,3' -dihydroxy-2,2' -dinor-~,~-carotene-4,4' -dione 5 ,6-epoxy-5 ,6-dihydro-~,~-carotene-3,3' -diol 3,3' -dihydroxy-~,~-carotene-4,4' -dione methyl hydrogen 9' -Z-6,6' -diapocarotene-6,6' -dioate ~,~-carotene-4,4' -dione 3,3' -dihydroxy-~,K-caroten-6' -one 3,3' -dihydroxY-K,K-carotene-6,6' -dione ~ ,E-carotene ~, ~-carotene ~,IjI-carotene

E,IjI-carotene E,E-carotene 7,8,7' ,8' -tetrahydro-IjI,IjI-carotene 5' ,6' -dihydro-5' -apo-18' -nor-~-caroten-6' -one 8,8' -diapocarotene-8,8' -dioic acid ~,E-caroten-3-ol ~,~-caroten-3-ol ~, ~-caroten-4-ol ~,~-carotene-4,4' -diol E,E-carotene-3,3' -diol ~,E-carotene-3,3' -diol 5 ,6-epoxy-5 ,6-dihydro-~ ,E-carotene-3,3' -diol IjI,IjI-carotene 5' ,6' -epoxy-6,7-didehydro-5,6,5' ,6' -tetrahydro-~,~-carotene-3,5,3' -triol 7 ,8-dihydro-IjI ,IjI-carotene 6,6' -diapocarotene-6,6' -dioic acid 7,8,11,12,7' ,8' ,11' ,12' -octahydro-IjI,IjI-carotene 7,8,11,12,7' ,8' -hexahydro-IjI,IjI-carotene 5,6:5' ,6' -diepoxy-5,6,5' ,6' -tetrahydro-~,~-carotene-3,3' -diol 2,2' -dinor-~, ~-carotene-3,4 ,3' ,4' -tetrone 7' ,8' -dihydro-E,IjI-carotene 7' ,8' -dihydro-~,IjI-carotene ~,~-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, respectively.

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 chlorophylls. The leaves of virtually all species contain the same main carotenoids, 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 wellknown 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 coloration.

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 longerlived 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 organisms 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 protoporphyria, a condition in which the patients are extremely sensitive to light because of an abnormality in haem synthesis, which leads to the accumulation 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 inflammation. ~-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 protection 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 conditions, ~-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


Monoterpenes •• ----

Mevalonic acid (MVA)

! Isopentenyl diphosphate e'UD"r--

Geranyl diphosphate

e," (GD'l---m'

Dimethylallyl diphosphate (DMADP)CS

Sterols

Tr'"" Sesquiterpenes •• --- Farnesyl diphosphate

CI5(FDP) ---.. t Squalene

Diterpenes

Phytol ----I Chlorophyll

!--IDP

Geranylgeranyl diphosphate C20(GGDP) r m

,

Long-chain polyprenols Ubiquinone Plastoquinone

Phytoene

I 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 biosynthesis 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

MVA

E"" ...... ---------- 1 GGDP

CPhytoene ----------- 1 ormatIOn

Phytoene

D"" .. "'oo---------- 1 Lyeopene

c,,';,.;;oo ---------- 1 f3-Carotene

""''''''''°"'''---------1 Xanthophyll.

Figure 7.4 Summary of the main stages of carotenoid biosynthesis.

205

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 azeacarotenes. 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 biosynthesis. 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 -

Ca

rote

ne

(:J

-C

aro

ten

e

I L

yco

pe

ne

j

0-

Ca

rote

ne

E

-C

aro

ten

e

Fig

ure

7.6

The

bio

synt

hesi

s o

f m

onoc

ycIi

c an

d di

cycI

ic c

arot

enes

fro

m l

ycop

ene.

a -

Ca

rote

ne

i :2 ~ s


•

/.\ ~.. ec--. Cc'"

{3 -ring €-ring r-ring

Figure 7.7 Mechanisms for the alternative formation of ~-, "Y- and E-rings from the acyclic precursor, Iycopene.

H )

~~P _-..c;..)_ Ii

H ~' , p

,,.:::; , 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 Capsicum or tomato fruit, the carotenoid content increases several fold during ripening.

CAROTENOIDS 209

~ ~~, ~' ~"O I / 0

~ ... M' " I ~ L ~ ... ~c~·-·

HO~~~~ HoM-oH

/ ~

/ ~ '.

--OH '.

Figure 7.9 Scheme for the formation of different end groups of plant carotenoids.

Carotenoid contents and compositions in plants are affected by environmental 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 invertebrate 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 unsubstituted 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 (C22 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 highdensity 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, carotenodermia 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 bisgentiobioside (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 introduced onto the market in 1954. Several companies now manufacture l3-carotene by processes based on Grignard reactions, enol ether condensations, 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 waterdispersible 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 overemphasized. 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 lipoxygenase 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 carotenoproteins, 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 geometrical 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 subsequently 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 contaminants by the polymerization of acetone. Details of recommended procedures 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 chromatography 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 characterization.

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 corresponding 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-epoxycarotenoids. 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 virtually 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

Table 7.2 Solvents suitable for the elution of different groups of carotenoids in column chromatography on alumina, grade III

Solvent system

Light petroleum (petrol) (or diethyl ether

Carotenoids separated

Carotenes

Carotene epoxides Xanthophyll esters Monooxocarotenoids

217

(ether)):petrol (1:99) Ether:petrol (5:95) Ether:petrol (10:90) Ether:petrol (20:80) Ether:petrol (1: 1) Ether (or ethanol:ether) (5:95) Ethanol:ether (1:4)

Dioxocarotenoids, Monohydroxy carotenoids Dihydroxy-more polar xanthophylls 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-toluenepetrol (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 UVvisible 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. 'dihydroxyfraction') 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% triethylamine) 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 carotenoids 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 throughout 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 chromophore 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 absorption 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 Zdouble 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 determination 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% b I em

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% b 1 em

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. b The 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 determined, 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 carotenoids 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 structural 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-hydroxygroup 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 enormously. 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 waterdispersible. The main oil-dispersible forms consist of solutions or suspensions of micronized carotenoid crystals in a vegetable oil. These preparations 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 waterdispersible 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 concentration. 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, popcorn, potato products and many others.

Water-dispersible forms of l3-carotene, 8'-apo-l3-caroten-8'-al and canthaxanthin, 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 considered.

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 goldenyellow 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 accumulate 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 investigations 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 commercial 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 Haematococcus, 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 canthaxanthin) 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 carotenogenic 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

o

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

o

80 astaxanthin


HOOC "

bixin

o

canthaxanthin

o

~OH

HO capsanthin

o

o capsorubin

a-carotene

CAROTENOIDS 235

B-carotene

y-carotene

6-carotene

€-carotene

<:-carotene

o

citranaxanthin


COOH HOOC "

crocetin

HO a-cryptoxanthin

HO B-cryptoxanthin

OH isocryptoxanthin

OH

OH isozeaxanthin

HO

HO

HO

~.' M;0H

CAROTENOIDS

lactucaxanthin

lutein

lu tei n-5,6-epoxide

Iycopene

neoxanthin

237

OH .. '

... OH

OH .. '

OH


neurosporene

COOH HOOC

nor-bixin

phytoene

pbytofluene

retinol

CAROTENOIDS 239

08

80 violaxanthin

o

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. MathewsRoth 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. LiaaenJensen 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. LiaaenJensen 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. 1 B: 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,

eds), Pergamon, Oxford (1982) pp. 135-153. 86. Buchecker, R. and Noack, K. in Carotenoids, Vol. IB: Spectroscopy (G. Britton,

S. Liaaen-Jensen and H. Pfander, eds), Birkhauser, Basel, Boston and Berlin (1995) pp.63-116.

87. Bernhard, K. and Grosjean, M. in Carotenoids, Vol. IB: Spectroscopy (G. Britton, S. Liaaen-Jensen and H. Pfander, eds), Birkhauser, Basel, Boston and Berlin (1995) 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,

eds), Pergamon, Oxford (1982) pp. 309-328. 92. Brush, A.H. in Carotenoids as Colorants and Vitamin A Precursors (J.C. Bauernfeind,

ed.), Academic Press, New York (1981) pp. 539-562. 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, ed.), Academic Press, New York (1981) pp. 745-754.

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, temperature, O2 , 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, biotechnology) that would make purer and more stable preparations available. 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 betaxanthins, 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 antilipoperoxidant 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, biosynthesis, 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 comprehensive 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 betalains 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 pigments 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 respiration [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 C6C3C6 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 C6C3C6 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 anthocyanidin, 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 consequence, 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 phydroxybenzoic 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 anthocyanins 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

e.g. flavone

acyl group sugar

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 glycosides 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 plantsa

Botanical name Common name (organ)

Anthocyanins present References

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 3-rutinosides

88

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-sophoroside, 3-

109-113

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-diglucosides and 3-rutinosides

120, 121

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

a Abbreviations: 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 accumulation 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-Bphotoreceptor [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 methyltransferases [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 O2 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 reduction of flavanones to flavan-4-0Is, precursors of the unusual 3-

ANTHOCY ANINS AND BETALAINS

I POOtooyn-1 co, + H,O

......... ~

co, t-o® " CH,

CHO I

HCOH I

HCOH

tH,O®

~/ I ShIdmI< Acid Pathway I

~

6;~6~ CInnamiC AcId

COOH Co-SCoA

C4H 0 .4CL 0 - -OH OH

L -Phenytolanlne p-Coumar1cAcid p-Coumar1cAcid

HOWO : I OH CHI

I --~

HO

r

OH 0

Flavanone (colo!1ess)

F3H 1

OH 0

FlaVan 3-d (colo!1ess)

OH

Anthocyanin (coloUred)

OH

OFR -+H,

~J FGT +--

HO

HO

r

Chalcone (yelloW)

OH OH

leucocyanldln (colorless)

FOH l-H,O

OH

Anlhocyanldln (coloUred)

OH

255

3AcelylCoA

3CO'-i

3 MaIonyf-CoA

CHS

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 putative dehydratase [46, 147, 171]. Since the naturally occurring anthocyanidins 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. Uridine 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 aromaticCoA 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, dihydroflavonoids, are further transformed to anthocyanidins by membraneassociated 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 anthocyanoplasts 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 instability; 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 knowledge 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 interdependence 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 fundamental 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 quinonoidal (anhydro) bases (collectively denoted by A), the red flavylium cation (AH+) and the colourless carbinol (pseudo)bases or hemiacetals (collectively denoted by B; the equilibrium concentration of B4 is usually considered to be negligible) and chalcones (CE and Cz , collectively denoted by C). At neutral or slightly acidic pH the anthocyanins exist predominantly in their non-coloured forms. The colour of anthocyanincontaining 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 acidbase or proton transfer (Ka '), hydration (K{,) and ring-chain tautomeric (KT ) reactions, where:

K; = ([A]/[AH+])aH+

K': = «[B] + [C])/[AH+])aH+

KT = [C]/[B]

(8.1)

(8.2)

(8.3)


ccB° w90 I . ~ ....:....: o ~

" I 4 • ~I ....: ~ ~ 0GIy 0GIy

0GIy 0GIy

A· ~. 7

*00 * 4 ~ I ...-::: ~ ~ 0GIy 0GIy

OGIy 0GIy

A, A,.

~~ A

t OGIy 1 ~

AH'

~ +H,O/-H"Y ~+H'OI-H" OH

OH

HO

0GIy

B,

B.

-11 OIl

OH

very slow GIyO OH 4 •

0GIy

OH C,

C,

Figure 8.4 Structural transformations of anthocyanins (i.e. cyanidin 3,5 diglycoside): flavylium cation (AH+); carbinol (pseudo)bases or hemiacetals (B4, B2); chalcones (CE , Cz); neutral quinonoidal bases (A4" A7); ionized quinonoidal bases (A4" A7-). Gly = glycosyl

moiety.

100

c: 0 ;:; IV ~ -c: Q) (.) c: 0 50 0 Q)

> ;:; IV Cii 0::: ~ 0

0

ANTHOCYANINS AND BETALAINS

Flavylium ion predominates

Mixture of flavyllum cation and quinonoidal

bases

4

B

c A

5

Neutral qulnonoidal bases predominate

259

6

pH

I

I Mixture of : neutral and I ionized quinonoidal 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; pKh ' = 3.38; p(Kh ' 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 Kh ' 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 Ka') 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. A4"

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. A4,-, A7-).


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 hemiacetal (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 (CE ) is extremely rapid, i.e. in the range of 1 s or less. However, the cis-trans equilibrium of CE with the corresponding (Z)-form occurs very slowly [192]. The hydration-<iehydration equilibrium is usually characterized by pKh ' values (pKh ' = -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 reduction in the value of Kh ' is sufficient for such stabilization, whereby the neutral and/or ionized quinonoidal bases are both the kinetic and thermodynamic products of equilibrium reactions. In vivo, copigmentation (see section 8.4.4.8) strongly favours the anthocyanin chromophores by reducing Kh ' 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 pHstability (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 ~ CE ) 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 carboxyaldehydes (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 browncoloured 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 kinetics. The mechanism of anthocyanin degradation appears to be temperature 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 temperaturedependence of thermal degradation.) It would appear, then, that anthocyanins 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 maximum 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

262

~y

Enzymattc


O OH Degradatton I Y OxIdized an1hocyanln - products

h OH

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 (HzOz), formed during aerobic oxidation of the ascorbic acid. Significant bleaching of anthocyanincontaining juices occurs when HzOz is added [231]. Nucleophilic attack at the C-2 position of anthocyanins by HzOz cleaves the pyrylium ring to yield a chalcone, which subsequently breaks down to various colourless esters and coumarin derivatives [59, 232]. In acid solution, HzOz was shown to oxidize acylated 3,5-diglycosides directly to acylated O-benzylphenylacetic acid esters [233]. The oxidation products may partake in further degradation 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]. Copigmentation 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 effectively inhibited by S02 [258], bisulfite, dithiothreitol, phenyl-hydrazine and cysteine [248, 259] and gallotannin [251]. The stabilizing effect of gallotannin 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 ftavylium salts with nucleophiles occurs readily when the C-5 position is unsubstituted [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 compounds 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 purpurinidin 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 accelerated 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 (aw )' Reduced aw is associated with a reduced rate of anthocyanin degradation [217]: the hydration of anthocyanin chromophores to colourless species becomes less favourable as water becomes limiting. Indeed, dried anthocyanin powders (aw ::; 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 anthocyanins 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 anthocyanins 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 copigmentation [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 facilitates good molecular contact with a copigment possessing similar structural features.

The extent of the bathochromic shift and increase in maximum absorption (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 Aureusidina 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-glucosidea 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

a Formed 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 copigmentation 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 absorption 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) +/EAH+) 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 copigmentation (Figure 8.2) [27, 31, 53, 70, 188, 197,291,295]. The chiral stacking provides an efficient means of protecting the anthocyanin chromophore 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, copigmentation 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 competition can take place between co-pigmentation and self-association reactions. 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 selfassociation 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 selfassociation of neutral quinonoidal bases or of flavylium cations. Selfassociation 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 nonenzymetreated 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 conjunction with anthocyanins, likely because of their similar biosynthetic pathways. 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 anthocyanins are progressively and irreversibly replaced by polymeric pigments through self-association reactions. Such polymeric material is less pHsensitive 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, ethylene glycol, propylene glycol, cold acetone, acetone/methanol/water mixtures 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 purification or use as a food colorant. Depending on the means of extraction, decreasing the ratio of extraction solvent to plant material, as in continuous 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 (BA W) 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-methanolaqueous 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 nondestructive 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 spectroscopic 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 attachment 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 twodimensional (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 available, MS techniques are the preferred methods for analysis of anthocyanin structure ,[312]. These techniques are often interfaced with liquid chromatography 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, malvidin, 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 Harborne [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 (AICI3) 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 absorption 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 subsequent chromatography of hydrolysis products to facilitate their identification [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 structureretention relationships have been reviewed by Strack and Wray [32, 312). The chromatographic mobility (Rf 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 characterization by all chromatographic methods requires comparison with authentic anthocyanin, aglycone and/or sugar standards. Although anthocyanins 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 hydrolysed, 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 unaffected 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 conveniently 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 compounds 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]. Considerable 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 measurements 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 estimated 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


0.7 ~max 468

0.6

Q) 0.5 () c C'il

..0 ~

0 0.4 (J) ~ 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 concentration 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 concentration 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 chromatography. Spectral measurements are usually of limited use since the absorption spectra of most anthocyanins are so similar [360]; yet, as previously mentioned, molar absorptivities have been reported and used for quantitation of a number of anthocyanins. Quantitation of individual anthocyanins has traditionally involved their separation by paper or thin layer chromatography, followed by photometric determination of pigments in situ by


reflectance or transmittance densitometry [55, 111, 139, 406]. Reversedphase 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 synthetic 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 anthocyanins 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 cranberry [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]. Anthocyanin 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 condensed 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 cellaggregate 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 encourage reactions with endogenous constituents that can lead to their stabilization 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 applications 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 strawberry 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 reactions (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 characterized by having Rand R' groups that do not extend the conjugation of the 1,7 -diazaheptamethin system. However, if conjugation is extended, whereby Rand R' comprise a substituted aromatic ring (i.e. cyclodopa), the


COOH

H HOOC ....

Figure 8.8 Structural transformations of the betalain chromophore.

281

COOH

chromophore is red and characterized as a betacyanin [3, 4]. All betacyanins 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-hydroxynorvaline. 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].

Tab

le 8

.4 S

truc

ture

s o

f so

me

know

n be

tacy

anin

s

H

Com

poun

d S

ubst

itut

ion

patt

ern

Rs

Bet

anin

j3

-glu

cose

A

mar

anth

in

2' -O

-(j3

-glu

coro

nic

acid

)-j3

-glu

cose

C

elos

iani

n-I

2' -0

-[ O

-(p-

coum

aroy

l)-j

3-gl

ucur

onic

aci

dj-j

3-gl

ucos

e

Cel

osia

nin-

II

2' -

0-[

O-(

tran

s-fe

rulo

yl)-

j3-g

lucu

roni

c ac

idj-

j3-g

luco

se

Ires

inin

-I

2' -O

-(j3

-glu

curo

nic

acid

)-6'

-O-(

3-h

ydro

xy-3

-m

ethy

lglu

tary

l)-j

3-gl

ucos

e P

rebe

tani

n 6'

-0-(

S03

H)-

j3-g

luco

se

Phy

lloc

acti

n 6

' -0

-( m

alon

yl )-

j3-g

luco

se

Riv

inia

nin

3' -

0-(S

03H

)-j3

-glu

cose

L

ampr

anth

in-I

6'

-O-(

p-co

umar

oyl )

-j3-

gluc

ose

Lam

pran

thin

-II

6' -O

-fer

uloy

l-j3

-glu

cose

B

ouga

invi

llei

n-r-

I j3

-sop

horo

se

Gom

phre

nin-

I H

G

omph

reni

n-II

H

Gom

phre

nin-

III

H

Gom

phre

nin-

V

H

CO

OH

Bot

anic

al s

ourc

e

R6

H

Bet

a vu

lgar

is

H

Am

aran

thus

tri

colo

r H

C

elos

ia c

rist

ata

L.

Che

nopo

dium

rub

rum

H

C

heno

podi

um r

ub ru

m

H

lres

ine

herb

stii

H

Bet

a vu

lgar

is

H

Phy

lloca

ctus

hyb

ridu

s H

R

ivin

ia h

umil

is

H

Lam

pran

thus

spp

. H

La

mpr

anth

us s

pp.

H

Bou

gain

ville

a gl

abra

var

. sa

nder

iana

j3

-glu

cose

G

omph

rena

glo

bosa

6'

-O-(

p-co

umar

oyI)

-j3-

gluc

ose

Gom

phre

na g

lobo

sa

Bas

ella

rub

ra

6' -O

-(tr

ans-

feru

loyl

)-j3

-glu

cose

G

omph

rena

glo

bosa

B

asel

la r

ubra

6'

-O-(

tran

s-fe

rulo

yl)-

j3-g

luco

se

Gom

phre

na g

lobo

sa

Ref

eren

ce

375,

376

377,

378

380

498

509

380

379

380

510

381

381

382

383

497

511

497

511

383

N

00

N

~ Cl ~ t"" ~ o " (":l o S ~ ~


Table 8.5 Structures of some known betaxanthins

R'O~I .H ~ +.'

R,O N eOOH

H~~COOH I H

Compound 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 betalainproducing families have been identified: Aizoaceae, Amaranthaceae, BaseJlaceae, Cactaceae, Chenopodaceae, Didiereaceae, Holophytaceae, Nyctaginaceae, 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 accumulate 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 condensation 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 .H ~ HO '" . COOH

Tyrosine

HO~.H _____ > HOWH HOV H,NA COOH HO '" ~ COOH

H L-DOPA

1

5-Cyclodopo

I, ·.H Ho~~Acoo

285

[HO~'" J HO~ .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 .1 other than cyclodopa 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 accumulation 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 synthesis 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. Comparatively 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 bathochromic 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 degradation 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


G~WH

00 ~ \0= WZJlCOOH

I H

lsobetonln

1~~~ GuOWH

HO::-- 6+ 'COOH ~~OH •

H I HOOC COOH

I H

Betonin

l~ GluO~

~~H HO::-- 6COOH

+ CO,

H I H COOH

I H

Decorboxylated Betonln

GIuOW ::-- I + "H + HO N COOH

~ Cyciodopa-5-O-glucoside

~t~ I H

Betolamlc Acid

Figure S.lO Mechanisms of betanin degradation,

287

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 130aC 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 nucleophiles, 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 degradation 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 degradation 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, aluminium) accelerate the rate of betanin degradation [491, 492]. As electron donors or acceptors, depending on their oxidation state, metal ions may stabilize the electrophilic centre of betalains resulting in rearrangement of associated bonds, destruction of the chromophore and concomitant 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 (aw). 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 fourfold 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 aw 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 generally 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 NH4N03 , KN03 , 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 procedures have been used over the years to separate/purify betalains from a variety of sources. However, most of these procedures are timeconsuming. 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 identification 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 electrophoretic 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 cooccurring 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 considerably 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 chromatographic or electrophoretic behaviour [512]. For example, on polyamide or cellulose-based resins:


1. The retention of betacyanins decreases with increased glycosyl substitution.

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 nonacylated 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 spectrometers 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 chromatography 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 betaxanthins 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 identification of the resulting polyols; or

2. Pigment methylation with Mel/AgO in dimethylformamide and subsequent 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 absorbance 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 betacyanin 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 simultaneously. 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 colorants 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 gelfiltration 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 offflavours, (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 application 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 demonstrated 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 physicochemical properties. However, if current legislation restricting the use of more highly purified extracts and sources other than grapes (and grape byproducts) 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. selfassociation 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 application of biotechnology (e.g. cell suspension or tissue culture) for largescale 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 assistance.

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, New York (1982) Chapter 3.

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,

New York (1982) Chapter 1. 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. 65. Terahara, N., Saito, N., Honda, T., Toki, K. and Osajima, Y. Tetrahedron Lett. 31

(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

York (1982) Chapter 6. 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

York (1982) Chapter 2. 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. 299. Asen, S. Acta Hortic. 63 (1976) 217-223. 300. Asen, S. Acta Hortic. 41 (1974) 57-68. 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. 307. Hoshino, T. Phytochem. 31 (1992) 647-653. 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

York (1982) Chapter 7. 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. 361. Hoshino, T. Phytochem. 23 (1986) 829-832.


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, Compositae, 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 determination. 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 highperformance 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



(c)

LESS COMMON NATURAL COLORANTS

GLU-I CAFF

(d)

6<fLU-<AFF~LU 'CAFF-GL

6 lJ-CAFF

311

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--

400

/ /

/

500 -----

600

WAVELENGTH (nm)

313

l 700

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.

314

90

'" '" :J ~ 80

...l ... ~ I: :J 70 :r:

'" '" :J

80

~ 60

~ '" 40 t2 ... ~ I: 20 :J :r:

5

3

c '" C 2 o U

c '"


pH 5.5 pH 3.5

90

80

70

60

50

10 20 30 40 50 60 0 10 20 30 40 50 60

70

60

50

40

30

20

10

10 20 30 40 50 60 10 20 30 40 50 60

~ o+-~~~,-~,-~,-~,-~ 0: 0 10 20 30 40 50 60

Time(Weeks) o 10 20 30 40 50 60

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

LESS COMMON NATURAL COLORANTS 315

(a)

R

(b)

(c)

(d) R

Figure 9.4 Carotenoid pigments of annatto «a) bixin, R = CH3 ; and norbixin, R = H; (b) trans-bixin; and (c) C17 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 monomethyl 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 steamsoftened 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 inorganic 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 demineralized 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 applications 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 digentiobioside 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 alphatocopherol 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, anthocyanins, Iycopene, curcumin, chlorophyll and iridoid preparations could be stabilized by cyclodextrins. Cyclodextrins are cyclic non-reducing oJigosaccharides 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, geniposide, shanzhiside, gardoside, methyldeacetylasperuloside, genipingentiobioside, 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 (c) (d)

H OOCH3

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. Apparently, 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 Y oshizumi [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 manufactured 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 egglaying 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 colorants 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 'naturealike' ,it is likely that considerable testing for toxicity would be required.

Schul [96] discussed the production, chemistry, applications and particularly 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

H

(b)

323

Figure 9.9 Anthraquinone pigments in lac. (a) Laccaic acids. A: R = CH2-NH-COCH3 ;

B: R = CH20H; C: R = CH(NH2)COOH; E: R = CHr NH2• (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-Meether. 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 bioanthraquinones, 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 spraydrying 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


H

Figure 9.10 Pigments of turmeric. Curcumin: RI = Rz = OCH3 . Demethoxycurcumin: RI = H; Rz = OCH3 . Bisdemethoxycurcumin: RI = Rz = 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 individual estimation is important because the extinction values differ considerably 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 encourage even wider use of turmeric in foods. Turmeric is already widely used in the food industry in canned products, pickles, soups, mixes and confectionery. Turmeric is the usual yellow colorant in mustard and pickled cauliflower.


(b)

H

(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. Carthamin, 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 isocarthamidin.

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 = C7H1S.

(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 watersoluble 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 aminoacid 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)

M'~' (c)

~ '-~ HO i N

HO bH OH 0 Pr N'R'

(b) R1 (d)

Wi"' 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 production by selection of mutants induced by UV light or N-methyl-N-nitronitrosoguanidine [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 cholesterol 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 particular, 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 pigments. 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 considerable 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 isolation 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 corresponding 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 trichotomum. 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 cycloleucomelone, 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 Actinomycetes (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 Shreektoniah [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 Aspergillus 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, canthaxanthin, 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 lanthinobacterium 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 background 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 (Ipomoea 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 compounds 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. 3. Terahara, N., Saito, N., Honda, T., Toki, K. and Osajima, Y. Tetrahedron Lett. 31

(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. 162. Saito, K. and Fukushima, A. Food Chem. 29 (1988) 161. 163. Saito, K. Food Chem. 36 (1990) 243. 164. Saito, K. and Fukushima, A. Food Chem. 32 (1989) 297. 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, 164-166,173,181,185,186,191, 219-224, 276, 313

acceptable daily intake 114-116, 192 Actinidia equina 230 actinioerythrol 200, 222, 230, 233 Actinomycetes 333 ADI see acceptable daily intake Adonis spp. 102 Agaricales 31 Agaricus bisporus 32 Agrobacterium rhyzogenes 82, 89 Ajuga reptans 85 Albatrellus spp. 31 alcoholic drinks see drinks alfalfa 146 algae 11, 27, 76, 82, 93-98, 122, 170-190,200

Alkanna tinctoria Tausch. 322 Allium cepa 252 allomelanin 16 allophycocyanin 177-185 Alternaria porri 324, 333 Amanita muscaria 18, 32, 283, 284 amaranth 47,61 amaranthin 282 Amaranthus spp. 18, 282, 285 amino acids 267 amphibians 28 Anabaena spp. 175 Anabaenopsis spp. 175 Anacystis nidulans 175, 177, 178 animals 200,209 annatto 47-52, 63, 72, 84, 101, 197-239, 315

antheraxanthin 200, 232, 233 anthocyanin(id)ins 14, 27, 53-59, 67, 85-88, 120, 244-296

Anthracophylum 332 anthraquinones 15, 321-325 Aphanizomenon ftos-aquae 175 Aphanocapsa spp. 175 Aphanothece spp. 175, 190 Aphidoidea 321 apigeninidin 248 Aplysia spp. 172 apo-carotenoids 121, 199 apples 53, 252 application forms 45,49, 60, 66, 70, 227

applications 51,57,59,63,64,72, 228-230

Aralia cordata 279 Armenian red 65, 321 Arthrobacter 333 Arthrospira maxima 175 Ascomycetes 30, 31 Aspergillus spp. 103, 294, 319, 333 astaxanthin 91, 95, 97, 101, 200, 210, 233

aubergine 53, 253 aurantinidin 248 aurones 267

Bacillariophyta 141 Bacillus spp. 98, 319 bacteria 37, 46, 87, 200 bacteriochlorophy\ls 141, 148 bakery products see confectionery Basella rubra 282, 334 Basidiomycetes 31 beetroot 18,59-64,67,281-285,335 benzoic acids 267 benzoquinone 15 Berberis thumbergii 332 betacyan(id)in 18, 59-64 betalains 17,27,59-64,85-87,125, 280-296

betan(id)in 18,47,84, 103 Beta vulgaris 18, 281-285, 335 betaxanthin 18, 59-64 beverages see drinks bilberries 278 bile pigments see bilins bilins 170-193 bilirubin 162 biliverdin 162 bioprocessing 98-103 biotechnology 80-108, 188,231 birds 28, 200, 229 Bixa orellana 27,47,84,85, 101,201, 212, 315

bixin 47-52, 72, 85, 197-239 blackberry 253, 314 see also Rubus spp. blackcurrant 53, 55, 253 see also Ribes spp.

Blakeslea trispora 91, 231 blood orange 252 blueberry 253,278 see also Vaccinium spp.

344

Boletales 33 Boletopsis leucomelaena 332 Bougainvillea glabra 282 bougainvillein 282 Brackenridgea zanguebarica 332 Brassica spp. 252, 334 Brevibacterium spp. 91,97,232 brilliant blue FCF 191-192 Bromeliacceae 310 Bulgaria spp. 31 Butea monosperma 321 butter see dairy products

cake mix see confectionery Calothrix spp. 175 cancer 118, 230 Candida utilis 103, 294 candies see confectionery Cantharellus spp. 31,91 canthaxanthin 91,97, 121, 197,200, 222, 229, 233

capensinidin 248 capsaicin 85 capsaxanthin 200, 222, 233 Capsicum spp. 77, 85, 103, 211 capsorubin 77, 200, 222, 233 caramel 2, 77, 124 carbon black 78, 125 carcinogens 113 carmine ~9, 123 carminic acid ~9 carmoisine 61 carotenes 12,27,76-77,86,91,99, 100,

105, 120, 197-239 carotenodermia 210 carotenoids 76, 197-239 Carpobrotus ancinacriformis 281 carrot 76, 83, 85-88, 100 carthamin 327-328 Carthamus spp. 327, 335 Cassia mimosoides 324 cattle 209 Cedrela spp. 317 cell culture 88-90 Celosia cristata 282 celosianin 282 Centrospermae 18, 32 chalcone 258 Chenopodium spp. 282 cherry 53, 55, 252 Chlorella spp. 91, 93 chlorins see chlorophylls Chlorobiaceae 141 Chloroflexaceae 141 Chlorogloea fritschii 175 chlorophyllide 134 chlorophyllin 134 chlorophylls 27,74, 125, 131-155

INDEX

Chlorophyta 141 Chlorosplenium spp. 31,32 cholesterol 119 Chondrus crispus 193 Chromatiaceae 141 Chromobacterium spp. 37 chromones 34 Chroogomphus spp. 33 Chroomonas spp. 141 Chrysophyta 141 cinnamic acids 267 cinerarin 311 cis-bixin see bixin cis-norbixin see bixin citraxanthin 200, 222, 233 Citrus sinensis 252 Clerodendrum trichotomum 332 Clitoria tematia 266, 278, 310 cloning 85 Clostridium tetanomorphum 177 Coccochloris elabens 182 Coccoidea spp. 321 Coccus cacti L. 64 cochineal ~9, 123, 320 cocoa 125 colour blindness 5 Comanthus parvicirrus 324 Commelinaceae 310 Compositae 19,310 confectionery 52, 58, 64, 68, 73, 76, 227,322

co-pigmentation 267 copper chlorophylls 75 Coprinus spp. 33 Coptis japonica 332 com 253 coronary heart disease 119 cosmetics 65 coumarin 267 cranberry 55,253,278 crocetin 200,222,236 crocin 77, 85 Crocus spp. 27,85,212,317,335 Cryptomonas spp. 175 Cryptophyta 141 cryptoxanthin 200,223,236 Curcuma spp. 27,69,325 curcumin 47,69-74,76 cyanidin 53, 248 Cyanidium caldarium 171, 175, 179, 184 Cyanophyta 141 cyanobacteria see Cyanophyta Cyphomandra betacea 252 cytochromes 165

Dacrymycetales 31 Dactylopiidae 321 Dactylopius coccus costa 16,64, 321

INDEX 345

dairy products 51,58,60,63,68,72,76, flavonoids 13, 320 flavonol 267 Flexibacter elegans 332 fluorescence 173, 181 Fornes fomentarius 33

77,227,228 Dalbergia candenatensis 332 Daldinia spp. 31 Daucus carota 76, 83, 85-88, 100, 279 delphinidin 53 demethoxycurcumin 69 depsides 34 Dermocarpa violacea 175 Dermocybe spp. 32, 333 desserts 68 diferoylmethane 69 dihydrochalcone 267 Dioscorea alata 252 Discomycetes see Ascomycetes dopaxanthin 283 drin~ ~,~,M,~,~,77,n1 duhat 278 Dunaliella spp. 38,77,93,94, 104, 122, 188, 230, 231

dry mixes see mixes

eggs 228 Eichhornia crassipes 146 Elaeis guineensis 212 elderberry 55 Enterobacter spp. 334 E numbers 43 Erwinia spp. 204, 232 europinidin 248 erythrosine 191-192 Escherichia coli 105, 204, 232 EU see European Union European Union 43,50,61,67,71,77, 116, 192

Euglenophyta 141 eumelanin 19 Euphorbia milli 279 Evening Primrose 334 extinction coefficients 47, 48, 53, 54, 60, 61,65,69,148,165,167,183,192, 222-223

extraction 49,60, 66, 70, 149, 165, 182, 215,257,286

FDA see Food and Drug Administration fermentation 85, 90-98 Festuca pratensis 153--154 Ficus spp. 252 fig 252 fish 28, 52, 200, 229 flamingos 210 flavanone 267 flavans 267 flavin 20 Flavobacterium spp. 91, 96, 232 flavone 267

Food and Drug Administration 43, 44, 115-118

Fragaria spp. 252 Fremyella diplosiphon 175, 181 frozen products 59, 73 fungi 28, 200 Fusarium spp. 324, 333

Gardenia spp. 77, 125,317 genetic engineering 83, 105 Gentianaceae 310 gentiodelphin 311 Gibasis geniculata 314 Gibberella spp. 334 Gladiolus segetum 324 Glaucosphaera vacuolata 175 Glottiphyllum longum 283 Glycine maxima 252 Glycyrrhiza spp. 333 Gomphrena globosa 282 gomphrenin 282 grape(s) 53-59, 253, 314 grape extracts 53--59 grape skin 53--59

haem 157-170 haematin 159, 167 Haematoccus spp. 91, 93, 95, 96, 102, 104

Haematomma spp. 35 haemin 159, 167 health 202, 229 Heliotales 31 Helvella spp. 31 heme see haem Hemiselmis spp. 175,278,334 Hibiscus spp. 55 high-performance liquid chromatography 89, 118, 169, 187,211, 217-219,272,278,291,310,316

hirscutidin 248 Homarus gammarus 231 HPLC see high-performance liquid chromatography

huckleberry 253 see also Vaccinium spp.

Hydrocoleum spp. 175 Hygrocybe spp. 32, 284 hypericin 15 Hypericum 32 Hypomyces spp. 333 Hypoxylon spp. 31

346 INDEX

ices see dairy products indicaxanthin 18, 283 indigo 2,17 indigo carmine 191-192 Indigofera tinctoria 17 iodinin 22 invertebrates 28, 200 Ipomoea spp. 252, 278, 310, 312, 334 lresine herbstii 282 iridoids 125, 319 iresinin 282 lsatis tinctoria 17 isocryptoxanthin 200, 223, 236 isozeaxanthin 200, 223, 236

jams see preserves lanhinobacterium lividium 334 JECFA see Joint Expert Committee on

Food Additives Joint Expert Committee on Food

Additives 61, 116

Kermococcus spp. 16, 321 Kerria lacca 124 kidney bean 252 Krill 102

lac 124 Laccifera lacca 65, 124, 321 Lactarius spp. 30, 332 Lactuca sativa 201 lactucaxanthin 200, 223, 237 lampranthin 282 Lampranthus spp. 282 Lawsonia alba 16 LDso see lethal dose (50) legislation 42,50,55,61,67,71, 106,

116-Jl7, 192 lethal dose (50) 115 Leguminosae 310 leprarinic acid 33 Letharia vulpina 35 lettuce 201 leucopterin 21 lichens 33 Iichexanthone 36 Liliaceae 310 Litchi chinensis 252 Lithospermum spp. 85-87,324,335 litmus 35 Lobelia spp. 310 lobster 231 lucerne 146 luteins 76,91,200,223,229,237 luteolinidin 248 Iychee 252 Iycopene 119, 122, 198, ZOO, 206, 223, 237 Lycopersicum esculentum 212

Lyngbya lagerheimii l75 Lyophyl/um spp. 30

MAFF see Ministry of Agriculture, Fisheries and Food

Malus pumila 252 malv(in)idin 53, 248 mammals 28 Mangifera indica 252 mango 252 margarine see dairy products Margarodes polonicus 321 marigold 76, 102, 229 mass spectrometry 225, 311 Mastigocladus laminosus 175, 182 maximum tolerated dose 113 meat 52, 69, 77 Medicago sativa 146 Mercurialis leiocarpa 332 metalloproteins 23 Metatrichia vesparium 32 Ministry of Agriculture, Fisheries and Food 192

Mirabilis jalapa 283 miracle fruit 253, 278 miraxanthin 283 mixes 52, 59, 64, 73, 76 monardein 311 monascus 91,328-331,334 Monascus spp. 92, 104 Morchella spp. 31 morning glory - heavenly blue 314 moulds 82 MID see maximum tolerated dose musca-aurin 283 Musophagidae spp. 140 myoglobin 165

naphthaquinone 15 natural colours 42, 44, 46 nature identical colours 42 neoxanthin 200, 223, 237 nettles see Urtica dioica neurosporene 200,206, 208, 223,238 N-heterocyclic compounds 17 NOAEL see No Observed Adverse Effect

Nocardia spp. 333 No Observed Adverse Effect 114-116 Nopalea spp. 65, 321 norbixin see bixin Nostoc punctiforme 175 nuclear magnetic resonance spectroscopy 226

Nyctanthes arbor-tristis 317

Ochrolechia tartarea 34 Oenothera spp. 334

O-heterocyclic compounds 12-24 oil of turmeric 70 oosporeine 32 Opuntia spp. 281,283,321 orcein 35 orcin( 01) 35 Oscillatoria spp. 175 oxygen 51,57,62,67,169,210,261,

288

palm oil 76, 100, 212 pap.ri~a 77, 122 panetm 35 Passiflora edulis 252 passion fruit 252 pasta 228 patent blue V 191-192 patents 311-335 peach 251 pelargonidin 53, 248 Penicillium spp. 31,332 peonidin 53, 248 Perilla spp. 278, 334 petunidin 53, 248 pH 46,50,62,66,67,71,75,184,257, 270,271, 276, 286

phaeomelanins 19 phaeophorbide 134, 144 Phaeophyta (brown algae) 141 phaeophytin 74, 134, 144 Phaffia rhodozyma 31,91,97,98, 102, 103,231

Phallales 31 Pharbites nil 312 pharmaceuticals 81,317 Phaseolus spp. 252, 333 Phellodendron amurense 332 phenazines 20 phenoxazines 21 Phlebia spp. 31 Phormidium spp. 175 photo-oxidation see oxygen photosynthesis 135 Phragmobasidiomycetidae 30 phycobilins 27 phycocyanin 91, 173 phycoerythrin 173 Phycomyces blakesleeanus 231 phyllocactin 282 Phyllocactus hybridus 282 phytoene 200, 206, 223, 238 phytofluene 200, 206, 223, 238 Phytolaccaca spp. 285 phytomelin 20 phytoplankton see algae Plagioselmis prolonga 175 plant tissue culture 82-90 platyconin 311

INDEX

Plectonema spp. 175 plum 252 Polyporales 31 polyporic acid 36 Polyporus spp. 33 ponceau 4R 61, 191-192 PopUlus spp. 279

347

Porphyridium spp. 175, 179, 184, 188 porphyrins 138-140 Porphyrophera spp. 321 Portulaca grandiflora 283 portulacaxanthin 283 poultry feed 76 Prasinophyta 141 prebetanin 282 preserves 58, 68 Prochloron spp. 141 Prochlorophyta 141 Propionibacterium shermanii 177 protohaem see haem protoporphyrin (IX) 158 protozoa 28 Prunus spp. 252 Pseudomonas spp. 22,37 pterins 20 pu\Chellidin 248 Punica granatum 252 purification 149, 165, 182, 215, 270, 290 purines 20 purpurogallin 33 pyocyanine 22 Py"ophyta 141

Quercus coccifera 321 quinoids see qui nones quinones 15

radish 252 raspberry 53,253 red 2G 191-192 red cabbage 53, 55, 252, 278, 314 redcurrant 53, 253 red onion 252 reptiles 28 retinal see vitamin A Rhaphanus sativus 253, 333 Rheum rhaponticum 253 Rhizopus spp. 320 Rhodella violacea maculata 175 Rhodobacter spp. 203 Rhodococcus spp. 97,334 Rhodomonas lens 175 Rhodophyta (red algae) 141 Rhodospirillacae 141 rhubarb 253 Ribes spp. 253 riboflavin 119 Ricciocarpus natans 247

348

riccionidin A 248 Rivina humilis 282 rivinianidin 282 RaceUa tinctaria 34 rocou see annatto root cultures 89 roselle see Hibiscus spp. rosinidin 248 Rubia tinctorium 16 rubrolone 332 Rubus spp. 253 Russula spp. 30

Saccharomyces spp. 103, 232, 294 safety 112-126 saffron 77, 84,212,317 San Red RC 315 sauces 77 Schizothrix alpicola 175 Schleichera oleosa 321 Scleranthus perennis 321 secondary metabolites 82-90 Sepia officinalis 19 Serratia spp. 334 Setcreasea purpurea 310 see also

Tradescantia paUida shikonin 83-87 snacks 52 soft drinks see drinks Solanum spp. 253 soups 77 soy bean 252 Spirulina spp. 91,93, 123, 188, 193 Spongicoccum excentricum 91 stability 51,56,62,67,71, 151, 164,

168, 181, 184,213,257, 286 strawberry 53,252 Streptomyces spp. 22, 332, 333 Streptoverticillium spp. 333 Suillus spp. 33 sulphur dioxide (S02) 51, 57, 63, 67, 72 sunset yellow 47 sweet cherry 252 sweet potato 252, 314 sweets see confectionery Synechococcus 6301 see Anacystis

nidulans Synechosysytis 6701 180 Synsepalum dulcificum 253 Sysyium cumini 278 synthetic colours 42, 190, 212

Tagetes erecta 102,201,212 tamarillo 252 tartrazine 47, 72 tematin 311 tetrapyrroles 9-12

INDEX

tetraterpenoids 12 Thalictrum minus 335 theaflavin 14 thin layer chromatography 215, 274 tissue culture 83 titanium dioxide 125 TLC see thin layer chromatography tocopherols see vitamin E Tolypacladium infiatum 333 Tolypothrix distorta tenuis 175 tomato 212 toxicity 112-114 Tradescantia paUida 249,278, 313, 314 transbixin see bixin Tremallales 31 tricetinidin 248 turmeric 69-74, 119-123, 325-327 tyrian purple 17

Urtica dioica 146 usnic acid 36

Vaccinium spp. 253,278 Verbascum phlomoides 317 vertebrates 28 Viburnum dentatum 278 violaxanthin 200, 223, 239 violerythrin 200, 223, 230, 239 vitamin A 119 vitamin B12 139, 140 vitamin E 136, 137, 140, 154 Vitis spp. 55,85,251,253,279 vulgaxanthin 59, 283

water activity 63, 289

xanthommatin 21 xanthophylls 212 Xanthophyta 36, 141 xanthopterin 21 Xanthorrhia spp. 35 Xenorhabdus luminescens 324

yams 252 yeast see Saccharomyces spp. yoghurts see dairy products

Zea mays 253 zeacarotene 200, 223, 239 zeaxanthin 91,96,200,223,229,239 Zebrina spp. 312, 314

see also Tradescantia pallida zebrinin 311 Ziziphus mauritiana 321

natural food colorants

Documents