sign in sign up
<<
Home , Coumestrol, Daidzin, Formononetin, Genistin, Glycitein, Glycitin

PLEASE TYPE THE UNIVERSITY OF NEW SOUTH WALES Thesis/Project Report Sheet

Surname or Family name: HUTABARAT

First name: LAMB OK SAORITA Other name/s: Abbreviation for degree as given in the University calendar: Ph.D School : Applied Bioscience Faculty: Life Science Title: IN AUSTRALIAN AND INDONESIAN FOODS

Abstract 350 words maximum: (PLEASE TYPE)

ABSTRACT

There is a lack of data on phytoestrogens in Australian and Indonesian foods. These data are needed to help local medical scientists, nutritionists and pharmacologists to elucidate the role of dietary phytoestrogens tn protection against certain , distressing menopausal symptoms and other nutrition-related conditions.

There is no international standard method for analysis of phytoestrogens in foods and thi s makes the collection of reliable analytical data on phytoestrogens very difficult. Reliable methods are also required for labelling foods with their phytoestrogens content. Two isocratic HPLC methods have been developed in this study. The first method was for the separation and quantification of , and from soya . The second method was for the separation and quantification of the , daidzein, genistein, and biochanin A and the , from soya beans. Both HPLC methods used C8 and phenyl columns, 1% acetic acid-water (33 :67, v/v) and -water (33:67, v/v) as eluent to separate all compounds in less than 24 minutes.

The second method exhibited a high degree of precision and high linearity e > 0.999) with the delectability being 47, 82, 76, 75 and 224 nM for daidzein, coumestrol, genistein, formononetin and biochanin A, respectively, and a good recovery (- 100%) of standard comp::>unds from foods. Extraction with acid and heat was used to hydrolyse all the conjugate forms in foods with a maximum yield and a high level of efficiency. The method could identify isoflavones in a wide range of beans and products from Australia and Indonesia from 0.01 up to 75 mg/100 g dry weight basis. Levels varied between different brands, foods and countries. Since the method of analysis had been shown to be highly reliable, other factors such as agricultural conditions, cultivars, stage of maturation, food processing and preparation are likely to have been the dominant factors affecting the results. The high degree of variability in the levels of phytoestrogens makes the task of obtaining representative datas therefore very difficult. In the few cases where label claims were made, they could not be verified by the analytical results for the product.

Declaration relating to disposition of project report/thesis

I am fully aware of the policy of the University relating to the retention and use of higher degree project reports and theses, namely that the University retains the copies submitted for examination and is free to allow them to be consulted or borrowed. Subject to the provisions of the Copyright Act 1968, the University may issue a project report or thesis in whole or in part, in photostat or microfilm or other copying medium. I aiJ '"h"""' ili< ••bhc,.;oo :v University Microfilms of a 350 word abstract in Dissertation Abstracts International (applicable to doctorates only) . .._ _ l...... 1 L...... t..g/.~/.C1L ...... Signature Witness Date

The University recognises that there may be exceptional circumstances requiring restrictions on copying or conditions on use. Requests for restriction for a period of up to 2 years must be made in writing to the Registrar. Requests for a longer period of restriction may be considered in exceptional circumstances if accompanied by a letter of support from the Supervisor or Head of School. Such requests must be submitted with the thesis/project report. FOR OFFICE USE ONLY Date of completion of requirements for Award: 1'-:------;--::~~ 'l...7(3(o2_ Registrar and Deputy Principal THIS SHEET IS TO BE GLUED TO THE INSIDE FRONT COVER OF THE THESIS MBT 613.28 68 CERTIFICATE OF ORIGINALITY

I hereby declare that this submission is my own work and to the best of my knowledge it contains no materials previously published or written by another person, nor material which to a substantial extent has been accepted for the award of any other degree or diploma at UNSW or any other educational institution, except where due acknowledgement is made in the thesis. Any contribution made to the research by others, with whom I have worked at UNSW or elsewhere, is explicitly acknowledged in the thesis.

I also declare that the intellectual content of this thesis is the product of my own work, except to the extent that assistance from others in the project's design and conception or in style, presentation and linguistic expression is acknowledged.

(Signed} •••ic__-----.JI - m mm PHYTOESTROGENS IN

AUSTRALIAN AND INDONESIAN FOODS

by

Lambok Saorita Hutabarat

A thesis

submitted to The University of New South Wales as fulfillment of the requirements for the degree of

DOCTOR OF PIDLOSOPHY

August 2002 UNSW 1 7 OCT 2002 LIBRARY ii ABSTRACT

There is a lack of data on phytoestrogens in Australian and Indonesian foods. These

data are needed to help local medical scientists, nutritionists and pharmacologists to

elucidate the role of dietary phytoestrogens in protection against certain cancers,

distressing menopausal symptoms and other nutrition-related conditions.

There is no international standard method for analysis of phytoestrogens in foods and

this makes the collection of reliable analytical data on phytoestrogens very difficult.

Reliable methods are also required for labelling foods with their phytoestrogens

content. Two isocratic HPLC methods have been developed in this study. The first

method was for the separation and quantification of daidzein, genistein and

biochanin A from soya beans. The second method was for the separation and

quantification of the isoflavones, daidzein, genistein, formononetin and biochanin A

and the coumestan, coumestrol from soya beans. Both HPLC methods used C8 and

phenyl columns, 1% acetic acid-water (33:67, v/v) and acetonitrile-water (33:67, v/v)

as eluent to separate all compounds in less than 24 minutes.

The second method exhibited a high degree of precision and high linearity (? >

0.999) with the detectability being 47, 82, 76, 75 and 224 nM for daidzein,

coumestrol, genistein, formononetin and biochanin A, respectively, and a good

recovery (~ 100%) of standard compounds from foods. Extraction with acid and heat was used to hydrolyse all the conjugate forms in foods with a maximum yield and a high level of efficiency. The method could identify isoflavones in a wide range of beans and bean products from Australia and Indonesia from 0.01 up to 75 mg/100 g iii dry weight basis. Levels varied between different brands, foods and countries. Since the method of analysis had been shown to be highly reliable, other factors such as agricultural conditions, cultivars, stage of maturation, food processing and preparation are likely to have been the dominant factors affecting the results. The high degree of variability in the levels of phytoestrogens makes the task of obtaining representative datas therefore very difficult. In the few cases where label claims were made, they could not be verified by the analytical results for the product. IV

PUBLICATION ARISING FROM THE WORK

RELATED TO THE THESIS

Papers

Hutabarat, L. S., Mulholland, M. & Greefield, H. (1998) Development and validation of an isocratic high-performance liquid chromatographic method for quantitative determination ofphytoestrogens in soya bean. J. Chromatogr. 795: 377- 382.

Hutabarat, L. S., Greenfield, H. & Mulholland, M. (2000) Quantitative determination of isoflavones and coumestrol in by column liquid chromatography. J. Chromatogr. 886: 55-63.

Hutabarat, L. S. Greenfield, H. & Mulholland, M. (2001) Isoflavones and coumestrol in and soybean products from Australia and Indonesia. J. Food Comp. Analysis 14: 43-58.

Abstract

Hutabarat, L. S., Greenfield, H. & Mulholland, M. (1997) Prelimenary studies of phytoestrogens in Australia foods. Proceeding of The Nutrition Society ofAustralia, 2Ft Annual Scientific Meeting. Briesbane, Australia.

Hutabarat, L. S., Greenfield, H. (1999) Phytoestrogens in soybean products from Indonesia.2nd South-West Pacific Nutrition and Dietetic Conference. Auckland, New Zealand.

Hutabarat, L. S., Greenfield, H. & Mulholland, M. (1999) Isoflavones and coumestrol in soybeans and soybean products available in Australia and Indonesia. Third International Food Data Conference. Rome, Italy. v

Hutabarat, L. S., Greenfield, H. & Mulholland, M. (1999) Isojlavones and coumestrol in soybeans and soybean products available in Australia and Indonesia. gth Asian Congress of Nutrition. Seoul, Korea. vi ACKNOWLEDGEMENT

I wish to express sincere gratitude to my immediate supervisor, Associate Professor

H. Greenfield, for providing invaluable encouragement, guidance and, most

importantly, inspiration, to me during this project and in the production of this thesis.

I also wish to thank my co-supervisor, Dr M. Mulholland, for her guidance and

advice in the early stages of this project.

My sincere gratitude also goes to Ms Eileen Emmerson for her invaluable assistance

concerning the technical aspects of this project. I wish to express my particular

appreciation to Mrs Yvonne El-Ghetany and Mr Camillo Taraborrelli for their help

regarding various technical aspects, also Mrs Christine Locke and Mrs Sharon

Debreczeni for their willing assistance.

To my fellow PhD students, Rosario Sagum, Manihar Situmorang and Paulina Taba,

many thanks for all their advice and help in the project and, as importantly, for

friendship. I would also like to thank my friends, Mr Sijabat, Mrs Denise Lynch and

their families for advice, support and assistance during my stay in Australia.

Finally, I would like to extend my heart-felt thanks to my husband, Mr Bonardo

Sitanggang, my daughter, Alvina Ester Greatica Sitanggang, and my parents, also my brothers, my sister and their families, for their support during my studies. They were very supportive. It is to my family, therefore, that I dedicate this thesis. vii

TABLE OF CONTENTS

ABSTRACT 11

PUBLICATIONS ARISING FROM THE WORK RELATED TO THE THESIS iv

ACKNOWLEDGEMENT vi

TABLE OF CONTENTS Vll

TABLE OF TABLES X

TABLE OF FIGURES XIV

CHAPTER 1. INTRODUCTION 1 1.1. Background 1 1.2. Objectives 3 1.3. Significance ofthe project 4 CHAPTER 2. LITERATURE REVIEW 5 2.1. Chemical structures of phytoestrogens 5 2.1.1. 5 2.1.2. Isoflavones 9 2.1.3. 21 2.1.4. Resorcyclic acid lactones 22 2.2. Methods of analysis of phytoestrogens 24 2.2.1. Identification 24 2.2.2. Extraction 33 2.3. Role of phytoestrogens in 35 2.4. Occurrence of phytoestrogens in forages and foods 37 2.4.1. Forages 37 2.4.2. Foods 37 2.4.3. Factors affecting phytoestrogens in foods 39 2.5. ofphytoestrogens 42 2.5.1. Lignans 42 2.5.2. Isoflavones 45 2.5.3. Factors affecting metabolism ofphytoestrogens 50 2.6. Effects of phytoestrogens in humans and animals 51 2.6.1. Experimental animals 51 2.7. 63 2. 7 .1. Cancer and phytoestrogens 65 2.8. Menopausal symptoms 67 2.9. Osteoporosis 68 2.1 O.Coronary heart disease 69 2.11.Antithyroid effect of isoflavones 70 2.12. 71 viii

2.12.1.General 71 2.12.2.Soya beans (Glycine max (L.) Merrill) 74 2.13.Cashew (Anarcadium occidental L.) 77 CHAPTER 3. MATERIALS AND METHODS 79 3.1. Apparatus 79 3.2. Materials 80 3.2.1. Reagents 80 3.3. Methods 81 3.3.1. Analytical method development 81 3.3.2. Isolation of daidzein, genistein, formononetin, biochanin A and coumestrol from foods 87 3.4. Stabilisation of daidzein, genistein, formononetin, biochanin A and coumestrol 94 3.5. Recovery of the isoflavones and coumestrol. 99 3.5.1. Food preparation 99 3.5.2. Analytical methods 99 3.6. Qualitative and quantitative analysis ofphytoestrogens in food 100 3.6.1. Sampling 100 3.6.2. Food varieties 101 3.6.3. Sample handling and storage 102 3.6.4. Food preparation 103 3.6.5. Analytical methods 106 3.6.6. Laboratory quality control 108 CHAPTER 4. RESULTS AND DISCUSSION- METHOD DEVELOPMENT 110 4.1. Separation and quantification of three isoflavones (daidzein, genistein and biochanin A) in soya beans 110 4.1.1. Separation of a mixture of standard compounds 110 4.1.2. Analytical performances 117 4.2. Separation and quantification of isoflavones ( daidzein, genistein, formononetin and biochanin A) and coumestans (coumestrol) in soya beans 121 4.2.1. Optimisation of the stationary phases 121 4.2.2. Optimisation of solvent system 128 4.2.3. Analytical performances 137 4.3. Optimisation of the isolation ofphytoestrogens from soya beans, cooked soya beans, canned soya beans, and soya 140 4.3 .1. Solvent, pH, and time 140 4.3.2. Temperature 149 4.3 .3. Optimisation of the extraction for cooked and canned soya beansandtofu 150 4.4. Stabilisation of daidzein, genistein, formononetin, biochanin A and coumestrol during storage at various levels of pH 155 4.5. Recovery 160 4.6. Conclusion 164 CHAPTER 5. RESULTS AND DISCUSSION- QUANTITATIVE ANALYSIS OF PHYTOESTROGENS IN FOODS 167 ix

5.1. Soya beans and soya bean products 167 5.1.1. Raw soya beans 167 5.1.2. Canned soya beans 172 5.1.3. Non fermented soya bean products 172 5.1.4. Dried soya milk curd or kembang tahu 183 5 .1.5. Fermented soya bean products 184 5.1.6. Second-generation soya foods 191 5.1.7. Laboratory cooked whole soya beans 196 5.1.8. Laboratory fried tofu, , and 197 5.2. Other legumes and products 199 5.2.1. Berlotti beans (Phaseolus vulgaris) 199 5.2.2. Black beans (Vigna unguiculata) 199 5.2.3. Black eye beans (Vigna unguiculata) 201 5.2.4. Broad beans (Viciafaba) 201 5.2.5. Canellini beans (Phaseolus vulgaris) 203 5.2.6. Chick peas (Cicer arietinum) 203 5.2.7. Green beans (Phaseolus vulgaris) 204 5.2.8. Groundnuts/ peanuts (Arachis hypogaea) 207 5.2.9. Haricot beans (Phaseolus vulgaris) 208 5.2.10.Lentils (Lens culinaris M.) 208 5.2.11.Lima beans (Phaseolus lunatus L.) 208 5.2.12.Mungbeans (Vigna radiata) 210 5.2.13.Peas (Pisum sativum) 212 5 .2.14.Petai (Parkia speciosa) and petai cina (Leucaena spp.) 217 5.2.15.Red kidney beans (Phaseolus vulgaris) 217 5.3. (Medicago sativa) 218 5.4. Cashew (Anacardium occidentale L.) 220 5.5. Conclusion 220 CHAPTER 6. CONCLUSION 222 6.1. Successful method development for analysis ofphytoestrogens in foods 222 6.1.1. Identification methods 223 6.1.2. Isolation methods 225 6.2. Determination of phytoestrogens in legumes and legume products available at retail level in Australia and Indonesia 225 6.3. Recommendations for further research 227

BIBLIOGRAPHY 230 APPENDICES 256 X

TABLE OF TABLES

Table 2.1. Elution time (tr, retention time, minutes) of daidzein, genistein, coumestrol, formononetin and biochanin A with different HPLC methods. 27 Table2.2. Deaths from in Australia, 1990-1996. 64 Table 2.3. The production of beans, peas, and groundnuts in the world, Australia and Indonesia in 1998. * 72 Table 2.4. The consumption of beans, peas and groundnuts in the world, Australia and Indonesia in 1998. * 73 Table 2.5. The production and consumption of soya beans in the world, Australia and Indonesia in 1997 and 1998*. 75 Table 2.6. Average per capita daily consumption of soya beans and soya bean products in Indonesia (g). 76 Table 2.7. Mean daily legume and pulse products consumption in Australia (g/person). 77 Table 4.1. Capacity factor (k') of daidzein, genistein, biochanin A, and flavone (internal standard) with C18 column and acetonitrile- aqueous acetic acid (pH 1% and 10%) (33:67, v/v) as eluent. 114 Table 4.2. Capacity factor (k') of standard daidzein, genistein, biochanin A, and flavone with different solvent composition. 115 Table 4.3. Calibration parameters of daidzein, genistein, and biochanin A using the proposed method and that of Franke et al. (1994, 1995). HPLC conditions: isocratic elution with acetonitrile-1% aqueous acetic acid (33:67, v/v) as eluent. 118 Table 4.4. Precision of HPLC technique for separation of a mixture of daidzein, genistein, and biochanin A standards. HPLC conditions: isocratic elution with acetonitrile-1% aqueous acetic acid (33:67, v/v) as eluent with six replicate injections. 119 Table 4.5. Spiked recovery for daidzein, genistein, and biochanin A from soya beans. HPLC conditions: isocratic elution with acetonitrile- 1% aqueous acetic acid (33:67, v/v) as eluent. 120 Table 4.6. The separation of daidzein, genistein, formononetin, biochanin A and coumestrol using the cyano, C8, C18 and phenyl columns. 122 Table4.7. Separation of daidzein, genistein, formononetin, biochanin A and coumestrol by different stationary phases and mobile phases. 127 Table 4.8. Separation of daidzein, genistein, formononetin, biochanin A and coumestrol by phenyl column and a variety of eluents. 129 xi

Table 4.9. Capacity factor (k') of daidzein, genistein, formononetin and biochanin A and coumestrol with different compositions of eluent. 130 Table 4.1 0. Quantitative analysis of daidzein, coumestrol, genistein, formononetin and biochanin A at different wavelengths (mean of 6 injections, mg/L ± SD) 132 Table 4.11. Coefficients of variation of the retention times of daidzein, genistein, formononetin, biochanin A and coumestrol (mean of 6 injections within- and between-assay). 138 Table 4.12. Coefficients of variation of the quantitative analysis of daidzein, genistein, formononetin, biochanin A, and coumestrol. 138 Table 4.13. Coefficient of determination(?) of standard daidzein, genistein, formononetin, biochanin A, and coumestrol using a phenyl column and acetonitrile-water (33:67, v/v) as eluent. 140 Table 4.14. Daidzein and genistein in dried soya beans at different extraction temperatures (mg/100 g, wet weight basis). 149 Table 4.15. Daidzein and genistein in soya milk at different extraction temperatures (mg/100 g, wet weight basis). 150 Table 4.16. Spiked recovery of daidzein, coumestrol, genistein, formononetin, and biochanin A in soya beans using a phenyl column with an isocratic elution system with acetonitrile-water (33:67, v/v) as eluent. 162 Table 4.17. Spiked recovery of daidzein, coumestrol, genistein, formononetin and biochanin A in flat beans using a phenyl column with an isocratic elution system with acetonitrile-water (33:67, v/v) as eluent. 163 Table 4.18. Spiked recovery of daidzein, coumestrol, genistein, formononetin and biochanin A in snow peas using a phenyl column with an isocratic elution system with acetonitrile-water (33:67, v/v) as eluent. 164 Table 5 .1. Mean levels of daidzein and genistein in dried soya beans purchased in Indonesia and Australia (mg/100 g, wet and dry weight basis). 168 Table 5 .2. Levels of daidzein, genistein, formononetin, biochanin A and coumestrol in soya beans (mg/100 g, wet weight basis). 170 Table 5.3. Daidzein and genistein levels in fresh soya beans purchased in Indonesia and Australia (mg/100 g, wet and dry weight basis). 171 Table 5.4. Daidzein, genistein and coumestrollevels in soya milk products from Australia (mg/100 g, wet and dry weight basis). 174 Table 5.5. Daidzein, genistein and coumestrollevels in soya milk products from Indonesia (mg/100 g, wet and dry weight basis). 175 Table 5.6. Daidzein and genistein levels in tofu products from Australia (mg/100 g, wet and dry weight basis). 180 x.ii

Table 5.7. Daidzein and genistein levels in tofu products from Indonesia (mg/100 g, wet and dry weight basis). 181 Table 5.8. Daidzein and genistein levels in dried soya milk curd products from Indonesia (mg/100 g, wet and dry weight basis). 183 Table 5.9. Daidzein and genistein levels in tempeh products from Indonesia (mg/100 g, wet and dry weight basis). 186 Table 5.10. Daidzein and genistein levels in fermented soya paste (taucho) products from Indonesia (mg/100 g, wet and dry weight basis). 188 Table 5.11. Daidzein and genistein levels in fermented sweet soya sauce (kecap manis) and unsweetened soya sauce products purchased in Indonesia (mg/100 g, wet weight basis). 191 Table 5.12. Daidzein and genistein levels in second-generation soya food products from Australia (mg/100 g, wet weight basis). 192 Table 5.13. Daidzein and genistein levels in cereal products from Australia (mg/100 g, wet and dry weight basis). 194 Table 5.14. Daidzein and genistein levels in soya bread products from Australia (mg/100 g, wet and dry weight basis). 196 Table 5.15. Daidzein and genistein levels in fried tofu, tempeh and oncom (mg/100 g, wet weight basis). 198 Table 5.16. Daidzein, genistein, coumestrol, formononetin and biochanin A in berlotti beans, black beans and black eye beans purchased in Australia and Indonesia (mg/100 g, wet weight basis). 200 Table 5.17. Daidzein, genistein, coumestrol, formononetin and biochanin A in dried broad beans purchased in Australia, in broad beans grown in the USA, and canellini beans purchased in Australia (mg/100 g, wet weight basis). 202 Table 5.18. Daidzein, genistein, coumestrol, formononetin and biochanin A in chick peas and products from Australia (mg/100 g, wet and dry weight basis). 204 Table 5.19. Daidzein, genistein, coumestrol, formononetin and biochanin A in whole fresh, canned, frozen and freeze-dried green beans purchased in Australia and Indonesia (mg/100 g, wet weight basis). 205 Table 5.20. Daidzein, genistein, coumestrol, formononetin and biochanin A in fresh beans, laboratory boiled and laboratory freeze-dried green beans (mg/100 g, wet and dry weight basis). 207 Table 5.21. Daidzein, genistein, coumestrol, formononetin and biochanin A in red lentils purchased in Australia and grown in the USA and lima beans purchased in Australia and Indonesia (mg/100 g, wet weight basis). 209 Table 5.22. Daidzein, genistein, coumestrol, formononetin and biochanin A in mungbeans purchased in Australia and Indonesia (mg/100 g, wet weight basis). 211 xiii

Table 5.23. Daidzein, genistein, coumestrol, formononetin and biochanin A in fresh, dried and frozen peas purchased in Australia and Indonesia (mg/100 g, wet weight basis). 213 Table 5 .24. Daidzein, genistein, coumestrol, formononetin and biochanin A in canned and freeze-dried peas purchased in Australia and Indonesia (mg/100 g, wet weight basis). 215 Table 5.25. Daidzein, genistein, coumestrol, formononetin and biochanin A in petai and petai cina purchased in Indonesia (mg/100 g, wet weight basis). 217 Table 5.26. Isoflavones and coumestrol in sprouted alfalfa from Australia (mg/100 g, wet weight basis). 219 xiv

TABLE OF FIGURES

Figure 2.1. Chemical structure oflignan. 6 Figure 2.2. Chemical structure of matairesinol. 6 Figure 2.3. Chemical structure of secoisolariciresinol. 7 Figure 2.4. Chemical structure of pinoresinol. 7 Figure 2.5. Chemical structure of isolariciresinol. 8 Figure 2.6. Chemical structure of . 8 Figure 2. 7. Chemical structure of . 9 Figure 2.8. Chemical structure offlavones. 12 Figure 2.9. Chemical structure ofisoflavones. 13 Figure 2.1 0. Chemical structure of daidzein. 13 Figure 2.11. Chemical structure of genistein. 14 Figure 2.12. Chemical structure of formononetin. 14 Figure 2.13. Chemical structure ofbiochanin A. 15 Figure 2.14. Chemical structure of -1713 (natural estrogenic ). 15 Figure 2.15. Chemical structure of (synthetic ). 16 Figure 2.16. Chemical structure of . 16 Figure 2.17. Chemical structure of . 17 Figure 2.18. Chemical structure of . 17 Figure 2.19. Chemical structure of . 18 Figure 2.20. Chemical structure of6"-0-acetyldaidzin. 18 Figure 2.21. Chemical structure of 6"-0-acetylgenistin. 19 Figure 2.22. Chemical structure of 6"-0-acetylglycitin. 19 Figure 2.23. Chemical structure of 6"-0-malonyldaidzin. 20 Figure 2.24. Chemical structure of 6"-0-malonylgenistin. 20 Figure 2.25. Chemical structure of6"-0-malonylglycitin. 21 Figure 2.26. Chemical structure of coumestans. 22 Figure 2.27. Chemical structure of coumestrol. 23 Figure 2.28. Chemical structure of . 23 Figure 2.29. Formation of enterolactone and enterodiol by human faecal flora (Kurzer & Xu 1997). 44 Figure 2.30. Metabolic pathway of daidzein and genistein by human gut and the formation of (Kurzer & Xu 1997). 46 XV

Figure 2.31. Rumen metabolism offormononetin in sheep (Price & Fenwick 1985). 47 Figure 2.32. Rumen metabolism ofbiochanin A in sheep (Price & Fenwick 1985). 47 Figure 3.1. extraction procedures from dried soya beans. 91 Figure 3.2. Phytoestrogen extraction procedures from soya milk. 92 Figure 3.3. Phytoestrogen extraction procedures from dried soya beans at 80°C and 100°C. 93 Figure 3.4. The schema of the addition of sodium hydroxide to the extracted samples for storage trials. 96 Figure 3.5. The schema of the addition of sodium hydroxide in the standard solutions for storage trials. 97 Figure 3.6. The schema of the addition ofhydrochloric acid in the standard solutions for storage trials. 98 Figure 4.1. Chromatogram of standard daidzein (D), genistein (G), biochanin A (B), and internal flavone standard (F), with C18 bonded phase column and acetonitrile-I 0% aqueous acetic acid (33:67, v/v) as eluent. 112 Figure 4.2. Chromatogram of standard daidzein (D), genistein (G), biochanin A (B), and internal flavone standard (F), with C18 bonded phase column and acetonitrile-!% aqueous acetic acid (33:67, v/v) as eluent. 113 Figure 4.3. (A) chromatogram of standard and (B) chromatogram of soya bean spiked with standards daidzein (D), genistein (G), biochanin A (B), and internal flavone standard (F) with C18 bonded phase column and acetonitrile-I% aqueous acetic acid as eluent. Al & Bl (30:70, v/v), A2 & B2 (33:67, v/v), A3 & B3 (35:65, v/v), A4 & B4 (38:62, v/v), AS & BS (40:60, v/v). 116 Figure 4.4. Capacity factor (log k') of standard daidzein, genistein, biochanin A and internal flavone standard from different compositions of solvent (acetonitrile). 117 Figure 4.5. The LC chromatogram of a mixture of standards, daidzein, genistein, formononetin, and biochanin A, coumestrol, and internal flavone standard, by a cyano column and acetonitrile- }% aqueous acetic acid (33:67, v/v) as eluent. 123 Figure 4.6. The LC chromatogram of a mixture of standards, daidzein, genistein, formononetin, biochanin A, coumestrol, and internal flavone standard, by C8 column and acetonitrile-I% aqueous acetic acid (33:67, v/v) as eluent. 124 Figure 4. 7. The LC chromatogram of a mixture of standards, daidzein, genistein, formononetin, biochanin A, coumestrol and internal flavone standard, by a C18 column and acetonitrile-I% aqueous acetic acid (33:67, v/v) as eluent. 125 xvi

Figure 4.8. The LC chromatogram of a mixture of standards daidzein, genistein, formononetin, biochanin A, coumestrol and internal flavone standard, by a phenyl column and acetonitrile-!% aqueous acetic acid (33:67, v/v) as eluent. 126 Figure 4.9. Capacity factor (log k') of standard daidzein, coumestrol, genistein, formononetin, and biochanin A from 30 to 40 % acetonitrile in water. 130 Figure 4.1 0. Absorption maxima of daidzein, genistein, formononetin, biochanin A and coumestrol at different wavelengths. 132 Figure 4.11. UV chromatogram of standards daidzein, coumestrol, genistein, formononetin, and biochanin A. The analytes were separated on a phenyl column with isocratic system: acetonitrile-water (33:67, v/v) as eluent. 133 Figure 4.12. UV chromatogram of daidzein (peak 1), coumestrol (peak 2), genistein (peak 3), formononetin (peak 4), and biochanin A (peak 5) standards obtained from LC-APCI-HN-MS analysis using a phenyl column and water-acetonitrile (67:33, v/v) as eluent. 135 Figure 4.13. Positive ion mass spectra of peaks 1, 2, 3, 4, and 5 observed in Figure 4.12. Peak 1 is daidzein with m/z of254.6. Peak 2 is coumestrol with m/z of269.0. Peak 3 is genistein with m/z of 270.8. Peak 4 is formononetin with m/z of260.2. Peak 5 is biochanin A with m/z of284.8. 136 Figure 4.14. The LC chromatogram and UV spectra of standard daidzein, coumestrol, genistein, formononetin and biochanin A. 142 Figure 4.15. The LC chromatogram and UV spectra of daidzein and genistein isolated from dried soya beans. 143 Figure 4.16. Peak areas of daidzein and genistein in soya beans extracted with different concentrations of phosphoric acid and at different periods of reflux. 144 Figure 4.17. Levels of daidzein and genistein in raw dried soya beans at different reflux times and pH (mg/100, dry weight basis). 145 Figure 4.18. The LC chromatogram and UV spectra of daidzein and genistein isolated form soya milk. 14 7 Figure 4.19. Daidzein and genistein levels in soya milk with different reflux times. 148 Figure 4.20. Levels of daidzein and genistein in cooked dried soya beans at different refluxing times and pH (mg/100 g, dry weight basis) 152 Figure 4.21. Daidzein and genistein levels in canned soya beans at different reflux times and pH (mg/100 g, dry weight basis). 153 Figure 4.22. Daidzein and genistein levels in tofu at different reflux times and pH (mg/100 g, dry weight basis). 154 xvii

Figure 4.23. Recovery of daidzein from extracted soya beans during storage for 1, 4, 14, 21 and 28 days at various pH levels. 156 Figure 4.24. Recovery of standards daidzein during storage for 1, 4, 14, 21 and 28 days at various pH levels. 156 Figure 4.25. Recovery of genistein from extracted soya beans during storage for 1, 4, 14, 21 and 28 days at various pH levels. 157 Figure 4.26. Recovery of standard genistein during storage for 1, 4, 14, 21 and 28 days at various pH levels. 158 Figure 4.27. Recovery of standard coumestrol during storage for 1, 2, 3, 14, 21 and 28 days at various pH levels. 159 Figure 4.28. Recovery of standard formononetin during storage for 1, 2, 3, 14, 21 and 28 days at various pH levels. 159 Figure 4.29. Recovery of standard biochanin A during storage for 1, 2, 4, 14, 21 and 28 days at various pH levels. 160 1 CHAPTERl.

INTRODUCTION

1.1. Background

Phytoestrogens are naturally occurring substances produced by plants. The chemical

structure of phytoestrogens is similar to the natural estrogen (17J3-estradiol) of

humans and animals.

Phytoestrogens have been recognized as one of the cancer-preventing

constituents. Animal and in vitro studies have demonstrated that phytoestrogens possess a wide range of biological activities. Epidemiological studies have indicated that populations consuming a diet high in phytoestrogens have a low incidence of, and mortality from, certain cancers and many other so-called "Western" diseases.

Asian populations consuming high levels of phytoestrogens derived from soy-based diets have a lower incidence of breast cancer and malignancies than Western populations.

It is thought that the weak estrogenic activity possessed by phytoestrogens can be used as natural estrogen therapy instead of synthetic hormone replacement therapy

(HRT). HRT has been used for preventing menopausal symptoms in women.

Menopausal symptoms are manifestations of estrogen deficiency. Effects include hot flushes, decreased density, disturbance in sleep patterns, and increased risk of cardiovascular disease. It has been found that synthetic estrogen replacement therapy 2

also involves risks as not all women are able to tolerate it. Dose and duration ofHRT

could stimulate breast cancer, endometrial hyperplasia, or thromboembolism.

Clarification of the role of phytoestrogens in preventing cancer and determination of

the additional effects in humans is necessary. Phytoestrogens are known to induce

in livestock. These compounds were first found to induce infertility in

sheep that were fed a particular species of clover containing these substances

(Bennetts et al. 1946). Phytoestrogens caused reproductive failure in captive

cheetahs who were fed a commercial feline diet containing high levels of plant

(Setchell et al. 1987) and caused infertility in quails in California when the

availability of phytoestrogens in forage was increased (Leopold et al. 1976).

Phytoestrogens may increase the risk of malignancies and may induce production of

(Medlock et al. 1995; Burroughs et al. 1985; Burroughs et al. 1990; Whitten

eta!. 1992b).

Research on the influence of dietary phytoestrogens on cancer and hormone-related

conditions in humans is lacking. Methods for dietary assessment of phytoestrogens

intake are limited. Analytical methods for the isolation and identification of

isoflavones conjugates and unconjugates ( aglycones) in plants and foods are

continually studied in order to develop a simple, accurate, rapid and sensitive

standard technique. Methods for analysis of aglycones in foods are needed

for food tables, while methods for analysis of all compounds (isoflavone conjugates

and unconjugates) are needed for physiological studies e.g. absorption of these

compounds by the body. Although daidzein and genistein are the most biologically active compounds and are found at high levels in soya beans, coumestrol is also 3

considered important due to its adverse effect on humans and animals. Coumestrol is

also found in soya beans and other foods and feeds. Therefore, a method for

identification and quantitation of isoflavones and coumestrol simultaneously in soya

beans and other foods has to be developed. Quantitative date for phytoestrogens in

foods is available only in some countries, including USA, Finland, Singapore, Japan,

Brazil and Korea.

Soya beans and soya bean products are known to contain higher levels of

phytoestrogens than other legumes. Indonesia has the highest consumption of soya

beans and soya bean products (8.9 kg/year) in the world with Japan second (8

kg/year) (FAO 1997). However, the phytoestrogens intake of Indonesians has not yet

been studied. Data for phytoestrogens content in soya beans and soya bean products

from Indonesia are not available hampering research on their effects.

1.2. Objectives

The focus of the work was for food labelling and other related purposes such as food

composition tables and dietary surveys. A global phytoestrogens database with

quantitative and qualitative data in foods is clearly needed. This will make it possible to assess phytoestrogen intake in certain countries through dietary assessment methods. It will also facilitate research to clarify the role of dietary phytoestrogens in preventing cancer, reducing menopausal symptoms and other health aspects. To fill the gap in information, this project studied two different countries, Australia and

Indonesia, two large neighbouring countries with very different health and food 4 consumption patterns (Arndt 1986). Australia is a Western country while Indonesia is a developing country in Asia.

Two main objectives of the work described in this thesis are:

To develop a reliable analytical method for the isolation and identification of four

isoflavones (daidzein, genistein, formononetin and biochanin A) and the

coumestan, coumestrol, in foods and food products.

To determine the levels of the five phytoestrogens in plant foods and food

products available at the retail level in Australia and Indonesia, using the method

developed.

1.3. Significance of the project

The analytical results are needed for incorporation in national and international food composition databases, which currently contain mostly data for nutrient composition of foods. The results of the project will also benefit government, public health, private and intellectual sectors. It may aid the development of appropriate diets for preventing cancer and for improvement of postmenopausal symptoms. The information produced may also aid the private sector concerned with the production of phytoestrogen-enriched food products. 5 CHAPTER2.

LITERATURE REVIEW

2.1. Chemical structures ofphytoestrogens

There are four groups of plant substances categorised as phytoestrogens. They are

lignans, isoflavones, coumestans, and resorcyclic acid lactones (Verdeal & Ryan

1979; Price & Fenwick 1985).

2.1.1. Lignans

Lignans are the products of the linking (dimerization) of two phenylpropene or

phenylpropane precursors (Waterman & Mole 1994). The C6-C3 dimeric skeleton of

lignans is associated with lignin (Figure 2.1, overleaf). The plant lignans,

matairesinol, secoisolariciresinol, pinoresinol, and isolariciresinol (Figures 2.2-2.5, pages 6-8) have been reported to be potent phytoestrogens and are precursors of the

mammalian lignans enterolactone and enterodiol (Figures 2.6 & 2.7, pages 8 & 9).

Enterolactone and enterodiol are formed by the removal of two methyl groups and two hydroxy groups from secoisolariciresinol and matairesinol. They both are in the oxydiarylbutane and diarylbutyrolactone classes of lignans. Isolariciresinol is a tetrahydronaphthalene and pinoresinol is a furofuran lignan. 6

HO

OH HO OH

Figure 2.1. Chemical structure of lignan.

HO

OH

Figure 2.2. Chemical structure of matairesinol. 7 H OH OH HO

OH

Figure 2.3. Chemical structure of secoisolariciresinol.

OH

Figure 2.4. Chemical structure of pinoresinol. 8

OH OH HO

OCH3 OH

Figure 2.5. Chemical structure of isolariciresinol.

HO

Figure 2.6. Chemical structure of enterolactone. 9

H HO OH OH

HO

Figure 2.7. Chemical structure of enterodiol.

2.1.2. lsoflavones

Flavones are natural compounds produced by plants. The basic skeleton of

consists of three phenolic rings and 15 carbon atoms (Figure 2.8, page 12). Ring A

contains a C6 fragment and ring B contains a C6-C3 fragment. compounds

occur in all part of plants: root, stem, leaf, flower, pollen, fruit, seed, wood and bark.

The have a very limited distribution in the plant kingdom, and are almost entirely restricted to the subfamily Papilionoideae of the Leguminosae. Even in the subfamily Caesalpinioideae and Mimosoideae of the Leguminosae, only one or two plants have been reported to contain isoflavonoids. The limited distribution of the isoflavonoids in the plant kingdom is restricted to the presence of several for synthesis of compounds (Harbome & Mabry, 1982). 10

Isoflavonoids are a subclass of the . All classes of flavonoids derive their

carbon skeleton from compounds of intermediary cell metabolism through the action

of two consecutive (general phenylpropanoids and flavonoid ) pathways.

Phenylpropanoid units derived from the shikimate pathway are common structural

elements of all flavonoid compounds and of various other classes of

phenylpropanoids, such as lignin, stilbenes and cinnamate esters. The sequence of

reactions converting phenylalanine into the CoA ester derivatives of substituted

cinnamic acids is therefore termed 'general phenylpropanoid metabolism'. The CoA

ester derivatives of substituted cinnamic acids are substrates in the biosynthesis of

flavonoids. Three enzymes catalyzing the individual steps are phenylalanine

ammonia-lyase, cinnamate 4-hydroxylase and 4-coumerate:CoA ligase (Ebel &

Hahlbrock 1982).

Isoflavonoids are synthesized by an aryl migration mechanism from the 2-

phenylchroman skeletons of the flavonoids (Figure 2.9, page 13). Two chalcones

(2 ',4' ,4-trihydroxychalcone and 2' ,4' ,6' ,4-tetrahydroxychalcone) act as substrates.

An oxidative reaction is involved producing the 4-hydroxyl sites. According to Pelter

et al. (1971), phenolic oxidation occurs in the chalcone leading to a spirodienone

intermediate. Decomposition of this intermediate via a proton produces daidzein and

genistein. Methylation via S-adenosylmethionine forms formononetin and biochanin

A. is also a precursor of isoflavonoids. An isoflavone synthase transforms the flavanone substrates (2S)- or (2S)- into the isoflavones genistein or daidzein, respectively. 11

Synthesis of isoflavonoids is induced as a response to microbial attack. Synthesis of

isoflavonoids is marked by large increases in the activities of several enzymes of

both general phenylpropanoid metabolism (phenylalanine ammonia-lyase, cinnamate

4-hydroxylase and 4-coumerate:CoA ligase) and flavonoid biosynthesis (chalcone

synthase and chalcone isomerase). A 4'-0-methyl-transferase is a specific

for catalyzing 4' -0-methylation of ring B of isoflavonoids after aryl migration to form 4'-hydroxy-isoflavones, such as daidzein (7,4'-dihydroxyisoflavone) or

genistein.

A large number of isoflavones have been isolated from plant species, but only a small number have been shown to possess estrogenic activity. Moreover, not all isoflavones isolated from plants that are known to affect estrus are active. Daidzein, genistein, formononetin, and biochanin A (Figures 2.10-2.13, pages 13-15) are isflavonoid compounds that have been shown to have effects resembling estradiol-

1713 (natural estrogenic steroid) and diethylstilbestrol or DES (a synthetic estrogen)

(Shutt et al. 1972) (Figures 2.14 & 2.15, pages 15-16). These compounds have a low molecular weight based on the flavan nucleus, an intense absorption rate of approximately 255 to 275 nm and a less intense band or inflection at approximately

310 to 330 nm. Nairn et al. (1973) were the first to identify and determine the structure of glycitein found in soya beans. Glycitein isolated from soybean was shown not to be uterotrophic (Nairn et al. 1974) but has recently been shown to possess some weak estrogenic activity (Song et al. 1999) (Figure 2.16, page 16). 12

In the intact plants, daidzein, genistein, and glycitein commonly occur in a bound form and are present in a variety of structural forms. One or more hydroxyl groups from these compounds are bound to a sugar or sugars. is the sugar bound with daidzein to form daidzein 7-0-glucoside or daidzin; with genistein to form 7-0- glucoside or genistin; and with glycitein to form glycitein 7-0-glucoside or glycitin

(Figures 2.17, 2.18 & 2.19, pages 17 & 18). Daidzein, genistein, and glycitein are also present as acetyl that are lmown as 6 -0-acetyldaidzin, 6-0- acetylgenistin and 6 -0-acetylglycitin, respectively (Figures 2.20, 2.21 & 2.22, pages

18 & 19). As malonyl forms, daidzein, genistein and glycitein are lmown as 6 -0- malonyldaidzin, 6 -0-malonylgenistin and 6 -0-malonylglycitin (Figures 2.23, 2.24 and 2.25, pages 20 & 21).

5 4

Figure 2.8. Chemical structure of flavones. 13

5'

Figure 2.9. Chemical structure ofisoflavones.

HO

H

OH

Figure 2.1 0. Chemical structure of daidzein. 14

HO

H

OH OH

Figure 2.11. Chemical structure of genistein.

HO

Figure 2.12. Chemical structure of formononetin. 15

HO

H

OH

Figure 2.13. Chemical structure ofbiochanin A.

Figure 2.14. Chemical structure of estradiol-17~ (natural estrogenic steroid). 16

OH

HO

Figure 2.15. Chemical structure of diethylstilbestrol (synthetic estrogen).

HO

H

H 0

Figure 2.16. Chemical structure of glycitein. 17

OH H H OH

Figure 2.17. Chemical structure of daidzin.

OH H OH OH

Figure 2.18. Chemical structure of genistin. 18

OH

Figure 2.19. Chemical structure of glycitin.

OH H H OH

Figure 2.20. Chemical structure of 6"-0-acetyldaidzin. 19

OH H OH OH

Figure 2.21. Chemical structure of 6"-0-acetylgenistin.

OH

Figure 2.22. Chemical structure of 6"-0-acetylglycitin. 20

OH H H OH

Figure 2.23. Chemical structure of 6"-0-malonyldaidzin.

OH H OH OH

Figure 2.24. Chemical structure of 6"-0-malonylgenistin. 21

OH

Figure 2.25. Chemical structure of 6"-0-malonylglycitin.

2.1.3. Coumestans

Coumestrol was frrst isolated from forage crops and shown to have estrogenic activity by Bickoff et al. (1957). It belongs to a subclass of coumestans. The coumestan skeleton has five heterocyclic rings and oxygen at the C-4 and the 2'- carbon of the B-ring (Figure 2.26, overleaf). Coumestans are derived from 3- arylcoumarins and have a carbonyl function adjacent to the C-ring oxygen, which is equivalent to C-2 in the isoflavone system. Coumestrol has an empirical formula of

C1sHsOs (Figure 2.27, page 23). This substance has been reported to have a bright blue either in neutral solution or acid solution and appears greenish- yellow in strong alkali solution (Bickoff et al. 1957). Coumestrol melts and decomposes at 385°C. Its absorption maxima are at 208, 243, and 343 mf.!. 22

2.1.4. Resorcyclic acid lactones

Resorcyclic acid lactones are not intrinsic components of plants but secondary mould metabolites of fungal species (mainly ). The potent compound from this class is zearalenone (Figure 2.28, page 23). The chemical structure of zearalenone consists of 6-(1 0-hydroxy-6-oxo-trans-1-undecenyl)-13-resorcyclic acid-J..t-lactone.

Zearalenone (CtsHzzOs) is a white crystalline material, has a molecular weight of

318, m.p. of 164-165°C, and .lvmax in of235 and 316 nm.

0--.....a..: 11 10

Figure 2.26. Chemical structure of coumestans. 23

HO

H c H 0-______., OH

Figure 2.27. Chemical structure of coumestrol.

HO 0

Figure 2.28. Chemical structure of zearalenone. 24

2.2. Methods of analysis of phytoestrogens

2.2.1. Identification

Various methods have been developed to isolate, identify and quantify

phytoestrogens from plants, plant foods, and food products; and from urine, faeces

and biological fluids of humans and animals.

2.2.1.1. Thin Layer Chromatography (TLC)

Thin layer chromatography methods were developed for the separation of coumestrol from soya beans (Lookhart et al. 1978, 1979) and ofzearalenone and zearalenonel in grains such as com, and by Swanson et al. (1984). TLC methods could identify coumestrol at levels of 0.05 f.!g/g (5 f.!g/100 g) in whole soya beans,

0.05 J.!g/g (5 f.!g/100 g) in soya bean cotyledons and 0.20 J.!g/g (20 f.!g/100 g) in soya bean hulls (Lookhart et al. 1978, 1979) and zearalenone with the sensitivity of

80 ng/g (8 f.!g/100 g).

2.2.1.2. High Performance Liquid Chromatography (HPLC)

High-performance liquid chromatography (HPLC) is the method most often used for the separation and the quantitative analysis of phytoestrogens in plants and foods.

HPLC methods for the separation of daidzein, genistein, coumestrol, formononetin and biochanin A from soya beans or soya bean products have been developed by 25 vanous authors and are summarized in Table 2.1 (pages 27-29). Early HPLC

methods were applied to the separation of 4', 6, 7-trihydroxyisoflavone and genistein

from soya beans (West et al. 1978); zearalenone from corn (Ware & Thorpe 1978)

and coumestrol from alfalfa (Livingstone et al. 1961); the separation of daidzein,

genistein, genistin and coumestrol from soya beans and soya bean products (Murphy

1981); and, daidzein, genistein, glycitein and their glucoside forms and coumestrol from defatted soya bean meal, soya protein isolates and soya protein concentrates

(Eldridge 1982a, 1982b).

The USDA-IOWA State University Isoflavones Database (1999) used the method developed by Murphy et al. (1997) as the reference method for the separation and identification of isoflavone aglycones and conjugates in foods. Murphy et al. (1997) used the YMC-pack ODS-AM-303 column and a linear gradient elution system composed of 0.1% aqueous acetic acid and 0.1% acetic acid in acetonitrile as eluent at a flow rate of 1.0 mL/min. This method has mainly been used for the separation of daidzein, genistein, glycitein and their conjugates in soya-based infant formulas with a total elution time for all compounds of 45 minutes.

The proliferation ofHPLC methods (West et al. 1978; Lookhart et al. 1978; Murphy

1981; Eldridge 1982b; Seo & Morr 1984; Kitada et al. 1986; Setchell et al. 1987;

Wang et al. 1990; Coward et al. 1993; Wang & Murphy 1994a; Barnes et al. 1994a;

Franke et al. 1994, 1995; Futukake et al. 1996; Murphy et al. 1997) reflects the different compounds of interest (conjugates and aglycones) and the search for a universally satisfactory method. However, many of these published methods provide inadequate detail about validation. 26

Different columns and solvents have been used resulting in variations in the retention time of compounds (Table 2.1, overleaf). The retention time for daidzein ranges from

5 to 43 minutes, for genistein from 7 to 54 minutes, for coumestrol from 12 to 55 minutes, for formononetin from 10 to 13 minutes and for biochanin A from 12 to 18 minutes. Most authors have applied gradient HPLC methods (Murphy 1981; Eldridge

1982a, 1982b; Coward et al. 1993; Seo & Morr 1994; Wang & Murphy 1994a;

Barnes eta!. 1994a; Franke et al. 1994, 1995; Murphy et al. 1997). These methods are not considered more efficient and reliable than isocratic HPLC methods (West et al. 1978; Kitada et al. 1986; Setchell & Welsh 1987; Wang et al. 1990) because dramatic changes in concentration of the mobile phase and the associated pressure changes adversely affect the retention of the compounds and the column. 27

Table 2.1. Elution time (tr, retention time, minutes) of daidzein, genistein, coumestrol, formononetin and biochanin A with different HPLC methods.

Compound fr Elution Column Mobile phase (v/v) Reference condition Daidzein 29 Linear DuPont Zorbax 15% aqueous (A) and Eldridge gradient ODS (25 x 0.46 65% aqueous methanol (B). 25% (1982) em I.D.) to 50% A in B. Daidzein 18 !socratic LiChrosorb RP-8 Acetonitrile (A) and 0.05M Kitada et (5 mm; 250x4 potassium dihydrogenphosphate al. (1986) mm I.D.) (pH 2.0) (B). 15% A in B. Daidzein 16 Linear Brownier Acetonitrile (A) and 1% aqueous Cowardet gradient Aquapore C8 -RP trifluoroacetic acid (B). 0 to al. (1993) (30 x 0.45 em 46.4% A in B by 2.25%/min I.D.) Daidzein 40 Gradient YMC-pack ODS- 0.1% glacial acetic acid in water Wang& AM-303 (5 J.tm, (A) & 0.1% glacial acetic acid in Murphy 25 x 4.6 mm I.D.) acetonitrile (B). 15% to 35% Bin (1994a) A over 50 min, and held at 35% B in A for 10 min. Daidzein 11.5 Non- Ultrasphere-ODS Methanol (A) and water (B). 0 to Murphy linear (250x4.6mm 100% A in Bin 15 min. (1981) gradient I.D.) Daidzein 5.4 Non- Nova-Pak RP- 23% acetonitrile (A) in 10% Franke et linear C18 aqueous acetic acid (B) linearly to al. (1994, gradient 70% A in B for 8 min. and held 1995) at 23% A in B for 12 min. Daidzein 5 !socratic Hypersil ODS (5 Methanol-0.1 M ammonium Setchell et J.tm; 25 x 4.6 mm acetate buffer pH 4.6 (60:40, v/v) al. (1987) I.D.) Daidzein 36.9 Linear Ultrasphere-ODS Water-acetic acid (95:5, v/v) (A) Seo& gradient C-18 RP (4.6 X & methanol-acetic acid (95:5, v/v) Morr 150 em I.D.) (B). 0% to 100% Bin A for a (1984) subsequent 55 min. Daidzein 8 !socratic J.t-Bondapak Cl8 Methanol-1mM ammonium Wanget (10 J.tm; 3.9 mm x acetate (60:40, v/v) al. (1990) 30 em I.D.) Daidzein 30.3 Linear YMC-pack ODS- 0.1% aqueous acetic-0.1% acetic Murphy et gradient AM-303 acid in acetonitrile a!. (1997) Genistein 7 !socratic Hypersil ODS Methanol-0.1 M ammonium Setchell et acetate buffer (pH 4.6) (60:40, al. (1987) v/v) Genistein 35 Linear DuPont Zorbax 15% aqueous methanol (A) and Eldridge gradient ODS (25 X 0.46 65% aqueous methanol (B). 25% (1982) em I.D.) to 50% A in B. 28

Table 2.1, continued

Compound t, Elution Column Mobile phase (v/v) Reference. condition Genistein 32 !socratic LiChrosorb RP-8 Acetonitrile {A) & 0.05M Kitada et (5 mm; 250x4 potassium dihydrogenphosphate al. (1986) mml.D.) (pH 2.0) (B). 15% A in B. Genistein 18 Linear Brownier Acetonitrile (A) & 1% aqueous Cowardet gradient Aquapore C8 -RP trifluoroacetic acid (B). 0 to al. (1993) (30x 0.45 em 46.4% A in B by 2.25%/min. I.D.) Genistein 54 Linear YMC-pack ODS- 0.1% glacial acetic acid in water Wang& gradient AM-303 (5 J.1m, (A) & 0.1% glacial acetic acid in Murphy 25 x 4.6 mm I.D.) acetonitrile {B). 15% to 35% Bin {1994a) A over 50 min, and held at 35% B in A for 10 min. Genistein 32 Linear ODS (5 mm; 4.6 x Acetonitrile (A) & 0.1% Futukake gradient 250 mm I.D.). trifluoroacetic acid (B). 0 to 60% etal. A in B. (1996)

Genistein 12 Non Ultrasphere-ODS Methanol (A) & water (B). 0 to Murphy linear (250x4.6mm 100% A in B for 15 min. (1981) gradient I.D.)

Genistein 8.3 Gradient Nova-Pak RP- 23% acetonitrile (A) in 10% Franke et C18 aqueous acetic acid (B) linearly to al. (1994, 70% A in B for 8 min. and held at 1995) 23% A in B for 12 min. Genistein 41.6 Linear Ultrasphere-ODS Water-acetic acid (95:5, v/v) (A) Seo& gradient C-18 RP (4.6 X & methanol-acetic acid (95:5, v/v) Morr 150 em I.D.) (B). 0% to 100% Bin A for a (1984) subsequent 55 min. Genistein 23 Linear C8 Aquapore RP 0.1% acetic acid or 10 mM Barnes et gradient (25 x4.6 mm ammonium acetate (A) & al. (1994a) I.D.) acetonitrile {B). 0 to 50% B in A

Genistein 9.5 !socratic J.t-Bondapak C18 Methanol-1mM ammonium Wanget (10 J.tm; 3.9 mm x acetate (60:40, v/v) al. (1990) 30 cml.D.) Genistein 14 !socratic Partisil-10 ODS Water-acetonitrile (4:1, v/v) Westetal. (1978)

Genistein 7 !socratic Hypersil ODS (5 Methanol-0.1 M ammonium Setchell et J.tm; 25 x 4.6 mm acetate buffer (pH 4.6) (60:40, al. (1987) I.D.) v/v) Genistein 41.9 Linear YMC-pack ODS- 0.1% aqueous acetic-0.1% acetic Murphyet gradient AM-303 acid in acetonitrile al. (1997) 29

Table 2.1, continued

Compound tr Elution Column Mobile phase (v/v) Reference condition Coumestrol 46 Linear DuPont Zorbax IS% aqueous methanol (A) Eldridge gradient ODS (25 x 0.46 em and 65% aqueous methanol (I9S2) I.D.) (B). 25% to 50% A in B. Coumestrol IS Non Ultrasphere-ODS Methanol (A) & water (B). 0 Murphy linear (250 x 4.6 mm I.D.) to IOO% A in B for IS min. (I9SI) gradient

Coumestrol I2 !socratic DuPont Zorbax Methanol-water (65:35, v/v) Lookhart ODS (5 f,lm; 25 x eta/. 0.46 em I.D.) (I97S) Coumestrol I4.5 !socratic f,l-Bondapak CIS Methanol-ImM ammonium Wanget (IO f,lm; 3.9 mm x acetate (60:40, v/v) a/. (I990) 30 cml.D.) Coumestrol 9 I socratic Hypersil ODS (5 Methanol-O.I M ammonium Setchell et f,lm; 25 x 4.6 mm acetate buffer (pH 4.6) a/. (I9S7) I.D.) (60:40, v/v) Formononetin IO.S Gradient Nova-Pak RP-CIS 23% acetonitrile (A) in IO% Franke et aqueous acetic acid (B) al. (I994, linearly to 70% A in B for 8 I995) min. and held at 23% A in B for I2min. Formononetin 11 !socratic Hypersil ODS (5 Methanol-O.I M ammonium Setchell et f,lm; 25 x 4.6 mm acetate buffer (pH 4.6) a/. (19S7) I.D.) (60:40, v/v) Formononetin I3 !socratic f,l-Bondapak CIS Methanol-ImM ammonium Wanget (10 f.lm; 3.9 mm x acetate (60:40, v/v) al. (1990) 30 em I.D.) BiochaninA 12.6 Gradient Nova-Pak RP-CIS 23% acetonitrile (A) in IO% Franke et aqueous acetic acid (B) al. (1994, linearly to 70% A in B for S 1995) min. and held at 23% A in B for 12min. BiochaninA I7 !socratic Hypersil ODS (5 Methanol-O.I M ammonium Setchell et f,lm; 25 x 4.6 mm acetate buffer (pH 4.6) a/. (19S7) I.D.) (60:40, v/v) BiochaninA IS Isocratic f,l-Bondapak CIS Methanol-ImM ammonium Wanget (10 f,lm; 3.9 mm x acetate (60:40, v/v) a/. (I990) 30 em I.D.) 30

2.2.1.3. Gas Liquid Chromatography (GLC)

Nairn et al. (1974) developed a GLC method for the isolation and separation of isoflavones from soya bean meals. A combination of ion exchange chromatography,

GC-MS selection ion monitoring (SIM technique) with deuterated internal standards has been developed for the isolation and the identification of enterolignans and isoflavones from human plasma, faeces and urine, as well as animal urine; and from foods (Setchell et al. 1981a; Adlercreutz et al. 1986a, 1986b; Wahala et al. 1986;

Fotsis & Adlercreutz 1987; Wahala & Hase 1991; Adlercreutz et al. 1993;

Adlercreutz et al. 1995a). Mazur et al. (1996) modified the method for the quantitative determination of isoflavones, coumestrol, and lignans in plant-derived foods. They used isotope dilution gas chromatography-mass spectrometry in the selected ion monitoring mode (ID/GC/MS/SIM).

2.2.1.4. Capillary Electrophoresis

Shihabi et al. (1994) developed capillary electrophoresis methods for the separation of isoflavones and coumestrol from soya beans and soya bean products. All compounds were separated in less than 10 minutes and the minimum detectable limit was 0.4 mg/1 (0.04 mg/100 g).

2.2.1.5. Combined methods (TLC, GLC, HPLC)

All methods can be used to identify zearalenone in foods. However, none has been found to be superior. In 1974, Mirocha et al. developed TLC, GLC, 31

chromatography, GLC-mass spectrometry and a combination of all of these methods for the detection of zearalenone in and barley. They found that the limit of detection and sensitivity using TLC was 0.1 J.tg/kg (0.01 J.tg/100 g) and 50 J.tg/kg

(5 J.tg/100 g), respectively, while sensitivity, using GLC, was less than 50 J.tg/kg

(5 J.tg/100 g). The detection limit ofUV spectrophotometry was between 0.1 J.tg/ml

(10 J.tg/ 100 g) and 0.5 J.tg/ml (50 J.tg/100 g).

Scott et al. (1978) developed TLC, HPLC, and GLC/High Resolution Mass

Spectrometry methods for the determination of zearalenone in com-based food. The detection limit of TLC and HPLC methods was 20 J.tg/kg (2 J.tg/1 00 g) and 5 J.tg/kg

(0.5 J.tg/100 g), respectively.

Mabry et al. (1970) studied the UV spectra of isoflavones and flavones dissolved in different reagents and identified that the UV spectra of isoflavones was in the range of 245 to 270 nm. The individual UV spectra of isoflavones daidzein, genistein, formononetin, biochanin A in methanol were 249, 261, 248, 261 nm, respectively, and confirmed by NMR spectroscopy and paper chromatography.

2.2.1.6. Summary of analytical methods for phytoestrogens

None of the published methods developed for analysis of phytoestrogens has been shown to be superior to any other, and few authors have fully reported their method validation. HPLC methods are considered less time-consuming than GC methods because, in many cases, they do not need derivatization or multistep extraction 32 procedures. The detection limit of coumestrol, daidzein, and genistein using an

HPLC method has been reported by Setchell et al. (1987) to be 5, 10, and 15 pg, respectively, and by Franke et al. (1995) to be 5.15 nM (0.13 mg/100 g), 8.75 nM

(0.24 mg/100 g), 7.25 nM (0.19 mg/100 g), 13.0 nM (0.37 mg/100 g) and 25.70 nM

(0.69 mg/100 g) for daidzein, genistein, formononetin, biochanin A and coumestrol, respectively. HPLC method can also detect genistein up to 0.05 ng in 1 g

(5 ng/100 g) of soya bean samples (K.itada et al. 1986). Most existing HPLC methods do not detect isoflavones and coumestrol in less than 0.1 mg/100 g of foods (Barnes et al. 1994a).

ID/GC/SIMIMS method using a deuterated internal standard was able to detect isoflavones and lignans in food samples of 50 mg (Mazur et al. 1996). The sensitivity ofiD/GC/SIMIMS method was approximately 2-3 J.Lg/100 g (Mazur et al.

1996). However, Wahala et al. (1995) suggested that the deuterated internal standards of isoflavones were unreliable because one or two deuterium labels of isoflavones genistein and daidzein were easily lost. This is due to the hydroxy substitution pattern in the ring A of genistein and daidzein. The use of deuteration is also very expensive.

There is no universally accepted standard method for isoflavones and coumestrol that has been adopted by the AOAC or any other international organization. 33

2.2.2. Extraction

Various solvents and methods have been developed for the extraction of conjugate forms (glucosides, malonylglucosides and acetylglucosides) and unconjugated forms

(aglycones). Various solvents and methods were also developed for the hydrolysis of the conjugate forms to obtain the total conjugate forms and coumestrol from plants and foods as isoflavones occur essentially in the intact plants as conjugate forms.

Nairn et al. (1974) reported that about 99 percent of the isoflavones in soya beans were in the glucoside form. Eldridge (1982b), and Seo and Morr (1984) also indicated that the majority of the isoflavones in soya bean flour and defatted soya flakes were in the glucoside form.

Eldridge (1982a) found that 80% of aqueous methanol gave the maximum extraction for conjugated and unconjugated isoflavones compared with ethanol (50%, 80% or

100%), methanol (50% or 100%), or acetonitrile. Coward et al. (1993) found that the extraction of conjugated and unconjugated isoflavones from soya protein concentrates, with 80% aqueous methanol at room temperature gave higher recovery of the compounds compared to 60% aqueous acetonitrile. Seo and Morr

(1984) used 2N NaOH for the extraction of conjugated and unconjugated isoflavones at room temperature for six hours. Wang et al. (1990) suggested that acetonitrile and

1M HCl were the best solvents for the extraction of isoflavones in whole soya bean when compared with a mixture of acetonitrile--water or

80% ethanol-hydrochloric acid-water or 80% methanol-hydrochloric acid-water

(80:0.5:19.5, v/v/v). Wang et al. (1990) and Franke et al. (1994, 1995) suggested that 34 the best method for the hydrolysis of isoflavone conjugates in soya beans and its processed products was 1M HCl with heating for 2 hours at 98-1 00°C.

Kudou et al. (1991) found that the extraction of soya beans with 70% aqueous ethanol at room temperature resulted in 6" -0-malonyl isoflavones as a major compound, and extraction at 80°C resulted in glucoside isoflavones as a major compound. Barnes eta!. (1994a) also indicated that the extraction of soya beans with

80% aqueous ethanol and heat led to the conversion of isoflavones 6"-0- acetylglucosides and 6"-0-malonylglucosides to their p-glucosides.

Various solvents and methods have been developed for defatting foods prior to analysis. Lookhart et al. (1978) used pentane for defatting soya beans prior to coumestrol analysis, but later (1979) found that petroleum ether was the best solvent available for defatting soya beans for this purpose. West et al. (1978) conducted an initial defatting of soya beans with petroleum ether before the extraction of isoflavones. Wang eta!. (1990) used a Sep-Pak C18 column to clean up the extract before analysis because petroleum ether used for defatting food did not increase the amount of daidzein, genistein, and coumestrol significantly. Coward et al. (1993) found that the initial clean up was unnecessary when using HPLC and the initial extraction with did not alter the qualitative and quantitative aspects, but was just useful in prolonging the life of the column.

Aglycones or unconjugated isoflavones are the bioactive forms whereas the glucosides, malonylglucosides, or acetylglucosides isoflavones are not bioactive.

However, since the isoflavones occur almost exclusively in plants or plant foods in 35 the conjugated form they have to undergo acidic or enzymatic hydrolysis and demethylation to yield the free forms ( aglycones) in the same way as performed by the human gut microflora (Xu et al. 1994) or processing techniques. The bioavailability of isoflavones is affected by the human gut microflora because the capacity of humans to hydrolyse conjugated isoflavones differs from person to person (CV ± 30%) (Xu et al. 1994). Therefore, identification and quantification of the total free forms of isoflavones ( aglycones) is, in this respect, very important in order to represent the potential bioactivity of the foods.

2.3. Role of phytoestrogens in plants

Phytoestrogens are isolated from leaves, stems, roots, flowers and seeds. The production of isoflavonoids by plants is induced in response to microbial attack.

Plants produce antibiotics, the pterocarpan phytoalexins (e.g. ) when they are infected by fungal pathogens. The study by Graham et al. (1990) of all soybean seed organs showed that the isoflavones are present in large quantities as conjugates, the 7-0-glucosyl- and 6"-0-malonyl-7-0-glucosyl-isoflavones. These conjugates were rapidly hydrolyzed to free daidzein and genistein when the soybean tissues were infected with Phytophthora megasperma f. sp. glycinea. Thus high levels of glyceollin subsequently accumulated. The production of phytoestrogens by plants increases in response to mechanical injury, ultraviolet light, freezing, and drought

(Price & Fenwick 1985). Phytoestrogens are also endogenous regulators of growth

(Smith et al. 1986). 36

Levels of isoflavones and coumestans differ according to the parts of plants from which they are derived (Vetter 1995; Kudou et al. 1991; Eldridge & Kwolek 1983;

Lookhart et al. 1978). Lookhart et al. (1978) found that levels of isoflavones in soya bean hulls or seed coat were higher than those found in whole soya beans and endosperms (cotyledons). This is in contrast to the study by Eldridge & Kwolek

(1983) and Kudou et al. 1991, levels of isoflavones in hypocotyl were higher compared with those in hulls (seed coat) and endosperms (cotyledons). Levels increased substantially during germination (Wong et al. 1965) and maturation of seeds (Kudou et al. 1991). Medigo species grown in gravelly soil had higher levels of coumestrol than those grown in sand. Levels were also higher in stems than in pods.

Levels in Burr medic leaves increased after infection by leaf rust (Uromyces striatus)

(Francis & Millington 1970).

Zearalenone and its derivatives are not produced by the plant per se, but are that are synthesised by a number of species of Fusarium, particularly by

F. roseum (Gibberelle zea) (Mirocha et al. 1974) associated with plants. Zearalenone has been found in many commodities such as wheat, maize, barley, , , sesame, and hay (Blaney et al. 1984; Hesseltine et al. 1978). 37

2.4. Occurrence of phytoestrogens in forages and foods

2.4.1. Forages

Bradbury and White (1954) first discovered genistein, biochanin A, and formononetin in subterranean clover (Trifolium subterraneum), and Bickoff et al.

(1957) first discovered coumestrol in forage crops (strawberry clover, ladino clover, and alfalfa). Subsequently, many studies found isoflavones and coumestrol in several forages, such as formononetin and genistein in alfalfa and ladino clover, and biochanin A and coumestrol in alfalfa (Guggolz et al. 1961; Livingstone et al. 1961).

Daidzein, genistein, formononetin and biochanin A were also found in many different parts of several Trifolium species by Vetter (1995).

2.4.2. Foods

2.4.2.1. International

Isoflavones and coumestrol have been found in a range of plant products frequently consumed by humans. These compounds were found most abundantly in the plant family Leguminosae and in several cereals. Walz (1931) first isolated the isoflavone genistin from soya bean meal, and found that hydrolysis of genistin resulted in the production of genistein and glucose. Walter (1941) confirmed some ofWalz's work and presented the evidence that the sugar produced by the hydrolysis of genistin was d-glucose. Soya beans are known to contain the highest isoflavones compared with other legumes consumed by humans (Franke et al. 1994, 1995; Mazur et al. 1998). 38

Several studies have identified isoflavones and coumestrol in many varieties of plant foods and food products. Coumestrol was identified in Soja hispida (Zilg &

Grisebach 1968); daidzein in Cicer arietinum (Wong et al. 1965); zearalenone in com (Shotwell et al. 1970; Stoloff et al. 1976; Bennett et al. 1976), wheat

(Hesseltine et al. 1978) and maize (Blaney et al. 1984).

The lignans secoisolariciresinol and matairesinol have been found in many varieties of foods but were particularly high in flaxseed (Linum usitatissimum) (Mazur et al.

1996). Lignan isolariciresinol, pinoresinol, secoisolariciresinol and matairesinol have been isolated from flaxseed meal by Meagher et al. (1999). The isoflavones content of soya beans and soya foods from several countries has been reported. These were the USA (Dwyer et al. 1994; Wang & Murphy 1994b; Murphy et al. 1997; Franke et al. 1994, 1995; Wang et al. 1990; Coward et al. 1993), Japan (Futukake et al. 1996;

Franke et al. 1994, 1995), and Hawaii and Singapore (Franke et al. 1999).

Secoisolariciresinol, matairesinol, formononetin, biochanin A, daidzein, genistein, and coumestrol have also been identified in tea (Mazur et al. 1998), and in beer

(Lapcik et al. 1998).

2.4.2.2. Australia

Dalais et al. (1997a, 1997b) reported the daidzein, genistein, daidzin, and genistin content of soya beans grown in Australia and in soy foods purchased from Australian markets but the brands and varieties were not specified. Knight et al. (1998) reported the daidzein and genistein content of cow's milk, yoghurt, soya drinks, casein-based infant formulas and soya-based infant formulas purchased in Australia, but did not 39

report any other isoflavones or coumestrol. King and Bignell (2000) provided values

for daidzein, genistein, glycitein, and their conjugates in a range of Australian soya

beans and soya foods (varieties and brands not specified).

2.4.3. Factors affecting phytoestrogens in foods

The levels of isoflavones and coumestans in plants are affected by factors such as

geographical location, differences in variety, condition of culture, harvesting date, maturation stage, and storage methods (Alexander & Rossiter 1951; Ferrando et al.

1961; Wong et al. 1965; Francis & Millington 1971; Smith et al. 1986; Kudou et al.

1991; Eldridge & Kwolek 1983; Murphy 1982; Smith et al. 1986; Vetter 1995). A summary of data for isoflavones and coumestrollevels in legumes and other products from many countries is presented in Appendix 1.

2.4.3.1. Agricultural conditions

The levels of isoflavones in foods are significantly affected by variety, harvest date, crop year, and storage. Eldridge and Kwolek (1983) found that levels of isoflavones in soya beans from different varieties ranged from 116 to 309 mg/100 g on a wet weight basis. The levels varied in the same variety were grown in different locations.

The levels ranged from 46 to 195 mg/100 g on a wet weight basis. The levels also varied from year to year when soya beans were grown in the same location. The variation was from 1,176 to 3,309 mg/100 g for different crop years and from 1,176 to 1,749 mg/100 g for different locations on a wet weight basis. 40

Total isoflavone levels varied in eight varieties of American soya beans with the

levels ranging from 2,053 to 4,216 mg/100 g on a wet weight basis (Wang & Murphy

1994b). The levels in different varieties of soya beans from Japan ranged from 2,041

to 2,343 mg/100 g (Wang & Murphy 1994b). The levels in the same varieties of soya

beans but different crop years ranged from 1,261 to 1,417 mg/100 g on a wet weight

basis (Wang & Murphy 1994b).

Smith et al. (1986) indicated that the concentration of formononetin, biochanin A,

and genistein present in subterranean clover was influenced by harvest dates and not

by storage. Vetter (1995) also indicated that the levels of isoflavones in Trifolium

species varied in different varieties and different parts of the plants.

2.4.3.2. Processing and preparation

Processing decreases the levels of glucoside forms of phytoestrogens in foods because hydrolysis and further breakdown may occur during processing. The production of tofu did not change the total isoflavones concentration, but did cause a

change in the composition (Wang & Murphy 1996). Genistin and daidzin (glucoside isoflavones) levels decreased significantly in tofu, but genistein and daidzein (free isoflavones) remained unchanged (Wang & Murphy 1996). Although isoflavones are easily hydrolysed chemically and enzymatically during processing, isolation or analysis (Price & Fenwick 1985), these compounds are relatively stable when heated.

Boiling causes only a small loss of genistin (Murphy 1982) and does not destroy daidzein and genistein in soya beans (Franke et al. 1994). Although roasting causes 41 losses of 15% and 21% of daidzein and genistein, respectively, in soy beans, the levels of daidzin and genistin are unaffected (Franke et al. 1994).

Fermentation processes hydrolyse phytoestrogens in foods. Isoflavones, in the form of glucosides, exist originally in soya beans before being processed into tempeh.

During the fermentation process, glucosides are hydrolysed to aglycones by microorganism activity. Daidzein and genistein are major aglycones in tempeh

(Murakami et al. 1984). The study by Ikeda et al. (1995) found both isoflavone glucosides and their aglycones in tempeh after fermentation for 40 hours at 31 °C but only aglycones in mamemiso after fermentation for 30 days at 30°C.

2.4.3.3. Method of extraction

Soya protein concentrates produced by aqueous leaching contain higher levels of isoflavones as compared with extraction by aqueous alcohol (Eldridge 1982b).

Various protein isolation methods such as acid precipitation, dialysis, ion exchange and activated charcoal treatment decreased the concentration of phytoestrogens in soya products (Murphy et al. 1982; Seo & Morr 1984). Extraction of oil from soya beans with hexane did not remove the isoflavones or the isoflavone glucosides because isoflavones are not soluble in hexane (Eldridge & Kwolek 1983). 42

2.5. Metabolism of phytoestrogens

Intestinal bacteria rapidly modify phytoestrogens structurally after ingestion to form compounds that have estrogenic and antiestrogenic activities. These are then absorbed into the circulation. Metabolites of phytoestrogens have been detected in urine, blood plasma, semen, milk and faeces of many animals and humans (Marrian

& Haslewood 1932; Richardson & Mattarella 1977; Axelson et al. 1982a: Dehennin et al. 1982; Bannwart et al. 1984a, 1984b; Adlercreutz et al. 1986b). The different metabolites and pathways are described in the following sections.

2.5.1. Lignans

Plant lignans are metabolised by intestinal bacteria after ingestion to the mammalian lignans. Secoisolariciresinol is metabolised to enterodiol (2,3-bis (3-hydroxybenzyl) butane-1, 4-diol, and matairesinol is metabolised to enterolactone (trans-2, 3-bis (3- hydroxybenzyl)-y-butyrolactone) (Axelson et al. 1982b). Enterodiol and enterolactone are formed by the removal of two methyl groups and two hydroxyl groups from secoisolariciresinol and matairesinol. Enterolactone is further oxidised to enterolactone (Borriello et al. 1985; Setchell et al. 1981b). Formation of enterolactone and enterodiol by human intestinal bacteria is shown in Figure 2.29

(page 44) (Kurzer & Xu 1997).

The metabolites, enterolactone and enterodiol, have been found in human urine, faeces, and plasma (Stitch et al. 1980; Setchell et al. 1980, 1981a; Borriello et al.

1985). Matairesinol has also been identified in human urine (Bannwart et al. 1984a; 43

1984b). In a recent study reported by Jacobs et al. (1999), six metabolites of enterolactone were discovered in the urine of female and male humans after ingestion of flaxseed for five days. These metabolites were the products of monohydroxylation at the para-position and at both ortho-positions of the parent hydroxyl groups in the aromatic ring of enterolactone. Jacobs et al. (1999) also found three metabolites of enterodiol in the urine of female and male humans after ingestion of flaxseed for five days, which were formed through aromatic monohydroxylation at the para- and ortho-positions. 44

Secoisolarfcirellinol-diglucoside

~OH·

.~OH

OH Sec:oisolariciresi1ol Mataitesinol l

Enterolactone

Figure 2.29. Formation of enterolactone and enterodiol by human faecal flora (Kurzer & Xu 1997). 45

Urinary excretion of enterolactone and enterodiol correlates significantly with the intake of foods containing plant lignans. A study by Lampe et al. (1994) found that levels of enterolactone and enterodiol excretion increased in omnivorous women after consumption of their usual diet, supplemented with flaxseed powder ( 10 g/ day). The level of enterolactone increased from 3.16 ± 1.47 Jlmollday to 27.79 ± 1.50

Jlmol/day, while enterodiol increased from 1.09 ± 1.08 Jlmol/day to 19.48 ± 1.10

Jlmol/day in urine. In the same study, Kurzer et al. {1995) found that faecal enterolactone and enterodiol excretion increased an average of 16- and 32-fold, respectively. Matairesinol excretion increased 1.6-fold. Faecal enterolactone increased from 0.08 ± 0.08 Jlmollday to 2.56 ± 3.10 Jlmollday, enterodiol increased from 0.64 ± 0.48 Jlmol/day to 10.3 ± 7.58 Jlmollday, and matairesinol increased from

0.007 ±0.01 Jlmol/day to 0.012 ± 0.008 Jlmol/day.

2.5.2. Isoflavones

The metabolism patterns of isoflavones in humans are similar to those of lignans and also to those in animals. Isoflavones are metabolised by bacteria in the . The metabolism of coumestans has not yet been characterised.

Daidzein is metabolised to dihydrodaidzein, which is further metabolised to both equol and 0-desmethylangolensin (0-DMA). Genistein is transformed to dihydrogenistein and is further metabolised to 6"-hydroxy-0-DMA (Kurzer & Xu

1997). The metabolic pathway of daidzein and genistein catabolism in humans is shown in Figure 2.30 (overleaf) (Kurzer & Xu 1997). 46

F ormononetin is metabolised mostly to equol, to daidzein and to 0-

desmethylangolensin (Price & Fenwick 1985). The metabolic pathway of

formononetin (Figure 2.31, overleaf) in humans is similar to that in sheep

(Adlercreutz et al. 1987). In the rumen of sheep, formononetin is demethylated to

daidzein and methyl equol. This is then converted to equol by hydrogenation and ring fission. Other active metabolites of formononetin are also formed as angolensin

and 0-desmethylangolensin (Price & Fenwick 1985). A study by Shutt et al. (1970) found that 70% of the formononetin ingested by sheep was converted to equol. A

study by Cayen et al. (1965) found that the injection of formononetin into hen, produced urine containing formononetin, daidzein and equol.

Daidzein Genistein

Dihydrodaidzein Dihydrogenistein

H

0-desmethylangolensin (0-DMA) Equol 6'-hydroxy-0-DMA

Figure 2.30. Metabolic pathway of daidzein and genistein by human gut bacteria and the formation of equol (Kurzer & Xu 1997). 47

OCH~ ..

HO HO r Angolensin r0-desmethylang<>lensin OCH:a H ....

ttO ·Ho l Fonnononetin J Daidzein OCH:a

.... HO HO

4' -methyl ether of equol Equol

Figure 2.31. Rumen metabolism offormononetin in sheep (Price & Fenwick 1985).

Blochanln A Genistein p-Ethyl phenol

Figure 2.32. Rumen metabolism ofbiochanin A in sheep (Price & Fenwick 1985). 48

Biochanin A is metabolised via demethylation to genistein and then via ring cleavage to the estrogenically inactive p-ethyl phenol and organic acids. The metabolic pathway of biochanin A (Figure 2.32, previous page) was studied in the rumen of sheep (Price & Fenwick 1985). A study by Batterham et al. (1965) of ewes treated with biochanin A (5 g per day for 4 days), fmmd that p-ethyl phenol was the major metabolite excreted, whilst biochanin A and genistein were the least. Cayen et al.

(1965) reported that injection ofbiochanin A into hens resulted in the appearance of genistein and equol in their urine.

Isoflavones or lignans undergo deconjugation and conjugation steps if they are absorbed and utilised by the body. The sugar moiety of both isoflavones and lignans is hydrolysed by human gut bacterialj3-glucosidases, gastric hydrochloric acid and

13-glucosidases of foods (Kelly et al. 1993). The metabolites are then absorbed by the and are further conjugated with glucoronic acid and sulfated by hepatic phase II enzymes (UDP-glucoronosyltransferases and sulfotransferases).

These conjugates are excreted in the urine and in the bile undergoing enterohepatic circulation. After excretion into the bile, conjugated isoflavones and lignans can be deconjugated again by gut bacteria. Deconjugation may promote reabsorption, further metabolism and degradation in the lower intestine (Xu et al. 1995).

Equol was found in human urine (Axelson et al. 1982a) and related significantly to the ingestion of foods containing equol precursors. A study by Axelson et al. (1984) found that urinary excretion of equol in humans increased by 100 to 1000 times after ingestion of a single meal containing soya protein, which is a source of isoflavones

(Axelson et al. 1984). In a study by Kelly et al. (1993), 12 healthy Caucasian men 49

and women consumed 40 g soya flour for two consecutive days. Their urinary

excretion of daidzein, genistein, and glycitein increased four, six and 12 times, respectively, over a three day period. However, only 7 to 30% from the total amount of isoflavones present in soya flour was recovered in urine. In the same study, Franke et al. (1995) found that urinary levels of daidzein, genistein and the metabolite, 0- desmethylangolensin in women increased dramatically the first 24 hours after consuming 40 to 96 g of roasted soya beans containing 42-100 mg of genistein and

32-77 mg of daidzein. Zhang et al. (1999) found that the average 48-h urinary levels of glycitein, daidzein and genistein were 55, 46 and 29%, respectively, of the dose ingested in trial with seven women and seven men.

Levels of daidzein and genistein in the plasma of 12 adult women increased from negligible levels to 0.79 - 2.24 J.!M and to 0.74- 2.15 J.!M, respectively, 6.5 hours after receiving a single dose of 0.7, 1.3 or 2.0 mg ofisoflavones/kg of body weight.

The average recovery of daidzein and genistein found in urine was approximately 21

± 8% and 9 ± 5% of the total amount consumed, respectively. The average recovery of daidzein and genistein found in faeces was approximately 1 to 2% of the total amount ingested (Xu et al. 1995). Four volunteers consumed beverages containing

20 g of isolated soya protein (21.0 g of genistein and 13.6 mg of daidzein) twice for

14 days. The plasma levels of genistein and daidzein taken at 6.5 hours after consumption ranged from 496-644 nM and 289-424 nM, respectively (Coward

1996). 50

2.5.3. Factors affecting metabolism of phytoestrogens

There were various metabolic responses of dietary lignans and isoflavones in humans

(Kelly et al. 1993; 1995; Lampe et al. 1994; Lundh 1995). About 30-40% of the subjects studied excreted significant quantities of equol after the consumption of isoflavones (Setchell et al. 1984). Human excretion of isoflavone metabolites is variable. Variation in levels found in urine and plasma is a reflection of the ability of individuals to ferment dietary isoflavones per se. Some subjects (about 33%) may not be able to metabolise or excrete them as biologically active forms (Setchell et al.

1984). Factors such as the composition of intestinal microflora, intestinal transit times, variability in redox level of the large intestine, composition of the basal diets, contamination levels of diet with xenobiotics, as well as antibiotics, could influence levels ofphytoestrogens in urine and faeces (Adlercreutz et al. 1986a; Setchell et al.

1981b).

Several studies have investigated urinary excretion of lignans and isoflavones in various population and dietary groups (omnivorous, vegetarians, !acto-vegetarians, adherents of macrobiotic diets, and those consuming a traditional Japanese diet). The highest concentration of urinary isoflavones was found in American adherents of macrobiotic diets, !acto-vegetarians, and Japanese men, with levels of 3,412-8,870,

885-2,188, and 1,829-3,630 nmollday, respectively. Japanese women frequently had very high concentrations of equol, daidzein and 0-desmethylangolensin with a large variation in the levels (347-6,610 nmollday) (Adlercreutz et al. 1991a). However,

Japanese women and men had a very low urinary excretion oflignans (Adlercreutz et al. 1991a). Urinary excretion of lignans and equol (although not statistically 51 significant for equol) was lower in postmenopausal Finnish breast cancer patients

(1,302-2,835 nmol/day) than in control, omnivorous, and vegetarian women

(Adlercreutz et al. 1982.). Faecal excretion ofisoflavones and lignans of omnivorous women was lower than that of vegetarian women living in Boston (Adlercreutz et al.

1995a).

2.6. Effects of phytoestrogens in humans and animals

2.6.1. Experimental animals

2.6.1.1. Estrogens and

Estrogenicity is defined as a physiological response of a compound that induces estrus in vivo. This causes an increase of uterine weight. The estrogenic activities of phytoestrogens exhibit affinities for binding an (ER).

Phytoestrogens bind unfilled estrogen receptor sites in the cytoplasm to form complexes. The cytoplasm receptor-phytoestrogen complexes are then translocated to the nucleus and bind nucleus receptors to form nucleus receptor-phytoestrogen complexes. These are then processed in a manner analogous to estradiol-bound receptors, which have an appropriate biological response in stimulating growth

(Noteboom & Gorski 1963; Folman & Pope 1969; Shemesh et al. 1972; Shutt & Cox

1972; Lee et al. 1977; Martinet al. 1978; Kitts 1986; Miksicek 1993).

Under certain conditions, phytoestrogens show their antiestrogenic activity.

Phytoestrogens either decrease cytoplasmic estrogen receptor concentration, 52 producing an insensitivity of the target tissue to estrogen stimulation, or interact with estrogen receptors to form complexes, which are unable to initiate the biosynthetic events leading to the growth of target tissue. The estrogenic and antiestrogenic effects ofphytoestrogens are determined by bioassay or uterotrophic assay.

2.6.1.2. The relative estrogenic potency of isoflavones and coumestrol

The relative binding affinity of phytoestrogens has been evaluated by their affinities to compete with eHJ estradiol for binding unfilled cytoplasmic estrogen receptor or unfilled nucleus estrogen receptor sites in in vivo and in vitro studies. Various relative molar binding affinities of phytoestrogens have been reported by different authors. The variation may have been due to the differences in cell cultures or animals used and tested.

Coumestrol was reported to be the most potent estrogenic compound compared with genistein and diethylstilbestrol based on uterine weight assay in mice fed with these compounds (Bickoff et al. 1957). However, the relative binding affinity of coumestrol was about 20% that of estradiol, based on bovine serum albumin assay, and on calf uterine estrogen receptor assay (Lee et al. 1977), but was similar to that of zearalenone using human breast cancer cell lines (MCF-7) (Martin et al. 1978).

Genistein was reported more potent compared with formononetin based on uterine weight assay in immature female mice fed with forage containing these compounds

(Bickoff et al. 1962), but less potent compared with coumestrol and zearalenone using human breast cancer cell lines (MCF-7) (Martin et al. 1978). Using transient 53 transfection tests in HeLa cell cultures, genistein and daidzein were shown inherently more active than biochanin A and formononetin. They exerted their hormone-like activities at concentrations ranging from 0.1 to 10 J.LM (Miksicek 1995). However, according to Cheng et al. (1954), the relative estrogenic activity of genistein and biochanin A was less than that of daidzein, based on uterine weight assay in mice fed with these compounds. Makela et al. (1995) using MCF-7 or T-47D breast cancer cell cultures demonstrated that the estrogenicity of phytoestrogens was in the following order: 17f3-estradiol > zearalenone > coumestrol > genistein > biochanin

A. The estrogenic activities of genistin and daidzin were evaluated in the B6D2F1 strain of mouse by Farmakalidis et al. (1985). The estrogenic response of 1.5 mg genistin and 3.8 mg daidzin was equivalent to that of 1 mg genistein.

The estrogenic activity of coumestrol is associated with the presence of two hydroxyl groups in both the 4' and 7 positions and in a furan ring (Bickoff et al. 1960).

Relative estrogenic potencies of coumestrol, evaluated by uterine weight response of immature mice, showed that its activity decreased to two-thirds by esterification of both hydroxyl groups. It decreased to one-sixth after the removal of one of the two hydroxyl groups, and responses were completely eliminated through the removal of both hydroxyl groups. The presence of the additional hydroxyl groups in coumestrol greatly diminishes its estrogenic activity (Bickoff et al. 1960). Free hydroxyl groups in positions 7 and 12 (4 ') of isoflavones are reported to be essential for effective interaction with estrogen receptors (Shemesh et al. 1972).

The relative estrogenic activities of genistein, biochanin A, daidzein, and formononetin present in Montgomery red clover relative to estradiol were 1.5, 1.0, 54

0.4, and very little or none, respectively, based on a mouse uterine weight assay

(Wong & Flux 1961). Concentrations of coumestrol and isoflavones applicable to exert their estrogenic effects were tested in vitro bioassays utilizing human endometrial adenocarcinoma Ishikawa-V ar I line cell culture. The concentration was estimated as estrogenic potency relative to estradiol (defined as 100 for estradiol): coumestrol 0.202, genistein 0.084, daidzein 0.013, biochanin A <0.006 and formononetin <0.0006 (Markiewicz et al. 1993). While it is clear that phytoestrogens are considerably less active than estradiol, on a weight-for-weight basis, they may be present in foods and feeds at high enough concentrations to exert estrogenic effects.

2.6.1.3. Reproductive organs and sexual differentiation

The estrogenic activity of phytoestrogens was first recognised in relation to a syndrome known as clover disease in sheep (Bennetts et al. 1946; Branden et al.

1967; Shutt et al. 1968). Phytoestrogens caused infertility in ewes consuming clover which resulted in the failure of sufficient sperm transport (Lightfoot et al. 1967).

Phytoestrogens also caused an increase of uterine weight amongst fattening lambs fed clover hay (Cheng et al. 1953).

Subsequently, many studies on the effects of phytoestrogens on the reproductive organs of experimental animals were reported. Genistein and coumestrol caused an increase in uterine weight of rats after a single injection of genistein and coumestrol diacetate (Noteboom & Gorski 1963). Levy et al. (1995) reported the effects of different dosages of genistein given to pregnant Charles CD rats. Twenty five mg of genistein caused a significantly smaller birth weight of female offspring compared to 55

controls. Five mg of genistein caused a significantly shorter anogenital distance of

both male and female offspring and a significantly later onset of vaginal opening of

female offspring compared to controls. Five mg of genistein decreased the sexually

dimorphic nucleus of the medial (SDN-POA) volume of female

offspring, though not significantly.

A study by Medlock et al. (1995) showed that 100 J.lg of coumestrol injected into

Sprague-Dawley pups on days one to five after birth, caused premature uterine gland

development and increased uterine weight. Burroughs (1995) observed the long-term

morphological alterations in the reproductive tract of mice exposed neonatally to

coumestrol. A dose of 0.001 J.lg coumestrol caused precocious vaginal opening and a dose of 100 J.lg coumestrol exhibited long periods of cornification at ages of 60 and

150 days. This elicited ovary-independent persistent vaginal cornification. A dose of

2: 5 J.lg coumestrol resulted in ovary-dependent uterine squamous metaplasia.

Coumestrol reduces ovary weight and causes a development of persistent vaginal estrus and anovulatory ovaries of female pups injected subcutaneously with a single dose of coumestrol (10 and 100 mg/kg) (Leavitt & Meismer 1968). Universal disruption of the cycle of persistent-estrus type in the female offspring of mother rats fed with coumestrol in feed (1 g coumestrol/100 g feed) during lactation was observed (Whitten et al. 1993). Markaverich et al. (1995) demonstrated that coumestrol given subcutaneously as multiple injections or orally via drinking water to rats did not stimulate uterine cellular hyperplasia and did not increase uterine

DNA content, but significantly increased uterine weight. 56

The estrogenic activity of formononetin in red clover silage (3.5 g formononetin/3.5 kg red clover silage) caused clinically significant changes in the reproductive organs of ewes fed with red clover silage. It changed the colour of the to pink and red, enlarged the vulva, and udder, increased teat length and circumference, and also stimulated secretion of milky fluid (Nwannenna et al. 1995).

2.6.1.4. Cell proliferation and differentiation

The estrogenic effects of phytoestrogens have been studied in vitro. They have been reported to stimulate or inhibit cell proliferation and differentiation. Phytoestrogens have been found to stimulate growth of estrogen-dependent MCF-7 human breast cancer cells (Martin et al. 1978; Welshons et al. 1987; Wang & Kurzer 1996).

Coumestrol, genistein, biochanin A, daidzein, enterolactone enhanced cell proliferation at concentrations below 1-10 JJ.M but at concentration of 0.0018,

0.0025, 0.04, and 0.3 JJ.M, coumestrol, genistein, biochanin A, and daidzein, respectively, enhanced half maximal growth (Martin et al. 1978; Welshons et al. ·

1987; Makela et al. 1994). Formononetin enhanced mammary gland proliferation in female mice administered with 40 mg/kg/day over a five day period (Wang et a/.1995).

Barnes et al. (1990) showed that 5%, 10% and 20% of soya bean protein concentrate in diet (782 JJ.g/g of daidzein and 510 JJ.g/g of genistein) reduced mammary tumour growth ofN-methyl-N-nitrosourea (NMU) and 7,12- dimethyl-benz (a) anthracene

(DMBA) induced breast cancer in rat models. Inhibitory effects of isoflavones on the serum-stimulated growth ofMDA-468 human breast cancer cell lines not containing 57 estrogen receptors were observed. Genistein was found to be the most potent inhibitor (ICso values from 6.5 to 12.0 J.t.g/ml). Biochanin A and daidzein were weak

(ICso values from 20 to 34 J.t.g/ml), whereas genistin and daidzin were the weakest

(ICso values> 100 J.t.g/ml) (Peterson & Barnes 1991).

Genistein and biochanin A inhibited serum and EGF (epidermal growth factor)­ stimulated growth of LNCaP and DU-145 human prostate cancer cell lines (ICso values from 8.0 to 27 J.t.g/ml for serum and 4.4 to 15 J.t.g/ml for EGF). Neither genistein nor biochanin A inhibited the EGF receptor tyrosine (Peterson & Barnes

1993). Genistein and daidzein also inhibited the growth of MCF-7 human breast cancer cells with ICso values of 10.5 and 28 J.t.g/ml, respectively (Barnes et al.

1994b). Genistein reduced mammary tumour cases to 12% in female offspring of

Sprague-Dawley rats with chemically-induced mammary tumours. The dose of genistein administered to offspring on days 2, 4, and 6 postpartum was 5 mg

(Lamartiniere et al. 1995). Genistein stimulated the expression of estrogen­ responsive pS2 mRNA at concentrations as low as 0.01 J.t.M and induced a proliferation ofER-positive MCF-7 cells but not ER-negative MDA-MB-231 cells at concentrations between 0.01 and 1.0 mM (Wang et al. 1996). Yu et al. (1999) found that doses of 15 and 40 mM genistein and daidzein, respectively, reduced the density of cell surface charge of human colon tumour (HCT) cell cultures using 3-(4,5- dimethylthiazol-2-yl)-2,5-diphenyltetra-zolium bromide calorimetric assay, and inhibited the growth rate ofHCT cells by 50%. 58

The inhibitory effects of genistein, daidzein, and zearalenone were also evaluated in

isolated rat adipocytes. Doses of 0.01, 0.1 and 1 mmol/L of genistein, daidzein, and

zearalenone, respectively, inhibited insulin stimulated lipogenesis at basal and 1

mmol/L concentrations. A dose of higher than 1 mmol/L of genistein and daidzein

enhanced basal levels of lipolysis. A dose of 1 mmol/L and > 0.1 mmol/L of

genistein and daidzein, respectively, inhibited 1 J..Lmol/L of epinephrine-stimulated lipolysis. A dose of 0.1 mmol/L daidzein augmented epinephrine-stimulated lipolysis while a dose of =:::: 1 mmol/L zearalenone reduced epinephrine-stimulated lipolysis

(Kandulska et al. 1999).

Folman and Pope (1962) found that the inhibitory and the enhancement effects of coumestrol on the eHJ estradiol uptake by the uterus and vagina were more influenced by the retention of the compound. Coumestrol enhanced the [3H] estradiol uptake on the first day after being given to mice and reduced the eH] estradiol uptake after the third day. Coumestrol could produce significant agonistic actions in several estrogen-dependent tissues. Doses ranging from 0.01 to 0.1% of coumestrol induced uterine growth and progestin receptor biosynthesis in immature female rats.

Doses ranging from 1 to 10 J..Lg of coumestrol increased the basal levels of luteinizing hormones in neonatal female rats injected with coumestrol. A dose of 10 J..Lg coumestrol prevented an increase of the GnRH-induced LH (gonadotrophin releasing hormone-induced luteinizing hormone) (Register et al. 1995). Doses ranging from 1 to 10 J..Lg of coumestrol also increased nuclear estrogen binding in rats in in vivo and in vitro assays (Whitten et al. 1992a, 1992b, 1994). 59

A recent study by Ko and Malison (1999) demonstrated that genistein did not display its estrogenic effect on the reproductive function of female yellow perch (Perea jlavescens) but did have a positive effect on growth.

Genistein has been used as a cancer chemoprevention agent and drug to inhibit, reverse and retard the cancer process. In an in vitro study, genistein significantly reduced (up to 34.1%) aberrant crypts in an azoxymethane-induced colon carcinogenesis model (Steele et al. 1995).

2.6.1.5. Protein tyrosine kinases

Protein phosphorylation is a key component of many signal transduction systems for diverse cell functions, cellular differentiation and malignant transformations. Serine, threonine, and tyrosine are amino acid residues involved in a phosphorylation process in euk:aryotes, and histidine and aspartate are in prokaryotes. To support these activities, the cellular receptors for growth factors such as epidermal growth factor (EGF), PDGF, insulin, and insulin-like growth factor I and CSF-I are required.

Oncogenes are mutated forms of normal cellular protein involved in growth factor­ stimulated signal transduction. Tyrosine protein kinases (PTKSs) are retroviral oncogenes of 'src' gene family products that play a key role in the tumorigenesis processes. The activities of tyrosine protein kinases can be inhibited by genistein, daidzein, and biochanin A (Markovits et al. 1989). A dose of 110 mM genistein inhibited 50% of protein histidine kinases in cell extract of yeast Saccharomyces cereviseae (Huang et al. 1992). 60

Tyrosine-specific protein kinase activity of the epidermal growth factor (EGF) receptor (pp6ov-src and pp110gag-fes) was inhibited in vitro by genistein (Akiyama et al.

1987). The inhibition was competitive with ATP and noncompetitive with phosphate acceptor (histone H2B). Genistein barely inhibited the enzyme activities of serine­ and threonine-specific protein kinases. Genistein was able to block mutagenic effects mediated by EGF on Nlli-3T3 cells (ICso value of 12 JlM), insulin (ICso value of 19

JlM) or thrombosin (ICso value of 12 JlM), although the thrombosin receptor did not involve a protein tyrosine kinase activity. Genistein also prevented the stimulation of specific protein kinase (S6-kinase) activity and in situ S6 phosphorylation of cells treated by EGF (Linassier et al. 1990). Doses of one or five Jlmol genistein inhibited

TPA-induced proto-oncogenesis expression (c-fos) which caused prolonged tumour latency and decreased tumour multiplicity to 50% in mouse skin tumorigenesis model (Wei et al. 1995).

2.6.1.6. Angiogenesis and methylation

Angiogenesis is the formation of new blood vessels and is essential for repair of wounds. Angiogenesis is a tightly-regulated and self-limited process. In healing of wounds, the endothelial cells can undergo rapid proliferation during a spurt of angiogenesis. However, many diseases are driven by persistent unregulated angiogenesis (Folkman & Shing 1992). In these conditions, only 10% or more of the vascular endothelial cells can proliferate actively (Denekamp 1990). In an in vitro study, genistein inhibited basic fibroblast growth factor (FGF) which induced the invasion of bovine microvascular endothelial (BME) cells (Fotsis et al. 1995). 61

Methylation IS an epigenetic mechanism involved m the activation

(hypomethylation) or inactivation (hypermethylation) of cellular genes that are lmown to play a role in carcinogenesis. Lyn-Cook et al. (1995) found that coumestrol and equol caused hypermethylation of the c-H-ras protooncogene in neonatal rats fed with a single dose of coumestrol and equol (100 J..tg).

2.6.1.7. Leukemia

B-cell precursor (BCP) leukemia is the most common form of acute lymphocytic leukemia (ALL) in children and the second most common form of ALL in adults.

Leukemia is a group of disorders in which proliferation, maturation and release of leucocytes and related cells are no longer restricted by the normal physiological control mechanism. Acute lymphocytic leukemia is a haematopoietic malignancy involving the progressive infiltration and replacement of normal bone marrow and lymphatic tissue by abnormal lymphoid precursors with immunocyte-specific determinants. Clinical symptoms of ALL are malaise, fatigue, weight loss, pallor, easy bruising or bleeding and finally death (Stass 1987).

In 1995, Uckun et al. found that genistein, with monoclonal antibody B43 immunoconjugate, bound with high affinity to BCP leukemia cells, inhibited selectively CD 19-associated tyrosine kinases, and triggered rapid cell death.

Genistein reduced the proliferation of promyelocytic (HL-60) leukemia, myelogenous (K0562) leukemia and SK-:MEL-131 melanoma cell cultures. A dose of 10 J..tg/ml genistein reduced Ill.,-60 and K-562 cells by 20% to 50% and a dose of

15 J..tg/ml genistein reduced SK-:MEL-131 melanoma cells (Constantinou & 62

Huberman 1995). Doses ranging from 0 - 200 1-1m of biochanin A inhibited the

growth of myeloid leukemia cell lines WEHI-3B (JCS) and induced the

morphological differentiation of JCS cells (Fung et al. 1997).

2.6.1.8. Non-mutagenic and antimutagenic activity

Daidzein, genistein, formononetin and biochanin A, coumestrol and zearalenone are

non-mutagenic compounds (Bartholomew & Ryan 1980), but have antimutagenic

activity. A study by Plewa et al. (1999) found that isoflavones (daidzin and genistin),

when isolated from concentrate (PCC) soya beans, repressed the

genotoxicity and mutagenicity of 2-acetoxy-acetylaminofluorene (2AAAF)-induced

DNA damage in Chinese hamster lung cells, Chinese hamster ovary cells, and human

lymphocytes.

2.6.1.9. Antioxidant activity

Isoflavones possess antioxidant activities. In an in vitro study, Wei et al. (1995) found that H202 formation by TPA-activated HL-60cells was inhibited by isoflavones. Genistein was the most potent inhibitor, daidzein was second and biochanin A was the weakest. The inhibiting effects of genistein were exerted by the presence of the hydroxyl groups at position 4' and 5, and the second aromatic ring at the position C-3. Isoflavones also inhibited the formation of superoxide anion by xanthine/xanthine oxidase. Genistein increased antioxidant enzyme activity in skin and the small intestines of mice fed with 250 mg/kg of genistein (Wei et al. 1995).

Murakami et al. (1984) reported that daidzein and genistein were the main 63

components responsible for the stability of tempeh throughout the oxidation process.

This result was obtained from incubating the tempeh oil.

2.7. Cancer

The differences in the incidence of and death rates from cancers across countries are

mostly caused by environmental differences, and particularly by diet. Coleman et al.

(1993) reported that the death rates from many forms of cancer in Asian countries

were lower than those in Western countries. The highest death rates for all cancer

sites for males and females among 50 countries (1986-1988) were in Hungary (235.4

per 100,000 population) and Denmark (239.4 per 100,000 population). The cancer mortality rates for Australia were in the middle or ranked 24th for males (162.8 per

100,000 population) and 25th for females (102.0 per 100,000 population) (Parkin

1992). Death rates from all malignant neoplasms in Australia in 1996 were reported to be 225.5 and 136.9 per 100,000 population for males and females, respectively

{Australia's Health 2000). Incidence rates from all cancer sites in Semarang,

Indonesia (1980-1981) were much lower at 51.4 and 78.4 per 100,000 population for males and females, respectively (Parkin 1986), but more recent data are not available.

The highest death rates from breast cancer in the world, in 1995, were in Iceland

(35.3 per 100,000 population) and Malta (31.5 per 100,000 population), and the lowest were in Asian countries, Republic of Korea (3.8 per 100,000 population) and

Japan (7.5 per 100,000 population). Deaths from breast cancer in Australia have remained stable within the last six years (1990-1995), with rates of approximately 64

20.1 per 100,000 female population (WHO Mortality Databank 1991-1995) (Table

2.2). Deaths from breast cancer in Australia, in 1996, reported by Australia's Health

(2000) were higher than those in 1991 to 1995 as reported by WHO Mortality

Databank (1991-1995), with the rate of 24.9 per 100,000 population (Australia's

Health 2000) (Table 2.2). Incidence rates of breast cancer in Semarang, Indonesia

were low at 13.0 per 100,000 population in 1980-1981 (Parkin 1986), with no recent

data available.

Table 2.2. Deaths from breast cancer in Australia, 1990- 1996.

Year Deaths Population Rate ASR(W)

1990* 2,421 8,553,900 28.3 20.5

1991* 2,524 8,668,600 29.1 20.9

1992* 2,428 8,771,600 27.7 19.4

1993* 2,611 8,861,200 29.5 20.4

1994* 2,669 8,953,700 29.8 20.4

1995* 2,532 9,076,400 28.6 19.3

1996** 2,619 8,668,627 17.6 24.9

Rate: Crude rate per 100,000 ASR (W): Age Standardized Rate (World population) per 100,000 *WHO Databank (1990-1995) ** Australia's Health 2000 (2000)

Prostate cancer is one of the most common hormone-related cancers amongst men.

The lowest mortality rates from prostate cancer in the world (1995) were in Republic of Korea (1.8 per 100,000 population), and the highest were in Barbados (50.6 per 65

100,000 population) (WHO Mortality Databank 1995). Death rates from prostate

cancer in Australia were 18.5 per 100,000 population in 1995 (WHO Mortality

Databank 1995) and 33.0 per 100,000 population in 1996 (Australia's Health 2000).

There are no recent data for death rates from prostate cancer in Indonesia. Incidence

rates of prostate cancer occurring in 1980-1981 in Semarang, Indonesia were very

low at 1.5 per 100,000 population (Parkin 1986).

2. 7.1. Cancer and phytoestrogens

The cancer protective properties of phytoestrogens have been shown by their

activities to stimulate the synthesis of the binding globulin (SHBG) in

the liver and to reduce the biological effects of sex hormones. An increase of SHBG

results in the lowering of the percentage of free (%FT), the percentage

of free estradiol (%FE), and the albumin-bound fractions of the sex hormones. A low

level of the free fraction of sex hormones causes a reduction in their metabolic

clearance rate (MCR), thus lowering their biological activity, as a suppressant of

hormone-dependent cancer (Adlercreutz et al. 1990).

2.7.1.1. Ecological studies

A high intake and urinary excretion of phytoestrogens has been reported to be

associated with the reduction of breast cancer risk. Asian populations, display a lower incidence of, and death rates from, cancer (particularly breast and prostate cancers) than Western populations, and are reported to have higher levels of phytoestrogens in their diet (Rose et al. 1986). A study by Hirayama (1986) showed 66

that a reduced risk of breast cancer amongst Japanese women was associated with a

high intake of soya beans, especially soya bean-paste () soup. Women in

Shanghai have also been reported to have substantially lower incidence rates of

hormone-dependent cancers than their Caucasian counterparts (Parkin 1997),

possibly due to the consumption of high amounts of soya foods. The mean intake of

soya foods among healthy women in Shanghai was 89.1 g/capita/day, equivalent to

39.26 mg of total isoflavones per capita per day (Chen et al. 1999).

2.7.1.2. Case control studies

A case control study conducted by Adlercreutz et al. (1989) found that the total

urinary isoflavonoid excretion of premenopausal Finnish women with breast cancer

(279 nmol/24 h) was lower than omnivores (391 nmol/24 h) and vegetarians

(665 nmol/24 h). In the same study, Adlercreutz et al. (1989) found that the total

urinary isoflavonoid excretion of postmenopausal Finnish women with breast cancer

(94.2 nmol/24 h) was also lower than of omnivores (95.3 nmol/24 h) and vegetarians

(323 nmol/24 h). Lee et al. (1991) reported a significant inverse association between

breast cancer risk and soya bean intake of Singapore Chinese women. Women with

breast cancer had lower daily intakes of total soy products than controls. The daily

intake of total soy products of women with breast cancer ranged from 16.9-48 g,

compared with 20.3-55 g for controls (Lee et al. 1991).

A case control study of women living in Perth (Western Australian) conducted by

Ingram et al. (1997) found that urinary excretion rates of isoflavonoids, lignans and equol of women with breast cancer were lower than those of controls i.e. 782.9 vs 67

913.4 nmol/24 h for daidzein, 97.2 vs 108.6 nmol/24 h for equol, 282.0 vs 316.5

nmol/24 h for enterodiol, 1,973.4 vs 3,097.7 nmol/24 h for enterolactone, and 28.9 vs

29.3 nmol/24 h for matairesinol, respectively. A similar study conducted by Zheng et

al. (1999) in urban areas of Shanghai (1996-1997) showed that the urinary excretion

of total isoflavonoids of Chinese women with breast cancer was 71% lower than that

of controls. Strom et al. (1999) reported a protective effect on prostate cancer of

Caucasians related to a high amount of phytoestrogens in the diet. The case control

study conducted by Strom et al. (1999) found that the daily phytoestrogens intake of

prostate cancer risk vs control were 19.8 vs 29.7 J..Lg for genistein, 14.2 vs 22.8 fJg for

daidzein, 2.1 vs 2.0 J..Lg for formononetin, 57.1 vs 61.6 fJg for biochanin A and 44.9

vs 67.5 J..Lg for coumestrol.

2.8. Menopausal symptoms

Menopause is a permanent cessation of menstruation, resulting in the loss of ovarian

follicular activity. The ovaries produce the hormone estrogen. When the ovaries run

out of eggs, ovulation ceases, as does the production of estrogens. A decline in

estrogen production causes a number of physiological changes that result in numerous symptomatic and asymptomatic manifestations, including hot flushes, atrophic vaginitis, vasomotor symptoms, osteoporosis, heart disease, bladder and vaginal dryness, and cardiovascular disease. Since phytoestrogens possess an activity similar to estradiol, they may be able to replace the loss of hormones in menopausal women, thus reducing menopausal symptoms (Wilcox et al. 1990). 68

Hot flushes are the most common and distressing symptom of . Hot flushes are described as a sudden sensation of heat and redness over the head and upper body lasting seconds or several minutes accompanied by sweating, dizziness, palpitations and anxiety. Hot flushes occur when the vascular system changes as a result of estrogen deficiency (Thomas 1999). Hormone replacement therapy is commonly used to eliminate flushes. However, soy flour supplementation can also reduce flushes. A double blind and randomized controlled trial by Murkies et al.

(1995) found that the number of hot flushes in 58 postmenopausal women who experienced at least 14 hot flushes per week, decreased by 40% after taking soy flour

(45 g/day) over 12 weeks while controls only decreased by 25% with wheat flour ( 45 g/day). A study by Albertazzi et al. (1999) found that serum levels of genistin, daidzin and equol were significantly higher in postmenopausal women after taking a soya diet supplementation for three months in a double-masked, parallel, and placebo controlled trial. However, the increase in phytoestrogens levels in blood and serum in the soya groups did not correlate with fewer hot flushes.

2.9. Osteoporosis

Bone mass is influenced by a number of factors including genetics, nutrition, hormonal conditions, exercise and life style. Osteoporosis is defined as a condition in which the amount of bone per unit volume (bone mass) decreases, though the composition remains unchanged (Paoletti et al. 1999). Osteoporosis has been related to a decrease of gonadal steroid production. This causes an imbalance in bone formation and resorption. The bone becomes porous, causing structural failures and predisposition to fracture. Osteoporosis in women is associated with the menopause. 69

The lack of estrogens (estrogen deficiency) after menopause accelerates bone loss

(Bingham et al. 1998).

Phytoestrogens may be useful for preventing bone loss though most evidence currently available only derives from animal experiments. Blair et al. (1996) reported that genistein administered in the diet could increase the mass of weight-bearing in rats. Similarly, Ishimi et al. (1999) found that genistein regulated B­ lymphopoiesis and bone metabolism in female mice administered with genistein, preventing bone loss. Anderson et al. (1998) found that genistein has effects similar to Premarin, an estrogen with bone-retaining properties, in maintaining trabecular bone tissue in rats.

Coumestrol may be utilized as medication against osteoporosis, judging from results from experimental animal studies. Coumestrol inhibits bone resorption-stimulating agent-induced bone resorption, but did not inhibit basal bone resorption in rat limb bone cultures (Tsutsumi 1995). Coumestrol also stimulated bone-mineralising activity in chick embryonic femurs (less than 24 hours old) in organ cultures

(Tsutsumi 1995).

2.10. Coronary heart disease

Coronary heart disease is a multi-factorial disease. Increasing risk of this disease among postmenopausal women is associated with the decrease in estrogen levels

Paoletti et al. (1999). This causes a reduction in plasma concentrations of high density lipoprotein (HDL) cholesterol and an increase in plasma concentrations of 70

low density lipoprotein (LDL) cholesterol and triglyceride. Phytoestrogens may act

as estrogen antagonists by producing effects on lipoproteins similar to those caused

by estrogen. Both hormone replacement therapy and the anti-estrogen

decrease LDL-cholesterol levels and increase HDL levels. Antioxidant activity of

phytoestrogens may prevent oxidative damage of serum cholesterol. Hodgson et al.

(1996) indicated that oxidation of LDL was inhibited by the metabolic product of

daidzein ( equol and 0-desmethylangolensin) in an in vitro study.

Phytoestrogens may be able to reduce serum cholesterol and hyperlipidemia if

animal experiments are considered. Siddiqui and Siddiqui (1976) reported that

biochanin A and formononetin in chickpeas (Cicer arietinum) were responsible for

lowering lipid levels in plasma of albino rats. Sharma (1979) also found that

biochanin A and formononetin significantly reduced levels of serum cholesterol and

triglyceride in legume-treated male albino rats. However, neither compound reduced

lipid phosphorus levels.

2.11. Antithyroid effect of isoflavones

Soya beans and their products have been reported to induce goitre in humans and

animals. Goitre is caused by iodine deficiency. Inhibition of TPO-catalysed tyrosine iodination activity results in decreased levels of circulating thyroid hormones, leading to an increase in the secretion of TSH by the anterior pituitary, causing thyroid enlargement. A study by Divi et al. (1997) found that genistein and daidzein, isolated from soya beans, inhibited thyroid peroxidase {TPO) catalysed iodination of tyrosine reactions in the incubation of substrates of iodinated casein, or human goitre 71

thyroglobulin. The ICso values for inhibition of TPO-catalysed reactions by genistein

and daidzein were 1-10 J.lM.

2.12. Legumes

2.12.1. General

Food legumes, particularly grain legumes, or pulses, are important foodstuffs in all

tropical and subtropical countries. Legumes are consumed by human beings or

domestic animals usually as mature dry seeds, immature green seeds, or green pods

with the immature seed enclosed. Legumes have shown many physiologically

beneficial effects in humans and animals. Legumes are considered good sources of

protein (average 20-26%) and are second in importance only to cereals as a source of

vegetable protein. Legumes can be considered as a natural supplement to cereals,

since, although they are usually deficient in the essential amino acids methionine and

cystine, they contain adequate amounts of lysine, whereas cereals are deficient in

lysine, but contain adequate amounts of methionine and cystine.

The predominant fraction of the fatty acids in all legume species is unsaturated fatty

acids, with oleic acid the most predominant in peanuts. Beans of Phaseolus vulgaris

species contain largely linolenic acid. Beans supply protein, complex carbohydrates,

soluble fibre, and essential vitamins and minerals to the diet. Beans are also low in

fat and sodium, and are cholesterol-free. Recently, most beans and peas have been reported to contain at least one phytoestrogen compound (Franke et al. 1994; Mazur et al. 1998). 72

The dominant food legumes vary from country to country and from region to region.

There are approximately 27 food legume crops predominantly consumed throughout

the world. The major important legumes are peas (Pisum sativum ), haricot beans

(Phaseolus vulgaris), chick peas (Cicer arietinum), broad beans (Vicia faba),

cowpeas (Vigna unguiculata), mungbeans (Vigna radiata), soya beans (Glycine

max), groundnuts (Arachis hypogaea L.) and lentils (Lens culinaris).

Table 2.3. The production of beans, peas, and groundnuts in the world, Australia and Indonesia in 1998.*

Foods Production

(MT)

World Australia Indonesia

Beans (dry) 16,848,228 56,000 900,000

Beans (green) 43,480,048 36,000 144,633

Peas (dry) 12,230,087 298,000

Peas (green) 7,023,952 76,000

Chickpeas 8,779,075 199,840

Lentils 2,872,616 31,000

Broad beans 3,600,694 133,000 (dry) Broad beans 993,591 (green)

Groundnut 33,750,854 39,000 990,000 (shell)

-=no data;*= FAO (1998) 73

The production and consumption of a number of legwnes in the world, Australia and

Indonesia are shown in Table 2.3 (previous page) and Table 2.4. Only beans and groundnuts were reported by FAO (1998) for Indonesia. The world production of beans, peas, groundnuts (in shells) and soya beans in 1998 was estimated to be

16,835, 12,228, 33,745, and 159,743 thousand metric tonnes, respectively (FAO

1998). Peas were the legwnes with the highest production levels among other legwnes in Australia in 1998, and broad beans were the least. The production of beans in Australia in 1998 was only 6% of that in Indonesia, and the consumption of beans per year per capita in Australia was only half of that in Indonesia. The production of groundnuts (in shells) in Australia in 1998 was about 4% of that in

Indonesia while the consumption of groundnuts in Australia was 40% of that in

Indonesia.

Table 2.4. The consumption of beans, peas and groundnuts in the world, Australia and Indonesia in 1998.*

Foods Consumption

(Kg/year/caput)

World Australia Indonesia

Beans 2.3 1.4 3.8

Peas 0.6 2.6 0.0

Groundnuts 1.3 1.1 2.8 (shells)

* = FAO (1998). 74

Phytoestrogens contents in various peas, beans, and groundnut varieties from the

USA have been reported by Franke et al. (1994, 1995) and Mazur et al. (1998), but

there have been no reports from Indonesia or Australia.

2.12.2. Soya beans (Glycine max (L.) Merrill)

Soya beans (Glycine max (L.) Merrill) are classified in the family Leguminosae,

subfamily Papilionoideae, and genus Glycine, L. Soya beans originated in China.

They have the highest protein content (around 40%) and the second highest oil content (around 20%) amongst legumes. Soya beans contain many valuable components including phospholipids, vitamins, minerals, and minor substances, which are lmown to be biologically active, such as trypsin inhibitors, phytates and oligosaccharides, as well as isoflavones.

Soya beans are processed into various foods for human consumption. However, soya beans are also used for oil, defatted meals and animal feed. People from Southeast

Asia, especially Indonesia, have used soya beans as a staple in their diet for centuries. Soya bean is one of the most important dietary protein and energy sources in the Indonesian diet. Soya bean products such as tempeh (fermented soya bean) and tahu (soya bean curd) are a familiar part of the diet throughout Indonesia because they are very acceptable in price and taste, and are easy and fast to prepare.

The production of soya beans in Indonesia is approximately 1,500 thousand metric tonnes per year. Table 2.5 (overleaf) shows that the soya bean production in

Australia in 1997 and 1998 is about 5% of that in Indonesia (FAO 1997, 1998). Soya 75 bean production in Indonesia is high because of the strong demand for soya beans.

The average per capita per year consumption of soya beans in Indonesia in 1997 and

1998 was about 73- and 89- fold, respectively, of that in Australia (F AO 1997, 1998)

(Table 2.5). The average per capita daily consumption of soya beans and soya bean products in Indonesia reported by the FAO (1998) was lower than those reported by sources from Indonesia. Kompas Daily Newspaper (1998) reported that the soya bean consumption per capita was 13.4 kg/year or 257 g/week or 36.7 g/day.

Table 2.5. The production and consumption of soya beans in the world, Australia and Indonesia in 1997 and 1998. *

Year Production Consumption

(1000MT) (kg/year/caput)

World Australia Indonesia World Australia Indonesia

1997 74 1,357 0.1 8.9

1998 159,823 54 1,306 2.3 0.1 7.3

* = FAO (1997, 1998).

Table 2.6 (overleaf) shows a breakdown of the type of soy products consumed in

Indonesia, as reported by the Central Bureau of Statistics of Indonesia (1990). The consumption of soya beans is higher among urban people than those from rural areas.

The consumption of tempeh is highest among the whole group of soya beans and soya bean products. According to Winarno (1976), the consumption of tempeh in

Indonesia, particularly in Java, was approximately 30 to 120 g/day. Tempeh is generally served as a meat substitute in the grain-centered diet. The estimated annual nationwide production of tempeh in Indonesia is 75,600 tonnes. Therefore, 76 approximately 14% of the annual Indonesian soya bean production is made into tempeh.

Table 2.6. Average per capita daily consumption of soya beans and soya bean products in Indonesia (g).

Items of foods 1990a 1991-1998b

Urban Rural Urban +rural Urban+ rural

Soya beans 0.1 0.4 0.3 ND Tofu 17.9 9.7 12.1 ND Tempeh (fermented 16.3 11.1 12.7 ND soya bean) Taucho (fermented 0.3 0.1 0.1 ND soy paste) 0.7 0.6 0.6 ND Oncom (fermented residues soya bean) 21.9 25.8 36.7 Total soya beans 35.3 and soya bean products

ND:Nodata a. Central Bureau of Statistics, Jakarta, Indonesia, ISBN 979-462 b. Kompas Daily Newspaper (1998)

The production of soya beans in Australia is different. Soya beans in Australia are mainly used to produce oilseed. There is little evidence of soya bean consumption in

Australia. The average daily consumption of soya beans in Australia reported by

Australian Bureau of Statistics (1995) (Table 2.7, overleaf) was very low compared to Indonesia. The consumption of soya beans in Australia is included in the group of legumes, pulses, and dishes. Included in these groups are also pappadam, baked beans, and vegetarian sausages. 77

Table 2. 7. Mean daily legume and pulse products consumption in Australia (g/person)*

Gender Age group (years)

2 - 3 4- 7 8- 11 12- 15 16- 18 19-24 25-44 45 - 64

Males 7.1 8.9 5.3 13.6 16.2 12.0 11.2 15.2

Females 6.7 5.6 2.8 6.7 9.0 9.1 8.4 8.0

* =Australian Bureau of Statistics (1995).

2.13. Cashew (Anacardium occidental L.)

Cashew is classified in the family Anacardiaceous, subfamily Papilionoideae, and genus Anacardium. Cashew originated in Brazil. Cashew is one of the most versatile tree crops and is considered a horticultural crop of commercial importance. The kernel is rich in protein, carbohydrate, unsaturated fats, minerals like calcium phosphorous and iron, and vitamins. Cashew proteins are complete with all essential amino acids. Hence, it can be considered equal to peanuts and soya beans for protein and to meat, milk and eggs for protein substances. The kernel supplies 6,000 calories of energy per kg as against 3,600 from cereal (Bose 1985). The world production of cashew kernel in 1997 and 1998 was 1,065,086 and 1,199,395 MT, respectively

(FAO 1998, 1999). The production of cashew kernels in Indonesia in 1998 and 1999 was 69,027 MT, but there were no data reported for Australia. There has been no study of the levels of phytoestrogens content in cashew kernels so far. Analysis of 78 the levels of phytoestrogens in cashew kernels is essential, because the cashew kernel is an important horticultural crop. 79 CHAPTER3.

MATERIALS AND METHODS

3.1. Apparatus

The liquid chromatography apparatus from Waters Australia (Rydalmere, NSW,

Australia) consisted of an automatic sample injector model 717 plus; a pump controller model 600; and a photodiode-array detector model 996. Columns used were: Alltima cyano 100A (4.6 x 150 mm I.D.; 5 mm) column from Alltech

(Baulkham Hills, CA, USA); symmetry (C8) (3.9 x 150 mm); C18 Nova Pak (3.9 x

150 mm I.D.; 4 mm) and phenyl Nova-Pak (3.9 x 150 mm I.D.; 4 mm) reversed­ phase columns from Waters (Milford, MA, USA). The grain mill (Fritsch

Mill/Pulverisette) was from Fritsch Giv.IBH (ldar, Oberstein, Germany); centrifuge model Hettich EBA 12 from Scientific (Tulingen, Germany); Eppendorf tube from

Bonnet (Taren Point, NSW, Australia). Other apparatus was: PTFE micro filter

(Dublin, CA, USA); electric balance from Sartorius (Germany); freeze-dryer model

Lyovac GT2 from Leybold-Heraeus GMBH (Koln, Germany); coffee grinder from

Moulinex (Ireland); Waring blender from Waring Products Division (New Hartford,

CONN, USA); and, pH meter model Cyberscan 500 from Eutech (Singapore).

LC-tandem mass spectrometry (MS-MS) analysis was performed on an API 300 triple-quadrupole mass spectrometer from Perkin Elmer SCIEX (Thornhill, ON,

Canada) equipped with an Apple Macintosh System 8.0 computer for data analysis. 80

3.2. Materials

3.2.1. Reagents

Genistein, biochanin A, flavone (internal standard), and butylated hydroxytoluene

(BHT) were purchased from Sigma Chemicals (StLouis, MO, USA); daidzein from

ICN (Aurora, Ohio, USA); formononetin and coumestrol from Indofine Chemicals

(Belle Mead, NJ, USA); dimethylsulfoxide (DMSO) from Merck (Kilsyth, VIC,

Australia). Acetonitrile (HPLC grade), methanol and isopropanol were purchased from Malinckrodt (Clayton, VIC, Australia); ethanol from Chem-supply (Gillman,

SA, Australia). Hydrochloric acid, sodium hydroxide, acetic acid, and sodium acetate were obtained from Ajax Chemicals (Auburn, NSW, Australia); Millipore water was used for all mobile phases.

3.2.2. Foods used for method development

Bowyer cultivar of soya beans (Glycine max) were purchased from Allgold Foods

(Leeton, NSW, Australia); dried soya beans from M:Kenzie's (Altone, VIC,

Australia), canned soya beans from Master Foods (Wyong, NSW, Australia), soya milk from Vitasoy International (Tuen Mun, NT, Hongkong), Hard Tofu from Soya King (Campsie, NSW, Australia), unbranded fresh green beans

(Phaseolus vulgaris L.) and unbranded raw snow peas (Pisum sativum L.) were purchased from local supermarkets in Sydney, Australia. 81

3.3. Methods

3.3.1. Analytical method development

Two analytical methods were developed, one for the separation and the

quantification of the three isoflavones (daidzein, genistein, biochanin A) in soya beans (see 3.3.1.1), and the other for the separation and quantification ofisoflavones

(daidzein, genistein, formononetin, biochanin A) and coumestrol in soya beans (see

3.3.1.2, page 84).

3.3.1.1. Separation and quantification of three isoflavones (daidzein,

genistein, and biochanin A) in soya beans

The separation of the isoflavones ( daidzein, genistein, and biochanin A) in soya beans was optimised in two stages; optimisation of solvent strength, and optimisation of pH control. A mixture of standard compounds ( daidzein, genistein, and biochanin A) was used for the trials.

3.3.1.1.1. Purity of standard

Daidzein, genistein, biochanin A, and flavone (internal standard) were used as reference materials. Approximately 10 mg of each compound was dissolved separately in 20 J.d DMSO and diluted to 1 L with 96% ethanol. Thirty J.!l solutions were individually injected into the LC and monitored for their maximum absorption 82

rates. Purity (%) was calculated by dividing the peak area of the compound by all

peak areas in the chromatogram and was multiplied by 100. Compounds of> 95%

purity were used for all experiments.

3.3.1.1.2. Preparation of stock standard solutions

Duplicate stock standard solutions were prepared according to Franke et al. (1994).

Approximately 10 mg each of daidzein, genistein, biochanin A, and flavone were

accurately weighed and dissolved in 20 Jll DMSO, diluted to 1 L with 96% ethanol

and stored at -20°C.

3.3.1.1.3. Chromatographic conditions

Thirty Jll of stock standard solution was injected into the LC under isocratic elution

conditions with 1% aqueous acetic acid-acetonitrile (33:67, v/v) at a flow rate of0.80 ml/min. Analytes were monitored at wavelengths ranging from 200 to 400 nm.

3.3.1.1.4. Calibration curves

Seven different concentrations, ranging from 11 J.I.M - 800 J.I.M, were prepared by diluting each stock standard solution with a mobile phase. Calibration curves were obtained by plotting standard concentrations against peak areas obtained from LC analysis using 30 Jll injections, and repeated six times. 83

3.3.1.1.5. Extraction and hydrolysis of foods

Dried soya beans were ground to a fine powder and extracted using the method

described by Franke et al. (1994) with slight modifications. Samples were prepared

in triplicate. One gram of finely ground soya beans was placed in a 50 m1 round­

bottomed flask, 10 m1 of2 M HCl and 40 m1 of96% ethanol containing 0.05% BHT

and 20 ppm flavone were added. The mixture was then stirred and ultra-sonicated for

20 minutes, then refluxed at 80°C for one hour. The extract was cooled to room

temperature, transferred into a 50 m1 volumetric flask, and diluted to volume with

96% ethanol. An amount of 1.2 m1 of this mixture was placed in a 1.5 m1 Eppendorf

tube and centrifuged at 800 g for 20 minutes. The clear supernatant was filtered

through a PTFE micro filter and kept in a vial. The sample was stored at -20°C until

LC analysis was carried out.

3.3.1.1.6. Recoveries of standards

In duplicate, three different concentrations of stock standard solution were each ' added to one gram of finely-ground soya beans in 50 m1 round bottom flasks. They were then extracted and hydrolysed under conditions established for food extraction and hydrolysis. Thirty ~-tl of a spiked sample was injected into the LC and the recoveries of the analytes were determined using the method developed. 84

3.3.1.2. Separation and quantification of isoflavones ( daidzein, genistein,

formononetin and biochanin A) and coumestans ( coumestrol) in soya

beans

3.3.1.2.1. Purity of standards and preparation of standard solutions for

calibration

Preparation of stock standard solutions and checking the purity of standard daidzein,

genistein, formononetin, biochanin A and coumestrol used procedures as described in 3.3.1.1 (page 81). Approximately 11 to 16 mg of the five standards was used to make up duplicate stock standard solutions. Each stock standard solution was then diluted with a mobile phase to nine different concentrations ranging from 0.02 mg!L to 16 mg!L. Five J.ll of each concentration was injected into the LC system using the method developed.

3.3.1.2.2. Optimisation of stationary phases and mobile phases

Four stationary phases including Alltima Cyano 100A, symmetry (C8), C18 Nova

Pak, and phenyl Nova-Pak reversed-phase and seven solvents including acetonitrile,

1% and 10% aqueous acetic acid, isopropanol, methanol, ethanol, acetate buffer (pH

2.6) and water were evaluated to optimise the separation of a mixture of daidzein, genistein, formononetin, biochanin A and coumestrol standards.

The solvents evaluated were: acetonitrile-1% aqueous acetic acid (20:80, 25:75,

30:70, 33:37, and 35:65, v/v); acetonitrile-water (20:80, 25:75, 30:70, 35:65, and 85

40:60, v/v); acetonitrile-10% aqueous acetic acid (20:80, 25:75, 30:67, 33:67, and

35:65, v/v); acetonitrile-acetate buffer (pH 2.6) (33:67, v/v); acetonitrile-methanol

(30:70, 40:60, and 65:35, v/v); methanol-water (35:65 and 40:60, v/v); methanol-1% aqueous acetic acid (30:70, 35:65, and 40:60, v/v); acetonitrile-1% aqueous acetic acid-isopropanol (15:75:10, 20:70:10, and 25:65:10, v/v/v); 1% aqueous acetic acid­ water-methanol (70:10:20, v/v/v); 1% aqueous acetic acid-water (70:30, v/v); acetonitrile-1% aqueous acetic acid-methanol (20:70: 10, 22:70:8, and 10:70:20, v/v/v); acetonitrile-acetate buffer (pH 2.6) (33:67, v/v); acetonitrile-1% aqueous acetic acid-water (20:70: 10, v/v/v).

Five J.Ll of stock solution was injected into the LC with an isocratic elution system at a flow rate of 0.8 ml/min. Analytes were monitored using the dual channel diode array detector at wavelengths ranging from 200 to 400 nm.

To confirm the identity of the peaks obtained by LC, a mixture of five standards was analysed by the LC/MS/MS. Five J.Ll of stock solution was injected into the

LC/MS/MS with a phenyl column and acetonitrile-water (33:67, v/v) as eluent at a flow rate of 0.8 ml/min. Positive ions from eluted solutes were introduced into the mass spectrometer following their generation by atmospheric pressure chemical ionization (APCI) caused by a corona discharge needle in the heated nebulizer interface of this instrument. Selected peaks were then extracted using positive acquisition mode with m/z ranging from 150 to 600. 86

3.3.1.2.3. Precision

To determine the precision of the LC method, a mixture of standards was injected

into the LC system on three occasions, using six injections on each occasion.

Repeatability and reproducibility of the methods were determined by calculating the

coefficient of variation (CV) of retention times and of the quantitative analyses of

individual compounds.

3.3.1.2.4. Detection limit

The minimum detectable quantity (MDQ) or detection limit is the minimum quantity

of sample for which the detector will give a visible response. The detection limit of

the method used in this study was measured three times using a noise signal

(Robards et al. 1994). Noise is the random perturbation in signal produced by a

detector in the absence of any sample.

3.3.1.2.5. Capacity factor (k)

Capacity factor (k') was used to characterize the retention of a compound. The capacity factor was calculated by the ratio of the adjusted retention time (t' r) with the dead time (tm)· The adjusted retention time is the difference between the retention times of analytes (tr) and the dead time. Low k' value indicates that a compound elutes near to the dead time (tm)· As a consequence, the separation is insufficient. In contrast, high k' value results in a good separation. However, a long analysis time 87 with corresponding peak-broadening could lead to low detectability. The k' value is optimized by the selection of suitable solvents and should lie between 1 and 10.

3.3.2. Isolation of daidzein, genistein, formononetin, biochanin A and

coumestrol from foods

The method for extraction of foods was optimised by varying the concentrations of phosphoric acid or hydrochloric acid, refluxing times and temperatures. Dried soya beans were used for the evaluation of the effects of concentration of phosphoric acid and hydrochloric acid in solid foods. Soya milk was used for the evaluation of the effects of concentration of hydrochloric acid in liquid foods. Optimisation of the extraction was also evaluated in processed foods, including canned soya beans, cooked soya beans and tofu.

3.3.2.1. Food preparation

Dried soya beans (~Kenzie's, Australia) from each of three different purchases were ground to fine powder, pooled, and then mixed together before analysis.

Three packs of dried soya beans from ~Kenzie's (Australia) were purchased, and

100 g of each were cooked in boiling water for two hours, drained, blended, frozen and freeze-dried. The water loss during freeze-drying was recorded. Dried samples were then ground. Finely ground samples from three purchases were pooled and mixed together before analysis. 88

Water was drained from three purchases of canned soya beans from Master Foods

(Australia). Approximately 100 g of each was then homogenized, frozen and then

freeze-dried. Foods were then prepared using procedures as described above for

boiled dried soya beans.

The packing water was drained from three purchases of Hard Tofu from Soya King.

Further preparation used the procedure, as described above, for boiled dried soya

beans.

3.3.2.2. Analytical methods

3.3.2.2.1. Moisture

Approximately 10 mg of the sample was dried a overnight in a vacuum oven at 70°C to a constant weight. The moisture content was determined by the weight differential before and after drying (James 1995).

3.3.2.2.2. Extraction solvents - phosphoric acid

The effects of these three different concentrations of phosphoric acid (0.1, 1 and 2M) on the levels of isoflavones and coumestrol extracted from dried soya beans were evaluated. One gram of finely ground soya beans was placed into each of three 50 m1 round-bottom flasks and diluted with 40 m1 of 96% ethanol. Ten ml of 0.1, 1 or 2M phosphoric acid were added into samples and then sonicated for 20 minutes. The mixture was refluxed by heating on a water bath at 80°C for 2 hours. The extract was 89

then adjusted to 50 m1 with 96% ethanol. 1.2 m1 of the mixture was placed in a 1.5

m1 Eppendorf tube and centrifuged at 800 g for 20 minutes. The clear supernatant

was filtered through a PTFE micro filter prior to LC analysis. Twenty J.Ll of sample

was injected into the LC and separated using a mixture of acetonitrile-water (33:67,

v/v) at a flow rate of 0.8 ml/min. The analytes were scanned at wavelengths ranging

from 200 to 400 nm. The peak area given by the response of the diode array detector

was then calculated using the standard curves to estimate the levels of isoflavones

and coumestrol. The effects of phosphoric acid on the levels of isoflavones and

coumestrol were then analysed using a one way analysis of variance (ANOVA) and

Tukey's test for determining the differences amongst means, using the Excel 95

computer software program (Snedecor et al. 1989).

3.3.2.2.3. Extraction solvents - hydrochloric acid

The effects of hydrochloric acid concentration and reflux time on the levels of

isoflavones and coumestrol extracted from soya beans were evaluated. The procedure

for extraction used the method as described for extraction solvent - phosphoric acid

(see 3.3.2.2.2, previous page). The sample was extracted with and without hydrochloric acid, and refluxed for 0, 1, 2, 3 or 6 hours. Extraction and hydrolysis procedures for dried soya beans are given in Figure 3.1 (page 91).

The extraction procedure for soymilk was similar to that for dried soya beans.

However, the sample volume used was 25 m1 with six different concentrations of hydrochloric acid added for four different periods of hydrolysis. The extraction procedure for soymilk is shown in Figure 3.2 (page 92). 90

3.3.2.2.4. Extraction temperature

The effects of temperature on the levels of daidzein, genistein, formononetin, biochanin A and coumestrol in dried soya beans and soya milk were evaluated. The extraction and identification procedures used the method as described for the extraction of solvent- hydrochloric acid (see 3.3.2.2.3, previous page). Two different temperatures were used, 80° C and 100°C (Figure 3.3, page 93). The extraction procedure for soya milk was similar to that for dried soya beans, though on this occasion 25 ml of soya milk were used (Figure 3.2, page 92). 91

Dried soya beans 1 gsample

Add 10 ml ofHCI (2M) I I No HCI ~/ I Add EtOH (96%) I ~~ No reflux I I Reflux (80°C)

I

Volume adjustment to 50 ml with EtOH (96%)

HPLC, 0.8 mVmin, 5 1-11 volume, phenyl column, photo diode array detector

Figure 3.1. Phytoestrogen extraction procedures from dried soya beans. 92

25 m1 soya milk

AddHC1(2M)

Add EtOH (96%)

Volume adjustment to 50 ml with EtOH

Centrifuge at 800 g for 20minutes.

Filtration with 0.4 !lm PTFE

HPLC, 0.8 ml/min, 5 Ill volume, phenyl column, photo diode array detector

Figure 3.2. Phytoestrogen extraction procedures from soya milk. 93

1 g sample

Add 10 ml HCI (2M)

Add EtOH (96%)

Refluxing for6hours

Volume adjustment to 50 ml with EtOH (96%)

Centrifuge for 20 minutes at800 g

Filter with 0.4 J.lffi PTFE

HPLC, 0.8 ml/min, 5 J.ll volume, Phenyl column, photo diode array detector

Figure 3.3. Phytoestrogen extraction procedures from dried soya beans at 80°C and 100°C. 94

3.3.2.2.5. Phytoestrogen extraction from canned and cooked soya beans, and

tofu

Experiments were carried out by extracting phytoestrogens from the samples, with or

without hydrolysis, and with or without refluxing, for four different periods.

Extraction and identification procedures for cooked and canned soya beans, and tofu

were similar to those for raw soya beans (Figure 3.1, page 91).

3.4. Stabilisation of daidzein, genistein, formononetin,

biochanin A and coumestrol

The effects of storage at different pH on the stability of isoflavones and coumestrol

in dried soya beans and on a mixture of standard compounds were evaluated.

One gram of dried soya bean powder was placed into a 50 ml round-bottom flask and then extracted with 40 ml of 96% ethanol and 10 ml of 2 M hydrochloric acid. The extract was then sonicated for 20 minutes and refluxed by heating on a water bath at

l00°C for 6 hours. The extract solution was then adjusted to 50 ml with 96% ethanol.

The pH of the extracted sample was adjusted to 1-7 by the addition of sodium hydroxide solution into the extracted soya beans. An amount of 1.2 ml of the mixture was placed in a 1.5 ml Eppendorf tube and was then centrifuged at 800 g for 20 minutes. The clear supernatant was filtered through a PTFE micro filter and stored in the freezer at -20°C for up to 28 days prior to LC analysis. The recoveries of the four 95

isoflavones and coumestrol were determined by injecting 5 J..tl of solution into the LC

and analysing using the LC method. The procedure is given in Figure 3.4 (overleaf).

The stability of the four isoflavones and coumestrol during storage in alkali or acid was studied using the standard compounds. These experiments were for comparison purposes only. Sodium hydroxide or hydrochloric acid was added into a mixture of

standard compounds (pH 7). The mixture was then centrifuged at 800 g for 20 minutes and was stored in the freezer at -20°C for up to 28 days. The recoveries of the four isoflavones and coumestrol were determined by injecting 5 J..tl of solution into the LC and analysing using the LC method. The procedures are shown in

Figures 3.5 and 3.6 (pages 97 & 98). 96

I Extracted sample I .------. r------.1NoNaOH . I Add NaOH (0.0954 M) I * ~

Stir 1 minute

Centrifuge at 800 g 20minutes

Filter 0.4 J.tm PTFE

Store at -20°C

~ I 2 days I I 4 days I 14 days I 118 days I I 21 days I I 28 days

Identification in HPLC for 30 min 5 J.tl, flow rate 0.8 ml/min 33% ACN in water

Figure 3.4. The schema of the addition of sodium hydroxide to the extracted samples for storage trials. 97

1 m1 standard solution

Centrifuge at 800 g 20minutes

1 day I I 2 days I I 4days I I 14 days I I 18 days I I 21 days I I 28 days t Identification in HPLC for 30 min 5 p.l, flow rate 0.8 ml/min 33% ACN in water

Figure 3.5. The schema of the addition of sodium hydroxide in the standard solutions for storage trials. 98 1 ml standard solution

105 J.tl 1 110.5 J.tl 1 11.05 J.tl 1 1 o.ws J.tl 1 ~~ ~~--S-tir_1_m_in_u-te___,l

Centrifuge at 800 g 20minutes

Filter 0.4 J.tm PTFE

Store at -20°C

2days 28 days

Identification in HPLC for 30 min 5 J.tl, flow rate 0.8 ml/min 33% ACN in water

Figure 3.6. The schema of the addition of hydrochloric acid in the standard solutions for storage trials. 99

3.5. Recovery of the isoflavones and coumestrol

To measure the efficiency of the LC method and extraction procedures, the

recoveries of daidzein, genistein, formononetin, biochanin A and coumestrol were

studied in dried soya beans, fresh whole green beans (Phaseolus vulgaris L.) and

fresh whole snow peas (Pisum sativum L ).

3.5.1. Food preparation

Three purchases of dried soya bean were each ground down to a fine powder, placed together and mixed thoroughly before analysis. Three different purchases of beans and snow peas were prepared as described in section 3.3.2.1 (page 87). The moisture content of samples was measured using the method as described in section 3.3.2.2.1

(page 88).

3.5.2. Analytical methods

3.5.2.1. Dried soya beans

To determine the recovery of isoflavones and coumestrol, three different concentrations of stock standard solution (0.40-3.6 mg/L) were each added to one gram of finely ground soya beans and then diluted with 40 ml of 96% ethanol. The mixture was extracted with 2 M HCl, stirred and ultra-sonicated for 20 minutes, followed by refluxing at 100°C for six hours. The extract was cooled to room temperature and adjusted to 50 ml with 96% of ethanol and to pH 4 with the addition 100 of sodium hydroxide. At that stage, 1.2 m1 of the mixture was placed in a 1.5 m1

Eppendorf tube and was centrifuged at 800 g for 20 minutes. The clear supernatant was filtered through a PTFE micro filter and kept in a LC vial. The sample was stored at -20°C until LC analysis was carried out. The recoveries of the four isoflavones and coumestrol were determined by injecting 20 Jll of solution into the

LC and analysing using the LC method.

3.5.2.2. Fresh green beans and fresh snow peas

The extraction and identification procedures were used as described above for dried soya beans (section 3.5.2.1, previous page).

3.6. Qualitative and quantitative analysis of phytoestrogens in

food

3.6.1. Sampling

Legumes and products commercially available in Australia and Indonesia were collected for analysis. Soya beans and soya bean products were purchased from retail

Asian retail shops and supermarkets in Sydney, Australia. Other legumes and products were mostly purchased from Coles and Franklin supermarkets. Each food was collected on three occasions from suburban supermarkets in the centre, east and south of Sydney, Australia. 101

In Indonesia, unbranded legume products were purchased mostly from street vendors

and open markets. Branded legume products were purchased from supermarkets.

Each food was collected from three different areas in Indonesia, Lampung (Sumatra),

Bogor (West Java) and Jakarta (the capital city).

3.6.2. Food varieties

Food varieties collected from Australia consisted of 12 different species of dried legumes; 14 different canned legume products from six different manufacturers; 11 different brands of soya milk; five different brands of tofu; four different second­ generation soya products from different brands; seven different brands of cereal; five different brands of bread; three different brands of freeze-dried products; nine different unbranded fresh products; two different brands of frozen peas. Details of foods and photographs are shown in Appendices 2 and 4.

Food varieties collected from Indonesia consisted of three different brands of soya milk; four different brands of tofu, two packages of unbranded tofu; two different brands of dried soya milk curd; five different brands of fermented soya paste and one unbranded fermented soya paste; six different brands of soya sauce; one unbranded oncom; six unbranded fresh pea and bean products, six canned pea and bean products from four different manufacturers; one branded frozen peas; unbranded petai and unbranded petai cina. Details of foods and photographs are shown in Appendices 3 and4. 102

Botanical name and taxonomy of foods were identified using Handbook of World

Food Legumes (Salunkhe & Kadam 1989) and Plant Resources of South-East Asia 1

(Vander Maesen & Somaatmadja 1989).

3.6.3. Sample handling and storage

Foods were analysed as raw, processed and cooked (according to package directions, where available). Preparation of foods purchased in Australia was carried out at the

Department of Food Science and Technology at the University ofNew South Wales,

Sydney, Australia. Preparation of foods purchased in Indonesia was carried out at

Bogor Agricultural University, Bogor, Indonesia.

Fresh samples from street vendors, local markets and supermarkets in Indonesia were placed in a sterile plastic pouches, kept in a cool box and then transported to Bogor

Agricultural University, Bogor, Indonesia. At the University, some foods were separated for raw and cooked analysis. Foods for raw analysis were then homogenised, frozen and freeze-dried. Water loss during freeze-drying was calculated. See section 3.6.4.4 (page 104) for preparation of cooked foods.

All foods were placed in vacuum-sealed plastic bags and stored in the freezer at

-20°C while awaiting transport to Australia. Foods in the vacuum-sealed plastic bags were transported by air to the Department of Food Science and Technology at the

University of New South Wales, Sydney, Australia for analysis. 103

Foods including freeze-dried mungbean sprouts, yard long beans, red kidney beans, lima beans, petai, petai cina, soya beans, dried ground nuts, dried red kidney beans, and fresh beans were quarantined in the School of Geography at the University of

New South Wales, Sydney, Australia. Foods were labelled with the name, weight, and date of arrival, and stored in dry places at room temperature until analysis. Foods were weighed in the quarantine facility and then taken to the Department of Food

Science and Technology for further analysis.

3.6.4. Food preparation

3.6.4.1. Fresh foods

Fresh beans and peas of many varieties, yard long beans, tofu, tempeh were analysed raw, processed and cooked. Alfalfa and bean sprouts were analysed raw.

All raw fresh foods were homogenized and frozen in the freezer at -20°C for 24 hours. Frozen foods were then freeze-dried for two to three days until completely dry. Dried foods were then placed in plastic bags and heat-sealed. Samples were then stored in the freezer at -20°C whilst awaiting analysis. Water loss after freeze-drying was calculated by weighing the fresh product and then weighing the dried product.

The water loss was determined by:

Water loss(%)= [Weight of fresh product]- [Weight of dried product] x 100

[Weight of fresh product] 104

3.6.4.2. Dried products

Dried products including soya beans, chick peas, green split peas, peas, yellow split peas, red kidney beans, haricot beans, black eye beans, cannellini beans, berlotti beans, red lentils, whole green lentils and yard long bean seeds were separately ground to a fine powder prior to analysis.

3.6.4.3. Canned and bottled foods

Canned foods including soya beans, red kidney beans, beans, and peas were drained, homogenized, frozen and freeze-dried. Bottled foods and fermented soya paste were homogenized, frozen, and freeze-dried.

3.6.4.4. Boiling

Dried legumes and raw fresh legumes were processed as they are commonly consumed. Three different purchases of dried legumes including soya beans, chick peas, green split peas, peas, yellow split peas, red kidney beans, haricot beans, black eye beans, cannellini beans, berlotti beans, red lentils, whole green lentils and yard long bean seeds were separately cooked in boiling water (100 g samples in 500 m1 water) for 2 hours, drained, homogenised, frozen and then freeze-dried. The water loss was recorded. Freeze-dried samples were placed in heat-sealed plastic bags and stored in a freezer at -20°C whilst awaiting analysis. 105

Three different purchases of fresh soya beans and a variety of fresh beans and peas

were separately cooked in boiling water. They were considered cooked when they became soft. Boiling time for soya beans, whole honey snap peas, peas, snow peas, whole fresh yard long beans, and green beans was 10 minutes; for fresh berlotti seeds was 20 minutes; for baby beans was 15 minutes; and for whole fresh continental beans, mungbean sprouts, petai and petai cina was 5 minutes. Cooked foods were drained, homogenised, frozen, and then freeze-dried. The water loss after freeze­ drying was calculated. Dried food was then placed in a plastic bag, heat-sealed and stored in a freezer at -20°C whilst awaiting analysis. Food products purchased in fudonesia were placed in vacuum-sealed plastic bags and stored in the freezer at

-20°C whilst awaiting transport to Australia.

3.6.4.5. Frying

Three different purchases of tempeh, tofu, oncom kedelai, and oncom kacang were separately sliced into approximately 1 em cubes. The slices were then deep-fried with Bimoli palm oil (fudonesia) for three to four minutes until brown. The slices were placed on paper towel to absorb excess oil. Fried products were homogenised, frozen, and then freeze-dried. The freeze-dried foods were placed in vacuum-sealed plastic bags and stored in the freezer at -20°C whilst awaiting transport to Australia.

Three different purchases of groundnuts and cashew kernels were separately deep­ fried in Bimoli palm oil (fudonesia). Groundnuts and cashew kernels were fried for approximately 5 minutes until brown and were then placed on paper towel to absorb excess oil. They were placed in plastic bags and vacuum-sealed. 106

3.6.4.6. Milling and grinding

All freeze-dried foods were ground with a coffee grinder to a fine powder and fresh

foods were blended prior to extraction and analysis.

3.6.5. Analytical methods

3.6.5.1. Moisture

Moisture content of raw products, dried seeds, and processed products was analysed

using the procedures as described in section 3.6.4.1 (page 103).

3.6.5.2. Extraction

All dried legumes and selected fresh legumes were analysed raw. Fresh, wet and

cooked products were analysed after freeze-drying except soya milk and soya sauce.

3.6.5.2.1. Dried products

One gram of finely ground foods was placed in a 50 m1 round bottomed flask and diluted with 40 m1 of 96% ethanol. Ten m1 of 2M HCl was added to the mixture and this was sonicated for 20 minutes. The mixture was then hydrolysed by refluxing and heating in a water bath at 100°C for 6 hours. The extract was made up to 50 m1 with

96% ethanol and adjusted to pH 4 with sodium hydroxide, then centrifuged at 800 g 107

for 20 minutes. The clear supernatant was passed through a 0.20 JJm PTFE micro

filter before LC analysis.

3.6.5.2.2. Fresh products

Ten grams of the homogenized fresh products were extracted using the procedure

described in section 3.6.5.2.1 (page 106).

3.6.5.2.3. Canned products

One gram of the finely ground foods was analysed using the procedure described in

section 3.6.5.2.1 (page 106). A refluxing time of two hours was used for these

samples.

3.6.5.2.4. Liquid foods

Twenty five m1 of soya milk or soya sauce was analysed using the procedure

described in section 3.6.5.2.1 (page 106).

3.6.5.3. Identification

Twenty JJl of sample solution was injected into the LC and separated using a phenyl column with acetonitrile-water (33:67, v/v) as eluent at the flow rate of 0.8 ml/min.

All chromatographic procedures were performed at 25°C. The analytes were then monitored with a photodiode-array detector at wavelengths ranging from 200 to 600 108

nm and the UV absorption spectra of the individual compounds were recorded.

Levels of isoflavones and coumestrol were estimated from standard curves.

3.6.6. Laboratory quality control

Quality control measures were carried out prior to and during the analysis of foods.

Quality control measures were:

• Analysis of the extracted soya beans and standards on the day they were

5 extracted and after the 2nd, 4th, 14th, 18th, 21 \ and 28th day of storage. This was to

confirm that the LC system operated correctly and consistently and to evaluate

the effects of storage on the levels of isoflavones and coumestrol.

• A mixture of standard compounds was analysed six times within and between

assays. This was to evaluate precision and repeatability of the LC system.

• A mixture of standard compounds was analysed with each batch of food samples.

• The LC system and extraction procedures were evaluated by determining the

recoveries of analytes in soya beans, beans and peas.

• Effects of freeze-drying on the levels of isoflavones and coumestrol were

evaluated in fresh whole beans from Australia. 109

The data are presented on a wet weight basis as mean± SD. Differences between means were analysed using analysis of variance (ANOVA) based on dry weight.

Tukey's test was used for determining differences between means (Snedecor et al.

1989). 110

CHAPTER4.

RESULTS AND DISCUSSION

METHOD DEVELOPMENT

4.1. Separation and quantification of three isoflavones

(daidzein, genistein and biochanin A) in soya beans1

4.1.1. Separation of a mixture of standard compounds

A mixture of the three standard compounds daidzein, genistein and biochanin A,

could be separated using the isocratic elution system with a mixture of acetonitrile

(30-40%) in 1% aqueous acetic acid (v/v). Solvent strength was selected by reducing the amount of acetic acid in water from 10% as used by Franke et al. (1994, 1995) to

1% because of possible adverse effects of acid on the column.

The mobile phase for previous gradient methods (Franke et al. 1994, 1995) was acetonitrile (A) and 10% aqueous acetic acid (B). Their elution system was gradient with 23 % A in B (v/v) linearly to 70% A in B in 8 minutes followed by holding at

23% A in B for 12 minutes. This change to 23% from 70% acetonitrile occurred before all the peaks had eluted. This is not a sensible way to apply gradients, and the

These results were published by L.S. Hutabarat, M. Mulholland, H. Greenfield in Journal of Chromatography A, 795 (1998) 377-382, "Development and validation of an isocratic high­ performance liquid chromatographic method for quantitative determination of phytoestrogens in soya bean". 111 column is adversely affected by such a dramatic change to concentration of the mobile phase and the associated pressure changes. It seems likely that this may have been a typographical error in the original paper and the intention was to retain acetonitrile at 70% rather than 23% for 20 minutes. However, all subsequent papers using this method also cited these unsuitable conditions.

In the present study, the lower aqueous acetic acid concentration of up to 1% did not cause a noticeable deterioration in method performance. Mobile phase pH showed an effect on the retention time of the analytes. A decrease in mobile phase pH decreased the retention time. The use of I% aqueous acetic acid (pH 2. 7) as eluent led to a slightly different resolution of the isoflavones compared to 10% aqueous acetic acid

(pH 2.2) as shown in Figures 4.1 and 4.2 (pages 112 & 113).

4.1.1.1. UV absorbance of phytoestrogens

The maximum absorbance of each compound was obtained from the maximum response of each compound plotted on the wavelength range 200 to 400 nm. This facility is available on the HPLC system employed for this work. However, a single wavelength of 280 nm could be used. The maximum absorption of daidzein was

249.8 nm with a shoulder at 301.9 nm, for genistein and biochanin A was 259.2 nm, and for flavone was 254.5 nm with a shoulder at 297.1 nm. 112

(D) (G)

'-"~

\

' . 5 10 15 20 25 Time (minutes)

Figure 4.1. Chromatogram of standard daidzein (D), genistein (G), biochanin A {B), and internal flavone standard (F), with C18 bonded phase column and acetonitrile­ tO% aqueous acetic acid (33:67, v/v) as eluent. 113

(D) (G)

...... ,~ ·srJl ~ ll) § -8 0 rJl (B) < (F)

u \~----

I

5 10 15 20 25 Time (minutes)

Figure 4.2. Chromatogram of standard daidzein (D), genistein (G), biochanin A (B), and internal flavone standard (F), with C18 bonded phase column and acetonitrile­ !% aqueous acetic acid (33:67, v/v) as eluent. 114

The capacity factors (k') of the isoflavones were equal to the ratio of the difference of the relative retention time of analytes with the dead time. The dead time was estimated using the clear solvent front peak. The capacity factors and RSD values of daidzein, genistein, biochanin A, and flavone using acetonitrile-I 0% aqueous acetic acid (33:67, v/v) and acetonitrile-1% aqueous acetic acid (33:67, v/v) as eluent are shown in Table 4.1.

Table 4.1. Capacity factor (k') of daidzein, genistein, biochanin A, and flavone (internal standard) with C18 column and acetonitrile-aqueous acetic acid (pH 1% and 10%) (33:67, v/v) as eluent.

Concentration Capacity factor (k') ± RSD of acetic acid

Daidzein Genistein BiochaninA Flavone

1% 0.92±0.92 2.24 + 0.66 12.76 ± 0.66 14.16 + 0.59

10% 0.60+4.69 1.24 ± 1.52 6.68 ± 1.28 9.22 + 1.34

Once the solvent was selected, the solvent compositions were optimised by measuring the retention time over a range of acetonitrile and 1% acetic acid compositions. The chromatograms for both the standard solution and the spiked soya bean sample at different solvent compositions are shown in Figure 4.3 (page 116).

There was no separation which had an ideal k' range of 1-10, however, a compromise solution with a k' range of 0.92 - 14.16 was selected. The optimum composition was acetonitrile-!% aqueous acetic acid (33:67, v/v). Table 4.2 (page overleaf) shows the calculated capacity factors for the phytoestrogens together with 115 their error estimate. A plot of log k' against solvent strength is displayed in Figure

4.4 (page 117).

The separation of all analytes was achieved in less than 24 minutes with acetonitrile-

1% aqueous acetic acid (33:67, v/v) as eluent. Daidzein eluted at 3 minutes

(k'=0.92), genistein at 5 minutes (k'=2.24), biochanin A at 21 minutes (k'=12.76) and flavone (the internal standard) at 23 minutes (k'=l4.16). This overall run time of the method is considerably more rapid than for gradient methods described by other investigators (up to 60 minutes) (Table 2.1, page 27).

Table 4.2. Capacity factor (k') of standard daidzein, genistein, biochanin A, and flavone with different solvent compositions.

Solvent Capacity factor (k' ± RSD)

Acetonitrile-water Daidzein Genistein BiochaninA Flavone (v/v)

30:70 1.30 + 0.84 3.31 + 0.11 20.16 + 0.31 21.07 + 0.35

33:67 0.92 + 0.92 2.24+0.66 12.76 + 0. 66 14.16 + 0.59

35:65 0.85 + 0.42 1.97 + 0. 23 11. 01 + 0. 39 12.41 ± 0.56

38:62 0.65 + 0.52 1.46 ± 0.67 7.42 + 0.66 8.86 ± 0.65

40:60 0.56 + 0.88 1.19 ± 0.57 5.67 + 0.85 7. 01 + 0.71 116

It

..0 .....

.IQ.Jio · Bt ~!'t fl!~ :i: .

< ~- .- •

A2 1111 • ... B2 ~ tl G· 'li "'' Ill ... .. • 0 ..; 0 .., ~ .. ... 011

"'0 f,: ' 'A3 .. Q..,, 83 C>· .. .. «·"" ...... "' 0 ...... ; 0 10

:__ u_ ~LJ

A B4 0 ~ .... e Ill g 0 .,. co:· .. "!...... Ill ~- rl .. "'! u...... __----~

Q· B5 ...... ;;~ -: ..." ~- .. C'I ~- .g, ~

1:--- . . '""'" u Figure 4.3. (A) chromatogram of standard and (B) chromatogram of soya bean spiked with standards daidzein (D), genistein (G), biochanin A (B), and internal flavone standard (F) with C 18 bonded phase column and acetonitrile-I% aqueous acetic acid as eluent. Al & Bl (30:70, v/v), A2 & B2 (33:67, v/v), A3 & B3 (35:65, v/v), A4 & B4 (38:62, v/v), A5 & B5 ( 40:60, v/v). 117

1.4 ....,._ Thid2ein 12 ...... Genistein --+-- BiochaninA 1 -Flawoe

0.8

::.: 0.6 bl) Q ~ 0.4

0.2

0

-0.2

-0.4

% Acetooitrile

Figure 4.4. Capacity factor (log k') of standard daidzein, genistein, biochanin A and internal flavone standard from different compositions of solvent (acetonitrile).

4.1.2. Analytical performances

4.1.2.1. Linearity

Linear curve fits were obtained from seven different concentrations using duplicate standard solutions and three replicate injections. Each analyte showed a linear fit with a coefficient of variation of> 0.999. This compares well with other methods

(Franke et al. 1994) (Table 4.3, overleaf). Linearity of internal standard (flavones) was not analysed because isocratic method used will not affect to the retention of the compounds. Therefore, the use of internal standard was discontinued. 118

Table 4.3. Calibration parameters of daidzein, genistein, and biochanin A using the proposed method and that of Franke et al. (1994, 1995). HPLC conditions: isocratic elution with acetonitrile-1% aqueous acetic acid (33:67, v/v) as eluent.

Compound Concentration range Equation Study (J.tM)

Daidzein 12.73-815.01 y=-0.21 ± 7.86x 0.999 Present study

Genistein 11.76-765.27 y=-0.22 ± 9.54x 0.999 Present study

BiochaninA 11.00-709.88 y=-0.35 ± 9.14x 0.999 Present study

Daidzein 0.7-35.0 Y- 0.008 + 1.53x 0.996 Franke et al. (1994, 1995)

Genistein 1.2-52.0 Y- 0.012 + 0.80x 0.998 Franke et al. (1994, 1995)

BiochaninA 1.0-50.0 Y- 0.028 + 1.05x 0.989 Franke et al. (1994, 1995)

4.1.2.2. Precision

The proposed LC method showed acceptable repeatabilities for each of the isoflavones; daidzein, genistein and biochanin A (Table 4.4, overleaf). Coefficients of variation (RSD) were obtained from six injections of a mixture of phytoestrogens within- and between-assay and the largest values were 1.26% and 1.51 %. 119

Table 4.4. Precision of HPLC technique for separation of a mixture of daidzein, genistein, and biochanin A standards. HPLC conditions: isocratic elution with acetonitrile-1% aqueous acetic acid (33:67, v/v) as eluent with six replicate injections.

Compound Number of Mean (area) RSD(%) RSD(%) injections within-assay between-assay

Daidzein 6 12.25 1.24 1.27

Genistein 6 14.02 0.96 1.20

BiochaninA 6 13.00 1.26 1.51

4.1.2.3. Recovery

The method showed good recoveries at close to 100%. The mean recoveries of the individual standards ranged from 100-127% for daidzein, 101-114% for genistein and 97-135% for biochanin A (Table 4.5, overleaf). The higher recoveries (>100%) were obtained for the lower levels of the spiked standard. This is probably due to small amounts of co-eluting impurities in soya beans, which were more significant at the lower spike levels. 120

Table 4.5. Spiked recovery for daidzein, genistein, and biochanin A from soya beans. HPLC conditions: isocratic elution with acetonitrile-1% aqueous acetic acid (33:67, v/v) as eluent.

Compound n J.tg/ml spiked J.tM spiked Recovery (%) RSD(%) Study (mean)

Daidzein 6 21.92 86.22 100.52 1.63 Present study

4 20.72 81.50 104.87 1.65 Present study

3 8.77 34.49 106.09 3.21 Present study

3 8.30 32.60 109.84 5.66 Present study

3 2.60 10.17 127.21 0.92 Present study

4 44.8 176.37 104.7 5.1 Franke eta/. (1995)

Genistein 6 20.32 75.19 101.36 3.09 Present study

4 20.68 76.53 107.45 4.36 Present study

3 8.13 30.08 108.33 1.03 Present study

3 8.27 30.61 114.68 1.32 Present study

3 2.58 9.50 109.63 2.76 Present study

4 40.5 150.00 93.7 4.6 Franke et al. (1995)

BiochaninA 6 20.18 70.98 97.82 2.18 Present study

4 20.04 70.49 100.26 2.69 Present study

3 8.07 28.39 105.70 0.99 Present study

3 8.02 28.20 108.68 1.97 Present study

3 2.50 8.80 135.98 2.78 Present study

4 30.7 108.09 101.1 2.5 Franke et al. (1995) 121

4.2. Separation and quantification of isoflavones (daidzein,

genistein, formononetin and biochanin A) and coumestans

(coumestrol) in soya beans1

4.2.1. Optimisation of the stationary phases

The separation of daidzein, genistein, formononetin, biochanin A and coumestrol using the cyano, C8, C18 and phenyl columns is shown in Table 4.6 (overleaf). Only four compounds (Figures 4.5-4.7, pages 123-125) could be separated on the C8, C18 and cyano columns using an eluent of acetonitrile-!% acetic acid (33:67, v/v) as described in section 4.1 (page 11 0) for the separation of three isoflavones (daidzein, genistein, and biochanin A). However, all five compounds were well separated on the phenyl column (Figure 4.8, page 126). At this stage, the identification of the individual compounds was not evaluated.

These results were published by L.S. Hutabarat, H. Greenfield, M. Mulholland in Journal of Chromatography A, 886 (2000) 55-63, "Quantitative determination of isoflavones and coumestrol in soybean by column liquid chromatography". 122

Table 4.6. The separation of daidzein, genistein, formononetin, biochanin A and coumestrol using the cyano, C8, C18 and phenyl columns.

Column Eluent Eluent (33:67, v/v) No. ofpeaks system

Cyano !socratic Acetonitrile - 1% aqueous acetic acid 4 C18 !socratic Acetonitrile - 1% aqueous acetic acid 4 C8 !socratic Acetonitrile - 1% aqueous acetic acid 4 Phenyl !socratic Acetonitrile - 1% aqueous acetic acid 5 123

N N(5) Ul m Ul m 'Jl :n 111D .D I I.::! I It .n ~ J\ Flavone

Time (minutes)

Figure 4.5. The LC chromatogram of a mixture of standards, daidzein, genistein, formononetin, and biochanin A, coumestrol, and internal flavone standard, by a cyano column and acetonitrile-!% aqueous acetic acid (33:67, v/v) as eluent. 124

Flavone

II (\J ..-! 1.11 >t !a ~ Ill ". ' N '!" ~ !-" ~

: '-"~

-e0 <

L I~ \Lt t II T T I v \J...... ·I HI II J 1r·r rr· I I I

Time (minutes)

Figure 4.6. The LC chromatogram of a mixture of standards, daidzein, genistein, formononetin, biochanin A, coumestrol, and internal flavone standard, by C8 column and acetonitrile-!% aqueous acetic acid (33:67, v/v) as eluent. 125

Flavone

Time (minutes)

Figure 4. 7. The LC chromatogram of a mixture of standards, daidzein, genistein, formononetin, biochanin A, coumestrol and internal flavone standard, by a C18 column and acetonitrile-I% aqueous acetic acid (33:67, v/v) as eluent. 126

!1'1 N N:l \1) !D !'- :m ~ !'- :D '1)1) I . " ""I "" ~ '" ~ Flavone "' ~ I"' I~ ~ I• ~ ..;f .~

'-'~ "' ·ap C1) Ul 0 § 1'- -€ ~ 0 • <"' :"II

\l II _l I I I I II. I l I I --r,,-. -¥-~r. rhr...~r-, '-~,~---,.r-.~..,..r 1~ . .J-.rrur_r..J_

I I I T

Time (minutes)

Figure 4.8. The LC chromatogram of a mixture of standards daidzein, genistein, formononetin, biochanin A, coumestrol and internal flavone standard, by a phenyl column and acetonitrile-I% aqueous acetic acid (33:67, v/v) as eluent. 127

Other experiments were also carried out by using three different columns (C8, C18 and cyano) with a variety of mobile phase systems for the separation of four isoflavones and coumestrol as shown in Table 4.7. One compound could not be resolved on a cyano column with an eluent of acetonitrile-I% aqueous acetic acid

(25:65, 30:70, 33:67, v/v) or acetonitrile-water (20:80, 25:75, 30:70, v/v), or on a C8 column with an eluent of acetonitrile-!% aqueous acetic acid (33:67, v/v), or on a

C18 column with an eluent of acetonitrile-!% aqueous acetic acid (33:67, v/v) or acetonitrile-10% aqueous acetic acid (33:76, v/v) or acetonitrile-acetate buffer (pH

2.6) (33:67, v/v). The phenyl column was the only one that could resolve all five compounds.

Table 4. 7. Separation of daidzein, genistein, formononetin, biochanin A and coumestrol by different stationary phases and mobile phases.

Stationary LC Mobile phases Composition No. of phases condition (v/v) peaks Cyano !socratic Acetonitrile-I% aqueous acetic acid 25:75, 30:70, 4 33:67

Cyano I socratic Acetonitrile-water 20:80, 25:75, 4 30:70

C8 I socratic Acetonitrile-I% aqueous acetic acid 33:67 4

C18 !socratic Acetonitrile-I% aqueous acetic acid 33:67 4

C18 I socratic Acetonitrile-I 0% aqueous acetic acid 33:67 4

C18 !socratic Acetonitrile-acetate buffer (pH 2.6) 33:67 4 128

4.2.2. Optimisation of solvent system

To optimise the separation, the selection of mobile phases using a phenyl column was studied. Summaries of the separation of the five compounds using a variety of solvents are shown in Table 4.8 (overleaf). Isoflavones and coumestrol were definitely not resolved with an eluent of acetonitrile-methanol or acetonitrile-ethanol

(in varying proportions), 1% aqueous acetic acid-water-methanol (70:10:20, v/v/v), and water-methanol (35:65, v/v). The five compounds were not completely separated using solvents isopropanol, methanol, acetate buffer (pH 2.6), and 10% aqueous acetic. They were well separated on a phenyl column with an elution system of acetonitrile-1% aqueous acetic acid (33:67, v/v) or acetonitrile-water (in varying proportions). The use of eluent with acid was avoided in this study in order to preserve the column. Therefore, acetonitrile in water was chosen for the mobile phase system throughout this study.

Altering the composition of acetonitrile in water markedly affected the resolution and the retention time. Thirty percent of acetonitrile in water almost separated coumestrol from genistein, while 35% acetonitrile in water completely separated coumestrol from genistein. All compounds could be eluted within 25 minutes, but this led to tailing of biochanin A. Thirty three percent of acetonitrile in water separated the five compounds satisfactorily. This system produced a sharp separation of coumestrol from genistein with the elution time for all compounds achieved within 22 minutes.

Daidzein eluted at 3.7 minutes (k'=l.1), coumestrol at 6.0 minutes (k'= 2.4), genistein at 6.5 minutes (k'=2.7), formononetin at 9.4 minutes (k'=4.3) and biochanin

A at 21.4 minutes (k'=11.1). Table 4.9 (page 127) shows the calculation of the 129 capacity factors (k') for the five compounds with their error estimate. A plot oflog k' against solvent strength is displayed in Figure 4. 9 (overleaf).

Table 4.8. Separation of daidzein, genistein, formononetin, biochanin A and coumestrol by phenyl column and a variety of eluents.

Solvent Composition No. of peaks

Acetonitrile-methanol 30:60, 65:35, 30:70 0

Acetonitrile-!% aqueous acetic acid 20:80, 25:75, 30:70, 33:67, 5 35:65

Acetonitrile-! 0% aqueous acetic 20:80, 25:75, 30:70, 35:65 4 acid

Acetonitrile-vvater 20:80, 25:75, 30:70 5

Methanol-vvater 35:65, 40:60 0, 3

Methanol-1% aqueous acetic acid 30:70, 35:65, 40:60 1, 1, 4

Acetonitrile-!% aqueous acetic acid- 15:75:10, 20:70:10, 2, 4, 4 isopropanol 25:65:10

1% aqueous acetic acid-vvater- 70:10:20 0 methanol

1% aqueous acetic acid-vvater 70:30 1

Acetonitrile-!% aqueous acetic acid- 20:70:10, 22:70:8, 10:70:20 4 methanol

Acetonitrile- acetate buffer pH 2.6 33:67 4

Acetonitrile-!% aqueous acetic acid- 20:70:10 2 vvater 130

Table 4.9. :ap~ity factor (k') of daidzein, genistein, formononetin and biochanin A and coumestrol with different compositions of eluent.

Acetonitrile Capacity factor -water (v/v) (k'i, SD)

Daidzein Coumestrol Genistein Formononetin BiochaninA

30:70 1.43i_0.09 3.17i_0.12 3.37i_0.07 5.68i_0.08 14.66+0.15

33:67 1.12i_0.07 2.41i_O.ll 2.69i_O.ll 4.32i_0.16 11.09i_0.35

35:65 0.99i_0.01 2.03i_0.03 2.20i_0.01 3.49i_0.03 8.37+0.16

1.4 1.2 1 ---+- Daidzein ~ 0.8 -II-Coumestrol C) *--- __...,._Genistein 0 0.6 )( ...J -- ~ Formononetin 0.4 ~ ~ Biochanin A 0.2 0 30 33 35 %Acetonitrile

Figure 4.9. Capacity factor Oog k') of standard daidzein, coumestrol, genistein, formononetin, and biochanin A from 30 to 40% acetonitrile in water.

The wavelength of the PDA detector was set from 200 to 400 run. To determine the optimum absorption of individual compounds, the area of the individual compounds was measured at three different wavelengths, i.e. 248.6, 259.2 and 342.1 nm. 131

Figure 4.10 (overleaf) shows the absorption of daidzein, coumestrol, genistein, formononetin and biochanin A at three different wavelengths. The maximum absorption of daidzein was achieved at wavelength 248.6 nm with a shoulder 300.5 nm. At 342.1 and 259.2 nm, the absorptions were 23% and 96%, respectively, less than those at 248.6 nm. The highest absorption of formononetin was at 248.6 nm with a shoulder 301.7 nm, while at 342.1 and 259.2 nm, the absorption was 22% and

98%, respectively, less than those at 248.6 nm. The maximum absorption of genistein was at wavelength 259.2 nm and 260.3 nm and shoulder 326.6 nm, but the absorption decreased to 31% and 76% at 248.6 and 342.1 nm, respectively.

Biochanin A was optimally absorbed at wavelength 260.3 and 259.2 nm, but at 248.6 and 342.1 the absorption decreased to 3% and 81%, respectively.

342.1 nm was the wavelength giving the maximum absorption for coumestrol.

However, the absorption decreased to 8% and 27% at wavelengths 248.6 and 259.2 nm, respectively. The quantitative analysis of daidzein, genistein, formononetin, biochanin A, and coumestrol at different wavelengths is given in Table 4.10

(overleaf). The UV chromatogram of the five compounds is shown in Figure 4.11

(page 133). 132

120

100 r---- ,...... - ..----1---r---- r--- ,...... - t--

...-.. 80 ::R ,__ ._..0 ...___ ,__ C:> Q) . > 60 0 0 Q) 0: 40

,...... - 1-t- 20 t--

0 248.6 259.2 343.1 Wavelength (nm)

D Daidzein D Coumestrol D Genistein D Formononetin D Biochanin A

Figure 4.10. Absorption maxima of daidzein, genistein, formononetin, biochanin A and coumestrol at different wavelengths.

Table 4.10. Quantitative analysis of daidzein, coumestrol, genistein, formononetin and biochanin A at different wavelengths (mean of 6 injections, mg/L ±. SD).

Species Compound Wavelength

248.6 nm 259.2nm 342.1 nm

Isoflavones Daidzein 15.7 ±. 1.3 15.0 + 1.6 3.6 ±. 0.4

Genistein 8.4 ±. 0.2 12.2 + 0.6 2.9 ±. 0.1

Formononetin 22.3 ±. 2.8 21.9 ±. 3.9 5.0 ±. 0.9

BiochaninA 14.6 _±1.0 21.8 + 1.7 4.2 ±. 0.5

Coumestans Coumestrol 10.4 ±. 0.3 8.2 + 0.3 11.3 ±. 0.4 133

0.070- , .t 0.065 I c ~c 0.060 §

e0 0.055 u.

0.050

0.045 <( ·~ 0.040 gro co ~ 0.035 I 0.030

0.025

0.020

0.01S.:

0.010

I I I I I I I I. I I I I" I I I I' I' I I' I I I I" I I I I I I I I I. I I I I I I I I I I I I I I I I I I I' I 2.00 4.00 6.00 8.00 10.00 12.00 14.00 16.00 18.00 20.00 22.00 24.00 26.00 28.00 30.0\

Time (minutes)

Figure 4.11. UV chromatogram of standards daidzein, coumestrol, genistein, formononetin, and biochanin A. The analytes were separated on a phenyl column with isocratic system: acetonitrile-water (33:67, vfv) as eluent. 134

Correct peak assignment of daidzein, coumestrol, genistein, formononetin, and

biochanin A was reconfirmed by LC-MS-MS. Solutes entering the mass

spectrometer via HN-APCI interface in the positive modes were analysed over a rn/z

range from 150 to 600. A full scan of the UV chromatograms of a mixture of five

standards obtained from LC-MS-MS is shown in Figure 4.12 (overleaf). No

significant difference in the chromatographic resolution was observed between LC­

UV detection and LC-MS-MS. The retention times produced by LC-MS-MS were

very close to those produced by LC-UV. Selected peaks from Figure 4.12 were then

extracted using positive acquisition mode over rn/z ranging from 150 to 600. Peaks 1

to 5 showed a major ion product at rn/z 254.6, 269.0, 270.8, 269.2, and 284.8, respectively, as shown in Figure 4.13 (page 136). Peak 1 was identified as daidzein

(Mw=254), peak 2 was coumestrol (Mw=268), peak 3 was genistein (Mw=270), peak

4 was formononetin (Mw=268), and peak 5 was biochanin A (Mw=284). The values accorded with the established knowledge of the molecular weight of these five compounds (Setchell & Welsh 1987; Franke et al. 1994; Barnes et al. 1994, 1998).

The elution order of daidzein, coumestrol, genistein, formononetin, and biochanin A was different from that of Setchell and Welsh (1987) and Franke et al. (1994) who reported the elution order as daidzein, genistein, coumestrol, formononetin, and biochanin A. The stationary phases and an eluent used by these authors were C 18 with methano-0.1 M ammonium acetate buffer pH 4.6 (60:40, v/v) and, C18 with a gradient system of acetonitrile and 1% acetic acid, respectively. 135 - .. ·-·· ...... -.... - ... -...... 2.6e7 - 1. 1 2.65e7 cps Ptak4 2.487 +. Peak I 2.2e7 8.78 + 3.49 2.0&7

1.887

1.6&7

;

1.0e7 .+

PW2 19.17 8.0e6 \

4.0t8

2.086

Time (minutes)

Figure 4.12. UV chromatogram of daidzein (peak 1), coumestrol (peak 2), genistein (peak 3), formononetin (peak 4), and biochanin A (peak 5) standards obtained from LC-APCI-HN-MS analysis using a phenyl column and water-acetonitrile (67:33,v/v) as eluent. 136

TIC of +01: from Standard 2.65e7 cps

:::~.~16..c.:.;;;._....12 ~+------.:24.09 5 10 15 20 Time, min ' +01: 3.53 min from Standard 1.65e6 cps "' 1.696 25 .6 fr • 1.0e6 ~c Q) 5.0e5 £ 531.0 186.2 337.0 371.2 466.0 573.4 200 300 400 600 mlz, amu +01: 5.69 min from Standard 2.69e6 cps 26 .0 "'fr 2e5

1e5 :l57iJ.4 .. 185.2 212.8 A 413.6 469,6 493,6 529.2 1• 200 300 400 500 mlz, amu +01: 5.97 min from Standard 5.21e5 cps 27 .8 "'fr 4e5 ;; ~ 2e5 Q) I £~.. __~1~85~-~2~~- 2_4~L·-4~u--~~~--~----~3~9~~;·0~----~4~59~-~4--~5~0~5~.2~----56~3~.0~~ 200 300 400 500 m/z, amu +01: 8.70 min from Standard 1.68e6 cps

fr 1.5e6J ' 26[.2 ;Eo 1.096 ;;; c .! 5.0e5 I 5 5 E L-~1~85~·=2~l~13~-~0--~~~~~Ti ______3~9~2~f~------~4~9~0~.?~----- _~~--0~~i 200 300 400 600 mlz, amu +01: 19.21 min from Standard 5.30e5 cps

Ill 26 .8 fr 4e5

2e5

164.6 ! 337.6 382.8 413.2 507.4 540.4 200 300 400 500 m/z, amu

Figure 4.13. Positive ion mass spectra of peaks 1, 2, 3, 4, and 5 observed in Figure 4.12. Peak 1 is daidzein with rnlz of 254.6. Peak 2 is coumestrol with rn!z of 269.0. Peak 3 is genistein with rn!z of 270.8. Peak 4 is formononetin with rn!z of 260.2. Peak 5 is biochanin A with rn/z of 284.8. 137

4.2.3. Analytical performances

4.2.3.1. Precision

The LC system with a phenyl column and an isocratic elution system with acetonitrile-water (33:67, v/v) as eluent showed a high degree of precision. This is shown by the coefficients of variation of six injections on one occasion (within­ assay) which were 0.3-0.6 % for the retention times and 0.7-5.7% for quantitative analysis of the individual compounds. The coefficients of variation of injections on three different occasions (between-assay) were 1.0-2.1% for the retention times and

2.4-6.8% for quantitative analysis of the individual compounds.

The coefficients of variation within- and between-assay for the retention times of each compound ranged from 0.3 to 0.6 and 1.0 to 2.1, respectively. The between­ assay coefficient of variation was higher than that within-assay (Table 4.11, overleaf). The coefficients of variation within-assay of the quantitative analysis of the individual compounds are shown in Table 4.12 (overleaf). The coefficients of variation of the quantitative analysis of the individual compounds using the developed LC technique within-assay ranged from 0.7 to 5.7. The coefficients of variation of between-assay ranged from 2.4 to 6.8. The between-assay coefficient of variation was higher than the within-assay coefficient of variation. 138

Table 4.11. Coefficients of variation of the retention times of daidzein, genistein, formononetin, biochanin A and coumestrol (mean of 6 injections within­ and between-assay).

Species Compound Mean cv cv retention time within-assay between-assay (minutes) (%) (%)

Isoflavones Daidzein 3.76 0.29 0.99

Genistein 6.54 0.56 1.56

Formononetin 9.48 0.45 1.64

BiochaninA 21.49 0.63 2.08

Coumestans Coumestrol 6.10 0.36 1.45

Table 4.12. Coefficients of variation of the quantitative analysis of daidzein, genistein, formononetin, biochanin A, and coumestrol.

Species Compound HPLC cv cv concentration within-assay between-assay (mg/L) (%) (%)

Isoflavones Daidzein 15.26 3.79 3.97

Genistein 13.01 3.10 6.41

Formononetin 22.00 0.74 2.39

BiochaninA 22.89 3.92 6.77

Coumestan Coumestrol 11.29 5.68 3.97 139

4.2.3.2. Linearity

The LC method, utilising a phenyl column and an isoc1'r.~~c elution system with acetonitrile-water (33:67, v/v), showed a E3h degree of linearity. The detector response was linear with the coefficients of determination (ft) being more than 0.999 over the concentration range from 0.02 mg/L to 11 mg/L ex~ected from the food extracts (Table 4.13, overleaf).

4.2.3.3. Detection limits

Detection limits for daidzein, coumestrol, genistein, formononetin, and biochanin A were found to be 47, 82, 76, 75, and 224 nM, respectively. This study used a UV detector for coumestrol. Wolfbeis and Schaffner (1980) used a fluorescence detector, which can detect coumestrol down to concentrations as low as 50 nM. However, UV detection of coumestrol at a concentration of 82 nM in ethanol is acceptable. 140

Table 4.13. Coefficients of determination (?) of standard daidzein, genistein, formononetin, biochanin A, and coumestrol using a phenyl column and acetonitrile-water (33:67, v/v) as eluent.

Compound Concentration range Equation

(mg/L)

Daidzein 0.02-15.10 y=3207 .43±33558.26x 0.999

Genistein 0.02-15.20 y=2697 .11±.40730.09x 0.999

Formononetin 0.03-16.20 y=1415.59±32244.41x 0.999

BiochaninA 0.03-18.20 y=1418.73±31029.66x 0.999

Coumestrol 0.02-11.20 y=557 .57±2552.91x 0.999

4.3. Optimisation of the isolation of phytoestrogens from soya

beans, cooked soya beans, canned soya beans, tofu and

soya milk

4.3.1. Solvent, pH, and time

As expected, dried soya beans were found to contain only the isoflavones daidzein and genistein among the five compounds of interest. The LC chromatogram and the

UV absorbance spectra of daidzein and genistein isolated from soya beans were identical to those of standard compounds (Figures 4.14 & 4.15, pages 142 & 143).

Statistical analysis with six replications showed that there was no significant difference in the peak area of daidzein and genistein at different concentrations of phosphoric acid and reflux times (p>0.05) (Figure 4.16, page 144). Therefore, 141

phosphoric acid was not considered a suitable and effective solvent for hydrolysis of

conjugated isoflavones in soya beans.

Since phosphoric acid was not applicable, ethanol with hydrochloric acid was

examined as the solvent for the hydrolysis of the conjugates from soya beans. The

optimum pH of acid and the length of refluxing were then investigated.

The effects of acid hydrolysis and reflux times on the levels of daidzein and genistein

in dried soya beans are shown in Figure 4.17 (page 145). Levels of daidzein and

genistein in dried soya beans were significantly affected by pH and reflux times

(p<0.05). The levels were significantly higher with acid hydrolysis (pH 0.6) than

without acid hydrolysis (pH 7). The figure shows that the longer the reflux times, the

higher the yield of daidzein and genistein obtained. These results agree with those of

Wang et al. (1990) and Franke et al. (1994, 1995) who suggested that the best method for the hydrolysis of isoflavone conjugates in soya beans and its processed

products was 1M HCI with heating for 2 hours at 98-100°C. 142

Speclnln Index Plot

Dllllztkl·l.m Ca1111111DI· t.Ut Gt'*fllln ·1.55S Fe1110-lil·1.554 ltl2ch11i A. 2Utl 22UI 1111 220.01 ftlll 221.00 IM UUO 11 221.tl lit UUI 311.00

3'2.1

•• ~

C) DJ5 ...,.... ~ I D.N .

c D.l2

1.11 . ~ 1.10 ~._ \.... \.

2.'- .... e.to ilo 101 12Jt 14JI 11:11 .. :.. »:o~ 22M 24M 21~~ 21At ~u Time (minutes)

Figure 4.14. The LC chromatogram and UV spectra of standard daidzein, coumestrol, getristein, formononetin and biochanin A using a phenyl column and acetonitrile-water (33:67, v/v) as eluent. 143

Sptctnun Index Plot

1.540 8nllltkl· 5.74 t 22UI 1111 22tJO All 381 380.01

282.7

1.21

0.21

l.t5 ~ O.tl i

O.H

O.M

Time (minutes)

Figure 4.15. The LC chromatogram and UV spectra of daidzein and genistein isolated from dried soya beans using a phenyl column and acetonitrile-water (33:67, v/v) as eluent. 144

Daidzein

700000

-0-1-2 600000 __,.._ 3 -::::::;- 4

500000 ...... -5 --e--6

ctl Cl> 400000 ctl ~ ctl '\..._ _! Cl> 300000 ..,. a. "" - 200000

100000

0 0 0.5 1 1.5 2 2.5 Phosphoric acid (M)

Genistein

500000~------~ 450000 400000 350000 ctl C1) 300000 ctl ~ 250000 ctl :.. 200000 150000 --+--1-2

100000 -~-3 --+--4 50000 -5 --CI--6 o+------~------r------,------~====~ 0 0.5 1.5 2 2.5

Phosphoric acid (M)

Figure 4.16. Peak areas of daidzein and genistein in soya beans extracted with different concentrations of phosphoric acid and at different periods of reflux. 145

Daidzein

60~------~ §- 40 ·-.... ECI:S c:oe o, 3o Fc=o.Gl 8 ~ 20 ~ c 0') 8 §. 10 • • • • 0 1 2 4 6 Refluxing Time (hours)

Genistein

35 c- 30 :8 ~ 25 CI:S 1.. b g 20 6 c 0 Cl> 15 I 0:!:::.. :-~- 1 c 0') 10 o-____, '--" _____. 8 §. 5 • • • • 0 0 1 2 4 6 Refluxing Time (hours)

Figure 4.17. Levels of daidzein and genistein in raw dried soya beans, refluxed with ethanolic HCl at different reflux times and pH (mg/100, dry weight basis). 146

The LC chromatogram and the UV absorbance spectra of daidzein and genistein isolated from soya milk were identical to the chromatogram and absorbance patterns of the authentic compounds (Figure 4.18, overleaf). The levels of both compounds were significantly affected by reflux time and acid hydrolysis (p<0.05). Figure 4.19

(page 148) shows that daidzein levels remain unchanged without hydrolysis (pH 7) for various reflux times or with hydrolysis but without refluxing. The levels slightly increased at pH 1.42 and 0.96 after refluxing for up to 6 hours. Extraction of phytoestrogens at pH 0.76, 0.57 or 0.44 for 1, 2, or 4 hours refluxing gave moderate levels. The highest level of daidzein was achieved after refluxing for 6 hours at pH

0.76, 0.57, or 0.44.

No significant changes were detected in the level of genistein without hydrolysis for various reflux times or without refluxing at various levels of pH. The level of genistein was significantly lower at pH 1.42 and 0.96 for various reflux times and at pH 0.44 for 1, 2, or 4 hours refluxing (p<0.05). The levels increased at pH 0.96 and

0.76 at various reflux times. Refluxing for 6 hours at pH 0.76 gave the highest level of genistein followed by refluxing for 6 hours at pH 0.57 and 0.44. 147

lptctrum Index Plot

. t.m Uti 2.121 UIS Dllizeti·Ult . Gellllttl·5,752 220M 11 221M Rll 221.10 •111 2zt.tt Rill 22UI lilt 221M 1111 "'

·~r------~ l.ft

1.11 l.f4 1.12· :c: :I ..... c .....

'·"· I 1.14 1: 'I ·i U2 ~ ~ ~

l.tt f---' ~~.___,...____,&..._ ~ ______~

2.Gt 4.M. IM llt fill f2~et f4JO 11JO 11~tl 21Jt 22Jt u:H 21:u 21~tl 3t.H Time (minutes)

Figure 4.18. The LC chromatogram and UV spectra of daidzein and genistein isolated from soya milk. 148

Daidzein

-+-7.22 -c-1.42 --.!.--0.96 --'; :-0.76 ---?IE--0.57 -e-0.44

0 1 2 4 6 Reflux time (hours)

Genistein

-+-7.22 -~-1.42 --&--0.96 _;:. i-0.76 ---?IE- 0. 57 -e-0.44

0 1 2 4 6 Reflux time {hours)

Figure 4.19. Daidzein and genistein levels in soya milk, refluxed with ethanolic HCI at different reflux times and pH (mg/100, dry weight basis). 149

4.3.2. Temperature

Mean levels of daidzein and genistein extracted from dried soya beans at different

temperatures are shown in Table 4.14. This table shows that daidzein and genistein levels were significantly different at different temperatures (p<0.05). The highest levels of both compounds were obtained by extraction at 100°C. The levels were found to be three fold higher than when the extraction temperature was 80°C.

Table 4.14. Daidzein and genistein in dried soya beans at different extraction temperatures (mg/100 g, wet weight basis).

Temperature Daidzein Genistein

35.60 ±5.99 24.19 ± 1.58

110.69 ± 1.18 75.20 ± 3.47

Values are means±. SD of duplicate analysis

Mean levels of daidzein and genistein in soya milk at different extraction temperatures are given in Table 4.15 (overleaf). The levels were significantly affected by temperature during the extraction process (p<0.05). The levels of daidzein and genistein in soya milk extracted at 100°C were about ten times higher than those extracted at 80°C. 150

Table 4.15. Daidzein and genistein in soya milk at different extraction temperatures

(mg/100 g, wet weight basis).

Temperature Daidzein Genistein

0.53 .±. 0.06 0.40 .±,0.07

5.73 .±,2.03 3.76 .±. 1.40

Values are means:!:. SD of duplicate analysis

4.3.3. Optimisation of the extraction for cooked and canned soya beans and

tofu

4.3.3.1. Cooked and canned soya beans

The effects of acid hydrolysis and reflux times on the levels of daidzein and genistein in cooked dried and canned soya beans are shown in Figures 4.20 and 4.21 (pages

152 & 153). The levels of both compounds in cooked dried and canned soya beans were significantly affected by pH and reflux times (p<0.05). The levels were significantly higher with acid hydrolysis (pH 0.6) than without acid hydrolysis

(pH 7).

Dried and cooked soya beans showed a similar trend of increasing the amounts of daidzein and genistein extracted with the increase of refluxing times. The levels were the same without refluxing or after 1 or 2 hours refluxing. The levels started to increase after 4 hours refluxing. Maximum levels were obtained after 6 hours 151 refluxing. Canned soya beans had the highest levels of daidzein and genistein after 4 hours refluxing although the levels slightly decreased after 6 hours refluxing.

4.3.3.2. Tofu

Tofu was found to contain only the isoflavones daidzein and genistein among the five analytes studied. Effects of acid hydrolysis and reflux times on the level of daidzein and genistein in tofu are shown in Figure 4.22 (page 154). Both daidzein and genistein in tofu were significantly affected by pH and reflux time (p<0.05). The levels were higher with acid hydrolysis (pH 0.6) than without acid hydrolysis (pH 7).

Daidzein and genistein levels in tofu increased significantly up to 2 hours reflux and slightly increased after hydrolysis for 2 to 4 hours but sharply increased from 4 to 6 hours reflux. 152

Daidzein 100 c- .2 E 80 ... C'CI ...-C'CI oD')- 60 Co CD,_ I ,·~~.61 o- 40- c D') r o-o E 20- 0 1:---~-~ ~ ~ 0 1 2 4 6 Refluxing Time (hours)

Genistein 12 ...------. ·-S-10E /D 1ii ~ 8 1-D') ..-c:: 1:CD,_ g 6 /--- 0- 4 c D') ~ 8 s 2 w:~=------.... __.... ----- ....·-- ...... 0+---~----~--~----~--~ 0 1 2 4 6 Refluxing Time (hours)

Figure 4.20. Levels of daidzein and genistein in cooked dried soya beans, refluxed with ethanolic HCI at different refluxing times and pH (mg/100 g, dry weight basis). 153

Daidzein 120 -.------, c- o E 100- :; e 8o A] J.. C) ~g 60 ~ ~ (.) :!:::.. 40 fJ l=!=IJ 8E 20 L/ 0 -1--*==:;::::::::.~::::;::::~::::=;:.~~:;=::L-1 0 1 2 4 6 Refluxing Time (hours)

Genistein

80~------~

~ l=!=IJ

0 1 2 4 6 Refluxing Time (hours)

Figure 4.21. Daidzein and genistein levels in canned soya beans, refluxed with ethanolic HCl at different reflux times and pH (mg/100 g, dry weight basis). 154

Daidzein c-80-.------. ~ ~ 60 ~0, , __~/r::J c cc 40 - =(' Cl) or- '"~ g Ol 20 ----~J o E u (.) - 0 +-__:...•-..---·--.--·-..---·--.--· --l 0 1 2 4 6 Refluxing Time (hours)

Genistein

r=;:o.6l ~

0 1 2 4 6 Refluxing Time (hours)

Figure 4.22. Daidzein and genistein levels in tofu, refluxed with ethanolic HCI at different reflux times and pH (mg/100 g, dry weight basis). 155

4.4. Stabilisation of daidzein, genistein, formononetin,

biochanin A and coumestrol during storage at various

levels of pH

The pH of the samples after extraction was found to be less than 1. This pH was not

in the range of the recommended pH (2-8) for operating the phenyl column.

Therefore, the pH was adjusted to a range of 3-6 by adding sodium hydroxide into

the extracted samples prior to the LC analysis. Since only daidzein and genistein

were found in soya bean samples, the study on the stability of isoflavones and

coumestrol in the samples at various levels of pH and storage times can only be

discussed in terms of these two analytes. Formononetin, biochanin A and coumestro~

however, are discussed based on the study of standards.

Figure 4.23 (overleaf) shows the mean recoveries of daidzein at various levels of pH

and at different storage times. There were no significant differences in the mean recovery of daidzein in the extracted samples after the addition of sodium hydroxide

(pH 1-6). Daidzein was stable for up to 28 days storage at -20°C and at pH 3-6. This result was supported by the study on the stability of the standard daidzein during storage in alkali and acid (Figure 4.24, overleaf). The mean recovery of standard daidzein did not change significantly after the addition of sodium hydroxide (pH 10-

12) or hydrochloric acid (pH 3-6) into a mixture of standards (p>0.05). Standard daidzein was also stable after storage at -20°C for up to 28 days (p>0.05). 156

pH 150 ~1 125 ~"-2 ~ 100 ~· -----3 .....>. 4 =I Q) -4 > 75 0 (.) ...... -s Q) 50 a: --+--6 25 --+-7 0 1 4 14 21 28 Storage (days)

Figure 4.23. Recovery of daidzein from extracted soya beans during storage for 1, 4, 14, 21 and 28 days at various pH levels.

pH 125 ~3 .--.. ~4 :::Re.... -----s ~ -6 ~ 100 0 • (.) 2±iiiiP _._10 Q) ':c:::t==-~ a: -+-11 --+-12 75 1 2 4 14 21 28 Storage (days)

Figure 4.24. Recovery of standards daidzein during storage for 1, 4, 14, 21 and 28 days at various pH levels. 157

There was no effect from the addition of sodium hydroxide to the extracted soya bean on the mean recovery of genistein (p>0.05). The levels of genistein also did not change after storage for up to 28 days at pH 1-7. Mean recoveries of genistein from extracted samples at various levels of pH and storage times are shown in Figure 4.25.

This result was supported by the study on the stability of standard genistein. Standard genistein was stable after the addition of sodium hydroxide or hydrochloric acid into the mixture of standards. Standard genistein was also stable during storage for up 28 days (p>0.05) (Figure 4.26, page 158).

pH 125 """'*--1 ~ 100 --.:il- 2 :>li!0 ~ -3 ~ 75 =~ Q) ·~ > -4 0 0 50 -.-s a:Q) -e--6 25 --+-7 0 1 4 14 21 28 Storage (days)

Figure 4.25. Recovery of genistein from extracted soya beans during storage for 1, 4, 14, 21 and 28 days at various pH levels. 158

pH 125 ~3

~ -tr---4 ~ i::' ---5 ~ 100 -6 0 •==::~ 0 Is~ ---.....-10 Ql a: ~11 -+--12 75 1 2 4 14 24 28 Storage (days)

Figure 4.26. Recovery of standard genistein during storage for 1, 4, 14, 21 and 28 days at various pH levels.

The mean recoveries of standard coumestrol, formononetin and biochanin A (Figures

4.27, 4.28, 4.29, pages 159 & 160) were not significantly different after the addition

of sodium hydroxide or hydrochloric acid into the standard solutions (p>O.OS). There

were no effects of storage for up to 28 days on the mean recoveries of standard coumestrol, formononetin or biochanin A (p>0.05). 159

pH 125 ~3 .--. -1-4 ;,R e.... ~5 >. '- Q) 100 ,..~ --c--6 > .-... -- 0 -~~-~~~~ "' ~- ...... 0 -D-10 Q) 0: -:~11 --+-12 75 1 2 4 14 21 28 Storage (days)

Figure 4.27. Recovery of standard coumestrol during storage for 1, 2, 3, 14, 21 and 28 days at various pH levels.

pH 125 ~3 .--. -1-4 ~ -5 -~ ~ 100 ~--=! .;so "-~ ----.r--6 0 ·~~P- L-1 0 ~ -=-:-10 Q) 0: -::-11 --+-12 75 1 2 4 14 21 28 Storage (days)

Figure 4.28. Recovery of standard formononetin during storage for 1, 2, 3, 14,21 and 28 days at various pH levels. 160

pH 150 ...... -::~3 ..._.,~ ---4 --+--5 ~ ¥ :-t Q) 100 C-=tt ,..,e- > "-./ """'*""" 6 0 ~ () -::J-10 a:Q) -:r-11 -:r12

50 I 1 2 4 14 21 28 Storage (days)

Figure 4.29. Recovery of standard biochanin A during storage for 1, 2, 4, 14, 21, and 28 days at various pH levels.

4.5. Recovery

The mean recoveries of standards added into soya bean samples ranged from 70 to

107% (Table 4.16, page 162). Mean recoveries of each compound were 104% for daidzein, 94% for coumestrol, 93% for genistein, 99% for formononetin, and 89% for biochanin A. The mean recoveries of daidzein spiked with 2.14, 1.07 or

0.43 mg/L was slightly over 100%. The mean recoveries of coumestrol spiked with 2 or 1 mg/L were less than 100%, but coumestrol was not detected when spiked with

0.40 mg/L. The mean recoveries of genistein, and formononetin spiked with 0.4 and

0.65 mg/L, respectively, were less than 100%. Spiking with 0.73 mg/L produced very low mean recoveries ofbiochanin A (69.6%). 161

Only genistein was found in fresh whole green beans. The level was very low compared to those found in dried soya beans. The mean recoveries of standards added to fresh whole green beans ranged from 90 to 106% (Table 4.17, page 163).

Mean recoveries of each compound were found to be 93% for daidzein, 98% for coumestrol, 100% for genistein, 99% for formononetin, and 93% for biochanin A.

No isoflavones and coumestrol were found in fresh snow peas. The mean recoveries of standards added into fresh snow peas ranged from 92 to 106%. Mean recoveries of each compound were 96% for daidzein, 99% for coumestrol, 95% for genistein, 96% for formononetin and 94% for biochanin A (Table 4.18, page 164). 162

Table 4.16. Spiked recovery of daidzein, coumestrol, genistein, formononetin, and biochanin A in soya beans using a phenyl column with an isocratic elution system with acetonitrile-water (33:67, v/v) as eluent.

Compound N mg/L spiked Recovery (%) (mean)

Daidzein 2 2.14 107.40

2 1.07 101.18

2 0.43 104.64

Coumestrol 2 2.02 94.69

2 1.01 92.48

2 0.40 ND

Genistein 2 2.18 107.49

2 1.09 101.33

2 0.44 70.69

Formononetin 2 3.42 102.43

2 1.62 105.64

2 0.65 87.56

BiochaninA 2 3.64 100.34

2 1.82 97.27

2 0.73 69.56

ND= not detected 163

Table 4.17. Spiked recovery of daidzein, coumestrol, genistein, formononetin and biochanin A in flat beans using a phenyl column with an isocratic elution system with acetonitrile-water (33:67, v/v) as eluent.

Compound N mg/Lspiked Recovery (%) (mean)

Daidzein 2 2.14 96.0

2 1.07 92.6

2 0.43 91.0

Coumestrol 2 2.02 102.7

2 1.01 93.1

2 0.40 98.6

Genistein 2 2.18 105.9

2 1.09 97.5

2 0.44 97.3

Formononetin 2 3.42 106.3

2 1.62 95.9

2 0.65 92.3

BiochaninA 2 3.64 96.2

2 1.82 93.5

2 0.73 90.7 164

Table 4.18. Spiked recovery of daidzein, coumestrol, genistein, formononetin and biochanin A in snow peas using a phenyl column with an isocratic elution system with acetonitrile-water (33:67, v/v) as eluent.

Compound N mg/Lspiked Recovery(%) RSD (%) (mean)

Daidzein 2 2.14 99.1 0.19

2 1.07 95.1 0.07

2 0.43 93.9 2.37

Coumestrol 2 2.02 106.5 0.33

2 1.01 93.7 0.34

2 0.40 99.5 0.08

Genistein 2 2.18 98.3 0.71

2 1.09 95.7 1.90

2 0.44 92.1 0.69

Formononetin 2 3.42 94.0 0.25

2 1.62 97.1 0.33

2 0.65 95.3 0.05

Biochanin A 2 3.64 95.2 2.5

2 1.82 100.04 0.29

2 0.73 91.7 1.47

4.6. Conclusion

The LC-UV system with C18 reversed-phase column and isocratic elution with 1% aqueous acetic acid-water (33:67, v/v) could be used for the separation and 165 identification of the isoflavone compounds ( daidzein, genistein and biochanin A) only. This method is superior to published gradient LC methods for the separation and quantification of daidzein, genistein, and biochanin A because it is rapid and simple. Separation can be achieved within 24 minutes.

The LC-UV system with C18 reversed-phase column and isocratic elution with 1% aqueous acetic acid-water (33:67, v/v) as eluent, can not be used for the separation of isoflavones and coumestrol. However, the LC-UV system with a phenyl column and isocratic elution system with acetonitrile-water (33:67, v/v) as eluent, can be used for the separation and quantification of daidzein, genistein, formononetin, biochanin A and coumestrol. The separation using this method can be achieved within 24 minutes which is similar to the method used for the separation of daidzein, genistein, and biochanin A.

This method is considered superior to other HPLC methods for the aglycones because it is isocratic, simple, more rapid and avoids the use of acid. It was therefore used to determine the five analytes of isoflavones and coumestrol in a wide range of

Australian and Indonesian foods (Chapter 5, page 167).

Based on the results, it was decided to use the following method for extraction and hydrolysis. One gram of solid foods (dried soya beans, cooked soya beans and tofu) or 10 ml of liquid foods (soya milk) were extracted with 10 ml of 2M HCl and diluted with 96% ethanol up to 50 ml. Foods were sonicated for 20 minutes and hydrolysed with refluxing in the water bath at 100°C for 6 hours. This method was 166

used throughout for the analysis of isoflavones and coumestrol in dried beans, liquid

foods, and cooked foods.

One gram of canned soya beans was extracted with 10 ml of 2 M HCl and diluted with 96% ethanol up to 50 mi. Foods were sonicated for 20 minutes and hydrolysed with refluxing in a water bath at 100°C for 4 hours. The extraction method for canned soya beans was then also used for the analysis of isoflavones and coumestrol in all canned products.

Based on these results, it was decided that foods would be adjusted to pH 4 with sodium hydroxide after extraction throughout the analysis of isoflavones and coumestrol in foods. pH 4 produced optimum results and was within the recommended range for the column. Foods would be stored at -20°C if the LC analysis could not be carried out on the same day as extraction.

The recoveries of standards extracted from soya beans, fresh whole green beans and fresh snow peas were optimal. It was assumed that since the method recovered the analytes well (-100%) therefore analytical results for foods needed no adjustment for recovery factors. 167 CHAPTERS.

RESULTS AND DISCUSSION

QUANTITATIVE ANALYSIS OF PHYTOESTROGENS

INFOODS1

5.1. Soya beans and soya bean products

The soya beans and soya bean products analysed in this study were classified into four groups: raw soya beans, non-fermented soya bean products, fermented soya bean products, and second-generation soya foods.

5.1.1. Raw soya beans

Dried soya beans from Australia, Indonesia, or imported to Indonesia from the USA, were found to contain only daidzein and genistein among the five compounds studied

(Table 5.1, overleaf). The levels differed between varieties and sources on a dry weight basis (p<0.05). The varieties of soya bean studied were unknown except for

Bowyer soya beans from Australia. Dried soya beans purchased in Indonesia contained higher levels of daidzein and genistein compared to those from Australia or to those imported from the USA to Indonesia.

1 Some of these results were published by L. S. Hutabarat, H. Greenfield, M. Mulholland in Journal of Food Composition and Analysis 14 (2001) 43-58, "Isoflavones and coumestrol in soybeans and soybean products from Australia and Indonesia" 168

Two different soya bean varieties from Australia had different levels of daidzein and

genistein. Soya beans from McKenzie's (variety unknown) contained higher daidzein

levels than Bowyer from Allgold Foods but lower genistein levels. The level of

genistein in Bowyer soya beans from Allgold Foods (Australia) was similar to those

imported from the USA. Dried soya beans from ~enzie' s were the only branded

dried soya bean products available in supermarkets in Sydney, Australia, while

Bowyer dried soya beans had to be purchased from the wholesaler (Allgold Foods,

Australia). Both were originally grown in Australia.

Table 5.1. Mean levels of daidzein and genistein in dried soya beans purchased in Indonesia and Australia (mg/1 00 g, wet and dry weight basis).

Foods Basis % Daidzein Genistein Water

Soya beans, dried (imported Wetwt 10.2 30.8±4.3 72.3 ± 2.4 from the USA to Indonesia) Drywt 34.4 ± 4.8d 80.5 ± 0.7b

Soya beans, dried (Indonesia) Wetwt 8.8 127.7 ±27.0 83.4 ± 8.9 Drywt 140.0 ± 29.2a 91.4 ± 10.5a

Soya beans, dried Wetwt 9.3 96.4 ± 12.8 61.4 ± 13.5 (McKenzie's, Australia) Drywt 106.3 ± 14.2b 67.6 ± 14.9c

Bowyer soya beans, dried Wetwt 9.1 65.4 ± 11.4 72.0 ± 16.8 (Allgold Foods, Australia) Drywt 72.0 ± 12.6c 79.2± 18.5b

Values are means± SD of duplicate analyses of three purchases. Dissimilar superscripts (a=highest) within column indicate significant differences on a dry weight basis (p<0.05) 169

The differences in the levels of daidzein and genistein in soya beans analysed would

have been influenced by factors such as cultivars and agricultural conditions (Dalais

et al. 1997a; Eldridge & Kwolek 1983), however, these could not be specified for

these retail purchases, except for Bowyer.

Dalais et al. (1997a), and King and Bignell (2000) reported the levels of isoflavones

in some soya bean varieties from Australia. Dalais et al. (1997a) did not analyse

Bowyer soya beans. They analysed daidzin, genistin, glycitin and their aglycones

levels in 15 different strains of soya beans. King and Bignell (2000) analysed

daidzin, genistin, glycitin and their acetyl, malonyl and aglycone forms, and total

aglycones in six different cultivars of soya beans. Dalais et al. (1997a) and King and

Bignell (2000) did not specify the variety of soya beans they analysed. Proper

comparison with their studies cannot be made for this reason, however, the levels of

total daidzein and genistein in the soya bean varieties analysed in this study were

higher than those analysed by King and Bignell (2000).

Levels of daidzein and genistein in dried soya beans analysed in this study were

considered higher than those in dried soya beans from the USA, Hawaii, Japan and

Singapore as listed in Table 5.2 (overleaf) (Murphy 1982; Wang & Murphy 1994a,

1994b; Franke et al. 1995, 1999; Mazur et al. 1998; Futukake et al. 1996). Mazur et al. (1998) also found coumestrol (0.8-0.18 mg/100 g), formononetin (0.04-0.12 mg/100 g) and biochanin A (0.0-0.02 mg/100 g), all on a wet weight basis, in some soya bean varieties from the USA. The differences in the isoflavones and coumestrol content in soya beans reported by this study and others, could be due to the different analytical methods used and to the variety of soya beans analysed. This study, and 170

that of other authors, used a range of HPLC methods (mostly gradient), while Mazur

et al. (1998) used the ID/GC/MS/SIM method.

Table 5.2. Levels of daidzein, genistein, formononetin, biochanin A and coumestrol in soya beans (mg/100 g, wet weight basis).

Soya bean Daidzein Genistein Coumestrol Formononetin Biochanin Authors A

Dried 91.3 76.3 ND ND ND Franke et (Singapore) al. (1999) Fresh, pods 12.2 14.4 ND ND ND Frankeet (Hawaii) al. (1999) Five varieties, 0.3-1.0 0.8-1.1 NA NA NA Dalais et dried al. (1997a) (Australia) Four varieties, 7.3-66.9 26.4-87.9 NA NA NA King& dried Bignell (Australia) (2000) Four varieties, 10.5-56.0 26.8-84.1 0.0-0.18 0.04-0.12 0.0-0.02 Mazuret dried (USA) al. (1998) Two varieties, 0.1-2.2 2.2-4.0 NA NA NA Murphy dried (USA) (1982) Two varieties, 0.7-2.6 1.7-2.9 NA NA NA Wang& dried (USA) Murphy (1994a) Seven 0.4-3.8 1.5-4.5 NA NA NA Wang& varieties, dried Murphy (USA) (1994b) Three tr-0.4 0.8-0.9 NA NA NA Wang& varieties, dried Murphy (Japan) (1994b)

NA = not analysed; ND =not detected; tr =trace

Fresh soya beans also contained only daidzein and genistein among the analytes studied (Table 5.3, overleaf). The levels of daidzein and genistein in fresh soya beans bought in Indonesia were similar to those bought in Australia (p>0.05). However, the fresh soya beans purchased in Australia were frozen products imported from China. 171

Table 5.3. Daidzein and genistein levels in fresh soya beans purchased in Indonesia and Australia (mg/100 g, wet and dry weight basis).

Foods Basis % Daidzein Genistein Water

Soya beans, fresh (Indonesia) Wetwt 67.0 19.8 ± 8.7 7.6±2.2 Drywt 58.0 ± 18.4 23.1 ±4.9 Soya beans, fresh, frozen Wetwt 66.3 14.3±0.9 8.9 ± 0.6 (Australia, imported from Drywt 42.4±2.8 26.4 ± 1.7 China) Soya beans, fresh, from Asian Wetwt 64.3 9.0 9.2 store in Hawaii (Franke et al. Drywt 25.2 25.7 1995) Soya beans, fresh, frozen, from Wetwt 61.8 28.2 31.5 Taiwan (Franke et al. 1995) Drywt 73.9 82.6

Values are means± SD of duplicate analyses of three purchases.

Although both fresh and dried soya beans were found to contain daidzein and genistein, their levels were different when considered on a dry weight basis. Fresh soya beans obtained from fresh pods in Indonesia contained daidzein and genistein at levels 25%-50% lower than those in dried soya beans. The levels were 140.0 mg/

100 g for daidzein and 91.4 mg/100 g for genistein, respectively, in dried soya beans

(Table 5.1, page 168) and 60.0 mg/100 g for daidzein and 23.1 mg/100 g for genistein, respectively, in fresh soya beans. The differences may be due to the maturation stage since phytoestrogen levels increase with germination or maturation of seeds (Kudou et al. 1991).

Fresh soya beans bought in Australia were imported from China as frozen products.

According to Smith et al. (1986) frozen storage does not affect the total 172 concentration of isoflavones. They found that levels of isoflavones in subterranean clover did not change after frozen storage. The levels of daidzein and genistein in fresh soya beans from Indonesia and Australia (imported from China) were lower than those from Taiwan (Franke et al. 1995) but were higher than those from Hawaii

(Franke et al. 1995) (Table 5.3, previous page).

5.1.2. Canned soya beans

Canned soya beans from Master Foods (Wyong, NSW, Australia) contained daidzein and genistein at levels of 25.8 ± 2.3 mg/100 g and 18.1 ± 1.9 mg/100 g wet weight

(72.0% water). Canning was not studied directly, however, the levels of daidzein and genistein in canned soya beans were similar to the levels in dried soya beans from

McKenzie's (Australia) on a dry weight basis (p>0.05). It is assumed that the variety of soya beans used for canned soya beans was grown in Australia but the manufacturer did not reply to queries about this. The levels in dried vs canned soya beans were 106.3 ± 14.2 mg/100 g vs 92.1 ± 2.3 mg/100 g on a dry weight basis for daidzein, respectively, and 67.6 ± 14.9 mg/100 g vs 64.5 ± 7.0 mg/100 g on a dry weight basis for genistein, respectively.

5.1.3. Non fermented soya bean products

5.1.3.1. Soya milk

All soya milk products bought in Australia and Indonesia contained daidzein and genistein. Coumestrol was not detected in any soya milk products bought in Australia 173

(Table 5.4, overleaf), but was detected in two soya milk products bought in Indonesia

(Table 5.5, page 175). Statistical analysis revealed that the variation in the levels of

daidzein and genistein within brands was very low in most soya milk products except

for Soy Natural soya milk. However, analysis of variance showed a significant

difference in daidzein and genistein levels across the brands on a dry weight basis

(p<0.05).

The levels varied up to 2.5-fold. Soy Drink (Soya King) and Vitasoy (Vitasoy

International) contained the highest levels of daidzein and (Sanitarium) the lowest on a dry weight basis. Instant soya powder from Nestle (imported from

Indonesia) contained the highest levels of genistein whilst Vitalife (Natural Foods) contained the lowest.

Levels of daidzein, genistein and coumestrol were significantly different between soya milk products purchased in Indonesia on a dry weight basis (p<0.05) as shown in Table 5.5. Susu Kedelai Mony from Salim Graha had the highest daidzein and genistein levels. Susu Kedelai Mony (Indonesia) and Soya Bean Milk, Yeo's

(Singapore) were found to contain coumestrol at levels of 8.0 mg/100 g and

3.3 mg/100 g on a wet weight basis, respectively. These levels would be of biological significance for consumers given the high estrogenic activity of coumestrol. 174

Table 5.4. Daidzein, genistein and coumestrol levels in soya milk products from Australia (mg/100 g, wet and dry weight basis).

Foods Ingredient Basis % Daidzein Genistein Coumestrol Water

Soya milk So Good (Sanitarium) Soya isolate Wetwt 87.1 2.3±0.2 2.9 ± 0.1 ND Drywt 17.8 ± 1.8f 22.7 ± 0.8r So Good Lite Soya isolate Wetwt 90.8 2.8±0.5 3.1 ± 0.2 ND (Sanitarium) Drywt 30.9 ±5.7d 34.2±2.4d Good Life (Berrivale Soya isolate Wetwt 89.2 3.3 ±0.3 2.4 ± 0.1 ND Orchards Ltd) Drywt 30.9 ±2.5d 22.7 ± 0.9r Soy Drink (No Frills) Soya isolate Wetwt 92.5 1.6 ± 0.2 2.7 ±0.3 ND Drywt 26.0±2.3e 35.6±3.7d Soy Drink (Sungold) Soya isolate Wetwt 89.1 4.8±0.7 3.0± 0.3 ND Drywt 44.1 ± 5.9c 27.6±2S Vitalife (Natural Whole soya Wetwt 87.5 2.9±0.2 2.2±0.2 ND Foods) beans Drywt 22.9 ± 1.4"f 17.7 ± 1.6g Vitasoy (Vitasoy Whole soya Wetwt 93.5 3.6±0.7 3.8±0.6 ND International) beans Drywt 55.2± 10.7a 59.1 ±8.7b Nature's (Pureharvest) Whole soya Wetwt 86.4 4.5±0.6 3.5±0.5 ND beans Drywt 32.9 ± 4.4d 25.8± 3.4cf So Natural (Natural Whole soya Wetwt 87.0 5.7 ± 2.0 3.8± 1.4 ND Foods) beans Drywt 44.1 ± 15.6c 28.9 ± 10.8" Soya Drink (Soya Whole soya Wetwt 93.8 3.6±0.2 3.0± 0.1 ND King) beans Drywt 58.8 ± 3.3a 49.1 ± 1.4c Instant soya powder Instant Soy Powder Soya isolate Wetwt 3.7 50.8±4.2 76.4±6.5 ND (Carnation, Nestle, Drywt 52.8± 4.4b 79.3 ± 6.8a Indonesia)

Values are means± SD of duplicate analyses of three purchases. Dissimilar superscripts (a=highest) within column indicate significant differences (p<0.05). ND=not detected 175

Raw materials are factors contributing to the variation of daidzein, genistein, and

coumestrol in levels in soya milk products. Traditionally soya milk is made from

whole soya beans by soaking the whole beans in water overnight, then washing and

grinding them with fresh water. The slurry is then filtered through a cloth and the

residue, known as soya pulp or , is separated. The filtrate is boiled and cooled

for a few minutes before serving. Soya protein isolate is also used to produce soya

milk. Soya protein isolates are principally prepared by defatting soya meal using

aqueous or mild alkali extraction (pH 7-10) of protein and soluble carbohydrates.

The insoluble residue, mostly carbohydrate, is then removed by centrifugation, followed by precipitation of soya protein at its isoelectric point (pH 4.5).

Table 5.5. Daidzein, genistein and coumestrol levels in soya milk products from Indonesia (mg/100 g, wet and dry weight basis).

Foods Ingredient Basis % Daidzein Genistein Coumestrol Water

Susu Kedelai Mony Soya Wetwt 90.2 2.1 ± 0.3 2.9 ± 0.3 8.0± 1.0 (Salim Graha) isolates Drywt 21.6± 2.5" 29.7 ±3.2" 81.6 ± 10.1°

Soya Bean Milk (Yeo's, Soya Wetwt 86.3 1.9 ± 0.1 3.2±0.2 3.3 ± 0.3 Singapore) isolates Drywt 14.6± l.Ob 23.2± 1.4b 42.6 ± 2.3 b

Susu kedelai (unbranded) Whole Wetwt 86.9 2.8±0.4 1.7 ± 0.3 ND soya bean Drywt 21.3 ± 3.5" 13.4 ± 2.0c ND

Values are means± SD of duplicate analyses of three purchases. Dissimilar superscripts (a=highest) within column indicate significant differences (p<0.05). ND=not detected.

The precipitated protein is separated by mechanical decanting, washing and neutralising to about pH 6.8 and then spray-drying (Waggle et al. 1989). According to Wang and Murphy (1996), the retention and distribution of isoflavone isomers 176 changes from the initial soya flour to the final product (soya protein isolates), and the overall concentration of total isoflavones in the final product is lower than in the initial soya bean flour.

Soya protein isolate and whole soya beans are used to produce soya milk in

Australia. This study indicated that the starting material did not affect the levels of daidzein and genistein in soya milk products on a dry weight basis (p>0.05). This could be due to the high variability between purchases of the same products. The mean levels of daidzein from soya bean milk and soya isolate milk products were

43.4 ± 15.7 mg/100 g and 30.2 ± 9.2 mg/100 g on a wet weight basis, respectively.

The mean levels of genistein from soya bean milk and soya isolate milk products were 36.1 ± 17.3 mg/100 g and 28.6 ± 6.1 mg/100 g on a wet weight basis, respectively.

The process used to manufacture soya milk could be a factor contributing to the differences in isoflavones and coumestrol levels. A number of new methods or techniques for soya milk preparation has been developed over the past several decades to minimize off-flavour and colour defects in soya milk. Daidzein and genistein are two of the compounds in raw soya beans contributing to off-flavour and colour defects in soya milk, which are undesirable (Okubo et al. 1992). Matsuura et al. (1989) reported that the intensity of off- flavor in soya milk paralleled the concentration of isoflavone aglucones daidzein and genistein, formed by the hydrolytic action of ~-glucosidases on glucosidic isoflavone precursors 177

Daidzein and genistein are formed in soya milk by the hydrolytic action of 13-

glucosidases on isoflavone glucoside precursors. Processes are being developed to maximise J3-glucosidases inhibition. Soaking during the production of soya milk

could enhance the conversion of daidzin and genistin to their aglycones, thus resulting in a more unpleasant flavour in soya milk. Heat treatments employed during the production of soya milk inactivate lipoxygenase (80°C-154°C), eliminates beany flavour (80°C) and sterilises the products (121 °C). Ha et al. (1992) found that soaking soya beans in boiling 0.25% sodium bicarbonate (NaHC03) could reduce the formation of daidzein and genistein by up to 22 times compared to soaking in distilled water at 50°C. These treatments could contribute to the lowering of the original levels of daidzein and genistein in soya milk products.

According to Wang and Murphy (1996), soya milk production did not lead to losses in isoflavones. The distribution of isoflavones of the raw material is into the okara

(12%), whey (44%) and a little in the soaking water (0.5%). Heat treatment did not destroy the isoflavones but changed their profile. Heat treatment during the cooking of the soya slurry converted the malonyl glucoside forms from the raw soya beans to their corresponding aglycones. Thus, malonyl glucoside forms decrease in concentration but the aglycones daidzein, genistein, and glycitein and the glucosides daidzin and genistin increase (Kudou et al. 1991; Wang & Murphy 1996). These authors described a traditional process for making soya milk. This did not use any treatment which could have inhibited the formation of free isoflavones through hydrolytic action of J3-glucosidases in raw soya beans. 178

The mean levels of daidzein and genistein in soya milk from Australia and Indonesia

were similar on a wet weight basis (p>0.05). The levels were 3.4 ± 1.2 mg/100 g and

2.3 ± 0.5 mg/100 g on a wet weight basis for daidzein, respectively, and 2.9 ± 0.8

mg/100 g and 2.6 ± 0.8 mg/100 g on a wet weight basis for genistein, respectively.

Extraction of soya milk with acid at 100°C during the analysis would have

hydrolysed the conjugate forms in soya milk, thus increasing the levels of total

daidzein and genistein.

Daidzein and genistein levels in So Good, So Natural, and Vitasoy analysed in this

study compared well with those reported for the same Australian products by Knight

et al. (1998). Levels of genistein in soya milk products (brands not specified) from

Australia reported by King and Bignell (2000) were considerably higher than those reported by this study, while levels of daidzein were lower (on a wet weight basis).

Levels of daidzein and genistein in soya milk products (brands not specified) reported by Dalais et al. (1997b) were lower than those found in this study on a wet weight basis. The method of analysis for identification of isoflavones in soya milk used by this study and the other studies was HPLC. However, the elution system used was different. King and Bignell (2000) used the gradient HPLC method with

15% acetonitrile in 0.1% trifluoroacetic acid-acetonitrile as eluent, Dalais et al.

(1997b) did not describe the method they used, while Knight et al. (1998) used the gradient HPLC method with 23% acetonitrile in 10% acetic acid as eluent developed by Franke et al. (1995) (see Table 2.1, page 27).

Soya milk products from Australia and Indonesia had greater daidzein and genistein levels (on a wet weight basis) compared to those from Ohio (Coward et al. 1993), 179

Minnesota (Dwyer et al. 1994), Singapore and Hawaii (Franke et al. 1999). The

levels compared well to those in products from the United States (Murphy et al.

1999) when considered on a dry weight basis.

5.1.3.2. Tofu

Tables 5.6 and 5.7 (pages 180 & 181) list the tofu products from Australia and

Indonesia, respectively, analysed in the present study. All tofu products were found

to contain daidzein and genistein. The levels (on a dry weight basis) were

significantly different (p<0.05) across brands purchased in Australia. The variation in

the levels was up to six-fold. Silken Firm Tofu from Joyce had the highest levels of daidzein and genistein, while Firm Tofu from Joyce and Organic Tofu from Earth

Star Foods had the lowest.

The levels of daidzein and genistein in tofu products from Indonesia were also different across brands on a dry weight basis (p<0.05). The variation in the levels was up to two-fold. Silken tofu from Kong Kee and Tahu Tau Kwa from Miko Sejati had the highest levels of daidzein and genistein while Silken Tofu Sakura from

Harum Sari Food had the lowest.

The mean levels of daidzein and genistein in tofu products bought in Indonesia were similar to those bought in Australia on a dry weight basis (p>0.05). The mean levels in tofu products from Indonesia and Australia were 51.8 ± 13.0 mg/100 g and 53.9 ±

40.4 mg/100 g for daidzein, respectively, and 56.9 ± 19.5 mg/100 g and 44.9 ± 24.2 mg/1 00 g for genistein, respectively. 180

Table 5.6. Daidzein and genistein levels in tofu products from Australia (mg/100 g, wet and dry weight basis).

Foods Basis % Daidzein Genistein Water

Hard Tofu (Soya King) Wetwt 73.6 14.1±0.6 14.3±0.8 Drywt 53.2±2.2b 54.1 ± 3.1c

Silken Firm Tofu (Joyce) Wetwt 87.0 16.6±2.0 10.3 ±2.2 Drywt 128.4 ± 14.2a 78.9 ± 15.5b

Silken Tofu (Soya King) Wetwt 89.4 13.7 ± 0.6 9.7 ±0.7 Drywt 129.4 ±5.6a 91.9 ± 6.6a

Smoked Tofu (Blue Lotus Wetwt 70.5 7.5 ± 0.5 5.6±0.5 Foods) Drywt 25.4 ± 1.6e 18.8 ± 1.8r

Firm Tofu (Joyce) Wetwt 84.2 6.0 ±0.4 5.6±0.8 Drywt 38.3 ± 2.1cd 35.7 ± 4.5e

Tofu with Tempeh (Nutrisoy) Wetwt 73.1 9.8 ± 0.6 11.2 ± 1.4 Drywt 36.4 ± 2.6d 41.0 ± 3.1d Tofu (Blue Lotus Foods) Wetwt 77.6 6.7 ± 0.7 7.5 ± 1.0 Drywt 29.9 ± 3.3e 33.4±4.6e

Tofu Cutlets (Blue Lotus Wetwt 68.3 11.0 ± 0.3 12.9 ±0.7 Foods) Drywt 34.0 ±2.5d 40.4 ± 2.0e

Nigari Tofu (Joyce) Wetwt 71.2 12.5 ± 1.3 11.7 ± 1.1 Drywt 43.5 ± 5.2b 40.6 ± 5.1d Organic Tofu (Earth Star Wetwt 71.5 6.0 ± 0.3 4.2±0.3 Foods) Drywt 21.2± 0.9e 14.8 ± 0.9g

Values are means± SD of duplicate analyses of three purchases. Dissimilar superscripts (a=highest) within column indicate significant differences on a dry weight basis (p<0.05). 181

Table 5.7. Daidzein and genistein levels in tofu products from Indonesia (mg/100 g, wet and dry weight basis).

Foods Basis % Daidzein Genistein Water

Tofu (unbranded) Wetwt 77.0 10.3±0.8 12.2±2.7 Drywt 44.6±4.5bc 53.4±13.6b

Silken Tofu (Kong Kee) Wetwt 83.8 11.0 ± 0.7 13.6 ± 0.8 Drywt 68.0±4.3a 84.0±6.5a Tahu Tau Kwa (Miko Sejati) Wetwt 74.8 15.2±2.3 15.2± 3.1 Drywt 62.4_±9.5a 64.5±13.0b Tofu Sakake (Mitra Boga Segar) Wetwt 82.5 8.2±0.0 9.1 ± 0.1 Drywt 47.0±0.0b 51.9±0.8c Silken Tofu Sakura (Harum Sari Wetwt 82.0 6.6±0.2 5.5 ± 1.2 Food) Drywt 36.8±1.4c 30.5±6.7d

Values are means± SD of duplicate analyses of three purchases. Dissimilar superscripts (a=highest) within column indicate significant differences on a dry weight basis (p<0.05).

The variation of daidzein and genistein levels in tofu would have been due to the variety of soya beans used and to the processing method. The procedures to produce tofu are similar to those used to produce soya milk, and only a single step is needed after the production of soya milk to produce tofu. Soya milk is heated before the addition of a coagulant (calcium salts) to form a curd. The curd is then pressed to separate whey from tofu (Liu 1997). This study found that the levels of daidzein and genistein in tofu were about double those in soya milk if the data were expressed on a dry weight basis (p<0.05). The levels in tofu and soya milk were 53.2 ± 33.2 mg/

100 g and 33.9 ± 14.5 mg/100 g for daidzein, respectively, and 47.2 ± 20.9 mg/100 g and 33.3 ± 18.0 mg/100 g for genistein, respectively. 182

According to Wang and Murphy (1996), the coagulation step causes the largest

losses of isoflavones in tofu. The coagulant (calcium salts) reacts with the various

proteins in soya milk and precipitates them to form curds. Some protein-associated

isoflavones are released into the whey, thus tofu contains only 33% (based on dry

basis) of the total isoflavones in the starting material, raw soya beans, meaning that

the levels of isoflavones in tofu should be lower than in soya milk. Therefore, if·was

about 67% from the total isoflavones in the starting raw soybeans were discarded in the processing liquid. However, the findings in this study are not consistent with the findings of Wang and Murphy (1996). A direct comparison of isoflavone levels between soya milk and tofu could not be made in this study because commercial products were analysed in which the variety of soya beans and processing techniques used were unknown.

The findings of this study also differed from those of Dwyer et al. (1994) who found similar levels of isoflavones in both commercial soya milk and tofu. Dwyer et al.

(1994) used an aseptic style of tofu, where the commercial tofu is coagulated in the package with 8-gluconolactone and no whey is removed. Wang and Murphy (1996) used a traditional type of tofu, where the whey is discarded after curd production to produce the product. This study analysed only the tofu with the liquid in the package discarded. These manufacturing differences would have contributed to the differences in isoflavones levels observed in the three different studies.

There is no consistent pattern in the levels of isoflavones in tofu reported by different authors (see Appendix 1). Comparisons between studies are difficult since the food brands and methods of analyses differed. Tofu products bought in Australia and 183

Indonesia had higher daidzein levels but lower genistein levels than those reported

by Murphy et al. (1999) when compared on a dry weight basis. The levels of

daidzein and genistein in these tofu samples from Australia and Indonesia were

lower than those from Singapore and Hawaii (Franke et al. 1999).

5.1.4. Dried soya milk curd or kembang tabu

Two different dried soya milk curd products purchased in Indonesia were analysed in this study. These were Kembang Tabu from Pelita, Indonesia and Dried Bean Curd imported from China. The variability of the levels of isoflavones for the triplicate samples was much lower for dried soya milk curd from Pelita (Indonesia) than for the product imported from China. The levels of daidzein and genistein in the two products were not significantly different (p>0.05) as shown in Table 5.8.

Table 5.8. Daidzein and genistein levels in dried soya milk curd products from Indonesia (mg/1 00 g, wet and dry weight basis).

Food Basis % Daidzein Genistein Water

Dried soya milk curd Wetwt 6.9 43.1 ± 6.2 41.6±3.4 (Pelita, Indonesia) Drywt 46.3±7.2 44.6±3.6

Dried soya milk curd (China Wetwt 6.5 38.6 ± 13.6 43.3 ± 19.7 brand, China) Drywt 41.3 ± 14.5 46.3 ± 21.0

Values are means ± SD of duplicate analyses of three purchases. 184

Dried soya milk curd is made by heating soya milk in a flat open pan to near boiling

point (about 80 to 90°C). A curd gradually forms on the liquid surface. After the curd has stiffened it can be lifted with sticks or by passing a rod underneath it, then dried by hanging on a line or spreading on galvanized wire mesh. The curd is removed continuously from the surface of soya milk until no further film formation occurs.

The level of daidzein was not statistically different between soya milk, dried soya milk curd and tofu (p>0.05), but the level of genistein was significantly different

(p<0.05) on a dry weight basis. The levels in dried soya milk curd, soya milk and tofu were 45.3 ± 11.4 mg/100 g, 33.9 ± 14.5 mg/100 g and 53.2 ± 33.2 mg/100 g for daidzein, respectively, and 45.5 ± 14.5 mg/100 g, 33.3 ± 18.0 mg/100 g and

47.2 ± 20.9 mg/100 g for genistein, respectively. The varieties of soya beans used to produce soya milk, tofu, and dried soya milk curd were unknown. Therefore, it is difficult to identify the source of the differences. This dried soya milk curd product does not appear to have been analysed before, therefore comparative data are not available.

5.1.5. Fermented soya bean products

5.1.5.1. Tempeh

Tempeh is a fermented soya bean food, which is indigenous to Indonesia. The levels of daidzein and genistein in tempeh were significantly different between the products of the two countries on a dry weight basis (p<0.05). Tempeh purchased in Indonesia contained 4.5 times more daidzein and 2. 7 times more genistein than tempeh 185

purchased in Australia on a dry weight basis (Table 5.9, overleaf). The differences in

the isoflavones levels would have been due to the variety of soya beans used and

processing factors. Variation of fermentation times and temperature leads to changes

in isoflavone levels in tempeh (Wang & Murphy 1996; Ikeda et al. 1995).

Traditionally, tempeh is made by cleaning dried soya beans and then cooking them in

boiling water for 30 minutes. The hulls are removed and the dehulled beans are

soaked overnight to allow full hydration and lactic acid fermentation. The soaked,

dehulled beans are cooked again for 60 minutes, drained using woven bamboo

baskets, and spread on a flat surface for cooling to room temperature. The cooled

treated beans are then inoculated with a starter culture of Rhizopus oligosporus or an

inoculum from a previous batch, wrapped in banana leaves or in perforated plastic

bags. Fermentation is allowed to occur at room temperature for up to 18 hours, or

until the beans are bound by a white mycelium.

According to Wang and Murphy (1996), removal of seed coats of soya beans in

tempeh manufacture caused 1.4% loss of isoflavones, but cooking caused a total loss

of isoflavones. The cooking water was not incorporated into the product but was discarded, thus the isoflavones leached into the cooking water and were lost.

Fermentation in tempeh manufacture does not cause a significant loss of isoflavones, but generates a different distribution of isoflavones. The hydrolytic action of ~­ glucosidase derived from the fungi liberated aglycones from glucosides, thus enhancing the levels of daidzein and genistein, but decreasing the levels of the corresponding glucoside forms. 186

Table 5.9. Daidzein and genistein levels in tempeh products from Indonesia (mg/ 100 g, wet and dry weight basis).

Food Basis %Water Daidzein Genistein

Tempeh (unbranded, Wetwt 70.9 21.5 ± 3.4 24.5 ±3.3 Indonesia) Drywt 74.5 ± 12.1a 85.4±16.1a Tempeh (Nutrisoy, Wetwt 73.2 4.7 ±0.6 9.0 ± 1.2 Australia) Drywt 17.6 ± 2.2b 33.3 ±4.6b

Values are means± SD of duplicate analyses of three purchases. Dissimilar superscripts (a=highest) within column and between countries indicate significant differences on a dry weight basis (p<0.05).

The levels of daidzein and genistein in tempeh from Indonesia are comparable to the

average levels in tofu from seven cities in the United States as described by Murphy et al. (1999), but higher than those in tofu from Iowa, United States as reported by

Wang and Murphy (1994b). However, the levels in tempeh from Australia were lower than those reported by Murphy et al. (1999) and Wang and Murphy (1994b).

The specific varieties of soya beans used to produce these tempeh samples are not known; therefore, differences in the isoflavones levels could not be interpreted.

5.1.5.2. Fermented soya paste

Soya paste is one of the most important fermented Oriental soya foods. It is commonly known as jiang (Mandarin) or chiang (Cantonese) in China; miso in

Japan; jang in Korea; taucho in Indonesia; and taotsi in the Philippines. Six different taucho products purchased in Indonesia were found to contain only daidzein and genistein of the five compounds studied (Table 5.10, 188). The levels showed little 187 variation within brands, except for the unbranded taucho. However, analysis of variance showed significant differences across brands (p<0.05). Traditional

Indonesian taucho, Salted Soy Bean Soja Sale (Yeo's, Malaysia) and Taucho Medan

Mekar (Kujang) were found to contain the highest daidzein levels, while Taucho

No.1 Macan (Pulau Seribu) had the lowest. The highest levels of genistein were in traditional Indonesian taucho and Salted Soy Bean Soja Sale (Yeo's, Malaysia) and the lowest in Taucho Asli No.1 (Gajah Dua) and Taucho No.1 Macan (Pulau Seribu).

Traditional taucho (unbranded) contained more daidzein and genistein than commercial products (Table 5.10, 188), probably because it had undergone more moderate treatment during processing, and less addition of other foods. Traditionally, soya paste or taucho is prepared by cleaning whole soya beans, crushing and soaking them for 3 to 5 hours and then draining. Beans are then steamed for 40 minutes and cooled before mixing with koji (0.3-0.5% ). Koji is produced from washed, soaked, cooked, and cooled which has been inoculated with a mixture of strains of

Aspergillus oryzae, Aspergillus soyae, Pediacoccus halophylus, or Saccharomyces rouxii and incubated. The inoculated mixture of soya beans is then fermented for several days until ripe. The ripe products are blended and mashed. 188

Table 5.10. Daidzein and genistein levels in fermented soya paste (taucho) products from Indonesia (mg/100 g, wet and dry weight basis).

Food Basis % Daidzein Genistein Water

Taucho (unbranded, Indonesia) Wetwt 63.7 27.6±7.5 23.4 ± 3.2 Drywt 76.7 ±22.5a 64.6 ± 10.4a

Salted Soy Bean Soja Sale (Yeo's, Wetwt 62.7 26.8 ± 1.3 23.7 ± 1.5 Malaysia) Drywt 71.9 ± 2.4a 63.5 ± 2.8a

Taucho Asli No.1 (Gajah Dua, Wetwt 60.2 19.7 ± 1.3 11.4± 2.2 Medan, Indonesia) Drywt 49.5 + 3.3b 28.8± 5.8c

Taucho Medan Mekar (Kujang, Wetwt 68.5 23.1 ±2.3 17.7 ± 1.7 Indonesia) Drywt 73.3 ± 6.3a 56.1 ± 4.8b

Taucho Medan Harum Sedap Wetwt 60.4 20.4± 2.8 17.9 ±2.9 (Medan, Indonesia) Drywt 51.5 ± 6.8b 45.3 ± 6.9b

Taucho N o.l Macan (Pulau Wetwt 59.5 12.8 ± 3.2 10.1 ± 1.4 Seribu, Indonesia) Drywt 31.5 ± 7.7c 24.8±3.lc

Values are means± SD of duplicate analyses of three purchases. Dissimilar superscripts (a=highest) within column indicate significant differences on a dry weight basis (p<0.05).

Aglycones are the major form of isoflavones in soya bean paste as a result of the hydrolysis process occurring during fermentation (Wang & Murphy 1994a).

Different inocula provide different results from hydrolysis. Some malonyl isoflavones are converted to glucosides, and then to aglycones through the hydrolysis process by glucosidases in soya beans (Matsuura & Obata 1993). Raw material, inocula and processing could affect the levels of daidzein and genistein in soya bean paste. Taucho from Indonesia analysed in this study contained similar levels of 189

daidzein and genistein to those in miso from the United States (Wang & Murphy

1994b; Murphy et al. 1999).

5.1.5.3. Fermented soya cake or oncom

Oncom is one of the most popular fermented foods in West Java, Indonesia. Fresh

oncom contained daidzein and genistein at levels of 6.6 mg/100 g and 3.1 mg/100 g wet weight (74.4% water), respectively. The results from this study show that the levels of daidzein and genistein in oncom were about half those in tofu or in soya milk on dry weight basis. The levels in oncom, tofu and soya milk were 25.8 ±

3.2 mg/100 g, 53.2 ± 33.2 mg/100 g, and 33.9 ± 14.5 mg/100 g for daidzein, respectively, and 12.1 ± 1.6 mg/100 g, 47.2 ± 20.9 mg/100 g, and 33.3 ±

18.0 mg/100 g for genistein, respectively.

Oncom is made by the fermentation of residues from soya milk and tofu production using the mould Neurospora sitophila. Freshly prepared oncom is a cake-like product, with a pleasant flavour. It is covered with, and completely permeated with red to orange mycelium. Wang and Murphy (1996) conducted a controlled study of the levels of isoflavones at each processing step of soya milk and tofu production.

The isoflavone levels in the residues were 6% and 14% (on a dry weight basis) of daidzein and genistein, respectively, of the levels in soya milk and 26% and 35% (on a dry weight basis) of daidzein and genistein, respectively, of the levels in tofu. The findings in this study are not consistent with the findings of Wang and Murphy

(1996) because retail foods were analysed and not experimentally produced products. 190

5.1.5.4. Fermented sweet soya sauce (kecap manis) and unsweetened soya

sauce

This study was unable to identify any isoflavones or coumestrol in the four

fermented sweet soya sauces or kecap manis products produced by Indonesia (Table

5.11, overleaf). However, this study was able to identify daidzein and genistein in

unsweetened soya sauce from Kikkoman (Japan). Murphy (1982) was unable to

detect isoflavones in US soya sauce products, while Coward et al. (1993) did find

daidzein and genistein in unsweetened soya sauce from Kikk.oman (Japan) at levels

of 1.4 and 0.9 mg/100 g wet weight, respectively. The levels found in this study were higher than those in the same products as reported by Murphy et al. (1999) who found daidzein and genistein at 0.6 mg/100 g and 0.3 mg/100 g wet weight, respectively.

Soya sauce is a very poor source of isoflavones as the fermentation organisms have probably degraded them (Murphy 1982). Processes used in sweet soya sauce production would also result in further lowering of isoflavones content, because of addition of other ingredients, e.g., sugar, flour, fat and water. Kecap manis or sweetened soya sauce is a traditional soya food made by fermentation. Boiled, drained soya beans are mixed with a small amount of wheat flour. The mixture is then inoculated with Aspergillus sp. and fermented in sunlight for several weeks. The resulting cake is then squeezed to produce the dark brown sauce (Liu 1997). 191

Table 5.11. Daidzein and genistein levels in fermented sweet soya sauce (kecap manis) and unsweetened soya sauce products purchased in Indonesia (mg/100 g, wet weight basis).

Food % Daidzein Genistein Water

Kecap Manis ABC (ABC Central Foods, 22.5 ND ND Indonesia) Kecap Manis Bango (Rina Sari, Indonesia) 21.1 ND ND Kecap Manis Indofood (Indosentra Pelangi, 23.1 ND ND Indonesia) Kecap Manis Kedele (Aneka Food Tatarasa, 24.6 ND ND Indonesia) Dark (Amoy) 62.1 ND ND Kikkoman Soy Sauce (Japan) 71.2 2.3 ± 1.8 0.5 ±0.2

Values are means± SD of duplicate analyses of three purchases. ND= not detected

5.1.6. Second-generation soya foods

The levels of isoflavones in second-generation soya foods analysed were considerably lower than in the other soya products (Table 5.12, overleaf). This was anticipated because soya beans are only one of several ingredients used in their production. Second-generation soya foods are made by mixing soya ingredients with a wide variety of other ingredients. Soya beans maybe used as an animal protein replacement and to reduce fat content in foods. The addition of soya ingredients is limited because of the need to maintain similarity to the original foods. Statistical analysis was not performed on these products because of the variation in ingredient levels.

\_I 192

Table 5.12. Daidzein and genistein levels in second-generation soya food products from Australia (mg/100 g, wet weight basis).

Food % Daidzein Genistein Water

Soy Sausage Rolls (Blue Lotus Foods, 50.8 2.7 ±0.2 ND Australia)

Tofu Dessert (Nutrisoy, Australia) 81.4 6.7 ± 1.2 5.4± 0.9

Tofu Burger (Nutrisoy, Australia) 56.5 13.3 ± 1.8 14.5 ±2.4

Soy Cheese (Simply Better Foods, A~stralia) 49.5 ND 2.3±0.5

Values are means ± SD of duplicate analyses of three purchases. ND=not detected

Sausage Rolls were found to contain only daidzein at 2. 7 mg/1 00 g wet weight; genistein was not detected. Daidzein levels in this product were comparable with those in soya bacon, but lower than those in soya hot dogs (Wang & Murphy 1994a).

Soya cheese was found to contain only genistein at the level of 2.3 mg/1 00 g wet weight. Anderson and Wolf (1995) reported only trace amounts of daidzein and low levels of genistein in soya cheddar cheese, 0.4 to 0.9 mg/100 g on a wet weight basis.

Coward et al. (1993) also detected daidzein and genistein in soya cheese at low levels of 0.1 mg/100 g and 0.2 mg/100 g wet weight, respectively. Franke et al.

(1999) found 7.8 mg/100 g of daidzein and 8.8 mg /100 g of genistein wet weight in

Jalapeno Monterey soya cheese. Wang and Murphy (1994b) reported 0.2 and 3.4 mg/100 g of daidzein and 0.5 and 4.0 mg/100 g of genistein on a wet weight basis in two soya cheddar cheese products. 193

The mean levels of daidzein and genistein in tofu dessert were 6.7 mg/100 g and

5.4 mg/100 g wet weight, respectively. Tofu burgers contained daidzein and

genistein at levels of 13.3 mg/100 g and 14.5 mg/100 g on a wet weight basis,

respectively. This study found higher levels than those of Murphy et al. (1999) in

burger products (soya meat analogues) and in institutional soya-extended hamburgers

from the USA. They found the mean levels of daidzein and genistein were 0.6

mg/100 g and 0.9 mg/100 g, respectively, for burger (soya meat analogues) (64%

water), daidzein was not detected but genistein was at 0.04 mg/100 g in soya­

extended hamburgers (62% water) on a wet weight basis.

The levels of daidzein and genistein in tofu dessert and tofu burger were lower than

those in tofu because of increased water content. Tofu dessert is tofu mixed with fruit

juice, therefore it probably contains very little tofu. Levels of daidzein and genistein

were lower in tofu burgers than in tofu because they contain non-soya bean

constituents. They are made by adding flour, salt, spices and either oil or fat to soya

beans. Comparisons cannot be made as no data for these types of products were found in the literature.

5.1.6.1. Soya cereals

Of the six different Australian soya cereals analysed, two contained no isoflavones while the remaining four were found to contain only genistein among the analytes of interest (Table 5.13, overleaf). The highest level of genistein was found in Soy

Flakes from Lowan Foods, Australia. The genistein level was 79.9 mg/100 g wet weight basis. Soy-Tana's phytoestrogens content as analysed was much lower than 194 its label claim. Dilution with non-soya constituents may have resulted in the lowering of isoflavone levels in the product below detectability levels. Soya flour or soya flakes or soya isoflavones concentrate are used for making soya cereal products, however it is not known how much soya bean is used. King and Bignell (2000) found both daidzein and genistein in soya flakes from Australia at levels higher than those reported by this study. King and Bignell (2000) did not identify the brands of soya flakes they analysed. Therefore, comparison with their data is impossible.

Table 5.13. Daidzein and genistein levels in soya cereal products from Australia (mg/100 g, wet and dry weight basis).

Food Basis %Water Daidzein Genistein

Soy & Flakes (Norganic Wetwt 3.1 ND 38.7 ±9.3 Foods, Australia) Drywt 39.9 ± 9.6b Phytoestro gens-Soy Wetwt 9.1 ND ND (Sunsol, Australia) Drywt Soy-Tana (Specialty Wetwt 5.8 ND 3.8 ± 3.2 Cereals, Australia)* Drywt 4.0± 3.4c Hi Bran & Soy Weet-Bix Wetwt 4.0 ND ND (Sanitarium Health Food, Drywt Australia) Soy Flakes (Lowan Whole Wetwt 5.6 ND 75.2± 19.5 Foods, Australia) Drywt 79.7 ±20.6a Soy-Tasty (Sanitarium Wetwt 6.4 ND 28.2 + 31.4 Health Foods, Australia) Drywt 30.1 .± 33.5b

Values are means± SD of duplicate analyses of three purchases. Dissimilar superscripts (a=highest) within column indicate significant differences on a dry weight basis (p<0.05). ND=not detected * Label claim "one serve provides 20 mg of phytoestrogens" 195

5.1.6.2. Soya breads and muffins

All seven soya bread products purchased in Australia were found to contain

genistein, but only four were found to contain daidzein (Table 5.14, overleaf). The

levels were significantly different across the brands (p<0.05). Molenberg Soy &

Linseed Bread (Quality Bakers Australia) was found to contain the highest levels of

daidzein and genistein.

Soya bread is made by adding kibbled soya or a soya flour or a low fat granule soya

into a bread mix. The concentration of soya beans added was approximately 6 to 9%

of the total ingredients. The findings in this study were different from those for soya

and linseed bread from Australia as reported by King and Bignell (2000). They found daidzein at levels ranging from 2.1 to 7.0 mg/100 g and genistein from 5.8 to

11.9 mg/100 g wet weight. Genistein levels found by this study at 6.6 to 24.4 mg/

100 g were considerably higher than those found by King and Bignell (2000).

Insufficient information is given by King and Bignell (2000) about the brands of soya bread they analysed, therefore direct comparisons cannot be made with their findings. 196

Table 5.14. Daidzein and genistein levels in soya bread products from Australia (mg/100 g, wet and dry weight basis).

Food Basis %Water Daidzein Genistein

Soy & Linseed (Mobile Rise, Wetwt 38.0 1.3 ± 0.5 13.0 + 3.3 Australia) Drywt 2.1 ± 0.8c 22.0 ±5.4 b Soy & Linseed (Uncle Tobys, Wetwt 40 1.6 ± 0.2 20.8±3.0 Australia) Drywt 2.6± 0.2c 34.2±5.3a Soy & Linseed (Helga's, Wetwt 40.5 ND 6.6 ± 1.3 Australia) Drywt 11.1 ± 2.2c Soy & Linseed Molenberg Wetwt 42.2 16.5 ± 3.2 24.4 ± 7.1 (Quality Bakers Australia, Drywt 28.6 ±5.5a 42.1 ± 12.2a Australia) Soy & Linseed (Burgen, Wetwt 51.6 5.5 +2.3 7.2±0.6 Australia) Drywt 11.3 ±4.7b 14.9 ± 10.4c Soy & Linseed (Vogel, Wetwt 35.6 ND 11.0 ± 4.9 Australia) Drywt 17.0 ± 7.6c Soy & Linseed Muffin Wetwt 40.9 ND 3.5 + 1.2 (Buttercup, Australia) Drywt 6.0 .± 1.2d

Values are means± SD of duplicate analyses of three purchases. Dissimilar superscripts (a=highest) within column indicate significant differences on a dry weight basis (p

5.1. 7. Laboratory cooked whole soya beans

In the orient, whole dry mature soya beans are usually consumed after soaking and cooking in boiling water until tender. Fresh whole soya beans generally require approximately 10 minutes boiling before consumption. The levels of daidzein and genistein in fresh soya beans do not decrease significantly (p>0.05) after cooking with boiling water. The mean levels of daidzein and genistein in fresh soya beans were 50.2 ± 14.8 mg/100 g and 24.8 ± 5.0 mg/100 g on a dry weight basis, 197 respectively, and levels after cooking were 44.3 ± 14.3 mg/100 g and 23.4 ±

9.2 mg/100 g on a dry weight basis, respectively. Boiling reduced levels of daidzein and genistein in dried soya beans but not significantly (p>0.05). The levels of daidzein and genistein in dried soya beans were 112.0 ± 26.0 mg/100 g and 72.4 +

15.9 mg/100 g on a dry weight basis, respectively, and levels after cooking were 92.5

± 21.2 mg/100 g and 68.8 ± 14.2 mg/100 g on a dry weight basis, respectively.

Daidzein and genistein are considered to be very heat stable but malonyl isoflavones are heat-labile and very unstable. Heat treatment transforms some malonyl isoflavones into acetyl isoflavones, or into glucoside isoflavones (Farmakalidis &

Murphy 1985; Kudou et al. 1991). Levels of isoflavones might also be reduced by leaching into boiling water (Franke et al. 1999).

5.1.8. Laboratory fried tofu, tempeh, and oncom

Tofu, tempeh, and oncom can be served with virtually any other foods. They are served in soups or cooked with meat and/or vegetables. In Indonesia, people prefer to consume them as fried products. This study indicated that the levels of daidzein and genistein in tofu did not change after frying (p>0.05). The levels in fresh and fried tofu were 44.6 ± 4.5 mg/100 g and 52.7 ± 13.8 mg/100 g dry weight for daidzein, respectively, and 53.4 ± 13.6 mg/100 g and 40.0 ± 12.8 mg/100 g dry weight for genistein, respectively. The levels of daidzein and genistein in fried tofu purchased from street stalls in Indonesia (Table 5.15, overleaf) were not significantly different to those fried in the laboratory (p>0.05). 198

Fried tempeh has a nutty flavour, pleasant aroma, and a crunchy texture. This study

determined that deep-frying decreased the levels of daidzein and genistein in tempeh

by about 30% (p<0.05). The levels in fresh and fried tempeh were 74.5 ± 12.1

mg/100 g and 50.8 ± 5.4 mg/100 g dry weight for daidzein, respectively, and 85.4 ±

16.1 mg/100 g and 60.0 ± 10.7 mg/100 g dry weight for genistein, respectively.

Oncom is often served deep-fried as a main dish or meat substitute, or as a side dish cooked with other ingredients. This study identified that levels of daidzein and

genistein reduced considerably after frying (p<0.05). The levels in fresh and fried oncom were 25.8 ± 3.2 mg/100 g and 14.8 ± 1.9 mg/100 g for daidzein, respectively, and 12.1 ± 1.6 mg/100 g and 2.8 ± 1.9 mg/100 g for genistein, respectively.

Table 5.15. Daidzein and genistein levels in fried tofu, tempeh and oncom (mg/ 100 g, wet weight basis).

Food %Water Daidzein Genistein

Laboratory fried tofu 62.8 18.0 ± 5.5 14.5 ±4.0

Laboratory fried tempeh 35.1 32.9 ±4.3 39.9 ± 7.6

Laboratory fried oncom 62.4 5.5 ± 1.6 1.0 ± 0.6

Street fried tofu (unbranded, 67.1 17.2 ± 13.1 19.9 ± 16.9 Indonesia)

Values are means± SD of duplicate analyses of three purchases. 199

Isoflavones are known to possess an antioxidative activity, therefore tofu is remarkably stable to lipid oxidation (Murakami et al. 1984). The levels of daidzein

and genistein in tempeh and oncom decreased after frying, thus they appear to be

degraded during the frying process. An explanation for this is that the heat applied during frying in oil causes decarboxylation of the compounds. This study differs from the study by Murphy et al. (1999) who observed no differences in the levels between raw and cooked tempeh.

5.2. Other legumes and legume products

5.2.1. Berlotti beans (Phaseolus vulgaris)

Dried berlotti beans purchased in Australia and cooked berlotti beans did not contain any isoflavones or coumestrol. Fresh berlotti beans purchased in Australia were found to contain formononetin at a low level (Table 5.16, overleaf). Comparison cannot be made because no other study has been reported for these beans.

5.2.2. Black beans (Vigna unguiculata)

Two black bean products, canned Black Beans (S&W Fine Foods) imported from the

USA to Indonesia and bottled salted black beans (Amoy) imported from Hong Kong to Indonesia were analysed. Only canned Salted Black Beans (Amoy) were found to contain daidzein and genistein at levels of 17.0 mg/100 g and 37.2 mg/100 g on a wet weight basis, respectively (Table 5.16 overleaf). Franke et al. (1995) could not detect 200

any isoflavones and coumestrol in dried black beans. Salted Black Beans are made

by fermentation of black beans using Aspergillus sp. The fermentation process may

assist the formation of the conjugate compounds in the raw black beans.

Table 5.16. Daidzein, genistein, coumestrol, formononetin and biochanin A in berlotti beans, black beans and black eye beans purchased in Australia and Indonesia (mg/100 g, wet weight basis).

Foods % Daidzein Genistein Coumestrol Formononetin Biochanin Water A

Berlotti beans Australia Berlotti beans, dried ND ND (McKenzie's) 10.3 ND ND ND Berlotti beans, fresh 59.0 ND ND ND 0.07 ±0.00 ND (unbranded) Black beans Indonesia Canned Black Beans 69.7 ND ND ND ND ND (S&W) Bottled Salted Black 57.4 17.0±4.9 37.2±6.4 ND ND ND Beans (Amoy) Black eye beans Australia

Black eye beans, 10.5 ND ND ND ND ND dried (McKenzie's)

Values are means± SD of duplicates analyses of three purchases ND=not detected 201

5.2.3. Black eye beans (Vigna unguiculata)

Dried and cooked black eye beans did not contain any isoflavones or coumestrol

(Table 5.16, previous page). However, Mazur et al. (1998) found daidzein, genistein,

formononetin, biochanin A and coumestrol in black eye beans at levels of 0.02 mg/100 g, 0.01 mg/100 g, 0.005 mg/100 g. 0.007 mg/100 g and 0.007 mg/100 g on a

wet weight basis, respectively. The differences in the findings of this study and that

of Mazur et al. (1998) might have been due to variety and growth conditions. The isocratic HPLC method used in this study could not identify isoflavones or coumestrol in black eye beans but they could be identified by the ID/GC/MS/SIM method used by Mazur et al. (1998). The detection limits of the isocratic HPLC method used in this study are 0.001 mg/100 g for daidzein, 0.002 mg/100 g for coumestrol, 0.002 mg/100 g for genistein, 0.002 mg/100 g for formononetin and

0.006 mg/100 g for biochanin A (see page 139). Levels of isoflavones and coumestrol present in these samples of black eye beans might have been below the detectability of this method.

5.2.4. Broad beans (Viciafaba)

Broad beans have been reported to contain daidzein, formononetin and biochanin A as shown in Table 5.17 (overleaf) (Mazur et al. 1998). However, this study did not identify any isoflavones or coumestrol in broad beans (Table 5.18, page 201).

Similarly, Futukake et al. (1996) also did not detect genistein in broad beans. Franke et al. (1995) found genistein and formononetin in fried broad beans from the USA

(Table 5.17), while Mazur et al. (1998) found daidzein and formononetin, and trace 202

amounts of genistein and biochanin A in two varieties of broad beans from the USA

(Table 5.17).

Variety, growth conditions and especially the method of analysis may have affected

the levels found for isoflavones and coumestrol in broad beans. Levels are low and may be below the detectability level of the isocratic HPLC method used in this study.

Table 5.17. Daidzein, genistein, coumestrol, formononetin and biochanin A in dried broad beans purchased in Australia, in broad beans grown in the USA, and canellini beans purchased in Australia (mg/100 g, wet weight basis).

Foods % Daidzein Genistein Coumestrol Formononetin Biochanin Water A

Broad beans

Australia Broad beans, dried 10.1 ND ND ND ND ND (McKenzie's) USA Broad beans, dried, NR ND 1.29 ND 0.021 ND fried (Franke et al. 1995) "Herz Freya" broad NR 0.016 tr ND 0.039 Tr beans, dried (Mazur et al. 1998) "Diana" broad beans, NR 0.032 tr ND 0.006 tr dried (Mazur et al. 1998)

Canellini beans

Canellini beans, 10.5 ND ND ND ND ND dried (McKenzie's)

Values are means ± SD of duplicate analyses of three purchases ND=not detected tr=trace NR=not reported 203

5.2.5. Canellini beans (Phaseolus vulgaris)

Dried canellini beans purchased in Australia did not contain any isoflavones or coumestrol. (Table 5.17, previous page). There is no literature to report on the levels of phytoestrogens in canellini beans, therefore a comparison cannot be made.

5.2.6. Chick peas (Cicer arietinum)

Table 5.18 (overleaf) shows isoflavones and coumestrollevels in dried chick peas and products purchased in Australia. Genistein was the only compound from the five compounds of interest detected in dried and cooked chickpeas from McKenzie's, and in canned chick peas from Master Foods and Edgell. No isoflavones and coumestrol were detected in canned chick peas from Farmland. Levels of genistein in dried chick peas were not significantly different after cooking (p>0.05) (Table 5.18). Chick peas have long been known to contain isoflavones. These include daidzein, coumestrol, genistein, biochanin A and formononetin (Wong et al. 1965; Zilg & Grisebach

1969). However, Wong et al. (1965), Zilg and Grisebach (1969) only reported a qualitative analysis of isoflavones and coumestrol present in chick peas. The findings in this study differed from those of Mazur et al. (1998) who detected daidzein, genistein, formononetin, biochanin A and coumestrol in chick peas grown in the

United States. Variety and method of analysis used by this study differed from

Mazur et al. (1998). They used ID/GC/MS/SIM method and chick peas grown in the

USA, while this study used an isocratic HPLC method and chick peas grown in

Australia. 204

Table 5.18. Daidzein, genistein, coumestrol, formononetin and biochanin A in chick peas and products from Australia (mg/100 g, wet and dry weight basis).

Foods Basis % Daidzein Genistein Coumestrol Formononetin Biochanin Water A

Dried chick peas Chickpeas, Wetwt 6.3 ND 0.87 ±0.30 ND ND ND (McKenzie's) Drywt 0.93±0.34

Chickpeas, Wetwt 51.8 ND 0.41 ± 0.21 ND ND ND dried, cooked (McKenzie's) Drywt 0.84±0.53

Canned chick peas Chickpeas Wetwt 65.2 ND 0.04±0.03 ND ND ND (Master Foods) Chickpeas Wetwt 66.5 ND 0.20±0.04 ND ND ND (Edgell) Chick peas Wetwt 66.1 ND ND ND ND ND (Farmland)

Values are means± SD of duplicate analyses of three purchases ND=not detected

5.2.7. Green beans (Phaseolus vulgaris)

Table 5.19 (overleaf) shows the levels of isoflavones and coumestrol in whole fresh green beans purchased in Australia and Indonesia. Three varieties of fresh beans, two from Australia and one from Indonesia, did not contain isoflavones or coumestrol.

Fresh whole green beans ("continental" and ''baby'' types) from Australia contained small amounts of genistein and formononetin. Fresh whole green beans (''flat" beans) contained genistein only. 205

Table 5.19. Daidzein, genistein, coumestrol, formononetin and biochanin A in whole fresh, canned, frozen and freeze-dried green beans purchased in Australia and Indonesia (mg/100 g, wet weight basis).

Foods % Daidzein Genistein Coumestrol Forrnononetin Biochanin Water A

Australia Green beans, fresh 89.9 ND ND ND ND ND ''Flat" beans, fresh 92.5 ND 0.08±0.07 ND ND ND "Continental" 92.0 ND 0.06±0.07 ND 0.19 ± 0.10 ND beans, fresh Yard long beans, 91.1 ND ND ND ND ND fresh "Baby" beans, 91.9 ND 0.07 ±0.02 ND 0.17 ± 0.11 ND fresh Sliced Green 91.5 ND ND ND ND ND Beans, canned (Edgell) Sliced Green 90.5 ND ND ND ND ND Beans, canned (Golden Circle) Whole Beans, 87.9 ND 0.02±0.01 ND ND ND frozen (McCain) Sliced Green 86.9 ND 0.40±0.50 ND ND ND Beans, frozen (Birds Eye) Sliced Green 90.3 ND ND ND ND ND Beans, frozen (Hy Peak) Sliced Beans, 6.7 ND ND ND ND ND freeze-dried (Home Brand) Indonesia Green beans, fresh 91.2 ND ND ND ND ND Yard long beans, 90.2 ND ND ND ND ND fresh

Values are means± SD duplicates analyses of three purchases ND=not detected

The levels and presence of isoflavones and coumestrol in fresh green beans found in this study differed from those found by Franke et al. (1995). They found 206 formononetin in raw green beans at the level of 0.15 mg/100 g wet weight, and coumestrol and biochanin A in trace amounts.

Two brands of canned green beans analysed in this study contained no isoflavones or coumestrol (Table 5.19, previous page). Two of the three frozen green bean products purchased in Australia contained genistein (Table 5.19). They were Whole Beans from McCain (Australia) at a level of 0.02 mg/100 g wet weight, and Sliced Green

Beans from Birds Eye (Australia) at a level of 0.37 mg/100 g wet weight. Isoflavones and coumestrol were not detected in the freeze-dried bean products analysed as shown in Table 5.19.

Fresh beans are normally cooked before consumption. To identify the effect of cooking on the levels of isoflavones and coumestrol, one variety of bean was cooked with boiling water before analysis. The level of genistein decreased significantly after cooking (p<0.05) (Table 5.20, overleaf). Franke et al. (1995) also reported that the level of formononetin in raw green beans decreased after cooking. This was probably because some formononetin was destroyed or leached into the cooking water.

All fresh bean products were freeze-dried before analysis for isoflavones and coumestrol content. To identify the effect of freeze-drying on the level of isoflavones and coumestrol, beans were analysed before and after freeze-drying. The levels of isoflavones and coumestrol in fresh beans were found not to be significantly different from freeze-dried beans (p>0.05) (Table 5.20). 207

Table 5.20. Daidzein, genistein, coumestrol, formononetin and biochanin A in fresh beans, laboratory boiled and laboratory freeze-dried green beans (mg/100 g, wet and dry weight basis).

Foods Basis % Daidzein Genistein Coumestrol Formononetin Biochanin Water A

Green Wetwt 92.5 ND 0.18 ± 0.0 ND ND ND beans, fresh Drywt 2.37±0S Green Wetwt 91.9 ND 0.15 ± 0.0 ND ND ND beans, Drywt 1.84 ± o.oa laboratory freeze-dried Green Wetwt 92.0 ND 0.05±0.0 ND ND ND beans, Drywt o.58± o.ob laboratory boiled

Values are means ± SD of duplicate analyses of three purchases ND=not detected Dissimilar superscripts (a=highest) within column indicate significant differences on a dry weight basis (p<0.05).

5.2.8. Groundnuts/ peanuts (Arachis hypogaea)

No isoflavones or coumestrol were detected in dried groundnuts. Fermented peanuts

(oncom) also did not contain isoflavones or coumestrol. Fermented peanuts or oncom is made by fermentation of peanuts using Neurospora sitopila. The findings in this study differed from Mazur et al. (1998) who found daidzein, genistein, formononetin, and biochanin A in groundnuts from the USA at levels of

0.049 mg/100 g, 0.083 mg/100 g, 0.007 mg/100 g and 0.007 mg/100 g on a wet weight basis, respectively. Variety and method of analysis used in this study differed from those used by Mazur et al. (1998). The isocratic HPLC method used in this study could not identify isoflavones and coumestrol in groundnuts, possibly due to the detectability limits of the method. 208

5.2.9. Haricot beans (Phaseolus vulgaris)

Isoflavones and coumestrol were not detected in dried or cooked haricot beans purchased in Australia and in canned baked bean products purchased in Australia and

Indonesia. Mazur et al. (1998) detected daidzein, genistein, and biochanin A in navy beans (haricot) grown in the USA at levels of 0.014 mg/100 g, 0.408 mg/100 g and

0.004 mg/100 g, respectively. Franke et al. (1995) detected a trace amount of biochanin A in dried white navy bean seeds from the USA. It is considered that levels of isoflavones and coumestrol in haricot beans or baked beans were below the detectability level of the HPLC method used in this study.

5.2.10. Lentils (Lens culinaris M.)

Red lentils and whole green lentils purchased in Australia did not contain isoflavones or coumestrol (Table 5.21, overleaf). Similarly, Franke et al. (1995) did not find isoflavones and coumestrol in red lentils from the USA, while Mazur et al. (1998) did detect these compounds in two varieties of red lentils grown in the USA (Table

5.21). Level of isoflavones and coumestrol in lentils may be below the detectability limits of the HPLC method used in this study.

5.2.11. Lima beans (Phaseolus lunatus L.)

Lima beans were reported to contain coumestrol, formononetin and biochanin A

(Franke et al. 1995; Mazur et al. 1998). However, this study could not identify 209 isoflavones and coumestrol in dried lima beans purchased in Australia and in canned lima beans imported from the USA to Indonesia (Table 5.21).

Table 5.21. Daidzein, genistein, coumestrol, formononetin and biochanin A in red lentils purchased in Australia and grown in the USA and lima beans purchased in Australia and Indonesia (mg/100 g, wet weight basis).

Foods % Daidzein Genistein Coumestrol Formononetin Biochanin Water A

Lentils

Australia Whole green lentils, 9.1 ND ND ND ND ND dried (McKenzie's) Red lentils, dried 9.1 ND ND ND ND ND (McKenzie's)

USA (Mazur et al. 1998) "Jack Rabbit", dried NR 0.010 0.019 ND 0.011 Tr

"Masoor dahl", dried NR 0.003 0.003 0.007 0.006 0.007

Lima beans

Australia Lima beans, dried 9.2 ND ND ND ND ND (McKenzie's) Indonesia Lima Beans, canned 68.9 ND ND ND ND ND (S&W, CA, US) USA (Mazur et al. 1998) "Henderson's Bush" NR 0.01 0.01 ND 0.01 0.002 lima beans, dried "Sieva" pole lima NR 0.01 0.01 tr 0.01 0.003 beans, dried ''Large Seaside" lima NR 0.09 0.02 0.01 0.01 ND beans, dried

Values are means ± SD of duplicate analyses of three purchases ND=not detected tr: trace NR=not reported 210

Franke et al. (1995) found coumestrol at levels of 1.48 mg/100 g on a wet weight

basis and a trace amount of formononetin in lima beans grown in the USA at levels

of 1.48 mg/100 g on a wet weight basis. However, coumestrol was not identified in lima beans after boiling or in cooked lima beans after freeze-drying. Mazur et al.

(1998) found isoflavones and coumestrol at different levels in three different varieties of lima beans grown in the USA (Table 5.21, previous page). Variety and

growth conditions were considered to affect the presence and the levels of isoflavones and coumestrol in lima beans. Methods of analysis were also considered to affect the levels detected. Isoflavones and coumestrol in lima beans might be below the detectability limits of the method used in this study.

5.2.12. Mungbeans (Vigna radiata)

This study did not detect isoflavones and coumestrol in dried and cooked mungbean seeds purchased in Indonesia. However, genistein and coumestrol were detected in mungbean sprouts from Australia and Indonesia (Table 5.22, overleaf). Mungbean sprouts and Bean Salad sprouts from Country Fresh (Australia) and local markets

(Indonesia) contained genistein and coumestrol, while mungbean sprouts from supermarkets in Indonesia and Crunchy Combo sprouts from Australia contained genistein only. 211

Table 5.22. Daidzein, genistein, coumestrol, formononetin and biochanin A in mungbeans purchased in Australia and Indonesia (mg/100 g, wet weight basis).

Foods % Daidzein Genistein Cournestrol Formononetin Biochanin Water A

Mungbean seeds Indonesia Mungbeans, dried 9.8 ND ND ND ND ND seed Mungbean sprouts Australia

Beans Salad, 65.9 ND 0.97 ± 0.57 4.60±4.94 ND ND sprouts (Country Fresh) Crunchy Combo, 65.9 ND 0.3±0.2 ND ND ND sprouts (Country Fresh) Mungbean, sprouts 67.6 ND 0.20±0.09 4.12± 0.71 ND ND (Country Fresh) Indonesia

Mungbean, sprouts 75.9 ND 0.26± 0.12 14.3 ± 9.4 ND ND (unbranded) Mungbean, sprouts 87.6 ND 0.51 ±0.12 ND ND ND (packaged)

Values are means± SD of duplicates analyses of three purchases ND=not detected

Coumestrol levels in mungbean sprouts from local markets in Indonesia were higher than those in mungbean sprouts from Country Fresh (Australia). Bean salads were a mixture of mungbeans, caloona peas and lentil sprouts, while Crunchy Combo is a mixture of garden peas, mungbeans, caloona peas and lentils. The findings in this study differed from the study of Franke et al. (1995) and Mazur et al. (1998). Franke et al. (1995) found formononetin at levels of 0.61 mg/100 g on a wet weight basis in dry mungbean seeds grown in the USA and a trace amount of formononetin in 212 mungbean sprouts from the USA. Mazur et al. (1998) found 0.01 mg/100 g of daidzein, 0.365 mg/100 g of genistein, 0.008 mg/100 g of formononetin, 0.014 mg/

100 g of biochanin A and a trace amount of coumestrol, respectively, in mungbean seeds grown in the USA.

5.2.13. Peas (Pisum sativum)

Table 5.23 (overleaf) shows the levels of isoflavones and coumestrol in peas and pea products purchased in Australia and Indonesia. In this study, three different varieties of whole fresh peas and one of pea sprouts from Australia, and one variety of peas from Indonesia were analysed. Levels of isoflavones and coumestrol in peas differed among the varieties.

Two of the three varieties of peas from Australia (peas and snow peas) were found to contain formononetin and coumestrol and one variety of peas from Indonesia was found to contain only coumestrol. Snow pea sprouts from Australia were found to contain genistein. The levels of coumestrol, formononetin and genistein in fresh peas from both Australia and Indonesia were very low, at less than 1 mg/100 g on a wet weight basis. 213

Table 5.23. Daidzein, genistein, coumestrol, formononetin and biochanin A in fresh, dried and frozen peas purchased in Australia and Indonesia (mg/100 g, wet weight basis).

Foods % Daidzein Genistein Coumestrol Formononetin Biochanin Water A

Australia Honey snap peas, 86.3 ND ND ND ND ND fresh (unbranded) Peas, fresh 81.8 ND ND 0.92±0.21 0.13±0.12 ND (unbranded) Peas, fresh, cooked 85.2 ND ND ND 0.05±0.01 ND (unbranded) Snow peas, fresh 85.5 ND ND 0.06±0.02 0.15±0.10 ND (unbranded) Snow peas, fresh, 87.6 ND ND ND ND ND cooked (unbranded) Snow Peas, sprouts 88.9 ND 0.06±0.09 ND ND ND (Country Fresh) Peas, dried 6.6 ND ND ND 1.19±0.01 ND (McKenzie's) Peas, dried, cooked 51.9 ND ND ND ND ND (McKenzie's) Green split peas, 9.6 ND ND ND ND ND dried (McKenzie's) Yellow split peas, 8.9 ND ND ND ND ND dried (McKenzie's) Mint Peas, frozen 77.2 ND 0.15±0.14 ND ND 0.30±0.16 (Birds Eye) Baby Peas, frozen 79.9 ND ND ND ND ND (McCain) Frozen Peas 76.3 ND 0.06±0.02 ND ND 0.46±0.37 (McCain) Garden Peas, 76.6 ND 0.02±0.01 ND ND ND frozen (Watties) Frozen Peas, 81.6 ND ND ND ND ND cooked (Watties) Indonesia Green peas, fresh 78.4 ND ND 0.07±0.01 ND ND (unbranded) Green Peas, frozen 77.7 ND 0.48±0.03 ND ND ND (Farm Valley) Green peas, fresh 78.4 ND 0.23±0.24 ND ND ND (unbranded)

Values are means± SD duplicates analyses of three purchases ND=not detected 214

The presence and level of isoflavones and coumestrol may be expected to vary on the

basis of seed maturity and other factors. Levels in dried pea products were different

from those in fresh products. This study indicated that one of three different varieties

of dried pea products was found to contain formononetin. These were dried peas

from M'l<.enzie's (Australia). No isoflavones and coumestrol were found in green

split peas and yellow split peas from McKenzie's (Australia) (Table 5.23, previous

page).

Factors such as freeze-drying, canning, and freezing might affect the levels and the

presence of isoflavones and coumestrol in peas. Five frozen pea products from either

Australia or Indonesia (Birds Eye, McCain, Watties, Farm Valley and unbranded peas) were found to contain genistein at low levels. Two of the products were found to contain biochanin A only (Birds Eye and McCain) (Table 5.23).

The presence and the level of isoflavones and coumestrol in canned peas varied between brands. Garden Peas from Edgell (Australia) contained biochanin A, Green

Peas from Golden Circle (Australia) contained daidzein and genistein, Young Sweet

Peas from S&W (imported from the USA to Indonesia) and Sweet Peas from Del

Monte (imported from the USA to Indonesia) contained genistein, while Processed

Peas from Ayam Brand (imported from Malaysia to Indonesia) contained genistein and formononetin (Table 5.24, overleaf). 215

Table 5.24. Daidzein, genistein, coumestrol, formononetin and biochanin A in canned and freeze-dried peas purchased in Australia and Indonesia (mg/100 g, wet weight basis).

Foods % Daidzein Genistein Coumestrol Formononetin Biochanin Water A

Australia Canned Garden Peas 79.8 ND ND ND ND ND (Mountain Maid) Garden Peas (Edgell) 80.2 ND ND ND ND 1.23±1.46 Green Peas 78.9 0.03±0.01 0.06±0.02 ND ND ND (Golden Circle) Processed Dried Peas 70.8 ND ND ND ND ND (PMU) Freeze-dried

Garden Peas 4.6 ND ND ND ND ND (Continental) Garden Peas, cooked 79.7 ND 0.02±0.01 ND ND ND (Continental) Garden Peas (Home 3.0 ND 2.48±0.24 ND ND ND Brand) Indonesia Canned Green Peas (Maling) 61.2 ND ND ND ND ND Young Sweet Peas 80.1 ND 0.55±0.54 ND ND ND (S&W) Sweet Peas (Del 81.1 ND 0.29±0.23 ND ND ND Monte) Processed Peas 70.2 ND 0.08±0.00 ND 0.19±0.00 ND (Ayam Brand)

Values are means± SD duplicates analyses of three purchases ND= not detected

The presence and the level of isoflavones and coumestrol in freeze-dried pea products were also different across the brands as shown in Table 5.24. Freeze-dried 216

Peas from Home Brand (Australia) contained genistein but freeze-dried Garden Peas from Continental (Australia) did not contain isoflavones or coumestrol.

The method of analysis and variety of peas might influence the presence and the level of isoflavones and coumestrol. The findings in this study were also different from those of Mazur et al. (1998). They found that levels of isoflavones in two green split pea samples were different. One sample contained daidzein (0.005 mg/100 g) and biochanin A (0.003 mg/100 g), while the other sample contained daidzein

(0.011 mg/100 g), formononetin (0.005 mg/100 g), biochanin A (0.003 mg/100 g), a trace amount of genistein and no coumestrol. Split peas (Chann dhal) contained the five isoflavones (0.008 mg/100 g of daidzein, 0.023 mg/100 g of genistein, a trace amount of coumestrol, 0.005 mg/100 g of formononetin and 0.006 mg/100 g of biochanin A). Yellow split peas contained daidzein (0.004 mg/100 g), a trace amount of genistein, formononetin (0.001 mg/100 g) and biochanin A (0.004 mg/100 g) and no coumestrol. Franke et al. (1995) found biochanin A (0.9 mg/100 g) in yellow split peas, daidzein (7.3 mg/100 g) and a trace amount offormononetin in green split peas.

Franke et al. (1995) also detected biochanin A (0.9 mg/100 g) in cooked Chinese peas and coumestrol (8.1 mg/100 g) in round split peas.

Peas are rarely consumed raw. They are normally cooked by boiling or stirring with other ingredients. This study determined the effect of cooking on the levels of isoflavones and coumestrol in dried, freeze dried, frozen and fresh peas. Boiling was the only method used for cooking. This study found that the presence and the levels of isoflavones and coumestrol in peas changed after cooking. Formononetin in fresh peas decreased after boiling (Table 5.23, page 213). It may have leached into the 217 boiling water. Genistein was detected in freeze-dried Garden Peas from Continental

(Australia) after cooking with boiling water for 2 hours. The level was 0.02 mg/100 g on a wet weight basis (Table 5.24, page 215).

5.2.14. Petai (Parkia speciosa) and petai cina (Leucaena spp.)

Isoflavones and coumestrol were not found in petai and petai cina purchased in

Indonesia. There in no literature reported on isoflavones and coumestrollevels in petai and petai cina (Table 5.25), therefore a comparison cannot be made.

Table 5.25. Daidzein, genistein, coumestrol, formononetin and biochanin A in petai and petai cina purchased in Indonesia (mg/100 g, wet weight basis).

Foods % Daidzein Genistein Coumestrol Formononetin Biochanin Water A

Petai, fresh 76.3. ND ND ND ND ND (unbranded)

Petai cina, fresh 66.8 ND ND ND ND ND (unbranded)

Values are means± SD of duplicate analyses of 3 purchases ND=not detected

5.2.15. Red kidney beans (Phaseolus vulgaris)

Isoflavones and coumestrol were not detected in dried, cooked and canned red kidney beans purchased in Australia or Indonesia. The finding of this study was similar to the study by Franke et al. (1995). They also did not detect any isoflavones 218

and coumestrol in red kidney beans. However, Mazur et al. (1998) found daidzein,

genistein and biochanin A in red kidney beans at levels of 0.008 mg/100 g,

0.007 mg/100 g and 0.007 mg/100 g, respectively. The differences in the findings

reported by this study and other studies were predominantly caused by the methods

of analysis used. This study and Franke et al. (1994) used LC methods while Mazur

et al. (1998) used the ID/GC/MS/SIM method. Detectability of the HPLC method

used in this study and Franke et al. (1995) were above the levels of isoflavones found

in red kidney beans. Therefore, those compounds could not be detected by these

methods. The variety of red kidney bean may also have contributed to the

differences. Both Franke et al. (1995) and Mazur et al. (1998) analysed red kidney

beans grown in the USA, while this study analysed those grown in Australia.

5.3. Alfalfa (Medicago sativa)

Two types of fresh alfalfa sprout, each mixed with onion or radish flavouring, were evaluated. Alfalfa was found to only contain formononetin of the five compounds studied, as shown in Table 5.26 (overleaf). Formononetin levels in both products were 9.5 ± 3.4 mg/100 g and 7.5 ± 4.2 mg/100 g dry weight for alfalfa onion and alfalfa radish, respectively. This study could not identify daidzein, genistein, coumestrol, and biochanin A in alfalfa but other authors could (Livingstone et al.

1961; Knuckles et al. 1976; Guggolz et al. 1961; Franke et al. 1995)

Method of analysis is one of the factors causing a difference in the levels. Hanson et al. (1965) reported that climatic conditions and biological attacks (foliar disease and insect damage) could affect the production of coumestrol in plants. Livingstone et 219

al. (1961) used a paper chromatographic fluorometric technique to separate

coumestrol in alfalfa. Alfalfa was soaked in ethanol overnight before analysis.

Knuckles et al. (1976) used a two-dimensional paper chromatography technique for

the separation of coumestrol from alfalfa. The sample was soaked in 75% ethanol for

16 hours before analysis. They found coumestrol at levels ranging from 11 to 118

!lg/g (1.1 to 11.8 mg/100 g) on a dry weight basis. Murphy et al. (1999) with HPLC

method identified formononetin and coumestrol in all alfalfa-labeled products at

levels ranging from 20 to 440 !lg/g (2.0 to 44.0 mg/100 g) on a dry weight basis.

USDA (1999) reported the levels ofcoumestrol (446 mg/100 g), formononetin (1771 mg/100 g) and biochanin A (2946 mg/100 g) on a wet weight basis in alfalfa.

Table 5.26. Isoflavones and coumestrol in sprouted alfalfa from Australia (mg/100 g, wet weight basis).

Product % Daidzein Genistein Coumestrol Formononetin Biochanin A Water

Alfalfa 90.5 ND ND ND 0.5 ± 0.10 ND onion Alfalfa 92.1 ND ND ND 0.6± 3.60 ND radish

Values are means ± SD of duplicate analyses of three purchases ND=not detected

Guggolz et al. (1961) detected daidzein (<1 ppm or <0.1 mg/100 g), formononetin

(14 ppm or 1.4 mg/100 g), genistein (<1 ppm or <0.1 mg/100 g), biochanin A (1-5 ppm or 0.1-0.5 mg/100 g) and coumestrol (57 ppm or 5.7 mg/100 g) in alfalfa. They used a combination of paper and silicic acid chromatography methods. Franke et al.

(1995) could not detect daidzein, genistein and biochanin A in alfalfa sprouts, but 220 could detect coumestrol and formononetin in alfalfa. The presence of formononetin

and coumestrol in alfalfa products was also found by Murphy et al. (1999) at levels ranging from 20 to 440 J.Lg/g (2.0 to 44.0 mg/100 g) on a dry weight basis. USDA

(1999) also reported coumestrol (446 mg/100 g), formononetin (1,771 mg/100 g) and biochanin A (2,946 mg/1 00 g) on a wet weight basis in alfalfa. The method for analysis of isoflavones and coumestrol in alfalfa used in this study differed from that of the other authors. This study did not soak the samples before analysis but refluxed them in a bath of boiling water for 6 hours during the extraction.

5.4. Cashew (Anacardium occidentale L.)

This study did not find isoflavones and coumestrol in dried and fried cashew nut purchased in Indonesia. These nuts were grown in Sulawesi Tenggara, Indonesia.

There was no literature reported on the isoflavones contained in cashew nut, therefore comparison cannot be made.

5.5. Conclusion

This study found many different types of soya bean products in Australia and

Indonesia, with a wider range of branded soya milk products available in Australia than in Indonesia. Soya bean paste and sweetened soya sauce were much more available in Indonesia, while a wider variety of second-generation soya foods was available in Australia. The second-generation soya foods are mostly made by simulating Australian foods. These are manufactured to replace animal protein and/or 221 to reduce fat, and/or to incorporate the currently fashionable soya ingredients.

However, similarity to the original foods was maintained by addition of soya ingredients in limited amounts. This results in the lowering of iso:flavone levels in the product compared to raw soya beans.

The large variation in the levels of iso:flavones across brands and across each soya food product is influenced by the variety of raw material used, the processing methods, and the amount of non-soya ingredients used.

There was a variation in the iso:flavones and coumestrol content in legumes and legume products analysed in this study. Of the 18 species of legumes and their products analysed, 13 contained no isoflavones or coumestrol. The levels were significantly different between brands, between legume products and between countries. Agriculture and growth conditions, processing and preparation, and method of analysis would have been among the factors affecting the results. As a consequence, representative data for these compounds in foods would be extremely difficult to derive.

It is of concern that some soya breads and cereals analysed contained no isoflavones.

This raises the question of the introduction of labelling regulations for phytoestrogens in foods in Australia. 222

CHAPTER6.

CONCLUSIONS

The objectives of this study (Chapter 1) were fulfilled as summarised below.

6.1. Successful method development for analysis of

phytoestrogens in foods

The methods developed are an improvement on current methods available for

identification and quantitation of isoflavones in foods, except where the presence of glycitein is suspected. These methods are an improvement on other published LC methods for the following reasons:

• The methods are rapid. The separation time for both methods is less then 24

minutes.

• The methods are more sensitive and capable of detecting the analytes at very low

levels (20J..Lg/100 g).

• The methods are simple and safe. They do not require lengthy equilibration

times and require only a simple pump. 223

• The isocratic conditions and the organic solvent used could prolong a column

life.

• Sample preparation for LC is faster than for GC, since it does not require prior

derivatisation.

• Neutralization of the sample prior to identification protects the column from

damage without altering the concentration of isoflavones in the foods.

The developed methods are preferred for routine use over the "gold standard" method, isotope dilution gas chromatographic-mass spectrometry of Mazur et al.

(1996) because they are cheaper, safer, require lower levels of analytical skills, and because HPLC is more accessible to laboratories everywhere. The isotope dilution gas chromatographic-mass spectrometric method is able to measure lower concentrations of the analytes in foods (<2~g/100 g), but this is probably not of practical significance for uses such as for food labelling.

6.1.1. Identification methods

Two HPLC methods were developed in this study. The first method was for the identification of isoflavones; daidzein, genistein and biochanin A. The method used an isocratic HPLC with C8 column and acetonitrile-aqueous acetic acid (33:67, v/v) as eluent with a flow rate of 0.8 ml/minutes. The three compounds were separated in less than 24 minutes. The second method was for the simultaneous determination of daidzein, genistein, formononetin, biochanin A and coumestrol. The method used an 224

isocratic HPLC with a phenyl column and acetonitrile-water (33:67, v/v) as eluent

with a flow rate of 0.8 ml/minutes. The five compounds were well separated in less

than 24 minutes. Correct peak assignment of daidzein, genistein, formononetin,

biochanin A and coumestrol was successfully confirmed by LC-MS-MS. The second method showed a high degree of precision with the coefficient of variation of the retention time between- and within-assay being 0.6% and 2.1 %, respectively, and the

coefficient of variation of the quantitative analysis of the individual compounds between- and within-assay being 5.7% and 6.8%, respectively. The method also showed a high linearity with coefficients of determination(?) higher than 0.999. The detectability of the method was 47, 82, 76, 75 and 224 nM for daidzein, coumestrol, genistein, formononetin and biochanin A, respectively. The method showed a good recovery (-100%) of standard compounds from foods such as soya beans, beans and peas. The results showed that both methods have the same separation time (less than

24 minutes). The values obtained from the analytical performances of both HPLC methods showed that both methods ful:filled reliability criteria very well.

Method A is simpler than method B. Method A is useful and recommended for the identification of isoflavones in foods that contain mostly daidzein and genistein, such as soya beans or soya bean products. Method B is more useful than method A for simultaneous identification and quantitation of isoflavones and coumestrol in food products, and especially for mixed or formulated foods, e.g. breads and breakfast cereals containing several different sources of these compounds. Both methods are suitable for the production of data for the purpose of food labelling. 225

6.1.2. Isolation methods

Methods were developed to extract and hydrolyse the conjugated forms of the

isoflavones and coumestrol from soya beans, cooked soya beans, canned soya beans,

tofu and soya milk. The method used acid (2 M HCl) and heat (1 00°C) in a water

bath. The amounts of soya beans, cooked soya beans, canned soya beans analysed were 1 mg/50 mi. The amount of soya milk analysed was 10 ml/50 ml solution. All extracted samples were adjusted to pH 4 with sodium hydroxide before identification and quantification.

6.2. Determination of phytoestrogens in legumes and legume

products available at retail level in Australia and Indonesia

• Despite the limitations of this study, it has produced the first comprehensive set

of data on the phytoestrogens content of Australian and Indonesian foods.

• This study clearly demonstrated that levels of isoflavones and coumestrol differ

between a variety of legumes, and brands of processed and non-processed

legume products from Australia and Indonesia. More than 50% of the legumes

and legume products analysed contained no isoflavones and coumestrol.

• This study identified that daidzein and genistein were the only isoflavones found

in soya beans and soya bean products, except for two soya milk products from

Indonesia that contained coumestrol. Raw soya beans always contained higher 226

levels of isoflavones than processed soya beans in whatever form. Levels of

daidzein and genistein within soya beans, and soya bean products of different

brands varied up to 22-fold on a wet weight basis. Levels of daidzein and

genistein in second-generation soya foods ranged from 0.0-13.3 mg/100 g and

0.0-14.5 mg/100 g, respectively. Levels of daidzein and genistein in some soya

beans and soya bean products diminished after boiling or frying (dried soya

beans, tempeh and oncom). However, cooking did not affect the levels of

daidzein and genistein in tofu.

• This study also identified that daidzein, genistein, and formononetin were the

only isoflavones found in beans and peas, whilst biochanin A was also found in

peas. Genistein and coumestrol were the only compounds found in bean sprouts.

Formononetin was the only compound found in alfalfa. Levels of daidzein,

genistein, formononetin, coumestrol and biochanin A in a variety of beans, bean

sprouts and peas, alfalfa, and their products were very low. Levels were less than

1 mg/100 g on a wet weight basis.

• These results lead to conclude that the levels of isoflavones and coumestrol in

legumes and their products were produced by nature (plant species, plant parts,

maturation, growing conditions) and by the manufacturing process, and were not

artefacts of the method of analysis due to its overall reliability.

• Most of the variability in isoflavones and coumestrollevels in legumes and their

products could not be interpreted precisely because most of the foods were 227

obtained from retail outlets and information, such as variety and processing

conditions was not available.

• Since the results obtained from this study showed that levels of phytoestrogens

were variable, it is concluded (e.g. for compositional tables) that the derivation of

representative levels of phytoestrogens in foods would be very difficult.

• Using data from Table 2.4, the total isoflavones available for consumption in

Australia and Indonesia are 200 and 550 mg/year/caput, respectively.

• This study has shown that the levels of isoflavones in soya bean and products

from Australia and Indonesia are similar in general to those from other countries.

Nevertheless, these data are recommended for use in Australian and Indonesian

food tables, because of some compositional differences. The data for mixed foods

such as breads and breakfast cereals are recommended for inclusion in Australian

food composition tables because of specific local formulations.

6.3. Recommendations for further research.

• Method A is simpler than method B. Method A is useful and recommended for

the identification of isoflavones in foods that mostly contain daidzein and

genistein, such as soya beans or soya bean products. Method B is more useful

than method A for simultaneous identification and quantitation of isoflavones

and coumestrol in food products, and especially for mixed or formulated foods, 228

e.g. breads and breakfast cereals containing several different sources of these

compounds. Both methods are suitable for the production of data for the propose

of food labelling.

• This study has shown that the levels of isoflavones in soya bean and products

from Australia and Indonesia are similar in general to those from other countries.

Nevertheless, the use of these data produced is recommended for use in

Australian and Indonesian food tables, because of some compositional

differences. The data for mixed foods such as breads and breakfast cereals are

recommended for inclusion in Australian food composition tables because of

specific local formulations.

• This study could not detect any isoflavones and coumestrol in some legumes and

their products. It developed a unique extraction method for soya beans and soya

bean products. However, it is suggested that the extraction method for soya beans

and soya bean products may not be applicable for other legumes and their

products. Isoflavones and coumestrol patterns naturally differ in individual

legume species. Therefore, a particular legume species and its products could

have different extraction efficiencies. This method needs to be tested for its

specificity for individual legume species and products and, where necessary, it

should be modified for particular species and products. The ratio of solvent

volume (milliliters) to food material (grams) needs to be examined, especially for

foods with very low levels of isoflavones and coumestrol. 229

• Methods for measuring conjugates and hydrolysates separately in legumes and

legume products need to be studied in the future because isoflavone and

coumestrol patterns in raw legumes differ between species and not all people can

convert the conjugated forms into free forms, which are the biologically active

compounds.

• It is important to develop new methods in order to incorporate glycitin and

glycitein. No study yet has shown the biological activity of glycitin and glycitein.

However, through processing, these compounds may be degraded into

biologically active compounds.

• Further research is required to relate measured dietary levels of phytoestrogens to

levels of metabolites in blood and/or urine for future epidemiological research on

phytoestrogens and health, especially in making comparisons between Australian

and Indonesian populations, given their very different dietary patterns.

• Collaborative trials are required to determine the most reliable method for

determining the content of phytoestrogens in foods, with a view to nominating a

standard method for universal use. 230

BIBLIOGRAPHY

Adlercreutz, H. (1990) Western diet and Western diseases: some hormonal and biochemical mechanisms and associations. Scand. J. Clin. Lab. Invest. 50, Suppl. 201: 3-23.

Adlercreutz, H. (1995a) Lignans and isoflavonoids: Epidemiology and a possible role in prevention of cancer. Proceedings of the Laboratory Medicine '95. 1 fh !FCC European Congress of Clinical Chemistry. Tempere, Finland.

Adlercreutz, H., Fotsis, T., Bannwart, C., Wahala, K., Makela, T., Brunow, G. & Rase, T. (1986b) Determination of urinary lignans and phytoestrogen metabolites, potential antiestrogens and anticarcinogens, in urine of women on various habitual diets. J. Steroid Biochem. MoZee. Biol. 25: 791-797.

Adlercreutz, H., Fotsis, T., Bannwart, C., Wahala, K., Brunow, G. & Rase, T. (1991a) Isotope dilution gas chromatographic-mass spectrometric method for determination of lignans and isoflavonoids in human urine including identification of genistein. Clin. Chim. Acta 199:263-278.

Adlercreutz, H., Fotsis, T., Heikkinen, R., Dwyer, J. T., Goldin, B. R., Gorbach, S. L., Lawson, A. M. & Setchell, K. D. R. (1981) Diet and urinary excretion of lignans in female subjects. Med. Biol. 59: 259-261.

Adlercreutz, H., Fotsis, T., Heikkinen, R., Dwyer, J. T., Woods, M., Goldin, B. R. (1982) Excretion of the lignans enterolactone and enterodiol and equol in omnivorous and vegetarian postmenopausal women and in women with breast cancer. Lancet 11: 1295-1298.

Adlecreutz, H., Fotsis. T., Hockerstedt, K., Hamalainen. E., Bannwart. C., Bloigu. C., Valtonen. A. & Ollus, A. (1989) Diet and urinary estrogen profile in premenopausal omnivorous and vegetarian women and in premenopausal women with breast cancer. J. Steroid Biochem. 34: 527-530.

Adlercreutz, H., Fotsis, T., Kurzer, M. S., Wahala, K. & Makela, T. (1995) Isotope dilution gas chromatographic-mass spectrometric method for the determination of unconjugated lignans and isoflavonoids in human feces, with preliminary results in omnivorous and vegetarian women. Anal. Biochem. 225: 101-108. 231

Adlercreutz, H., Fotsis, T., Lampe, J., Wahala, K., Makela, T., Brunow, G. & Hase, T. (1993) Quantitative determination of lignans and isoflavonoids in plasma of omnivorous and vegetarian women by isotope dilution gas chromatography-mass spectrometry. J. Clin. Lab. Invest. 53: 5-18.

Adlercreutz, H., Goldin, B. R., Gorbach, S. L., Hockerstedt, K. A. V., Watanabe, S., Hamalainen, E. K., Markkanen, M. H., Makela, T. H., Wahala, K. T., Hase, T. A. & Fotsis, T. (1995c) Soya bean phytoestrogen intake and cancer risk J. Nutr. 125: 757S-770S.

Adlercreutz, H., Hockerstedt, K., Bannwart, C., Bloigu S., Hamalainen, E., Fotsis, T. & Ollus, A. (1987) Effect of dietary components, including lignans and phytoestrogens, on enterohepatic circulation and liver metabolism of estrogens and on sex hormone binding globulin (SHBG). J. Steroid Biochem. 27: 1135-1144.

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

Adlercreutz, H., Mousavi, Y., Clark, J., Kockerstedt, K., Hamalainen, E., Wahala, K., Makela, T. & Hase, T. (1992) Dietary phytoestrogens and cancer: in vitro and in vivo studies. J. Steroid Biochem. MoZee. BioZ. 41: 3-8.

Adlercreutz, H., Musey, P. I., Fotsis, T., Bannwart, C., Wahala, K., Makela, T., Brunow, G. & Hase, T. (1986a) Identification of lignans and phytoestrogens in urine of chimpanzees. Clin. Chim. Acta 158: 147-154.

Adlercreutz, H., van der Wildt, J., Kinzel, J., Attalla, H., Wahala, K., Makela, T., Hase, T. & Fotsis, T. (1995b) Lignan and isoflavonoid conjugates in human urine. J. Steroid Biochem. MoZee. BioZ. 52: 97-103.

Akiyama, T., Ishida, J., Nakagawa, S., Ogawara, H., Watanabe, S., ltoh, N., Shibuya, M. & Fukami, Y. (1987) Genistein, a specific inhibitor of tyrosine-specific protein kinases. J. Biol. Chern. 262: 5592-5595.

Albertazzi, P., Pansini, F., Bottazzi, M. & Bonaccorsi, G. (1999) Dietary soy supplementation and phytoestrogen levels. Obstet. Gynecol. 94: 229-231. 232

Alexander, G. & Rossiter, R. C. (1951) The effects of fertilizer treatments on the oestrogenic potency of Trifolium subterraneum L. Aust. J. Agric. Res. 3: 24-28

Amin, El. S., A wad, 0., Samad, M. A., Iskandaer, M. N. (1969) Isolation of from moghat roots and from pollen grain of Egyptian date palm. Phytochemistry 8: 295-297.

Anderson, J. J. B., Ambrose, W. W. & Gamer, S. C. (1998) Biphasic effects of genistein on bone tissue in the ovariectomised, lactating rat model. Proc. Soc. Exp. Biol. Med. 217: 345-350.

Anderson, R. L. & Wolf, W. J. (1995) Compositional changes in trypsin inhibitors, , saponins and isoflavones related to soya bean processing. J. Nutr. 125: 581S-588S.

Arndt, H. W. (1986) Australia and Indonesia - neighbour forever. Nedlands, WA, Australia: Department of Economic, University of Western Australia.

Australian Bureau of Statistics. (1995) Australia's Food and Nutrition. Canberra: ABS.

Australia's Health 2000 (2000): The seventh biennial health report of the Australian Institute ofHealth and Welfare. Canberra: Australian Institute.

Axelson, M., Kirk, D. N., Farrant, R. D., Cooley, G. & Lawson, A. (1982a) The identification of the weak oestrogen equol [7-hydroxy-3-(4'-hydroxyphenyl) chroman] in human urine. Biochem. J. 201: 353-357.

Axelson, M., Sjovall, J., Gustafsson, B. E. & Setchell, K. D. R. (1982b) Origin of lignans in mammals and identification of a precursor from plants. Nature 298: 659- 660.

Axelson, M., Sjovall, J., Gustafsson, B. E. & Sethcell, K. D. R. (1984) Soy - a dietary source of the non-steroidal oestrogen equol in man and animals. J. Endocrinol. 102: 49-56. 233

Bannwart, C., Adlercreutz, H., Fotsis, T., Wahala, K., Hase, T. & Brunow, G. (1984a) Identification of 0-desmethylangolensin, a metabolite of daidzein, and of matairesinol, one likely plant prcursor of the animal lignan enterolactone, in human urine. Finn. Chem. Lett. 45: 120-125.

Bannwart, C., Fotsis, T., Heikkinen, R. & Adlercreutz, H. (1984b) Identification of the isoflavonic phytoestrogen daidzein in human urine. Clin. Chim. Acta 136: 165- 172.

Barbetti, M. J. & Nichols, P. G. H. (1991) Effect of Phoma medicaginis and Leptosphaerulina trifolii on herbage and see yield and coumestrol content of annual medicago species. Phytophylatica 23: 2223-2237.

Barnes, S., Coward, L., Kirk, M. & Sfakianos, J. (1998) HPLC-Mass spectrometry analysis of isoflavones. Proc. Soc. Exp. Biol. Med. 208: 254-262.

Barnes, S., Grubbs, C., Setchell, K. D. R. & Carlson, J. (1990) Soybeans inhibit mammary tumors of breast cancer. Pariza, M. (ed.) and carcinogens in the diet. New York: Willey-Liss: 239-253.

Barnes, S., Kirk, M. & Coward, L. (1994a) Isoflavones and their conjugates in soy foods: Extraction conditions and analysis by HPLC-Mass spectrometry. J Agric. Food Chem. 42: 2466-2474.

Barnes, S., Peterson, G., Grubbs, C. & Setchell, K. (1994b) Potential role of dietary isoflavones in the prevention of cancer. Jacobs, M.M. (ed.). Diet and cancer. New York: Plenum Press; 135-145.

Bartholomew, R. M. & Ryan, D. S. (1980) Lack of mutagenic of some phytoestrogens in the salmonella/mammalian microsome assay. Mutation Res. 78: 317-321.

Batterham, T. J., Hart, N. K. & Lamberton, J. A. (1965) Metabolism of oestrogenic isoflavones in sheep. Nature 1: 509.

Bennett, R. D., Ko, S. & Heftmann, R. (1966) Isolation of estrone and cholesterol from date palm, Phoenix dactylifera L. Phytochemistry 5: 231-235. 234

Bennett, G. A., Peplinski, A. J., Brekke, 0. L. & Jackson, L. K. (1976) Zearalenone: distribution in dry-milled fractions of contaminated corn. Cereal Chem. 53: 299-307.

Bennetts, H. W., Underwood, E. J. & Shier, F. L. (1946) A specific breeding problem of sheep on subterranean clover pastures in Western Australia. Aust. Vet. J. 22:2-12.

Bickoff, E. M., Booth, A. N., Lyman, R. L., Livingston, A. L., Thompson, C. R. & DeEds, F. (1957) Coumestrol, a new estrogen isolated from forage crops. Science 126: 969-970.

Bickoff, E. M., Booth, A. N., Lyman, R. L., Livingstone, A. L., Thompson, C. R. & Kohler, G. 0. (1958) Isolation of a new estrogen from ladino clover. J. Agric. Food. Chem. 6: 536-539.

Bickoff, E. M., Livingstone, A. L. & Booth, A. N. (1960) Estrogenic activity of coumestrol and related compounds. Arch. Biochem. Biophys. 88: 262-266.

Bickoff, E. M., Livingstone, A. L. & Booth, A. N. (1964) Tricin from Alfalfa­ isolation and physiological activity. J. Pharm. Sci. 53 (11): 1411-1412.

Bickoff, E. M., Livingstone, A. L., Hendrickson, A. P. & Booth, A. N. (1962) Relative potencies of several oestrogen-like compounds found in forages. J. Agric. Food Chem. 10:410-412.

Bingham, S. A., Atkinson, C., Liggins, J., Bluck, L. & Coward, A. (1998) Phytoestrogens: where are we now? Br. J. Nutr. 79: 393-406.

Biro Pusat Statistik (Central Bureau of Statistics). (1990) Pengeluaran untuk konsumsi penduduk Indonesia 1990 (Expenditure for consumption of Indonesian 1990). Jakarta: BPS.

Blaney, B. J., Moore, C. J. & Tyler, A. L. (1984) and fungal damage in maize harvested during 1982 in Far North Queensland. Aust. J. Agric. Res. 35: 463- 471.

Blair, H. C., Jordan, S. E., Peterson, T. G. & Barnes, S. (1996) Variable effect of tyrosine kinase inhibitors on ovarian osteoclastic activity and reduction of bone loss in overiectomised rats. J. Cell. Biochem. 60: 1761-1769. 235

Booth, A. N., Bickoff, E. M. & Kohler, G. 0. (1960) Estrogen-like activity in vegetable oils and mill by-products. Science 131: 1807-1808.

Borriello, S. P., Setchell, K. D. R., Axelson, M. & Lawson, A.M. (1985) Production and metabolism oflignans by the human faecal flora. J. Appl. Bacterial. 58: 37-43.

Bose, T. K. (1985) Fruits of India. India: B. Mitra.

Bradbury, R. B. & White, D. E. (1954) Estrogens and related substances in plants. Vitam Horm. 12: 207-233.

Branden, A. W. H., Hart, N. K. & Lamberton, J. A. (1967) The oestrogenic activity and metabolism of certain isoflavones in sheep. J. Agric. Res. 18: 335-348.

Burroughs, C. D. (1995) Long-term reproductive tract alterations in female mice treated neonatally with coumestrol. Proc. Soc. Exp. Biol. Med. 208: 78-81.

Burroughs, C. D., Bern, H. A. & Stokstad, E. L. R. (1985) Prolonged vaginal cornification and other changes in mice treated neonatally with coumestrol, a plant estrogen. J. Toxicol. Environ. Health 15: 51-61.

Burroughs, C. D., Mills, K. T., & Bern, H. A. (1990) Long-term genital tract changes in female mice treated neonatally with coumestrol. Reprod. Toxicol. 4: 127-135.

Cassidy, A. & Bingham, A. (1995) Biological effects of isoflavones in young women: importance of the chemical composition of soya bean products. Br. J. Nutr. 74: 587-601.

Cayen, M. N., Tang, G. & Common, R. H. (1965) Urinary conversion products of biochanin A and formononetin in the fowl Biochim. Biophys. Acta 3: 349-457.

Chen, Z., Zheng, W., Custer, L. J., Dai, Q., Shu, X-0., Jin, F. & Franke, A. A. (1999) Usual dietary consumption of soy foods and its correlation with the excretion rate of isoflavonoids in overnight urine samples among Chinese women in Shanghai. Nutr. Cancer 33: 82-87. 236

Cheng, E., Story, C. D., Payne, L. C., Yoder, L. & Burroughs, W. (1953) Detection of in alfalfa and clover hays fed to fattening lambs. J. Anim. Sci. 12: 507-514.

Cheng, E., Yoder, L., Story, C. D. & Burroughs, W. (1954) Estrogenic activity of some isoflavones derivatives. Science 120: 575-576.

Coleman, M.P., Esteve, J., Damiecki, P., Arslan, A. & Renard, H. (1993) Trends in cancer incidence and mortality. Lyon: IARC.

Constantinou, A. & Huberman, E. (1995) Genistein as an inducer of tumor cell differentiation: possible mechanisms of action. Proc. Soc. Exp. Biol. Med. 208: 109- 115.

Coward, L., Barnes, N.C., Setchell, K.D.R. & Barnes, S. (1993) Genistein, daidzein, and their ~- conjugates: antitumor isoflavones in soya bean foods from American and Asian diets. J. Agric. Food Chem. 41: 1961-1967.

Coward, L., Kirk, M., Albin, N. & Barnes, S. (1996) Analysis of plasma isoflavones by reversed-phase HPLC-multiple reaction ion monitoring-mass spectrometry. Clin. Chim. Acta 247: 121-142.

Dalais, F. S., Wahlqvist, M. L. & Rice, G. E. (1997a) Variation in isoflavonoid phytoestrogen content in soya beans grown in Australia. Proc. Nutr. Soc. Aust. 21:161.

Dalais, F. S., Wahlqvist, M. L. & Rice, G. E. (1997b) Isoflavonoid phytoestrogen content in soya bean products in the Australian market. Proc. Nutr. Soc. Aust. 21:123.

Dean, P. D. G., Exley, D. & Goodwin, T. W. (1971) Steroid oestrogens in plants: re­ estimation of oestrone in pomegranate seeds. Phytochemistry 10: 2215-2216.

Dehennin, L., Reiffsteck, A., Jondet, M. & Thibier, M. (1982) Identification and quantitative estimation of a lignan in human and bovine semen. J. Reprod. Fertil. 66: 305-309.

Denekamp, J. (1990) Vascular attack as a therapeutic strategy for cancer. Cancer Metastasis Rev. 9: 267-282. 237

Department of Health and Family Service (1995) National Nutrition Survey Australia. Selected Highlights. Canberra: ABS.

Dewick, P. M. (1988) Isoflavonoids. Harborne, J. B. (ed.) The Flavonoids. New York: Chapman & Hall; 125.

Divi, R. L., Chang, H. C. & Doerge, D. R. (1997) Anti-thyroid isoflavones from soya bean. Biochem. Pharmacal. 57: 1087-1096.

Dwyer, J. T., Goldin, B. R., Saul, N., Gualtieri, L., Barakat, S. & Adlercreutz, H. (1994) Tofu and soy drinks contain phytoestrogens. J. Am Diet. Assoc. 94: 739-743.

Ebel, J. and Hahlbrock, K. Biosynthesis. In Harborne, J. B. & Mabry, T. J. (Eds) (1982) The Flavonoids: Advances in Research. London: Chapman & Hall.

Eldridge, A.C. (1982a) Determination of isoflavones in soya bean flours, protein concentrates, and isolates. J. Agric. Food Chem. 30: 353-355.

Eldridge, A.C. (1982b) High performance liquid chromatography separation of soya bean isoflavones and their glucosides. J. Chromatogr. 234: 494-496.

Eldridge, A. C. & Kwolek, W. F. (1983) Soya bean isoflavones: Effect of environment and variety on composition. J. Agric. Food Chem. 31: 394-396.

Elghamry, M. I. & Hansel, R. (1969) Activity and isolated phytoestrogen of shrub palmetto fruits (Serenoa repens Small), a new estrogenic plant. Experientia 25: 828- 829.

Farmakalidis, E., Hathcock, J. N. & Murphy, P. A. (1985) Oestrogenic potency of genistin and daidzin in mice. Food Chem. Toxic. 23: 741-745.

Farmakalidis, E. & Murphy, P. A. (1985) Isolation of 6'-0-acetyldaidzein and 6'-0- acetylgenistein from toasted defatted soy flakes. J. Agric. Food Chem. 33: 385-389.

Farnsworth, N. R., Bingel, A.S., Cordell, G.A., Crane, F. A. & Fong, H. H. S. (1975) Potential value of plants as sources of new antifertility agents II. J. Pharm. Sci. 64: 717-754. 238

Ferrando, R., Guilleux, M. M. & Guerrillot-Vinet, A. (1961) Estrogen content of plants as a function of conditions of culture. Nature 23: 1205.

Fitzpatrick, L. A. (1999) Selective estrogen receptor modulators and phytoestrogens: new therapies for the postmenopausal woman. Mayo Clin. Proc. 74: 601-607.

Folkman, J. & Shing, Y. (1992) Angiogenesis. J. Biol. Chem. 267: 10931-10934.

Folman, Y. & Pope, G. S. (1969) Effect of acetate, dimethylstilboestrol, genistein and coumestrol on uptake of eHJ oestradiol by uterus, vagina and skeletal muscle of immature mice. J. Endocrinol. 44: 213-218.

Food Agriculture Organisation (1997) Food Balance Sheet http://apps.fao.org/lim 500/nph-wrap.pl?FoodBalanceSheet&Domain=FoodBalanceSheet.

Food Agriculture Organisation (1998) Food Balance Sheet http://apps.fao.org/lim 500/nph-wrap.pl?FoodBalanceSheet&Domain=FoodBalanceSheet.

Fotsis, T. & Adlercreutz, H. (1987) The multicomponent analysis of estrogens in urine by ion exchanges chromatography and GC-MS-I. Quantitation of estrogen after initial hydrolysis of conjugates. J. Steroid Biochem. 28: 203-213.

Fotsis, T., Pepper, M., Adlercreutz, H., Rase, T., Montesano, R. & Schweigere, L. ( 1995) Genistein, a dietary ingested isoflavonoid, inhibits cell proliferation and in vitro angiogenesis. J. Nutr. 125: 790S-796S.

Francis, C. M. & Millington, A. J. (1971) The presence of methylated coumestans in annual Medicago species: response to a fungal pathogen. Aust. J. Agric. Res. 22: 75- 80.

Franke, A.A., Custer, L.J., Cerna, C.M. & Naral, K.K (1994) Quantitation of phytoestrogens in legumes by HPLC. J. Agric. Food Chem. 42: 1905-1913.

Franke, A.A., Custer, L.J., Cerna, C.M. & Naral, K.K (1995) Rapid HPLC analysis of dietary phytoestrogens from legumes and from human urine. Proc. Soc. Exp. Biol. Med. 108: 18-26. 239

Franke, A.A., Hankin, J. H., Yu, M. C., Maskarinec, G., Low, S. & Custer, L. J. (1999) Isoflavone levels in soy foods consumed by multiethnic populations in Singapore and Hawaii. J. Agric. Food Chem. 47: 977-986.

Fukutake, M., Takahashi, M., Ishida, K., Kawamura, H., Sugimura, T. & Wakabayashi, K. (1996) Quantification of genistein and genistin in soya beans and soya bean products. Food Chern. Toxicol. 34:457-461.

Fung, M. C., Szeto, Y.Y., Leung, K. N., Wong-Leung, Y. L. & Mak, N. K. (1997) Effects of biochanin A on the growth and differentiation of myeloid leukemia WEHI-B (JCS) cells. Life Sci. 61: 105-115.

Gavaler, J. S., Rosenblum, E. R., Deal, S. R. & Bowie, T. B. (1995) The phytoestrogen congeners of alcoholic beverages: current status. Proc. Soc. Exp. Biol. Med. 208: 98-102.

Gawienowski, A. M. & Gibbs, C. C. (1969) The isolation of estrone from apple seeds. Phytochemistry 8: 685-686.

Gildersleeve, R.R., Smith, G.R., Pemberton, I. J. and Gilbert, C.L. (1991) Detection of isoflavones in seedling subterranean clover. Crop Sci. 31: 8 89-892.

Graham, T. L., Kim, J. E. & Graham, M. Y. (1990) Role of constitutive isoflavone conjugates in the accumulation of glyceollin in soybean infected with Phytophthora megasperma. MoZee. Plant Microbe Interact. 3: 157-166

Guggolz, J., Livingston, L. & Bickoff, E. M. (1961) Detection of daidzein, formononetin, genistei, and biochanin A in forages. J. Agric. Food Chem. 9: 330- 332.

Ha, E. Y. W., Morr, C. V. & Seo, A. (1992) Isoflavones aglucones and volatile organic compounds in soybeans: effects of soaking treatments. J. Food Sci. 57: 414- 417.

Hanson, C. H., Loper, G. M., Kohler, G. 0., Bickoff, E. M., Taylor, K. W., Kehr, W. R., Stanford, E. H., Dudley, J. W., Pedersen, M. W., Sorensen, E. L., Carnahan, H. L. & Wilsie, C. P. (1965) Variation in coumestrol content of alfalfa, as related to lacation, variety, cutting, year, stage of growth and disease. United States Department of Agriculture, Agricultural Research Service, Technical Bulletin 1333: 1-72. 240

Harborne, J. B., Ingham, J. 1., King, L. & Payne, M. (1976) The isopentenyl isoflavones as a pre-infectional antifungal agent in the genus Lupines. Phytochemistry 15: 1485.

Harborne, J. B. & Mabry, T. J. (Eds) (1982) The Flavonoids: Advances in Research. London: Chapman & Hall

Heftmann, E., Ko, S. & Bennett, R. D. (1966) Identification of estrone m pomegranate seeds. Phytochemistry 5: 1337-1339.

Hesseltine, C. W., Rogers, R. F. & Shotwell, 0. (1978) Fungi, especially , and zearalenone occurrence in wheat. Mycologia 70: 14-18.

Hirayama, T (1986) A large scale cohort study on cancer risk by diet- with special reference to the risk reducing effects of green-yellow vegetable consumption. Hayashi, Y., Nagao, M. & Sugimura, T. (eds) Diet, Nutrition, and Cancer. Utrecht, the Netherlands: Tokyo Scientific Press; 41-53.

Hodgson, J. M., Croft, K. D., Puddey, I. B., Mori, T. A. & Beilin, L. J. (1996) Soybean isoflavonoids and their metabolic products inhibit in vitro lipoprotein metabolism. J. Nutr. Biochem. 7: 664-669.

Huang, J., Nasr, M., Kim, Y. & Matthews, H. R. (1992) Genistein inhibits protein histidine kinase. J. Bioi. Chem. 267: 15511-15515.

Hutchins, A. M., Lampe, J. W., Martini, M. C., Campbell, D. R. & Slavin, J. L. (1995) Vegetables, fruits, and legumes: effect on urinary isoflavonoid phytoestrogen and lignan excretion. J. Am. Diet. Assoc. 95: 769-774.

Ikeda, R., Ohta, N. & Watanabe, T. (1995) Changes of isoflavones at various stages of fermentation in defatted soya beans. Nippon Shokuhin Kogaku Kogaku Kaishi 42: 322-327.

Ingram, D., Sanders, K., Lolybaba, M. & Lopez, D. (1997) Case-control study of phyto-oestrogens and breast cancer. Lancet 350: 990-994. 241

Ishirni, Y., Miyaura, C., Ohmura, M., Onoe, Y., Sato, T., Uchiyama, Y., Ito, M., Wang, X., Suda, T. & Ikegami, S. (1999) Selective effects of genistein, a soya bean isoflavone, on B-Lymphopoiesis and bone loss caused by estrogen deficiency. Endocrinology 140: 1893-1900.

Jackson, D. L. & Jacobs, S. W. L. (1985) Australian Agricultural Botany. Sydney: Sydney University Press.

Jacobs, E., Kulling, S. E. & Metzler, M. (1999) Novel metabolites of the mammalian lignans enterolactone and enterodiol in human urine. J. Steroid Biochem. MoZee. Bioi. 68: 211-218.

James, C. S. (1995) Analytical Chemistry of Foods. London: Blackie Academic & Professional.

Ioannou, G. E., Kelly, G. E., Reeder, A. Y., Waring, M. & Nelson, C. (1995) A urinary profile study of dietary phytoestrogens. The identification and mode of metabolism ofnew isoflavonoids. J. SteroidBiochem. MoZee. Bioi. 54: 167-184.

Kandulska, K., Nogowski, L. & Szkudelski, T. (1999) Effect of some phytoestrogens on metabolism of rat adipocytes. Repro. Nutr. Dev. 39: 497-501.

Keen, N. T. Zaki, A. I. & Sims, J. J. (1972) Biosynthesis of hydroxyphaseollin and related isoflavanoids in disease-resistant soya bean hypocotyls. Phytochemistry 11: 1031-1039.

Keinanen, M. (1993) Comparison of methods for the extraction of flavonoids from birch leaves (Betula pendula Roth.) Carried out using high-performance liquid chromatography. J. Agric. Food Chern. 41:1986-1990.

Kelly, G. E., Ioannou, G. E., Reeder, A. Y., Nelson, C. & Waring, M. A. (1995) The variable metabolic response to dietary isoflavones in humans. Proc. Soc. Exp. Biol Med. 208: 40-43.

Kelly, G. E., Nelson, C., Waring, M. A., Joannou, E. & Reeder, A. Y. (1993) Metabolites of dietary (soya) isoflavones in human urine. Clin. Chim. Acta 223: 9- 22. 242

King, R. A. & Bignell, C. M. (2000) Concentrations of isoflavones phytoestrogens and their glucosides in Australian soya beans and soya foods. Aust. J. Nutr. Diet. 57: 70-78.

Kitada, Y., Ueda, Y., Yamamoto, M., Ishikawa, M., Nakazawa, H. & Fujita, M. (1986) Determination of isoflavones in soy bean by high-performance liquid chromatography with amperometric detection. J. Chromatogr. 366: 403-406.

Kitts, D. D. (1986) Competitive binding of phytoestrogens in coffee extracts with 17~-estradiol to uterine cytosol receptors. Can. Inst. Food Sci. Technol. J. 19: 191- 194.

Knight, D. C., Eden, J. A. Huang, J. L. & Waring, M. A. (1998) Isoflavone content of infant foods and formulas. J. Paediatr. Child Health 34: 135-138.

Knuckles, B. E., deFremery, D. & Kohler, G.O. (1976) Coumestrol content of fractions obtained during wet processing of alfalfa J. Agric. Food Chem. 24: 1177- 1180.

Ko, K. & Malison, J. A. (1999) Effect of genistein on the growth and reproductive function of male and female yellow perch Perea flavescens. J. World Aquaculture Soc. 30:73-77.

Kompas (1998) Swasembada kedelai, bukan tak mungkin (It is not possible for Indonesia to be selfsufficient in soya beans). Jakarta, Indonesia.

Kudou, S., Fleury, Y., Welti, D., Nagnolato, D., Uchida, T., Kitamura, K. & Okubo, K. (1991) Malonyl isoflavone in soya bean seeds (Glycine max. MERRILL). Agric. Bioi. Chem. 55:2227-2233.

Kulling, S. E. & Metzler, M. (1997) Induction of micronuclei, DNA strand breaks and HPRT mutations in cultured Chinese hamster V79 cells by the phytoestrogen coumestrol. Food Chem. Toxicol. 35: 605-613.

Kurzer, M. S., Lampe, J. W., Martini, M. C. & Adlercreutz, H. (1995) Fecallignan and isoflavonoid excretion in premenopausal women consuming flaxseed powder. Cancer Epidemiol. Biomark. Prev. 4: 353-358. 243

Kurzer, M. S. & Xu, X. (1997) Dietary phytoestrogens. Annu. Rev. Nutr. 17: 353- 381.

Lamartiniere, C. A., Moore, J., Holland, M. & Barnes, S. (1995) Neonatal genistein chemoprevents mammary cancer. Proc. Soc. Exp. Bioi. Med. 208: 120-123.

Lampe, J. W., Martini, M. C., Kurzer, M. S., Adlercreutz, H. & Slavin, J. L. (1994) Urinary lignan and isoflavonoid excretion in premenopausal women consuming flaxseed powder. Am. J. Clin. Nutr. 60: 122-128.

Lane, G. A. & Newman, R. H. (1987) Isoflavones from Lupinus Angustifolius root. Phytochemistry 26: 295-300.

Lapcik, 0., Hill, M., Hampl, R., Wahala, K. & Adlercreutz, H. (1998) Identification of isoflavonoids in beer. 63: 14-20.

Leavitt, W. W. & Meismer, D. M. (1968) Sexual development altered by non­ steroidal oestrogens. Nature 218: 181-182.

Lee, H. P., Gourley, L., Duffy, S. W., Esteve, J., Lee, J. & Day, N. E. (1991) Dietary effects on breast-cancer risk in Singapore. Lancet 337: 1197-1200.

Lee, Y. J., Notides, A. C., Tsay, Y. & Kende, A. S. (1977) Coumestrol, NDB­ Norhexestrol, and dansyl-norhexestrol fluorescent probes of estrogen binding proteins. Biochemistry 16: 2896-2901.

Leopold, S. A., Erwin, M., Oh, J. & Browning, B. (1976) Phytoestrogens: adverse effects on reproduction in California quail. Science 191: 98-100.

Levin, E., Burns, J. F. & Collins, V. K. (1951) Estrogenic, androgenic and gonadotrophic activity in wheat germ oil. Endocrinology 49: 289-301.

Levy, J. R., Faber, K. A., Ayyash, L. & Hughes, C. L. (1995) The effect of prenatal exposure to the phytoestrogens genistein on sexual differentiation in rats. Proc. Soc. Exp. Bioi. Med. 208: 60-66. 244

Lightfoot, R. J., Croker, K. P. & Neil, H. G. (1967) Failure of sperm transport in relation to ewe infertility following prolonged grazing on oestrogenic pastures. Aust. J. Agric. Res. 18: 755-765.

Linassier, C., Pierre, M., Le Pecq, J. & Pierre, J. (1990) Mechanisms of action in NIH-3T3 cells of genistein, an inhibitor of EGF receptor tyrosine kinase activity. Biochern. Pharm. 39: 187-193.

Liu, K. (1997) Soya Beans. New York: Chapman & Hall.

Livingston, A. L., Bickoff, E. M., Guggolz, J. & Thompson, C. R. (1961) Quantitative determination of coumestrol in fresh and dried alfalfa. J. Agric. Food Chern. 9: 135-137.

Lookhart, G. L. (1979) Note on an improved method of extracting and quantitating coumestrol from soya beans. Cereal Chern. 56: 386-388.

Lookhart, G. L., Jones, B. L. & Finney, K. F. (1978) Determination of coumestrol in soya beans by high performance liquid and thin-layer chromatography. Cereal Chern. 55: 967-972.

Lundh, T. (1995) Metabolims of estrogenic isoflavones in domestic animals. Proc. Soc. Exp. Biol. Med. 208: 33-39.

Lyn-Cook, B. D., Blann, E., Payne, P. W., Bo, J., Sheehan, D. & Medlock, K. (1995) Methylation profile and amplification of pro-oncogenes in rat pancreas induced with phytoestrogens. Proc. Soc. Exp. Biol. Med. 208: 116-119.

Makela, S., Poutanen, M., Lehtimaki, J., Kostian, M. L., Santti, R. & Vihko, R. (1995) Estrogen-specific 17~-hydroxysteroid oxidoreductase type 1 (B.C. 1.1.1.62) as a possible target for the action of phytoestrogens. Proc. Soc. Exp. Biol. Med. 208: 51-59.

Markaverich, B. M., Webb, B., Densmore, C. L. & Gregory, R. R. (1995) Effects of coumestrol on estrogen receptor function and uterine growth in ovariectomised rats. Environ. Health Perspect. 103: 574-581.

Markiewicz, L., Garey, J., Adlercreutz, H. & Gurpide, E. (1993) In vitro bioassays of non-steroidal phytoestrogens. J. Steroid Biochern. MoZee. Biol. 45: 399-405. 245

Markovits, J., Linassier, C., Fosse, P., Couprie, J., Pierre, J., Jacquemin-Sablon, A., Saucier, J., Le Pecq, J. & Larsen, A. K. (1989) Inhibitory effects of the tyrosine kinase inhibitor genistein on mammalian DNA topoisomerase II. Cancer Res. 49: 5111-5117.

Marrian, G. F. & Haselwood, G. D. A. (1932) Equol, a new inactive phenol isolated from ketohydroxyostrin fraction of mares urine. Biochem. J. 26: 1226-1232.

Martin, P. M., Horwitz, K. B., Ryan, D. S. & Mcguire, W. L. (1978) Phytoestrogen interaction with estrogen receptors in human breast cancer cells. J. Endocrinol. 103: 1860-1867.

Matsuura, M. & Obata, A. (1993) ~-glucosidases from soybeans hydrolyse daidzin and genistin. J. Food Sci. 58: 144-147.

Matsuura, M., Obata, A., & Fukushima, D. (1989) Objectionable flavor of developed during the soaking of soybeans and its control. J. Food Sci. 54: 602.

Mazur, W., Fotsis, T., Wahala, K., Ojala, S., Salakka, A. & Adlercreutz, H. (1996) Isotope dilution gas chromatographic-mass spectrometric method for determination of isoflavonoids, coumestrol, and lignans in food samples. J. Anal. Biochem. 233: 169-180.

Mazur, W. M., Wahala, K., Rasku, S., Salakka, A., Hase, T. & Adlercreutz, H. (1998) Lignan and isoflavonoid concentrations in tea and coffee. Br. J. Nutr. 79: 37- 45.

Meagher, L. P., Beecher, G. R., Flanagan, V. P. & Li, B. T. (1999) Isolation and characterization of the lignans, isolariciresinol and pinoresinol, in flaxseed meal. J. Agric. Food Chern. 47: 3173-3180.

Medlock, K. L., Branham, W. S. & Sheehan, D. M. (1995) Effects of coumestrol and equol on the developing reproductive tract of the rat. Proc. Soc. Exp. Biol. Med. 208: 67-71.

Miksicek, R. J. (1993) In situ localization of the estrogen receptor in living cells with the fluorescent phytoestrogens coumestrol. J. Histochem. Cytochem. 41: 801-810. 246

Miksicek, R. J. (1995) Estrogenic flavonoids: structural requirements for biological activity. Proc. Soc. Exp. Biol. Med. 208:44-50.

Mirocha, C. J., Schauerhamer, B. & Pathre, S. V. (1974) Isolation, detection, and quantitation of zearalenone in maize and barley. J. Assoc. Off. Anal. Chern. 57: 1104- 1110.

Mohanty, G. B. & Roy, A. B. (1955) Report on wax from several species of Tillandsia and fromAnanas comosus (L.) Merr. Science 121: 42.

Murakami, H., Asakawa, T., Terao, J. & Matsushita, S. (1984) Antioxidative stability of tempeh and liberation of isoflavones by fermentation. Agric. Biol. Chern. 48: 2971-2975.

Murkies, A. L., Lombard, C., Strauss, B. J. G., Wilcox, G., Burger, H. G. & Morton, M. S. (1995) Dietary flour supplementation decreases post-menopausal hot flushes: effect of soy and wheat. Maturitas 21: 189-195.

Murphy, P.A. (1981) Separation of genistin, daidzin and their aglucones, and coumestrol by gradient high-performance liquid chromatography. J. Chromatogr. 211: 166-169.

Murphy, P. A. (1982) Phytoestrogen content of processed soya bean products. Food Techno!. 36: 60-64.

Murphy, P. A., Farmakalidis, E. & Johnson, L. D. (1982). Isoflavone content of soya-based laboratory animal diets. Food Chern. Toxicol. 20: 315-317.

Murphy, P. A., Song, T., Buseman, G. & Barua, K. (1997) Isoflavones in soy-based infant formulas. J. Agric. Food Chern. 45: 4635-4638.

Murphy, P. A., Song, T., Buseman, G., Barua, K., Beecher, G. R., Trainer, D. & Holden, J. (1999) Isoflavones in retail and institutional soy foods. J. Agric Food Chern. 47: 2697-2704.

Musey, P. I., Adlercreutz, H., Gould, K. G., Collins, D. C., Fotsis, T., Bannwart, C., Makalela, T., Wahala, K., Brunow, G. & Hase, T. (1995) Effect of diet on lignans and isoflavonoid phytoestrogens in chimpanzees. Life Sci. 57: 655-664. 247

Nairn, M., Gestetner, B., Kirson, I., Birk, Y. and Bondi, A. (1973) A new isoflavone from soya beans. Phytochemistry 12: 169-170.

Nairn, M., Gestetner, B., Zilkah, S., Birk, Y. & Bondi, A. (1974) Soya bean isoflavones. Characterisation, determination, and antifungal activity. J. Agric. Food Chern. 22: 806-810.

Newsome, F. E. & Kitt, W. D. (1975) A method for separating estrogens and phytoestrogens on Shepadex LH-20 columns. Steroids 26: 215-217.

Noteboom, W. D. & Gorski, J. (1963) Estrogenic effect of genistein and coumestrol diacetate. Endocrinology 73: 736-739.

Nwannenna, A. I., Lundh, T. J-0., Madej, A., Fredriksson, G. & Bjornhag, G. (1995) Clinical changes in overiectomized ewes exposed to phytoestrogens and 17~­ estradiol implants. Proc. Soc. Exp. Biol. Med. 208: 92-97.

Obermeyer, W.R., Musser, S.M., Betz, J.M., Casey, R.E., Pohland, A.B. and Page, S.W. (1995) Chemical studies of phytoestrogens and related compounds in dietary supplements: flax and chaparral. Proc. Soc. Exp. Biol. Med. 208: 6-12.

Ogawara, H., Akiyama, T., Watanabe, S., Ito, N., Kobori, M. & Seoda, Y. (1989) Inhibition of tyrosine protein kinase activity by synthetic isoflavones and flavones. J. Antibiotic. 42: 340-343.

Ohta, N., Kuwata, G., Akahori, H. & Watanabe, T. (1979) Isoflavonoid constituents of soya beans and isolation of a new acetyl daidzin. J. Agric. Biol. Chern. 43: 1415- 1419.

Olah, A. F. & Sherwood, R. T. (1973) Glycosidase activity and flavonoid accumulation in alfalfa infected by ascochyta imperfecta. Phytopathology 63: 739- 742.

Paoletti, R., Crosignani, P. G., Kenemans, P., Wenger, N. K. & Jackson, A. S. (1999) Women's health and menopause: risk reduction strategies improved quality of health. Dordrecht; Boston: Kluwer Academic Publishers. 248

Parkin, D. M., Arslan, A., Bieber, A., Bouvy, 0., Muir, C. S., Owor, R. & Whelan, S. (Eds). (1986) Cancer occurrence in developing countries. Lyon: IARC.

Parkin, D. M., Muir, C. S., Whelan, S. L., Gao, Y. T., Ferlay, J. & Powell, J. (1992) Cancer incidence in five continents. Lyon: IARC.

Parkin, D. M., Muir, C. S., Whelan, S. L., Gao, Y. T., Ferlay, J. & Powell, J. (1997) Cancer incidence in five continents. Lyon: IARC.

Pelter, A., Bradshaw, J. & Warren, R. F. (1971) Oxidation experiments with :flavonoids. Phytochemistry 10: 835-850.

Peterson, G. & Barnes, S. (1991) Genistein inhibition of the growth of human breast cancer cells: independence from estrogen receptors and the multi-drug resistance gene. Biochem. Biophys. Res. Commun. 179: 661-667.

Peterson, G. & Barnes, S. (1993) Genistein and biochanin A inhibit the growth of human prostate cancer cells but not epidermal growth factor receptor tyrosine authophosphorylation. Prostate 22: 335-345.

Plewa, M. J., Wagner, E. D. Berhow, M. A., Conway, A., Rayburn, A. L. & Anderson, D. (1999) Antimutagenic activity of chemical fractions isolated from a commercial soya bean processing by-product. Teratogenesis, Carcinogenesis Mutagenesis 19: 121-135.

Price, K. R. & Fenwick, G. R. (1985) Naturally occurring oestrogens in foods- A review. J. FoodAddit. Contam. 2: 73-106.

Register, B., Bethel, M. A., Thompson, N., Walmer, D., Blohm, P., Ayyash, L. & Hughes, C. (1995) The effect of neonatal exposure to diethylstilbestrol, coumestrol, and ~-sitosterol on pituitary responsiveness and sexually dimorphic nucleus volume in the castrated adult rat. Proc. Soc. Exp. Bioi. Med. 208: 72-84.

Richardson, T. & Mattarella, N. (1977) Hormonal substances in human milk, cow's milk and dairy products. J. Food Protection 40: 57-64.

Robards, K., Haddad, P. R. & Jackson, P. E. (1994) Principles and practice of modern chromatograpic methods. London: Academic Press. 249

Robinson, T. J. (1948) Oestrogenic potency of subterranean clover (T. Subterraneum L. var. Dwalganup ): the preparation and assay of extracts. Australian. J. Exp. Biol. Med. Sci. 27: 297-305.

Rose, D. P., Boyar, A. P. & Wynder, E. L. (1986) International comparisons of mortality rates for cancer of the breast, oveary, prostate, and colon, and per capita food consumption. Cancer 58: 2363-2371.

Saloniemi, H., Wahala, K., Nykanen-Kurki, P., Kallela, K. & Saastamoinen, I. (1995) Phytoestrogen content and estrogenic effect of legume fodder. Proc. Soc. Exp. Biol. Med. 208: 13-17.

Salunke, D. K. & Kadam, S. S. (Eds) (1989) Handbook of world food legumes: nutritional chemistry, processing technology, and utilisation. Boca Raton, Florida: CRC Press.

Scott, P. M., Panalaks, T., Kanhere, S. & Miles, W. F. (1978) Determination of zearalenone in cornflakes and other corn-based foods by thin layer chromatography, high pressure liquid chromatography, and gas-liquid chromatography/high resolution mass spectrometry. J. Assoc. Off. Anal. Chern. 61: 593-600.

Seo, A., & Morr, C.V. (1984) Improved high-performance liquid chromatographic analysis of phenolic acids and isoflavonoids from soya bean protein products. J. Agric. Food Chern. 32: 530-533.

Setchell, K. D. R., Borriello, S. P., Hulme, P., Kirk, D. N. & Axelson, M. (1984) estrogens of dietary origin: possible roles in hormone-dependent disease. Am. J. Clin. Nutr. 40: 569-578.

Setchell, K. D. R., Lawson, A. M., Borriello, S. P., Harkenss, R., Gordon, H., Morgan, D. M. L., Kirk, D. N., Adlercreutz, H., Anderson, L. C. & Axelson, M. (1981b) Lignan formation in man-microbial involvement and possible roles in relation to cancer. Lancet 2: 4-7.

Setchell, K. D. R., Lawson, A. M., Conway, E., Taylor, N. F., Kirk, D. N., Cooley, G., Farrant, R. D., Wynn, S. & Axelson, M. (1981a) The definitive identification of the lignans trans-2,3-bis (3-hyroxybenzyl)-g-butyrolactone and 2,3-bis (3- hydroxybenzyl) butane-1,4-diol in human and animal urine. Biochem. J. 197: 447- 458. 250

Setchell, K. D. R., Lawson, A. M., Mitchell, F. L., Adlercreutz, H. & Kirk, D. N. (1980) Lignans in man and animal species. Nature 287: 740-742.

Setchell, K. D. R. & Welsh, M. B. (1987) High-performance liquid chromatographic analysis of phytoestrogens in preparations with ultraviolet, electrochemical and thermospray mass spectrometric detection. J. Chromatogr. 386: 315-323.

Sharaf, A. & Negm, S. (1973) Oestrogenic activity of Zea mays oil. Qual. Plant. Mater. Veg. 22: 249-252.

Sharma, R. D. (1979) Isoflavones and hypercholesterolemia in rats. Lipids 14: 535- 540.

Shemesh, M., Lindner, H. R. & Ayalon, N. (1972) Affinity of rabbit uterine oestradiol receptor for phyto-oestrogens and its use in a competitive protein-binding radioassay for plasma coumestrol. J. Reprod. Fertil. 29: 1-9.

Shihabi, A. K., Kute, T., Garcia, L. L. & Hinsdale, M. (1994) Analysis of isoflavones by capillary electrophoresis. J. Chromatogr. 680: 181-185.

Shotwell, 0. L., Hesseltine, C. W., Goulden, M. L. & Vandegraft, E. E. (1970) Survey of corn for , zearlenone, and . Cereal. Chern. 47: 700-707.

Shutt, D. A. & Braden, A. W. H (1968) The significance of equol in relation to the oestrogenic responses in sheep ingesting clover with a high formononetin content. Aust. J. Agric. Res. 19: 545-553.

Shutt, D. A. & Cox, R. I. (1972) Steroid and phyto-oestrogen binding to sheep uterine receptors in vitro. J. Endocrinol. 52: 299-310.

Shutt, D. A., Weston, R. H. & Hogan, J. P. (1970) Quantitative aspects of phyto­ oestrogen metabolism in sheep fed on subterranean clover (Trifolium subterraneum cultivar Clare) or red clover (Trifolium pratense ). Aust. J. Agric. Res. 31: 713-722.

Siddiqui, M. T. & Siddiqui, M. (1976) Hypolipodemic principles of Cicer arietinum Biochanin A and formononetin. Lipids 11: 243-246. 251

Smith, D. A. & Banks, S. W. (1986) Biosynthesis, elicitation and biological activity ofisoflavonoid phytoalexins. Phytochemistry 25: 979-995.

Smith, G. R., Randel, R. D. & Bradshaw, C. (1986) Influence of harvest date, cultivar, sample storage method on concentration of isoflavones in subterranean clover. Crop Sci. 26: 1013-1016.

Snedecor, G. W. & Cochran, W. G. (1989) Statistical methods. Iowa: Iowa State University Press.

Song, T. T., Hendrich, S. and Murphy, P. A. (1999) Estrogenic activity of glycitein, a soybean isoflavone. J. Agric. Food Chem. 47: 1607-1610.

Stass, S. A. (ed) (1987) The acute leukemias, biologic, dianostic, and therapeutic determinats. NY: Marcel Dekker.

Steele, V. E., Pereira, M., Sigman, C. & Kelloff, G. (1995) Cancer chemoprevention agent development strategies for genistein. J. Nutr. 125: 713S-716S.

Stitch, S. R., Toumba, J. K., Groen, M. B., Funke, C. W., Leemhuis, J., Vink, J. & Woods, G. F. (1980) Excretion, isolation and structure of a new phenolic constituent offemale urine. Nature 287: 738-740.

Stoloff, L. & Dalrymple, B. (1977) Aflatoxin and zearalenone occurrence in dry­ milled com products. J. Assoc. Off. Anal. Chem. 60: 579-582.

Stoloff, L., Henry, S. & Francis, 0. J. (1976) Survey for and zearalenone in 1973 crop com stored on farms and in country elevators. J. Assoc. Off. Anal. Chem. 59: 118-121.

Storm, S. S., Yamamura, Y., Duphome, C. M., Spitz, M. R., Babaian, R. J, Pillow, P. C. & Hursting, S. D. (1999) Phytoestrogen intake and prostate cancer: case-control study using a new database. Nutr. Cancer 33: 20-25.

Swanson, S. P., Corley, R. A., White, D. G. & Buck, W. B. (1984) Rapid thin layer chromatographic method for determination of zearalenone and zearalenol in grains and animal feeds. J. Assoc. Off. Anal. Chem. 67: 580-584. 252

Tang, B. Y. & Adams, N. R. (1980) Effect of equol on oestrogen receptors and on synthesis of DNA and protein in the immature rat uterus. J. Endocrinol. 85: 291-297.

Thomas, J. P. (1999) Hormone replacement therapy for menopausal women. Amsteldijk: Harwood academic publishers.

Thompson, L. U., Robb, P., Serrino, M. & Cheung, F. (1991) Mammalian lignan production from various foods. Nutr. Cancer 16: 45-52.

Tsutsumi, N. (1995) Effect of coumestrol on bone metabolism in organ culture. Biol. Pharm. Bull. 18: 1012-1015.

Uckun, F. M., Evans, W. E., Forsyth, C. J., Waddick, K. G., Ahlgren, L. T., Chelstrom, L. M., Burkhardt, A., Bolen, J. & Myers, D. E. (1995) Biotherapy of B­ cell precursor leukemia by targeting genistein to CD19-associated tyrosine kinases. Science 267: 886-891.

USDA-Iowa State University database on the isoflavone content of foods, release 1.1 (1999) http://www. nal. usda.gov/fnic/foodcomp/Data/isoflav/isoflav. html.

Vander Maesen, L. J. G & Somaatmadja, S. (1989) Pulses. Wageningen: Pudoc

Verdeal, K. & Ryan, D. S. (1979) Naturally-occurring estrogens in plant foodstuffs­ A review. J. Food Protection 42: 577-583.

Vetter, J. (1995) Isoflavones in different parts of common Trifolium species. J. Agric. Food Chem. 43: 106-108.

Waagle, D. H., Steinke, F. H. & Shen, J. L. (1989) Isolated soy proteins. Matthews, R. H. (ed.) Legumes. New York: Marcel Dekker Inc.; 99-138.

Wahala, K. & Hase, T. (1991) Expedient synthesis of polyhydroxyisoflavones. J. Chem. Soc. Perkin Trans 1: 3005-3008.

Wahala, K., Hase, T. & Adlercreutz, H. (1995) Synthesis and labeling of isoflavone phytoestrogens, including daidzein and genistein. Proc. Soc. Exp. Rio. Med. 208: 27- 32. 253

Wahala, K., Makela, T., Backstrom, R., Brunow, G. & Hase, T. (1986) Synthesis of the (2H)-labelled urinary lignans, enterolactone and enterodiol, and the phytoestrogen daizein and its metabolites equol and 0-desmethylangolensin. J. Chem. Soc. Perkin Trans I: 95-98.

Walter, E. D. (1941) Genistin (an isoflavone glucoside) and its aglucone, genistein, from soybeans. J. Am. Chem. Soc. 63: 3273-3276.

Walz, E. (1931) Isoflavon und saponin-glucoside in soja hispida. Ann. 489: 118-155.

Wang. G., Kuan, S. S., Francis, 0. J., Ware, G. M. & Carman, A. S. (1990) A simplified HPLC method for determination of phytoestrogens in soya bean and its processed products. J. Agric. Food Chem. 38: 185-190.

Wang, H. & Murphy, P. A. (1994a) Isoflavone content in commercial soya bean foods. J. Agric. Food. Chem. 42: 1666-1673.

Wang, H. & Murphy, P. A. (1994b) Isoflavone composition of American and Japanese soya beans in Iowa: Effects of variety, crop year, and location. J. Agric. Food Chem. 42: 1674-1677.

Wang, H. & Murphy, P. A (1996) Mass Balance study of isoflavones during soya bean processing. J. Agric. Food Chem. 44: 2377-2383.

Wang, W., Tanaka, Y., Han, Z. Higuchi, C. M. (1995) Proliferative response of mammary glandular tissue to formononetin. Nutr. Cancer 23: 131-140.

Ware, G. M. & Thorpe, C. W. (1978) Determination of zearalenone in corn by high pressure liquid chromatography and fluorescence detection. J. Assoc. Off. Anal. Chem. 61: 1058-1062.

Waterman, P. G. & Mole, S. (1994) Analysis of phenolic metabolites. Oxford: Blackwell Scientific Publications.

Wei, H., Bowen, R., Cai, Q., Barnes, S. & Wang, Y. (1995) Antioxidant and antipromotional effects of the soya bean isoflavone genistein. Proc. Soc. Exp. Bioi. Med. 208: 124-126. 254

Welshons, W. V., Murphy, C. S., Koch, R., Calaf, G. & Jordan, V. C. (1987) Stimulation of breast cancer cells in vitro by the environmental estrogen enterolactone and the phytoestrogens equol. Breast Cancer Res. Treat. 10: 169-175.

West, L. G., Birac, P. M. & Pratt, D. E. (1978) Separation of the isomeric isoflavones from soya beans by high-performance liquid chromatography. J. Chromatogr. 150: 266-268.

Whitten, P. L., Lewis, C. & Naftolin, F. (1993) A phytoestrogens diet induces the premature anovulatory syndrome in lactationally exposed female rats. Bioi. Reprod. 49: 1117-1121.

Whitten, P. L. & Naftolin, F. (1992a) Effects of a phytoestrogen diet on estrogen­ dependent reproductive processes in immature female rats. Steroids 57: 56-61.

Whitten, P. L., Russell, E. & Naftolin, F. (1992b) Effects of normal, human­ concentration, phytoestrogens diet on rat uterine growth. Steroids 57: 98-106.

Whitten, P. L., Russell, E. & Naftolin, F. (1994) Influence ofphytoestrogens diets on estradiol action in the rat uterus. Steorids 59: 443-449.

Wilcox, G., Wahlqvist, M. L., Burger, H. G. & Medley, G. (1990) Oestrogenic effects of plant foods in postmenopausal women. Br. Med. J. 301: 905-906.

Winarno, F. G. ( 1979) Fermented vegetable protein and related foods of Southeast Asia with special reference to Indonesia. J. Am. Oil Chern. Soc. 56: 363-366.

Wolfbeis, 0. S. & Schaffner. (1980) Solvent and acidity dependence of the absorption and fluorescence of coumestrol. Photochem. Photobiol. 32: 143-148.

Wong, E. & Flux, D.S. (1962) The estrogenic activity of red clover isoflavones and some of their degradation products. J. Endocrinol. 24: 341-348.

Wong, E., Mortimer, P. I. & Geissman, T. A (1965) Flavonoid constituents of Cicer arietinum. Phytochemistry 4: 89-95.

World Health Organization (1991) WHO Mortality Databank http://www.dep. iarc.fr/cgi-bonlcgisqVwho2 country.idc. 255

World Health Organization (1992) WHO Mortality Databank http://www.dep. iarc.fr/cgi-bonlcgisq1/who2 country.idc.

World Health Organization (1993) WHO Mortality Databank http://www.dep. iarc.fr/cgi-bonlcgisq1/who2 countcy.idc.

World Health Organization (1994) WHO Mortality Databank http://www.dep. iarc.fr/cgi-bonlcgisq1/who2 countcy.idc.

World Health Organization (1995) WHO Mortality Databank http://www.dep. iarc.fr/cgi-bonlcgisq1/who2 country.idc.

World Health Organization (1996) WHO Mortality Databank http://www.dep. iarc.fr/cgi-bonlcgisq1/who2 country.idc

Xu, X., Wang, H., Murphy, P. A., Cook, L. & Hendrich, S. (1994) Daidzein is a more bioavailable soymilk isoflavone than is genistein in adult women. J. Nutr. 124: 825-832.

Xu. X., Harris, K. S., Wang. H., Murphy, P. & Hendrich, S. (1995) Bioavailability of soybean isoflavones depends on gut microflora in women. J. Nutr. 125: 2307:2315.

Young, I. J., Knights, B. A. and Hilman, J. R. (1977) Oestradiol and its biosynthesis in Phaseolus vulgaris L. Nature 267: 429.

Zheng, W., Dai, Q., Custer, L. J., Shu, X., Wen, W., Jin, F. & Franke, A. A. (1999) Urinary excretion of isoflavonoids and the risk of breast cancer. Cancer Epidemiol. Biomark. Prev. 4: 353-358.

Zilg, H. & Grisebach, H. (1968) Biosynthesis of isoflavones-XVII. Identification and biosynthesis of coumestans in Soja hispida. Phytochemistry 7: 1765-1772.

Zilg, H. & Grisebach, H. (1969) Coumestanes in Cicer arietinum. Phytochemistry 8: 2261-2263. Appendix 1. Levels of iso:flavones and coumestrol in soya beans and soya bean products (mg/100 g, wet weight basis).

Foods Daidzin Genistin Glycitin Daidzein Genistein Glycitein Coumestrol Formononetin Biochanin Author A Soya beans

Australia 5 varieties 34.9-68.4 45.8-84.7 ND 0.3-1.0* 0.8-1.1* NA NA NA NA Dalais et al. (1997)

4 varieties 17.3-48.7 26.4-63.2 6.0-9.7 36.5-112 1.3-3.2 11.7-18.8 NA NA NA King & Bignell (2000) USA

Soya beans 85.3 139.0 0.0 6.6 5.4 NA NA NA NA Coward et al. (1993)

4 varieties NA NA NA 10.5-56.0 26.8-84.1 NA 0.0-0.18 0.02-0.12 0.0-0.02 Mazur et al. (1998)

2 varieties 0.0, 11.7 74.4, 102.4 NA 0.1, 2.2 2.4, 4.0 NA ND NA NA Murphy (1982)

2 varieties 18.0, 69.0 39.4, 85.2 5.3, 5.6 24.0, 60.0 64.8, 95.4 8.2, 10.7 NA NA NA Wang & Murphy (1994b) Soya beans, 46.0 55.1 6.8 56.3 86.9 19.3 NA NA NA Wang & Murphy dried, (1994b) roasted Soya beans, 45.1 43.0 4.8 54.6 72.9 7.9 NA NA NA Wang & Murphy fresh (1994b) Singapore

Soya beans, 25.7 25.1 3.5 91.3 76.3 11.4 NA NA NA Franke et al. (1999) dried Soya beans, 22.7 21.8 3.3 64.9 54.1 9.2 NA NA NA Franke et al. (1999) dried, cooked

256 Hawaii

Soya beans, 41.3 76.8 1.9 75.0 131.7 7.5 NA NA NA Franke et al. (1999) dried Soya beans, 34.7 39.8 4.1 83.0 89.9 7.4 NA NA NA Franke et al. (1999) dried, roasted Soya beans, 1.6 2.0 1.3 7.6 7.1 4.6 NA NA NA Franke et al. (1999) fresh, cooked Japan Soya beans, NA 20.1 NA NA 0.5 NA NA NA NA Futukake et al. (1996) dried Soya nuts NA 96.8 NA NA 1.2 NA NA NA NA Futukake et al. (1996)

Soya milk

Australia Soya milk 2.4 5.8 NA 0.2 0.5 NA NA NA NA Dalais et al. (1997)

Soya milk, 1.4-7.1 2.4-14.3 0.5-0.7 1.6-7.8 3.4-15.2 0.5-0.7 NA NA NA King & Bignell 5 brands (2000)** Powdered 8.6-20.2 13.3-41.0 1.7-8.4 24.8-42.5 42.7-86.6 3.4-15.3 NA NA NA King & Bignell soya drink (2000)** mixes, 4 brands Cow's milk, NA NA NA 0.01-0.09* 0.0-0.01* NA NA NA NA Knight et al. (1998) 3 brands Soya milk, NA NA NA 1.08-3.15* 1.21-4.80* NA NA NA NA Knight et al. (1998) 5 brands

257 Japan

Soya milk, NA 9.5, 13.3 NA NA 0.2, 1.2 NA NA NA NA Futukake et al. (1996) 2 brands Soya powder NA 46.4 NA NA 1.8 NA NA NA NA Futukake et al. (1996)

USA Soya milk, 2.0-9.4 2.3-13.0 004-1.6 3.70-6.30 5.00-9.20 0.4-1.1 ND ND ND Murphy et al. (1999) 12 brands Instant soya 40.4-52.5 67.4-74.5 6.8-7.7 29.5-40.7 56.0-66.5 10.5-11.1 NA NA NA Wang & Murphy milk, (1994b) 4 brands Soya-based 3.1-7.6 6.5-13.4 1.7-2.1 5.7-7.8 12.6-15.4 1.7-3.0 NA NA NA Murphy et al. (1997) infant formula, 6 brands Soya milk, NA NA NA 0.02-0.70 0.10-2.10 NA NA NA NA Dwyer et al. (1994) 4 brands Soya milk 10.3 13.0 NA 1.1 0.7 NA NA NA NA Coward et al. (1993)

Hawaii Soya protein 7.1 10.1 2.2 17.0 23.6 6.0 NA NA NA Franke et al. (1999) drink Soya shake 20.7 36.7 3.2 37.3 61.7 10.6 NA NA NA Franke et al. (1999)

Soya energy 3.0 44.7 0.6 8.7 14.2 4.7 NA NA NA Franke et al. (1999) shake Soya diet 11.4, 5.1 17.7, 7.4 2.3, 1.0 37.1, 14.3 59.6, 22.6 14.3, 12.3 NA NA NA Franke et al. (1999) shake, 2 brands Soya milk, 0.5, 2.3 0.4, 2.8 0.2, 0.5 2.4, 0.5 3.1, 0.4 0.2, 0.5 NA NA NA Franke et al. (1999) 2 brands

258 Tofu

Japan Tofu NA 13.8 NA NA 1.4 NA NA NA NA Futukake et al. (1996) USA Extra firm, 5.9 9.9 1.8 9.1 13.8 2.2 ND ND ND Murphy et al. (1999) composite Extra firm, 5.0 8.3 1.5 7.4 11.1 1.7 ND ND ND Murphy et al. (1999) raw Extra firm, 5.8 9.8 1.7 8.0 12.8 2.0 ND ND ND Murphy et al. (1999) cooked Firm tofu, 6.6, 6.8 9.0, 9.7 1.4,1.6 11.3, 12.2 14.6, 14.8 2.0, 2.1 ND ND ND Murphy et al. (1999) 2 brands Firm tofu, 7.4, 7.8 10.0, 11.3 1.8, 2.1 12.7, 13.0 15.8, 16.5 2.4, 2.5 ND ND ND Murphy et al. (1999) cooked, 2 brands Firm, 4.0-10.5 6.4-13.5 1.1-1.8 8.4-14.6 12.1-17.8 1.7-2.2 ND ND ND Murphy et al. (1999) composite, 3 brands Regular 6.4 10.2 1.5 9.8 14.5 1.8 ND ND ND Murphy et al. (1999)

Firm, silken 10.5 14.3 2.9 8.6 12.9 2.4 ND ND ND Murphy et al. (1999)

Tofu, 0.0-3.5 5.1-10.4 NA 0.0-1.3 2.9-7.8 NA NA NA NA Murphy (1982) 3 brands Tofu 2.5 8.4 0.8 14.6 16.2 2.9 NA NA NA Wang & Murphy (1994) Firm tofu, NA NA NA 7.6 21.3 NA NA NA NA Dwyer et al. (1994)

Soft tofu, NA NA NA 7.3, 9.8 18.7, 21.6 NA NA NA NA Dwyer et al. (1994) 2 brands

259 Silken tofu NA NA NA 8.6 20.7 NA NA NA NA Dwyer et al. (1994)

Tofu 12.1-20.0 24.9-26.9 NA 1.5-1.6 1.5-3.1 NA NA NA NA Coward et al. (1993)

Singapore Tofu, raw 5.2 6.1 0.8 13.9 14.1 1.7 NA NA NA Franke et al. (1999)

Tofu, cooked 4.6 5.5 0.7 12.4 11.9 1.5 NA NA NA Franke et al. (1999)

Taukwa, 5.0 6.2 0.8 13.6 13.9 2.0 NA NA NA Franke et al. (1999) raw Taukwa, 5.2 6.4 0.8 13.7 14.1 2.1 NA NA NA Franke et al. (1999) cooked Taupok, raw 4.7 7.0 0.6 16.3 19.0 1.6 NA NA NA Franke et al. (1999)

Taupok, 2.1 3.1 0.2 7.2 8.3 0.7 NA NA NA Franke et al. (1999) cooked Australia Tofu, 1.2-10.0 3.2-12.9 0.7-2.0 2.4-11.1 7.0-18.8 1.3-2.5 NA NA NA King & Bignell 6 brands (2000) Tofu mix 21.6 34.3 7.6 93.8 153 21.5 NA NA NA King & Bignell (2000) Tofu 11.2 14.4 NA 1.3 1.3 NA NA NA NA Dalais et al. (1997)

Hawaii Firm tofu 6.2 6.5 1.4 16.9 15.5 3.1 NA NA NA Franke et al. (1999)

Soft tofu 5.1 5.3 1.4 13.7 12.9 3.0 NA NA NA Franke et al. (1999)

Taukwa, 8.8 11.5 2.7 23.8 25.0 5.7 NA NA NA Franke et al. (1999) raw Taupok, 3.3 4.9 0.5 10.3 13.8 1.4 NA NA NA Franke et al. (1999) cooked

260 Dried soya milk curd

Singapore Foojook, 72.3 92.6 13.5 116.0 131.7 18.4 NA NA NA Franke et al. (1999) raw Foojook, 13.7 19.0 2.6 23.2 28.8 3.5 NA NA NA Franke et al. (1999) cooked Hawaii Foojook, 4.6 6.9 0.8 10.1 10.5 1.6 NA NA NA Franke et al. (1999) cooked Fermented soya beans

Australia Soya sauce ND ND NA 0.9 0.4 NA NA NA NA Dalais et al. (1997)

Miso 5.3 12.4 NA 5.8 7.1 NA NA NA NA Dalais et al. (1997)

Soya sauce, 0.0, 0.2 ND ND 0.2,0.6 0.0, 0.4 NA NA NA NA King & Bignell 2 brands (2000) Singapore Fermented ND ND ND 23.2, 26.9 26.2, 31.3 4.9 NA NA NA Franke et al. (1999) tofu, 2 brands Hawaii Miso, 1.5-3.5 1.8-6.8 0.1-0.3 3.5-17.1 5.1-21.0 0.4-3.6 NA NA NA Franke et al. (1999) 3 brands Natto 34.2 39.5 16.8 55.3 49.0 19.8 NA NA NA Franke et al. (1999)

261 USA Soya sauce, ND ND ND 0.1, 0.6 0.0, 0.3 0.0, 0.5 ND ND ND Murphy et al. (1999) 2 brands Miso, 6.2, 6.6 8.9, 9.2 1.0, 1.3 8.8, 8.9 11.7, 12.1 2.3, 2.5 ND ND ND Murphy et al. (1999) 2 brands Tempeh, raw 9.3 20.6 1.4 20.1 31.6 2.2 ND ND ND Murphy et al. (1999)

Tempeh, 10.5 22.6 1.4 19.3 31.6 2.2 ND ND ND Murphy et al. (1999) cooked Soya sauce ND ND NA 0.0 0.0 NA ND NA NA Murphy (1982)

Tempeh 0.2 6.5 1.4 27.3 32.0 3.2 NA NA NA Wang & Murphy (1994b) Miso 7.2 12.3 1.8 7.9 17.7 3.8 NA NA NA Wang & Murphy (1994b) Bean paste 0.0 9.6 2.1 27.2 24.5 7.7 NA NA NA Wang & Murphy (1994b) Tempeh 4.0 11.3 NA 11.3 16.4 NA NA NA NA Coward et al. (1993)

Miso 3.5 4.3 NA 34.5 49.7 NA NA NA NA Coward et al. (1993)

Rice miso 0.0 19.8 NA 7.1 13.6 NA NA NA NA Coward et al. (1993)

Barleymiso 14.2 15.5 NA 18.5 23.9 NA NA NA NA Coward et al. (1993)

Shiromiso 16.3 26.7 NA 10.8 17.0 NA NA NA NA Coward et al. (1993) soup mix Akamiso 25.4 31.9 NA 13.6 17.3 NA NA NA NA Coward et al. (1993) sou mix

262 Soya bean 4.4 7.8 NA 19.7 25.1 NA NA NA NA Coward et al. (1993) paste Soya bean 8.5 6.6 NA 10.3 10.8 NA NA NA NA Coward et al. (1993) paste/rice Soya bean 9.4 11.0 NA 10.5 12.4 NA NA NA NA Coward et al. (1993) paste/wheat Soya sauce NA NA NA 1.4 0.9 NA NA NA NA Coward et al. (1993) (Kikkoman) Japan Miso, NA 7.2, 19.0 NA NA 5.6, 22.9 NA NA NA NA Futukake et al. (1996) 2 brands Nattos, NA 28.3, 49.3 NA NA 3.9, 6.4 NA NA NA NA Futukake et al. (1996) 2 brands Soya sauce, NA 1.0, 2.0 NA NA 0.3 NA NA NA NA Futukake et al. (1996) 2 brands Soya/meat analogues

USA Soya 3.0 7.8 0.8 3.5 7.9 0.9 ND ND ND Murphy et al. (1999) chicken, raw Soya 3.5 9.2 0.9 4.4 9.4 0.9 ND ND ND Murphy et al. (1999) chicken, cooked Meatless 1.1 1.7 0.5 1.0 2.1 0.3 ND ND ND Murphy et al. (1999) frank, raw Meatless 1.1 1.6 0.6 1.4 2.0 0.4 ND ND ND Murphy et al. (1999) frank, cooked Soya burger, 0.8-1.2 1.7-2.5 0.5-0.6 2.4-3.7 4.2-6.6 1.0-1.2 ND ND ND Murphy et al. (1999) raw, 3 brands Soya burger, 0.8-1.0 1.6-2.3 0.5-0.6 2.3-3.0 4.0-5.4 0.8-1.1 ND ND ND Murphy et al. (1999) cooked, 3 brands

263 Soya link, 0.7 1.2-1.5 0.4-0.5 0.8-1.5 2.3-2.7 0.3 ND ND ND Murphy et al. (1999) raw, 3 brands Soya link, 0.7 1.4-1.6 0.5 0.7-0.8 2.7-2.8 0.3 ND ND ND Murphy et al. (1999) cooked Soya beef 0.2-0.3 0.3-0.6 0.0-0.1 0.2-0.6 0.4-1.1 0.0-0.1 ND ND ND Murphy et al. (1999) burger, raw, 4 brands Soya beef 0.3-0.60 0.5-1.0 0.0-0.3 0.3-1.1 0.5-1.7 0.0-0.2 ND ND ND Murphy et al. (1999) burger, cooked, 4 brands Soya hot dog 3.5 6.7 1.5 3.4 8.2 3.4 NA NA NA Wang & Murphy (1994b) Soya bacon tr 2.7 1.4 2.8 6.9 2.4 NA NA NA Wang & Murphy (1994b) Tempeh 3.6 15.8 1.8 6.4 19.6 3.0 NA NA NA Wang & Murphy bur er (1994b) Soya bread

Australia Soya bread ND ND NA 0.7 0.7 NA NA NA NA Dalais et al. (1997)

White bread, ND 0.0-0.1 ND 0.0-0.2 0.1-0.3 NA NA NA NA King & Bignell 3 brands (2000)** Soya& 0.5-2.4 1.3-3.5 0.0-0.8 2.1-7.0 5.8-11.9 0.0-1.4 NA NA NA King & Bignell linseed (2000)** breads, 3 brands

264 Soya flakes

Australia Soya flakes 17.2 26.9 1.9 2.9 2.8 1.9 NA NA NA King & Bignell (2000)** Defatted ND ND NA 0.1 5.1 NA ND NA NA Murphy (1982) soya flakes, toasted Soya flour

Australia Soya flour, 16.4-509 25.1-77.0 7.5-13.3 56.8-93.1 107-160 18.3-22.9 NA NA NA King & Bignell 3 brands (2000)** Soya flour 89.4 140 NA 0.4 0.7 NA NA NA NA Dalais et al. (1997)

USA Soya flour, 104.6-116.1 130.0-144.8 NA 1.9-4.4 2.4-4.4 NA NA NA NA Coward et al. (1993) 4 brands Defatted NA NA NA 8.0-48.0 4.0-40.0 NA NA NA NA Eldridge (1982) soya flour Soya flour ND ND ND 22.6 81.0 ND NA NA NA Wang & Murphy (1994b) Hawaii Soya flour, 20.4, 45.5 23.9, 35.7 10.3, 21.3 71.5, 149.6 87.6, 115.5 30.6, 59.3 NA NA NA Franke et al. (1999) 2 brands

265 Texture vegetable protein

Australia Texture 27.8, 29.5 36.0, 56.1 10.9, 12.2 50.6, 69.8 71.8, 150 19.9, 25.2 NA NA NA King & Bignell vegetable (2000) protein, 2 brands USA Texture ND ND ND 47.3, 48.4 70.2, 71.7 ND NA NA NA Wang & Murphy vegetable (1994b) protein, 2 brands Texture soya ND ND NA 3.0 6.7 NA ND NA NA Murphy (1982) rotein Canned soya beans

Australia Soya beans, ND ND ND 27.2 46.3 ND NA NA NA King & Bignell (2000 canned Soya sprouts

USA Am soy ND ND NA 1.9 7.8 NA 0.7 NA NA Murphy (1982)

Soya protein isolate

USA Soya isolate, 23.2, 27.8 43.0, 58.9 NA 7.3, 10.2 10.5, 18.9 NA NA NA NA Coward et al. (1993) 2 brands

266 Protein ND ND ND 7.7-12.2 27.3-39.3 ND NA NA NA Wang & Murphy Technologies (1994b) International Soya isolate, ND ND NA ND 7.7 NA ND NA NA Murphy (1982) acid Soya protein concentrate

USA Water 11.8 140.0 NA 3.9 3.3 NA NA NA NA Coward et al. (1993) extracted Alcohol 6.4, 10.2 8.7, 22.7 NA 0.4, 4.5 0.4, 6.9 NA NA NA NA Coward et al. (1993) extracted, 2 brands Soya bean NA NA NA 2.0-20 1.0-22.0 NA NA NA NA Eldridge (1982) protein concentrate 'Grain tr 1.8 3.1 tr 1.3 4.2 NA NA NA Wang & Murphy Processing (1994b) Cor . (lA)' Soya cheese

USA Soya cheese 2.1 2.8 NA 0.1 0.2 NA NA NA NA Coward et al. (1993)

Soya tr tr 0.0 1.5 0.8 4.1 NA NA NA Wang & Murphy parmesan (1994b) Mozzarella 0.7 3.3 1.5 1.1 3.6 3.0 NA NA NA Wang & Murphy cheese (1994b) Cheddar tr,l.6 0.0,4.6 1.2, 1.7 0.2, 3.4 0.5,4.0 2.7, 3.5 NA NA NA Wang &Murphy cheese, (1994b) 2 brands

267 Hawaii Soya cheese 2.3 3.5 0.3 7.8 8.8 2.1 NA NA NA Franke et al. (1999)

Soya yoghurt

USA Tofu yoghurt 4.2 8.0 1.2 5.7 9.4 1.2 NA NA NA Wang &Murphy (1994b) Hawaii 2 brands 8.8-11.5 11.2-14.6 2.2-4.5 23.0-30.9 24.0-29.8 4.3-9.4 NA NA NA Franke et al. (1999)

Values are for one brand unless otherwise stated NA = not analysed ND = not detected tr =trace * = mg/100 m1 **=reported after work for this thesis completed

268 269

Appendix 2. Foods purchased in Australia

Foods Sources

Bread

Soy & Linseed Mobile Rise (Chatswood, NSW, Australia)

Soy & Linseed Helga (Australia)

Soy & Linseed Uncle Tobys (Moorebank, NSW, Australia)

Soy & Linseed Molenberg (Moorebank, NSW, Australia)

Soy-Lin LoafBurgen Burgen Bakers (Chatswood, NSW, Australia)

Soy & Linseed Vogel (Moorebank, NSW, Australia)

Soy & Linseed Muffin Quality Bakers Australia (Malaga, WA, Australia)

Cereals

Hi-bran & Soy Weet-bix Sanitarium Health Foods (Berkeley Vale, NSW, Australia)

Phytoestrogens-Soy Breakfast Start Sunsol (K.eperra, QLD, Australia)

Soy Flakes Lowan Whole Foods (Picton, NSW, Australia)

Soy-Tana Speciality Cereals (Mt Kuring-gai, NSW, Australia)

Soy & Oat Flakes Norganic Foods (Bethania, QLD, Australia)

Canned

Baked Beans "Wattie's" Heinz Wattie's (King Street North, Hasting, New Zealand)

Baked Beans "Heinz" Heinz Wattie's (King Street North, Hasting, New Zealand)

Chick Peas (Cicer arietinum) Edgell (Melbourne, VIC, Australia)

Chick Peas (Cicer arietinum) Farmland (Tooronga, VIC, Australia)

Chick Peas ( Cicer arietinum) Master Foods (Wyong, NSW, Australia)

Garden Peas (Pisum sativum L.) Mountain Maid (Batlow, NSW, Australia)

Garden Peas "Edgell" (Pisum sativum Simp lot Australia (Melbourne, VIC, Australia) L.) 270

Appendix 2, continued

Foods Sources

Garden Peas (Pisum sativum L.) Golden Circle (Northgate, QLD, Australia)

Processed Dried Peas "PMU" (Pisum H.J. Heinz (Dandenong, VIC, Australia) sativum L.)

Red Kidney Beans (Phaseolus vulgaris Master Foods (Wyong, NSW, Australia) L.)

Red Kidney Beans "Edgell" Simp lot Australia (Melbourne, VIC, Australia) (Phaseolus vulgaris L.)

Sliced Green B.eans "Edgell" Simp lot Australia (Melbourne, VIC, Australia) (Phaseolus vulgaris L.)

Sliced Green Beans (Phaseolus Golden Circle (Northgate, QLD, Australia) vulgaris L.)

Soya beans (Glycine max L.) Master Foods (Wyong, NSW, Australia)

Dried legumes

Berlotti beans (Phaseolus vulgaris L.) M"Kenzie's (Altona, VIC, Australia)

Black eye beans (Vigna unguiculata) M"Kenzie's (Altona, VIC, Australia)

Bowyer soya beans (Glycine max L.) Allgold Foods (Riverina, NSW, Australia)

Broad beans (Viciafaba LJ M"Kenzie's (Altona, VIC, Australia)

Canellini beans (Phaseolus vulgaris L.) M"Kenzie's (Altona, VIC, Australia)

Chick peas (Cicer arietinum) M"Kenzie's (Altona, VIC, Australia)

Green split peas (Pisum sativum L.) M"Kenzie's (Altona, VIC, Australia)

Haricot beans (Phaseolus vulgaris L.) M"Kenzie's (Altona, VIC, Australia)

Lima beans (Phaseolus lunatus L.) M"Kenzie's (Altona, VIC, Australia)

Peas (Pisum sativum L.) M"Kenzie's (Altona, VIC, Australia)

Red kidney beans (Phaseolus vulgaris M"Kenzie's (Altona, VIC, Australia) L.) 271

Appendix 2, continued

Foods Sources

Red lentils (Lens culinaris) McKenzie's (Altona, VIC, Australia)

Soya beans (Glycine max L.) M"Kenzie's (Altona, VIC, Australia)

Whole green lentils (Lens culinaris) M"Kenzie's (Altona, VIC, Australia)

Frozen

Baby Peas (Pisum sativum L.) McCain Foods (Washdyke, Timaru, New Zealand)

Garden Peas (Pisum sativum L.) Heinz Wattle's (King Street North, Hastings, New Zealand)

Beans (Phaseolus vulgaris L.) M"Cain Foods (Washdyke, Timaru, New Zealand)

Mint Peas (Pisum sativum L.) McCain Foods (Washdyke, Timaru, New Zealand)

Peas (Pisum sativum L.) McCain Foods (Washdyke, Timaru, New Zealand)

Sliced Green Beans (Phaseolus Simp lot Australia (Melbourne, VIC, Australia) vulgaris L.)

Sliced Green Beans (Phaseolus Hy Peak (Australia) vulgaris L.)

Freeze-dried

Garden Peas (Pisum sativum L.) Home Brand (New Zealand)

Sliced Beans (Phaseolus vulgaris L.) Home Brand (New Zealand)

Raw

Alfalfa Onion, sprouts (Medicago Country Fresh (Australia) sativa)

Alfalfa Radish, sprouts (Medicago Country Fresh (Australia) sativa)

Baby beans (Phaseolus vulgaris L.) Unbranded

Berlotti beans (Phaseolus vulgaris L.) Unbranded

Bean Salad, sprouts (Phaseolus Country Fresh (Australia) vulgaris L.) 272

Appendix 2, continued

Foods Sources

"Continental" green beans (Phaseolus Unbranded vulgaris L.)

Crunchy Combo, sprouts Country Fresh (Australia)

Green beans (Phaseolus vulgaris L.) Unbranded

Honey snap peas (Pisum sativum L.) Unbranded

Yard long beans/snake beans (Vigna Unbranded unguiculata)

Mungbean, sprouts (Vigna radiata) Country Fresh (Australia)

Peas (Pisum sativum L) Unbranded

Snow Peas, sprouts Country Fresh, Australia

Soya milk

"Carnation" Instant Soy Powder Nestle (imported from Indonesia)

"Good Life" Berrivale Orchards (Berri, NSW, Australia)

''Nature's" Pureharvest (East Bentleigh, VIC, Australia)

''No Frills" Soy Drink Franklins (Chullora, NSW, Australia)

"So Good Lite" Sanitarium Health Foods (Barkeley Vale, NSW, Australia

"So Good" Sanitarium Health Foods (Barkeley Vale, NSW, Australia

"Sungold" Soy Drink Australia Co-operative Foods (Wetherill Park, NSW, Australia)

"So Natural" Australian Natural Foods (Tarent Point, NSW, Australia)

"Soya King" Tixana (Campsie, NSW, Australia)

"Vitalife" Australian Natural Foods (Tarent Point, NSW, Australia)

"Vitasoy Light V anita" Vitasoy International (Tuen Mun, N.T., Hongkong) 273

Appendix 2, continued

Foods Sources

Tofu

Finn Tofu Joyce (Riverstone, MLB, Australia)

Hard Tofu "Soya King" Tixana (Campsie, NSW, Australia)

Nigari Tofu Joyce (Riverstone, MLB, Australia)

Organic Tofu Earth Star Foods (Murwillumbah, NSW, Australia)

Silken Finn Tofu Joyce (Riverstone, MLB, Australia)

Silken Tofu ''Soya King" Tixana (Campsie, NSW, Australia)

Smoked Tofu Blue Lotus Foods (Kilsyth, VIC, Australia)

Tofu Blue Lotus Foods (Kilsyth, VIC, Australia)

Tofu Cutlets Blue Lotus Foods (Kilsyth, VIC, Australia)

Tofu with Tempeh Nutrisoy (Botany, NSW, Australia)

Tempeh

Tempeh Nutrisoy (Botany, NSW, Australia)

Others

Soy Cheese Mild Flavour Simply Better Foods (Australia)

Soy Sausage Rolls Blue Lotus Foods (Kilsyth, VIC, Australia)

Tofu Burger Nutrisoy (Botany, NSW, Australia)

Tofu Dessert Nutrisoy (Botany, NSW, Australia) 274

Appendix 3. Foods from Indonesia

Foods Sources

Canned & bottled products

Black Beans (Vigna unguiculata) S&W Fine Foods (San Ramon, CAL, USA)

Green Peas "Mating" (Pisum sativum L.) China National Cereals (China)

Lima Beans (Phaseolus lunatus L.) S&W Fine Foods (San Ramon, CAL, USA)

Processed Peas (Pisum sativum L.) Ayam Brand (Malaysia)

Salted Black Beans "Amoy" The Am.oy (Hongkong)

Sweet Peas (Pisum sativum L.) Del Monte Foods (San Francisco, SA, USA)

Young Sweet Peas (Pisum sativum L.) S&W Fine Foods (San Ramon, CAL, USA)

Dried legumes

Cashew kernel U.D. Layar Dufa (Bau-Bau, SULTENG, Indonesia)

Groundnuts (Arachis hypogaea) Unbranded

Yard long bean seeds/snake bean seeds Unbranded (Vigna unguiculata)

Red kidney beans (Phaseolus vulgaris L.) Unbranded

Soya beans (Glycine max L.) Unbranded

Soya beans (Glycine max L.) Imported from USA by Primamakmur Langgeng Perkasa (Jakarta, Indonesia)

Dried soya milk curd

Dried Bean Curd Imported from China

Kembang Tabu Pelita (Indonesia)

Fresh

Baby peas (Pisum sativum L.) Unbranded

Green beans (Phaseolus vulgaris L.) Unbranded 275

Appendix 3, continued

Foods Sources

Green peas (Pisum sativum L.) Unbranded

Yard long beans (Vigna unguiculate) Unbranded

Mungbean, sprouts (Vigna spp) Unbranded (supermarket)

Mungbean, sprouts (Vigna spp) Unbranded (street vendors)

Petai (Parkia speciosa) Unbranded

Petai cina (Leucaena spp.) Unbranded

Soya beans (Glycine max L.) Unbranded

Fermented products

Kecap "ABC" ABC Central Food (Indonesia)

Kecap "Bango" Rina Sari (Jakarta, Indonesia)

Kecap "Indofood" Indosentra Pelangi (Cibitung, Indonesia)

Kecap ''Kedele" AnekaFood Tatarasa(Probolinggo, Jawa Tengah, Indonesia)

"Kikkoman" soya sauce The Amoy (Hongkong)

Oncom kacang Unbranded (Indonesia)

Oncom kedele Unbranded (Indonesia)

Salted Soybean "Soya Sale" Imported from Yeo's (Malaysia)

Taucho Unbranded (Indonesia)

Taucho "AsliNo.l" Gajah 2 (Medan, Indonesia)

Taucho "Medan Harum Sedap" Medan (Indonesia)

Taucho "Mekar" Kujang (Bogor, Indonesia)

Taucho ''No.1 Macan" Pulau Seribu Indonesia (Indonesia)

Tempeh Unbranded (Indonesia)

Dark Soy Sauce "Amoy" Amoy (Hongkong) 276

Appendix 3, continued

Foods Sources

Frozen

Green Peas (Pisum sativum L.) Valley Farm (Bristol, N.B., Canada)

Green peas (Pisum sativum) Unbranded

Soya milk

Mony Susu Kedelai Monysaga Prima (Bekasi, Jawa Barat, Indonesia)

Susu kedelai Unbranded

Yeo's SoyaBeanMilk Imported from Singapore by (Singapore)

Tofu

Tahu TauKwa Mico Sejati (Indonesia)

Tofu Sakake Mitra Boga Segar (Indonesia)

Tahugoring Unbranded from street vendors

Silken Tofu Sutra Kong Kee (Jakarta, Indonesia)

Silken Tofu Sakura Harum Sari Foods (Tanggerang, Jawa Barat, Indonesia) 277

Fi gure 1. Dtied legume analy eel.

Figure 2. Fre h ya b ans analysed. 278

Figure 3. Example of fresh tempch analy ed.

Figur 4. Example of fre h nc m analysed _79

Figure 5. Au tndian ya milk brands analy eel.

Figure 6. Sec nd-gen rati n ya food analysed. 280

Figur 7. Example f oya bread analy d.

Figur 8. Pete (petai) analy ed. t•ctnl ('inn

Figure 9. Petai cina analy ed.

Figur 10. B an and p a products analy d. 282

Figure 11. Bean and p a pr ul analy eel. The following articles have been removed from the digital copy of this thesis. Please see the print copy of the thesis for a complete manuscript.

Title: lsoflavones and Coumestrol in Soybeans and Soybean Products from Australia and Indonesia Authors: L. S. Hutabarat, H. Greenfield and M. Mulholland Journal: Journal of Food Composition and Analysis

Title Quantitative determination of isoflavones and coumestrol in soybean by column liquid chromatography Authors: L.S. Hutabarat, H. Greenfield, M. Mulholland Journal: Journal of Chromatography A,

Title: Phytoestrogens in Soybeans and Soybeans Products from Indonesia Authors: LS Hutabarat, H Greenfield Journal: 2nd South West Pacific Nutrition and Dietetic Conference

Title: Good Nutrition for all - New Era for Nutrition Rights. Authors: 8th Asian Congress of Nutrition AUG. 29- SEPT. 2, 1999 Seoul, Korea Journal: Hutabarat, Lambok Saorita., Greenfield, Heather., Mulholland, Meny.

Title: Isoflvones and Coumestrol in Soybeans and Soybean Products Available in Australia and Indonesia Authors: L.S. Hutabarat, H. Greenfield1 and M. Mulholland Journal: Third International Food Data Conference

Title: Preliminary studies of phytoestrogens in Australian foods Authors: L. S. Hutabarat, H. Greenfield, M. Mulholland Journal: Preliminary studies of phytoestrogens in Australian foods

Corrected- Ll B